Anda di halaman 1dari 267

Taggants in Explosives

April 1980

NTIS order #PB80-192719


Library of Congress Catalog Card Number 80-600070

For sale by the Superintendent of Documents, U.S. Government Printing Office

Washington, D.C. 20402 Stock No. 052-003-00747-9

This assessment was made in response to a request from the Senate Committee
on Governmental Affairs that OTA examine the issues surrounding a proposal to re-
quire that commercial explosives and gunpowders be manufactured with “tag-
gants” as an aid to law enforcement. Two types of taggants are contemplated:
● “identification taggants” would be designed to survive an explosion, and would
carry a code which would enable those who recovered such taggants from the
debris of a criminal bombing to assemble a list of the last legal purchasers of
the batch of explosives used to make the bomb;
● “detection taggants” would be designed to emit a vapor which would escape

from a suitcase, package, etc., so that a taggant-sensing machine at an airport

or public building could detect the presence of concealed explosives.
The proposal to require taggants is generally viewed as helpful by the law en-
forcement community, and opposed by the manufacturers of explosives (and some
others) on the grounds that taggants would be ineffective, unsafe, and too costly.

The report addresses four major questions. First, it reviews the program to de-
velop such taggants, and addresses the question of whether taggants would in fact
work. Second, it assesses the question of whether adding such taggants to explosives
and gunpowders might create a safety hazard. Third, the cost of a taggant program
(on the assumption taggants work and are safe) is calculated, and the major parame-
ters which would affect its costs are identified. Finally, the study assesses the likely
value of such a program (assuming that taggants work, are safe, and are available at
a reasonable cost) to law enforcement.

The project was directed by Dr. Peter Sharfman, Program Manager for Interna
tional Security and Commerce within OTA’S Energy, Materials, and International
Security Division, headed by Assistant Director Lionel S. Johns. The principal inves
tigator was David Garfinkle of Science Applications, Inc.
OTA is grateful for the assistance of its Taggants in Explosives Advisory Panel,
as well as for the assistance provided by the Bureau of Alcohol, Tobacco, and Fire-
arms of the U.S. Department of the Treasury, the Institute of Makers of Explosives,
the Sporting Arms and Ammunition Manufacturers Institute, the 3M Company, and
the Federal Aviation Administration.


Taggants in Explosives Project Advisory Panel

Sanford Kadish, Chairman

Dean, School of Law, University of California, Berkeley

Tom Ashwood Robert E. Hodgdon

Air Line Pilots Association president, Hodgdon powder Co., Inc., and
Pyrodex Corp.
Jerome S. Brower
president, /. S. Brower & Associates, Inc. Neal Knox
Executive Director, Institute for
H. J. Burchell Legislative Action
president, At/a Powder Co. National Rifle Association

Charles E. Cal fee Lynn Limmer

Special Agent Director, Department of Public Safety
Federal Bureau of Investigation Dallas-Fort Worth Airport

Robert R. Dimock, Jr. Alexander v. d. Luft

Utah Copper Division Director, Internationa/ Operations
Kennecott Copper Explosives products Division
E. 1. du Pent de Nemours & Co.
Ernest H. Evans
Brookings Institution Hugh M. McGowan
New York City Police Department
Henry Eyring
Department of Chemistry William T. Poe
University of Utah Louisiana State Police

Eugene H. Eyster Theodore J. Sullivan

Los Alarnos Scientific Lab. Naval Surface Weapons Center

Rona M. Fields Robert W. Van Dolah

Consultant in Psychology Pittsburgh, Pa.

Gary L. Hendrickson Charles O. Williams

Dane Count y Sheriff’s Department Olin Corp.
Madison, Wise.

NOTE: The advisory panel provided advice and critique throughout the assessment, but does not necessarily
approve, disapprove, or endorse the report, for which OTA assumes fulI responsibility.
Taggants in Explosives Project Staff

Lionel S. Johns, Assistant Director, OTA

Energy, Materials, and International Security Division

Peter Sharf man, Program Manager

International Security and Commerce Program

David R. Garfinkle, Principal /investigator

(under contract with Science Applications, Inc.)

Administrative Staff

Dorothy Richroath Jacqueline Robinson

GIoria Proctor Helena Hassell


Edward James, Lawrence Livermore Laboratory Marvin Liebstone, Science Applications, Inc.

Steve Kornish Rowland B. Shriver, J r., Science Applications, Inc.

Roland R. Franzen, Physics Internationa/ Co. Susan Katznelson

James A. Henderson, Jr. Mark Starinsky

OTA Publishing Staff

John C Holmes, Publishing Officer

Kathie S. Boss Debra M. Datcher Joanne Heming

Chapter Page
1. Summary. . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . 3

Il. Detailed Findings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

I l l . Taggant Research Review . . . . . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . 51

Iv. Taggant Safety and Compatibility Review . . . . . . . . . . . . . . . . . . . . . 75

v. Taggant Cost Review. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

V1. Taggant Utility Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137


A. Letterof Request. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175

B. Detection and identification Taggants and Criminal Bombings–
Summary and Questionnaire. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
C. OTA Recovery Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
D. Products Liability implications of Legally Requiring the Inclusionof
Taggants in Explosives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
E. Suitability of ANFO as aFiller for Criminal Bombs. . . . . . . . . . . . . . . . . 227
F. Derivation of Bombing Statist Tables . . . . . . . . . . . . . . . . . . . . . . . . 233
G. Glossary . . . . . . . . . . . . . . ,... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
Chapter I

Chapter L- SUMMARY


Introduction ***. *.. **. **. **** **. **** *.. .*. **e**************** 3
ResearchApproach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Some Project Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
SummaryofFindings .**.*..* . . . . . . . . . .***..*.. .**.**.*. ● **..*.. 9
Taggant Utility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
TaggantCost. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......11
Technical Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......12
Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......12
ContinuingControversies . . . . . . . . . . . . . . . . . . . . . . . ...0..0. ..0..0.. 13
Significance of Compatibility Testing to Date . . . . . . . . . . . . . .......13
Countermeasures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......14
Blasting Agents(ANFO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......14
Survivability and Recovery of Taggants . . . . . . . . . . . . . . . . . . .......15
Development Time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ....16
Congressional Options . . . . . . . . ..*..*.. ....***** ..0..... * * O . * 16
● ● ●



l. MajorSourcesof information. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2. Current Status of Taggant Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3. Summary of Current Status of Taggants . . . . . . . . . . . . . . . . . . . . . . . . . 9
4, Minimum Bombing Incidents Statistics Summary . . . . . . . . . . .......10
5. Praportions of Bombings Attributed to Groups of Perpetrators. .. ....10
Chapter I


At the request of the Senate Committee on Governmental Affairs, the Office of

Technology Assessment has undertaken an analysis of the proposal to mandate the
use of taggants in explosive materials manufactured for commercial use. A “tag-
gant” is a material that might be added to explosives and gunpowders* at the time
of manufacture, as an eventual aid to law enforcement. This study assesses the ex-
isting taggant technology in order to assist Congress in its decision whether to adopt
legislation which would require taggants in explosives and gunpowders.
Two different kinds of taggants are being developed for possible incorporation
in chemical explosives, and it has been proposed that both be required. Identification
taggants are designed to survive the detonation of an explosive, and to be retrieved
from the debris. They would contain a code identifying the batch of explosives or
gunpowder used in a particular bombing. The intent of those advocating the devel-
opment of such taggants is that law enforcement officers investigating a criminal
bombing would retrieve identification taggants and decode them, could then begin
their investigation knowing what kind of explosive material had been used, and
would be able to obtain a list of the last legal purchasers of these explosives and
gunpowders. At the present time the leading contender for an identification taggant
is a color-coded microscopic plastic chip which has been developed by the 3M Co.
Detection taggants are designed to be sensed by a suitable detection machine
even when contained in a package. The intent of those developing detection tag-
gants is that detection machines at airports, public building entrances, and other ap-
propriate sites would signal any effort to introduce explosive materials into the
area. In facilities not normally protected by such devices, portable detection sen-
sors could be used to search the facility in response to a threat. The leading con-
tender for a detection taggant is a microcapsule which would emit small quantities
of a vapor whose molecules are so distinctive that a suitable sensing instrument
(which is under parallel development) could detect a parts-per-trillion concentra-

The Bureau of Alcohol, Tobacco, and Fire- Iy developed and tested; the detection taggant
arms (BATF) of the Department of the Treas- effort is less advanced than the identification
ury, which is the executive agency that has taggant effort.
jurisdiction over most crimes involving high ex-
Legislation proposed in the U.S. Senate
plosives, has sponsored a program to develop
would make it unlawfuI (in the words of S. 333)
taggants, Most of the effort has been carried //
. . . for any person or persons to manufacture
out or supervised by the Aerospace Corp., un-
any explosive material which does not con-
der contract to BATF. Neither identification
tain . .“ both detection taggants and iden-
taggants nor detection taggants have been ful-
tification taggants, a n d w o u l d r e q u i r e t h a t
manufacturers and distributors keep records
● The term gunpowder includes black and smokeless powders
and pyrodex (a registered trademark of thePyrodex Corp) a showing the distribution chain for each batch
black powder substitute of explosive material that carried a separate

4 ● Taggants in Explosives

identification taggant code. (Similar legisla- 5. the effects on cost and utility of excluding
tion has been proposed in the House of Repre- certain explosive materials from the tag-
sentatives. ) The Secretary of the Treasury gant program;
wouId issue regulations implementing this re- 6 the removal of taggants from tagged ex-
quirement, and such regulations would be plosives; and
phased in as testing was completed and tag- 7 alternatives to a taggant program.
gants became avaiIable in sufficient quantity.
The text of the request letter is included as ap-
At hearings on this proposal, representatives pendix A.
of the explosives and gunpowder industries
The proposal to require that taggants be
and others expressed opposition to this pro-
added to commercial explosives at the time of
posal on the grounds that:
manufacture has aroused intense controversy.
● it is premature to consider explosives tag- While OTA believes that this report will serve
ging legislation while development and test- to narrow many of the areas of controversy,
ing of taggants have not been completed; there are a number of issues on which the
● taggants may be unsafe, since they would available data do not permit a scientifically
require adding a foreign substance to the ex- conclusive finding. OTA has therefore made a
plosive materials; number of judgments based on the available
● a taggant program would be extremely cost- evidence where conclusive proof was lack ing.
ly; and In some cases these judgments, and the reason-
● a taggant program would not, in fact, have ing underlying them, have proved unpersua-
much utility for law enforcement. sive to one side or another in the controversy.
Therefore, the final section of this chapter
Proponents of a taggant program have coun-
calls attention to the major areas in which one
tered that:
or more affected parties may disagree with the
● taggants are inert materials, no more unsafe OTA findings.
than current additives to explosives and gun-
● a taggant program need not be unduly cost- Research Approach
ly; and In order to assess the impacts of a taggant
● bombings are extremely difficult crimes to program, a two-stage approach has been nec-
prevent or solve using existing methods, and essary. As the first stage, an analysis has been
taggants would provide an extremely useful made of the safety and technical efficacy of
tool to law enforcement agenices. the taggants at the current state of develop-
The Senate Committee on Governmental Af- ment, since cost and utility are moot points if
fairs has requested that OTA review the avail- the taggants are not safe and do not work. As
able data on explosive taggant technology, the second stage, an assumption has been
and conduct an assessment which would ad- made that the taggants work and are safe and
dress; a parametric analysis of costs and utility made
as a function of the specific implementation
1. the safety of adding taggants to explo- plan.
2. the postdetonation survivability and re- Due to severe time constraints, OTA did lit-
coverability of identification taggants; tle original research; instead, an intensive re-
3. the cost impact of a taggant program on view of existing research was supplemented by
the explosives industry and users; discussions with manufacturers, distributors,
4. the utility of a taggant program to law en- and users of explosives and gunpowders, and
forcement; with law enforcement personnel and experts
Ch. l—Summary ● 5

on terrorism. Table 1 summarizes the major tion, are integrated into the analysis in chapter
sources consuIted. VI, and reported in detail in appendix B.
I n addition, OTA sent a questionnaire to ap-
proximately 950 members of the International OTA also directed a series of tests on the re-
Association of Chiefs of Police (IACP) asking coverability of the 3M identification taggant.
them to assess the utility of taggants. (The The Aerospace Corp. had conducted a large
IACP membership list was chosen because it number of laboratory tests on the survivability
constituted a broad cross section of the law en- of the 3M identification taggants, but the only
forcement community. ) The questionnaire was information on the recovery of taggants under
sent to a random sample of the IACP members, field conditions came from poorly docu-
and the low response rate (about 15 percent) mented demonstrations and training tests, con-
probably created a bias towards those with in- ducted by BATF, the Federal Bureau of investi-
terest in, and knowledge of, the subject. (A gation, and other organizations. These tests,
possible misconception may have been intro- and others conducted by the Institute of Mak-
duced by the explanatory material introducing ers of Explosives, had produced conflicting
the questionnaire, which inadvertently indi- and contradictory results. OTA planned and
cated that identification taggants could iden- supervised a limited series of tests of the post-
tify the last legal purchaser of explosives used detonation recovery process of taggants from
in a bombing, rather than identifying a list of automobiIes. The resuIts of these tests are inte-
last legal purchasers. ) The results of the ques- grated into the findings, and described in de-
tionnaire, interpreted with considerable cau- tail in appendix C.

Table.–Major Sources of Information

Manufacturers Explosives users–continued

Explosives manufacturers (Du Pent, Atlas, Independent, Goex, Hercules) Coal Mine (Webster Coal Co. )
Gunpowder manufacturers (Hercules, Goex, Olin, Pyrodex ) (Quarries (Tri-State, Rockville Crushed Stone)
Manufacturer of identification taggants (3M Co. ) Construction firm (Guy Atkinson)
Trade organizations Blasting contractor (Tri-State Explosives)
Institute of Makers of Explosives(I ME) Law enforcement personnel
Sporting Arms and Ammunition Manufacturers’ Institute (SAAMI) New York, N.Y,
Consumer organizations San Mateo County, Calif,
National Rifle Association (NRA) Dallas-Fort Worth Airport, Tex,
National Muzzle Loaders Association (NM LA) Summit County, Ohio
Washington, DC.
Organizations developing a taggant program
Bureau of Alcohol, Tobacco, and Firearms of the U S. Treasury Experts on terrorists and terrorism
Department (BATF) Experts from foreign and domestic law enforcement agencies
Aerospace Corp. (BATF contractor) Writers on the subject (Dr. Ernest Evans, Dr. Rona Fields,
Dr. Robert Kupperman)
Organizations involved in taggant research
Management Sciences Associates Foreign law enforcement sources
Institute for Defense Analyses West Germany
Lawrence Livermore Laboratories England
Explosives and gunpowder distributors Interpol
B, F Hodgdon
Tri-State Explosives U.S. Federal agencies
Federal Bureau of Investigation
Gunpowder retailer Federal Avation Administration
The Bullet Hole Bureau of Mines
Explosives users Department of Transportation
Copper mines (Bingham Canyon open pit mine. Crow Fork U.s. Army (Corps of Engineers, Criminal Investigation Division,
underground mine) Development and Research Command)

SOURCE Office of Technology Assessment

6 ● Taggants in Explosives

Photo credit Kennecott Copper Co.

[explosives are utilized extensively at the Bingham Canyon open pit copper mine

Some Project Limitations with detection taggants. Evidence has been

found of reactivity (using high taggant concen-
There are three general limitations to the trations at elevated temperatures) between the
completeness of this analysis of the proposal 3M identification taggants and one type of
to legislate the use of taggants in explosive ma- smokeless powder, as well as one booster ma-
terials. The primary limitation is caused by the terial. This reactivity creates a presumption of
preliminary nature of the taggant research– incompatibility. Until this presumed incom-
much data are simply not available. Additional patibility is resolved, taggants cannot be safely
information is req Jired on all aspects of the added to these explosive materials. Resolution
analysis—technical efficacy, safety, cost, and of the problem may result in significant
utility. Table 2 summarizes the research con- changes in the taggants, requiring a new set of
ducted to date. compatibility tests and perhaps changing the
basis of the cost analysis. If the problem is re-
Preliminary safety testing has been con- solved, more data still need to be generated.
ducted on only a portion of the materials to The lack of data on long-term effects, in terms
which identification taggants would be added, of safety, stability, and performance, especial-
and compatibility testing has barely begun ly on products such as gels and slurries, is par-
Ch. l—Summary ● 7

Photo credit U S Department of fhe Treasury

Photograph of automobiles utilized in the OTA taggant recovery test

Table 2.–Current State of Taggant Researcha

ID taggants Detection taggants

Compatibility Survival recovery Compatibility
C a p s e n s i t i v e . Preliminary finished Preliminary finished Preliminary underway
B o o s t e r s Preliminary underway–compatibility problem identified Preliminary underway Testing initiated
Detonators . ., Prelimmary underway Preliminary underway Testing initiated
Blastlng agents ., ., ., None None None
Detonating cord ., . . None Testing initiated Testing initiated
Black powder ., . . . . Preliminary finished Preliminary underway Preliminary underway
Smokeless powder ... . . Preliminary underway–compatibility problem identified Preliminary underway Testing initiated
Military explosives ., ., None None None
As of mid-January 1980
SOURCE Office of Technology Assessment
8 . Taggants in Explosives

ticularly important. As a result of this uncer- has reached a courtroom. (Those investigating
tainty, not even preliminary indications of and prosecuting the case considered evidence
safety are possible at this time, much less the from taggants very helpful.) Quantification of
demonstrations recessary before a taggant the utility of taggants (identification as well as
proposal could safely be implemented. detection) is therefore simply not possible, par-
ticularly given the inadequacy of bombing sta-
While preliminary research has been con-
tistics. Experience with the date-shift code
ducted on the survivability and recoverability
(which facilitates tracing of undetonated ex-
of the 3M identification taggants, only a por-
plosives) provides useful data, as does the ex-
tion of the explosive materials which might be
perience of foreign countries, but the available
tagged was tested, and that research is poorly
information on the utility of taggants is pre-
documented. Hundreds of possible detection
ponderantly qualitative in nature.
taggants have been screened to yield five can-
didate materials, but detailed testing of the
properties of those materials is barely under- A second general limitation to the complete-
way. Similarly, three candidate detection sen- ness of the analysis, imposed by Iimits on avail-
sors have been identified, and Iimited Labora- able time and resources, is that only a limited
tory testing of preliminary or “breadboard” sample of the population concerned with the
models completed. Methods of air sampling study could be contacted. As a result, cost data
are also at a preliminary stage. Thus, estimates derived from a detailed analysis of one or two
of technical efficacy can only be made on the companies have been assumed to be represent-
basis of preliminary data. ative of an entire segment of an industry, such
as underground coal mining or retail sale of
As a result of the pilot test program, reason- gunpowders. Similarly, processes for adding
able data are available for the analysis of the taggants, reworking of waste material, quality
cost impact of adding taggants during the control, compatibility testing, and storage,
manufacture of cap-sensitive high explosives, which are applicable to a segment of the man-
at least for those companies which partici- ufacturers of explosive materials, have been
pated in the program. The data, however, on assumed to be universal for the purpose of
the cost impact of adding taggants during the generating cost estimates. A more serious man-
manufacture of the other types of explosive ifestation of the limited sample size is that in-
materials (for exalmple, gunpowder) are less depth discussions of the utility of identifica-
adequate. While firm estimates of the cost of tion and detection taggants to law enforce-
unencapsulated identification taggants are ment and security personnel could only be
available from 3 M under a variety of imple- held with a small number of organizations. As
mentation conditions, little data are available the bomber threat varies considerably from
for the cost of encapsulated identification tag- one part of the country to another, it is diffi-
gants (a more likely baseline case) or for the cult to generalize the results of those discus-
cost of detection taggants. Only the grossest sions.
estimates have been made of recordkeeping
costs, and the estimates by both the propo-
The third limitation on the analysis is caused by
nents and opponents are open to some ques-
the language of the draft legislation, S. 333. The
tions of objectivity. Rule-of-thumb engineering
bill calls for tagging of all “explosive materi-
estimates have been made for the candidate
als, ” which does not appear practicable if the
sensor systems costs, but the accuracy of those
phrase is strictly interpreted to include the tag-
estimates cannot be very precise as neither
ging of blasting agents that are mixed the same
production rate, tc)tal production, nor specifi-
day they are detonated, and otherwise offers
cations have been established.
no guidance for the implementation regula-
So far, identification tagging of explosives tions which the Secretary of the Treasury
has played a part in only one criminal case that would promulgate.
Ch. l—Summary ● 9

This assessment distinguishes between an cation and detection taggants had successful Iy
evaluation of the present state of development completed the development process, including
of taggants and a projection of the cost and a resolution of the safety issues. These anal-
utility of a taggant program if and when the yses are contained, respectively, in chapters V
necessary development and testing are suc- and VI. Details of these and other findings are
cessfully completed. A detailed evaluation of given in chapter 11. The principal findings are
the development status of the identification shown in table 3 and briefly summarized be-
and detection taggants is contained in chapter low.
I I 1. A crucial factor in the development status
evaluation concerns the safety of adding tag-
gants to explosives; the safety and general Taggant Utility
compatibility analysis is contained in chapter
IV. OTA then separately evaluated the cost Assuming, for purposes of analysis, that stabil-
and utility of a program to add taggants to ity questions are successfully resolved and that
commercial explosive materials. For this anal- technical development is successfully completed,
ysis, it was assumed that the baseline identifi- both identification taggants and detection tag-

Table 3.–Stimmary of Current Status of Taggants

Idenhf!cation taggants Detection taggants
Dynamites, gels. slurries, No change in sensitivity, stability No reported data; testing initiated
Black powder. No change in sensitivity, stabddy No reported data; testing initiated
S m o k e l e s s p o w d e r Reactivity with Herco’ powder observed, No reported data, testing initiated
incompatibility presumed
B o o s t e r m a t e r i a l s Reactivity with Composition B observed, No reported data; testing initaited
incompatibility presumed
Blasting agents ., No data No data
Performance . . . . . . . . . . . . Limited testing No data
Favorable conditions. Yes N/A
Fire ., ., Probable N/A
Confinement : Insufficient data N/A
Field recovery ., Probable if survive N/A
F i e l d r e a d i n g Unlikely N/A
Laboratory reading Almost all conditions NIA
Sensor development. . . . . . . N/A Early stages
Low-value targets Little Virtually none
High-value targets, no
countermeasures High High
High-value, Including
c o u n t e r m e a s u r e s High, due to Increased risks High for all but most sophisticated
Cost, $ millions/year Identification Detecioon Both
Low-level program (ID tag code for each product changed
e a c h y e a r . A N F O e x c l u d e d )a . $15 $22 $30
Baseline program (ID tag code for each product changed
for each date/shift, ANFO excluded) . 25 25 45
High-level program (ID tag changed for each 10,000-lb
batch, ANFO Included) . . 215 65 268
N/A not applicable
aThese programs are defined in detail in ch v

SOURCE Off Ice of Technology Assessment

10 ● Taggants in Explosives

gants would be useful law enforcement tools greatly in their motivation, skill, training, re-
against most terrorist and other criminal bomb- sources, and ability to respond to a changing
ers. Their utility against certain types of bombers enforcement environment. They are defined
would probably be quite high; their utility against and their proportions estimated in table 5.
the most sophisticated of terrorists and profes- Note that despite the tendency for some
sional criminals is open to question. groups to claim “credit” for a bombing, a
motive was established for only 23 percent
● Data on the number and kinds of bombings
of the bombings reported to BATF in 1977
committed are dispersed and inconsistent.
and only 38 percent in 1978; table 5 is based
Table 4 gives an idea of the magnitude of
on the assumption that the distribution of
the problem; its significance is discussed in
motives was the same for the numerous inci-
chapter II and the derivation of the figures
dents in which law enforcement officials
in appendix F. OTA diligently sought to find
were unable to assign a motive.
or reliably derive data from which one could
calculate the number of bombings that a Identification taggants would facilitate the in-
taggant program would solve or deter, and vestigation of almost all significant criminal
found this an impossible task. bombings in which commercial explosives
were used. Due to the need for laboratory in-
● Criminal bombings are committed by a wide
volvement in the taggant recovery process,
range of perpetrators, including both “individu-
the taggants would probably not enter into
als and groups. It is helpful to group criminal
investigations of bombings that produce no
bombers into four categories, which differ
casualties and I ittle property damage.
● Detection taggants would be very effective in
Table 4,–Minimum Bombing Incidents Statistics Summary a
protecting those high-value targets where pro-
BATF FBI tection by detection taggant sensors is feasible.
Item 1977 1978 1977 1978 The improvement in protection of such po-
Explosive bombings, number. ., ... l,037b 896b 867 768 tential targets would be quite substantial.
Undetonated explosive bombs, number. 319 287 118 105 However, most current bombings take place
Incendiary bombings, number . ., 339 446 248 349
Unignited incendiary bombs, number ., 81 71 85 79 against targets that are unlikely to be pro-
Criminal accidents, number ., ., ., ., 21 67 – – tected by detection taggant sensors.
Property damage from bombings,
millions of dollars c d . ., . . . . . . $ 10 $ 17 $ 9 $ 9 ● Adding taggants to blasting agents would have
Injuries ., ., . . . . . . . . . . . . 180 185 162 135 some utility, but the incremental utility would
People killed by bombingsc ., , ., , 38 23 22 18
a BATF reported 3.177 total Incidents in 1977 and 3,256 in 1978 Total incidents include ac -
be small compared to the utility of tagging cap-
cidents, threats deized and recovered explosives, and hoaxes as well as axtual explosives and in- sensitive high explosives, gunpowders, and
cendiary bombings The OTA stud/ was concerned only with explosive bombings
of these 953 in 1977 and 787 m 1978 were against substantial targets detonators (and the incremental cost would be
c lncludes both explosive and incendiary bombings OTA was unable to obtain Separate figures for
the number of criminal accidents, injuries, deaths, and property damage caused by Incendiary
high). A taggant program that did not in-
bombs Incendiary bombs and bombings would not be affected by taggant program clude gunpowders would be of relatively
d Actual value probably considerably higher due 10 lack Of data file updates
limited utility as pipe bombs filled with gun-
SOURCE SOURCE: BATF 1978 Explosives Incidents Report. FBI Uniform Crome Report. Bomb Report.
1978 See app F for a dicussion of the derivation of these figures powder are used in a substantial number of

Table 5.–Proportions of Bombings Attributed to Groups of Perpetrators (average for years 1974-78)

Percentage Estimated number

Bomber type Characteristics of bombings in 1978a
Terrorists. . . ., ., ., ... Highly motivated, varied skill levels, act in groups, continuing involvement 12 107
Criminals . . . ., . . . . . . Varied motivations, varied skill levels, act alone or in small groups, 11 98
repeated activities, specific targets
Mentally disturbed ., ., ., . . . . Highly motivated, poorly trained, act alone, seldom repeat crimes 38 340
Vandals and experimenters, Limited motivation, poor training, limited resources, do little damage 39 348
asee app F for dervations of these estimates
Ch. l—Summary ● 11

bombings; i t o n l y h i g h e x p l o s i v e s w e r e against considerations of law enforcement

tagged, criminals could shift to pipe bombs ut i I it y.
rather easily.

● The utility of both identification and detection A low-level taggant program, in which a
taggants would be decreased because some unique taggant species would be used to
bombers would take countermeasures. Explo- identify each year’s production of a specific
sives experts have suggested a number of product, and 800 detection sensors would be
possible countermeasures to the proposed deployed, would cost $3o million per year.
taggant technology which would be avail-
able to those bombers with the requisite A “baseline” program identified by OTA (de-
knowledge and resources. Most available scribed in detail in ch. V) would cost approxi-
countermeasures would increase the risk to mately $45 million per year,adding approxi-
the bomber of personal injury or arrest, or mately 12 percent to the cost of cap-sensi-
decrease the reliability of the bomb. Law en- tive explosives and slightly under 8 percent
forcement officials and experts on terrorism to the cost of gunpowder, Cap-sensitive
agree that most bombers would not utilize high explosives, boosters, detonators, deto-
the available countermeasures. A taggant nating cord, and gunpowder would be
program would retain substantial utility tagged. A unique taggant species would be
even though some criminal bombers would used for a shift’s production of each product
attempt countermeasures, and these coun- and size. Fifteen hundred detection sensors
termeasures would be effective whenever would be deployed. The bulk of this cost
they were carried out with sufficient knowl- would eventually falI on users of explosives
edge and skil1. and on users of products produced with the
aid of explosives; the costs of detection tag-
● The utility of taggants to law enforcement per-
gant sensors would presumably be borne by
sonnel is not adequately quantifiable, due to
the owners or users of protected facilities. It
the paucity of data on taggants or similar
is not expected that costs of this magnitude
control mechanisms, the difficulty of ana-
would lead to any major shifts in the pat-
lyzing the currently collected statistics on
terns of production and use of explosives.
bombings, and the fact that it is difficult to
quantify how much any single clue adds to
an investigation or prosecution. Generally Separate baseline identification and detection
speaking, law enforcement techniques are taggant programs would cost approximately
seldom subjected to cost-benefit analysis, $25 million per year each, including public
and the data which exist do not lend them- overhead costs.
selves to such effort. Similarly, OTA was un-
able to quantify the deterrent effect tag- A high-level program, in which a unique tag-
gants may have, although the apparent ef- gant would be used for each 10,000-lb batch
fectiveness of airport screening procedures of explosives or 2,000-lb batch of gunpow-
in reducing the number of hijacking at- der, in which blasting agents would be
tempts suggests that detection taggants may tagged, and in which 5,000 detection sensors
have a considerable deterrent value. would be deployed, would have an estimated
cost of $268 million per year.
Taggant Cost
The cost estimates assume that the taggant
The cost of a taggant program would vary material costs do not differ appreciably from
enormously depending on the nature of the pro= current estimates for mass-produced taggants.
gram. Costs are likely to be reasonable if and Chapter V discusses the causes and the ex-
only if any taggant legislation requires regula- tent of the uncertainties surrounding these
tions to be written in a way that weighs costs cost estimates.
02 ● Taggants in Explosives

Technical Development Safety

The development of taggants is not yet com- The tests so far conducted create a presump-
plete. Further developmental effort, particu- tion that there are no incompatibilities between
larly resolution of the questions regarding the the 3M identification taggant and dynamites, slur-
stability of smokeless powder and cast boost- ries, gels, emulsions, or black powder. Neverthe-
ers to which taggants have been added, and less, a full-scale qualification program is neces-
successful completion of a variety of tests, sary before taggants can be added to all such ma-
would be required before it would be appropri- terials.
ate to begin adding taggants to commercial ex- ● The addition of 3M identification taggants to

plosives. one brand of smokeless powder (Herco” *)

and one variety of booster material (Composi-
● The identification taggants developed by 3M
appear to survive the detonation of commer- tion B) produces a chemical reaction at ele-
vated temperatures and high taggant concen-
cial explosives under ideal conditions. Con-
finement and fire may adversely affect sur-
trations. The taggants must be considered in-
compatible with such explosives unless or
vival, although test data is very limited. Re-
until: 1 ) the composition of the taggant is
covery of the taggants appears to be a func-
tion of the specific incident conditions changed in a way that eliminates this chemi-
(weather, type of target, firefighting activ- cal reaction, or 2) a determination is made
ities) as well as the training and care of the that the reaction takes place only under cir-
field and laboratory investigators. A trained cumstances that can be prevented from aris-
ing in commercial production, distribution,
team can probably recover debris from
which a laboratory can separate taggants and use. If the incompatibility remains, then
Congress could, if it chose, require that
under most incident conditions.
these particular explosives either be them-
● There is little basis for judging whether the de- selves modified, withdrawn from the mar-
tection taggant system, based on machine ket, or granted an exemption from tagging.
sensing of microencapsuiated vapors, which (OTA believes that exemption of smokeless
appears to show promise under laboratory powders could significantly diminish the
conditions, would function reliably under con- utility of a tagging program; exemption of
ditions of mass production and field use, or cast boosters would diminish this utility to a
how soon sucn a system would be available. somewhat lesser extent. ) If compatibility is
established, completion of a qualification
program would still be necessary.
● There is little evidence regarding the safety of
detection taggants, or of the combination of
identification and detection taggants, as testing
has only recently been initiated and no results
have yet been reported.
● Analysis, and the limited testing so far con-
ducted, indicate that the performance of ex-
plosive material would not be degraded by the
addition of taggants. However, preliminary
tests suggest that abnormally high concen-
trations of taggants might decrease the bal-
listic performance of smokeless powder.
Testing, including long-term effects, would
be necessary, however, before the question
could be fully resolved.

*A registered trademark ot Hercules, Inc

Ch. l—Summary ● 13

Some of OTA’S findings have been chal- should be reviewed and possibly augmented.
lenged by one or more of the participants in However, Aerospace points out that while not
the controversy that surrounds the proposal to every test has been conducted with every
require that commercial explosives be tagged, brand of every explosive, the program suc-
The nature of these challenges is outlined here cessfully carried out was designed by industry
and was considered sufficiently thorough so
that several major firms were willing to distrib-
Significance of Compatibility u t e p i l o t q u a n t i t i e s of tagged explosives
Testing to Date through their normal commercial distribution
channels. With regard to smokeless powders
A large number of tests have been carried
and cast boosters, Aerospace takes the view
out to determine whether the 3M identifica-
that no safety hazard has been demonstrated,
tion taggant is compatible with commercial ex-
but that the failure of the tagged explosive to
plosives. More tests are required, and the Aero-
pass certain extreme tests means that compati-
space Corp. (under contract to BATF) is spon-
bility has yet to be demonstrated, and the pos-
soring a continuing testing program. The tests
sibility that some changes will be required to
completed to date are described in chapter IV.
ensure safety cannot be ruled out.
OTA found that the testing done to date cre-
Representatives of the explosives industry take
ates a reasonable presumption that the 3M iden-
the view that taggants cannot be considered
tification taggant is compatible with dynamites,
compatible with explosives until all the testing
gels, slurries, emulsions, and black powder. On
that ought to be carried out has been successfully
the other hand, there is evidence of increased re-
completed. They maintain that untiI safety has
activity, and thus a presumption of incompatibil-
been conclusively demonstrated, it would be
ity, with at least one form of smokeless powder,
premature to consider whether to legislate a
and at least one cast booster composition. It is
requirement that commercial explosives be
not yet possible to arrive at presumptions
tagged. Explosives industry representatives
about the compatibility of the 3M taggant with
also make a distinction between the pilot pro-
blasting caps or detonating cord, or about the
gram so far carried out and normal commer-
compatibility of detection taggants with any
cial production. They maintain that the tagged
commercial explosive. OTA further found that,
explosives manufactured under the pilot pro-
even for products such as dynamites where no
gram received unusual care and attention dur-
evidence of incompatibility exists, further test-
ing the manufacturing process, and were
ing is required before it can be definitely con-
distributed to a limited number of selected
cluded that taggants are compatible with, and
distributors. The manufacturers also believe
can safely be added to, al I such explosives.
that the terms of the pilot program relieved
The Aerospace Corp. takes the view that the them of liability for accidental explosions due
compatibility tests with dynamites, gels, slurries, to taggants, a point which the Aerospace Corp.
emulsions, and black powder generally are suffi- contests. Some explosives industry represent-
cient to permit implementation of a program to atives take the view that the failure of the mix-
tag these substances. Aerospace recognizes ture of taggants with one brand of smokeless
that there is a need for Mine Safety and Health powder and one cast booster composition to
Administration approval of tagged permissible pass one safety test means that the 3M taggant
dynamites, that final qualification of produc- should be viewed as unsafe unless or until it is
tion-line 3M taggants must be made to ensure redesigned, and point out that any such rede-
that they match those used in the pilot test, sign would require repeating all other tests
and that the black powder ballistics testing previously carried out.
14 . Taggants in Explosives

Countermeasures gram to counter aircraft hijacking; since that

program began, thousands of weapons have
It is clear that it would be possible for terrorists been detected each year, while there have
or other criminals to take measures to defeat the been no cases of aircraft hijacked with wea-
impact of a tagging program, by making or ac- pons smuggled onboard, despite the fact that
quiring untagged explosives. OTA found that mechanisms can be postulated for smuggling
such countermeasures would require a consid- weapons past the screening apparatus. OTA
erable degree of technical knowledge and believes that while countermeasures to a tag-
skil1, and that in most cases countermeasures gant program would be available and would be
would either require the commission of an ad- effective if correctly used, most bombers
ditional crime (with some added risk of ap- would not make effective use of such counter-
prehension), or else manufacturing or modify- measures. OTA believes that taggants, if success-
ing explosives in a way that would risk either a fully developed, could have significant law en-
premature explosion or a misfire of the bomb. forcement utility even if some terrorists or other
The law enforcement experts whom OTA con- criminals successfully employed countermeas-
sulted predict that many terrorists and other ures.
criminals would probably not avail themselves of
countermeasures that were theoretically avail-
able to them. Blasting Agents (ANFO)
Representatives of the explosives industry take Blasting agents are the most widely used
the view that one should assume that an avail- type of commercial explosive; the most com-
able countermeasure will in fact be employed. mon type of blasting agent consists of mixtures
They point out that the most sophisticated of prilled ammonium nitrate and fuel oil; these
bombers, who are most likely to be willing and explosives are collectively known as ANFO.
able to employ countermeasures, are those ANFO can be mixed in a factory, or mixed di-
which may pose the greatest threat. They fear rectly at the site where blasting is to take
that a taggant program would fail to be effec- place. Ammonium nitrate fertilizer can be
tive because of widespread use of counter- mixed with ordinary fuel oil to create a rather
measures, and that law enforcement officials insensitive ANFO.
would then wish to counter the countermeas-
ures by extending the range (and hence the Because of the very large volume of ANFO
cost) of the taggant program. that is used commercially, a tagging program
which included ANFO would be substantially
OTA has noted a consistent pattern of dis-
more costly than one from which ANFO was
agreement on this point. Experts in the explo-
excluded. Chapters I I and V present detailed
sives industry and Government explosives ex-
information on this point. One of the reasons
perts almost unanimously believe that coun-
for the wide gap between BATF and the explo-
termeasures exist which would enable bomb-
sives industry cost estimates for a tagging pro-
ers to evade the effects of a taggant program,
gram is that the industry read the draft legisla-
whether the countermeasures take the form of
tion (S. 333) as requiring that ANFO and other
removal of taggants from tagged explosives,
blasting agents be tagged, while BATF was
use of untagged blasting agents, theft of explo-
planning for a taggant program that would not
sives, fabrication of “homemade” explosives,
include ANFO.
or use of incendiary devices. Law enforcement
experts, and experts on terrorists and terrorism, Representatives of the explosives industry
almost unanimously believe that most bomb- have taken the position that exclusion of
ers, including terrorists, would fail to take the ANFO would greatly diminish the law enforce-
steps necessary to evade a taggant program, ment utility of a taggant program, because
even though the necessary equipment and bombers could and would use untagged ANFO
knowledge is not too difficult to obtain. A pos- in place of tagged, cap-sensitive explosives or
sible analogy is the effectiveness of the pro- tagged gunpowders. OTA believes that it is in-
Ch. l—Summary ● 15

deed the case that an effective bomb, suitable At the present time, ANFO is seldom used in
for almost all criminal or terrorist purposes, pipe bombs despite the fact that it is cheaper
can be manufactured from ANFO if the crimi- and, if properly detonated, considerably more
nal has adequate time, skill, knowledge, and energetic than smokeless powder. Whether the
motivation. The critical area about which judg- tagging of cap-sensitive high explosives and
ments differ is the extent to which terrorists and powders would in fact lead many criminals to
other criminals would in fact make use of ANFO switch to the use of ANFO is a question that
bombs if other commercially available explosive cannot be answered with certainty. However,
materials were tagged. as in the case of other countermeasures, OTA
has found that explosives experts tend to expect
OTA does not consider it appropriate to de-
that criminals would switch to ANFO, while law
scribe here how one would go about manufac-
enforcement experts and experts on terrorism
turing an AN FO-filled bomb. The process in-
tend to doubt that this would happen in many
volves more steps, a greater number of materi-
als and components, and more opportunities
for error than a bomb made from a cap-sensi-
tive explosive; however, it would be easier and
Survivability and Recovery of Taggants
safer than fabrication of a bomb from “raw
chemical s.” The ANFO commercially avail- The testing done to date on the conditions
able in the United States would not be reliably under which identification taggants would in
detonated by an ordinary detonator (#8), even fact survive an explosion, and surviving tag-
in a pipe bomb. ANFO can be readily deto- gants could in fact be recovered, is not ade-
nated by using a smalI high-explosive booster, quate to sustain firm conclusions. Much of the
but such boosters would be tagged, and a large available data is anecdotal rather than system-
booster or several small ones would make an atic. Part of the problem is that it is difficult
efficient bomb without the use of ANFO. to arrange for testing under realistic but con-
ANFO can also be detonated using materials trolled conditions. Faced with inadequate and
that wouId not be tagged (if the bomber knows somewhat contradictory data, particularly
how to wire them), but an ANFO pipe bomb is with respect to the recovery question, OTA ar-
substantialIy harder to detonate than a smoke- ranged for a very Iimited test program to sup-
less-powder pipe bomb or a stick of dynamite. plement the previous tests; appendix C reports
on this effort.

OTA feels that prior testing supports the

presumption that taggants would probably sur-
vive most bomb detonations under most condi-
tions. However, survivability decreases with
the size of the explosive charge and its power.
The survivability of individual taggants in
large explosive charges or in extremely power-
ful explosives (such as booster material and
military explosives) has not been demon-
strated. Pressed pellets, fabricated from the in-
dividual taggants, do survive the detonation,
but recovery has not been adequately demon-
strated, and compatibiIity tests on pellets re-
main to be accomplished. OTA found that the
Photo credit U.S Department of the Treasury
taggants surviving most bombs could probably
be recovered under most conditions. However,
A typical pipe bomb. Such bombs are normally filled with
black and smokeless powder, but a bomber with sufficient field investigators might well find it impossible to
knowledge and skill could use ANFO separate the taggants from the debris, identify in-
15 ● Taggants in Explosives

dividual taggants, and read the codes in the field; judgment about the utility of an explosives
instead the field team would have to gather de- tagging program.
bris likely to contain taggants, and a laboratory
could thereafter separate and read the taggants.
Development Time
Such a laboratory need not be elaborate, and
could be installed in a truck if onsite taggant OTA believes that the further development
reading was considered desirable. and testing that would be required before an
identification taggant program couId be imple-
BATF maintains that, on the contrary, the 3M
mented are likely to take until 1983. If an iden-
identification taggant can be recovered and read
tification taggant program were legislated
in the field by investigators who have received a
early in 1980, it would be at least late 1984 be-
reasonable amount of training.
fore all commercial explosives could be manu-
Some industry representatives maintain that factured with taggants. Even if the sensor de-
there is considerable doubt as to whether tag- velopment and detection taggant programs are
gants would actually survive and be recovered successful, OTA feels it would be at least 1985
from a bomb. Such doubts should, they hold, before full implementation could occur. BATF
be cleared up before attempting to reach any maintains that these times are too pessimistic.

Given the present state of development of cision process to reside in the Treasury De-
taggants, OTA’S data and analysis appear to be partment.
consistent with any of three possible courses ● Defer legislative action on taggants, but en-
of action. (No significance is intended in order
courage BATF to continue taggant develop-
of listing. )
ment, with a view to consideration of legisla-
tion when development and testing are com-
● Pass legislation requiring taggants, and set up
a procedure to determine if and when the
technical development and testing have pro- ● Take no legislative action on taggants, and en-
gressed to a point where implementation courage the executive branch to search for
can begin. Given the active involvement of other ways of improvin g the effectiveness of
BATF in the development of taggants, it may law enforcement against terrorist and other
be inappropriate for the implementation de- criminal bombers.
Chapter II


Overview of the Problem . * . * . . * * . * . . . * * * . . . * * * * * * * * * * * * * * * * * * * * 19
The Bombing Threat. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......19
Manufacturer to User Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . .......21
Explosives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......21
Gunpowders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......22
TechnicalEfficacy ● *****.* ● *****9** ● ***.**** ● *******O .******** ● 24
Compatibility ofTaggant wtih Explosive Materials. . . . . . . . . .. . . .......28
Cost of aTaggant Program * * * * * * * *
● ● ******** * * * * . . . * . * * 32

Utility of Taggants . . . . . . . . . . . . . . * * * . * * * * * *
● * * * * * * **.**.**
● 36 ●

Terrorists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......36
Common Criminals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......37
Mentally Disturbed.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......37
Vandals and Experimenters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....38
program implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. ....46


6. Minimum Bombing Incidents Statistics Summary . . . . . . . . . .......20
7. Identified Explosive Fillers Used in Bombs. . . . . . . . . . . . . . . .......20
8. BombingCasualtiesand Damage in 1978 byTypeof Bomb . . .......21
9, BATF Recovery Demonstrations. . . . . . . . . . . . . . . . . . . . . . . .......25
10. Recovery Tests Participated in by Summit County Sheriff’s Office ...25
11. OTA Recovery Test Results. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..26
12. Elements of a Suggested Compatibility Qualification Program ......29
13. Cost of a Taggant Program as a Function of implementation Plan. ...32
14. Attributes of Criminal Bomber Croups. . . . . . . . . . . . . . . . . . .......37
15. Taggant Utility Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .....38



I. Explosive Distribution Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....22

2. Gunpowder Distribution Chain. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23
3. Photo micrographs of Recovered Taggants. . . . . . . . . . . . . . . . .......26
4. Size Comparison of the 3M Identification Taggantand
Some Smokeless Powders. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .....41
S. Marginal Cost-Utility Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....47
Chapter II

This chapter presents the findings of the study in some detail, along with a
sketch of the data and analytical methods used to arrive at them. The full analyses
on which these findings are based are found in the subsequent chapters and the ap-
The analysis proceeded in two stages, which were conducted simultaneously.
The first stage assessed the technical efficacy of the taggants, and their compatibili-
ty with explosive materials. Definitive judgments on these points must await the re-
sults of further technical development and testing. The second stage estimated the
cost and law enforcement utility of taggants, assuming that taggants can be made
which work and are safe. It should be clearly understood that a taggant program is
only appropriate if all the conditions are met: it must be technically sound, it must
be safe, it must have value for law enforcement, and the costs must be reasonable in
the light of this law enforcement value.
The analysis and discussion of technical efficacy and safety were conducted as
if it had been established that taggants are useful in relation to their cost. The analy-
sis and discussion of cost and utility were conducted as if it had been established
that taggants work and are safe.
Because a variety of implementation plans are possible, costs and utility are eval-
uated parametrically in order to show how the choices made in writing regulations
would lead to variations in cost and law enforcement value.


In order to appreciate the potential benefits of the data, and the lack of updating proce-
and shortfalls of a tagging program it is neces- dures, make accurate analysis difficult. Appen-
sary to understand the magnitude of the cur- dix F explains in some detail which data
rent and projected future bombing threat, as sources were used, and why. While BATF and
well as the processes involved in the manufac- FBI data differ in the absolute values (e. g.,
ture, distribution, and sale of the various ex- number of bombings in a year), both sets of
plosive materials. data support the OTA findings. Most tables in
this report make use of BATF data because its
format appeared more amenable to analysis.
The Bombing Threat
The BATF 1978 Explosives Incidents Report
Both the Federal Bureau of Investigation includes over 3,000 incidents for both 1977 and
(FBI) and the Bureau of Alcohol, Tobacco, and 1978. The incidents include accidents, threats,
Firearms (BATF) maintain national bombing seized and recovered explosives, and hoaxes,
data information centers which collect statis- as well as actual explosive and incendiary
tics on bombings and other explosive inci- bombings. Of these incidents, 1,377 repre-
dents. The data are not consistent between the sented actual explosive detonations, acciden-
two centers, however, and many bombings are tal detonations by criminals, or recovered
not reported to either center. The formatting bombs that failed to detonate in 1977, with

20 ● Taggants in Explosives

1,250 the corresponding number for 1978. At base. Management Sciences Associates (MSA)
least 953 of these in 1977 and 787 in 1978 rep- conducted a detailed study of the data in the 5
resent actual detonation of explosive bombs years from 1972 through 1976 without discov-
against substantial targets (mailbox and open- ering any significant trends. Many experts on
area bombings are not included). terrorism believe that the United States may
experience an increase in bombings, particu-
During 1977, BATF estimates that 38 people
larly catastrophic bombings, in the years
were killed and 180 wounded by explosive and
ahead. However, this belief is based on an as-
incendiary bombs, while the numbers in 1978
sessment of U.S. vulnerability to bombings and
were 23 and 185, respectively. Due to the way
the observation that the United States has
initial estimates of property damage are made,
recently had less of a terrorist problem than
and the lack of updating, only the crudest
other developed countries; there is no evi-
property damage estimates can be made.
dence that this increased threat has material-
There was at least $10 million in direct proper-
ized. In looking at bombing statistics, one
ty damage due to explosive and incendiary
should bear in mind that a single incident in-
bombs in 1977, and at least$17 million in 1978.
volving an aircraft exploding in flight could
In 1977, 35 of the 38 reported deaths and 20
produce more deaths than have occurred in
of 23 reported in 1978 were from bombings
any year to date.
against vehicles, residences, and commercial
establishments. Similarly, about 80 percent of Data on the types of fillers used in bombs
the injuries from bombing of known targets in are also not consistent between the FBI and
1977 and about 70 percent in 1978 were caused the BATF data banks. It is instructive to look at
by bombings of those three types of targets. two BATF data sources, however, as shown in
The 1977 and 1978 statistics are summarized in table 7. The second column represents 1978
table 6, and discussed in more detail in appen- data for the fillers identified in the field for all
dix F. explosive bombs that were detonated, bombs
recovered undetonated, and criminal acci-
The available data do not sustain any con-
dents. The first column represents 1978 data
clusions about trends in the bombing threat;
for only those fillers that were identified in the
both the number of incidents and the extent of
laboratory from postdetonation analysis. The
deaths, injuries, and property damage vary
third column averages the first two. In both
from year to year, and from data base to data
cases, black and smokeless powders and cap-
sensitive high explosives all occur with high
Table 6.–Minimum Bombing Incidents Statistics Summary a frequency. Table 8 shows a breakout of the
minimum number of significant bombing inci-
BATF FBI dents, deaths, and injuries occurring during
Hem 1977 1978 1977 1978 1978 by explosive material fillers. The average
Explosive bombings, number. . . . . . . . . 1,037b 896b 867 768 column in table 7 was multiplied by data on
Undetonated explosive bombs, number. 319 287 118 105
Incendiary bombings, number . . 339 446 248 349
Unignited incendiary bombs, number 81 71 85 79
Criminal accidents, numberc . . . . . 21 67 – –
Table 7.–ldentified Explosive Fillers Used in Bombs
Property damage from bombings, millions
of dollarsc d . . . . . . . . . . . ... $ 10 $ 17 $ 9 $ 9
Injuries c ., . . . . . . . . . . . . . . . . . . 180 185 162 135 Lab identified All Identified
People killed by bombingsc . . . . . 38 23 22 18 fillers 1978 fillers 1978 Average
Black powder ., . 13% 21% 17!40
aBATF reported 3,177 total incidents in 1977 and 3,256 in 1978 Total incidents include ac- Smokeless powder . 16 19 17.5
cidents threats, seized and recovered explosives and hoaxes as well as actual explosive and in-
cendiary bombings The OTA study was concerned only with explosive bombings
Military ., ., ., ... 2 7 4.5
b of these 953 in 1977 and 787 m 1978 were against substantial targets Cap sensitive ., ., 32 30 31
includes both explosive and incendiary bombings OTA was unable obtain separate figures for Blasting agents, – 1 .5
number of criminal accidents, injuries deaths, and property damage caused by explosive and in- Chemicals . . – 1 .5
cendiary bombings Incendiary bombs and bombings would not be affected by a taggant pro-
Others, ... ., 36 21 285
gram value probaly Considerably higher due to lack Of data file updates
See app F for derivation of these numbers
SOURCE : BATF 1978 Explosive Incidents Report, FBI Uniform Crime Report: Bomb Report
1978 See app F for a discussion of the derivation of these figures SOURCE BATF data
Ch. n-Detailed Findings ● 21

Table 8.–Bombing Casualties and Damage in 1978 by Typo of Bomb

Number of
bombings against Property damage
Filler material substantial targets Deaths Injuries $ millions’
All fillers. ., 1,298 23 185 $17.2
I n c e n d i a r y . . . . , 428 3 13 3.7
Black powder 148 4 19 .2
Smokeless powder : : : : : : : 152 3 23 .2
Military explosives. 39 0 7 —
Cap sensitive. . . . . . . . . . . . 270 7 26 3.3
Other . . . . . 3 40 2.4
Unknown ., ., 3 57 7.4
Total for those fillers which
would be directly tagged b. 570 14 68 3,7
Value probably higher due to lack 01 data update
b cap-sensitive explosives black powder and smokeless Powder would be tagged
SOURCE BATF data See app F for a derivation of these figures

total bombing to generate the table 8 esti- packaged before another batch is started on
mates. See appendix F for details. that production line. In a semicontinuous proc-
ess, the mixed batch is fed into an intermediate
M a n u f a c t u r e r to User Chain hopper from which packaging takes place,
while another batch is mixed in parallel to the
Explosives packaging of the first batch. In a continuous
process, the material is continuously added to
Approximately 4 billion lb of explosives are
the mixer, processed, and packed in a con-
manufacturered and used annually in the
tinuous flow.
United States. Of this amount, approximately
600 million lb are standard explosives and 3.4 If taggants were added to standard explo-
billion lb are blasting agents, primarily am- sives, they would be added at the mixing stage.
monium nitrate-fuel oil mixtures. Of the 600 Taggants could also be added to packaged or
million lb of standard explosives, about half bulk form manufactured blasting agents at the
are cap-sensitive (will reliably be detonated by mixing stage. If the ammonium nitrate used to
a #8 detonator) dynamites, emulsions, gels, make onsite-fabricated blasting agents were to
and slurries, and about half are non-cap-sensi- be tagged, identification taggants could be
tive gels, slurries, and emulsions. Most of the added during the “prilling” process, while de-
standard explosives are manufactured in a tection taggants, which are not batch specific,
plant, packaged in cartridges, and shipped, could be added with the fuel oil.
either directly to a large user such as a coal
Boosters are generally fabricated by pouring
mine or to a distributor, although some are
a molten, high-energy, cap-sensitive explosive,
processed essentially onsite. Some of the blast-
such as TNT, into containers. Taggants could
ing agent products are prepared by a manufac-
be added during the cooling process of the ex-
turer and sold in packages, some are prepared
by a manufacturer and sold in bulk [tanker
truck), while some are mixed onsite and used Detonators and detonating cord are manu-
the same day they are prepared. factured products in which the product is built
up around an explosive core in an assembly-
Standard explosives are made by mixing to-
Iine process. In both cases, the taggants would
gether the fuel and oxidizer ingredient and
be added during the assembly process, rather
feeding the mixed product into the final car-
than directly to the explosives.
tridges by a batch, semicontinuous, or continu-
ous process. I n a batch process, the ingredients All of the products have a similar flow from
for a particular batch are first mixed and then manufacturer to ultimate user, as shown in fig-

12 ● Taggants in Explosives

ure 1. Some of the products are sold directly those of explosives. Approximately 2 1/2 million
by the manufacturers to large users, such as a lb of black powder and 20 million lb of smoke-
less powder are produced for commercial use
Figure 1.— Explosive Distribution Chain each year. Most of the smokeless powder is
used in fixed ammunition for rifles, pistols, and

I Manufacturer
shotguns, would not be sold to users as an end
product, and would not be tagged under S.
333. Approximately 5 million lb per year would
I 1 be sold to the end user, primarily for handload-
I Distributor I ing of ammunition. Of the black powder pro-
1 I duction, approximately 2 million lb are used as
an intermediate product in the manufacture of
. 1
fuzes and other finished products and would
Retailer not be tagged; approximately 400,000 lb per
b year are sold for use in muzzle-loading guns
and would be tagged if a taggant program
- User User m were legislated.

