Table of Contents

THE CONSEQUENCES OF COMPRESSOR OPERATION IN OVERLOAD

by
James M. Sorokes
Manager of Development Engineering

Harry F. Miller
Product Manager - Marketing of Turbo Products
and
Jay M. Koch
Manager of Aero/Thermo Design Engineering
Dresser-Rand Company
Olean, New York

James M. (Jim) Sorokes is Manager of Jay M. Koch is Manager of Aero/Thermo

Development Engineering, with Dresser- Design Engineering, at Dresser-Rand, in
Rand Company, in Olean, New York. He is Olean, New York. He has been employed
involved in all aspects of product there since 1991. Prior to joining Dresser-
development for Dresser-Rand worldwide. Rand, he was employed by Allied Signal
He previously spent 28 years in the Aerospace. He spent 14 years in the
Aerodynamics Group, becoming the Aerodynamics Group, before being
Supervisor of Aerodynamics in 1984 and promoted to his current position in 2005.
being promoted to Manager of During Mr. Koch’s time in the
Aero/Thermo Design Engineering in 2001. Aerodynamics Group, his responsibilities
During Mr. Sorokes’ time in the included the development, design, and analysis of all aerodynamic
Aerodynamics Group, his primary responsibilities included the components of centrifugal compressors. Additionally he was
development, design, and analysis of all aerodynamic components responsible for the development of software used to select and
of centrifugal compressors. His professional interests include: predict centrifugal compressor performance.
aerodynamic design, aeromechanical phenomenon (i.e., rotating Mr. Koch holds a B.S. degree (Aerospace Engineering) from Iowa
stall), and other aspects of large centrifugal compressors. State University. He has authored or coauthored 10 technical papers.
Mr. Sorokes graduated from St. Bonaventure University (1976).
He is a member of AIAA, ASME, and the ASME Turbomachinery
Committee. He has authored or coauthored more than 30 technical ABSTRACT
papers and has instructed seminars and tutorials at Texas A&M
and Dresser-Rand. He currently holds two U.S. Patents and has This work describes the potential consequences associated with
two other patents pending. operating a centrifugal compressor in overload. Nomenclature is
offered to explain what is meant by overload operation, and
methods that are used by original equipment manufacturers
Harry F. Miller is the Product Manager - (OEMs) and end users to define overload limits are presented. The
Marketing of Turbo Products at Dresser- paper also describes the conditions that can lead to overload
Rand, in Olean, New York. His career in operation. Computational fluid dynamics (CFD) results are used to
turbomachinery began 31 years ago with illustrate the forces acting on an impeller when it operates at very
Dresser Clark, and Mr. Miller has held high flow. Finally, this paper suggests considerations that should be
a variety of design engineering and addressed when designing (or selecting) an impeller that could be
marketing positions, most recently, being subjected to extended overload operation.
Manager of Development Engineering and
Leader of the DATUM Development Team. INTRODUCTION
His prior work experience consists of four It is common knowledge that operating centrifugal compressors in
years as a mechanical construction engineer for the Pennsylvania surge can have detrimental effects on the mechanical integrity of
Power & Light Company. parts. Prolonged excursions into surge have caused damage to
Mr. Miller received a BSME degree from Northeastern University, impellers, bearings, seals, and other rotating components. The violent
and an MBA degree from Lehigh University. His areas of expertise nature of surge events makes it necessary to install sophisticated
include turbocompressor and gas turbine design and application. control systems to prevent compressors from operating in surge.
63
 Table of Contents
64 PROCEEDINGS OF THE THIRTY-FIFTH TURBOMACHINERY SYMPOSIUM • 2006

