PERFORMANCE | RELI ABI LI TY | EFFI CI ENCY | ASSET MANAGEMENT

&
Compressors
compressed air systems
2005 - 2006 collection
JANUARY 2005 WWW.PLANTSERVICES.COM 42
RELIABILITY
Compressors
K
NOWN AS THE FOURTH UTILITY, compressed air is used in many
facets of manufacturing. Many plants use it in one capacity or another and,
for the most part, these systems provide similar outputs. However, not all com-
pressed air is identical. In several situations, specially engineered air powers a variety
of machinery and tools used across several different industries, from automobile man-
ufacturing to beverage processing.
Engineered air is a term that describes
compressed air that has been tailored
to meet at least three specific industry
needs 100% oil-free, particulate-free
and completely reliable. It goes beyond
identifying only a specified pressure.
Engineered air provides the right type
of air for the application. Engineered air
reliability is a function of system redun-
dancy, accessibility and maintenance, and
technological advances in control and
monitoring.
Consider redundancy
When an air compressor is vital to plant
operations, a redundant system mini-
mizes the possibility of system failure
that can ruin production quotas. Exam-
ine your operation to see where air sys-
tem redundancies might be useful. For
example, the air compressors oil pump,
the heart of the lubrication system, keeps
the machine running smoothly. If the
pump fails, the machine comes to a
grinding halt. A redundant system would
include two full-capacity, full-pressure
pumps, one a shaft-driven main pump
and the other a motor-driven auxiliary.
During regular operations, the shaft-
driven main pump operates, while the
motor-driven auxiliary is on perpetual
standby for emergency situations.
Such a system provides full capacity
backup. Without this redundant system,
the entire compressor would need to be
shut down in the event of an oil sys-
tem malfunction.
Make repairs straightforward
Although it requires additional time and
attention from plant professionals, com-
pressor cleaning and maintenance rep-
resent a sound investment. As with other
plant assets, clean, well-maintained com-
pressed air systems are less likely to break
down. Simply put, less downtime allows
for more production. In addition, con-
sistent cleaning and maintenance min-
imize wear and tear, which saves money
in replacement parts.
Many vendors design compressors with
accessibility and maintenance in mind.
The intercooler is a good example. For
instance, using either U-shaped or straight
intercooler tubes is standard industry
practice. However, unlike the U-bend
design, straight tubes are easier to clean.
A plant engineer can simply remove the
cooling water lines, unbolt the water box
and rod the tubes in place. Rodding
isnt possible with U-bend tubes used in
some compressors. In addition, intercool-
er tubes with a water-in-tube design are
easier to clean and maintain than those
with an air-in-tube design that requires
wire brush or chemical bath cleaning.The
longer it takes to clean the intercooler,
the longer the engineered airflow is
unavailable.
Journal bearings are another good
example of important compressor parts
that benefit from diligent maintenance
AIR
Maintaining reliability requires attention to three key
points that minimize downtime
By Addison Kelley
ENGINEERED
and cleaning. Properly installed and
maintained, these bearings can last for
extended periods of time. Horizontally
split bearings are easier to maintain,
clean and replace because accessing them
requires only removing the top half of
the gear case. No other disassembly is
required. Some compressors, on the
other hand, have one-piece bearings that
require complete compressor disassem-
bly for cleaning and maintenance.
Interchangeable parts that save time
and money are another factor that
simplifies compressor maintenance. For
example, multistage compressors use a
bull gear and pinion system to power the
impellers. The quality of the bull gears
directly determines whether parts are
interchangeable. The American Gear
Manufacturers Association (AGMA)
provides established gear quality ratings
that range from 3 to 15.
If a compressor gear train features
AGMA Quality Level 12 (or less) and
either a bull gear or one of the pinions
fail, all three will need to be replaced. On
the other hand, AGMA Quality Level
13 gears, otherwise known as aircraft-
quality gearing, are generally regarded as
high-precision gears. They produce less
noise and, under normal operating con-
ditions, have a longer life. More impor-
tantl y, though, gears and pinions
manufactured to this standard are inter-
changeable.The plant maintenance tech-
nician onl y needs to swap out the
damaged piece in question, which saves
maintenance time and money.
Monitor and control
Plants that use multiple compressors to
feed a single air system need to coordi-
nate, monitor and control compressor
operation. An initial investment in mon-
itoring technology can ultimately pay
for itself.
A PLC-based automatic sequencer
can allow for as many as eight compres-
sors to communicate with one another
and operate as a team as it follows a pro-
grammed schedule.The sequencers mon-
itor and match compressor supply to
demand. For example, they can select
which compressors to use, shutting down
those not necessary to plant operations,
even choosing backup units as needed.
An automatic sequencer can ensure a sta-
ble system pressure, allowing your entire
operation to run as efficiently as possi-
ble, saving both time and money.
PLC-based modular control systems
can allow your plant operations engineers
to monitor and perform diagnostic
checks on your compressed air systems
remotely, helping to predict and prevent
systems malfunctions that could result
in engineered-air downtime. These con-
trol systems should be easy to operate,
resulting in less training time.
Engineered air is produced with
increased reliability and efficiency with
a specific plant application in mind. Rec-
ognizing that engineered air is not iden-
tical to compressed air is the first step
to minimizing downtime and achieving
better productivity. The systems that
produce engineered air must, themselves,
be engineered to operate as reliably as
possible. Three points can help your
plant engineers ensure the most reliable
flow of engineered air possible:
System redundancy, such as a dual-
pump oil system, ensures engineered
air is always available at full capacity.
Accessibility and maintenance reduces
the wear and tear on key machine parts,
keeping the system running smoothly.
Technological advances in control
and monitoring provide the infor-
mation needed to keep engineered
airflow stable. p
Addison Kelley is vice president of global
customer support at FS-Elliott. Contact him
at akelley@fs-elliott.com and (724) 600-8900.
Figure: FS-Elliott
JANUARY 2005 WWW.PLANTSERVICES.COM 43
RELIABILITY
Compressors
Figure 1. This compressor lubrication flow diagram features a redundant oil pump.
Go with the flow
Driver
Main
oil pump
PAP
compressor
Oil
mist
eliminator
Filter Cooler
Oil
reservoir
Auxiliary
oil pump
Oil
temperature
transmitter
Oil
pressure
transmitter
Panel
Check
valve 3
Check
valve 2
Check
valve 1
Relief
valve
Thermal
valve
Bleed
line
FEBRUARY 2005 WWW.PLANTSERVICES.COM 57
A common error we see in compressed air
systems, in addition to poor piping practice,
is line sizes too small for the desired air flow.
This isnt limited to the interconnecting piping from compres-
sor discharge to dryer to header. It also applies to the distrib-
ution lines conveying air to production areas and within the
equipment found there. Undersized piping restricts the flow
and reduces the discharge pressure, thereby robbing the user of
expensive compressed air power. Small piping exacerbates poor
piping practices by increasing velocity- and turbulence-induced
backpressure. (See Theres a Gremlin in your air system Its
name is turbulence, Plant Services, July 2002, p. 37).
Pipe size and layout design are the most important variables
in moving air from the compressor to the point of use. Poor
systems not only consume significant energy dollars, but also
degrade productivity and quality. How does one properly
size compressed air piping for the job at hand? You could ask
the pipefitter, but the answer probably will be, What we always
do, and often thats way off base.
Another approach is matching the discharge connection of
the upstream piece of equipment (filter, dryer, regulator or com-
pressor). Well, a 150-hp, two-stage, reciprocating, double-acting,
water-cooled compressor delivers about 750 cfm at 100 psig
through a 6-in. port. But most 150-hp rotary-screw compres-
sors, on the other hand, deliver the same volume and pres-
sure through a 2-in. or 3-in. connection. So, which one is right?
Its obvious which is cheaper, but port size isnt a good guide
to pipe size.
Charts and graphs
Many people use charts that show the so-called standard pres-
sure drop as a function of pipe size and fittings, which sizes
the line for the what is referred to as an acceptable pressure
drop. This practice, too, can be misleading because the charts
THERES NO SUCH THING AS TOO
LARGE A COMPRESSED AIR LINE
PERFORMANCE
Compressors
The secret
is in the
pipe
By Hank van Ormer, Don van Ormer and Scott van Ormer
cant accommodate velocity- and flow-induced turbulence.
Think about it. According to the charts, a short run of small-
bore pipe exhibits a low total frictional pressure drop, but the
high velocity causes an extremely large, turbulence-driven pres-
sure drop. Then theres the question of the meaning of accept-
able pressure drop. The answer to this question often isnt
supported by data, such as the plants electric power cost to
produce an additional psig.
Weve audited many plants during the past 20 years
and found the unit cost of air for positive-displace-
ment compressors runs f rom several hundred dollars
per psig per year to several thousand dollars per psig
per year. At current energy costs, you dont want the
pipe to be a source of pressure drop.
Shooting blind
Not knowing the energy cost of lost pressure as a function of
line size can lead to a blind decision. Unfortunately, this is what
we find in most of the air piping systems installed during the
past 30 years. Older systems that were designed with care are
often right on the mark, except if theyve been modified after
the original installation.
Some might call pipe sizing a lost art, but we see the
issue as a lack of attention to detail, basic piping prin-
ciples and guidelines. Read on to learn how to size air
pi pi ng usi ng vel oci t y, whi ch, when combi ned wi th
appropriate piping practice, ensures an efficient com-
pressed-air distribution system. As compressed-air sys-
tem consul tants and troubl eshooters, we use these
guidelines to design new piping systems and to analyze
exi sti ng system performance and opportuni ti es for
improvement.
Interconnects and headers
The interconnecting piping is a critical element that must deliv-
er air to the distribution headers with little pressure loss, if any.
This isnt only an energy question. It also ensures the capacity
controls will have sufficient effective storage to allow them to
react to real demand and translate less air usage to a compara-
ble reduction in input electrical energy.
The mai n di stri buti on headers not onl y move ai r
throughout the plant, they also supply the appropriate
local storage that ensures the process feeds have ade-
quate entry pressure and flow. The main header system
represents storage that supports the operating pressure
band for capacity control. You want the pressure drop
between compressor discharge and point of use to be
significantly less than the normal operating control band
(10 psig maximum).
The targets
The objective in sizing interconnecting piping is to transport
the maximum expected volumetric flow from the compressor
discharge through the dryers, filters and receivers to the main
distribution header with minimum pressure drop. Contempo-
rary designs that consider the true cost of compressed air tar-
get a total pressure drop of less than 3 psi.
Beyond this point, the objective for the main header is
to transport the maximum anticipated flow to the pro-
duction area and provide an acceptable supply volume for
drops or feeder lines. Again, modern designs consider an
acceptable header pressure drop to be 0 psi.
Fi nal l y, f or t he drops or f eeder l i nes, t he obj ec-
ti ve i s to del i ver the maxi mum anti ci pated fl ow to
t he work st at i on or process wi t h mi ni mum or no
p r e s s u r e l o s s . Ag a i n , t h e l i n e s i z e s h o u l d b e
si zed f or near-zero l oss. Of course, t he cont rol s,
regul at or s , act uat or s and ai r mot or s at t he wor k
stati on or process have requi rements f or mi ni mum
i nl et pr es s ur e t o be abl e t o per f or m t hei r f unc-
t i ons . I n many pl ant s , t he s i ze of t he l i ne f eed-
i ng a work stati on of ten i s sel ected by peopl e who
don t know t he f l ow demand and aren t aware of
how t o si ze pi pi ng.
In our opinion, new air-system piping should be sized
according to these guidelines. For a system that does-
nt meet the criteria, the cost of modification must be
weighed against the energy savings and any improve-
ments in productivity and quality.
Obviously, the lower the pressure drop in transporting air,
the lower the systems energy input. Lower header pressure
also reduces unregulated air flow (including leaks) by about
1% per psi of pressure reduction.
Eliminate the drop
Most charts show frictional pressure drop for a given flow
at constant pressure. Wall friction causes most of this loss,
which is usually denominated as pressure drop per 100 ft. of
pipe. Similar charts express the estimated pressure loss for fit-
tings in terms of additional length of pipe. When added to
the length of straight pipe, the sum is called total equivalent
length. These charts reflect the basic calculations for pressure
loss, which include:
Air density at a given pressure and temperature.
Flow rate.
Velocity at pipeline conditions.
The Reynolds number.
Other factors, including a friction factor based on the size
and type of pipe.
The calculations and chart data help to identify only the
probable minimum pressure drop. Internal roughness and scal-
ing dramatically affect the pipes resistance to flow (friction
loss). Resistance increases with time as the inner wall rusts,
scales and collects more dirt. This is particularly true of black
iron pipe.
Pressure drop is proportional to the square of the velocity.
Any high-volume, intermittent demand produces dramatic
pressure drop during peak periods. Ignoring this fact affects
every process connected to the header. For more detail, see
The compressed air receiver: The endless question, Plant
FEBRUARY 2005 WWW.PLANTSERVICES.COM 58
PERFORMANCE
Compressors

Services, May 1997, p. 49, and Appendix 1, Tables and Out-
line from DOE/CAC Air Master Training Manual. For a
given size pipe:
At constant pressure, the greater the flow, the greater the
loss per foot of pipe.
At constant flow rate, the lower the inlet pressure, the greater
the loss per foot of pipe.
At any condition, smooth-bore pipe (copper, stainless steel)
exhibits lower friction losses.
Air velocity
The most overlooked idea in piping layout and design is air
velocity. Excessive velocity can be a root cause of backpressure,
erratic control signals, turbulence and turbulence-driven pres-
sure drop.
The British Compressed Air Society suggests that a veloc-
ity of 20 fps or less prevents carrying moisture and debris past
drain legs and into controls. A velocity greater than 30 fps is
sufficient to transport any water and debris in the air stream.
Thus, the recommended design pipeline velocity for intercon-
necting piping and main headers is 20 fps or less, and never
to exceed 30 fps. Field testing reveals that, under these condi-
tions, air stream turbulence is generally negligible. Line
drops, feed lines or branch lines less than 50 ft. long work
well at a velocity of 30 fps, but velocity here should not
exceed 50 fps.
Crunching numbers
First, look at the velocity at maximum anticipated flow con-
ditions using the following equation:
V = 3.056 * Q/D
2
(Eqn 1)
Where V = air velocity (ft./sec.)
Q = volumetric flow rate (cfm)
D = conduit inside diameter (inches)
Although this method of determining the minimum pipe
size on the basis of air velocity is easy, you also must consider
that the compressed air volume is expressed in cubic feet per
minute of free air, which is the air volume at ambient atmos-
pheric conditions, not the compressed volume.
To adjust the inlet air volumetric flow rate to actual pipeline
conditions, youll need to divide the volume of free air by the
compression ratio using the following equation:
CR = (P+P
a
)/P
a
(Eqn 2)
Where P = line pressure (psig)
P
a
= average atmospheric pressure at your elevation (psi)
Tabl e 1 shows t he compressi on r at i o as a f unc-
t i on of gauge pressure f or a l ocat i on at sea l evel ,
whe r e t he a t mos phe r i c pr e s s ur e i s 14. 7 ps i . At
hi gher el evat i ons, t he aver age at mospher i c pres-
s ur e dr ops and t he compr es s i on r at i o r i s es . For
exampl e, Fl agst af f, Ar i z. , at a 7, 000-f t . el evat i on,
has an aver age at mospher i c pressure of about 11
psi . At 100 psi g, t he compressi on rat i o i s equal t o
10 ( i . e. 111/11) .
To determine the pipeline velocity at conditions, merely
divide the velocity given in Equation 1 by the compression
ratio given in Equation 2. After selecting the minimum
pipe size on the basis of velocity, check any long runs for exces-
sive pressure drop using an appropriate drop chart. For exam-
ple, a velocity of 25 fps in black iron pipe represents about
0.25 psi loss per 100 ft. of run. Although this is a little
above the recommended minimum of 20 fps and, depend-
ing on the layout, would probably be acceptable from a tur-
bulence standpoint, a high total frictional loss might dictate
using a larger pipe.
This might seem to be somewhat complicated at first, but
its the most accurate way to avoid problems in sizing com-
pressed air piping. Once you get the hang of it, its easy to use.
After carefully selecting a conduit size that eliminates unnec-
essary loss, be sure to pay the same attention to downstream
items such as quick disconnects, regulators, filters, controls,
fittings, number of drops from a given header and number of
connections per header, so as not to offset the gains made with
the pipe. Good piping performance is not an accident it
takes planning. p
Hank van Ormer, Don van Ormer and Scott van Ormer are owners of
AirPower USA, Pickerington, Ohio. Contact them at
HankvanOrmer@aol.com and (740) 862-4112.
Figures: AirPower USA
FEBRUARY 2005 WWW.PLANTSERVICES.COM 59
PERFORMANCE
Compressors
Table 1. Compression ratios at gauge pressures
psig Compression ratio
60 5.05
70 5.76
80 6.44
90 7.12
100 7.8
110 8.48
120 9.16
130 9.84
140 10.52
150 11.20
200 14.5
The calculations and chart
data help to identify only
the probable minimum
pressure drop.

A 2005 .PLANTSERVICES. 41
PERFORMANCE
Compressors
Air compressors are key components in many
manufacturing and process industries. Teyre interesting
from an engineering standpoint because of the many disci-
plines involved in their design and application. Also, theyre
interesting historically because they are among the earliest
machines, and most people have an intuitive understanding
of compressor operation. Anyone who has used a bicycle
pump or a bellows has operated an air compressor, and they
know that work is required to compress air. Anyone knows
that compression heats air.
Big picture
Te major compressor classes are positive-displacement and
dynamic. An example of the positive-displacement class is
the bicycle pump or replace bellows, both of which change
the volume of a chamber to compress air. If a piston inside
a cylinder forms the chamber, the compressor is known as
a recipro-
cating type.
Tese are fur-
ther subdivided into single-
acting and double-acting. In a
single-acting type, only one piston face
compresses the air; double-acting types use both faces alter-
nately. Reciprocating compressor sizes range from fractional
horsepower to more than 600 hp.
Another type is the rotary positive-displacement com-
pressor, in either helical screw or sliding vane varieties.
Te helical screw compresses air between a meshing rotat-
ing rotor and screw assembly. Helical screw compressors
are available in sizes from about 3 hp to several thousand
horsepower.
Te sliding-vane compressor uses a set of sliding vanes
p
ressure
Applying
the
By Ben J. Sliwinski
These are the elements that
afect air-compressor
performance
P
h
o
t
o
:

I
n
g
e
r
s
o
l
l
-
R
a
n
d
A 2005 .PLANTSERVICES. 42
PERFORMANCE
Compressors
placed in slots on a rotor eccentrically mounted in a cylindri-
cal casing. As the rotor spins, centrifugal force presses the
vanes against the casing wall to compress air between the
vanes and the casing.
Te second major compressor class is the dynamic type,
which compresses by converting air velocity into air pressure
using blades mounted on a rotating shaft. In centrifugal com-
pressors, air enters near the base of the impeller blades, accel-
erates along the blade and exits near the ends of the blades at
the circumference of the compressor case. Centrifugal com-
pressors range in size from about 100 hp to several thousand
horsepower.
In an axial-ow compressor, the air enters and exits along
the axis of the shaft, usually after passing through several
stages of rotor blades. Each set of rotating blades is separated
from the next by nonrotating stator blades. Air compressors
in aircraft jet engines are a common example of the axial
type. Axial-ow compressors are available in sizes from a few
hundred horsepower to several thousand horsepower.
Both positive-displacement and dynamic compressors can
be single- or multistage. Multiple-stage compressors need
two or more stages to reach the nal output pressure; the
output of one stage being the input to the next. Cooling the
air between stages improves compressor eciency.
Each compressor type reciprocating, screw, rotary
vane, centrifugal and axial has typical operating charac-
teristics. Tere is, however, overlap and, for a given appli-
cation, one might have a choice of types. Some important
characteristics are ow, pressure, capacity control and
lubrication.
The output
Te higher operating speed and continuous ow through
dynamic compressors gives them the greatest ow capacity.
Axial units provide the greatest ow capacity, but theres
overlap in ow capacity between centrifugal and axial com-
pressors. A rough ranking of the ow capacity of the posi-
tive-displacement compressors from highest to lowest would
be screw, sliding vane and reciprocating, but theres a great
deal of overlap.
Output pressures from positive-displacement units are sim-
ilar, with reciprocating units developing the greatest pressure.
Within the dynamic compressor family, centrifugal compres-
sors provide greater pressure capabilities than the axial type.
Figure 1 shows the approximate range of ow and pressure
for various compressor types.
Holding back
Some methods of capacity control are unique to a compres-
sor type, whereas other methods apply to all types. Cylinder
unloading is uniquely applied to reciprocating compressors.
It controls capacity by delaying the closing of the suction
valves so that air drawn into the cylinder can leak back into
the suction plenum before compression starts. Keeping the
valve open through the entire compression stroke completely
unloads the cylinder. More sophisticated systems that allow
the valve to close at any time during the compression stroke
achieve 100% to 0% capacity variation.
Other control schemes for reciprocating compressors
include start-stop, variable-speed and bypass control (in
which compressed air is bypassed to the suction). Vibra-
tion and bearing lubrication might limit variable-speed
control in reciprocating compressors to about 40% of nomi-
nal speed.
Slide-valve control is unique to rotary-screw compressors.
Te slide valve varies compressor displacement by returning
air back to the suction. Some slide valve applications also vary
the discharge port location, which varies the volume ratio.
Lift valve unloaders also allow air to return to the suction.
Te xed location of the lift valves results in stepped capacity
control as opposed to a slide valves stepless control.
Start-stop, suction throttling and variable-speed operation
also can control rotary-screw compressor capacity. Several
manufacturers oer VFD-drive screw compressor packages.
Centrifugal compressors use inlet-vane control, which
pre-rotates the incoming air to alter the compressors per-
formance curve. Variable-speed control also is eective for
centrifugal compressor capacity reduction. Output pres-
sure, however, is proportional to the square of the rota-
tional speed. Inlet-vane capacity control results in less of
The lay of the land
Figure 1. This chart shows the typical operating regimes for
diferent compressor types.
Reciprocating
Oil-ooded screw
Sliding-vane
Dry rotary-screw
Axial
Centrifugal
1
1
10
100
1,000
10,000
100,000
10 100 1,000 10,000 100,000 1,000,000
acfm
p
s
i
a
Some methods of capacity control
are unique to a compressor type,
whereas other methods apply
to all types.
F
i
g
u
r
e
:

