Principles of Flow Control

Most pumps operating today were selected to meet a maximum system

demand, or potential future demands. This means that most pumps are
oversized, rarely operating at their full design capacity. In addition,
pumps are often installed in systems with multiple operating points that
coincide with process requirements. Flow control is essential to system
performance. Sufficient flow ensures that equipment is properly cooled and that tanks are
drained or filled quickly. Sufficient pressure and flow must be guaranteed to satisfy system
requirements; this creates a tendency to oversize pumps and the motors that run them.
Because systems are designed with flow control devices to regulate temperature and
protect equipment from overpressurization, oversizing pump system can burden these
flow control devices with high energy dissipation loads. pumps are often operated over a
wide range of conditions. For example, many cooling systems experience variable loads
caused by changes in ambient conditions, occupancy, and production demands. To
accommodate demand changes, flow should be controlled
There are four primary methods for controlling flow through a system or its branches:
throttle valves, bypass valves, pump speed control, and multiple pump arrangements. The
appropriate flow control method depends on the system size and layout, fluid properties,
the shape of the pump power curve, the system load, and the system’s sensitivity to flow
rate changes. Throttle valves provide flow control in two ways: by increasing the upstream
backpressure, which reduces pump flow, and by directly dissipating fluid energy. By
increasing the backpressure on a pump, throttle valves make a pumping system less effi-
cient. In low-static-head systems, variable speed operation allows the pump to run near its
best efficiency point (BEP) for a given head or flow
A throttle valve chokes fluid flow so that less fluid can move through the valve, creating a
pressure drop across it. Throttle valves are usually more efficient than bypass valves,
because as they are shut, they maintain upstream pressure that can help push fluid through
parallel branches of the system. Bypass lines provide accurate flow control while
avoiding the danger of “deadheading” a pump. Deadheading is the condition in which a
pump’s
flow is completely choked off by closed down-stream valves. Unfortunately, bypassing
flow is usually the least energy-efficient flow control option
Bypass lines allow fluid to flow around a system component. A major drawback of bypass
valves is their detrimental impact on system efficiency. The power used to pump the
bypassed fluid is wasted. In static-head-dominated systems, however, bypass
valves could be more efficient than throttle valves or systems with adjustable speed drives
(ASDs).

Adjustable speed drives

Pump speed control includes both mechanical and electrical methods of matching the
speed of the pump to the flow/pressure demands of the system. ASDs, multiple-speed
pumps, and multiple pump configurations are usually the most efficient flow control
options, especially in systems that are dominated by friction head, because the amount of
fluid energy added by the pumps is determined directly from the system demand. Pump
speed control is especially appropriate for systems in which friction head predominates.
Both ASDs and multiple-speed motors provide efficient system operation by driving
pumps at different speeds according to system needs. During a period of low system
demand, the pump is operated at low speeds. The primary functional difference between
ASDs and multiple-speed motors is the degree of speed control available. ASDs typically
modify the speed of a single-speed motor through mechanical or electrical methods, while
multiple-speed motors contain a different set of windings for each speed. ASDs are
practical for applications in which flow demands change continuously.
ASDs allow pump speed adjustments to be made over a continuous range, avoiding the
need to jump from speed to speed. ASDs control pump speeds using several different
types of mechanical and electrical systems. Mechanical ASDs include hydraulic clutches,
fluid couplings, and adjustable belts and pulleys. Electrical ASDs include eddy current
clutches, wound-rotor motor controllers, and variable frequency drives (VFDs). VFDs
adjust the electrical frequency of the power supplied to a motor to change the motor’s ro-
tational speed. VFDs are by far the most popular type of ASD.
Pump speed adjustments are not appropriate for all systems, however. In applications with
high static head, slowing a pump could induce vibrations and create performance problems
that are similar to those found when a pump operates against its shutoff head. For systems
in which the static head represents a large portion of the total head, however, operators
should use caution in deciding whether to use ASDs. Operators should review the
performance of ASDs in similar applications and consult ASD manufacturers to avoid the
damage that can result when a pump operates too slowly against high-static-head
conditions.