While there are numerous publications in the open literature

addressing the consequences of operating in surge, there are few, if
any, articles on the potential dangers associated with operating a
compressor in overload—the high flow region of a performance
map. In the vast majority of applications, overload operation does
not represent any cause for concern. In fact, many compressors
spend the majority of their life operating in the high flow end of
their performance envelope. Yet, in some applications, overload
operation can be just as damaging as operation in surge.
This paper will discuss the potential consequences of operation
in overload. The primary focus will be the potential damage that
could occur in impellers, as they are typically the most expensive
and difficult component to replace in a centrifugal compressor. The
discussion will begin with definitions of the terms “overload” or
“deep overload.” These terms mean different things to different
people. Therefore, it is important to understand the meaning of
these terms in the context of this paper. The focus will next move to
the types of situations that can lead to operation in overload. This is Figure 2. Setting Overload Limit.
followed by an overview of some “real-world” examples of damage
caused by overload operation. Computational fluid dynamics In the broadest sense, the term “overload” is used to reference
(CFD) and finite element analysis (FEA) results will then be used any condition under which the compressor’s inlet flow exceeds its
to provide a clearer picture of the forces imposed on impellers when design flow rate (Figure 2). However, most compressors regularly
they operate well above their intended design flow rates. operate at flow rates higher than design and most users do not
Suggestions are offered on factors that should be considered when consider this operating in overload. To them, this is normal
designing impellers that may be subjected to prolonged overload operation. Instead, this group deems overload to be operating at
operation. Application of the general guidelines outlined in this any condition that exceeds the safe or recommended flow limit for
discussion will in no way ensure the long-term mechanical integrity the machine. This hazardous region to the extreme right on the
of the impellers. Given the unique factors associated with each performance map is often called “deep overload.”
situation, it is imperative that you seek the guidance and direction End users typically defer to the original equipment manufacturer
from engineering. However, proper application of these guidelines (OEM) to establish an overload operating limit if the OEM
may increase the impellers’ resistance to the forces imposed by believes such a limit is warranted. This is where the difficulties
overload operation that could lead to unwanted consequences. really begin because, unlike surge, compressor OEMs, in general,
Finally, conclusions will be offered on the need to be cognizant of do not specify an overload limit. Because compressor performance
the consequences of overload operation when establishing the drops off rapidly in the overload region, it is frequently assumed
operating envelope for a new compressor application. that an end user will not want to operate the compressor in that
regime or will lack the driver capacity to do so. As will be seen, this
DEFINITIONS can be a risky assumption to make.
Unlike surge, there is no commonly accepted definition The majority of OEMs address the issue of overload operation by
for overload or “operation in overload.” The compressor putting a clause in the operator’s manual stating that off-the-map
aeroperformance maps given in Figures 1 and 2 will be used to operation must be avoided. In the strictest interpretation, this clause
illustrate the various definitions. The maps show the head covers both the low-flow and high-flow ends of a performance map,
coefficient and polytropic efficiency plotted against the normalized i.e., do not operate at flow rates lower or higher than shown on the
inlet flow coefficient. map. Still, surge control systems are installed to ensure no (or
limited) operation in surge. It is somewhat uncommon to find
control systems that limit overload operation. Again, this is quite
often the case because the compressor driver cannot provide
sufficient power to maintain prolonged operation in overload.
On occasion, OEMs will provide an overload limit on the right-
hand side of the performance map (refer to Figure 2). This limit is
set based on a variety of methods that are described in the
following paragraphs. In fact, in rare instances, anti-overload
control systems have been provided that operate based on
algorithms very similar to an anti-surge control system, except that
a discharge throttle valve is utilized instead of a recycle valve.
Percent of Design Flow
One of the simplest methods for establishing the overload limit is
to specify the maximum amount of flow as a percentage of the design
flow. Such limits are typically set based on experience and are a
function of the machine Mach number (U2/A0), gas handled, number
of stages, etc. It is well known that a compressor that runs at high
machine Mach number or is handling high mole weight gases will
have less overall flow range than a compressor that runs at low U2/A0
or handles low mole weight gases (Figure 3). Therefore, the allowable
percentage increase from design flow must vary. For example, one
might impose an overload limit of 120 percent of the compressor
design flow rate for a high mole weight machine and allow as much
Figure 1. Definition of Overload. as 140 percent of design flow for a low mole weight application.
 Table of Contents
THE CONSEQUENCES OF COMPRESSOR OPERATION IN OVERLOAD 65

