Encyclopedia of Energy Engineering and Technology - 3 Volume Set

Compressed Air EnergyStorage *
(CAES)
David E. Perkins
ActivePower,Inc., Austin, Texas,
U.S.A.
Abstract
Compressed-air energy storage (CAES) is currently being deployed as an alternative to lead-acid batteries
for uninterruptible power supplies. These systems use compressed air supplied from either transport
cylinders delivered by local gas services, or from stationary cylinders refilled from on-site compressors to
drive avariety of economical expansion engine topologies. Several factors make these systems feasible for
use in small-scale CAES systems for load leveling in conjunction with wind or solar energy generation
while opportunities exist for improving cycle efficiency.
INTRODUCTION Thermal and Compressed Air Storage (TACAS) tech-

nology. [ 4 ] Several companies, including amajor supplier
Large-scale CAES has been successfully used as ameans of photovoltaic cells have expressed interest in the use of
of peak shaving as an alternative to peaking gas turbines. this technologyfor electricitystorage generatedby
Two such systems are the 290-MWunit in Huntorf, renewable sources. At this time,however,systems based
Germany [ 1 ] and the 110-MW unit in McIntosh, Ala- on TACAS technology achieve cycleefficiencies between
bama. [ 2 ] These systems allow independent operation of the 10 and 15%, sinceitcurrently relies on oil-lubricated
compression and expansion processes commonly found in reciprocating compressors,which are nearly isothermal.
conventional gas turbines. These systemshave been called This paper presentsthe current embodimentofthe TACAS
hybrid CAES; in that they continue to use fuel in the technology for UPS and discussesproposed enhancements
expansion process, the benefit being that the turbine need to improvecycle efficiency for electricity storage.
not producepower to drive the compressor when operating
from the compressed air reserve.Adisadvantage of these
systemsisthat they require the use of fuel, which results in THERMALAND COMPRESSED AIR
CO 2 emissions, and the heat of compressionisdiscarded STORAGE
FOR
thus compromising the cycle efficiency. UPS
Onesolutiontothese issues is to useadiabatic ACAES system with thermal energy suppliedbythe grid
compression and thermal energy storage (TES) in place has been developed for UPS applications. The outputof
of the combustion process associated with hybrid CAES. this system is 85 kW and is capable of deliveringpowerfor
Systems employing adiabatic compression and energy up to 15-min (21 kWh). Aschematic of the system
storage were explored in the 1970s and have received architecture is shown in Fig. 1 .
renewedinterest. One program titled Advanced Adiabatic The TACAStechnology system uses high-pressure gas
Compressed Air Energy Storage (AA-CAES) [ 3] is under-
cylinders for air storage since volumetric energy density
way in Europe, which is focused on zero-emission storage should be compatible with batteries for UPS applications
technology for centralized storage as well as modular and the use of caverns is impractical. Theuse of high
products for distributed storage. temperature TES heated by the grid allowshigher turbine
Several companies are now offering CAESsystemsin inlet temperature than would, otherwise, be available from
various configurations as environmentally friendly alterna- direct expansion of the air from cylinders.
tives to the lead-acid batteries found in uninterruptible The specific energy available from compressed air is a
power supply(UPS) systems. Earlier this year, active function of the turbine pressure ratio and inlet temperature.
power-introduced products for the UPS market based on its System sizing then becomes atradeoffbetweenthe mass
of gas stored in the cylinders and the mass of TES required
to heat the gas beingdelivered to the turbine to achieve a
*
This entry was originallypresented as “Compressed-Air Energy desired inlet temperature. Cost optimization was per-
Storage for Renewable Energy Sources” at the World Energy Engineering formed to balance the size of the cylinderbanks and TES
Conference(WEEC), 13–15 September 2006, Washington DC, U.S.A. as afunction of theturbine pressure ratio, inlet
Reprinted with permission from AEE/Fairmont Press.
temperature, anddischarge temperature.The turbine
Keywords:Energy storage; CAES; Uninterruptible power supplies; UPS; pressure ratio was constrained by manufacturing capab ili-
Air turbine generators; Thermal energy storage; Renewable energy. ties and the discharge temperature was constrained to be
214
© 2007 by Ta yl or & F r anc i s G r oup, LL C
10.1081/E-EEE-120043197—4/5/2007—19:02—RATHNAS—225284—XML Taylor &Francis Encyclopedias

Compressed Air Energy Storage (CAES) 215
expensive permanent magnet material and is acompro-

mise between several machine design factors.
The system-optimized turbine showninFig. 2operates
at 385-psig and 2308 Cinlet temperature and delivers
100-kW shaftpoweratanefficiency of approximately
72% (small turbines typicallyhave efficiencies between 70
and 80% whereaslarger designs can approach 90%).
Nearlythe entire temperature drop occurs during
expansion across the supersonic converging-diverging
nozzles, so the temperature reachingthe bladed disk
approaches ambient. This allowsthe use of inexpensive
rotor materials with integrally machined blades.
The single-stage turbine readily adapts to overhung-
rotor architecture. The short overhang distanceand short
overall rotor length considerably simplify rotor-dynamic
Fig. 1 Thermal and compressed air storage (TACAS) tech- issues. In thisparticularconfiguration, rolling-element
nology for UPS.
bearings with elastomer dampers are employed sincethe
operational duty cycle for UPS is low. Thedesign is easily
suitablefor discharge into an interior building space. configurable to foil-type gas bearings for higherduty
Maximum storage pressure was constrained by the most service. Blade reaction is close to zero so axial bearing
economicalcylinders andcompressors as well as loadsare very low and easily managed. Cooling of the
diminishing returnsdue to significanteffectsofgas bearing closest to the turbine inlet plen um is achallenge,
compressibility effects ( O 10%) above 310 bar [4500 psi]. but effective routing of air released from the control
Small-scale electric power and storage systems are regulators helpstomitigatethis. In fact, cooling of the
generally more expensive to produceper kW and/or kWh turbine-PMG and its powerconverter is also provided by
output, so the optimization described above is essential. the process air ow.
Furthermore, in order to meetsystem-cost targets, the Becausethe system is presently designed for UPS,
simplest designs must be adopted. transient response is of paramount importance. It takes
about 1s for the turbine-generator to reach full speed
while carrying increasing load during the acceleration
event. Asmall ywheel is used to “bridge” the gap
EXPANSION TURBINE AND betweenutility outage and full-loadturbine output. This
CONTROL
ywheel is configured with abi-directional converter,
For small-scalesystems, single-stage axial- ow impulse which allows the ywheel to absorb step unloadsand
turb ines offer an efficient and cost effective solution for eliminates theneed forunloading valves. During
high-pressure ratio expansion applications, particularly discharge, the ywheel also managessmall power
whendirect-coupled to ahigh-speed permanent-magnet uctuations so that turbine speed is held constant.
generator (PMG). Adesign speed of 70,000 rpm mini- Simple turbines basedonfixednozzle geometry
mizes the PMG size helping to reduce the mass of achievepower regulation through inletpressureand
temperature control. For this system, turbine inlet pressure
and temperature are controlled by apair of dome-loaded
pressure regulators using aunique control scheme, [ 4, 5 ]
which routes air through or aroundthe TES. Referring to

Fig. 2, Main Regulator 340 controlsair ow going to the
Heat Exchanging System 350 (in this case,the TES).
Bypass Regulator 320 controls air ow aroundthe TES.
Orifice 330 in the bypass path provides control stability.
By combining ows through regulators 320 and 340 in
varying proportions,constant uid discharge temperature
and pressure can be achieved throughout the sliding
temperature range of the TES.
Regulators 320 and 340 identified in Fig. 3 are called
dome-loaded regulators. This type of regulator is often
employed whenahigh ow coefficient is needed. Adome-
loaded regulator provides adischarge pressure equal to the
Fig. 2 100-kW single-stage axial- ow air turbine generator. pressure signal applied to the dome. Dome pressure
© 2007 by T ayl or & Fr a nci s Gr oup, L L C

216 CompressedAir Energy Storage (CAES)
Fig. 3 Method of controlling constant turbine inlet temperature and pressure.

Source: From Ref. 5.
control is provided by acombination of solenoid valves cost is relatively expensive, manufacturing processes are
1050 and1052 in Fig. 3that, respectively,either simplifiedbyintegrating the thermal storage and pressure
pressurizesthe dome with air supplied by an upstream retention functions and an annular channel configuration [7 ]
regulatorand accumulator, or depressurizes the dome by achieves very high heat transfer coefficients and heat
venting the dome to atmosphere. Operation of these two extraction efficiencies ( Fig. 5 ).
valves is provided by adigitalcontroller with feedback Advanced micro-porous insulation is used to prevent
from downstream sensors. For reliable termination of excess heat loss and provides atemperature gradient of
dome pressure in emergency situations, aredundant NO over 6008 Cwith approximately25-mm thickness. Maxi-
solenoidvalve 1060 is in line with an Emergency Power mum operating temperature of 7008 Cischosenbased on
Off(EPO) circuit and is powered closed during normal life considerationsfor replaceablecartridge heating
operation. Faults such as sensed turbine overspeed or elements and creep considerations for the storage material
manual EPO will de-power the valve and vent the dome based on estimated time at pressure and temperature. In the
causing the pressure regulator to close (Fig. 4). case of UPS, the expected cycling is low, but low-cycle
fatigue due to dilatation stresses in the inlet piping when
THERMAL ENERGY cold air from the regulators is introduced were investi-
STORAGE gated. These were found to be non-issues for UPS. For
High-temperature thermal storage using 304stainless steel more extensive cycling that would be requiredfor wind
provides extremelycompact, robust, and low-riskTES or solarstorage,further investigationisneeded. In
when designed in accordancewith American Society of addition, work is ongoing to identify more economical
Mechanical Engineers (ASME) rules. Although material TES designs.
Fig. 4 Method of controlling con-

stant uid pressure.
Source: From Ref. 6.

Compressed Air Energy Storage (CAES) 217
Fig. 5 High energy thermal energy

storage for
UPS.
COMPRESSEDAIR STORAGE IN at fee. Therefore, someusers in the UnitedStates must

VESSELS pay twice or more for air storage than others.
Most large-scale CAES systems will probably continue to The regulatory picturefor compressed air storage in
use underground caverns for pressurized air storage where vessels in Europe is muchmore favorable with the recent
available. However, for small-scale systems or where harmonization of the European Union and introduction of
storage caverns are unavailable, small-storage vessels thePressure EquipmentDirective (PED). Through
offer economiesofscale duetomassproduction appropriate Notified body channels, it is astraightforward
techniques employed. These cylinders are produced on matter to re-qualifydesigns originally intendedfor one
automated process lines by backward extrusion of billet as application into another application so long as the types of
opposedtolarger vessels produced using seamless pipe service are similar.
with integrally forged heads. One manufacturer claims that
the productionrate for billet-formed cylinders is less than
30-s per cylinder. MODIFICATIONS TO
In theUnited States, transportcylinders having TACASTECHNOLOGY
Department of Transportation (DOT) exemption certifi- One of the most sign ificant adaptationsneededtoimprove
cation for 310-bar [4500-psi] service have been found TACAStechnology for electricity storage is the develop-
mosteconomical. Specific costs of less than $0.015 per mentofahigh-pressureadiabatic compressor and aslight
bar-liter [$0.034per psi-cu ft] can be obtained, whereas modification to the system architecture to allow heat
the cost of larger vessels has been found to be twice that or recovery. Aschematic of the proposed architecture is
more. For the turbine inlet conditions being employed for showninFig. 6.
UPS, this leads to energy cost for stored air of around The exact form of the high-pressure compressor is
$350/kWh. Energy storage costs for compressed air will under consideration. The high discharge temperatures
decrease with increasing turbine inlet temperature. This imposedbyadiabatic compressioneliminates reciprocat-
mustbebalanced with the cost of the TES needed to ingcompressors. Large-scalemulti-stagecentrifugal
achieve higher discharge temperatures. compressors with direct-drive induction motors have
In theUnitedStates,disagreementabounds on
regulatory issues surrounding the on-site generation and
storage of compressed air. Historically,stationary pressure
vessels are designed to ASME standards, and installation
and operationare regulated by the individual states.
Transport vessels fall under the jurisdiction of the U.S.
DOT and are exempted from state control. Many states
allow the use of DOT vessels for stationary storage of air,
but some do no tand somehave no regulatio ns on pressure
vessels. Attempts to qualify the most economical DOT
designs through standardASME channels have been
fruitless sincethe two organizations’ design rules are not
harmonized, even thoughthe service seen by the cylinders
is comparable.Furthermore, some jurisdictions assess Fig. 6 Schematic arrangement of modifications to thermal and
permitting fees on aper vessel basis, whereas others will compressed air storage technology architecture for improved
consider abank of vessels as one installation and asses a cycle efficiency.

218 CompressedAir Energy Storage (CAES)
been used for high-pressure applications in the oil and gas † Thermal energy storage cost reductionand or increase
industry. Scaling dynamic compressor designs for small- in allowableoperating temperature.
scaleapplications andadaptingfor high-discharge
temperatures poses significant technical challenges, but Focus on these initiatives will allow consideration of
will be crucial if high cycle efficiency is to be achieved. TACAS technology for near-term electricity storage in
Afurther enhancement to the TACAS technology conjunction with wind or solar power generating systems
wouldallowhigherturbine inlettemperatures since with lower upfront capital expenditure compared to large-
limitations on turbine-exhaust temperatures imposed by scale systems.
indoordischarge requirements of UPS couldberelaxed.
Increasing the discharge temperature reduces the TES
discharge temperature range and so mustbecompensated REFERENCE
by higher storage temperatures or additional TES mass for S
agiven output. Further advances in TES materialselection 1. Crotogino, F.; Mohmeyer, K.; Scharf, R. Huntorf CAES:
and design will be neededtoachieve economic viability. More than 20 Years of Successful Operation,Proceeding
Finally,inapplications for solarand wind energy SMRI Spring Meeting, Orlando, FL, U.S.A., 15–18 April
storage,itmay be possible to eliminate the bridging 2001.
2. Daly, J.; Loughlin, R.M.; DeCorso, M.; Moen, D.; Davis, L.
ywheel since the turbine hassuchrapid response.
CAES—Reduced to Practice,ASME 2001-GT-397, 2001.
However, the excellent load regulationand speed control
3. Bullogh, C.; Gatzen, C.; Jakiel, C.; Koller, M.; Nowi, A.;
attributes of the ywheel in tandem with the turbine make Zunft, S. Advanced Adiabatic Compressed Air Energy
this element of the architecture favorable if power quality Storage for Integration of Wind Energy,Proceeding of
enhancements are desired. European Wind Energy Conference (EWEC), 22–25
November 2004.
4. Pinkerton, J.F. Thermal and Compressed Air Storage
System. U.S. Patent 7086231, 8August 2006.
SUMMAR 5. Schuetze, K.T.; Hudson, R.S. Systems and Methods for
Y Controlling Temperature and Pressure of Fluids. U.S. Patent
Compressed-air energy storage and TES system have been Application 2006/0076426A1, Filed 5February 2003.
successfully integrated for UPS applications. Selected 6. Schuetze, K.T.; Weaver, M.D. Systems and Methods for
Controlling Pressureof Fluids. U.S. Patent Application 2006/
features of the TACAStechnology are presented and
0060246A1, Filed 17 September 2004.
enhancements for improving storage cycleefficiency are
7. Perkins, D.E.; Hudson, R.S. Thermal Storage Unit and
discussed. These enhancements include: Methods of Using the Same to Heat aFluid. U.S. Patent
6955050, 18 October 2005.
† Development of asmall-scale high-pressureadiabatic
compressor with high discharge temperature capa-
bility.
† Modification to thermal storage charge and discharge BIBLIOGRAPH
ow path. Y
† Turbine bearing upgrades for longer operational life. 1. Vosburgh, K. Compressed Air Energy
† Turbine modifications for increased inlet and discharge Storage,AIAA
Conference on New Options in Energy Technology, San
temperatures. Francisco, CA, 2–4 August ,1977.

Compressed Air Leak Detection and
Repair
Robe rt E. Wilson
ConservAIR Te chnologies Co., LLP,Kenosha, Wisconsin, U.S.A.
Abstract
Compressed air is amajor cost component in manufacturing. As such, it offers one of the largest savings
opportunities. The investment in compressing air to energize it and then letting it escape from the system
through leaks, without doing any useful work, is acomplete waste. This waste can be minimized by
implementing aprogram of leak detection and repairs. This entry covers the subject of how to use a
handheld ultrasonic leak detector to locate leaks and the procedures required to implement repairs. The
documentation and informational database required to ensure that leak waste is controlled and that new
leaks are prevented is summarized. Different application technologies for controlling leaks are presented,
and time and cost controls reviewed.
INTRODUCTION desired task. In reality, however,many of the openings are

unintentional, wasteful leaks.
Th is entrycovers the basicmethodsoffinding and The leak volume is directly proportionate to the area of
repairing leakages. the opening, the resistanceto ow, and the applied
The sectio nonLeak Detection presents the most pressure differential. The larger the area of the unrestricted
commonly used procedures to identify leakages.The opening and the higherthe supply air pressure, the greater
sound signature of air leaks and the use of ahandheld the leak ow. Achartshowing the discharge of air through
ultrasonic acoustic detector to locate leaks are explained. an orifice is includedasanappendix. Note that the values
Suggested procedures for tagging and documenting air listed are based on a100% coefficient of ow and should
leaksare presented. The benefit of entering information be adjusted for otherorifice configurations as suggested
into adatabase for historical trending is reviewed. ( Table 1 ).
The sectiononleakrepairs covers the most commonly When the air expandsback to atmospheric pressure,it
found air leaks and the procedurestorepair them. The transitions from ahigh-pressure laminar ow to alow-
logistical problems that create obstacles to expeditious pressure turbulent ow. The escape velocities become
repairs are discussed. The need for after-control, rechecks, extremeasthe air volume expands. This results in afull
and establishing standards are reviewed. sound spectrum of noise, ranging from audible to high
The sectiononleak control and prevention presents frequencyinaudible.
various approachesfor managing leaks to minimize waste One common method of detecting leakages is to usea
and ensurethat the savings realized are ongoing. The soap-like liquid that forms bubbles. Products specifically
application of ow monitoring and pressure regulationis formulated for high viscosity and film strength exaggerate
presented. the bubble effect to enhancethe detection capabilities. The
Leak costs are discussed and asummary of the entry is liquid is poured, sprayed, squirted, or brushed on asuspect
presented. area, and the formation of the bubbles is visually observed.
This method allows the detection of leaks that cannot
otherwisebeheard or felt in thenormaloperating
production environment,but bubble detection is time
LEAK consuming and messy. It requires the inspection of every
DETECTION connection to the air system,and the foaming agent may
In acompressed air system,the pressurized air confined by require materialapproval before it can be used in a
the pipes and vessels escapes from the system through particular facility. It also is not practical for checking
openings as it expandsback to atmosphericpressure. overheadceiling pipesorunder, behind, or inside
Ideally, all of these openings are createdintentionally to operating machinery.
extract energy from the compressed air in performing a The more commonly accepted method for detecting
leaks is to use ahandheld, ultrasonic acoustic detector that
can register the high frequencysound signature associated
Keywords:Industrial compressed air system leak detection; Repair;
with gas leaks and translateitinto an audible signal. Air
Control and prevention. leaks have adefinitive ultrasonicsound component in their
219