The basic process for the manufacture of

gunpowders involves the following steps:

mixture of ingredients, which may include

the raw ingredients as well as surplus and
reworked powders;
granulation, where the “dough” is ex-
truded, chopped, or otherwise granulated
SOURCE: Off Ice of Technology Assessment
to form the various grains;
mine or large construction company. Such screening of grains into designated sizes;
sales may represent an entire day’s production. and
The rest is sold to distributors, who may buy blending of various batches to get the de-
portions of several production batches, entire sired ballistic characteristics.
batches, or even several batches. The distribu- In the smokeless powder manufacturing
tors in turn sell to retail stores, supply explo- process, nitroglycerine, nitrocellulose, and
sives directly to some users (such as a quarry or other additives are combined to make various
construction site), and may also do explosive grades before the blending process. Smokeless
contracting themselves. A particular uniquely powder grades therefore differ due to size dif-
tagged batch of explosives may, therefore, go ferences and composition differences (various
directly to one user, may go to one distributor, amounts of nitroglycerine), while black pow-
or may be sold to a number of users and dis- der and black powder substitutes such as Pyro-
tributors. From the distributor it may again go d e x@ * vary only by grain size. In a given grade
to one of several users, sometimes with a fur- of powder, variations in density and other fluc-
ther distribution level (retailer) involved. A list tuations during the manufacturing process can
of the ultimate purchasers of one specific cause considerable variations in the ballistic
batch of explosives could, therefore, contain properties of the final powder. As the hand-
one name, or up to a hundred names for a Ioader generally has no means of controlling
worst case example, although generally the his ballistics other than the weight or volume
number would be at the low end of that range. of powder added, the ballistic properties of a
particular grade of powder must be carefully
Gun powders controlled by blending. A given brand name
The manufacture and distribution processes
for gun powders are significantly different from *A registered trademark of Pyrodex Co
Ch. n-Detailed Findings “ 23

product may therefore contain parts of several stage might be necessary. For some products, it
batches, blended like brandy to give the de- may be possible to add taggants to the dough,
sired ballistic properties. Several sequential although this may affect the granulation proc-
blending operations may be necessary before ess and present blending problems.
the product meets the required specifications.
If the ballistic properties of a particular batch
The distribution network from gunpowder
or blended lot are too far off, the material
manufacturers to users differs markedly from
must be reprocessed or used for something
that of explosives, since there is a very large
other than hand loading.
number of ultimate users, each of whom con-
If taggants are added to gunpowders, they sumes a smalI amount of powder. The network
may have to be added at different stages in the is shown schematicalIy in figure 2. The manu-
manufacturing process for different manufac- facturer has several master distributors, each
turers, due to the differences in blending and of whom supplies a number of distributors.
reworking processes. As an example, at one Each distributor supplies a number of retailers,
smokeless powder factory that makes powder who sell the product, often in lots as small as 1
for both handloading and fixed ammunition, lb. A 2,000-lb uniquely tagged batch of prod-
taggants could be added during the blending uct “A” may therefore ultimately be sold to
stage; blended batches that were still not over a thousand customers. Not only does this
satisfactory could be used for fixed ammuni- produce a much larger list of last legal pur-
tion. At another factory, due to their large chasers, but considerably more record keeping
rework factor, an additional taggant-mixing would be involved at the retail level.

Figure 2.—Gunpowder Distribution Chain


u Master

u Master

SOURCE Off Ice of Technology Assessment

24 ● Taggants in Explosives

The issues to be addressed here include the When conditions are less than ideal, survival
survivability of the identification taggants and decreases. The number of surviving taggants
the status of the detection taggant materials decreases sharply as the size of the charge in-
and sensors. A detailed discussion of the re- creases, although sufficient taggants have
search program related to technical efficacy is been recovered even from a 25-Ib Power
in chapter I I 1; chapter IV discusses in detail the Primer charge to establish a definite identifica-
research related to safety. tion. The number of taggants also decreases if
the explosive is confined, for example, in a
The identification taggants developed by 3M
pipe bomb. Hundreds of taggants survive a
appear to survive the detonation of commercial
black powder pipe bomb; tens of taggants
explosives under ideal conditions. Confinement
have been recovered, under nonideal recovery
and fire may adversely affect survival, although
conditions, f r o m s m o k e l e s s p o w d e r p i p e
the test data are very limited. Recovery of the
bombs. Only one test seems to have been con-
taggants appears to be a function of the specific
ducted with cap-sensitive high explosive in a
conditions in which the explosion and taggant re-
pipe bomb; scores of taggants were recovered
covery take place, as well as the training of the
from a pipe bomb filled with 60-percent Extra,
field and laboratory investigators.
a low-energy explosive.
A large number of laboratory survival tests
The recoverability of the taggants under
have been conducted to establish the postdet-
real-world conditions is less well-established.
onation survivability of the 3M identification
The vast majority of the tests of recovery have
taggants. In many of these tests, the chamber
been demonstrations and training exercises,
used to recover the taggants was not ideal, re-
with IittIe attempt at scientific controls, pro-
sulting in low r e c o v e r y r a t e s . F o r e x a m p l e ,
cedures, or documentation. Table 9 shows the
when relatively small steel-walled chambers
results of 10 demonstrations using explosives
were used, the impacting taggants either broke
tagged during the manufacturing process with
up upon impact, or flowed plastically due to
encapsulated taggants at a 0.05 percent by
the impact pressure pulse. When the explosive
weight tagging level. The number of taggants
charges were detonated in large chambers, or
recovered is shown in each case; in some cases
on a large open pad, however, several hundred
heroic recovery efforts were required. Statisti-
tags were recovered from a single, one-half-lb
cal analysis by the Aerospace Corp. indicates
stick of the cap-sensitive explosives, including
that it is highly desirable to recover 20 tag-
Atlas Power Primer, the most energetic of the
gants; that many were not recovered in each
standard commercial explosives. Similarly, the
case. In some tests, particularly the last one,
taggants should survive the detonation of
recovery was halted after the reported number
black and smokeless powders, which have
was found. Table 10 shows the results of 14
much lower energy than the more energetic ex-
similar tests, conducted without the assistance
plosives, under idea I conditions. The indi-
of the Aerospace Corp. and the BATF labora-
vidual taggants are not expected to survive the
tory team. These tests were significantly less
detonation of high-energy explosives, such as
the TNT used in boosters or military explo-
s ives; Aerospace Corp. calculations have Due to the apparent inconsistency of the
shown that the taggant material would be test results and the lack of documentation,
raised above the taggant decomposition tem- OTA had a limited series of five recovery tests
perature in these explosives. Survival in these conducted. The purpose was twofold: to get a
energetic explosives has been demonstrated feel for the recovery process and its difficul-
when the taggants are pressed into large pel- ties, and to generate a limited number of data
lets (one-fourth inch), but no definitive re- points for which the testing, recovery, and
covery testing has been conducted. anaIysis w e r e w e l l c o n t r o l l e d a n d d o c u -
Ch. n-Detailed Findings 25

Table 9.–BATF Recovery Demonstrations

Place Time Target Explosive Test conditions Taggant recovery

Birmingaham, Ala. February 1977 Car 1 1/2 -lb Power Primer Against engine, fire, firefighting 35 from soil sample in laboratory
House 11/2 -lb Coalite-8S Table, near front hall Hundreds, at scene
House j/4-lb, 60% Extra in pipe Outside house, near wall Scores, at scene
Donaldson, Pa March 1977 Borehole in 101/2 -lb Coalite-8S 7 each, 1 I/2-lb packages in 20 from coal in laboratory
coal mine separate boreholes
Seneca, Md. June 1977 House 2-lb Coalite-8S Exterior room Dozens at scene
Car 2-lb Coalite-8S Passenger compartment Few at scene

Fort McArthur, Calif, November 1977 House 1/2 -lb Powerdyne Many at scene
Los Angeles, Calif. August 1978 Open 1-lb Powerdyne In suitcase 20 at scene
Otis AFB, Mass. October 1978 Open 1-lb. Tovex 220a Three shots, 1 lb each Less than 10
Fort Belvoir, Va March 1979 Car 2-lb. Coalite Z Trunk 3 in field
a Undetonted stick had only 10 percent of expected taggants Data Indicates that this explosive was from end of a batch
SOURCE Office of Technology Assessment

Table 10.–Recovery Tests Participated in by Summit County (Ohio) Sheriff’s Office

Date Explosives Target Conditions Recovery results

May 2, 1978 Total of 41/z-lb permissibles Two cars, — 2-hour field search (night), 10 men, 4 taggants in one car,
ground no tags from other targets.
May 11, 1978 2-lb permissibles Car — 2-hour field search (night) by 2 men. No taggants
May 17, 1978 3-lb permissibles, 1 black 3 cars, pipe 1 car fire 1-hour field search (daylight with blankets). No taggants.
powder pipe bomb bomb in open
Oct 12, 1978 2-lb permissibles Car — 2-hour field search (night). by 2 men. No taggants.
May 16, 1979 1/2 permission Car — 1 I/2 -hour field search (daylight with blankets), 20 men.
No taggants
May 17, 1979 2-lb permissibles Car — 2-hour field search (night) by 2 men. No taggants
Aug 14, 1979 2-lb water gel Car Under driver seat 3-hour field search (dark), 6 men Found 3 taggants from water
1 3/4-lb gelatine dynamite Car Under driver seat gel. Laboratory analysis of 60-lb debris from each car Found
2-lb permissible Car Under driver seat 5 more taggants from water gel.

SOURCE Office of Technology Assessment

mented. The results of the tests are summar- conditions were not ideal. Field recovery
ized in table 11 and described in detail in ap- and identification of the taggants may be
pendix C. Sample photomicrographs of recov- more Iikely on paved surfaces.
ered taggants are shown in figure 3. Although 2. Under ideal conditions (no fire, subse-
these tests were extremely limited in scope, quent firefighting activities, or adverse
and covered only one type of target (automo- weather), sufficient debris can be gath-
bile), they provided a great deal of insight into ered in a short time (less than 1 hour) by
the recovery process and suggest a reconcilia- an untrained team to produce a positive
tion of the prior test results. However, a full- taggant identification (more than 20 tag-
scale test program must be completed before a gants) in the laboratory. Only a moderate
definitive assessment of taggant recovery is (1 to 2 hour) laboratory effort is necessary
possible. With that caveat, the foIlowing tenta- by a highly trained laboratory team to iso-
tive observations may be made: late and identify the taggants. This prob-
ably holds for all classes of unconfined
1. The recovery process does not appear to commercial explosives (excluding very
be a field-readable process under the high-energy explosives such as boosters or
tested conditions. No taggants were spot- military explosives). The laboratory need
ted, and identified as such, in any of the not be elaborate and could well be trans-
five tests, under daylight or night condi- portable to the bombing site.
tions, without the use of a laboratory sep- 3. Under conditions of confinement (bomb
aration procedure. However, the recovery placed between the engine block and the
26 ● Taggants in Explosives

Table 11 .–OTA Recovery Test Resutts

Target Placement Dynamite Test condition Taggant recovery

Auto Under driver seat 2-lb Collier C 5-gal gas in tank; no fire 28 taggants in 1 M-hour lab time
Auto Under driver seat 2-lb Unigel 5-gal gas in tank; no fire 23 taggants, 1 contaminant in %-hour lab time
Auto Under driver seat 2-lb Power Primer 5-gal gas in tank; no fire 21 taggants in 1 1/2-hour lab time:
12 of type A
oftype ~ d u a l tagged
Auto Under driver seat 2-lb Collier C 1-gal gas adjacent to bomb, 23 taggants in3-hour lab time
fire, firefighting
Pickup Between engine 2-lb Power Primer Dry tank, no fire 26 taggants, plus one contaminant in 4 hours lab time, 5-hour
and firewall induction time preceded the search time due to confusion caused
by equipment contamination,
SOURCE Off Ice of Technology Assessment

Figure 3.—Photomicrographs of Recovered Taggants


\ \
a) Scale— 1 division = 1 mm b) Collier C under driver, 8 taggants
on slide

e) Collier C under driver, 17 taggants

on slide
Photo credits’ U.S Department of the Treasury

firewall), sufficient taggants can still be 4. Taggants can be recovered from an auto-
recovered for a confirmed identification, mobile bombing with a low-power explo-
although somewhat more effort is prob- sive, even after a gasoline fire and subse-
ably necessary, both in the field and in the quent firefighting efforts. Tests would be
laboratory. This tentative conclusion necessary to determine if taggants would
would hold for all cap-sensitive commer- survive a postdetonation fire in conjunc-
cial explosives (excluding boosters and tion with a more energetic explosive. It
miIitary explosives). should be noted that no fire occurred in
Ch. n-Detailed Findings 27

the three tests in which gasoline was airport environment has indicated that the
placed in the gas tank. Fire had to be spe- spectrometer can differentiate molecules of
cifically induced (a gallon of gasoline was mass similar to the vapor taggants from the
placed adjacent to the bomb) for the burn ambient environment. Similarly, laboratory
test. testing of the continuous electron capture de-
5. The results of the automobile tests may tector has indicated its ability to discriminate
well be generalizable to other test condi- taggant-like molecules.
tions (buildings, open areas), but testing
These limited tests, however, are a long way
would be required before that claim could
from demonstrating that the sensors can distin-
be made.
guish the specified vapor taggant species from
6. No substantive recovery data are avail a-
other molecules, particularly those in the same
ble for large charges, explosives in pipe
mass range. The ion mobiIity spectrometer and
bombs, tagged boosters, detonators or
mass spectrometer have an active separation
detonating cord, or charges consisting of
mechanism to preclude interference with mol-
an untagged blasting agent with a tagged
ecules that differ significantly in mass; the
booster and detonator. Taggants were re-
continuous electron capture spectrometer
ported recovered from a large bomb con-
must rely on a far less reliable passive breakup
sisting of an untagged blasting agent and
a tagged booster, conducted in December
1979, but the test specifics have not yet No estimates have been made of the time re-
been examined by OTA. quired to produce fielded units, once a feasi-
bility demonstration has been made (none of
The technology for detection sensors has been
the three candidates has yet progressed that
demonstrated in the laboratory, but at least
far). The only time estimate so far made is an
several years of development would be neces-
estimate by the Aerospace Corp. that it would
sary before field models would be available.
take 14 months from demonstration of feasibil-
Three types of sensors are being considered for
ity to the completion of the prototype stage
use with the microencapsulated vapor detec-
for the ion mobility spectrometer. This esti-
tion taggants. Each type is capable of sensing,
mate is quite optimistic for an instrument that
under properly controlled conditions, in the
would be produced in large numbers by a
parts-per-trillion regime envisioned for the sys-
small company. OTA feels it would be at least
tem. The mass spectrometer sensor is a simpli-
3 years, and probably more like 5, before a tag-
fied version of a standard laboratory instru-
gant sensor could be fielded. The estimate is
ment. The spectrometer, however, must be cal-
based on generalizing from other commercial
ibrated regularly, requires ski lied scientists to
and military instrument development exper-
operate and maintain it, is large, and is quite
expensive. The ion mobility spectrometer has
been commercially available for approximate- The candidate detection taggant vapors ap-
ly 5 years, with approximately 50 machines be- pear promising, but more research is necessary.
ing used in laboratory analyses. It shares the Several hundred candidate chemicals have
laboratory instrument characteristics of the been screened in a search for a vapor that ex-
mass spectrometer. The continuous electron hibits the desired properties of scarcity in
capture detector has been produced as a labo- nature, long-term stability, chemical inertness,
ratory instrument, but in limited numbers. Lab- vapor pressure, penetration, and nonadhesion
oratory and controlled-environment testing to surfaces likely to be present in containers
with the three types of instruments has shown used to conceal bombs. The five candidate
promising results. For example, a less sensitive perfluorinated cycloalphones appear promis-
mass spectrometer is currently operating in an ing on the basis of early tests. (No long-term
online process mode at Libby-Owens-Ford, stability data are available, however, nor are
maintained by regular maintenance personnel. there data on the long-term stability of the dif-
Testing of the ion mobility spectrometer in an fusion rate through the encapsulating materi-
28 . Taggants in Explosives

al). Additional problems, such as ease of manu- ficult to manufacture, require highly special-
facture, specificity with respect to the de- ized equipment, a n d w o u l d b e h a r d f o r
tector, and compatibility, have not yet been bombers to make or acquire for use as counter-
addressed. Ease of manufacture is a double- measures. Once the equipment is operational,
edged problem — if manufacture is too diffi- unit costs should not be unreasonable. A prob-
cult, then costs will be high; if it is too easy, lem which probably applies to all varieties of
then illegally manufactured material can be vapor taggants is that seals can be made that
used as a countermeasure to the detection sen- are taggant proof — although apparently com-
sors. The most promising candidates are dif - mon seals are insufficient.


The compatibility of explosive materials dence of any decreased stability, or other sig-
with the specific identification and detection nificant changes, was found in any of the tests
taggant materials is addressed here. Compati- with dynamites, gels, slurries, or black powder.
bility has two connotations: the first concerns
The tests with tagged cast booster materials
the safety during manufacture, transportation,
showed some indications of instability at ele-
storage, and use of explosive material due to
vated temperatures. A mixture of RDX and
the addition of taggants; the second concerns
TNT (Composition B) showed evidence of reac-
changes in the performance of the explosive
tion and probable decomposition at tempera-
materials to which taggants have been added.
tures of 120° C when taggants were added to
Such compatibility must be demonstrated by
the booster mix; significantly less reaction oc-
specific tests. Generalization of the results to
curred without taggants. Tests with Octol
other hypothetical taggants is hazardous at
showed little reaction whether taggants were
present or not. Pentolite showed little evi-
Safety tests conducted to date with the encap- dence of reaction with taggants in one test at
sulated 3M identification taggants have shown no 1 20 0 C; the gas evolution from untagged pen-
incompatibilities with dynamites, gels, slurries, tolite was too high for comparative testing on
emulsions, or black powder, allowing a presump- a second series.
tion that comprehensive testing would show that
Similarly, the stability of one type of Hercu-
these taggants are compatible with these explo-
les smokeless powder has been shown to be
sives. High concentrations of taggants do react
significantly decreased by the addition of the
with one kind of smokeless powder and one type
3M identification taggants at elevated temper-
of cast booster material at elevated tempera-
atures and taggant concentrations. (Although
tures, and consequently incompatibility must be
Hercules tested only Herco * powder, Her-
presumed pending further research. A large
cules believes that their other brands of pow-
number of paired safety tests have been con-
der designed for the reloading market are so
ducted comparing the sensitivity and stability
similar to Herco @ that similar test results could
of commercial explosives and gunpowders
be expected.) Tests were conducted at temper-
with and without identification taggants
atures ranging from 80° to 120° C and at tag-
added. Safety tests included mechanical im-
gant concentrations of 50 percent. Tests at the
pact, thermal stability, thermal impact, fric-
Lawrence Livermore Laboratories appear to in-
tion, electrical properties, and chemical reac-
dicate that the incompatibility is between
tivity, although no single explosive has been
some element of the powder and the basic
subjected to al I of the above tests. I n no case
melamine/alkyd material of the taggants,
did the addition of encapsulated taggants sig-
rather than with the encapsulant or a pigment.
nificantly increase the sensitivity of the explo-
sive materials to the test conditions. No evi- *A registered trademark of Hercules Inc.
Ch. n-Detailed Findings 29

Both the smokeless powder and booster ma- Accidents involving explosives can have ex-
terial tests took place at high temperatures, tremely severe consequences to these thou-
and, in most of the tests, at high taggant con- sands of people; therefore, safety must be
centrations. The temperature used for the demonstrated. A carefully administered quali-
smokeless powder test was higher than would fication program of analysis, safety testing,
be expected in actual manufacture, storage, or manufacturing procedures control, and experi-
use; the temperature used for the cast booster ence is necessary before a new explosive, or an
is sometimes reached in manufacturing proc- explosive with a significant change in compo-
esses. In each test, a taggant concentration of sition, can be considered safe. I n addition,
50 percent was used rather than the 0.05-per- each type of explosive product requires indi-
cent tagging concentration suggested for rou- vidual evaluation and testing, The type of
tine use. The tests, nonetheless, indicate that qualification program considered necessary
the stability of the materials has decreased, before safety can be demonstrated is shown in
due to the addition of taggants, and that a re- table 12 and discussed in detail in chapter IV.
action is taking place between elements of the A particularly important aspect of that qualifi-
taggant and elements of the explosive mate- cation testing is the effect of long-term stor-
rial. Standard qualification test procedures re- age.
quire that such evidence be considered a sign
While the qualification program outlined in
of an existing incompatibility between the ma-
table 12 must be performed before taggants
terials. Carefully controlled testing and exten-
sive analysis must be completed before it can
be determined if the observed evidence of in- Table 12.–Elemertts of a Suggested Compatibility
(qualification Program
compatibility does, in fact, indicate a potential
safety problem during the manufacture, stor- ● unique with each manufacturer
age, transportation, and use of the tested ma- ● analysis to define the new explosive or Ingredient
● laboratory testimg
terials. Unless demonstrated otherwise, it must —impact, friction, thermal, chemical composition
be assumed that it is unsafe to add the tag- –electrical, aging, chemical interactions, performance
● pilot production
gants to that smokeless powder or to the
committee and management review
booster material. Until the elements of the in- ● early production and review

compatibility have been identified, a question special tests

remains as to the safety of adding the taggants ● experience

to similar smokeless powders and booster ma- SOURCE Off Ice of Technology Assessment
terials, although tests with other smokeless
powders and boosters have shown no evidence can be safely added to explosive materials, the
of incompatibility. apparent incompatibility with the Herco”smoke-
The tests so far conducted are only a small
Iess powder must be resolved before it makes
fraction of the total number of tests that must
sense for the taggant compatibility qualification
be performed before it can conclusively be de-
program to proceed. Resolution of this problem
is pertinent for the entire identification tag-
termined whether taggants are compatible
gant program, not simply for smokeless pow-
with commercial explosives and gun powders.
ders or for Herco @ . As discussed in detail in
Even if the current question of the stability chapter Vl, smokeless powders are used in a
of smokeless powder and boosters is resolved, significant number of criminal bombing inci-
it is not possible to generalize from the results of dents and account for a significant fraction of
the limited test program so far completed and bombing casualties. If smokeless powders are
conclude that the testing has demonstrated that not controlled, then more bombers may well
taggants can be safely added to explosives. switch to their use, resulting in an even greater
Thousands of people come into contact with smokeless powder born bing problem. The reso-
explosives every day during the manufacture, lution could take any of several forms, includ-
storage, transportation, and use of explosives. i ng:

30 ● Taggants in E x p l o s i v e s

Reformulation of the 3M taggant — this composition rate was within the normally
could require starting essentially from accepted range for temperature regimes
scratch in the taggant testing program, as and concentrations which reflect worst
the reformulated taggant would undoubt- case actual use conditions, then it may be
edly exhibit different compatibility, as possible to add taggants to the smokeless
well as survivability, properties. powder, particularly if no further incom-
patibilities surface. Demonstration of
. Reformulation of the particular reactant
safety would have to be quite convincing,
smokeless powder— this may or may not
however, to overcome the currently per-
be easily accomplished, once the element
ceived incompatibility.
or elements that react with the taggant
are isolated. This option would be viable Similarly, the apparent incompatibility with
only if no other smokeless powders were one cast booster material should be resolved be-
found to show incompatibilities. fore the taggant compatibility qualification pro-
gram should proceed. Booster material is rarely
● Exe/us ion of the reacting smokeless pow-
used as a bomb filler, but it is used to initiate
waler from the taggant program– the eco-
blasting agents. The current BATF plan would
nomic effects on competition could need
be to not directly tag blasting agents, but to
to be carefully considered, as would alter-
tag the booster and detonators used to initiate
nate control mechanisms.
the blasting agent. Exclusion of boosters from
● Exclusion of smokeless powders from the the taggant program may well require an alter-
identification taggant program — such an nate control mechanism for blasting agents.
exclusion would rely on the fact that Given the extremely large quantity of blasting
smokeless powders would be less effec- agents produced (3.4 billion lb annually), any
tive than cap-sensitive high explosives and other control mechanism may have serious
that the detonators would be tagged. OTA cost consequences.
believes that this last approach may not
The limited number of tests conducted, the
be viable–too many people are currently
conditions under which some of the tests were
killed or injured by bombs using smoke-
conducted, and the preliminary manner in
less powders and the numbers would al-
which the tests have been reported, make it
most certainly increase if this approach
difficult to definitely assess the extent of the
were adopted. Alternate control mecha-
potential compatibility problem. If definitive
nisms for smokeless powders could also
test results do show an increased decomposi-
be adopted.
tion rate, at least for RDX/TNT explosive mate-
● Development of a different type of tag- rials, the incompatibility will have to be re-
g a n t f o r u s e w i t h H e r c o@ , or with all solved before those booster materials can be
smokeless powders, while retaining the ex- tagged. Most of the mechanisms for resolution
isting taggant for high explosives. This of the smokeless powder incompatibility are
would somewhat complicate field investi- applicable to booster materials, with the same
gation of bombings. consequences and caveats.

● Demonstration that the observed stability While the testing program conducted to
problem does not constitute a safety haz- date gives an indication that the identification
ard. The observed decreased stability oc- taggants may well be compatible with most
curs at elevated temperatures and taggant commercial explosives and gunpowders, little
concentrations 1,000 times greater than data are available as to the potential compatibili-
“normal. ” As the decomposition rate is ty of detection taggants with explosive materials.
both temperature and concentration sen- Compatibility testing with gunpowders and
sitive, it may be that no safety hazard ex- cap-sensitive high explosives has recently been
ists under realistic conditions. If it could initiated under contract to the Aerospace
be positively demonstrated that the de- Corp.; however, no compatibility testing has as
Ch. n-Detailed Findings . 31

yet been reported. As indicated above, each with gunpowders that differ in both size and
change to an explosive composition must be density from the taggants, the taggants and
evaluated separately. Successful completion powder fines tend to separate from the larger
of the preliminary detection compatibiIity pro- powder grains. Tests with smokeless powder
gram would indicate the need for a full qualifi- matched in size with the taggants, but differ-
cation program. As some compounding of sen- ent in density, were inconclusive. Testing is re-
sitivity may occur with both types of taggants quired to determine both the extent of segrega-
present, the full qualification testing program tion which could be expected if tagged gun-
should address that issue. powders went through extreme but plausible
conditions of transportation and storage, and
Compatibility testing includes performance
also the statistical probability that segregation
testing, as well as the safety testing discussed
to this degree would adversely affect ballistic
above. In most cases, the performance of ex-
performance or in-gun safety.
plosive materials is unlikely to be significantly af-
fected by the addition of small amounts of tag- The Winchester Western Division of the
gant materials. Performance proof-testing must Olin Corp. recently conducted a series of tests
be completed, however, before a definitive state- to evaluate the effects of segregation and high
ment could be made. The energy density and taggant concentration on the ignition proper-
rate of energy release are the two most impor- ties of smokeless powder. Significantly re-
tant performance attributes of commercial ex- duced ballistic performance was noted on one
plosives. Energy density is a fundamental round, fired at – 30 0 C with four times the sug-
chemical property of the explosive formula- gested taggant concentration. The other
tion. The rate of energy release is a function of rounds fired in this test series showed accept-
the materials involved and the physical prox- able performance (velocity, chamber pressure,
imity of the fuel and oxidizer components. The and ignition time).
presence of taggants, in the few hundreths-of-
a-percent by weight basis being considered, is Olin-Winchester conducted additional tests
unlikely to directly affect either of those per- using 100-percent segregation of taggants from
formance characteristics. Similarly, the pres- powder grains, a condition so extreme that no
ence of taggants in the suggested concentra- conclusions can be drawn (see ch. IV).
tion is unlikely to affect the ballistic properties OTA believes that although testing is indeed
of gunpowders. The few tests conducted s o required to establish the ballistic effects, if
far, including tests of the basic properties of any, of adding taggants to smokeless powder,
explosive materials, such as detonation veloci- it is necessary first to establish (by testing and
ty, cap sensitivity, chamber pressure, and pro- by statistical analysis) the extent to which
jectiIe velocity, support that conclusion. variation in taggant concentrations and segre-
Physical segregation of the taggants is one gation of taggants in normal conditions of
mechanism which could affect performance. If transportation and use could be expected.
the gunpowder grains segregate from the tag- Taggant clumping (10 to 15 taggants) some-
gant, then it is statistically possible that a times occurs when the taggants are added to
clump of taggants could cause uneven burn- explosive materials. I t i s u n l i k e l y t h a t t h e
ing, prevent ignition, or result in a hazardous
clumping would affect performance or safety,
hangfire condition. Similarly, in some specialty but that type of anomalous behavior should be
explosive products, such as shaped charges investigated, particularly as the physical chem-
used for oiI welI perforators, migration of the istry of some of the explosive products, partic-
taggants to the explosive-metal interface could ularly the gels and slurries, is so poorly under-
cause poor jet formation. Testing with gun- stood.
powder has shown that migration apparently
does occur, at least u,nder vibration conditions As for the possible performance degrada-
consistent with truck transportation. I n tests tions in shaped charges due to taggants, OTA
32 ● Taggants in Explosives

estimates, based on tests conducted by the could cause some degradation to occur, but it
U.S. Army Ballistics Research Laboratory, in- is difficult to envision a mechanism which
dicate that a clump as large as 0.02 inch would would allow that large a clump to accumulate,
not affect performance, even for precision- as that would represent all of the taggants in
shaped charges, unless the clump contained a approximately one-half lb of explosives.
large hollow center. Clumps as large as 0.1 inch


Estimates can be made of the total cost of a Table 13.–Cost of a Taggant Program as a Function
taggant program, the cost impact on manufac- of lmplemmtation Plan
turers and users of explosives, the effects of a Program level
legislated monopoly, and the possibility of Cost parameter Low Baseline High
added liability of manufacturers due to the in- Added cost per pound to cap-sensitive explosives 3.5$ 6.0$ 9.6c
clusion of taggants in explosives. In the above Added cost per pound to gunpowders . . 3,5c 65.8c $1.04
safety and efficacy discussion, the status of Public overhead cost, millions of dollars per year $5,3 $8.5 $24.5
Total program costs, millions of dollars per year $30.5 $45 $268
the current identification and detection tag-
gant systems was evaluated. In the following SOURCE Office of Technology Assessment

cost section, an assumption is made that the

taggants work and are safe, and cost estimates total program cost for separate implementa-
are generated parametrically as a function of tion of identification and detection taggant
the implementation plan. It is specifically as- programs is included in the discussion of each
sumed that the resolution of the smokeless case. The low, baseline, and high cost esti-
powder and booster material incompatibility mates do not correspond to different estimates
questions, and any subsequent questions of the same program; rather they refer to dif-
which may arise, do not have significant cost ferent tagging levels, different explosives tag-
impacts. I n the case of the smokeless powder ged, and different numbers of sensors. Chapter
and booster materials, this assumption is prob- V contains a detailed discussion of the cost
ably justified, as the cost of the taggant materi- estimates and a discussion of the sensitivity of
als represents only a small fraction of the total the costs to the accuracy of the cost element
cost added by a taggant program. estimates. To compare the program costs for a
constant number of detection taggant sensor
The primary finding of the cost analysis is
locations, it is only necessary to adjust the
that the cost of a taggant program can vary by
high- and low-program cost figure by $4,370 for
almost an order of magnitude, depending on the
each sensor deployed.
implementation plan. A baseline program is iden-
tified that would increase the cost of explosives 1. The low-level program would use a unique
and gunpowders to the ultimate user by approxi- identification taggant for each manufac-
mately 10 percent. The primary variables af- turer, type of product, and year of manu-
fecting the total program costs are the class of facture. A total of 800 detection sensors
explosive materials to be tagged, the uniquely would be deployed, one for passengers
tagged batch size, and the number of locations and one for baggage at each airport loca-
at which the detection sensors would be de- tion currently deploying magnetometers
ployed. Cost estimates for total program cost, and hand baggage X-ray units. Cap-sensi-
added cost per pound of explosive or gunpow- tive high explosives, detonators, boosters,
der, and public overhead costs are shown in detonating cord, and smokeless and black
table 13 for three implementation levels. The powders would be tagged with both iden-
cost estimates include the costs for both iden- tification and detection taggants. Blasting
tification and detection taggant programs. The agents would not be directly tagged. The
Ch. n-Detailed Findings 33

cost of separate low-level identification plosive materials, including blasting

and detection taggant programs would be agents, would be directly tagged. Am-
approximately $15 million and $22 mil- monium nitrate fabricated for use in blast-
lion, respectively. ing agents would be tagged, but not ferti-
lizer-grade ammonium nitrate. Approxi-
2. The baseline program would tag the same mately 5,000 detection taggant sensors
materials as the low-level program, but would be deployed at every major trans-
would use a unique identification taggant portation facility, controlled-access utili-
for each shift of each product –analo-
ty, Government facility, and other poten-
gous to the current date-shift code mark- tial high-value targets such as campus
ing on the exterior of explosives. Tracea-
computer locations. Portable units would
bility to the list of last legal purchasers
be routinely available to police bomb
would be maintained, as the taggants squads. The taggant level and types of ex-
would contain all the information needed plosives to be tagged in the high-level pro-
for a BATF trace (date, shift, product, and
gram correspond to a strict interpretation
size). Approximately 2,500 detection tag- of S. 333, as propounded by the Institute
gant sensors would be deployed at air- of Makers of Explosives (IME). The cost of
ports and major controlled-access facili- separate high-level identification and de-
ties such as powerplants, refineries, and tection taggant programs would be ap-
Government buildings. Major police proximately $214 million and $65 million,
bomb squads would operate portable respectively.
This baseline program differs from the The identification taggant cost figures used
program proposed by the BATF/Aero- in alI three levels of the analysis are based on
space Corp. team in only two respects. price estimates furnished by 3M, for specific
The most important is that a full shift of implementation guidelines. 3M furnished man-
the same product (a different cartridge agement-approved cost estimates for unencap-
size would be treated as a different prod- sulated taggants for three different quantities
uct) would be tagged with the same tag- of explosives to be tagged, assureing a firm
gant, rather than an arbitrary 10,000 to order for 2 years (costs would remain the same
20,000 lb. The practical utility result is for a 5-year contract). These cost estimates rep-
that a potentially longer list of last pur- resent the firmest figures possible short of an
chasers would be produced by a trace, at actual contract. Assuming linear interpolation
least for those lines that make more than between data points furnished, the unencapsu-
10,000 to 20,000 lb of a product in a single Iated taggants would cost between $93 and
shift. The second difference concerns re- $114/lb for the amount of taggants necessary
work. It has been assumed that a special for the baseline level case (41 9 million lb of ex-
taggant will be added to material with plosive equivalent). The first figure represents
more than 10-percent cross-containina- a cost goal and the second a worst case esti-
tion; such a taggant would indicate that mate. 3M technical people also furnished an
the material used was a composite and estimate of encapsulating cost, but were un-
that taggant codes other than the specific able to estimate the cost of the opaque encap-
sulation assumed by OTA as the baseline prod-
composite code should be ignored.
The cost of separate baseline identifica- uct. Based on the above data, OTA estimated
tion and detection taggant programs that it would cost approximately $55/lb for
would be approximately $25 million for opaque encapsulated taggants; as the baseline
each. tagging level is 0.05 percent by weight of en-
capsulated taggants, and the encapsulating
3. The high-level program would uniquely material weighs the same as the unencapsu-
tag each 10,000-lb batch of explosive and Iated taggants, this corresponds to 2.75 cents/-
each 2,000- I b batch of gunpowder. AlI ex- lb of cap-sensitive explosives for the identifica-
34 . Taggants in Explosives

tion tagging material ($93 for 1 lb of unencap- ticular mining operation uneconomic. Similar-
sulated taggants plus $17 for 1 lb of encapsu- ly, that type of increase in the cost of cap-
lating material plus the process equals $110 for sensitive explosives, boosters, detonators, and
2 lb of encapsulated taggants, or $55/l b.) OTA detonating cord in a large, open pit copper
estimated the same cost for taggants at the mine would increase the cost of producing
other two implementation levels. Chapter V in- copper only 0.03 percent. As blasting agents
cludes an analysis of how changes in the cost are currently used whenever possible in that
and/or concentration of the taggants them- mine (cap-sensitive explosives are used only
selves wouId affect the cost of the program. for secondary breakup), no shift in explosive
products used would take place. The cost of a
All other cost figures are estimates based on
recent explosive-intensive dam construction
specific inputs submitted to OTA by manufac-
project would increase 1 percent under the
turers, distributors, and end users. Detailed
baseline program, a larger percentage, but not
treatment of the cost elements is contained in
enough to be significant or force alternate
chapter V.
uses. A price differential of approximately
The cost impact to end users of explosives five-to-one currently exists in favor of blasting
can be considerable. Implementation plans agents over cap-sensitive high explosives,
that do not take into account the impact on which has caused most users of explosive ma-
manufacturers and users of explosives could terials to consider blasting agents, and shift
drive a number of manufacturers and users out where feasible; an increase in that differential
of the market; could make some classes of to six-to-one is unlikely to significantly change
finished products, like copper, uncompetitive the current status.
in the world market; and could force entire
segments of industries to radically change As a final example, consider the cost impact
operating procedures, such as shifting under- on handloaders. Handloaders load their own
ground coal mining from explosive mining to ammunition for two reasons —economy and
mining machines. Detailed discussions and the hobby aspect. A less-than-l O-percent cost
analysis, however, indicate that it is quite un- increase in expendable material is unlikely to
likely that a taggant program similar to the affect a hobby for which hundreds of dollars in
“baseline” would eliminate any current uses of costs have already been incurred (hand loading
explosive materials, although marginal com- equipment and guns). As powder is only one of
panies and product lines might be eliminated. As several materials on which a handloader saves
indicated above, the baseline program differs costs (cartridge cases, projectiles, wadding),
from the BATF-proposed implementation only and additional cost-savings are realized from
in that batch size takes into account the nor- labor and by eliminating the excise tax on pur-
mal production processes and quantities of the chased ammunition, an 8-percent increase in
explosives and gunpowder manufacturers. This powder cost would translate into an even
finding is based on detailed discussions with a smaller increase in total reloading costs. It is
limited number of users and manufacturers possible, however, that manufacturers would
about current costs and the possible impact of shrink the range of available product lines in
cost increases. order to minimize the startup costs of tagging.
A smaller choice of products would bean addi-
Some examples are illustrative. Increasing tional “cost” to the handloader.
the cost of cap-sensitive high explosives the 12
percent projected would increase the cost of The identification taggants currently pro-
extracting coal in a particular modern under- posed to be used are manufactured only by 3M
ground mine by only 0.1 percent. Such a small and are a proprietary product manufactured
increase would not be significant to this inten- by a proprietary process. In addition, a signifi-
sive user of cap-sensitive explosives, and cant public overhead cost would have been in-
would be quite unlikely to cause a shift to curred before the compatibility of explosive
mechanical mining machines or render a par- materials with the taggants could have been
Ch. n-Detailed Findings ● 35

demonstrated. Mandating the addition of iden- dressed by this study; if a multiyear contract is
tification taggants to explosive materials an acceptable mechanism, there may be some
would, therefore, ensure a monopoly of the advantage to a single contracting agency (pre-
Government-mandated market for 3M, at least sumably within the Government), rather than
for a period of several years. Under such cir- separate contracts with each manufacturer of
cumstances, development of a mechanism to explosives and gunpowders. I n addition to sav-
regulate the virtual monopoly of the identifica- ing the cost of multiple contracting, the single
tion taggant market which 3M would enjoy is contract concept would Iimit the amount of in-
highly desirable. formation on numbers of product lines and
production quantities of explosives available
A number of mechanisms are available to
to 3M, a matter of some sensitivity to the ex-
regulate the price of taggants, including:
plosive manufacturers.
1. a price level set by Congress in the en-
abling legislation, A final cost-related issue merits attention.
2. regulation as a public utility, The legislation of a taggant program might
3. licensing by 3M of competitors, change the extent to which manufacturers are
4. a multiyear, fixed-price contract, and held liable for accidental explosions. In the event
5. a free-market price, regulated only by the that an accidental explosion takes place, those
possibility of competition or sanctions if injured may attempt to hold the manufacturer
prices get too high. of the explosives, the seller of the explosives,
or the manufacturer of the taggants liable. The
The free-market mechanism may be unac-
addition of taggants to explosives could
ceptable to manufacturers of explosives and change the existing situation in several possi-
gunpowders, given the long time needed to
ble ways:
either develop and qualify an alternative tag-
gant or enact sanction legislation. Legislation ● The use of taggants would make it easier
of a price or use of a regulation mechanism to identify undetonated explosives from
similar to that used for public utilities would the same batch as those involved in the
be an awkward, time-consuming process for a accident, thus facilitating proof or dis-
product whose total annual value would be on proof of allegations that the explosive, the
the order of $10 milIion. taggant, or both were incorrectly manu-
Licensing is not only disagreeable to 3M, but factured.
it is probably not cost-effective. The cost of ● Evidence that incorrectly manufactured
the taggant material includes a component for taggants had been involved in an accident
amortization of the taggant production facili- would probably subject the taggant manu-
ty, as a new facility must be built and the facturer to liability, regardless of any
primary market for identification taggants disclaimers made at the time of sale.
would Iikely be the mandated explosives mar- ● Evidence that taggants had been incor-
ket. The process which 3M plans to implement rectly added to explosives (e. g., an ex-
is capital intensive. Licensing of other manu- cessive concentration) might expose the
facturers would therefore require the construc- explosives manufacturer to Iiability, if
tion of facilities for the licensee, in addition to evidence could be presented that such a
a new 3M facility, resulting in a substantially high concentration posed a danger.
higher total cost. ● There should be no cases in which the
A long-term contract is a potentially attrac- evidence shows that taggants were unsafe
tive mechanism, In fact, the 3M cost estimates if made and used correctly, due to the ex-
are conditional on firm orders for a 2-year tensive qualification program required to
period, although 3M is willing to consider con- demonstrate taggant safety. In any event,
tracting periods of up to 5 years. The details of the fact that Federal law required the use
the regulating mechanism have not been ad- of taggants would be a defense.
36 . Taggants in Explosives

● If, however, taggants actualIy create a Iegedly due to taggants. Alternatively, by legis-
hazard but there is no evidence that they lating a presumption that taggants are safe or
do so, the manufacturers of explosives simply by granting immunity to manufactur-
might be exposed to liability based on an ers, Congress could shift the cost of any tag-
(incorrect) assumption that the manufac- gant-caused accidents to explosives users. A
turing process was somehow at fault. third possibility would be to legislate in a way
that would make taggant and/or explosives
manufacturers liable for accidents caused by
Furthermore, Congress could include in the taggants despite legislative coercion to use
legislation mandating a taggant program provi- them. A final option would be to divide the
sions directing who should bear the costs of ac- costs of accidents by legislative limits on the
cidents. For example, Congress could shift the dollar amount of claims arising from accidents
cost to the Government by allowing suits allegedly caused by taggants. The issue of
against the Government for accident losses al- liability is treated in detail in appendix D.

