THE MOST ADVANCED

VOVA.GE DA.TA. RECORDERS

IN THE WORLD

Most advanced features in the industry

Preferred choice of the cruise line industry
Unbeatable performance record
Years of operational success and proven
data retrieval
Largest international team dedicated
exclusively to VDR services
Worldwide type approvals

RUTTER TECHNOLOGIES INC.

ST . ..JOHN'S, N L, CANADA
TEL: + 170936S42 13/4
FAX: +1709-36S1337
WWW. RUTTE RTEC H. C C M n:073510 2 RUTTER
 ~

INTERTANKO

A Guide to Crude Oil Washing

and Cargo Heating Criteria

May 2004

 G)

I NTERTANKO

International Association of Independent Tanker Owners

Acknowledgements

INTERTANKO is grateful to the author of 'A Guide to Crude Oil Washing

and Cargo Heating Criteria', Timothy Gunner, and the INTERTANKO
Safety Technical and Environmental Committee (ISTEC) for their valuable
contribution to this publication.

All rights reserved. No part of this publication may be reproduced in any material form (including
photocopying or storing it in any medium by electronic means and whetheror nottransiently or incidentally
, to some other use of this publication) without the written permission of the copyright owner. Applications
for the copyright owner'swritten permission to reproduce any part of this publication should be addressed
to the publisher.
INTERTANKO 2004

Whilst every effort has been made to ensure that the information contained in this publication is correct,
neither the authors nor INTERTANKO can accept any responsibility for any errors or omissions or any
consequences resulting therefrom.

No reliance should be placed on the information contained in this publication without independent
verification.

INTERTANKO OSLO INTERTANKO LONDON

P.O. Box 5804 Majorstuen St Clare House 30-33 Minories
0308 Oslo, Norway London EC3N 100. UK

Tel: +47 22 122640 Tel: +44 20 7977 7010

Fax: +4722122641 Fax: +44207977 7011
Email: Oslo@intertanko.com Email: London@intertanko.com

INTERTANKO ASIA INTERTANKO NORTH AMERICA

5 Temasek Boulevard 801 North Quincy Street
#12-07 Suntec City Tower 5 Suite 200, Arlington
Singapore 038985 VA 22203, USA

Tel: +65 6333 4007 Tel: +1 7033732269

Fax +65 6333 5004 Fax: +1 703841 0389

Email: Singapore@intertanko.com Email: Washington@intertanko.com

www.intertanko.com
 A Guide to Crude Oil Washing
and Cargo Heating Criteria

May 2004

TABLE OF CONTENTS

1 Introduction 1

2 The COW Manual 2

2.1 Section 9 of the COW Manual 3

3 What is crude oil? 6

4 The physical behaviour of crude oil 6

4.1 The physical behaviour of aromatic crude oils 7

4.2 The physical behaviour of paraffinic crude oils 9

5 The Cloud Point temperature determination of a crude oil 10

5.1 Methods for determining the Cloud Point temperature onboard a tanker 11

5.1.1 Bondi Test method 11

5.1.2 Primary Cloud Point test method 13

5.1.3 Density method 16

5.1.4 The Cloud Point temperature model 17

5.2 The predicted extent of precipitated sludge before COW 18

6 The practical implications of the Cloud Point determination 19

6.1 MARPOL regulations and commercial implications 19

6.2 The crude oil washing of sludge and some practical implications 21

7 COW equipment and procedures 23

7.1 COW machines and tank cleaning nozzles 23

7.2 COW pipeline systems 25

7.3 Stripping systems 26

7.4 Crude oil washing concurrent with cargo discharge 27

7.5 "Closed cycle" crude oil washing 28

8 Crude oil washinp of sludges 29

8.1 The theory and background 29

8.2 Practical circumstances affecting the crude oil washing of sludges 31

8.3 An alternative, practical programme for the crude oil washing of sludge 34

9 Closing remarks 36

1 Introduction

Regulations laid down in the 1978 Protocol to the 1973 Marine Pollution Convention (MARPOL
73/78) require the cargo tanks of crude oil tankers to be cleaned using a procedure called crude oil
washing (COW). With the COW procedure the crude oil cargo itself is used as the cleaning medium.
During the 196,6s it was discovered that crude oil, when applied to the cargo still remaining on tank
floors and c1in~ing to the tank structures, using tank cleaning machines, effectively dissolves and
dilutes these residues and mixes it in with the rest of the cargo which is being discharged ashore by
the cargo pumps.

Prior to the advent of COW, cargo tanks were washed with sea water on their ballast voyage to the
next loading port. The mixture of oil and cleaning water resulting from this type of cleaning operation
could settle out in the tanker's slop tanks with the decanted water being discharged overboard into
the ocean. Consequently, this operation resulted in inevitable operational discharges of oil-water
mixture into the sea.

However, the use of crude oil to COW the tanks means that the solvent action of the crude oil makes
the process far more environmentally friendly than when water is used. Additionally, after undertaking
COW, th volume of cargo residues left in the tanks is greatly reduced removing the subsequent risk
of opera ional discharges at sea.

Modern ankers are designed with segregated ballast tanks (SST), and there are only a few stipulated
occasio~s on which seawater comes into contact with the oil cargo system during the course of
normal tbnker operations. The requirement for new crude oil tankers to be built with double hulls,
introduc~d in the 1990s, has further improved the efficiency of COW operations because more of
the structural support members are placed outside the cargo tank and on these types of ships, the
amount bf crude oil residues left in the cargo tank following discharge is much reduced. Overall, the
COW prbcedure and ship design changes have greatly reduced the need for operational discharges
from tankers.

RegUlati~n 13S of Annex I of MARPOL 73/78 requires that the COW installation and arrangements
onboardla tanker should comply with the provisions of the "Specifications for the Design, Operation
and Conkol of Crude Oil Washing Systems"adopted by the International Maritime Organization (IMO)
in 1978'fThe COW regime requires that before departure on a ballast voyage, after the complete
discharg of cargo, sufficient tanks shall have been crude oil washed to preclude the ballasting of a
cargo ta~k without it having been crude oil washed. On SST ships approximately 25 per cent of the
crude oi carrier's cargo tanks need to be washed, in the prescribed manner, on every voyage for
sludge c ntrol purposes provided that no tank need be crude oil washed for sludge control purposes
more th n once in every four months. For tankers with insufficient SST capacity, the number of tanks
to be crude oil washed has to be increased above this minimum level in order to render sufficient
cargo ta~ks "clean" enough (as defined by the regulations) to take onboard enough water ballast to
achieve he tanker's required sailing ballast draught for the voyage.

Page 1 INTERTANKO - A Guide to Crude Oil Washing and Cargo Heating Criteria - May 200
 In addition to the regulatory controls governing the use of COW, cornmercial or charter party
requirements may require the tanker operator to carry out a greater or lesser degree of COW than
the specified rninimum in order to rnaximise the discharge of the crude oil cargo. Notwithstanding
these commercial pressures for the extent of COW to be undertaken, at no time should a tanker
undertake less than the minimum levels specified in paragraph 6 of Section 1 of the mandatory
onboard COW Manual.

Although the MAR POL 73/78 COW regime has proved to be eminently successful in minimising
tanker operational discharges and improving cargo outturns during the last two decades of the 20th
century, the tanker industry has also been learning more about the behaviour of crude oil cargoes
over the period. A number of research projects' have led to a better understanding of the COW
process and how it could be further improved. As a result of this work and at the initiative and
suggestion of INTERTANKO, in 1999 the IMO adopted amended COW requirements that are laid
down in the revised "Specifications for the Design, Operation and Control of Crude Oil Washing
Systems". These revised Specifications can be found in the 2000 Edition of the "Crude Oil Washing
Systems" publication, issued by the IMO.

From an operational perspective the changes provide a more realistic and accurate way of determining
the suitability of a crude oil for use in crude oil washing.

2 The COW Manual

The 2000 Edition of the "Crude Oil Washing Systems" publication contains the following:
-Revised "Specifications for the Design, Operation and Control of Crude Oil Washing Systems"
(IMO Resolution A.446(XI), as amended by Resolutions A.497(XII) and A.897(21));
-Standard format for the "Crude Oil Washing Operations and EqUipment Manuals" (IMO Resolution
MEPC.3(XII), as amended by Resolution MEPC.81(43));
-Examples of "Crude Oil Washing Operations and Equipment Manuals";
-"Guidelines for In-port Inspection of Crude Oil Washing Procedures".

It is a MARPOL requirement that every crude oil carrier has onboard a reference manual, known
as the "Crude Oil Washing Operations and Equipmen: Manual", in order to assist the responsible
tanker officer in carrying out the prescribed COW operation correctly. Each tanker's COW Manual is
customised to meet the specific requirements of the particular ship. The COW Manual is comprised
of the following 17 sections:

1. The text of the revised specifications.

This section contains the complete text of the revised "Specifications for the Design, Operation
and Control of Crude Oil Washing Systems" and the agreed interpretations of certain of the
provisions of the revised Specifications.

1 T.J. Gunner, "Pollution Control for Crude Oil Tankers and the Accuracy in the Measurementof theirCargoes"
- University of Wales .1993

INTERTANKO - ,A, Guide io Crude Oil \jVElsh\\lQ <.H",C: CatTiO Heahng Criteria - ~/ay 2004 Page 2
2. COW system drawings

This section contains line drawings showing the following aspects:

(a) crude oil washing lines and valves;

(b) cargo pumps, lines and valves;
(c) ballast systems (where fitted);
(d) stripping systems;
(e) position of tank washing machines;
(f) position of facilities for hand dipping and tank gauges; and
(g) inert gas deck distribution system.

3. Description of the COW system and operational and equipment parameters. This section contains
a description of the cargo, ballast, washing and stripping systems. In addition, it specifies the
following details:
(a) types of tank washing machines and their standpipe length inside the tanks;
(b) revolutions of the machines;
(c) methods of checking the operation of tank washing machines;
(d) minimum operation pressure for crude oil washing;
(e) maximum permitted oxygen level in cargo tanks;
(f) duration of tank washing machine cycles;
(g) results of tests performed in accordance with paragraph 4.2.1 O(b) of the revised

Specifications and the method of computation; and

(h) method of preventing entry of oil/vapours into the engine room.

4. Dangers of and precautions against oil leakage

5. Use and control of inert gas
6. Precautions against electrostatic hazards
7. Personnel requirements (training and qualifications)
8. Methods of communication during COW
9. Determination of the suitability of a crude oil for use in crude oil washing
10. Crude oil washing checklists
11. Approved methods and programmes for crude oil washing
12. Typical crude oil wash programmes
13. The method for draining cargo tanks
14. The method and procedures for draining cargo pumps and lines
15. Typical procedures for ballasting and method of preventing hydrocarbon vapour emissions
16. Compliance procedures for Regulation 9 of MARPOL Annex 1
17. Inspections and maintenance of equipment

2.1 Section 9 of the COW Manual

Each tanker's COW Manual contains all the necessary information regarding the technical operation
of the crude oil washing equipment and associated onboard systems. Of particular interest in the
context of this publication is Section 9, which provides information on the crude oil washing medium.