CONDITIONS LEADING TO OVERLOAD OPERATION

Common circumstances that lead to prolonged operation in
overload are:
• Loss of parallel compressor or compression train,
• Performance degradation (i.e., fouling) within a compressor,
• An undersized compressor (either due to changes in flow
requirements for an existing compressor or misapplication of a new
machine),
• Alternate operating conditions (i.e., summer and winter conditions
in a pipeline application),
• Unanticipated changes in gas characteristics (mole weight, etc.),
• Process upsets.
Parallel Compressors
Figure 3. Variation in Flow Range with Mole Weight. Many facilities have duplicate compressors or trains of
compressors and the process flow is divided equally (or nearly
Impeller Mach Number equally) between these compressors or trains. Each machine in the
Some OEMs limit the overload capacity based on calculated train is sized such that the compressors operate near the middle of
values for impeller inlet relative Mach number. Calculations are their performance map under normal conditions. However, if for
performed with increasingly larger flow rates and the performance some reason (process upset, regular maintenance, etc.) one or more
map is truncated (or “cut off”) when the Mach number in any stage trains are taken offline, plant operators may attempt to make up for
within the compressor exceeds the specified level. Typically, the the missing train(s) by forcing more flow through the remaining
limit is a calculated Mach number of 1.0, the flow rate at which the operational trains. This causes those remaining trains to operate at
impeller throat is choked. However, to be conservative, some significantly higher flow rates, pushing them into the overload
OEMs set the overload limit based on an inlet relative Mach region of their performance maps. Assuming the drivers have
number of 0.96 or lower. sufficient power (and many do not), it would be possible to
It is noteworthy that all OEMs do not use the same methods to operate in the overload region for extended periods of time. The
calculate inlet relative Mach numbers. Further, inlet relative Mach performance would suffer (i.e., low efficiency) but production
number can be calculated at various locations on the impeller leading requirements would be maintained.
edge, i.e., at the shroud, at the mean. Therefore, agreement should be
reached between the end user, contractor, and OEM regarding the Excess Driver Power
exact definition of the Mach number being considered. It is considered normal practice by many users to rate gas turbine
Another approach often employed by end users is to specify a drivers for the highest expected ambient air temperature expected,
maximum allowable inlet relative Mach number at the design or at least an “average” of the historical high temperatures for that
condition for any impeller in a compressor. This method assumes location. This is done so that the compressor and hence the “process
that limiting the inlet relative Mach number at design will ensure a plant” may be operated at close to design output on a hot day, or
sufficient amount of overload capacity. However, this method does throughout the hot season. With typical aeroderivative gas turbines
not ensure that the compressor is not operated in a risky portion of rated for a 120ºF inlet, the power output will increase from 25
the performance map. percent to 75 percent when the ambient drops to 30ºF. It is also not
uncommon for steam turbines to be rated for 110 percent of the
Drop in Head Coefficient/Efficiency compressor rated power with “minimum” steam conditions, only to
Another popular method for establishing overload limit is based be capable of providing 50 percent or more power when operated at
on the rate of change of head coefficient or efficiency as a function “maximum” steam conditions.
of flow coefficient. This method takes advantage of the rapid drop
in the compressor’s performance in the overload region as seen in Performance Degradation
Figures 1, 2, and 3. Often referred to as “stonewall” (because of its In a multistage compressor, it is possible to drive the latter stages
vertical nature and the fact that it limits compressor flow like a into overload if there is a reduction in the performance of the
stone wall), the high flow end of the map has a highly negative earlier stages or if there is a decrease in the mole weight of the gas
slope. Some OEMs will limit the overload capacity to the flow being compressed. With regard to the former, the impellers or
rate corresponding to an “X” percent drop in head coefficient or stages in a compressor are aerodynamically matched such that
efficiency for a 1 percent increase in flow. Common values of “X” when all are functioning correctly, each operates near the middle of
fall between 10 percent and 20 percent. its respective performance map. However, if the performance of
An alternate form of this approach limits the overload capacity the early impellers or stages in the compressor begins to degrade,
to the flow rate at which the efficiency or head coefficient is those impellers or stages will not provide the expected volume
a given percentage of the design flow level, i.e., 15 percent, 20 reduction. Therefore, any subsequent stages will be forced to
percent, or 30 percent of the design efficiency or head coefficient. “swallow” more flow than expected. If the performance
Beyond this point, the performance would be so low as to be degradation is substantial, the latter stages will operate in overload.
unusable in most processes. Again, assuming there is sufficient driver capability, it would be
possible for the stages to operate in overload for prolonged periods.
Section Summary
Again, the horsepower consumption would increase
Clearly, it is important that the OEM and end user have a significantly but production could be maintained.
common understanding of how the overload limit was established The same scenario occurs if there is a change in the gas
and the potential implications of violating this limit. The latter will characteristics such that the volume reduction or overall flow range
be addressed in subsequent sections herein. of the individual stages within the compressor drops. This will
 Table of Contents
66 PROCEEDINGS OF THE THIRTY-FIFTH TURBOMACHINERY SYMPOSIUM • 2006

