Breadth of Efficiency

SEPTEMBER 2013 ISSUE

THIS TOPIC SPONSORED BY
SHARE THIS PAGE
by Joe Evans
In 2012, I wrote a six-part series on centrifugal pump efficiency. In Part Five, I
touched on the importance of the breadth of efficiency and how it can be more useful
than the peak best efficiency point (BEP). During the past year, I have received a
number of requests to revisit this topic and show some examples. I hope that this
column will answer your questions.

Definition & Importance

I define breadth of efficiency as the range of flow at BEP and within a couple
percentage points on either side of the BEP. This range is important for two reasons.
Let me address the first one with a question. How often does a pump, running at full
speed, operate at its BEP? The answer is seldom. Simple systems will often operate at
BEP. However, more complex systemsupgrades to existing systems with older pipe
and build/design systemsusually do not. The second reason is that a wide
efficiency range will allow higher operating efficiency when pump speed is reduced
by a variable frequency drive (VFD). Peak BEP efficiency is also important, but
losing one or two points to attain a wider range of efficiency can be a reasonable trade
off. As a rule of thumb, higher flow pumps usually have broader BEP ranges than
lower flow pumps. However, some low-flow designs offer a much broader range of
efficiency than do others. Sometimes, one manufacturers pump will have a broader
range than those of its competitors. In other cases, a single manufacturer will have
several different models that produce similar flows at different efficiency ranges.

Examples
Lets take a look at two lower flow pumps with different efficiency ranges. These are
real pumps, and I downloaded the pump curves from the manufacturers electronic
catalogs. Figure 1 shows the variable speed curves for a 3x4x13, end-suction pump
(Pump A) operating against an ideal system curve that intersects the center of the BEP
range at 60 hertz. The red dots show the BEP range. The BEP remains at 75 percent
from approximately 465 gallons per minute (gpm) to 530 gpm (a 65-gpm range). At
full speed, the efficiency drops to about 68 percent at 400 gpm.

Figure
1. The variable speed curves for a 3x4x13, end-suction pump (Pump A)
Under VFD control, the efficiency drops to about 71 percent at 400 gpm
(approximately 54 hertz), and at 300 gpm (50 hertz), the efficiency drops to 65
percent. In a variable speed application, designed for flows from 250 to 500 gpm,
pump efficiency will range from 60 percent to 75 percent. Figure 2 shows the
performance curve for a 3x4x12, end-suction pump (Pump B) plotted against the
same system curve. The BEP efficiency is also 75 percent, but the range is more than
twice that of the example in Figure 1425 gpm to 575 gpm (a 150-gpm range). Full
speed efficiency drops just 2 points at 400 gpm and just 1 point when
under VFD control. At 300 gpm, variable speed efficiency remains at 71 percent. In a
variable speed application designed for the same range of flow as the example in
Figure 1 (250 to 500 gpm), Pump Bs efficiency ranges from 68 percent to 75 percent.

Figure
2. The performance curve for a 3x4x12, end-suction pump (Pump B)
Pump B would be the best choice for both across the line (full speed) and variable
speed operation. Its increased breadth of efficiency provides a better chance of
actually hitting the BEP when operating at full speed. It will also offer an overall
higher efficiency when operating at variable speed. If you look at pumps across the
industry, you will find many examples in which this holds trueespecially with lower
flow pumps. We have many choices when selecting pumps. Take your time and
compare several models and manufacturers.
- See more at: http://www.pumpsandsystems.com/topics/pumps/centrifugalpumps/breadth-efficiency#sthash.xYQOHRH3.dpuf

Calculating Energy Savings and Payback

THIS TOPIC SPONSORED BY
PRINT PAGE
SHARE THIS PAGE
by Joe Evans, Ph.D
Energy efficiency and reduced consumption are important issues in the pump and
motor marketplace. Over the long term, electricity costs will continue to increase
regardless of our actions. This will be due to increasing fuel costs and inflation, but if
programs that limit CO2 emissions are enacted, energy costs will skyrocket. I belong
to that group of those "not-so-green" scientists who do not believe that anthropogenic
CO2 emissions have any measureable impact on global temperatures. That said, I am

completely in favor of reducing energy consumption, but for a different reasoneconomics.