Page 3 INTERTANKO - /1, Guide to Crude Oil Washing and Cargo Heatmg Criteria - May 200
This section was the subject of major amendment by IMO in 1998. Following detailed research into
the behaviour of a wide range of different crude oil cargoes, it was agreed that it would be more
appropriate to specify criteria to help determine the suitability of a crude oil for COW rather than the
previous approach of lncludinq a long list of crude oils deemed to be unsuitable for COW As a result,
the heading of Section g has been changed from "List of crude oils unsuitable for crude oil washing"
to "Determination of the suitability of a crude oil for use in crude oil washing".

Section g, like all sections of the COW Manual with the exception of the Specifications, is
recommendatory. The revised contents of section g of the COW Manual is worded as follows:

QUOTE

(1) For a tanker not fitted with heating coils

This tanker is not fitted with heating coils in the cargo tanks and should not carry cargoes which will
require to be heated either to obtain minimum pumpabiiity criteria or to avoid excessive sludging of
both the designated clean ballast tanks and the vessel's tanks to be washed for sludge control.

Notes and Definitions

(1) The pumpability criterion is determined by the oil's kinematic viscosity at the observed temperature
of the cargo prior to its discharge. In order to attain optimum efficiency for discharge this viscosity
should not exceed 250 centistokes (cSt) and never be in excess of 600 cSt.

(2) The excessive sludging criterion is determined primarily by the crude oil's temperature throughout
its transportation and storage. If the cargo's temperature is likely to drop below the crude oil
cargo's Cloud Point temperature, then sludging of cargo tanks is to be expected. The Cloud
Point temperature is the temperature at which the crude oil's wax and associated solid phase
separates from the bulk liqUid phase of the cargo.

(2) For a tanker only fitted with heating coils in slop tank(s)
This tanker is fitted with heating coils only in the slop tank(s) and should not carry cargoes which
will require to be heated for pumpabiiity.

Notes
(1) If sludge deposition is suspected or determined in the main cargo tanks, then an alternative
crude oil washing programme should be utilised. Recommendations regarding this programme
and procedure are to be found in Section 11 of this COW Manual.

(3) For a tanker fitted with heating coils

This tanker is fitted with heating coils in all the cargo tanks and, SUbject to the iimitations of the
cargo heating system, can carry cargoes which require heating for either pumpability or sludge
deposition/control.

General Guidance
Attention is drawn to the difficulties which may be encountered with certain crude oils. During the
discharge of crude oils that exhibit criteria that would create either pumpability problems or sludqe

INTER IANKO - A Cude to Crude Oil VVashing and Ca-ge Heating Criteria Mav 2004 Page 4
 deposition, crude oil washing of each tank scheduled for such an operation should be carried out
concurrently with the discharging of the particular tank in order to minimise the affect on the crude
oil residues of cooling. Cooling will increase both the kinematic and dynamic viscosities of the tank
residues and, therefore, affect the efficiency of the crude oil washing programme.
As a general guidance to the suitability of an oil for crude oil washing on board this tanker, the
following criteria should be used:

(1) For aromatic crude oils whose kinematic viscosity is the temperature controlling characteristic,
the kinematic viscosity of the oil used for crude oil washing should not exceed 60 cSt at the oil
wash medium temperature

Or

(2) For paraffinic crude oils whose Pour Point temperature is the controlling characteristic, the
temperature of the cargo to be used for crude oil washing should exceed its Cloud Point
temperature by at least 10C if excessive sludging is present and should only be used once in a
"closed cycle" washing programme.

The approximate Cloud Point temperature of an oil may be calculated with the following formula,
where the Pour Point temperature (x) of the crude oil is known:

Cloud Point temperature eC) = 20.2 (100.00708. 0.1157714) + 8'

Where:
x is the crude oil's Pour Point temperature degree C

UNQUOTE

As mentioned, the above advice is contained in Section 9 of the COW Manual which, in turn, is
contained in the 2000 Edition of IMO's COW Systems publication. The advice shows that the crude
oil itself is key to the effectiveness of all COW programmes.

Crude oil is a generic name for a broad range of hydrocarbon liquids. Crude oils range from the wax
rich, paraffinic cargoes with high Pour Point temperatures to the aromatic high-viscosity types. The
densities of the full spectrum of crude oil cargoes range from about 700 to 980 kg/m'. Some crude
oils contain large amounts of gaseous/volatile components Which, in turn, may contain hydrogen
sulphide. In contrast, other crudes contain little or no gaseous component. To achieve the desired
goal of adequately cleaning cargo tanks carrying crude oil, it is essential to understand the nature
and behaviour of a crude oil prior to using it as a washing medium.

2 For a worked example using this equation and a corresponding graph for Cloud Point temperature determination,
please see section 5.1.2 below

Page 5 INTERTANI<O A GUide to Crude Oil Washing and Cargo Heating Criteria May 200
 3 What is crude oil?
Crude oil is a naturally formed liquid/substance that is, probably, the most complex liquid carried by
sea. The term "crude oil" loosely describes a vast number of varying types of liquids that contain the
basis of supply of all hydrocarbons required by society. To the general population, this substance
is perceived to be the black sticky material that is very occasionally found washed up on beaches
or seen in media coverage of pollution incidents covering sea birds and mammals. In fact, such a
perception is wrong as crude oil is a heterogeneous mixture of hundreds of differing hydrocarbon
species, all in differing proportions depending upon the specific type of crude oil and even perhaps
the unique consignment of the oil. In nature it is far from the black sticky substance, except for certain
types of crude oil, and is normally a brown/black liquid whose viscosity is similar to that of water.

In the context of crude oil washing, the MARPOL 73/78 definition of crude oil is as follows:
"Crude oil means any liquid hydrocarbon mixture occurring naturally in the earth, whether or not
treated to render it suitable for transportation, and includes:
(a) crude oil from which certain distillate fractions may have been removed; and
(b) crude oil to which certain distillate fractions may have been added."

The treatment referred to is the preparation of the crude oil for transport by stabilising it. In this
process the dissolved gas content within the crude oil, including the toxic hydrogen SUlphide gas, is
reduced. Reference to the removal of distillate fractions expands the number of liquids capable of
being used for crude oil washing to include "topped crude oils", i.e. partially refined crude oil. In the
extreme case this definition could apply to atmospheric residue, i.e. the residue remaining follOWing
the distillation of crude oil as the definition does not contain any percentage of the fractions that
may have been removed. Reference to the addition of certain distillate fractions further augments
the number of COW liquids covered by the definition to include spiked crude oils and, again in the
extreme case, reconstituted (recon) crude oils.

4 The physical behaviour of crude oil

As a basic rule for all substances their physical state varies with the temperature of the substance
that reflects the heat energy supplied to the substance or retained by the substance. The physical
states of a substance fall into three primary phase categories namely, their solid, liquid and gaseous/
vapour phases. In the case of crude oil as a heterogeneous mixture of many individuai hydrocarbon
substances, its normal visual state is that of a liquid.

Crude oil as a mixture contains numerous species of hydrocarbons that will also, if separated from
the original mixture, exhibit the same three phases, as described above, at varying temperatures. If a
pure individual hydrocarbon specie were to be studied, then precise temperatures may be ascertained
for the formation of each of the relevant phases. However, by mixing the numerous hydrocarbon
types together to generate the complex mixture which constitutes a crude oil, a vast array of differing
physical properties is created. The determination of the behaviour of the overall mixture and the
interaction of the various hydrocarbon species upon each other become very important when

INTERTANKO -/\ Guide to Crude Oil Washing and Cargo Heatirg Criteria - May 2004 Page 6
considering the most effective methods for the removal of any separated sludge from a crude oil, i.e.
the interaction of liquid solvent phase upon a partially solidified phase. However, this publication will
not attempt to describe such behaviour in detail but will concentrate upon that behaviour that would
impact the transportation criteria and the subsequent use of a crude oil as a washing medium.

In terms of their composition and their general cargo-handling attributes, crude oils can be considered
to fall into one of three categories, i.e. low-wax, higher viscosity aromatic crude oils; highly waxy,
paraffinic crudes; intermediate/naphthenic crude oils that fall between these two extremes.

Paraffinic crude oils, for example, Bu Attifel, Widuri and Sarir, have relatively high Pour Point
temperatures due to the relatively large amounts of paraffinic wax they contain but much lower
viscosities than aromatic crudes in the liquid phase. Paraffinic crude oils are temperature-sensitive
and require continuous heating in order to prevent phase separation and "solidification" during
transportation and for efficient discharge.

Aromatic crude oils, for example, Boscan and Tia Juana Pesado, have very much higher kinematic
viscosities than paraffinic crude oils. As such, without heating they may not be able to meet the 250
cSt criterion laid down in Section 9 of the COW Manual and would need to be carried in tanks with
heating colis. Because they do not have the phase separation characteristics of paraffinic crude oils,
aromatic crude oils pose fewer difficulties with regard to any phase separation characteristics when
being transported by sea and during cargo-handling and COW operations. When heating is required
for aromatic crude oils, it will be for viscosity control and pumpabllity rather than to prevent phase
separation and sludge deposition.

In the past, the two characteristics of crude 011 identified as the primary determinants of its carriage
conditions in tankers were its Pour Point temperature and kinematic viscosity. However, as reflected
in the 2000 Edition of the COW Systems publication, and as will be explained in this Guide, the
Pour Point temperature has been replaced by the Cloud Point temperature as the second of the two
primary criteria in respect of crude oil washing.

4.1 The physical behaviour of aromatic crude oils

Like paraffinic crude oils, aromatic crudes also contain the relevant series of hydrocarbons that
would be gaseous, liquid and solid phase products in their own right under normal transportation
conditions. While the gaseous and liquid phases are readily detectable in aromatic crudes, the
partially solidified phase is less noticeable. An aromatic crude oil's true Pour Point temperature
is difficult to determine accurately. As aromatic oils contain only very limited amounts of paraffinic
waxes, attempts to determine this temperature precisely can yield varying results and will depend
upon the analyst's observational skills. Whatever the final result, the real Pour Point temperature for
aromatic oils is normally very low.

Why should this be so? As stated above, aromatic crude oils are temperature-dependant due to
their higher kinematic viscosities compared to those of paraffinic crude oils. As viscosity varies with
temperature, i.e. a log linear variation, the viscosity of aromatic crude oils increases as the oil cools.

Page 7 INTERTANKO - A GUide to Crude Oil Wa~':ihjng and CarQD Heating Criteria - May' 200
 Advanced tank cleaning

Alta Laval can now add tort er value for our customers by offering
complete tank cleaning systems, after acq uiring Gunclean Toftejorg
the world lender in advanced tank cleaning.

With our joint networks wo can guarantee the best systerns and
most efficie nt aftersales support, wherever you are ,

Alfa Laval +46 8 530 650 00 Alta Laval Tank Equipm ent +4b 1\;,55 8699 www al alav, , com
For the analyst, the Pour Point temperature is that at which the oil ceases to move within the testing
apparatus. For aromatic crude oils, the Pour Point is reached as kinematic viscosity slowly becomes
so large that movement within the test sample slows to such an extent as not to be detectable. In
contrast, determining the Pour Point temperature of a paraffinic crude oil is more straightforward
because the point at which the waxes crystallise is relatively easy to observe and creates a more
rapid transition from liquid to near solid state. At their Pour Point temperatures aromatic crude oils
still remain fairly homogenous mixtures, with no observable phase separation.