Before the utility of identification and de- involvement in catastrophic, illegal activities
tection taggants to law enforcement, security, against society. These characteristics make the
and other regulatory agencies can be assessed, terrorist particularly dangerous to society and
it is first necessary to examine the bomber a particularly appropriate target for anti bomb-
threat in some detail. The utility against each ing controls. Terrorists can be roughly divided
segment of the bomber population can then be into political, reactionary, and separatist
assessed, together with the possible responses groups. Political groups are primarily inter-
of the criminal bombers, and be compared to ested in attracting attention to, and sympathy
the utility of other control methods. identifica- with, their cause. For that reason they engage
tion taggants may also have utility for pur- in spectacular events, such as bombings, but
poses other than tracing of criminal bombers. generally attempt to avoid or limit injury and
death resulting from their bombings. Political
The bomber population of the United States is
terrorists often have considerable resources
extremely heterogeneous, with varying motives,
available to them, due to the significant num-
resources, skills, and ability to adapt to a chang-
ber of people who support their aim, if not nec-
ing control environment. For ease of discussion,
essarily their means. The leadership of most of
bombers are divided into four categories
these groups are of above-average inteliigence,
which differ from each other in most charac-
and have either had specialized training or
teristics. These categories include terrorists,
have studied extensively in terrorist activities.
common criminals, the mentally disturbed,
They are thus able to adapt to a changing envi-
and vandals and experimenters. The character-
ronment, although the range of responses
istics of the various types of bombers are sum-
available to them may be Iimited by their polit-
marized in table 14 and briefly described
ical aims. Such political groups have been rela-
tively inactive in the United States in recent
Separatist groups, such as FALN (a Puerto
The terrorist groups active in the United Rican terrorist group), generally hope to gain
States vary widely in ability, resources, train- their aims by generating a reaction to their ac-
ing, and adaptability. They share the common tivities, rather than a sympathy to their aims.
characteristics, however, of high motivation, They are therefore generally less concerned
action as a part of a group, and a continuing with public revulsion to bombings that cause
Ch. n-Detailed Findings ● 37

Table 14.–Attributes of Criminal Bomber Groups

Experience Individual
Perpetrator and training Resources Motivation or group Reaction capability Frequency
U n s o p h i s t i c a t e d L L M I M Mu It!
S o p h i s t i c a t e d . H M H I H Multi
Political ., M-H M-H M-H G M-H Multl
S e p a r a t i s t M-H M H G H Multi
R e a c t i o n a r y L L H G L-M Multi
Mentally disturbed
D i s e n c h a n t e d L L L-M I L Single
V e n g e f u l L L M-H I L-M Single
Pathological . , L-M L H I L-M Varies
V a n d a l s L L L-M I L Single
E x p e r i m e t t o r M L L-M I L-M Single

L Low M-Moderate H-High I Individual G. Group

SOURCE Oftice of Technology Assessment

s u b s t a n t i a l i n j u r y a n d d e a t h s . Separatist readiIy adapt to a changing enforcement envi-

groups have been credited with more than 25 ronment. The only major characteristic he
percent of catastrophic bombings—those re- shares with the professional bomber is that his
sulting in major property damage, injuries, and targets are generally individuals or small com-
deaths. The resources of domestic separatists mercial establishments, unlikely to be pro-
vary from group to group, but are generally tected by a detection taggant sensor. The pro-
less than for comparable groups of political fessional bomber is highly trained and moti-
terrorists. vated and generalIy has considerable re-
sources available to him, either directly or
Reactionary groups, such as the Ku Klux
through his “employer.” Criminals share with
Klan and the American Nazi Party, share some
terrorists the characteristics of engaging in re-
of the characteristics of the political terrorists, peated bombings, but differ in that the profes-
but generally do not possess the same levels of sional criminal bomber usualIy works alone,
training, motivation, and resources, and are
rather than as part of a group. Criminals as a
not as capable of reacting effectively to a group are responsible for approximately 6 per-
changing control environment. They also differ
cent of bombing incidents. Most incidents are
in that their bombings are usualIy directly tar-
limited to specific targets and do not generally
geted at the individual or group they intend to
cause substantial injury or death to innocent
influence, rather than simply at a spectacular
Terrorists have been responsible for approxi-
Mentally Disturbed
mately 12 percent of those bombing incidents
in the past 5 years to which law enforcement The mentally disturbed bomber differs from
agencies assigned a motive. terrorists and criminals in that he generally
does not engage in multiple bombings, al-
though exceptions such as the Los Angeles “Al-
Common Criminals
phabet Bomber” certainly exist. He generally
Criminals range from the petty operator who is poorly trained, has Iimited resources, and
utilizes a bomb for extortion to the profes- acts alone. He is often highly motivated, but
sional bombers of organized crime. The petty perhaps only for short periods of time, in direct
operator is generally poorly trained, is not very response to some stimulus. He is extremely
motivated, has limited resources, and cannot limited in his ability to respond to changin g
38 ● Taggants in Explosives

control situations, either through lack of care Table 15.–Taggant Utility Summary
of consequences or belief in his invincibility. Specific bombing
As his motives are hard to identify, it is dif - conditions Identification taggants Detection taggants
ficult to predict his targets. Low-value targets Limited utility Limited utility
High-value targets,
no bomber High utility Extremely high utility
Vandals and Experimenters countermeasures
High-value targets, Countermeasures Countermeasures
vandals and experimenters share the charac- bomber costly due to require technical
countermeasures increased risk knowledge,
teristics of poor training, limited motivation, planning
and limited resources. They generally work
SOURCE Office of Technology Assessment
alone or in small groups, and do not generally
intend to harm people or cause extensive dam-
age. Their targets are often of little value, like ual search of the area by untrained law en-
mailboxes or outhouses, but some acts of van- forcement personnel is unlikely to turn up
dalism can cause extensive damage to build- identification taggants. Similarly, detection
ings such as schools. While accounting for taggant sensors are unlikely to be present be-
over 40 percent of the reported bombing in- fore the detonation. The lack of utility in these
cidents, they are responsible for little damage cases, however, does not greatly diminish the
and few casualties. overall utility of a taggant program, as the in-
Given the diversity of the criminal bomber tent of the program is not to prevent this type
population, the range of targets involved in of bombing, but to help prevent significant
bombings, and the choice of explosives avail- bombings and to help in the arrest and convic-
able to the bombers, it is difficult to assess the tion of the perpetrators of such bombings.
utility of taggants to law enforcement agen- Identification and detection taggants would
cies. The assessment is made particularly dif- provide a quantum increase in utility in combat-
ficult by the lack of experience with taggants,
ing bombings of high-value targets, assuming the
although the McFillan case (recently tried in
absence of effective bomber responses.
Baltimore) provides one example where identi-
fication taggants were an extremely important The current procedure for the apprehension
piece of evidence linking a suspected perpetra- and control of criminal bombers consists of
tor to the crime. Inferences can also be made three phases:
from experience with the date-shift code and
with the X-ray machines and magnetometers 1. the postdetonation search of the area for
used at airports to prevent hijackings. A useful physical evidence;
construct for viewing the findings is shown in 2. the investigation, based on the results of
table 15, the discussion of which follows. the analysis of the physical evidence; and
3. intelligence gathering on, and surveil-
Both identification and detection taggants lance of, suspected perpetrators or ex-
would have limited utility in combating bombings pected targets.
of low-value targets. Due to limitations on law
enforcement time and resources, minor bomb- The search for evidence phase includes a de-
ings, such as a vandalism bombing of a mail- tailed analysis to try and determine the type of
box, do not warrant as thorough an investiga- explosive used (successful approximately 5 0
tion as bombings involving casualties or signif- percent of the time) and examination of what-
icant property damage. In New York, for exam- ever parts of the bomb, such as elements of the
ple, such cases are generally handled at the in- timing device, may have survived the detona-
dividual precinct level, without the use of the tion. This evidence, together with any evidence
trained bomb squad, bombing investigators, of the presence of the perpetrator (such as hair
and forensic laboratories. As evidenced by the or footprints) serves as the startin g point for
results of the recovery demonstrations, a vis- the investigative phase.
Ch. I/—Detailed Findings ● 39

The investigative phase consists primarily of The value of the list of last legal purchasers
trying to generate some type of lead to the per- will depend somewhat on the length of the Iist.
petrators from the physical evidence gathered, A trace which indicates that the full taggant
as well as tracking leads provided by inform- batch of explosives was sold directly to a mine
ants or witnesses and attempts to correlate the by the explosives manufacturer obviously pro-
characteristics of the bombing with similar in- vides a more useful lead than a trace which
stances. A great deal of effort may be ex- shows a large number of purchasers of a lot of
pended, for instance, in investigating the smokeless powder. Even for the smokeless
sources of a common c ock used as the timing powder case, the list of names would probably
mechanism. not be excessively long. The types of bombings
likely to warrant a detailed investigation are
The addition of ident fication taggants to ex- unlikely to be caused by 1 or 2 lb of gunpow-
plosives would aid the investigatory efforts of der, eliminating most purchasers from the list
law enforcement personnel in a number of or providing multiple traces of the multiple 1-
ways, provided that tagged explosives are Ib lots used to make up the filler.
used, the taggants survive the detonation, and
the taggants are recovered from the explosive The utility of detection taggants in protect-
debris. The taggants provide a good starting ing high-value targets is obvious. The current
point for an investigation as they directly in- procedures for protection of potential high-
dicate the type of explosive used, manufac- value targets vary with the type of the facility
turer, time of manufacture, and provide a list and the time since the last perceived threat.
of the last legal purchasers. This information Airports are protected by requiring all carry-on
may lead directly to a bomber who purchased luggage to go through inspection (usually X-
the explosives legally. In some cases, the ray) and all passengers to walk through a mag-
bomber would not otherwise be identified with netometer. Search of checked baggage is not
the bombing; in others, as was the case with routinely required, although spot checks,
the McFillan incident in Baltimore, the tag- sometimes with trained dogs, do occur, par-
gants add a strong link in a chain of evidence, ticularly when the perceived threat is high.
which may help to obtain a conviction. Tag- Many Government buildings and other con-
gants may provide intelligence information, trolled-access facilities require a package or
such as linking a series of bombings, or linking briefcase check as well as personnel identifica-
a suspect to a theft of explosives by establish- tion to gain entry. The airport instruments are
ing that one of the legal purchasers reported a operated and inspection checks conducted pri-
theft at the time the suspect was in the city in marily by personnel who are poorly trained,
which the theft occurred. Finally, bombers poorly paid, and subject to the problems of
may be deterred from committing bombings maintaining alertness over long periods while
by the knowledge that the chances of their be- performing a dull job. The magnetometers are
ing apprehended are increased by a taggant useful solely to detect metal, and information
program. from the X-ray machines must be interpreted
by the attendant. The use of a self-calibrating
sensor, which would reliably give an alarm at
In order for the taggant information to be the presence of explosives in hand baggage,
useful, however, the bombing must be of suffi- checked baggage, or on a person would offer
cient importance (in terms of property dam- an enormous increase in utility over current
age, notoriety generated, or casualties pro- methods.
duced) to warrant a thorough investigation. In
such cases, identification taggants will provide Many of the criminal bombers who would be
much more information, and more reliable in- likely to attack a high-value target would be
formation, than present methods, and this in- deterred by the knowledge that the target was
formation will require much less effort by the protected by a sensor that would detect the ex-
investigating team. plosives in their bombs (assuming no effective
40 . Taggants in Explosives

countermeasures by the bomber). The deter- ● “spookng” of taggant sensors, or

rence might work to redirect the bomb against ● resorting to another unlawful activity,
another target, to cause a less vulnerable part such as assassination or kidnapping.
of a target to be attacked, or (perhaps infre-
quently) to deter the attack altogether. Those The baseline 3M identification taggants con-
who were not deterred would have their bombs tain both a magnetic layer and a fluorescent
intercepted, protecting that target and pro- layer to aid in recovery after a detonation. The
viding security personnel with additional clues taggants could therefore be removed from
to the perpetrator. powdery explosives by using a magnet; the
Detection taggants would only provide util- process would be both easy and safe, and
ity to those targets that were protected by a would require less than an hour for a typical
detection taggant sensor. Portable detection bomb. In order to hinder this countermeasure,
taggants sensors would also be quite valuable taggants have been manufactured without a
in locating a bomb whose approximate loca- magnetic layer. If a powdery explosive were
tion was known and in determining if a sus- tagged with a mixture of magnetic and non-
pected package contained explosives. magnetic taggants, then the use of a magnet
would enable a criminal to remove only a por-
In summary, identification taggants would tion of the taggants; the remainder would be
provide a quantum increase in utility for those present after an explosion, although they
bombings significant enough to warrant a thor- would be somewhat more difficult to recover
ough investigation, while detection taggants than the baseline taggant. If the criminal were
would provide that increased utility in protect- deterred from attempting magnetic removal
ing those potential targets sufficiently impor- by the knowledge that about half the taggants
tant to warrant a detection taggant sensor. were nonmagnetic, then postdetonation recov-
ery would be only marginally more difficult
The above discussion assumes that the crim- than the baseline case.
inal bombers do not respond to the introduc-
tion of a taggant program. However, counter-
Another possible technique for removing
measures exist which would enable bombers to
taggants from an explosive is to use a black
evade the effects of a tagging program. The avail-
light to identify the taggants by their fluores-
able countermeasures require varying degrees of
cence, and then remove them with a tweezer.
specialized knowledge, and some of them in-
This process is safe, but more difficult than
volve significant risks. Because most bombers
magnetic separation, and would probably re-
would probably not avail themselves of the possi-
quire many hours of painstaking effort for a
ble countermeasures, a taggant program would
typical bomb. Unlike magnetic separation, it
probably retain substantial law enforcement util-
could be used to remove taggants from explo-
sives that are tacky rather than powdery. It has
Bombers seeking to respond to a taggant been proposed that the encapsulation of the
program by using countermeasures can use taggants be made opaque, and matched to the
any ot several approaches: color of the explosive, in order to render such
removal impossible. Since the encapsulant
● removal of the taggants, would be melted by the heat of a detonation,
● fabrication of homemade explosives, postdetonation recovery would not be af-
● use of incendiary bombs, fected. Although it should not be difficult to
● theft of explosives, develop an opaque encapsulant, this has not
● black-market purchase of explosives, yet been done. Opaque encapsulation would
● use of explosives manufactured before make quality control, both of manufacturing
the taggant program commenced, taggants and mixing them with explosives,
● use of blasting agents, more difficult, and its cost impact has not
● sealing of detection taggants, been evaluated.
Ch. /l—Detailed Findings . 41

In order to remove a nonmagnetic taggant Figure 4.—Size Comparison of the 3M Identification

with an opaque encapsulant from an explo- Taggant and Some Smokeless Powders
sive, the explosives could be acetone dis-
solved, the taggants and other SoIid material
removed by filtering, and the explosives recon-
stituted. This complex operation wouId require
specialized knowledge, be roughly equivalent 3M identification
in danger and difficulty to fabrication of ex- taggants
plosives from raw materials, and would result
in less reliable (less likely to detonate) explo-

Taggant removal from some gunpowders

could be significantly easier than from explo-
sives, as some gunpowder grains are consider-
ably larger than the identification taggants, as
shown in figure 4. Separation from these pow- Hercules Bullseye
ders may therefore be accomplished simply by
screening, even if the taggants are nonmag-
netic. Tests with several Du Pont IMR powders
have shown that it would be difficult to sepa-
rate the taggants from the chips and fines con-
tained in the gunpowder package, but all small
particles could easily be separated from the in-
tact grains by screening. It has been proposed
to alleviate this problem by agglomerating the Du Pont IMR 4350
taggants into clumps whose size roughly
matches the specific powder grain size. The
cost impact of such a solution was not ad-
dressed during this study.
Removal of the detection taggants would
not be feasible.
Fabrication of explosives may be accom-
plished by a variety of means, but a consider- W-W 452AA
able degree of expertise is required to avoid
the risk of premature detonations, and to en-
sure high reliability. It should be noted that
fabrication of detonators is significantly more
difficult than fabrication of the explosive
A substantial number of bombing incidents
involve the use of incendiary bombs; it is quite
Hercules Red Dot
impractical to tag the wide range of materials
from which incendiary bombs could be fabri-
cated. It may be more difficult, however, to
fabricate a reliable delay fuze for an incendi-
ary bomb. In addition, while incendiary bombs
may be effective in destroying structures and
jeopardizing groups of people, explosive bomb

F 1-4 1’) “1 - - 93 - L
42 . Taggants in Explosives

fillers offer a better chance of killing, injuring, agents, if it were judged desirable, would
or intimidating a particular individual. greatly increase the cost of a taggant program.)
Effective bombs can be fabricated from
A significant fraction of the explosive cur-
ANFO; to do so requires a certain level of skill
rently used for fabricating bombs is stolen. A
to ensure reliable detonation and the assembly
taggant program may well increase the theft of
of a number of components, some of which
explosives; however, additional explosive secu-
may not be readily available. The risk of pre-
rity could reduce the incidence of theft. Tag-
mature detonation is small for a bomber with
gants from stolen explosives would not provide
adequate knowledge and patience, but may be
a direct clue to the purchaser, but would help
significant for bombers without those charac-
law enforcement officials to establish patterns
teristics. Blasting agents are infrequently used
and I inks between crimes, improving the
at present in criminal bombings.
chances of apprehending the criminals. The
bomber who steals explosives further increases The effectiveness of detection taggants can
the risk of apprehension by committing an ad- be severely limited by creating a seal between
ditional crime. Finally, taggants could pinpoint the explosives and the detection taggant sen-
locations from which explosives were stolen, sor as the vapor could not escape the package
providing a guide to tightening security in to trigger the sensor. Such a seal can be con-
those places most vulnerable to theft. structed with the appropriate industrial materi-
Explosives could be purchased on the black als and equipment, but a reliable seal would
market or illegally imported from abroad. Both be very cliff i cult to fabricate with the resources
courses of action subject the bomber to in- normally available to individuals. Hence spe-
creased risk of capture, from informants or un- cialized knowledge, advance planning, and the
dercover agents in the former case and as a re- resources to buy the required material, would
sult of smuggling, in the latter. Both courses of be needed to defeat the detection taggants.
action would require substantial resources and
Detection taggant sensors could be purpose-
the ability to plan in advance.
ly triggered or “spooked” by placing detection
Explosives manufactured before the imple- taggants, or other materials so similar chemi-
mentation of a taggant program could be used cally to the detection taggant that the sensor
to fabricate bombs. There is some evidence could not make the distinction, in nonexplo-
that a considerable stockpile of explosives cur- sive materials. If several suitcases or packages
rently exists in the hands of criminal bombers, within a short period of time triggered the de-
and this stockpile could be expanded in the tection taggant sensor for no apparent reason,
time between legislation and implementation those operating the sensor might well con-
of a taggant program. Acquisition and storage clude that it was malfunctioning, and discon-
of the explosives for a period of time require nect it. It would then be possible to introduce
considerable advance planning and resources, tagged explosives into the protected area. This
however, and increase the risk to the bomber countermeasure would require that the
of discovery of the explosives. While the use of bomber obtain a supply of the detection tag-
explosives manufactured prior to a taggant gant material; access to detection taggants
program may be an effective countermeasure could and should be made difficult.
initial I y, m o s t e x p l o s i v e m a t e r i a l s h a v e a
Finally, bombers can turn to other crimes,
limited shelf-life. Gels, slurries, and emulsions
such as murder, assassination, or kidnapping.
are generally reliable for less than 1 year; the
These crimes, however, are often not as spec-
sensitivity of dynamites tends to increase with
tacular as bombings and all involve greatly
age; gunpowders and booster materials have a
higher risk to the perpetrators than do bomb-
long shelf-life.
ings. I n addition, a direct action against a visi-
Blasting agents, such as ANFO, are not ble target requires more motivation and a dif-
among the explosive materials BATF plans to ferent temperament than does an indirect
directly tag. (OTA finds that tagging blasting crime such as a bombing.
Ch. n-Detailed Findings ● 43

Dynamite bomb with nails OTA consulted numerous explosives ex-

perts, all of whom agreed that countermeas-
ures such as these are possible. However, the
experts on law enforcement and terrorism
which OTA consulted agreed that criminal
bombers would fail to make use of countermeas-
ures, even when the necessary knowledge and
equipment could be obtained without enormous
efforts. However, some terrorists and profes-
sional criminals would make use of countermeas-
ures. This judgment appears to be based on an
assessment of the type of personality that is
generally involved in this kind of criminal ac-
Pipe bomb
tivity. Bombings are currently a low-risk, rel-
atively simple type of criminal activity. Each
added element of risk, or additional stage
necessary to fabricate a bomb, will decrease
the likelihood of the prospective bomber ac-
tually committing the bombing. An instructive
analogy is aircraft hijacking. It is possible to
smuggle a weapon on to an airplane by a num-
ber of means, but, in fact, since the antihijack-
. .. ing program started there have been thousands
of weapons found annually by the screening
process, hundreds of weapons found aban-
doned near the controlled boarding gates, and
. .-
few or no cases of aircraft hijacked with the
Molotov cocktail, dynamite, and grenade
use of smuggled weapons.

Consequently, OTA believes that counter-

measures are not likely to greatly diminish the
law enforcement utility of a taggant program,
despite their potential to do so.

The above discussion has been essentially

qualitative, as little quantitative data is availa-
ble. However, an attempt was made to draw in-
ferences from similar programs. The data avail-
able from the date-shift program suggests that
identification taggants may prove effective in in-
creasing the arrests and convictions of criminal
bombers. However, the data base is too small to
be more than suggestive. Similarly, data on the
reduction of hijackings after the introduction of
an antihijacking program suggests that detection
taggants would prove an effective deterrent. The
program most directly analogous to the pro-
posed identification taggant program is the re-
quirement that the date and shift of cap-sensi-
Photo credits US. Department of the Treasury tive high explosives be clearly printed on each
Various types of explosives used by terrorists stick. For undetonated bombs the date-shift
44 ● Taggants in Explosives

code provides the same information as identi- The above discussion dealt with the utility
fication taggants would provide for the post- of taggants for the control of criminal bomb-
detonation case. No total review of the cases ers. There exist other approaches to the problem
involving explosives recovered from malfunc- of control of criminal bombers which could be
tioning bombs has been conducted. A limited used in conjunction with, or instead of, a tagging
set of 55 cases was examined, however, by program. Some of the methods, however, may
BATF. In that sample, six cases were forwarded be unpalatable or not cost-effective. Other ap-
for prosecution (10.9 percent). That is twice the proaches, some of which have been imple-
percent forwarded in cases that did not in- mented in areas facing a more severe bomber
clude date-shift code data. Similar results were threat, particularly from separatist terrorist
obtained by MSA during a review of the BATF groups, include:
data. Of the 10 bombing attempts MSA re-
● alternate detection approaches,
viewed, the date-shift code proved useful in 40
● control of explosive materials,
percent of the cases, was not useful in 50 per-
● better security,
cent of the cases, and was of questionable util-
. more coordinated police response, and
ity in 10 percent. While the results were posi-
● harsher judicial response.
tive in both cases, the extremely small sample
size makes it impossible to draw significant The Aerospace Corp., the Federal Aviation
conclusions. I ME reported to OTA that manu- Administration, and the military are currently
facturers are seldom requested to appear in investigating, or have investigated, a large
court to testify regarding a date-shift trace; in number of techniques for detection of untag-
recent years less than 1 percent of the traces ged explosives. Methods investigated have in-
requested led to a court appearance. cluded X-ray fluorescence, gamma ray excita-
tion, nuclear magnetic resonance, both fast
The most direct analog of the detection tag- and thermal neutron activation, dual energy
gant program is the antihijacking program ini- tomography, detection of the characteristic
tiated in 1971. There was an average of 27 hi- vapors of explosives, and deactivation of blast-
jackings from domestic origins in the 4 years ing caps. Some of the approaches are prom is-
preceding full implementation of the program. ing, although all but the last two would be
In the next year (1973), hijackings decreased to limited to checked baggage. However, none of
a single incident, and have averaged only four the approaches, with the exception of non-
per year since. It should be noted that a num- tagged vapor detection, has progressed as far
ber of countermeasures are possible that as the detection taggant research and most ap-
would evade the currently used magnetom- pear to be significantly more expensive, both
eters and X-ray machines. However, essentially for the instrument and for personnel to man
no incidence of the use of these countermeas- the instrument. Commercial vapor detectors
ures have occurred since the inception of the are currently marketed for explosive detection,
anti hijacking program. but their sensitivities and flexibility fall far
short of the goals of the taggant vapor detec-
Numerical estimates of the numbers of
tion devices. Research on the promising ap-
bombers who would be arrested and the num-
proaches should continue; it may be most ef-
ber who would be deterred by a taggant pro-
fective to deploy a detection taggant system in
gram were made by MSA in order to generate
conjunction with one of the other systems.
input to their cost-effectiveness analysis of the
taggant program. The numbers they used in the Control of explosive materials could range
analysis were a 50-percent increase in the ar- from uniform procedures for the purchase of
rest rate (from 8 to 12 percent) and a 5-percent explosives to the total control by the military
detergency rate. These numbers are simply or police of all explosives, from manufacture
guesses and OTA has no data that would allow to the legal detonation. In some States, explo-
it to make guesses or assess the accuracy of sives are tightly controlled. For instance, in
the MSA guesses. Louisiana all users or transporters of explosives
Ch. /l—Detailed Findings ● 45

must be licensed by the State police. I n some systems which would be necessary. A reason-
other States, however, explosives may be pur- able alternative may be to increase the securi-
chased over-the-counter simply by providing ty of military explosives.
identification and presenting a Federal permit
A more coordinated law enforcement re-
or filling out a form. Uniform tight control
sponse to the bomber threat would be effec-
would make it more difficult to purchase ex-
tive, whether a taggant program were insti-
plosives for illegal use and would be particu-
tuted or not. At present, “major” bombings
larly effective in combating the less sophisti-
must be reported to either the FBI or BATF.
cated bombers. Complete control of explo-
However, no uniform definition of “major” ex-
sives, to the point of requiring police or mili-
ists. Other agencies, including some State
tary personnel to physically be at the site of a
agencies, also collect bombing statistics. Ex-
legal use of explosives and be responsible for
amination of the statistics shows a significant
each detonator, as is the case in Ireland, would
lack of uniformity in what is reported to each,
essentially eliminate the use of domestically
the information available on each incident re-
produced commercial explosives in bombings.
ported, the retrievability of information from
Sophisticated bombers would be forced to fab-
the data bank, and the methods for updating
ricate their own explosives (or purchase
the files. One responsible center, to which all
“homemade” explosives on a black market),
bombing information would be required to be
while the unsophisticated bomber would be
reported in a uniform, easily updated, easily
eliminated. Such a program would entail ex-
accessed format, would be an obvious aid to
tremely high costs however, both in monetary
law enforcement efforts against criminal
terms and in terms of the disruption to indus-
tries that currently use explosives.
Better coordination and communications
Better security is possible, both to protect
between the forensic laboratories and the field
potential targets and to protect explosive ma-
investigators would also be helpful. Agents in
terials from theft. It would be possible, as an
the field are sometimes not sensitive to what
example, to hand-search all checked luggage
information or what physical evidence would
being loaded on an airplane; in fact, EL AL (na-
be useful to the laboratory. This coordination
tional airline of Israel) does conduct such
will be particularly important if an identifica-
searches. Similarly, it would be possible, al-
tion taggant program is introduced, as the re-
though extremely time-consuming, to search
covery of the taggants appears to be a labora-
every person entering the Rose Bowl for the
tory-intensive procedure.
Rose Bowl game. However, detection taggants
appear a more reasonable alternative. Finally, control of the physical site of the
bombing by a single responsible individual
Protection of explosives from theft could be
would be extremely useful. A major incident
improved, however, and may well have to be
may involve several levels of law enforcement
to prevent a wholesale shift to theft as a
agencies, several levels of elected representa-
source of explosive material if a taggant pro-
tives, and other activities such as first aid and
gram is instituted. All of these controls have
fire control. Uncoordinated activity by all
cost impacts which have not been calculated
these people could well destroy valuable phys-
in this study; a match must be made between
ical evidence. Excessive use of water by fire-
their cost and their marginal utility in the face
fighters is a potentially serious problem if iden-
of the current bomber threat. As an example, if
tification taggants are used, as they might be
the use of military explosives in criminal
washed totalIy away from the bombsite.
bombings increases markedly it may become
necessary to counter that threat. Tagging of The utility of a harsher judicial response to
military explosives would be extremely costly, criminal bombers is a particularly sensitive is-
due both to the large amount produced and to sue, with little technological insight available,
the requalification cost of all current munition and is mentioned only for completeness.
46 ● Taggants in Explosives

Program Implementation plosives in criminal bombings, tagging of com-

mercial explosives may shift the expected fu-
Given the current development state of the
ture frequency. Similarly, tagging of black and
identification and detection taggants, a num-
smokeless powders is of critical importance to
ber of options are available regarding the
an overall taggant program.
method of implementation of a taggant pro-
gram. Among the issues are what, if any, tag- Some mechanisms to tag blasting agents
gant program should be legislated; if a taggant may also be desirable. However, the cost of
program is legislated, what materials should be directly tagging the agents would be extremely
tagged, what level of tagging should apply, high. The BATF plan to tag the detonators,
and what is the procedure for making deci- boosters, and detonating cord normally used
sions not specifically resolved by the legisla- with blasting agents may be a reasonable com-
tion. promise, particularly as blasting agents are
now rarely used in criminal bombings and ap-
One of the first issues needing resolution is
proximately half of the blasting agents are
what explosives should be tagged. The analysis
mixed and used onsite in the same day.
conducted showed that criminal bombers tend
to use the most readily available source of explo- As indicated above, various levels of im-
sives. Therefore the tagging program with the plementation of a taggant program are possi-
highest utility would include provisions for tag- ble, each with an associated cost of implemen-
ging of commercial explosives and gunpowders. tation. The most reasonable way to determine
the optimum program to implement may be to
Table 7 showed the frequency-of-use distri-
consider the marginal additional cost of each ad-
bution of explosives for bombings, including
ditional element of utility. This approach is il-
explosives identified both in the field and in
lustrated in figure 5, where the identification
the BATF laboratory. While the completeness
taggant utility function is varied. Qualitative
of these statistics may be open to interpreta-
estimates of marginal utility are shown to ap-
tion, it is clear that a wide variety of materials proximate scale, along with quantitative esti-
are used as bomb fillers. Discussion with both
mates of the cost of implementing a program
domestic and foreign law enforcement offi-
that would yield that level of utility.
cials has stressed the fact that all types of
bombers will use the most readily available The lowest implementation option would
source of explosives, although sophisticated tag cap-sensitive explosives, boosters, detona-
bombers would be more likely to limit their tors, detonating cord, and gunpowders, but not
use to materials that are efficient for the in- blasting agents. A unique identification tag-
tended purpose. As an example, a relatively gant would be used for each manufacturer,
small amount of a powerful explosive was ap- type of product, and year of manufacture. This
propriate for the La Guardia Airport bombing, program corresponds to the low-level program
as it would cause extensive damage and be previously discussed. That level of implemen-
concealable in a relatively small package. The tation would directly provide most of the phys-
amount of gunpowder needed to do as much ical evidence information that current meth-
damage would occupy a much larger volume, ods attempt to provide. However, it would not
and might be noticed; it would therefore not directly provide a list of last legal purchasers.
be an appropriate choice for a sophisticated The relatively modest cost for that program
bomber. would be approximately $15 million per year, *
probably less than is currently expended in an
If one type of explosive material is not as
attempt to provide the same information by
highly controlled, then bombers will tend to
current means, although the cost would be
shift toward that material. For that reason, it
shifted to manufacturers and users of explo-
may be desirable to tag or otherwise control
military explosives. Although current statistics
*The cost estimate In this sectton IS for an Identification tag-
show a relatively infrequent use of military ex- gant program only
Ch. n-Detailed Findings ● 47

Figure 5.— Marginal Cost-Utility Function somewhat finer grain of intelligence informa-
Marginal utility Marginal cost tion. However, the cost increase of $20 million
per year would be fairly substantial.
Additional marginal utility could be gained
by tagging blasting agents. This would be of
value in two cases —the case in which the iden-
tification taggants from the detonator and
booster used to ignite the blasting agent did
not survive (or were not recoverable) from the
debris of an explosion, or the case in which a
bomb was fabricated that used some other (un-
tagged) means of detonating the blasting
agent. There is no body of test data to indicate
the likely frequency of the first condition;
while the second condition is certainly possi-
ble, almost all bombers capable of detonating
a blasting agent without commercial detona-
tors and boosters would also be capable of ob-
taining or fabricating untagged explosives in
the first place. At present blasting agents are
infrequently used for bombings — averaging
two BATF sources suggests that blasting agents
are used in about 0.5 percent of bombings, and
account for a small percentage of the property
damage and casualties. Since the cost of tag-
ging blasting agents would be approximately
$170 million per year, several times that of all
the other elements of a tagging program com-
bined, the marginal utility of doing so appears
SOURCE Office of Technology Assessment
relatively low.

In short, the implementation of a taggant

The next o p t i o n w o u l d b e t o p r o v i d e a program would require unambiguous decisions
unique taggant code for each shift of each about which materials required taggants, and
product manufactured and to keep a record of what the applicable regulations would be. It
the movement of explosives from the manu- would be desirable if any legislation on the
facturer to the last purchasers, in a manner subject either made these determinations or
analogous to the date-shift code currently unambiguously delegated authority to do so.
marked on the casings of explosives. This op-
tion corresponds to the OTA-identified base-
Given the present state of development of tag-
line program, and would provide a list of last
gants, OTA’S data and analyses appear to be con-
legal purchasers and additional intelligence in- sistent with any of three possible courses of con-
formation, at a program cost increase of ap-
gressional action:
proximately $10 million per year. 1 Pass legislation requiring taggants, and set
A further implementation option would be up a procedure to determine if and when
to uniquely tag each 10,000-lb batch of explo- the technical development and testing
sives and each 2,000-Ib batch of gunpowder. have progressed to the point where imple-
This would lead to a somewhat smaller Iist of mentation can begin.
last legal purchasers, which would mean fewer 2. Defer legislative action on taggants, but
places that must be investigated, as well as a encourage (inter alia by appropriating
48 . Taggants in Explosives

adequate funds) BATF to continue tag- til such a resolution is accomplished, and
gant development, with a view to consid- neither smokeless powders nor b o o s t e r s
eration of legislation when development should be excluded from a tagging pro-
and testing are complete. gram.
3. Take no legislative action on taggants, ● When and if a sufficient probability of
and encourage the executive branch to survival and postdetonation recovery of a
search for other ways of improving the ef- given identification taggant has been
fectiveness of law enforcement against demonstrated to justify adding that tag-
terrorists and other criminal bombers. gant to a given type of explosive.
● When and if a detection sensor has dem-
If Congress chooses the first of these op-
tions, it should recognize that even though the onstrated adequate sensitivity, low false-
legislation can define precisely what materials alarm rate, ease of operation, ease of
would require taggants and provide guidance maintenance, and acceptable unit cost
on the stringency of regulations, there will re- under field conditions to be considered
main some determinations which it is not yet sufficiently “available” to justify requir-
possible to make: ing the addition of detection taggants to
● When and if an adequate number of suc- ● When and if a detection taggant has dem-
cessful compatibility tests have been con-
onstrated adequate shelf-life, nontoxicity,
ducted. Particularly pertinent in this re-
and penetrativity to be considered “avail-
gard would be a determination of what
able. ”
constitutes a resolution of the current in-
compatibility between the 3M identifica- In view of the fact that BATF has become the
tion taggants and one type of smokeless major proponent of the use of taggants in ex-
powder or the RDX-based booster mate- plosives, there is much to be said for entrusting
rial. The 3M identification taggants can- such determinations to an official or proce-
not safely be added to these materials un- dure outside the Treasury Department.
—. —

Chapter Ill
—. .——


* * * * * * * * *.*0 **** 9*** **** **** ***0 *0** **** **0 51 * * * * * 9

Taggant Development History ● **** **** **** 900* **4* *a **a********** 51

Identification Taggants. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Predetonation Only. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Radiological Tracers. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Chemical Assay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Physical Taggants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Detection Taggants. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
Vapor Taggants. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Detection Taggant Sensor Systems. . . . . . . . . . . . . . . . . . . . . . .......59
untagged Detectoin **** a*** **** **** e.** a*** **** *e* be********* 62

Vapor Detection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..62

Differential Contrast Radiography . . . . . . . . . . . . . . . . . . . . .. .....63
Excitation lnduced Emissions. . . . . . . . . . . . . . . . . . . . . . . . . . .......64
Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......65
Current BAFT/Aerospace Tagget ● * * * * 65
Program Status ... , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Projected Schedule. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ............ 68
Implementation Philosophy. . . . . . . . . . . . . . . . . . . . . ., . . . . .......69
Identification Taggant Surviaval Teasting **0* **** **m* **09 *b** **em **0** 70

Boosters, Military Explosives . . . . . . . . . . . . . . . . . . . . ., . . . . .......71

Black and Smokeless Powders . . . . . . . . . . . . . . . . . . . . . . . . . .......72
Detonators and Detonating Cord . . . . . . . . . . . . . . . . . . . . . . . .......72
Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......72
Chapter Ill



The idea of adding material to explosives to Identification Taggants
enhance the predetonation detection and the
postdetonation identification of explosives has Ideas for tagging materials to be used for
been considered by various military and civil- identification of the source of explosives used
ian agencies for at least 15 years. Some of the in criminal bombings and bombing attempts
suggested material, such as radioactive iso- can be generally grouped into the following
topes, would perform both functions, some four classes:
could only perform one. A number of the con- 1. addition of materials that would not sur-
cepts which have been proposed during that vive the detonation, but which would pro-
time are briefly described in the following sub- vide information if a bomb were recov-
sect ions. ered undetonated;

52 ● Taggants in Explosives

2. addition of materials that would physical- not legally usable in New York. This data is not
ly survive the detonation and be recov- helpful to police in tracking bombers but does
ered intact; assist in control of legal uses of dynamite
3. addition of materials to the explosives within New York.
that would be detected in an assay of the
The English apparently use a method some-
debris; and
what better than the date-shift code in that the
4. addition of radioactive isotopes.
identifying code consists of colored threads
within the explosives. The threads do not sur-
Predetonation Only vive the detonation, but the information con-
Since 1970, the date, shift, manufacturer, tent is not lost by discarding the cartridge, as is
and product have been printed on the car- the case with the date-shift code; it may not be
tridge of cap-sensitive high explosives. The possible, however, to encode sufficient infor-
manufacturer keeps records, by that date-shift mation for U.S. needs by that method.
code, and can tell to whom each batch of ma-
terial was sold; distributors also are required to Radiological Tracers
keep records of sale. It is possible, from the
Addition of small amounts of radioactive
date-shift code, to compile a list of last legal
isotopes to explosives during the manufactur-
purchasers of explosives from a lot with the
ing process is particularly attractive as it pro-
same date-shift code. I n fact, BATF maintains
vides a mechanism for both identification of
a National Explosives Tracing Center, whose
the explosive materials from the postdetona-
function is to coordinate that activity. A typi-
tion debris and a simple detection mechanism.
cal trace would start with the recovery of an
There are a large number of radioisotopes, so
undetonated bomb by a BATF special agent.
an identification scheme could certainly be de-
He would call into the tracing center with the
veloped that would provide sufficient unique
information, and the data would be forwarded
code species.
to the manufacturer who would provide the
list of consumers or distributors; if explosives The two primary objections to this often-
from that lot were sold to a distributor or dis- proposed solution are public reaction and
tributors, they would be contacted for a list of safety. Given the present widespread antipathy
retail purchasers. to anything involving radioactivity, it is doubt-
ful if the public would accept such a solution,
The date-shift code information has proven
even if there were no safety hazards.
useful in investigations of criminal bombings,
although its utility is Iimited to instances Two potential safety hazards exist, one hav-
where the explosive is recovered before deto- ing to do with sensitization of the explosive
nation, or in some cases, where a low-order materials, and the other with the effects of
detonation does not destroy the cartridge. In low-level radiation. Addition of foreign materi-
addition, the information is only on cap-sensi- als to explosives poses a potential sensitivity
tive high explosives, and on the packages of hazard. However, the amount of radioisotopes
detonators, black powder, and detonating required would be far smalIer than the mate-
cord. No trace data is available for other ex- rial necessary for other tagging mechanisms,
plosive material, such as smokeless powder, in- so explosive sensitization would probably be
dividual detonators, or even cap-sensitive high no more of a problem than with other types of
explosives that have been removed from the taggants.
The hazards of low-level exposure to radia-
Smaller amounts of information are given by tion are not well-defined; the current trend is
other systems that do not survive the detona- toward severe limitation of exposure. Thou-
tion. For instance, all dynamite legally coming sands of people come into direct contact with
into New York must be red. I f dynamite is re- explosives every day at the manufacturers, dis-
covered that is not red, it indicates a purchase tributors, and users level, so a large number of
Ch. 111—Taggant Research Review . 53

people would have some exposure. Primary ethanol solutions. By using several rare earths
concern would be at the manufacturing level, and by varying concentrations, a sufficient
where workers would have more continuous number of unique codes could be constructed.
exposure than, for instance, a user. Aside from The taggants were recovered from the debris
the adverse psychological effect the use of with ethanol-dampened cotton swabs. The
tracers might have on such workers, and the swabs were then assayed in the laboratory by
possible long-term effects of low-level expo- ion-exchange methods; analysis was accom-
sure, there would be a large cost impact due to plished by X-ray excited optical luminescence
the need for specially trained personnel, as techniques.
welI as storage, handling, and decontaminat-
ing equipment. If it were necessary for the Drawbacks to the Ames taggants included
Nuclear Regulatory Commission to control the sensitization of the explosives by the ethanol
shipment of the explosives and to license and carrier, a high background level, particularly
otherwise supervise all explosive users, addi- for detonations taking place near or on the
tional major costs and inconvenience would ground, and a rather specialized laboratory
occur. procedure necessary for the taggant assay and
A final drawback is that reading of the in-
formation encoded in the postdetonation de-
Physical Taggants
bris would be a fairly complicated laboratory
procedure involving sample preparation, radia- This class of taggants is designed to survive
tion counting, and radioisotope identification. the detonation in its original physical form, to
Only a limited number of laboratories in the be separated from the debris, and to be de-
country have the trained personnel and facili- coded, either in the field or in the laboratory.
ties; police f o r e n s i c l a b o r a t o r i e s a r e n o t Several types of materials have been sug-
among them. gested. Physical taggants must meet the same
requirements as the chemical taggants, how-
Chemical Assay ever, in addition to physical survival, so the
number of serious candidates is somewhat lim-
A number of approaches have been pro-
ited. Three taggants remain promising candi-
posed that have in common the addition of
chemicals to the explosives that would be re-
covered from the postdetonation debris and be 3M COLOR-CODED TAGGANT
identified by a laboratory assay of the debris. More research has been conducted with the
While the number of chemical materials is al- 3M identification taggant than with any other.
most limitless, a successful chemical taggant It is the baseline taggant proposed by BATF for
must have the following properties: implementation if a taggant program is legis-
● inertness, lated, and is the taggant used for the OTA cost,
● nonsensitization of the explosives, safety, and utility analyses.
● not present in background material,

The taggant consists of an irregular chip of
able to survive the detonation,

thermosetting melamine alkyd, approximately
long-term stability,

0.12 mm thick and about 0.40 mm in its great-
not a health hazard, and

est dimension. Figure 6 shows the eight-layer
sufficient variation must be possible to
construction; variation of the sequence colors
form a large number of unique codes.
provides the necessary library of codes. A total
The chemical taggant with which the great- of approximately 6 million unique codes is
est amount of research has been conducted available, when al Iowances are made for cer-
was developed by the Ames Laboratories in the tain forbidden adjacencies (colors too difficult
early 1970’s, I n this method, rare earths were to distinguish) and other restrictions. One face
added to explosives as oxides or as nitrates in of the taggant visably fluoresces when illumi-
54 . Taggants in Explosives

Figure 6.—3M Color-Coded Identification Taggants

I 4 ---
. “ : . 3 - .
I m

r - . ., -. --4
Ch. 111—Taggant Research Review ● 55

nated with black Iight (366 nanometers) as an fluoresces in the visible range when illumi-
aid in recovery, either in the field or labora- nated by shortwave ultraviolet radiation (254
tory. The other face contains iron powder, al- nanometers) and magnetic particles, both of
lowing the taggant to be picked up by a which assist in the recovery process.
magnet, another recovery aid.
Due to the limited number of rare-earth
I n theory, the taggant can be recovered from compounds available, and the fact that the in-
the debris by use of a magnet and a black light, dividual components are not ordered like the
read in the field by a low-power microscope, 3M taggant layers, the Iibrary of possible codes
and traced through the BATF tracing center. I n is only approximately 3,000, even with three
fact, laboratory separation may be needed in distinct spotting phosphors. Use of different
most bombings; the recovery and laboratory concentrations or pairing of two different tag-
procedures are quite simple, however, and can gants to form a unique species can significant-
be performed in the field with little equipment ly increase the library, with approximately
and train i ng. 600,000 codes available for the paired taggant
Several variations of the basic concept have
been tried, some including a polyethylene en- A significant number of compatibility tests
capsulant and some including SIightly different have been conducted with the taggant, as have
chemical and physical properties of the indi- a small number of survivability-recoverability
vidual layers. The safety, survivability, utility, tests. Due to the ceramic nature of the taggant,
and cost aspects are discussed in great detail it is extremely survivable and does not ther-
elsewhere in this report. mally degrade in high-energy explosives (such
WESTINGHOUSE CERAMIC TAGGANT as boosters), as does the 3M taggant. In addi-
tion, since the rare-earth doping is homoge-
The Westinghouse taggant consists of a mix- neous throughout the material, the full code
ture of rare-earth compounds, bound together can be read from even a small recovered tag-
into a ceramic-1ike particle, whose appearance gant chip. The Westinghouse taggant is ex-
is similar to a grain of sand, and whose largest tremely gritty, and has been shown to sensitize
dimension is approximately 0.2 mm. Each of explosives if not encapsulated in a polyethyl-
the rare-earth compounds fluoresces at a char- ene coating.
acteristic wavelength when illuminated by ul-
traviolet radiation (325 nanometers). A scan- No additional effort is currently underway
ning monochronometer is used to read the with the Westinghouse taggant, due to a West-
wavelength of the various rare-earth com- inghouse concern over liability should some
pounds, and thus to identify the taggant code. taggant not be fully encapsulated and thus
The 10 rare earths that have been evaluated, cause sensitization of an explosive material.
and their characteristic emission wavelengths, From the limited data available, it would ap-
are: pear that the Westinghouse taggant shows in-
teresting potential, particularly due to its high
survival rate, a l t h o u g h s o l u t i o n s m u s t b e
Strontium chlorophosphate. europium
Yttrium vanadate thulium. 476 sought to ensure 100-percent encapsulation. I n
Yttrium phosphate cerium, terbium 546 addition, some further limitations are imposed
Yttrium vanadate erblum 555 by the relatively small code library available
Yttrium vanadate: dysprosium 575 and by the rather complex laboratory identifi-
Yttrium vanadate: samarium 608-648
Yttrium vanadate: europlum 618
cation procedure required.
Yttrium oxy sulfide europlum 626
Strontium fluoroborate. europlum,
samarium 687 The Curie point taggant consists of a collec-
Strontium fluoroborate europlum 375 tion of five distinct ferrites, packaged with an
As in the 3M taggant, the Westinghouse tag- ultraviolet sensitive spotting phosphor in a
gant incorporates a spotting phosphor which binder of potassium silicate. Ferrites exhibit
56 ● Taggants in Explosives

the property that their ferromagnetism disap- Detect ion Taggants

pears when the temperature of the ferrite is
raised above a specific temperature, desig- Four general types of detection tagging ap-
nated the Curie point temperature. identifica- proaches are described in the literature, i n -
tion of a particular taggant is thus accom- CI uding:
plished by placing the recovered taggant in a 1. radioisotopes,
temperature-controlled chamber and record- 2. vapors,
ing the magnetism as a function of tempera- 3. electromagnetic (E/M) taggants, and
ture. 4. activation of nonradioactive isotopes

Radioisotopes for use as detection taggants

Approximately 50 ferrites have been identi-
possess the same drawbacks as they do for use
fied whose Curie point falls in a laboratory
as identification taggants; the above discus-
practical temperature range. The 50 ferrites,
sion need not be repeated here.
used in combinations of s at a time, yield a li-
brary of approximately 2 million unique spe- Electromagnetic taggants incorporated into
cies. a detonator, such as the passive harmonic ra-
dar taggant investigated by the Aerospace
As the taggants are ceramics, their surviva- Corp., offer the possibility of detection at a
bility in high-energy explosives, such as boost- distance with a relatively low rate of false
ers, should be good. Very preliminary tests alarms. All of the concepts so far proposed,
have demonstrated the survivability of the tag- however, can be easily defeated by wrapping
gant in boosters and high-power commercial explosives in metal foi1. I n addition, inclusion
explosives such as Power Primer. of such devices would probably have a signifi-
cant effect on the procedures used to manu-
The Curie point taggants share the potential facture detonators, on detonator cost, and sig-
sensitization problem of the Westinghouse nificant false alarms could be caused by com-
taggants, and must therefore be encapsulated mon diodes from radios, calculators, and other
with 100-percent certainty. The Curie point electronic instruments.
taggants have another serious drawback: mag- A variation of the idea of electromagnetic
netic separation from powdery materials such taggants has been proposed, called detonator
as gunpowders and powdery dynamite would deactivation. In this concept, a reed switch is
be an obvious simple countermeasure. connected in series with a detonator bridge
wire. illumination of the detonator by a switch-
able electromagnetic source would cause the
reed to open. A number of methods are possi-
The 3M taggant, which has been the most ble to ensure that the reed could not be subse-
thoroughly researched identification taggant, quently closed. The advantages of the concept
appears to be the most viable candidate, al- are twofold:
though the Westinghouse taggant exhibits a ● the necessary illuminator could probably
good deal of promise at this early stage of de-
be made quite inexpensively, allowing it
velopment. The other candidates exhibit tech-
to be used to protect far more targets than
nical, cost, countermeasure, or public accept-
would be possible with other detector
ance problems, or require elaborate laboratory
concepts; and
separation and analysis to yield the identifica- ● the deactivator process is passive — no op-
tion code. However, as other sections of this
erator is necessary.
report make clear, the 3M taggant is not yet
fully developed or tested, and could not be Disadvantages include the fact that deacti-
generally used unless and until several remain- vation rather than detection of bombs would
ing problems are resolved. offer no help in finding the would-be criminal
Ch. 111—Taggant Research Review . 57

bombers; significant (and possibly costly) im- Vapor Taggants

pacts on current processes of manufacturing
vapor taggants have received the bulk of
detonators; and the risk of accidentally deacti-
the research on detection taggants. vapor tag-
vating detonators, resulting in their failure for
gants share the common taggant requirements
normal use. No research beyond initial concep-
of stabiIity, inertness, compatibility with ex-
tualization has been conducted for this con-
plosives, and absence from normal materials.
In addition, they must have a vapor pressure
An interesting taggant concept has been sug- sufficient to produce enough molecules to be
gested by the Franklin Institute, based on the sensed, but not so high that a large initial mass
idea of using Moss bauer active isotopes as tag- would be required to ensure continued opera-
gants. The technique involves the addition of tion when placed in explosives that have a
nonradioactive trace taggants to explosives, shelf-life of several years. They must have a
followed by the gamma ray excitation of the relatively steady molecuIe emission rate over a
Mossbauer isotopes and the measurement of 5- to 10-year shelf-life, must not produce an en-
the characteristic absorption spectrum of vironmental hazard, and must not readily ad-
those taggant isotopes. The Mossbauer effect here to surfaces with which they are likely to
has been measured in numerous common ele- come into contact.
ments, including iron, tin, and nickel. In a Several hundred different vapor sources
Mossbauer isotope, gamma rays, whose energy have been considered, with almost 200 having
corresponds to the transition energy between been investigated in the laboratory. Avenues
nuclear levels, may be resonantly absorbed of approach have included the use of dispro-
upon excitation, producing a sharp absorption portionating salts, the direct adsorption of
spectrum characteristic not simply of the vapor taggants into the elastomeric plug mate-
Mossbauer element, but of the chemical com- rial of detonators, and the microencapsulation
pound of the element. This effect is due to the of taggant materials.
small perturbations of the nuclear levels by the
surrounding electrons. For use as a taggant, a DISPROPORTIONATING SALTS
chemical compound not found in nature or A number of the salts of weak acids and
used in industry would be manufactured. Due bases, such as boron trifluoride adduct com-
to the low excitation level required, little pounds, disproportionate or separate into two
shielding of the source wouId be necessary. or more constituent parts, some of which sub-
limate at room temperatures, theoretically
Mossbauer taggants are simply a concept at providing a possible stable vapor emission
this stage, however, so little judgment can be source. Tests conducted by the Aerospace
made of its practicality, cost, or safety in ex- Corp. indicated that no compounds investi-
plosives An Aerospace Corp, analysis ques- gated had the proper balance of vapor pres-
tions the practicality of the technique. A sig- sure, emission rate, desired Iifetime, and pro-
nificant I imitation to the use of the Moss bauer jected detection limit by a sensor to allow the
and other activation techniques is that they use of a sufficiently small amount of taggant
cannot be used to search people, due to the ac- material. It is possible to control the emission
tivation radiation rate of a high vapor pressure salt by the use of
a microencapsulation membrane; use of such
A number of other activation taggant tech-
a membrane allows the consideration of a
niques have been suggested, including the dop-
large number of more easily handled liquid
ing of explosives with material that would en-
taggants, however, as described below.
hance the effectiveness of X-ray or similar de-
vices These concepts al I lack specificity, how- ELASTOMERIC ADSORPTION OF
ever, and could cause the X-ray to be triggered VAPOR TAGGANTS
by many common items, resulting in an unac- The adsorption of the vapor detection mat e-
ceptable faIse a I arm rate rial directly into the elastomer used to fabri-

61-401 0 - 80 - 5
58 ● Taggants in Explosives

cate the end plug of detonators offers a num- sive, easy to use with the candidate taggant
ber of advantages, including removal of the materials, compatible with the explosive mate-
necessity for additional steps or changes in the rials, and form membranes that account for
detonator fabrication process. Research has only 10 to 20 percent of the microencapsu-
therefore been conducted to evaluate the ef- Iated taggant weight. Figure 7 shows a photo-
fectiveness of various elastomer/taggant pairs. graph of a canadidate microencapsulated vapor
Taggants evaluated include sulfur hexafluo- detection taggant, with a needle to indicate
ride, and hologenated alkanes, amines, aero- relative size.
batics, esters, and ketones. A number of com-
binations appear feasible, although useful life- Emission rate studies are currently under-
times may be shorter than the 5-year minimum way with a number of membrane materials.
desirable. A more severe limitation, however, Early tests were very encouraging; a number of
is that the elastomerically adsorbed taggants more recent test results show variations in
would be useful only on detonators, and pos- emission rate from lot to lot and as a function
sibly with detonating cord. None of these tag- of ambient relative humidity and temperature.
gants appears to be as successful as other can- Tests have not yet started on long-term emis-
didates when microencapsulated for use with sion behavior, especially in the presence of ex-
other explosive materials. Use of separate tag- plosives. Tests have only recently started on
gants for detonators for other explosives the compatibility of explosive materials with
would lead to the development of two sensors either the taggant vapors or the membrane ma-
or to the requirement for dual-mode sensing in terials.
a single sensor, an unnecessary sensor develop-
ment constraint.
Although a wide range of detection taggant
Approximately 180 vapor materials have
materials have been proposed, the need for
been screened in the laboratory as candidate
long life, stability, specificity, and absence of
microencapsulated vapor taggants. In addi-
easy countermeasures has caused the bulk of
tion, several hundred other materials were re-
these to be rejected, at least given the current
jected after a thorough analytical review. Five
state-of-the-art. The most promising concept is
candidate perfluorinated cycloalkane com-
the microencapsulation of perfluorinated cy-
pounds have been extensively tested, and have
cloalkane compounds, although the direct ad-
successfulIy completed barrier penetration,
sorption of taggants into the detonator plug
mutagen, toxicity, and atmospheric impact
elastomer appears promising for that applica-
testing. The five candidate vapor taggants and
tion. A number of preliminary tests have been
their chemical properties are shown in table
conducted with five candidate taggants; com-
patibility testing has just been initiated. Deto-
A parallel research effort has been under- nator deactivation is a possible alternate ap-
way to find an appropriate microcapsuIe mate- proach, although little research has been ac-
rial. The optimum material would be inexpen- compl i shed.