There are several ways to reduce energy costs in pumping applications. The first, and
probably most important, is the application design. Well designed systems are usually
far more efficient than poorly designed ones. Increasing pump hydraulic efficiency at
the H/Q point reduces the BHP required. A reduction in BHP reduces the energy
required per gallon pumped. Motor efficiency can also have a significant impact.
Increased motor efficiency reduces the energy required to produce a certain BHP.
When taken together, application design and wire-to-water efficiency can significantly
reduce electrical consumption.
There are times, however, when pump efficiency can take a back seat to other
important issues. For example, the efficiency of a vortex sewage pump can be 25
points lower than a standard non-clog pump with a similar flow and head (see my
August 2007 column for more information). But, if plugging is a problem and the
non-clog has to be pulled weekly, the maintenance costs will far exceed the increased
power cost of the vortex pump. Even if a lower efficiency pump is the best choice for
an application, a higher efficiency motor will still decrease the overall operating costs.
Figure 1 is a screen shot of my Excel-based, Wire-to-Water Energy Calculator. It is
available for download in the "Pump Sizing & Selection Tools" section of my website
(www.PumpEd101.com). The calculator allows evaluation of the electrical
consumption of various pumps with the same motor, various motors with the same
pump or various combinations of each. It also provides a simple "payback" analysis
when comparing two different pump and motor combinations.

ADVERTISEMENT
Figure 1

After entering the required data, the calculator will produce a number of results
including the BHP required, wire-to-water efficiency and annual energy cost. KW and
cost per thousand gallons pumped are also displayed. When you compare two
different pump/motor combinations and enter the cost of each, the calculator will
produce a simple payback analysis that displays annual savings and payback in years.
Payback is simply the pump/motor cost differential divided by annual savings.
Although a present value analysis may be needed in some instances, simple payback
will usually provide the information needed to make a selection decision.
- See more at: http://www.pumpsandsystems.com/topics/motors/calculatingenergy-savings-and-payback#sthash.mz6wJrFL.dpuf

Centrifugal Pump EfficiencyWhat Is Efficiency?

by Joe Evans, Ph.D.

Editor's Note: This is the first in a six-part series on centrifugal pump

efficiency. For other articles in this series, click here.
In this multi-part series, we will investigate several aspects of
centrifugal pump efficiency. First, I will define efficiency and give
some examples. Next, I will examine some of the design criteria that
ultimately dictate the efficiency exhibited by a particular pump.
I will also try to make that somewhat nebulous quantity, known as
specific speed, more meaningful. I will illustrate its effect on the
shape of a pumps performance curve and overall pump efficiency.
Next, I will explain the contributions of individual pump components
to a pumps overall efficiency and show why the combined efficiency
of a pump and its driver is the product, not the average, of the two
efficiencies.
How pump efficiency can be preserved by changing impeller speed
rather than reducing it diameter will also be examined. Then I will
compare the value of peak efficiency versus the breadth of efficiency
over a range of flow. The discussion will end with the importance, or
sometimes unimportance, of efficiency as it relates to a particular
application or process.
Advertisement

What Is Pump Efficiency?

When we speak of the efficiency of any machine, we are simply
referring to how well it can convert one form of energy to another. If
one unit of energy is supplied to a machine and its output, in the
same units of measure, is one-half unit, its efficiency is 50 percent.

As simple as this may seem, it can still get a bit complex because the
units used by our English system of measurement can be quite
different for each form of energy. Fortunately, the use of constants
brings equivalency to these otherwise diverse quantities.
A common example of such a machine is the heat engine, which
uses energy in the form of heat to produce mechanical energy. This
family includes many members, but the internal combustion engine is
one with which we are all familiar. Although this machine is an
integral part of our everyday lives, its effectiveness in converting
energy is far less than we might expect.
The efficiency of the typical automobile engine is around 20 percent.
To put it another way, 80 percent of the heat energy in a gallon of
gasoline does no useful work. Although gas mileage has increased
somewhat over the years, that increase has as much to do with
increased mechanical efficiency as increased engine efficiency itself.
Diesel engines do a better job but still max out around 40 percent.
This increase is due, primarily, to its higher compression ratio and the
fact that the fuel, under high pressure, is injected directly into the
cylinder.
Advertisement
In the pump industry, much of the work involves two extremely
simple, yet efficient, machinesthe centrifugal pump and the AC
induction motor. The centrifugal pump converts mechanical energy
into hydraulic energy (flow, velocity and pressure), and the AC motor
converts electrical energy into mechanical energy.
Many medium and larger centrifugal pumps offer efficiencies of 75 to
93 percent and even the smaller ones usually fall into the 50 to 70
percent range. Large AC motors, on the other hand, approach an