 Pump Operating Efficiency Improvements

For many systems, VFDs can help to improve pump operating efficiency despite changes
in operating conditions. The effect of slowing pump speed on pump operation is illustrated
by the three curves in Figure 18. When a VFD slows a pump, its head/flow and brake
horsepower (bhp) curves drop down and to the left, and its efficiency curve shifts to the
left. This efficiency response provides an essential cost advantage; keeping the operating
efficiency as high as possible across variations in the system’s flow demand can reduce the
energy and maintenance costs of the pump significantly. VFDs can also be used with
positive displacement pumps.
 System Operating Efficiency Improvements
VFDs can provide operating cost reductions by increasing a pump’s operating efficiency.
However, the majority of savings derive from the reduction in frictional or bypass flow
losses.
Using a system perspective to identify areas in which fluid energy is dissipated in
nonuseful work often reveals opportunities for reducing operating costs. For example, in
many systems, increasing flow through bypass lines does not have a noticeable impact on
the backpressure on a pump. Consequently, in these applications, pump efficiency does
not necessarily decline during periods of low flow demand. However, analyzing the entire
system allows operators to identify the energy lost in pushing fluid through bypass lines
and across throttle valves. Figure 19 depicts energy losses attributable to bypass valve
operation; Figure 20 depicts energy losses attributable to throttling. (Note that Figures 19
and 20 are illustrative and do not represent actual pump curves.) One major benefit of
VFDs is that they can reduce energy losses by lowering the overall system flow or head.
By slowing down the pump and reducing the amount of fluid energy imparted to the
system when it is not needed, VFDs offer substantial savings with respect to the cost per
gallon of liquid pumped. Another system-related benefit is that VFDs provide a soft-start
capability. During start-up, most motors experience in-rush currents that are 5 to 6 times
higher than normal operating currents. This high current fades when the motor achieves
normal speed.
VFDs allow the motor to be started with a lower start-up current—usually only about 1.5
times the normal operating current. This reduces wear on the motor and its controller.
 Maintenance Requirements
As added system equipment, VFDs require maintenance and repairs. However, in many
applications, VFDs lower the maintenance requirements for the pump, system piping, and
components. The principal factors behind these maintenance savings are the reduced load
on the pump and the lower static and dynamic fluid forces imparted to the system.
By reducing a pump’s operating speed, a VFD often shifts the BEP to the left of the BEP
corresponding to the pump’s normal operating speed. In these cases, since the bearing
loads on a pump are lowest when the pump is operating at its BEP, this shift of the BEP
during periods of low flow allows the pump to operate with lower bearing loads and less
shaft deflection. Most pump bearings are roller- or ball-type; their design operating life is
a function of the cube of the load. Consequently, using a VFD can extend the interval
between bearing maintenance tasks.
In addition, VFDs reduce stress on pipes and piping supports. When the system flow far
exceeds equipment demands, excess fluid energy is dissipated in the form of noise and
vibration. Vibrations help to loosen mechanical joints and cause cracks in the welds in
pipes and pipe hangers. By reducing the fluid energy, VFDs lessen system wear. For more
information on indications of excessive system flow and ways to correct it, see the fact
sheet in this section titled Indications of Oversized Pumps.
 Limitations of VFDs
Although using VFDs can help to reduce operating and maintenance costs, they are not
appropriate for all applications. As a pump’s speed decreases, it generates less pressure. In
high-static-head applications, the use of VFDs can slow a pump down so that it operates at
or near shut-off head conditions. The pump thus experiences the same harsh conditions
that the manufacturer attempts to guard against when setting a minimum flow rate, which
usually corresponds to the pump’s rated speed. The consequences include greater shaft
deflection, high vibration levels, and high bearing loads.
Power quality can also be a concern. VFDs operate by rectifying the alternating current
(ac) line power into a direct current (dc) signal, then inverting and regulating this signal
into ac power that is sent to the motor. Often, the inverter creates harmonics in the power
supplied to the motor. These harmonics can cause motor windings to operate at higher
temperatures, which accelerates wear in insulation. To account for the added winding heat,
motors are typically derated 5% to 10% when used with VFDs. A classi-fication of motors
known as “inverter-duty” has been developed to better match VFDs with motors.
In some electrical systems, the harmonics creat-ed by the inverter can be picked up by
other electrical lines that have common connections with the VFD. Systems that are
sensitive to minor disturbances in power supply should be served separately from the VFD
power supply.
In some applications, VFDs contribute to reduced bearing life. The interaction between the
three phases of the power supply from a VFD inverter sometimes induces a small voltage
across the motor bearings. As a result, these bearings can experience pitting and
accelerated wear. VFD manufacturers are familiar with this problem, and several methods
can be used to correct it. These methods include insulating certain bearings, grounding the
shaft, and conditioning the power supply.
Finally, anticipated energy savings are not realized in some applications because some of
the losses associated with VFD installation were not taken into consideration. The VFDs
themselves are approximately 95% to 97% efficient, and motor efficiency generally begins
to decrease at less than 75% of full load. In addition, the quality of electric power supplied
to the motor can affect both its efficiency and its power rating.
Although VFDs are an attractive option in many applications, all these considerations
should be incorporated into a feasibility study before VFDs are installed.