R
e
s
e
a
r
c
h

A
s
s
o
c
i
a
t
e
s
A 2005 .PLANTSERVICES. 43
a reduction in pressure output than capacity control using
variable-speed control.
Below a minimum ow, air bypass might be necessary to
avoid surge conditions. Other types of control methods for
centrifugal compressors include suction throttling, adjustable
diuser vanes and movable diuser walls.
Axial compressor capacity can be controlled with vari-
able-speed drives or with adjustable stator vanes. Continu-
ously variable vanes with automatic control are usually sup-
plied on constant-speed applications with frequent changes
or uctuations in operating conditions. With continuously
variable stators, a drive ring adjusts the orientation of vanes
simultaneously.
Put them to work
Industry nds many uses for compressed air, including air-
driven tools, assembly line actuators and drives, powering
mold presses, injection molding, process machinery, material
transfer, painting, cleaning, blowing, dehydration, vacuum
packing and cooling.
Small- to medium-sized plants probably use reciprocat-
ing, rotary-screw and rotary-vane compressors. Laboratories
that require oil-free air might opt for oilless rotary-vane or
oilless reciprocating compressors. Plants with high air vol-
ume requirements will favor rotary-screw, centrifugal and
axial compressors. Dry rotary-screw, centrifugal and axial
compressors can provide high volumes of oil-free air. Many
larger-capacity applications for centrifugal and axial com-
pressors are found in industries where a process consumes
air. Tis includes air used for combustion, blast furnaces,
sewage treatment, compressed-air energy storage, air separa-
tion plants and ammonia production.
No seizing allowed
Air compressors can use any of several lubricants: petro-
leum-based oil, petroleum oil, synthetic blends and com-
pletely synthetic lubricant. Lubricant selection depends on
the compressor type, service and air quality requirements.
Some plants require a lube that is USDA-approved for H-1
application (lubricants with incidental food contact).
Te compressor lubrication system is dependent upon
compressed air quality requirements. Lubricant-free recip-
rocating compressors dont allow lubricant within the com-
pression chamber. Tese compressors have heat-resistant,
self-lubricating pistons, riders and rings. A distance piece
between the crankcase and cylinders prevents crankcase
oil from entering the compression chamber. Oilless recip-
rocating compressors are similar, but without lubricant in
the crankcase.
Rotary-screw compressor options are lubricant-injected,
dry or water-injected. Lubricant-injected units use the oil
to seal the space between the rotating screws, to remove
heat and to lubricate the rotors and bearings. Dry-type
screw compressors need no lubricant for sealing purposes,
operate at higher speeds and provide oil-free air. Water-
injected types use water to seal compression chambers
internal clearance and to remove heat. Te lubricated bear-
ings and gears in both dry and water-injected types are
isolated from the compression chamber. Centrifugal and
axial compressors use pressure-lubricated bearings and
drive gears. Shaft seals isolate the bearings from the com-
pression chamber so that centrifugal and axial compressors
can provide oil-free air.
Prime movers for driving air compressors include electric
motors, turbines (steam and gas), natural gas, diesel and gaso-
line engines, in constant-speed or variable-speed varieties.
Fuel for engines and turbines includes natural gas, landll
gas and sewage treatment gas. Steam turbine drives can be
used, particularly if the waste heat from some exothermic
chemical process can produce steam.
PERFORMANCE
Compressors
www.dresser-rand.com/default.asp
www.compair.com.au/pages/compservindustvane.html
www.knowpressure.org
www.gardnerdenver.com
www.gepower.com/prod_serv/products/compressors/en/
index.htm
www.manturbo.com/en/index.php
Books, journals and magazines
Improving Compressed Air System Performance: A Sourcebook
for Industry, Compressed Air Challenge, U.S. Department of
Energy, 2005.
Van den Braembussche, Veress, Arpad, Inverse Design and
Optimization of a Return Channel for a Multistage Centrifugal
Compressor, ASME Journal of Fluids Engineering, September,
2004.
Hanlon, Paul, C., Compressor Handbook, McGraw-Hill, 2001.
Compressed Air Systems: The Fourth Utility, David McCulloch,
Energy User News, August 2000.
Developments in Industrial Compressors and Their Systems-
European Conference, Proceedings of the Institution of
Mechanical Engineers, IMechE, 1994.
Gresch, M. Theodore, Compressor Performance, Butterworth-
Heinemann, 1991.
MATERIAL FOR SELF-STUDY
Compressor type Technology development
Reciprocating Valves, rings, packing-type seals,
capacity control and lubricants.
Screw Reductions in internal leakage,
increased bearing life, seals, capacity
control and lubricants.
Centrifugal/axial Further extension of stable
operating range through
improvements in impeller,
blade, stator, inlet and difuser
aerodynamics and materials;
labyrinth seals and bearings.
TABLE 1
A 2005 .PLANTSERVICES. 45
PERFORMANCE
Compressors
The crystal ball
Air compressors are a mature technology characterized
by incremental improvements in specic components and
subcomponents. Technology trends are strongly driven by
user requirements and by a manufacturers desire to increase
market share. A manufacturer might focus on a single com-
pressor type and work to expand its capabilities or might
oer a range of compressor types covering the entire market.
Some manufacturers might focus on niche industrial markets
(manufacturing, oil and gas, chemical) or by performance
range (very high pressure, very high ow rate, and so on).
Te competition is a complex mix; key factors are ow capac-
ity, output pressure capability, air quality, eciency and, of
course, price.
Manufacturers hold topics of current research closely, but
they are logically driven by the characteristics of each com-
pressor type. Table 1 shows the projected areas of technology
development by compressor type. Computers have been an
enabling technology in compressor development and applica-
tion. Finite-element analysis nds application in positive-dis-
placement and dynamic compressor technology development.
Computational uid dynamic software applied in 2-D and
3-D analysis is providing impetus to continuous improve-
ments in dynamic compressor performance. Computer-based
controls are standard on many compressor systems and often
include network and Internet capabilities.
Tis thumbnail sketch of compressor technology should
provide a useful jumping-o point for those who would like
to learn more about compressors. Te sidebar lists a number
of books, journals and Web sites for your reference. p
Ben J. Sliwinski owns Research Associates, Urbana, Ill. Contact him
at benj@prairienet.org and (217) 367-2270.
Figures: Research Associates and Ingersoll-Rand Co.
The compressor lubrication system
is dependent on compressed-air
quality requirements.
M 2005 .PLANTSERVICES. 28
MANAGEMENT
Compressed Air
M 2005 .PLANTSERVICES. 29
T
he easiest but potentially the most expensive way to improve your compressed
air operations is to hire someone to do a compressed air audit. A team of
experienced, professional compressed air wizards will visit your plant, spend
days measuring pressures and examining your system, and give you a list of all the
things that are wrong. In many cases, along with repairing leaks, poor piping and
other relatively minor problems, theyll advise you spend tens of thousands of dol-
lars on new compressors and controls.
ON
Te
HUNT
THE TOP 10 TARGETS OF A
COMPRESSED AIR AUDIT
By Rich Merritt, senior technical editor
And theyll be right. Following the recommendations of
an audit usually pays for itself in a short time by saving those
tens of thousands of dollars in operating expenses.
You are likely to benet from such an audit, but it may make
sense to know what the audit team is likely to nd so you can
identify the typical problems yourself. To that end, we asked
some top compressor manufacturers and service companies to
tell us what they usually nd. Listed below are their 10 most
typical, highest-payback audit items.
One caveat, though: Fixing these before doing a full sys-
tem audit can make it more dicult to justify the higher-cost
improvements. Often, many parts of a system upgrade that
improve the quality, reliability and repeatability of the system
are nanced in conjunction with the reductions in waste,
says Mark Krisa, air audit manager at Plant Air Technology
(www.plantair.com). Energy reductions associated with your
eorts cannot be incorporated into future return on invest-
ment projects.
In other words, if you pick all the low-hanging fruit and
do an audit, the payback on the investment wont be as high.
Its a numbers game, but it might be important in determin-
ing who pays what, and whose budget it comes out of. Te
numbers game may decide which you do rst: the professional
audit or your own list of low-cost repairs.
1. Plug away at leaks
One of the most common problems is leaks, says Wayne Perry,
technical director, Kaeser Compressors (www.kaeser.com).
Studies indicate that as much as 35% of the compressed air
produced in the market today is wasted to leaks, and everyone
has leaks. Identifying and correcting them might save not only
the purchase price of a compressor, but reduce the amount of
energy needed to run the compressor.
It has been our experience that plants that have no for-
mal, monitored, disciplined, compressed air leak-manage-
ment program will have a cumulative leak level equal to 30%
to 50% of the total air demand, adds Henry van Ormer,
engineer and owner of Air Power USA (www.airpoweru-
sainc.com). Every 8 cfm to 12 cfm leak can cost you $800
to $1,200 per year.
Van Ormer suggests setting up a short-term leak inspec-
tion program so that every sector of the plant is inspected
once each quarter to identify and repair leaks. Inspections
should be conducted with a high-quality ultrasonic leak loca-
tor during production and nonproduction, he recommends.
A record should be kept of all ndings, corrective measures
and overall results.
Afterward, he suggests setting up programs to monitor the
air ow to each department and making each department
responsible for identifying its air usage
as a measurable part of the expense for
that area.
If you get rid of leaks, you might
cause other problems. Elimination of
waste, such as leakage and articial
demand, may result in reduced loading
on compressors that are not equipped
to turn down eciently, says Mike
Bakalyar, manager, enhanced services,
Gardner Denver (www.gardnerdenver.
com). Dynamic eciency may actually
degrade, resulting in very little positive
eect on energy usage (Table 1). Waste
has been reduced, but the cost recovery
shifts to compressor controls.
Te campaign to reduce leaks must
be complemented with conguration
and control improvements that will
allow the air generation to turn down
with the reduced demand, he says.
2. Take down overpressurization
Excessive pressure increases leaks
and wastes money. Some end users
will actually increase pressure in an
attempt to compensate for capacity
issues, Perry says. In fact, increas-
ing pressure will have the opposite
eect on air ow and often exacerbate
the problem. Tere is also a propor-
tional relationship between pressure
and power consumption for every
10 psi in excess pressure there is a 5%
increase in power cost.
Norm Fischer, president, Centrifu-
gal Equipment Service (www.cescon-
trols.com), says too-high pressure will
amplify problems, not solve them. Te
easy answer to many system problems
is to jack up the pressure. Unfortu-
nately, the leaks will leak more, and
unregulated users will waste more air
and more energy.
Lowering the pressure may solve
problems. Lower system pressures
mean less mass required, therefore
fewer compressors running, Fischer
adds. Compressors are usually more
ecient when run at lower pressures.
But you have to convince produc-
tion, Fischer says. Often, the greatest
struggle is gaining the condence of
the production people that the system
is reliable enough to respond when
required, so they will lower the pres-
sure requirement closer to the actual
design requirement.
3. Zero in on air requirements
Often, production overestimates the
amount of air it needs. If production is
allowed to dene their own compressed
air requirements based on as much as
they want whenever they want it at any
pressure, the system will never operate
eciently, Krisa says.
Dave Booth, systems specialist at
Sullair (www.sullair.com), agrees.
Te entire paradigm under which the
compressed air system operates must
change, he advises. We must shift
from the principal goal of maintaining
a minimum pressure and that higher
pressure is OK to the goal of maintain-
ing a consistent and stable pressure.
Plants must change their focus from
maintaining air supply to supplying
air to meet demand. More air and more
pressure is simply more cost.
Van Ormer says, More often than
not, it is one process that needs a cer-
tain minimum pressure. Tese claims
should always be reviewed. In one audit,
the rest of the plant could run on 80
psi but the compressed air system had
to run at 98 psi because the grinding
area with only 20% of the demand
required it. Testing revealed that the
actual inlet pressure to the tool was 63
psig at load. In other words, we had a
35 psig pressure loss from the header
to the tool. Further tests indicated that
the optimum inlet pressure for these
particular tools was 75 psig. Te plant
installed a larger feed line and a regula-
tor to deliver full ow to the grinders
at feed pressure. Te header pressure
was lowered to 85 psi. Results after 18
months showed that tool repair went
down for the grinders, production
increased by 30% and total air demand
fell from 1,600 to 1,400 cfm. Total
cost for the regulator, piping changes
and adding quick disconnects on nine
grinders was $1,362. Annual electrical
savings are about $18,000 per year.
In those cases where you have a small
area that actually needs high pressure,
van Ormer suggests setting up a sec-
ondary, smaller, high-pressure unit
or an appropriate booster, rather than
drive the entire plant system at the
higher pressure. Expecting the supply
system to support a black hole is not a
realistic design criterion, Krisa adds.
M 2005 .PLANTSERVICES. 30
MANAGEMENT
Compressed Air
Table 1. Demand can afect efciency
Leaking system Tight system Tight plus controls
Process demand* 1,500 cfm 1,200 cfm
Demand reduction 300 cfm (20%)
Power 259 kW 243.5 kW 206.9 kW
Dynamic efciency 5.8 cfm/kW 4.9 cfm/kW 5.8 cfm/kW
Annual energy cost* $108,780 $102,270 $86,898
Net savings $6,510 (6%) $21,882 (20%)
*Example system at 90 psi and $0.05/kWh
Smooth it out
Flatline compressed air header pressure
to stabilize performance of pneumatic
equipment.
Not for knuckleheads
Get a competent, unbiased compressed
air system audit.
Baselining a compressed air sys-
tem
Save thousands with simple measure-
ments, a pencil and paper, a calculator
and a telephone.
Energy conservation efort saves
Conbraco $860,000
Keys were correcting compressed air
leaks, power factor and billing errors.
www.plantservices.
com/0505_extras
MORE ON
COMPRESSED
AIR AUDITS AT
PLANTSERVICES.COM
4. Eliminate Mr. Tee
One of the simplest xes in a compressed air system is to
replace tee connections with directional angle entry connec-
tions. In a piping system where a feed line of compressed air is
trying to feed into another air line, the turbulence caused by
a 90 entry often causes a 3 psi to 5 psi pressure loss. Such a
loss can cost you $1,200 per year at every one of those tees.
More important, the back pressure sends a false unload signal
to the controls, causing premature unloading or extra compres-
sors to be on line, van Ormer says. Using a 30 to 45 direc-
tional angle entry instead of a tee will eliminate this pressure loss.
Te extra cost of the directional entry is usually negligible.
Even worse is a dead-head tee connection, where com-
pressed air enters at opposite ends of the tee, causing
extreme turbulence. In one instance, the pressure loss was
almost 10 psi, van Ormer says. This is 300 hp worth of
air, or about $12,000 annual power cost. To avoid such
a situation, he suggests using two directional angle con-
nections spaced so the incoming air does not cause such
turbulence.
5. Set sights on bad piping
Convoluted piping, piping restrictions, old pipes and incor-
rect pipe sizes often lead to pressure loss. In a well-laid-out
system, the interconnecting piping from the compressed air
supply to the process and header distribution piping should
produce no pressure loss. In many cases, it is easy to simply
replace a section of pipe to gain eciency.
Booth looks at it more simply. If you cannot walk up to your
compressed air piping system and in a brief glance obviously
gure out how the air gets from the compressors through the
contaminant removal system and to the plant and then on to
the points of use, you probably have a problem, he says. Look
at your piping. Is it logical? Does it make sense? Would you
install it that way?
Piping is a major consideration, especially in older
facilities or shops that have grown and expanded, Perry
says. Cast-iron piping will rust over time, releasing rust
and scale into the compressed air and creating buildups
at various points in the system. Tis not only degrades air
quality, but reduces the eective internal diameter of the
pipe and obstructs air ow creating unwanted pressure
drops and velocity problems.
Measuring pressure loss in piping sections will identify
the worst culprits. If you nd a severe pressure drop through
some convoluted sections, or determine that the pipe is
too small, the cost of changing the pipe often pays back
quickly. Upgrading to copper or aluminum piping provides
excellent value for money and ideal delivery characteristics,
Perry says. When upgrading, ensure that the physical pip-
ing diameter is sized to deliver the required air ow with
minimum pressure drop.
Interconnecting piping between two or more compressors
often needs attention. Tis is the piping area where we nd
the most opportunities for improvement, van Ormer says,
particularly in systems installed after the late 1970s. Older
systems were put in more carefully.
6. Blow away obsolete restrictions
Clogged lter elements, forgotten manual drain traps and
neglected separator cartridges can cause signicant drops in
pressure and negatively impact capacity and reliability, not
to mention creating air-quality issues.
In one example, a pet food plant was running a 150-hp
rotary screw that produced 750 cfm. Te discharge pressure
was 120 psi, and actual pressure at packaging was 90 psi,
van Ormer says. Investigation of the main header from the
compressor room to process found an old, forgotten inline
lter full of rust and scale. Te lter was removed, the dis-
charge pressure was reduced to 100 psig, and this produced
an annual electrical energy savings of $6,570.
7. Spot small-caliber storage
Perry says insuff icient storage is a common problem.
Across the board in manufacturing and processing, the
value of an appropriately sized air receiver and appropriate
compressed air piping is underestimated, he says. Tese
tanks provide a rst stage of moisture separation to help
maintain compressed air quality. However, their primary
function is storing and delivering compressed air to help
meet periods of peak demand and to prevent excessive
compressor cycling.
All air systems will do better with storage between the
user and the process. Te amount of eective storage for
any system is where the operating control band is equal-
ized by the back pressure in the system. In one example, a
280-hp, two-step controlled, lubricant-cooled rotary screw
compressor was running 24 hours per day, seven days a
week at a relatively level load of 70% ow. Te unit had
very low storage capacity and would unload, idle for 15 to
25 seconds, then reload.
Te bleed-down time for this unit was one minute to reach
full unloaded power. Te unit did not stay o long enough to
reach the low power point and spend time there, van Ormer
explains. Correcting the eective storage to almost 10 gals
per cfm created a two-minute idle allowing full blowdown to
the low idle input power and a full one-minute run at this low
power before reloading. Tis resulted in an annual electricity
cost reduction of more than $14,000.
8. Shoot down inappropriate use
Unregulated use of compressed air, and using compressed
air for inappropriate purposes, wastes a lot of energy. Con-
sidering that it costs eight times as much to use air as it does
to use electricity, you may want to reevaluate unregulated
air-powered cabinet coolers, blow-os, vacuum generators,
mechanical pumps, air motors and hoists, vibrators, aeration,
spraying and a host of other equipment.
Compressed air is readily available in a plant, and the cost
of using it is not always understood. Terefore, when a need
M 2005 .PLANTSERVICES. 31
MANAGEMENT
Compressed Air
was identied, air was usually the easy answer, Fischer says.
Sometimes its even used for cooling people at workstations,
blowing dust, or to power vortex-type coolers and air to keep
food clean.
Open blow, refrigeration and vortex cooling may all be
replaceable with heat tube cabinet coolers with a potential
savings of 3.5 kW to 4 kW each on a 30- x 24- x 12-in. aver-
age cabinet, van Ormer says. Te initial cost is usually in the
$700 to $750 range with a potential resultant power savings
of $1,000 to $2,000 per year each.
He also suggests using venturi air amplier nozzles or
air inducers whenever possible, which will reduce blow-o
compressed air by 50% or more.
9. Pump away at pumps
Air-operated diaphragm pumps tolerate aggressive conditions
relatively well and can run dry, which makes them a favorite
with plant personnel. But is an air-operated pump the best
solution? Electric motor-driven diaphragm pumps are read-
ily available, and may work just as well. A 2-in., air-operated
diaphragm pump, pumping water at 130 gpm, will use 25
hp worth of compressed air at a cost of $9,947 per year. A
3-hp electric pump may well do the same, at an energy cost
of $1,989 per year.
If air-operated pumps must be used, consider adding con-
trols to shut the pumps o when they are not needed. Pumps
waste the most air when they are pumping nothing. Also,
check to see if the pump is running at the lowest possible pres-
sure. Simple controls can increase pressure when needed.
10. Get a clean shot
Poor air quality adversely aects overall plant operations. What
you want is air that is clean and dry, and that requires maintain-
ing the lters, separators and driers. Neglecting recommended
maintenance can let oil get into the plant air and cause produc-
tion problems from dripping tools to sheyes in paint systems.
Poor maintenance also aects eciency. Van Ormer says
they did an audit and found three 150-hp compressors with
9.5 psi inlet pressure instead of the normal 14.2 psi. Tis
reduced the eective output from 725 cfm to 501 cfm, or
a 31% loss. Te plant had to run all three compressors at
full load to supply the 1,400 cfm demand. Investigation
discovered dirty and restrictive inlet conditions. Correct-
ing the problem resulted in almost $45,000 per year in
energy savings.
Change air/oil separators, lters and other components
at the optimum time, not when they clog up and cause a
pressure loss.
Bring in the big guns
Most of the lasting benets and big opportunities identi-
ed in air audits are really common-sense solutions, Booth
says. Most involve simple maintenance issues, misappli-
cations and general problems caused by neglect and not
fully understanding the consequences of mismanaging a
compressed air system.
But many plants can benet from more sophisticated
analysis by professional auditors who might recommend, for
example, changes to the control system. Te most com-
mon problem identied in complete air system audits is the
improper application or at worst, the complete lack of
compressor controls, Perry says.
Te pros cite symptoms such as compressors ghting
each other, too many compressors running, compressors
running just in case they might be needed, and uctuat-
ing plant pressures.
Tose problems are more dicult to nd and x than the
leaks, inappropriate equipment and rusty pipe problems
described here. Unless you are a compressed air wizard
yourself, you may need an audit to tell you whats wrong
with your system controls and overall design. p
M 2005 .PLANTSERVICES. 34
MANAGEMENT
Compressed Air
Several of the compressor wizards we interviewed prefaced their list of audit
items with a note bemoaning the sad state of afairs in most plants. Mark Krisa,
air audit manager at Plant Air Technology (www.plantair.com), said it best: The
common issue that exists in almost every facility is the attitude toward acceptance
of responsibility for the problems. If the intention is to correct the problems, the
organization as a whole has to take responsibility for the problems.
Krisa says that in many plants the production department determines compres-
sor requirements. This is classically based on keeping production happy, he says.
If the operating goal for the compressed air system is to keep production from
complaining, then production has no involvement in resolving problems, only in
creating them. Production may overestimate its needs for compressed air, misuse
it and misapply equipment.
Without formal changes to how the system is approached and an assignment
of responsibility, the system will ultimately return to the initial state of operation,
regardless of what eforts are made to purchase equipment to make the system
better, Krisa advises.
Dave Booth, systems specialist at Sullair (www.sullair.com), says its a lack of
understanding. Most plants really do not understand what it truly costs them
to operate their system and what efects it can and does have on their overall
production process and quality, he laments. If you dont know what it costs or
how it operates, how can you even begin to consider evaluating savings or other
potential improvements and changes?
LACK OF OWNERSHIP
S
C
ompressed air is too costly to use as a prime mover.
Consider the fact that the price of 100-psig air is in
the range of 18 to 32 cents per 1,000 standard cubic
feet of free air. In the automobile industry, compressed
air is a signicant part of the energy cost, ranging
from 10% in component plants to as much as 40%
in stamping plants. In a typical Ford plant, this can
represent anywhere from several hundred thousand
dollars to well over a million dollars per year.
One way to reduce this cost is by applying best
practices and a systems approach to improve com-
pressed air system eciency. Analyzing the case from only
the supply side limits the opportunities for improvement.
Focus on air user demands because that is what drives system
requirements. Concentrating on proper end-use application,
design, operation and maintenance ensures higher operating
eciency, lower cost and reduced production losses. Review
these aspects of your current air system:
Consider electro-technology conversion.
Align supply side with demand side.
Reduce system pressure.
Improve maintenance.
Eliminate inappropriate uses.
Tink in terms of life-cycle cost.
Electrotechnology conversion
Te history of compressed air in the auto industry goes back
to Henry Fords day. Ten, it was a byproduct of electricity
J 2005 .PLANTSERVICES. 38
RELIABILITY
Compressors
q ueezi n g
mo n ey
o ut
o f
t h i n
ai r
It pays to apply
best practices and a
systems approach to your
compressed air network
By Joe Ghislain
J 2005 .PLANTSERVICES. 39
production: waste steam from the generators turbines pow-
ered the steam engine-driven compressors that produced
compressed air. Electricity was in its infancy and couldnt
yet duplicate what could be accomplished with compressed
air. But times have changed.
Electricity now produces compressed air, and it can take
8 hp of input power to deliver only 1 hp of work where com-
pressed air is being used. At that rate, its obvious that it can
be more economical to use electricity to drive mixers, dryers
and blowers. Even direct-current nut runners are replacing air
tools, not just because of the energy eciency, but because of
increased quality by being able to tie torque feedback to the line
operation. Te advances in electrotechnology now oer many
ecient options for replacing compressed air applications.
Aligning supply with demand
System demand drives the supply requirements in any com-
pressed air system. You need to know the true air demands
and how to fulll them using proper compressor operation
(number and total horsepower, duration, pressure and ow).
Because the system is dynamic, it requires monitoring and
controlling both the compressors and air users.
First, develop a pressure prole that quanties system demand
characteristics. Take pressure readings after the main supply
components, at the beginning and end of the main distribution
system and at several points of use. Spread your readings out over
a period of time to establish the high, low and average system
demand. Te pressure variation you document indicates how
the system and compressor react to the demands.
Te adage, If you cant measure it, you cant manage it,
applies to establishing your baseline. While temperature and
dew point are useful air system measurements, the key metrics
are pressure, rate of air ow and electrical consumption. Tis
trio helps to determine the cost, monitor system operation
and establish a baseline for evaluating future modications.
Determine real-time air system eciency using the ow
rate (cfm free air) and power (kW). Let system size, compo-
nent location and estimated air ow range determine the ow
meter type and its location. Get your electrical consumption
by calculating kW or from a kWh meter. For smaller systems,
use voltage and current readings and apply the motor power
factor to estimate power consumption. Convert your kW/cfm
reading to cost by applying your electrical rate. Converting
compressed air usage into dollars puts the system operation and
improvements into terms that everyone can understand.
Apply controls to the compressors and other supply-side
components and to air users that have the greatest eect on
the system. Te type of compressor control and operation
depends on compressor type and system dynamics.
Control of an individual compressor requires consideration
of demand variation and control of air users to minimize
their eect on the system. Operate a minimum number of
compressors necessary to base load (operate at full capacity),
and use only one trim compressor to track the overall vary-
ing load. If you have multiple compressors of the same type,
use sequencing controls to run all but one at full capacity.
Tese sequencers not only control trim compressor turn-
down, but also will start and stop compressors according to
system demand.
For systems with multiple compressor types, it may be
benecial to separate the control for each type. Sophisticated
sequencing controllers and global systems now available can
control more than one compressor type. When using these con-
trol schemes, dont ignore compressor type. For example, rotary
compressors with modulating, or load/unload, capacity control
should be run fully loaded; variable-speed rotary compressors
should be used only for trim; and centrifugal units have rela-
tively ecient but limited, reduced capacity modulation.
Primary and secondary storage also can help align supply
with demand by minimizing the eects that air users have on
the system. Air receivers are vessels that store air thats needed
to meet peak demand events with minimal eect on changes
in pressure. Primary storage, located close to the compressors,
reacts to any system event. Secondary storage, located close
to an end use, minimizes the eect that a local high-volume,
low time-duration event has on the upstream system.
In conjunction with storage, an application that requires a
narrow pressure band can be equipped with a pressure/ow
controller that monitors downstream pressure and reacts
quickly to maintain line pressure stability.
As you can see, proper control and monitoring aligns air
supply with demand. Te correct control system must be
able to handle a compressed air system that is almost always
dynamic. If your production process or operating schedule
changes, verify your baseline numbers again to ensure the
change hasnt degraded your system dynamic.
Pressure reduction
Compressed air systems often operate at excessive output
pressure to compensate for pressure uctuations caused by
changes in end use (high intermittent volume). Operating
at elevated pressure increases the rate of air leaks, air con-
sumption at users and energy consumption. Te benets of
reducing your supply pressure follow the same logic that applies
to pressure drop, except in reverse: every 2 psi increase in pressure
costs an additional 1% in power. For example, running a 100-hp
compressor at 80 psi rather than 100 psi saves approximately
$3,500 per year at 5 cents per kWh. Operating a compressed air
system at the lowest possible pressure is well worth the eort.
Often, only a small number of end uses require high pressure.
Tese need to be addressed individually. Sometimes the need for
high pressure is merely a perception that entered plant lore when
someone once said, We have problems with the equipment if it
drops below this pressure. Question everything.
Any number of things can cause problems, including pressure
drop and swings in the line feeding the equipment. If you suspect
perception-based needs, address the cause. If a user truly requires
high pressure, either modify the equipment or isolate it.
Because modication is equipment-specic, it cant be
addressed in this article, but there are techniques to iso-
RELIABILITY
Compressors
late high-pressure loads. Air boosters or intensiers can be
used for intermittent loads. Booster compressors or separate,
smaller compressors can be used for continuous or high-duty-
cycle loads. Finally, if several loads require high pressure, it
may be possible to separate them from the main system and
supply them from one compressor, thus allowing the main
system to be run at a lower pressure.
Maintenance
Proper supply-side and demand-side maintenance is critical to
ecient operation. Often, system maintenance is considered
a necessary evil, one of the rst cuts to hit the budget, but it
may be the wrong place to start.
On the supply side, pressure drops across dryers and lters
can have adverse eects on system operation. Te concept
that 2 psi costs 1% in power applies, so its critical to change
lters and maintain dryers to minimize pressure drops. When
ignored, inlet air lters will load up and reduce compressor
capacity and eciency. A good air lter guideline is that a pres-
sure drop of 4 in. WC is equal to 1% of compressor capacity.
Air leaks are the biggest maintenance loss in any system.
Te Department of Energy suggests that a tight system
still has a 10% leak rate. Its common to nd industrial com-
pressed air systems with 20% to 30% leakage. Air leaks cause
eciency losses in several areas.
Te obvious one is the leak itself. At 5 cents per kWh, the
equivalent of a quarter-inch hole burns $8,382 per year. Te
additional rate of ow for compressed air leaks decreases
system pressure. Te resulting articial load requires the
system to operate at elevated pressure and can even prompt the
purchase and running of more unneeded compressor capacity.
Air leaks cause supply side equipment to cycle too often, thus
increasing maintenance and reducing equipment life. Te only
way to reduce these eects is to implement an aggressive and
ongoing air leak program that identies and xes air leaks.
Inappropriate uses
Compressed air isnt always the most appropriate energy source.
Many times its used because its convenient, but this is a costly
convenience. Blowing, drying and sparging are examples in
which air may have been selected because it was easy or was
a quick x for a production problem. Blowing and drying are
usually done at excessive pressure, which often can be reduced
drastically by regulating it and using high-eciency nozzles.
Low-pressure electric blowers are a viable option.
Cooling workers and cabinets are two other examples of
incorrect compressed air use. Purchasing a fan or a cooling unit
can provide a payback in less than a year, perhaps within several
months. Vacuum generation, diaphragm pumps and vacuum
venturis also are applications that you should review.
Think life-cycle cost
While this may be a basic concept, its often overlooked.
System design and the equipment purchased to implement
it determine 80% to 90% of the ultimate operational costs.
Total life cycle cost and benets must be weighed carefully
before selecting the most cost-eective option, not only for
the compressed air supply system but also for the end uses.
Where is the sense in making air compressor purchases based
on rst cost, while the units life-cycle cost is less than 10%
hardware and more than 80% energy?
Rarely is pressure drop a consideration when purchasing or
designing equipment and systems, yet the pressure drop across
dryers, lters and piping systems has a dramatic eect on energy
costs. Youd be wise to analyze the incremental cost of increasing
hardware size to reduce the pressure drop. Often, the incremen-
tal cost is small compared to the ongoing energy cost.
Specify air users that operate at the lowest possible pres-
sure. I know of one instance where two identical large presses
were purchased for two locations. One plant specied a 60-
psi operating pressure, the other let the supplier dictate the
operating pressure. Te result was a press operating at 60 psi
and the second operating at 80 psi. Te dierence in operat-
ing cost was more than $300,000 per year. Tis illustrates
that using life-cycle cost to drive design, specication and
purchasing is critical to ecient long-term operation.
Concentrating on proper application, design, operation
and maintenance ensures the highest operating eciency
and lowest cost. It improves energy eciency while reduc-
ing production losses. Reducing compressed air costs, like
reducing any energy cost, has a direct eect on the bottom
line. Making compressed air systems more ecient reduces
costs and makes a company more competitive. p
Joe Ghislain is business strategy manager for vehicle operations at
Ford Motor Company in Dearborn, Mich. Contact him at jghislai@
ford.com or (313) 594-2695.
Te Compressed Air Challenge (CAC) is a national collaborative
of public and private organizations dedicated to increasing the
understanding and improving eciency of compressed air systems
within U.S. industry. Te CAC oers Fundamentals of Com-
pressed Air Systems and Advanced Management of Compressed
Air Systems training, and in cooperation with the US Department
of Energy, the Compressed Air Systems Sourcebook for Industry
as well as the Qualied AIRMaster+ Specialist training. CAC
has built a reputation for being a reliable resource for cost-eective
solutions and unbiased information, including the recent publica-
tion Best Practices for Compressed Air Systems, a comprehensive
and detailed reference for plant personnel. For more information
about CAC training and publications call (301) 751-0115 or visit
www.compressedairchallenge.org.
J 2005 .PLANTSERVICES. 40
RELIABILITY
Compressors
While temperature and dew point
are useful air system measurements,
the key metrics are pressure, ow
and electrical consumption.
A 2005 .PLANTSERVICES. 43
EFFICIENCY
Compressors
COMPRESSED AIR
IS FOR PUSHING,
NOT PULLING
Improve energy efciency by
restructuring vacuum generators
By Dan Bott
V
acuum generators powered by
compressed air represent one
of the most inecient uses of
that valuable utility. Behind
every quiet, vibration-free, low-cost,
environmentally friendly venturi vac-
uum pump is an expensive, energy-con-
suming, large-footprint air compressor.
In many applications, electric motor-
driven vacuum pumps can achieve the
same performance as vacuum generators
while using one-fourth to one-tenth the
energy. In fact, replacing compressed air
vacuum generators might be one of the
last methods remaining for increasing
production energy eciency and taking
overworked air compressors o-line.
The simple mechanism
Venturi-style vacuum pumps, also
called vacuum generators or compressed
air ejectors, produce vacuum by passing
high-velocity compressed air through a
venturi or nozzle. Performance depends
on the nozzles shape and size, com-
pressed air pressure and ow, and the
desired vacuum level. Te greater the
vacuum being maintained, the lower
the f low of induced air.
Other motive f luids include steam,
vapor, water and other liquids.
Compressed air vacuum generators
are common in industry. Palletizers,
material-handling systems, pick-and-
place operations, drum-type vacuum
cleaners and packaging applications
are just a few examples. Each genera-
tor is mounted in close proximity to the
point of use, with supply tubing con-
necting the vacuum device to a central
compressed air system.
Vacuum generators are reliable, com-
pact, lightweight and quiet. Tey have no
moving parts and can be mounted directly
on production machinery. Teir mainte-
nance requirements are minimal. Teyre
available in aluminum, plastic and corro-
sion-resistant construction for harsh appli-
cations. Replacement or repair is simple
and requires no special tools or training.
Air versus electricity
A vacuum generator, by itself, is equiv-
alent to an engineless automobile.
Neither makes any noise nor requires
maintenance. Neither has an operating
cost. Te drawback, of course, is that
neither does any useful work. Without
an air compressor engine operating
under the equipment room hood, the
vacuum generator can do no work. One
cant evaluate a vacuum generator with-
out accounting for the air compressor in
the calculation. An objective evaluation
compares the relative eciencies of elec-
tric-driven vacuum pumps and vacuum
generator-compressor combinations.
Vacuum generator literature uses
two key terms: induced airow and air
consumption. Induced airow is the air
being evacuated from inside the vacuum
system. Air consumption refers to the
compressed air the vacuum generator
requires. Tese ows combine and dis-
charge through an exhaust port.
A vacuum pump driven by an electric
motor, on the other hand, uses a varying
rotational swept volume to produce a suc-
tion that induces ow from inside the vacuum system. Te rotor
compresses the induced ow and discharges it to an exhaust port.
Motor-driven vacuum pumps consume no compressed air.
Analyzing the question
Production demands dictate vacuum pump size and serve as
the basis for evaluation. Continuous vacuum applications seek
to maintain a xed vacuum level. In cyclic applications, on the
other hand, a chamber at atmospheric pressure is evacuated to a
target vacuum level for a period of time and then vented. Given
the two application types and the two vacuum technologies,
which pairing is most energy ecient?
Te answer lies rst in determining the cost of compressed air
and how much induced ow a vacuum generator develops. Table 1
illustrates a representative continuous application with 20 vacuum
generators. Te values represent the average performance of typical
industrial units. Each generator requires 20 scfm of motive air to
induce a vacuum ow that is a function of vacuum level.
Nearly every vacuum generator application uses 100-plus psig
air from the central compressed-air system and regulates it down
to the recommended 30 psig to 90 psig for the venturi. Rarely
is low-pressure air generated specically for these applications.
Tis regulation in itself is a major source of ineciency.
A typical compressed-air system produces no more than 4
scfm output for every input horsepower. While a standalone
air compressor is more ecient, losses through ancillary equip-
ment, headers and the partial loading of compressors reduce
overall system eciency. So, we need nearly 100 compressor
horsepower to drive the 20 vacuum generators.
Table 2 highlights typical performance ratings for an elec-
tric motor-driven vacuum pump that is equivalent to the 20
vacuum generators. Rotary lobe blowers are for vacuum levels
below 15 in. HgV, and rotary vane vacuum pumps are used
for higher vacuum levels.
Te data in the tables reveal that the vacuum generator
system requires nearly 100 compressor horsepower while the
motor-driven vacuum pump system needs 15 hp. For any
level of vacuum, an electric motor-driven vacuum pump is at
least 6.5 times more ecient than a compressed-air vacuum
generator.
Te real kicker is that, in many cases, pressurized air
ows through the vacuum generator even when no vacuum
is needed. Most vacuum generator installations have built-in
shuto valves to avoid this situation, but bypassed or defective
valves add signicant waste.
Up and down repeatedly
Vacuum generator specications typically include a table
showing pumpdown time needed to achieve a targeted vac-
uum in a volume of 1 cu. ft. Tis pump selection information
is used for applications requiring vacuum pickup or parts
movement in production machinery. Locating the vacuum
generator close to the point of use reduces the volume of
piping to be evacuated. Smaller chamber volumes result in
faster cycling.
Te next example highlights a cyclic application with 20
use points, each consuming 30 scfm of compressed air. Te
total air demand is 600 scfm, which represents about 150 com-
pressor horsepower. Figure 1 shows the horsepower required to
pumpdown a volume of 20 cu. ft. Pumpdown times range from
fractions of a second for 5 in. HgV to more than 20 seconds
for 27 in. HgV. Many production applications require shorter
pumpdown times, but this chart is intended to illustrate the
relative eciencies of each vacuum pump technology over a
wide range of conditions.
Te disparity between vacuum generator and electric motor-
driven pumps energy eciency in cyclic applications is quite
remarkable. A closer look at cyclic applications reveals that dur-
ing an average evacuate/hold/release cycle, compressed air might
be used only one-third of the time. In the absence of functional
shuto valves, compressed air ows needlessly during two-thirds
of the cycle.
Even with shuto valves in place, energy comparisons
between motor-driven vacuum pumps and vacuum generators
are still valid. Motor-driven vacuum pumps can be downsized
to meet the actual application requirement. Figure 1 illus-
trates the apples-to-apples comparison when both technolo-
gies are operating at full load.
Te dierences in eciency are alarming for both continu-
ous ow and closed applications. In addition, the eciency
gap widens as the required vacuum level increases. Te bot-
tom line is that from an energy perspective, compressed-air
vacuum generators are not environmentally friendly.
A 2005 .PLANTSERVICES. 44
EFFICIENCY
Compressors
Table 2: Electricity doing the work
Vacuum
(in. Hg)
Approx
hp reqd
Induced fow
(scfm)
Induced fow
(acfm)
5 10 540 648
10 10 220 330
15 15 140 280
20 10 60 180
25 7.5 20 120
Typical rotary vacuum pump efciencies; acfm = scfm * P
1
/P
2
Table 1: Pulling a load
Vacuum
(in. Hg)
Air consumption
(scfm)
Induced fow
(scfm)
Induced fow
(acfm)
5 400 540 648
10 400 220 330
15 400 140 280
20 400 60 180
25 400 20 120
System of 20 typical venturi vacuum pumps; acfm = scfm * P
1
/P
2