cause the impellers or stages to operate at successively higher compressor had operated in overload for long periods. In this case,
capacity relative to design until the latter stages are operating at or the dynamic forces because of incidence caused a portion of the
near choke. impeller blade’s leading edge to fracture and separate from the
impeller. The loss of material produced an unbalance on the rotor
Undersized Compressor/Alternate and the compressor had to be taken offline for repairs.
Operating Conditions/Unexpected Changes
At times, compressors are purposely or inadvertently undersized
in the selection process. For example, an end user may anticipate a
sizeable reduction in flow rate during the life of the compressor.
This user may choose, in the beginning, to operate the compressor
at the high-flow end of its map, knowing that in coming years the
flow rate will be reduced and the compressor will operate nearer
the center of its map.
Occasionally, end users purchase equipment before finalizing
their operating requirements. When the compressors are put into
operation, they discover that the inlet conditions and/or gas
mixtures are not as expected or that their production requirements
exceed those originally projected. These changes can cause the
compressor to operate at flow rates that are much higher than
anticipated. In short, the compressor operates in the overload
region of its performance map.
Finally, process conditions may also change during the life of a
compressor, causing the flow rate through the machine to increase.
The end user may lack sufficient funding to revamp or upgrade the Figure 4. Schematic of Blade Leading Edge Fracture
compressor, choosing instead to operate existing equipment in
overload despite the low performance. While the mechanisms that ultimately led to the impeller
fractures were different (more on this in the discussion to follow),
generally speaking, the root cause for the fractures was the same—
POTENTIAL CONSEQUENCES operation in overload.
OF OVERLOAD OPERATION
Two examples are presented to illustrate the potential DESCRIPTION OF FORCES
consequences of overload operation. In both cases, the compressors Although the forces resulting from operation in overload have a
were known to have operated for extended periods in the overload similar root cause to those found in surge or stall, the true nature of
region. Note: Before proceeding further, it must be noted that end those forces is radically different. The violent forces associated
users and OEMs are often very reluctant to provide details on with surge are typically very low frequency (6 Hz or less) and
difficulties resulting from off-design operation. End users do not result from the flow reversal through the compressor when the
want details of their operating practices made public and OEMs impeller or impellers can no longer overcome the downstream
typically do not want any suggestion that their products are static pressure. When in surge, the inability of the compressor to
anything but perfect. For that reason, limited details will be overcome such pressure is directly related to the increase in
provided on the two sample cases. incidence or other losses in the compressor components, i.e.,
The first example is taken from a high-pressure gas reinjection impellers, diffusers, return channels, etc. That is, as the flow rate in
compressor. The compressor experienced repeated impeller a compressor is reduced from design toward surge, the angle at
fractures. The welded impeller was a low flow coefficient design, which the flow impinges on the bladed or vaned components
having a very short blade height relative to the impeller diameter increases, thus increasing the incidence (or delta angle between the
(low b/r). Essentially, the impeller disk was fracturing at or around flow angle and the blade/vane angle) (Figure 5). At some point, the
the leading edge region of the impeller. In the worst case, the inner incidence angles lead to flow separation or other anomalies that
portion of the impeller disk separated from the outer portion. cause very high losses within all of the compressor components,
Following an investigation by the OEM and end user, it was making it impossible for the compressor to overcome the
reported that the impeller had been run extensively in the overload downstream pressure. Because flow moves from the region of
region of its performance map when a parallel train was taken higher pressure to a region of lower pressure, the flow reverses
offline. Dynamic forces caused by the combination of high direction and goes backward through the compressor. The resulting
impeller leading edge incidence, and a nonuniform pressure forces on the compressor internals can be destructive.
distribution caused by the downstream discharge volute, were
sufficient to initiate cracks in the impeller, leading to the fractures.
The compressor had to be taken out of service for several days to
allow installation of the spare rotor.
It should be noted that the forces that led to the fractures were
significantly lower at the design flow condition. That is, at design
flow, the dynamic forces due to incidence and the volute pressure field
were not sufficient to initiate the cracking. Further details on this case
can be found in Borer, et al. (1997), and Sorokes, et al. (1998).
The second example is from a compressor processing heavy
hydrocarbons. The subject impeller was a high-flow coefficient
design, implying that the leading edge was quite tall (high b/r).
Blade fractures occurred near the impeller’s leading edge. In a few
of the blades, a portion of the blade broke away as indicated by the
crosshatched area in Figure 4. Again, based on an analysis
conducted by the end user and the OEM, it was found that the Figure 5. Impeller Incidence.
 Table of Contents
THE CONSEQUENCES OF COMPRESSOR OPERATION IN OVERLOAD 67

