Introduction
Technical professionals understand a variety of fluid-transfer
performance concepts. The principles have much to do with
evaluating if an individual pump, on an individual/micro
scale, will succeed in accomplishing its fluid-transfer duties
with a reasonable degree of dependability. This includes
evaluation of inlet/discharge conditions, flow, speed and
power requirements, as well as durability.
This article explores a segment of the positive displacement
(PD) pump arena where precise flow control is needed from
a rotary-style PD pump (Figure 1). Despite the PD style of
operation for these pumps, their use in precise metering
applications has to be approached with caution because
of the potential for excessive slip, which induces errors.
Historically, instead of rotary pumps, reciprocating-type
pumps have been favored in these types of applications.
However, some processes cannot accept reciprocating-type
pumps because of their inherent pulsation, cost, automation
complexity, or other parameters. The NPSHr (net positive
suction head requirements) and stuffing needs of
reciprocating pumps are also a challenge.
The Challenge
Figure 1 outlines the learn concepts to evaluate the
suitability of rotary positive displacement pumps for
applications needing precise flow control with variable
process conditions, while factoring in pump wear.
The application of advanced fluid-transfer concepts on a
macro (or process) scale will enable entire processes to
become efficient in addition to aiding the efficiencies of any
specific pump. Users can find ways to produce a product at
the least cost considering factors such as plant-wide labor, floor space, capital investment, cleaning infrastructure and
total process energy usage (Figure 2).
For instance, users could replace batch-blending processes
with continuous in-line blending processes. New pumps
with good metering/predictable flow performance are
enabling this process method switch.
In its simplest form, a batch process (Figure 3) first involves
sending ingredients in the correct amounts to a processing
tank. Subsequently, and possibly in a distinct step, the
products are mixed within the tank to produce the desired
blended product. In contrast, with an in-line continuous-blend
process (Figure 4), the ingredients are fed in proportionally
correct amounts and instantly combined as they are
transferred within a common manifold. This manifold
may also contain in-line mixing devices to make sure
the ingredients are properly blended.
A full analysis of the benefits and drawbacks of truly
continuous over batch processes are not possible in the
scope of this article. In summary, however, continuousbatch
processes can yield:
■ Large reductions in floor space (no multi-stage
blend tanks needed)
■ Possible quicker product-formulation changes to
match needs
■ Reduced cleaning surfaces (eliminating
multi-stage tanks)
■ Capability of high degree of automation
(recipe control)
■ Reduced product losses and waste treatment
Several drawbacks in the use of continuous-blend processes
have been caused by limitations in the pumping technology
employed. Past systems and some existing systems can be
effective, but cannot accommodate wide changes in process
parameters like flow rates (affecting proportion limits) and
viscosity (ingredient flexibility). Additional issues with
existing continuous in-line blending processes include
stability as a result of startup/shutdown conditions,
equipment aging and process upsets.
New pump technologies, as well as correct selection of
existing technologies, are now enabling the wider use of
continuous-blending processes that require more flexibility
and stability.
Details on Pump Performance
A pump’s “performance band” is the family of duty points
(pump speed versus delivered flow rate) resulting from
pump slip for a range of possible process conditions,
including viscosity, back pressure, temperature and even
pump wear during its lifetime. The pump performance band
can be described as either tight or loose, which indicates
how much the flow can change (think of slack) for a fixed
pump speed. The performance band can also be described
as wide or narrow to indicate the possible range of speeds
the pump can run while producing flow.
From a practical standpoint for in-line blending applications,
the tighter the pump-performance band, the better the
metering accuracy under varying process conditions. At the
same time, the wider the performance/flow rate band, the
more flexibility in handling formulations that require a
wide range of possible ingredient input flows.
This article explores these new concepts that take pump
performance to the next level. In addition to the pump just
simply working, the correct application of these pump
concepts allows refinement of the transfer process, permitting
new, enhanced applications that were previously not possible
or reliable. The in-line blending process described above is
one example. Other examples include coating, spray drying,
filling, filtering and heat-exchange processes that require
controlled flow with tight pump performance bands.
Tight Versus Loose Pump Performance
The root issue with rotary PD pumps is that the flow
performance on all pumps is to some degree affected by
internal clearances that result in slip. The degree of slip
changes with:
■ Viscosity changes
■ Differential pressure changes
■ Clearance allowances for temperature change
■ Wear (resulting in an increase in clearance)
Given these product/process variables, tight performance
occurs when the pump maintains close to its theoretical
displacement independent of changes to the above
variables. The definition of a PD pump is a pump that
transfers a set displacement per unit operation, such as
revolution or stroke.