Multiple-speed motors are practical for systems in which the flow demands change
between distinct, discrete levels that feature lengthy periods of operation. One of the
drawbacks to multiple-speed motors is the added cost of equipment. Since each speed has
its own set of motor windings, multiple-speed motors are more expensive than single-
speed motors. Also, multiple-speed motors are slightly less efficient than single-speed
ones. Multiple pump arrangements typically consist of pumps placed in parallel in one of
two basic configurations: a large pump/small pump config-uration, or a series of identical
pumps placed in parallel. In the large pump/small pump case, the small pump, often called
the “pony pump,” operates during normal conditions. The large pump is used during
periods of high demand. Because the pony pump is sized for normal system operation, this
configuration operates more efficiently than a system would that relies on the large pump
to handle loads far below its optimum capacity
Pony Pumps
Pumping systems have a wide range of flow needs. In many applications, there is a large
difference between the flow required during normal system operation and that required
during peak load conditions. For example, some cooling system and rainwater collection
applications require a relative-ly low flow rate. Occasionally, however, a heavy storm or a
large heat load caused by a sudden increase in production demand creates a need for
greater pumping capacity.
If pumps are sized to handle a peak flow or worst-case conditions, they could operate at
substantially less efficient levels for long periods during times of high demand. Oversized
pumps in applications like these tend to waste energy, and they require frequent
maintenance because they operate far from their best efficiency points.
In applications such as sewage treatment plants, the normal operating demands on pumps
may be relatively low. During storms, however, the amount of fluid that must be drained
from holding ponds or tanks increases dramatically. So pumps that maintain holding pond
levels must be able to handle storm conditions.
To avoid the high friction losses and maintenance problems that accompany continuous
operation or frequent starts of oversized pumps, a plant can install smaller ones, called
“pony pumps,” to handle normal operating conditions. The large pumps would then be
used occasionally only to handle severe load conditions, providing considerable cost
savings.
 When To Consider Pony Pumps
Indicators of a need for a smaller pump to handle normal operating conditions include the
following:
■ Intermittent pump operation
■ Excessive flow noise, cavitation, and piping vibrations that disappear during heavy
demand periods. (If these conditions persist, then the primary pump may need to be
downsized.)
 Costs of Intermittent Pump Operation
Intermittent pump operation is caused by an unbalanced set of system flows. For example,
a pump’s high flow rate drains the tank or reservoir to the point where the low-level switch
de-energizes or turns off the pump. When the fluid level in the tank rises and activates the
high-level switch, the pump is re-energized, turning back on to drain the tank (see Figure
16 on page 48).
Repeatedly stopping and restarting a pump wears out the motor controllers and dynamic
surfaces in the pump/motor assembly, and it can lead to unreliable pump operation. This
problem is especially severe for large pumps, because of their high starting currents. Each
repeated closing and opening of high-voltage contacts also creates a danger of sparking
that can damage the contact surfaces. In addition, discontinuous loading of the
transformers and switchgear often shortens their operating lives. Some pump/motor
assemblies are specially designed to handle repeated starting and stopping. For such
applications, this more expensive type of equipment should be specified.
Many pumps do not respond well to start-ups and shutdowns. The mechanical seals used
in many pumps rely on a lubricating film of system fluid. This film requires a revolution
or two to develop and, over time, repeated start-ups accelerate seal wear. Similarly,
bearings that are subjected to cyclical loading tend to have shorter operating lives than
those in constant-use applications.