(absolute)
A 2005 .PLANTSERVICES. 45
Restructure it
Vacuum generator popularity is derived from its low
capital cost. OEMs favor first cost over operating cost.
Regardless, retrofitting each venturi vacuum pump with
a dedicated electric motor-driven vacuum pump would
be cumbersome.
Te alternative is a central vacuum system. Like compressed
air, vacuum can be generated at a central location and distributed
through a network of headers and drops. Unlike compressed air,
vacuum supply piping can be made of light, exible, inexpensive
and easy-to-install plastic.
Installing a duplex vacuum pumping station to provide 100%
backup can put reliability issues to rest. If the lead pump needs
servicing, the backup pump takes over automatically. Duplex
vacuum pump packages with alarms, automatic sequencing,
PLC interfacing and manual overrides are standard products.
Tey can be installed in the same location as the existing extra
air compressor.
A central vacuum system retains the advantages of indi-
vidual compressed-air vacuum generators. Teres no pump
or motor noise at the point of use. Vacuum tubing takes up
about the same space as compressed-air supply tubing. Tere
are no heat problems or oil mist. Te servicing schedule for
an electric motor-driven vacuum pump is usually identical
to that of an air compressor.
Heed the numbers
Economic evaluation is straightforward. First, determine
the cost of compressed air and the total air consumption for
the vacuum generators. Amortize the cost of maintenance,
oor space, repairs and the like, and add it to the base electric
cost. In addition, determining the total air leakage in the
compressed air system is sometimes sucient to initiate a
leak repair program for the entire site.
With current costs identied, evaluate proposed costs to
determine if switching to vacuum pump technology is jus-
tied. Dont forget inlet ltration for those rotary vacuum
pump technologies that require it. Life expectancy for some
rotary technologies is closely related to the eciency and care
of the inlet ltration system.
Focus on the system with the lowest energy and mainte-
nance cost for the required production throughput. Com-
pute the payback period if a new vacuum pump is involved.
Te annual cost for a 100-hp air compressor, including costs
for cooling, air treatment, maintenance, depreciation and
the like, exceeds $50,000 at $0.06/kWh. A typical 15-hp
electric motor-driven vacuum pump, on the other hand, has
an annual operating cost of $7,700.
Many applications use hundreds of compressor horse-
power to generate vacuum. Replacing these systems with
dedicated electric vacuum pumps can save thousands of
dollars annually. How vacuum is generated is irrelevant to
the production process, as long as vacuum is at the required
level when needed.
Not every application is a candidate for electric vacuum
pump replacement. But, if demands on the compressed-air
system are suggesting the need for an additional compressor,
its worthwhile to investigate alternate vacuum technology. Te
eort can result in signicant cost and energy reductions. p
Dan Bott is owner of Dan Bott Consulting, an independent indus-
trial vacuum and compressed-air system auditor in Loxley, Ala.
Contact him at (251) 960-1026 or at www.danbottconsulting.com.
What at rst appears to be a
winning way to produce vacuum
turns out to be a technology with
inadequate performance.
No more imput than necessary
150
35 40 40
30
25
20
Vac. generator hp
Rotary vac. hp
160
140
120
100
80
60
40
20
150 150 150 150 150
5 10 15
Inches of mercury vacuum
20 25 27
H
o
r
s
e
p
o
w
e
r
Figure 1. Closed-system pumpdown energy requirements to
evacuate 20 cu. ft.
EFFICIENCY
Compressors
O 2005 .PLANTSERVICES. 41
M
ost compressed air systems are relatively mod-
est at rst, but grow and develop as production,
inappropriate uses and leaks increase over time.
Even assuming that air leaks and inappropriate
uses have been investigated and reduced, adding production
equipment raises demand for compressed air. Some equipment
may need a dierent operating pressure, and the required air
quality may change. Tese modications represent potential
problems for the plant engineer who wonders:
What is the required volumetric ow rate and the pressure
of compressed air, and what size of air compressor should
be added?
What type of compressor and control system oers greatest
reliability and lowest life-cycle operating costs?
Where should the compressor be located?
Is sucient power, ventilation and cooling capacity available?
What type of system capacity controls would be best?
Is primary compressed air storage sucient?
Is distribution piping adequate?
Is secondary compressed air storage sucient?
Is the current compressed air quality satisfactory?
Does this project require professional help?
Compressor size
Te current average and peak compressed air ow rates, in
cubic feet per minute (cfm), should have been established
before considering any proposed additions. Te rated output
of the existing compressor(s) also should have been estab-
lished. Te specications for the proposed equipment that
needs additional compressed air should state the required
ow rate, pressure and air quality. Tis information provides
the new total ow rate. Deducting the rated output of the
existing compressor(s) gives the additional air ow required
from a new compressor.
Tis calculation ignores potential leakage and increased
inappropriate compressed air use. Also, it doesnt account
for dierences in the frequency of operation of each piece of
production machinery. Individual consumption peaks might
not occur simultaneously. Nevertheless, youll need to deter-
mine the average and peak ow rates (Table 1).
By David M. McCulloch and Bill Scales, P.E.
COMPRESSORS
Reliability
RELIABILITY
Compressors
Planning
air SyStEM uPgraDES
10
Ten steps to
successful system improvements
Dont add a fudge factor to the required air f low
when calculating the rated capacity for the additional
air compressor(s) because it could result in the new unit(s)
operating at less than full capacity most of the time, rob-
bing eff iciency. Should you decide to oversize the addi-
tional compressor by 20% or more, select the compressor
that combines the best full-load and part-load economy
to minimize the operating cost over the full range of the
compressed air requirements.
Another important consideration is standby capacity in
case of compressor malfunction or needed repair. Tis may
require at least one additional compressor. Conventional
wisdom says that three 50%-capacity compressors are bet-
ter than two 100% compressors, because this provides more
exibility without sacricing system reliability, particularly
during periods of reduced consumption. Its also benecial
to operate the smallest total compressor horsepower, par-
ticularly for periods of reduced capacity requirements, such
as a second- or third-shift operation. Tese factors, com-
bined with reliable compressor service, are keys to main-
taining energy and production eciencies and protable
outcomes.
Compressor type
Each type of compressor has its advantages, disadvantages
and preferred range of capacity and pressure. Table 2 pro-
vides a simple method for comparing dierent compressor
types. Life-cycle cost analysis always is recommended, and
should include specied maintenance.
Compressor cooling is a major consideration. If water-
cooled, the important issues include availability and quality
of cooling water, disposal or recirculation, possible treat-
ment and overall cost. If compressors are air-cooled, adequate
room ventilation is essential. Heat recovery also is a potential
opportunity.
Compressor size and type determines the electrical
power requirements. Additional ancil lary equipment
may require a different voltage and current. Consider,
too, the availabil it y of the required electrical supply
and its support equipment. Ensure that proper circuit
protection is provided for the added electrical load.
Location, location, location
Many plants have a compressor room and, in some cases, that
room is shared with other equipment. Several factors must
be considered if a compressor is to be added. Tese include
the need for a foundation, space for maintenance activities,
space for drying and ltration equipment, room ventilation
to handle the added heat release and the sound level.
It might make sense to install the new compressor in a dif-
ferent location, perhaps closer to the point of greatest demand
or at the application requiring the highest pressure.
Capacity controls
Tere are several types of capacity control for individual com-
pressors, sequencing controls for multiple compressors and
pressure and ow controls for compressed air systems. Te
correct selection of each determines system eciency over
the anticipated operating ranges.
Two rules for achieving optimum eciency are (1) only
the number of compressors needed to maintain the required
system pressure should be in operation at any given time, and
(2) all but one, a trim compressor, should be running at full
capacity and pressure. Te trim compressor should have an
ecient capacity-control mode. If its a reciprocating com-
pressor, this could be unloading in a series of capacity steps.
For a rotary compressor, variable-speed control or variable
displacement is most ecient.
Storage and piping
Te size and location of the primary air receiver aects the
eciency of your capacity control. Ecient system control
may require changes in air storage volume, but adding receiver
O 2005 .PLANTSERVICES. 42
RELIABILITY
Compressors
Dont add a fudge
factor to the required
air ow when calculating the rated
capacity for the additional air
compressor.
Standby capacity in case of
compressor malfunction or needed
repair may require an
additional compressor.
A typical compressed air system
Figure 1. Compressor efciency is a function of pipe
geometry, leaks, operating pressure and other factors.
O 2005 .PLANTSERVICES. 43
volume wont compensate for insucient compressor capacity
or inadequate distribution piping.
Many compressed air distribution systems originate at
a primary air receiver in the compressor room where dis-
tribution piping moves the air throughout the plant. As
production increases, another one or more buildings might
be erected, but the supply of compressed air still passes
through the original distribution piping. If the piping isnt
adequate for the increased demand, the result can be exces-
sive pressure losses and increased energy consumption.
Intermittent high-volume demand can cause severe
dynamic pressure uctuations in the entire compressed air
system that can upset manufacturing processes. Many uc-
tuations can be softened with an appropriately sized and
located secondary air receiver that can provide enough air to
satisfy the intermittent demand without compromising the
pressure in the main system.
Air quality
Each piece of production equipment requires compressed
air at a given ow rate, pressure and air quality. Tese con-
siderations may vary signicantly. A cardinal rule is to avoid
drying and ltering compressed air any more than is needed
for the specic application. Going overboard can result in
increased pressure losses and energy consumption. Consider
the idea of satisfying the major compressed air requirements
centrally and supplementing these requirements locally,
where needed.
Many industrial applications can be served well with a
pressure dewpoint of 35F to 38F, which can be achieved
with a refrigerated dryer. Standard regenerative desiccant
dryers can drop the pressure dewpoint to 40F, and more
specialized dryers can bring it down to 100F. Dry the air
only to the requirements of the end users or to meet local
RELIABILITY
Compressors
Table 1. Basic demand worksheet
End user
identity
Minimum
fow (cfm)
Average
fow (cfm)
Peak fow
(cfm)
Cycle time
(seconds)
#1
#2
#3
#4
#5
#6
Totals
Note: In some cases, the minimum fow rate may be very low or
zero (cycle time - of) until an intermittent operation (demand
event) occurs, when there is a large demand (peak fow rate) for a
time (cycle time on). The combination of these determines the
average rate of fow. End users having a constant demand should
be tabulated by the average fow rate. Peak fow events may require
additional primary storage and secondary storage.
Table 2. Simple matrix for comparing compressor types
1
Compressor
characteristic
2
Compressor type
Two-stage, double-
acting reciprocating
Lubricant-injected screw Lubricant-free screw Centrifugal
Size and weight 3 1 2 2
Compact size and
complete package
3 1 1-2 1-2
Can be located close
to points of use
4 2-3 2-3 3
Maintenance cost 3 2 2 1
Foundation requirements 4 1 1 1-2
Reduced
capacity efciency
3
1-2 1-4 1-3 1-3
Lubricant-free air delivery
- lube/lube-free
4/1 2 1 1
Lubricant carryover
- lube/lube-free
4/1 3 1 1
Lubricant changes or
makeup - lube/lube-free
4/1 3 1 1
First cost,
including installation
4 1 2 2
Full-load operating
cost, kW/100 cfm
4
15 to 16 16 to 19 18 to 22 16 to 20
1
These evaluations are general in nature and might not cover specifc features of a given compressor type or manufacturer. Theyre intended
to provide a general guide for comparing compressors. Its important to evaluate each point in any comparison of quoted equipment. Other
factors to be considered include relative size and cost, warranty and service.
2
Each compressor type is rated from 1 to 4. Key: 1 = very good; 2 = good; 3 = fair; 4 = poor.
3
Refer to the section on compressor controls. Its important to compare kW/100 cfm at all reduced capacities.
4
Operating costs are based on full capacity at a discharge pressure of 100 psig; a full-load motor efciency of 92% and 0.746 kW/bhp.
ambient conditions.
Improving your air quality also requires lters to remove
particulates and might require coalescing and adsorption l-
ters to remove liquids and other contaminants. Tese added
lters will result in increased pressure losses and maintenance
requirements.
Outside help
In most cases, seeking professional help is a good idea. Equip-
ment distributors with good local service capabilities can be
helpful. An alternative approach is hiring an independent
compressed air consultant to provide a product-neutral opin-
ion or solution. p
Bill Scales, P.E. owns Scales Air Compressor Corp. in Carle Place,
N.Y. Contact him at bscales@scalesair.com and (516) 248-9096
ext. 611. David M. McCulloch owns Mac Consulting Services in
Shalimar, Fla. Contact him at macconsulting@cox.net and (850)
651-4540.
The Compressed Air Challenge (CAC) is a national collabora-
tive of public and private organizations dedicated to increasing
the understanding and improving efficiency of compressed air
systems within U.S. industry. The CAC offers Fundamentals of
Compressed Air Systems and Advanced Management of Com-
pressed Air Systems training, and in cooperation with the U.S.
Department of Energy, the Compressed Air Systems Sourcebook
for Industry as well as the Qualified AIRMaster+ Specialist train-
ing. CAC has built a reputation for being a reliable resource for
cost-effective solutions and unbiased information, including the
recent publication Best Practices for Compressed Air Systems, a
comprehensive and detailed reference for plant personnel. For
more information about CAC training and publications call (301)
751-0115 or visit www.compressedairchallenge.org. The authors
of this article, David McCulloch and Bill Scales were also the
authors of Best Practices for Compressed Air Systems.
O 2005 .PLANTSERVICES. 44
RELIABILITY
Compressors
The size and location of the primary
air receiver aects the eciency of
your capacity control.
]aNuavv :oo6 www.PIANTSIRVICIS.cox 41
ir Liquide Large Industries U.S. LP is part of Air Liq-
uide Group, which produces industrial and medical gases
and is headquartered in Paris. In the U.S., Air Liquide
maintains more than 125 production facilities and 700 customer
installations spread across some dicult to reach geographies.
Before 2002, Te U.S. company used a legacy vibration program
that was inconsistent in its application of technology and wasnt
producing the desired results.
Late in 2002, Air Liquide partnered with Rockwell Automa-
tion to provide vibration analysis services to 32 plants on a trial
basis. Te program expanded quickly in early 2003 to include
vibration monitoring at 107 primary production facilities. In
August 2004, Air Liquide recognized the need to expand its
predictive maintenance (PdM) program to include oil and infra-
red analysis, and again partnered with Rockwell.
Transitioning from legacy systems
Air Liquides needs and aggressive long-term strategy didnt
leave time for incremental continuous improvement. Te key
program objectives included:
Transition to state-of-the-art information solutions
Understand how reliability aects protability
Recognize the need for uniformity of predictive technologies
Leverage technology in geographically challenging areas
Analyze results and setting goals for improvement
Before 2000, maintenance and reliability functions were decen-
tralized as was the responsibility for approximately 100 plant sites
throughout the U.S. Since that time, Air Liquide in the U.S. has
centralized these functions under a new maintenance department
and regional reliability centers. Te new department deployed a
maintenance management process, a computerized maintenance
management system (CMMS) and preventive maintenance pro-
grams. With these systems in place, the department turned its
attention to its predictive maintenance programs.
Te sites used vibration, infrared and oil condition moni-
toring, but because no corporate standard existed for applying
predictive technologies. Applications were inconsistent and
couldnt be integrated. Similar data was taken at dierent
frequencies with dierent tools and at dierent locations for
like equipment. Data and reports varied in format and detail,
and information couldnt be compared and analyzed.
Because a company-wide CMMS didnt exist, predictive nd-
ings couldnt be linked to traceable work orders. Compliance of
corrective actions versus predictive ndings was unknown.
Several plants eventually used common vibration service
provider, but the program had problems. Te contractor owned
the data it collected. While the contractor provided some stan-
dard reports, Air Liquide had to pay for ad hoc analysis it could
have performed itself. But, more importantly, the contractor
had only one oce in the far southeast corner of the U.S. Given
RELIABILITY
Compressors
CONTROLLING THE COMPRESSORS
A
How Air Liquide integrated a successful predictive maintenance program
By Mark E. Lawrence, P.E., CMRP, and George F. Hofer
Air Liquides vast geography in the U.S., more than 80% of the
costs the contractor charged were incurred for travel.
By 2001, the infrared scanning program was probably the
closest to being national. Electrical standards had been devel-
oped and applied, and an internal resource was used for data
collection and report writing. While the program was eective,
it was used primarily for electrical devices and didnt include
any applications to identify process, xed equipment or rotating
equipment problems. And given Air Liquides geographic dis-
persal and the travel it required, having a single resource dedi-
cated to the program didnt seem viable in the long term.
Although it was implemented at several sites, oil condition
monitoring was probably the least used of the technologies. Tere
seemed to be signicant potential benets to increasing its use.
Customer relationships and proftability
Many of Air Liquides products are commodities. Tis places a
premium on our unit availability and equipment reliability. Not
only must there be robust maintenance programs to ensure high
reliability, but also a way to see problems far in advance. Tats
why predictive maintenance programs play such a large role in
Air Liquides reliability strategies, which play a signicant role
in its business strategy.
Before the 1990s, Air Liquide was primarily an air separation
company that produced oxygen, liquid argon and both liquid
and gaseous nitrogen. Tese air separation units could store
large quantities of liquid nitrogen and oxygen to sustain several
days of downtime. Te primary strategy was to reduce our mean
time to repair. Tis meant focusing on stocking spare parts and
having resources ready to react.
Air Liquide also operates four large cogeneration units
near Houston to provide electricity and steam to customers
in the area. Because theres no eective way to store backup
quantities of these products, Air Liquide adopted a conserva-
tive maintenance strategy that required signicant costs and
downtime to ensure predictable uptime.
In the late 1990s, Air Liquide expanded its business into
hydrogen, a product that cant be stored easily in quantities
needed to ride through any signicant downtime. Te stan-
dard maintenance strategy didnt apply because even small
trips caused huge problems. Major breakdowns could be
costly. For hydrogen, the conservative maintenance strategy
required too much planned downtime to be feasible. Te
strategy had to provide interruption-free production with
optimum downtime for planned maintenance.
In the meantime many customers had already embarked
on their own journey to reliability improvement. As they
eliminated reliability problems, the reliability of their gas,
steam and electricity suppliers gained greater visibility. Tey
demanded more from suppliers. A world-class predictive
maintenance program was becoming an important element
of our improvement strategy.
Unity of predictive technologies
Implementing a world-class, best-practice PdM program
begins with identifying the required measurement vari-
ables and appropriate technology for capturing reliability
data. For Air Liquide, the nature of the equipment monitored,
predominant failure modes and mean time between failures
necessitated using vibration, oil and thermographic technologies.
Failure modes, MTBF data and industry best practices dictated
a monthly interval for vibration data collection, quarterly for oil
analysis, and annually for infrared thermography scans.
Before August 2004, the Air Liquide PdM program consisted
of a decentralized approach to vibration monitoring, oil analysis
and thermography. Neither a unied nor integrated approach
was used in managing the data. Reports and recommendations
werent linked to the CMMS or return on investment data.
Improvements started late in August 2004 have demonstrated
an integrated PdM program through the following steps:
Standardized reporting process and information ow
Centralized and access through a Web interface
Centralized PdM technologies, reports and analysis
Recommendations and reporting linked to CMMS data
An interface for live-time, closed-loop progress measurement
Standardizing the reporting process and information ow
involved establishing a natural link between the reports and
recommendations submitted for vibration, oil and thermography.
Tis also included switching oil and thermography programs to
a single-source provider managed at the corporate level.
Rockwell Automation provided a common platform for inte-
grating vibration and oil data, and partnered with Predictive
Service Corp. to provide infrared thermography. A common
reporting platform was used as a report-generation tool installed
on each eld service engineers personal computer.
Because infrared thermography was provided on an annual
basis and generally only used for electrical components, this
technology was linked via the Web interface only. Rotating
equipment is added to the thermography scan when excep-
tions are noted in vibration and oil.
Centralizing the data storage into a common server platform
enabled Air Liquide to leverage the CMMS database, PdM
software database for reporting vibration, oil and thermographic
]aNuavv :oo6 www.PIANTSIRVICIS.cox 42
Where the efort went
Figure 1. Distribution of PdM work orders for a single year.
August '04 - August '05 PdM Data
Vibration
Oil
IR
IR Compliance
50%
17%
30%
3%
RELIABILITY
Compressors
]aNuavv :oo6 www.PIANTSIRVICIS.cox 43
scans. A Web interface formed a dynamic link among the three
data repositories.
When lab results are reported, oil condition data is added
to the PdM database and incorporated into the overall rec-
ommendations made regarding asset health. Viscosity, wear
particle analysis (WPA), analytical ferrography and other
diagnostics are reported along with vibration data for com-
plete machine condition status.
Since January 2005, we identied more than 148 cases
of viscosity breakdown or improper lubrication. In at least
ve cases, WPA revealed signicant particulate counts in
conjunction with increased vibration measurements. Fol-
low-up samples verify that proper lubrication was restored
and machinery repairs have been made. A direct link to the
recommendation and repair action taken enables these to be
tracked and linked to PdM program performance.
Linking recommendations from the three PdM technolo-
gies with the computerized maintenance management system
established the basis for measuring PdM program success. Once
completed and properly distinguished, key PdM program per-
formance indicators are tracked and measured on a real-time
interface. ROI data is linked directly to individual facilities, by
zone, by business class, and even summarized as a whole for Air
Liquide senior management.
A Web interface gives senior management a status at a glance
indicator, requiring properly classifying work orders and entering
nancial data into each action taken from PdM recommenda-
tions. Leveraging our CMMS capabilities, each PdM work order
was classied using the following critical components:
Work class: PDM denotes any work order initiated as a direct
result of a PdM program recommendation
Activity type: Denotes predictive technology used to identify
problem using one of four tags: Predictive, Vib, Predictive,
Oil, Predictive, IR or Predictive, Elec
Actual repair cost: Installation and repair cost of repair
required
Estimated savings: Te reliability engineers evaluation of
problem reported and potential cost savings averted by avoid-
ing catastrophic failure
Failure class: Type of machinery aected: motor, com-
pressor, etc.
Problem code: Detailed denition of problem component;
e.g., coupling failure
Developing the interface for monitoring real-time progress of
maintenance work orders and PdM recommendations involved
leveraging the existing CMMS database backbone and the cen-
tralized database storage architecture. Partnering with Predictive
Services, Rockwell Automation designed and developed a Web
interface capable of supplying a PdM Web management tool for
tracking the closed-loop PdM process. Te interface links, tracks
and reports progress of any maintenance action initiated from the
PdM program technologies. Tis tool enables senior management
to track program KPIs, maintenance activity bottlenecks and
overall program eectiveness quickly and eciently.
Fundamentally, the PdM programs integration and unity estab-
lished a direct link between maintenance repair recommendation
and maintenance action taken. It permits measuring and track-
ing nancial data, metrics and program success to provide ROI.
Unless a direct correlation is established between maintenance
action taken and recommended repair, program success cant be
measured fully. An overwhelming majority of PdM programs fail
because they lack this tie, and work orders and repairs are made
without regard to the PdM recommendation.
Leveraging technology
Managing maintenance activities on a national level is a chal-
lenge many large organizations face. Geography and disparity
of equipment implementation and plant design makes stock-
ing of spares and planning maintenance activities dicult.
Many OEMs, suppliers and key contractors have equipment
in the eld that will ultimately require maintenance.
Implementing a system capable of tracking manufacturer
type and reliability information simultaneously enabled Air
Liquide to isolate problematic equipment manufacturers. Tis
information can be used in supply-management negotiations
and, more importantly, designing and engineering new plant
construction. By leveraging the Web-based technology and
interface, any reliability center manager, reliability engineer or
maintenance technician can search and sort reliability problems
by manufacturer type, equipment type or installed locations.
We can do this at the plant level, by zone, business class or
summarized for Air Liquide overall. Leveraging this data dur-
ing contract negotiations can save Air Liquide a signicant
amount of time, money and eort.
Linking this interface to data in the CMMS database and
PdM software database also enables better management of
reported problems. Before visiting any Air Liquide facility,
a preventive maintenance work order initiates data collec-
tion in the CMMS. Any recommended repair or follow-up
work generated as a result of the PdM visit is then classied
as a PdM work order type, linked to the PM work order
RELIABILITY
Compressors
Confirming the work
Figure 2. Tracking work order compliance ensures that the
feedback loop is working.
and distinguished with the appropriate
activity type and problem.
Because follow-up work orders are
linked to the original PM visit, reports
reveal if any follow-up work was com-
pleted. Tis enables better tracking
of problems throughout the approval
process, enables reliability engineers
to budget appropriately and empowers
the PdM engineer with the results of
closing the loop.
The bidirectional gateway and
exchange of data between the PdM
software database and the CMMS
database is another example of leverag-
ing technology to substantiate program
success. Tere is now a direct correla-
tion between vibration data and work
completed. Te dynamic link enables
up-to-date status information and the
resultant cause/eect on newly acquired
PdM data.
Analyzing results and setting
goals
Aligning the PdM program vision
statement with monthly tracking met-
rics (KPIs) establishes an overall indica-
tor of the relationship among Rockwell
Automation, Air Liquide and contract
performance. KPIs are reviewed in the
form of a monthly compliance report
submitted to supply management and
reviewed with the senior maintenance
and reliability team. Because KPIs
are reviewed monthly, PdM program
success is tracked and recorded to jus-
tify program savings and ROI data.
An example of the monthly compli-
ance data and KPIs for Air Liquide
include:
Sites visited
Samples taken
PdM saves
Warranty claims
Customer care issues
Program costs and payment informa-
tion
Because each technology is a unique
component of the overall PdM pro-
gram, and work orders can be classi-
ed as such, further detail and tracking
of individual PdM saves can be mea-
sured on a monthly and year-to-date
basis. Figure 1 represents a breakdown
of PdM activity by type for the period
from August 2004 through August
2005 (one year of implementation).
Also, we track and measure work
order compliance to ensure the program
derives actionable repairs from the PdM
program recommendations. Tis vari-
able is a measure of the number of new
problems reported versus work orders
initiated. Tracked by month, Figure 2
shows an example of one reliability zone.
Future goals include tracking and
monitoring turbine efficiency and
machine performance as well as rening
ROI data. One capability currently being
implemented is the addition of pressure,
ow and temperature measurements to
the PdM program vibration routes. Te
PdM software uses these variables to
calculate thermal and mechanical e-
ciencies. Trending can then be used to
predict expander replacements based on
eciency savings.
Te partnership between Rockwell
Automation and Air Liquide pro-
vides benet to both Air Liquide and
its customers. For starters, the nearly
2,000 interventions before equipment
breakdown have avoided countless unit
shutdowns. Tis isnt only a benet to
our customers. It saves Air Liquide
considerable costs by attacking prob-
lems while theyre still relatively small.
However, this is just the start of what
can be done with better information.
One can now look across common
equipment and determine which OEM
provides equipment with the lowest
levels of vibration. Couple this with
work order and reliability data from the
CMMS, and we can provide quantita-
tive information about what equipment
to buy to improve MTBR.
Were getting to the point where,
armed with data and information that
gets to the root of our problems, we can
call in an OEM to discuss equipment
issues. Instead of anecdotal opinions
driving the discussions, the information
from our systems now allow us to focus on
improving long-term reliability.
Any engineer, specialist or expert,
anywhere in the world, with access to our
Internet site, can look at data and infor-
mation on any piece of equipment in the
U.S. and can help us troubleshoot prob-
lems at even the most remote sites. p
Mark E. Lawrence, P.E., CMRP, is director of
maintenance and reliability at AirLiquide
Large Industries U.S. LP, Houston. Contact
him at mark.lawrence@airliquide.com and
(713) 624-8585. George F. Hofer is corporate
program manager at Rockwell Automa-
tion, Houston. Contact him at gfhofer@
ra.rockwell.com and (713) 402-2288.
]aNuavv :oo6 www.PIANTSIRVICIS.cox 44
RELIABILITY
Compressors
Equivalent annual cost Repair or replace
Air system design Keep it simple
Air system upgrades Planning air system upgrades
Valving for compressed air Shutof and special duty valves
Optimum receiver location The compressed air receiver: the endless question
For more, search www.plantservices.com using the keywords database, interface and
vibration
MORE AT WWW.PLANTSERVICES.COM/THIS MONTH
Ivnvuavv :oo6 www.PIANTSIRVICIS.cox 58
CATEGORY
Subcategory
A
lmost every compressed air system uses ex hose to
make the nal connection to production machin-
ery. Proper selection and application of this air hose
and the quick disconnects is critical to achieving optimum
performance.
Compressed air system audits often uncover signif i-
cant opportunity for savings at such locations. Typi-
cally, total system pressure is unnecessarily high to off-
set pressure drops in small-diameter hose and incorrect
quick disconnects.
Te most important sizing data for any process is the
air ow and minimum pressure required at the tool entry.
If you dont know these data, its easy for system analysts
to measure them on-site. In areas where the pressure or
ow are critical to productivity or quality, economical
mass owmeters and pressure gauges can be rigged for
continuous machine monitoring.
Working by hand
Air-driven tools can illustrate the eect of hose and con-
nector selection on productivity and quality. Most air tools
are designed for a hose feed pressure of 90 psig. Te tool
designer really sizes for full ow at about 80 psig for opti-
mum performance. Depending on the tool, pressure signi-
cantly higher than 90 psig may not increase performance,
but lower pressure certainly will reduce it. In many cases,
out-of-range air pressure can damage tools and reduce the
time between rebuilds.
Standard impact tools, screwdrivers, grinders, chippers
and banders prefer a constant 80 psig to 90 psig inlet pres-
sure. Te phrase at rest pressure has no meaning. Table
1, abstracted from selected air tool technical data sheets,
clearly shows the general magnitude of performance loss
at low pressure. At 70 psig, most tools will still oper-
ate, but below rating. At 60 psig, performance is seriously
degraded and probably will be unacceptable. Operating
below 60 psig isnt really a viable option. However, unless
specically stated, no tool is designed for inlet pressure
greater than 100 psig. Table 2 shows the approximate
performance losses at various inlet pressures in 1-hp to
3-hp vane motor grinders and sanders. Te power drops
Hosing
down
your losses
How to specify cost-effective
air hoses and connectors
By Hank van Ormer
Never select air hose unless
you know the air ow and
hose length the tool requires.
Ivnvuavv :oo6 www.PIANTSIRVICIS.cox 59
may preclude eective job performance. Along with the
loss in power, which is most important, theres also a loss
in speed. Both factors aect productivity.
Beware of 38-in. hose
Never select air hose unless you know the air ow and
hose length the tool requires. Te most common hose siz-
es for plant use range from
3
/8 in. to
3
/4 in. and handle 300
psig. Hose choice is often left to the operator, who usually
wants
3
/8-in. hose, regardless of application, because:

3
/8-in. hose appears to be the lightest and easiest to handle.
A 50-ft. length of
3
/8-in. or
1
/2-in. hose weighs about
13 lbs., depending on grade but a 50-ft. length of
3
/4-in.
hose weighs 22 lbs.
Te operator might not be trained regarding the hose
size required to run the tool.
A
3
/8-in. hose isnt a viable supply hose for industrial
tools. Te smallest size you should use is
1
/2 in. Table 3
refers to premium black industrial air hose. Te data leads
us to specic conclusions:

1
/2-in. hose in 50 ft. lengths is suitable only for 1 hp
or smaller tools (approximately 30 cfm/hp).

3
/4-in. hose is acceptable for 2 hp to 3 hp (60 scfm to
90 scfm), depending on the length of run.
For runs greater than 50 ft., use larger hose or pipe,
supported on the walls or ground as required, to elimi-
nate pressure drop.
For more comfort and easier operation, adding an
8-ft. to 10-ft. whip hose to the larger
3
/4-in. or 1-in. main
line will have minimal eect on performance, but still
gives the operator the feel of a lighter hose.
Dont use any more hose than necessary. Coiling the
extra just adds pressure drop. Cut the hose to the proper
length and install ttings.
Dont forget about OSHA safety requirements. Going
from a
3
/8-in. to
1
/2-in. hose still allows personnel to handle
Table 1. Performance data for air-operated tools
Typical vane air motor performance at various inlet pressures (actual results will vary by manufacturer and model)
Inlet air
pressure (psig)
hp hp 1 hp 1 hp 2 hp 3 hp
60
rpm at max
load
8,500 5,809 3,810 5,550 3,730 3,900
Max hp 0.35 0.47 0.765 0.927 1.74 2.32
scfm at max hp 20 20.1 27.5 30 51 67
Max torque
ft-lb
0.36 0.88 1.67 1.67 3.7 5.0
70
rpm at max
load
9,000 6,184 4,060 5,900 3,975 4,160
Max hp 0.41 0.58 0.95 1.15 2.16 2.88
scfm at max hp 21 53 32 35 60 78
Max torque
ft-lb
0.42 1.0 1.95 1.95 4.3 5.8
80
rpm at max
load
9,500 6,429 4,250 6,190 4,160 4,350
Max hp 0.5 0.69 1.13 1.38 2.58 3.44
scfm at max hp 22 27 36 40 68 89
Max torque
ft-lb
0.5 1.2 2.2 2.2 4.9 6.7
90
rpm at max
load
10,000 6,700 4,400 6,400 4,300 4,500
Max hp 0.6 0.8 1.4 1.5 3.0 4.0
scfm at max hp 24 30 39 42 76 100
Max torque
ft-lb
0.55 1.3 2.5 2.5 5.5 7.5
100
rpm at max
load
10,500 6,888 4,520 6,580 4,415 4,630
Max hp 0.6 0.9 1.5 1.8 3.4 4.6
scfm at max hp 26 33 45 50 85 111
Max torque
ft-lb
0.6 1.4 2.8 2.8 6.1 8.3
PERFORMANCE
Compressors
Ivnvuavv :oo6 www.PIANTSIRVICIS.cox 60
smaller hose without the mandatory automatic air shuto
valve or safety velocity fuse. Tese fuses are an excellent
safety device when applied correctly. Refer to U.S. Depart-
ment of Labor, Occupational Safety and Health Adminis-
tration Power Operated Tools 1926.302, page 2, para-
graph 1926.302(b)(7), which mandates a safety velocity fuse
on all hoses larger than
1
/2 in. inside diameter.
A real-world example
More often than not, a process requires some minimum
pressure. Trace these so-called requirements to their ori-
gin to determine if they are actual OEM specications or
simply an operators perception.
A recent client was running the plant headers at 100
psig to 110 psig because a critical hand-tool grinding
process was believed to require 98 psig to run correctly.
Terefore, they reasoned, the system should run at 98
psig or more.
When you hear things like this, dig for more informa-
tion. If the system header pressure falls below 98 psig,
the grinders probably dont work well. Production per-
sonnel probably dont know the actual pressure at the tool
or how much air the tool uses. Te rest of the plant could
have run at 80 psig, but it operated at 98 psig because the
grinding area supposedly required it. Grinding account-
ed for only 20% of the demand, so 80% of the plant was
supplied with air at a much higher pressure than needed.
We didnt calculate how much the higher pressure was
costing, but intuition says it amounts to thousands of
dollars a year.
Testing with a needle gauge at full operation revealed
that the actual inlet pressure to the tool was 63 psig at
load, but the header pressure stayed at 98 psig. In other
words, there was a 35-psig pressure drop between the
header and each grinder. Further testing revealed that the
grinders only needed 75 psig for optimum performance.
Te operators argued that they found the recommend-
ed
3
/4-in. hose to be too heavy, so they used
3
/8-in. hose
instead. Te smaller hose restricted the air ow, which
produced a substantial pressure drop. Furthermore, the
3
/8-in. hose used standard quick disconnects, which add-
ed their own 23-psi pressure drop.
We changed the standard
3
/8-in. quick disconnects to in-
dustrial quick disconnects costing only $2.50 per pair a
whopping $5 per station to reduce the pressure drop to
5 psig. Ten, we replaced the
3
/8-in. hose with 1-in. pipe
routed to the base of the work stations at a cost of $30 each.
Next, we installed a regulator that delivered full ow to the
grinders at 75 psig with 80-psig feed pressure. Finally, we
reduced the header pressure to 85 psig. About 18 months
later, grinder repair costs had decreased and production
throughput increased by 30% with the addition of more
equipment. Te cost of materials to implement these chang-
es was $1,362 for nine grinders. Even with the production
increase and new equipment, the average total air demand
fell from 1,600 to 1,400 cfm.
Te key to this success was monitoring the workstation
inlet pressure while simultaneously monitoring header
pressure. If the header pressure stays steady, and the pro-
cess inlet pressure falls, then the restriction is in the feed
line from the header to the process.
Break down the connection
Tis case study demonstrates that small hose represented
only 12 psid while the quick disconnects represented 23
psid. Often, but not always, a quick disconnect is the best
answer for overall productivity. But, size the quick dis-
connect for the maximum expected ow and the allow-
able pressure loss. Read the manufacturers performance
data sheet.
Never select by connection size select by accept-
able performance at specied ow and entry pressure.
If you want to use the same quick disconnect every-
where for exibility, do it. But, size them for the single
largest ow demand at the lowest expected pressure.
Table 2. Of-design performance
Design pressure Actual pressure Performance loss
100 psig 90 psig 7% to 17%
90 psig 80 psig 7% to 16%
80 psig 70 psig 17%
70 psig 60 psig 20%
100 psig 60 psig 50%
90 psig 60 psig 39%
80 psig 60 psig 33%
Table 3. Pressure drops at flow rates
Tool size (hp) Flow (cfm)
Pressure drop per 50 ft. (psi)
in. hose in. hose
1 30 2.4 0.4
2 60 14.9 2.2
3 90 41 4.6
(Assumes 90 psig supply, does not include fttings)
If the header pressure stays steady,
and the process inlet pressure falls,
then the restriction is in the feed
line from the header to the process.
PERFORMANCE
Compressors
Ivnvuavv :oo6 www.PIANTSIRVICIS.cox 61
Remember that each feed has at least two quick dis-
connects.
Use quick disconnects that shut o the ow when
disconnected to eliminate potential hose whipping.
Consider ISO 4414 exhaust-type quick disconnects that
bleed o the air trapped inside the connection to eliminate
blasting compressed air onto the operator at disconnect. Its
easier to uncouple a depressurized tting.
Quick disconnects should have proper safety latches,
wires and keepers or be of a design that wont open when
dragged over the ground, oor or machinery.
Seek tested performance curves
Dont assume that because couplers appear similar the per-
formance is similar. Review the performance curves or, even
better, measure the pressure loss at specic ows. On a recent
audit to help select the proper disconnect for a major tool op-
eration, we tested the pressure drop on two specic types of
quick disconnect a lock-ring coupler with a ball-check
nipple versus an exhaust-type coupler with a standard nipple.
Both had 1
1
/4 in. diameter coupler bodies and ports sizes of
3
/8
in.,
1
/2 in. and
3
/4 in. Te
3
/8-in. nipple on the lock-ring type
coupler didnt have a ball check to shut o the air. Te
1
/2-in.
and
3
/4-in. units did. Te exhaust-type couplers had the full
shuto and exhaust to allow disconnect at zero pressure.
Figure 1 shows the three sets of performance curves that
reect the measured pressured drop of each quick disconnect
at various ows and inlet pressures. Te results will probably
vary by manufacturer. Te key is to optimize performance by
investigating.
We found a signicant pressure drop dierence between the
1
/2-in. quick disconnects. Te exhaust coupler could work in an
acceptable manner from less than 30 cfm to as much as 60 cfm
and still maintain 100 psig inlet with 80 psig to the tool or 90
psig inlet with 70 psig to the tool.
Te
1
/2-in. lock ring/ball check nipple quick disconnect
appears acceptable at 30 cfm but probably wont be accept-
able at 60 cfm.
Te
3
/4-in. quick disconnects are closer in perfor-
mance, but the lock ring/ball check type introduces 30%
to 40% more pressure drop.
Te
3
/8-in. lock ring/ball check nipple quick disconnect
tested didnt have the ball check valve in the nipple, which
accounts for its lower pressure drop compared to the
1
/2-in.
lock ring coupler, which did. Tis, of course, means that the
safety feature to control potential hose whip isnt incorpo-
rated into the
3
/8-in. lock ring set.
Tis test data isnt intended to recommend one disconnect
over another. For the particular application investigated,
with many grinders and impact tools using between 60 scfm
and 90 scfm, the exhaust-type quick disconnect exhibited
the best overall performance and economics. On a dier-
ent application, testing may well dictate another choice.
Te important point is to select quick disconnects, hose and
pipe with diligence and attention to detail. Although dis-
connects are a relatively inexpensive piece of equipment, if
misapplied, they can be costly. p
Hank van Ormer is owner of AirPower USA, Pickerington, Ohio. Con-
tact him at HankvanOrmer@aol.com and (740) 862-4112.
Lock-ring type with ball-check nipple
Push the lock-ring coupler to connect. Turn the lock ring about
20 to disconnect. This feature prevents accidental discon-
nects.
Nipple with ball check seals the air in the hose or tool con-
nected to the nipple to eliminate blowback and possible un-
controlled hose whip.
The disconnect will be made under some pressure with variable
flow dependent on the installation.
Flow check-type nipples are more expensive than a standard
industrial interchange nipple, which will work in many manu-
facturers couplers.
Exhaust-type
These quick disconnects use a common standard industrial in-
terchange nipple. When comparing cost, its important to con-
sider that in many operations, there are usually three or four
nipples for every coupling.
Exhaust-type couplings are push-to-connect, exhaust-style
action with a self-locking sleeve to guard against accidental
disconnection.
To connect, push the nipple into the coupler. The locking
sleeve slides forward automatically to lock the nipple in place.
No air flows through the coupling at this point. Rotate the valve
sleeve to open flow and engage the sleeve-lock mechanism.
To disconnect, rotate the valve sleeve in the other direction to shut
of the air fow and vent downstream air to atmosphere. The lock-
ing sleeve can then be retracted and the nipple removed.
The valve sleeve acts as an integral shutoff valve that allows
connect and disconnect at zero pressure. The valve sleeve is
operated independently of the locking sleeve. When the sleeve
is moved to stop air flow, it automatically vents downstream
pressure so disconnect can be performed at zero pressure.
Exhaust couplers eliminate the need for flow-check nipples
and still meet safety issues by connecting and disconnecting
at zero pressure.
HOW QUICK DISCONNECTS WORK
Sizing and selecting FRLs Operation peak performance
Baselining Baselining the compressed air system
Proper line sizes The secret is in the pipe
Equivalent annual cost Repair or replace
For more, search www.plantservices.com using the keywords
compressor, performance or pressure.
MORE AT WWW.PLANTSERVICES.COM/
THISMONTH
PERFORMANCE
Compressors
F
l
o
w