Tight versus loose pump performance is the extent to
which, under a given range of conditions, the pump
maintains high volumetric efficiency. High volumetric
efficiency is the extent (ratio) in which the true
displacement of the pump approximates its theoretical
displacement for given process/product conditions. Pump
slip is the difference between the theoretical displacement
and the actual displacement. Therefore, the lower the pump
slip in any condition, the tighter the pump’s performance
under conditions of changing viscosity, pressure, temperature
or wear.
Classifying a pump as simply positive displacement
without quantifying the tightness of its performance band
can greatly affect the desired results in an application. The
extreme example is one in which, regardless of the pump
speed, the slip is 100%. That is, all fluid that is pumped
forward then flows (slips) back through the pump’s internal
clearances to produce no net fluid transfer. While sounding
dramatic, it is not uncommon that a pump reaches this
point (total loss of flow) before it is taken out of service to
be repaired or replaced.
To understand slip for traditional PD pumps, see Figure 5. It
illustrates the possible loose-performance range (the yellow
area) of a typical PD pump when operating in variable
conditions (changes in viscosity, back pressure, temperature,
and wear). This graph shows how flow for a given pump
speed (A) can vary from the theoretical (intersection BA) to
an extreme (intersection EA) which indicates no flow. This
condition occurs in pumps with worn pumping elements,
for example.
Even in non-extreme cases such as when needing a flow
rate of (B), the pump would need to be accelerated from
(A) to (F) in order to achieve the flow (B). This can prove
to be an automation challenge and result in a reduction
of reliability. If the automation system does not have a way
to compensate for loss of flow and the pump remains at the
same speed (A), the flow rate (D) would be inadequate. An
actual curve for such a pump with 0.153 gallons/revolutions
can be seen in Figure 6.
Most users specifying pumps realize this and attempt to
control the extreme variabilities of viscosity, pressure,
temperature and wear simultaneously.
In many applications, this variation is sufficient to produce
a challenging operational scenario. In some cases, advanced
automation can help, such as using flow meters with speed/
pressure control loops and back pressure stabilization
valves. However, there are cases for which the possible
variation cannot be compensated without recalibration or
retuning the processes. These methods can prove costly or
unfeasible, and could also increase system complexity (thus
reducing reliability).
Figure 5 illustrates a tight performance band, which is
shown as the green performance band range superimposed
on the same graph. Even with large variations in pumping
conditions within its published performance limits, the
maximum variation in flow versus pump speed would be
between (B) and (C) instead of (B) and (E), illustrated by
the yellow loose-performance band. An actual curve band
for such a pump can be seen in Figure 7. Both pumps
(Figures 6 and 7) have a theoretical displacement of 0.15
gallons/revolution, but the curve in Figure 6 shows how
loose the pump’s performance is at 250 rpm, producing as
much as 28 gallons/minute of slip while attempting to
pump 38 gpm. The pump shown in Figure 7 has only 4
gpm of slip under the same conditions.
Today’s advanced pump manufacturers provide the tools that
permit evaluating the possible slip for a given application.
Curves are supplied that demonstrate how to down-rate the
flow given changes in back pressure, viscosity or change of
internal component clearance to handle certain temperature
ranges. These tools are helpful for compensating for the
performance. At times, however, these performance changes
can’t be adequately or reliably compensated and may not
produce optimal control.
The Effects of Pump Component Wear
To further complicate matters, pump component wear
invalidates most pump-performance curves. In highly
variable conditions, wear cannot be accurately modeled or
predicted. For processes that require tight and predictable
performance over time, the solution is pumps that have
tight performance ratios to begin with and are either
immune to wear or can compensate for wear. Pumps can
also be repaired to like-new condition. In doing this, there
still remains the risk that the pump’s performance will
degrade before the anticipated rebuild point and cause
production issues. Repair or replacement to regain proper
pump performance can result in high costs for rotary PD
pumps. In other words, the pump works mechanically just
fine, but needs to be repaired to regain performance, which
can be costly.
Loose pump performance also has associated side effects.
These include an increased amount of shear that is
imparted on the fluid, greater power requirements (and
reduced efficiencies) of the pump and heat generation
Narrow Versus Wide Performance Band
This is not to be confused with tight and loose. In fact, in
many cases a pump with a tight performance band gives it
the ability to handle a wide flow performance range. The
width of the pump’s performance band describes the range of speeds in which the pump can produce acceptable flow
for the application. This is also sometimes referred to as the
effective turn-down ratio of the pump, borrowed from
terminology used in conjunction with motors or variablespeed
drives.