The phase separation factor is a distinguishing feature when considering the transport of aromatic
and paraffinic crude oils, in particular for their respective heating requirements and COW operations.
Phase separation will be discussed further below when other physical properties of crude oils are
considered. However, the General Guidance as supplied in the revised Section 9 of the COW Manual
with regard to kinematic viscosity should be noted. Poorer pumping performance will commence
with centrifugal pumps when the kinematic viscosity of the cargo exceeds 250 cst. Poorer stripping
performance wili occur when the kinematic viscosity of the cargo exceeds 600 cst. Thus, the crude
oil cargo temperature at which these two operation parameters are reached becomes an essential
pumping criterion for effective discharge of all the crude oil cargo and should be made available in
the cargo's Material Safety Data Sheet (MSDS). These data sheets are now required prior to the
delivery onboard of both oil cargoes and ship's bunkers as from 2 0 ' June 2003 as a result of the IMO
Marine Safety Committee Resolution NO.150 (77).

4.2 The physical behaviour of paraffinic crude oils

Phase separation and the effect of temperature

The examination of the behaviour of numerous crude oils has shown that under certain temperature
conditions extensive sludge/residues can separate from the crude oil. These sludges would, under
normal circumstances, remain onboard tankers after cargo discharge and the mandatory (or charter
party-specified) crude oil washing of the ship's tanks.

Some of the crude oils that experienced such phase separation, for example Gulf of Suez mix, Flotta,
Iranian Heavy, Iranian Light and Kuwait Export Blend (on occasions and subject to the blend), were
not traditionally expected to behave in this manner. Samples of the residues from actual discharges
were obtained for physical examination and analysis, together with numerous samples of a wider
selection of crude oils. The investigation was undertaken in order to gain a greater understanding of
the behaviour and effectiveness of the crude oil as a medium for a COW programme.

This investigation enabled the preparation of guidance as part of achieving the overall objective, as
mandated in the regulations, of cleaning tanks in preparation for the possible loading of baliast and to
control sludge build-up in the cargo tank spaces, not least on segregated ballast tankers. An added
benefit of the investigation, particularly for the cargo owner, is that it provided an understanding of
how to maxlrnise the volume crude oil cargo discharged.

Page 9
The investigation was in order to achieve the following specific aims:

(1) To identify the critical temperature at which a specific crude oil will begin to precipitate sludge
forming hydrocarbons, as noted onboard the various tankers.

(2) To determine a method to calculate the extent of the likely deposition; for the purpose of accuracy
a comparison was undertaken between the observed and measured volumes of sludge onboard
the vessels.

(3) As a result of gaining a better understanding of the behaviour of the sludge, to determine the
Iikeiy effectiveness of COW operations and develop suitable methods for the removal of the
sludge by COw.

It is clear from the investigation that the phase that creates the greatest difficulty for crude oil washing
is that of paraffinic sludge (sometimes, erroneously referred to as "sedirnents'"). Sludge is the partially
SOlidified hydrocarbon deposits of paraffinic waxes and other species from a crude oil that also has
entrapped within its complex structure a variety of hydrocarbon species from the crude oil as a liquid
phase. Predominantly, the composition of sludge is a selection of the various wax species as contained
within a crude oil cargo. Once the sludge has separated it is nearly impossible, given the normal
facilities on board a tanker, to return this waxy material to its original liquid phase within the crude oil's
liquid phase.

In the investigation the deposits/sludges were also examined to determine their flow/pumpability
properties under differing conditions and dilutions. This was done to determine whether it was possible
to develop alternative COW techniques that would efficiently remove the siudge. The temperature at
which the phase separation, or sludge deposition, phenomenon occurs has been termed the Cloud
Point temperature of a crude oil.

5 The Cloud Point temperature determination of a crude oil

The Cloud Point temperature of a crude oil may be defined as the temperature at which the waxes
in the oil change from their liquid phase and become suspended, partially solidified particles within
the bulk of the remaining cargo. In other words, it is the temperature at which a phase separation is
created within the oil.

For transparent oils, such as gas oil, the Cloud Point temperature has traditionally been obtained by
visual examination of samples SUbjected to a cooling programme and the identification of a cloud or
haze of precipitated wax crystals at the critical temperature. It is to be noted that this "new" Cloud Point
temperature criterion is the start ofthe phase separation whereas the traditional Pour Point temperature
is the "completion" temperature for the phase separation.

3 Reference: ISO TR 8338 for relevant definition - see definitions in paragraph 6.2.1.

I~ TERTANKO ..A Guice to Crude Oil Waslling and Cargo Healing Criteria - May 2004 Page 10

Clearly, for operational purposes it is more important to determine the temperature at which a waxy
phase separation starts rather than the one at which it finishes. This is particularly important because
once this phase separation has occurred onboard, tankers do not normaliy have the facilities or
heating capacity to re-Iiquefy the partialiy solidified phase. Once this phase has started, it wili have
developed its own physical parameters, such as its Pour Point temperature, that are different from
the original crude oil's parameters.

5.1 Methods for determining the Cloud Point temperature onboard a tanker

5.1.1 Bondi Test method

The visual technique used to determine the Cloud Point of a transparent oil like gas oil cannot be
utilised for crude oils due to their opalescence. The first method for determining the Cloud Point
temperature of a crude oil that was investigated, the Bondi Test method, focuses on the physical
behaviour of an oil during a controlied cooling programme that could be undertaken safely and easily
onboard a tanker. The method relies upon the detection of the oil's stable temperature transition
through its latent heat loss to partial "solidification".

This method can be accomplished as foliows:

(1) Obtain a representative sample of the cargo. Do not aliow the cargo sample to cool before
commencing the test procedure.

(2) Obtain a glass container (preferably nearly spherical) that is made of heat! temperature resistant
glass and fili the container with the sample.

(3) Insert a thermometer into the centre of the sample volume and heat the sample to a
temperature of at least 30"C above the Pour Point temperature of the crude oil in a hot water
bath. Do not tightly stopper the container during this process.

(4) Prepare an alternative bath with water whose temperature is at least 15"C below the sample's
Pour Point temperature. Have a clock or watch available with a second hand.

(5) Immerse the heated sample container into the cold bath and record at very regular intervals (at
least every 30 seconds) the temperature of the sample.

(6) Graphicaliy plot the sample temperature against time elapse from the start of the test. Determine
the temperatures at which there was no loss of temperature with time elapse. The first such
instance as detected wili be the approximate Cloud Point temperature and the second instance,
if testing time is extended to the required period, will be the approximate Pour Point temperature
of the crude oil.

Upon completion of the test and the plotting of the measured results, the graph wili roughly resemble
the shape shown in Figure 1, overleaf.

Page 11 INTERTANKO -1\ Guide to Crude Oil \iVashing and Cargo Heating Criteria - May 200
 Up to now, a great manypeople in classification circles saem to have thouqht: "Oh, ir s just a tanker. Simple design
simple to class." But in r cent years shipowners have begun to ask questions about the fatigue problems affec ng
bulk oil tankers attar Dilly , few years in service or the problems chemical tankers are conlrontad with when carry
ma certain cargoes. Ma)lIe tankers ar n't mat simple after all. Not surprisrnqlv, the focus IS switching to issues such
as operational durahility and corrosion resistance, and to a tanker's service life. Because at GL w e are applying all
our expe rience and expertise to the serious business of improvmq operational durability and safety standardsfor tan
kers. At GL we have the expertise to help your tankers achieve this goal - over a long service life.

@
Germanischer Lloyd Ak(iengesellschaft Vorsetzen 35 020459 Hamb urg
Phone +49(O)40 -:J61 49-0 Fax +49(0l40-361 49-200 G rmarus h loyd
e-mail: headoffice@g l-group.com . www.gl-group.com p RAT I N 2 4 I 7
Figure 1 shows a temperature drop of an oil sample against time. There are two positions along the
temperature gradient where no apparent temperature loss is observed. However, heat loss (energy)
will be occurring at these positions, which is termed "latent heat". It is these positions that determine
the phase transition from liquid to solid for the paraffinic waxes within the oil sample. The highest
temperature "plateau" is the Cloud Point temperature of the sample (in this case approximately 59'C)
while the lowest is the Pour Point temperature (approximately 43'C).

"'

65

fr--------

55

i -

45 +---+-- - -
1--- 1/;-----1

Time In Minutes

Figure 1 Example of a Bondi Test on a crude oil sample

The Bondi test is recommended for the determination of the Cloud Point temperature of high wax
content opaque hydrocarbons. For these oils with a high Pour Point temperature, the Bondi Test
provides reasonable indications of the critical Cloud Point temperature, as shown in Figure 1.

However, the overall results for all of the crude oils tested did not meet the accuracy requirements of
the investigation, i.e. within 3'C of the equivalent observed Cloud Point temperature.

5.1.2 Primary Cloud Point test method

The second method relies on the potential non-linear (logarithmic) increase in the kinematic viscosity
of the crude oil with temperature decrease, a phenomenon caused by the precipitation of the wa,xl
sludge phase. Although this method sounds complicated, it relies upon a basic dynamic physical
parameter of a liquid which is well-known to most Chief Engineers. The Cloud Point temperature is
that point on the graph where the kinematic viscosity shows an inflection point as the temperature
decreases. This is the point at which the oil sample becomes non-Newtonian or is a mixture of
materials rather than a homogenous substance.

The quidlnq test method used for this determination is the standard kinematic viscosity test method
as used in all oil laboratories. Although this testing procedure cannot normally be carried out onboard
a tanker, a formula has been developed from the numerous results obtained. This formula allows the
Cloud Point temperature to be calculated to a degree of accuracy sufficient to ensure the reliability
of the results for most crude oils.

Page 13 INTERTAr-,IKO - A Guide to Crude Oil Washing and Cargo Heating Criteria - rvlay 200
The results of this developed testlnc method were then applied in the testing of some 73 different
crude oils types for their kinematic viscosities to determ ine their Cloud Point temperatures . Log
Kinemat ic. Viscosity graphs for a selection of these crude oils are shown in Figure 2. As mentioned ,
the indiv idual Cloud Point temperatures may be identified by the inflection points in the various
graphs .

4 ,
Bach Ho
Bo n ny Light
3.5 - ' - Bront Blend
- -- Cabinda

3 'I Daq lng

.- Djano
~.
DraLigen
\~ . - Escravo9
2.5 ,

~
. - E~ Si de r

i0
(.)
, ".\.\ ." I '. "" . '\ .... i,\ .
Foreados
...... Gulf of SU~l
"2
:;; "
. \', .
\.. 1
.....--. \

" .
Kuwait Export
'. lalang
" \ 'Ii'\ "' : :
\. .. . . '. , ',-. -. -"--- """_. ".
- - - Minas
lower Zakum

1.5 .. . : . ..: . . . ... : ., .. . .. .:.;-..:..__ _ _ ..._

. .
..J
Murban
.~ Nemba
- ..- Oseberq
' .. ,...: . Qualhoe
0,5
a - Rab i
- -.&.- - Widuri

o
o 10 20 30 40 50 60 70 80
Temperate \c>;nl igm dn)

Figure 2 Cloud Point graphs for a variety of crude oils

As examples , the Cloud Point and Pour Point temperatures for a number of crude oils , as found by
analysis , are given in Table 1.