The basic driver of the forces associated with overload

operation is also incidence. However, the frequency of the force
mechanism tends to be high frequency. Whereas surge
phenomenon has a frequency of 6 Hz or less, the forces
associated with overload tend to be more on the order of blade
passing frequency (1 to 3 kHz), i.e., the number of impeller
blades times compressor running speed. This distinction is
important as one considers the nature of the potentially damaging
forces.
As with surge, when the compressor is operated at flow rates
above design, there is an increase in blade incidence on the
impeller. Eventually, the incidence becomes high enough that, as
in surge, the flow separates from the impeller blade. In or near
surge, this separation occurs on the suction surface; while in
overload, it occurs on the pressure surface of the blade. The flow
separation causes a reduction in the effective throat (or
minimum) area in the impeller, leading to a significant increase
in the flow velocity. If the velocity gets high enough, shock
waves will form and choke will occur. The pressure fields
associated with shock waves are highly dynamic and cause
excessive forces on the impeller blades and walls. These forces
alone could be sufficient to damage an impeller. However, when Figure 6. CFD Results Showing IGV Wakes.
these forces are further exacerbated by other nonuniformities
within the compressor flow path, one can rapidly reach The situation with downstream diffuser vanes is somewhat
conditions that can quickly lead to impeller fractures. Two such different. Rather than interacting with wakes, the excitation forces
situations are described in the following discussion. associated with vaned diffusers are a consequence of the pressure
fields forming around the diffuser vanes because of potential flow
Blade/Vane Interactions effects (Figure 7). Again, as the impeller blades pass the diffuser
vanes they pass through the “lobes” in the pressure distribution.
It is well known that interactions occur between impellers and The resulting variation in pressure imposes a dynamic force on the
stationary vanes upstream (inlet guidevanes or IGVs) and/or impeller that could excite an impeller’s natural frequency. As with
downstream (diffuser vanes). Numerous technical papers have the IGV wakes, the magnitude of the pressure variation surround-
been written on this subject including Kushner (1980), Fisher ing the diffuser vanes will be greater during overload operation
and Inoue (1981), and Eckert (1999). Summarizing, the upstream than at design. This is due in large part to the increased incidence
or downstream vanes are surrounded by fields of varying on the diffuser vanes and probable separation of the flow from the
pressure through which the impellers rotate causing pressure vane pressure surface. The increased velocities resulting from the
fluctuations and, therefore, varying forces on the impeller blades. higher flow rates also increase the kinetic energy or velocity
The frequency of the pressure disturbance is determined by pressure (1/2ρV2). This causes even higher loads across the impeller
multiplying the rotational speed of the impeller by the number blades. As a result, the dynamic forces in overload will be greater
of stationary vanes. Of course, if the upstream component and might be sufficient to initiate a fracture whereas those at design
comprises multiple vane rows, the vane wakes from all of the are not.
upstream vanes must be considered because it may be possible
for the wakes from the first row to propagate through the second
row and reach the impeller.
Problems arise when the frequency of the pulsations align with
a natural frequency in the impeller, thereby exciting the impeller,
resulting in a resonance condition. When the forces are of
sufficient magnitude, it would be possible to initiate a fracture in
the impeller.
Blade/vane interactions during overload operation are of
particular concern because of the high level of energy in the gas
stream. In overload, the higher-than-nominal flow rate causes the
velocities within the aerodynamic components to be higher than
at design. Therefore, the static pressure variation at the exit of the
upstream inlet guide vanes will be less uniform than it is at
design flow. That is, the high core flow velocity in the IGV
passages will cause regions of low static pressure, while the
static pressure in the stagnated (or wake) region immediately
downstream will be very high (Figure 6). If the inlet guide vanes
are sufficiently close to the impeller blade leading edges (as is
often the case with full inducer-style impellers), each blade will
be subjected to this highly nonuniform static pressure field.
Because the magnitude of the pressure variation is higher in
overload than at design, the forces during overload operation
may be sufficient to create problems, even though the forces at
design are not. Figure 7. CFD Results Showing Pressure Field Around LSD Vanes.
 Table of Contents
68 PROCEEDINGS OF THE THIRTY-FIFTH TURBOMACHINERY SYMPOSIUM • 2006

Interaction with Other Pressure Nonuniformities Wake Effects

The forces acting on the impeller during overload operation There is much conjecture as to whether it is possible for an
can be further exacerbated by the presence of other pressure impeller to excite itself into resonance. Although there is no
nonuniformities in the compressor flow path. A prime example of definitive proof that this can occur, it has been theorized that the
this is the nonuniformity caused by a discharge volute or collector. blade wakes shed by an impeller could provide the mechanism for
Again, several technical papers have been published that document self-excitation. Under normal operation, the wake regions are
this nonuniformity, including Hagelstein, et al. (1997), and sufficiently small and of reasonably limited pressure magnitude (i.e.,
Sorokes and Koch (2000). Essentially, an impeller upstream of a the difference between the static pressure in the wake region and the
discharge volute or collector will be operating in a nonuniform static pressure in the core flow is small). Therefore, there is not
circumferential pressure field as seen in Figure 8. The impeller can enough force or “delta pressure” to cause any difficulties. However,
be somewhat shielded from the effect of the volute though the use as in the inlet guide vane, the difference between the static pressure
of vaned diffusers. However, the nonuniformity persists nonetheless. in the core flow and static pressure in the secondary zone and blade
wakes is higher. It is possible that an impeller blade could interact
with the wake shed by the preceding blade (Figure10), resulting in
an excitation at blade passing or other frequency (i.e., number of
impeller blades times the rotational speed).

Figure 10. Impeller Blade Wakes.