efficiency of 97 percent, and any motorten horsepower and above

can be designed to break the 90 percent barrier.
The overall efficiency of a centrifugal pump is simply the ratio of the
water (output) power to the shaft (input) power and is illustrated by
the equation below:
Ef = P W / P S
Where:
Ef= efficiency
Pw= the water power
Ps= the shaft power
In the U.S., Ps is the power provided to the pump shaft in brake
horsepower (BHP) and Pw is:
Pw = (Q x H) / 3960
Where:
Q= Flow (gallons per minuteGPM)
H= Head (feet)
The constant (3,960) converts the product of flow and head (GPMfeet) into BHP. These equations predict that a pump that produces
100 GPM at 30 feet of head and requires 1 BHP will have an overall
efficiency is 75.7 percent at that flow point. An extension of the
second equation also allows the computation of the BHP required at
any point on a pumps performance curve if we know its hydraulic
efficiency. I will show some examples of this later in this series.
How Is Pump Efficiency Attained?
The overall efficiency of a centrifugal pump is the product of three
individual efficienciesmechanical, volumetric and hydraulic.
Mechanical efficiency includes losses in the bearing frame, stuffing
box and mechanical seals. Volumetric efficiency includes losses due
to leakage through the wear rings, balancing holes and vane

clearances in the case of semi-open impellers. Hydraulic efficiency

includes liquid friction and other losses in the volute and impeller.
Although mechanical and volumetric losses are important
components, hydraulic efficiency is the largest factor. The centrifugal
pump has a lot in common with the induction motor when it comes to
the design phase. The commonality is that both have only two major
components that can be modified by the designer. In the case of the
motor, it is the rotor and the stator. For the centrifugal pump, it is the
impeller and the volute (or diffuser). Lets start our investigation of
centrifugal pump efficiency with the impeller.
The affinity laws tell us quite a bit about the inner workings of an
impeller. We know that, for any given impeller, the head it produces
varies as the square of a change in speed. Double the speed and the
head increases by a factor of four. If you keep speed constant, the
same rule holds true for small changes in its diameter.
The flow through an impeller follows a similar rule, but in this case, its
change is directly proportional to the speed or diameter change
double the speed or diameter, and the flow doubles. Actually, a
change in rotational speed or impeller diameter refers to its
peripheral speed or the speed, in feet per second, of a point at its
periphery. It is this speed that determines the maximum head and
flow attainable by any impeller.
The head produced by an impeller is almost entirely dependent upon
its peripheral velocity, but flow is influenced by several other factors.
Obviously, the width and depth (cross sectional area) of the flow
passages (vanes) and the diameter of the impeller eye are important
considerations as they determine the ease with which some volume
of water can pass through the impeller.
Other factors such as vane shape also influence an impellers
performance. But where do you start if you want to design an impeller
from scratch? Do you take a wild guess about dimensions and
shapes, make some samples and then test them?
In the early days, that is exactly what pump designers did. Today,
however, they can draw on years of experience and, at a minimum,

find a suitable starting point for the design. That starting point is
specific speed. Next month, we will investigate specific speed and
how it can predict the performance of a particular impeller.
Where Can You Learn More?
There are many great pump books available today, but one of the
classics is now available as a free download at Google Books.
Pumping Machinery was authored by Arthur M. Green, a professor of
mechanical engineering at Rensselaer Polytechnic Institute and was
published by John Wiley & Sons over 100 years ago. It begins with a
comprehensive history of pumps and ends with a detailed review of
centrifugal pumps and their advances over the previous twenty years.
You will be impressed at the level of knowledge possessed by the
author. The number of illustrations is amazing and accounts for a
significant portion of its 725 pages. This one gets an A+ rating from
me.

Pumps & Systems, February 2012

About the Author
Joe Evans is responsible for customer and employee education at
PumpTech, Inc., a pump & packaged system manufacturer &
distributor with branches throughout the Pacific Northwest. He can be
reached via his website www.PumpEd101.com. If there are topics
that you would like to see discussed in future columns, drop him an
email.
Centrifugal Pump EfficiencySpecific Speed
by Joe Evans, Ph.D., P&S Editorial Advisory Board

Editor's Note: This is the second in a six-part series on centrifugal

pump efficiency. For other articles in this series, click here.
Early centrifugal pump design was heavily influenced by turbine
technology during the mid to late 1800s. Specific speed was first
applied to centrifugal pumps in the latter 1800s and was a modified
version of one developed for water turbines. Many pump designers
see specific speed as the most important contributor to centrifugal
pump design. It allows the use of existing design and test data to

design similar higher and lower flow pumps because the specific
speed of a pump is independent of its size.
An Index Number
As Terry Henshaw stated in Centrifugal Pump Specific Speed
(Pumps & Systems, September 2011), the definition of specific speed
can be confusing. It is best to think of it as an index number that can
predict certain pump characteristics. Viewed this way, specific speed
can be useful when selecting a pump for a particular application and
predicting premature failure due to off best efficiency point (BEP)
operation.

Figure 1 shows the relationship of the numerical value of specific

speed to an impellers geometric profile.