Table 16.–Candidate Vapor Taggant Properties

Empirical Molecular Boiling point Melting point Specific Vapor pressure

Chemical name Abbreviation formula weight “c “c gravity (300° K = 27° C)
Perfluoro-1 1-2-dlmethyl-cyclobutane PDCB C, F,, 300 45 -32 1.67 390
Perfluoromethylcy clohexane ., PMCH C, F,, 350 76 -37 1.79 106
Perfluoro-1,3 3-dimethylcyclohexane PDCH C, F,, 400 101-2 -70 185 35
P e r f l u o r o d e c a l l n PFD C 0 F ,8 462 141-2 0 193 6.6
Perfluorohexylsulf sulfur-pentafluonde L-4412 CSFI,SF, 446 118 -31 1.89 195

SOURCE The Aerospace Corp

P h o t o credtt Aerospace Corp

Detect ion Taggant Sensor Systems prior to insertion of the air into the sensor. If
the vapor taggant is present, an alarm indica-
tion is registered; if none is present, then the
The development of a system to detect the
item passes through with no delay. A detailed
emitted vapors is proceeding in parallel with
procedure has not been developed to deal with
the development of vapor-emitting detection
alarms, but the procedure would probably in-
taggants. A schematic block diagram for the
clude a recycle through the sensor to eliminate
operation of such a system is shown in figure 8.
the chance of an equipment transient being re-
Air, from the vicinity of the item being in-
sponsible, followed by a suspected bomb dis-
spected, is collected and delivered to a sensor,
posal procedure if the alarm persists.
after first being conditioned. The sample col-
lector can simply consist of a gust of air for in- Work is progressing on three candidate de-
spection of boarding passengers, or can in- tection sensors. Very little effort has been ex-
clude a small pressure pulse to a piece of pended by the Aerospace Corp. on the other
checked baggage to introduce more of the air elements of the system, although some prelimi-
from the interior of the baggage into the air nary design identification work has taken
sample stream. For some of the concepts the place on the air sampling process and on meth-
free oxygen and water vapor must be removed ods of enhancing the original sample. A U.S.
60 ● Taggants in Explosives

Figure 8.— Detection Taggant Sensor System time so that one can turn-on, or gate, the de-
Block Diagram tector to respond only to a specific molecular
species or group of species such as the taggant
Sample vapors.
The taggant molecules being considered all
have long drift times and are easily separated
Air from common gasses in the IMS. Additional
sampler specificity is gained by the toughness of the
taggants; most other large molecules fragment
Sampler in processing through the detector.
IMS devices have been commercially avail-
able for approximatelys years, with about sO
currently in use for various applications. Tests
have been run with a commercial IMS unit at
airports to examine ambient air for the pres-
Alarm No Alarm ence of molecules in the critical drift time re-
gion; no molecules which would have triggered
a false alarm were detected.
Calibrator Inspection
While the laboratory tests are promising, it is
not possible to extrapolate to estimates of IMS
SOURCE: Office of Technology Assessment. performance in the field, in a real-life envi-
ronment, when maintained by normal airport
maintenance people, and when using an inter-
Customs Service device has been tested, for in-
nal calibration source.
stance, which exerts a gentle force on baggage,
causing an exhalation of the baggage interior CECD can be conceptually viewed as an IMS
air into the sampling network. device without a drift tube. It simply consists
of the conditioner and reaction chamber; the
The three candidate detection sensors are, in
decrease in current in the reaction chamber is
order of increasing complexity and cost, the
a sign that the taggant molecules are present
continuous electron capture detector (CECD),
and have been ionized. As described, CECD
the ion mobility spectrometer (lMS), and the
would have less specificity than IMS, and
mass spectrometer (MS). Figure 9 shows a sche-
would probably be triggered by a wider range
matic diagram of the operation of IMS. Gas is
of interference sources. The key to the device
introduced from the sampling device into the
is in the conditioning chamber; the chamber is
conditioner. After the free oxygen and water
a catalytic reactor that contains hydrogen gas
vapor are removed, the sampled gas molecules
and palladium metal plated onto a number 5A
are drawn into the ionization region where
molecular sieve and operating at 1400 C. The
many molecular species, including the taggant
reactor removes oxygen and water vapor, frac-
molecules if present, form negatively charged
tures some o t h e r p o t e n t i a l i n t e r f e r e n c e
ions. The negative ions are then gathered and
sources, while still others are removed by
injected into a drift tube where an electric
reduction or combustion. The number of mole-
field causes them to flow against a counter-
cules that will survive the conditioning cham-
flowing drift gas stream. By virtue of the ion
ber is limited, but the taggants may well not be
molecule reactions between the negative ions
the only survivors of the passive screening
and the neutral drift gas molecules, the ions
are separated into spatial clumps of like spe-
cies. Each species, depending on the strength CECD devices have been used as a labora-
of the ion-molecule interaction, traverses the tory instrument by the Brookhaven National
length of the drift tube in a different length of Laboratory for the past several years. A bread-
Ch. 111—Taggant Research Review . 61

Figure 9.— Cutaway View of the Phemto”Chem 100 Sensor Cell in the Ion Mobility Spectrometer


7 ml ol d
SAMPLE INLET 7 = - 7 !
. — = ’ ) :
- L . — . —————— -— ~ - M o L = E :
— —
NIr2x. AZ—= + - ‘ REKJON R EG I O< y :

100 ml Vol


100-3000 ml/min

SOURCE David Williams BATF

board device was recently shown to be quite The time required to develop instruments of
successful in detecting vapor-tagged dummy this type is a pertinent subject for discussion,
blasting caps in baggage on a conveyer belt. even assuming that the technical problems can
be solved. The milestones in a development
The MS is a standard laboratory instrument,
process include:
easily capable of resolving the taggant mole-
cuIes from other species. Current MSS, how- ● demonstration of technical feasibility,
ever, are usualIy expensive, relatively sensitive ● generation of specifications for a proto-
laboratory instruments. The challenge is to de- type,
sign and develop a low-cost, field-usable in- ● prototype development,
strument that will detect taggant molecules in ● generation of specifications for the instru-

a parts-per-tril I ion concentration level. ment,

● pilot production of the instrument, and
The limited laboratory testing of detection
● ful l-scale production.
sensors that has taken place has demonstrated
that the technology exists for sensors which None of the detection sensor concepts has
could detect the taggant vapors. These tests yet passed the technical feasibility demonstra-
have not yet demonstrated, however, the abili- tion milestone. The only time estimate which
ty of the instruments to distinguish between has been made is an extremely optimistic es-
the taggant materials and similar materials timate of 14 months from demonstration of
which may exist in the environment or may be technical feasibility to completion of a proto-
deliberately introduced into the environment type. The estimate assumed no technical, con-
as a countermeasure. It has also not been dem- tractual, or other problems, and may well be
onstrated that any of the instruments can suc- off by a factor of two. Given the fact that these
cessfully detect the taggants in the required instruments would be produced in quantity (up
parts-per-tri I I ion concentration level under to several thousand), must be self-calibrating,
field-use conditions. maintained by routine maintenance people,
62 ● Taggants in Explosives

and detect at the state-of-the-art parts-per- effective; the operating costs and false alarm
trillion level, it is unlikely that production rates would be negligible while the detection
could be underway in less than 5 years. rate would ensure essentially no successful
penetration of the sensor system.
If the instruments can be developed to per-
form as desired, however, they should be quite


Three general methods have been explored ors, and boosters. A detection device would
for detecting explosives that do not have de- thus have to be able to detect a significant va-
tection taggants added. These include vapor riety of vapors (and thus either be quite slow or
detection of the characteristic vapors present expensive) or it would be subject to a high rate
in the explosives, the use of differential con- of false alarms if it could be triggered by the
trast radiography, and the use of excitation in- spectrum of materials that would be spanned
duced emissions. Some of the specific tech- by the vapors from the common explosive ma-
niques investigated a r e b r i e f l y d i s c u s s e d terials.
A second significant problem is the amount
of vapor actually available for detection.
Vapor Detection While the equilibrium concentrations of the
vapors are high enough to ensure detection,
A great deal of research effort has been ex- the actual amount of vapor present will be sig-
pended in the field of detection of the charac- nificantly degraded by the container that con-
teristic vapors emitted by explosives. Table 17 tains the explosive, particularly if an effort is
shows the physical properties of the vapor made to create a vapor barrier. The explosive
phase of a number of explosive materials, vapors do not have the properties of penetra-
while table 18 shows some of the methods tion and nonadsorption of the vapor taggant
used to detect the explosive vapors. ’ Much of materials discussed in the previous section.
the effort has been concerned with character- Concentration of the vapors could help alle-
izing the vapors that are present in explosives, viate this problem, but that might cause suffi-
looking for vapors common to a number of ex- cient concentration of ambient interference
plosive materials, and quantifying the prob- molecules to generate a high false alarm rate.
lems of vapor detection. While the equilibrium
concentrations of the vapors shown in table 17 These defects must be balanced against the
major advantage that detection of the charac-
are within the detection capabilities of much
teristic vapors of explosives has over the detec-
of the instrumentation depicted in table 18,
several problems limit the utility of vapor de- tion of taggant vapors —only those explosives
tection. that have been tagged can be detected if the
sensors are designed to look for the vapor tag-
One of the primary problems is the lack of a gant.
common vapor in the various explosive mate-
rials. Either nitroglycerine or EGDN is often As shown in table 18, a large number of
present in dynamites, and in smokeless pow- physical principles have been used to detect
ders, but neither are present in the other ex- the vapors. The most successful, however, are
plosive materials used in criminal bombings, the ionization mechanisms exploited for detec-
tion of taggant vapors. Continued research is
such as gels, slurries, black powder, detonat-
primarily devoted to these sensors.
‘From “Explosive Vapor Detection Instrumentation, ” by j R
Hobbs, prtnted In the Proceedings of the 1 979 Electro Profes- Animal detection deserves a specific com-
sional Program, New York, April 1979 ment. Although less sensitive than the other
Ch. 111—Taggant Research Review . 63

Table 17.–Vapor Pressures of Selected Explosives

Vapor pressure Composition Mole fraction

Compound Molecular weight Temperature 0
C mm Hg gm/cm 3 (V. P./76O)
EGDN–ethylene glycol dmltrate 152 25 2,8 X 10-2 23 x 1 0 7 37 ppm
N G – n i t r o g l y c e r i n e . 227 25 2,4 X 1O-s 29 x 10-10 32 ppb
PETN–pentaerythrltol tetranltraie’ ., . . 316 25 54 x 10”6 9.2 X 10 II 7 ppb
AN–ammonium nitrate 80 25 5.0 x 10”6 2.2 x 1011 7 ppb
DNT–dinitrotluene , . 182 25 1 4 x 10”4 1,4 x 109 184 ppb
TNT–2, 4, 6,-trinitrotoluene. ~ 227 25 30 x 10”6 37 x 10”11 4 ppb
R D X 222 25 1.4 x 109 1.7 x 1014 2 ppt

SOURCE J R Hobbs Explosive Vapor Dection Instrumentations`

Table 18.–Explosive Vapor Detection Techniques

Optical Ionization Animals Other

Infrared Electron capture Bioluminescence Plezoelectnc
Ultraviolet Gas chromatography Dogs Thermoionic
Microwave Mass spectrometry Gerbils Condensation nuclei
Fluorescence Gas chromatography/ Enzymes
Laser- raman mass spectrometry
Two-photon absorption Plasma chromatography
Laser optoacoustical

SOURCE J R Hobbs Explosive Vapor Detection Instrumentations

sensors (by orders of magnitude), animals have density. Most dynamites have a specific gravi-
some potential advantages. If small animals ty of approximately 1.6; booster materials and
such as rats and gerbils can successfulIy detect military explosives are SIightly higher (up to
explosive vapors, then the cost of an animal 1,8); gunpowders have a bulk density of less
backup system would be quite small. Dogs are than 1.0.
more expensive to train and work with, but
The current imaging systems at airports are
have the advantage of being used for other law
operator-monitored and therefore dependent
enforcement work such as patrols.
on the ability of the poorly trained operator to
discriminate small density differences. Most
Differential Contrast Radiography
recent research has been concerned with auto-
Differential contrast radiography takes ad- mating the radiographic scanning systems.
vantage of the fact that different materials at- Due to the wide span in density of explosive
tenuate the strength of a source to a different materials, and the large density overlap be-
degree, depending primarily on density and tween explosives and other materials, it is
atomic number. Common clinical X-rays and necessary to include other means of discrim-
the imaging X-ray detectors used to screen ination in the detection algorithm. Shape is the
hand baggage at airports work on this princi- other discriminant currently used. The pattern
ple. Similar devices have been fabricated using recognition algorithm in a computer reacts
gamma radiation and neutrons as the beam when the proper density and shape pattern are
source. This method is quite effective for de- detected. Such a system is sensitive to orienta-
tecting materials whose density is significantly tion, arrangement, and shape of the high explo-
greater than other materials in the environ- sive as well as to the mass of the high explo-
ment, such as a steel gun (specific gravity of sive. The breadboard laboratory models so far
7.8) in a briefcase containing books or clothes developed can incorporate only a limited
(specific gravity less than 1,0), but is much less number of shape-density combinations and are
effective in detecting smaller differences in able to detect only certain shapes of C-4 explo-
64 “ Taggants in Explosives

sive and certain shapes of dynamite bombs. duced radiation whose energy may be a func-
While they could detect a 2-lb C-4 charge tion of the element itself or of the specific
shaped like a package of butter, they would compound, due to the interaction of the orbit-
not detect the same charge shaped as a sphere, al electrons with the nuclear material. The
cylinder, pancake, or sausage, or even another Mossbauer isotope taggants described in the
explosive of slightly different density shaped previous section were an example. Several
in the butter package shape. As the devices methods of utilizing induced emissions have
scan from only one axis, a 2-inch-thick slab been investigated for detection of explosives,
with a specific gravity of 0.5 looks much Iike a including the use of thermal neutrons, X-ray
l-inch-thick slab of density 1.0. Such a lack of fIuorescence, and nuclear magnetic reso-
specificity not only generates high false nance.
alarms, but explosives arranged in an unusual
shape would not be detected. The thermal neutron detection concept uti-
Two avenues of approach are being pursued lizes the capture of thermal neutrons by nitro-
to try and alleviate the discrimination specific- gen with the subsequent prompt emission of a
ity problem. The first is to use more than one 10.8 MeV gamma ray. Explosives are rich in ni-
energy level for the radiation source. Each trogen and should be easily detected in an un-
type of material has a different opacity to dif- shielded suitcase, but so are a large number of
ferent radiation energies. If more than one en- other materials, such as wool, orlon, nylon,
ergy source is used to illuminate the object, and leather. Coupling the system to a pattern
then additional information about the material recognition computer might be sufficient to
is gained. Some recent work indicates substan- discriminate between a solid block of explo-
tial gains in information are possible using two sives and a couple of orlon sweaters (although
carefully chosen energy levels. test results were marginal), but discrimination
between these sweaters and a bomb in which
The second approach is to illuminate the single dynamite sticks are connected by deto-
package along more than one scanning direc- nating cord, for instance, would be extremely
tion. The information gained can help generate difficult. Processing times for this concept are
a better idea of both the package shape and its also rather long for efficient transport of bag-
density. In a technique called tomography, the gage.
images formed by scanning from several direc-
tions are computer processed and used to gen- Nuclear magnetic resonance (NMR) is a
erate a three-dimensional image of the pack-
technique with considerably greater specifici-
age in the computer. Any two-dimensional pro-
ty. In NMR detection, an applied radio fre-
jection can then be generated as well as an ac-
quency magnetic field, with the correct fre-
curate density value. This image can be com-
quency, induces energy level transitions in hy-
pared to all possible conformations of com-
drogen, with the subseq~ent prompt reradia-
mon explosive materials by the computer,
tion of energy in a manner specific to the
yielding a much higher probability of detec-
chemical compound containing the hydrogen.
tion as well as a lower false alarm rate. Aero-
A sensor, tuned to receive the signals that
space Corp. is currently sponsoring research on
would be emitted by the hydrogen in various
dual-energy tomography, which would com- explosive materials, could theoretically detect
bine the additional information available from
any type of explosive, even when present in
both multiple directional scans and multiple small quantities. A major problem with the
energy scans.
utilization of this technique for explosive
detection would be the fact that metal inter-
Excitation-Induced Emissions feres with the NMR performance, thus shield-
ing the explosive. The unit would also have to
Many materials absorb radiation of a specif- be quite large (and thus expensive); the magnet
ic wavelength and subsequently emit an in- for an NMR unit large enough to scan a suit-
Ch, 111—Taggant Research Review . 65

case would weigh several tons. Another prob- suggested, and extremely Iimited testing has
lem is the rather slow response cycle time. been conducted on some of them. All of the
untagged detector concepts contain signif-
icant problems in terms of adaptation to field
use. Instrumentation for many of the concepts
A number of techniques have been de- would be large and expensive; many are easily
scribed for the detection of untagged explo- countermeasure and none, with the except ion
sives. Preliminary testing has been accom- of the vapor detection devices, could be used
plished on most of the techniques discussed; to screen passengers.
few concepts have progressed as far as the
studies on detecting vapor taggants, with the Granting the many problems in nontagged
exception of the use of animals to detect the detection, there may still be a significant po-
characteristic vapors of explosive materials. tential payoff. If an explosive detection instru-
Some explosive detection devices are currently ment or technique could be fielded, it could
on the market, although their performance is detect all explosives, not just those to which
not satisfactory. Other techniques have been taggants had been added.


I n 1976, the Aerospace Corp. was designated and to demonstrate their use in explosive ma-
by BATF as the system technical manager of terials. Details of the taggant and sensor devel-
the taggant program. Prior milestones leading opment programs were given above; the status
to the current taggant program development of the compatibiIity testing program is de-
effort were: tailed in chapter IV; the status of survivability
and recovery testing is reviewed in the follow-
1973.–Joint establishment by BATF and
ing section and in appendix C; some details of
FAA of an ad hoc committee on explo-
the analysis and pilot testing status are re-
sives seeding.
viewed in chapter V. This information is briefly
1973.–Formation of the Advisory Commit-
summarized below, as is a description of the
tee on Explosives Tagging chaired by
BATF implementation philosophy.
BATF for coordination of Federal agen-
cies involved with tagging and the control
of the illegal use of explosives. Program Status
1973.–Lawrence Livermore Laboratory
study to determine feasibility of identifi- The status of the taggant development ef-
cation tagging with Aerospace Corp. act- fort is summarized in table 19 for identifica-
ing as the program technical manager and tion taggants and in table 20 for detection tag-
LEAA as sponsor. gants. In the tables, “Technical feasibility”
1976.–National Implementation Model refers to a demonstration or analysis which in-
and Pilot Test Plan for Identification Tag- dicates the concept is feasible, “Technical
ging developed by the Aerospace Corp. read i ness ” refers to a demonstration or anal-
u rider contract to the Bureau of Mines, ysis that the concept will work in the manner
1977.–Aerospace Corp. designated the suggested, and “Practical readiness” indicates
system technical manager for the tagging that the full spectrum of analyses and tests has
program by BATF. been completed which shows that the concept
is ready for full-scale implementation.
Since 1977, Aerospace has been engaged in
an ongoing program of analysis and testing to The ability of the 3M Co, to produce the
develop identification and detection taggants color-coded taggants has been demonstrated,
66 ● Taggants in Explosives

Table 19.–ldentification Taggant Program Status

Accomplished I Planned or required

Technical feasibility Technical readiness Practical readiness 1 Technical feasibility Technical readiness Practical readiness
Color-coded taggant development
● Initial survivability ● Pilot production ● Leadyime study ● Tooling-up period/
compatibility compatibility testing
testing Optimize hues
● Environmental im-

pact assessment
● Health impact

Cap-sensitive packaged explosives (dynamite, water gels, slurries,
and emulsions)
● Initial compatibihty Online tagging ● Pilot test produc- Comprehensive Comprehensive sur- Analysis/optimiza-
● ● ●

testing demonstrated tion-level tagging compatibility testing vivability testing tion of approach
● Initial survivability ● Tagging methods ● Record/tracing Long-term compati-
testing selected/evaluated methods demon- bility
● Manufacturing strated
process reviewed
and practicality
Black powders
● Initial compatibility ● Online tagging ● Some ballistics Comprehensive com- ● Comprehenswe sur- ● Ballistics testing
testing ● Additional compati- testing patibility testing vivability testing ● Online tagged sur-
● Hand-mix survwa- bility (electrosta- vivability testing
bility testing tic) testing Long-term segre-
● Manufacturing ● Transport/vibration gation
process reviewed segregation testing ● Long-term compati-
and practicality bility
Cast boosters
● Initial compatibility ● Online tagging Solution of problem ● Comprehensive sur- ● Pilot testing, produc-
testing ● Tagging methods posed by reactivity vivability testing tion-level tagging
● Initial survivability selected/evaluated (and presumed in- ● Long-term compatt-
testing compatibility) with bility
● Manufacturing
I Composition B ● Comprehensive sur-
process reviewed Comprehensive com- vivability testing
and practicahty patibility testing ● Record/tracing
assessed Recovery testing methods
I ● Analysis/optimiza-
tion of approach
Detonating cord
● Taggants added by 1 0 Recovery testing ● Tagging station Comprehensive sur-
hand, initial development vivability/compati -
surwvability demon- ● Online tagging bility testing
strated ● Pilot testing
● Manufacturing

process studied and

tagging practicabil-
ity assessed
Smokeless powders
● Hand-mix surviva- Solution of problem ● Evaluation testing Ballistics testing
bility testing I posed by reactivity of sequential lots Pilot testing
(and presumed in- ● Production hazard
compatibility) with and acceptance
Herco” powder testing
● Compatibility and ● Comprehensive sur-
hazards analysis vivability testing
. Compatibility and ● Online tagging

acceptance testing
Detonators I
— I Full range 01 tests and process evaluation required
SOURCE Office of Technology Assessment
Ch. 111—Taggant Research Review ● 67

Table 20.–Detection Taggant Program Status

Accomplished I Planned or required

Technical Practical I
Technical feasibility readiness readiness 1 Technical feasibility Technical readiness Practical readiness
Microcapsule development I
Production and eval- 10 Intial compatibility studies ● Pilot production of capsules ● Competitive award /leadtime
uation of test batches 10 Complete health and atmos- studies
● Health and atmospheric
I pheric Impact assessment ● Development and testing
Impact assessment 10
Taggant selection of production
● Full-scale production capability
Dynamite, slurries, and water gels
● Compatibility testing —
Black powder
● Compatibility testing —
Cast Boosters I
● Compatlblllty testing — —
initiated The full range of analyses and tests detailed for identification taggants must be accomplished
Smokeless powder for the detection taggants, with the exception of postdetonation surivability
● Compatibility testing— — and recovery testing
initiated I
Detonating cord I
— — — I
Detonators I
● Compatibility testing I
Initiated I
Continuous electron capture detector I
● Successful bread- — — Instrument character- ● Design prototype ● Prototype field test
board demonstration I ization (in process) ● Fab and lab test ● Prototype design changes
● Instrument charac- 10 Calibration (in process) evaluation ● Final production drawings
terization (initiated) ● Aerospace lab test ● Production pilot release
● Callbration system ● Production pilot complete
(initiated) ● Field support function setup
● Training and field test
IMS detector I
● Initial feasibility Demonstration ● Design prototype ● Prototype design changes
studies (Imminent) ● Fab and lab test ● Production drawings
I prototype ● Manufacture and checkout
● Aerospace lab test engineering
Prototype field test ● Production pilot release
● Production pilot complete
● Support functions setup
● Training and field test
MS detector I
● High-cost laboratory 10 Development and bread- Prototype design, ● Prototype design changes
system testing I board demonstration fabrication, and test ● Production drawings
● Development and to be completed ● Manufacture and checkout
breadboard demon- I engineering
stration—in process I Production pilot release
I Production pilot complete
Support functions setup
I Training and field test
SOURCE Off Ice of Technology Assessment

although some hue and color code optimiza- in detail, this initial testing has revealed ap-
tion remains, as well as construction of a facili- parent incompatibilities between the 3M tag-
ty to produce the taggants. Initial compatibili- gant and one type of smokeless powder and
ty and survival testing has been completed for also between the 3M taggant and one cast
the cap-sensitive high explosives, as has pilot booster material. If and when these presump-
production of tagged explosives and activation tions of incompatibility are removed, compre-
of the tracing network. As chapter IV describes hensive compatibility and survivability testing
68 . Taggants in Explosives

must then be completed and decisions made fort; the estimate is shown in table 21. T h i s
on implementation levels before readiness is schedule does not take into account, however,
demonstrated. A similar level of testing and the need for additional compatibility and sur-
analysis has been accomplished for black pow- vivability recovery tests, particularly the res-
der, while significantly less has been accom- olution of the current smokeless powder and
plished for smokeless powder and cast boost- booster material reactivity issues, and the need
ers. One of the key remaining booster issues is for the evaluation of long-term effects of tag-
the recoverability of the taggants when gants on explosive material safety and per-
pressed into large pellets (survivability has formance. These efforts would probably add
been demonstrated). Methods of approach at least 1 year, and possibly more, to the devel-
have been explored for tagging detonators and opment time. It is unlikely that the effort to
detonating cord, but little testing has oc- demonstrate the use of identification taggants
curred. in cap-sensitive high explosives, the type of ex-
plosives with which the research effort has pro-
The significant accomplishments in identifi-
gressed farthest, could be completed prior to
cation taggant compatibility testing which
early 1981. The research on identification tag-
have so far occurred have been made possible
gants in detonators, including pilot-plant tool-
by cooperation between the Aerospace Corp.
up and testing, would not likely be finished
and the explosives and gunpowder industries.
before late 1983; the research on other explo-
Unfortunately, this working arrangement has
sive materials would probably fall between
broken down in the past few months, and the
these dates. These estimates assume a success-
industry has, for a number of reasons, with-
ful completion of each development stage.
drawn its cooperation. The result of this
Technical problems may occur that add sub-
change in the prior working relationship has
stantially to the estimate delays; continued
been a significant delay in the program, par-
lack of industry participation could make pilot
ticularly with regard to compatibility testing of
testing impossible; even resolution of contrac-
the detection taggants. The results of these de-
tual problems could add months of delay.
lays, together with an originally planned lag of
approximately 1‘A years between the identifi-
cation and detection taggant development ef-
forts, are evident in the current status of the Table 21 ,–Revised Schedule Estimates for the
development program, Identification Tagging Program
detection taggant
shown in table 20. Aerospace preliminary
Program element estimated completion date a
Development of candidate detection tag-
Identification taggants
gants is continuing. Taggants have only recent- C o l o r - c o d e d t a g g a n t Early 1983
ly been added to explosive materials for com- Cap-sensitive packaged explosives ., Early 1980
patibility testing and process evaluation. As Black powders. ., .,
Cast boosters ., ., ., Mid-1981
described previously, development of three Detonating cord ... ., ., ., ., Mid-1980
candidate sensors is also continuing, with lab- S m o k e l e s s p o w d e r s . , Mid-1983
oratory-type tests showing prom i sing resuIts. Detonator ., . ., ., ., Late 1983
Detection taggants
Mlcrocapsule development ., Mid-1981
Cap-sensitive packaged explosives. ., Mid-1981
Projected Schedule B l a c k p o w d e r Not critical
Cast boosters ., . 7
As a result of withdrawal of industry coop- Smokeless powder, ... Late 1981
eration, technical problems which have oc- Detonating cord ., ., Not critical
Blastlng caps-micocapsules. ., ?
curred, and the uncertainty of funding for out- CECD ., ., Mid-1982
year efforts, a firm schedule for the remaining IMS detector. ., ., Late 1981
development effort is not available. An esti- MS detector ., ., . ., Mid-1982
mate was made by Aerospace of the revised a Estimated by Aerospace October 1979
schedule for the remaining development ef - SOURCE Office of Technology Assessment
Ch. 111—Taggant Research Review ● 69

3M has indicated that it would need a lead- sives containing detection taggants could
time of at Ieast 22 months after receipt of a probably not be underway until mid-1 982, with
firm order before substantial quantities of tag- sometime in 1984 a more reasonable estimate.
gants could be delivered. It is unlikely that a
As indicated previously, the estimated de-
firm order would be given before resolution of
velopment schedule for the detection taggant
all technical problems, including uncertainties
sensors is extremely optimistic; a more realistic
regarding long-term effects. If a mid-1 983 date
estimate would be that production of the sen-
is assumed for resolution of al I identification
sors couId be underway by late 1984.
taggant efficacy and compatibility questions,
then explosives tagged with the 3M identifica- In summary, by early 1985 it is possible that
tion taggant could be in full-scale production all explosives manufactured could be tagged
by late 1985. with both identification and detection tag-
gants, and that detection taggant sensors
A decision could be made to implement tag-
could be in fulI production. This schedule is
ging as soon as all technical uncertainties are
realizable only if no major development prob-
resolved for some portion of the explosive ma-
lems occur and a taggant program is mandated
terials, such as cap-sensitive explosives. Under
by legislation,
those circumstances, 3M could receive firm
orders by early 1981 and tagged explosives
could therefore be in full-scale production as Implementation Philosophy
early as 1983,
BATF has publicly stated’ that it feels tag-
The detection taggant development has
gants should be included only in those explo-
lagged that of identification taggants; the
sive materials that constitute a present or ex-
development cycle may be shorter, however,
pected threat of use by criminal bombers.
due both to the learning experience of the
They feel that explosive materials that do not
identification taggant tests and to the fact that
constitute a threat could be excluded. Among
no survivability demonstration is necessary.
the materials which BATF considers appropri-
The Aerospace Corp. estimates are probably
ate for exclusion are:
quite optimistic, however, for development
and test times of both the detection taggant 1. explosives manufactured for U.S. Govern-
and the detection sensors, Few compatibility ment agencies other than the military
tests have yet been conducted. These tests, (e.g., National Aeronautics and Space Ad-
particularly the effects of long-term storage, ministrate ion); military explosives are spe-
will take at least 2 years. No specific taggant or cificalIy excluded in S.333;
encapsulation method has been chosen. Pilot- 2. special fireworks such as used for 4th of
plant production of the taggant is likely to J uly displays;
take a considerable time, as the manufacturing 3, industrial tools such as explosive bolts,
processes are complex and the reagents used switches, and air bag in flaters;
quite reactive. It is unlikely that solving the 4. blasting agents. It is the BATF intention to
technical problems and constructing proper fa- tag the boosters and detonators normally
ciIities for the large-scale production of detec- used to initiate the blasting agents. The
tion taggants can be accomplished in a signifi- explosives industry maintains that if cap-
cantly shorter period than that required for the sensitive explosives are tagged but blast-
identification taggants. Assuming completion ing agents are not, the use of ANFO by
of the compatibility tests, pilot-plant testing of bombers will increase, and BATF will then
detection taggants in the explosive materials
could be accomplished by early 1983, and
assuming 22 months from that time to the
Proposed Guidelines for Exemptions to the Requirements for
availability of production quantities of detec- Tagging Explosive Materials Bureau of AIcohol, Tobacco, and
tion taggants, full-scale production of explo- Firearms, June 7, 1978
70 ● Taggants in E x p l o s i v e s

wish to tag ANFO. See chapters 1, 11, and team has concentrated on taggant research for
VI for a discussion of this issue; and cap-sensitive high explosives (dynamites, gels,
5. explosives which are raw materials used in emulsions, slurries), boosters, detonating cord,
a fabrication process, such as the black black and smokeless powders directly con-
powder used infuzes. sumed by the public (primarily for handload-
ing), and detonators. Blasting agents would not
In addition to the categories eligible for ex-
be directly tagged; rather the detonators and
emption, certain types of explosive materials
boosters normally used to initiate the blasting
are currently exempted from regulation, and
agents would be tagged.
are viewed by BATF as inappropriate for tag-
ging, including: A strict interpretation of S. 333, at least in
the opinion of the Institute of Makers of Ex-
1. explosives used in medicine;
plosives, would not allow the Secretary of the
2. fireworks soId to the public;
Treasury to exempt explosives simply because
3. propellant-activated industrial devices,
they do not constitute a significant threat.
such as nail guns; and
Resolution of this issue may be facilitated by
4. fixed small arms ammunition.
more specific wording in the final proposed
Given that philosophy, the BAT F/Aerospace legislation.


The 3M identification taggant would have to muddy or covered the taggants with a layer of
survive the detonation of the explosive and be water, severely decreasing the efficiency of the
recoverable from the postdetonation debris to magnetic pickup.
be useful in identifying the source of the ex-
The survival test results for cap-sensitive
plosive. It is useful to separate the survival and
high explosives, under the varying conditions,
recovery discussions. Recovery of taggants un-
are gathered in table 22. That table includes all
der real-life conditions is discussed in detail in
the survival tests conducted by Aerospace with
chapter I I and in appendix C. Survival of the
taggant is briefly reviewed here. uniformly tagged explosives. Earlier tests, in
which the explosive stick was split down the
To assess the survivability of taggants in ex- center and salted, are not realistic and are not
plosives, the tests should be carried out so that discussed here. Some of the tests used unen-
recovery is maximized. ideally, tests would capsulated taggants (so indicated on the
take place on a large concrete pad or in a very table); as no difference was observed, they are
large bunker with steel or concrete walls and lumped together in the discussion.
floor. Unfortunately, few of the survivability
tests carried out by the Aerospace Corp. were Aging time was another variable tested, with
done under conditions that enhanced recov- the material being aged up to 6 months before
ery. A majority of the tests were carried out in testing; again, no effect was observed and all
a 4-ft-diameter steel-walled chamber. For all the tests are lumped together.
but the lowest power explosives, the taggants
Given the diversity of test sites and condi-
either shattered upon impact or flowed plasti- tions, it is difficult to assess each test. How-
cally due to the large impact pressure pulse ever, several trends appear clear:
(estimated by Aerospace to be between 10 and
40 kilobars (kb)). Many of the other tests were 1, Under optimum recovery conditions,
carried out in a chamber with a cracked rock using small explosive charges, many hun-
floor, or in the open on a dirt and cinder floor. dreds of taggants survive, even for Power
I n several cases rain made the open area quite Primer, the most powerful cap-sensitive
Ch. 111—Taggant Research Review . 71

Table 22.–3M Identification Taggant Survival Testing

pressure Number of Tags recovered
Explosive K bars Explosive weight, lb Test site tests (averaqe)
Independent K -10-40 1/2 4-ft diameter steel chamber 2 1,000
Coalite 8S 30-40 1/2 4-ft diameter steel chamber 10 1,000
10 (part of composite 25-lb charge) Open air, dirt, cinder floor 1 180
Gel coal -25-40 3/4 4-ft diameter steel chamber 7 75
10 10-ft cube concrete chamber, rock floor 1 4
Gel power A-2 -40 1 4-ft diameter steel chamber 8 115
10 10-ft cube concrete chamber, rock floor 1 10
12x 20x 8 ft concrete bunker 3 1,450
600/o Extra 50 ‘/2 4-ft diameter steel chamber 9 1,160
5 (Part of composite 25-lb charge) Open air, dirt, cinder floor 1 58
Tovex 800 70 ‘/2 12x 20x 8 ft concrete bunker 6 1,390
400/o giant gelatin 75 ‘/2 4-ft diameter steel chamber 5 16
(some tests with
encapsulated, some
‘/2 12x 20x 8 ft concrete bunker 6 545
Specially sensitized
emulsion 100 ‘/2 4-ft diameter steel chamber 12 620
Power Primer 135 1/2 4-ft diameter steel chamber 11 16
1/2 12x 20x 8 ft concrete bunker 13 510
1 4-ft diameter steel chamber 6 3
1 500 x 100 ft concrete pad 6 530
10 (part of composite 25-lb charge) Open air, dirt, cinder floor 1 4
25 Open air, muddy, cinder floor 1 0
25 500 x 100 ft concrete pad, rainy day 1 26
SOURCE Office of Technology Assessment

commercial explosive (excluding boost- Boosters, Military Exp osives

2. As the size of the charge increases, the Commercial boosters are normally made
percent of surviving taggants decreases from cast TNT or TNT-based explosives. These
sharply, particuIarly for the most power- explosives have higher detonation pressures
ful explosives, Under optimum condi- than even the most powerful cap-sensitive
tions, however, dozens of taggants still commercial explosives (180-200 kb v. 135 kb).
survive; even under rainy conditions 26 Calculations by the Aerospace Corp. show that
taggants were recovered from the 25-lb taggants will be raised above 4000 C, their de-
Power Primer tests. composition temperature, by booster explo-
3. Confinement sharply decreases survival, sives. Testing showed fewer than two taggants
even under optimum recovery conditions. recovered per pound of booster, even for tests
Only one test has been conducted with ex- conducted under ideal conditions on a large
plosives confined in a pipe bomb (see concrete pad. The Aerospace solution to the
chapter I I discussion); in that test scores problem is to press the individual taggants and
of taggants were recovered from 60 Per- polyethylene into a large pellet (one-fourth
cent Extra Dynamite. When that result is inch). Tests show that approximately 65 tag-
compared to the chamber survival tests gants survive in a pound booster when pelle-
(in which over 1,000 taggants were recov- tized into a one quarter-inch-diameter pellet.
ered from 60 Percent Extra) it appears Iike- Initial recovery tests indicate that the taggants
Iy that considerably fewer taggants would from boosters can be recovered, but far too
survive in pipe born b detonations using few tests have been completed to allow a de-
one of the more powerful explosives. finitive judgment.
72 ● Taggants in Explosives

Military explosives are generally at least as Detonators and Detonating Cord

energetic as boosters, presenting even more se-
vere survival problems for the taggants. Due to Only the most rudimentary tests have been
the survival issue , the excessive cost of tagging conducted of the survival of identification tag-
military explos ives and their low frequency of gants when placed on a detonator and none
use in criminal bombings, BATF does not pIan conducted with detonating cord. As the tag-
to include military tary explosives in the taggant gants are placed outside of the explosive in
program. both cases, sufficient taggants should survive
to enable a positive trace to be made. How
likely the taggants are to be recovered in real-
world s ituat ions, however, cannot be ascer-
Black and Smokeless Powders tained without testing.
Black and smokeless powders are much less
energetic than the least energetic dynamite. Summary
Gunpowders are normally used as fillers for
pipe bombs, however, so the effect of confine- In summary, the 3M identification taggants
ment is expected to be considerable, Tests with survive the detonation of cap-sensitive high ex-
both black and smokeless powders were con- plosives in large numbers for small charges
ducted in a 20-ft semicircular chamber having which are unconfined. Survival decreases as
steel walls but a sand floor. Due to the poor the charge size increases, but sufficient tag-
recovery conditions, only 2 to 3 dozen tag- gants should survive even a large charge of the
gants were recovered for the black powder most energetic commercial explosive. The ef-
bombs, and from O to 3 for the smokeless pow- fect of confinement significantly reduces tag-
der. When black powder bombs were deto- gant survival, but taggants can probably sur-
nated under near ideal recovery conditions, vive pipe bombs filled with low-energy explo-
using the 8 ‘ x 12 ‘ x 20 ‘ bunker, an average sives and gunpowder; their survival in pipe
of 1,100 taggants survived 1 lb of the FFFg bombs filled with higher energy explosives is
powder. No ideal recovery tests have been uncertain. Individual taggants do not survive
conducted with smokeless powders, but the booster detonation but pellets made from the
one pipe bomb test with explosives gives an in- taggants do. Taggants wou Id probably survive
dication that scores to hundreds of taggants the explosion of detonators and detonating
should survive. cord, but there is little or no test data.

Chapter IV
* .

9 * . * * . *
Chapter IV

The addition of identification and detection taggants to explosive materials

would constitute a significant change to the material qualification program is
therefore necessary to investigate the compatibility of the explosive materials with
the taggants. This chapter briefly discusses the involved in compatibili-
ty, describes qualification procedures in industry and for defense applica-
tions, suggests the form. that a qulification program should take to demonstrate the
compatibility of taggants with explosives and gunpowders , and describes the com-
patibilitytesting that has been reported to date.


Explosive materials are chemical systems compatibility of the explosive materials with
that I iberate a large amount of energy in an ex- the taggants, it is necessary to show that there
tremely short time. The detailed physical and is no significant change in these parameters as
chemical behavior of these reactants is not a result of the addition of taggants. The prin-
well-understood, due to the complexity of cipal parameters include:
some of the reactants and the very short reac-
tion time scale. However, the principal meas- ● energy density and rate of release,
urable parameters of the materials and their ● sensitivity,
reactions are well-known. To demonstrate ● chemical stability,

76 ● Taggants in Explosives

● electrical properties, energy density and sensitivity of the explosive

● generalized mechanical properties, and material. The balance that yields idealized
● toxicity. combustion products generally yields the
highest energy and most sensitive explosives.

Energy Density and Rate of Release The rate of energy release cannot be pre-
dicted quantitatively from basic physical and
The energy density and rate of energy re- chemical considerations but it can be esti-
lease are the two most important performance mated in a qualitative way. Energy release rate
attributes of commercial explosives and gun- can be measured accurately but the test meth-
powder. Energy density is a fundamental ods can be quite expensive and difficult. A
chemical property of the explosive material few hundredths of a percent by weight of tag-
formulation. The available energy of a given gants should not affect the energy release rate.
explosive material is well-understood, and it
can be measured with a high degree of accu-
racy and reliability. It can also be calculated Sensitivity
quite accurately from the basic chemical
Sensitivity is an ill-defined term which has
knowledge of a particular formulation. The
meaning in a safety sense, but is not definable
presence of the small amounts of taggants that
with simple direct physical constants. One rel-
are currently recommended should have only
ative sensitivity scale can be developed from
a minute effect. Limited testing has borne out
impact and friction tests, another scale from
this conclusion. 1 2
electrochemical reactions, and still another
Generally speaking, the higher energy densi- from thermal considerations. All aspects of
ty explosives tend to be easier to initiate and reactions to external stimuli must be consid-
tend to progress to a fast energy release or ered and judged with respect to practical ex-
detonation more quickly. Primary explosives perience. Then with a variety of “sensitivity “
used in caps are an exception. They are easy to numbers and functions a systems safety esti-
initiate, and build to detonation very rapidly, mate is made — not always totally scientificalIy
but do not always have a high energy density. but with an additional input from experience
and common sense.
The rate of energy release is a function of
the materials involved and the physical prox- Sensitivity tests are referenced and dis-
imity of the fuel and oxidizer components. cussed in other sections of this report, but the
When the fuel and oxidizer are in the same individual numbers are not in themselves the
molecule, as in nitroglycerine, the explosive final criteria. It is their sum total plus experi-
can release its energy on a millionth of a sec- ence which determines sensitivity.
ond time scale. Ammonium nitrate/fuel oil
mixtures, on the other hand, contain rather
Chemical Stability
large, separated fuel and oxidizer components
and thus release their energy on a much slower Chemical stability is a critical safety param-
time scale. The physical proximity of the com- eter, of paramount importance in the handling,
ponents also tends to affect sensitivity; the in- transportation, and storage of the raw materi-
timately connected materials are generally als that go into making explosives and gunpow-
more sensitive than the gross mixtures. The der and in the manufacture, handling, trans-
balance of fuel to oxidizer directly affects the portation, storage, and use of the final explo-
sive product. The stability of the explosive
1 etter, R E L u n n ( D u P e n t ) t o C [30yars ( A e r o s p a c e ) , “Tag-
products cannot be adequately predicted ana-
ging — Du P e n t Pilot Te~t S~tety a n d stabil Ity T e s t s , ” M a r 6 ,
1978, pp 5-17, 5-41, 5-42 ‘.Sa/ety a n d P e r f o r m a n c e Tests for Qual/t/cat/on oi Exp/o\/ves,
‘C Hovar$, CornpatIbI/Ity of /derrf/f/cat/on Taggant\ W ItfI Ex- IKablk, (NSWC, W O ) , R Strefau (Stre$du Ldboratorles, I nc ),
p/o~Ive\, A e r o s p a c e r e p o r t N o A T E!-78( 1 8 6 0 - 0 2 } 1 N D , August K R Hamilton (NWC), J Jones, (NWC), N~vord 0[1 44811, VOI 1,
1978 January 197.2
Ch. /V— Taggant Safety and Compatibility Review 77

Iytically, but must be confirmed by tests that Most, but not all, commercial explosives are
demonstrate the stability behavior of the prod- rather soft granuIes, rubbery or gelatinous sub-
ucts, such as long-term rates of decomposition, stances, or sometimes Iiquid-like.
interact ions between the explosive compo-
When soft substances are subjected to im-
nents, and reaction with materials into which
pact the mechanical forces are not concen-
they are likely to come into contact during
trated in a smalI volume and they dissipate as
manufacture, packaging, and end use. As an
low-level thermal waves. Stiff, brittle materials
example, picric acid and ammonium picrate,
experience strong fast compression or shock
rather powerful high explosives, which are in-
waves under impact conditions that locally
sensitive and generally quite safe, were once
produce high-energy concentrations. Local
used extensively. When these explosives come
high-energy concentrations create hot spots.
into contact with copper or copper salts, how-
This means that a hot spot can be a center of
ever, they become quite sensitive; their use is,
intense chemical reaction and therefore, in an
therefore, now quite limited.
explosive composition, a region of fast energy
release. Thus, an initiation center is created
Electric Properties when the rate of energy release exceeds its
dissipation. Grit or hard substances can create
The sensitivity of initiation of explosives by local hot spots under handling conditions pres-
static electricity and/or induced currents has ent in the mixing and packaging processes, and
always been a major concern. There are sever- especially in operations such as explosive
al modes of initiation due to electrical energy. tamping in the bore hole. As an example, a
One, inductive coupling, is serious enough to small number of hard particles has been dem-
preclude the use of electric blasting caps in onstrated to critically sensitize certain military
some operations. Direct initiation by static explosives in United Kingdom la boratories. 4
spark discharges is another mode, The energy The danger of hot-spot creation may be even
of an electric field can be coupled to an explo- greater for more, brittle explosives, such as
sive device in other ways, for example, by ther- those used in cast boosters.
mal heating of a wire or capacitance effects,
The primaries, lead styphnate and lead azide, The effects of adding taggants to explosives
are extremely sensitive to electric effects. Dry could be simulated using complex hydro-elas-
nitrocellulose and black powder are also very tic-plastic computer codes, but the calcula-
sensitive. Most cap-sensitive high explosives tions would be quite expensive. In a d d i t i o n ,
and generalIy used blasting agents are not par- lack of sufficient data on the detailed physical
ticularly sensitive to electric forces. Addition properties of the various materials would tend
of taggants to the explosive materials could to limit the reliability of such calculations. Ex-
cause a change in their electrical properties; perimental testing must therefore be under-
buildup of a static charge during the addition taken.
of the taggant to the mix could be one mode.
As analytical methods are not adequate to Toxicity
handle the problem, tests are normally con-
ducted. The decomposition products of explosive re-
actions are generalIy toxic; standard precau-
Generalized Mechanical Properties tionary measures must be taken to avoid ex-
cessive exposure. The materials used in the
The relationship of mechanical properties to taggants are generally not mutagenic or car-
explosive safety has only recently been under- cinogenic. Tests must be conducted to eval-
stood to be of paramount importance, Experi- uate the toxicity of any taggant materials
ence and intuition led the industry into ex-
plosive formulations that were not ideal chem-
ically, but have proven safe and economical.
78 ● Taggants in Explosives

whose properties are not well-known, and to tional toxicity as a result of the addition of tag-
determine if the end-product gases show addi- gants to explosive products.

A new explosive compound or formulation NAVORD Report OD 44811 specifies safety
must be subjected to an extensive series of and performance tests for qualification of ex-
tests before it can be qualified for use and plosives for the Navy. There is also a Joint
manufacture. The number and nature of the Service Safety and Performance Manual used
tests differ between various manufacturers of by all three services. The DOE procedures are
commercial explosives and between commer- similar to the DOD ones but are not docu-
cial manufacturers and Government develop- mented in a single manual. Each plant and lab-
ers such as the Department of Defense (DOD) oratory has its own rules and specifications ap-
and the Department of Energy (DOE). Tests are proved by the director. There are certain pro-
specifically designed for the explosive prod- cedures and test methods that are common to
uct, the environment it will be subjected to, all, however, which are briefly discussed in this
and its end use. It follows that an extensive sect ion.
battery of tests are required for each explosive.
The initial testing is done on small quantities
Interpretation of the tests, including the validi-
on a laboratory scale, usually less than a gram.
ty of some prescribed ones, is not straightfor-
Drop weight impact tests are always done, fol-
ward and a single number derived from a test
lowed by friction and thermal test such as
or tests cannot alone define its safety. The
DTA, DSC, Taliani, or others. The results of a
closest that one can come to a measure of ex-
statistically significant number of tests are
plosive safety is the long-term accident record.
then compared with known standard explo-
[t is important to realize that experience plays
sives. If the tests give satisfactory results, then
a role equal to good scientific understanding
a laboratory or plant level management deci-
and execution of prudent, conservative prac-
sion, usually backed up by a safety committee
tices. The decisionmaking process as to
review, will give a go ahead to make limited
whether or not the new explosive and process
quantities sufficient to do the preliminary per-
of manufacture are safe is therefore unique to
formance tests such as detonation velocity,
each organization.
detonation pressure, and shock sensitivity.
These tests usually require several pounds of
In general, the qualification procedures de-
the new explosive to complete. At this stage
scribed in this section are those followed by
more elaborate chemical compatibility and
agencies or companies that routinely develop
thermal stability tests are also run along with
new explosives or significant modifications of
some accelerated aging tests. The small-scale
existing explosives, including Government
laboratory tests are repeated at this stage and
agencies such as DOD and DOE and some
compared with the original results. Unless all
manufacturers of commercial explosives.
test results are satisfactory, further work on
Companies that rarely develop new products
the new explosive will be stopped.
do not generally need a comprehensive qual-
ification program. Within those organizations If results are satisfactory and if the per-
that do have a comprehensive program, the formance is as desired then a management de-
complexity, qualification time, and cost vary cision beyond the laboratory level will gener-
considerably, due to differing manufacturing ally be made to proceed with Iimited pilot pro-
procedures and end uses. As an example, com- duction. As much as several hundred pounds
plete qualification of a new military explosive may be involved. It is at this stage that manu-
can take several years with a total cost of facturing hazards are assessed. Special tests
many m i I I ions of dolIars. will usually evolve at this stage that will relate
Ch. IV—Taggant Safety and Compatibility Review ● 79

to the actual manufacturing equipment such category of the explosives used in the pro-
as pipe diameter in which a liquid explosive or gram. 5
slurry will or wilI not propogate a detonation.
The aspects of quality control are addressed
Exact details of equipment and controls are
during the pilot phase of development. Chem-
then reviewed. In the case of addition of tag-
ical and physical test specifications are estab-
gants there is the possibility of buildup of the
lished to control all component raw materials.
material in some part of the mixing or car-
Incoming taggants must be examined for for-
tridge-loading machinery. Consideration is
eign material and their code verified. If the
given to fail-safe controls in the event of power
taggants are gritty, such as the Westinghouse
failures or other equipment failures. Transpor-
ceramic particles, there must be assurance
tation of raw materials and finished product
that each taggant is properly coated with the
within the plant is planned. Barricades and re-
desensitizing polyethylene or wax. Similarly,
mote control are planned where required. For
sampling and test schemes for product quality
example, the pressing of booster pellets of
assurance are set up at this stage.
Tetryl or PETN is a hazardous operation and
must be done by remote control and the press In some cases a company’s management
itself barricaded so that no personnel are ex- may decide that the change involved in the
posed in case of an accidental explosion. Stor- new explosive is smalI and complete requal-
age in magazines must also be planned. if ication is not required. The extensive experi-
ence the management has developed in the
If the new product has passed its perform-
history of its plant and products makes this, in
ance and safety requirements in the pilot
many cases, an acceptable procedure. Al-
study, a parallel effort of evaluating the new
though taggants would be added in only a
explosive in its use environment is made. Here
smal I amount by weight, their use in explosives
DOD and DOE differ significantly from indus-
is sufficiently different from other constituents
try. Military weapons are subjected to many
that it is the general consensus of manufac-
extreme environments and the finished weap-
turers and other parties that addition of tag-
on with the new or modified explosive must
gants will require complete requalification of
undergo special safety testing to qualify it.
al I tagged explosives.
Commercial explosives generally are used in
somewhat more benign environments and the
end-use safety testing is more limited and less Description of Qualification Tests
expensive. End-use testing is required for per- Normally Performed
missible explosives (i. e., explosives that have
been approved by the Bureau of Mines for use Testing of explosives involves a wide variety
in underground coal mining operations). Their of tests which must ascertain chemical compo-
cap sensitivity, toxic fume production, and sition, performance, sensitivity, and stability.
failure diameter must be established. For ex- Chemical composition analysis is a dominating
ample, the minimum size bore hole required factor since it is obvious that the manufacturer
for a particular permissible explosive to func- and user must know what he is using and what
tion properly must be determined, as well as he has made. Chemical analysis methods are
the safety of use in the underground coal envi- not the direct concern here, as taggants
ronment (incendivity testing). change the composition little, but it is to be
emphasized that knowledge of the chemical
Samples from pilot production must, at this composition must be a part of qualification
stage, be submitted to the Department of assessment.
Transportation (DOT) for determination of
shipping category. DOT has stated that addi- ‘Letter, P J Student (As$oc of Amer Railroads) to R B Moler
tion of taggants does not change the shipping (Aerospace), June 27,1977
80 ● Taggants in Explosives

There is a large number of tests that are spe- diameter at which 50 percent of the tests prop-
cific to evaluation of an explosive product. agate to a high-order detonation is the critical
The details of these tests are given in several or failure diameter.
sources. 6 10 The most commonly used tests are
The chamber pressure of gunpowder is
briefly described below.
measured by the use of spherical copper crush
Performance gauges or by transducers placed in the cham-
ber. Burn rate is measured by a variety of
Performance is determined by measuring methods, often by placing the powder in aV-
detonation velocity, detonation pressure, pres- groove, igniting one end, and measurin g t h e
sure rise rate, shock sensitivity, and failure di- velocity by high-speed camera, thermocouple,
ameter in explosives and ballistic properties or pressure transducers. The muzzle velocity
such as burn rate, muzzle velocity, and cham- of the propelled projectiles can be measured
ber pressure in gunpowder. The addition of by a variety of methods, includin g p h o t o g r a -
smalI amounts of inert material to an explosive phy and make or break switches.
probably will not effect its performance sig-
n if i cant I y; however, performance must be Impact
demonstrated. Detonation velocity measure-
ments consist of placing electric probes in Impact tests, although variable in nature
precisely measured positions, detonating the and sometimes difficult to interpret, are criti-
explosive, and measuring the time that it takes cally important; their relationship to safety is
the detonation front to pass between the obvious. They quickly provide information
probes with high-speed electronic equip- that categorizes the level of hazard of an ex-
ment. 11 12 Initiation or shock sensitivity tests plosive composition. They normally are used
are done by separating a donor explosive from to tell if significant differences exist between
the test acceptor explosive by a measured gap. explosive samples. Impact tests are not infalli-
The gap is varied until a 50-percent probability ble and the results must be considered in rela-
of explosion of the acceptor explosive is estab tion to other type testing.
Iished. Impact tests range from laboratory-scale
Detonation pressure and pressure rise rate tests involving less than 35 mg to large-scale
are measured by inserting transducers into the drop tests amounting to as much as 50 kg. As
explosive material and recording the resultant indicated previously, the initial tests would be
pressures on fast response rate electronic laboratory-scale tests.
equipment. Critical diameter testing, to estab-
All laboratory impact machines are similar
lish the failure diameter of an explosive mate-
in principle. The energy source is a free-falling
rial, is accomplished by attempting to deto-
weight which impacts the explosive sample
nate varying diameters of the explosive. The
through a mechanical linkage. Criteria are es-
6“Safety and Performance Te\t\, op c it tablished for distinguishin g between positive
‘)o/nt Service .Sa/et y anci Performance Manual tor Qua//i/cat/on
and negative responses. The criteria differ for
of ~ xp/o JILw\ for M///tar y U\P (Ch ln,~ Lake, Ca I If Nava I Wea pens
Center, September 1971 ) various laboratories so comparisons are only
‘G R Walker, CARDE, Canada, E G Whltbread, ERDE, United valid when made in a single laboratory. The
K Ingdom, D C Horning, NSWC/WO, U S A , The Technica/ Co- tests consist of dropping the weight from vary-
operation Program Manual of Sensitiveness Tests, TTCP P a n e l
0-2, February 1966 ing heights onto samples of test explosives
“K R tlecker, C M Ma\on, ~nd R W Wat\on, B u r e a u o f placed between them — sample weights are
M)nef /n~trurnentec/ /mpact Tester (Bureau ot M i n e s ) RI 7 6 7 0 ,
usualIy about 50 to 100 mg. The results are re-
‘“R W Wdtson, (arci(;ap a n d Pro/ecti/e /mpact Senslflvity corded as a go or no-go. A statistical analysis
Measurements, a compilation, 1 C 8605, 1971 of the data determines the relative stimulus
‘ ‘Safety and Performance Tes(~, op clt level correspondin g to a chosen level of prob-
‘‘~ M Mason and t G Alken, Methods for [ valuatlng Explo
~lve$ and Ha/ardou\ Mater/a/$ (Pltt\burg Mlnlng and Safety Re-
abiIity that the explosive will react to give a
search Center, Bureau of Mine\), report No 1 (“ 8541, 1971 positive result accordin g to the arbitrary cri-
Ch. /V—Taggant Safety and Compafibility Review . 81

teria, 13 14 15 Some manufacturers report a 50- Stability

percent probability height, but most report a
Stability testing may be divided into two
threshold height.
general categories. One is simply long-term
Bullet tests are done by firing buIlets or pro- storage in which samples are removed period-
jectiles, usually .22, .30, or 50 caliber, into the icalIy and retested to see if a significant
test explosive. Powder loads are varied to ob- change has occurred. The second category in-
tain a range of projectile velocities. The test volves accelerated aging, which generally
explosive may either be essentialIy unconfined means subjecting the test sample to extreme
in an ice cream carton, or highly confined in a temperature environments and then measuring
heavy steel pipe, The minimum velocity re- the effects of the environment. Stability tests
quired to obtain a reaction is reported, 6 normalIy conducted include the above-de-
scribed friction and performance tests, plus
Friction tests which are basically thermal in nature.
These thermal tests provide a measure of some
I n the manufacture, handling, and use of ex-
physical chemistry parameters of the explosive
plosives there are many situations where fric-
as well as being measurements of stabiIity.
tional forces either are or could be present.
Several test methods have been devised over Among the stability tests widely used are:
the years and two of them have been used ex-
Differential thermal analysis [DTA) in which
tensively in evaluating the taggants. In the
identical containers, one containing the sam-
Bureau of Mines tester a sample is placed on
ple and the other a standard reference materi-
an anviI and subjected to the glancing, rubbing
al, are set up in identical thermal geometries
motion of a weighted shoe attached to the end
with temperature sensors arranged so as to
of a pendulum that swings freely over the an-
give both the temperature in each container
vil. The shoe is either mild steel or a specified
and the difference in temperature between the
phenolic resin-bonded composite. The other
containers. The data are displayed as a DTA
test, developed by commercial industries, uti-
thermogram in which this temperature differ-
lizes a 2-kg torpedo which is released to slide
ence is plotted against the temperature of the
down a V track and obliquely impact the test
sample. Such a plot is almost a straight line if
sample. Both the height and angle of impact
the sample has no rapidly changing thermal
are independent variables, 17
behavior. Excursions below or above the base-
A new precision instrument developed in line are due to endothermic, that is heat ab-
West Germany and known as the BAM (after sorbing, or exothermic, that is heat releasing,
the Bundesanstalt fur Material prufung which reactions. The DTA analysis permits the inter-
developed it) seems to demonstrate improved pretation of phase changes, decomposition,
discrimination. Some of the permissible will and melting points; from these, some kinetic
be tested on this new machine at the Bureau of information on thermal stability can be ob-
Mines. 18 The friction surfaces in this device are tained. Sample sizes are in the order of 20 mg.
ceramic. The load on the moving friction sur- Since the temperature of the thermal event is
face is varied until a response level is estab- dependent, to some extent, on the heating
lished. rate, various heating rates are normally used.
The standard rates are 100 C/rein and 20 C/rein.