Comp Day
Table 1 Discharge of air through an orifice
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
Ar e a s q. 1/64 1/32 3/64 1/16 5/64 3/32 7/64 1/8 9 /6 4 5/32 3/16 7/32 1/4 9/32 5/16 3 /8 7/16 1 /2 9/16 5 /8 3 /4 7/ 8 1
i n. p re ss .00019 .00077 . 0 0 17 3 .0 0 3 0 7 .0 0 4 7 9 .0 0 6 9 0 .0094 .01227 .01553 .01973 .02761 .03758 . 04 9 0 9 . 0 62 1 3 .0 7 6 7 0 .1 1 0 4 5 .1 5 0 3 3 .1 9 6 3 5 .2 4 8 5 0 .3 0 6 8 0 .4 4 1 7 9 .6 0 1 3 2 .78540
1 0.0 2 8 0 . 1 1 2 0 . 2 53 0 .4 5 0 0 .7 0 0 1 .0 6 1 .4 8 1 . 8 0 2 . 2 7 2 . 8 0 4 .0 5 .5 7 . 2 9 .1 1 1 .2 1 6 .2 2 2 .0 2 8 .7 3 6. 3 4 4 . 8 6 4 . 7 8 8 1 1 5
2 0.0 4 0 0 . 1 5 8 0 . 3 56 0 .6 3 3 0 .9 8 9 1 .4 2 1 .9 4 2 . 5 3 3 . 2 0 3 . 9 5 5 .7 7 .7 1 0 . 1 1 2 .8 1 5 .8 2 2 .8 3 1 .0 4 0 .5 5 1. 0 6 3 . 4 9 1 . 2 1 2 4 1 6 2
3 0.0 4 8 0 . 1 9 4 0 . 4 36 0 .7 7 5 1 .2 5 1 .7 4 2 .3 7 3 . 1 0 3 . 9 2 4 . 8 2 6 .9 9 .5 1 2 . 4 1 5 .7 1 9 .2 2 7 .8 3 7 .8 4 9 .5 62 . 5 7 7 . 0 1 1 1 . 0 1 5 2 1 9 8
4 0.0 5 6 0 . 2 2 3 0 . 5 02 0 .8 9 2 1 .3 9 2 .0 0 2 .7 3 3 . 5 6 4 . 5 0 5 . 5 5 8 .0 1 0 .9 1 4 . 3 1 8 .1 2 2 .2 3 2 .1 4 3 .5 5 7 .0 72 . 0 8 8 . 9 1 2 8 . 0 1 7 5 2 2 8
5 0.0 6 2 0 . 2 4 8 0 . 5 60 0 .9 9 3 1 .5 5 2 .2 3 3 .0 4 3 . 9 7 5 . 0 2 6 . 1 9 8 .9 1 2 .2 1 5 . 9 2 0 .1 2 4 .7 3 5 .7 4 8 .5 6 3 .5 80 . 1 9 9 . 3 1 4 3 . 0 1 9 5 2 5 4
6 0.0 6 8 0 . 2 7 2 0 . 6 12 1 .0 9 1 .7 0 2 .4 5 3 .3 2 4 . 3 4 5 . 4 9 6 . 7 5 9 .8 1 3 .3 1 7 . 4 2 2 .0 2 7 .1 3 9 .1 5 3 .0 6 9 .5 87 . 9 10 8 . 0 1 5 6 . 0 2 1 3 2 7 8
7 0.0 7 3 0 . 2 9 3 0 . 6 95 1 .1 7 1 .8 2 2 .6 3 3 .5 8 4 . 6 8 5 . 9 0 7 . 2 9 1 0 .5 1 4 .3 1 8 . 7 2 3 .6 2 9 .2 4 2 .2 5 7 .3 7 5 .0 94 . 7 11 6 . 0 1 6 8 . 0 2 3 0 3 0 0
8 0.0 8 3 0 . 3 3 1 0 . 7 41 1 .3 2 2 .0 6 2 .9 6 4 .0 5 5 . 3 0 6 . 7 0 8 . 2 4 1 1 .9 1 6 .2 2 1 . 2 2 6 .9 3 3 .0 4 7 .7 6 4 .7 8 4 .7 1 0 6. 0 13 2 . 0 1 9 1 . 0 2 6 0 3 3 9
1 2 0 .0 9 5 0 . 3 7 9 0 . 8 56 1 .5 2 2 .3 7 3 .4 1 4 .6 5 6 . 0 7 7 . 6 6 9 . 4 2 1 3 .6 1 8 .6 2 4 . 3 3 0 .7 3 7 .8 5 4 .6 7 4 .1 9 7 .0 1 2 2. 0 1 51 . 0 2 1 8 . 0 2 9 7 3 8 8
1 5 0 .1 0 5 0 . 4 2 0 0 . 9 45 1 .6 8 2 .6 2 3 .7 8 5 .1 5 6 . 7 2 8 . 5 0 1 0 . 4 8 1 5 .1 2 0 .5 2 6 . 9 3 4 .0 4 1 .9 6 0 .5 8 2 .5 1 0 8 .0 1 3 6. 0 16 8 . 0 2 4 2 . 0 3 2 9 4 3 0
2 0 0 .1 2 3 0 . 4 9 1 1 . 1 00 1 .9 6 3 .0 5 4 .4 0 6 .0 0 7 . 8 6 9 . 9 2 1 2 . 1 2 1 7 .6 2 4 .0 3 1 . 4 3 9 .8 4 8 .8 7 0 .7 9 6 .0 1 2 6 .0 1 5 9. 0 1 96 . 0 2 8 3 . 0 3 8 5 5 0 3
2 5 0 .1 4 0 0 . 5 6 2 1 . 2 6 2 .2 5 3 .5 0 5 .0 5 6 .8 8 8 . 9 8 1 1 . 3 8 1 3 . 9 9 2 0 .2 2 7 .4 3 5 . 9 4 4 .5 5 6 .0 8 0 .9 1 1 0 .0 1 4 4 .0 1 8 2. 0 22 4 . 0 3 2 3 . 0 4 4 0 5 7 5
3 0 0 .1 5 8 0 . 6 3 3 1 . 4 2 2 .5 3 3 .9 4 5 .6 8 7 .7 1 0 . 1 1 2 . 7 7 1 5 . 7 0 2 2 .7 3 1 .0 4 0 . 5 5 1 .3 6 3 .0 9 1 .1 1 2 4 .0 1 6 2 .0 2 0 5. 0 25 3 . 0 3 6 5 . 0 4 9 6 6 1 8
3 5 0 .1 7 6 0 . 7 0 3 1 . 5 8 2 .8 1 4 .3 8 6 .3 1 8 .6 1 1 . 3 1 4 . 2 6 1 7 . 6 0 2 5 .3 3 4 .5 4 5 . 0 5 7 .0 7 0 .0 1 0 1 .0 1 3 7 .0 1 8 0 .0 2 2 7. 0 28 1 . 0 4 0 5 . 0 5 5 1 7 2 0
4 0 0 .1 9 4 0 . 7 7 4 1 . 74 3 .1 0 4 .8 4 6 .9 7 9 .5 1 2 . 4 1 5 . 6 5 1 9 . 3 1 2 7 .9 3 8 .0 4 9 . 6 6 3 .0 7 7 .0 1 1 2 .0 1 5 1 .0 1 9 8 .0 2 5 0. 0 31 0 . 0 4 4 6 . 0 6 0 7 7 9 3
4 5 0 .2 1 1 0 . 8 4 5 1 . 90 3 .3 8 5 .2 7 7 .6 0 1 0 .3 1 3 . 5 1 7 . 0 5 2 1 . 0 0 3 0 .4 4 1 .4 5 4 . 1 6 8 .0 8 4 .0 1 2 2 .0 1 6 5 .0 2 1 6 .0 2 7 3. 0 33 8 . 0 4 8 7 . 0 6 6 2 8 6 5
5 0 0 .2 2 9 0 . 9 1 6 2 . 06 3 .6 6 5 .7 1 8 .2 2 1 1 .2 1 4 . 7 1 8 . 6 0 2 2 . 9 0 3 2 .9 4 4 .9 5 8 . 6 7 4 .0 9 1 .0 1 3 2 .0 1 8 0 .0 2 3 5 .0 2 9 6. 0 36 5 . 0 5 2 8 . 0 7 1 8 9 3 8
6 0 0 .2 6 4 1 . 0 6 2 . 38 4 .2 3 6 .6 0 9 .5 0 1 2 .9 1 6 . 9 2 1 . 4 0 2 6 . 3 5 3 7 .9 5 0 .8 6 7 . 6 8 5 .0 1 0 5 .0 1 5 2 .0 2 0 7 .0 2 7 1 .0 3 4 2. 0 42 2 . 0 6 0 9 . 0 8 2 8 1 , 0 8 2
7 0 0 .3 0 0 1 . 2 0 2 . 69 4 .7 9 7 .4 5 1 0 .5 3 1 4 .7 1 9 . 2 2 4 . 2 5 2 9 . 9 0 4 3 .0 5 8 .6 7 6 . 7 9 7 .0 1 2 0 .0 1 7 3 .0 2 3 5 .0 3 0 7 .0 3 8 8. 0 47 9 . 0 6 9 0 . 0 9 3 9 1 , 2 2 7
8 0 0 .3 3 5 1 . 3 4 3 . 01 5 .3 6 8 .3 3 1 2 .0 4 1 6 .4 2 1 . 4 2 7 . 1 0 3 3 . 3 3 4 8 .1 6 5 .5 8 5 . 7 1 0 8 .0 1 3 1 .0 1 9 3 .0 2 6 2 .0 3 4 3 .0 4 3 3. 0 53 7 . 0 7 7 1 . 0 1 ,0 5 0 1 , 3 7 1
9 0 0 .3 7 0 1 . 4 8 3 . 33 5 .9 2 9 .2 5 1 3 .3 4 1 8 .2 2 3 . 7 3 0 . 0 0 3 6 . 9 0 5 3 .0 7 2 .3 9 4 . 8 1 2 0 .0 1 4 7 .0 2 1 3 .0 2 8 9 .0 3 7 9 .0 4 7 8. 0 59 2 . 0 8 5 3 . 0 1 ,1 6 1 1 , 5 1 6
1 0 0 0 .4 0 6 1 . 6 2 3 . 65 6 .4 9 1 0 .5 0 1 4 .5 8 1 9 .9 2 6 . 0 3 2 . 8 0 4 0 . 5 0 5 8 .0 7 9 .0 1 0 4. 0 1 3 2 .0 1 6 2 .0 2 3 4 .0 3 1 6 .0 4 1 5 .0 5 2 3. 0 64 9 . 0 9 3 4 . 0 1 ,2 7 2 1 , 6 6 1
1 1 0 0 .4 4 1 1 . 7 6 3 . 96 7 .0 5 1 1 .0 0 1 5 .8 2 2 1 .5 2 8 . 2 3 5 . 6 0 4 3 . 9 0 6 3 .0 8 6 .0 1 1 3. 0 1 4 3 .0 1 7 6 .0 2 5 4 .0 3 4 5 .0 4 5 2 .0 5 7 0. 0 70 2 . 0 1 ,0 1 6 . 0 1 ,3 8 3 1 , 8 0 6
1 2 0 0 .4 7 6 1 . 9 1 4 . 29 7 .6 2 1 1 .4 0 1 7 .1 5 2 3 .4 3 0 . 5 3 8 . 5 1 4 7 . 5 0 6 8 .0 9 3 .0 1 2 2. 0 1 5 4 .0 1 9 0 .0 2 7 4 .0 3 7 3 .0 4 8 8 .0 6 1 6. 0 71 2 . 0 1 ,0 9 7 . 0 1 ,4 9 4 1 , 9 5 1
1 2 5 0 .4 9 4 1 . 9 8 4 . 45 7 .9 0 1 2 .3 0 1 7 .7 9 2 4 .2 3 1 . 6 4 0 . 0 0 4 9 . 2 5 7 0 .0 9 6 .0 1 2 6. 0 1 6 0 .0 1 9 6 .0 2 8 4 .0 3 8 6 .0 5 0 6 .0 6 3 8. 0 78 9 . 0 1 ,1 3 8 . 0 1 ,5 4 9 2 , 0 2 3
1 5 0 0 .5 8 2 2 . 3 7 5 . 31 9 .4 5 1 4 .7 5 2 1 .2 0 2 8 .7 3 7 . 5 4 7 . 4 5 5 8 . 2 5 8 4 .0 1 1 5. 0 1 5 0. 0 1 9 0 .0 2 3 4 .0 3 3 8 .0 4 5 9 .0 6 0 0 .0 7 5 8. 0 91 0 . 0 1 ,3 1 5 . 0 1 ,7 8 9 2 , 3 3 8
2 0 0 0 .7 6 1 3 . 1 0 6 . 94 1 2 .3 5 1 9 .1 5 2 7 .5 0 3 7 .5 4 9 . 0 6 2 . 0 0 7 6 . 2 1 1 0 .0 1 5 0. 0 1 9 6. 0 2 4 8 .0 3 0 5 .0 4 4 1 .0 6 0 0 .0 7 8 4 .0 9 9 0. 0 1 , 22 5 . 0 1 ,7 6 4 . 0 2 ,4 0 1 3 , 1 3 6
2 5 0 0 .9 3 5 3 . 8 0 8 . 51 1 5 .1 8 2 3 .5 5 3 4 .0 0 4 6 .2 6 0 . 3 7 6 . 1 5 9 4 . 0 1 3 6 .0 1 8 4. 0 2 4 1. 0 3 0 5 .0 3 7 6 .0 5 4 2 .0 7 3 8 .0 9 6 4 .0 1 , 2 1 8. 0 1 , 50 8 . 0 2 ,1 6 9 . 0 2 ,9 5 2 3 , 8 5 6
3 0 0 0 .9 9 5 4 . 8 8 1 0 .9 5 1 8 .0 8 2 8 .2 5 4 0 .5 5 5 5 .0 7 1 . 8 9 0 . 6 1 1 1 . 7 1 6 1 .0 2 2 0. 0 2 8 7. 0 3 6 4 .0 4 4 6 .0 6 4 6 .0 8 8 0 .0 1 , 1 4 8 .0 1 , 4 5 4. 0 1 , 79 5 . 0 2 ,5 8 3 . 0 3 ,5 1 5 4 , 5 9 2
4 0 0 1 .2 2 0 5 . 9 8 1 3 .4 0 2 3 .8 1 3 7 .1 0 5 3 .4 5 7 2 .4 9 4 . 5 1 1 9 . 4 1 4 7 . 0 2 1 3 .0 2 8 9. 0 3 7 8. 0 4 7 9 .0 5 9 0 .0 8 5 1 .0 1 , 1 5 5 .0 1 , 5 1 2 .0 1 , 9 1 5. 0 2 , 36 0 . 0 3 ,4 0 2 . 0 4 ,6 3 0 6 , 0 4 8
5 0 0 1 .5 1 9 7 . 4 1 1 6 .6 2 2 9 .5 5 4 6 .0 0 6 6 .5 9 0 .0 1 1 7 . 3 1 4 8 . 0 1 8 2 . 5 2 6 4 .0 3 5 8. 0 4 6 9. 0 5 9 3 .0 7 3 0 .0 1 , 0 5 5 .0 1 , 4 3 0 .0 1 , 8 7 6 .0 2 , 3 6 0. 0 2 , 93 0 . 0 4 ,2 2 1 . 0 5 ,7 4 5 7 , 5 0 4
7 5 0 2 .2 4 0 1 0 . 9 8 2 4 .6 0 4 3 .8 5 6 6 .1 5 9 8 .5 1 3 3 .0 1 7 4 . 0 2 2 0 . 0 2 7 1 . 0 3 9 2 .0 5 3 1. 0 6 9 6. 0 8 8 1 .0 1 , 0 8 4 .0 1 , 5 6 6 .0 2 , 1 2 5 .0 2 , 7 8 4 .0 3 , 5 1 0. 0 4 , 35 0 . 0 6 ,2 6 4 . 0 8 ,5 2 5 1 1 , 1 36
1 0 0 0 2 .9 8 5 1 4 . 6 0 3 2 .8 0 5 8 .2 1 9 1 .0 0 1 3 0 .5 1 7 7 .0 2 3 1 . 0 2 9 1 . 5 3 6 0 . 0 5 2 0 .0 7 0 8. 0 9 2 4. 0 1 , 1 7 1 .0 1 , 4 4 0 .0 2 , 0 7 9 .0 2 , 8 2 0 .0 3 , 6 9 6 .0 4 , 6 5 0. 0 5 , 79 0 . 0 8 ,3 1 6 . 0 1 1 ,3 1 8 1 4 , 7 84
Tableisbased on 100% coefficient of ow. For well-rounded orifice, multiply by 0.97. For asharp-edged orifice, amultiplier of 0.65 will give approximate results. Values calculated by approximate formula
p ); where: W ,discharge (lb/s); A ,area of orifice (in.
proposed by S.A. Moss. W Z 0.5303(ACP = T 2 ); C ,coefficient of ow; P ,upstreampressure (PSI, abs.); T ,upstreamtemperature ( 8 F, abs.); Values used in
calculating table: C Z 1; T Z 5308 R(708 ); P Z Gage pressure plus14.7 psi;weights converted to volumes using density factor of 0.07494 lb/ft 3 (correct for dryair at 14.7 psi abs. and 708 F); values from 150 to
1000 psi calculated by Compressed Air Magazine and checked by Test Engineering Dept. of Ingersoll-Rand Co.
© 2007 by T a ylor & Fr a nc is Gr oup, L L C

Compressed Air Leak Detection and Repair 221
noisesignature that is beyond the hearing threshold of the Different color tags can be used to visually indicate the
humanear. The ultrasonicleakdetector translates the severity of leaks and establish priorities. Typical classi-
ultrasonic noise of the leak signature into an audible sound fications might include three levels:
heardinthe earphones worn by the leak surveyor. Some
instruments are also equipped with display meters and Level 1: Not audible in any environment without an
indicator lights that visually register the magnitude of the ultrasonic detector.
air leak. Adistinctive, loud rushingsound is produced in Level 2: Audible in aquietenvironment but not in an
the earphones when the leak detector sensor probe is operatingfacility.
aligned with aleak. With the productionbackground noise Level 3: Serious leaks requiring immediate attention.
suppressed and filtered out by the headphone set, the
leakage hissing is heard. Level 1leaks cannot be felt or heardunder any
The sound wavegenerated by an air leak is directional conditions and require the use of the previously described
in transmission. The intensityofthe leak noise is based procedures to detect. They are less than 1scfm and are
upon the shape of the orifice opening, the distance to the assigned no value, sincethe cost of the associated logistics
senso rprobe,and the differential expansion pressure.The and labor do not economically justify the repair, unless it is
sound level is loudest at the actual point of the leakage very simple, such as the ubiquitous push lock fitting on
exit. Theprocedure for detecting leaksultrasonically uses plastic tubing. Level 1leaks are tagged and documented
this characteristic to locate the actual leaks. Initially, the for future recheck, sinceair leaks never fix themselves and
leak detector is set at the maximum practical sensitivity only grow larger over time. Thecumulative effect of the
consistent with the specific environment of the area being Level 1leaks on the compressed air system can be better
inspected. Asweep of the general area is performed as the controlled by maintaining astable delivered air pressure at
surveyor walks the system.When aleakisdetected, the the lowest optimum level through the applications of
direction of the leak is determined by scanning the area pressure/ ow control and regulating use points.
until the loudest noise levelregisters. With the probe Level 2leaks are in the 2scfm range and can typically
pointing in the direction of the noisesource, the surveyor be felt but not heard without the use of an ultrasonic leak
movestowards the leak, adjusting the sensitivity of the detector. Repairs are economically justifiable and should
leak detector accordingly. The intensityofthe sound be performed within a60-dayperiod.
increases in the proximity of the leak and is loudest at the Level 3leaks in excessof2scfm can typically be felt
actual point of air exit. Extensiontubesorcones attached and sometimes heard by the human ear. These require
to the sensorprobe focus the sound and pinpoint the immediate attention, since they not only waste air but
location of smaller leaks. The bigger, moreserious leaks impact the operational efficiency of the compressed air
can be felt. Afurther test using abubblesolution can system.Leak owisareal demand that adds to the
augment the process by visual enhancement of the exact filter/dryer loading, increases the pressure drop throughout
location. One such productisformulated to producean the system, and createspressure uctuations that impact
ultrasound shockwav easthe bubbles burst, so the surveyor production.
gainsthe benefitsofboth the visual observation and While the true owfor any specific leak cannot be
ultrasonic detection. measured practically, the surveyorcan assign values based
Competing sounds often mask aleakorotherwise upon the chosenleak volume associated with the various
distortthe directional transmission. If possible, the best leak levels.These can then be totaled at the end of the
way to eliminate acompeting sound is to shut the system surveytoestimate the cumulative system leak waste. The
off. If that is not possible, shielding techniques can be surveyorwilltypically overestimateabout thesame
applied. The angle of the probe extension can be changed. amount of leakages that are underestimated, so the final
The competing sound can be blocked using the body or figure givesagood portrayal of the total leak waste. As
othersolid barrier like apiece of cardboardorclipboard. long as the survey procedures are replicated during the
Cupping the hand over the leak ,orusing arag, can often re-check, the comparative value for trending becomes an
isolatethe true sourceofthe sound. Bubble tests can accuratemeasure forevaluatingthe remedial repair
pinpoint the location regardless of the competing sound. It actions taken. Acost figure can be assigned for use in
is imperative when workinginand aroundoperational the financial analysis. Take into account power cost and
machinery that safety be most important. Common sense associated compressor maintenance and repair costs, plus
dictates the extent of effort that should be expended to thecosts to operate andmaintainall theauxiliary
identify and quantifyaspecific leak. equipment, when determining the real value of the leak
The first step in the preparation for performing aleak waste.
surveyistoestablish apattern for surveying the facility to Efforts have been madetoestimate the actual volumeof
ensurethat all the piping, connected use points, and an air leak based upon pressure and the decibel level
workstations in an area are inspected. Detected leaks are registered at aspecific distance. People have assembled
identified and tagged during the surveillance of the system. test stands usin gthe most common orifice configurations

222 Compressed Air Leak Detection and Repair
Fig. 1 Noise vs leak loss at various pressures.
found in compressed air systems, and then have measured repairs. Regardless of theconfiguration of thetag,
air ow and decibel noise at different pressuresand information sufficient to allow revisiting an individual
distances. One such Chart,publishedbyUESystems of leak for repair,even if the tag falls offorismissing, should
Elmsford, NY, is presented in Fig. 1. Note the disclaimer be recorded on aseparate worksheet. This typically
that the values are not stated as “factual CFM”and are consists of:
provided as a“general guideline.” Aleak signature is
affected by manyfactors, and the loudness of the noise † Recording the unique, sequential tag number assigned
generated is by itselfnot the sole measureofthe volume of to the specific leak.
the leakage. For example, ahigh-pitched whistle will † Defining its workplace location in away that is
sound alot louderthan alow-level whoosh sound, but the meaningful to the air user.
whistle will consume less air. At best, the leak detection † Identifying the specificitem that is leaking.
process will provide an estimatefor use in planning the † Identifying the actual point of air exit on the leaking
priorities of the remedial repair proceduresand avalue for item.
evaluating trends. Classifying the degree of leakage so priorities for
†
remedial action are established.
LEAK The tag can be used for after-control by providing aplace