Figure 1. Pump profile comparisons (courtesy of

www.pumpfundamentals.com)
The lower values (500 to 1,500) on the left of the figure describe the
geometry of the radial vane impeller while the higher values (9,000
and higher) on the right of the figure equate to true axial flow
impellers (propellers). A radial vane impeller discharges 100 percent
of its flow perpendicular to its suction, usually with a low flow-to-head
ratio.
Advertisement
An axial flow impeller discharges 100 percent of its flow along the
same axis as its suction with a high flow-to-head ratio. Mixed flow
impellers (4,000 to 8,000) exhibit both radial and axial characteristics,
discharging between the radial and axial angles with a high-flow-tomoderate-head ratio.

Those between radial and mixed flow (1,700 to 3,500) are known as
Francis vane impellers. This design discharges radially, but the
transition from inlet to outlet is more gradual and results in the
highest efficiency. The cross sectional pictures in Figure 1 show that,
as specific speed increases, the impeller inlet or eye diameter
increases and eventually approaches or equals that of the vane
outlet. The flow passages also increase in size at a corresponding
rate.
Pump Design
While this is a nice comparison, pump designers may question its
usefulness. An equation (shown below) that relates specific speed
and its corresponding geometry to real application values of head,
flow and rotational speed was developed.

Advertisement
Ns = n x Q / H0.75
Where:
Ns= specific speed
n= pump rotational speed (rpm)
Q= flow (gallons per minute)
H= head (feet)

This equation can be used to determine which impeller design best

matches the application requirements.
Impeller Geometric Design
An impeller is needed that will produces 900 gallons per minute
(gpm) at 190 feet of head. If these values are entered as Q and H in
the equation with a motor speed of 3,600 rpm, the specific speed is

2,110. The geometry would be similar to the Francis vane impeller

(Figure 1 at the 2,000 point).
If the motor speed is lowered to 1,800 rpm, an impeller with a specific
speed of 1,055 would be required for the same flow and head. Its
geometry would be similar to the radial vane impeller (beneath the
1,000 point). At 1,200 rpm, specific speed is 703, and the impeller
would look like a hybrid of the two impellers seen to the left of Figure
1. Specific speed is directly proportional to rotational speed when
head and flow remain constant. However, the specific speed of a
single impeller design does not change with a change in rotational
speed. It remains constant because flow and head change in
accordance with the affinity laws.
Pump Performance
Figure 2 illustrates how specific speed can provide predictions about
the performance of an impeller design. Experience dictates that a
pumps efficiency reaches its maximum at specific speeds between
2,000 and 3,000, although favorable efficiency can occur at almost
any specific speed. Also, the area around the BEP tends to be flatter
and broader as specific speed decreases.

Figure 2. Specific speed and pump performance curves (courtesy of

www.pumpfundamentals.com)
Pump efficiency also increases with pump rotational speed,
especially high speeds. The increase is not as pronounced at 3,600
rpm and below. Specific speed also affects head-capacity curve
shape. Low specific speeds (500 to 1,500) produce flat curves. High
speeds (6,000 and higher) produce steep curves. Intermediate

speeds produce curves between the extremes. Curve shape will be

discussed further in another column.

Power Curve
Specific speed provides another predictionthe characteristics of the
power curve. At specific speeds below 4,000, power drops as flow is
reduced and is at its minimum at shut off head. The power curve
remains relatively flat, across the head-capacity curvebetween
4,000 and 4,500and rises toward shut off at specific speeds of
5,000 and higher. At speeds above 9,000, the power curve becomes
extremely steep and almost parallels the head-capacity curve.
Impeller Dimensions
Once geometry is chosen for the impeller, the pump designer can
conduct a mathematical analysis that will allow him to derive all the
impeller dimensions and angles needed to meet the design point.
This is an arduous task. To review a comprehensive example of how
this is done, see pages 2.23 - 2.31 of the second edition of the Pump
Handbook (McGraw-Hill).
Conclusion and Other Resources
The Ns ranges cited in this column are not cast in stone. They can be
narrower or wider or vary based upon the design characteristics of a
pump. They are, however, a good rule of thumb.

A great book on centrifugal pumps was written by John Richards,

editor of the San Francisco based journal Industry, in 1894. He was
the first writer to point out that centrifugal pumps do not operate via
centrifugal force since there is no such force in nature. Centrifugal
Pumps: An Essay on Their Construction and Operation, and Some
Account of Their Origin and Development in This and Other
Countries is available as a free download at Google Books. Some of
the knowledge about impeller vane shape from the 1850s is amazing.
About the Author
Joe Evans is responsible for customer and employee education at
PumpTech, Inc., a pump & packaged system manufacturer &
distributor with branches throughout the Pacific Northwest. He can be

reached via his website www.PumpEd101.com. If there are topics

that you would like to see discussed in future columns, drop him an
email.
Centrifugal Pump EfficiencyIndividual Efficiencies
by Joe Evans, Ph.D., P&S Editorial Advisory Board

Editor's Note: This is the third in a six-part series on centrifugal pump

efficiency. For other articles in this series, click here.
Hydraulic Efficiency
The shape and spacing of the impeller vanes have an effect on
overall pump efficiency. Although the ideal impeller would have an
infinite number of vanes, the real world limits us to five to seven for
clear water pumps and even fewer for pumps that handle larger
solids.