Crude Oil Pour Point Temperature C Cloud Point Temperature C

Arab Light - 57.0 10.0
Iran Heavy - 12.0 14.0
Brent Blend _3.0 0 15.0 0
DubaiExport - 60 .0 0 16.0
0
Flotta Crude ..
- 8.0 18.0
Kale Melange o.o- 24 ,0
0
G.S.M . -I- 3.0 26.0
Ardjuna , + 27 .0
0
30 .0
Sarir
I + 27 .0"
c
38 .0"
Bu Attifel + 35 .0 50S

Table 1 Pour Point and Cloud Point temperatures for a variety of crude oils

It was recognised that if the formula is to be capable of use onboard vessels then the input data
to the formula would have to be readily available to a vessel. A rough "rule of thumb" correlation
exists between the Cloud Point and Pour Point temperatures of certain refined petroleum products.

I NTERTAI'4"'~ . alln,..l emf ,[j /l1!).1 Paae 1

For example, for middle distillates there is a relatively stable gap between the two temperatures of
approximately 15'C. However, there is no similar, simple correlation for crude oils.

The Pour Point temperature of a crude oil was thought to be the most suitable starting point, given
its common reference within documents, when inputting data in order to determine a crude oil's
approximate Cloud Point temperature. Thus, the equation capable of predicting a crude oil's Cloud
Point is based on the use of the crude oil's Pour Point temperature (x) as the key parameter.
This equation is as follows:

Cloud Point temperature eC) = 20.2(10.00708'-0.1157714) + 8

where:
x is the crude oil's Pour Point temperature degree C

For ease of understanding, a worked example of the use of this equation for a crude oil that has a
Pour Point temperature of 15 C is:

Cloud Point temperature eC) = 20.2(10.00708'-0.1157714) + 8

= 20.2 x (1010.00708"51-0."57714) + 8

= 20.2 x (1001062-0.1157714) + 8

= 20.2 X (10-00095714) + 8

= (20.2 x 0.9782) + 8

= 19.76 + 8

Cloud Point temperature (OC) = 27.8 0

This equation may be represented in a graphical format for ease of use, as shown in Figure 3.

50~-----

45

40

30

25

2O.jC:'-----t-------+-----+------t------+-------!
-20 10 10 20 30 40
Pour POlnl Temperature (centigrade)

Figure 3 A graph used to determine the Cloud Point of a crude oil

Page 15 lNTERT/l..NKO - A Guide to Cn.de Oil'v".la.<jnmp and Cargo Heating Criteria - May 200
5.1.3 Density method
This third Cloud Point temperature prediction method utilises the parameter of density and is based
on a considerable amount of research. Crude oil density is a parameter well understood by tanker
deck officers and the method can be safely undertaken onboard ship to determine the Cloud Point
temperature, provided the basic equipment is available . This density method was also used in
this investigation of crude oil behaviour to correlate the results with the Cloud Point temperatures
determined by viscometry under laboratory conditions .

A more sophisticated piece of analytical equipment was used during the research to determine the
density for a range of crude oils over a typical carriage temperature range . By plotting densities
against the respective temperatures. an inflection in the graph was noted at the previously determined
Cloud Point temperature. Figure 4 shows the respective results for the Cloud Point temperature of
Bonny Light crude oil obtained by both the density and kinematic viscosity techniques .

0 .B 6 0 . _ . 5 UO

"-50

I
C BEi O
I 400
0 .3 5:;
" -~~".~l--.._~.~
" ', <, I
I
, 3 50

1300 ~
~
I D . r1 S lt y

-I~" ,-
?; a R~O
~ 250 V I$-C O S It y c s !
I - ,-

1 20 0

I'"
:lI::

C> C B" :; 1

~
I
. Ba)

\ ,--_1I I::"
om I 10 10 20 25 30 30

Ternoerat ure (centIgrade )

Figure 4 Density Relationship with KViscosity for Bonny Light Crude Oil Sample No.1 032

Figure 4 Cloud Point temperature for Bonny Light crude oil , 8S determined by density and kinematic
viscosity methods ".

It is believed that the density method provides a viable way of determining the Cloud Point temperature
of crude oil which can be undertaken onboard ship.

The density method procedure is comprised of the following steps:

(1) Obtain a sample of the crude oil cargo .

(2) Obtain a set of standard petroleum hydrometers and thermometers to cover the range of
temperatures likely to be encountered during the voyage .

, The KED line inserted on the graph identifies the temperature at which the density plot is no longer linear whereas the BLUE vertical
line identifies the temperature at which the kinemati c vlscosity IS no longer Newtonian . These two temperatures are the same and
identify the Cloud Point temperature of the crude oil.

'NT ER Tl\ NKO - A GUIde llJ Crw

Pa
7.2 COW pipeline systems

Figure 5 shows a simplified schematic plan of a pipeline system used to deliver crude oil to the COW
machines . An examination shows the following three sources of crude oil supply :

(1) Directly into the main crude oil wash line on deck via a crossover on deck from the main crude oil
discharge pipelines . This is achieved by opening the relevant valves connecting the main crude
oil pipelines with the main cow pipeline on the crossover pipeline.

(2) By isolating one cargo pump from the bulk discharge for the purpose of taking suction on a crude
oil volume in a specific cargo tank . This is achieved by opening the relevant valve on the COW
pipeline crossover (e.g. connecting the green crude oil line) to the COW main and closing the
main green crude oil block valve on deck.

(3) By isolating one cargo pump (e.g. the green pump) from the bulk discharge for the purpose of
taking suction on a prepared and designated crude oil wash stock located in the vessel 's slop
tanks. This is achieved by the same method as above but taking direct suction on the slop tank
using the green pump.

- COW Mil'"
- Red C" rgo Line
- Blue Cargo lInc
- Gfo'n c.~rgo line
- Slop Tank Sucuon hnes
- COW Machlnea

Figure 5 Outline of a COW pipeline system

These differing crude oil supply methods allow flexibil ity when compiling a specific COW programme
in order to achieve optimum effectiveness and cleanliness in the cargo tanks to be washed .

The crude oil supply is delivered into a main fore and aft pipeline running the full length of the ship 's
cargo tanks space; this is the crude oil wash main pipeline. At suitable locations along the length
of this main pipeline smaller diameter transverse pipelines are connected to enable delivery of the
crude oil supply to specific COW machines. The COW pipeline system, as required by the MARPOL
regulations, is a fixed system. In contrast , prior to the usc of COW, the water washing systems
employed onboard tankers were based on the use of portable hoses and machines.

25 I. : E .I ~ ~ . ""I . 'Ill l' I ruu II'. .l I, l I . . ', _ j,

(3) Heat or cool the sample onboard by means of marginal cooling and heating of the sample using
water in a safe environment (being alert to risks posed by crude oil vapour emissions).

(4) Record both the observed density, to at least three decimals, and the corresponding sample
temperature, to at least O.5'C.

(5) Data recorded can be plotted to reveal the inflection point on the plot similar to that shown in Figure 4.

5.1.4 The Cloud Point temperature model

As an alternative to measuring crude oil density at differing temperatures and using the database
of the Cloud Point temperatures obtained by the viscometric method, a mathematical equation was
developed to provide Cloud Point temperatures that fall within the accuracy limits required by the
research programme, i.e. 3'C, for all crude oils. In order to obtain a reasonably simple model, only
the three main parameters that affect the temperature and extent of precipitation of waxes from a
crude oil were used in developing the equation. Although very pertinent to the precipitation of crude
oil sludges, these parameters may not be readily available to all, including ships' crews, who would
wish to determine the value of this parameter.

The three parameters are as follows:

(1) Total wax content of the crude oil.

(2) Melting point/congealing point temperature of the various types of waxes contained in the crude
oil and their respective amounts.

(3) Extent of "solvent" present in the oil which prevents, or inhibits, the precipitation until it is
"saturated".

The equation developed to determine the Cloud Point temperature of a crude oil is as follows:

Predicted Cloud Point temperature eC) = 104.26(lnln F -1.55)

where: F = a + b+ c
2T
and where:
a = 4(W , x M , x Csso
)
b= 2(W2 x M2 x C S09
)

c = W 3 X M3 X C
369
T= L C , 49 , C 232 , C342 , C36 9 , CS09 ' Csso

and where:
W is the wax content (% weight) of the relevant C refining "cut"
M is the relevant melting point of the wax content W
C is the percentage weight of the signified distillation/refining "cut" at
the subscript temperatures

Page 17 !NTERTP.,NKO - A Guide to Crude Oil \/\i3Shi~1~1 and Cargo Heating Cnteria - May 200
Although the equation looks complicated, it can be easily computerised for onboard use. However,
the calculation relies upon the availability and knowledge of the relevant parameters for a crude oil
so that they may be inserted into the equations at the appropriate point. Such parameters may be
found in the more complete types of crude oil assays compiled by oil companies for their internal
use. This calculation method will provide an accurate temperature to which a cargo should be heated
in order to prevent the phase separation/deposition of potentially extensive sludges in a crude oil
tanker's cargo tank.

5.2 The predicted extent of precipitated sludge before COW

One negative result of heating an oil cargo, particularly for a volatile crude oil, is the increased
probability of volumetric loss due to breathing/evaporation of volatile fractions from the crude oil.
The loss of these volatile fractions from the oil will, in turn, reduce the potential "solvency" of the
remaining crude oil for washing purposes. As a result, it is necessary to exercise close control over
the heating of crude oil in a cargo tank, i.e. within a narrow temperature band of say 3'C, during
transport in order to minimise volumetric loss of vapours from the heated cargo and, therefore,
maximise the volume of crude oil discharged.

A further preliminary equation has been developed to predict the likely extent of sludge deposition
from a crude oil given a cargo temperature below the Cloud Point temperature. This equation could
be of use to a tanker's Chief Officer during the preparation of a COW programme, helping with the
answers to questions such as: How many cycles are required? Should the COW medium be heated?
How much COW medium is needed? Should the COW concentrate on a bottom wash only In order
to remove the majority of the precipitated sludges?

The sludge deposition equation is as follows:

Sludge deposited (% weight of total cargo] =WTIc WTIo

(Tc Tpl
where:
T p is the Pour Point temperature eC) of the crude oil
T c is the Cloud Point temperature eC) of the crude oil
To is the observed temperature eC) of the crude oil

WT is the total wax content (% weight) of the crude oil - this unknown can be calculated using the
equation for wax content In section 6.2

The sludge deposition equation has been tested for its accuracy on a number of occasions and has
shown reasonable accuracy, i.e. within 0.075 per cent of the true/observed residual onboard (ROB)
volume percentage. As a recent example, a tanker which discharged a 337,000 m' crude oil cargo
at an average temperature of 20SC, and whose cargo Cloud Point temperature was 24'C, was left
with an ROB volume of 1,333 m' prior to COw. The deposition model predicted an ROB/unpumpable
volume of 1,146 rn", or 0.34 per cent of the cargo volume. The actual ROB volume observed was
0.39 per cent of the full cargo volume.

INTERTANKO -A Guide to Crude Oil Washing and Cargo Heating Cnleria - May 2004 Page 18
6 The practical implications of the Cloud Point determination

6.1 MARPOL regulations and commercial implications

As a result of the introduction of MARPOL 1973f78, crude oil tankers are obliged to undertake the
cleaning of a selection of their tanks utilising their crude oil cargo during its discharge for either the
subsequent carriage of ballast water or de-sludging the cargo tanks. Within the commercial world
of chartering and trading of vessels/cargoes this procedure was originally deemed to be financially
advantageous given that all the vessel's cargo, save for, perhaps, up to 0.1 %, would be discharged
including the traditionally unpumpable material.