REPAIRS to enterthe date and repairperson’s name. The supervisor
Detected leaks must be visually tagged for future repair. should check to ensure that the repair has been properly
Some tags are configured to enable you to tear offacopy to completed before signing offand removing the tag. The
give to the maintenance supervisor responsible for the leak repair actions and associated time should be recorded on the

orig inal worksheet and entered into adatabase to establish LOGISTICAL PROCEDURES AND
time and cost control accou nting procedures. OVERHEAD
ASSOCIATED WITH LEAK
The surveyor should record complete information on REPAIRS
the worksheet to describethe leak. The probable cause of 1. Meetings and Planning
theleak, such as aging, wear,damage, looseness, 2. Maintenancerequisitions
mishandling, breakage, or otherreasons, should be noted 3. Purchaser—product and supplier identification
with an explanatory note if required.Determinewhether 4. Order costs—cost per placed order
the leak should be repaired or apart replaced, and note it 5. Transportation
on the worksheet. If replacement is recommended, the 6. Control of receipt—administration
surveyor should collect enough information about the item 7. Storage—space and logisticscosts
to allow for purchasing the repair part or replacement unit. 8. Labor schedule—days/weeks
Someone will have to do this if the leakisgoing to be 9. Leakage cost per week/month
fixed, so the surveyor shouldmake the extraeff ort to 10. Time control—verification and administration
record the information at the same time the leak is
identified. Theair user will also need to know if the leak is
repairable without having to shut down the associated LEAK
machinery. The worksheet should have areas for helpful MANAGEMENT
comments and field notes to facilitate remedial actions or Air leaks grow bigger over time, and repaired leakages
to alert peopleabout other issues and opportunities that usually reappear within six months to one year after they
cometothe attention of the surveyor. are fixed. Steps mustbetakentocontrol the growth rate of
Detected leaks must be repaired in order to realize any leaksand to preventreoccurrences afterrepairs are
savings. Since mostleaks occur at the operating machinery completed. Thekey to managed leakage control and
in theproduction area,repairprocedures tend to be prevention is rechecks anddocumentation.Periodic
repetitive. Stressesare applied to all the various hoses and rechecks at predetermined intervals ensurethat the leak
couplings, tubing connections, and pipe joints because of rate is stabilized at alow level. Through documentation,
machinery vibration and movement of the connected tools the trends become obvious and developing patterns, both
and pneumatically driven devices. Over time, leaks develop good and bad, are identified. Problemsare recognized
at sealing areas. These are easily fixed by reconnecting the beforecreatingissues that are more serious.Takin g
hose or reinstalling the pipe fitting after inspection and appropriate actions drives the leak trend downward until
cleaning. Worn couplings or quick disconnectsare it reaches the target established by management,typically
replaced. The plastic components of point-of-use devises, 5%–10% of thetotal air demand.Thishistorical
such as filters,regulators, and lubrications, tend to age and information is used to institute leak prevention measures
crackover time.These must be replaced. Gaskets and seals and for calculating the mosteconomical interv al for
dryout andbecomebrittle,sotheynolongerseal rechecks to ensurethat the gains realized are maintained in
effectively. Valve stem packing and sealingrings, manifold the future. Establishing standards and good practices
gaskets, hose reel rotary joints, and cylindershaftseals wear minimizesfutureleakage. With the time andcosts
over time and need to be replaced. Clamps, pipe unions, documented, controls can be put in place to properly
anges,and pipe groove seals often require re-tightening. administer aleak management program.
Leaksinthe compressorroomare foundaroundair An alternative approach to implementing afullleak
treatment equipment, condensatedrains, receiverman- management program is to simplyfixthe leaks immedi-
holes, and control tubing. ately upon discovery, assuming asystem is checked for air
Leaks on pipe joints are relatively easy to fix by either leaks on aregular basis. The technician brings alongatool
tightening or reinstalling aconnection. Clean all surfaces tote with the appro priate equipment needed to fix the most
beforereassembly. Use anon-hardening sealing paste for commonly found leaks. Usually, only the more serious
threadedconnections to prevent the possiblecontami- leaks are addressedinthe simple seek and fix approach.
nation of the air system from torn or frayed Te on e Little, if anything,isdocumented.
tape. [1 ] Leaks in main headers and branchlines often The total leakage for afacility can be estimated using
require lifts or special rigging equipment to gain access, techniques that measurepressure degradation over time
and may require special plumbing skills to repair. Advance whenthere are no production demands on the system.
planning and scheduling will be necessary for coordinating One such method is to measurethe load/unload cycle
the repairs on machinery not accessible during production. time of compressors whenproduction is shut down and the
The largest obstacletorepairing leaks is the logistics only air demand on the system is leakage. Start the
involved in planning andimplementingthe repair compressor(s) and record the on-load time and off-load
procedures. These logistical problemsoften take months time over asampling period long enough to provide a
to resolve and sometimes impede the process entirely. A representative average. Calculate the leakage lost as atotal
typical scenario follows. percentage of compressor capacity usingthe formula:

224 Compressed Air Leak Detection and Repair
Leakage ð % Þ Z ½ðT ! 100Þ = ð T C t Þ point. Electricallyactivated designs require apower

source. While this is sometimes inconvenient,electric
where: T Z average on-load time, and t Z average off-load units have the advantages of (1) indicator lights that show
time. the operational status of the drain and (2) contacts that can
In systems configured with compressor controls other be interfaced with abuilding management system for
than load/unload, leakage can be estimated based upon the remote monitoring.
total system capacitance.The total estimated volume ( V ) High efficiency blowing devices are available to entrain
of all air receivers, the main piping distribution, and other surrounding ambient air in the primary air stream to
significant air containment vessels must be calculated in increase the impingement force, so that less compressed
cubic feet. Pressure in the main header must be measured air is required to perform the equivalent task. Air knives,
at the startand endofthe evaluation testperiod. nozzles, and jets are offered with avariety of different
Production mustbeshut down so that the only demand air ow patterns to better suit aspecific task.Air
on the system is leakage. The compressors are then started amplification ratios as high as 40:1 over open blowing
in order to pressurize the system to its normal operating are achievable. [ 2] Supply pressure at the point of use can
pressure ( P 1 ). The compressors are turned off, and the time often be regulated to alower level to further reduce
( T )ittakes the system to drop to apressure equal to half compressed air consumption and the associated noise.
the normal start pressure ( P 2 )ismeasured. The leakage is Air volume amplifiers are available to create directional
estimated usingthe formula: air motionintheir surroundings and to efficiently move air
and light materials. Asmall amount of compressed air is
Leakage ð cfm of free airÞ used as apower sourcetoamplify the owofentrained
ambientair. Air ow is directional, with an inlet and outlet,
ZðV!ðP KP Þ = T ! 14: 7 Þ ! 1 : 25
1 2 to exhaustand/or sweep an area in ashaped pattern. Air
volume amplifiers can create output ows up to 25 times
[2]
where: V is the volume in cubic feet, P 1
and P 2 are in psig, the compressed air consumption.
and T is the time in minutes.
Because air escapes from thesystematarate
proportional to the supply pressure,the leakvolumerate LEAK CONTROL AND
at the normal start pressure will be much greater than the PREVENTION
leak volume rate at the end of the timedcyclewhen Leaks of all sizes, both intentional and unintentional, can
the pressure is half. A1.25 correctionfactor is applied to be controlled by supplying them, at minimum, an
compensate for the difference in the leak rate and to acceptable, delivered air pressure.Anair system that has
provide amoreaccurate estimation of the loss. acumulative equivalent of a5/16 0 0 leakage orifice, for
Installing a ow meter to measureand record the actual example, is illustrated in Fig. 2.

ow improves the accuracy of the air leak estimateover The applicationofPressure/Flow Controlinthe
usingacalculatedcapacitance based upon estimated compressed air system primary is agood method for
system volume. Aproperly configured ow meter can minimizing leakwaste.The smaller leaks, determined to be
also be used to monitor the system consumption in order to uneconomicallyrepairable, leak less at the lower delivered
(1) establish abaseline for evaluating the performance pressure.The Pressure/Flow Control also prevents the
improvements realized from any remedial actions taken,
(2) verify trends, and (3) verify that the gains continue to
return the investmentinto the future.
Many of the leaks in an industrialcompressed air
system are intentional or planned. Condensate drainage
and disposal, spot cooling, fume venting and exhausting,
material conveyingand blowing off, and drying are
examples of intentional leaks. Devicesare available to
eliminate or mitigate the air used to perform these typesof
assigned tasks.
No air loss condensatedrains collect the condensation
in avessel, until it fills with water.A oat or sensordetects
the high liquid leveland opens adrain port, allowing
compressed air to displace the water and forcing it to
discharge from the vessel. The sensor shutsoff the drain
port before the vessel is completely drained, so that no air
is lost with the water discharge. Some designs are entirely
pneumatic, so no electric powerisrequired at the use Fig. 2 Cumulative system leak demand at different pressures.

supplypressure from rising because of the lower demand leak survey shouldbeperformed several timesayear in a
that stems from leak repairs. Without some method of system recheck. Results should be entered into adatabase
supplyside pressure control, the system pressure increases and analyzed. Flow monitoring systemsare available to
inversely with demand, forcing leaks and other unregulated measureactual ow. These can interface with manage-
use points to consumemore air. The savings achieved by ment informationsystems that have remoteaccess.
lowering the leak demand are offset by air that is shunted Measuring real ow allows the true cost of the delivered
out elsewhere in the system because of the rising pressure. air to be calculated in $/mmcf (Dollarsper millioncubic
The application of point-of-use pressure regulationis feet). Charting the savingsinreports for management
another method for minimizing leakages. Settingand ensures continuedsupport for the program.
securing pressure regulators to supply air at the lowest
minimal acceptable pressure maintains the respective leak
losses at their lowest possible level. SUMMAR
Y
Shuttingoff the air to non-productive workstations and
assembly lines is another good method of leak control. In summary, agood leak control and preventionprogram
Lock out valves and isolationvalves can be installed to for acompressed air system starts with aleak survey.
completely shut offthe air to machinery that is shut down. Ultrasonic leak detectors are the best tool to find air leaks
The procedure can be manual or automated through the and pinpoint their location. The information about the air
installation of actuated shut offvalves that are activated by leaks is recorded in aworksheet and documented in a
an external signal. Automation eliminates the dependency database for use in generating reports and identifying
on ahuman action to stop the waste. trends. Savings are only realized if leaks are repaired.
Repairproceduresmustbeestablished and an investment
madeinsatisfying all of the logistical obstacles before the
LEAK actual remedial actions can be taken. Consideration should
COSTS
be giventocontracting outthe repairs, alongwith the
The cost of leaks mustbedetermined to allow manage- logistical requirements, to expeditethe process and realize
menttomake properdecisions about the compressed air the savings as soon as possible. Rechecks and monitoring
system. TheChart in Fig. 3illustrates the cost of air are necessary to drive the leak trend down and keep it at
consumed by leaks. the targeted rate.
In addition to thepower cost showninFig.3,
consideration should be giventoother associated costs,
such as labor to logdailyoperations, scheduled REFERENCE
maintenance, repair services, and periodic major over- S
hauls. Leaks are areal demand that require real air ow to 1. Ultraprobee Instruction Manual Volume 3. UE Systems,
satisfy. There is an added cost burdenthat resu lts from Inc., Elmsford,
treating the leak air, removing conden sation, additional NY.Corporation Catalog No. 20, 11510 Gold Coast Drive,
2. Exair
compressor wear, and increased powertocompensate for Cincinnati, OH.
the greater pressure drop because of the higher ow. The
final cost figures can more than doublethe cost based
solely upon electrical power. BIBLIOGRAPH
An effective leak control and prevention program Y
requirescontinuousmonitoring and verification that the 1. ConservAIR Technologies Company, LLP, Kenosha, WI,
gainsrealized are ongoing into the future. At aminimum, a Industrial Seminar Series, Compressed Air Management.
2. Improving Compressed Air System Performance aSource-
book for Industry,U.S. Department of Energy, Energy
Air consumed by leaks at 100 psig
Efficiency and Renewable Energy, Prepared for the
Compressed Air Challengew and the United States Depart-
D i a me te r i n . SC F M Le a k a ge A nn ua l v o l ume C os t pe r y ea r *
ment of Energy, Prepared by Lawrence Berkeley National
1 / 6 4( . 01 6 ) 0 . 4 1 2 1 5 , 49 6 c f/ y r $ 47 . 4 1
Laboratory, Washington, DC, and Resource Dynamics
1 / 3 2( . 03 2 )1. 6 2 8 5 1 , 47 2 c f/ y r $ 18 7 . 3 2
Corporation, Vienna VA, Section 2–3 Compressed Air
1 / 1 6( . 06 3 ) 6 . 4 9 3 , 4 1 1 , 14 4 c f/ y r $ 75 0 . 4 5 System Leaks.Nov 2003.
3 / 3 2( . 09 4 ) 1 4. 6 7 , 6 7 3 , 76 0 c f/ y r $ 1 , 6 88 . 2 3 3. Taranto, T. Data Power Inc., Energy Efficiency in
1 / 8 (. 1 25 ) 2 6. 0 13 , 6 6 5 ,6 0 0 c f/ y r $ 3 , 0 06 . 4 3 Compressed Air Systems,1998, Chapter 6-Compressed Air
5 / 3 2( . 15 6 ) 4 0. 5 21 , 2 8 6 ,8 0 0 cf / yr $ 4 , 6 83 . 1 0 Systems, Performance.
4. Te one is atrademark of DuPont.
1 / 4 (. 2 50 ) 11 3 . 0 5 9 , 3 9 2 ,0 8 0 c f/ y r $ 1 3 , 06 6 . 4 2
5. Wilson, R. PEMCO Services, St. Petersburg, FL,
*B as e d u p o n r a te u s e d b y US D e p t. o f E n e rg y, EER E, Presen-
tation, 2005.
A So u r ce b o o k f o rIn d u st r y. C =1.0
6. Winkler, D. LeekSeek International Ltd., Training Presen-
Fig. 3 The typical cost of air leaks. tation, 2005.

Compressed Air Storage and
Distribution
Thomas F. Taranto
ConservAIR Te chnologies, LLP,Baldwinsville, New Yo rk, U.S.A.
Abstract
Consistent stable operation of an industrial compressed air system is achieved when compressed air
ow supplied to the system equals compressed air demand. Energy distributed to the system is available
from two sources; rotating energy of the air compressors, and energy from compressed air storage.
Optimum system energy efficiency is possible when the proper amount of energy is available from
compressed air storage. Presented here are the physical and mathematical relationships that may be used
to assess system performance and determine compressed air storage requirements. These relationships
are also applied to design the air storage volume and distribution pressure profile necessary for effective
compressed air storage.
INTRODUCTION than the average air demand. For reliable operation of

the system,peak air demands must be supplied as they
Compressed air systemshave historically used on-line occur. If air supplyfalls short of demand, system pressure
air compressorcapacitysufficient to supply peak air will decrease. When peak air demands are not supplied,
demands. Whereair receivers have been installed ,the system pressure can fall belowthe minimum acceptable
system’s pressure profileand lack of storage control limit operating pressure.Thisoften leads to lost productivity.
the effectivenessofcompressed air energy storage.Today’s The supplyofair is available from two sources; rotating
energy costsand competitive world economyrequire that on-line compressed air generation capacity and com-
inefficient, wasteful practices of the past must end. This pressedair energy storage. Operating excessiveair
entry develops the necessary calculations to assess com- compressor capacity to supply peak air ow is inefficient
pressed air energy demand and calculate the usable com- and expensive. Properly engineered compressed air energy
pressed air energy in aproperlyengineered storage system. storage will supply peak air demand and reduce rotating
Methods of design, application, control,and optimization on-line energy. The result is improved overall system
of compressed air energy storage are introdu ced. efficiency and reduced power cost.
Compressed Air Generation Efficiency

ENERGY FLOW IN COMPRESSED AIR
SYSTEMS Generation efficiency is highest when an air compressor
Compressed airsystems canbeviewedashavingtwo is operating at full-load capacity. It is commonfor
parts: supply anddemand. Thesupplysystemincludesair compressors to operate at less than full-loadcapacity, a
compressors, dryers,filters, andequipment foundinthe condition referred to as part-load operation. During part-
powerhouse/compressor room.The demand side includes load operation, acompressor consumes less than full-load
perhapshundredsofuse points,including tools, actuators, power as its compressed air output is reduced. “Specific
processuse,blowing,cooling,material transport, and power” measured as power per unit of air produced (kW/
sparging (a processwhereby airisinjectedintoatank of 100 scfm), however,isgreater during part-load operation.
liquid,resulting in abubblingofthe solution to providea The resu lt is that part-load operation is less efficient than
desiredaction).The amount of energy necessarytodrive full-loadoperation. Compressors may alsooperate in the
productive airdemands is changing constantly as equipment no-load or unloaded condition. When unloaded, the air
andprocesses startand stop.Duringnormalmachine cycles, compressor produces no compressed airatall yet
compressed airuse is oftencyclicalratherthancontinuous. consumes 25%–30% of full-load power. Running com-
Normaldemand variations and diversity of applica- pressors unloaded for along period greatly reduces a
tions result in an air ow profile with peaks and valleys in system’s overall generation efficiency.
air ow rates that are at times significantly more or less
Compressed Air SystemEfficiency
Keywords:Air; Compressed air; Storage; Receiver; Pneumatic; Supply; Supplying peak air ow demand with rotating on-line
Demand; Air storage; Energy; Compressor. generation requires one or morecompressors to operate in
226
10.1081/E-EEE-120042157—4/5/2007—20:34—VELU—223878—XML Taylor &Francis Encyclopedias

Compressed Air Storage and Distribution 227
the part-load or unloaded condition. As the peak demand The ideal balance between generation and demand is
occurs, the compressor(s) will load for ashort time during achieved when Q g en Z Q d mn d .
the demand event and then return to part-load or unloaded
operation. The result is poor overall system efficiency. IdealAir System Energy Balance
For acompressed air system to achieve maximum Qg en Z Q ð2Þ
d mn d
operating efficiency, the compressed air supply should
Only whensystem pressure Z constant
incorporate both compressed air generation and storage.
The goal is to supplyaverage air demand with generation Eq. 2Ideal compressed air system energy balance.
(on-line rotating energy) and to supply peak air ow The first law of thermodynamics for achange in state of
requirements from storage (stored compressed air energy). asystem, [2 ] or the law of conservation of energy, states
that energy is not created or destroyed. This implies that
Compressed Air System EnergyBalance
the air ow rate of generation mustequal the air ow rate of
demand. From practical experience, it is observed that
The energy delivered to and consumed in acompressed
generation and demand are not always equal, resulting in
air system is afunction of the weight or mass owofair
changing system pressure. When Q g en
O Q d mn d ,system
moving through the system. Themassofcompressed air
pressureincreases,and when Q gen
! Q d mn d ,system
dependsonpressure and temperature. Increasing pressure
pressure decreases.The energy imbalancebetween
increases the density and, therefore, the mass of air.
generation and demand is either absorbed into or released
Increasing air temperature will decrease the air’s density,
from storage ( Q s to ).
decreasing the mass of air. This relationship is stated in the
IdealGas Law. [ 1] With Compressed Air Enteringand Exiting Storage
pV Qgen G Q ZQ
sto d mn d
ð1Þ
T Z aconstant ð for afixed mass of gasÞ Entering storage pressure increases
Eq. 1IdealGas Law for fixed massofgas. Exiting storage pressure decreases
Compressed air is often measured in terms of its
ð3Þ
volume—ft 3 ,for example. The volumetric measureofair is
irrelevant with respecttothe airmassunlessthe Eq. 3Actual compressed air system energy balance.
temperature and pressure of the air volume are also Defin ition: Q sy s .Air ow rate of the system is the
known. Therefore, standards are adopted to express the compressed air mass ow rate (scfm) produced by the
massofair under “Standard” conditions, resulting in the rotatingon-line compressorcapacity(Q g en
)atany
definition for aStandard Cubic Foot of air (scf). Standard moment,minus the air ow absorbed into storage ( K Q s to
conditions adopted by CAGI (Compressed Air and Gas for increasing pressure) or plus air ow released from
w
Institute) and Compressed Air Challenge (CAC) are storage ( C Q s to for decreasing pressure) (Fig. 1).
14.5 psia,68 8 F, and 0% relative humidity.
Compressed air energy transfer can be expressed as the Practical Air SystemEnergy Balance
mass ow rateofair at agiven operating pressure in Qs ys Z Q d mn d
stan dard cubicfeet per minute (scfm). This is both a ð4Þ
measure of volumetric ow rate and the mass or weight Qs ys Z Q gen
GQ sto
ZQ d mn d
ow rate of compressed air. Higher- ow-rate scfm accounts for changing system pressure
deliversahigherpower rate, and the time duration of
ow determines the energy transferred. Eq. 4Energy balance of compressed air systems.
Compressed air power enters the system from the air
compressors and exits the system through air demands,
including productive demand, leaks, and all points where
compressed air leaves the system,expanding back into the
atmosphere.Power delivered from the compressors is
measured as mass ow rate Q (scfm)from generation, or
Q ge n ,and power leaving the system also is measured as
mass ow rate of air demand, or Q dm n d
.
Definition: Q g e n .Air ow rate of generation is the com-
pressed air mass owrate(scfm) produced by the rotating
on-line compressor capacity at any moment.
Definition: Q dm n d .Air ow rate of demand is the
compressed air mass ow rate (scfm) escaping from the Fig. 1 Air ow relationship: generation, storage, and system
compressed air system to the atmosphere at any moment. ows.