 Costs of High Flow Velocity

An additional penalty for using an oversized pump is the added friction losses which occur
during pump operation. Higher flow rates create higher flow velocities which, in turn, lead
to higher friction loss. The relationship between velocity and friction loss is provided by
the Darcy-Weisbach equation:
hf = f L V2
D 2g
where
hf = head loss
f = pipe friction coefficient
V = fluid velocity
g = gravitational constant
D = inner diameter of the pipe
L = length of pipe.
The V2 term shows that pressure loss through a pipe is proportional to the square of the
fluid velocity. Consequently, given the same size pipe, a flow rate that is two times higher
endures four times more friction loss. This means that it costs much more to pump a gallon
of fluid at a higher-than-necessary flow rate.
 Recovering the Costs of Installing a Smaller Pump
Installing a smaller pump to run parallel to an existing one can provide substantial energy
and maintenance cost savings. A simple economic analysis can demonstrate the cost of
current power consumption and maintenance intervals in com-parison to the capital cost
and projected savings associated with operating a smaller, more efficient pump.
Energy-saving alternatives to a pony pump include reducing the impeller size, replacing
the existing pump/motor assembly with a smaller one, and installing an adjustable speed
drive (ASD) on the pump motor. Depending on the requirements of the application,
impeller adjustments and the smaller pump/motor assembly could compromise the
capacity of the existing pump during worst-case situations. Although ASDs in general
allow a pump to run at lower capacity, variable frequency drives (VFDs) are more suitable
for varying demand rather than for continuously low demand.
The VFDs themselves introduce efficiency losses. If normal operation is far below the full
load rating of the motor for long operating periods, the cost of these losses can be
considerable. A VFD can also introduce harmonics in the motor wind-ings, which
increases the winding temperature. Over an extended period of time, this increase in the
motor winding temperature accelerates the breakdown of insulation. A project undertaken
by the city of Milford, Connecticut, provides a practical example of the successful use of a
pony pump. By adding a pony pump to the city’s Welches Point Sewage Lift station,
Milford realized substantial energy savings and reduced maintenance costs.
Multiple Pump Arrangements

With a series of identical pumps placed in parallel, the number of operating pumps can be
changed according to system demands. Because the pumps are the same size they can
operate together, serving the same discharge header. If the pumps were different sizes, the
larger pumps would tend to dominate the smaller pumps and could cause them to be
inefficient. If the proper pumps are selected, each pump can operate closer to its highest
efficiency point. An added flow control benefit of parallel pumps is that a system curve
remains the same whether one or several pumps are operating; what changes is the
operating point along this system curve.
Multiple pumps in parallel are well suited for systems with high static head. Another
advantage is system redundancy; one pump can fail or be taken off line for maintenance
while the other pumps support system operation. When identical parallel pumps are used,
the pump curves should remain matched; therefore, operating hours should be the same for
each pump, and reconditioning should be done at the same time for all of them.