In return for these increased out!urns, tanker owners were normally allowed additional time of, typically,
12 hours to accommodate cargo discharge along with the COW procedure. However, today many
standard, modern charter party forms allow substantially less additional time and only when COW
is specifically requested over and above the mandatory minimum to comply with MARPOL 73/78.
The following charter party examples highlight the current state of play:

(1) BEEPEEVOY 4, clause 19, allows additional time, only when additional COW (additional to the
MARPOL minimum) is requested, at 25 per cent of the total discharge time stipulated for the
pumping warranty, e.g. 6 hours for a 24-hour discharge. No time is allowed for the basic MARPOL
minimum COw.
(2) SHELLVOY 5, clause 20, allows additional time if COW is required "by charterers or any other
competent authority". The question remains: Which authorities fall within the definition of 'a
competent authority'? For example, is IMO a competent authority? If so, should undertaking the
mandated minimum COW to comply with MARPOL be compensated with additional time? The
time allowance stipulated by the clause is 0.75 hours per cargo tank washed.

These charter party examples of additional discharge time would enable a tanker operator to carry
out one cycle of washing of the cargo tanks to render them clean and acceptable for the receipt of
ballast; to desludge the requiredlrotation tanks for that loaded voyage; and to maximise the discharge
volume. In the case of the SHELLVOY charter party the extra time allowance would only compensate
for the normal time required for a bottom wash with minimal stripping.

Given the possibility of sludge deposition from a crude oil cargo, it has been necessary to consider
the likely effects/efficiency of a typical crude oil washing programme for the removal of the deposited
sludges generated when the crude oil's temperature drops below the Cloud Point temperature. As
an extreme example, the extent of deposition witnessed and measured when the oil temperature
drops below the Cloud Point can be greater than 10,000 barrels of oil on a single laden voyage for
the larger crude oil carriers. However, before considering the efficiency and expectations of COW,
given the foregoing time criteria, the physical criteria of the crude oil should be reviewed to discover
whether the oil is suitable for the washing process.

The amended requirements of Section 9 of the COW Manual remove from consideration the concept
of the "SUitability" of a crude oil for its purpose as a washing medium. All crude oils within the MARPOL

Page 19 Ir--HERTA.NKO - A Guide to Crude Oil lNa::.hing and Carqo lteatinq Criteria - May 200
 definition would be suitable as a washing medium, provided the crude oil or the proposed washing
stock can be pre-treated to meet the physical requirements stated above. However, with respect
to certain crude oils, particularly the aromatic types due to their individual behaviour, it is doubtful
that significant additional benefit can be accrued from crude oil washing over the traditional efficient
stripping of the cargo tanks.

In addition, if the tanker is offered a crude oil wash medium from ashore, it is necessary to check the
following aspects

- Is it to be delivered heated and, if so, at what temperature?

- What volume or tonnage is available?
- What is the quality of the material being offered?
- Is the crude oil wash material compatible with the likely residues to be found in the vessel's
tanks? If the oil is aromatic, what impact will it have on paraffinic sludge residues?

It has been stated that in order for a crude oil to remain in a "totally" liquid phase a temperature above
the Cloud Point temperature must be maintained. It is important to ensure that this temperature
requirement is observed in order to avoid havinq to use a heterogeneous crude oil mixture for crude
oil washing. Such a mixture contains an amount of quasi-solid suspended or phase-separated
sludqe. The behaviour and characteristics of this sludge, which could accumulate in a vessel's tank
or COW piping system, are very different from the "mother" crude.

The efficiency of crude oil washing with a liquid cargo containing separated sludge will not be
as good as that with a "totally" liquid phase crude oil because the reduced solvent powers of the
heterogeneous crude oil mixture make the removal of deposited sludges more difficult. Laboratory
simulations have shown that liquid cargo containing separated sludge has a limited impact as a
solvent. This is to be expected as crude oil in such a state and at such a temperature, below the
Cloud Point temperature, would be in the process of precipitating some of its own constituent waxes
into a suspended sludge phase.

With such high Pour Point temperature materials in suspension, i.e. typically between 45'C and
60'C, when a paraffinic crude oil's temperature drops below the crude oil's Cloud Point temperature,
additional deposition in the tank being washed is probable.

The provisions of the additional clauses usually contained in tanker charter parties, often entitled
"cargo retention" or "cargo delivery" clauses, place a responsibility on the vessel's owners for the
discharge of the entire tank contents, including residues, unless it is commercially or physically
impracticable to do so. Failure to comply with such provisions will result in the payment of the value of
the "retained" quantity by way of a deduction from freiqht". The normal commercial criteria associated
with a balanced set of charter party clauses relevant to this topic are as follows:

(1) The tank residues are liquid, or should have been removable by efficient COW and/or maintaining
the specified/required heating of the cargo.

5 Harvey Williams, "A Guide to Tanker Charters" INTERTANKO 2001.

lNTERTANKO - A Guide to Crude Oil Waslling and Cargo Heating Criteria - ~"lay 2004 Page 20
 (2) The residues are reachable by the vessel's pumps; i.e. they will flow to the aft end of the vessel's
tank and the pump, through the length of suction piping' , with the energy induced by the trim of
the vessel.

(3) The residue is pumpable by the vessel's pumps.

(4) The circumstances are certified, where appropriate, by an independent surveyor.

In order to meet the foregoing commercial requirements, three primary criteria must be present. In
order for the COW programme to succeed and extensive volumes of sludges and oils to be removed
from the cargo tanks, it may be necessary to bring the three criteria into play simultaneously.

The three conditions which must be met for a successful COW operation are as follows:

(a) The crude oil wash medium must have adequate residual solvency. This can be achieved by
heating, and avoiding the use of sludge contaminated wash medium i.e. extensive recirculation
during "closed cycle" crude oil washing. Typically, the minimum temperature required for crude
oil washing of a precipitated paraffinic sludge phase would be 30C.

(b) The crude oil wash jet must be applied with sufficient force to move the settled sludge phase.
Therefore, the COW pressure must be closely monitored, particularly at the forward end of the
crude oil wash supply pipeline. The input of warm "fresh" crude oil will help to create a dilution
by adding a liquid phase to the static sludge phase to change the crude oil's overall "dynamic
behaviour",

(c) The obtaining of a suitable trim of the vessel to induce flow to the suction positions in a tank.
Subject to stress and bending moment restrictions, a minimum trim of 8 metres over the length of
VLCC is normally recommended.

6.2 The crude oil washing of sludge and some practical implications

What is sludge?
In ISOITR 8338:1988 standard definitions are given for both "sludge" and the more commonly, but
potentially erroneously, used alternative term of "sediment". These definitions are as follows:

Sludge

That element of material in a ship's cargo tank that is essentially not free-flowing. It consists of

hydrocarbon waxes and may contain waterloil emulsions and sediment.

Sediment

That element of non-free-flowing material left in a ship's tank which is essentially inorganic in nature,

for example sand, rust particles, etc. It is not soluble in hydrocarbon oil.

6 Net Positive Suction Head (NPSH) explanations - see "An Explanation and Guideline for Pumping Calculations"
- INTERTANKO, March 2001.

Page 21 !NTERTANI<O - A Guide to Crude Oil Washing and Cargo HeatlrJ[J Criteria - f\..1ay 200
Given the foregoing and before considering the dynamic behaviour of sludge, it is important to
record the physical properties of sludqe so they may be related to the ISO definitions above. The
various sludges, as sampled from cargo tanks upon completion of a crude oil wash programme,
normally exhibit Pour Point temperatures in the 45'-60'C range, although very much higher Pour
Point temperatures have occasionally been observed, irrespective of the paraffinic type of crude oil.
A set of test methods and physical properties of a typical sludge can be found in Table 2 below.

Test Method Result

Density @ 15C kg/I I.P.189/190 0.9225
Wax Content %(m/m) UOP 46/85 mod 25.51
Pour Point C I.P.15 + 66
Asphaltenes %(m/m) I.P.143 0.42

Table 2 Test results and physical properties for a typical sludge

Due to the difficulty in accurately determining the total hydrocarbon wax content of an oil by analysis,
a model has been developed which correlates the total wax content of a crude oil with its Pour Point.
The various analyses that have been undertaken have shown that the percentage of wax normally
determined is only a proportion of the total wax in a crude oil sludge. In addition, only a proportion of
the paraffinic types of wax and none of the non-paraffinic waxes have been detected.

The formula for determining the total hydrocarbon wax content of an oil is as follows:

Log (wax content % weight) = 100.00708,.0 115n'4

where:
x is the Pour Point temperature of the oil in 'C

This formula indicates that the associated wax contents of sludges having the Pour Point temperatures
quoted above would be in the 33-54% range. The balance of the sludge volume is normally comprised
of entrapped oils, precipitated asphaltene/polycylic aromatic hydrocarbons (PAH), water and other
foreign matter (rust and sand) in the wax-rich crystalline matrix. Depending on the extent of water
present in the sludge, a wax-rich "tight" emulsion with high dynamic viscosities is created.

The dynamic behaviour of sludge does not correspond to the normal behaviour of highly viscous
substances. When flow is induced by a force on highly viscous substances, i.e. their yield stress
has been exceeded, in normal circumstances such substances will continue to flow thereafter. The
yield stress of a substance is the necessary force that has to be applied to the substance to achieve
deformation due to failure of its structure.

In contrast, waxy crude oil sludges are so called "memory liquids". In other words, although the
required force is applied to the sludge to promote flow, after a short period the sludge will re-adopt
its original form and cease flowing. It remembers its original molecular form! This will clearly create a
problem when attempting to remove sludge from a vessel's tanks because, as the COW jet rotates
around the tank, it is applying force for only a moment at a specific point in the tank.

INTERTANKO ~ A Guide to Crude Oil Waslling and Cargo Heating Criteria - May 2004 Page 22
Preliminary investigations to determine the actual dynamic behaviour of sludge samples when
simulating the practical conditions onboard a tanker revealed important physical data. Not least, the
yield stress of these precipitated sludges at the corresponding environmental temperature was in the
order of 1,000 to 3,000 Pascals.

In addition, the dynamic viscosity of one sludge sample was found to be as high as 2,000,000
centipoise (mPa s). With a density of 1,000 kg/m3, this dynamic viscosity would be equivalent to
2,000,000 cSt at the simulated environmental temperature. With such yield stresses and dynamic
viscosities it is clear that these sludges, on their own, could not be classed as liquid, pumpable or
reachable by the vessel's pumps. In other words they would not readily flow under normal gravitational
forces, as experienced when the ship is trimmed, to the pump suction.

A substance with these types of properties is similar to that of a "wet concrete". With these properties
the programming for the crude oil washing of a vessel's tanks to remove the deposited sludge and
its oils must be carefully considered. This exercise must take into account the physical and chemical
properties of the wash medium and the energy input options available to the tanker operator.

7 COW equipment and procedures

In Section 2 the basic objectives and relevant equipment for crude oil washing, as specified in COW
Manuals, are listed. Previous sections in this Guide have described the behaviour of the various
crude oil wash media. This section will cover the types of equipment and cargo systems typically
used in COW operations, as well as the procedures/programmes most commonly employed to
carry out COW. The following descriptions will not be appropriate for the equipment and procedures
employed on every crude oil tanker, simply because of the vast array of differing types of equipment,
cargo system designs and pipeline systems available on the market. Instead, generalisations are
made in order to provide a simple explanation of the key COW equipment and procedures.