In Figure 8, notice the point at which the green or yellow
pump curves are at greatest slip point and cross the zero/no
flow (x axis line). These are points (A) and (B) respectively.
These are points in which the green and yellow bands,
representing respective pumps, begin to produce flow under
the greatest slip condition possible for the process. The
pump that starts to produce flow at point (A) will use the
total range of pump speed more effectively (revolutions per
minute) than the pump starting at point (B).
The performance band width of a pump is also affected by
the ability to drive the pump at low to high speeds. Torque
requirement, gear reduction, motor cooling and variablespeed
drive capabilities all play a part and are not in the
scope of this article. Motor and variable-speed drive
capabilities, for example, set lower and upper limits.
For an actual illustration of performance band width, refer
back to Figure 6. Notice that a pump, in this case a typical
lobe pump with a 0.153 gallon/revolution theoretical
displacement, effectively has a narrow performance
envelope. That is because under an arbitrary worst
condition—in this case pumping 1 cP (water-like viscosity)
fluid against 75 psig—the pump only begins to produce
flow at 185 rpm. This means that speeds between 50 rpm to
185 rpm, which are considered good speeds for ensuring the
long life of rotary PD pumps, are not available to the
pumping process. The performance band is therefore narrow
as it ranges from 185 rpm (instead of 0 rpm) to the
maximum mechanical speed capability of the pump, or
some other process limitation like NPSHr versus NPSHa, or
the abrasiveness of product.
In comparison, refer back to Figure 7, which shows the
actual performance graph of a pump with a wide performance
envelope. Notice that under the same conditions as Figure 6—
pumping 1 cP product against 75 psig—flow begins to be
produced at 15 rpm (instead of 185 rpm). In this case, on
the low-RPM range, the pump in Figure 7 produces flow at a
much wider range of RPMs than the pump in Figure 6.
The lobe pump curve shown in Figure 6 does not show
how performance degrades as the pump wears. It is only a
“snapshot” of the pump performance when it is new. This is
the case with most PD pumps. If wear occurs in this pump,
the manufacturer-supplied performance curve no longer
applies and actual performance is unknown, unless verified
in the field. In Figure 6, the point at which the pump begins
to produce flow under wear conditions could be even
greater than 185 rpm and prompt repairs.
In sharp contrast, the pump illustrated in Figure 7
compensates for wear by maintaining as-new clearances.
Therefore, slip does not change, and the pump performance
remains tight with a wide range of flow capabilities. Both
the Blackmer® sliding vane pumps and Mouvex® eccentric
disc pumps share this phenomenon.
Our example application that exploits these needs—the
continuous in-line blending process—benefits from pumps
that have a high turn-down ratio. This is because the recipe
to produce the final product can be highly variable as far as
the content percentage of each ingredient. In other words,
the wider the flow rate range that is achieved by the pump,
the wider the variation of recipes that can be produced with
the system.
Conclusion
Good flow control from rotary PD pumps offers options for
more advanced processes, like in-line blending, that can
have far-reaching influence on a production facility’s overall
capital and operating costs. Respected pump manufacturers
offer performance curves that can be evaluated to determine
if the performance band is comfortably suitable for the
application. If not, alternative pumping technologies should
be studied and considered.
Most curves do not show the effects of wear on performance.
Therefore, if wear is anticipated during the expected life span
of the pumps and their parts, more subjective analysis is needed.
Some curves do model wear, so look for those. Even better,
some pump technologies, such as Blackmer® sliding vane
pumps and Mouvex® eccentric disc technology, compensate
by eliminating clearances caused by wear. Therefore, determine
if these pumps are applicable for the application.
Table 1 is a guide that compares different rotary PD pump
technologies and how they compare regarding their
performance bands and other criteria that may be important.
Basically, the most important criteria for the process should
be heavily weighted, but none of the criteria cause a
disqualification.
In-line blending systems are already common in the
beverage industry where the variation of ingredient
viscosities can be controlled. Several suppliers specialize in
these processes. Finding examples of more complex in-line
blending processes that demand pumps with tight and wide
performance bands is more elusive since they have been
developed under proprietary restrictions and confidentiality.
After all, these systems, when successful, give a clear
advantage to the processor, one which they rightfully desire
to keep and exploit.