(
S
C
F
M
)
Pressure drop (PSID)
190
180
170
160
150
140
130
120
110
100
90
80
70
60
50
40
30
20
10
0
0 5 10 15 20 25
*
*
*
*
*
*
*
*
*
*
" Parker EZ 12HB
@90 cfm 2.5 psid
@65 cfm 1 psid
@30 cfm Neg
" Hansen B/C
" Hansen 12 HB
@90 cfm 3.9 psid
@60 cfm 2.5 psid
@30 cfm 1 psid
" Parker EZ 8HB
@90 cfm 11 psid
@60 cfm 5 psid
@30 cfm <2 psid
" Hansen 8HB
@90 cfm 22.5 psid
@60 cfm 10 psid
@30 cfm 3 psid
F
l
o
w

(
S
C
F
M
)
Pressure drop (PSID)
Pressure drop (PSID)
190
180
170
160
150
140
130
120
110
100
90
80
70
60
50
40
30
20
10
0
0 5 10 15 20 25
*
*
*
*
*
*
*
*
*
*
F
l
o
w

(
S
C
F
M
)
190
180
170
160
150
140
130
120
110
100
90
80
70
60
50
40
30
20
10
0
0 5 10 15 20 25
*
*
*
*
*
*
*
*
*
*
" Parker EZ 12HB
@90 cfm 3.4 psid
@60 cfm 1.4 psid
@30 cfm <1 psid
" Hansen 12 HB
@90 cfm 5.4 psid
@60 cfm 3 psid
@30 cfm 1.5 psid
" Parker EZ 8HB
@90 cfm 14.5 psid
@60 cfm 6.2 psid
@30 cfm 2.8 psid
" Hansen 8HB
@90 cfm >30 psid
@60 cfm 11 psid
@30 cfm 4 psid
" Hansen HB
" Parker EZ 12HB
@90 cfm 4 psid
@60 cfm 2 psid
@30 cfm <1 psid
" Hansen 12 HB
@90 cfm 5.5 psid
@60 cfm 3.2 psid
@30 cfm 1.5 psid
" Parker EZ 8HB
@90 cfm 16 psid
@60 cfm 6.5 psid
@30 cfm 2.8 psid
" Hansen 8HB
@90 cfm >30 psid
@60 cfm 12 psid
@30 cfm 4.4 psid
" Hansen 6HB
Figure 1. These measured performance characteristics of several quick disconnects show
fow as a function of pressure drop.
Ivnvuavv :oo6 www.PIANTSIRVICIS.cox 62
PERFORMANCE
Compressors
Measure before you act
R
egenerative desiccant dryers are used in com-
pressed air systems that require dew points to
be below the minimum that refrigerated dryers
can produce (generally 40F). Tree types of regenera-
tive desiccant dryers are widely used throughout industry:
heatless, heated and blower purge.
Te following discussion doesnt address heat-of-com-
pression (HOC) desiccant dryers, even though they require
the least amount of energy to operate. Te use of HOC dry-
ers is limited to lubricant-free compressors.
Many plants require air quality that only regenerative
desiccant dryers can produce. Unfortunately, in too many
cases, the decision about which type of regenerative dryer to
purchase is based on initial capital cost alone. Tis decision
basis ignores the cost of energy that will be required to oper-
ate the dryer. Including energy cost can alter the economics
of a purchase decision dramatically.
Regenerative desiccant dryers use a desiccant medium to re-
move moisture from the compressed air stream. Wet air passes
directly through the desiccant medium, which then adsorbs
the moisture. Te desiccant medium has a nite capacity for
adsorbing moisture before it must be dried out, or regenerated.
To do this, the tower containing saturated desiccant medium
is depressurized and the accumulated water is driven o using
purge air, heat or a combination of both. Desiccant dryers are
generally of a twin-tower construction, with each tower con-
taining its own desiccant bed. Tis allows one bed to dry com-
pressed air as the other undergoes regeneration.
Te energy cost associated with operating a desiccant dryer
depends primarily on how the desiccant is regenerated us-
ing purge air, heat or a combination of both. Desiccant dry-
ers are categorized by their method of regeneration, the three
primary types of which are:
Heatless, which uses only compressed air as a purge.
Heated, which uses both heat and compressed air to purge
moisture.
Blower purge, which uses air from an external blower,
heat and minimal compressed air.
Te energy cost to operate each of these dryers depends
on the amount of purge air used, the heater size and the
blower motor size.
Heatless desiccant dryer
Heatless desiccant dryers use treated, unheated compressed
air to regenerate the desiccant bed. Tis dryer is the most ex-
pensive to operate because of the large amount of compressed
air thats consumed for purging during the regeneration cycle.
Typically, 15% of the rated ow capacity of the desiccant dryer
is consumed as purge air. Tis is a signicant amount of expen-
.PLANTSERVICES.
The energy cost associated with
operating a desiccant dryer depends
primarily on the means by which the
desiccant is regenerated.
h
u
n
g

o
u
t


t
o

d
r
y
h
u
n
g

o
u
t


t
o

d
r
y
D
O
N

T

G
E
T
D
O
N

T

G
E
T
Insight into
the economics of
operating regenerative
desiccant dryers
By Noel Corral and Andrew Sheaffer
A 2006 55
.PLANTSERVICES.
sive compressed air to be used in this manner. In many facilities
that operate heatless units, the air dryers are the largest single
user of compressed air at the facility. Although the cost to oper-
ate this type of dryer is high, the heatless type is often selected
because it has the lowest initial cost of any desiccant dryer.
Heated desiccant dryer
Generally, heated desiccant dryers have the lowest energy
operating cost of the three types. A heating element supple-
ments the drying action of the purge air. Te heater can be
mounted internally to heat the desiccant bed directly or ex-
ternally to heat the purge air thats blown through the bed.
Internal heating elements shorten the life cycle of desic-
cant beds and may cause dew point spikes. Terefore, they
are not recommended and are not covered here. Like heat-
less dryers, the purge is compressed air supplied from the
system. However, the amount of air required is reduced be-
cause the heat supplements the airs drying action.
Te amount of purge air an externally heated dryer requires
varies by manufacturer, but is typically about 7% of the dryers
rated ow capacity. Te combination of heat and purge air is more
energy ecient than purge air alone, yielding lower energy costs.
Blower purge desiccant dryer
Like heated-type desiccant dryers, blower purge-type dryers
supplement the drying action of purge air with heat. In a blow-
er purge desiccant dryer, however, the purge air is ambient air
supplied by an electric blower. Te typical blower supplies the
entire purge air load, which is typically 20% of the dryers rated
ow. Depending on the dryer design, blower purge-type dryers
may still consume a small amount of compressed air that cools
the bed at the end of the regeneration cycle. Blower purge-type
dryers are generally more expensive to operate than the heated
type, but less expensive to operate than the heatless variety.
Determining your own energy costs
So, we know that heatless desiccant dyers are the most ex-
pensive to operate in terms of energy costs, heated dryers are
the least expensive, and blower purge dryers are somewhere
in between. But, how can plant management quantify these
costs so they can be used to make an informed decision when
purchasing a desiccant dryer? Table 1 shows the appropriate
equations you can use to calculate the energy cost to operate
each type of desiccant dryer.
Tese equations can be used with dryer manufacturer
specications to determine the energy cost to operate the
unit. Te equations, with generalized manufacturer dryer
specications, were used to develop Figure 1, which sum-
marizes the energy cost associated with the operation of
heatless, heated and blower purge desiccant dryers.
Te heatless desiccant dryer is the most expensive type
to operate at any capacity, while the heated type is the least
expensive to operate at any capacity. Te magnitude of dif-
ference between the heated and blower purge type is fairly
PERFORMANCE
Compressors
Heatless
EC =
RDC x 15% x 0.746 x t x $/kWh
CGE
Heated
HC =
(375-T ) x RDC x 7%
3,160
EC =
RDC x 7% x 0.746
CGE
+ (0.75 x HC)
) (
x t x $/kWh
Blower
Purge
EC = + (0.75 x HC)
HP
b
x 0.746
eta
b
) (
x t x $/kWh
HC =
(375-T ) x RDC x 20%
3,160
EC = Annual energy
cost to operate ($)
HC = Electrical
capacity of heater
(kW)
RDC = Rated dryer
capacity (cfm)
t = Annual operating
time (hrs)
$/kWh = Utility rate
($/kWh)
CGE = Compressor
generation
efciency (cfm/hp)
T = Temperature of
regeneration air (F)
HP
b
= Blower
horsepower (hp)
eta
b
= Blower motor
efciency (%)
EC = Annual energy cost to operate ($)
HC = Electrical capacity of heater (kW)
RDC = Rated dryer capacity (cfm)
Heatless
EC =
RDC x 15% x 0.746 x t x $/kWh
CGE
Heated
HC =
(375-T ) x RDC x 7%
3,160
EC =
RDC x 7% x 0.746
CGE
+ (0.75 x HC)
) (
x t x $/kWh
EC = Annual energy
cost to operate ($)
HC = Electrical
capacity of heater
(kW)
RDC = Rated dryer
capacity (cfm)
t = Annual operating
time (hrs)
$/kWh = Utility rate
($/kWh)
CGE = Compressor
generation
efciency (cfm/hp)
T = Temperature of
regeneration air (F)
HP
b
= Blower
horsepower (hp)
eta
b
= Blower motor
efciency (%)

Blower purge
HC =
(375-T ) x RDC x 20%
3,160
EC = + (0.75 x HC)
HP
B
x 0.746
) (
x t x $/kWh
Table 1. Use these formulae to calculate your own operating costs.
EC = Annual energy cost
to operate ($)
HC = Electrical capacity
of heater (kW)
RDC = Rated dryer
capacity (cfm)
t = Annual operating
time (hrs)
$/kWh = Utility rate ($/kWh)
CGE = Compressor
generation efciency
(cfm/hp)
T = Temperature of
regeneration air (F)
HP
B
= Blower
horsepower (hp)

B
= Blower motor
efciency (%)
0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
50,000
55,000
60,000
5,000 4,000 3,000 2,000 1,000 500
Tally the cost
Figure 1. Annual operating cost of desiccant dryers. Data
assumes 8,760 operating hours per year, $0.05/kWh utility
rate, compressor generation efciency of 4.5 cfm/hp and
compressed air and ambient air temperature of 100F.
A 2006 56

B
small, indicating that users should base any decisions be-
tween these two types on other factors (initial cost, mainte-
nance costs, excess capacity for purge air, etc.)
What about controls?
Manufacturers of desiccant dryers provide optional con-
trols that can reduce the dryers operating cost. Te most
widely used control is the dew point demand controller. It
uses a probe that continuously monitors the compressed air
leaving the drying tower. Its congured to delay regenera-
tion until the exit air reaches the users maximum allow-
able dew point. Te controller bases regeneration on dryer
performance instead of a timed cycle, thereby reducing
the number of regeneration cycles. Controls such as these
can yield 30% to 80% in energy savings, depending on the
maximum allowable dew point. Savings achievable using
these controls must be calculated on a case-by-case basis.
Controls reduce cycling frequency, resulting in lower over-
all maintenance and longer desiccant and valve life. Although
the cost of controls increase the dryers purchase price, the
benets they provide can pay back the initial cost quickly.
Desiccant air dryers are a proven and eective technology
that can provide extremely dry air. However, they can be ex-
pensive to operate in terms of energy costs, and every plant
professional who needs a desiccant dryer should be aware of
these costs before going out for bids. Using the equations
laid out here, you can determine what your energy costs will
be and make informed decisions about what type of desic-
cant dryer is best for your application. p
Andrew Sheafer and Noel Corral are research engineers at the En-
ergy Resources Center at the University of Illinois at Chicago. Con-
tact Corral at ncorra2@uic.edu and (312) 413-8546. Contact Sheafer
at sheafer@uic.edu and (312) 413-3615.
Figures: Energy Resources Center
.PLANTSERVICES.
PERFORMANCE
Compressors
Air system design Keep it simple
Dew point monitors Fog in the pipeline
System engineering Engineered air
Treatment and storage Compressed air quality
HOC calcs What does Mother Nature say about cooling
hot air?
For more, search www.plantservices.com using the keywords
dryer, CAC and desiccant.
More resources at www.PlantServices.com/this month
A 2006 57
M 2006 .PLANTSERVICES. 32
C
ompressed air isnt free. Youd
think that would be obvious to
maintenance and operations personnel in
plants where some of the largest motors are
harnessed to compressors, howling and yammer-
ing away 24/7, sucking up, in many cases, the larg-
est percentage of an increasingly outrageous utility
bill. Wasting upwards of 30% of that energy should
probably be illegal, but its not, and apparently most folks
are oblivious.
If water or oil are leaking, you know it, says Len Bishop,
manager, Draw Professional Services (www.drawproser-
vices.com). If gas is leaking, you smell it. You can see steam.
But compressed air? Its not a safety issue until someone
complains that its too loud. So unless its knocking your
hat o, you ignore it.
Statistics from the Compressed Air Challenge (CAC)
and DOE are conrmed by the compressed air system
experts: Te average facility has 30% to 35% leakage if
it hasnt taken any recent action. And a survey by the
Oce of Industrial Technologies says
57% of facilities have taken no ac-
tion during the past two years.
Air is a clear uid that doesnt
make a mess, says David Booth,
system specialist, Sullair (www.
sullair.com). It can leak forever
and not directly aect anything
but cost. So, unfortunately, leaks are a big component that no
one is willing to x permanently.
Even a /-in. diameter leak can cost big bucks (Figure 1).
But you neednt get out a caliper. If you can hear it without
an ultrasonic leak detector, its at least 8 cfm to 10 cfm at
$300/year per cfm, says Bishop. It may not be a safety issue,
but its costing you money.
Te excess cost goes beyond wasted energy. Leaks lead to
other plant problems:
Fluctuating system pressure: Inconsistent or faulty performance
of air tools and other air-operated and powered equipment.
Excess compressor capacity: higher than necessary equip-
ment and maintenance costs. Not only are you wasting en-
ergy, youre likely to be mismatching the compressor, says
John Bartos, vice president of engineering and new product
development, Cooper Compression (www.coopercameron.
com). You buy based on calculated demand, and leaks can
really aect the actual performance.
Excess load on supply equipment: Increased maintenance
costs, decreased service life.
Twarting other system eciency eorts: Its impossible to
optimize system pressure and compressor control schemes
with excessive leaks. In cases where the total leak load ex-
ceeds 10%, the articial demand created by leaks must be
addressed in order to obtain an accurate air demand prole
for the plant, says Wayne Perry, technical director, Kaeser
Compressors (www.kaeser.com). Only then can other rec-
ommendations and improvements take place.
M 2006 .PLANTSERVICES. 33