7.1 COW machines and tank cleaning nozzles

COW-specific equipment found on a tanker comprises the crude oil washing machines and their
"guns". The different types of COW machines available fall into one of three main categories, i.e.
programmable, non-programmable and submerged machines. The first two categories are deck
mounted equipment with the tank cleaning nozzles, or guns, suspended on fixed, rigid pipes about
5 metres below the vessel's main deck. The submerged machine is tank bottom-mounted and is
non-programmable but "spot-located" to assist in the washing of shadow areas of the tank where
direct impingement of the main jet from the deck-mounted guns is deficient. Operation of submerged
machines is similar to that of deck-mounted, non-programmable machines.

Programmable machines allow the operator to decide on the scope of movement of the washing
gun and the rate of rotation of the nozzle in the cargo tank. The scope is controlled by the angle
of elevation of the gun, with O' being directly beneath the gun position in the tank. Therefore, for a
bottom wash cycle, for example, the elevation of the gun will be limited to, say, 40', and the cycle

Page 23 INTERTANKO - A C~Jld8 to Crude Oil Washing ario Cargo Heating Cntena ~ May 200
will be determined as being say, 0'- 40'- 0'. likewise, with a side or top wash the scope of elevation
of the crude oil wash machine will be set at, say, 40'- 110'. Although programmable machines offer
a number of advantages (these will be discussed below), such machines are more expensive and
more maintenance intensive than the non-programmable alternative.

As suggested by their name, it is not possible to control the wash programme with non-programmable
machines. The starting location of the machine nozzles is normally unknown and the operator has
to assume that the whole tank has been washed in the operation manual's stipulated time for one
complete cycle of the machine (bottom and side/top wash).

Another major difference between the two types of equipment is the number of nozzles per machine.
A typical programmable machine has one nozzle of between 25 and 35 mm diameter whereas non
programmable machines have two nozzles of smaller diameter. Some new programmable machines
have a smaller diameter nozzle. It is said that such nozzles provide a more effective COW operation
and require less motive pressure to adequately operate the machine. However, a smaller nozzle
does suggest less COW medium is used in the COW operation and this will therefore have an impact
upon the extent of solvency available to remove the precipitated sludges.

The final major difference between the types of COW equipment is the manner in which the machines
and their nozzles rotate. The different approaches are described in the following paragraphs:

Programmable machines
The machine head rotates horizontally, with the nozzle being set at the programmed angle. With
each completed horizontal rotation, the angle of the single nozzle is either lifted or depressed in a
vertical direction by the programmed number of degrees. This movement of the machine head and
nozzle creates a form of helical pattern with tighter rotation tracks closer to 0' (vertically below the
machine). The speed of rotation of the head is determined by either the crude oil supply pressure or
a compressed air-driven motor attached to the machine's control mechanisms.

Non-programmable machines
These machines have both a horizontal rotation of the machine head and a near vertical rotation of
the twin (directly opposing) nozzles. As the head of the machine rotates horizontally, the twin nozzles
rotate in the near vertical plane, thereby creating a form of hatched lattice type wash pattern for both
butkheads, deckhead and tank floor simultaneously. Such machines have no method of control, save
for the supply valve for crude oii and the supply pressure that controls the rate of rotation.

The number of machines in each cargo tank varies and is dependent upon the extent of internal
structure within the tank and the size of the tank. Typically, the "effective" jet length from a nozzle,
operating at its correct supply pressure, would be about 40 metres. Therefore, nonnally each tank is
provided with 2 or 3 machines so that the extent of shadow area (non-contact or impingement area)
is limited and meets the requirements of the MARPOL requlations.

INTERT.~NKO - A Guide to Crude Oil Washing and Cargo Healing (>te,.ia . May 2004 Page 24
7.2 COW pipeline systems

Figure 5 shows a simplified schematic plan of a pipeline system used to deliver crude oil to the COW
machines . An examination shows the following three sources of crude oil supply:

(1) Directly into the main crude oil wash line on deck via a crossover on deck from the main crude oil
discharge pipelines. This is achieved by opening the relevant valves connecting the main crude
oil pipelines with the main COW pipeline on the crossover pipeline.

(2) By isolating one cargo pump from the bulk discharge for the purpose of taking suction on a crude
oil volume in a specific cargo tank . This is achieved by opening the relevant valve on the COW
pipeline crossover (e.g. connecting the green crude oil line) to the COW main and closing the
main green crude oil block valve on deck.

(3) By isolating one cargo pump (e.g. the green pump) from the bulk discharge for the purpose of
taking suction on a prepared and designated crude oil wash stock located in the vessel's slop
tanks. This is achieved by the same method as above but taking direct suction on the slop tank
using the green pump.

- GOW Mall'
- Red c aroo Line
- Blu e Cargo Un e
- Gw en Cargo Line
- Slop Ten~ Suction lines
- COW Mach ines

=--------------
.-
From
} Cargo
Tanks

Figure 5 Outline of a COW pipeline system

These differing crude oil supply methods allow flexibility when compiling a specific COW programme
in order to achieve optimum effect iveness and cleanliness in the cargo tanks to be washed .

The crude oil supply is delivered into a main fore and aft pipeline running the full length of the ship's
cargo tanks space; this is the crude oil wash main pipeline. At suitable locations along the length
of this main pipeline smaller diameter transverse pipelines <=Ire connected to enable delivery of the
crude oil supply to specific COW machines. The COW pipeline system , as required by the MARPOL
regulations, is a fixed system . In contrast, prior to the use of COW, the water washing systems
employed onboard tankers were based on the use of portable hoses and machines .

Page ~J ~ r 'A .I .' ) J '. IJ . l J JIlin il .1 '. 'ill. I J. I ut . . I .

7.3 Stripping systems

COW performance and its effectiveness also depend on a well-designed and efficient stripping
system. This is used for the removal of the crude oil wash medium from the washed tanks together
with any recovered oil/sludges. There are two basic types of stripping system, namely the main
cargo pump "vac strip" or similar system and the cargo eductor, sometimes referred to as "ejector".

Vac-strip systems operate by applying a partial vacuum to the "suction" side of the vessel's main
centrifugal pumps in order to evacuate any gases or vapour in the pump and aid the pump in gaining
or maintaining suction on the residue liquids within the relevant tank' . Alternative systems to the vac
strip approach generate a column or head of liquid in a chamber close to the cargo pump, ensuring
that the pump is "primed" at all times. This type of stripping system uses the main cargo pipeline
system and suction, which may be located as much as 150 mm above the tank floor. However,
certain pipeline designs also have a separate stripping suction connected to the main pipeline that
is only about 50 mm above the tank floor. With a suitable trim this type of stripping system will
efficiently drain a tank of liquid leaving only small amounts of liquid residues in the aft bay of the tank
as allowed for under paragraph 4.4.4. of the tanker's COW Manual.

The eductor obtains a very efficient form of suction using Bernoulli's principle. The principle of an
eductor is that a fluid is pumped past a restriction somewhere along the length of the pipeline,
thereby causing the rate of flow to increase. At the point of the restriction an alternative pipeline is
inserted such that its open end is in the direction of flow of the liquid. At the open end of the inserted
pipeline a vacuum is generated, thereby creating a suction within the pipeline. Crude oil, supplied by
one of the main cargo pumps, is used as the driving fluid for the operation of the eductor. When using
an eductor, it is essential that the pressure of the driving liquid entering the eductor is high enough,
i.e. normally about 11 kg/cm 2 , to create the design suction.

This system differs significantly from the action of a centrifugal pump as it supplies a suction to the
stripping system. Contrary to popular belief, cargo pumps do not have any suction capacity and only
apply additional energy to the liquid entering the pump. Therefore, the liquid entering the pump must
flow there due to an external head (energy source) overcoming the net positive suction head (NPSH)
of the suction pipeline systems.

Eductors have two additional major advantages over a conventional pumping/stripping system, i.e.

(1) Even though suction from a tank may be lost or the eductor begins drawing in vapour or gas,
the eductor will continue to operate without danger or the risk of damage. The eductor has no
moving parts and does not have to be transferred immediately to an alternative tank containing
liquid.

(2) The suction supplied by an eductor system operating in its design mode is very efficient and will
draw not only liquids into the system but also higher viscosity substances that are very near to
the tank suction point.

7 Reference: "An Explanation and GUideline for Pumping Calculations" -INTERTANKO, March 2001.
B Reference: "An Explanation and Guideline for Pumping Calculations" - INTERTANKO, March 2001.

INTERTAW<O - A Guide to Crude Oil \Nashing and Cargo He8ting Criteria - MaV 2004 Page 26
Having briefly discussed the main COW equipment, it is necessary to discuss the commonly used
COW techniques. In broad terms there are two very dissimilar techniques, i.e. concurrent washing
(with discharge) and closed cycle washing.

7.4 Crude oil washing concurrent with cargo discharge

COW concurrent with cargo unloading is undertaken during the main discharge programme. The
method utilises the crude oil stream being discharged, together with its associated back pressure,
to supply the COW pipeline with crude oil. It can be achieved either by a "bleed-off' system from
the main crude oil discharge pipelines for supply to the COW pipeline (see Figure 5 and the deck
crossover to the COW pipeline) or by separate supply using a designated cargo pump. If supply is
achieved by the bleed-off from the bulk discharge, then sufficient back pressure has to be maintained
at all times. The normal minimum back pressure required for the operation of the COW guns is 8
kg/cm'.

This procedure cannot be properly used by tankers equipped with non-programmable COW machines
if a full crude oil wash programme involving all the vessel's tanks is required. This is due to the lack of
controllability of the nozzle and the inability of a non-programmable machine to wash selected areas of
the cargo tank to be washed. The programme will commence when the first tanks to be washed are at
about one-half to two-thirds discharged. This will represent the top or side wash in the relevant tank.

Advantages of the COW concurrent with cargo discharge method are as follows:

(1) It does not necessarily require a designated pump to supply the COW system and, therefore,
it does not significantly slow the overall discharge rate unless one main cargo pump is re
designated for the wash programme.

(2) It utilises "clean" crude oil throughout the programme as the crude oil is used only once in the crude
oil wash programme. Thus, any recovered sludge or tank washings are more homogeneously
mixed with the bulk discharge rather than consolidated for discharge at the end of the discharge
programme as a high Pour Point temperature mixture.

NOTE: High Pour Point temperature crude oil mixtures can block pipelines/COW systems in adverse
temperature environments.

(3) As the crude oil is used only once in the programme, the "solvency" effect of the crude oil and its
lower kinematic viscosity assists the potential for the removal of precipitated paraffinic sludges
(see further discussion below).

However, COW concurrent with cargo discharge is not without its disadvantages, as follows:

(1) The programme requires a significant amount of planning and discipline to ensure that
sufficient crude oil is present to complete the required COW programme without stopping the
bulk discharge.

Page 27 INTERTANKO A Guide to Crude Oil Washing and Cargo Heating Criteria- May 200
 (2) The programme requires continuous monitoring of tank levels, particularly slop tanks, to avoid
overfilling. Slop tanks receiving the stripped washing medium and recovered sludge should be
discharged ashore at reasonable time intervals to avoid significant loss of safe/spare Ullage.

(3) The efficiency of the COW system and the satisfactory operation of the COW cleaning machine
nozzles rely upon the COW pipeline pressure. If the shore back pressure is below 8 kg/cm',
then increased back pressure on the main discharge pipeline delivering to the COW pipeline
would have to be achieved or the COW programme must be stopped as required Within the
COW Manual guidelines. This is normally undertaken by "throttling", or partially closing, the deck
pipeline master/block valve to increase the pressure at the "bleed off' position to the COW main
(see Figure 5 above). However, the action of throttling may cause damage to the valve or its
seating if it is not designed for such purposes.