Also, flow would always be exactly parallel to the vane surfaces, but
that does not happen either. Oddly enough, if the designer follows
some well-documented rules, impeller vane efficiency losses remain
relatively flat (about 2.5 percent) across a specific speed (Ns) range
of 500 to 7,000.
Disk friction, which is caused by contact between the fluid and the
impeller shrouds and hub surfaces, can reduce impeller efficiency
another 4 to 15 percent at specific speeds below 2,000 but
decreases to 2 percent or less at a specific speed of 3,000 and
higher. Depending on its design, the impeller can reduce overall
pump efficiency by as little as 4.5 percent or as much as 17.5
percent.
The volute also plays an important role in pump efficiency. At specific
speeds below 2,000, friction losses range from 1 to 2.5 percent, but
losses can approach 10 percent at a specific speed of 5,000 and
higher. Typically, volute design begins with the throat (see Figure 1).
Advertisement
Its cross sectional area will determine the flow velocity out of the
volute. Flow through the throat and other portions of the casing
follows the law of constant angular momentum. Therefore, the

designer will try to avoid abrupt changes in its nearly circular

geometry while gradually increasing its volume.

Figure 1. Pump volute

Another critical area of the volute is the clearance between the outer
circumference of the impeller and that of the volute tongue or
cutwater. As this distance becomes larger, an increasing volume of
liquid escapes entry into the volute throat and recirculates back into
the volute case. The smallest distance that does not give rise to
pressure pulsations during vane passing will produce the best
efficiency. As a general rule, 5 to 10 percent of the impeller radius
tends to be a safe value. The next section explains this in more
detail, comparing the efficiencies that result from trimming an impeller
versus changing its rotational speed.
Volumetric Efficiency
Whether the volumetric efficiency of a centrifugal pump is a function
of the volute or the impeller is debatable (it is probably both), but I will
include its effect here. Volumetric efficiency represents the power lost
due to flow leakage through the wearing rings, the vane front
clearances of semi-open impellers and the balancing holes in the
rear shroud.

Leakage usually increases with a decrease in specific speed, flow or

a combination of the two. For example, at a specific speed of 500
and a flow of 100 gallons per minute, leakage can account for as
much as 7 percent of the total power consumed. At 2,000 gallons per
minute, it is reduced to about 2 percent. At higher specific speeds
and flows, volumetric losses can be as low as 1 percent.

Advertisement
Mechanical Efficiency
The final piece of the pump efficiency puzzle is that of mechanical
losses, although some of these losses are not always included in
published efficiency curves. In the case of a frame-mounted pump,
these losses are caused by the shaft bearings and the mechanical
seal or packing. For close-coupled pumps, bearing losses are figured
into the motor efficiency. Again the rule of thumb follows that of
volumetric efficiency, and losses increase as flow and/or specific
speed decrease.

If the same values of specific speed and flow as in the volumetric

example above are used, we could expect losses of 5 percent and 1
percent for a frame-mounted pump. At higher specific speeds and
flows, mechanical losses can drop below 1 percent.

Figure 2. Simplified total efficiency

Combined Efficiency
When looking at the overall efficiency of a pump in operation, the
efficiency of the driver must be included, and in many instances, that
driver will be an electric motor. When the Energy Independence and
Security Act of 2007 went into effect in December 2010, it raised the
bar on motor efficiency. Today, all new motors must meet premium
efficiency standards. Obviously, a higher efficiency motor will
increase the overall efficiency of a pumping system, but by how
much? How do we calculate the combined efficiency of pumps and
motors?