By Paul Studebaker, CMRP, Editor in Chief



M 2006 .PLANTSERVICES. 34
Wet air: While letting air out, leaks let moisture in.
Valves on drain legs are left cracked open because too
much water is coming into the equipment, says Scott
Stroup, president, Airometrix Manufacturing (www.ai-
rometrix.com). But unless the air dryer is malfunction-
ing, the air is leaving it at a -40 dewpoint. Wheres the
water coming from? It comes in at the air leaks.
Its counterintuitive, but Ficks Law explains why water
migrates from high-humidity outside air into dry com-
pressed air even though air is coming out the leak. Its hard
and for some people, impossible to believe, but its true,
Stroup says. Is your water problem bigger when the hu-
midity is high or its raining? Tats why.
So with all these potential benets, why is it that, as Jan
Zuercher, director, air systems, Quincy Compressor (www.
quincycompressor.com), says, Most people do very little leak
management a majority do it infrequently or not at all.
Some plants are ignorant, more feel they cant devote time
and eort to leak management, and plenty see it as pointless:
xing some leaks makes the others leak more, and youre back
to zero (see sidebar, Exercise in futility?). But understand-
ing where most leaks occur, ecient ways to detect them, and
their eects on system pressure and performance can help you
implement an energy-saving leak management program thats
simple, ecient, rewarding and, dare we say, almost fun.
Modi operandi
Before you start shopping for leak detection equipment or
get bogged down seeking management support, its help-
ful to know where many of the largest leaks are likely to
be found (Table 1). Drawing on his extensive experience,
Stroup provides this ranking:
1. Hoses and hose ttings: On hoses, the tting-to-hose
connections are most likely due to improper clamping or
working loose. Next comes the hose itself, cut or gashed,
Stroup says, then the quick-disconnects.
2. Pressure regulators and lter/regulator/lubricators
(FRLs): Te stems on the lter bowls leak or are left open, O-
rings leak at the bowl-to-housing connection, lter bowls are
cracked, and on the regulators themselves, O-rings, gaskets,
and the piping connections between elements often leak.
3. Plastic tubing: Tere are so many of them, the sheer
number adds up, Stroup says. Te leaks are mainly at
push-to-connect ttings.
Zuercher agrees that tubing connectors are a common cause of
leaks, but says, Teyre typically too small to be worth repairing.
Fix the largest leaks rst to get the biggest bang for the buck.
4. Header and distribution piping: Welded steel and sol-
dered copper pipe are generally pretty tight, Stroup says.
Treaded pipe is more likely to leak.
5. Leaks within equipment: In some plants, these are a
major concern. A small hole there can go undetected and
cause a pressure drop that leads to an equipment malfunc-
tion, says Stroup. People end up replacing a cylinder or so-
Figure 1: At 24/7 operation and $.08/kWh, even a small leak with
an equivalent diameter of / in. steals more than $700 per year.
Grand larceny
$13,867
$11,427
$31,200
$25,733
$55,333
$45,733
$720
$865
80 psig
100 psig
$60,100
$50,100
$40,100
$30,100
$20,100
$10,100
$100
1/2
3/8
1/4
1/16
Leak size, inches
Cost
Couplings
Hoses
Tubing
Fittings
Pipe joints
Qucts
Filter/regulator/lubricator
(FRL) units
Condensate traps
Valves
Flanges
Packings
Thread sealants
Point-of-use devices
Open condensate traps
Open shut-of valves
Table 1: Common culprits
MANAGEMENT
Compressed Air Systems
0
100
200
300
400
500
600
700
800
900
1000
0
100
200
300
400
500
600
700
800
900
1000
5 10 15 20 25 30 35 40
Compressed air leakage (cfm)
Rolling average
Figure 2: Even solid progress in reducing leaks over time will be
uneven, as shown by this chart from a wood-products plant.
Tough cookies
Weeks
M 2006 .PLANTSERVICES. 35
Everybody knows they have leaks and most have a lot of them,
says Jan Zuercher, director, air systems, Quincy Compressor
(www.quincycompressor.com). Its a bit of a vicious circle. As
you fx leaks, pressure rises and the remaining leaks leak more.
A lot of plants reach the point where they say, Whats the use?
Most maintenance stafs are running very lean. Theyre focused
on keeping the plant running. Who has the time?
Its true that most calculations of potential savings from leak
repairs are optimistic, and for two reasons:
1. Estimates assume that the system is 100% efcient: saved
cfm are converted directly to less horsepower consumed in the
compressor room. In reality, this is never the case, and its usu-
ally not even close, says Robert Horneman, marketing manag-
er, Industrial Air Solutions, Ingersoll Rand Industrial Technolo-
gies (www.irco.com).
For example, reducing the air load on a modulating compres-
sor will cause the inlet valve to close more, drawing a vacuum
under the valve. A modulating machine at 70% load will still
pull 93% of its full-load horsepower, Horneman says. Similarly
a centrifugal compressor can only back down so far depending
on its surge points before it starts to blow of. If its close to its
surge point and you fx leaks, it will simply blow of air and you
have gained nothing. You need a good handle on the air system
as a whole and how each part afects it, especially how the sys-
tem reacts to part loading.
2. Estimates assume that fxing a leak has no efect on the rest
of the system. In practice, fxing a leak often causes pressure
to rise in the vicinity of the repair. This may not be all bad -- ac-
cording to Scott Stroup, president, Airometrix Manufacturing
(www.airometrix.com), Most of our customers would do back-
fips for fve more psi.
But higher pressure means an exponential increase in fow
through unregulated uses, including leaks. This will very
quickly return the leak load to original levels, Horneman says.
The solution is to tightly control the demand pressure so the
decrease in leak load will not increase the local pressure, and to
control the compressor output so the decreased demand will
be refected in online horsepower. Pressure can be controlled
with an expander valve between supply and demand. This al-
lows you to fx leaks without afecting the rest of the system by
holding pressure within two psi, says Horneman. This valve
can react to dynamic demand changes quicker than the com-
pressor controls, allowing the system to respond to events with
stored compressed air rather than compressor horsepower.
I am not advocating throwing a piece of machinery into
any system for a resolution, Horneman says. Compressed air
systems are unique and components must be understood in
their relation to the whole before an efective solution can be
identifed. A detailed analysis of the system is essential to avoid
improper allocation of capital.
EXERCISE IN FUTILITY?
MANAGEMENT
Compressed Air Systems
lenoid valve trying to x an intermittent equipment prob-
lem thats actually caused by an undetected leak.
For example, a packaging machine wouldnt work proper-
ly when other equipment in the area was running. Mechan-
ics replaced a solenoid valve with no eect. When we got
next to it, we could hear a leak, Stroup says. We crawled
under it and found where a chain had worn a hole in an air
line. Fixing that xed the problem with the machine.
Pneumatic systems inside panels often share a common,
ducted exhaust, masking the presence of leaks in the normal
airow. An internal failure of a solenoid valve can blow a
lot of undetected air, whether the machines running or not.
We also often nd missing faceplates and uncapped lines
in solenoid cabinets, Stroup adds.
Dont ignore useful leaks. For example, compressed air
used to position paper in an envelope machine should be
turned o when the machines not running. Add a shut-o
valve interlocked with the on-o switch on the machine. A
manual valve isnt reliable because operators dont want to
worry about it and will simply leave it on.
Get on the case
Detection methods range in sophistication from listening
and running hands over pipes and equipment to soap bubbles
and, todays big gun, the ultrasonic leak detector (ULD). All
these tools have their place, but the most eective and least
familiar is the ULD.
Ultrasonic is a critical element, says Zuercher. Te old-
school thinking was that you had to detect leaks on nights
and weekends when the plant is quiet so you can hear the
leaks, and most people dont want to do that. With ultra-
sound, you can detect leaks during normal production.
Handheld ULDs locate leaks both in compressed air
and vacuum systems, and search areas such as valve seats,
drain traps, tanks and piping. USDs can detect internal
leaks and leaks that are near the ceiling, outdoors or in
noisy environments. For example, a blown diaphragm on
a valve in a baghouse can leak a lot of air where it is easily
overlooked, Zuercher says. Good ultrasonic can detect
leaks from over 50 feet away.
A good ULD can nd leaks quickly, but acionados dis-
agree on whether they can be used to judge leak rate. Zu-
ercher says the size of a leak is proportional to the volume of
the sound given o by the ultrasonic leak detector, but Ben
Fried, product development and support, CTRL Systems
(www.ctrlsys.com), says, Te number-one misconception
about ULDs is that they can measure the intensity of a leak
and correlate it to dollars lost. Tis is 100% false.
Many factors change the intensity of a leak, includ-
ing pressure; temperature; humidity; distance from the
receiver; size, shape and location of the hole; proximity
of hole to other objects; battery power of receiver; etc.,
Fried says. There is no way to determine amount of cfm
loss based on a decibel reading or some other relative
measurement.
In general, ULD capabilities vary, and experts have strong
words about making sure the one you buy will do what you
expect (see sidebar, ULD: Pick from a lineup).
M 2006 .PLANTSERVICES. 36
Experts have observed a number of efective practices that can
help you enlist operations and more easily implement an efec-
tive leak management program:
Be ready to fx as you go: Carry a wrench when you look for
leaks. A lot of them can be repaired on the spot.
Tag leaks with a three-part tag: Bring in one stub to make
a work order. Bring in the second stub when the leak is fxed.
Have the repair checked, then bring in the rest of the tag.
Leaks can be hard to fx or come back right away, says Da-
vid Booth, system specialist, Sullair (www.sullair.com). Use a
three-part tag fnd, fx, verify and have someone check
the repair and bring in the tag before you say its done.
Enlist management: One guy takes a camera out in the
plant, gets pictures, brings them into management meetings
and says, This is a $500 problem, says Booth. Hes taken
pictures of leaks with two and three leak ID tags that havent
been repaired. Until he did this, no one took him seriously.
Educate operators and get their buy-in: Otherwise, peo-
ple will stand at a machine for years with air blowing on their
knees, Stroup says. Theyre already wearing earplugs so
they arent even bothered by the noise. But if you approach
their machine with an ultrasonic instrument and tell them
what youre doing, theyll point out the leaks for you.
Like safety, everyone can be aware of the cost of air leaks
and be on the lookout. The day after you do a survey, you
can get a large leak, says Jan Zuercher, director, air systems,
Quincy Compressor (www.quincycompressor.com). If the
employees care and report those leaks, it can be a good part
of the program. Management can provide incentives.
One plant decided to share the proft. They pay opera-
tors $5 apiece for identifying leaks, says Booth. It pays of.
Operators can be the worlds largest detection force.
See which equipment holds up: Depending on operating
conditions, there can be a lot of variation in diferent types
and brands of regulators, FRLs, quick-disconnects, etc. Toss
the ones you replace in a bucket, fll it up and look at them.
Quit buying stuf that doesnt survive. One client picked the
best quick-disconnect and replaced 10,000 all at the same
time that is an automotive plant with 3,200 employees,
all using hand tools, says Scott Stroup, president, Airome-
trix Manufacturing (www.airometrix.com). They did it three
years ago and it paid of.
Use a trade-in system: One plant uses a two-barrel trade-
in system near the machines on the plant foor to replace
leaking hose assemblies. Operators can quickly and easily
eliminate a leak by tossing a leaky assembly into a red barrel
and taking a refurbished one from a green barrel. It saves a
fortune in air and maintenance labor, Stroup says.
Add leak surveys to compressor maintenance contracts:
If it makes sense, divide the plant into quarters and survey
one quarter with each visit, Booth says.
Keep everyone aware: Like a safety program, hang signs
such as, -in. leak = $8,000/year.
START A NEIGHBORHOOD WATCH
Table 2: Critical steps
1. Evaluation (leaks, pressures and compressor controls)
2. Detection (involve operators)
3. Identifcation (tagging)
4. Tracking
5. Repair
6. Verifcation
7. Evaluation
MANAGEMENT
Compressed Air Systems
Tose same experts agree that, although the learning
curve is short, some training is required. Get a demo,
says Zuercher. Have someone come in and show you
how they work.
Tough not yet a common use, increasingly sophisticated
infrared imagers can detect the thermal plume of a compressed
air or gas leak. We have used this technique in the past for
pinpointing leaks on certain pieces of equipment, especially
when it is in an environment where you cannot enter safely,
says Stroup.
Stop repeat ofenders
Vigilance is your best defense against a growing leak popu-
lation, but a few key practices can help. First, set aside the
Teon tape when assembling threaded pipe and ttings.
Teon tape seals the contact points, but it doesnt ll the
voids, says Tom Buckley, application engineer, Henkel
Corp. (www.loctite.com). Teres a lot of space left for
leak paths. Sealants in the form of thick liquids or pastes
fully ll the spaces in the threads.
On compressors, our extensive testing shows Teon tape is
not an answer, says Booth. You need a quality sealant com-
patible with the compressor uid.
Before the sealant has cured, pipe threads can be adjusted
for alignment by loosening or tightening without compromis-
ing the seal. Sealants that wont cure while exposed to air can
be applied to all the pipes and ttings in one operation before
assembly. Fittings can be adjusted for several hours after as-
sembly and once cured, the sealant adds mechanical strength
and vibration resistance to the joint.
Fluid compatibility is an issue with the elastomers in FRLs,
non-threaded pipe joints and connected equipment. Te di-
ester lubricants can be extremely aggressive, Booth says. If
youre using a diester, be sure youre using the correct elasto-
mers and hoses.
Tough piping ranks low on the frequency list, leaks there
can be the most dicult to repair. Joints and ttings can
loosen or wear over time and should be routinely checked,
says Perry. Old or outdated piping of any material should be
immediately replaced.
While many prefer copper and stainless steel, there are several
modular aluminum piping products that oer a cost-eective al-
ternative. Not only can these products be fully integrated into
existing systems, they oer push-to-t connections that elimi-
nate brazing and welding joints and ttings the most likely
place piping leaks occur, Perry adds.
Lightweight modular systems can be installed quickly
without specialized skills. Our push-to-connect sys-
tem uses high-nitrile seals and is rated to 232 psi, says
Tim Deal, manager of customer service and sales, Legris
Transair (www.transair.legris.com). Te seal material is
the same that is commonly used in compressors so it will
resist all compressor uids. Its a good way to replace steel
or copper so it doesnt corrode and it leak again.
M 2006 .PLANTSERVICES. 37
MANAGEMENT
Compressed Air Systems
lenoid valve trying to x an intermittent equipment prob-
lem thats actually caused by an undetected leak.
For example, a packaging machine wouldnt work
properly when other equipment in the area was running.
Mechanics replaced a solenoid valve with no effect.
When we got next to it, we could hear a leak, Stroup
says. We crawled under it and found where a chain had
worn a hole in an air line. Fixing that fixed the problem
with the machine.
Pneumatic systems inside panels often share a common,
ducted exhaust, masking the presence of leaks in the normal
airow. An internal failure of a solenoid valve can blow a
lot of undetected air, whether the machines running or not.
We also often nd missing faceplates and uncapped lines
in solenoid cabinets, Stroup adds.
Dont ignore useful leaks. For example, compressed air
used to position paper in an envelope machine should be
turned o when the machines not running. Add a shut-o
valve interlocked with the on-o switch on the machine. A
manual valve isnt reliable because operators dont want to
worry about it and will simply leave it on.
Get on the case
Detection methods range in sophistication from listening
and running hands over pipes and equipment to soap bubbles
and, todays big gun, the ultrasonic leak detector (ULD). All
these tools have their place, but the most eective and least
familiar is the ULD.
Ultrasonic is a critical element, says Zuercher. Te old-
school thinking was that you had to detect leaks on nights
and weekends when the plant is quiet so you can hear the
leaks, and most people dont want to do that. With ultra-
sound, you can detect leaks during normal production.
Handheld ULDs locate leaks both in compressed air and vac-
uum systems, and search areas such as valve seats, drain traps,
tanks and piping. USDs can detect internal leaks and leaks that
are near the ceiling, outdoors or in noisy environments. For
example, a blown diaphragm on a valve in a baghouse can leak
a lot of air where it is easily overlooked, Zuercher says. Good
ultrasonic can detect leaks from over 50 feet away.
A good ULD can nd leaks quickly, but acionados disagree
on whether they can be used to judge leak rate. Zuercher says the
size of a leak is proportional to the volume of the sound given
o by the ultrasonic leak detector, but Ben Fried, product devel-
opment and support, CTRL Systems (www.ctrlsys.com), says,
Te number-one misconception about ULDs is that they can
measure the intensity of a leak and correlate it to dollars lost.
Tis is 100% false.
Many factors change the intensity of a leak, including
pressure; temperature; humidity; distance from the receiv-
er; size, shape and location of the hole; proximity of hole
to other objects; battery power of receiver; etc., Fried says.
Tere is no way to determine amount of cfm loss based on a
decibel reading or some other relative measurement.
In general, ULD capabilities vary, and experts have strong
words about making sure the one you buy will do what you
expect (see sidebar, ULD: Pick from a lineup).
Tose same experts agree that, although the learning curve
is short, some training is required. Get a demo, says Zuerch-
er. Have someone come in and show you how they work.
Tough not yet a common use, increasingly sophisticated
infrared imagers can detect the thermal plume of a compressed
air or gas leak. We have used this technique in the past for
pinpointing leaks on certain pieces of equipment, especially
when it is in an environment where you cannot enter safely,
says Stroup.
Stop repeat ofenders
Vigilance is your best defense against a growing leak pop-
ulation, but a few key practices can help. First, set aside the
Teon tape when assembling threaded pipe and ttings.
Teon tape seals the contact points, but it doesnt ll the
voids, says Tom Buckley, application engineer, Henkel
Corp. (www.loctite.com). Teres a lot of space left for
leak paths. Sealants in the form of thick liquids or pastes
fully ll the spaces in the threads.
On compressors, our extensive testing shows Teon tape is
not an answer, says Booth. You need a quality sealant com-
patible with the compressor uid.
Before the sealant has cured, pipe threads can be adjusted
for alignment by loosening or tightening without compromis-
ing the seal. Sealants that wont cure while exposed to air can
be applied to all the pipes and ttings in one operation before
assembly. Fittings can be adjusted for several hours after as-
sembly and once cured, the sealant adds mechanical strength
and vibration resistance to the joint.
Ben Fried, product development and support, CTRL Systems
(www.ctrlsys.com) ofers these suggestions for choosing an
ultrasonic leak detector (ULD):
1. Sensitivity: Several ultrasonic leak detectors (ULDs) on
the market ofer a wide range of sensitivity. Rule of thumb:
You get what you pay for.
2. Selectivity: This relates to dynamic range of a ULD.
Some detectors have wide bandwidth, thereby picking up
more white noise or static. Additional noise decreases the
users ability to detect a leak.
3. Volume: Some ULDs feature adjustable sensitivity and
volume. Changing the volume increases the output ampli-
tude, but doesnt improve sensitivity.
4. Price: ULDs range in price from $100 to $10,000. Gener-
ally, higher-price ULDs are capable of doing condition moni-
toring as well. The lowest-priced ULDs arent very efective
in a manufacturing facility and can miss most leaks. Ultra-
sonic leak detector quality is all over the map, adds Jan
Zuercher, director, air systems, Quincy Compressor (www.
quincycompressor.com). Someone wanting to purchase a
leak detector should talk to an experienced user or test a
handful of diferent instruments in their plant.
ULD: PICK FROM A LINEUP
lenoid valve trying to x an intermittent equipment problem
thats actually caused by an undetected leak.
For example, a packaging machine wouldnt work
properly when other equipment in the area was running.
Mechanics replaced a solenoid valve with no effect.
When we got next to it, we could hear a leak, Stroup
says. We crawled under it and found where a chain had
worn a hole in an air line. Fixing that fixed the problem
with the machine.
Pneumatic systems inside panels often share a common,
ducted exhaust, masking the presence of leaks in the normal
airow. An internal failure of a solenoid valve can blow a
lot of undetected air, whether the machines running or not.
We also often nd missing faceplates and uncapped lines
in solenoid cabinets, Stroup adds.
Dont ignore useful leaks. For example, compressed air
used to position paper in an envelope machine should be
turned o when the machines not running. Add a shut-o
valve interlocked with the on-o switch on the machine. A
manual valve isnt reliable because operators dont want to
worry about it and will simply leave it on.
Get on the case
Detection methods range in sophistication from listening
and running hands over pipes and equipment to soap bubbles
and, todays big gun, the ultrasonic leak detector (ULD). All
these tools have their place, but the most eective and least
familiar is the ULD.
Ultrasonic is a critical element, says Zuercher. Te old-
school thinking was that you had to detect leaks on nights
and weekends when the plant is quiet so you can hear the
leaks, and most people dont want to do that. With ultra-
sound, you can detect leaks during normal production.
Handheld ULDs locate leaks both in compressed air
and vacuum systems, and search areas such as valve seats,
drain traps, tanks and piping. USDs can detect internal
leaks and leaks that are near the ceiling, outdoors or in
noisy environments. For example, a blown diaphragm on
a valve in a baghouse can leak a lot of air where it is easily
overlooked, Zuercher says. Good ultrasonic can detect
leaks from over 50 feet away.
A good ULD can nd leaks quickly, but acionados disagree
on whether they can be used to judge leak rate. Zuercher says the
size of a leak is proportional to the volume of the sound given
o by the ultrasonic leak detector, but Ben Fried, product devel-
opment and support, CTRL Systems (www.ctrlsys.com), says,
Te number-one misconception about ULDs is that they can
measure the intensity of a leak and correlate it to dollars lost.
Tis is 100% false.
Many factors change the intensity of a leak, including
pressure; temperature; humidity; distance from the receiv-
er; size, shape and location of the hole; proximity of hole
to other objects; battery power of receiver; etc., Fried says.
Tere is no way to determine amount of cfm loss based on a
decibel reading or some other relative measurement.
In general, ULD capabilities vary, and experts have strong
words about making sure the one you buy will do what you
expect (see sidebar, ULD: Pick from a lineup).
Tose same experts agree that, although the learning curve
is short, some training is required. Get a demo, says Zuerch-
er. Have someone come in and show you how they work.
Tough not yet a common use, increasingly sophisticated
infrared imagers can detect the thermal plume of a compressed
air or gas leak. We have used this technique in the past for
pinpointing leaks on certain pieces of equipment, especially
when it is in an environment where you cannot enter safely,
says Stroup.
Stop repeat ofenders
Vigilance is your best defense against a growing leak population,
but a few key practices can help. First, set aside the Teon tape
when assembling threaded pipe and ttings. Teon tape seals
the contact points, but it doesnt ll the voids, says Tom Buckley,
application engineer, Henkel Corp. (www.loctite.com). Teres
a lot of space left for leak paths. Sealants in the form of thick
liquids or pastes fully ll the spaces in the threads.
On compressors, our extensive testing shows Teon tape is
not an answer, says Booth. You need a quality sealant com-
patible with the compressor uid.
Before the sealant has cured, pipe threads can be adjusted
for alignment by loosening or tightening without compromis-
ing the seal. Sealants that wont cure while exposed to air can
be applied to all the pipes and ttings in one operation before
assembly. Fittings can be adjusted for several hours after as-
sembly and once cured, the sealant adds mechanical strength
and vibration resistance to the joint.
Fluid compatibility is an issue with the elastomers in FRLs,
non-threaded pipe joints and connected equipment. Te di-
ester lubricants can be extremely aggressive, Booth says. If
youre using a diester, be sure youre using the correct elasto-
mers and hoses.
Tough piping ranks low on the frequency list, leaks there
can be the most dicult to repair. Joints and ttings can
loosen or wear over time and should be routinely checked,
says Perry. Old or outdated piping of any material should be
immediately replaced.
While many prefer copper and stainless steel, there are several
modular aluminum piping products that oer a cost-eective al-
ternative. Not only can these products be fully integrated into
existing systems, they oer push-to-t connections that elimi-
nate brazing and welding joints and ttings the most likely
place piping leaks occur, Perry adds.
Lightweight modular systems can be installed quickly with-
out specialized skills. Our push-to-connect system uses high-
nitrile seals and is rated to 232 psi, says Tim Deal, manager of
customer service and sales, Legris Transair (www.transair.legris.
com). Te seal material is the same that is commonly used in
M 2006 .PLANTSERVICES. 38
MANAGEMENT
Compressed Air Systems
M 2006 .PLANTSERVICES. 39
MANAGEMENT
Compressed Air Systems
Make a major crackdown
Quantifying and systematically controlling compressed air
leakage starts and ends with periodically evaluating your
systems leak rate (Table II), preferably as a percentage of
use. Today, most plants are at 30% to 40%, and Ive been in
several that were at 50% or more, says Stroud.
Te best way to determine energy use and savings is to mea-
sure electricity consumption of compressors. Te Compressed
Air Challenge (CAC, www.compressedairchallenge.org) oers
other methods to baseline and evaluate system leakage. Once
you know where you are, you can estimate potential savings
and measure progress over time. No one is going to get to zero
leaks, Stroud says, but even in heavy industry with inaccessible
equipment, you should be able to achieve 15% or less. Cleaner
facilities should reach 10% and the best are at about 5%, though
I know of two that made it to 3%.
Te inspirational dollar amounts shown in Figure 1 arent
very useful for estimating real-life potential savings, as typical
leaks are not perfect orices at constant, known pressure. We
test air systems and meter the ow volumes to quantify the
leaks, says Stroup. When we do that at dierent pressures,
we dont see a one-to-one correlation.
For example, he says, the charts will show a -in. hole
at 100 psig leaks 104 cfm and costs about $10,000 per year,
but thats based on a perfect orice. A real leak is gener-
ally a gash or cut or other irregular shape, so you use a
-in. equivalent area, but you cant put 104 cfm through
the tube, so you have to take it with a grain of salt, Stroup
says. I usually cut it in half for a more realistic estimate.
Ready to nd and x leaks? First, prioritize. Start in the
area where the most air is used or where the most leaks are.
Stroup says, Most plants have a feel for that.
Te rst time through, focus on 20 cfm leaks and ignore
1 cfm leaks. Start with your ears and hands, then get to
ultrasonic, Stroup says. You cant quantify a leak with ul-
trasonic or any other method, but you can estimate.
Stroup says, initially, a large leak is more than 10 cfm, me-
dium is 5 cfm to 10 cfm, and small is less than 5 cfm. Do
the large leaks rst, and the medium ones if you can, he says.
Next time around, the medium leaks will be large reclassi-
fy them and do it over. Eventually the leaks will be too small to
be worth xing. Ten watch them and x the ones that grow.
Dont be daunted by the results of your rst survey. You
do a survey and three days later, you have 150 tags, says
Bishop, You gag on it. Te answer, he says, is dont try to
do it all at once do it one day a month. After a year, issue
ve open work orders a week from your CMMS to nd and
x a leak and dont come back until its xed.
Repeat the survey at regular intervals. In a high-vibration
and shock facility like a sawmill, check for leaks monthly,
Stroup says. In most facilities, such as food and light manu-
facturing with stationary machines, quarterly is good. Clean,
quiet industries like pharmaceuticals can go semi-annual,
but, he warns, Nobody should go longer than six months.
Te key is to keep at it. Te benets dont come in one
month. Youll x the leaks, then x more, Stroup says.
For most plants it takes six months to a year to see benets.
Te progress of a typical facility is shown in Figure 2.
Ultrasonic, feel, listen, whatever youve got, says Bartos.
Te important thing is to do it regularly. You spend so much
money on a compressor and energy, its going to pay o.
Figures: Airometrix, Kaeser
T
he compressed air industry is taking a step that will
clear the air about the performance of its equipment.
For the rst time, many compressed air equipment
manufacturers will submit their 50 hp to 200 hp lubricated
rotary-screw compressors and 200 cfm to 1,000 cfm refrig-
erated air dryers to an independent laboratory for perfor-
mance testing. Te Compressor Distributors Association
(CDA) led the charge that moved the industry to this deci-
sion. Te CDA, a group of industrial air compressor dis-
tributors representing dierent manufacturers, was formed
to sponsor the Compressed Air Challenge (CAC) and to
have a voice in its operations. Te CDA is represented on
both the CAC Board of Directors and its Product Develop-
ment Committee.
After several meetings, the CDA began a series of dis-
cussions about how the organization might improve the
industry. Te group agreed that a need existed for ac-
curate and independently veried performance data on
the equipment they provide to end users. Te motivation
is that distributors are being pressured to provide much
more accurate system analysis and
recommendations than have been
required in the past.
With the ability to be accurate
comes an increasing need for accurate
and dependable input information. The
commitment to this need prompted CDA to fund
an independent, blind testing program for seven manu-
facturers 50-hp lubricated rotary-screw compressors
against their published performance information. Find-
ing a capable, well-equipped laboratory was a formidable
task, as was securing the use of seven competitive ma-
chines for the short time window available.
Test results are presented as percentage variance from the
published data. With test data in hand, CDA approached
the Compressed Air Challenge Board and the Compressed
Air and Gas Institute (CAGI) to illustrate the need for
universal testing. While there had been many discussions
within the industry about independent testing, nobody
seemed to be making tangible moves in that direction.
CAGI members had adopted a Compressor Data Sheet to
report performance data in accordance with ISO 1217, but
many manufacturers hadnt made this information readily
available. After CDAs presentations, these groups, along
with Association of State Energy Research and Technol-
ogy Transfer Institutions (ASSERTI), endorsed the value
of outside verication of product performance.CAGI took
the lead and worked with other organizations to formulate
the testing program. Tree years later, after a lot of coop-
eration and hard work, the program is now in place.
J 2006 .PLANTSERVICES. 46

Tird-party certication will power better decisions
By Ken Byrd
Once a machine tests successfully,
compressor and air dryer
manufacturers will be entitled to use
the CAGI Performance Label.
J 2006 .PLANTSERVICES. 47
Whats being tested?
The CAGI compressor performance testing
includes rated capacity (cfm) at full-load
operating pressure (psi), total input power
at rated capacity at full operating pressure
(kW) and specific package power (kW/100
cfm). Refrigerated air dryers are tested for f low (scfm),
outlet pressure dewpoint (deg. F), pressure drop (psid),
total input power (kW) and specific package power
(kW/100 cfm).
Te lab will administer the program, contact the man-
ufacturer and select a specic machine to be tested. Teir
random selection will come either directly from manu-
facturer inventory or from a distributors stock. After
the test, the manufacturer will be advised if the machine
passed or failed.
Once a machine tests successfully, compressor and air
dryer manufacturers will be entitled to use the CAGI
Performance Label. CAGI manufacturers whose equip-
ment fails the test will have an opportunity to correct
the problem. If it cant be corrected, the manufacturer
isnt entitled to use the CAGI label. Te CAGI Web site
will post a notice of non-performance, thus putting the
manufacturer at a signicant marketing disadvantage.
Te program also is being oered to manufacturers that
arent CAGI members.
What it means to you
Tis testing program allows users to evaluate the data pre-
sented with new equipment bids and to feel more condent
with purchasing decisions. In the most basic comparisons of
machines, end users ask:
How much compressed air will this machine provide
compared to its capital cost?
How much energy will it consume per unit of compressed
air (cfm/bhp)?
How much will it cost to operate?
Answering these and similar questions always has
been a fundamental step in the decision to purchase a
compressor or dryer. In the past, end users could only
rely on answers based solely on the information the man-
ufacturer provided. This typically included only the full-
load volume (cfm), horsepower required at full load and
horsepower at an unloaded state. In sales presentations,
this information most often is coupled with discussion
of the control system in an attempt to evaluate the ma-
chines efficiency while operating at partial load, often
leaving room for substantial doubt. Soon, most of these
performance claims will be substantiated and reported
on the CAGI Web site.
After purchase, a machines full-load and partial-load
efficiency has been difficult to verify and often faded to
a point of little interest. The result has been
many, if not most, systems are unmonitored
and operating with unnecessary energy con-
sumption. Further, system changes with air
users being added or removed, pressure in-
creased or reduced, production hours increas-
ing or decreasing, and evolving leaks can make dramatic
changes in the energy consumed. Without a reliable
baseline for comparison, its been difficult to quantify
the effects of these changes.
A system analyst now can easily evaluate a machines
performance with modern data collection tools, but reliable
equipment performance information is still essential. With
good data, knowledgeable compressed air system analysts
can verify system performance and easily determine if the
system is operating eciently. Tis data also can determine
if system changes would lead to substantial energy savings.
An example
Consider a plant thats operating two 125-hp xed-speed
single-stage rotary air compressors and one 200-hp xed-
speed single-stage rotary screw air compressor. Tis plant
operates 24 hours per day with a load that varies greatly
throughout the day. Te machines operate o individual
controls that sense discharge pressure. Plant maintenance
personnel decide when to run the 200-hp as the lead ma-
chine instead of a smaller unit.
Te air system energy analysis found that during most
of the day one 125-hp unit provides more than enough air
to maintain pressure. For a few hours each day, the sys-
tem required either both 125-hp units to operate partially
loaded or the 200-hp unit to run partially loaded.
Most of the time, the plant runs two partially loaded
125-hp units, an inefficient approach given that fixed-
speed compressors are inherently inefficient when oper-
ated that way. The plant considered buying a 160-hp vari-
able-speed single-stage rotary air compressor to improve
efficiency. The unit would be able to handle the entire
load for most of each day, varying its speed to maintain
pressure. During the short periods when the 160-hp unit
wouldnt be capable of maintaining pressure, one of the
125-hp units would start and run fully loaded.
Because a variable-speed compressor is more eff icient
than a partially loaded f ixed-speed machine, the 160-
Many, if not most, systems are
unmonitored and operating with
unnecessary energy consumption.
RELIABILITY
Compressors
hp unit would serve as a trim machine after the smaller
unit loads up. Once system demand drops into the 125-
hp units range, the smaller f ixed-speed machine will
unload and the 160-hp variable speed unit takes over to
satisfy demand.
In this case, the estimated yearly system operating cost
with the two 125-hp xed-speed single-stage rotary air
compressors and 200-hp xed-speed single-stage rotary
air compressor was about $68,000. Installing the 160-hp
variable-speed single-stage rotary air compressor into the
system could reduce power costs about $36,000 per year.
Payback on the project was less than two years.
As with many things, these recommendations are
only as good as the input. Accurate performance data
for the existing compressors and the machine being rec-
ommended are critical for the desired outcome. Before
suggesting any changes, the analyst must have conf i-
dence in information being used for the decision. This
same information continues to be helpful for verifying
savings and in monitoring the effects of future changes
in the system.
Te CDA members are proud to have been the impetus
for industry taking this step forward and continue to look
for opportunities to promote and improve our industry. For
further information, contact CAGI at www.cagi.org.