(4) This type of programme produces more volumetric loss of hydrocarbon vapour from the
cargo, because of the continuous use of "fresh" crude oil. Therefore, an apparent increase in
retained vapours onboard may be recorded, particularly for the more volatile types of crude
oil, i.e. those with a Reid vapour pressure (RVP) in excess of 6 psia" , thereby generating the
possibility of an increased outturn loss of volume.

(5) When heated wash material is required in order to comply with guidance notes in Section 9
of the revised COW Manual and to assist with the removal of precipitated sludqes, then this
method may not be usable if the cargo is transported and discharged unheated.

7.5 "Closed cycle" crude oil washing

The closed cycle COW method utilises a fixed and prepared (de-watered) volume or stock of crude
oil washing liquid, usually contained within the vessel's slop tank(s). One of the tanker's main
cargo pumps is used both to supply this oil to the main COW pipeline and to drive the eductor
for the simultaneous stripping of the washed tanks. The tank strippings are returned back to the
slop tank COW stock and this standard stock, together with washing recoveries, is continuously re
used throughout the total COW programme, or for as long as washing efficiency is maintained, in a
closed cycle. The content of the slop tank/s is finally discharged ashore at the end of the discharge
programme.

The advantages of closed cycle COW are as follows;

(1) The programming of the crude oil washing is very much simplified and does not require
judgement regarding the rate of discharge in order to be sure that sufficient crude oil
wash stock remains within the bulk to be discharged in order to complete a required COW
programme. The exception is when the existing wash stock needs replacing due to a drop in
efficiency of the COW performed. This is mainly due to the extent of "contamination" of the wash
stock by recovered sludges.

9 Reference: T J Gunner, "The Physical Behaviour of Crude Oil influencing Its Carriage by Sea", ISBN 827860-042-2. It was found by
analysis of data that vapour loss expectancy increased when a crude oil had an RVP over 6 psia.

INTERTANKO - A Guide to Crude Oil Washing and Cargo Heating Criteria - May 2004 Page 28
(2) Overfilling of the vessel's slop tanks is much less likely, subject to the extent of residue recovery
from the COW programme.

(3) Given that a vessel's pump is designated as the motive power for sole delivery to both the
COW system and for the eductor drive, closer control over delivery pressures can be aintained for
optimum efficiency of the COW programme and eductor drive pressure (stripping efficiency).

(4) The time allowed by the charter party for COW can be cioselyadhered to and thewash programme
modified or stopped at any time.

(5) Less vapour loss will be created from the cargo volume because the volumetric vapour loss will
be mainly from the COW stock only and not the bulk cargo volume.

(6) Pre-treatment of the COW stock can be accomplished if heating of the crude oil over its ambient
temperature is required for an efficient COW programme. This condition is subject to the proper
functioning of the heating coils in the slop tank(s).

There are disadvantages to the use of closed cycle COW, as follows:

(1) Because one cargo pump is dedicated to the COW programme, the tanker's full pumping
capacity is not available. Therefore, the overall discharge rate will be reduced, particularly when
undertaking "top washing" using this technique.

(2) If the fixed stock of crude oil is required for an extensive wash programme and particularly for
the recovery of large amounts of paraffinic sludges, then its efficiency, both as a solvent and
for retaining the sludges in suspension, will be quickly reduced.

(3) The crude oil wash stock may become very wax-rich and have an associated high Pour
Point temperature. In cold environments, therefore, there is a risk that the COW pipeline and
machines could become blocked with precipitated sludges, impairing the efficiency of the COW
programme.

(4) The efficiency of the fixed COW stock may be reduced by its "saturation" with sludges, and the
last tanks to be washed might not benefit from the COW operation. In such circumstances it
will be necessary to obtain a fresh "charge" of crude oil. There could be associated delays in its
preparation, especially if it is required to be heated, before COW can be resumed.

8 Crude oil washing of sludges

8.1 The theory and background

The primary objective when removing deposited sludge is to reduce the viscosity of the sludge to a
pumpable state that is also capable of adequate flow to the suction positions within the tanks. The

Page 29 INTES:TANKO - A Guide to Crude O:l \Nashing and Cargo Heating Criteria - May 200
controlling parameter in seeking to achieve this goal is the crude oil's kinematic viscosity that is used
to blend and dilute the sludges to an acceptable viscosity for pumping. In order to strip the cargo
tank of sludge efficiently, the sludge's viscosity has to be reduced to below 600 cSt. This is roughly
equivalent to the viscosity of a heavy fuel oil at a temperature of about 40"C.

To assist in achieving the objective of effective tank cleaning, the concept of "solvency" of the crude
oil wash medium upon the sludges and "dilution" of the sludges by the COW medium should be
considered. This is the primary way to reduce the sludge's viscosity to an acceptable level to enable
its collection from the cargo tank for ultimate discharge.

The "solvency" powers of the COW medium upon sludges depend upon the crude oil retaining an
adequate quantity of its lighter, more volatile components. These lighter components act as solvents
on the heavier components to be found within the precipitated sludges, reducing the viscosity of the
overall mixture. For example, tanker men know that kerosene is a good medium for cleaning greasy
or "oil-caked" instruments. More volatile hydrocarbon liquids, such as "white spirit", make even better
cleaning fluids. However, the volatility of the components in a crude oil which make good solvents
demands care and attention, especially when the crude oil wash medium is heated in preparation
for its use. Under such conditions, the highly volatile components may escape as vapour, reducing
the solvent powers of the crude oil wash medium and, hence its effectiveness as a cleaning agent.
Therefore, it is important that, when planning the proposed COW programme and pre-treating/
heating the COW stock (assuming a "closed cycle" programme), a balance is obtained between the
benefits of heating the COW medium and the potential loss of solvency of the medium. It should be
noted that the application of heat may melt the paraffinic waxes binding the sludge mass, but once
the heat energy is dissipated throughout the sludge mass, then the waxes will reform. Under such
circumstances, there is generally little to be gained by overheating the COW medium, say beyond
45"C, as a loss of solvency of the COW medium will ensue. However, with certain specific crude oils,
due to their characteristics, this general "rule of thumb" does not apply.

The concept of "dilution" is more closely related to the physical reduction of the viscosity of the
sludge rather than the dissolving of the paraffinic waxes in the context of "solvency". In the crude oil
washing of sludges dilution should be considered in two different ways. By introducing liquid phase
crude oil under pressure into a tank containing precipitated paraffinic studqes, the yield stress of
the sludges in way of the crude oil jet will be exceeded due to the force of impact of the jet upon the
sludges. This action starts the movement and break-up of the sludge mass and its form.

Having achieved "movement" of the sludge mass, the next objective is to "stir up" the precipitated
sludges into the crude oil wash volume, as introduced by the COW jet, and allow the combined
slurry, which is now of reduced viscosity, to flow aft to the tank suction pipeline. However, given the
viscosity of the crude oil wash medium, the amount of crude oil required to undertake this "blending"
can be significant, depending upon the volume of sludges in the tank. Further information regarding
this aspect will be discussed below when suggested alternative techniques for the crude oil washing
of sludges are considered.

INTERTANKO - A Guide to Crude Oil \Jilashing and Cargo Heating Criteria - May 2004 Page 30
8.2 Practical circumstances affecting the crude oil washing of sludges

The two sets of circumstances in which dilution may assist with the crude oil washing of sludges
relate to the potential delays before COW can be undertaken in a specific tank. Such delays may be
caused, for example, when the stripping suction shares the main pipeline system to the cargo pumps.
Thus, if the main cargo pipeline is used for bulk discharge from alternative tanks in the same pipeline
group of tanks, it would not be possible to use the stripping suction for the tank to be washed.

(1) If significant volumes of sludges are believed to exist in a tank due to the cargo temperature and
the calculated Cloud Point temperature of the cargo, the sludge volume can now be estimated
using the equation given earlier. If the tanker operator expects large volumes of sludge to be
generated, contrary to the general advice given in a COW manual, the tank bottom should not
be stripped dry prior to commencement of COw. A volume of the original cargo should be left in
the tank to assist with the slurrifying of the precipitated sludge phase.

(2) If there are likely to be significant delays before COW is commenced, e.g. because
of a two-port discharge schedule, then, given the high Pour Point temperatures of typical
paraffinic type sludges as shown above, it would be more beneficial, particularly in a cold water
environment, if the relevant cargo tanks are not stripped to dryness in the primary discharge
programme as normally suggested in a COW manual. A volume of the original cargo should
be left in the tank to maintain a slightly higher overall Pour Point temperature for the mixture
contained in the tank. This will be beneficial and limit the time required to undertake an efficient
COW programme in the tanks.

The motion of a crude oil wash machine will not always assist with the objective of removing sludges.
The motion of a programmable crude oil wash gun operating in a tank, especially when undertaking
a bottom wash, is best described as helical. As the machine rotates through its horizontal arcs, for
50% of the arc the applied force from the crude oil wash jet upon the sludges pushes the sludges
away from the desired position, i.e. the suction pipeline position in the aft bay of the tank. This action
decreases the effectiveness of the operation and needs to be considered when designing a COW
programme for the removal of extensive sludges.

Sludge's behaviour will vary with its temperature. As paraffinic wax is a relatively good temperature
insulator, the behaviour of sludge is not constant throughout the depth of a sludge volume. Within the
vertical profile of the sludge, the closer the sludge volume is to the colder outside surface of the tank
the more viscous the material will become. The surface of the precipitated sludge in contact with the
liquid phase oil will be only slightly viscous and can be removed by one cycle of crude oil washing.
This observation applies to roughly 80% of the depth of a precipitated sludge volume, the actual
percentage depending upon the environment temperature and the temperature of the original crude
oil cargo. The remaining 20% of the sludge volume close to the external surface of the tank, which is
in contact with either the sea in single-hull tankers or ballast water double-hull tankers, will have the
highest viscosity. Therefore, a precipitated sludge volume is not a homogenous volume exhibitinq
standard behavioural characteristics throughout its depth.

Page 31 INTERTA~II\O - A Guide to Crude Oil Washin[J and Cargo Heating Criteria - May 200 .
With a normal set of operating parameters for a programmable COW machine on a large crude oil
carrier, in one second of operat ion the crude oil jet will be directed at approximately 1 m2 of tank
surface and approximately 30 litres of oil will be used in the cleaning of that area . Calculations were
undertaken based upon the crude oil washing equipment operational criteria to establish the rate of
input into a tank per machine per second . These calculations established a possible maximum dilution
of 19% for a depth of sludge of 15 centimetres for each transit of the washing jet over an area of 1
mL .

Given the foregoing calculated dilution, the effects of solvency of fresh crude oil upon the sludges
were visually observed during laboratory testing when creating differing dilutions/blends of the sludges
with fresh crude oil. The blends constructed were for 10%, 20%, 40% and 60% at a temperature of
60C in order to represent the dilution of sludges for 1, 2 and 3 washing cycles . The blends were ,
thereafter, cooled at a rate of 1' C per hour to just above the prevailing ambient sea temperature of
13C (a possible sea temperature at the time of COW). The viscosity of the mixtures was thereafter
o-15
continuously measured within the temperature range 10 'C. The results of this work can be seen
in Figure 6.