228 CompressedAir Storage and Distribution
Fig. 2 Compression of air from atmospheric pressure to 100 psi gauge (114.5 psi).
COMPRESSED AIRENERGY Assumethat the compressor intakes 7.9 ft 3 of dry
STORAGE atmospheric air at 14.5 psia.Multiplying by r Z 7.9 results

The compressed air system engineer must analyze the in apressure of 114.5 psi absoluteor100 psi gauge (Fig. 2).
dynamicsofcompressed air energy storage.This section The application of the combined gas law is the basisof
develops the mathematical expressions necessarytostudy compressed air storage calculation. More advanced forms
the relationship among air system generation, storage,and of the calculation can be used to assess many aspects of
demand. To optimize system operation, system energy compressed air system performance.
3
supply mustremain balanced with air demand. Further- If an air receiver of 1ft volume is pressurized to
more, system energy supply must be optimizedbetween 114.5 psi absolute, and its dischargevalve is opened to the
3
generation (rotating on-line energy) and storage (stored atmosphere, how many ft of air (at atmosphere) will be
compressed air energy). discharged from the receiver? The form of the ideal gas
law used to solvethis problem is
Application of the Combined Gas Law D Pr ec

Vg as Z V rec ð7Þ
Pat m
The combined gas law states that the pressure of an ideal
gas multipliedbyvolume divided by temperature is a Eq. 7Gas volume-receiver volumerelationship (assuming
constant. temperatureZ constant).
The volume (at atmosphere) of air ( V g as )released from
V1 V2 the receiver is equal to the air receiver’s volume ( V )
P1 ZP ð5Þ re c
T1 2 T2 times the pressure change of the receiver ( D P r ec
)divided
by atmospheric pressure ( P a tm
). Furthermore, the recei-
Eq. 5Combined Gas Law.
ver’s pressure change ( D P re c )isequal to the final pressure
The compression process usually begins with air at
( Pf )minus initial pressure ( P i
)giving ( D P r ec Z P f K P i ).
ambientconditions of pressure and temperature (assuming
Substituting:
no moisture content). Alarger volume of air is compressed
to areduced volume, increasing the pressure of the air. Pf K P i
During the compression process, the air’s temperature Vg as Z V rec
Pat m
increases. Assuming the air’s temperature is ultimately
returnedtoambient, as aresult it can be saidthat the end ð 14: 5psia K 114: 5psiaÞ
ð8Þ
Vg as Z 1ft 3
pressure of the compression process is equal to the initial 14: 5psia
pressure times the ratio of beginningvolume to ending
volume (compression ratio). ZV Z K 6 : 89 ft 3
g as
V1 1 Eq. 8Gas volume released from areceiver.

P1 ZP 2
; where V Z r ð Compression RatioÞ
V2 V2 In the previous discussion of air compression,itwas
P1 r Z P shown that reaching apressure of 114.5 psia requires
2
compressing 7.9 ft 3 of air. The receiver calculations above,
ð6Þ 3
however,show that onlyK 6.89 ft of air are released from
3
Eq. 6Compression ratio. the 1ft receiver above.
3
If during compression,7.9 ft of air are reduced to a What happened to the other cubic foot of air? The other
3
volume of 1ft ,the resultant pressure is 7.9 times the cubic foot of air is still inside the air receiver because the
initial pressure or the ratio ( r Z 7.9 times). receiver’spressure remains at 14.5 psia (1 atm).

sy s rec CV p i pe
Cp n Z V ZV ð9Þ
Pa Pa
Eq. 9Pneumatic capacitance.

For every1atm change of system pressure,increaseor
decrease, the system will absorborreleaseone timesits
volume (ft 3 )ofcompressed air.Eq. 9above givesthe
capacitance of the system in terms of ft 3 per atmosphere of
pressure. Assume that the atmosphere is at standard

conditions of pressure and temperature, and that the piping
volume is negligible.
s ys
Vg as Z C pn
DPZV DP
Pa
Fig. 3 Air receiver pressure change.
Cp n Z 100 cu ft ð 10Þ
It is important to note that the receiver’s pressure 1atm Z 100 scf atm
change is the slope of the line calculated from final
pressure minus initial pressure. Lookingatthe receiver Vgas Z 100 scf
pressure throughout time, when pressureisfalling atm ! 1atm Z 100 scf
(negative slope), air is owing from the receivertothe
system. The xy plot in Fig. 3shows pressure ( y )and time Eq.10Volume of gasasafunction of pneumatic
( x ). Storageair ow Q s to
hasaninverse relationship capacitance ( C p n ).
betweenthe storage receiverand system—that is to say, Eq. 10 showsthata100 ft 3
-volumeair receiver
air leaving the receiver ( K Q s to )isair entering the system changing pressure y 1atm will displace 100 scf of air
( C Q s to ). into or out of the vessel. The pneumatic capacitance of the
system is 100scf/atm. If the D P is positive (i.e., initial
Pneumatic Capacitance pressure is lower than the final pressure), the air displaced
is 100 scf, and air is absorbed into the air receivertank.If
Pneumatic capacitance of acompressed air system ( C pn
) the D P is negative(i.e., initial pressure is higherthan the
represents the compressed air energyabsorbed into or final pressure), the air displaced (100 scf) is delivered from
released by acompressed air system as its pressure the air receivertank.
increases or decreases. It is expressed in terms of the mass Becausecompressedair system pressure is often
of air/un it change in pressure—for example, Standard measured in psi, it is desirable to express the capacitance
Cubic Foot/Atmosphere (scf/atm) (Table 1). of compressed air systemsinterms of scf per psig (scf/psi).
Table 1 Definition of variables and units of measure
Cp n Pneumatic capacitance (scf/atm) or (scf/psia)

Vrec Receiver volume (cu ft)
Vpi p e Piping volume (cu ft)
Vsy s System volume (cu ft)
Pa Atmospheric pressure (psia)
Pi Initial receiver pressure (psig)
Pf Final receiver pressure (psig)
D P Storage pressure delta ( P f
K Pi )
rs Storage pressure ratio ( P f
K P i )/P a
Vgas Compressed air volume (scf) standard cubic feet
P l oad Compressor load pressure (psig)
P un l oad Compressor unload pressure (psig)
Q sy s Air ow rate for the system
(scfm)
Q gen Air ow rate from generation compressor(s)
(scfm)
Q st o Air ow rate of storage
(scfm)

Considering that ( P f psigK P i psig)yieldsabsolute pipe, it is often insignifican t. For example, 1000 ft of 2-in.
3
pressure difference D P (psia):Substituting P f K P i (psia) schedule 40 pipe has avolume of 23.3 ft or 174 gal
3
for the Storage Pressure Delta ( D P atm). (7.48 gal/ft ), and 1miof1-in. schedule 40 pipe is only
31.7 ft 3 (237 gal). Theavailable pressure delta for storage
Vs ys ð cu ftÞ
Vgas Z is determined by the system pressure profile.
Pa ð atmÞ D P ð atmÞ
Vs y s ð cu ftÞ Air System Pressure Profile and Storage Delta
Vg as Z KP ÞðpsiaÞ ð 11Þ
Pa ð psiaÞ ð P f i
The highest pressure available in the system is usually
sy s
ð cu ftÞ determined by the maximum working pressure of the air
Therefore : C ZV
pn
P a ð psiaÞ compressors. Thelowest acceptable operating pressure is
determined by manufacturing requirements. With an air
Eq. 11 Capacitance and storage pressure delta (scf/psia). compressor rated at 125 psig maximum workingpressure
Therefore, pneumatic capacitanceofacompressed air and arequired use-pointpressure of 75 psig, for example,
system ( C p n )isafunction of the total volume of the system amaximum 50 psig pressure differential is available. Only
and atmosphericpressure, which can be expressedas: aportion of this differential can be used for storage,as
there are unrecoverable pressure losses as compressed
CV
Cpn Z V sys
ZV r ec p ip e
ð 12Þ air owsthroughthe system. The pressure profilein Fig. 4
Pa Pa allows for 15-psig control pressure band, 5-psig treatment
Eq. 12 Capacitance ( C pn ) Z scf/psia. pressure drop, 2-psigloss through distributionpiping, and
The V g as of the system and the C p n in units of scf/psia 8-psig differential in the point-of-use connection piping.
are directly related by the change in system pressure ( D P ) Given aminimum demand-side use-pointpressure of
in units of psia. 75 psig, and including the imposedpressure delta (10 psi)
through distribution plus point-of-use piping, the pressure
Vg as Z C pn
DP ð 13Þ profile in Fig. 4shows that the lowest optimum target
pressure of the supply-side headeris85psig. The normal
Eq. 13 Gas volume as afunction of capacitanceand
supply-side header pressure is 105psig. This profile
delta P .
3 allowsfor primarystoragepressure differential of
For the system above with V s y s Z 100 ft ,the capaci-
20 psig available to the system.
tance is 6.896 scf/psia. If the system pressure delta is
3 There are costsassociated with compressed air energy
14.5 psia (1 atm), the stored V g as is 100ft (see Eq.14),
becau se atmosphere (14.5 psia)isthe condition defined for storage.The discharge pressure at air compressors must
Standard Gas Conditions V g as Z 100 scf. be increased to provide storage pressure differential.
Increased energy is about 1% for each 2-psigincrease in
Vsy s 100 cu ft compressor discharge pressure (for positive displace-
Cpn Z Z ment-typecompressors).Also, increasingthe com-
Pa 14: 5psia Z 6 : 896 cf= psia
pressed air system pressure increases the air demand of
Vg as Z C pn
DP the system.Compressedair leaves the system through
various openings to the atmosphere, such as the open
For D P Z 14: 5psiað 1atmÞ the Gas Volume is : port of acontrol valve, ablowing nozzle or open
blowing tube,oraleakinthe piping. Any opening in the
Vg as Z C pn
D P Z 6 : 896 ! 14: 5 system that does not have apressureregulator
controlling the applied pressure will blow an increased
Vg as Z 100scf amount of air ow as the system’s applied pressure is
ð 14Þ increased. This increasedair owiscalled artificial
demand. System air demand is increased by approxi-
Eq. 14 Calculating gas volume.
mately 1.0%–1.3%for every 2-psig increaseinsystem
pressure.For systemswith little effective use-point
UsableCompressed Air EnergyinStorage pressure regulation, artificial demand will be greater.
Definition:Artificial demand. Artificial demand is the
In the previous discussion, it is apparent that two factors additional compressed air ow demand consumed by the
determinethe amount of compressed air energy storage: system due to actual applied air pressure being greater than
the receivervolume and pressure delta (initial minus final the minimum required target pressure.
pressure). The volume of acompressed air system is For the pressure profileshown in Fig. 4, the storage
determined primarily by the number and size of air pressure differential is 20–35 psig as compressor controls
receivers in the system. Piping volume adds to the total, cycle between their load and unload set points. If the
but unless there are several hundredfeet of large-diameter average storagepressuredifferential is 28 psig,the

Fig. 4 Compressed air system pressure profile with uncontrolled storage.
resulting power increaseis14%, and waste to artificial sys r ec

C V pi p e
demand is between14and 18% of the system’s air ow. Cpn Z V ZV Z 1000ð galÞ = 7 : 48ð gal= cu ftÞ
Pa Pa 14: 5 ð psiaÞ
As shownabove, creating compressed air energy
storage increases the system’s energy requirement and Cpn Z 133: 67ð cu ftÞ
powercost. Therefore, it is unwise to create more storage 14: 5 ð psiaÞ Z 9 : 2 ð cu ft= psiaÞ
than the system requires. Proper design will minimize the 0 1
increasedcompressordischarge pressure andpower cu ft
Vgas Z C pn D P Z 9 : 2 @ A 20ð psiaÞ Z 184ð scfÞ
requirement. Also, proper control of compressed air psia
energyinstorage can virtually eliminate artificial
ð 15Þ
demand.Thistopic is discussed in “Maximize and
ControlCompressed AirEnerg yStorage,” presented
later in this entry. Eq. 15 Pneumatic capacitance ( C pn
)and usable storage
( Vg as ).
But first, the compressed air system engineer must
consider various system requirements for stored energy
andengineer the storage system appropriately. The Calculating Peak Air Demand
following sections demonstrate common applications of
the pneumatic capacitancecalculation to solve various During alarge demand event, the supply pressure in the
system energy storage requirements. system above is observed to draw down from 105 to
80 psig in 30 s. What is the air ow rate for Q s to (scfm)
during the demand event?
Calculate UseableCompressed Air in Storage Adding time to Eq. 12 for gas volume (scf) allows
solving for gas owrate Q s to (scfm).
Compressed air energy in storage is of use to the air system
only if storage pressure is greater than the minimum Vg as Z C DP
pn
system supply pressure.Given the pressure profile shown
in Fig. 4, and assuming that the air receiver volume of the dP ð 16Þ
system is 1000gal, the amount of usablecompressed air Qst o Z C pn
d T Wheretime T Z minutes
energy in storage can be calculated.
For an air receiverof1000-gal volume with 20 psia
primary storage delta P ,see Fig. 4; solvefor the usable Eq. 16 Solve for storage air ow rate Q s to (scfm).
compressed air energ ystorage V ga s (scf) (assuming that the Solving for the peak air ow rate from storage ( P f ,final
air is at standard temperature and relative humidity). pressure; P i ,initial pressure),

dP Pf K P i
period that might range from many seconds to minutes
Qs t o Z C pn pn depending on the type of compressors and controls. Most
dTZC dT
0 1 air compressors must start in an unloaded state to allow the
electricmotor to accelerate to normal running speed.For a
Qs t o Z 9 : 2 cf@ft A 80ð psigÞ K 105ð psigÞ
typical lubricant-injected rotary screw compressor with
psia 0 : 5 ð minÞ
part winding or Y -Delta, for example, starting might
Qs t o Z K 460ð scfmÞ require 5–10 sfor transition to full running torque. When
the permissive time is past, the compressor’s controls must
The storage air owrate is 460 scfm, which is equal to
open the inlet to begincompressing air.Then the internal
approximately100 hp of rotating on-line compressor
piping, oil sumpreceiver, and possibly after-cooler must
capacity.
be pressurizedbefore the compressor’s internal pressure
Calculating Required ReceiverVolume for exceeds the system pressure and the first cubic foot of air is
Demand Events forced through the compressor’s discharge check valve
into the air system.
How much air receivervolume should be added to support Consider the air system shown in Fig. 5 , including
the demand event while maintaining supply pressure at three compressors operating in abaseload ,trim capacity,
85 psig minimum? Solve Eq.15for pneumatic capaci- andstandby controlconfiguration.The system air
tance ( C p n ); then converttogal and solve for additional demand is 700 scfmrequired at 85 psig minimum
receivervolume. pressure.The base-load and standby compressors are
fixed-speed load/unload compressors with rated capacity
dP of 400 scfm. The trim compressor is avariable-speed
Qs t o Z C pn
dT drive (VSD) compressor rated at 500 scfm capacity. The
VSD trim compressor is set to maintain atarg et pressure
dT dT
Cp n Z Q of 90 psig.Assume that the standby compressor is set to
st o sto
dPZQ Pf K P i start automatically at apressure of 88 psig and requires
3
0 : 5 ð minÞ a15spermissive startup time to deliver its first ft of air
Cp n Z K 460ð scfmÞ into the system.
ð 85 K 105ÞðpsiaÞ
With the unanticipated shutdown of the base-load
Cp n Z 11: 5 ð scf= psiaÞ compressor, whatsize air receiver (gallons) is necessary to
ensure that the system pressure does not fall below85psig
Vrec Z C pn
Pa during thepermissivestartup time of thestandby
0 1 compressor?
scf First, calculatethe air ow required from storage after
Vrec Z 11: 5 @ A 14: 5 ð psiaÞ Z 166: 8 ð cu ftÞ shutdownofthe base-load compressor. System air demand
psia
is 700 scfm with shutdown of the base-load compressor;
0 1
the VSD trim compressor will increaseits air delivery
Convert to gallons : 166: 8 ð cu ftÞ ! 7 : 48 gal@ A to full capacity of 500 scfm. Theremainingair deficit
cu ft of 200 scfm must be supplied from storage(Q ).
sto
Z 1248ð galÞ The pressure profilefor the event will result in afall of
pressure to 88 psig before the standby compressor is
Additional Receiver Volume: 1248 K 1000 signaled to start. Theminimumpressurefor the
receiver is 85 psig. Therefore, the initial receiver pressure
Z 248ð galÞ ð 17Þ
( Pi )is88psig, and the final receiverpressure ( P f
)is
88 psig. The permissive startup event duration is 15 s
Eq. 17 Solve for additional air receiver volume (gal). or 0.25 min.
dT dT
Calculating Air Storage for Compressor Permissive Cp n Z Q sto sto
dPZQ Pf K P
Start-up Time i
Storage of compressed air energy is also necessaryto Cp n Z K 200ð scfmÞ 0 : 25ð minutesÞ
85 K 88ð psiaÞ
supportcompressed air demand duringvarious supply-side
events. One common supply-side event is the unantici- Cp n Z 16: 7 ð scf= psiaÞ
pated shutdownofanair compressor (due to amotor
overload or ahigh-temperature condition, for example). Vrec Z C pn
Pa
The startup of reserve compressor capacity requiresa

Fig. 5 Base-load, trim capacity, and standby

compressed air system.
0 1
Higher pressure also adds energy cost to the system’s
scf
Vr ec Z 16: 7 @ A 14: 5 ð psiaÞ Z 242: 2 ð cu ftÞ air demand. Increasing supply-side pressure createsa
psia corresponding demand-side pressure increase. The result
0 1
gal is additional energy consumption of the system through
Convert to gallons : 242: 2 ð cu ftÞ ! 7 : 48 @ A an air system loss called artificial demand. Simply stated,
cu ft
if the compressed air pressure applied to leaks and
Z 1812ð galÞ ð 18Þ unregulated airuse points is increased, theair ow
consumed will alsoincrease. Artificial demand is the
Eq. 18 Permissive startup—solve for pneumatic capaci- additional compressed air ow demand consumed by the
tanceand receiver volume. system when the actual applied air pressure is higherthan
Pneumatic capacitance calculations can be applied to the minimum requiredtarget pressure. Artificial demand
solveavariety of compressed air storage requirements. in asystem withoutany effective point-of-use pressure
regulation will increase the system’s energy demand by
2% for each 2-psig increaseinpressure. In a“typical”
compressed airsystem, it is common to find that
MAXIMIZE AND CONTROL 35%–50% of all air demands have effective pressure
COMPRESSEDAIR
ENERGY regulation. Therefore,artificial demand typically
STORAGE increases by 1.0%–1.3% for every2-psig increasein
For compressed air energy in storage to be effective, the applied system pressure.
storage pressure mustbehigherthan the demand-side Artificial demand can be eliminated by controlling
target pressure.Assupply-side pressure increases, the the demand-side target pressure at an intermediate point
powerrequiredbypositivedisplacement compressors also separating the supply and demand sides of the system.An
increases. Theair compressor’s powerincrease is intermediate ow control valveisinstalled as shown in
approximately1%for every 2-psig increaseindischarge Fig. 6 , downstream of the primary storage air receiver
pressure.Increasedstorage pressure or increased air at the beginningofthe distribution piping. This separates
receiver volume increases usable air in storage.The the supplyside from the demand side of the system.
economic tradeoffisthe capital cost of increased air Flow control is used to control the energy (air ow)
receiver volumevsthe increased compressor supply-side entering the system while maintaining areal-time energy
energy cost of higher storage pressure. balance betweensupply and demand. An intermediate

uctuating pressure, so compressed air energyfrom

storage must be available for instantaneous expansion
into the system to maintain the supply/demand energy
balance.The immediate energysupplycannotbe
dependent on rotating on-line generation. Air must be
stored during dwells in the demand cycle whenexcess
compressor capacity is available. Then stored energy is
released by the intermediate owcontrol to satisfy the
peak demands.
The intermediate ow control,like all components of a
compressed air system, has someunrecoverable pressure
loss, which represents an energy cost to the system.
Intermediate ow controls are designed to operate with
low unrecoverablepressure loss—typically, less than
Fig. 6 Intermediate ow control separates supply and demand. 5psig. For the pressure profile showninFig. 6, the
intermediate owcontrolhas an unrecoverable
pressure loss of 3psig, which represents 1.5%increased
ow control is apackaged assembly of one or more ow
energy cost at the compressor. In Fig. 4 , it is shown that
controlvalves in amanifold arrangementwith an
waste to artificial demand is between14and 18% of the
automatic bypass or fail-safe open overridedevice. It is
system’s airdemand. The netsavingsachieved by
normally installed in the compressor room at the beginning
eliminating artificial demand with application of inter-
of the main piping distribution system.Adequate-size
mediate ow control is 12.5%–16.5% of the system’s
receiver(s) installedupstreamofthe ow control provide
energy input.
compressed air energy storage for controlled releaseinto
the system. The ow control senses air pressure at its
discharge. Changes in the demand-side energy require-
ments cause uctuating pressures. The ow control senses SUMMAR
these changes and increases or decreases the air ow from Y
storage as neededtomaintain the system’s energy balance Compressedair systems constantly undergodynamic
betweensupplyand demand. The result is stable system changes in their energy demand. For reliable,consistent,
pressure set at the lowest optimum target pressure— and efficient system operation, energy supplyand demand
normally, to within G 1psi (Fig. 7). must be balanced in real-time performance.Operating a
The upstream compressed air energy storage is crucial system without properenergy storage results in part-load
to the satisfactory application of an intermediate ow or no-load operation of compressors,which decreases
control.The valvepackage responds immediately to the system efficiency and increases energy cost.
Fig. 7 Compressed air system pressure profile with controlled storage.

Compressed air energystorage can provide the CONCLUSIO

necessary energy to meet peak demands. Pneumatic N
capacitance calculations derived from the Ideal Gas Law The compressed air system engineer must assess the
and First Law of Thermodynamics for systems allow dynamic energycharacteristics of theair system.
mathematical modeling of compressed air energy storage. Compressed air energy storage requirements to support
Usable air storage is afunction of two factors: available normal system eventsmust be calculated. Energy storage
storage volume, andthe pressure difference between mustbeoptimized with an appropriate system pressure
storage pressure and demand-side target pressure. profileallowing the necessary storage pressure differential
Demand-side target pressure should be the lowest opti- and adequate air receiver storage volume. Proper control
mum pressure requiredtosupportproductive air demands. of stored energy and the resultant control response of the
Increasing storage pressure increases the compressor’s system’s air compressors, whenoptimized, will provide
energy use by 1% per 2psig.Uncontrolled compressed air the best possible system operating efficiency.
storage pressure also increases the applied demand-side
pressure, resulting in waste to artificial demand. Artificial
demand typicallywastes1.0%–1.3%ofthe system’s
REFERENCE
air ow for every2-psig increaseinsystem pressure. S
Artificial demand can be eliminated through the appli-
1. Resnick, R.; Halliday, D. Kinetic Theory of Gasses: Physics
cation of intermediate ow control to separatesupply and for Students of Science and Engineering;Wiley: New York,
demand. Theintermediate ow control paces the energy 1966; pp. 57–574.
ow from supply to demand, maintaining areal-time 2. Van Wylen, G.; Sonntag, R. The First Law of
energy balance. The result is reliable, consistent,and stable Thermo-
dynamics: Fundamentals of Classical Thermodynamics,2nd
operation at an appropriate demand-side target pressure. Ed.; Wiley: New York, 1973; pp. 90–97.