Ken Byrd is on the board of directors of the Compressed Air Chal-
lenge, the Compressor Distributor Association as well as the North
American Association of Compressor Distributors. Contact him at
kbyrd@airservicesco.com and (630) 766-7900.
J 2006 .PLANTSERVICES. 48
RELIABILITY
Compressors
Controls Compressor capacity controls
Baseline measurements Vital signs
Dryers Controlling compressed air moisture
Energy surveys So, you want a compressed air
energy survey?
TCO Look beyond the sticker price
Efciency issue Load em up
Dryers Theres more than one way to dry the air
For more, search www.plantservices.com using the keywords
dryer, survey and efciency.
More resources at www.PlantServices.com/thismonth
The CAGI Web site will post a notice
of non-performance, thus putting
the manufacturer at a signicant
marketing disadvantage.
or the rst half of the 20th century and through the
1950s and 1960s, the predominant industrial air
compressor was the positive-displacement double-
acting, water-cooled reciprocating design. Te predominant
control methodology allowed for three to ve control load
settings: 0% 50% 100% or 0% 25% 50%
75% 100%. Tese settings responded to the systems dy-
namic demand and the power consumption was quite close
to the load factor, while the unloaded horsepower ran be-
tween 18% and 20% of maximum.
Te mechanisms that accomplished this essentially
changed the geometry of the compression chamber cylin-
der with a variety of devices that held a set of valves open,
which allowed air to move in and out of the chamber with-
out compression. With respect to eciency, this design
generally consumed 15 kW to 16 kW per 100 CFM of
compressed air. Todays compressors and controls barely
reach that level of eciency.
Ten, why have reciprocating designs fallen out of favor rela-
tive to the rotary compressor that is so predominant today?
Problems with pistons
A double-acting reciprocating compressor is a belt-driven
cast iron water-cooled compression chamber. It generates
unbalanced forces that require large concrete foundation
pads. Installation is arduous. Tink of an Erector set having
a compressor, motor, starter, belt drive, aftercooler, receiver,
condensate drain traps and dozens of minor pieces. Te
A 2006 .PLANTSERVICES. 48
A 2006
working device also requires more oor space than a rotary-
screw unit. Te only cooling option was water. Maintaining
the valves, crossheads and main bearings required substan-
tial mechanical skills lacking in todays labor pool.
Compressed air delivery wasnt smooth. Tis caused a
slight but real interruption in the compressed air ow. Large
receiver tanks were required to tame it into a smooth air
ow. And, dont forget the temperature gain of about 300F
across the compression cycle.
The rise of rotaries
Te 1960s saw the advent of rotary compressors. From the
outset, these packaged machines received a warm market re-
ception. Tey were generally available as prewired, prepiped
units. Vibration was minimal and a rotary unit could be
placed on any structure that could support its weight.
Other features included zero unbalanced forces, only
about 200F heat gain, no receiver required, no startup or-
deal, no run-in time, and lower installed cost. Tree to ve
gallons of storage capacity per CFM worked for even the
largest machine (Figure 1).
Rotary compressor eciency was only 18 kW to 19 kW
per 100 CFM, or about 18% worse than the reciprocating
design. In those days, electricity cost only about $0.025
per kWh, erroneously considered insignicant. After all,
air is free, isnt it?
Tese gures apply to single-stage lubricant-ooded com-
pressors. Two-stage lubricant-ooded compressors are more
ecient and often pay for themselves in a reasonable time.
Te extra investment was justied. Te operating eciency
of two-stage units is 16 kW to 17 kW per 100 CFM.
Only load/unload controls were available. At a preset
pressure, the compressor unloaded and idled. Generally,
unloaded horsepower is 25% of full-load horsepower.
Pay adequate attention to the timing of the load and
unload cycles on lubricant-ooded rotary screw compres-
sors. Te cycle should allow ample time for the sump to
blow down, often as long as 60 seconds or more. Each ma-
chine is dierent. Measure blow down times individually.
Short cycling can cause the lubricant to foam and dimin-
ish control system life expectancy. It also risks foaming
the lubricant out of the compressor inlet onto the oor.
Inlet modulation is inefcient
Te load/unload approach was supplanted by modulation
controls that feature a gradually closing inlet buttery valve.
Te weakness in this design is miserly little power savings at
part load (Figure 2). Note that at:
100% capacity, 100% of the power is required
90% capacity, the power required is 97%
80% capacity, the power required is 95%
70% capacity, the power required is 90%
60% capacity, the power required is 85%
50% capacity, the power required is 83%
1 gal/cfm
10 gal/cfm
3 gal/cfm
5 gal/cfm
Percent capacity
1 gal/cfm
3 gal/cfm
5 gal/cfm
10 gal/cfm
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
Load/unload and storage size
No blowdown
With blowdown
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
Percent capacity
Inlet valve modulation
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
Percent capacity
Rotary compressor performance with variable displacement
Variable displacement
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
10
0
Percent capacity
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
% kW input vs % capacity
With unloading
With stopping
Variable-speed drive
Total kW input and/or specific power over the full
operating range must be analyzed for a proper
comparison with other types of capacity control.
Figure 1. Input power versus capacity for a lubricant-injected
rotary screw compressor using load/unload control.
On/of
1 gal/cfm
10 gal/cfm
3 gal/cfm
5 gal/cfm
Percent capacity
1 gal/cfm
3 gal/cfm
5 gal/cfm
10 gal/cfm
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
Load/unload and storage size
No blowdown
With blowdown
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
Percent capacity
Inlet valve modulation
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
Percent capacity
Rotary compressor performance with variable displacement
Variable displacement
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
10
0
Percent capacity
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
% kW input vs % capacity
With unloading
With stopping
Variable-speed drive
Total kW input and/or specific power over the full
operating range must be analyzed for a proper
comparison with other types of capacity control.
Figure 2. Input power versus capacity for a lubricant-injected
rotary screw compressor controlled by inlet modulation.
Throttling
1 gal/cfm
10 gal/cfm
3 gal/cfm
5 gal/cfm
Percent capacity
1 gal/cfm
3 gal/cfm
5 gal/cfm
10 gal/cfm
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
Load/unload and storage size
No blowdown
With blowdown
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
Percent capacity
Inlet valve modulation
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
Percent capacity
Rotary compressor performance with variable displacement
Variable displacement
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
10
0
Percent capacity
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
% kW input vs % capacity
With unloading
With stopping
Variable-speed drive
Total kW input and/or specific power over the full
operating range must be analyzed for a proper
comparison with other types of capacity control.
Figure 3. Input power versus capacity for a lubricant-injected
rotary screw compressor using variable-displacement control.
Flexible cylinders
49 .PLANTSERVICES.
EFFICIENCY
Compressors
40% capacity, the power required is 80%
Zero capacity requires 25% power. For best energy e-
ciency run multiple rotary compressors at full load and use
one for trim. Operating several inlet modulation control
systems at part load is a terrible waste of money.
Pushing the frontiers
Technology brought us variable-displacement machines
in which the geometry of the compression chamber could
change. Some machines used a spiral valve, while others
used poppet valves that shorten the eective length of the
compression chamber. Both approaches allow much better
part-load performance (Figure 3). Note that at 100% capac-
ity, 100% of the power is required, but:
90% capacity needs 92% power
80% capacity needs 83% power
70% capacity needs 78% power
60% capacity needs 68% power
50% capacity needs 63% power
40% capacity needs 60% power
Te next big thing in the world of rotary compressors
was the introduction of variable-speed drives (Figure 4).
As before, 100% capacity requires 100% power. But:
90% capacity requires 91% power
80% capacity requires 81% power
70% capacity requires 71% power
60% capacity requires 61% power
50% capacity requires 51% power
40% capacity requires 42% power
Control for efciency
Centrifugal compressors rely on modulation and blow-o,
which generally limits turndown capacity to between 70% and
75% of total capacity. When demand falls, the control goes to
blow-o mode, dumping the full capacity to atmosphere. When
operating multiple machines, the same suggestions about hav-
ing only one trim unit and modulating multiple units apply.
A variety of system controls for multiple compressor instal-
lations are available. Single master or sequencing controls use
one control to sequence or lead/lag the compressors. Multiple
master or network controls allow the system to alternate, unload
or turn o automatically as system demand changes. Control
systems can be integrated with computer-based overall plant
energy control systems that monitor HVAC, lights and other
items. Pressure ow control devices are separate from com-
pressor controls and are intended to deliver a relatively constant
static pressure supply to the air distribution system.
The end of modulation?
It appears likely that the industry will serve the rotary screw,
both lubricated and dry, with less complicated, more ener-
gy-ecient controls. Youll probably see modulation elimi-
nated, leaving only load/unload or variable speed as the two
most energy-ecient mechanisms available today.
A major caution is in order. At full load, variable-speed
units generally require more power. Tus, in multiple rota-
ry-compressor installations, operate only one variable-speed
unit on trim duty if you want the promised energy savings.
Other surprises are on the horizon. Consider the global-
ization of the compressor business. Fifty years ago, it was
inconceivable for an oshore automobile manufacturer to
own half the U.S. car market. Te same is happening in the
compressor industry.
Te entrepreneurial creativity demonstrated by the young
folks coming into the market today is far beyond anything I
could imagine. Te Electronic Age has everything moving
at supersonic speed. We aint seen nothin yet.
Henry Kemp is principal at Strategic Air Concepts in St. Petersburg,
Fla. Contact him at henrykemp@strat-air.com and (727) 867-4044.
Figures: Compressed Air Challenge
A 2006 .PLANTSERVICES. 50
EFFICIENCY
Compressors
1 gal/cfm
10 gal/cfm
3 gal/cfm
5 gal/cfm
Percent capacity
1 gal/cfm
3 gal/cfm
5 gal/cfm
10 gal/cfm
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
Load/unload and storage size
No blowdown
With blowdown
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
Percent capacity
Inlet valve modulation
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
Percent capacity
Rotary compressor performance with variable displacement
Variable displacement
110
100
90
80
70
60
50
40
30
20
0 20 40 60 80 100 10 30 50 70 90
10
0
Percent capacity
P
e
r
c
e
n
t

k
W

i
n
p
u
t

p
o
w
e
r
% kW input vs % capacity
With unloading
With stopping
Variable-speed drive
Total kW input and/or specific power over the full
operating range must be analyzed for a proper
comparison with other types of capacity control.
Figure 4. Input power versus capacity for a lubricant-injected
rotary screw compressor using variable-speed control.
Motor drives
Capacity controls
Compressor capacity controls
Baseline measurements
Vital signs
Integration issues
Integrated compressor controls
Smart compressors
Air compressors and microprocessors
Efciency and loading
Load em up
For more, search www.plantservices.com using the keywords
compressor, modulation and load/unload.
More resources at www.PlantServices.com/thismonth
D
uring the past several years, the variable-speed-drive
(VSD) compressor has become a frequent choice for
new compressor purchases. Te VSD compressors
popularity is partly due to rising energy prices and its e-
ciency as a trim compressor. Unfortunately, much of the VSD
compressors popularity is a result of marketing spin. For ex-
ample, sales personnel oer free audits during which they
install a black box that might as well ash the phrase on the
screen, Buy a VSD compressor. Often, the free audits exag-
gerate paybacks by including savings the end user can achieve
without purchasing a VSD compressor. Te marketers also fail
to provide the necessary engineering details required to maxi-
mize the purported benets. Its time to eliminate the spin and
point out both the limitations and engineering required if you
are to derive the benets of a VSD compressor.
The spin
A trim compressor is a cycling compressor that matches
supply capacity to uctuating system demand, while a base-
load compressor is a constant-speed compressor operating
at its maximum eciency.
Over much of its capacity range, the VSD compressor is
the most ecient part-load rotary screw compressor. On
the other hand, the VSD compressor isnt a panacea for all
your compressed air system problems, as the spin doctors
would like you to believe. Some of what they say about
VSD compressors includes:
Compressor eciency: You get maximum eciency
while peaking at 100%, or only operating at the average,
50% to 70% of capacity. In graphical form, the spin doc-
tors always show a power consumption comparison of op-
erating modes in a graph of percent power versus percent
capacity (Figure 1).
Energy savings: One ad claims, Cuts energy costs up to
50%, while another ad by the same manufacturer says, 40%
reduction in energy costs possible. Another states Energy
savings up to 35%. Yet another manufacturer says, Slash
power use resulting in a life-cycle cost savings of 22%.
Power factor: Te spinners claim, No penalties for spikes
or low power factor.
Free energy analysis: Tey want us to believe that a free en-
ergy analysis is the same as a compressed air system audit.
.PLANTSERVICES. O 2006 45
.PLANTSERVICES.
Turndown: Spinners want us to believe that every VSD
compressor has a turndown of 80% or more.
Nameplate horsepower: Te claimed horsepower often
isnt the required horsepower.
Constant pressure: One manufacturer claims, Line
pressures held within 1 psi, while another says, Line
pressures held within 1.5 psi. Still another states, High
process stability. Tis leads compressor sales personnel
to make the statement, If you install a VSD compressor,
you wont need a pressure/ow controller.
Storage: Because the VSD compressor responds to sys-
tem events with online horsepower, the spinners tell us,
A large receiver isnt required.
While, in a limited perspective, one can interpret these
claims as technically true, the spin doctors have gone for the
maximum rotational speed. From this point forward, how-
ever, the discussion about VSD compressors enters a No Spin
Zone, where we peel o the spin to leave only the facts.
Compressor efciency
VSD compressor literature uses the percent power versus
percent capacity graph to show how much more efficient
the VSD compressor is than a constant-speed compres-
sor operating in either modulation or load/unload mode.
The VSD and VSD II curves (Figure 1) represent ef-
ficiency curves from two manufacturers. The problem
with this graph is that it hides the benefits of constant-
speed compressors.
For example, at full load, a VSD compressor isnt as
ecient as the comparable load/unload compressor from
the same manufacturer because the variable-frequency
drive increases power draw by 2% to 4%. To remove the
spin, simply plot the graph using actual air capacity in-
stead of percent capacity. As you approach 100% power,
the constant-speed compressor is more ecient than the
VSD compressor (Figure 2). In addition, constant-speed
compressors generally have a higher capacity. Tis gure
shows that the constant-speed compressor delivers 10% or
45 cfm more capacity at full load than the VSD compres-
sor, while the VSD II compressor delivers 4 cfm less.
Graphing compressor eciency (input kW/100 cfm) versus
cfm (Figure 3) properly portrays the eciency dierences.
Te load/unload eciencies assume 10 gal./cfm of storage,
based upon the trim compressors capacity. Te graph shows
that as the load on the compressor decreases, the VSD com-
pressor eciency is relatively at between approximately 46%
and 78% of its capacity, while the load/unload eciency de-
creases rather quickly. Furthermore, below 15% to 20% ca-
pacity, the VSD compressors eciency follows a downward
trend, while the load/unload compressors eciency, because
of unloaded horsepower, remains at a higher level.
Te second compressor eciency graph (Figure 4)
zooms in on the crossover point, above which the con-
stant-speed compressor is more ecient. At full load, the
constant-speed compressor is approximately 12% more
ecient than the VSD compressor and approximately 5%
more ecient than the VSD compressor operating at its
most ecient operating point.
Energy savings
Statements such as cuts energy costs up to 50%, 40% reduction
in energy costs possible, energy savings up to 35%, and slash
power use and reduce life-cycle cost by 22% lead plant person-
nel to believe that just by installing a VSD compressor, they can
reduce their total annual energy cost by as much as 50%. Te spin
doctors reinforce this notion by oering a free energy analysis or
audit. However, unless you operate only one compressor, a 50%
reduction isnt possible merely by installing a VSD compressor.
MANAGEMENT
Compressors
105 psig
adjustable
range
PID
setpoint
always in trim
VSD-100 hp
500 cfm
compressor
dryer
5 psid
93 psig 91 psig
filter
tank
5 psid 5 psid
dryer dryer
75 hp
370 cfm
compressor
75 hp
370 cfm
compressor
base 1 base 2
95 psig
95 psig
104 psig
93 psig
102 psig
91 psig
90 psig
Header pressure
controlled by VSD
pressure dips to start
#1 base and base #2
75
cfm
485
cfm
885
cfm
max
1225
cfm
Flow
increasing
95 psig
105 psig
adjustable
range
PID
setpoint
always in trim
VSD-100 hp
500 cfm
compressor
dryer
5 psid
filter
tank
5 psid 5 psid
dryer dryer
75 hp
370 cfm
compressor
75 hp
370 cfm
compressor
base 1 base 2
95 psig
104 psig
93 psig
102 psig
91 psig
90 psig
pressure rises 10 psi
if below min flow
VSD cycles but
pressure never
drops below 95
psig setpoint
pressure rises 4 psi
to unload base #2
pressure rises 2 psi
to unload base #2
75
cfm
445
cfm
815
cfm
max
1225
cfm
Flow
decreasing
pressure pressure pressure
upstream drop across downstream
Flow (cfm) of dryer (psi) dryer (psi) of dryer (psi)
494 93.5/96.5 5.0 88.5/91.5
99 93.5/96.5 0.2 93.3/96.3
98 (with 65.2/100.0 0.05 65.1/99.9
120 gal
storage)
Pressure swing with local control signal, a 5-psi rise in
pressure to the stop point and a 20-sec. blowdown time
VSD compressor
with 80% turndown,
95-psi setpoint,
494-cfm
full load capacity.
Total
system
storage
dryer
Pressure swing within
turndown range is 7.8
psi. With 120 gal. of
system storage, the
pressure swing below
the turndown
range is 34.8 psi.
Power versus capacity
% Capacity (cfm)
100%
80%
60%
40%
20%
10%
0%
0% 60% 10% 20% 30% 40% 50% 70% 80% 90% 100%
90%
70%
50%
30%
%

P
o
w
e
r

(
k
W
)
VSD
Modulation
Load/unload
VSD II
Power versus capacity
Capacity (cfm)
100%
80%
60%
40%
20%
10%
0%
0 300 50 100 150 200 250 350 400 450 500
90%
70%
50%
30%
%

P
o
w
e
r

(
k
W
)
VSD
Modulation
Load/unload
VSD II
Compressor efficiency curves
cfm
35
31
27
23
19
17
15
0 300 50 100 150 200 250 350 400 450 500
33
29
25
21
E
f
f
i
c
i
e
n
c
y

(
k
W
/
1
0
0

c
f
m
)
VSD
Variable
Displacement
Load/unload
VSD II
Compressor efficiency curves
E
f
f
i
c
i
e
n
c
y

(
k
W
/
1
0
0

c
f
m
)
cfm
24
23
22
21
20
19
18
17
250 300 350 400 450 500
VSD
Variable
Displacement
Load/unload
VSD II
Average annual capacity of load/unload compressor (%)
A
n
n
u
a
l

e
n
e
r
g
y

s
a
v
i
n
g
s

(
%
)
45%
40%
35%
20%
15%
10%
5%
0%
5%
30%
25%
77% 14% 25% 40% 53% 67% 80% 86% 100%
70% 20% 10%
0
30% 40% 50% 60% 80% 100%
VSD % annual
savings
90%
Figure 2. To take out the spin in Figure 1, look at percent power
as a function of volumetric fow.
A fair comparison
105 psig
adjustable
range
PID
setpoint
always in trim
VSD-100 hp
500 cfm
compressor
dryer
5 psid
93 psig 91 psig
filter
tank
5 psid 5 psid
dryer dryer
75 hp
370 cfm
compressor
75 hp
370 cfm
compressor
base 1 base 2
95 psig
95 psig
104 psig
93 psig
102 psig
91 psig
90 psig
Header pressure
controlled by VSD
pressure dips to start
#1 base and base #2
75
cfm
485
cfm
885
cfm
max
1225
cfm
Flow
increasing
95 psig
105 psig
adjustable
range
PID
setpoint
always in trim
VSD-100 hp
500 cfm
compressor
dryer
5 psid
filter
tank
5 psid 5 psid
dryer dryer
75 hp
370 cfm
compressor
75 hp
370 cfm
compressor
base 1 base 2
95 psig
104 psig
93 psig
102 psig
91 psig
90 psig
pressure rises 10 psi
if below min flow
VSD cycles but
pressure never
drops below 95
psig setpoint
pressure rises 4 psi
to unload base #2
pressure rises 2 psi
to unload base #2
75
cfm
445
cfm
815
cfm
max
1225
cfm
Flow
decreasing
pressure pressure pressure
upstream drop across downstream
Flow (cfm) of dryer (psi) dryer (psi) of dryer (psi)
494 93.5/96.5 5.0 88.5/91.5
99 93.5/96.5 0.2 93.3/96.3
98 (with 65.2/100.0 0.05 65.1/99.9
120 gal
storage)
Pressure swing with local control signal, a 5-psi rise in
pressure to the stop point and a 20-sec. blowdown time
VSD compressor
with 80% turndown,
95-psi setpoint,
494-cfm
full load capacity.
Total
system
storage
dryer
Pressure swing within
turndown range is 7.8
psi. With 120 gal. of
system storage, the
pressure swing below
the turndown
range is 34.8 psi.
Power versus capacity
% Capacity (cfm)
100%
80%
60%
40%
20%
10%
0%
0% 60% 10% 20% 30% 40% 50% 70% 80% 90% 100%
90%
70%
50%
30%
%

P
o
w
e
r

(
k
W
)
VSD
Modulation
Load/unload
VSD II
Power versus capacity
Capacity (cfm)
100%
80%
60%
40%
20%
10%
0%
0 300 50 100 150 200 250 350 400 450 500
90%
70%
50%
30%
%

P
o
w
e
r

(
k
W
)
VSD
Modulation
Load/unload
VSD II
Compressor efficiency curves
cfm
35
31
27
23
19
17
15
0 300 50 100 150 200 250 350 400 450 500
33
29
25
21
E
f
f
i
c
i
e
n
c
y

(
k
W
/
1
0
0

c
f
m
)
VSD
Variable
Displacement
Load/unload
VSD II
Compressor efficiency curves
E
f
f
i
c
i
e
n
c
y

(
k
W
/
1
0
0

c
f
m
)
cfm
24
23
22
21
20
19
18
17
250 300 350 400 450 500
VSD
Variable
Displacement
Load/unload
VSD II
Average annual capacity of load/unload compressor (%)
A
n
n
u
a
l

e
n
e
r
g
y

s
a
v
i
n
g
s

(
%
)
45%
40%
35%
20%
15%
10%
5%
0%
5%
30%
25%
77% 14% 25% 40% 53% 67% 80% 86% 100%
70% 20% 10%
0
30% 40% 50% 60% 80% 100%
VSD % annual
savings
90%
Figure 1. The spin doctor will show percent power as a
function of percent capacity, which masks the efciency of the
constant-speed compressor.
One-sided view
O 2006 46
To reduce the total annual energy cost in a multi-compressor
system by these large percentages, the VSD compressor supplier
also must congure the system to base load each compressor
except the VSD unit. Of course, you can base load all of your
compressors, except one, without installing a VSD compressor.
Terefore, the only actual savings attributable to the VSD com-
pressor is the dierence between it and the operating mode for
the constant-speed trim compressor. In fact, because its easy
and inexpensive to convert most modulating compressors to
load/unload, the savings attributable to the VSD compressor is
only the dierence between its annual energy cost and that of a
load/unload compressor, not a modulating compressor.
Figure 5 shows the annual savings percentage available by
installing a 125-hp VSD compressor versus a 125-hp constant-
speed compressor. Note that its more expensive to operate a
VSD compressor near full load. Te graph shows that the VSD
compressor wont provide any savings unless the load/unload
compressor operates below 83% of its capacity. In addition, you
can achieve most of the savings shown in the lower capacity
range by replacing the oversized constant-speed compressor
with a smaller unit. Generally, if you expect that the constant-
speed trim compressor will operate above 80% of its capacity,
its the more ecient choice. On the other hand, if the constant-
speed trim compressor is operating below 80% of its capacity,
then replacing it with a VSD compressor will provide additional
savings. However, the question you must ask is, Will installing
a VSD compressor generate enough savings to provide an ac-
ceptable payback? Tis isnt an easy question to answer.
For example, in a multi-compressor system, we want the
VSD compressor to operate in the trim mode. When demand
decreases, we must unload a base-load compressor rather
than stopping the VSD compressor. To prevent the base-load
compressors from short cycling, which can damage them, the
VSD compressors turndown capacity must be equal to or
greater than the capacity of the largest base-load compres-
sor that will unload. Terefore, given an 80% turndown,
the VSD compressor must be a minimum of 1.25 times the
size of the largest base-load compressor that will unload.
If the VSD compressor has only a 50% turndown, then it
must be twice the size of the largest base-load compressor that
will unload. To use a smaller VSD compressor and still pre-
vent short cycling the base-load compressor, you must either
provide sucient storage or reduce the size of the base-load
compressors. As you can see, the payback for a VSD compres-
sor depends on its turndown and system storage volume. In ad-
dition to the VSD compressor cost, you must account for other
costs such as line reactance and a remote transducer.
Two clients decided to install a constant-speed trim com-
pressor rather than a VSD compressor because the simple pay-
.PLANTSERVICES.
MANAGEMENT
Compressors
105 psig
adjustable
range
PID
setpoint
always in trim
VSD-100 hp
500 cfm
compressor
dryer
5 psid
93 psig 91 psig
filter
tank
5 psid 5 psid
dryer dryer
75 hp
370 cfm
compressor
75 hp
370 cfm
compressor
base 1 base 2
95 psig
95 psig
104 psig
93 psig
102 psig
91 psig
90 psig
Header pressure
controlled by VSD
pressure dips to start
#1 base and base #2
75
cfm
485
cfm
885
cfm
max
1225
cfm
Flow
increasing
95 psig
105 psig
adjustable
range
PID
setpoint
always in trim
VSD-100 hp
500 cfm
compressor
dryer
5 psid
filter
tank
5 psid 5 psid
dryer dryer
75 hp
370 cfm
compressor
75 hp
370 cfm
compressor
base 1 base 2
95 psig
104 psig
93 psig
102 psig
91 psig
90 psig
pressure rises 10 psi
if below min flow
VSD cycles but
pressure never
drops below 95
psig setpoint
pressure rises 4 psi
to unload base #2
pressure rises 2 psi
to unload base #2
75
cfm
445
cfm
815
cfm
max
1225
cfm
Flow
decreasing
pressure pressure pressure
upstream drop across downstream
Flow (cfm) of dryer (psi) dryer (psi) of dryer (psi)
494 93.5/96.5 5.0 88.5/91.5
99 93.5/96.5 0.2 93.3/96.3
98 (with 65.2/100.0 0.05 65.1/99.9
120 gal
storage)
Pressure swing with local control signal, a 5-psi rise in
pressure to the stop point and a 20-sec. blowdown time
VSD compressor
with 80% turndown,
95-psi setpoint,
494-cfm
full load capacity.
Total
system
storage
dryer
Pressure swing within
turndown range is 7.8
psi. With 120 gal. of
system storage, the
pressure swing below
the turndown
range is 34.8 psi.
Power versus capacity
% Capacity (cfm)
100%
80%
60%
40%
20%
10%
0%
0% 60% 10% 20% 30% 40% 50% 70% 80% 90% 100%
90%
70%
50%
30%
%

P
o
w
e
r

(
k
W
)
VSD
Modulation
Load/unload
VSD II
Power versus capacity
Capacity (cfm)
100%
80%
60%
40%
20%
10%
0%
0 300 50 100 150 200 250 350 400 450 500
90%
70%
50%
30%
%

P
o
w
e
r

(
k
W
)
VSD
Modulation
Load/unload
VSD II
Compressor efficiency curves
cfm
35
31
27
23
19
17
15
0 300 50 100 150 200 250 350 400 450 500
33
29
25
21
E
f
f
i
c
i
e
n
c
y

(
k
W
/
1
0
0

c
f
m
)
VSD
Variable
Displacement
Load/unload
VSD II
Compressor efficiency curves
E
f
f
i
c
i
e
n
c
y