Differential/ scanning calorimetry is very

similar to DTA except the energy difference
(calories) between the standard reference ma-
terial and the explosive is recorded during the
time-temperature program.
Vacuum stability is measured by placing a 5-
mg sample in a gas burette and then evacuat-
82 ● Taggants in Explosives

ing the burette. The flask containing the sam- there for a number of hours. Visual evidence of
ple holder is then heated to an appropriate decomposition is sought as well as the meas-
temperature for 20 to 48 hours. The gas urement of endothermic and exothermic reac-
evolved is measured by the manometer con- tions.
nected to the sample flask and then normal-
T h e abel heat t e s t c o n s i s t s o f h e a t i n g
ized to standard temperature and pressure.
samples in contact with methyl violet paper,
Test temperatures specified for military ex-
usually at 71 0 C. The elapsed time before the
plosives are 1000 C and 1200 C. Dynamites and
paper changes color is recorded. The test is ap-
slurries are less temperature-resistant and usu-
plicable only to explosives containing nitrate
ally contain volatile compounds; therefore,
ester. A similar test, the German test is done at
the test is really only useful for candidate
1200 C and a minimum time of 40 minutes
booster materials, gunpowders, and explosive
allowed before a color change.
components of detonating cord.
When the stability of an explosive is being
The Taliani test is almost exactly the same as
compared to the stability of that explosive
the vacuum stability test except that the test is
after an additive (such as the taggant) has been
usually run in a nitrogen atmosphere at 750 C
incorporated, the tests are normally conducted
at some laboratories and 93.30 C at others; tag-
with significantly increased concentration of
gant tests in one laboratory were run at 1200 C.
-that additive. Thus, while only 0.05 percent by
At the end of 1 or 2 hours, the apparatus is
weight of taggants is proposed to be added to
vented to 1 atmosphere to eliminate the effect
explosives, stability tests are conducted with
of the vapor pressure of water and the expan-
taggant concentration as high as so percent.
sion of the original gas. The pressure change
between 2 and 5 hours is measured.
In the chemical reactivity test (CRT) a sample Incendivity Testing (The Gallery Test)
of the explosive, approximately 0.25 g, is usual- Incendivity testing is done to certify ex-
ly heated under a helium blanket at 1200 C for
plosives and blasting assessories for use in
22 hours. Tests have been conducted at other
underground mines. Permissible explosives are
temperatures and times; tests with the West-
those that pass the proscribed incendivity test.
inghouse taggants in dynamites were run at
An explosive charge, which is loaded into a
1000 C for 4 hours. A cryogenic gas chromatog- steel cannon (mortar), is fired directly into the
raphy unit is then used to measure the individ-
gallery chamber containing a flammable mix-
ual volumes of the product gases, including
ture of natural gas and air or natural gas, air,
such species as nitrogen oxide, carbon monox-
and coal dust. There are two large gallery tests
ide and dioxide, water, and other gases as may
for explosives. on one test the incendivity is
be determined necessary. This test is used prin-
measured in mixtures of coal dust and natural
cipally to determine the reactivity of explo-
gas in which the gas concentration (4 percent)
sives with other materials, i.e., a compatibility
is below the explosive limit of the mixture. In
the other, the incendivity of explosives is meas-
In the hot bar test a bar is heated to 2500 C ured in the presence of an 8-percent natural
and test samples of explosive are dropped on gas-air mixture.
it. In the hot tip test, a 7\8-inch square by 1\8
The gallery represents a coal mine face, and
inch-thick piece of steel is heated to white heat
is a 6-ft, 4-inch diameter steel tube, 80 ft long.
by means of a Presto-Lite torch and dropped The first 20 ft are charged with the flammable
on a test sample.
air/gas mixture and isolated by a thin mem-
The stability bath test measures an exotherm brane from the remaining 60 ft of tube which is
and, therefore, decomposition at elevated filled with air and acts as an expansion vol-
temperatures. It is similar to the DTA, but uses ume. I n the 4-percent concentration test, 1‘A -
larger samples. The sample is generally heated lb charges of the explosive are fired in the can-
to a predetermined temperature and retained non under specified conditions. Ten trials are
Ch. IV—Taggant Safety and Compatibility Review ● 83

made; if any explosion occurs the explosive chute (grounded stainless steel, 2 ft long), and
has failed the test. In the 8-percent concentra- an ungrounded stainless steel catch container
tion version, the amount of explosive that is with a known capacitance connected to an
being treated is varied from shot to shot to es- electrostatic volt meter. The taggants were
tablish the weight required to cause a 50-per- poured from a polyethylene container, down
cent probability of ignition. 19 the chute into the catch container. The charge
developed is calculated from the voltage. The
Cap Sensitivity relaxation time is determined by the time re-
quired for the charge to dissipate. The charge
This test provides a simple means for differ-
generated, and relaxation time, can then be
entiating an explosive from a blasting agent. A
compared to materials commonly added to ex-
No. 8 detonator is inserted into a sample of
plosive materials, such as aluminum powder.
given size and fired. If the sample is initiated
to detonation, the material is classified as an
explosive. A material that is not initiated to Elements of a Taggant Compatibility
detonation is classed as a blasting agent. The Qualification Program
test is used by the Bureau of Explosives to
establish its shipping classification. The sam- Taggants are a sufficient departure from the
ple is put into a container at its approximate materials normally used in explosives and gun-
packaged density and a No. 8 detonator is in- powder to require full qualification of the
serted through the cover. The assembly is new taggant-explosive material composition.
placed on soft ground in an isolated, safe- While the taggants are fabricated from quite
guarded area, and the detonator is fired. If a inert materials and are to be added i n amounts
crater is formed, the sample is considered to of only a few hundredths of a percent by
be cap-sensitive. The sample container is a 1- weight, the conservative safety philosophy of
qt, spiralwound, paperboard cylinder with the explosives industry makes requalification
cover, of the type used commercially for food necessary. As the detailed physical chemistry
packaging. Any commercial No. 8 blasting cap of the explosive reactions is not completely
may be used as the detonator. understood, it is not possible to safely conduct
a few spot tests and generalize to alI explosive
Spark Sensitivity materials from these tests. Table 23 outlines
the elements of the type of qualification test
T h e m e t h o d o f d e t e r m i n i n g sensitivity to program considered adequate by the OTA
spark initiation is to subject the material to study team.
single discharges from a capacitor charged to
a high voltage. The maximum energy of the In principle, the manufacture of explosive
spark discharge to which the material can be materials consists simply of adding together
subjected without being ignited is a criterion the fuel, oxidizer, sensitizers, and stabilizers,
of its sensitivity. Results are expressed as the mixing the components and packaging them in
maximum energy, in jouIes at 5,000 v, at which
the probability of an ignition is zero. 20
Table 23.–Elements of a Suggested Compatibility
Qualification Program
Charge Generation
● Unique with each manufacturer.
Taggants are electrically nonconductive. A
● Analysis to define the new explosive or ingredient
charge can be generated on them by pouring ● Laboratory testing+impact, friction, thermal, chemical composition,

the taggant into the mixer; a charge generation electrical aging, chemical Interaction, performance
● Pilot production
test was therefore devised by one manufac- ● Commitee and management review
turer. The test apparatus consists of an angled ● Early production and review

Special tests.
● Experience
l k e r , et dl 0 1 ) c It
‘“K w Wdtw)ll, ()[) ( It SOURCE Off Ice of Technology Assessment
84 ● Taggants in Explosives

a casing (most explosives) or granulating the manufacturing process and end use, the results
mixture (gunpowder). I n practice, however, of the analysis, and the standard procedure of
each explosive mixture of ingredient is com- the manufacturers. At a minimum, tests must
bined and processed in ways that differ sig- be conducted to demonstrate that the addition
nificantly for each manufacturer. The number of taggants to explosive materials does not in-
of ingredients used can vary from 2 (for ANFO) crease their impact and friction sensitivity;
to 10 or more for some explosives and smoke- does not detrimentally alter the thermal,
less powders. The mixing process used can vary chemical, electrical, or storage properties of
from the simple mixing of ammonium nitrate the materials; does not decrease stability; does
and fuel oil to form ANFO to a complex proc- not alter the chemical interactions involved
ess involving preparation of the basic ingredi- (by eliminating interactions originally present
ents (one manufacturer grinds all ingredients or by introducing new interactions); and does
to a 300 mesh powder for instance) and several not adversely affect the performance of the ex-
mixing and processing stages. The equipment plosive material,
used also varies widely, from the wooden mix-
After the small-quantity laboratory tests and
ing equipment used by one manufacturer of
the analysis are successfully completed, pilot-
nitroglycerine-based dynamites to the complex
plant scale production should be initiated to
continuous process equipment used by one
investigate potential problems involved in the
manufacturer of emulsions. End uses also vary;
manufacturing, packaging, and storage of the
soft dynamites are often dropped or otherwise
tagged explosives and gunpowder. This test-
subjected to impact forces which would be un-
ing should simulate, as nearly as possible, the
safe if used with more brittle explosives such
actual manufacturing processes to be used if
as TNT boosters. For these reasons, the qualifi-
tagged explosives were to be produced.
cation program must be unique to each manu-
facturer, and must reflect the exposure ex- Reviews, both technical and managerial, are
pected during the manufacture, storage, trans- an integral part of the qualification process.
portation, handling, and use of that particular Substantive special reviews would probably be
product. held at the end of the small-scale laboratory
testing phase and at the end of the pilot pro-
While it is true that the state of the art and
laboratory instrumentation of physical chem-
istry are not sufficiently advanced to provide a Through their qualification process the man-
detailed understanding of the process involved ufacturer would gain a great deal of experi-
in all explosive reactions, it is certainly true ence in handling and workin g with the tagged
that a careful and thorough analysis of the explosives. This experience, and the general ex-
probable effect of adding taggants to explo- perience gained by working with the untagged
sive materials can provide a great deal of in- explosives, and with other explosives, repre-
formation. This information can be used as a sent an important, although qualitative, part
preliminary screen to eliminate obviously dan- of the qualification evaluation process. For
gerous explosive-taggant combinations, such this reason, it is desirable for the manufac-
as taggants placed directly in primary explo- turers to conduct at least a large part of the
sives or the use of gritty taggants. In addition qualification process. Some manufacturers do
the analysis can suggest critical tests and pro- not have the requisite facilities and personnel
vide insight into the expected result and their to conduct the initial analyses and laboratory
interpretation. Proper analysis must therefore testing. This testing can be accomplished by
be considered the first element of any com- outside agencies. It is obviously necessary for
patibiIity qualification program. the manufacturer to participate in the pilot-
scale testing phase.
Laboratory testing must obviously play the
central role in a qualification program. The ex- In the taggant compatibility testing which
act tests to be performed are a function of the has taken place (presented below), the manu-
Ch. IV—Taggant Safety and Compatibility Review ● 85

facturers were asked to suggest critical tests minimum program, such as described above,
that were required before the pilot test manu- must be conducted; additional tests, suggested
facturing and distribution program could take by the manufacturer, may be made a part of
place. That process is not sufficient for a for- the program.
mal compatibility qualification program. A


Several hundred individual tests have been mine the nature and cause of the reaction, the
conducted in an effort to define the compati- extent of the safety hazard created, and what
bility of identification taggants with explosive remedial steps may be feasible. Extremely Iim-
materials. These tests have generally been ited testing has indicated no significant change
paired tests in which the reaction of a specific in balIistic velocity or chamber pressure when
explosive material to a specific test is com- the 3M taggants are added to smokeless pow-
pared to the reaction of that material when ders, even at extremely high taggant concen-
identification taggant have been added. Mate- trations.
rials tested include dynamite and other cap-
The hard 3M taggants (types B and D) did
sensitive high explosives, cast boosters, black
cause significantly increased sensitivity in cap-
powder, and smokeless powder.
sensitive explosives, as did the Curie-point tag-
Several varieties of identification taggants gant and the unencapsulated Westinghouse
have been tested, including the current 3M taggant.
baseline taggant in both encapsulated (type C)
Compatibility testing for the detection tag-
and unencapsulated (type A) form; a harder,
gant materials has been recently initiated with
more highly cross-1 inked variety of the taggant
black powder and cap-sensitive high explo-
(type B); a higher melting point variety (type
sives. No data has been formally reported; tox-
D); the Westinghouse ceramic taggant; and the
icity and mutogenacity tests of the materials
Curie-point taggant.
themselves have been negative.
No tests have shown increased explosive
The following paragraphs briefly summarize
sensitivity due to the addition of the baseline
the tests so far conducted. The extent of test-
3M taggant (either encapsulated or unencapsu-
ing described in the tables includes those
Iated). Similarly, no changes in electrical,
whose results had been formally reported by
general mechanical, or toxicity characteristics
March 1, 1980. However, OTA has reviewed all
have been noted, Decreased chemical stability
testing about which information was received,
was noted, however, for one type of smokeless
whether or not formal reports have been
powder (Herco @ ); 21 2 2 decreased stability was
issued. Tests are continuing.
also noted in one type of booster material
(Composition B). The tests conducted to date
clearly show that some chemical reaction
takes place when Herco @ powder or Composi-
tion B is mixed with a high concentration of The paired compatibility tests conducted
3M taggants and then heated to a high tem- with dynamite and with EDCN are summarized
perature; further research is required to deter- in table 24, In this table and those which fol-
low in this section, an asterisk by the taggant
type indicates a sensitization or other indica-
tion of noncompatibility The other symbols
are defined in the legend. As can be seen from
the table, no significant differences in re-
sponse to the various tests evaluated were ob-
86 ● Taggants in Explosives

Table 24,–Summary of Compatibility Tests Conducted With Dynamite and Dynamite Ingredients

Test type
Drop Sliding 5-kg static Chemical
Type of dynamite w e i g h t Friction r o d impact discharge Heat D
( aTbAo l ) reactivity pH
V i b r o g e l A,C A,C c c
R e d H A A,C A,C A,C c
Tamptite gelatin extra 6 0 % A. , C A,C A,C c
U n i g e l A,C A,C A,C c
G e l o b e l A A A,B* x * A,B* ,W*
EGDN C,W, X* A, B*, C, C , W * C, W, X,D, E W.x ”
W, X,D, E x A’
Nitroglycerin. c c A’ c A’
9 0 / 1 0 E G D N / N G : : C,Y,Z* D* c
60% ammonia gelatin w w w
60% semigelatin w w w w
4 0 % s p e c i a l w w
8 5 % h y d r i v e . w w
850/o gelatin w w w w
Gelatinous permissible w w w w
60/40 NG/EGDN w w A’
Power Primer A’ ,C’ Y*, E,A*, C A*, C,D* A’
Y*, Z*
A– unencapsulaled 3M laggant X–unencapsulaled Westinghouse ceramic taggant
B–unencapsulated hard cross.linked 3M taggant Y–encapsulated Curie-point Iaggant
C–encapsulated 3M Iaggant Z–unencapsulah?d Cunepolnt faggant
D–encapsulated higher meltlng point 3M taggant ‘ indicating Irradiated taggant
E –unencapsulated higher melting point 3M taggant “–md[caled noncompatibility
W–encapsulated Westinghouse ceramic taggant

SOURCE Office of Technology Assessment

served for any of the dynamites into which other signs of decreased stability appeared in
either the encapsulated or unencapsulated the other tests.
baseline 3M taggants were added. Unencapsu-
Iated hard or gritty taggants of various sorts
Gels and Slurries
caused sensitization under impact testing.
A smaller number of tests was conducted to
addition to those tests shown in the table,
compare the response of tagged and untagged
a small number of drop weight tests were con-
gels, slurries, and emulsions. These tests are
ducted in which the 3M taggants (both base-
summarized in table 25. I n no case tested was
line and the cross-linked varieties) were encap-
there an indication of changes in sensitivity or
sulated in several high melting point resins.
stability due to the presence of taggants. Tests
Sensitization of both Power Primer and 90/10
were also conducted to determine if the addi-
EGDN/NG were noted for most combinations
tion of taggants to the gels and slurries would
affect performance as the explosive materials
A final series of tests examined the stability aged. Tests included initiation sensitivity and
of tagged Power Primer, Coalite-8S, and EC DN detonation velocity as well as visual observa-
under both accelerated aging (higher tempera- tion of gel quality. Both ambient and acceler-
ture) and ambient aging conditions. The Power ated aging tests were conducted. No changes
Primer showed a significant decrease in stabili- in these properties were observed. Cap-sensi-
ty as measured in the Abel test after 2 months tivity tests at low temperature were also con-
aging at 400 C. Unfortunately, no control test ducted with special sensitized emulsions con-
was conducted with untagged Power Primer, taining a combination of the baseline 3M and
so no compatibility judgment can be made. No the Westinghouse taggants. The performance
Ch. IV—Tagganf Safety and Compatibility Review ● 87

Table 25.–Summary of Compatibility Tests Conducted With Gels and Slurries

Test type
Weight Electro-
Drop Sliding Projectile Chemical Thermal loss under Hot Hot static
Type gel or slurry weight rod impact Friction stability stability Taliani heat tip bar disch
G e l - p o w e r A - 2 A,C A,C c
● H20, MMAN, SN, AN A c
Mixture of tovex 700, tovex 800, tovex 320 C c c c c
G e l - c o a l c c c c c
Gel-powder c c c c c
Permissible (unspecified) W W w

A–unencapsulaled 3M taggant Y–encapsulated curie-poml taggant

B–unencapsulated hard cross linked 3M taggant Z–unencapsulated curie point taggant
C–encapsulated 3M taggant ‘ indicating iradiated taggan
D–encapsulated higher melting point 3M taggant “MMAN –monomethylamme nitrate
E–unencapsulated higher melting point 3M taggant SN –sodium nitrate
W–encapsulated Westing ceramic taggant AN–ammonium nitrate
X –unencapsulated Westinghouse ceramic taggant

SOURCE Office of Technology Assessment

of the tagged explosives was superior to the this nomenclature difference, the tagged com-
untagged control samples. It should be noted position B showed significantly more severe
that the reason for any change in performance degradation at the 120° C test temperature
should be carefully investigated. than did the untagged composition B at a 1300
C test temperature. As no control tests were
Cast Boosters conducted with an untagged batch of explo-
sives for the Octol and Pentolite tests, it is im-
The tests comparing the sensitivity and sta- possible to ascertain if the taggants were re-
bility of tagged and untagged cast boosters are sponsible for the observed reactions. While
summarized in table 26. The 3M taggant did testing is often conducted at temperatures
not affect the sensitivity of any of the cast above those encountered in normal use, it is
boosters explosives in any of the paired test- extremely dangerous to heat common booster
ing. Evidence of decreased stability was ob- materials to temperatures above 1200 C. The
served in tests conducted of molten booster test serves as an indication of a potential com-
material to which 3M taggant had been added. patibility problem. More carefully controlled
I n a series of tests, Goex heated booster explo- tests are currently underway at the Naval Sur-
sives to temperatures between 1200 and 1650 face Weapons Center, White Oak, Md. Prelimi-
C for a period of 16 hours. ” Evidence of de- nary indications are that a 50-50 mixture of un-
composition of the explosives occurred, in- encapsulated taggants and TNT undergoes a
cluding bubbling, dislocation, and the appear- chemical reaction at 1200 C; research is con-
ance of voids. Pentolite (50/50 PET N/TNT), Oc- tinuing to determine the nature, cause, and
tol (25/75 TNT/HMX), and an explosive mixture safety significance of this apparent incompat-
similar to Composition B were tested. The only ibi I it y.
paired test was with the Composition B-like
material. Composition B normally contains On July 15, 1979, an explosion and fire oc-
just under 30 percent TNT and just under 60 curred at the Goex factory in Camden, Ark.,
percent RDX, with the rest being wax. The causing damage which Goex has estimated at
Goex mixture used A-3 instead of pure RDX. As $2 million. The explosion took place in a melt-
pour operation in which scrap high explosives
A-3 contains approximately 9 percent wax, the
were being melted. Goex, Inc., asserts that the
composition of the Goex Composition B dif-
scrap materials avaiIable for melting down in-
fers from standard Composition B. Ignoring
I I L~tt~r j w H~rOn (GO~X, I n c ) t o S []erda (Aemspdc ~), cluded some materials containing 3M identi-
“StdtlJs ot Tdgg, !ng P r o g r a m , ” A e r o s p a c e p u r c h a s e order W - 0 2 5 , fication taggants. Goex further asserts that the
l a b rept DTL) 10479 explosion began in a way that resembled the
88 . Taggants in Explosives

Table 26.–Summary of Compatibility Tests Conducted With Cast Boosters

Test type
Vacuum BAM Pendulum Thermal
Type of booster Drop weight stability friction friction Sliding rod stability
PETN , A, B. C,X*,W A,B, X w C,W
P e n t o l i t e A, B,X* A, C,Y,Z A.B, X w
50/50 pentolite w w w
C o m p o s i t i o n B . w w w c*
T N T w w w
R D X w w w
A–unencapsulated 3M taggant ..
X– unencapsulaled Westinghouse ceramic taggant
B–unencapsulaled hard cross linked 3M taggant Y–encapsulated Cune-point Iaggant
C –encapsulated 3M taggant Z–unencapsulated Curie-poml taggant
D–encapsulated higher melting point 3M Liggant ‘ –idicating Irradiated taggant
E - unencapsulaled higher melting point 3M taggant ‘ –Indicated noncompatibility
w–encapsulated Westinghouse ceramic t a g g a n t

SOURCE Off Ice of Technology Assessment

reaction of tagged booster material in the Table 27.–Summary of Compatibility Tests

Conducted With Black Powder
above tests. Goex claims that the explosion
must have been caused by the taggants. The Test type
Aerospace Corp. asserts that no tagged booster Drop BAM
material was located at the Camden factory at Type of powder weight frlctlon
this time, and that furthermore the low con- FFFg A,B, X A,B, X
Tailngs A.B, X A,B, X
centrations which Goex asserts were present
could not have initiated an explosion; the tests A - unencapsulated 3M tagant
B–unencapsulated hard cross-linked 3M taggant
to which Goex refers involved extremely high X –unencapsulated Westinghouse ceramic taggant
taggant concentrations, OTA is not familiar SOURCE Off Ice 01 Technology Assessmem

with the facts regarding the possible presence

of taggants, and is not aware as the report goes
to press of any experimental data on the possi- Smokeless Powders
ble destabilizin g effects of low concentrations
of taggants mixed with TN T/RDX mixtures. The compatibility tests conducted with
smokeless powders are summarized in table
As would be expected, the more gritty tag-
28. Only the encapsulated 3M taggant (type C)
gants clearly showed evidence of sensitizin g was tested. Tests were originally conducted by
the booster explosives. In the case of the Curie- Hercules, Olin, and Du Pent on their own
point taggant, sensitization occurred even for smokeless powders. 24 25 No evidence of sensiti-
encapsuIated taggants; these are the onIy tests
zation or change in electrostatic properties
showing sensitization with encapsuIated tag-
was observed. I n the case of the Herco@ p o w -
gants. der, however, the Taliani and German heat
tests both indicated a significant decrease in
stability due to the addition of the taggants (in
Black Powder a 50-percent concentration) to the smokeless
powder. (Although Hercules tested only
The black powder compatibility test results
H e r c o @ powder, Hercules believes that their
are summarized in table 27. Neither the black
powder nor the black powder tailings are sen-
sitive to either the friction or impact tests con-
ducted, even for the gritty taggants, However,
no stabiIity tests were conducted.
Ch. IV—Taggant Safety and Compatibility Review “ 89

Table 28.–Summary of Compatibility Tests Conducted With Smokeless Powders

— —-. — ————
Test type
Electro- Crltlcal
static Impmge- height to German Balllstlc Balllstlc
——— of powder Impact Frlctlon discharge ment explosion DSC Tallanl heat veloclty pressure
Hercules HPC c
Hercules bullseye c
Hercules m
Herco’ ’ C c c c c c c* c,
Du Pent H1-skor c c c c c c
Du Pent PB c c c c c c
Du Pent IMR 3031 c c c c c c
Du Pent IMR 4064 c c c c
O l l n 2 3 1 c
Olln 296 c
O l l n 4 5 2 c
Olln 540 c
Olln 473 c
Olin 571 c
Olln 680 c
Olln 748 c
O l l n 7 6 0 . c
Olln 785 c
O l l n W C 5 7 1 c c
— —
C–encapsulated 3M taggant
“–lndlcaled noncom pall bllrly
SOURCE Off/cc of Technology Assessment

other brands of powder designed for the re- At the present time, there appears to be an
loading market are so similar to Herco” that incompatibility between the 3M taggants and
similar test results could be expected. OTA the Herco” smokeless powder. Hercules has
believes that this is highly likely for the four indicated that it does not consider the com-
other Hercules brands that are chemically bination safe and has stopped all work on it.
identical to Herco”; it may not be the case for OTA feels that, on the basis of the tests just
the three Hercules brands with different com- described, the conclusion must be drawn that
positions.) As no changes were noted for the the 3M taggants cannot be safely added to the
Du Pent or Olin Abel tests, the Herco @ t e s t s H e r c o @ powder unless the present incompati-
were repeated at the Naval Ordnance Station, bility is resolved. Some justification exists for
Indian Head, Md. The decreased stability was questioning the validity of tests using severely
confirmed. A more carefully controlled series increased concentrations of the taggant mate-
of tests was then conducted by the Lawrence rials (5o percent in the tests v. 0.05 percent of
Livermore Laboratory (LLL) for the Aerospace encapsulated material in the proposed taggant
Corp. in an attempt to isolate the element or program), but it has not been demonstrated
elements of the taggant materials which are that there is a threshold concentration below
responsible for the incompatibility. zb Briefly, which the problem disappears, and that such a
the tests indicated that there exists an in- threshold would never be exceeded in prac-
compatibility between something in the tice.
Herco” and the melamine/alkyd which forms
Preliminary ballistic tests have been con-
the basic matrix of the 3M taggants. It may be
ducted on tagged WC 571 shotgun powder
a basic reaction with the melamine/alkyd or
manufactured by Olin. Ballistic velocity,
with the catalyst used to speed up the cure
chamber pressure, and time to initiate burning
time. There may also be reactions occuring be-
were measured. Tests were conducted at three
tween the taggant pigments and the Herco @
temperatures ( –30° C, 20° C, and 50° C) and
powder. The LLL tests are continuing in an at-
four taggant concentrations (2, 4, 10, and 20
tempt to resolve the issue.
times the recommended concentrations), both
‘“[l 5f’,]ton A I’,]vne Iettt’r, OIJ c It with the taggants mixed in the powder and

61-401 9 - 80 - 7
90 ● Taggants in Explosives

with the taggants separated and placed direct- concentration would vary in actual use by
ly over the primer flash hole. handloaders, which can and should be estab-
lished by careful testing and statistical analy-
The Olin rationale for such extreme tests
sis. At the low-temperature condition two
condition (up to 20 times the nominal concen-
anomalous test results occurred. Evidence of
trations, 100-percent segregation) was an at-
improper ignition occurred in 1 of the 20 fir-
tempt to evaluate the worst-worst case condi-
ings at the 20 times normal concentration, 100-
tions that might appear due to segregation of
percent segregation condition. Improper igni-
the taggants from the powder during manufac-
tion would constitute a safety hazard as the
ture, transportation, and storage.
round might not clear the barrel, Significantly
No deviation from acceptable ballistic per- reduced bal I istic performance occurred on 1
formance was noted for the ambient- and high- of the 20 tests at 4 times nominal taggant con-
temperature tests. A steady decrease in veloci- cent rat ion, with the taggants and powder
ty and pressure was noted with increasin g tag- mixed. No other performance degradation was
gant concentration. The practical signifance of noted, even under conditions of higher taggant
this depends on the extent to which taggant concentrate ion.


Several hundred tests have been conducted and experience of the organization conducting
to investigate the compatibility of explosive the tests. If the candidate explosive product
materials with identification taggants. Most of fails to pass any of the critical tests in the
the tests have been conducted with the base- series, it is judged to have failed the qual ifica-
line 3M taggants and variations of these tag- tion test program. If a flaw can be corrected,
gants; a large number of tests, however, have then the tests can continue, but the material
also been conducted with several other candi- must pass al I of the critical tests, not just a ma-
date taggant materials. Compatibility tests jority or a certain fraction.
have included those designed to indicate in-
creased sensitivity, decreased stab i I it y, There is no indication that the 3M taggants
changed electrical properties, and changed are incompatible with dynamites, gels and SIur-
performance, Explosive materials have in- ries, or black powder.
cluded dynamites, gels, emulsions and slurries,
Composition B booster material and
cast boosters, bldck powder, and smokeless
Herco” smokeless powder do show significant-
powders. A full set of qualification tests has
ly reduced stability in the presence of the 3M
not been completed on any single explosive
identification taggants. Furthermore, careful
product and only a small fraction of the hun-
testing appears to indicate that the incompati-
dreds of products has had any testing. Given
bility is with the basic melamine/alkyd materi-
these limitations, it is still possible to draw
al of the taggants, rather than with a particular
some tentative conclusions on the compat ibi I i-
pigment or the polyethylene encapsulate.
ty of taggants with explosive materials (which
Tests, similar to those conducted with Herco”,
may change as more data becomes available)
were conducted with other smokeless pow-
and to discuss the implications of these results
ders; no loss in stability was noted for other
for the taggant program,
Hercules powders, or for the Olin or Du Pent
First, it is important to realize the purpose of smokeless powders. The reaction, therefore,
~ cornpatibiIity qua I if i cat ion testing program. probably is between the melamine/alkyd and
In brief, a set of tests is established on the one of the sensitizers or stabilizers of the
basis of analysis, the projected manufacturing, H e r c o @ . As the formulations of both Herco @
storage, transportation, and end-use process- and the 3M identification taggants currently
ing of the material, and the normal procedures stand, the two are not compatible. Further in-
Ch. IV— Taggant Safety and Compatibility Review ● 91

vestigation may isolate the element of incom- mix taggants with smokeless powder in such a
patibility, and it may be possible to replace way as to avoid extreme variations in taggant
elements in either the Herco @ or the taggants concentration from one round to the next.
to remove the incompatibility. It is not yet Testing is required to establish how great a
possible to tell whether the booster material variation in concentration could be expected
incompatibility is with the basic melamine/al- using reasonable manufacturing methods, and
kyd or with one of the components of the tag- normal transportation, storage, and loading
gants. procedures. The Olin tests did show one case
of poor performance (at four times the sug-
Both the smokeless powder and booster ma-
gested taggant concentration), but perform-
terial tests took place at high temperatures,
ance anomalies sometimes occur without tag-
and, in most of the tests, at high-taggant con-
gants, and a single anomaly is not enough to
centrations. The temperature used for the
justify a prediction as to whether taggants
smokeless powder test was higher than would
would increase the frequency of such occur-
be expected in actual manufacture, storage, or
rences. The segregation tests were conducted
use; the temperature used for the cast booster
with 100-percent segregation, which appears
is sometimes reached in manufacturing proc-
quite unreal istic. Testing is needed to establish
esses. In each test, a taggant concentration of
the extent of segregation which might occur
so percent was used rather than the 0.05-per-
before a realistic worst case can be defined.
cent tagging concentration suggested for rou-
Unlike the Herco @ and Composition B cases,
tine use. The tests, nonetheless, indicate that
the Olin ballistic property tests do not appear
the stability of the materials has decreased,
to OTA to constitute sufficient evidence to re-
due to the addition of taggants, and that a re-
quire presumption of an incompatibility. It re-
action is taking place between elements of the
mains true, however, that no presumption of
taggants and elements of the explosive mate-
compatibility can be made until adequate bal-
rial. Standard qualification test procedure re-
listics tests have been conducted.
quires that such evidence be considered a sign
of an existing incompatibility between the
This raises the question of the value of a tag-
materials. Careful Iy controlled testing, and ex-
gant program from which smokeless powders
tensive analysis must be completed before it
and cast boosters were excluded. As noted in
can be determined if the observed evidence of
chapter Vl, smokeless powders are used in a
incompatibiIity does, in fact, indicate a poten-
significant percentage of criminal bombings
tial safety problem during the manufacture,
(approximately 20 percent) and cause 10 to 20
storage, transportation, and use of the tested
percent of deaths and injuries. As also noted in
materials. Unless demonstrated otherwise, it
chapter Vl, criminal bombers are Iikely to re-
must be assumed that it is unsafe to add the
act to a taggant program, If smokeless pow-
taggants to that smokeless powder or the
ders are not tagged, then a logical reaction
booster material. Until the elements of the in-
would be for a large number of bombers to
compatible Iity have been identified, a question
switch to the use of smokeless powders. Al-
remains as to the safety of adding the taggants
though bombs using smokeless powder are
to similar smokeless powders and booster ma-
considerably less efficient (lower specific
terials, although tests with other smokeless
energy) than those using cap-sensitive high ex-
powders and boosters have shown no evidence
plosives, smokeless powder bombs are respon-
of incompatibility.
sible for a considerable number of injuries and
The significance of the 01 in ballistic proper- deaths. Effective controls over smokeless pow-
ty tests cannot be fully assessed at this time. der by means other than taggants may be pos-
The Olin tests indicated that increasing tag- sible but appear unlikely. Booster material is
gant concentrations lead to a reduction in ve- rarely used as a bomb filIer. It is used, how-
locity and pressure, and this could create a ever, to initiate blasting agents. The current
problem if and only if it proves impossible to BATF plan would be to not directly tag blast-
92 ● Taggants in Explosives

ing agents, but to tag the booster and detona- As noted above, much compatibility testing
tors used to initiate the blasting agent. Exclu- remains to be accomplished. Identification
sion of boosters from the taggant program may taggants have undergone comprehensive test-
well require an alternate control mechanism ing with a representative sample of dynamites,
for blasting agents. Given the extremely large gels, slurries, cast booster materials/smokeless
quantity of blasting agent produced (3.4 billion powders, and black powder; even after the res-
lb annually), any other control mechanism may olution of the compatibility questions which
have serious cost consequences. testing so far has revealed, it would eventually
be necessary to test taggants with all such ma-
The above discussion concerned the results
terials before instituting a comprehensive tag-
of the tests to investigate the compatibility of
g i n g p r o g r a m . In the case of detonators and
the baseline 3M taggants with explosive mate-
detonating cord, compatibility testing has not
rials. Tests were also conducted using hard or
been completed even with a representative
gritty taggants, In all cases, the unencapsu-
sample. Compatibility testing of detection tag-
Iated hard taggants caused increased sensitivi-
gants started only recently, and with the excep-
ty to the drop weights, and, in most cases, to
tion of testing with detonators it is less far ad-
the sliding rod tests. The ceramic Curie-point
vanced than compatibility testing of identifi-
taggants caused increased sensitivity in some
cation taggants.
cases even when encapsulated, although no in-
compatibility was noted for the Westinghouse It is necessary to resolve the incompatibility
or hard-core 3M taggants when encapsulated observed between the 3M identification tag-
with polyethylene. When a hard resin was used gants and the Composition B booster material
as an encapsulant, the 3M taggants showed a as well as the Herco@ powder however, before
clear sensitization of PETN. The implications it makes any sense to finish the rest of the tests
of these tests are obvious. l-lard or gritty tag- with other materials. The resolution of the
gants must be encapsulated. The encapsulated smokeless powder incompatibility could take
material should not only be soft but it should any of several forms, including:
also be a heat sink. The use of a soft additive is ● Reformulation of the 3M taggant– this
a common desensitizer in military explosives.
could require starting essentially from
Composition B and other RDX-based explo-
scratch in the taggant-testing program, as
sives include approximately 1 percent wax
the reformulated taggant would un-
with a softening point in the 800 F range.
doubtedly exhibit different compatibility,
The tests show that encapsulated gritty tag- as well as survivability properties.
gants, such as the Westinghouse ceramic tag- ● It might be possible to develop a different
gant, may be alternatives to the baseline 3M taggant that proved compatible with
taggant. As even a small amount of the unen- smokeless powders, and to use the exist-
capsulated material (0.01 percent) causes in- ing 3M taggant for explosive materials
creased sensitivity, however, great care must with which it is compatible.
be exercised to ensure essentially 100-percent” ● Reformulation of the Herco” powder—
encapsulation; this may seem to create an im- this may or may not be easily accom-
possible quality control problem. However, plished, once the element or elements
the problem may not be as difficult as it first that react with the taggant are isolated.
appears. If 99 percent of the taggants are en- This option would only be viable if no
capsulated, then unencapsulated taggants other smokeless powder were found to be
would constitute only .00025 percent by incompatible.
weight of the explosive, almost two orders of ● Exclusion of Herco” from the taggant
magnitude less than the amount demonstrated program —the economic effects on com-
to cause increased sensitivity. Tests of those petition would need to be carefully con-
extremely low levels might welI show no in- sidered, as would alternate control mech-
creased sensitivity. an isms.
Ch. IV—Taggant Safety and Compatibility Review . 93

● Exclusion of smokeless powders from the rate is both temperatu r-e and concentra-
identification taggant program — such an tion sensitive, it may be that no s a f e t y
exclusion wouId rely on the fact that hazard exists under realistic conditions If
smokeless powders would be less effec- it could be positively demonstrated that
tive than cap-sensitive high explosives and the decomposition rate was within the
that the detonators would be tagged. OTA normal I I y accepted range for temperature
believes that this last approach may not regimes and concentrations which reflect
be viable– too many people are currently worst case actuaI use conditions, then it
killed or injured using smokeless powders may be possible to add taggants to the
and the numbers wouId almost certainly smokeless powder, particuIarly if no fur-
increase if that approach were adopted. ther incompatibiIities surface. However,
Alternate control mechanisms for smoke- demonstrate ion of safety wouId have to be
less powders would be required, quite convincing to overcome the current-
● Demonstrate ion that the observed stabiIity ly perceived incompatibility.
problem does not constitute a safety haz-
ard. The observed decreased stabiIity oc- A resolution of the booster incom pat bility
curs at elevated tern peratures and at more problem could be accomplished by a s milar
than two orders of magnitude higher tag- set of methods, once the elements of t l e i n -
gant concentration, As the decomposition compatibiIity have been identitied.
Chapter V

Page Page

Overview ● * . * . * . . , . * , . * . * * * * . . * , . . . , . * * , 97 Recordkeeping Costs . ..................111

T-t Material Costs........., . . . . . . . . . . . 101 Markup. . . . . . . . . . . . . . . . . . . . . . ........112
Identification Taggants, . ................101 Summary of Manufacturing Costs Added .. .112
Detection Taggant Materials Costs . .......102 Distributor Costs . . . . . . . . . ................113
Cost and Supply Guarantees , . ...........103 Recordkeeping at 13istribution Levels .. ....113
Sensor-Related Costs. . . . . . . . . . . . . . . . . . . . . .104 Storage . . . . . . . . ......................115
Numbers of Sensors Needed . . . . . . . .. ....105 Summary Cost IncludingMarkup. . ........116
Sensor System Related Costs . ............105 UserCostImpacts. . ......................117
Mix of Sensors . . . . . . . . . . . . . . . . . . . . .. ..106 Increased Material Costs . ...............117
False Alarm Costs. . . . . . . . . . . . . . . . . . . . . .107 Commercial IJsesof Explosives and
ExpMvesandGunpowder Manufacturing Costs . 1(I7 Cunpowcfers-Genera\. . ............118
Revised Processes, Tooling, and Facility Underground Mines. . ................118
costs. . . .....’... . ................108 Quarries . . . . . . . . . ..................118
Cap-Sensitive PackagedExpkMVes .. ....108 Open PitMines. . ....................119
Cast Boosters, Smokeless Powcfersg and Construction. . . . . . . . . . . . . . . . . . . . . . . .119
Black Powder., . . . . . . . . . . . . . . . ....108 Handloading. . . . . ...................120
DetonatingCord. . ...................109 o t h e r c o s t l m p a c t s . . . . .o . . . 0 . . . . . . . . . . . . . 1 2 1
BlastingCaps ..+. . . . . . . . . . . . . . . . . .:.109 Government investigation Costsand Program
Labor . . . . . . . . . . . . . . . . . . . . . . . ... ...’. .109 Administration. . ..................121
Cap-Sensitive Packaged Explosives .. ....109 InvestigativeCosts. . ...................121
Cast Boosters.. . . . . . . . . . . . . . . . ... ,..?09 EffectsofCompetition-Substitution . . . . ...I22
Black Powder... . ..................,109 Effects on Fixe&PriceCommodities. . ......122
Smokeless Powder . ...............,..110 Possibie RemovaiofSome Gunpowders From
DetonatingCord and Blasting .. 110 the Market . ......................123
Productivity . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 0 - b w * $ - . . . . . . . ● . . . . . . 123
Cap-SensitivePackagedExplosiv~ .. ....110 “ Identification Taggant Material
O t h e r E x p l o s i v e C a t e g o r i e s . ...........111 Costs. . . . . . . . . . . . . . . . . . ..........123
InventoryCosts . ......................111 Detection TaggantProgram Material Costs. .123
QualityControl . . . . ...................111 Manufacturing Level Program Costs . . . . . . 124
Safety . . . . . . . . . . . ....................111 Distribution Network Program Costs. ., ....124
Public Overhead Program Cost . ..........124 47. Identification Taggant and Detection
Taggant Program Baseline Cost Estimate. .. .125 Taggant Program Cost Comparisons -
Program Cost Versus Implementation Level .125 Basline Case. . . . . . . . . . . . . . . . . . . . . .. .126
Program Cost of Separate Identification and 48. Summary Program Costs Versus Level of
Detection Taggant Programs . ........126 Implementation ., . . . . . . . . . . . . . . . . . . . . 127
Comparison of OTA Cost Estimates With I ME 49. Comparison of the Estimates for ID
and Aerospace Corp. Estimates . . . . . .. 127 Tags . . . . . . . . . . . . . . . . . . . . . ..........127
Who Bears the Cost of a Taggant Program? .. 128 50. Comparison of OTA and Aerospace Program
Cost Analysis Precision. . . . . . . . . . . . . . . . . . . . . 129 Estimates . . . . . . . . . . . . . . . . . . . . . ......128
Cost Sensitivity Analysis . ...............129 51. Taggant Program Cost Impact by Who
Taggant Materials . ..................129 Will Bear the Cost. . . . . . . . . . . . . . . . . . . . . 128
Adequacy of Current Data. . ................133 52. Elements of Cost Uncertainty. . ..........130
Suggested Further Research . . . . . . . . . . . . . . . . 134
. 53. Annual Cost per Sensor for Various
Mixes ., . . . . . . . . . . . . . ...............132



29 Qualification of the Estimating Basis for

Taggants. . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 FIGURES
30 List of Taggant Program Cost Elements .. ..100
31 Baseline Taggant Program Configuration. .. 100 Page
32 Annual Taggant Requirements. . .........101 10. Schematic Illustration of General Cost
33 Qualification of Estimating Basis for Element Sources . . . . . . . . . . . . . . . . . . . . . . 99
Sensors. . . . . . . . . . . . . . . . . . . . . . .......105 11. BM Identification Taggant Cost
34 Vapor Taggant Detector System Cost .. ...106 Estimates: 5Ib Tag Lots, Unencapsulated,
35 Current Manufacturing Cost/Price 2-Year Minimum Required . . . . . . . . .. ....102
Data . . . . . . . . . . . . . . ...............108 12. General Functional Network for Vapor
36. Summary of Explosives and Gunpowder Taggant Detector, . . . . . . . . . . . . . . . . .. ..106
Manufacturing Costs Included. . .........112 13. Estimated Annual Vapor Taggant
37. Cost Summary of Cap-Sensitive Packaged Detector Cost v. Quantity Deployed .. ...,106
High Explosives Manufacturing Cost 14, Estimated Cost of False Alarms v. False-
Variations With Assumptions. . ..........113 Alarm Rates . . . . . . . . . . . . . . . . . . . . ... , .107
38. Estimated Cost for Powders at 15. Schematic of Cap-Sensitive High Explosive
Distribution Network . .................116 Distribution Network . .................114
39. Distribution System –Summary of Cost 16. Schematic Distribution Network of
Added and Markup. . ..................116 Gunpowder. . . . . . . . . . . . . . . . . . . . . ... , 114
40. Current High Explosives Cost Impact for 17. Recordkeeping Activity v. Tagging
Various User Classes . .................121 Level. . . . . . . . . . . . . . . . . . . . . . .........115
41. Identification Taggant Material Annual 16. Summary of Added Costs to Explosive Users
Costs, Baseline Program. . ..............124 Cost Per Unit of Explosives in Dollars .. ...117
42. Detection Taggant Material Annual 19. Bingham Canyon open Pit Copper
Costs. . . . . . . . . . . . . . . . . . . . . . .........124 Mine. . . . . . ., ., . . . . . . . . .............120
43. Manufacturing Cost Added . ............125 20$ Baseline Program Cost Sensitivity Impact
44. Distribution System Cost Added . ........125 With Changes in Identification Taggant,
45. Taggant Program Summary Annual Cost- Material Cost, and Concentration Level, .. .131
Baseline Program . . . . . . . . . . . . . . . . . . . . .125 21. Baseline Program Cost Sensitivity Impact
46. Taggant Program Summary Annual Cost With Changes in Detection Taggant,
Versus Implementation Level. . ..........125 Material Cost, and Concentration Level. .. .131
Chapter IV

A detailed review of the potential cost and economic impacts of the proposed
taggant program was conducted in parallel with the safety and utility segments of
the study. In this analysis, the assumption was made that the taggants work and are
safe to put in explosive materials. it was furthermore assumed that the current in-
compatibilities observed between the 3M identification taggant and one type of
smokeless powder, as well as one type of cast booster material, would be resolved in
a way which has no additional cost impact. The various cost elements were esti-
mated by: ‘
● drawing on existing studies and testimony; and
● interviewing the identification taggant manufacturer, explosive and gunpow-
der manufacturers and distributors, users of explosive materials, law enforce
ment personnel, and sensor instrumentation engineers.

Other important economic issues were ad- quent followup process. These costs must,
dressed in parallel with the development of the however, be compared with the cost of current
program cost. The addition of taggants to ex- law enforcement practices.
plosives has a potential cost impact to an in-
Detection taggants require a sensor and a
dustry in which explosive-type decisions are
system to sample and convey the air from the
frequently made on an economic, rather than
sample item to the sensor. The sensor and sam-
performance or brand loyalty, basis. An addi-
pling system requires operation and mainte-
tional taggant material cost issue is that raised
nance, although it is possible that current
by the probable monopoly of supply by one
security personnel could operate the addition-
company, particularly by 3M for the identifica-
al equipment at an airport, for instance. There
tion taggants. The question of assuring price
is an additional potential cost associated with
and taggant availability also required atten-
possible delays raised by false alarms in the de-
tion. Introduction of taggants into the explo-
tection system. Significant false alarms could
sive fabrication process will cause changes in
cause enough ill-will (in addition to high costs)
the manufacturing process, due both to possi-
to lead to the abandonment or curtailed usage
ble tooling costs and to the labor costs associ-
of detectors in situations such as airports.
ated with purchasing, controlling, and using
the taggants. Other, one-time costs are associ- A final cost aspect which must be consid-
ated with product requalification tests for ered is the economic effect of a taggant pro-
safety, potential costs for waste disposal gram in which only selected explosives are re-
equipment, and added plant capacity to make quired to be tagged. In the cost-conscious
up for lost productivity. commercial explosive industry, that could
eliminate certain products or companies from
Identification taggants require additional
the marketplace, perhaps resulting in signifi-
recordkeeping by the manufacturer, by whole-
cant local unemployment.
salers and distributors, and by the retail sellers.
There are law enforcement costs associated Due to the fact that the identification tag-
with the recovery and tracin g of identification gants have progressed further down the devel-
taggants from explosions and with the subse- opment path, the relative precision of the cost

98 ● Taggants in Explosives

estimates associated with their introduction agencies with whom detailed discussions were
into explosives is expected to be greater than held include the Bureau of Alcohol, Tobacco,
the estimates of detection taggant and related and Firearms (BAT F), the Federal Aviation Ad-
sensor costs. The precision of each estimate is ministration (FAA), the Department of Com-
indicated during the course of the cost analysis merce (DOC), the Bureau of Mines (BOM), and
discussion. various Department of Defense agencies.
This cost analysis by OTA has been an inten- Various degrees of uncertainty exist in cost-
sive, short-duration study. Of necessity, the ing out the taggant program, as little test data
study was accomplished by drawing on exist- exists and some potential manufacturing proc-
ing studies from a wide variety of sources and ess applications are undefined. Table 29 illus-
by a limited number of onsite interviews with trates the qualifications of the estimating basis
industry and Government. Discussion with in- for the taggant program, indicating the status
dustry included various explosives manufac- of pilot testing and the OTA understanding of
turers and BM, the taggant manufacturer. Vari- the manufacturing processes required to im-
ous user types such as mining companies (un- plement taggants. On the right side of table 29
derground and surface), construction firms, is set forth, in general terms, the method for
and quarry operators were also visited. Exten- estimating utilized, such as direct estimating,
sive discussions were also held with the Aero- Aerospace Corp. analysis and assumptions, the
space Corp. (the taggant program development Institute of Makers of Explosives (IME) mem-
contractor), with the Institute for Defense ber estimated inputs, Sporting Arms and Am-
Analysis, with Management Science Associ- munition Manufacturers’ Institute ( S A A M I )
ates, and with consumer groups such as the estimated inputs, etc. The particular methods
National Rifle Association and the National and data sources utilized are documented
Muzzle Loaders Association. Government throughout this study where appropriate.

Table 29.–(lualification of the Estimating Basis for Taggants

Etimating, basis
Taggant mfgr
Type explosive Pilot tested process Process labor Process tooling Other capital expenses
Cap-sensitive Yes Yes Direct/ estimate Direct estimate Direct estimate
packaged explosives Proprietary detail estimate Nonrecurring
available. Requalification of products.
I ME member inputs.
Cast boosters Yes Yes Aerospace analysis/ Waste disposal if
assumptions additional waste due to
‘‘unacceptable contaminated
tag batches
Smokeless powder Underway Yes Aerospace analysis Equipment required:
SAAMI estimate. storage bins, hoppers,
equipment for weighing,
packaging, transferring
tag samples,
Black powder Yes Yes Goex Study Investment offset losses
storing in productivity.
● security

● administrative & records

● mfgr, process cleanup

Detonating cord Planned No Aerospace assumptions, Tooling. ● Cost of taggant Inventory

Design required (no effective Including the cost of money
equipment currently available
Blasting caps Planned No Aerospace assumptions Significant cost ● expected–
new machine must be designed.