Compressed Air
Systems
Diane Schaub
Industrial and Systems Engineering, University of Florida, Gainesville, Florida, U.S.A.
Abstract
Compressed air is avaluable resource for manufacturers, allowing the use of pneumatic-driven hand tools,
which can be an ergonomic boon to employees. This resource comes with aprice, however, in the form of
higher energy costs. This article describes the use of compressed air and the creation and delivery of
compressed air from both asupply side and demand side approach. Amajor focus of this article is on the
costs associated with the generation of compressed air and ways to reduce the waste of this resource.
INTRODUCTION As these latter applications do not have the need for

portability, and can be performed morecheaply without
The first section of this entry focuses on the use of the additional process stepofcompressing air, their use is
compressed air and how it is generated. The section on fairly limited. Theeconomics of compressed air will be
generation is then separated into adiscussion of the supply discussedlater in this article.
side and demand side components of acompressed air
system.Finally, the costs associated with compressed air,
as well as sources of further information,are found at the
end of the entry. COMPRESSED AIR
SYSTEMS
Although arelatively simple-looking, self-contained air
OVERVIE compressor can be purchased at ahardwarestore, they are
W limited in size and thesesmall units (battery-powered,gas-
Compressed air systems couldbeconsidered aunique powered ,orplug-in models) are typically only to be used
source of energy despite the fact that they are actually to fill tires or in ate rafts. Our discussion from this point
powered by electricity.This similarity stems from the fact onward will focus on larger,commercial compressed air
that compressed air lines can be designed to allow modular systems.
tools to plug into the air lines, just like electrical devices The typical compressed air system is composed of:
can be powered by tapping into electrical outlets.
By far, the mostcommon use of compressed air is † One or more in-series compressors
to drive pneumatic tools, ranging from nail guns to † An air dryer and air filters
jackhammers to large drill presses. Pneumatic tools are † Areceiving tank (for storage)
favored over electric motor driven models because: † Piping
† Enduses
† They’re smaller,lighter, and more maneuverable.
† They deliver smo oth power and are not damaged by Compressed air systemsshouldbeperceivedas
overloading. possessing both asupply side and ademand side. Fig. 1
† They have the ability for infinite, variable speed and shows atypical block diagram of aindustrialcompressed
torque control, and can reach these very quickly. air system,with both the supply and demand side noted.
† They can be safer because they do not pose the These block diagramsare avery helpful first step in
potential hazards of electrical devices, particularly understandinghow to better manage compressed air
where water and gases are present. systems, as recommendedbyRef.1.
Additionaluses for compressed air in the manufactur- † Improvingand maintainingpeakcompressedair

ing sector may include:filtration or control systems, system performancerequiresaddressingboththe
driving conveyors, dehydration, aeration, or refrigeration. supplyand demand sides, as well as how the two
interact in order to have dependable,clean, dry, stable
air delivered at the properpressure. Awell-planned
Keywords:Compressor; Pneumatic; Supply side; Demandside; End-use balanced system will yield the cheapest and most
applications; Manufacturing. energy efficient results.
236

Compressed Air Systems 237
ASimplified Block Diagram
End End
Use Use
Compressor #1
DEMAND
SIDE
Control
Compressor #2 (Wet) Dryer Filter
Receiver
Receiver Dry
Compressor #3 SUPPLY
SIDE End End
Use Use
Includes compressor and air treatment

Includes
distribution,
storage and end use
Fig. 1 Schematic of acompressed air system.
COMPONENTSOFACOMPRESSE otherhand, lubricant-injected rotary screw compressors

D
AIR have the ability to trim to partialloadstomeet usage
SYSTEM needs,which can further save on their already lower
powercosts.
SupplySide
Another issuethat can greatlyimp act the energy
Athorough understanding of the end-use compressed air efficiency of air compressors is their control strategy.
needs,from both avolume and usage profile perspective, is Start/stop, load/unload ,and modulating(or throttling)
necessary in order to select the appropriate number and size control strategies can be used, depending on the facility’s
of air compressors. It is raretofind amanufacturing plant compressed air usage profile.
that has aconstant, uniform use of compressed air through- In order to deliver clean compressed air, filters are
out the day. Most manufacturing plants have cyclical installeddownstream from the air compressors. The filters
ow and volume demands due to production schedules, and remove particulates,somecondensate, andlubricant.
also desire back-up supply, so engineers typically plan for Regularreplacement of filters is necessary to prevent
morethan one air compressor to meet afacility’s needs. pressure drop, which results in athrottling effect. To illus-
Agoodstrategy is to sizeacompressor for abase load,and trate the filter’s importance, see the following example:
have one or more compressors staged to comeonline to
meetadditional compressed air demand. In designing a Example ( Replacement of aCompressed Air Filter
compressed air system,altitude,inlet air temperature, Element)
and relative humidity should be considered, as they impact
compressor capacity. More information on how to calculate Assumea100 hp compressor that operates continuously
the in uence of these design considerations can be found with an energy cost of seven cents/kWh, resulting in an
in Ref. 2, pp. 9–10. It may also be helpful to have different annual energy cost of $55,328. As the filter becomes
size compressors, so that they can be tailored to fit the clogged, assume the pressure drop increases to six psi
operating conditions. Additionally, asmall compressor or across the filter (as compared to atwo psi pressure drop for a
separate booster may be appro priate for off-shift operations new filter). Consider that this four psi increase can cost
or aspecial high pressure,periodic application. two percent of the annual required energy, or $1100,as
The vast majority of industrial compressors are of the compared to $375for anew filter element.
rotary screw variety, but double-acting reciprocating or Another component of acompressed air system is
centrifugal compressors are alsoavailable for specific the dryer(s). Thecompressingofair will condense outthe
applications. Rotaryscrewcompressors comeintwo moisture from the natural water vapor found in atmos-
configurations: lubricant-injectedorlubricant-free. Both pheric air. This liquid water can cause rust problemsinthe
have various pros and cons associated with their use. lines or, shouldcompressed air supply lines connect
Lub ricant-free rotary screw compressors require higher betweenbuildings, freeze in the winter. Compressed air
electrical demand, but assurenolubricant carryover. This shouldbedried to adew point at least 188 Fbelowthe
may be crucial when ultra-clean air is required. On the lowest ambient temperature of the demand side.

238 Compressed Air Systems
The various typesofdryers are: compressor. In fact, it isn’t unusualtofind apoorly

maintained system running acompressor that is only
† Refrigerated: This is the mostcommon type,with both feeding leaks. Some facilitieswillbury large portions of
low initial and operating costs. It can be subject to their distribution piping, which makefinding and repairing
freezing if operating at low capacities. leaks an expensive proposition. A3/16 0 0in. hole in a
† Regenerative desiccant: Typically operated in tandem system operating at 100 psig can cost over $5000 ayear.
between two twin dryers, with one operating and the Another operating consideration associated with the
other regenerating. The requiredvolumeofpurge air demand side is the cost of “normal production.” Decisions
needed to regenerate can increase the load or even cause to add additional applications shouldundergoarealistic
an idle compressor to be started. Heaters can be used in cost evaluation. Consider the following exampleofan
place of purge air, but present their own energy penalty. end-use application:
† Heat of compression: Similar to theregenerative
desiccant dryer, this type of dryer is available for Example (Addition of an End-Use Application)
lubricant-free rotary screw compressors and utilizesthe
hot dischargecompressed airtoregenerate the Aquarter inch orifice requiredtooperate apneumatic hand
desiccant. Their efficiency is affected by changing air tools at arecommended pressure of 100 psig was found to
temperatures and additional heat may be required for have a ow rate of 63.3 scfm (standard cubic feet per
low load situations. minute). After ayear of constant use, this equates to
† Deliquescent desiccant: Adissolvable desiccant is used. 33.3 MMcf (million cubic feet) of compressed air. If com-
Regular replacement of this resource is necessary, pressed air generation costs $300/MMcf, then the power
requiring laborand material costs. cost for this application will be approximately $10,000/
† Membran e-type: Aporous membrane separates water year. If we add additional operating costs of $170/MMcf to
vapor from the air and suppressesthe dew point. account for the operator maintaining the compressed air
Although thereisalow initial cost, these dryersare equipment and the maintenance, lubricant,and repair costs
appropriateonly for low-volume applications. for the system, we find that the cost of this new application
use is over $15,000/year. Compare this with less than
Airreceivers canbefound on either thesupplyside $2000/year to operate acomparable electrical tool.
(immediately afterthe compressor or thedryer)oronthe High costscan also be incu rred through the artificial
demand side,close to theapplication enduse.Air receivers demand associated with setting the compressor pressure
storecompressedair andhelpcover peak events of short level higher than needed. According to Ref. 3, p. 56,
duration.Ifsized properly,theycan greatlyreducethe supplying20% extra psig will force the system to consume
frequent loadingand unloadingofthe compressor,saving 20% moreair ow, resulting in 20% waste. Poor applica-
both energy andmaintenance costs. They also stabilize tions, such as stuck condensatedrains, personnel use of
system pressure,which improves performanceofthe enduse. compressed air for cooling or drying, or sparging (aerating
Other componentsassociated with the supply side may of liquids), also use up precious compressed air.
include aftercoolers or intercoolers (for lubricant-free
systems), moisture separators, and condensate drains.
Depending on the manufacturer, theselatter items may ESTIMATING NECESSARY
be packaged in asingle housing with the compressor itself. PRESSURE
SET
POINT
Demand Side The determination of the pressure set point for the air
compressors needs to be equated. Because of natural
Besides adownstream air receiver, the demand side pressure drops associated with the componentsofacom-
consistsofthe distribution system or pipin g, and the end- pressed air system,aswell as unrepaired air leaks, the final
use applications. Correct sizing of the distributionpiping point is more difficult to find than just dialing in the pressure
is acritical featureincompressed air system design in recommended by the end-use equipment manufacturer.
order to minimize energy costs. In fact, it is not unusual for plant personnel to reach the
The piping typically consists of rigid metal or plastic desired pressure by trial and error, increasing the set point
piping from the air compressor room to the general area of until equipment operators stop complaining about low
the end-use equipment. From this point, exible rubberor pressure.When possible, pressure measurements shouldbe
plastic tubing is used, which may be plumbed directly to made after each component of the compressed air system to
the end use, or have ashut-offvalvewith quick-co nnect monitor system performance. Flow or electrical readings
attachment points. This exible tubing may be subject to can also provide useful performancedata. More infor-
being run over by foot or equipment traffic and can wear mation on how to calculate optimum compressed air system
out over time. As aresult, air leaks can grow to epidemic settings can be found in Ref. 2, p. 205. Fig. 2 provides an
proportions, and greatly increasethe demand on the exampleofthe pressure drops that can occur alon gthe line.

Compressed Air Systems 239
Fig. 2 Estimating compressed air system pressure drop.
COSTS OF COMPRESSED it isn’t unusualtofind that compressed air can be the

AIR largest end user of electricity.
To operate aone hp air motor, seven to eight hp of
electrical energy are required. This largeenergy penalty,
alongwith thecommonemployeeperception that
compressed air is essentially afree resource, mak es it a ARTICLES OF FURTHER
challengetocontrolthe costsofcompressed air. INTEREST
Inadequate compressorcontrolschemes cancause The U.S. Department of Energy’s Industrial Technologies
multiplecompressors to run at partialloads, rather than Program sponsorscompressed air training andAir-
turn ing them off. Problems with poor maintenance can MasterC toolsthrough theirBestPracticespro grams.
increaseconsumption or cause pressure variability. In fact, Seehttp://www.oit.doe.gov/bestpracticesfor more

240 Compressed Air Systems
information. Theorganizationcharged with actually 2. Scales, W.; McCulloch, D. Best Practices for Compressed Air
deliveringthe compressed air training can be found at: Systems;The Compressed Air Challenge, 2003; 9–10.
http://www.compressedairchallenge.org . 3. Taranto, T. Compressed Air Management;ConservAIR
Technologies Industrial Seminar Series, Conservair Tech-
nologies Company, LLP, Kenosha, WI, ( www.conservair.
com )1998; 56.
REFERENCE
S
1. Fundamentals of Compressed Air Systems;Training Manual
for the Compressed Air Challenge, Prepared by Laurel and BIBLIOGRAPH
Associates, Ltd and Resource Dynamics Corporation and Y
presented by the trade association known as the Compressed 1. Advanced Management of Compressed Air Systems;Training
Air Challenge, ( www.compressedairchallenge.org )1999. Manual for the Compressed Air Challenge, 1999; 205.

Compressed Air Systems: *
Optimization
R. Scot Foss
IR Air Solutions, Davidson, North Carolina, U.S.A.
Abstract
This article will provide you with an complete action plan to optimize your compressed air system
including compressor optimization, demand management, density management, and storage in avariety of
different applications.
Compressed air represents one of the most critical regulation, which is adjustedlower than the lowest
utilities in mostproduction and process environments. The supply pressure. If it is not possibletoachieve this
efficiency of acompressed air system is 100% energy in with operator discipline, then you mustuse a
and, when perfect,produces 11% usefulworkout. demand controller or expander at acentral location
Understanding this,itwill cost more to operate a adjusted in the same man ner.
compressor in the first year than it costs to buy and install. 2. Reduce the pressure differentials on installation
Despitethisharsh information, power is thrown at componentssuch as filters, regulators, lubricators,
symptoms of undefinedproblemsevery day. The tube, hose, and disconnects on the demand side of
opportunities of reducing operating cost and energy in the system.The intent is to operate demand at the
air systems is typically more than 50%. This session will lowest possiblesupplypressure on criticalhigh-
carveout aplan of attack to optimize the supply and pressure applications.
demand systemically and yield the lowest demand at the 3. Flat line thehighrateof ow,intermittent
highest rate of standardcubic feet per kilowatt of energy. applications with dedicated storage and metered
There are anumber of essential actions that need to be recovery.This is muchlike abattery charger or
takentooptimize the compressed air systems. You need to water tower application. This can alsobeapressure
minimize demand,control theexpansion of theair, driver for the operating protocol. You will slightly
distributeitwhile minimizing energy loss, store potential increase the base usage and eliminate peaks.
energy,and compressthe airefficiently. Otherthan 4. Review and add as necessary general and control
operatingthe compressors, as efficientlyaspossible, storage to slow the rate of change in the system.
everything else seemstoelude most everyone. This This will allow youtomaintain ahigher point of
workcannot be done theoretically on each piece of use pressure if necessarywithout increasing the
equipment only in the compressor room. It must be done supply pressure. If there is any diligence used, you
systemically. More efficient compressors makemoreair can normally reducethe supplypressure simul-
with the same amount of power. They cost more and can taneously.
be an important part of awell-operated system. On the 5. Upgradethe quality of informationtotrack
otherhand, if you throw amoreefficient compressor at a progress and improv edecisionmaking. This should
highly inefficient system,you will waste more air at the include a ow meter and demand pressure monitor
sameoperating cost and save nothing. at the discharge of the demand controller or the
expander. If you do not use ademand controller,
recognizethat demand is only accurately displayed
CONTROLLING DEMAND IN THE when the demand exceeds the supply. This is
SYSTEM referred to as anegativerate of change.When
1. Control the expansion of the compressed air to the supply exceeds demand, which is apositive rate of
point of use. You mustcontrol 100% of userswith change, youare measuring supply response to
demand. The system will take whatever supply
power you throw at it.A450 scfm negativerate of
*
This entry originally appeared as “Optimizing the Compressed Air change will recover to the original pressure in
System” in Energy Engineering,Vol. 102, No. 4, 2005. Reprinted with 1min, if we respond with a200 hp compressor.If
permission from AEE/Fairmont Press. we throw a400 hp compressor at the event,itwill
recover in 15 satamore rapid rise in pressure.The
Keywords:Compressors; Compressed air; Storage; Demand controls;
Expanders; Metered storage; Leaks; Transient events; Air dryers and inefficiency is the part load energy of the larger
filters; Potential energy. compressor for the balance of the 45 s. If we match
241

242 Compressed Air Systems: Optimization
the event with a100 hp compressor, the pressure alternatives.Use electricity whenever possiblefor
will hold at the load pressure of the compressor its better wire to work energy relationship.
until the event stops, at which time, the pressure 4. Reduce the size of demand events as seen by the
will reco ver at the same rate of rise as the initial system including high rampapplications. This can
rate of decay. be accomplished by slowing down the introduction
6. Review and add as necessary general and control of these eventsinto the system. This can be done by
storage to slow the rate of change in the system. opening the demand valve slower manually or
This will allow you to maintain ahigher point of automatically. This reduces the “ramp in” rate of
use pressure,ifnecessary, without increasing the ow, so that the supply including control storage
supply pressure. If there is any diligence used, you can match the event limiting the ultimate pressure
can normallyreducethe supply pressure drop, which would result.
simultaneously. 5. Regulate all pointsofuse, even if youhave
installed ademand controller or expander in the
main supplysystem’s piping. Make sure that the set
pointsonthe regulators are equal to the minimum
supplypressureminus thepoint of usefilter
REDUCING DEMAND IN THE and regulator pressure drop or less. If you allow
SYSTEM for a2–3 psig margin below this value, small leaks
1. Developaleak benchmarking program on a and filter dirt loading will not cause frequent
gradual reduction of the tolerance volume. Select changesinprocess performance.
alevel at aknownlow load, and repair your way to 6. Limit the coincidence of events that cause peak
that level. Every several weeks, check the low load demands in the system.Thisincludesminimizing
and scan the system using an ultra sonic leak the blow duration on timer drains and adjusting
detector. Find and repair the largest leaks found to intervals seasonally for relative humidity. Move
bring the system back intobenchmark. When you large events to low load timeswhere possible.
are comfortable with this level, lower the level and 7. Shut offall air usingequipment whennot in use.
begin again. You will reach apoint where there are Make sure that the shut offvalves are ergonomi-
so many small leaks to fix during the benchmarking cally installed, so that operatorscan easily reach
period, the labor hours cannot be justified. At this them. If this does not work, install solenoidshut off
point return to the previoushighertolerance value. valves that are tied into the electrical shut offonthe
Record the types and nature of the leaks that you machine, work station, or process.
are fixing, so that you can leverage this information
into buyingmoreleak resistant componentsand
improving best practices installations. Notethat it
is important that the reduction of demand does not
cause the demand pressure to rise. If it does, then STORE POTENTIAL ENERGY TO
other unregulated users will increase at the elevated SUPPORT
TRANSIENT EVENTS INCLUDING
pressure. That is why it is so important to have A
COMPRESSOR FAILURE IN
demand controlsinstalledbefore youbecome THE
SUPPLYSYSTE
aggressive in demand reduction. It is also important M
to offload alinear amount of supplyenergyfor the 1. Convert enough kinetic energy to potential energy
demand reductions. so that you can handle largest event without turning
2. Eliminate all open compressed air blowing appli- on another compressor during normal operation.
cations and replace with low pressure centrifugalor If you do this, you will alsohandleall of the smaller
positive displacement blowers, if at all possible. If transient events that are notcontrolled from
it is not possibletouse blowers, applyspecialty air downstream. This can include the coincidental
volume reducing nozzles for the application. Take impact of athird to first shift startup. Remember
your time with these applications developing the that storage is afunction of the capacity to store air
thrust per squareinch as close as possibletothe timesthe useful differential across it. If you are
open blowing application. You will alsoneed to operating constant pressure compressor controls
filter the air for specialty nozzles, as they will easily and they operate correctly, no amount of capaci-
plug up with pipe debris. Whenever possible, use a tancewill generate any useful storage.
solenoid valvetoshutofthe aironcyclical 2. Storeenough air on the supply side of the system to
applications. manage adesired pressure drop, while bringin gup
3. Replace all applications, which are poor users of abackup compressor to replace afailedone. The
compressed air.Focus on operating cost intent would be that the event will have no impact