(
k
W
/
1
0
0

c
f
m
)
cfm
24
23
22
21
20
19
18
17
250 300 350 400 450 500
VSD
Variable
Displacement
Load/unload
VSD II
Average annual capacity of load/unload compressor (%)
A
n
n
u
a
l

e
n
e
r
g
y

s
a
v
i
n
g
s

(
%
)
45%
40%
35%
20%
15%
10%
5%
0%
5%
30%
25%
77% 14% 25% 40% 53% 67% 80% 86% 100%
70% 20% 10%
0
30% 40% 50% 60% 80% 100%
VSD % annual
savings
90%
Figure 3. Efciency as a function of volumetric fow.
Low fow is inefcient
105 psig
adjustable
range
PID
setpoint
always in trim
VSD-100 hp
500 cfm
compressor
dryer
5 psid
93 psig 91 psig
filter
tank
5 psid 5 psid
dryer dryer
75 hp
370 cfm
compressor
75 hp
370 cfm
compressor
base 1 base 2
95 psig
95 psig
104 psig
93 psig
102 psig
91 psig
90 psig
Header pressure
controlled by VSD
pressure dips to start
#1 base and base #2
75
cfm
485
cfm
885
cfm
max
1225
cfm
Flow
increasing
95 psig
105 psig
adjustable
range
PID
setpoint
always in trim
VSD-100 hp
500 cfm
compressor
dryer
5 psid
filter
tank
5 psid 5 psid
dryer dryer
75 hp
370 cfm
compressor
75 hp
370 cfm
compressor
base 1 base 2
95 psig
104 psig
93 psig
102 psig
91 psig
90 psig
pressure rises 10 psi
if below min flow
VSD cycles but
pressure never
drops below 95
psig setpoint
pressure rises 4 psi
to unload base #2
pressure rises 2 psi
to unload base #2
75
cfm
445
cfm
815
cfm
max
1225
cfm
Flow
decreasing
pressure pressure pressure
upstream drop across downstream
Flow (cfm) of dryer (psi) dryer (psi) of dryer (psi)
494 93.5/96.5 5.0 88.5/91.5
99 93.5/96.5 0.2 93.3/96.3
98 (with 65.2/100.0 0.05 65.1/99.9
120 gal
storage)
Pressure swing with local control signal, a 5-psi rise in
pressure to the stop point and a 20-sec. blowdown time
VSD compressor
with 80% turndown,
95-psi setpoint,
494-cfm
full load capacity.
Total
system
storage
dryer
Pressure swing within
turndown range is 7.8
psi. With 120 gal. of
system storage, the
pressure swing below
the turndown
range is 34.8 psi.
Power versus capacity
% Capacity (cfm)
100%
80%
60%
40%
20%
10%
0%
0% 60% 10% 20% 30% 40% 50% 70% 80% 90% 100%
90%
70%
50%
30%
%

P
o
w
e
r

(
k
W
)
VSD
Modulation
Load/unload
VSD II
Power versus capacity
Capacity (cfm)
100%
80%
60%
40%
20%
10%
0%
0 300 50 100 150 200 250 350 400 450 500
90%
70%
50%
30%
%

P
o
w
e
r

(
k
W
)
VSD
Modulation
Load/unload
VSD II
Compressor efficiency curves
cfm
35
31
27
23
19
17
15
0 300 50 100 150 200 250 350 400 450 500
33
29
25
21
E
f
f
i
c
i
e
n
c
y

(
k
W
/
1
0
0

c
f
m
)
VSD
Variable
Displacement
Load/unload
VSD II
Compressor efficiency curves
E
f
f
i
c
i
e
n
c
y

(
k
W
/
1
0
0

c
f
m
)
cfm
24
23
22
21
20
19
18
17
250 300 350 400 450 500
VSD
Variable
Displacement
Load/unload
VSD II
Average annual capacity of load/unload compressor (%)
A
n
n
u
a
l

e
n
e
r
g
y

s
a
v
i
n
g
s

(
%
)
45%
40%
35%
20%
15%
10%
5%
0%
5%
30%
25%
77% 14% 25% 40% 53% 67% 80% 86% 100%
70% 20% 10%
0
30% 40% 50% 60% 80% 100%
VSD % annual
savings
90%
Figure 4. Efciency as a function of volumetric fow
(expanded detail).
Zoom in on the crossover point
O 2006 47
105 psig
adjustable
range
PID
setpoint
always in trim
VSD-100 hp
500 cfm
compressor
dryer
5 psid
93 psig 91 psig
filter
tank
5 psid 5 psid
dryer dryer
75 hp
370 cfm
compressor
75 hp
370 cfm
compressor
base 1 base 2
95 psig
95 psig
104 psig
93 psig
102 psig
91 psig
90 psig
Header pressure
controlled by VSD
pressure dips to start
#1 base and base #2
75
cfm
485
cfm
885
cfm
max
1225
cfm
Flow
increasing
95 psig
105 psig
adjustable
range
PID
setpoint
always in trim
VSD-100 hp
500 cfm
compressor
dryer
5 psid
filter
tank
5 psid 5 psid
dryer dryer
75 hp
370 cfm
compressor
75 hp
370 cfm
compressor
base 1 base 2
95 psig
104 psig
93 psig
102 psig
91 psig
90 psig
pressure rises 10 psi
if below min flow
VSD cycles but
pressure never
drops below 95
psig setpoint
pressure rises 4 psi
to unload base #2
pressure rises 2 psi
to unload base #2
75
cfm
445
cfm
815
cfm
max
1225
cfm
Flow
decreasing
pressure pressure pressure
upstream drop across downstream
Flow (cfm) of dryer (psi) dryer (psi) of dryer (psi)
494 93.5/96.5 5.0 88.5/91.5
99 93.5/96.5 0.2 93.3/96.3
98 (with 65.2/100.0 0.05 65.1/99.9
120 gal
storage)
Pressure swing with local control signal, a 5-psi rise in
pressure to the stop point and a 20-sec. blowdown time
VSD compressor
with 80% turndown,
95-psi setpoint,
494-cfm
full load capacity.
Total
system
storage
dryer
Pressure swing within
turndown range is 7.8
psi. With 120 gal. of
system storage, the
pressure swing below
the turndown
range is 34.8 psi.
Power versus capacity
% Capacity (cfm)
100%
80%
60%
40%
20%
10%
0%
0% 60% 10% 20% 30% 40% 50% 70% 80% 90% 100%
90%
70%
50%
30%
%

P
o
w
e
r

(
k
W
)
VSD
Modulation
Load/unload
VSD II
Power versus capacity
Capacity (cfm)
100%
80%
60%
40%
20%
10%
0%
0 300 50 100 150 200 250 350 400 450 500
90%
70%
50%
30%
%

P
o
w
e
r

(
k
W
)
VSD
Modulation
Load/unload
VSD II
Compressor efficiency curves
cfm
35
31
27
23
19
17
15
0 300 50 100 150 200 250 350 400 450 500
33
29
25
21
E
f
f
i
c
i
e
n
c
y

(
k
W
/
1
0
0

c
f
m
)
VSD
Variable
Displacement
Load/unload
VSD II
Compressor efficiency curves
E
f
f
i
c
i
e
n
c
y

(
k
W
/
1
0
0

c
f
m
)
cfm
24
23
22
21
20
19
18
17
250 300 350 400 450 500
VSD
Variable
Displacement
Load/unload
VSD II
Average annual capacity of load/unload compressor (%)
A
n
n
u
a
l

e
n
e
r
g
y

s
a
v
i
n
g
s

(
%
)
45%
40%
35%
20%
15%
10%
5%
0%
5%
30%
25%
77% 14% 25% 40% 53% 67% 80% 86% 100%
70% 20% 10%
0
30% 40% 50% 60% 80% 100%
VSD % annual
savings
90%
Figure 5. Savings as a function of capacity.
There can be a payof
.PLANTSERVICES.
back exceeded a 24-month requirement.
In one case, one compressor manufac-
turer quoted a $22,660 (74% premium)
for a VSD compressor having the same
horsepower rating as the constant-speed
compressor, while another manufac-
turer wanted a $10,785 (30% premium),
but the constant-speed compressor cost
$5,187 more. Te second project re-
quired a VSD compressor larger than
the constant-speed unit. Te premium
for this VSD compressor was $27,720,
or 2.4 times the cost of the constant-
speed compressor.
On the rst project, we project-
ed the VSD compressor to oper-
ate at 50% capacity and provide an
annual savings of approximately
$7,000, giving a simple payback of
47.5 months. On the second project,
operating at 70% capacity, the VSD
compressor had a simple payback of
83.1 months. Both projects involved
retrotting existing systems.
When we design new systems,
we often nd that installing appro-
priately sized constant-speed com-
pressors provides the same or lower
annual energy costs than a system
containing a VSD compressor, and
at a lower capital cost.
Part 2 of this two-part article will
appear in the November 2006 issue. It
will discuss power factor, energy analysis,
turndown, horsepower, storage and con-
trol schemes.
Chris E. Beals is president of Air System Man-
agement, Inc., in Denver, Colo. Contact him
at cbeals@earthlink.net or (303) 771-4839.
Figures: Air System Management, Inc.
MANAGEMENT
Compressors
O 2006 48
CA as a process variable Treat compressed air as a process variable
Speed-regulated drives Drive down the cost of compressed air
Control systems Control options in rotary screw compressors
Condensate What does Mother Nature say about compressor condensate rates?
Flow monitoring and zoning Building a case for better air
For more, search www.plantservices.com using the keywords VFD, audit and storage.
More resources at www.PlantServices.com/thismonth
The payback for a VSD
compressor depends
on its turndown and
system storage volume.
Part 1 of this two-part article appeared in the October, 2006
issue. It discussed dierent ways to quantify compressor eciency
and operation schemes that can lead to energy savings.
M
uch of the VSD compressors popularity is a
result of marketing spin. Often, free audits ex-
aggerate paybacks by including savings the end
user can achieve without purchasing a VSD compressor. Te
marketers also fail to provide the necessary engineering de-
tails required to maximize the purported benets. Its time
to eliminate the spin and point out the limitations and engi-
neering required to derive the benets of VSD.
Power factor
Te spinners claim there are no penalties for spikes or low
power factor. In the context of a compressor, this statement
is true. Te VFD provides a soft start similar to a Wye-
Delta or electronic starter and the power factor for a VSD
compressor motor is close to unity over the complete speed
range. A unity power factor is a benet and it can reduce
energy costs if the VSD motor constitutes a signicant por-
tion of the plants motor load. On the other hand, if the local
utility doesnt have a power factor penalty or the VSD com-
pressor constitutes only an insignicant portion of the mo-
tor load, then a unity power factor oers little cost benet.
Free energy analyses
Te spinners oer free energy audits. Te old adage, you
get what you pay for, applies here. Many compressor manu-
facturers gear free audits to get their sales personnel into a
plant to sell equipment. Free air audits seldom address pres-
sure drop, piping, air quality, risk of an outage or demand-
side issues. Nor do they oer any non-equipment solutions.
Te whole process is automated so it produces an equip-
ment quote for the sales personnel regardless of expertise or
experience in the industry. When clients ask us to review
these free audit reports, we generally nd that the salesper-
son has misinterpreted the data, the proposals are incom-
plete, and plant personnel can achieve the majority of the
projected savings without purchasing any compressor.
Turndown goes up and down
At a 100-psi operating pressure, many lubricated VSD com-
pressors can regulate capacity from 100% down to 15% before
.PLANTSERVICES.












u
n
w
i
n
d
i
n
g

t
h
e

s
p
i
n



















































(
P
a
r
t

2
)
N 2006 57
105 psig
Adjustable
range
PID
setpoint
95 psig
104 psig
93 psig
102 psig
91 psig
90 psig
93 psig 91 psig
always in trim
VSD-100 hp
500 cfm
compressor
Dryer
5 psid filter
tank
5 psid 5 psid
Dryer Dryer
75 hp
370 cfm
compressor
75 hp
370 cfm
compressor
base 1 base 2
95 psig
Header pressure
controlled by VSD
Pressure dips to start
#1 base and base #2
75 cfm
485
cfm
885
cfm
max
1225
cfm
Flow
increasing
105 psig
Adjustable
range
PID
setpoint
95 psig
104 psig
93 psig
102 psig
91 psig
90 psig
VSD cycles but
pressure never
drops below 95
psig setpoint
75 cfm
445
cfm
815
cfm
max
1225
cfm
Flow
decreasing
95 psig
always in trim
VSD-100 hp
500 cfm
compressor
Dryer
5 psid filter
tank
5 psid 5 psid
Dryer Dryer
75 hp
370 cfm
compressor
75 hp
370 cfm
compressor
base 1 base 2
Pressure rises 10 psi
if below min flow
Pressure rises 4 psi
to unload base #2
Pressure rises 2 psi
to unload base #1
Pressure Pressure Pressure
upstream drop across downstream
Flow (cfm) of dryer (psi) dryer (psi) of dryer (psi)
494 93.5/96.5 5.0 88.5/91.5
99 93.5/96.5 0.2 93.3/96.3
98 (with 65.2/100.0 0.05 65.1/99.9
120 gal
storage)
Pressure swing with local control signal, a 5-psi rise in
pressure to the stop point and a 20-sec. blowdown time
VSD compressor
with 80% turndown,
95-psi setpoint,
494-cfm
full load capacity.
Total
system
storage
Dryer
control signal
Pressure swing within
turndown range is 7.8
psi. With 120 gal. of
system storage, the
pressure swing below
the turndown
range is 34.8 psi.
Power versus capacity
% Capacity (cfm)
100%
80%
60%
40%
20%
10%
0%
0% 60% 10% 20% 30% 40% 50% 70% 80% 90% 100%
90%
70%
50%
30%
%

P
o
w
e
r

(
k
W
)
VSD
Modulation
Load/unload
VSD II
Power versus capacity
Capacity (cfm)
100%
80%
60%
40%
20%
10%
0%
0 300 50 100 150 200 250 350 400 450 500
90%
70%
50%
30%
%

P
o
w
e
r

(
k
W
)
VSD
Modulation
Load/unload
VSD II
Compressor efficiency curves
cfm
35
31
27
23
19
17
15
0 300 50 100 150 200 250 350 400 450 500
33
29
25
21
E
f
f
i
c
i
e
n
c
y

(
k
W
/
1
0
0

c
f
m
)
VSD
Variable
Displacement
Load/unload
VSD II
Compressor efficiency curves
E
f
f
i
c
i
e
n
c
y

(
k
W
/
1
0
0

c
f
m
)
cfm
24
23
22
21
20
19
18
17
250 300 350 400 450 500
VSD
Variable
Displacement
Load/unload
VSD II
Average annual capacity of load/unload compressor (%)
A
n
n
u
a
l

e
n
e
r
g
y

s
a
v
i
n
g
s

(
%
)
45%
40%
35%
20%
15%
10%
5%
0%
5%
30%
25%
77% 14% 25% 40% 53% 67% 80% 86% 100%
70% 20% 10%
0
30% 40% 50% 60% 80% 100%
VSD % annual
savings
90%
Table 1. Operating a VSD compressor outside of its turndown
range can promote large pressure swings.
Turndown range
.PLANTSERVICES.
they must operate in either start-stop or load-unload mode.
Spinner literature suggests that all VSD compressors have the
same turndown capability, but the fact is that at 100 psi, some
can only regulate output capacity from 100% down to 55%.
In addition, turndown range decreases as discharge pres-
sure increases. For example, at 125 psi, the turndown can be
as little as 40%, from 100% capacity to 60% capacity.
Turndown range is important because a smaller turndown
decreases the savings, increases the size and frequency of
pressure swings, and can result in short cycling. Or, it can
increase the cost to integrate the compressor into the system
properly. In summary, turndown varies signicantly among
compressor manufacturers, models and pressure settings.
Horsepower: the nameplate game
Te size of a VSD compressor can be confusing because
spinners often play games with the motor nameplate data.
For example, a motor might be a 125-hp unit with a 1.15 ser-
vice factor or a 100-hp motor with a 1.4 service factor. Te
motor manufacturer can nameplate the motor either way.
Tis practice has resulted in end users installing undersized
electrical circuit breakers. Its important to determine the
actual power applied at full load at the operating pressure to
understand eciency and installation requirements.
Constant pressure varies
Te spinners claim that VSD compressors hold pressure
within 1.5 psi. Te truth is that VSD compressors can do
that only at their discharge over their turndown range. You
must properly integrate the unit into your system if you want
to prevent larger swings.
In one example, a VSD compressor requires a signal from
a pressure transducer located downstream of the cleanup
equipment if it is to prevent the cleanup equipment from
increasing system pressure swings. If the control signal is on
the compressor, the smallest swing possible is 7.8 psi within
the VSD compressors turndown range. To reduce the pres-
sure swings, we can either oversize the dryer or just move
the control signal to the downstream side of the cleanup
equipment, which can reduce the maximum pressure swing,
within the compressors turndown range, to 3.0 psi.
If youre operating within the turndown range, storage
isnt required, except to ride through sizeable system events.
As Table 1 shows, you cant prevent larger pressure swings
from occurring when the unit operates outside its turndown
range or when base-load compressors load and unload.
You still need enough storage
Adding sucient storage minimizes the pressure swings
that occur when the compressor operates below minimum
turndown or when base-load compressors load and unload.
Understanding why swings occur requires an explanation of
how VSD compressors operate below minimum speed.
Consider a single-VSD-compressor system that stops when
demand is less than its minimum capacity. As demand de-
creases, the compressor reduces motor speed to its minimum.
If demand continues to fall, the pressure rises to the stop
MANAGEMENT
Compressors
N 2006 58
105 psig
Adjustable
range
PID
setpoint
95 psig
104 psig
93 psig
102 psig
91 psig
90 psig
93 psig 91 psig
always in trim
VSD-100 hp
500 cfm
compressor
Dryer
5 psid filter
tank
5 psid 5 psid
Dryer Dryer
75 hp
370 cfm
compressor
75 hp
370 cfm
compressor
base 1 base 2
95 psig
Header pressure
controlled by VSD
Pressure dips to start
#1 base and base #2
75 cfm
485
cfm
885
cfm
max
1225
cfm
Flow
increasing
105 psig
Adjustable
range
PID
setpoint
95 psig
104 psig
93 psig
102 psig
91 psig
90 psig
VSD cycles but
pressure never
drops below 95
psig setpoint
75 cfm
445
cfm
815
cfm
max
1225
cfm
Flow
decreasing
95 psig
always in trim
VSD-100 hp
500 cfm
compressor
Dryer
5 psid filter
tank
5 psid 5 psid
Dryer Dryer
75 hp
370 cfm
compressor
75 hp
370 cfm
compressor
base 1 base 2
Pressure rises 10 psi
if below min flow
Pressure rises 4 psi
to unload base #2
Pressure rises 2 psi
to unload base #1
Pressure Pressure Pressure
upstream drop across downstream
Flow (cfm) of dryer (psi) dryer (psi) of dryer (psi)
494 93.5/96.5 5.0 88.5/91.5
99 93.5/96.5 0.2 93.3/96.3
98 (with 65.2/100.0 0.05 65.1/99.9
120 gal
storage)
Pressure swing with local control signal, a 5-psi rise in
pressure to the stop point and a 20-sec. blowdown time
VSD compressor
with 80% turndown,
95-psi setpoint,
494-cfm
full load capacity.
Total
system
storage
Dryer
control signal
Pressure swing within
turndown range is 7.8
psi. With 120 gal. of
system storage, the
pressure swing below
the turndown
range is 34.8 psi.
Power versus capacity
% Capacity (cfm)
100%
80%
60%
40%
20%
10%
0%
0% 60% 10% 20% 30% 40% 50% 70% 80% 90% 100%
90%
70%
50%
30%
%

P
o
w
e
r

(
k
W
)
VSD
Modulation
Load/unload
VSD II
Power versus capacity
Capacity (cfm)
100%
80%
60%
40%
20%
10%
0%
0 300 50 100 150 200 250 350 400 450 500
90%
70%
50%
30%
%

P
o
w
e
r

(
k
W
)
VSD
Modulation
Load/unload
VSD II
Compressor efficiency curves
cfm
35
31
27
23
19
17
15
0 300 50 100 150 200 250 350 400 450 500
33
29
25
21
E
f
f
i
c
i
e
n
c
y

(
k
W
/
1
0
0

c
f
m
)
VSD
Variable
Displacement
Load/unload
VSD II
Compressor efficiency curves
E
f
f
i
c
i
e
n
c
y

(
k
W
/
1
0
0

c
f
m
)
cfm
24
23
22
21
20
19
18
17
250 300 350 400 450 500
VSD
Variable
Displacement
Load/unload
VSD II
Average annual capacity of load/unload compressor (%)
A
n
n
u
a
l

e
n
e
r
g
y

s
a
v
i
n
g
s

(
%
)
45%
40%
35%
20%
15%
10%
5%
0%
5%
30%
25%
77% 14% 25% 40% 53% 67% 80% 86% 100%
70% 20% 10%
0
30% 40% 50% 60% 80% 100%
VSD % annual
savings
90%
Figure 1. An integrated VSD compressor in a multi-compressor system (courtesy Dean Smith at airmanagement@mindspring.com).
The left sides represents increasing demand, the right side represents falling flow demand.
Optimizing the whole system
.PLANTSERVICES.
point, which is normally set 5 psi above
the target pressure, and the VSD com-
pressor shuts down.
Some lubricated VSD compressors
blow down the sump when they stop.
When the pressure drops to the target
pressure, the compressor starts immedi-
ately, but the VFD needs approximately
10 sec. to ramp up to minimum speed.
If the VSD compressor doesnt
blow down when it stops and pressure
drops to the target pressure, a blow-
down valve then opens and the sump
pressure must drop to 55 psi or 60 psi
before the VSD compressor starts.
Te blowdown timer is adjustable so
the startup requires between 15 sec.
and 25 sec. Te size of the storage
required to minimize the pressure
swings is directly proportional to
startup time and inversely propor-
tional to turndown.
Compressor manufacturers recom-
mend 1 gal. to 2 gal. of storage per cfm
for VSD compressors, but this means
theres a dierent guideline for each
compressor type. To eliminate this prob-
lem, use the following rule for all rotary-
screw compressor operating modes:
Given a 10-psi dierential, storage
must equal 5 gal. to 10 gal. per cfm of that
portion of the trim compressors capacity
operating load/unload or start/stop.
Until recently, weve always rec-
ommended 10 gal. per cfm, but re-
ceiver prices require us to consider
5 gal. per cfm, the minimum required
to prevent short cycling. For example,
if the 494-cfm compressor in the table
were a load/unload compressor, wed
recommend 2,500 gal. to 5,000 gal. of
storage. Te reason is that when the
compressor unloads, it unloads 100%
of its capacity. However, because the
compressor is a VSD compressor hav-
ing an 80% turndown, it only unloads
98 cfm of the compressors capacity
when it starts and stops, so we would
recommend 500 gal. to 1,000 gal. of
storage. Tis guideline doesnt con-
sider system events such as the failure
of the largest compressor.
In a perfect system having multiple
base-load rotary screw compressors
operating in load/unload mode and a
VSD trim compressor, the VSD unit
would never operate below its mini-
mum speed and the capacity of the
base-load compressors, which would
be less than the turndown capacity
of the VSD compressor. Tis design
points out the importance of turn-
down on VSD compressors.
To prevent the VSD compressor
from starting and stopping when, at
times, the VSD compressor is the only
unit operating, the low load demand
must exceed the minimum speed ca-
pacity of the VSD compressor. Tis
minimizes storage but increases capital
costs because it requires at least one ad-
ditional base-load compressor. On the
other hand, if we use base-load com-
pressors with a capacity larger than the
turndown capacity of the VSD com-
pressor, we might short cycle the base-
load compressor unless we add storage.
For example, assume the base-load
compressor capacity is 500 cfm and
the capacity and turndown of the VSD
compressor are 470 cfm and 80%, re-
spectively. When the base-load com-
pressor unloads, the VSD compressor
can supply only 376 cfm (470 cfm x
0.8). Tis means that demand exceeds
the online capacity by 124 cfm. Based
on a minimum cycle time of 90 sec.,
which most rotary screw compressor
manufacturers recommend to prevent
short cycling, and a 5-psi dierential,
we must have minimum system stor-
age of 2,045 gal.:
We can increase the cycle time to 3 min.
by increasing the stop pressure 5 psi to
10 psi above the VSD compressor target
setpoint or by doubling the storage.
You need storage for a VSD com-
pressor in a system having multiple
compressors if you want to prevent
pressure sags during base-load unit
startup. Also, storage prevents short
cycling the base-load compressors,
which can occur when their capacity
exceeds the turndown capacity of the
VSD compressor.
VSD vs. pressure/fow controller
Spin doctors often equate a VSD pres-
sure control to a pressure/ow control-
ler with storage. But, we must point out
that a VSD compressor responds to all
events using brute strength horsepower
and cant keep the next compressor from
starting. On the other hand, a pressure/
ow controller with sucient storage
responds to events with stored air, which
can prevent the startup of the next com-
pressor while maintaining the system
pressure within 1% of its setpoint.
Integrating the VSD unit
Figure 1 shows how to stage multiple
base-load compressors with a VSD
compressor to minimize pressure uc-
MANAGEMENT
Compressors
The spinners
seldom mention our
favorite uses for VSD
compressors.
N 2006 59
CA as a process variable Treat compressed air as a process variable
Speed-regulated drives Drive down the cost of compressed air
Control systems Control options in rotary screw compressors
Condensate What does Mother Nature say about compressor condensate rates?
Flow monitoring and zoning Building a case for better air
For more, search www.plantservices.com using the keywords VFD, audit and storage.
More resources at www.PlantServices.com/thismonth
x 14.7 x 7.48
124 cfm x 90 sec./60 sec.
5 psi x 2
PS0611_57_61_MGT_Cmp_Reflow.indd59 59 2/21/07 2:20:14 PM
.PLANTSERVICES.
tuations. Note that this arrangement
requires:
Dedicated dryers and lters or tighter
control bands
A remote transducer
Keeping the control signal for the
base-load compressors on the package
An adjustable-range stop pressure
setpoint on the VSD compressor
As you add base-load compres-
sors, pressure variation will increase
unless automation controls the base-
load units. While we hope your com-
pressed air system is engineered this
well, the remaining question you
face is, Does the savings from us-
ing a VSD compressor with all this
additional equipment provide an ac-
ceptable payback?
Good uses for VSD
Te spinners seldom mention our fa-
vorite uses for VSD compressors: ex-
tending the turndown range on cen-
trifugal compressors and making more
ecient use of storage located upstream
of a pressure/ow controller. Using a
VSD compressor with networked cen-
trifugal compressors that are capable
of load-sharing signicantly extends
the turndown range of the centrifugal
units while maintaining a relatively
constant pressure.
Automating the base-load com-
pressors and using a VSD compressor
upstream of a pressure/ow controller
maximizes the available dierential in
the control storage, which reduces re-
ceiver size, achieves more dierential
to support system events, or provides
more startup time.
Chris E. Beals is president of Air System Man-
agement, Inc., in Denver. Contact him at
cbeals@earthlink.net and (303) 771-4839.
Figures: Air System Management, Inc.
MANAGEMENT
Compressors
N 2006 60