‘Aerospace estimates utilized and OTA survey inputs
SOURCE Off Ice of Technology Assessment
Ch.--Yaggant Cost Review . 99

The primary methodology utilized in this Cost estimates were also generated for the im-
cost analysis was to translate all program plementation program proposed by BATF.
costs, both nonrecurring one-time costs and re-
curring costs, to annualized values. Capital in- All cost data and program estimates in this
vestment costs were annualized over a 10-year report are stated in fiscal year 1979 dollars to
period at an interest rate of 10 percent. This assure consistent treatment. A list of taggant
method was utilized for all initial expenditures program cost elements was developed to per-
(requalification, waste facilities, etc.) with the mit a comprehensive framework for treating
exception of tooling costs estimated for deto- all potential costs and resources impacted by
nators and blasting caps, which were written the taggants program. Figure 10 illustrates the
off in a 5-year period at 10-percent interest. general sources of costs potentialIy involved in
the program, while a detailed list of potential
The taggant program costs vary substantial- cost elements is shown in table 30.
ly as a function of the level of implementation
of the program. In this study, an O T A i d e n t i - For purposes of exposition throughout this
fied baseline program was assumed for base- cost impact assessment, a baseline set of con-
line cost estimates, and the parametric varia- ditions or assumptions is utilized in the deter-
tion of the costs examined as a function of mination of a total program estimate. These
higher and lower level implementation plans. are shown in table 31. This baseline program

Figure IO.— Schematic Illustration of General Cost

Element Sources

Direct Costs Direct costs

added by added by
mabyfactyrer distributor

Taggant Direct costs

materials Incurred by
Program Cost

SOURCE Office of Technology Assessment.

100 . Taggants in Explosives

Table 30.–List of Taggant Program Cost Elements includes several provisions which, OTA be-
lieves, would do much to hold down costs
Taggant materials
Idenhflcahon taggants without a significant reduction in the utility of
Detection taggants the program: blasting agents are not tagged;
Detection sensor-related costs the identification taggant code is changed
Sensors only when the date, shift, or product changes
Sensor sampling and transport instrumentation
Operations and maintenance (resulting in some code numbers correspond-
Cost of false alarms ing to a large batch size and others to a small
Explosive and gunpowder manufacturing costs batch size); and a special “composite code” is
Nonrecurring cost used for taggants added to already tagged ma-
● Tooling

● Storage
terial (permitting rework without removal of
● Product requalification- safety testing previous tags). The special composite code
● Waste disposal facilities
taggant would be added to material with more
● New Investment to offset production losses

Recurring costs than 10-percent cross-contain ination; such a

● Manufacturing process labor taggant would indicate that the material used
● Record keeping was a composite and that taggant codes other
● Quality control
● Production losses than the specific composite code should be ig-
● Waste product line nored.
● Inventory costs
● Administration expense Although confidence levels are relatively
Markup high for certain elements of costs, particularly
Distributor costs for the identification taggant program, other
Record keeping program elements are subject to considerable
Markup uncertainty (particularly the number and types
User costs of sensors to be employed in the detection tag-
Other costs
gant program). Attention is called to the base-
Government administration line assumptions associated with each cost ele-
Taggant program development ment throughout the discussion of cost.
Investigative costs
In the following section the costs for the tag-
SOURCE Off Ice of Technology Assessment
gant materials are developed. This is followed
by detection taggant sensor-related program
Table 31 .–Baseline Taggant Program Configuration cost estimates. The potential cost increases oc-
curring during the explosive manufacturing
. Encapsulated identification taggants process and at the distribution level are then
● Explosive weight or units to be tagged and tagging concentration
addressed. The potential cost impact(s) to the
Category Units/yr Concentration
users of explosives are subsequently discussed.
Cap-sensitive packaged
explosives 325,000,000 lb .05% Other cost impacts, including the cost contri-
Boosters 6,000,000 lb .1 % bution by Government for administration, in-
B l a c k p o w d e r 400,000 lb .05% vestigation, and taggant program develop-
Smokeless powder 5,000,000 lb .05%
Detonating cord 500,000,000 ft 5 tags/in. ment, are set forth in the next section. A gener-
Blasting cap 84,000,000 units 50 mg 1 synthesis and summary of the taggant pro-
. Identification and detection taggants gram cost estimates follows, with the relative
. 1,500 sensors to be deployed precision or accuracy of the estimates dis-
● Sensor mix. M S 10°/0, I MS 90°/0

. 10% taggant contamination permitted

cussed after that, including aspects of cost un-
● “Composite tag” permits rework of previously tagged material certainty and program cost sensitivity. The
● Days production of each type/size explosive (date-shift basis) adequacy of the current cost data and sug-
● New taggant code for each
gested further research are briefly discussed in
SOURCE Office of Technology Assessment the last two sections, respectively,
Ch. V—Taggant Cost Review ● 101


The cost of both the identification and de- Table 32.–Annual Taggant Requirements
tection taggant material is heavily influenced — .
Annual taggant
by the amount of explosive material to be Concentration requirement
Quantity to be level pounds
tagged, the form of the tagging material, and (unencapsulated) (unencapsulated)
Explosive category lagged
the concentration levels. Material cost esti-
Cap-sensitive pack-
mates are developed for the baseline program aged high explosives 325.000.000 lb 0.025% 81,250
described above. Cast boosters 6.000,000 lb 005 % 3,000
Smokeless powder 5,000.000 lb O 025% 1,250
Black powder 400,000 lb 0.025% 100
Detonating cord 500,000,000 ft 5 tags/in, 160
Identification Taggants Blasting caps 84,000.000 caps 50 mg each 4,620
The annual quantity of explosives produced
in the United States, shown in table 32, was SOURCE Of ficeof Technology Assessment
estimated based on data obtained from IME,
BATF, Aerospace Corp., BOM, and DOC. An encapsulated taggants produced in 5-lb lots
unresolved problem exists with respect to the and assume a firm order for a minimum of 2
production of cap-sensitive packaged high ex- years. The 150,000-lb level is a result of a de-
plosives. The basic difficulty stems from the tailed Ieadtime study conducted under con-
method of reporting data in the surveys col- tract to the Aerospace Corp. The target price
lected by both BOM and DOC, Some “un- and worst case estimates for the 75,000- and
known” quantity (both permissible and other 100,000-lb levels were provided by 3M in re-
high explosives) of cap-sensitive explosives is sponse to an OTA request. The range of prices
reported as included in unprocessed ammonia reflects the fact that less time was available
nitrate and “al I other purpose” categories in for the 3M estimates than the original 150,000-
order to avoid disclosing individual company Ib level, resulting in some uncertainties. These
data. Since the data are masked to protect the target prices have all been through a rigorous
marketing positions of explosive manufactur- price review within the 3M corporate structure
ers, the uncertainty in annual quantity will per- and represent the firmest commitment possi-
sist, For purposes of this study, the quantity of ble short of a production contract.
325 million lb/year (as adopted by Aerospace)
Assuming linear extrapolation between the
wil I be used as the baseline condition.
data points, the price for unencapsulated iden-
A second variation concerns the level of tification taggant material was estimated by
black powder produced. Approximately 2.5 OTA (from figure 11) to be approximately $93/
m i I I ion lb of black powder are produced per lb for the estimated 90,000 lb of taggants to be
year in the United States, but the majority is required annualIy. This cost figure assumes
used as a raw material in other fabrication p r o d u c t i o n in 10,000-lb lots. In cases where
processes, s u c h a s f u z e s . Approximate y most lots are substantialIy smaller, taggant
400,000 lb are sold directly to the consumer; costs per pound of explosives might rise.
this amount is included in the explosive materi-
This figure is for unencapsuiated taggants,
als to be tagged. Table 32 shows the produc-
while the baseline OTA program assumes the
tion quantity, the concentration of unencapsu-
taggants are encapsulate? in an opaque poly-
Iated taggant material suggested by the BATF/
ethylene wax. The 3M technical people fur-
Aerospace team, and the resultant quantity of
n i shed an estimate of the cost of encapsuIating
unencapsulated taggants required annualIy,
the taggants in polyethylene wax, but were un-
Price estimates, obtained from 3M as a func- able to estimate the cost impact of using an
tion of annual taggant production, are shown opaque polyethylene wax. Based on the above
in figure 11 The estimates quoted are for un- data, OTA estimated that it would cost $55/lb
102~Tggants in Explosives

Figure 11 .—3M Identification Taggant and an additional library maintenance fee of

Cost Estimates $100/year per unique taggant species. This
identification taggant cost has been clearly
● 5-lb tag lots Unencapsulated

●2-year minimum required identified by 3M as the cost of taggants pro-

duced in their current pilot plant, which is
Cost per pound labor intensive, if there is no program legis-
of taggant lated to tag commercial explosives. It does not
in 1979 dollars
represent a potential cost figure if a taggant
program is legislated, Details of the cost of
taggants, a s a f u n c t i o n o f t o t a l q u a n t i t y
needed, were given above. No additional fee
would be required for Iibrary maintenance.

120 Detection Taggants Materials Costs

The Aerospace Corp., as part of its taggant
110 contract effort for BATF, has put considerable
effort into the development of molecules for
detection taggant purposes. As a result of in-
100 vestigation of the properties of several hun-
dred potential molecules, five chemicals are
currently considered excellent candidates for
90 the program. These perfluorinated cycloal-
phones are:

● PDCB — perfluorodimethyl cyclobutane,

● PMCH — perfluoromethyl cyclohexane,
● PDCH — perfluorodimethyl cyclohexane,

● PFD — perfluorodecal in, and

● PS P — perfIuorohexyl-suIfur-pentafIuo-


100,000 150,000
The final selection of a particular detector
taggant will depend on the results of compati-
Annual identification taggant production in pounds
bility testing, efficacy in conjunction with the
SOURCE: Office of Technology Assessment detection taggant sensor, price, and availabil-
for opaque encapsulated taggants, as the base- The microencapsulated detection taggant
line tagging level is 0.05 percent by weight of
would be directly incorporated as a free-flow-
encapsulated taggants, and the encapsulating ing powder in commercial explosives and gun-
material weighs the same as the unencapsu- powder. Since part of the chemical selection
Iated taggants. ($93 for 1 lb of encapsulated
criteria includes a low or negligible utilization
taggants, plus $17 for 1 lb of encapsulating ma- of these materials in standard manufacturing
terial, plus the process, equal $110 for 2 lb of (to minimize false alarms due to ambient air
encapsulated taggants, or $55/l b.) This corre-
background), standard cost/price data current-
sponds to 2.75 cents/lb of cap-sensitive explo- ly available was supplemented by requests by
sives for the identification tagging material.
the Aerospace Corp. to a number of companies
IME and a number of other individuals and for budgetary pricing-type estimates at quanti-
organizations have based their cost estimates ty levels of 200,000 lb/year. A range of esti-
on a price of $200/lb of encapsulated taggants mates was received for both the cost of the de-
Ch. V—Taggant Cost Review ● 103

tection taggants and for the encapsulation regulate the virtual monopoly of the identifi-
process. Taking these values into account, as cat ion taggant market that 3M would enjoy is
well as adjustments for process yield, the fol- highly desirable. While several suppliers a r e
lowing range of estimates was made by OTA. capable of supplying the vapor detection tag-
gant, production in the necessary quantity wilI
Lower end of range $22 20/lb
Medium 40.00\Ib
probably require significant capital invest-
Higher end of range 58.15/lb ment, much of which wouId be amortized by
the taggant program. It is therefore desirable
For purposes of the baseline study OTA has
to have a mechanism that will ensure the price
utilized the medium cost of $40/lb of encapsu-
of the vapor taggant material as welI
lated detection taggant. The Aerospace Corp.,
in their inflationary impact study, estimated A number of mechanisms are available to
conservatively a value of $65/lb, based on reguIate the price of taggants, incIuding:
early data. With the more recent quotes it is
. a price level set by Congress in the ena-
reasonable to estimate a lower value for detec-
bling legislation,
tion taggant material. Uncertainty as to the ● regulation as a public utility,
value chosen remains due to the following fac-
● licensing by 3M of competitors,
● a multi year, fixed-price contract, and
final taggant selection, ● a free-market price, regulated only by the

● final contract price, possibility of competition or sanctions if

● cost of encapsuIation, prices get too high.
● the weight effect of the encapsu I at ion
The free-market mechanism is probably un-
process, and
acceptable, given the long time needed to
the final yield ratio of the encapsulant ion
either develop and qualify an alternative tag-
gant or enact sanction legislation, Legislation
Since the detection taggant program re- of a price or use of a regulation mechanism
mains in the early stages of development, un- similar to that used for public utilities would
certainty will persist in this value. Variations be an awkward, time-consuming process for a
from this value will be examined in the cost product whose total annual value would be on
sensitivity analysis. The relative significance of the order of $11 million.
the variations of the detection taggants cost is
not expected to greatly perturb the overalI tag- Licensing is not only disagreeable to 3M, but
gant program cost estimates. it is probably not cost-effective. The cost of
the taggant material includes a component for
amortization of the taggant production facili-
Cost and Supply Guarantees ty, as a new facility must be built and the
primary market for identification taggants
The identification taggants currently pro- would Iikely be the mandated explosives mar-
posed to be used are manufactured only by 3M ket, The process that 3M plans to implement is
and are a proprietary product manufactured capital-intensive. Licensing of other manufac-
by a proprietary process. In addition, a signifi- turers would therefore require the construc-
cant public overhead cost would have been in- tion of facilities for the licensee, in addition to
curred before the compatibility of explosive a new 3M facility, resulting in a substantially
materials with the taggants could have been higher total cost.
demonstrated, Mandating the addition of iden-
tification taggants to explosive materials A long-term contract may be the most effec-
would, therefore, ensure a monopoly of the tive mechanism. I n tact, the 3M cost estimates
Government-mandated market for 3M, at least are conditional on firm orders for a 2-year peri-
for a period of several years. Under such cir- od, although 3M is willing to consider contract-
cumstances, development of a mechanism to ing periods of up to 5 years, The details of the
104 ● Taggants in Explosives

contracting mechanism have not been ad- 3. utilize the discretionary power of BATF to
dressed by this study, although there may be provide relief from the legislation in cases
some advantage to a single contracting agency of emergency induced interruption of sup-
(presumably within the Government), rather ply.
than separate contracts with each manufac-
A detailed tradeoff would be necessary to
turer of explosives and gunpowder. In addi-
decide the relative merits of options 1 and 2.
tion to saving the cost of multiple contracting,
Option 2 shares the cost impact of additional
the single-contract concept would limit the
capital-intensive construction identified for
amount of information available to 3M on
the licensing option considered above. The ac-
numbers of product lines and production
ceptability of option 1 to the explosives and
quantities of explosives, a matter of some sen-
gunpowder manufacturers may be heavily
sitivity to the explosives manufacturers.
weighted by who bears the cost burden of
Assurance of availability of a taggant supply maintaining the 6-month inventory. Option 3
is a related issue. A number of approaches are carries with it a possibility of weakening the
possible, including: utility of the taggant program, and would
probably be implemented only if necessary;
1. manufacture and maintain a large inven-
for instance, if a manufacturer ran out of tag-
tory of taggant materials, either by the gants and would otherwise be forced to stop
manufacturers directIy or by the Govern-
product ion.
ment acting as purchasing agent; a 6-
month supply should certainly be ade- In the OTA baseline costing estimate, the 6-
quate; month inventory option was assumed, and
2. develop redundancy by constructing a manufacturing cost estimates include the cost
backup manufacturing site for taggants; of the taggant inventory, as well as the cost of
and money to carry the inventory.

The detect ion taggant sensor program is in ity at the parts-per-tril ion level and low (0.01
the very early stages of development. To date, percent) false alarm rates. Parts lists for each
most of the effort in the detection area has of these systems have been identified and
been devoted to the vapor taggant selection priced by Aerospace Corp. instrumental ion
process. Because detection taggants are still in engineers and scientists, Commercial engineer-
an early development phase, a relatively high ing “rules-of-thumb” have been utilized in esti-
degree of uncertainty exists in several of the mating production price levels. Development
principal cost-driving factors. The sensor(s) de- cost budgets and outyear forecasts totaling on
velopment and product ion unit cost estimates the order of $2.5 million have been estimated
are one area, and the quantity of sensors to be for advanced engineering development. The
deployed is another. Table 33 sets forth the estimates, by the very nature of a development
major qua I if i cat ions which underlie cost esti- program, assume that development proceeds
mates of the sensor program. Three systems smoothly and without major redirect ion of de-
are current I y undergoing development by the sign activity. In addition to the total number of
Aerospace Corp.: the continuous electron cap- sensors Iikely to be deplovecl, uncertainty ex-
ture device (CECD), the ion mobiIity spectrom- ists in:
eter (IMS), and the mass spectrometer (MS).
Performance specifications are severe for each the development cost,
of these canal i date opt ions incIuding sensitiv- ● the production unit cost,
Ch. V—Taggant Cost Review ● 105

Table 33.–Qualification of Estimating Basis for Sensors

Continuos electron capture device Ion mobility spectrometer Mass spectrometer

General availabilty of technology Currently utilized in lab situation– Commercially available 5 years– High-cost laboratory model in use–no
Brookhaven Breadboard 50 currently in use commercially available that meets
cost and performance requirements
Taggant program status Design of field Instrument in progress Off-the-shelf PC-1OO Instrument IS Preliminary design underway for
being characterized for candidate low-cost field unit
Parts (materials) identified and
estimated by Aerospace Yes Yes Yes
Taggant sensor production cost
estimated with engineering rule-
of-thumb factor applied to material
costs Yes Yes Yes
Quantities to be Implemented in a
national program Quantities depend on scenario selection–also decision to purchase Instruments rests with a large
and varied user community–airports, courthouses, nuclear reactors, nuclear weapon centers,
military communication centers, national shrines, Government office buildings, etc –quantities are
uncertain and open-ended

SOURCE Off Ice of Technology Assessment

the system or systems actually employed, In the baseline program identified by OTA, a
and total of 1,500 sensors was assumed d e p l o y e d .
the relative mix of systems to be deployed That number would include one sensor each
if severaI successfuI canal i dates emerge. for passenger screening, carry-on baggage, and
checked Iuggage for each current X-ray ma-
chine stat ion, as welI as 300 for protection of
Numbers of Sensors Needed
other high-value targets. l-he low-level pro-
Estimates of the total quantity of sensors gram assumed 800 sensors, 2 each for each cur-
Iikely to be deployed in the field are further rent X-ray station. The high-level program as-
subject to a wide range of uncertainty, as the sumed 5,000 sensors, enough for aII centrolled-
decisions must be made individualIy by a large access transportation faciIIties nucIear power-
number of organizations, aIthough reguIatory plants, important Government buildings, and
a u t horities s u c h as FAA and the NucIear Reg- portable police use
uIatory Commission couId potentiaIIy repre-
sent customers for Iarge numbers of sensors, Sensor System Related Costs
The target to be protected must be high-valued
and subject to control ted-access. With the ex- The annual unit system cost for the sensors,
ception of checked baggage, it is unlikely that including installation, maintenance, and false
any Iocation that does not now have a guard alarms, is shown in table 34. Since each point
wouId employ a detection taggant sensor. of controlled access where detection sensors
Likely targets for bombers, and Iikely Iocations are contemplated is already manned by per-
for sensors, include airports, nuclear reactors, sonnel (who check entering personnel or
nuclear weapons centers, military communica- search baggage), direct operator costs are not
tions centers, Government buiId i rigs, and com- included for the baseline case. Excess false
puter centers. There are approximately 620 air alarm rates would possibly be a cause for add-
ports in the United States, using a p p r o x i m a t e l y ing personnel, Training would be accom-
400 X-ray machines to scan carry-on luggage plished by the detector instrumentation com-
There are 70 nuclear power station~, and thou- pany and occur either at the company as part
sands of Government buildings of one type or of an operator training seminar or at the time
another. PO Iice bomb squads may also use of equipment instaIIation. Maintenance costs
portable sensors for investigation of bomb for all for al I of the canidate systems are estimated
threats. at 10 percent of the hardware investment cost.

61-401 - 80 - 8
106 ● Taggants in Explosives

Table 34.–Vapor Taggant Dotoctor System Cost (annual cost per unit)

electron capture Ion mobility
Hardware investment device spectrometer Mass spectrometer
Cost per unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $12,355 $15,160 $35,270
Installation and checkout . . . . . . . . . . . . . . . . . . . . 500 500 500
Hardware subtotal a ., ., . . . . . . . . . . . . . . . . . . . 12,855 15,660 35,770
Annual cost of investment per unitb. . . . . . . . . . . . . . . . 2,082 2,537 5,795
Annual maintenance. . . . . . . . . . . . . . . . . . . . 1,236 1,516 3,433
Cost of false alarm@ .01% rate. . . . . . . . . . . . . . . . . . 0 0 0
Total annual cost per detector . . . . . . . . . . . . . . $3,318 $4,053 $9,228
a lncludes cost of training operating Personnel
b Estimated 10-year life and 10 percent interest rate
SOURCE Office of Technology Asessment

Figure 12.– General Functional Network for Figure 13.—Estimatecl Annual Vapor Taggant
Detector Cost v. Quantity Deployed
Vapor Taggant Detector
● Hardware cost only

Annual cost
in millions of
● Annual cost = P - S

where: P = first cost

(1 + i)n - 1
S = salvage value (assumed O)
+i + SI

FY 79$ n = equipment life (estimated 10 years)

i = interest rate (estimated 10 percent)

V a p o r 25

gas source 15


Total Cost A + B + C + D
SOURCE: Office of Technology Assessment.

Mix of Sensors 5001,000 2,000 3,000 4,000 5,000
Quantity of detectors deployed
Development of the CECD, IMS, and M S
sensors is expected to continue in a parallel
sumes a total of 1,500 sensors is deployed, 90-
fashion. A system type would be eliminated if
percent IMS and 10-percent MS.
demonstrated to be infeasible. A m i x o f p o s s i -
ble sensors in the field is likely (given feasibil- The annual cost per sensor for this mix is ap-
ity demonstration) since each instrument type proximately $4,580. In the cost synthesis sec-
would be found to offer advantages in given tion program costs have been estimated for
scenarios for performance (specificity, thresh- various levels of implementation of sensor sys-
old, etc. ) and costs (acquisition and operation tem to fit various utility levels examined in this
and maintenance). The baseline program as- study.
Ch. V—Taggant Cost Review ● 107

False Alarm Costs rates greater than 0.05 percent (1 in 2,000).

Since the performance design specification for
False alarm response costs have been exam- the taggant sensor false alarm rate has been es-
ined by FAA as a function of the false alarm tablished at 0.01 percent (1 in 10,000), no false
rate for various technical approaches includ- alarm costs are expected if this performance
ing explosive vapor detector schemes. The FAA goal is realized. Cost level impacts reflect the
study examined two airline operations at Lo- particular operational activity characteristics
gan Airport, Boston, as a basis for the opera- of Logan Airport and would not necessarily
tional scenario. As false alarm rates increase, reflect nationwide characteristics. Discussions
so do the number of hand-searchers required with FAA personnel indicate that nationwide
and, therefore, the cost of operation. The re- cost effects due to false alarms would be less
sults of that analysis, adjusted for the taggant than that reflected for the Logan scenario;
vapor sensor, are shown in figure 14, where costs of false alarms, on a national average,
estimated annual cost impact for each of the would probably not be significant at rates as
airlines is shown as a function of the vapor high as a few percent, the current false alarm
detector false alarm rate. Incremental costs rate for airport magnetometers.
are incurred in a stepwise fashion at alarm
The cost of false alarms can also be calcu-
lated as a function of the cost per bag
Figure 14. —Estimated Cost of False Alarms checked. At a rate between 0.05 and 0.175, the
v. False”Alarm Rate estimated cost of increased inspections due to
false alarms is approximately 2.8 cents/bag at
system Logan Airport. At an annual level of 300 mil-
costs lion checked bags per year in the United
States, the estimated cost of false alarms due
to checked baggage alone would be approxi-
Based on analysis of selected
airline activity @ Logan Airport mately $8.4 million. As noted, the cost esti-
processing 6.5 bags/rein in mate for Logan is considered high for purposes
3 explosives vapor-detection scheme
of estimating national levels; nonetheless, the
Detection taggant sensor ●
potential cost due to false alarms would be a
significant cost impact when considered in ab-
2 solute terms. Since the cost of security checks
at airports are ultimately passed on to the
airline customer, the direct per capita costs
1 would be minimal. At an average of 1.5 bags
checked per passenger the per capita annual
cost for the above conditions would be on the
order of 5 cents. A high false alarm rate could
.2 .4 .6 lead to delays in the departure of aircraft, with
Probability of false alarm s i g n i f i c a n t l o s s e s t o b o t h a i r l i n e s a n d the
SOURCE: Off Ice of Technology Assessment delayed passengers.


The value-added costs of the taggant pro- understood for cap-sensitive packaged high ex-
gram that occur at the explosive manufactur- plosives where pilot-plant tests have been ac-
ing level are addressed here. As has been al- compl ished. The tagging implications for deto-
Iuded to earlier, the manufacturing process im- nating cord and detonators, conversely, are
placations for tagging implementation are best only addressable in a general way. As no feasi-

108 ● Taggants in Explosives

ble designs have been set forth for the required higher markups, the added cost of taggants in
tooling, and engineering design and analysis the manufacturing process. This issue will be
have not been accomplished, the implications amplified later.
for blasting cap design remain uncertain. Be-
cause the OTA study effort was time-con-
strained, t h e m a j o r s u r v e y e m p h a s i s w a s Revised Processes, Tooling, and
placed in the area of cap-sensitive packaged Facility Costs
high explosives. The estimates for cap-sensitive
manufacturing costs are based on discussions Cap-Sensitive Packaged Explosives
with the major manufacturers. Some of these Requirements for additional tooling and
estimates are applied to other explosive types equipment to accommodate the tagging proc-
where appropriate. Preliminary estimates and ess in dynamites, emulsions, slurries, and gels
analysis by the Aerospace Corp. are also uti- consist of equipment for weighing, hoppers,
lized as a cost basis for certain explosive types means of transferring taggant samples, and
and associated cost elements where deemed storage bins for secured storage areas. The
appropriate. These cases will be cited and cost for equipment to add the taggants into the
commented on as to their reasonableness and explosive mixing process is small, as most
depth of treatment. manufacturers use a handmixing operation.
Based on data provided by one explosives
The following subsections address each of
manufacturer, OTA estimated the added cost
the manufacturing cost elements considered in
for these investments as a function of the
this study. The last subsection summarizes the
estimates of the various elements of manufac- unique batch size and other considerations re-
turing cost. garding waste and productivity. OTA assumed
a 10-year Iife, 10-percent interest rate in order
Estimates of the current cost for each of the to annualize this initial investment. Detailed
explosive product categories considered are requirements for other manufacturers of cap-
shown in table .35, along with the raw material sensitive packaged explosives were not made
costs. The difference between price and raw available for this study. OTA believes that
material costs is made up primarily of labor, these marginal cost requirements are represen-
overhead, and markup (profit). Specific data tative of the cap-sensitive explosives industry.
for these important elements of cost were not
available to this study, since this kind of data The Aerospace Corp. indicated that some
is considered extremely proprietary. The un- manufacturers might wish to install automatic
certainty in the specific division of the other taggant-dispensin g equipment, and concluded
costs and markups makes it difficult to assess that this cost should be similar to the cost of
the degree to which the explosives manufac- the labor it replaces and hence would be cov-
turer will either absorb, or pass on through ered under the Iabor cost element. OTA’S
Table 35.–Current Manufacturing Cost/Price Data study survey and site visits did not uncover any
particular requirement for automatic dispens-
Current cost of ing equipment at either gel or dynamite manu-
explosive raw Average current
Explosive product category materials a price per unnit
facturing facilities.
Cap-sensitive explosives. 15c/lb 50@/lb
C a s t b o o s t e r s 60cflb $1.5011b Cast Boosters, Smokeless Powder, and
Black powder . . 11 c/lb $6- $9/lbd Black Powder
Smokeless powder. NAc $6 - $9/lbd
D e t o n a t i n g c o r d 2c/ft 5/ft Specific tooling and equipment require-
B l a s t i n g c a p s 20c - 30c/cap 50c/cap
— ments for these product categories were not
‘Source I ME available. For estimating purposes the assump-
bAerospace Corp
c The ME reference did not contain thiss data It IS known that the military Pays on the order of tion was made that the estimate for cap-sensi-
88 cents/lb
d A leading manufacturer has recently quoted$9 of powder tive explosives should be a representative val-
SOURCE Officeof Technology Assessment ue until detailed requirements are established.
Ch. V—Taggant Cost Review ● 109

Detonating Cord would be required in various locations such as

the dope house, works control, laboratory (in-
Tooling designs must be developed in order
cluding works laboratory), and in the magazine
to provide tagging capability at each detonat-
area. Additional activities involved include
ing cord production line. Aerospace Corp. indi-
ordering, stocking, weighing, and supplying
cates that several pieces of hardware have
taggants to operators; collecting data, taggant
been tested but no effective equipment is cur-
samples, keeping records of codes; handIing in-
rently available. They further feel that a sta-
creased record keeping in magazine areas; and
tion configuration would apply both the identi-
examining the codes before use in the manu-
fication and detection taggants together with
facturing process. One contractor also indi-
an adhesive before the final assembly polyeth-
cated increased manpower costs due to code
ylene sheath is applied, and that a reasonable
confusions and returned shipments. It should
cost for a station having a 5-year Iife is $50,000.
be noted that incremental labor costs for the
Five such stations would be required by the in-
actual mixing operation of taggants and re-
dustry for an annual production of 500 mill ion
lated packaging are essentially zero. All addi-
ft. The estimated cost for detonating cord tool-
tional estimated labor costs are associated
ing is $ 2 5 0 , 0 0 0 . A m o r t i z i n g t h i s c o s t o v e r 5
with peripheral activities in coordinating, han-
years at 10-percent interest yields an annual
dling, and recordkeeping activities.
cost of $66,000 or $0.00013/ft.
The estimate for labor, as indicated by the
Blasting Caps manufacturers, is SIightly greater than 1 cent/lb
The process by which taggants would be of explosives, which reflects approximately
added to blasting caps has not yet been deter- five to six additional men at the plantsite.
mined; it may well vary from one manufac-
turer to another. Alternate possible ap- Cast Boosters
proaches are to place the taggants between
For the purposes of developing a baseline es-
two end plugs, embed the taggants within a
timate, the Aerospace Corp. analysis is utilized
single end plug, or add taggants to an existing
here. Assuming that this will be a manual proc-
interior polyethylene strip. Cost will vary con-
ess, two additional personnel were estimated
siderably depending on the process chosen
per assembly line, Given the four manufactur-
and the current cap assembly process. For pur-
ers (eight Iines) the estimated annual cost is
poses of the study, a conservative value of $2
$400,000 or $0.067/lb of explosives.
milIion per manufacturer was assumed. Amor-
tizing the $8 million cost (four manufacturers)
over 5 years yields an annual cost of Black Powder
$2,112,000 or $0.025/cap. This figure would be Labor costs associated with tagging black
high if one of the simpler methods of tagging powder were studied by the Goex Co. and ref-
detonators were adopted. However, the effect erenced in the Aerospace Corp. Inflationary y
on the total cost of a tagging program is small. Cost Impact Study. The estimated cost per
pound of black powder for manufacturing la-
bor of 1.5 cents is based on replacing the pres-
Labor ent date-shift code with a tagging material sys-
Cap-Sensitive Packaged Explosives tem. EIements include:

Manpower estimates by the manufacturers storing tagging materiaIs,

indicated a range of requirements varying security for storage and handling of tag-
from two to six additional men at a site, The ging materials,
variation results from differences among par- administrative and recordkeeping, and
ticular site layouts, processes, and procedures impact on the manufacturing process (as-
in use, For instance, in one company effort suming a cleanup would be required in
110 ● Taggants in Explosives

the glaze and packhouse operation each Productivity y

Cap-Sensitive Packaged Explosives
This cost is exclusive of taggant material costs. Potential productivity losses have been esti-
Based on the study by Goex, OTA estimated mated by the industry to be as high as 15 per-
the cost of labor for black powder to be 1.5 cent. The primary cause of such losses would
cents/lb. be halting production to change taggant codes
and avoid contamination. Consequently, the
Smokeless Powder extent of such losses depends on the degree of
taggant cross-contamination that would be
The Aerospace Corp. estimated labor effort
permissible and the taggant batch size. Vari-
added costs per pound of smokeless powder to
ous kinds of cost can impact the situation.
be on the order of 6.6 cents (including the
They are:
distribution system costs) and assumed that
much of this cost could be absorbed within the ● loss associated with scraping of hoppers,
current manufacturing and distribution organi- ● new investment to offset production
zation. The estimate is based on the following losses,
assumptions: ● loss of the market for mixed scrap, cur-

rently sold as an inexpensive explosive,

● 2,000 Ib/lot, and
2,500 different tag lots produced, and ● new investment for expanding waste dis-
● 100,000 cases/year (50-lb cases).
posal facilities.
Manufacturing costs were estimated to be As currently perceived by one major manu-
0.4 cents (of the total 6.6 cents). Since ade- facturer of cap-sensitive packaged high explo-
quate data are unavailable to validate the esti- sives, productivity losses wilI have a direct cost
mate, OTA estimated the cost of manufactur- impact in each of the areas noted above. Pro-
ing labor for smokeless powders at the same ductivity losses are estimated at 15 percent in
level as black powder, using the Goex estimate the condition where cross-contamination is not
of 1.5 cents/lb. permitted and on the order of 8 percent where
batch cross-contamination of 10 percent is
Detonating Cord and Blasting Caps permitted. Waste losses associated with scrap-
ing of hoppers every fourth mix were also esti-
The Aerospace Corp. estimate for detonat-
ing cord assumes that each assembly Iine mated. A significant amount of the mixed
scrap material is currently marketed as a low-
would require one additional person to main-
quality explosive. If this material could no
tain a tagging station and to operate it during
longer be marketed due to extensive taggant
production. At $25,000 per man, the five sta-
cross-contamination, there would be a further
tions would add an annual cost of $125,000 or
loss in profits. Current environmental regula-
$0.00025/f t of cord,
tions require that waste be disposed of by
Similarly, the Aerospace Corp. estimates are means other than burning in the open, in effect
used for blasting caps. Several additional requiring additional waste disposal facilities.
workers may be necessary to operate and In order to maintain the current production
maintain the new equipment required. A rea- and sales base, and thus maintain an adequate
sonable estimate is four per manufacturer profit level for the company, additional pro-
(there are four manufacturers) for an annual in- duction facility augmentation would be re-
crease of $400,000. The resulting cost per blast- quired to offset the expected losses in produc-
ing cap is $0.0048/cap. tivity.
Ch. V—Taggant Cost Review ● 111

The total cost due to losses in productivity bor would be required for cap-sensitive explo-
could thus add up to several cents per pound sives, as the batch size would be the same as
of explosives for the worst case condition. If a the current date-shift batch size. For the high-
10-percent taggant cross-contamination level Ievel program, with 10,000-lb maximum batch
were permitted (BATF assumes this level) the size, each batch would need to be separated
cost impact would drop dramatically. If a spe- by an access aisle from other batches, requir-
cial “composite code” were created, then tags ing additional space and labor. Access aisles
containing this code could be added to scrap would need to be maintained for inventory
material and any other material containing control and inspection.
cross-contamination in excess of 10 percent;
investigators findin g tags with the composite
code would know that any other tags should Quality Control
be ignored, This would essentially eliminate
Quality control cost estimates are included
costs for decreased productivity, The OTA in the labor costs element. Some level of effort
baseline program assumes that such a compos- is required to ensure the taggant code and tag-
ite code taggant is used, so that productivity gant quality prior to mixing. This effort would
losses are negligible.
take place in the plant lab or “works” lab, to
examine each code before use in the product.
Other Explosive Categories This appears to be a reasonable precaution
Since pilot testing of adding taggant mate- since the integrity of all substances entering
rial to boosters, gunpowder, detonating cord, the “mix” must be assured to maintain prior
and caps has not taken place, the effects on safety levels. In addition, occasional speci-
productivity are not apparent. For purposes of mens would be examined to assure that the
costing the baseline system, OTA assumed taggant-mixing specification (uniformity, shelf-
there would be no productivity losses. Iife, etc. ) was being achieved.

Inventory Costs Safety

Inventory costs, including the associated Requalifying all product lines with taggant
cost of money, are a function of supply held in materials would be a necessary safety testing
inventory. There is no reason to assume the requirement for the various explosives manu-
tagged finished product would be held longer facturers This one-time capital cost would in-
than is currently the case. It may be necessary, volve analysis and testing of each type of prod-
however, to stockpile a significant inventory of uct. To an extent uncertain at this time, the
the taggant material to ensure an uninter- pilot testing programs have and will contribute
rupted supply, particularly for identification to this requalification effort. Due to the uncer-
taggants, where there is likely to be only one tainty involved, OTA included the cost of safe-
supplier. For the baseline case, the quite con- ty requalification in the cost element esti-
servative assumtion was made that a 6-month mates. It should be pointed out that the abso-
inventory of both types of taggant materials lute cost levels of nonrecurring costs are not
would be stockpiled. The added costs for the insignificant. However, after amortizing these
various types of explosives wouId be: costs over the significant production weights
of explosive produced annuaIIy, the relative
(’alp Cap sensitive $0 0021‘1 b contribution of incremental costs to a pound
Boosters $0 0066/lb of explosives is quite smalI.
Smokeless powder $0 0021/1 b
Black powder $0. 0021 lb
Record keeping Costs
Space and added labor have been included in
the facility and labor costs detailed above. For I n order to maintain the integrity of the iden-
the baseline case, no additional storage or la- tification taggant tracing network, a certain
112 ● Taggants in Explosives

amount of additional or new recordkeeping Markup

must take place within the explosives distribu-
tion network. Current Federal requirements are To the extent that incremental taggant costs
that each explosive package and shipping case are passed on to distributors and users, markup
be marked with an identification code citing costs must be included as part of the final
the: product price. No specific data were available
to treat markup for most of the explosive prod-
● plant of manufacture, uct categories. For purposes of establishing a
● the date and shift manufactured, and baseline cost estimate, OTA assumed a IO-per-
● the type and grade of explosives. cent markup at the manufacturing level. This
explosives covered under this regulation are value may seem low, but all handling costs
the: have been specifically covered in other cost
elements, including an overhead allowance.
● cap-sensitive packaged explosives (dyna- Markup in that sense is essentially profit on the
mites, slurries, water gels, and emu Is ions), additional costs. Normal markups must cover
● cast boosters, al I of the handling costs.
● blasting caps,
● black powder, and In addition to manufacturing level markups,
● detonating cord. OTA considered the pyramid of markups that
occurs throughout the various echelons of dis-
Records of the identification code must be tributor and retailer levels. This is addressed in
maintained at the manufacturer level as well the next sect ion.
as each subsequent distributor. Smokeless
powders are currently exempt from this re-
quirement, although powders used to hand- Summary of Manufacturing
Ioad pistol ammunition must be recorded at Costs Added
the retail sales level.
Manufacturing costs elements and total cost
The cost of recordkeeping has been in- added as a result of the inclusion of identifica-
cIuded as part of the labor manufacturing cost tion and detection taggant materials in explo-
elements. sives are summarized in table 36. The added

Table 36.–Summary of Explosives and Gunpowder Manufacturing Costs included

— . — — . — .
Costs included
Baseline case Black Smokeless Detonating
Cost element cap sensitive Boosters powder powder cord Blasting caps
Nonrecurring costs
Tooling Yes Yes Yes Yes Yes Yes
Storage No No No No No No

P r o d u c t r e q u a l i f i c a t i o n Yes Yes Yes Yes NAa NA

Waste disposal facilities No No No No No No
New investment to offset product losses No No No No No No
Recurring costs
Manufacturing process labor. . . . . . . . . .
Recordkeeping . . . . . . Yes Yes Yes Yes Yes Yes
Quality control. . . . . . . . . . . . . }
Product losses. . . . . . . . . . . . . . . . . . . . . . . No No No No No No
Waste product line , . No No No No No No
Inventory costs Yes Yes Yes Yes Yes Yes
A d m i n i s t r a t i v e e x p e n s eb .
Bottom line cost per unit of explosives 1.03/lb 7 7c/lb 2c./lb 7,2./lb .04c/lb
— .
Data unavailable
lncluded in labor
SOURCE Office of Technology Assessment
Ch. V—Taggant Cost Review 113

costs include the estimated costs to the manu- not shown here. The detailed cost data were
facturer and associated markup as well as the anaIyzed and alternative ground rules were
markup placed on the cost of the taggant raw established to gain insight into cost effects
materials. where taggant batch size was varied; related
effects were taken into account regarding the
Manufacturing costs for cap-sensitive pack- productivity and waste issues. The cost ele-
aged high explosives are based on detailed in- ments incIuded in various assumptions, along
puts received from a major manufacturer. The with the bottom Iine cost per- pound of explo-
raw data are proprietary informat ion and are sives, are shown in table 37.

Table 37.–Cost Summary of Cap-Sensitive Packaged High Explosives Manufacturing Cost Variations With Assumptions

Costs Included
Case 1 — Case 2 Case 3 Case 4 Case 5
20,000-lb tag batch size
10000-12000 lb tag plus allow cross- Tag batch size equals
Cost elements batch size 20,000-lb tag batch size contamination day s production Plan! /year

Site manpower Yes Yes (less than case 1 ) Yes (less than case 1 ) ‘ No No
Production losses Yes Yes Yes No No
W a s t e Yes Yes (less than case 1 ) No No No
Requalification Yes Yes Yes Yes Yes
Waste disposal facilities Yes Yes (less than case 1 ) No No No
Equipment and storage Yes Yes (less than case 1 ) Yes (less than case 1 ) Yes (less than case 3) Yes less than case 3)
Investment to offset production losses Yes Yes (less than case 1 ) Yes (less than case 2) No No
Taggant Inventory costs Yes Yes Yes Yes Yes
Administrative Yes Yes Yes Yes No
Bottom line cost per pound of
explosives excluding markup 4.Oc/lb 2 3c/lb 1 4c/lb O 6c/lb O 3c/lb

SOURCE Office of Technology Assessment

A general schematic illustration of the dis- tioning of explosive products may be required
tribution network for explosives is s h o w n i n beyond that required by the date-shift code
figure 15 while the network for gunpowder is reguIations, which may or may not have an in-
shown i n figure 16 Detailed quantitative net- cremental cost effect at the distribution level.
works are not available; however, these illuss NO detailed studies of additional recordkeep-
trations serve to depict the manner in which ing elements which wouId be required, or the
transactions take place within the industry. time necessary, have been conducted to date.
Within the networks, potential cost impacts I IME assessment of new activity requirements
occur in the areas of recordkeeping, process- by the distributor includes.
ing and handling, storage, and further poten-
● comparing the lot numbers with
tial pyramiding of markup costs throughout
the bilI of lading with greater- frequency,
the distribution network.
● classifying each explosiveproduct by typeby produ ct by t ype

and taggant lot number to faciIitate locat-

Recordkeeping at Distribution Levels ing records,
● expanding storage space for the increaseci

Record keeping and control of packaged number- of books and records, and
high explosives are required by the present ● increasing the time to Iocate the proper
date-shift code regulation. AdditionaI part I- product and taggdnt lot number at sale
114 . Taggants in Explosives

Figure 16.—Schematic Distribution Network

of Gun powders


points ■



User Retailer


SOURCE: Office of Technology Assessment.

(due to the greater number of records that ● 1.46 codes per order (12,000-lb tagging lev-
must be searched). el) (based on Du Pont data), and
. 1.66 codes per order (7,900-Ib tagging Iev-
The Aerospace Corp. further considered:
el) (based on dynamite traces).
● segregating material on trucks and in
In effect these data indicate that the addition-
magazines to a smaller quantity; and
al recordkeeping, processing, and handling ef-
recording additional information in or-
forts for the finished explosives may be in-
ders, invoices, and inventory lists.
creased by up to 66 percent, depending on the
An analysis by the Aerospace Corp. of avail- tagging level. A plot of activity increases ver-
able BATF tracing records revealed that rec- sus tagging level is plotted in figure 17. This
ordkeeping entries on bills of lading would in- plot underscores the dramatic inverse relation-
volve: ship of recordkeeping activity with the unique
tagging batch level.
1.26 codes per order (20,000-lb tagging lev-
el) (based on 282 BATF traces of seven The Aerospace Corp. further reviewed the
manufacturers in 1976 and 1979), additional data entry requirements which
Ch. V—Tagganf Cost Review ● 175

Figure 17.— Recordkeeping Activity v. ● cap-sensitive explosives,

Tagging Level ● boosters,
● detonating cord, and
● Cap.sensitive packaged explosives
● Impact on distributor/retailer ● blasting caps.
● Based on Aerospace analysis of
BATF tracing record This conclusion is particularly appropriate for
the baseline case, in which the taggant batch
percent increase corresponds to the current date-shift code
in recordkeeping, batch size.
processing and
handling effort The impact on the distributors of black and
70 smokeless powders is somewhat different.
Black powder and pistol-grade smokeless pow-
der currently have significant recordkeeping
60 requirements, while the other smokeless pow-
der grades have no current recordkeeping re-
quirements. (Pyrodex”, a black powder substi-
50 tute, would be marketed and regulated like
smokeless powder, so incremental recordkeep-
ing costs would approximate those of smoke-
40 l e s s p o w d e r . ) An estimate was therefore made
of the additional cost of entering the currently
unregistered smokeless powder in, and detail-
30 ing it out of, the records at each distributor
level by taggant code. It was assumed that a
record for an “item” would take 2 minutes.
20 The further conservative assumption was
made that the average size of an “item” at the
master distributor level was 25 lb (primarily
case lots handled), was 10 lb at the distributor
5,000 10,000 15,000 20,000 level, and was 2 lb at the retail level. Since con-
Tagging level in pounds siderable recordkeeping requirements current-
ly exist for pistol-grade smokeless powder, the
SOURCE Off Ice of Technology Assessment
costs were assumed to be half those of the
other powders. A small additional cost for rec-
would be required on bills of sale. Tagged ex- ordkeeping was assumed at the retail level for
plosive materials would require approximately black powder. The cents per pound added by
25 percent more entries than the untagged ex- those costs are shown in table 38.
plosives for transactions at the distributor
level. This analysis was specifically for tagging
at the 20,000-lb level. At the retailer/explosive Storage
user level an 8.7-percent increase in data en-
Explosives are now generally separated by
tries were computed using Federal form 4710
date-shift code batches for magazine storage
and the biII of sale or delivery ticket.
at all levels in the distribution chain, as records
Aerospace did not quantify the absolute must be kept, and physical control maintained
cost impact as a result of this tracing analysis, by date-shift batch. For the baseline taggant
but did conclude, however, that the costs case, no changes would be necessary. If the
would be insignificant for cap-sensitive pack- taggant batch were smaller, then additional
aged high explosives. The OTA analysis as- storage space would be required for access. An
sumed that negligible added costs exist at the estimate was made of the cost of magazine
distributor retailer level for: space, based on two data points. The added
116 ● Taggants in Explosives

Table 38.–Estimated Cost Impact for Powders additional storage space at both the master
at Distribution Network (cents per pound)a distributor and distributor levels, but probably
Smokeless powder
not at the retailer level. Using the same data
Pistol loading Rifle and
base as above, the cost was estimated to be ap-
Distribution level Black powder grade shotgun grade proximately 0.2 cent/lb at each level, as shown
Master distributors in table 38.
Recordkeeping . . . ., 0 1 .2b 2.4C
Storage . . . . . . 0.2 0.2 0.2
Distributor/wholesale level Summary Cost Including Markup
Recordkeeping . 0 3d 6C
Storage . . . . . . . . . 0.2 0.2 0.2
Retail level Distribution level costs are summarized in
Recordkeeping. . . . . 1 I5e 30C table 39. Markup on total costs incurred
Storage ., . ., . . 0 0 0 through the distribution system for explosives
1,4 19,6 38.8 was assessed at 25 percent; for black and
Total cost through the distribution chain
Black, . ., . . ., . . ., ., ... .. 1,4$ smokeless powders a total markup of 80 per-
Pistol . . . . . ... . . . . ., .19,6$ cent was assumed. This estimate is based on
Other. . . ~ . . . . . . . . . . .. 38.8$ analysis of costs and price at each level, sup-
If pistol powder is assumed to be 25 percent of total smokeless powder, plied by an integrated powder distributor.
the average cost impact for smokeless powder is 33c/lb.
Table 39 sets forth the net cost added by the
aEstimate by Integrated master distributor wholesaler, retailer
bBased on I minute Average lot size 25 lb
distribution network and further summarizes
Assume 2 minutes/lot the net cost to explosive users from both
‘Assumed lot size IS 10 lb
‘Assumed lot size IS 2 lb manufacture and distribution for the various
SOURCE Off Ice of Technology Assessment explosive categories. To illustrate the effect
that the method of program implementation
cost per pound of explosives was less than 0.1 can have (taggant batch size and treatment of
cents, even for the case in which 10,000-lb waste), costs for the five cases previously
maximum lots were tagged. For black and defined for the cap-sensitive high explosives
smokeless powders, the assumption was made are shown. Case 4 is, as noted, the OTA base-
that separation by taggant lot would require line case.

Table 39.–Distribution System-Summary of Cost Added and Markup (cents per pound)
Total cost Ieaving Distribution Distribution Total cost added by Total added price
Explosive category manufacturing facility system cost added system markup distribution system to user
Cap-sensitive packaged high explosivesv
Case . ., 8.5 0.2 2.2 2,4 10,9
Case 2 . 66 01 1.7 1.8 8.4
Case 3 ., ... ., 5.6 0.1 1.4 1.5 7.1
Case 4 (baseline). . . . . . . ., . . ., 4.8 — 1,2 1,2 6.0
Case 5 ..., . . . ., . . 29 — 0.7 0.7 3.6
Boosters . . . . . . ., ., . . . 20.9 0.2 5.3 5.5 26,4
Black powder, 63 1,4 6.20 7.6 13.9
S m o k e l e s s p o w d e r 6.3 33.0 31.4 64.4 70,7
Detonating cord ., ., 0.6 — 0.2 0,2 0.8
Blasting caps ., 50 — 1.2 1,2 62

SOURCE Office of Technology Assessment

Ch. V—Taggant Cost Review ● 117


The cost increases estimated to occur as a ● distribution network cost added inc uding
result of the baseline taggant program are sum- markup.
marized and their impact on users analyzed.
The following increases are noted for the base-
line case:
Increased Material Costs Explosive category Percent cost increase
Cap-sensitive packaged high explosives. ., .. ..11.9
The net cost increase due to tagging e x p l o -
B o o s t e r s
sives is summarized here. Summary cost im- Black powder. 2.3
pacts include: Smokeless powder . . .. 11.8
Detonating cord .23.5
● the cost of identification taggant materi- Blasting caps . . . . 15
● the cost of detection taggant materials, The individual contributing cost elements to
● manufacturing costs added including the overall cost impact are illustrated in figure
markup, and 18 for the respective explosive categories.

Figure 18.—Summary of Added Costs to Explosive Users Cost

Per Unit of Explosives in Dollars

$/lb Dist.
9.70 .57 . cost
$/lb r
9.60 .062~ .56 .