Compressed Air Systems: Optimization 243
on the process or production serviced by the ow and storage keeps it functioning. Mass ow
system.The intent is to operate onlythe restrictions are differential pressures in the system,
supply that is requiredatany time with everything which change as asquare function of owchange.
else off. It is important to design or retrofit your system for a
Example :largest compressorZ 1600 scfm, maximumdifferentialathighest ow, highest
maximum allowablepressure drop from the load temperature, and lowest inlet pressure. This will
pressure on the back up compressorZ 10 psid, produce the highest differential pressure across the
permissive time to load the compressor from a componentsbeing evaluated. Althoughweare
cold start signal to full loadZ 15 s, atmospheric recommending aconservativeapproach towards
pressureZ 14.3 psia, gallons per standard cubic this process, the piping distribution system should
feetZ 7.48 gal not be made intentionally oversized or all the same
size for convenience. Oversized piping will not
1600 ! ð 15= 60Þ ! ð 14: 3 ! = 10Þ ! 7 : 48 provideeconomicalstorageand will makeit
difficult for supply to see demand efficiently. A
Z 4278: 6gal reasonable differential pressure would be 1–2 psid
from the discharge of the cleanup equipment at the
3. Create enough storage to control the maximum supply or the discharge of the demand controller, as
load cycles per time period on any trim compres- it applies to your system, to the farthest point in the
sor. It is safe to say that 3min load–unload cycles demand system at the previously discusseddesign
or longerwouldbedesirable on any positive conditions.
displacement compressor.This can get trickier on 3. In most systemsthat have distribution problems,
large dynamic compressors, but it is not imposs- you should minimize waste and at line transient
ible. users with dedicated storage and metered recovery
4. If the size of any event or compressor is too large to at the point of use before considering making
handle with control storage or you wanttoprotect changes in the pipingdistribution system.
the system and production against an electrical
outage, single phase, or brown out, of ine high- As little as a10%–20% demand reduction at the peak
pressure peak shaving would be the mostdesirable condition can be sufficient to eliminate themost
approach to minimize on board power. It would not distribution losses and the requirement for piping retrofits.
be unusualtostore30–40,000ft 3 of air in a
100 psig differential supported by a20hpcom-

pressor of ine. You would then introduce the air
back into the system on variety of different cues or REDUCE SUPPLYENERGY
logic patterns to supportthe various events. WHEREVER
POSSIBL
Note that it is the intent of all potential energy E
applications to either prevent the normal operation 1. When 100% of demand is at alower pressure than
of an additional compressor, extend the mechanical the lowest supply pressure,set up the supply
life of acompressor or compressors, or both. Well pressure to optimize the pound per kilowatt of
applied storage will increasethe base load in the compressed air energy for the on board compres-
system slightly, and eliminate the requirement for sors. Operate all compressors that need to be on at
added compressors during peak plus the inefficient out and optimized exceptone compressor trimming
part load in betweenpeaks. and all othercompressors offregardless of inlet
conditions or relative demand load.You must
optimize the compressor and the motor simul-
DISTRIBUTE THE COMPRESSEDAIR, taneously. Optimal means the mostpounds or
WHILE
MINIMIZING ENERGY standardcubicfeet at the optimal density (pressure
LOSSES andtemperature), whilemanagingthe highest
1. The concept of designorredesign shouldbeto power factor and motor efficiency simultaneously.
minimize the highest amount of air mass or volume In this scenario, the trim compressor is the only
of air and the distancethat the air must owto compromise to “optimal”assuming you can
supportany part of the system from supply to maintainarange of supply pressure across the
demand. range of load conditions that relates to optimal on
2. Resistance to ow is necessaryinthe system. the base load compressors. Another option is to
Without resistanceto ow there is no ow. As the trim with variablefrequencydrive compressors
system is open on both ends of the system all of usingstorage continuously,while adding and
the time to alarger or lesser degree, resistance to subtracting base load compressors.The Variable

244 Compressed Air Systems: Optimization
frequencydrive (VFD) compressor or compressors totaldifferentialacrossall cleanup equipment

will displace or fill in the removal or additionofa shouldnot in uence the total connected horse-
base. In this case, you will optimize both the base poweronthe compressors by morethan 4% at the
load compressors and the trim compressors at the worst case maintenance condition.
same time. 5. Use amastersignal for the compressors locatedin
Notethat this is calleda“Bellows Effect” the dry clean storage downstream of the contami-
operating protocol. nant control equipment. If the signals are in the
2. Base compressors should always be selected based compressors upstreamofthe cleanup equipment,
on the best energy efficiency. Trim compressors thecompressors controls will respond to the
should be selected first on operating speed to cold demand interpreted throughthe differential
or hot start and shut offcapabilities, and secondly pressure, which changes as asquare function of
on their exibility for automation interface.Ifyou owchange. This causesthe compressors to over
are trimming with VFD/s,the same requirements shot and under shot,which results in hunting. This
areapplicable. Thistypicallytranslates into requires excess energy to compensate.
smaller, less permissive compressors. You must Please note in the illustration that we have
be certainthat the total trim capacity (one, two,or installed athree-way valve so that you can return to
three trim compressors) is equal to or larger in localcontrol signals whenyou wish to isolate the
capacity than the largest base compressor in the compressor from the system.Itisalso important to
supply arrangement. This will assure that there are note that the adjustment of the compressor controls,
no gaping holesinsupply,sothat you can make with amaster control signal, shouldbebased on
smooth transitions from one powerlevel to the co ntrolling downstream of the cleanup.Ifthe
next.Supplysystems that do not have this pressure across the cleanup equipment is 10 psid,
capability end up running too muchpowerpart whenyou movedthe signal, you would also wantto
loaded all of the time to support the transitions. reducethe control set points on the compressor/s an
Rememberthat bigger is more expensive. additional 10 psid. This is because you will absorb
3. Develop an operating profile for the supplysystem, the differential when you move the signal and
which optimizesthe compressors based on afull without adjusting the operating set points for the
range of usage and conditions. In mostsystems, the compressor/s, you may overload the motor.
only time the system is remotely efficient is during 6. Develop an operating profilewhich takes control
peak load. It generally goes down hill during storage, set points of thecompressors, signal
lighter or low load. Also evaluatethe fullrange of locations, and differentials into account. Putit
system’s usage against the full range of ambient down on paper prior to implementing it and check
inlet and cooling conditions to determine how the the range of conditions to make sure it will work.
system will work before you make any final plans Do not put fudge factors into the profile. This is not
on equipment selection. Make every attempt to an art form. It is ascience.Ifyou are not sure of
manage peaks with potential energy instead of on whatyou are doing, contact atechnology firm who
line power. You must alsoevaluate the risk of a can assist yo u. Literally, 95% of all compressor
unit failure in order to have asolid curtailment profiles are not set up correctly. Most engineering
plan. If brownouts or black outs are common, you firmsthat design systemsselect the equipment and
must include this in your plan. never think throughthe operating protocolor
4. Unload all unnecessary ancillary power, such as profileprior to installation.
dryers, pumps, fans, etc. through the use of more 7. Finally, you must getthe system to operate
efficientcontrols andmotor drivers. Size all effectivelyand efficiently beforeyou automate it.
filtration and dryer equipment for atotal differ- More than 90%ofthe time, userstry to apply
ential of 3.5–5 psid. The differential should be at automation to asystem to get it to workproperly.
the highest inlet ow, highest inlet temperature, If youautomate asystem that does not work, you
and the lowest inlet pressure.The differential on will have an automated mess. You must be able to
the filtration should be in awet and clean condition. get it to work correctly on the local controls first.
Plan the additional differential from dirt loading When and if you automate, keep in mind that their
when selecting the compressionequipment, so that purpose is to refinethe operating cost and
you do not overload the motor drives as you will reliability issues across all conditions unattended.
absorb theadded differentialatthe air end Automate the operation based on at least rate of
discharge. We would recommend no more than change,storage, time, and pressure.You may even
an additional 1.5–2 psid on the total filters. There wish to add aselective rate of change protocol,
are filters available to accomplish this with a which chooses the correct compressor for the
change every 5–6 yearsatthis dirt loading rate. The situation. Take your time and testyourconcept

Compressed Air Systems: Optimization 245
prior to making the decision by preparing algo- SUMMAR

rithms including transitions of powerand demand Y
including failure scenarios. Keepinmind that you Acompressed air system is ahighlyinteractive configu-
do not have to match the event in the system. You rationwithall aspectsaffecting all otheraspects.
only need to slow it down so you can wait longer. Developing an action plan to improvethe efficiency and
The essence of amasterfully designed system is the reducethe operating cost can be rewarding, but mustbe
ability to control demand by matching transient done in the correct order to enjoy the success and avoid
events as quickly as possible with an expander or production inconvenience. It is aprocess of black and
demand controller serviced with potential energy. white with alot of gray in between. Far too many owners
wanttobuy asolution, rather than apply one. Problem
Oncethis is accomplished, the compressors’ control job definition, metrology, andcarefully planning areall
is managing control storage by replenishing it as slowlyas essential. When you have completed the action plan, do
possible. The longeryou can take,the lessenergyyou not forget to measurethe results. Validation is necessaryto
will use. supportyou return on investment strategy.

Cooling
Towers
Ruth Mossad
Faculty of Engineering and Surveying, University of Southern Queensland, Toowoomba,
Queensland, Australia
Abstract
Cooling is necessary to many industrial processes, such as power generation units; refrigeration and air
conditioning plants; and the manufacturing, chemical, petrochemical, and petroleum industries. As recently
as 20 years ago, cooling towers were more the exception than the rule in the industry because of their high
operating cost and the large capital required for their construction. Due to the recent stringent
environmental protections, cooling towers became more common. Cooling towers range in sizes and types.
Wet, dry, and hybrid are the main types, and each type has many variations in design according to the way
the uids aremoved through the system. Some of the advantages and disadvantages of these types, methods
of determining their performance, and some terminology common to the cooling industry are presented in
this entry.
INTRODUCTION hundreds of thousands of litres per minutesupplied in

pipes as huge as 5min diameter in alarge powerplant.
Most industrialproduction processes need cooling of the Cooling towersare believed to be only the direct
working uid to operate efficiently and safely. Refineries, contact type heat exchangers. They can be direct or
steel mills, petrochemical manufacturing plants, electric indirect, however, and they are also characterized in many
utilities, and paper mills all rely heavily on equipment or other different waysbased on the type of uid beingused
processesthat require efficienttemperature control. Cool- in the cooling process, the means by which the uids are
ing water systemscontrol thesetemperatures by transfer- moved, and the way the two uids (hot and cold)move
ring heat from hot process uids intocooling water. with respect to each other. Description of these types,
Through this process, the cooling water itself gets hot, and some of their advantagesand disadvantages, and methods
before it can be used again, it must either be cooled or be of estimation of their performance are discussed in the
replacedbyafresh supply of cool water. following sections. Some useful terms common to cooling
Acooling system in which the water used in cooling towers industry are also given. This information is a
processesorequipment is discharged to waste is called collectionofmaterials published in the list of references
once-through cooling. Characteristically, it involves large and Web sites at the end of this article.
volumes of water and small increases in water tempera-
ture. Once-through cooling is usually employed when
water is readily available in large volumeatlow cost. TYPES OF COOLING
Commonsources are rivers, lakes, and wells, where the TOWERS
only cost involved is that of pumping. But with today’s Cooling towersare classified mainlyonthe basis of the
need for water conservation and minimal environmental type of uid used in the cooling process—water or air.
impact, industry is turning more and more to recycling Theseare three types: wet(evaporative), dry(non-
water in whatare calledcooling towers. evaporative), and wet–dry(called hybrid).
Recently,cooling towersare becoming widelyused in
most industrial powergeneration units; refrigeration and Wet Cooling Towers
air conditioning plants; and the manufacturing,chemical,
petrochemical, and petroleum industries to discard waste The wet type is more common in large cooling towers,
heat to the environment. They range in sizes—the smallest such as in electrical powergeneration. It is adirect contact
cooling towers are designed to handlewater streamsof heat exchanger,inwhich hot water from the condenser and
only afew litres of water per minute suppliedinsmall cooling air come into direct contact.The water ows in
pipes like thoseinaresidence, whereasthe largest cool either open circuit or closed circuit. In open circuit,
cooling water is pumped into asystem of pipes, nozzles,
and sprayers within the tower,and is drawn by gravity into
Keywords:Cooling; Towers; Wet cooling; Dry cooling; Hybrid cooling; apond below ( Fig. 1 ). Air from the atmosphere enters the
Plume; Performance; Counter- ow;Cross- ow. tower from the bottom of the tower and ows upward
246

Cooling Towers 247
Air Air outlet

outlet
Pump
Eliminator
Fill
Waterinlet Water Inlet
Air
inlet Fill
Water
outlet Condenser Fan
Pond
Air inlet
Fig. 1 Open-circuit cooling tower. Water outlet
Fig. 3 Mechanical-draft counter- ow tower.

thro ugh the falling water. The two uids go through a
material that is provided to increasethe surface area of
contact betweenthem, which is calledpacking (or fill). the air and evaporation of some of the water) to the
The heated and moisture-laden air leaving the fill is atmosphere. Operation of the closed cooling towersisvery
discharged to the atmosphere at apoint remoteenough similar to the open cooling tower,therefore, with one
from the air inletstoprevent it from beingdrawn back into exception: the process uid beingcooled is contained in a
the cooling tower.The water is collected at the bottom of “closed”circuit andisnot exposed directly to the
the tower and then recirculated to remove moreheat atmosphere or the recirculated externalwater.
from the condenser. The temperature of the cold water Closed systemsoffer theadvantages of precise
entering the condenser will determine the steam condensate temperature control (which is criticalinmany process
temperature and, hence,the backpressure, which impacts applications) and low treatment cost. Becauseasecondary
the efficiency of the whole powergeneration system. cooling system and heat exchangersare neededtocool the
The closed-circuit cooling tower (Fig. 2) involves no closed system,highercapitaland operating costs are
disadvantages of this design.
direct contact of the air and the liquid—usually, water or a
glycol mixture—that is beingcooled. This cooling tower
has two separate uid circuits.Water is recirculated in an Mechanical and Natural Draft
external circuit outside aclosed circuitmade of tube
bundles or coils containing the hot uid beingcooled. Air In wet cooling towers, there are two types, based on the
is drawn through the recirculatingwater cascading over mechanism by which air is being circulated: mechanical
the outside of the hot tubes, providing evaporative cooling draft and natural draft. The mechanical draft uses fans (one
similar to an open cooling tower. In operation, the heat or more) to move large quantities of air throughthe tower.
ows from the internal uid circuit, through the tube walls The mechanical draft is againdivided into two types,
of the coils, to the externalcircuit and then (by heating of based on the location of the air fan: forced and induced. In
the case of the forced draft, the fan is located at the air
entry at the base of the tower;inthe induced draft, the fan
Air outlet is located at the air exit at the top of the tower. Theinduced
draft produces more uniform air ow,which enhances its
Pump
Water inlet effectiveness overthe forced draftand reduces the
Fill possibility of exhaust air recirculation.
There are manyconfigurations of mechanical draft
cooling towersthat depend on the way the two uids ow
Air with respect to each other, such as counter ow, cross ow,
inlet and mixed ow. In acounter- ow cooling tower,air
travelsupward through the fill opposite to the downward
Water outlet motionofthe water (Figs. 3and 4).
In across- ow cooling tower, air moves horizontally
through the fill as the water moves downward ( Fig. 5
and 6). In amixed- ow tower, air movesinadirection that
Cooling fluid
is acombination of acounter ow and across o w. Cross-
circuit
ow towershave greater air intake area,which results in
Fig. 2 Closed-circuit cooling tower. considerably lower towers. This means that they have low

248 Cooling Towers
Air outlet Air outlet Fan
Fan Water inlet Water inlet

Eliminator
Water inlet
Fill
Fill Air inlet Air inlet
Air inlet Air inlet

Water outlet
Fig. 6 Induced-draft cross- ow tower.

Water outlet
Fig. 4 Induced-draft counter- ow tower. back to the chimneythat would otherwise reducethe
performanceofthe cooling tower. The immediate vicinity
is not affected by plumes from the cooling tower,because
pressure drop in relation to their capacityand lower fan
the hot-airexit is situated at avery high elevation. The
power requirement, leading to lower energy costs, but the only drawbacktonatural-draft towers is that they are large.
risk of recirculation increases in tower exhaustair. On the There is also the assisted-draft tower, which is a
other hand, counter- ow arrangements occupy less oor natural-draft tower with some fans added at the air entry
space than cross- ow towersbut are taller for agiven that help reducethe size of the tower.Natural-draft towers
capacity,sothey require higherpump heads.Itshows are also divided into counter ow and cross ow, defined
better tower performance, since, the driest air contacts in asimilar fashion to the mechanical-draft towers.
the coldest water,producing higherdriving force to
the heat. Heat Exchange in Wet Cooling Towers
Anatural-draft tower is alarge chimneyand typically
has ahyperbolic profile, which is chosenfor its structural The type of heat rejection in awet cooling tower is termed
capability of withstanding wind-induced stressesand evaporative, in that it allows asmall portion of the water
vibration; also, it requiresless material. The design creates being cooled to evaporateinto amoving air stream to
achimneyeffect that causesair to move by natural provide significant cooling to the rest of that water stream.
convection through the fill region, which is located inside The heat from the water stream transferred to the air
the base of the chimney( Fig. 7 ). As the air gets warmer stream raises theair’stemperature andits relative
from the contact with the cooling water,itgets lighter; humidity to 100%, and then this air is discharged to the
buoyancyforces drive the air to the top of the tower and atmosphere. Theambientair wet-bulb temperature is the
into the atmosphere, and draw fresh air into the bottom of controllingfactor in recirculatedsystems andwill
the tower.The major economical advantage of natural- determinethe steam condensate temperature. Evapora-
draft cooling towers is the extremely low auxiliary power tive-heat rejection devices such as cooling towersare
consumption. Because thereare no rotatingparts, commonly used to provide significantly lower water
operationalsafety andlow maintenancecosts are temperatures than areachievablewith air-cooled or
maintained. The great distancebetweenair inlet and air “dry” heat rejection devices. The evaporative process
exit in the cooling tower prevents any hot-airrecirculation enhances the performance of wet cooling towers over dry
cooling towers severalfold due to the change in both
sensible and latent heats.
Water inlet
Fill
Consequences and Concerns for Wet
Cooling Towers
Fan Air outlet Wet cooling towers are the mostcommon type due to their
high effectiveness, but there are some drawbacks. If cooled
water is returned from the cooling tower to be reused, as in
the circulating systems, some water mustbeadded to
replace, or make up, theamount of thewater that
Water outlet
evaporates. Becauseevaporation consistsofpure water,
Fig. 5 Mechanical-draft cross- ow tower. the concentration of dissolved minerals and othersolids in

Cooling Towers 249
Air outlet
Eliminator
Water inlet
Air inlet Air inlet
Water outlet Fig. 7 Natural-draft cooling tower.
circulating water will tend to increase unless some means droplets. To wer maintenan ce and operation levels can also
of dissolved-solidscontrol (such as blow-down) is in uence the formation of drift,such as excessive water
provided. Blow-downisthe amount of the circulating ow, excessive air ow,orbypassing the tower drift
water that is removed to maintain the quantityofdissolved eliminators can increase drift emissions. Types of drift-
solids and other impurities at an acceptable level. Some eliminatorconfigurationsinclude herringbone(blade-
water is alsolost from droplets being carried out with the type), wave form, and cellular (or honeycomb).The
exhaustair (drift). The makeup amount mustequal the cellular units generallyare themostefficient. Drift
total of the evaporation, blow-down,drift, and other water eliminators aremadeofvarious materials, such as
losses (such as wind blow-out and leakage) to maintaina ceramics, fibre-reinforced cement,fibreglass, metal,
steady water level. Devicessuch as wind screens, louvers, plastic, and wood.
splash de ectors, and water diverters are used to limit Otherunfavourable environmentalimpacts of wet
theselosses. cooling are pollutant discharge—e.g., zinc, chlo rine, and
The mag nitude of drift loss is in uenced by the number chromium (chromium is used to protect cooling-system
and size of droplets produced withinthe cooling tower, equipmentfromcorrosion)—tothe atmosphere.The
which in turn are determined by the fill design, the air and spread of Legionnaires’ disease is due to the bacteria
water patterns, and other interrelated factors. Driftis that thrive at temperatures typical in wet cooling systems
typically reduced by installing baf elike devices, called and that can be transported through air aerosolsformed in
drift eliminators, through which the air must travel after cooling towers. Other impacts are mineral drift and the
formation of visual plumes.Under certainconditions, a
leaving the fill and sprayzones of the tower to collectthe
cooling-tower plume may present fogging or icing hazards
to its surroundings (Fig. 8). Some interesting pictures of
cooling towerscan be found at The Virtual Nuclear
Plumes Tourist: Nuclear Power Plants Around the World (www.
nucleartourist.com/systems/ct.htm).
Types of Fill(Packing)
To wers
The fill may consist of multiplevertical, wetted surfaces
upon which athin film of water spreads (film fill); several
levels of horizontal splashelements, which create a
cascade of many small droplets that have alarge combined
surface area (splash fill); or trickle, which is acombination
of the film-and splash-type fills. Awide varietyof
Fig. 8 Visual plume from cooling towers. materials and geometries have been used for packing, such