It was observed that, during the slow cooling programme , the sludges precipitated from the fresh
crude oil. This was to be expected when the crude oil reached a saturation temperature for the
additional types of hydrocarbons introduced into the mixture . At the lower test temperature of 10C
the fresh crude oil used in the testing provided little , if any, solvency effect.

The dynamic viscosit ies of the three dilutions /blends showed , in the main , significant variation for
flow and pumpability purposes from those measured for the unblended sludge. The results ranged
from 7,000 mPa s to 100 mPa s at 13'C , depending upon the extent of dilution . It was concluded that
even the tested blends, if created in the vessel 's tanks , would not flow to the pump suction position
after one or two washing cycles .

1600 0 r--...,....----,------,------,..-----,,....---;---...,....--~-____,_-____,_-____..,-_____,

/.Jriol lJ~ 1 1 : .nl o; t.qm t 10% ,-Nr 180, Jl O DO Vi ~ : IJ ~}' ;<:.:< Vl ~.:.a slty
12.Q ~j ;1 4 31 ~CI:4 ...... EI11i:I LJ;ali t ~ , ~ O (j" NI. 180. J l0 /II;}wn E:a:UD Li,,!l : ~ ; 0% . Nr 180 Jl 0
12 J..
"a.g;- 16,( 11 13.C::!.97. 08 39
V1iCl:lCai.... ~-:oC Vt::t: o:>lty VI S':: O~lty
Ar:i!l bl.R f ' E:&! l lto l~ I I I -=-: l lY-li. , Nr' ro Jl ::: Arb oll E. n;. Lq !'f;4, 60% Nr. l HO, .110
,; 2 . (O ~" 15 ~A 1:: s;":2.::'7,'16..4,2

Figure 6 Sludge from Forozon/Arabian Extra Light (AEL) crude oil, diluted with A.E.L.

IN [RTAN n , Ul d 1.1 .. Il . . I... - . I ~ I, .U p

Using the data from this investigation , the expected extent of dilution of fresh crude oil needed to
generate a pumpable slurry with the precipitated sludges that would reasonably flow in a tank with the
tanker trimmed 8 metres by the stern was between 60 and 70%. This is supported by additional and
subsequent dilutions of an alternative crude oil to sludges of up to 40% subject to two wash cycles .
The resulting dynamic viscosities were found to be about 8,000 mPa s. It should be remembered that
the final objective of COW is to reduce the viscos ity of the combined mixture of oil and sludges to a
maximum of about 600 mPa s or approximately 600 cSt.

00

rn ~.
3 5000 . -Q-V ...

~v
? "1. v.
i '11.11.
3 00000
.......v

~
, ..
~ ....
)00 / "'~ - !;IV 'jI' 'i
~
/" I .
"
I
I
..----.
2 000 00 .---
v..,. ~ I

I
. '
..
. a..
~/ ~ -..~ c=. , o !. ~

. .- I
..
1

150000
/

. ~ -
. ~ ='f ~
1

~V
"V 1

1000000

,I

sco 0
I

it
~ II< 10

Shear RJte
Figure 7 Yield stress

An investigation and analysis for the sludqe 's yield stress was undertaken. Figure 7 shows the
measurements of yield stress, i.e. the force required to move the sludge , over a range of temperatures
for a particular slucqe sample. The results revealed a typical yield stress of approximately 1,400,000
mPa . When the results were compared with the impact force of the crude oil wash jet , it showed that
the jet's force easily exceeded the yield stress of the sludge and would irnpinqe/penetrate to the full
depth of 15 cm for the type of sludge tested . However, it is believed that the sludge , given its nature ,
would be pushed /displaced to one side of the jet impingement area/track and create "furrows" in
the sludge mass. Indeed , on inspection of vessel 's tanks after a COW operation , this pattern was
witnessed.

The "furrowing" of the sludge s causes much of the Wash medium to drain through the furrow s aft to
the suction position within the tank . This clearly impacts the theoretical dilution of tho crude oil wash
medium with the extent of sludqes, As a result, the foregoing theoretical calculations are only best
estimates of efficiency. Further, with the loss of the majority of the crude oil wash medium in this
way the impact of solvency of the wash medium upon the precipitated sludges is also reduced . In
fact, it is believed that the wash medium only erodes the boundary sludges within the furrows. It was
predicted in one such circumstance that only about 5% of the sludges would have been removed in
one wash cycle due to the observed physical nature of the sludge .

P 9 .3 I J l.. ' .. 1 ' J .. . .' 11 II - I .r

8.3 An alternative, practical programme for the crude oil washing of sludge

A typical COW programme will accommodate the possibility of all the machines in one tank operating
at the same time. In practice such a scenario would have to take into account the number of machines
that can be operated simultaneously in a safe manner, given the supply pressure and the size of
the eductor used for the stripping of the tank during the washing programme. The programme used
could be based on either the "concurrent" or "closed cycle" method, or even a mixture of the two
techniques.

The use of all the tank cleaning guns in one tank simultaneously may not be the optimum way of
removing sludges due to, amongst other reasons, the action of the guns on the precipitated sludges.
There is a possibility that they would push the sludges away from the tank suction location for 50%
of a programmable machine's cycle.

When confronted with sludging of the cargo tanks, particularly in colder climatic conditions, a departure
from the typical or standard procedure of crude oil washing has been found to be more effective in the
removal of sludges. A suggested procedure can be described on a step-by-step basis, as follows:

(1) Utilise a closed cycle wash system and load both slop tanks to the 60% full level with dewatered
crude oil wash stock. Pre-treat the crude oil wash stock by heating, if possible, up to 10'C above
the oil's Cloud Point temperature. Shut the equalisation line valve between the slop tanks and
cycle back to the first slop tank used for the COW programme.

(2) Assuming that programmable machines are available and unless required to employ an alternative
approach, carry out only a bottom wash in each tank using a 5'_40'_0' cycle. By commencing at
0', a significant cumulative time is used "spot washing" the arc 0' _5' due to the helical movement
of the gun.

(3) Programme the discharge so that two tanks may be washed together. It is assumed that each
tank has two COW machines, and a maximum of two machines may be operated at anyone
time. It is further assumed that the vessel is equipped with at least one eductor and it may take
suction on every cargo tank with its delivery directly back to either slop tank. The cargo tanks to be
washed need not be stripped completely dry before commencement of COW and, subject to the
external environment/sea temperatures and the nature of the sludge, i.e. how "sticky" it is, oil to a
depth of about 30 cm can be at the aft sounding location in the tank before commencing COw.

(4) Commence the COW in the designated pair of cargo tanks by opening the aft machine in each
tank to wash and clear the aft bay. Depending upon the nature of the sludge and the status of
the washing medium (see below for further discussion), one cycle of washing with the warm oil
should be sufficient. For efficient draining of the cargo tank washings, the aft bay of the tank
around the stripping position must be cleaned first and kept clear of any accumulated wash
material throughout the programme.

(5) Start the forward two machines in each tank with the remaining machines stopped. This procedure
has two primary advantages, i.e.

INTERTANKO - A Guide to Crude Oil Washing and Cargo Heating Criteria - May 2004 Page 34
 (a) The jet of crude oil wash medium projected by the COW guns does not push the sludge
forward as it is constrained by the tank's forward bulkhead. Indeed, deflected "splash" will
start to direct the sludges aft without any contrary interference/forces from the operation of
the aft machine had it been in operation.

(b) The sludges remaining in the middle bays of the tank will force the warm wash oil to percolate
through the sludges in order to be stripped from the tank. The warmer oil will also warm the
unwashed sludges and use all its remaining "heat energy" more effectively.

The length of time required for this wash will again depend upon the nature and amount of sludges
in the tank. However, previous use of this technique suggests that a maximum of two cycles would
be adequate to clear the forward bays.

(6) Due to the importance of keeping the suction position clear, it may be necessary to rewash the
aft bay. This is not always necessary and will depend upon the nature and degree of degradation
of "solvency" of the washing medium.

(7) The final section of the washing programme requires the washing of the centre and aft bays
of the tank with the aft machine in the pair of tanks. This final stage of the programme should
remove the bulk of the sludges from the tank, provided their viscosity has been adequately
reduced to render them pumpable.

(8) Upon completion of the final stripping period after closing the aft machine, the tank shouid be
checked for its cleanliness by sounding in the prescribed four locations (paragraph 4.4.4 of the
COW Manual). If residue sludges are found in the tank at certain locations, "spot" washing to
remove them may be considered.

After a period oftime, as this process is repeated in other pairs of cargo tanks, the efficiency ofthe crude
oil wash medium will be reduced due to its saturation with the recovered sludges. This will become
evident when the expected effectiveness of a crude oil wash in a pair of cargo tanks is not apparent.
At this time the contents of the slop tank being used for the storage of the COW medium and recovery
sludges and oil should be immediately discharged. The already heated and prepared COW medium
in the alternative slop tank should thereafter be used. If there is excessive sludging onboard and a full
COW programme is required, then it may be necessary to recharge the original slop tank after it has
been discharged with fresh crude oil and prepare it in readiness to continue the COW programme.

Page 35 !NTERTANKO - A Guide to Crude Oil Washing and Cargo Heating Criteria ~ May 200 .
9 Closing remarks

All tanker operators and seagoing officers will have at some time in their career encountered sludging
problems onboard crude oil tankers. At the time it may not have been immediately clear either how
or why this problem had occurred or exactly what the impact of the sludging was. The aim of this
Guide is to help resolve some of these questions. By exploring the normal washing procedures used
onboard tankers, it is further hoped that some guidance and recommendations have been provided
on how best to avoid the sludging problem and, when sludging does occur, to minimise its effects on
tanker operations.

INTERTANKO - A Guide 10 Crude Oil Washing and Cargo Heating Criteria - May 2004 Page 36
 (D
INTERTANKO
The International Association of Independent Tanker Owners

FOR SAFE TRANSPORT, CLEANER SEAS AND FREE COMPETITION

Box 5804 Majorstua, 0308 Oslo, Norway

Fax No. +4722122641

I would like to order _ _ copy/ies of"A Guide to Crude Oil Washing and Cargo Heating

Criteria'

USD 60 (members)

USD 120 (non-members)

Please send the publication indicated above:

o Our Company is a Member/Associate Member of INTERTANKO

COMPANY:

ADDRESS:

Attention:

Date:

METHOD OF PAYMENT

o Please charge my credit card: Visa/Eurocard/Mastercard/Amex (delete as appropriate)

in the amount of USD .

Credit card no.: .

Expiry date: ..

Cardholder's name: .

Signature: .

USD paid into INTERTANKO's account"

Swift: HANDNOKK Handelsbanken NS, P.O. Box 755 Sentrum, N-0106 Oslo, Norway.

Account No. 8396.04.41497

All remittances should be free of Bank charges and commissions.

Publications will be dispatched upon receipt of payment.
 o e n ta k elea ing tee 010 y
with ag etlc tran i slon ,

WashTrac
Compl ete
ank cleaning monitoring system
range of
combining Scanjet's tank cleani ng
portable
machines with Saab TankRadar
equipment
from Saab Rosemount Marine.

Tank clean ing mon itoring systems are

recommended by INTERTA NKO.

.. ~ I//
_

P.o. Box 13051

SE402 5 1 GOTEBO G
Sw ed n
Te l. +46-31 84 93 30
Fax +46-31 84 93 50
www.scaruei.se
office@ canjet. e