Chemical Engineering May 2013 - 58

Engineering Practice
compressors offer better operation. For
instance, centrifugal compressors offer
a relatively low head rise for a given
flow change, compared to that of axial
or positive-displacement compressors.
The curve shape is a function of the
impeller geometry (mainly impellerblade
angles), and the process conditions.
Radial impeller blades result in
a (theoretically) near-flat curve (that
is, close to a flat line).
An impeller with a more-backward
angle will create a higher reaction
(from the compressor wheel) and thus
will yield a steeper curve. In general,
centrifugal compressors are also more
reliable compared to other compressors
types.
On any performance curve, the operational
area of a dynamic compressor
(either centrifugal or axial compressor)
is bounded on the left by the
surge line, on the right by the choke
line (stonewall effect), on the top by
the maximum speed line, and on the
bottom by the minimum speed line.
The orientation of the inlet piping
and its influence on the compressor
performance is also an important consideration
during compressor specification.
There should be neither prerotation,
nor anti-rotation, in the flow
suction of the dynamic compressors.
The flow should be free from any random
distortion.
Centrifugal compressors are somewhat
more forgiving in terms of flow
distortion in the inlet stream than
other compressors (such as axial compressors);
however, there are some limits.
Based on the compressor design,
gas velocities and process conditions,
a minimum length of straight pipe is
necessary ahead of any dynamic compressor
inlet. As a rough indication,
for the centrifugal compressors, this
straight pipe length should be around
four to six times the pipe diameter.
When the minimum straight length
of piping cannot be achieved, vaned
elbows or straighteners may be used
instead. However, these should only be
used in very special cases, as the last
solution because they may create additional
operational problems.
Performance testing
There are two different types of performance
tests available for dynamic
compressors (centrifugal
and axial compressors). A
Type-1 performance test is
carried out in the shop by
the vendor, prior to delivery
of the unit, under conditions
that closely match
those anticipated in the
actual installation. Specifically,
this test is conducted
using the same process gas
as will be found at the CPI
site [the same gas with
the molecular weight (MW) deviation
below ±2%]. Generally, permissible deviations
on pressures, temperatures,
compressor speed and capacity are
below ±4-8%.
By comparison, the Type-2 performance
test is completely different.
The Type-2 test permits the use of a
substitute test gas and allows for extensive
deviations between the shop
test conditions and the specified operating
conditions at CPI site. There
are only a few limits on some essential
gas dynamic parameters of the test
conditions compared to the specified
operating conditions. For a Type-2
test, the following limits are usually
considered:
* The volume ratio and the flow coefficient
are maintained within ±5%.
* The machine Mach number is maintained
within ±0.1 deviation
* The Reynolds number of the Type-2
test arrangement should be within
0.1 to 10 times of the Reynolds
number of the expected operating
condition
When using the Type-2 test, the test
speed, capacity, mass flow, pressures,
temperatures, compressor power, and
other operational details are often totally
different from the specified operating
condition speed.
For the Type-2 test, an alternative
gas (generally an inert gas) should
be selected - preferably, one that
does not lead to an excessive power
or a high discharge temperature and
is readily and cheaply available. The
substitute gases that are typically
used are air, nitrogen, carbon dioxide,
helium, or mixtures of these gases.
The safe operating-speed range, the
critical speeds, the maximum allowable
pressures, the allowable temperatures
and other machine limits are
56 CHEMICAL ENGINEERING WWW.CHE.COM MAY 2013
FIGURE 2. Shown here is an example of a compact
lubrication skid for a shop test-stand facility
considered when establishing suitable
test conditions.
As noted, the ASME-PTC-10 Type-2
test allows for considerable deviations
in the test conditions. For example,
a natural gas (MW=16) compressor
can be performance tested (Type-2
test) using CO2 (MW=44) with around
half the inlet flow, approximately 20%
of the mass flow, around 50% of the
speed, approximately 6% of the absorbed
power and much less pressures
(even less than 10%) compared to the
specified operating conditions.
The idea behind the Type-2 performance
test is to allow a test to be
performed by the vendor using different
gas and flow details (while still
maintaining the major fluid characteristics
within certain limits), and to
use available knowledge and formulations
of fluid mechanics to estimate
the compressor performance in the
specified operating conditions. The
flow patterns of a dynamic compressor
(centrifugal or axial compressor) are
mainly a function of the major fluid
characteristics, such as the volume
ratio, the flow coefficient, the machine
Mach number and Reynolds number.
Reynolds number. Fluid friction is
mainly affected by the machine Reynolds
number. In a Type-2 test, the
Reynolds number during the test may
deviate from that of the specified operating
condition; however, it should
stay within a range (for instance,
within 0.1 to 10 times of the Reynolds
number at the expected operating condition)
to keep the governing friction
model (and formulations) relevant.
Mach number. The machine Mach
number is a measure of the maximum
compressor capacity and is mainly associated
with the " stonewall " effects
(That is, the operation of a compres
http://WWW.CHE.COM
Chemical Engineering May 2013

Table of Contents for the Digital Edition of Chemical Engineering May 2013