250 Cooling Towers
as corrugated roofing sheets made of cement-based or In powergeneration applications, the heat transfer
plastic material, timber laths of triangularorrectangular between air and cooling water is achieved by convection,
cross section, plastic-impregnated paper honeycomb, and and the driving force of this cooling process is the
complexcellular geometries madeofthin plastic material. approach temperature (which is defined as cooling water
temperature at outlet and air temperature at inlet)—not by
Thermo-Fluid Dynamic Efficiencyin evaporation, as in awet cooling tower. Some of the
Cooling Towers advantagesofdry cooling towers are that they do not need
any makeup cooling water or water treatment and they do
To choose the most convenient fill, you need to find the not generateplume, fog, mineral drift, and disposal issues
one that produces the maximum heat transfer with the associated with wet cooling. Size is aconcern in dry
minimum pressure drop. Other factors to be considered cooling systems, however, because dry cooling is not as
are the physical and chemical characteristics requiredfor effective as wet cooling; the towers have to be much larger
the water to cool, fouling properties,suspended materials, to achieve comparable heat rejection. Another challenge to
etc. Sirena [1 ] has suggestedaThermo-Fluid Dynamic direct dry cooling is operational control in regard to how to
Efficiency that can be used to compare and select afill balancethe steam ow to keep thedesired steam
materialfor aparticular application. This efficiency is condensationtemperature (and, hence, turbineback-
defined as the ratio of the number of units of diffusion to the pressure), which varies with loading. Other disadvantages
pressure drop coefficient. In this paper,the pressure drop of dry cooling are increase in noise, plume recirculation,
coefficient is given for somecommercial fill materials. maintenance of manycomponents, and energy penalties
Al-Nimr [2 ] has studied the dynamic thermal behaviour of caused by the variation sofdaily temperature and increases
cooling towerscontaining packing material and was able to in air emissions.
predict closed form solutions for the transient and steady Dry-cooling performance dependsonthe ambientair
performance of acounter- ow cooling tower. dry-bulb temperature (i.e., the sensible heat)instead of the
wet-bulb temperature in the case of awet cooling tower.
Dry Cooling Dry cooling becomes more economical when the approach
temperature becomes considerably high. Other factorsthat
Dry cooling towers transfer heat to the atmosphere without affect the performance of dry cooling towersare the
the evaporative loss of water. Dry cooling is capable of crosswind speed and the way that the heat exchanger
only smaller temperature variations(around10 8 C), unlike bundles are arranged. The effect of crosswinds at different
wet cooling. Similar to wet cooling, there are twotypes of speeds and the effect of adding windbreak walls on the
dry cooling: direct and indirect. Direct dry cooling systems thermal performance of natural-draft dry cooling towers is
utilize air directly to condense steam, which is exhausted given by Al-Waked and Behnia. [3 ] Theeff ect of arranging
from the turbine into ducts and headers for distribution the heat exchanger bundles—either vertically aroun dthe
into rows of small-diameter finned tubes (i.e.,heat circumference of the tower or horizontally in the inlet
exchanger). Indirectdry cooling, which is also knownas cross sectionofthe tower—is givenbyduPreez and
the Heller System, utilises aclosed-cycle water cooling Krö ger. [4 ]
system as the primary coolant to condense steam, and the
heated water is air cooled. The cooling water ows Wet–Dry Cooling (Hybrid)
through bundles of tubeswith air owing over them, but
the cooling air never comes into contact with the cooling Combined wet–drycooling towerswere introduced due to
water. In both systems, direct and indirect, the ow rateof the recent stringent environmental protection laws. These
air required to achieve the samecooling capacity will have towers effectively suppress detrimental plume formation at
to be three or more timesgreater than in awet cooling an efficiency levelcomparable to that of wet cooling
tower,sothe tower will have to be muchlarger and more towers. In hybrid wet–drysystems, the hot water from the
expensive. power station condenser is cooled to the design discharge
Cooling towersare also characterizedbythe means by temperature as it passesinseries first through the dry
which air is moved. Similar to wet cooling towers, dry section and then throughthe wet sectionofthe tower.The
cooling towerscan be mechanical draft, natural draft, or low-humidity hot air stream from the dry system is mixed
fan-assisted draft. Mech anical-draft towersrely on power- with the moist warm air, leaving the tower at humidity
driven fans to draw or force the air through the tower. levels sufficiently low to prevent the formation of visible
Natural-draft cooling towers use the buoyancyofthe plumes.The wetand drycomponentscan be used
exhaustair rising in atall chimneytoprovide the draft. A separatelyorsimultaneouslyfor either water conservation
fan-assisted natural-draft cooling tower employs mechan- or plume abatement purposes. At low ambienttempera-
ical drafttoaugment the buoyancyeffect. Many early tures, the cooling tower can be operated as adry cooling
cooling towers relied only on prevailingwindtogenerate tower only, whereasathigh temperatures, it can be used as
the draft of air. awet cooling tower only to achieve the required cooling

Cooling Towers 251
without the risk of plume formation; the dry air is not put ( h Z KaV / L ), where K is theaverage mass transfer
into operation until the ambienttemperature startstofall. coefficient of condensed steam, a is the area of transfer
The design and construction of hybridcooling towers surface per unit volume, and V is the effective volume.
[ 5]
are more complicated, and according to Streng, the Manufactures supplycharts for their cooling towers that
following data need to be specified for winterand summer present the tower characteristics as afunction of L / G and
operation. These data are thermal performance; cooling the difference between the (outlet cooling water tempera-
water oworcooling range,which is the difference ture (CWT) and air wet-bulb temperature (WBT).
betweenthe water temperature at inlet and the water Cooling-tower performance can be specified from the
temperature at outlet; ambienttemperature; criteria for following parameters: water mass ow rate, inlet and exit
operating without plume; sound attenuation regulations; temperature of water,and atmospheric wet- and dry-bulb
and limitations with respect to the erection area or overall temperatures of air. Many researchers have attempted to
height and operating weight and water analysis of the analyze wetcoolingsystems to estimate theirper-
makeup water.Inhis work, [5 ] the construction, including formance. Abasic theory of wet-cooling-tower operation
material selection and automatic operation of the cooling was first proposed by Walker. [6 ] Thepractical use of basic
system, is discussedindetail. differential equations, however, was first presented by

When acombination of wet and dry cooling technology Merkel, [7 ] who combined the equations for heat and water
is used,depending on system configuration, water vapor transfer. He showed the utilityoftotal heat or
consumptioncan approach that of recirculatingwet enthalpy difference as adriving force to allow for both
systems or can be muchlower.Design stud ies have sensibleand latent heats. The basic approximations in
rangedfrom 30 to 98% reduction in water use compared Merkel’s theory are
with all wet recirculating systems. As the hybrid cooling
towers conformwelltothe stringentenvironmental † Theresistance for heat transfer in the liquid film is
protection requirements and to the standardoperation negligible.
reliability set for cooling systems, it is expected, therefore, † Themass ow rate of water per unit of cross-sectional
[ 5]
that they will become more widespread. area of the tower is constant (i.e., there is no loss of
water due to evaporation).
† Thespecific heat of the air-stream mixture at constant
PERFORMANCE AND RATING pressure is the sameasthat of the dry air.
OF
COOLING † TheLewis number (which relates heat transfer to mass
TOWERS transfer) for humid air is unity.
In power generation, lower turbine backpressures are † Theair exiting the tower is saturated with water vapor.
achieved whensteam condensate temperatures are lower.
Designing and operating acooling system that can remove It is importanttonotethatthe formulationand
the heat of condensation consistently and continually at implementation of Merkel’s theory in cooling-tower
thoselow temperatures is essential. Therefore, the cooling design and rating are presented and discussed in most
system should be considered to be an integral part of the textbooksonunit operations and process heat transfer.
powergeneration process that can have amajor in uence Asummary of some of the methods that attempt to
on overall powerplantperformanceand availability. evaluate wetcooling towers’performance hasbeen
The choice of an appropriate cooling tower for aspecial publishedbyKloppers and Kro¨ ger. [8 ] They compared
application dependsonmany factors, such as capacity, cooling-tower performanceobtainedbyMerkel,Poppe,

availabilit y, reliability, cost, andeffectiveness.The and e-NTU methods.Merkel applied the mass and energy
effectivenessofacooling tower is defined as the ratio of conservation laws to adifferential control volume that
the actual energy that is exchanged to the maximum includesthe air and the water in acounter- ow cooling-
energy that couldpossibly be exchanged. The number of tower arrangement, and derived the following differential
transfer units (NTU) is another parameter that measures relationships:
the heat transfer size of the cooling tower. The higherthe
NTUvalue, thecloserthe coolingtower is to its d ha hD afi A
thermodynamiclimit.Toestimatecooling-tower effec- Zðh as ; w
Kh a
Þ
dzZ m_a
tivenessand NTU,different analyses are used (similar to
the ones used for analysing heat exchangers) that depend d Tw ma 1 d ha
on the particular type under consideration. dzZ_ m_w cpw dz
The thermodynamicperformance of any wet cooling
tower is afunction of the geometry and the ratio of the
water ow rate ( L )tothe gas ow rate(G )—i.e., L / G .This where h ,enthalpy; h D ,masstransfer coefficient; a ,surface
valueisquantifiedbymeans of aparameter knownas area per unit volume; A ,frontal area; m ,mass ow rate; z ,
the tower characteristic or number of diffusion units h vertical direction; c p ,specificheat at constant pressure;

252 Cooling Towers
subscript: a, air;w,water;s,saturated; fi, fill;(w), where NTU is given by:

evaluated at water temperature T w.
Afterthe aboveequations were combined and 1 1Ke C
integrated over the whole length of the tower, the Merkel NTU Z
1 K C ln 1Ke:
equationwas derived:
Tð
C is the uid capacity rateratio and is defined as
a ALfi
D fi
a L
D f i fi
wi
cpw d Tw Cm in / Cm ax .Itistobenoted that the e-NTU method is
MeM Z h Zh Z applicable to the cross- ow arrangements, provided that
m_a Gw ð has; w K h a
Þ
T
wo
the air and water streams should be defined, whetherthey
where Me,Merkel number; L ,length; G ,mass velocity; are mixed, unmixed,oracombination.
subscript: M, according to Merkel approach; i, inlet; o, Khan and Zubair [1 1 ] presented an analysis to estimate
outlet. the effectivenessand NTU of acounter- ow wet cooling
The term on the right side is ameasureofthe cooling tower that matchedthe experimental data closely. They
requirement whereasthe term on the left side is ameasure includedintheir modelthe effect of the Lewis number,
of the performanceofthe packing. defined in asimilar fashion to Poppe’s as the ratio of the
Poppe includedthe effect of Lewis factor Le (definedas convective heat transfer coefficient to the convective mass
hc /hD cp a )and thereduction in water ow dueto transfer coefficient times the specific heat at constant
evaporation. He derived two equations for the Merkel pressure of moist air, the heat resistanceinthe air-water
number based on the state of the air at exit—unsaturated or interface,and the effect of water evaporation on the air
supersaturated. If the air is exiting as unsaturated, the states alongthe vertical length of the tower. In their
Merkel number can be obtained by an iterative procedure analysis, they assumed constant convective heat and mass
of integrating the following equation: transfer coefficient, and ignored the heat lost through the
tower walls, variation in specific heat properties,and water
d MeP h
lost by drift. Theyapplied themass andenergy
Zc = has ; w K h C ð Le K 1 Þfh Kh
d Tw pw a as ; w a conservation equations to adifferential volume to relate
i the change in enthalpy of moist air to its humidity ratio, in
K ð ws ; w K w Þ hv g K ð w s;w
K w Þ cpw Tw terms of Lewis number Le and otherproperties of moist
and saturated air. The outlet properties are obtainedby
In the above equation, w ,humidity ratio; the subscript numerically integrating the set of differential equations of
P, Poppe approach; v, vapor; a, air; s, saturated, and conservation of massand energy on an increment volume
w, water. of the cooling tower.
The Merkel numb er for air exits as supersaturated can They also gave the following definition of NTU and
be obtainedbyaniterative procedure of integrating the effectiveness:
following equation :
ðo
W
hD AV V dw
d MeP h NTU Z Z
Zc =h Kh C ð Le K 1 Þfh Kh Kðw m_a ws ; w K w
d Tw pw as ; w ss as ; w ss s;w
Wi
i
K ws ; a Þ vhg K ð w K ws ; aÞ pcw Tw C ð w K w s ; w Þ pcw Tw ho K h i
3Z :
hs ; w ; i K h i
The subscript ssZ supersaturated. Details can be found
AV ,surface area of water droplet per unit volume of
in Poppeand Ro¨ gener [ 9] and in Bourillot. [1 0 ]
the tower; V ,tower volume; subscript: o, outlet, and i,
According to the e-NTU method, in which the same
inlet.
simplification of Merkel is used, the Merkel number for
They also gave an empiricalequationfor NTU :
the case where dry air mass ow rate m a O m w cpw /(dh as w /
em
d Tw )can be obtained by: mw nC1

NTU em Z c _
cp w m_a
Mee Z NTU
d has; w = d wT c and n ,empirical constants specific to aparticulartower
design, and subscript em, empirical.
The subscript e Z the e-NTU approach.
Another approach is to return to the fundamental
The Merkel number for the case where dry air mass
equations of uid mechanics and heat and masstransfer,
ow rate m a ! m w cp w /(dh as w /dTw )can be obtainedby:
andarrive at numerical solutions with theaid of
computational uid dynamics technique(CFD).Some
m
_a NTU [1 2 ]
Mee Z ; examples are the work by Al-Waked et al. and Hasan
m_w et al. [1 3 ]
among many others. These solutions can in

Cooling Towers 253
prin ciple be used as the sole basis of design or they can be REFERENCE
used to examine, modify, and improveexisting simpler S
methods—such as work by Kloppersand Kro ¨ ger, [1 4 ] who 1. Sirena, J.A. The use of athermo- uid dynamic efficiency in
used the finite difference method to compare the three cooling towers. Heat Trans. Eng. 2002, 23 (2), 22–30.
approaches of Merkel, Poppe, and e-NTU. 2. Al-Nimr, M.A. Modelling the dynamic thermal behaviour of
cooling towers containing packing materials. Heat Trans.
Eng. 1999, 20 (1), 91–96.
3. Al-Waked, R.; Behnia, M. The effect of windbreak walls on
the thermal performance of natural draft dry cooling towers.
CONCLUSIO Heat Trans. Eng. 2005, 26 (8), 50–62.
N 4. du Preez, A.F.; Kro¨ ger, D.G. Effect of the heat
exchanger
arrangement and wind-break walls on the performance of
The different types of cooling towers—wet,dry, and
natural draft dry-cooling towers subjected to cross-winds.
hybrid—have been presented. Research and experience
J. Wind Eng. Ind. Aero. 1995, 58 (3), 293–303.
show that the hybrid cooling towersconform well to the 5. Streng, A. Combined wet/dry cooling towers of cell-type
stringent environmental protectionrequirements and to the construction. J. Energ. Eng. 1998, 124 (3), 104–121.
stan dard operation reliability set for cooling systems; it is 6. Walker, W.H.; Lewis, W.K.; McAdams, W.H.,;
expected, therefore, that they will become morewide- Gilliland,
E.R. Principles of Chemical Engineering;McGraw-Hill,
spread. Diff erentmethods to estimate cooling-tower Inc.: New York, NY,
performance are presented, based on someassumptions 1923. F. Verdunstungshuhlung. Zeitschrift des Vereines
7. Merkel,
that simplify the problem.Asthe systemsget more Deutscher Ingenieure (V.D.I) 1925, 70,123–128.
complicated, however, CFD is capable of pred icting 8. Kloppers, J.C.; Kro¨ ger, D.G. Cooling tower
performance
evaluation: Merkel, Poppe, and e-NTU methods of analysis.
performance and can be used as the sole basis of design,
J. Eng. Gas Turbines Power 2005, 127 (1), 1–7.
or it can be used to modify and improveexisting simpler
9. Poppe, M.; Ro¨ gener, H. Berechnung von Ru¨ ckku¨ hlwerken,
methods to makethem closer to reality.
VDI-Wa¨ rmeatlas;6.Au age, VDI Verlag GmbH
:Berlin,
Germany, 1991.
10. Bourillot, C. TEFERI: Numerical Model for Calculating the
Performance of an Evaporative Cooling Tower. EPRI Report
CS-3212-SR, Electric Power Research Institute: Palo Alto,
CA, 1983.
Glossary 11. Khan, J-U-R.; Zubair, S.M. An improved design and rating
analyses of counter ow wet cooling towers. J. Heat Trans.
Approach temperature: The difference between the temperature
2001, 123 (4), 770–778.
of the condenser water leaving the tower and the wet-bulb
12. Al-Waked, R.; Behnia, M. The performance of natural draft
temperate of the air entering the tower in the case of the wet
dry cooling towers under crosswind: CFD study. Int.
tower and the dry-bulb temperature of the air entering the
J. Energ. Res. 2004, 28 (2), 147–161.
tower in the case of the dry tower.
13. Hasan, A.; Guohui, G. Simplification of analytical models
Blow-down: The quantity of the circulating water that is removed
and incorporation with CFD for the performance prediction
to maintainthe amount of dissolved solids andother
of closed wet cooling towers. Int. J. Energ. Res. 2002, 26,
impurities at an acceptable level.
1161–1174.
Blow-out: Water droplets blown out of the cooling tower by
14. Kloppers, J.C.; Kro¨ ger, D.G. Acritical investigation into
wind—generally, at the air inlet openings. In the absence of
the heat and mass transfer analysis of counter ow wet-
wind, water may also be lost through splashing or misting.
cooling towers. Int. J. Heat Mass Trans. 2005, 48 (3–4),
Drift: Water droplets that are carried out of the cooling tower
765–777.
with the exhaust air.
Drift eliminator: Equipment containing acomplex system of
baf es designed to remove water droplets from cooling-tower
air passing through it.
Noise: The sound generated by the impact of falling water; the BIBLIOGRAPH
movement of air by fans; the fan blades moving in the Y
structure; and the motors, gearboxes, and drive belts. Some useful sites and references that wereused to collect some of
Plume: The stream of saturated exhaust air leaving the cooling the information in this article are:
tower. The plume is visible when the water vapor it contains 1. Hewitt, G.F., Shires, G.L., Polezhaev, Y.V., Eds.,
condenses in contact with cooler ambient air. Inter-
national Encyclopaedia of Heat and Mass Transfer,CRC
Range: The difference between the cooling water temperature Press: New York, NY,
entering the tower and the cooling water temperature leaving 1997.K.W. Handbook of Air Conditioning and Refriger-
2. Shan,
the tower. ation,2nd Ed.; McGraw-Hill, Inc.: New York, NY,
More glossary words are available at the Cooling Technology 2001. Technology Institute. Available at:
3. Cooling www.cti.org .
Institute(CTI) Website( www.cti.org/whatis/ 4. Gulf Coast Chemical Commercial, Inc. Available at: www
coolingtowerdetail.shtml ). gc3.com/techdb/manual/coolfs.htm .

254 Cooling Towers
5. U.S. Environmental Protection Agency. Available at: 8. Lenntech. Cooling Towers. Available at: www.lenntech.
www.epa.gov/waterscience/316b/technical/ch4 . com/cooling%20towers.htm
pdf#searchZ ‘wetdry%20cooling%20towers 9. Wikipedia. Available at: www.en.wikipedia.org/wiki/
6. Online Chemical Engineering Information. Available at: Cooling_tower
www.cheresources.com/ctowerszz.shtml 10. Legionella Control. Available at: www.legionellacontrol.
7. GEA Cooling Tower Technologies. Available at: www com/legionella-glossary.htm
bgrcorp.com/default-gct.htm .

Data Collection: Preparing
EnergyManagers
and *
Technicians
Athula Kulatunga
Department of Electrical and Computer Engineering Te chnology,Purdue University,West
Lafayette, Indiana, U.S.A.
Abstract
Energy audits can be used to provide hands-on activities related to an energy management course. After
learning the necessary background concepts, students need to be aware of what measurements must be
taken to evaluate an existing energy system. In industry and universities, one may find apprentices and
students from different educational backgrounds, such as electricians with no exposure to newer measuring
instruments and students with no ability to take electrical and/or mechanical measurements. By studying
the capabilities and limitations of measuring instruments, newcomers to the energy auditing may collect
reliable data. This article introduces several hands-on activities that could be replicated to teach students
how to take accurate measurements of electrical, light, and heat ow parameters, ultrasonic leak detection,
electronic combustion analysis, and simple data acquisition before conducting energy audits. Asample
laboratory activity includes adescription of the measuring instrument, factors that contribute to inaccurate
readings, safety concerns, and several practice measurements useful to energy audits.
INTRODUCTION data. With inaccurate data,the conclusion of an energy

improvement project holds no validity. [2]
It has been accepted that engineering technology courses Preliminary or walk-through energy audits are the most
shouldhave some hands-on activities such as labs,projects, suitable for beginners. Apreliminary energy audit is a
and other practical experiences. In the field of energy process during which an auditor examines an existing
management, energy audits have been used effectivelyto energy consuming system according to apredetermined
provide hands-on experiences. [ 1]
An energyaudit, also set of procedures. The procedures are outlined as aresult
knownasenergysurvey, energy analysis,orenergy of ahistorical data analysis of the targeted system and
evaluation, is aprocess that examines the current energy conversations the auditor had with the owner or the
consumption of aprocess or facility and proposes alter- operator of the system. These proceduresinclude taking
native waystocut down energy consumption or costs. One electrical andother measurements undercertain
aspectofthe energy auditing process is to collect specific conditions.
data of aprocess or afacility. Measuring temperature, ow
rates (heat, liquid,and air), intensity of light, electrical
current, voltage, power, power factor (PF), humidity, ELECTRICAL
MEASUREMENTS
pressure, or vibration may be requiredtodetermine the
Data collection of any system that consumes electrical
energy consumption and waste.New measuring equipment
energy requires at leastthree basicmeasu rements—
is pouringintothe measurement worldmakingdata
voltage,current,and PF—for energy analysis calculations.
collection easier, more accurate, and safer.
Utility meters collect all these data at abuilding service
Accurate data collection is paramount not only to
entrancepoint or at any other sub-metering location, if
analyze energy consumption, but also to evaluatethe
such meters are installed. By contacting theutility
effectivenessofproposed changessuggested in an energy
provider,one can easily obtain the historical data related
auditreport. Some energy-savingelectrical retrofits may
to the above parametersand more for agivenfacility.
introduce electrical power quality problems that may not
However, whenitcomes to individual systems within a
be accounted for by traditional meters, causingerroneous
facility, these data may notbeavailable, unless additional
utilitymeters are installedateach of the service entrance
points to individualsystems.Inamodern power
*
This entry originally appeared as “Preparing Energy Managers and distributionsystem, harmonicsofthe fundamental
Techniciansfor Energy Data Collection” in Energy Engineering,Vol. 102,
frequency, 50 or 60 Hz, appeardue to non-linear devices
No. 4, 2005. Reprinted with permission from AEE/Fairmont Press.
connected into the system. Power line harmonics cause
Keywords:Energy audits; Energy data; Energy measurements. erroneous readings,ifameter is not capable of measuring
255

256 Data Collection: PreparingEnergy Managers and Technicians
and current probesthat come with the meter.The metersare

capable of measuring true RMS, peak, and total harmonic
distortio n(THD) for voltage and current. They also display
true power (Watts), reactive power volt ampere reactance
(VAR),PF, displacement power factor (DPF), crest factor,
K -factor, and harmonics individually up to the 31st.
The meters are capable of displaying data in three
views—waveforms, barographs showing harmoniclevels,
and numeric values for voltage,current,and power—as
depicted in Fig. 2.
Each of the data displayed on Fig. 2givessomeclue
about the nature of power quality. For example, at-topped
voltage waveform is an indication of the presence of
currentharmonics. Onecan usethe THDlevelsto
determineifthey are within the specified limits. If not,
thesamedatacan be used to determinesizingof
transformers and harmonicfilters. The bar-graph display
of the current reveals the percentage comparison of odd-
numberedharmonics with respect to the fundamental
frequency. Ev en-numbered harmonics cancelout in a
power system;therefore their effects are not aconcern in a
power system.Finally, th emeter displaysnumerical
values such as total PF, DPF,kilovolt ampere, and
kilowatt useful in determining power-factor correction
methods.The data can be further analyzed by down-
loading FlukeView w software, which comes with the
meter. Fig. 3 depicts current waveform of anonlinearload
and downloaded data in atabular form.
Fluke 43B powerquality analyzer has all the features of
a41B as well as sample and storage capabilities. These
additional features allow auser to detect voltage/current
sags and transientsinapowersystem.
Fig. 1 (A) Fluke 41B and 43B power analyzer, (B) Using meter
probes for three-phase. (Courtesy of Fluke Corporation).
Practice Activity Outline
Atypical electrical power system may have one of the

true root-mean-square (RMS) values of the fundamental
following problems: (1) voltage sags, (2) current balance
frequencyand the harmonics at agiveninstant. [3 ]
and loading, (3) harmonics, (4) grounding, or (5) loose
Two power quality analyzersmade by Fluke Corpo-
connections. Taking measurements at the electrical serv ice
ration are showninFig. 1A. One of these meters can be
panelduringanenergyaudit, theinvestigator can
easily used to obtain all necessary data for an electrical
determinethe sources of these problems. In alaboratory
system.Fig. 1B showshow to obtain measurements for a
environment (or in aworkshop), each of the above
three-phase balance (5% or less imbalance) using voltage problemscan be replicated to demonstrate the effects.
Fig. 2 Waveform, bar graph, and numerical displays of data presented by Fluke 41B. (Courtesy of Fluke Corporation).