Figure 8. Nonuniform Pressure Field Due to Volute. ANALYTICAL METHODS
As reported in Sorokes and Koch (2000), such nonuniformities, Computational fluid dynamics can provide valuable insight into the
whether caused by a volute, compressor inlet, or other driving flow physics and the detrimental forces associated with operation
mechanism, further increase the level of dynamic forces acting on in overload. For example, Borer, et al. (1997), used CFD to investigate
the impeller because the circumferential pressure distribution possible excitation mechanisms in a problematic stage in a
tends to become even more skewed when operating in overload. high-pressure reinjection compressor. Their work showed that a high
The dynamic strains shown in Figure 9 are for an impeller static pressure load was occurring near the leading edge of the impeller
upstream of a volute operating at various flow rates. As can be while operating in overload. The static pressure distribution for the
seen, the strains are highest in the overload region of the impeller is shown in Figure 11. The root cause of the pressure load was
performance map. It should be noted that these strains were high negative incidence on the impeller blade caused by operating at a
obtained during an extensive test program to identify the root high flow rate (i.e., overload). When the impeller was operating nearer
cause of the impeller fracture. design flow, the pressure load was significantly smaller. In short, the
analytical results were helpful in identifying the adverse conditions
within the impeller that contributed to the fractures.

Figure 11. Impeller Static Pressure Distribution During Overload

Figure 9. Variation in Impeller Dynamic Strain with Flow. Operation.
 Table of Contents
THE CONSEQUENCES OF COMPRESSOR OPERATION IN OVERLOAD 69

CFD can also be used to develop pressure loads or net impeller blade
forces that can be applied when conducting finite element analyses to
assess the structural integrity of designs. CFD can also be used to
determine how said forces vary with flow rate. Historically, CFD was
limited to providing steady-state (or time averaged) pressure
distributions that could be applied as pressure loads in FEA studies.
However, in the past decade, advances in CFD have made it possible to
generate unsteady or transient pressure distributions. These unsteady
pressures can be translated into a force at a frequency and imposed as a
boundary condition in an impeller natural frequency analysis. This
offers designers the opportunity to assess possible excitation and
determine if changes are needed to avoid a resonance problem.
High-flow coefficient impellers that may be subjected to overload

defined herein as any impeller having a flow coefficient, φ, of 0.12 or

operation are of particular concern. For clarity, high-flow coefficient is

greater, where φ or phi is an internationally accepted, nondimensional

flow coefficient relating flow, speed, and diameter. By their nature,
such impellers have very tall leading edges, that is, the length of the Figure 13. Static Pressure Load on High Flow Coefficient
blade from hub to shroud at the leading edge is large. The excessive Impeller Operating at Overload.
length makes it possible for the blade to flex or flutter if it is subjected
to a sufficient dynamic force. The impeller blade leading edge is most As can be seen, the variation in pressure loads is much higher
susceptible to flexing because it is the tallest and often the thinnest when the impeller is operating in overload. One can note the wide
portion of the blade. If the magnitude of the vibrations becomes variation in pressure on the pressure side leading edge at overload
sufficiently large, a fracture could occur. Therefore, it is imperative that versus the considerably lower variation at design. Since the
designers understand the magnitude and the frequency of the forces pressure on the suction side of the leading edge is fairly consistent
acting on such impellers. between design and overload, it is also clear that the delta pressure
The above issue is of even greater concern on unshrouded or open suction to pressure surface is higher in overload than at design.
impellers. Without a cover or shroud to help hold the blades in position, Note further that the pressure distribution in Figure 13 bears a
the blades have significantly more freedom to flex or flutter. As a result, resemblance to the schematic of the blade leading edge fracture in
operation near stonewall on open impellers is strongly discouraged. Figure 4. This alone is not definitive evidence that a fracture would
As noted, operation in overload causes high levels of negative occur. However, were the pressure forces on the impeller unsteady
incidence on the impeller blade leading edges. This negative incidence (which they certainly would be), and were the unsteadiness to be at
causes unbalanced pressure forces between the two sides of the blades. a frequency that aligns with the natural frequency of the blade
If these forces become dynamic or unsteady because of the presence of leading edge (i.e., the blade leading edge mode), a blade fracture
upstream IGVs or some other circumferential asymmetry in the flow like the one shown in Figure 4 could occur.
field, the fluctuating pressure forces could be more than sufficient to
induce alternating stresses in the impeller. Assuming the excitation DESIGN OR OPERATIONAL CONSIDERATIONS
force does not change in frequency, and assuming the frequency aligns It is critically important that end users and OEMs discuss the full
with an impeller natural frequency, resonance could occur and range of conditions that a compressor will face in operation. For
ultimately result in the loss of mechanical integrity. example, if the process engineers or end user know that there is a
As an illustration, the pressures acting on the leading edge of a chance for extended operation in overload, the OEM may want to
high-flow coefficient impeller are shown in Figures 12 and 13. The undertake more detailed analyses of the proposed equipment.
analytical results given in Figure 12 are for the impeller operating at its These analyses may include detailed CFD and FEA analyses as
design flow rate while the results in Figure 13 are for overload described above to understand the potential ramifications of
operation. In both cases, the plot on the right is for the pressure surface overload operation. These analyses can be used to:
of the blade, while the plot on the left is for the suction surface. Also,
the pressure scales are consistent among all of the figures. • Establish the maximum safe operating limit,
• Help the end user understand the risks of operating in that flow
regime, or
• Provide guidance to the OEM regarding how to modify
component designs to minimize the risks by increasing the
robustness of the design.
Regarding the latter, it may be possible to change the number of
vanes in adjacent stationary hardware to avoid natural frequency
interference issues. Alternatively, the design or manufacturing
technique used for an impeller could be changed to make it less
susceptible to interference issues. For example, the OEM may
choose to apply more stringent criteria for welds (i.e., weld shape
requirements, more rigorous inspection techniques) or possibly
resort to single-piece impeller fabrications. Of course, a natural
frequency in the speed range is acceptable as long as its response
does not exceed allowable limits imposed by the end user or OEM.
In short, by having full knowledge of the potential modes of
operation, the OEM can undertake efforts to minimize the risks.
One additional example illustrates how this happened on a recent
Figure 12. Static Pressure Load on High Flow Coefficient Impeller compression project. The client, a major oil and gas producer,
Operating at Design Flow. needed more compression on an existing offshore platform to boost
 Table of Contents
70 PROCEEDINGS OF THE THIRTY-FIFTH TURBOMACHINERY SYMPOSIUM • 2006