. Mfgr.
.55 9.50 .55

.54 9.40 .54

$/lb $/lb
.53 1.80 9.15 9.30 1 .52
.52 1.70 9.10 .52

.51 1.60 9.05 .51 ID tag


.5( . -Current
.50 1.50 9.
‘ r i c e
Cap-sensitive Boosters Black Smokeless Detonating Blasting
packaged powder powder cord caps
SOURCE: Office of Technology Assessment
118 ● Taggants in E x p l o s i v e s

For the baseline case, the overall average in- Underground Mines
crease in costs due to tagging is on the order of
The Crow Fork (Anaconda) Mine near Toole,
12.8 percent, the weighted average for each of
Utah, is a large, deep underground operation
the above percentage contributions. The esti-
in hard-rock, mining for essentially high-grade
mate of absolute annual cost increase in explo-
ore. The mine will primarily produce copper,
sives is approximately $37 million.
although significant amounts of silver, gold,
and molybdenum are expected as byproducts.
Commercial Uses of Explosives and This mine is still under development and has
Gunpowders—General had no production of ore as yet. Mine reserves
are estimated at 20 years with an estimated
Who uses commercial explosives and gun- production output capacity of 10,000 tons of
powders? Over 55 percent of the total weight ore per day. The total use of explosives is pro-
of explosives and blasting agents is utilized in jected to be approximately 0.6 percent of total
the mining of coal, both in underground and operating costs. Approximately 80 percent of
surface mining operations. Quarrying and non- the explosives used are non-cap-sensitive gels
metal mining are next in rank (1 5.4 percent) fol- and blasting agents such as ANFO. The remain-
lowed closely by metal mining (14.6 percent). ing 20 percent of explosives, including dyna-
Construction work at 10.6 percent and “other mites, slurries, boosters, detonators, and deto-
uses” at 4.2 percent complete the spectrum of nating cord would be subject to a tagging re-
user classes as adopted by BOM’S annual quirement if taggant legislation were enacted,
“Mineral Industrial Survey s.” Onsite investiga- A 12.8-percent boost in the cost of tagged ex-
tions were conducted for each of the major plosives would translate into a 0.02-percent in-
user classes in order to determine the order of crease in the cost of mining, certainly an in-
magnitude cost and economic impact to the significant cost increase. The use of ANFO is
users of tagged high explosives. The selection currently related to clearing and aboveground
of users investigated included both under- excavation. Steady-state underground mining
ground mining and surface mining as each in the future can be expected to change the ex-
type differs in the relative utilization of high plosive mix and potentially increase the cost
explosives. Onsite investigations were con- increase noted above. If all explosives used in
ducted with the following users during the the future were the cap-sensitive types, a tag-
course of the study: gant program would increase mining costs less
than 0.1 percent.
Underground mining
Metal mining (copper) –Anaconda, The The cost impact on underground coal min-
Crow Fork Mine, Utah ing is somewhat higher. At present, the cost of
Coal mine– Webster Coal Co., Kentucky. the cap-sensitive slurry and detonators (the ex-
Quarry plosives used to mine the coal) represents ap-
Tri State, Maryland proximately 1.4 percent of the total cost of
Rockville Crushed Stone, Maryland bringing the coal out of the ground. The in-
Surface mining (open pit] crease in the cost of the explosives, due to tag-
Metal mining (copper)– Kennecott ging, would increase operating costs less than
“Bingham Canyon Mine, ” Utah 0.2 percent. Other economic factors far out-
Construction work weigh increases of this sort.
Guy F. Atkinson, California
The following sections describe the findings
of the limited number of intensive investiga- Discussion with the Rockville Crushed Stone
tions of the above explosive users, Quarry revealed that explosives contribute to
Ch. V—Tagganf Cost Review ● 119

slightly over 8 percent of the gross total costs explosives used per day or over 36 million
of operation. Between 1.5 million and 1.75 mil- lb/year. For every pound of explosives used, 4.2
lion lb of cap-sensitive (80 percent) and non- tons of material are mined. Cap-insensitive ex-
cap-sensitive (20 percent) explosives are uti- plosives predominate the utilization, consist-
lized annually at their location. Since the envi- ing of almost 80-percent ANFO and almost 20-
ronment is wet, no ANFO is currently utilized. percent cap-insensitive slurry. Explosive costs
The blasting activity is all contracted with a run from 3 to 5 percent of total operating
local blasting jobber, who provides the drilling, costs. High explosives, although a smalI per-
explosives, and blasting operation. The cost centage of the total weight of explosives used,
impact of an increase due to a tagging pro- account for 7 to 10 percent of costs for all ex-
gram is thus significantly higher in this ex- plosives used in the mine. Large amounts of
plosive-intensive operation, However, the in- primacord are used, together with boosters,
crease would still be less than 1 percent of op- detonators, and some dynamite for secondary
erating costs. If the costs of explosives, caused blasting (e. g., breaking up boulders). High ex-
by legislation of a tagging program, are much plosives therefore contribute on the order of
higher than estimated for the baseline pro- 0.3 percent of the total cost of operation. The
gram, then the quarry might investigate the cost increase for a baseline taggant program
cost potential of using inexpensive blasting would be on the order of 0.03 percent of oper-
agents, coupled with a water pumping opera- ating costs.
t ion.
A quite dissimilar situation is provided by
The study team discussed the impact of
the quarry operated by Tri State Explosives.
The Tri State Quarry produces “facing stone” tagged explosives with the Guy F. Atkinson Co.
in South San Francisco, Cal if., a large con-
in various grades. The use of explosives in the
operation is relatively insignificant, averaging tracting firm that utilizes large quantities of
explosives in both underground (tunnels, etc. )
from 10 to 15 blastings per year. Between 15 to
105 lb of explosives are used in each blasting, and aboveground construction operations. In
recent years this firm has utilized on the order
characterized as a “very precise operation. ”
The incremental cost of tagged explosives is of 20 million lb of explosives annually. In un-
derground applications, operating costs are
therefore trivial.
considered to be very sensitive to the cost of
powder. Values placed on underground opera-
Open Pit Mines
tions were:
The OTA study team visited the Kennecott
Pounds Of
“Bingham Canyon Mine” near Salt Lake City,
powder to
Utah. This open pit mine has many distinc- remove yd 3 Cost per yd’
tions, including: G e n e r a l 1/4 to- 1 3\4 lb 13 - 8 8
C o a l ‘/~ I b 17~
● the world’s largest manmade excavation, Hard-rock 1 lb 50~
● the first open pit mine in the copper indus-
try (started in 1904),
● the largest single mining operation ever In a recent tunnel application, Guy F. Atkin-
undertaken, and son used approximately 900,000 caps in the
● the holder of the largest copper produc- construction of a 22-mile tunnel. At an esti-
tion record of any individual mine in his- mated 50 cents/cap, the value of caps alone
tory. amounted to approximate y $500,000.
Figure 19 shows a photograph of the Bing- In aboveground work, Guy F. Atkinson re-
ham pit. Each vertical terrace is approximately cently utilized over 40 milIion lb of explosives
50 ft high. The mine is an extremely large user in the construction of the Maloney Dam in Cal-
of explosives, with approximately 105,000 lb of ifornia. This fixed-price contract was very
120 ● Taggants in Explosives

Figure 19.— Bingham Canyon Open Pit Copper Mine

Photo credit: Kennecott Copper Co

“powder intensive. ” The value put o n e x p l o - classes is shown in table 40. Explosives per-
sives was approximately 9 percent of operating centage contributions to operating costs vary
costs, consisting of 70-percent cap-sensitive ex- (dependent on user type) from less than 1 per-
plosives and 30-percent ANFO. cent (underground metal mining) to as high as
9 percent (dam construction example). As a re-
A baseline taggant program would increase
sult, the cost impact of an increase in the price
operating costs approximately 1 percent, a sig-
of cap-sensitive high explosives also varies,
nificant cost, but probably not sufficient to
particularly as these explosives represent vary-
cause a shift to alternative excavation meth-
ing portions of the total explosive mix used.
ods. One additional potential impact should
be noted. Such construction projects are nor-
mally long-term, fixed-price contracts. A sharp Hand loading
jump in the cost of explosives during the
The above cost impact calculations were for
course of the contract could significantly af-
industries that are generally able to pass on in-
fect profits.
creases in the cost of operations to their cus-
A summary of the findings on current explo- tomers. Handloaders, however, are the ulti-
sive cost contributions to the various user mate users of the product, and must absorb
Ch. V—Taggant Cost Review ● 121

Table 40.–Current High Explosives Cost Impact any increased cost due to a taggant program.
for Various User Classes’ Handloaders load their own ammunition for
Percent Increase in
two reasons —economy and the hobby aspect.
operating costs A less than 10-percent cost increase in expend-
Percent of due to baseline able material is unlikely to affect a hobby for
operating costs taggant program
which hundreds of dollars in costs have al-
Underground metal mining 0.2b 0.02
1 .4C 02
ready been incurred (hand loading equipment
Underground coal mining
Open pit metal mining 0.2 to 0.5d 003 and guns). As powder is only one of several
Quarries 8 0e 10 materials on which a hand loader saves costs
Aboveground dam construction 9.0 e
(cartridge cases, projectiles, wadding) and ad-
Excavation–general 2 to 3 ditional cost-savings are realized from labor
Tunneling 5 — and avoiding — paying the excise tax on pur-
dTheSe are single poml samples chased ammunition, an 8-percent increase in
bTo[al Operating COSIS mcludlng ref[mng were nol available For dlrecl mlnmg cost OPeratlOnS e~
powder cost would translate into a very few
ploslves accounted Ior less than 1 percent of costs
NOTE This data point reflects a htghly efflclent operalton percent increase in total reloading costs.
‘Excludes blasllng agents
‘Includes blastlng agents
SOURCE Otflce of Technology Assessment


Government Investigation Costs jected outyear costs are estimated at $4.6 mil-
and Program Administration lion.

BATF has estimated* a requirement of 11

man-years of effort annually to enforce the
Investigative Costs
provisions of S.333, primarily to establish
standards and monitor implementation of the Investigators of bombing incidents currently
taggants program. Estimated program costs in devote considerable time to examining explo-
fiscal year 1979 dollars for this level of effort sive debris for clues regarding the type and
are approximately $500,000. This would in- source of the explosive material. Further effort
clude several explosive specialists, chemists, is devoted to forensic analysis at the labora-
inspectors, and clerical help. Estimated costs tory level. If an identification taggant program
for actually investigating taggant-tracing serv- is implemented, collection of debris for a lab-
ices are expected to be marginal beyond cur- oratory search for taggants wiI I become part of
rent BATF personnel levels and are contained the standard bombing-scene investigatory pro-
in the above estimate. Their current tracing cedures. There should be little or no impact on
service personnel would require one additional the time required for a bombing-scene investi-
slot at a cost of approximately $30,000. The to- gation. Taggant recovery from the debris will
tal annual costs estimated for BATF are, there- be an additional laboratory exercise but it
fore, just over $500,000. could we I I replace the more time-consuming
Completing the spectrum of Government procedures now carried out to obtain less in-
level costs are those expenditures that are formation than would be furnished by tag-
budgeted and projected to complete the tech- gants. Similarly, it will take time to follow up
nical development of the taggants program by on the leads furnished to investigators by hav-
the Aerospace Corp. Total program costs (in- ing a list of last legal purchasers of the bomb
cluding sunk costs of $5.4 million prior to fis- filler material, but that time is probably less
cal year 1980) are $10.0 million budgeted; pro- than would be expended following up less di-
rect leads. For purposes of this study, the as-
sumption was made that a taggant program

61-401 0 - 80 - 9
122 . Taggants in E x p l o s i v e s

would have no net cost effect on investigation borehole dewatering (by pumping the hole out
time. and utilizing a borehole liner) coupled with
ANFO. This kind of substitution, for cap-sensi-
tive packaged high explosives to ANFO, was
Effects of Competition-Substitution also noted by an explosives jobber (operating
Depending on the ultimate rise in the price in a quarry environment) as a highly likely
of explosives to the user community due to the prospect should the cost of tagged explosives
addition of taggants, a variety of economic im- increase inordinately. l-he accuracy and objec-
pacts could occur. As has been pointed out tivity of this type of unsubstantiated estimate
earlier, the choices of the type of explosive are open to question, particularly as other
purchased by users are frequently made on a operators expressed opposite views. Safety, re-
basis of the lowest price rather than brand loy- liability, and ease of handling were cited as
alty. Since this is so, various kinds of potential reasons why a cost increase, such as would oc-
substitution threaten the explosives industry if cur for the baseline tagging program, would
the user perceives more economical choices not cause a product substitution. The exam-
available to him. For instance, in the under- ples do, however, highlight a very real poten-
ground mining of coal, the cost of explosives tial problem, particularly if the taggant pro-
can play a predominant role in the overall cost gram were to substantially increase the cost of
of operations, particularly so in marginal types cap-sensitive explosives, or if a program were
of mining operations. Substitution of mechani- adopted that included tagging some portion of
cal coal mining equipment could essentially a cost-competitive segment of the industry
eliminate the use of explosives in those mines. (such as tagging dynamite, but not gels and
The cost impact of the baseline taggant pro- slurries).
gram is unlikely to significantly affect that It is noted that the current annual utilization
type of choice, particularly given the capital of ANFO in this country is on the order of 3.4
investment in machinery that is currently used billion lb. It is estimated that the trend toward
to support explosive mining. A full economic utilization of ANFO has gone about as far as it
cost tradeoff analysis between mechanical can go, given the excel lent economies for
tools and the increased cost of explosives ANFO in a wide variety of circumstances. in-
would need to take place for a meaningful creasing inordinately the cost of explosives
sample size of users to determine the net ef- due to tagging could, however, further shift
fect on the explosives industry. current utilization from cap-sensitive pack-
Discussions with a dynamite and packaged aged explosives to ANFO.
slurry manufacturer revealed that in one case
a recent 5.4-percent increase in the price of a
Effects on Fixed-Price Commodities
slurry product resulted in several buyers shift-
ing to other products —a loss in sales of 6 m i l - There is a potentially important economic
lion lb of product for that manufacturer. Other spillover on the marketplace for fixed-price
estimated potential losses by substitution were commodities, due to taggants. Copper prices
suggested by the manufacturer. For instance, are established in a competitive worldwide
given a price increase of $1 0/1 00-weight in market setting. The Kennicott copper mine, for
their nitroglycerine-based products, that man- instance, competes in this environment, and as
ufacturer estimated that as much as 25 percent a result is limited in its. ability to pass on addi-
of their business would shift to other boost- tional costs of operations. Tagged explosives
er/slurry combinations. The manufacturer fur- could affect this situation, depending on the
ther estimated that if a 10-cent increase in the degree of tagging implemented and the cost of
price of packaged slurries occurred, they could tagging. The OTA analysis revealed that only
lose 50 percent of their slurry business to insignificant influences on cost of operation
ANFO, as mining operations w o u l d s u b s t i t u t e would take place due to cost increases from a
Ch. V— Tagganf Cosf Review ● 123

mandated taggant program. If ANFO and un- analysis has assumed would be amortized over
packaged slurries were also tagged, however, 10 years and passed along to the consumer in
the impact could be quite different. T h e p r i c e the form of somewhat higher prices. It is possi-
of ANFO could approximately double, raising ble, however, that some manufacturers of
the cost of operations as much as percent. black or smokeless powder might prefer to
Such an increase may well require a higher take some product lines off the market, so as
grade cutoff point for ore, resultin g in a signifi- to incur these startup costs for only a portion
cant decrease in the effective reserves of eco- of their existing product line. It is also possible,
nomically recoverable copper at that site. though perhaps less Iikely, that a manufacturer
might choose to halt all production for the
handloader market rather than be involved in
tagging such powders. If this should occur,
Possible Removal of Some handloaders would find their existing choice
Gun powders From the Market among powders reduced; this reduction in
choice would be a “cost” to handloaders,
The initiation of a tagging program involves though not one which can be expressed in dol-
startup costs to the manufacturer, which this lars.


In this section of the report, cost estimates lowing that are set forth the various aspects of
are established for implementing a baseline cost uncertainty in the study and a cost-sensi-
taggant program. This development of cost is tivity analysis of key uncertainty cost drivers
an accumulation of total program cost ele- or parameters intrinsic to the taggant program.
ments developed in prior sections of the re-
port. The program cost elements include:

Identification Taggant Program

. identification taggant material costs; Material Costs
● detection taggant material costs;
manufacturing level costs; Table 41 shows the buildup of identification
● distribution system costs; and taggant material costs. The calculations, which
● public overhead costs: are self-explanatory, are based on the program
— sensor-related production, units (weight, feet, caps) set forth in the earlier
— sensor development, section on “Taggant Material Costs, ” A price
— other taggant program development for polyethylene encapsulated tags of $55/lb is
costs, and utilized with the concentration noted. The to-
— BATF annual administration and trac- tal annual cost for this baseline condition is
ing activity, $11,200,000.

Subsequent to the buildup of the total base-

line taggant program costs, a series of alter- Detection Taggant Program
native implementation levels are examined for
Material Costs
their cost impact. Costs are estimated for a
total taggant program and for separate identi- Table 42 sets forth the buildup of detection
fication and detection taggant programs. Fol- taggant program material costs. The calcu la-
124 ● Taggants in Explosives

Table 41 .–Identification Taggant Material Annual Costs, Baseline Program

Annual cost
for taggant
Estimated Explosive Encapsulated/ Taggant Increased cost materials
annual average Taggant unencapsulated cost per per unit of Increase in (dollars in
production unit cost concentration (total pounds) pound explosives explosive cost thousands)
Cap-sensitive packaged
high explosives. 325 million lb $0.5011b o 05% Encapsulated $55 2.75u 5.5% $8,900
Cast boosters, ., , . 6 million lb $1 50/lb 0.1% Encapsulated 122 12.2C 8.1 % 732
pellets (6,000)
Smokeless powders 5 million lb $6 00/lb 0,05% Encapsulated 55 2.75$ 0 46% 137
Black powder, 400,000 lb $9.00/ lb 0.05% Encapsulated 55 2.75c o 30% 11
Detonating cord ., 500 million ft 5$/ft 5 taggants Encapsulated 25/batch 0.05$ 1 %0 250
per inch (160)
Blasting caps . 84 million units 50c each 50 mg Encapsulated 120 1.32$ ea. 2 64% 1,100
(9,240) ( +46)

Total program . . . , . ., .,, . ., . . .,, ,,, ,, . . . ,., $11,200

‘Allowance for cap materials
SOURCE Office of Technology Assessment

Table 42.–Detection Taggant Material Annual Costs

Taggant cost per unit Expected

Estimated annual Detection taggant level Detection taggant explosives total annual costs
Explosive category production concentration required, pounds (@$40 /lb taggant) (dollars in thousands)
Cap-sensitive packaged high
e x p l o s i v e s 325 million lb 0.025% by weight 87,500 1$ $3,250
Cast boosters 6 million lb 0.025% by weight 1,500 lc 60
Smokeless powders. 5 million lb 0.025% by weight 1,250 I 50
B l a c k p o w d e r 400,000 lb 0.025% by weight 100 1$ 4
D e t o n a t i n g c o r d 500 million ft 100 mg/ft 110,000 0.9C 4,500
B l a s t l n g c a p s 84 million units 200 mg per cap 36,960 1.76c 1,478
worst case set
Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $9,340
SOURCE Office of Technology Assessment

tions, which are self-explanatory, are estab- Distribution Network Program Costs
lished at the noted concentration levels and
weights, feet, and unit quantities common to The annual program cost attributable to the
the identification taggant program. At the esti- distribution network is $9,231,000. The calcula-
mated cost of $40/lb of detection taggant ma- tion, shown in table 44, is based on the quanti-
terial, the total annual program estimate is ties of explosives and distribution system in-
$9,340,000. cremental costs established in previous sec-
t ions.

Manufacturing Level Program Costs

Explosive manufacturing level program Public Overhead Program Cost
costs are delineated in table 43. The annual
Public overhead program costs are defined
cost estimate for the baseline program is
to include the folIowing cost elements:
$7,068,500. The costs are based on explosive
quantities and manufacturing incremental ● sensor-related deployment costs,
costs developed in previous sections. ● taggant program development, and
Ch. V—Taggant Cost Review ● 125

Table 43.–Manufacturing Cost Added Table 45.–Taggant Program Summary Annual

Cost-Baseline Program (millions of FY 1979 dollars)
Total program
Annual cost
cost added
Estimated annual Manufacturing (dollars m T a g g a n t m a t e r i a l s $2056
Explosive category producilon cost added/unit thousands) Identification taggants(11 22)
Detector taggants (9,34)
Cap-senstive packaged
high explosives ., 325 million lb 1 . 0 3a S e n s o r - r e l a t e d c o s t s a 683
Explosives manufacturers’ added costs 7.07
Boosters . . 6 million lb 7.72 463
400,000 lb Distributors’ costs 923
Black powder 2.57c 10
Smokeless powder 5 million lb Government costs. ., 1,68
2.57c 128
Detonating cord ., 500 million ft 470 Administration and tracing
Taggant program development
Blasting caps .84 million units 3.15C 2,650
Increased Investigative costs . 0
Total ., ., ., $7,068
Total baseline program annual cost $4537
Baseline condition
SOURCE Off Ice of Technology Assessment
a Assumed 500 units 90.percent IMS and 10 percent MS
SOURCE Off Ice of Technology Assessment

Table 44.–Distribution System Cost Added taggant materials (identification and detec-
tion), manufacturer-added cost, distributor-
Total program
distribution added costs, and public overhead (sensors, tag-
system gant development, and BATF administration).
Distribution cost added
Estimated annual system cost (dollars in
Explosive category production added/unit thousands)

Cap-sensitive packaged Program Cost Versus

high explosives ., 325 million lb 1. 19$a $3,869 Implementation Level
Boosters . ., 6 milllon lb 5.48 328
Black powder ., 400,000 lb 7.55$ 30
Smokeless powder 5 milllon lb 6443$ 3,222 Table 46 shows the major cost elements of
Detonating cord ., 500 million f t 015$ 750 the taggant program as a function of imple-
Blasting caps .84 million units 1.23c 1.033 mentation level. The low-level program would
Total ., ., ., $9,232 use a unique identification taggant for each
Baseline conditions manufacturer, type of product, and year of
SOURCE Office of Technology Assessment

Table 46. --Taggant Program Summary Annual Cost Versus

● BATF administrative costs, including trac- Implementation Level (millions of FY 1979 dollars per year)
ing activity. Low case Baseline High case
Summary cost elements program program program
The annual sensor program cost is $6.83 mil-
Taggant materials
lion for the baseline case in which 1,500 units Identification taggants $ 5.61a $11.22 $1122
are deployed in an assumed mix of 90-percent Detection taggants 934 934 934
IMS and 10-percent MS sensor types. As in- Explosive manufacturers’ added cost 5.26 b 7. 07C 19.41d
Distribution system added cost 5.02 e 9.23 16.55 f
dicated earlier, the annual BATF a d m i n i s t r a - P u b l i c o v e r h e a d 5329 8. 51h 24.5 ‘
tion cost is approximately $0.53 million, while Total program annual cost $3055 $4537 $810
the taggant program development annual cost (less ANFO)
is estimated at $1.15 million, for a total of ANFO $187.0
$8.51 million. $268.0

aOTA estimate of slmpllf!ed code based on halwng the baseline es~lmafe

bPlant/year tagging level
Taggant Program Baseline cDate-shlff Iagglng level
dlo 00010 f 2 lagglng level for cap-sens(tlve 2000 lb for powders
Cost Estimate ‘Inc’ludes markup costs only
flncludes Increase for adlustecj markups 75 mllllon lb of powders powder record keeping @
The total estimated cost for the baseline tag- gBased on 800 sensors
‘Based on 1 500 sensors
gant program is $45.37 million per year. The IBased on 5 000 sensors
IBased on 34 bllllon lb of ANFO lagged annually ID lag @ I 2C/lb of ANFO detection lag C @
calculation of this estimate is shown in table O 5c/lb of ANFO manufacturing @ 2c/lb of ANFO and recordkeepmg @I lc/lb of ANFO
45. It includes the estimated cost impact of SOURCE Office of Technology Assessment
126 ● Taggants in Explosives

manufacture. A total of 800 detection sensors for use in ANFO would be tagged, but not fer-
would be deployed, one f o r p a s s e n g e r s a n d tilizer-grade ammonium nitrate. Approximate-
one for baggage at each airport location cur- ly 5,000 detection taggant sensors would be
rently deploying magnetometers and hand deployed at every major transportation facil-
baggage X-ray units. Cap-sensitive high explo- ity, controlled-access utility, Government fa-
sives, detonators, boosters, detonating cord, cility, and other potential high-value targets
and smokeless and black powders would be such as campus computer locations. Portable
tagged with both identification and detection units would be routinely available to police
taggants. Blasting agents would not be directly bomb squads. The taggant level and types of
tagged. explosives to be tagged in the high-level pro-
gram correspond to a strict interpretation of
The baseline program would tag the same
S.333, as propounded by IME.
materials as the low-level program, but would
use a unique identification taggant for each
shift of each product — analogous to the cur- Program Cost of Separate Identification
rent date-shift code marking on the exterior of and Detection Taggant Programs
explosives. Traceability to the lists of last legal
The above discussion has been for a pro-
purchasers would be maintained, as the tag-
gram that includes both identification and de-
gant would contain all the information needed
tection taggants. Interest has been expressed
for a BATF trace (date, shift, product, and
in the cost of each program separately; the
size). Approximately 1,500 detection taggant
total cost and breakouts by cost elements are
sensors would be deployed at airports and ma-
discussed for each of the three implementa-
jor controlled-access facilities such as power-
tion levels. For the baseline set of conditions,
plants, refineries, and Government buildings.
the cost breakout is set forth in table 47. These
Major police bomb squads would operate port-
costs are, in summary:
able units.
I dentification taggant program $248 mllllon
This baseline program differs from the pro-
Detection taggant program $254 milllon
gram proposed by the BATF/Aerospace Corp. Total combined program $4537 mllllon
team in two respects. The most important is
that a full shift of the same product (a differ- Table 47.–identification Taggant and Detection Taggant
ent size would be treated as a different prod- Program Cost Comparisons–Baseline Case
uct) would be tagged with the same taggant, (millions of dollars per year)
rather than an arbitrary 10,000 to 20,000 lb. Identification Detection Baseline
The practical utility result of that change is taggant taggant combined
Program cost elements program program program
that a longer I ist of last legal purchasers would
Taggant materials ., ,$1122 $9.34 $20.56
be produced by a trace, at least for those lines
Sensor-related costs – 6.83’ 6.83’
that make more than 10,000 to 20,000 lb of a Manufacturers’ cost ., 6.0b 94 707
product in a single shift. The second difference Markup
concerns rework. It has been assumed that a
Labor and
special taggant containing a “composite tooling
code” will be added to material containing Distribution system cost, 6.66b 2.57 9.23
more than 10-percent cross-contain inat ion; Markup
such a taggant would indicate that other codes Labor and
in the explosive were contaminants and could tooling
Government cost
be ignored. 13C
Administration and tracing 53 53
Taggant program development ,34 81 1.15
The high-level program would uniquely tag
Total ., ., ., ., .$24.76 $25.44 $4537
each 1(),()0()-lb batch of explosive and each
2,000-Ib batch of gunpowder. All explosive a
For 1,500 sensors
b LesS markup on delecllon Iaggafll
materials, including blasting agents, would be CASSUMed 25 percent of combined Program
directly tagged. Ammonium nitrate fabricated SOURCE Off Ice of Technology Assessment
Ch. V— Taggant Cost Review ● 127

As one can note, the sum of individual pro- 2.5 m i l l i o n l b o f b l a c k p o w d e r w o u l d b e

grams is greater than the total combined pro- tagged. All but 400,000 lb of the black powder
gram. This follows from the fact that each of is used as a raw material input to other manu-
the programs share certain labor and capital factured items, such as fuzes, however, and so
resources in the combined program and each would not be tagged.
opt ion bears the tota I cost for these resources
For a taggant program with the scope as-
if only one of the programs would be imple-
sumed by IME, OTA estimates the cost would
mented. Shared resources in the combined
be $214 million, not $700 million. The major
baseline program are approximately $5 mil-
reasons for this difference are: I ME assumed
lion/year, The detection taggant program is
material cost for the identification taggants of
directly sensitive to the number of deployed
$200/lb (versus the OTA estimate of $55/lb), the
sensors; variation in this would affect the cost
inclusion of a Iibrary maintenance fee of $100/-
differentials significantly.
year per unique taggant (this fee would not be
Similar cost breakdowns were calculated for charged), and a concentration level of 0.05 per-
the separate identification and detector tag- cent for unencapsulated taggants versus the
gant programs at the low and high implemen- BAT F/Aerospace suggested level of 0.025 per-
tation levels; these separate costs for the three cent (equivalent to a 0.05-percent concentra-
implementation levels are summarized in table tion level for encapsulated taggants). As in-
48. dicated previously, the IME figures for the
material and library maintenance costs reflect
Table 48.–Summary Program Costs Versus a 3M quoted cost for taggants produced in a
Level of Implementation
pilot program.
Total combined
Identification Detection program a
Table 49 depicts the various cost elements
L o w $ 1493 $2192 $3055
for an identification taggant program that in-
Baseline 2476 2544 4537 cludes blasting agents. The three columns
H i g h 21454 6526 2688 show, respectively, the element cost estimates
ac~~tJ(”~d ~rOgram CoSIS are less—than the sum of the Indlwdual Programs because of shared made by I ME, the corresponding costs under
l a b o r tooling admlnlsfrallon etc
the same assumptions made by OTA, and the
SOURCE Otftceof Technology Assessment
actual cost elements, as estimated by OTA. It
must be clearly understood that these cost esti-

Comparison of OTA Cost Estimates

With I ME and Aerospace Corp.
Table 49.–Comparison of the Estimates for ID Tags
Estimates (millions of dollars per year)
In testimony before the Senate Governmen- OTA estimates OTA estimates
tal Affairs Committee, IME has estimated that I ME cost using I ME using OTA
the cost of the identification taggant program Cost elements estimate assumptions a assumtions
ID tag materials–non-ANFO $ 525 $ 1038 $ 112
would be on the order of $700 miIIion/year.
ID tag materials–ANFO 3400 68.0 680
That estimates includes the cost for the tag- Manufacturers’ costs–
gant materials, library maintenance fees, and n o n - A N F O — 172 18.47
record keeping costs. The estimate did not in- Manufacturing cost–ANFO
and recordkeeping – 1020 102,0
clude public overhead cost, manufacturing Distribution system cost – 80 13,98
added costs, costs through the distribute Public overhead – 87 87
chain, and markup. In addition, the I ME esti- Record keeping costs 195 in mfgr & in mfgr &
distribution distribution
mates for the quantity of cap-sensitive explo- Code reservation 291 1 — —
sives produced is lower than the OTA estimate T o t a l $703,1 $206.45 $21454
by 50 million lb, IME does not include the ef-
dAssumpllons 275 mllllon lb of cap senslhve p a c k a g e d e x p l o s i v e s 2 5 mllllon lb of black
fects of tagging 5 million lb of smokeless pow- powder smokeless powder not Included
der and assumes that the total production of SOURCE Ott[ceol Technology Assessment
128 ● Taggants in Explosives

mates are for the identification tagging pro- In s u m m a r y , t h e q u e s t i o n a s t o w h i c h c o s t

gram for the high implementation level. estimate is “correct,” that by Aerospace or
that by I M E , c a n n o t b e s i m p l y a n s w e r e d , a s
The Aerospace Corp. cost estimate of ap-
they are giving estimates for different levels of
proximately $48 million/year was for a differ- implemental ion. Both estimates contain val-
ent program —one in which ANFO and other ues for cost elements that are not currently
blasting agents are not directly tagged. As relevant, and these are clearly indicated in
noted above, the program for which the Aero- tables 49 and 50.
space Corp. cost estimate was given is quite
similar to the OTA identified baseline pro-
Who Bears the Cost of
gram, differing only in the size of the unique
taggant batch and in some assumptions on re-
a Taggant Program?
work material. For the baseline program set of conditions,
A summary of major differences between an analysis was made to determine which of
the Aerospace Corp. assumptions and the OTA the various segments affected would bear the
baseline case assumptions is as follows: costs of the taggant program. Table 51 shows
the cost breakout. Sensor-related costs would
Aerospace OTA reflect the perceived utilization of sensors at
assumptions assumptions
Detonating cord. ... 12,000,000 500,000,000 f t
airports for screening of personnel, hand-car-
Number of sensors ried baggage, and checked baggage. For the
deployed ., 5,000 1,500 baseline case of 1,500 sensors, 1,200 or 80 per-
Increased investigating cent are assumed to be employed at airports,
costs. .$5.4 million None
with 300 or 20 percent in Government build-
M a r k u p . , No Yes
ID tag material cost,
ings, courthouses, transportation centers, and
encapsulated . $50/lb tag $55/lb tag police bomb squads.
Detection tag material The users of explosives absorb the primary
cost . . . . . . $65/lb tag $40/lb tag impact of the program, assuming that all costs
Table 50 depicts the various cost elements associated with the taggants (material, manu-
for an identification and detection taggant facturing, and distribution), are passed on to
program that does not include blasting agents. the various classes of users examined. The ex-
The columns represent, respectively, the cost tent to which these costs will ultimately im-
estimates made by the Aerospace Corp. and pact consumers of goods produced by the ex-
the cost elements as estimated by OTA. plosive users is uncertain.
Public overhead costs of administration and
taggant program development are borne di-
Table 50.–Comparison of OTA and rectly by the taxpayer who would also bear
Aerospace Program (Option 2) Estimate some portion of the detection taggant sensor
deployment in the baseline case.
Cost elements estimates OTA estimate
Table 51 .–Taggant Program Cost Impact by Who Will Bear the
ID tag materials ., ., ., $8.58 $11.22 Cost (millions of dollars by impact segments)
Detection tag materials 7.86 9.34
L a b o r 2.05 —b
—b Users of Airline
Retooling. ., ., ., ., 1.65 Baseline program costs explosives Taxpayers users Total
Total instrumentation cost 22.50 6.83
Increased investigative costs . . . . 5.40 -o- Taggant materials . . $20.56 $20.56
Explosives manufacturing cost ., (c) 7.07 Sensor-related costs ., – $1–3 $5-53 6.83
Distribution system cost ., ., (c) 9.23 Explosive manufacturers’
Government costs, ., — 1.68 c o s t s 7,07 – – 7.07
Ditribution system costs. 9.23 – – 9.23
Total ., ... . . . ., $48.04 $4537 Public overhead –
— 1,68 – 1.68
aFr~m Explosives Tag9m9 Inflaflon Impact Analysls Total ., $36.86
Aerospace Corp April 1979 $2.98 $5.53 $45.37
blncluded In exploswes manulaclunng cost Percent, . 81 ,2% 6,6% 12 2%40
Clncluded In labor cost
SOURCE Office of Technology Assessment SOURCE Off Ice of Technology Assessment
Ch. V—Taggant Cost Review ● 129


I n the preceding narrative description of the Cost Sensitivity Analysis
taggant program cost analysis, OTA has set
forth the basis for estimating the various fac- The method used here essentially sets forth
tors in the total program cost equation. The the cost impact changes that occur due to vari-
relative certainty (or precision) of the esti- ations in cost-driving variables of interest. The
mates has been addressed to varying degrees. cost-impact variations from an established or
In this section, OTA specifically summarizes hypothesized baseline case is the traditional
concerns regarding the precision of the esti- method taken. Cost element changes in abso-
mates and the related implications for: 1 ) the lute or percentage terms are set forth and the
reasonableness of the estimates and 2) the impact on total program cost is noted. Since
prospects for cost-estimate growth or stability. the taggant program is in the early stages of
development, the factors in the total cost
equation need to be examined to determine
A precise evaluation of the costs of a tag- the potential ranges of variance from an estab-
gant program is not possible due to the current lished baseline. Table 52 includes a relatively
state of development of the taggants and sen- comprehensive Iist of elements that have an in-
sors and the uncertainties in how a taggant fluence on the program cost estimate. These
program would be implemented. Pilot testing include the various factors (both cost and re-
has been conducted between the identifica- lated requirements) for:
tion taggants and several of the types of ex-
taggant materials;
plosive materials proposed to be tagged (cap-
the manufacturing and distribution sys-
sensitive packaged explosives, boosters, and
black powder), testing is underway on smoke-
public overhead (sensors, administration,
less powder, and no pilot tests have been con-
taggant program development); and
ducted for detonating cord or blasting caps.
programmatic considerations.
Three candidate sensors are being evaluated,
but no system has progressed past the labora-
Taggant Materials
tory stage. Various implementation levels are
possible, each of which directly affects costs. IDENTIFICATION TAGGANTS
Examples of critical implementation decisions
Various factors can further influence the
include: which explosives will be tagged, what
cost of identification taggant material. The
would constitute a unique “batch” with a
best estimate from 3M is based on their recent
unique identification species, and how many
Ieadtime study, $75/lb of unencapsulated tag-
of which type of detection sensors would be
gants in 2.5- to 5-lb lots. This value is based on
tagging 600 million lb of explosives per year,
requiring a guarantee of manufacturing of
Several forms of cost uncertainty analysis 150,000 lb of taggants per year for a minimum
are possible. Given a baseline case, one can ex- of 2 years. Values utilized in the OTA study are
amine the cost effects of changes in individual based on lower quantities of encapsulated tag-
cost factors and note the perturbation on total gants. 3M has made their best estimate of this
program cost in a deterministic manner. This effect on cost; however, more detailed study
method is employed in the following section in would be required by them to provide an
order to highlight the primary cost drivers in equivalent confidence to the current $75/lb
the taggant program. Another method treats quotation. Encapsulated taggants estimates
costs in a probabilistic manner. Additional provided for this study are targeted at $55/lb of
data would be required to implement this pro- polyethylene-coated taggants for 90,000 lb of
cedure. taggants per year. Additional study of opaque-
130 ● Taggants in Explosives

Table 52.–Elements of Cost Uncertainty group to permit the cost uncertainty analysis
/dentiiication taggant material of the taggant program. The ultimate effect of
● Taggant cost dollars per pound the worst case condition would be to increase
–Encapsulation cost-opaque capsule identification taggant direct costs of materials
–Yield from encapsulation process
–Cost IS an estimate, not a contracted value
by 27 percent.
–Monopoly issue
● Taggant concentration level If one were to implement unencapsulated
● Quantity of explosives to be tagged
taggants, as was studied in some detail in the
–Cap-sensitive packaged explosive
–ANFO and other blasting agents Ieadtime study, the ultimate effect would be a
● Taggant waste reduction in the baseline program estimate
Detection taggant material from $11.2 million to $9.6 million, a reduction
● Molecule prices of approximately 14 percent.
● Encapsulation cost
● Concentration levels Other areas of cost uncertainty are:
● Quantity of explosives to be tagged

Sensor cost ● Monopoly issue–this is discussed in the

● Quantity of sensors to be deployed second section of this chapter.
● What type sensors WiII be successfully developed?

● What will be the mix of deployed sensors?

● Taggant concentration levels –the surviv-
● Development cost uncertainty
ability and recovery tests so far con-
● False alarm rate

● Production price uncertainty ducted have been at one concentration

Explosive manufacturers added cost
level, as have the safety tests. The tests
● Record keeping costs (particularly smokeless powders) have identified areas where the taggants
● Tooling and labor, etc . for explosive categories not pilot tested survive and areas where individual tag-
(powders, detonating cord, blasting caps)
● Batch size
gants do not survive (with a substantial
–Productivity grey area). Nonsurvival seems to be pri-
–Waste marily a function of the thermal or phys-
● Taggant inventory costs

● Markup and degree to which costs are passed on

ical decomposition of the taggant materi-
als, which would be essentially unaffected
Distribution costs
Recordkeeping by concentration level. If concentration
● Storage levels were changed, the cost of material
● Markup levels
would increase almost linearly (see
Cost of investigation below).
●Cost penalty v. cost savings
● Quantity of explosives to be tagged–
Government regulation and administration
greater quantities (over 325 million lb of
Implementation and programatic
● Level of Implementation cap-sensitive) of tagged explosive would
● Stand alone program costs decrease cost per pound of taggant mate-
–Identification taggant program rial; however, total program increases
–Detection taggant program
would not increase I i nearly.
SOURCE Office of Technology Assessment ● ANFO tagging—see the section on “Tag-
type encapsulation is r e q u red in order to re- gant Program Cost Synthesis” for esti-
fine the $55/lb estimate. 3M assessment of the mated effects. It is probable that if ANFO
worst case is $70/lb, to account for the uncer- were to be tagged, a taggant with addi-
tainty i n : tional layers would require development,
to permit the larger number of codes re-
● encapsulation and encapsulant ion process
quired by the large quantities of ANFO
yield (further research is required to de-
and other blasting agents.
finitize these parameters), and
● Taggant waste— the degree of taggant
● ultimate contractual conditions specified
waste (if any) in a production environment
(the only basis for “precise” quotations).
is unknown; this factor, which is not con-
3M believes that the worst case estimate is sidered significant, would tend to increase
highly unlikely and was provided to the study taggant material cost estimates.
Ch. V—Taggant Cost Review . 131

Summary baseline program cost sensitivity Concentration levels are another issue. Current
to variations in identification taggant material expectations are that 0.025-percent concentra-
costs or concentration levels is depicted in tions are adequate. Further development test-
figure 20. Cost-impact changes include the ef- ing is required in order to definitive this param-
fect of markup at the manufacturing level and eter. Baseline program cost sensitivity due to a
throughout the distribution network. range of variation in detection taggant materi-
al costs or concentration levels is set forth in
figure 21. Cost variations include the succes-
Figure 20.—Baseline Program Cost Sensitivity sion of markups that are estimated at the man-
Impact With Changes in Identification Taggant, ufacturing level and throughout the distribu-
Material Cost, and Concentration Level
tion network. It should be noted that the con-
Total program centration levels for identification and detec-
cost in millions of dollars tion tagging of detonating cord are inconsist-
ent, with a very small concentration of identifi-
cation taggants assumed and a very high con-
centration of detection taggants.


Taggant program cost estimates at the man-
ufacturing and distribution levels vary in their
degree of precision and are highly influenced
50 - by various assumptions that are required due
to the lack of substantive empirical data. Con-
fidence is relatively higher in the estimates
program Percent
change in
1 45 1 ID taggant Figure 21 .—Baseline Program Cost Sensitivity
2 500\o 100 ”/0 material Impact With Changes in Detection Taggant,
cost or Material Cost, and Concentration Level
Total program
cost in millions of dollars

Increase ● Decrease

SOURCE: Off Ice of Technology Assessment


Detection taggant materials are still in the

exploratory stage of development, with five
candidate molecules currently under consid-
eration. As shown in our discussion in the sec- 50;
ond section, estimates based on recent budget-
ary and pricing quotations vary depending on
the molecule and the spread in the submitted Percent
change in
cost estimates. The average value utilized in 45 _ detection
this study is $40/lb. The range of estimates is 50% 1000% taggant
from $22 to $58/Ib. The uncertainty in program material
dollar terms is as follows: cost or
estimate .,
million ❑ increase ■ Decrease
Worst case estimate . . . $13.54 million SOURCE: Office of Technology Assessment.
132 . Taggants in Explosives

where pilot testing has been accomplished Taggant inventory costs, which were as-
(e.g., cap-sensitive packaged explosives). The sessed as part of the manufacturers’ costs,
degree to which costs will be passed on, with were estimated at 10-percent interest for a tag-
associated markups through the distribution gant inventory supply of one-half year. Varia-
network, to the user of explosives is another tions from this assumption would have rela-
area of uncertainty. tively minor influence over total program cost
effects. Markup costs were estimated at 10
As a result of the pilot test program, reason-
percent at the manufacturing level and 25 per-
able data is available for the analysis of the
cent for the distribution network for explo-
cost impact of adding taggants to the manu-
sives, while 80-percent markup was utiIized for
facture of cap-sensitive high explosives, at
the black and smokeless powders for the distri-
least for those companies that participated in
bution network, based on estimated inputs
the program. No similar data is available,
from a manufacturer. Uncertainty exists in the
however, on the manufacturing impact of the
degree to which taggant program costs will be
other types of explosive materials that might
passed on to explosive users, since ultimately
be tagged. Only gross estimates have been
these markups would be determined in the
made for recordkeeping and storage costs.
Federal requirements for date-shift code rec-
ordkeeping currently pertain to cap-sensitive PUBLIC OVERHEAD
packaged explosives, boosters, black powder, Sensor-related costs.— Considerable uncer-
detonating cord, and blasting caps. Smokeless tainty exists in estimates of the sensor program
powders, currently exempt from the require- cost. These relate to:
ment, represent the largest uncertainty in rec- ● what type of sensors will be successfully
ordkeeping costs. OTA has treated this cost
element parametricalIy with the level of imple- ● what will be the mix of deployed sensors?
mentation analysis. For the three cases stud- ● how many will be deployed?
ied, the following cost estimates were utilized: ● development cost uncertainty,
● production price uncertainty, and
Low program no cost increase
B a s e l i n e 60.4/lb powder ● false-alarm rates.
High estimate ., ., 100$/lb powder
Table 53 delineates a set of cost possibilities
These estimates are based on preliminary as-
where sensor mix and quantity are varied. One
sessments; further refinements in the smoke-
can note the wide spread of resulting estimates
less powder recordkeeping estimate require a
given these variations in assumptions. OTA es-
data base reflecting pilot-testing experience
timated the sensor development costs of twice
and a detailed description of the distribution
the level of the Aerospace estimates to ac-
An analysis of manufacturing cost impact
for cap-sensitive packaged explosives revealed Table 53.–Annual Cost per Sensor for Various Mixes
the following cost sensitivity to program im- Total annual cost
plementation levels: (millions of dollars)
Annual cost
Manufacturers’ cost per per sensor 1,500 sensors 5,000 sensors
Tag batch size pound of explosives FY79 dollars
10,000 to 12,000 lb ., ., ., 4.O A l l C E C D $3,318 $5.0 $16.6
20,000 lb . ., 2.3 All IMS . 4,053 6.1 20.3
Shift production .. 0.6$ A l l M S 9,228 13.8 46.1
Plant year . 0.3 CECD 90% MS l0%. 3,909 586 19.5
CECD 75%; MS 25%. 4,796 72 24.0
Uncertainty in other particular explosive t y p e IMS 90%, MS 10%
c o s t elements will persist until a p a r t i c u l a r p r o - ( b a s e l i n e ) . 4,570 6.8 22.8
gram level is recommended for implementa- IMS 75%; MS 25%. 5,347 80 26.74
tion. SOURCE Off Ice of Technology Assessment
Ch. V—Taggant Cost Review ● 133

count for development program contingen- the development phase of most major hard-
cies. Production cost estimates confidence has ware system programs. This is so even for pro-
been stated by Aerospace as about * 25 per- grams where precedent-type data are available
cent. This production effect on the baseline (e.g., aircraft, missile, electronics). The taggant
case estimate would be as follows: program has no direct precedent as such and
analogous situations are limited. Historical
BaselIne (1, 500 sensors). $683 million
Low estimate $512 million
data are therefore severely limited and slowly
W o r s t c a s e $854 million evolved as pilot testing progresses. Traditional-
ly, as a program proceeds during development,
The effects of quantity and sensor mix are
new elements of costs are recognized that
more profound. Sensor costs could vary from
were poorly perceived at the onset of develop-
$5 million to $13.8 million (see table 53) for the
ment” in addition, program directions change
baseline quantity of 1,500 sensors depending
as ergineering and scientific problems are un-
on the ultimate mix of system deployed. Quan-
covered, resulting in scope changes and poten-
tity variations would also proportionately im-
tial for cost growth. Questions of scope, for in-
p a c t p r o g r a m c o s t s . High false-alarm rates
stance, include program implementation levels
(greater than 0.05 percent) in fielded sensors
which have been addressed in the cost synthe-
would have tangible cost impacts in the cost of
sis section. As noted, costs estimates can vary
operations and in creating ill will.
by significant degrees depending on the pro-
gram specification. Related to the scope issue
Programmatic considerations.— The overrid- are the individual identification and detection
ing uncertainty in the cost of the taggants pro- taggant programs as separate entities. Pursuing
gram stems from the nature of the present ear- either one of these objectives rather than pro-
ly phase of program development. Program ceeding jointly would have a significant im-
cost uncertainty is a profound problem during pact on cost.


The taggant program cost estimates are scope changes; cost-es i mating error contrib-
based on a limited empirical data base and utes to a lesser degree.
various analyses and assumptions. This situa-
ution is caused by the relatively early stage of Further pilot testing and sensor deveiop-
the development program, the limited number ment efforts are requi~ed in order to provide
of pilot tests conducted to date, and the lim- refined designs and requirements data for both
ited sample of organizations surveyed (manu- manufacturing processes (e. g., detonating cord
facturers, distributors, and users of explosives). and blasting caps) and sensors, which are nec-
The limitations in the data base and resultant essary for redefining the cost estimates. Until
assumptions have been underscored within the this progress is made, further refinements in
cost analysis section. Where assumptions were cost-estimate precision are not possible.
made, OTA has taken a conservative position
in order to provide a reasonable cost estimate Additional survey samples of manufactur-
for the program options. This is important be- ers, distributors, and explosive users would
cause cost growth normally ensues in typical provide higher confidence in certain of the
developmental efforts. Cost growth is predom- cost-element estimates and other cost impact
inately affected by redesign and program areas.
134 ● Taggants in Explosives


Additional cost analysis research would im- Other special stud es and analyses would
prove the ability to determine m o r e a c c u r a t e l y provide further value to the understanding of
and at a finer level of resolution the cost im- taggant program cos impact. Among these
pact of the taggant program. This research ef- are:
fort could take a number of avenues including: !!
Q cost/uncertainty probabiIity analysis; i
● development of a cost model, ● price elasticity for black powder, smoke- [
● development of an economic model, less powders, and cap-sensitive high ex- ,i
● application of design-to-cost principles plosives, etc.;
for the sensor development, and ● assessment of manufacturers’ “front end”

● special studies and analysis. costs and the related burden; and
● amplified cost and economic impact sur-
T h e OTA study effort on the costs of the tag-
veys of manufacturers, distributors, and
gant program was limited in time and re-
users of explosives.
sources. Various insights gained during this re-
search indicate that further research in the It must be clearly understood, however, that
above areas would contribute significantly to resolution of the basic program issues, such as
a better understanding of the multitude of cost level of implementation, as well as resolution
and economic tradeoffs and effects which of technical efficacy, safety, and utility is nec-
could guide the development of a taggant sys- essary before it makes sense to attempt a more
tem. The model developments (cost and eco- detailed cost analysis. The work reported in
nomic) would further this goal. AppIications of this chapter clearly indicates the order of mag-
formal design-to-cost principles to the devel- nitude of the cost impact that decisions con-
opment of sensors will further permit the pro- cerning taggant legislation would have on the
duction and implementation of cost-effective manufacturers, distributors, and users of ex-
systems. plosives and gun powders.
Chapter VI

Page Page
introduction . . . . . 0 . . . . . . . . . , . . . , . . . . . , . . 137 55. Explosive Bombing Incident Trends,
Problem Characterization . .................138 1972-78. . . . . . . . . . . . . . . . . . . . . . . . .. ...140
The Bombing Threat. . ..................139 56. identified Explosive Fillers Used in
Explosives Used in Bombs . ..............141 Bombs . . . .; . . . . . . . . . . . .............141
Types of Targets Bombed. . ..............142 57 Bombing Casualties and Damage in 1978
Characteristics of Criminal Bombers . ......142 by Type of Bomb . ...............,..,.142
Terrorists ... , ... , . . . . . . . . . . . . . . . . , .143 58 Bombings by Specific Targets for
Common Criminals. . .................145 1977-78 . . . . . . . . . . . . . . . . . . . . . . . . . ...143
Mentally Disturbed . .................145 59 Percent of Bomber Targets That WouldBe
Vandals and Experimenters . ...........145 Protected by a Dectection Sensor . .......143
Sources of Explosives. . .................146 60. Attributes of Criminal Bomber Groups. ....144
Current Security Measures. . .............148 61. Estimated Number of Significant
Sources of Explosives. . ...............148 Bombings by Group of Perpetrators. .., ...146
Potential Targets . ................,..149 62. Stolen and Recovered Explosive
Current Antibomber Procedures . .........154 Summary . . . . . . . . . . . . . . . . . . . . . ..., .,147
Discussion of TaggantlJtiiity. .. ..............155 63. Explosives Thefts by Method of Entry–
Deterrence. . . . . ......................155 Number of lncident sand Percentages
Bomb Detection–Target Protection . . . . . . .157 for 1977-78 . . . . . . . ...................148
Bomber Apprehension . ...........,.....159 64. Explosions Aboard U.S. Aircraft . .........150
Intelligence Concerning Criminal Bomber 65. Location of Explosions Aboard Aircraft,
Activities . . . . . . . . . . . . . . . . . . . . ....161 1949-76. . . . . . . . . . . . . . . . . . . . . . . . .. ...150
Prosecution of Criminal Bombers . ........162 66. Explosions and Devices Found at U.S.
Taggant Utility by Typeof Perpetrator .., ..163 Airports,1972-75 . ...................,151
Utility of Taggants to Update the Taggant 67 Results of Civil Aviation Security Program
Program . . . . . . . . . . . . . . . . . . . ., ,, , .163 Passenger Screening. . . . . . . . . . . . . . . . . . . 153
Nonbomber Control Utility of Taggants, ...,164 68 Premature Detonation Statistics . . . . . . . . . 155
PossIble BomberCountermeasures in Response 69 Commercial Airliner Hijacking Statistics
to a TaggantProgram . ...............,..164 by Year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
Fabrication of Homemade Explosives. .. ...167 70. Possible Perpetrator Response
Use of lncendiary Bombs. . ..............167 Countermeasures to Taggant Program. . . . . 165
Use of Blasting Agents. . ................168
Theft of Explosives. . ...................168
Illegal Sources. . .....................,169
Use of Explosives Manufactured Before a
Taggant Requirement. . ........,....169
Detection Taggant Seal . ................169 FIGURES
“Spooking’’of DetectionTaggant Sensors. ..l7O
Shift to Other Unlawful Activity . .........170 Page
Summary . ...........................170 22. Annual Bombing Statistics, 1972-77 . . . . . . . 141
Foreign Experience Controlof80mbers .. ....171 23. Schematic Airport Security System . . . . . . . 152
24 Passenger/Hand-Baggage Screening
TABLES Station . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
25 Schematic lllustration of Identification
Page Taggant Utility in Criminal Investigation. . . 160
54. Minimum Bombing Incidents Statistics 26 Size Comparison of the 3M Identification
Summary . ..........................139 Taggant and Some Smokeless Powders. , . . 166
Chapter VI

Bombings are a particularly heinous crime as they are normally indiscriminate
in their choice of victims, often involve innocent people, and have the potential for
producing large numbers of casualties and high property damage. Bombings are at-
tractive to the perpetrator as bombs can be placed at the bomber’s convenience and
set to detonate at a time when the bomber is elsewhere. Bombings are a quite spec-
tacular crime, easily drawing public attention when that is the perpetrator’s pur-

Bombings are particularly difficult crimes for law enforcement agencies to

handIe as the bomber is not usually near the scene of the crime, the physical evi-
dence is destroyed or damaged by the detonation, and the materials necessary to
fabricate even a quite catastrophic bomb are easily obtainable.