Data Collection: Preparing Energy Managers and Technicians 257
Fig. 3 Data table, current waveform,

and harmonic representation of anon-
linear load taken by Fluke View w
software.
For example, the circuit showninFig. 4may help The concept of upstreamdisturbance can be demon-
students learn howtodetect harmonics and locate the strated by switching the DPDT switch to position Band
harmonicproducing sources in apower system.The three- starting the drive. Asample display taken under this
phase power supplyshould be takenfrom aY-connected condition is showninFig. 5B. Notethat the voltage sag
transformer secondary where the common point is taken as occurssimultaneously with the current sag, indicating that
neutral. It is also necessary to ground the common point, thesource of thedisturbance is upstream of the
if concepts related to overload ing are introduced. Adouble- measurement point. [4 ]
pole double-throw(DPDT)switch allows switching
betweenthe lamp-only circuit and the lamp with adimmer Precautions
circuit. The lamp-only path wouldallow the student to
studythe waveform characteristics and harmoniccontent of First of all, studentsmust have athorough understandingof
alinear load. The lamp with adimmer can be used to the instrument’s limitations. They shouldmake voltage
observe the voltage and current waveformvariation when measurements after acircuitbreaker and wearsafety gloves
harmonics are present to the right of the point, where the and eyewear whentaking measurements of live panels.
Fluke 43B meter is connected. Using the sag and swell
mode of the Fluke 43B meter, asample can be taken when
DPDTisinposition Aand the Variable Speed Drive (VSD) LIGHT
is off. The waveform wouldlook similar to the one shown in MEASUREMENTS
Fig. 5A . Note that the voltage sag coincides with acurrent On average, lighting consumes 35% of energy used in
swell indicating that the disturbance has occurred down- commercial buildings and 25% in industrial facilities.
stream of the measurement point. Lighting levels directlyaffectthe productivityof
Fig. 4 Acircuit for learning the sources of

harmonics in an electrical power system.

Fig. 5 (A) Downstream disturbance,

(B) Upstream disturbance.
employees. However, manylighting systems are impro- foot-can dles (fc). The purposeofmeasuring light is to
perly designed and unattended over the years. Lighting is determinethe incidentlight on ahorizontal or vertical
one of the areaswhere companies can save energy with the surface.Illumination Engineering Society of America
least amount of capitalinvestments. Fig. 6illustrates the (IESA) publishesrecommendedlight levels for specific
before and after appearance of awarehouse, where 50% tasks.
more light was obtained.
Light meters are very easy to use. The meter in Fig. 7 Practice Activity Outline
measures light intensity in Lumenand light density in
Select alocation where students can move aroundwith a
ladder for at least 2h.First, students draw an accurate
layout of the space beingaudited including the location of
furniture,windows, cabinets, and lighting fixtures. The
location couldbeaclassroom, alab, acorridor, or agym.
Using alight meter, studentsrecord the available light
intensityinfconhorizontal and vertical surfaces with and
withoutartificial lights.Thiswould yield alight
distribution map.Now compare the available light with
the recommended lighttodetermine levels and locate
poorly illuminated or over-illuminated areas. Most class-
room and labs are illuminated by uorescent light fixtures.
The amount of light generated by uorescentlight
deteriorates with time.Students may replace existing
uorescent tubes with new ones to examinethe available
light.
Precautions
Students need to be aware that only the inciden tlight on

horizontal surfaces can be added. Light intensitymustbe
measured at the recommendedheight as specified by IESA
or any otherstandard. It is important to stayawayfrom the
detector to minimize the effect of body-re ected light
entering into the detector.
BUILDING ENVELOP
MEASUREMENTS
The heat loss or gains of abuilding depend on many
Fig. 6 Fifty percent more light obtained through lighting factors: R -value of walls and roof material, window
retrofits in awarehouse. characteristics, ambienttemperature, humidity level, and
Source: From Business News Publishing Company (see Ref. 5 ). heating and cooling degree days of agiven location where

THERMAL
IMAGING
During any energy conversion process, heat is produced as
aby-product.Poor insulation of buildings and pipes
Fig. 7 Astick-type light meter with aretractable sensor. carrying hot or cold liquids causesenergy wastes. New
(Courtesy of Omega Inc.). retrofitsand devices containing switching power electronic
circuits may introduce harmonics, which may overheat the
neutral conductors. Poor connections of an electrical
the building is allocated. When R -value and U -values are
distribution system are the major contributortosystem
known, adetailed analysis of abuilding envelop would
inefficiencies and may lead to acatastrophic fire. All of the
yield Britishthermal unit lost or gain ed through walls,
above couldbeavoided, if one could detect them in
windows, and otherheating and cooling sources. It is very advance. To detect such abnormalities, one requires a
difficult to determin ethe R -values of even aseveral-year thermal imager.Athermal imager, compared to an IR
old facility due to poor record keeping and later add-ons to thermometer, is capable of measuring temperature vari-
the structures. The OMEGA w OS-650energy conserva-
ation betweentwo adjacent points. In the past, the cost of
tion and plant maintenancekit (see Fig. 8) is very useful in thermal imagersprevented widespread use in energy
energy audits and general plantmaintenance. The kit management and plant-maintenance activities. Industries
consists of an infrared (IR)thermometercapableof hired aconsultant to surveyall electrical distribution
measuring temperatures from K 2to2008 Fand aheat panels and othercriticallocations in afacility to determine
ow meter, which is aspecially designed IR radiometer the thermal profile annually. But prices have comedown
capable of displaying heat ow though ascanned wall in significantly, under $10,000, over the past few years. A
terms of British thermal unit per squarefeet hour. [6 ] The
specially designed, hand-held thermal imager made for
kit is priced around $1600. If purchased separately, either energy audits and plantmaintenance is shown in Fig. 9 .
the thermometer or the heat ow meter would cost about The imager can hold up to 100images,which are
$800. stampedwith time and date as they are taken. Through an
Universal Serial Bus (USB) port,the images can be
downloaded to the accompanying software installedona
PC for further analysis. Images can be further analyzed by
assigning asingle color to atemperature range and
Alaboratory activitycan be developed, so studentswill be
creating athermal profile. Athermal profile represents the
able to (1) estimate R -value value of an unknowninsulator,
temperature at the x and y axes as the cursor moves aroun d
(2) measureheat owthrough walls and windows, and
the image. [7 ]
(3) makeenergy cost analyzes at the end of the activity. To
estimate R -value, net heat owshouldbedetermined by
usingthe heat owmeter. This is done by taking two
measurements across awall at the same height—one from Many interesting activities can be developed aroundthis
inside and the other from outside of the room.The meter device. Students couldmeasure thetemperature of
showsaC orK number, indicating the direction of heat different light sources and compare the power consump-
ow. tion of each light bulb and the surface temperature. The
internal heat distribution of aroom couldbedetected and
documented by hanging black-painted aluminum sheets.
Since this device measures the re ectedthermal energy,
the manufacturer recommendspainting highly re ective
surfaceswith somedark colortominimize reading errors.
Students coulddevelop thermal images of buildings on
campus and analyze the heat losses. In apowerlab,
students could measure the temperature of conductors
when motors aredriven with variable speeddrives
( Fig. 10 ).
Precautions
When taking an image, the focus is paramount. As with

any digitalcamera, lackoffocus blurs the image, thus
Fig. 8 OMEGAOS-650 EnergyConservationand Plant minimizing thedevice’s ability to distinguish the
Maintenance Kit. (Courtesy of Omega Corporation). temperature difference betweenadjacent pixels. Some

Fig. 9 Raytek ThermoViewe Ti30 portable imager and its display. (Courtesy of Raytek Corporation).
imagers have aminimum focusing distancethat must be applications, because the humaneye can detect variations
adhered to. Most current thermal imagersallow users to of grayscale better when thermal changesare subtle. In
view the object using different colorpalettes such as addition, students mustbeaware of concepts such as
rainbow,ironbow, and grayscale. Despite the popularity of qualitative vs quantitative temperaturemeasurements,
color palettes, it is recommended to use grayscale for most distance to targetratio, field of view,effects of
environmental conditions (steam, dust, smoke, etc.), and
effects of emissivity.
COMBUSTION
ANALYSIS
The effectivenessoffossil-fuel burning receiveslittle
attention during an energy audit process unless the auditor
is well trained to perform mechanical tests. The
mechanical test, atedious process, requires aset of data
from several different pieces of equipment such as draft
gauges, thermometers, carbon monoxide (CO) stain length
tubes, wet chemical absorption instruments, etc. Then the
data are entered into aslide ruler calculator to determine
the combustion efficiency. [8 ]
New electronic combustion analyzersprovide more
reliable and provable data than atraditional “eyeballing
the ame” analysis. It also produces faster professional
analysis than amechanical test. Fyrite w Tec 50 and 60
residential combustion analyzers by Bacharach Inc. shown

below ( Fig. 11 )falls under $500 and comes with many
features that make this instrument idealfor techniciansand
studentsalike.
Auser first selectsthe type of fuel being burnt. Thefuel
choices are natural gas, #2 oil, propane, and kerosene. The
meter can measure uegas oxygen content 0.0%–20.95%
O , ue gas temperature up to 9998 F, ambienttemperature
2
328 F–1048 F, and ue gas CO content 0–2000 ppm CO.
Based on thesemeasurements,the meter is capable of
calculating the following: combustion efficiency 0.1%–
99.9%, ue gas carbon dioxide content 0.0 to afuel
Fig. 10 Multi-imagedisplay andthermal profile using dependent maximum (in percent), uegas CO air-free
ThermoViewe Ti30 imager. (Courtesy of Raytek Corporation). content 0–9999ppm, and excess air 0%–400%.

Fig. 11 Fyrite w
Tec 50 and 60 residential combustion analyzers and amanufacturer’s comparison of combustion analysis techniques.
(Courtesy of Bacharach Inc.).
Practice Activity Outline ULTRASONIC

LEAKDETECTION
Practice activities for thismeter may include: (1) testing Aleak,whether it is acompressed air,steam, or
amesproduced by the four different type ames while conditioned air, is awaste. Hidden costs due to leaks can
varying the air intake into the ame,(2) experimenting be significant in very competitive market environments.
heat transfer characteristics of ajar containing water (or Some leaks are very easy to detect, while others are not.
any otherliquid applicable to acertainindustrial process) Most leaks go undetected due to ambientnoises. To detect
under different insulations and ambienttemperatures, and sometypes of leaks, one has to use ultraviolet (UV) dyes,
(3) simulating aspecific type of burnerused in an special lamps, bubblesolutions, etc.Leak detection takes
industrial facility to investigate efficiency improvement time and money. Any instrumentthat detectsdifferent
opportunities. Flames for activity-1 and activity-2 above typesofleaks within avery short time would be the most
can be set up by usingoff-the-shelf burners or commercial practical solution.
water heaters. Ultrasonic leak detectors minimize traditional problems
associated with leakdetection. Fig. 12 depicts aleak
Precautions detector made by Superior SignalCompany Inc., which
comes with aprice of less than $250.
When workingwith ammable gases in closed environ- According to the manufacturer’s specifications, the
ments, one must adhere to OccupationalSafety and Health meter can detect leaks of any type of refrigerant gas,
Administration (OSHA) safety guidelines. Students must vacuum, and pressure leaks. The reading is not affected by
be aware of every detail of the gas being used and wear saturated gas areasorwindy environments. An energy
appropriate protectionapparatus, which include burn- auditor may even detect leaks aroundfreezer and cooler
proof glovesand eye protection. doors with this meter. [ 9]

Fig. 12 AccuTrak VPE by Superior Signal Company, Inc.

(Courtesy of Superior Signal Company, Inc.).
Several simple activities can be developed to familiarize

this equipment in alabo ratory or workshop environment
where compressed air is available. Several holes, at least
one foot apart with different diameters, can be drilled in a
copper piping tube. Aconnector needs to be soldered to
one end and the other end mustbecapped. Once connected
to aregulated compressed air outlet, learners may trace air
leaks of each hole under different pressure to learn the
nature of the leaks and the corresponding meter responses.
The activitycan be further expanded by moving away
from the leaks and/or walking aroundthe pipe. Similarly,
to studythe nature of leaks due to doors and windows,
learners may take measurements aroundcommercial soft Fig. 13 DATAQ 4-channel data acquisition and chart recorder
drink coolers located in any facility. starter kit model. (Courtesy of DATAQ Instruments,
Inc.).
Precautions
system and the computers used to record the data.For
Peoplehave anatural tendency to clear the area around a laboratory applications and short-term data logging, one
suspected leak, especially pain ted or corroded surfaces, to may use the D1-194RS low-cost (less than $25) data
examinethe leak more closely. This may cause the hole to acquisition starter kit shown in Fig. 13.The kit includes
burst witho ut much warning, spewingdust and loose hardware, recorder software, and aserial port cable.
particlesall over. Therefore, studentsmust wear safety The hardwareisself-powered through the serial port of
glasses and dusk masks as appro priate whentesting leaks PC. Analog signals connectedtoany of the four channels
of any form. are digitized and saved into the hard drive while showing
on the screen as astrip-chart recorder. Each channel can be
sampledupto240 samples persecond with 12-bit
LOW COSTDATA resolution, which is adequate for most analog signals in
ACQUISITION the energy field. Each channelaccepts G 10 V, which is
Energy managers have to rely on historical data when large enough for most commercially available transducer
energy analyzes are performed. Atrend of ameasured outputs. The software includes Active-X control libraries
parameter such as current, voltage,temperature, etc. that allow the user to program the kit from any Windows
presentsabetter pictureofenergy-savingopportunities environment. Therecoded data may be play backedfor
than instantaneous readings.Mostdataloggers are later analysis. [1 0 ] The input signals are not optically
expensive, because they are designed to minimize potential isolated and measured with respect to acommon ground
damages to the front-end electronics of the data acquisition point.

Encyclopedia of Energy Engineering and Technology - 3 Volume Set

Diunggah oleh

Informasi Dokumen

Deskripsi Asli:

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

Encyclopedia of Energy Engineering and Technology - 3 Volume Set

Diunggah oleh

Hak Cipta:

Format Tersedia

Compressed Air EnergyStorage *

INTRODUCTION Thermal and Compressed Air Storage (TACAS) tech-

© 2007 by Ta yl or & F r anc i s G r oup, LL C

10.1081/E-EEE-120043197—4/5/2007—19:02—RATHNAS—225284—XML Taylor &Francis Encyclopedias

expensive permanent magnet material and is acompro-

which routes air through or aroundthe TES. Referring to

© 2007 by T ayl or & Fr a nci s Gr oup, L L C

10.1081/E-EEE-120043197—4/5/2007—19:02—RATHNAS—225284—XML Taylor &Francis Encyclopedias

Fig. 3 Method of controlling constant turbine inlet temperature and pressure.

Fig. 4 Method of controlling con-

© 2007 by Ta yl or & F r anc i s G r oup, LL C

10.1081/E-EEE-120043197—4/5/2007—19:02—RATHNAS—225284—XML Taylor &Francis Encyclopedias

Fig. 5 High energy thermal energy

COMPRESSEDAIR STORAGE IN at fee. Therefore, someusers in the UnitedStates must

© 2007 by T ayl or & Fr a nci s Gr oup, L L C

10.1081/E-EEE-120043197—4/5/2007—19:02—RATHNAS—225284—XML Taylor &Francis Encyclopedias

© 2007 by Ta yl or & F r anc i s G r oup, LL C

10.1081/E-EEE-120043197—4/5/2007—19:02—RATHNAS—225284—XML Taylor &Francis Encyclopedias

INTRODUCTION desired task. In reality, however,many of the openings are

© 2007 by T ayl or & Fr a nci s Gr oup, L L C

10.1081/E-EEE-120042156—4/5/2007—19:05—RATHNAS—221744—XML Taylor &Francis Encyclopedias

Table 1 Discharge of air through an orifice

© 2007 by T a ylor & Fr a nc is Gr oup, L L C

© 2007 by T ayl or & Fr a nci s Gr oup, L L C

10.1081/E-EEE-120042156—4/5/2007—19:05—RATHNAS—221744—XML Taylor &Francis Encyclopedias

Fig. 1 Noise vs leak loss at various pressures.

LEAK The tag can be used for after-control by providing aplace

© 2007 by Ta yl or & F r anc i s G r oup, LL C

10.1081/E-EEE-120042156—4/5/2007—19:05—RATHNAS—221744—XML Taylor &Francis Encyclopedias

© 2007 by T ayl or & Fr a nci s Gr oup, L L C

10.1081/E-EEE-120042156—4/5/2007—19:05—RATHNAS—221744—XML Taylor &Francis Encyclopedias

Leakage ð % Þ Z ½ðT ! 100Þ = ð T C t Þ point. Electricallyactivated designs require apower

Installing a ow meter to measureand record the actual example, is illustrated in Fig. 2.

© 2007 by Ta yl or & F r anc i s G r oup, LL C

10.1081/E-EEE-120042156—4/5/2007—19:05—RATHNAS—221744—XML Taylor &Francis Encyclopedias

© 2007 by T ayl or & Fr a nci s Gr oup, L L C

10.1081/E-EEE-120042156—4/5/2007—19:05—RATHNAS—221744—XML Taylor &Francis Encyclopedias

INTRODUCTION than the average air demand. For reliable operation of

Compressed Air Generation Efficiency

© 2007 by Ta yl or & F r anc i s G r oup, LL C

10.1081/E-EEE-120042157—4/5/2007—20:34—VELU—223878—XML Taylor &Francis Encyclopedias

© 2007 by T ayl or & Fr a nci s Gr oup, L L C

10.1081/E-EEE-120042157—4/5/2007—20:34—VELU—223878—XML Taylor &Francis Encyclopedias

COMPRESSED AIRENERGY Assumethat the compressor intakes 7.9 ft 3 of dry

STORAGE atmospheric air at 14.5 psia.Multiplying by r Z 7.9 results

Application of the Combined Gas Law D Pr ec

V1 1 Eq. 8Gas volume released from areceiver.

© 2007 by Ta yl or & F r anc i s G r oup, LL C

10.1081/E-EEE-120042157—4/5/2007—20:34—VELU—223878—XML Taylor &Francis Encyclopedias

Eq. 9Pneumatic capacitance.

pressure. Assume that the atmosphere is at standard

Table 1 Definition of variables and units of measure

Cp n Pneumatic capacitance (scf/atm) or (scf/psia)

© 2007 by T ayl or & Fr a nci s Gr oup, L L C

10.1081/E-EEE-120042157—4/5/2007—20:34—VELU—223878—XML Taylor &Francis Encyclopedias

© 2007 by Ta yl or & F r anc i s G r oup, LL C

10.1081/E-EEE-120042157—4/5/2007—20:35—VELU—223878—XML Taylor &Francis Encyclopedias

Fig. 4 Compressed air system pressure profile with uncontrolled storage.

resulting power increaseis14%, and waste to artificial sys r ec

© 2007 by T ayl or & Fr a nci s Gr oup, L L C

10.1081/E-EEE-120042157—4/5/2007—20:35—VELU—223878—XML Taylor &Francis Encyclopedias

© 2007 by Ta yl or & F r anc i s G r oup, LL C

10.1081/E-EEE-120042157—4/5/2007—20:35—VELU—223878—XML Taylor &Francis Encyclopedias