When I ask this question in my hydraulics classes, the common

sense response is that it is the average of the two efficiencies. After
all, the efficiency of two or more pumps, operating in parallel, is equal
to the average of their individual efficiencies at their operating points.
However, when machines operate in series, their combined efficiency
is different. Just as the total efficiency of a centrifugal pump is the

product of several different internal efficiencies, so is the total

efficiency of a pump and its driver.
The total efficiency of an electric motor and a centrifugal pump (wire
to liquid or wire to water efficiency) is the product of the individual
efficiencies. Sometimes, this is hard to visualize due to the units of
measure used to compute the two efficiencies. Figure 2 simplifies this
by using a single unit of measure that can be used to describe the
operation of any machine. That unit is a bag of energy.
In this example, the motor consumes 10 bags and produces 9 bags,
so its efficiency is 90 percent. The pump consumes the 9 bags
provided by the motor and produces 7, so its efficiency is 78 percent.
Therefore, for every 10 bags of energy consumed by the motor, 7
bags are produced by the pump, and the total efficiency of the
system is 70 percent (7 bags / 10 bags).
This confirms that the product of the two individual efficiencies is the
total efficiency of the system (90 percent x 78 percent = 70 percent).
If this system is controlled by a VFD with an efficiency of 97 percent,
the total efficiency will be reduced to 68 percent (at full speed, 60
Hz). This lower efficiency is the product of the three individual
efficiencies.
Next month, the column will cover the efficiency savings that can be
achieved by reducing the speed rather than trimming an impeller.
Pumps & Systems, April 2012
About the Author
Joe Evans is responsible for customer and employee education at
PumpTech, Inc., a pump and packaged system manufacturer and
distributor with branches throughout the Pacific Northwest. He can
be reached via his website, www.PumpEd101.com. If there are
topics that you would like to see discussed in future columns, drop
him an email.
Centrifugal Pump EfficiencyPreservation of Efficiency
by Joe Evans, Ph.D., P&S Editorial Advisory Board

Editor's Note: This is the fourth in a six-part series on centrifugal

pump efficiency. For other articles in this series, click here.
The performance of a centrifugal pump with a trimmed impeller will
follow the affinity laws as long as that trim is relatively small. Many
experts recommend a maximum of 10 percent reduction from the full
design diameter, but, when looking at a typical set of manufacturers
catalog curves, trims as great as 30 percent to 35 percent can be
found.
In Part 3 of this series, I examined the volutes contribution to
hydraulic efficiency. An important part of the volute is the tongue, or
cutwater. Its purpose is to maintain flow into the throat while
minimizing recirculation back into the case. The optimum clearance
between the tongue and the impeller periphery is the smallest
distance that does not give rise to pressure pulsations during vane tip
passing. A well-designed pump will have a full-size impeller that
meets these clearance criteria. When an impeller is trimmed, this
distance increases and allows more fluid to recirculate back into the
case. As recirculation increases, hydraulic efficiency decreases.
The Effect of Impeller Trim
Figure 1 shows the catalog curves for a centrifugal pump with a 9inch, full size impeller and several trims.At the best efficiency point
(BEP), hydraulic efficiency is 77 percent for the 9-inch impeller. When
trimmed to 8 inches (11 percent), BEP efficiency drops to 74 percent.
A 23 percent trim (7 inches) reduces BEP efficiency to 70 percent
and a 33 percent trim (6 inches) lowers it to just 62 percent. Although
not shown on the graph, if the trim were just 6 percent (8.5 inches),
BEP efficiency would remain at 77 percent. The trims shown on
themanufacturers catalog curves are the allowable trims, not
necessarily the most prudent ones.

Although an efficiency reduction of 3 percent could be acceptable in

certain applications, reductions of 7 percent and 15 percent should
seldom be acceptable. If you have to trim an impeller that much to
meet the requirements of the application, it is time to consider a
smaller pump.
Advertisement

Figure 1. Catalog curves for a centrifugal pump with a 9-inch,

full size impellar and several trims
Altered Speed
An alternative to trimming the impeller is to alter the impellers speed.
This can be achieved in many ways. Figure 2 shows the same pump
with four different speed curves that were selected to match the trims
shown in Figure 1. These speed changes were produced by a
variable frequency drive (VFD), so the curves are labeled in Hertz.
Other speed change options include belt drives and adjustable,
magnetic couplings.

The result is that BEP efficiency remains at the full-speed (diameter)

efficiency (77 percent) across a broad range of speeds. This occurs
because the impeller periphery to tongue clearance remains
unchanged and recirculation is limited to its design conditions. The
affinity laws also hold true across this range.
Next month, Part Five of this series will compare the impact of peak
BEP efficiency versus a broad range of high efficiency. It will also

show how a pump curves breadth of efficiency can affect both fixed
and variable speed operation.
Advertisement

Figure 2. The same pump from Figure 1 with four speed curves
Note:
In Part 2 of this series, Centrifugal Pump EfficiencySpecific
Speed (March 2012), I forgot to mention that parts of Europe use an
alternative method when computing specific speed (Ns) for double
suction pumps.
The European method uses half the BEP flow. In the U.S., we use full
flow regardless of the pump design. When Ns is calculated using half
the BEP flow, the result equals 0.707 that of the full flow calculation.
Pumps & Systems, May 2012
About the Author

Joe Evans is responsible for customer and employee education at

PumpTech, Inc., a pump & packaged system manufacturer and
distributor with branches throughout the Pacific Northwest. He can
be reached via his website www.PumpEd101.com. If there are topics
that you would like to see discussed in future columns, drop him an
email.