gas going to shore. Because of the declining pressure nature of the “overload” has many connotations, so it is important that all parties
field, the compressors were designed with both the present and adopt a common definition. Because one person’s “overload”
several future conditions in mind to allow for change-out of internal might well be another’s normal operation, a more rigorous
parts over the life of the field to accommodate the changing definition must be applied when specifying compressor flow range
operating conditions. Because the application required very high requirements or discussing how the compressor is being operated
horsepower drivers (>40 MW), and because the early years of in production. It might also be possible to adopt an industrywide
operation had the highest inlet pressures, only a few stages standard that defines “overload” as operating a compressor at
(impellers) were required to meet the specified discharge pressure. flow rates that exceed the maximum flow rates shown on the
This resulted in a fairly high horsepower per stage relative to the performance map provided by the OEM. This would put the onus
end user’s and OEM’s experience for this particular size unit. In on both the OEM to provide an accurate prediction of overload
contrast, operation in future years at much lower inlet pressures capability and the end user to properly assess their need to operate
would require more stages (impellers) to meet the same desired at such high flow rates.
discharge pressure. Therefore, in the later years of operation, the As seen herein, in some reported cases, prolonged operation in
horsepower per stage would drop into more a comfortable range. overload can lead to impeller fractures. Overload operation can
The end user also desired that if one unit had to be taken out of also exacerbate structural natural frequency interference issues that
service, the remaining units would need to accept more inlet flow; may exist within a compressor flow path. Forces and/or pressure
that is, operated in overload, a flow rate 30 percent to 50 percent nonuniformities tend to be greater when operating at flow rates
higher than the design flow rate. Because the stage power was high much higher than design. Such increases are caused in large part by
at the normal design point, the OEM was concerned about the the increased incidence levels on impeller blades or adjacent
prospect of even higher powers being absorbed in an off-design stationary vanes (i.e., vaned diffusers).
mode. When the end user consulted the OEM about the potential Advanced analytical tools such as computational fluid dynamics
operation at higher flows, the OEM responded with an engineering or finite element analysis help quantify the magnitude of the forces
study to determine the magnitude of the dynamic stress. The study associated with overload operation. Such analyses can also be used
included an innovative combination of transient CFD and FEA. to mitigate risk. Designs can be modified to reduce the potential for
The resulting analyses indicated that the original impeller geometry harmful interferences, or operating limits can be derived so as to
was acceptable but with relatively low safety margins. The OEM avoid risky portions of the performance map. However, the most
then determined that the calculated stresses could be significantly sophisticated analyses and most advanced manufacturing methods
reduced with very minor changes to the geometry of the impeller, cannot eliminate the risk of component failures due to overload
notably in the blade to shroud fillet weld. This substantially operation. Common sense dictates that the most effective way to
increased the factor of safety, and providing confidence that the eliminate such risk is to avoid high risk operating conditions.
impeller integrity would not be at risk due to overload operation. Simply put, if there is increased risk of mechanical failure by
If the end user had not consulted the OEM, it is uncertain running at a portion of the performance envelope, the risks must be
whether or not a problem would have occurred. However, by weighed against the potential gains and the cost of maintenance or
having an open dialog, and by being proactive in performing an replacement of the equipment.
analysis, the risk of a problem was much more remote. In conclusion, though not receiving as much attention as surge,
Of course, it may not be possible to eliminate all of the risks prolonged operation in overload can have very detrimental effects on
associated with overload operation. Even the most advanced a centrifugal compressor. End users and OEMs alike need to be
analyses are but approximations of the real world. Therefore, one cognizant of the potential risks associated with operating in this
cannot be assured that such analyses will capture all of the portion of the performance map. End users accept the risks associated
potentially damaging phenomena within the compressor flow path. with surge and, despite the extra horsepower consumed, often run
Further, even the most robust designs will fail if subjected to the their compressors on recycle so as to avoid surging the units. As
right excitation mechanism, i.e., one that aligns with the natural noted, the forces associated with surge and overload are similar, yet
frequency of the impellers. Even a solid ring would fail if subjected the industry has not taken steps to protect equipment from overload
to the right excitation. At some point, common sense must prevail. operation. Failure to recognize the risks associated with overload
The end user and OEM must face the reality that the safest operation can have a devastating impact on compression equipment,
approach is to avoid operating in the overload region of the production, profitability, and engineering careers.
performance map. If the compressor might break if you run there,
do not run thereóor be prepared to undergo regular overhauls to NOMENCLATURE
check for internal damage.
End users can employ overload control systems to ensure that a ACFM = Actual cubic feet per minute
compressor does not operate beyond some agreed upon maximum A0 = Sonic velocity of gas at impeller inlet
capacity. These are implemented by incorporating algorithms in b = Impeller blade height
the control system that limit driver operation (speed, load, etc.) or Cm = Impeller meridional velocity
restrict the movement of control valves to keep the compressor in D2 = Impeller exit diameter
a safe region on its performance map. Some argue that such IGV = Inlet guide vane
overload controls limit production and decrease profitability. LSD = Low solidity vaned diffuser
However, when this reduction is weighed against the costs and lost N = Rotational speed in rpm
production associated with equipment failure, limiting overload Q = Inlet volumetric flow in ACFM
operation does not seem to be a bad choice. r = Impeller radius
U1 = Impeller inlet peripheral velocity
CONCLUSIONS U2 = Impeller exit peripheral velocity
U2/A0 = Machine Mach number
The most important conclusion to be derived from this work is V = Gas velocity
that, contrary to many commonly held beliefs, operation in W1 = Impeller
. inlet relative velocity
φ
overload can subject a centrifugal compressor to adverse forces. In
700Q
Ν D23
fact, in some circumstances, overload operation can be just as =