An alternative to using one pump to serve the requirements of a system is to use several
smaller pumps in combination (parallel operation).
Wide variations in system demand preclude a single pump from consistently operating
close to its best efficiency point (BEP). Operating a pump away from its BEP can result in
higher operating and maintenance costs. In some systems, espe-cially those with high
static head requirements, energizing or de-energizing multiple pumps to meet demand
changes allows each pump to operate more efficiently, improving overall system
efficiency. However, this efficiency advantage depends on the pump curves, the system
curve, and the demand change that is being met.
Some of the advantages of multiple pump arrange-ments are flexibility, redundancy, and
the ability to meet changing flow needs efficiently in systems with high static head
components. In systems with high-friction components, alternatives such as adjustable
speed motors tend to be more efficient solutions to variable demand requirements.
Multiple pumps are usually parallel combinations of the same pump model. Placing an
additional pump on line adds flow to the system and shifts the operating point to the right
along the system curve (see Figure 14 on page 44).
Parallel pumps are usually identical, to provide balanced load-sharing when all the pumps
are operating at the same time. Using different-sized pumps could result in a condition in
which the largest pump dominates the system, forcing other pumps to operate below their
minimum flow ratings. If different-sized pumps must be config-ured in parallel, their
performance curves should be carefully reviewed to ensure that no pump operates below
its minimum flow requirement.
 Best Efficiency Point
Design characteristics for both performance and service life are optimized around a
capacity designated as the best efficiency point (BEP). Every centrifugal pump has a BEP
—the point at which its operating efficiency is highest and its radial bearing loads are
lowest. A pump’s BEP is a function of its inlet configuration, impeller design, casing
design, and pump speed. At the BEP, the hydraulic efficiency is at its maximum, and the
liquid enters the impeller vanes, casing diffuser (discharge nozzle), or vaned diffuser in a
shockless manner. Flow through the impeller and diffuser vanes (if the pump is so
equipped) is uniform, free of separation, and well controlled. The flow remains well
controlled within a range of capacities designated as the preferred operating region
(POR). Within this region, the service life of the pump will not be affected significantly
by hydraulic loads, vibration, or flow separation. The allowable operating region (AOR)
defines the precise limits for minimum and maximum flow in a pump.
Most centrifugal pumps are equipped with roller or ball bearings. Since the operating life
of these types of bearings is an inverse function of the cube of the load, selecting a pump
with a BEP that is close to the system’s normal operating range significantly extends the
interval between bearing replacements.
 Advantages of Multiple Pump Arrangements
There are many advantages to using combinations of smaller pumps rather than a single
large one. These advantages include operating flexibility, redundancy in case of a pump
failure, lower maintenance requirements, and higher efficiency.
Operating Flexibility. As shown in Figure 14, using several pumps in parallel broadens
the range of flow that can be delivered to the system. (Note that Figure 14 is illustrative
and does not represent actual pump curves.) In addition, energizing and de-energizing
pumps keeps the operating point of each one closer to its BEP (for systems with flat
curves). Operators should use caution when operating parallel pumps, however, to ensure
that the minimum flow requirement is met for each pump.
Redundancy. With a multiple pump arrangement, one pump can be repaired while
others continue to serve the system. Thus, the failure of one unit does not shut down the
entire system.
Maintenance. Multiple pump configurations allow each pump to be operated close to
its BEP (for systems with flat curves), which reduces bearing wear and permits the pumps
to run more smoothly. Other benefits include less reliance on energy-dissipating flow
control options such as bypass lines and throttle valves. The use of a single, large pump
during low-flow demand conditions forces the excess flow to be throttled or bypassed.
Throttling the flow wears the throttle valves and creates energy losses. Simi-larly,
bypassing the flow is highly inefficient, since all the energy used to push the excess flow
through the bypass lines is wasted. Variable speed drives can also be an efficient solution.
Efficiency. A potential advantage of using multiple pumps is higher overall efficiency,
since each pump can operate close to its BEP (for systems with flat curves). Energizing or
de-energizing pumps as needed to meet changes in system demand allows each pump to
operate over a smaller region of its performance curve—ideally, around the BEP. A single
pump would have to operate over a larger range, and thus farther away from its BEP at
times.
At a given head and flow, high-speed pumps tend to be more efficient than low-speed
pumps. Pumps with specific speed values greater than 3,000 are the exception; they tend to
be less efficient at higher speeds. However, this is not typical of most pumps. Since
smaller pumps require smaller motors, the use of multiple high-speed pumps can provide
an efficiency advantage over a single, low-speed pump. However, this efficiency
advantage should be balanced against the tendency of high-speed machines to require
more maintenance.