Centrifugal Pump Efficiency Curve Shape & Breadth of

Efficiency
by Joe Evans, Ph.D., P&S Editorial Advisory Board

Editor's Note: This is the fifth in a six-part series on centrifugal pump

efficiency. For other articles in this series, click here.
Curve Shape
In Part 2, I showed the effect of specific speed (Ns) on the shape of a
pumps performance curve and its maximum efficiency. A typical
performance curve is relatively flat at low values of Ns and becomes
steeper as Ns increases. Pump efficiency is lowest at low values of
Ns (500 and below) and increases as Ns increases. It reaches its
maximum in the mid-to-high 2,000 range and begins to decrease
above 3,000. However, the decrease above 3,000 is much smaller
than it is below 1,000.

Figure 1 compares the curve shape of four pumps with different

specific speeds. It also shows the percent increase in head from best
efficiency point (BEP) to shut off. As shown, the slope of the curve
increases with Ns. The black curve (Ns = 600) is relatively flat, and
this example actually decreases as it approaches shut off.

Figure 1. Comparison of four pumps' curve shapes

The blue curve (Ns = 3,600) also rises continuously but exhibits a 55
percent increase in head. The red curve (Ns = 5,700) is extremely
steep, and the head increases by 100 percent from BEP to shut off.
Steeper curves usually offer a greater range of control when
operated under variable speed control against some fixed elevation
or pressure head. These pumps can be problematic when running in
parallel or starting against varying system head conditions. The
green curve (Ns = 2,500) rises continuously as it approaches shut off,
and the head increases by about 35 percent from BEP to shut off.
Advertisement
Flatter curves work fine in across-the-line applications as long as the
static or pressure head remains relatively constant. They also work

well in closed-loop (and most open-loop) circulation applications

when operated under variable speed control.
Although the curves shown are representative of the shape that end
users can expect for those values of Ns, they are not cast in stone.
For example, pumps with a specific speed range of 900 to 1,200 can
be extremely flat, or they could exhibit a 15 to 20 percent rise in head
as they approach shut off. The same is true of pumps in the 1,500 to
2,000 range. They can be relatively flat or exhibit a head rise of 20 to
30 percent. The actual conditions will depend upon the individual
pump design.

Breadth of Efficiency
For example, a 2 x 2.5 x 8 centrifugal pump with a BEP efficiency of
70 percent at 200 gallons per minute (GPM) drops to 63 percent at
flows below 190 GPM or over 210 GPM. Since pumps seldom
operate at BEP, a wide range of high efficiency can allow for design
errors and potential changes in the system.All pumps reach their
peak efficiency at BEP, but the range of peak efficiency can vary
significantly from model to model. Some pumpsespecially lower
flow modelsexhibit a narrow BEP range, and once flow is out of
that range, efficiency drops quickly.

Figure 2 shows the variable speed curves for a pump with a specific
speed of 1,654. It exhibits a rise in head toward shut off of about 30
percent and a wide range of high efficiency.
Advertisement

Figure 2. Variable speed curves for a pump with a specific speed of 1,654

When running across the line (60 hertz curve), this pump will
maintain its BEP efficiency of 86 percent from 1,500 to 1,750 GPM.
The head change across this range is about 17 feet and allows for
incorrect duty point head calculations, as well as aging of the system.
It still performs at 85 percent from 1,375 to 1,875 GPM and will
maintain 84 percent from 1,250 to 1,900 GPM.
If this pump were operating in a variable speed application, against
the system curve shown in red, its control range would be about 12
hertz (49 to 60 hertz), and it would maintain 86 percent efficiency
from 1,375 to 1,625 GPM. Efficiency would still be at 85 percent
down to 1,250 GPM and would remain at a robust 84 percent down
to 1,050 GPM. This high efficiency across the range of flow enhances
the power reduction due to a change in speed.
Next month, I will end this series by looking at some examples of
when pump efficiency is important and when it is not so important.
About the Author

Joe Evans is responsible for customer and employee education at

PumpTech, Inc., a pump & packaged system manufacturer &
distributor with branches throughout the Pacific Northwest. He can be
reached via his website www.PumpEd101.com. If there are topics
that you would like to see discussed in future columns, drop him an
email.
Centrifugal Pump EfficiencyPreservation of Efficiency
by Joe Evans, Ph.D., P&S Editorial Advisory Board