ρ
detrimental to component structural integrity as surge.
Of course, it is critical that the end user and OEM come to an = Gas density
understanding on the meaning of overload operation. The term
 Table of Contents
THE CONSEQUENCES OF COMPRESSOR OPERATION IN OVERLOAD 71

REFERENCES Kushner, F., 1980, “Disc Vibration—Rotating Blade and Stationary

Vane Excitation,” ASME Journal of Mechanical Design, 102,
Borer, C., Sorokes, J. M., McMahon, T., and Abraham, E. A., 1997,
pp. 579-584.
“An Assessment of the Forces Acting Upon a Centrifugal
Impeller Using Full Load, Full Pressure Hydrocarbon Testing,” Sorokes, J. M. and Koch, J. M., 2000, “The Influence of Low
Proceedings of the Twenty-Sixth Turbomachinery Symposium, Solidity Vaned Diffusers on the Static Pressure Nonuniformity
Turbomachinery Laboratory, Texas A&M University, College Caused by a Centrifugal Compressor Discharge Volute,”
Station, Texas, pp. 111-122. ASME Paper No. 00-GT-454.
Eckert, L., 1999, “High Cycle Fatigue Cracks at Radial Fan Sorokes, J. M., Borer, C. J., and Koch, J. M., 1998, “Investigation
Impellers Caused by Aeroelastic Self-Excited Impeller of the Circumferential Static Pressure Nonuniformity Caused
Vibrations, Part 1: Case History, Root Cause Analysis, by a Centrifugal Compressor Discharge Volute,” ASME Paper
Vibration Measurements,” Proceedings of DETC99, ASME No. 98-GT-326.
Design Engineering Technical Conference.
Fisher, E. H. and Inoue, M., 1981, “A Study of Diffuser/Rotor ACKNOWLEDGEMENTS
Interaction in a Centrifugal Compressor,” IMechE Journal The authors acknowledge Mr. Donald Schiffer, Mr. Paul Geise,
Mechanical Engineering Science, 23, (3), pp. 149-156, and Mr. Edward Abraham for their support in the preparation of
London, England. this paper. The authors also recognize the efforts of Mr. Jason
Hagelstein, D., Van den Braembussche, R., Keiper, R., and Kopko, Mr. Robert Kunselman, and Mr. Edward Thierman in the
Rautenberg, M., 1997, “Experimental Investigation of the preparation of several of the figures included herein. Finally, the
Circumferential Pressure Distortion in Centrifugal Compressor authors thank Dresser-Rand Company for allowing them the
Stages,” ASME Paper No. 97-GT-50. opportunity to publish this paper.
 Table of Contents
72 PROCEEDINGS OF THE THIRTY-FIFTH TURBOMACHINERY SYMPOSIUM • 2006