It is the purpose of this chapter to review the utility of both identification and
detection taggants to law enforcement and security personnel. In order to assess the
utility of taggants, it is first necessary to understand the magnitude of the bomber
problem, including the types of bombers, the types of targets, the sources of explo-
sives, and current measures to control and combat bombers. This information is re-
viewed in the next section. The utility of taggants is then discussed, together with
possible responses by criminal bombers to a taggant program. The chapter con-
cludes with a short discussion of the experience of selected foreign countries in the
control of bombers.

In the analysis it is assumed that the t a g - The analysis is primarily qualitative. Data
gants have been demonstrated as safe to add exist on the numbers and types of criminal
to explosive materials; that the identification bombings which take place, but it is difficult
taggants survive the detonation of tagged ex- to analyze the data as it is not consistent from
plosives and can be recovered at the scene of one data bank to another and information re-
the crime, either directly or by laboratory trieval in any other than summary form is diffi-
separation of collected debris; and that sen- cult. Characterization of types of perpetrator,
sors exist which detect the detection taggant or of motives, is available in only a limited
vapor at a parts-per-tril I ion concentration in number of bombings; even identification of
air, with extremely few false alarms and with the explosive fiIler is not available for a sign if i-
no requirement for special maintenance or cant fraction of bombings.
skilled operators. These assumptions would
have to be verified before a taggant program No data exist that would allow a quantita-
could be implemented. tive assessment of the numbers of bombers


61-401 0 - 80 - 10
138 ● Taggants in Explosives

who would be deterred, arrested, or convicted Division, U.S. Army Development and Re-
as a result of a taggant program, or of the search Command); and
amount of property damage or casualties
● contractors (Management Sciences Asso-
which would be averted by such a program. An
analogy can be drawn between the utility of ciates (MSA) and Institute for D e f e n s e
the current date-shift information contained Analysis).
on explosive cartridge cases and the utility of A number of discussions were also h e l d , o n
identification taggants in apprehending and various subjects, with the Bureau of Alcohol,
convicting bombers, but the date-shift infor- Tobacco, and Firearms (BATF), the agency
mation utility data base is quite small. Simi- charged with explosives control.
larly, an analogy can be made between the
drop in hijackings that occurred after the intro- Unfortunately, it was not possible, given the
duction of antihijacking procedures and the time and money constraints of the OTA analy-
potential reduction to be expected in the sis, to meet with as many law enforcement per-
bombings of high-valued, controlled-access sonnel as would be desirable, particularly
buildings protected by detection sensors. Such given the large variations in types of bombers,
analogies are discussed in the text. The pri- types of targets, and local laws and procedures
mary source of data on the current bombings in the various parts of the country. To obtain a
threat, current means of combating that larger sample of expert opinion, a question-
threat, and the utility of taggants to law en- naire was sent to approximately 950 members
forcement personnel, however, comes from of the International Association of Chiefs of
the opinion of law enforcement personnel in police (l AC P), chosen at random from their
the field. directory. The IACP was chosen because of the
OTA desire to obtain input from a broad cross
In-depth discussions were held with a broad section of the law enforcement community—
cross section of law enforcement and security geographically, functionally, and by size of
personnel, including personnel from the fol- community. The results of the in-depth inter-
lowing agencies: views and questionnaire responses are inte-
grated in the discussion in this chapter. A
● domestic law enforcement and security detailed discussion of the questionnaire is
personnel. (New York City; San Mateo given as appendix B. Due to the small response
County, Calif.; Dallas-Fort Worth Airport; rate (approximately 15 percent) the sample
Summit County, Ohio; Washington, D.C.); may be biased. However, the bias is probably

toward those with knowledge of, and interest
foreign law enforcement personnel (West
in, the subject. An additional possible bias was
Germany, England, Republic of Ireland,
introduced by an error in the explanatory
material accompanyin g the survey, which in-
● Federal agencies (Federal Bureau of inves- dicated that the identification taggant trace
tigation (FBI), Federal Aviation A d m i n - would identify the last legal purchaser of the
istration (FAA), Bureau of Mines, Depart- explosives, rather than indicating that the
ment of Transportation, Corps of Engi- trace would produce a Iist of last legal pur-
neers, U.S, Army Criminal I n v e s t i g a t i o n chasers.

Approximately 3,000 incidents are reported we I I as actual explosive and incendiary bomb-
annually in the BATF Explosives Incidents ings. The BATF report contains a breakout by
Report. T h e i n c i d e n t s include accidents, target type and explosive filler used, but Iittle
threats, recovered explosives, and hoaxes, as information on the various types of perpetra-
Ch. V1—Taggant Utility Review ● 139

tors. The FBI compiles similar bombing statis- mates of property damage are made in the
tics at its National Bomb Data Center, which BATF data and the lack of updating, only the
are published quarterly and summarized annu- crudest property damage estimates can be
ally. The bombings are committed by a wide made. There was at least $10 million in direct
range of perpetrators, who differ in their skills, property damage due to explosive and incendi-
resources, motivations, and types of targets at- ary bombs in 1977, and at least $17 million in
tacked. Current security measures at most ex- 1978. Thirty-five of the thirty-eight reported
plosive manufacturers, distributors, and u s e r s deaths in 1977 and twenty of the twenty-three
are sufficient to dissuade casual outside theft, reported in 1978 were from bombings against
but cannot readily protect against thefts that vehicles, residences, and commercial estab-
are committed by or assisted by employees, or lishments. Similarly, about 80 percent of the
against a determined outside attempt to steal injuries from bombing of known targets in
explosives. Protection of some high-value po- 1977 and 70 percent in 1978 were caused by
tential targets against bomber threats is cur- bombings of those three types of targets.
rently adequate but some targets are essential-
The FBI data, as indicated above, are some-
ly unprotected against a serious bombing at-
what different, both in number of incidents re-
tempt. Finally, current law enforcement efforts
ported and in the breakout of categories. In
to control criminal bombings are not very ef-
1977, for instance, FBI data show 867 actual
fective. These topics are discussed briefly
explosive bombings and 118 attempted bomb-
ings. Similarly, the number of people reported
killed that year from both explosive and incen-
The Bombing Threat diary bombings was 22, while 162 were re-
ported injured. In 1978 there were 768 explo-
Both the FBI and BATF maintain national sive bombings and 105 attempted explosive
bombing data information centers which col- bombings. The pertinent 1977 and 1978 BATF
lect statistics on bombings and other explosive and FBI statistics are summarized in table 54.
incidents. The data are not consistent between
the two centers, however, and many bombings
Table 54.–Minimum Bombing lncidents Statistics Summary’
are not reported to either center. The format-
ting of the data, and the lack of updating pro- BATF FBI
cedures, make accurate analyses difficult. Item 1977 1978 1977 1978
Explosive bombings, number 1 , 0 3 7 b 8 9 6b 8 6 7 768
The BATF 1978 Explosives Incidents Report Undetonated explosive bombs, number 319 287 118 105
incIudes over 3,000 incidents for both 1977 and Incendiary bombings, number 339 446 248 349
1978. The incidents include accidents, threats, Unignited incendiary bombs, number 81 71 85 79
Criminal accidents, number 21 67 – –
seized and recovered explosives, and hoaxes as Property damage from bombings.
welI as actual explosive and incendiary bomb- m i l l i o n s o f d o l l a r sc d $ 10 $ 17 $ 9 $ 9
i rigs. Of these incidents, 1,377 represented ex- Injuries c... 180 185 162 135
People killed by bombings 38 23 22 18
plosive detonations, accidental detonations by .
criminals, or recovered bombs which failed to dEATF repel’fed 3 177 total Incidents [n 1977 and 3 256 In 1978 Total Incidents include ac
cldents threats seized and recovered explosives and hoaxes as well as actual explosive and (n
detonate in 1977, with 1,250 the corresponding cendlary bombings The OTA study was concerned only wllh explosive bombings
bof these 953 In 1977 and 787 In 1978 were agalnsl subslarm!drgek
number for 1978. At least 953 of these in 1977 clncludes both exp[oslve and Incendiary Domblngs OTA was unable to oblam separate f19ures tor
and 787 in 1978 represent actual detonations number of cnm!nal accidents lnjurles deaths and property damage caused by exploswe and In
cendlary bombs Incendiary bombs and bombings would not be affected by Ihe proposed laggant
of explosive bornbs against substantiaI targets
d~~l~~~~alue probably considerably higher due 10 lac~ of data file uPdates
(mailbox and open-area bombings are not in-
S O U R C E EM TF 1978 Exp/os/ves /nc{derrfs Report Ft3/ Um/rmrJ Cnrne Repor/ B o m b /7epor/
clided). 1978 See app F for a discussion ot me derivation 01 these hgures

During 1977, at least 38 people were killed An effort was made to resolve the differ-
and 180 wounded by explosive and incendiary ences in statistics compiled by FBI and BATF;
bombs, while the numbers in 1978 were 23 and according to the Explosives Enforcement Divi-
185, respectively Due to the way initial esti- sion of BATF:
140 ● Taggants in Explosives

● There is no Federal statute or law on the ● The normally higher number of incidents an-
books requiring local police officials to re- nually in BATF reports is a direct result of
port bombing incidents to either BATF or the the above.
It is of considerable interest to know
● Cooperation at the local level has led to an
whether the statistics for 1977 and 1978 are
informal procedure on the part of local po-
characteristic of the recent past, or if trends in
lice to report a bombing incident to either
criminal bombings are apparent. Table 55
BATF or FBI, who in turn will normally noti-
shows the bombing trend since 1972, from the
fy each other. (There are obviously some
FBI data. While the BATF numbers differ, the
breakdowns in this procedure).
● There is a statute giving BATF the “right of
rough trends are similar. Figure 22 shows the
trends graphically, with the total number of in-
inspect ion” at the site of any explosion;
cidents depicted in figure 22a, property dam-
therefore, whether BATF receives word of a
age in 22b, injuries in 22c, and deaths in 22d.
bombing from the local police, or whether a
The total incident numbers in figure 22a in-
local special agent reads of it in the local
clude both successful bombings and attempts;
paper, BATF can by law check it out.
s BATF requires each agent to report a l l the property damage and casualty figures may
include incendiary bombings as well as explo-
bombing incidents to its explosives data cen-
sive bombings. No long-term trend is detec-
ter in Washington, irrespective of the theo-
table from the data, although an unusually
retical importance, damage, casualties, or
high number of incidents and casualties oc-
jurisdiction since, among other uses, these
curred in 1975. This increase was primarily due
data are used by the Secret Service in ar-
to three incidents.
ranging security for the President when he is
travel ing. 1. On January 24 a bombing at the Fraunces
● There is a question of jurisdiction with refer- Tavern in New York .City killed 4 p e o p l e ,
ence to investigations. A memo of under- injured 53 others, and did extensive prop-
standing exists between BATF and the FBI. erty damage. Responsibility for the bomb-
Generally the FBI covers terrorist acts, at- ing has been claimed by FALN, the Puerto
tacks on airlines, attacks involving unions, Rican separatist terrorists.
college campus buildings, and Federal build- 2 A bomb detonated in the baggage claim
ings other than Treasury and Postal build- area at La Guardia Airport, on December
ings. BATF has primary jurisdiction over 29, killing 11 people with 70 additional
criminal bombings related to interstate com- serious injuries. No positive identification
merce, firearms violations, and Treasury of the exact type of explosives used has
buildings. Either the FBI or BATF may re- been made for this incident and no at-
spond to requests for aid from other jurisdic- tribution has been made.
tions. Conflicts are settled by mutual agree- 3 A bomb detonated at a sponge factory in
ment. Shelton, Corm., in March 1975, killing

Table 55.–Explosiv e 8 Bombing Incident Trends, 1972-78

Total actual and Total actual and
attempted ex- attempted incen- Property damage Personal
Year plosive bombings Actual Attempted diary bombings Actual Attempted (dollar value) injury Death
1972 ., 951 714 237 1,011 793 218 $ 7,992,000 176 25
1973 . . . . . . . 995 742 253 960 787 173 7,262,000 187 22
1974. , . . . . . 1,129 893 236 915 758 157 9,887,000 207 24
1975, . . . 1,326 1,088 238 748 613 135 27,004,000a 326’ 69a
1976 ..., . . . 1,040 852 188 530 405 125 11,265,000 212 50
1977, ... , 985 867 118 333 248 85 8,943,000 162 22
1978. , . . . . 873 768 105 428 349 79 9,161,000 135 18
alncludes three rnalor born~lrlg mcldents resulting In unusually high personal mjunes and dealhs and substanhal damage to ProPedY
SOURCE FBI UrJ//orrrJ Crmre Repwk t?wnb .%rnmary /978
Ch. V1—Taggant Utility Review . 141

Figure 22.—Annual Bombing Statistics, 1972-77 ticularly catastrophic bombings, can be ex-
t 1 1 I I I I I pected over the next few years. It should be
3,000 ~ noted that a single incident involving an air-
Including incendiary
2,000 - craft exploding in flight c o u l d p r o d u c e m o r e
deaths than have occurred in the United States
1,000 (1,301)
. r from bombings during this decade. Such inci-
I 1 1 I Explosive only~ 1 I dents have occurred in foreign countries and a
1972 1973 1974 1975 1976 1977 1978
near miss occurred recently in New York. On
a. Number of incidents (includes attempts) March 25, 1979, a TWA plane bound from New
30L I I I I I I -1 York to Los Angeles was delayed. A bomb
planted in the checked baggage exploded
while being transported to the aircraft on t h e
luggage truck. If the aircraft had taken off on
time the bomb might have caused the deaths
of most or all of the 166 people aboard.
1972 1973 1974 1975 1976 1977 1978

b. Property damage (includes incendiary bombings)

Explosives Used in Bombs
350f I I I d 1 I I
300 Data on the types of fillers used in bombs
.- 2 5 0 - are also not consistent between FBI and BATF
data banks. It is instructive to look at two
BATF data sources, however, as shown in table
150 -
(135) 56. The second column represents 1978 data
100 “ for the fillers identified in the field for all ex-
I 1 1 I 1 I
1972 1973 1974 1975 1976 1977 1978 plosive bombs that were detonated, bombs re-
c. Persona/ injury (includes incendiary bombings) covered undetonated, and criminal accidents.
The first column represents 1978 data for only
those fillers that were identified in the labo-
ratory from postdetonation analysis. In both
cases, black and smokeless powders and cap-
sensitive high explosives all occur with high
frequency. Table 57 shows a breakout of the
estimated number of significant bombing in-
‘ “ ~ ’ ‘ cidents, deaths, injuries, and property damage
occurring during 1978 by explosive material fil-
d. Deaths (includes incendiary bombings)
Ier. The average of the two frequencies col-
SOURCE: Drawn by OTA from FBI data umns shown in table 56 was used for the table
57 estimates. (See app. F for the derivation of
three people and injuring several others.
No attribution has been made for this inci-
Table 56.–ldentified Explosive Fillers Used in Bombs
Lab identified All identified
Using FBI and BATF data, the trend of both fillers 1978 fillers 1978 Average
total bombing incidents and catastrophic inci- Black powder . . . . . . 13% 21% 17Y0
dents was analyzed by MSA for the 5-year peri- Smokeless powder ., ., 16 19 17.5
od from 1972 through 1976. The data show no Military ., ., . . . . . . . . 2 7 4.5
Cap sensitive . . . . ., 32 30 31
significant change in incidents over that peri- Blasting agents. ., . . . . . – 1 .5
od, although 1975 and 1976 had significantly Chemicals . . . . . . –
higher injuries and deaths. In contrast to in- Others, ., . . . . ., 36 2: 28:;
ferences based on past statistics, many experts See app F for derivation of these numbers
believe a significant increase in bombings, par- SOURCE BATF data
142 ● Taggants in Explosives

Table 57.–Bombing Casulties and Damage in 1978 by Type of Bomb

Number of
bombings against Property damage
Filler material substantial targets Deaths Injuries $ millionsa
All fillers. ... . . . . . . 1,298 23 185 $17.2
Incendiary . . . . . ., 428 3 13 3.7
Black powder ., ., . . . . 148 4 19 .2
Smokeless powder, ., 152 3 23 .2
Military explosives. . . . . . . . . . . 39 0 7 —
Cap sensitive. . . . . . . 270 7 26 3.3
Other . . . . . . . . . . 3 40 2.4
Unknown . . 3 57 7,4
Total for those fillers which
would be directly tagged b, 570 14 68 37
avalue probably higher due to lack Of data uPdale
bcap-sensltlve explosives black Powder, and smokeless powder would be ta99ed
SOURCE BATF data See app F for a dertvatlon of these rlgures

these numbers. ) The table shows that a large incidents that happened at commercial facili-
percentage of the total bombings deaths and ties occurred at facilities unlikely to be pro-
casualties is caused by black powder, by tected by sensors, then 79 percent of the inci-
smokeless powder, and by cap-sensitive high dents, 89 percent of the injuries, and 94 p e r -
explosives. cent of the deaths from actual explosive and
incendiary bombings which happened in 1977
Types of Targets Bombed and 1978 occurred at places unlikely to be pro-
tected by detection taggant sensors.
The types of targets that attract criminal
bombers range from attacks on mailboxes and Data are not available that would allow sep-
outhouses by vandals and pranksters to at- aration of the explosive and incendiary bomb-
tacks on aircraft by terrorists. The targets most ings statistics. It is Iikely that a larger percent-
frequently attacked on a year-in, year-out basis age of the targets of explosive bombings would
are private residences, commercial facilities be of the type protected by a detection sensor,
(usually small operations), and vehicles. Table but probably not a large percentage.
58 is taken from the BATF 1978 Explosives inci-
dents Report. It shows the total number of ac-
tual bombings (both explosives and incendiary)
Characteristics of Criminal Bombers
for the years 1977 and 1978, the bombing Criminal bombings are committed by a wide
breakout by target type, the number killed and range of perpetrators, including both individu-
injured, and the estimated property damage, als and groups. While it is always difficult to
all by target type. The FBI data are somewhat place a heterogeneous population into well-de-
different, but show the same trends in that the fined categories with well-defined characteris-
majority of bombings, property damage, and tics, it is helpful to group criminal bombers
casualties occurs at residences, at commercial into four categories: terrorists, common crim-
facilities, and in vehicles. In table 59, these inals, mentally disturbed, and vandals and ex-
data are rearranged to explicitly show that perimenters. These groups vary greatly in moti-
most of the bombings and casualties would oc- vation, skill, training, resources, and ability to
cur at targets that are not likely to be pro- respond to a changing enforcement environ-
tected by detection sensors. It is extremely un- ment. It is also difficult to determine which
likely that such sensors would be placed in pri- group is responsible for a bombing, although
vate residences or in vehicles; most commer- “credit” is sometimes claimed, particularly by
cial establishments would also not have sen- certain terrorist groups. Of the bombings re-
sors. If the assumption is made that all of the ported in the BATF 7978 Explosives Incidents
Ch. V1Taggant Utility Review ● 143

Table 58. –Bombings by Specific Targets for 1977-78 (actual detonations or ignitions)

Total incidents No. killed No. injured Property damagea

Type target 1977 1978 1977 1978 1977 1978 1977 1978
Residential . . 352 294 17 7 66 57 $ 1,022.3 $2,982.2
C o m m e r c i a l 367 375 7 6 48 46 6,640.1 8,777.7
Airports/aircraft: . 7 l – 1 — ,2
Police facllltles/vehicle . . . . . 14 2: — — — — 5:; 70.4
Educational ., . . . . 106 97 — — 13 5 43.1 532.3
Government (local) 24 9 — 1 1 4 145.6 70.1
Government (Federal) 26 22 — — 4 1 2.4 6.6
Military installations 4 3 — — — 1 — 0.0
U t d i l i t i e s . . 51 57 l – 1 2 628.0 1,727.7
Banks . 22 18 — — — — 225.2 49.3
V e h i c l e s 216 252 11 7 24 25 363.3 2,119.4
O p e n a r e a s 36 40 1 2 8 13 .5 4.2
Mailboxes . 48 69 — — 1 2 25.8 2,1
O t h e r 90 137 — — 8 27 1,206.8 869.9
U n k n o w nb 34 2 — — 5 2 22.6 0.0
Total ., ., 1,397 1,409 38 23 180 185 $10,331.7 $17,212.1
ap(opefiy danlage figures are In thousands and are esmated
~Thls category includes those Incldenls where the type [argel was either unknown or not repofled
S O U R C E BATF T978 Exphswes Inc(derm Reporf

Table 59.–Percent of Bomber Targets That Would Be Protected by a Detection Sensor

Total bombings a Injuries Deaths

Average number of bombings of known, substantial targets . . . 1,175 150 29
Bombings of residences, vehicles. . . . . . . . . . . . . . . . . . . . . . . . 557 (47%) 86 (58%) 21 (72%)
Bombings of commercial establishments. . . . . . . . . . . . . . . . . . . . . 371 (32%) 47 (31%) (22%)
Total unlikely to have sensors . . . . . . 928 (79%) 133 (89%) 2 (940/0)
alncludes both Incendiary and exploswe bombings for 1977 and 1978
bopen f(elds and mailboxes are excluded from these data


Report, a m o t i v e w a s
e s t a b l i s h e d for only 2 3 into political, reactionary, and separatist
percent of the bombings in 1977 and only 38 groups. Political groups, such as the Weather
percent in 1978. Keeping in mind the above ca- Underground, are primarily interested in at-
veats, it is nonetheless useful to examine the tracting attention to and sympathy with their
characteristics of the various groups, which cause. For that reason they engage in spectacu-
are summarized in table 60 and briefly de- lar events, such as bombings, but generally at-
scribed below. tempt to avoid or Iimit injury and death result-
ing from their bombings. Political terrorists
often have considerable resources available to
them, due to a significant number of people
The terrorist groups active in the United who support their aim, if not necessarily their
States vary widely in ability, resources, train- means. The leadership of most of these groups
ing, and adaptability. They share the common are of above-average intelligence, and have
characteristics, however, of high motivation, either had specialized training or have studied
action as a part of a group, and a continuing extensively in terrorist activities. They are thus
involvement in catastrophic, illegal activities able to adapt to a changing environment, al-
against society. These characteristics make the though the range of responses available to
terrorist particularly dangerous to society and them may be limited by their political aims.
a particularly appropriate target for anti bomb- They may lack mechanical skilIs, however, and
ing controls. Terrorists can be roughly divided be more likely to be involved in accidental ex-
144 ● Taggants in Explosives

Table 60.–Attributes of Criminal Bomber Groups

Perpetrator and training Resources Motivation Individual or group Reaction capability Frequency
Unsophisticated . . . . . . . . L L M I M Multi
Sophisticated . . . . . . . . . . . . . . . . H M H I H Multi
Political ., ., ., . . . . . . . M-H M-H M-H G M-H Multi
Separatist . . . ., ... . . . . . M-H M H G H Multi
Reactionary . . . . . . . . . . . . . . . L L H G L-M Multi
Mentally disturbed
Disenchanted ... . . . . L L L-M I L Single
Vengeful ., . . . . . . . . . . . L L M-H I L-M Single
Pathological . . . . . . . . . . . L-M L H I L-M Varies
Vandals ., ., . . . . . . . . . . . . . . . L L L-M I L Single
Experimenter . . . . . . . . . . . . M L L-M I L-M Single

L-Low, M-Moderate, H-High, I-lndwldual, G. Group

SOURCE Ofllce ol Technology Assessment

plosions, either while fabricating bombs or not committed any bombings in New York
while placing them. Political terrorists have be- since that time. Their ability to react to a
come less visible in the United States in recent changing control environment is less than the
years. political terrorist groups, due to more limited
resources. If the goal of the separatist group is
Separatist groups, such as FALN, generally
viewed with sympathy by a large part of the
hope to gain their aim by generating a reaction
population, as is the case in Northern Ireland,
to their activities, rather than sympathy for
then the group can attract resources, attract
their aims. They are therefore generally less
recruits, and perfect skills. If, on the other
concerned with public revulsion to bombings
hand, the population is either not in sympathy
that cause substantial injury and deaths. Sep-
with the separatists or is not directly affected
aratist groups have been credited with more
by the cause of the separatists (as is the case of
than 25 percent of catastrophic bombings—
the Croatians in the United States or the South
those resulting in major property damage, in-
Moluccans in the Netherlands), then the group
juries, and deaths. The resources of domestic
will not be able to attract resources or other-
separatists vary from group to group, but are
wise grow.
generally less than for comparable groups of
political terrorists, as only a fraction of the Reactionary groups, such as the Ku Klux
population represents even potential support- Klan and the American Nazi Party, would ap-
ers. As an example, few people outside of the pear to share some of the characteristics of the
Yugoslavian exile community care whether or political terrorists, but generally do not pos-
not the Croatians achieve separation from the sess the same levels of training, motivation,
Yugoslavian federation; on the other hand, a and resources, and are not as capable of react-
group like the Weather Underground, that ing effectively to a changing control environ-
seeks to exploit discontent with the U.S. Gov- ment. They also differ in that their bombings
ernment, could seek support from a larger pop- are usually directly targeted at the individual
ulation. Separatist groups are often critically or group they intend to influence, rather than
dependent on a small cadre of leaders; loss or simply at a spectacular target. Generally, their
incapacitation of those leaders may shatter the purpose is intimidation; thus, fairly small, con-
group or considerably reduce their effec- tained bombs are used. Even when murder or
tiveness. As an example, FALN in New York injury is desired, the results are usually con-
lost their bombmaker over a year ago and have fined to the directly targeted individual. While
Ch. V/— Taggant Utility Review ● 145

the political terrorists are generally younger tion. Many of the individuals who become ter-
and well-educated, the reactionary terrorists rorists or criminals could fall into this broad
tend to be less well-educated and somewhat category; the term is Iimited here to the dis-
older. turbed persons who act alone and do not act
for profit.
Terrorists, as a group, have been responsible
for approximately 12 percent of those bomb- The mentally disturbed bomber also differs
ing incidents in the past 5 years for which the from terrorists and criminals in that he general-
FBI attributed a motive, ly does not engage in multiple bombings, al-
though exceptions such as the Los Angeles “al-
Common Criminals phabet bomber” certainly exist. He generally is
poorly trained, has limited resources, and acts
Criminals range from the petty operator who
alone. He is often highly motivated, but per-
utilizes a bomb for extortion to the profession-
haps only for short periods of time, in direct re-
al bombers of organized crime, The petty oper-
sponse to some stimulus. He is extremely Iim-
ator is generalIy poorly trained, not very moti-
ited in his ability to respond to changing con-
vated, has limited resources, and cannot readi-
trol situations, either through lack of care of
ly adapt to a changing enforcement environ-
consequences or belief in his invincibility. As
ment. The only major characteristics he shares
his motives are hard to identify, it is difficult to
with the professional bomber are that his tar-
predict his target.
gets are generalIy individuals or smalI commer-
cial establishments, unlikely to be protected The mentally disturbed account for approxi-
by a detection taggant sensor, and that he mately 38 percent of al I bombing incidents
generally works alone or as part of a small that can be attributed to a specific type of per-
group. The petty operator normally engages in petrator.
repeated bombings over a number of years.
Vandals and Experimenters
The professional bomber is highly trained
and motivated and generally has considerable Vandals and experimenters share the charac-
resources available to him, either directly or teristics of poor training, limited motivation,
through his “employer.” While the profession- and limited resources. They generally work
al generally works alone, he may be affiliated alone or in small groups, and do not generally
with a larger criminal structure, such as the or- intend to harm people or cause extensive dam-
ganized crime network in the United States. age. Their targets are often of little value, such
His target may range from bombs planted as a as mailboxes or outhouses, but some acts of
result of labor problems to murder-for-hire vandalism can cause extensive damage to
“hits.” The professional bomber and the more buildings such as schools. While accounting
sophisticated terrorists share many character- for 39 percent of the reported bombing inci-
istics and are the most difficult to control or dents, they are responsible for little damage
contain. and few casualties.

Criminals as a group are responsible for ap- The primary danger from this group is that a
proximately 11 percent of bombing incidents. harmless prank may accidently turn into a ma-
Most incidents are limited to specific targets jor bombing with subsequent significant prop-
and do not generally cause substantial injury erty loss and casualties. There is also the dan-
or death to innocent bystanders. ger that experimenters will learn their craft and
“graduate” to a more dangerous category of
Mentally Disturbed criminal bomber.
The category of mentally disturbed includes In summary, table 61 shows the approximate
psychopaths, those seeking revenge for a real number of significant explosive bombings (ex-
or imagined wrong, and those who may be tem- cluding mailboxes and detonations in the
porarily disenchanted with a particular situa- open) that would be attributable to each type
146 Taggants in Explosives

Table 61 .–Estimated Number of Significant Bombings by the explosives. Such an effort would be ex-
Group of Porpotrators (average of years 1974-78) pended if the recovered bomb had the poten-
tial to cause catastrophic damage, if the target
Estimated number
Perpetrator group of bombings was an important one, or if the pattern of the
Terrorists, ., ., ., ., 107 attempted bombing indicates that useful intel-
Criminals . . . . . . . 98 ligence information would be gathered by the
Mentally disturbed ., ., . . . . ., ., ., 340 trace. Devices recovered undetonated, which
Vandals and experimenters. ., . . . . 348
were small in size or which were to be used
SOURCE FBI data See app F for a deribation of these figures against relatively unimportant targets, may
welI never be reported to the BATF network,
of perpetrator, if the same relative distribution
by perpetrator held for unattributed bombings While it is impossible to determine precisely
as for attributed ones. To obtain these esti- the source of explosives used in most criminal
mates, OTA averaged F B I data from the 5 bombings, analysis of the existing data does in-
years 1974-78 (no 1979 data is yet available). dicate some trends. Examining table 56, it ap-
Year-to-year numbers vary due to changes in pears that homemade explosives are used very
the FBI categories and method for allocating infrequently in criminal bombings in the
bombings by motive. (See app, F for more de- United States, although they account for up to
85 to 90 percent of the explosives used in coun-
tail. )
tries such as West Germany and England,
No detailed data is available concerning the where commercial explosives are rigorously
number of deaths and injuries caused by the controlled. There also appears to be little use
various bomber groups. However, almost 40 of explosives imported from abroad, a judg-
percent of catastrophic bombings (those with ment supported by discussion with various law
casualties or serious property damage) are at- enforcement agencies. Both of these sources
tributed to separatist terrorists and the more could become more important, however, if a
professional criminals. taggant program were legislated.

I I legal purchases are primarily of stolen ex-

Sources of Explosives plosives, discussed below. That leaves legal
The explosives used in criminal bombings purchases and theft as the primary current
can come from a variety of sources, including: sources of explosives.

● legal purchase, Explosive materials can be purchased legally

● illegal purchase, in each State; the requirements vary from State
● theft, to State, and they vary for different explosive
● importation from abroad, m a t e r i a l s . In every State, gunpowder can be
● homemade, and purchased legally; identification may or may
● theft of some components, fabrication of not be required for smokeless powders and is
others. required for black powder. In some States,
cap-sensitive high explosives can be purchased
At present, a determination of the source of ex-
simply by showing identification and filling
plosives can rarely be made except in the case
out a form. In others, the explosives can only
of bombs that have been recovered undeto-
be legally sold to people with State or Federal
nated. The date-shift code information on the Iicenses.
cartridge label allows the source of the recov-
ered explosives to be traced. Such traces can, A general rule-of-thumb expressed by most
theoretically, locate the source of essentially law enforcement personnel was that criminal
all cap-sensitive high explosives recovered in bombers will use the most easily available
their original cartridges; however, investigative source. If explosives can be purchased legally,
effort is necessary to determine which of the the bombers will do so; the Weather Under-
last legal purchasers on the list is the source of ground apparently purchased much of their ex-
Ch. V/— Taggant Utility Review ● 147

● The amount of cap-sensitive explosives

plosives legally in New Hampshire. If explo-
sives are easy to steal, then stolen explosives stolen and recovered appears in rough
will be used. E x p l o s i v e s a r e m o r e p r e v a l e n t balance. Some of the recovered explo-
and easier to steal in the western States; a sives, however, include abandoned explo-
large theft from Colorado, for instance, fur- sives found in old mines and other places.
nished the explosives for a large number of A significant net amount is probably avail-
bombings in the Eastern States. able, and used, for criminal bombings.
● A large net number of blasting caps ap-
BATF keeps track of stolen explosives, as
well as explosives seized, recovered, or found. pears to be stolen each year, and to be
The data for 1977 and 1978 are summarized in available for use in criminal bombings.
table 62. While no firm conclusions as to out- This is not surprising as caps are generally
standing amounts of explosives can be made not as well secured as main charge explo-
on the basis of the data, several trends are ap- sives. If a taggant program is initiated, se-
parent. curity of detonators will require upgrad-
ing, as detonators are generally needed to
Little gunpowder is stolen. As gunpowder initiate explosives and the fabrication of
are easily purchased, there is little need detonators is a much more difficult and
for theft dangerous job than fabrication of the
Large amounts of blasting agents are main explosive charge.
stolen, and recovered, each year. Accord-
ing to table 56, however, little of it is used An additional analysis can be made of the
in criminal bombings. frequency with which explosives are stolen on
More military explosives seem to be re- a State-by-State basis and compared to the fre-
covered than stolen. This may be due to quency of criminal bombings. A high correla-
the inclusion of “souvenirs” as recovered tion appears between the number of thefts and
explosives, or to the reluctance of the mil- number of bombings. An even higher correla-
itary to report thefts. At any rate, the tion appears when the thefts from nearby
amounts stolen are small. Much of the States are included in the analysis. As an exam-
miIitary explosives u s e d b y c r i m i n a l ple, both California and New York have more
bombers is material acquired some years stringent regulations controlling the use and
ago. For instance, the Cuban exile terrorist storage of explosives than nearby States such
groups, such as omega 7, still primarily as New Jersey and Washington. Law enforce-
use C-4 given to them by the Central In- ment officials feel that many of the incidents
telligence Agency at the time of the Bay in New york and California use explosives
of Pigs invasion. stolen in New Jersey and Washington.

Table 62.–Stolen and Recovered Explosive Summary

Amount stolen Amount recovered

Tvpe 1977 1978 1977 1978
Blasting agents, pounds . . 20,834 42,172 21,260 23,623
Black powder, pounds. . . . . 145 379 277 723
Smokeless powder, pounds . . . 0 163 16 1,361
Boosters, pounds . 2,177 9,528 2,804 362
Military explosives, pounds ., 49 140 640 701
Cap-sensitive high explosive, pounds 36,498 44,316 43,738 41,097
Primer, units ., ., ., : : . 1,300 4,333 2,733 344
B l a s t i n g c a p s , u n i t s 61,531 66,614 40,719 44,456
Det. cord/safety fuse/lgnltor cord feet . 183,224 113,510 84,554 101,117
T o t a l , e x p l o s i v e s , p o u n d s 61,003 101,217 71,470 74,966
B l a s t i n g c a p s , u n i t s 61,531 66,614 40,719 44,456
Det cord/safety fuse/igniter cord, feet . . 183,224 113,510 84,554 101.117

SOURCE BATF 1978 Explosivea incidents Report

148 ● Taggants in Explosives

Current Security Measures Table 63.–Explosivos Thefts by Method of Entry–

Number of Incidents and Percentages for 1977-78
Sources of Explosives
Number Percentage
Current methods of securing explosives vary Entry method 1977 1978 1977 1978
somewhat from State to State; different types Locks cut. . . . . . . . . . . . 59 71 31.1 26.9
of explosives are also secured in different Locks pried ... . . ., 36 50 18.9 19.0
ways. In general, all cap-sensitive high ex- Door pried . . . . . . . . . . . 10 10 5.3 3.9
Key. ... . . . . . . . . . . . 14 23 7.4 8.8
plosives, including boosters and detonating Window entry. . . . . . . . 7 3 3.7 1.1
cord, must be stored in BAT F-approved maga- Inside help. ., . . . . . . . . 3 0 1.6 –
zines. The magazines require hardened locks Wall entry ... . . . . . . . 10 16 5.3 6.1
Burning. . . . . . . . . . . 2 1 1.0 .4
and lock-covers to protect the lock from direct Roof entry . . . . . . . . . . 7 3 3.7 1.1
access by hacksaws or from attempts to shoot Door blown. . . . . . . . . . . 1 2 .5 .8
off the lock. Detonators must be stored sepa- Floor entry . . . . . . . . . . 0 .4
Vent entry . . . . . . . . . . . 1 3 .5 1.1
rately, in magazines that are not as well pro- Other b. . . . . . . . . . . . . . 40 80 21.0 30.4
tected from theft as the high-explosive maga- Unknown. . . . . . . . . . . . 137 99 – –
zines. Blasting agents are not as well-regu- Total . . . . . . . . . . . . . 327 362 100 100
Iated; bulk ANFO is often stored in large hop- a These percentages do not Include 137 unknown method incidents for 1977 and the 99 Incident
pers for direct loading into trucks. Gunpow- b
for 1978
This figure reflects those incidents where the entry method could not be Placed in the above
der are stored in BATF-approved magazines, categories
at least at the manufacturer and distributor SOURCE BATF 1978 Explosives incidents Report

levels. At the retail sales level however, gun-

powders are just stacked on the shelves. Bingham Copper Mine, for instance, the maga-
The above provisions are for permanent zine is placed within the interior of the proper-
storage; some States allow overnight storage ty of the large open pit mine. The mine has a
of explosives in temporary magazines; at least limited number of access points, controlled by
one manufacturer keeps less than full-lot guards. As the mine is operated three shifts a
amounts of detonators in the detonator as- day, 7 days a week, it would be difficult for
sembly area overnight. anyone to gain illegal access to the magazine
area. A similar situation prevails for at least
The purpose of BATF and other regulations
one manufacturer. T h e e n t i r e p r o p e r t y i s
on the storage of explosives is primarily to pro-
fenced with cyclone fencing, topped by
tect against surreptitious or casual theft by
barbed wire. Inside the perimeter, and placed
outsiders, in much the same way that locking
strategically throughout the complex, is a
your car door protects the car from theft. The
microwave break-circuit alarm system. These
magazines, however, are fairly flimsy, often
facilities are in sharp contrast to others, in
simply a correlated frame building with addi-
which the magazines are located in areas re-
tional plywood or plank walls. Entry can still
mote from other operations, and accessible by
be gained by cutting or prying off the locks,
nearby roads.
forcing entry through the door, a window, the
roof, or a vent, or by help from an employee. Security of explosives on military reserva-
Table 63, from the BATF 1978 Explosives Inci- tions is stricter, w i t h m a g a z i n e s w i t h i n a
dents Report, tabulates the methods used to fenced area. Security lighting is provided, the
gain entry to explosives. An average of 48 per- magazines are either directly guarded or pro-
cent of known entries were by removing the tected by an alarm which would bring a re-
lock, another 16 percent were by forcing entry sponse within 15 minutes, security patrol in-
through the door, wall or vent, while almost 9 spections are held at frequent intervals, and
percent involved the use of a key or other in- access is only through secured access roads.
side help.
At present neither commercial nor military
Some magazines are well-protected by their installations can guard against theft by in-
placement in a facility or by guards. At the siders. While the theft of case lots would be
Ch. V—Taggant Utility Review ● 149

quickly discovered by inventory procedures, it Government buildings, banks, police sta-

would be difficult to detect the theft of small tions, and military establishments account for
amounts of explosives, whether by military less than 10 percent of bombing incidents and
troops or by a miner daily placing a couple of just over 3 percent of casualties. Most of these
sticks of dynamite in his lunch pail. targets have controlled access and maintain
some sort of guards. In times of increased
Transportation of explosives is another po- bombing threats, as happened in the late
tential point of theft. The primary purpose of 1960’s and early 1970’s, many of these facili-
regulations concerning the transportation of ties instituted checks of incoming people and
explosives is to protect those people who Iive packages. A similar situation exists with re-
along the route being traversed. For that rea- spect to high-value manufacturing facilities,
son trucks are clearly marked when they carry utilities, and high-value complexes within edu-
explosives. Commercial explosives are often cational facilities, such as computer centers.
transported by a single driver; military ex- Many of these facilities now require inspection
plosives normally have two drivers. In neither of any parcels (including briefcases and purses)
case is the driver normalIy armed. brought into the facility, as well as identifica-
tion of people entering. Detection sensors
Potential Targets could be easily installed in each of these facili-
ties, given sufficient threat.
A previous section discussed the wide varie-
ty of targets attacked by criminal bombers. Airports and aircraft represent another ma-
The security measures vary widely for each jor class of potential targets. While attacks on
type, in response to the perceived probability airports and aircraft represent well under 1
of attack and the perceived consequence of percent of incidents, the catastrophic conse-
such a bombing. Table 59 indicates that almost quences of an aircraft bombing make it an at-
half of the bombing incidents (and 60 percent tractive potential target for criminal bombers
of bombing casualties) result from attacks on and the subject of much current security ef-
private residences and vehicles. Security at fort.
these targets is almost nonexistent, unless the
Current large aircraft cost in the neighbor-
individual believes he is likely to be attacked;
hood of $2o million to $50 million each, and
except in certain cases, such as Government
carry several hundred passengers. A single air-
officials or witnesses, it is unlikely that law en-
craft bombing could, therefore, cause more
forcement officials play much of a security
property damage and more deaths than the
role with regard to those targets.
sum of all domestic bombings this decade.
Another 32 percent of the incidents, and 30 Table 64 lists the explosions that have oc-
percent of the casualties, occur in commercial curred aboard U.S. aircraft from 1949 through
establishments. Most of these establishments 1976. Table 65 lists the location of the explo-
have no security means at present and it is sive devices for the 19 U.S. aircraft listed in
unlikely that the development of detection table 64 and compares the location with the 63
taggants and sensors would significantly aircraft bombings worldwide in that time peri-
change that situation. Some large office build- od. Table 66 Iists the 26 incidents between
ings, with control led access, have provisions 1972 and 1976 in which explosive or incendiary
for checking people as they enter and leave devices were found at U.S. airports. All of the
the building and, in fact, institute checks in tables are from FAA report FAA-R D-77-28. The
off work hours. G iven a sufficiently severe tables show that no bomb has caused casual-
bombing threat, it would be possible to protect ties on a domestic flight since 1962; in fact,
the larger facilities by a detection sensor, but since 1962, all but one of the casualties, and all
the difficulties involved, the large number of deaths at U.S. airports or on U.S. domestic
facilities, and the cost of operators and equip- flights, were caused by bombs placed in
ment probably preclude such deployment. lockers.
150 ● Taggants in Explosives

Table 64.–Explosions Aboard U.S. Aircraft

Date Carrier Aircraft Aircraft location Bomb location Outcome Device

.... -- -- .11 minutes
. . after
,. 10
-- Baggage Airplane disintegrated–44 killed Dynamite
1 1/1/55 UAL UL-6B
7/25/57 WA CY-240 47 minutes after TO Lavatory Passenger thrown out of lavatory– Dynamite
hole in aircraft side; plane landed
1/6/60 NA DC-6B 184 minutes after TO Under seat passenger 34 killed, airplane disintegrated Dynamite, dry cells
5/22/62 co 707 39,000 ft Towel container in rear Tail blown off–45 killed Dynamite
11/12/67 AA 727 102 minutes after TO Rear baggage 3 bags destroyed; aircraft saved Black powder (?)
11/ 19/68 co 707 24,000 ft Lavatory Fire and explosion in lavatory; —
extinguished by crew; plane landed
8/29/69 TW 707 Ground after hijack Explosives thrown in No casualties from explosion Grenades &
(Damascus, Syria) cockpit after evacuation canister explosive
9/7/70 PA 747 Ground after hijack — Demolished after evacuation —
(Cairo, Egypt)
9/12/70 TW 707 Ground after hijack — Demolished after evacuation —
(Dawson Field, Jordan)
12/29/71 — Turbo Cmdr In hangar Seat in cabin Aircraft destroyed, hangar damaged; —
no casualties
3/8/72 TW 707 Parked on ground Cockpit No casualties (plane empty) c-4
9/21 /73 — Navion Parked on ground Engine manifold Not known —
12/17/73 PA 707 On ground, Rome Attack while loading Fire damage; 30 killed, White phosphorous
many injured grenades
8/26/74 TW 707 On ground, Rome Aft baggage compartment Fire, confined to local area; c-4
no casualties
9/8/74 TW 707 Over Ionian Sea Aft baggage compartment High-order explosion; 88 killed, —
aircraft lost
2/3/75 PA 747 In air, Burma Lavatory (suicidal Extinguished by crew; minimum Petrol and butane
passenger set fire) damage
12/19/75 — Alouette On ground Near fuel tank $10,000 damage to aircraft Blasting caps
7/2/76 EA Electra Parked next to fence External, near right Explosion and fire destroyed main Dynamite (8-10
landing gear fuselage sticks)
7/5/76 — Helicopter On ground External, under tail Extensive damage Dynamite

SOURCE FAA Civil Avaton Security Service

Table 65.–Location of Explosions Aboard Aircraft, 1949-76 from FAA report FAA-RD-77-28, shows a de-
tailed schematic of the flow of people and ma-
Worldwide U.S. aircraft
terial into the airport area.
Location of explosion Number Percent Number Percent
Stowed . . ... , ., . . 13 21 21 It is possible that bombs could be intro-
Baggage. . . . . . . . (8) — — duced through the mail, freight, air courier
Cargo or freight ., ... (5)
services, or f o o d services, as well as from
Ground attack. . . ., 5 8 4 21
External attachment. . . 7 11 3 16 checked baggage; or could be carried on by
Passenger or crew aircraft flight or service personnel or by pas-
compartment. . . . . 52 42
Lavatory. ... . . . . . — — sengers. Current security procedures assume
Passenger compartment (19) — (2) that personnel screening procedures will be
Cockpit. . . . . . (4) — sufficient to eliminate a serious threat from
Unknown, . . . . . . . . . 5 8 o airport or aircraft personnel and that air
Total . . . . . . . . 63 100 19 100
freight and mail service would not allow a
SOURCE Data supplied by FAA Civil Avation Security Service criminal bomber to be sure his bomb would be
aboard a particular aircraft. Current aircraft
Current airport security is based on an at- security procedures, therefore, concentrate on
tempt to separate the areas of public access passengers, carry-on baggage, and checked
from the secure air operations areas. Figure 23, baggage. Air courier services, in which a small
Ch. V—Taggant Utility Review ● 151

Table 66.–Explosions and Device Found at U.S. Airports, 1972-75

Date Airport
Location Effects Comment Device
,Kenneay No explosion Detected by dog c-4
Cockpit of TWA B-707
3/8/72 Seattle Baggage compartment (UAL No explosion Extortion attempt; timer Gelatin dynamite in aerosol
flight) stopped cans, blasting caps
11 /19/72 Denver Attache case carried by No explosion Indvidual stated intent to 8 sticks of dynamite
Individual blow up plane
3/24/72 San Carlos, Calif Hanging from belly of Hole in ground at remote Removed by police 3 sticks of dynamite, timer and
helicopter location detonators
12/1 /72 Grand RapIds, Paper towel container in No exploslion Device extinguished after —
Mich. terminal emithng smoke
12/31 /72 Austin Concession area Moderate damage — Incendiary (gasoline)
3/20/73 Los Angeles On runway during approach None Thrown by individual on field Molotov cocktail
of Continental Airlines plane
3/29/73 Milwaukee Locker 1 Injury–moderate damage Extortion attempt —
8/9/73 Los Angeles Locker Did not detonate Extortion attempt/located —
by dogs
11/30/73 Nashville Locker Did not detonate Extortion attempt Smokeless powder, timer,
3/1 /74 Kennedy Locker 3 injured–moderate damage — —
7/21 /74 New Orleans (unknown) No explosion Removed by bomb squad 3-m long bamboo with powder
and fuse
8/1 /74 Kennedy Cargo building No explosion Removed Cardboard container with
explosive powder, fireworks
8/6/74 Los Angeles Locker 3 killed, 34 injured — —
8/9/74 Johnstown- Hangar Hangar and aircraft destroyed — Probable incendiary (in 55-gal
Camoria, Pa. drum)
8/26/74 O’Hare Men’s room Commode damaged — Probably firecrackers
9/ 16/74 Boston Airline baggage room Substantial damage Bomb was in an unclaimed Incendiary (?)
suitcase destined for Tel Aviv
3/15/75 San Francisco Near ticket counter Minor damage — Probably firecracker
3/22/75 Honolulu Lost & found baggage area Did not detonate — Crude pipe bomb
3/27/75 Kingsford, Mich Storage area No explosion Removed —
7/22/75 Tampa Baggage cart 1 injured — Firecrackers
10/17/75 Miami Locker Lockers and ceiling destroyed — —
10/20/75 Miami Dominican Airlines Office No explosion Discovered by janitor; Time bomb
disarmed by bomb squad
11/6/75 Buffalo Baggage claim area (2 bags) No exploslon Checked bags unclaimed after Black powder and gasoline
flight, timers turned off
11 /27/75 Miami Bahamasair aircraft. Behind No explosion Removed —
wall panel in lavatory
12/29/75 La Guardia Locker 11 killed, 70 injured; — Dynamite and RDXa
substantial damage
a FAA estimate Other agencies diagree with this assessment
SOURCE FAA CIVI I Aviation Security Service

parcel can be placed aboard a specific aircraft eter, which will trigger an alarm upon detec-
for subsequent pickup, are treated in the same tion of a significant metal mass, such as a gun
way as freight or maiI by most airlines. or knife. If the alarm is triggered, the passenger
is instructed to remove any metal objects, s u c h
as keys, and repass through the magnetometer,
As a result of the hijacking threat in the mid- If an alarm still rings, he is searched by a hand-
1970’s, a set of procedures were developed to held magnetometer and subject to a patdown
deal with passengers, checked baggage, and search if the alarm persists. FAA estimates that
carry-on baggage. Figure 24 (from FAA report the probability of detection of guns or knives
FAA-R D-78-66) shows a schematic of the pas- by the magnetometer, hand magnetomer, and
senger and carry-on luggage-screening systems. patdown, are 0.90, 0.95, and 0.95, respectively,
Passengers must pass through a magnetom- adding up to an overall detection probability
152 ● Taggants in Explosives

13 NNOSU3d

i i
I i

u =\ I

~ ? ,
d d D
L =~ - - - - - - - k
- - - - - - I G r i l


~lE~,,,,,,,,--f: - - - - - - - - - - - - - - - - - - 1 6

— —

b = 4
13 iNNOSti3d
Ch. V—Taggant Utility Review ● 153

Figure 24.—Passenger/Hand-Baggage are nonoperable). FAA estimates that the prob-

Screening Station abiIity of detecting guns and bombs in carry-on
baggage is 0.81 and 0.19, respectively.

o FAA estimates are probably high, especially

for the X-ray detection of illegal materials in

o hand baggage. Magnetometers are set to a

wide range of sensitivities; one may trigger on
a small keyring while another may fail to trig-
ger on a sizable metal mass. X-ray attendants
are generally paid at, or near, the minimum
wage, have little training, and must deal with
the problem of maintaining alertness for long
hours while performing an extremely dull job.
While an attendant may well recognize a gun,
particularly at the start of a shift, it is doubtful

o= = Screening contractor personnel

that a carefully constructed explosive device
would be detected.

o Law enforcement officer Notwithstanding the above limitations, the

use of magnetometers and X-ray machines,
SOURCE: FAA report No. FAA-RD-78-88 coupled with a search profile of likely hi-
jackers, has resulted in the recovery of an im-
of 0.81. * The system is not designed to detect pressive amount of hardware, and the arrest of
bombs, but FAA estimates that the probability substantial numbers of people, as shown in
of detecting a bomb is 0.17. table 67 (from FAA report FAA-R D-77-28), as
Carry-on baggage is screened, either by an X- well as the virtual halt of hijackings of U.S.
ray examination or by visual hand search (only domestic airlines.
at small airports or when the X-ray machines
The current procedure for screening
● The total probability of detection must be less than the prob- checked baggage consists simply of ensuring
ability of detection by the magnetometer, as no subsequent that baggage can only be checked by a pas-
searches are conducted on those passengers who do not trigger
the magnetometer Total detection probability is thus senger with a valid ticket. When checking bag-
PDT = ( PD 1) (PD,) (PDN) gage at curbside or at the check-in counter, the

Table 67.–Results of Civil Aviation Security Program Passenger Screening

1972 1973 1974 1975

Passengers (millions) ., . . ., . . ., 192 203 201 202
Passengers denied boarding. ., ., 8,265 3,459 2,663 (a)
Referrals to law enforcement . . . ., ... (a) (a) (a) 12,270
Persons arrested. ... ... ., ., . . ., 3,658 3,156 3,501 2,464
Aviation offenses detected
Carrying weapons or explosives aboard aircraft 774 736 1,147 1,364
Giving false information ., ., ., ., . . . . 244 658 1,465 227
Weapons detected
Firearms. . . 1,313 2,162 2,450 4,783
Explosive devices . . ., ... ... ., 13 3,459 14,928 b 158
Ammunition, fireworks, ... ., ., ., . . (a) (a) (a)