Editor's Note: This is the fourth in a six-part series on centrifugal

pump efficiency. For other articles in this series, click here.
The performance of a centrifugal pump with a trimmed impeller will
follow the affinity laws as long as that trim is relatively small. Many
experts recommend a maximum of 10 percent reduction from the full
design diameter, but, when looking at a typical set of manufacturers
catalog curves, trims as great as 30 percent to 35 percent can be
found.
In Part 3 of this series, I examined the volutes contribution to
hydraulic efficiency. An important part of the volute is the tongue, or
cutwater. Its purpose is to maintain flow into the throat while
minimizing recirculation back into the case. The optimum clearance
between the tongue and the impeller periphery is the smallest
distance that does not give rise to pressure pulsations during vane tip
passing. A well-designed pump will have a full-size impeller that
meets these clearance criteria. When an impeller is trimmed, this
distance increases and allows more fluid to recirculate back into the
case. As recirculation increases, hydraulic efficiency decreases.
The Effect of Impeller Trim
Figure 1 shows the catalog curves for a centrifugal pump with a 9inch, full size impeller and several trims.At the best efficiency point
(BEP), hydraulic efficiency is 77 percent for the 9-inch impeller. When
trimmed to 8 inches (11 percent), BEP efficiency drops to 74 percent.
A 23 percent trim (7 inches) reduces BEP efficiency to 70 percent
and a 33 percent trim (6 inches) lowers it to just 62 percent. Although
not shown on the graph, if the trim were just 6 percent (8.5 inches),
BEP efficiency would remain at 77 percent. The trims shown on

themanufacturers catalog curves are the allowable trims, not

necessarily the most prudent ones.
Although an efficiency reduction of 3 percent could be acceptable in
certain applications, reductions of 7 percent and 15 percent should
seldom be acceptable. If you have to trim an impeller that much to
meet the requirements of the application, it is time to consider a
smaller pump.
Advertisement

Figure 1. Catalog curves for a centrifugal pump with a 9-inch,

full size impellar and several trims
Altered Speed
An alternative to trimming the impeller is to alter the impellers speed.
This can be achieved in many ways. Figure 2 shows the same pump
with four different speed curves that were selected to match the trims
shown in Figure 1. These speed changes were produced by a
variable frequency drive (VFD), so the curves are labeled in Hertz.
Other speed change options include belt drives and adjustable,
magnetic couplings.

The result is that BEP efficiency remains at the full-speed (diameter)

efficiency (77 percent) across a broad range of speeds. This occurs
because the impeller periphery to tongue clearance remains
unchanged and recirculation is limited to its design conditions. The
affinity laws also hold true across this range.
Next month, Part Five of this series will compare the impact of peak
BEP efficiency versus a broad range of high efficiency. It will also
show how a pump curves breadth of efficiency can affect both fixed
and variable speed operation.
Advertisement

Figure 2. The same pump from Figure 1 with four speed curves
Note:
In Part 2 of this series, Centrifugal Pump EfficiencySpecific
Speed (March 2012), I forgot to mention that parts of Europe use an
alternative method when computing specific speed (Ns) for double
suction pumps.

The European method uses half the BEP flow. In the U.S., we use full
flow regardless of the pump design. When Ns is calculated using half
the BEP flow, the result equals 0.707 that of the full flow calculation.
Pumps & Systems, May 2012
About the Author
Joe Evans is responsible for customer and employee education at
PumpTech, Inc., a pump & packaged system manufacturer and
distributor with branches throughout the Pacific Northwest. He can
be reached via his website www.PumpEd101.com. If there are topics
that you would like to see discussed in future columns, drop him an
email.
Centrifugal Pump EfficiencyWhen Is Efficiency Important?
by Joe Evans, Ph.D., PumpTech, Inc.

Editor's Note: This is the sixth in a six-part series on centrifugal pump

efficiency. For other articles in this series, click here.
The power required by a pump is directly proportional to both the flow
and the head that it produces. As flow and/or head increase(s) so
does the power required. Conversely, power is inversely proportional
to hydraulic efficiency. For the same flow and head, an increase in
efficiency reduces the power requirement. The two equations below
illustrate this relationship:
P = (Flow x Head) / Ef
BHP = ((Flow x Head) / 3960) / Ef

Where:
P

= Hydraulic power

BHP = Brake horsepower

Ef = Pump efficiency

Advertisement

Advertisement

Pumps that run continuously or for extended periods can experience

a substantial reduction in energy costs with a relatively small
increase in efficiency. Figure 1 shows two, 3,000-gallon-per-minute
pumps that operate 24/7. With efficiencies of 87 percent and 84
percent, the horsepower required is 130 and 135 respectively. The
electrical cost per thousand gallons is 5.7 and 5.9 cents. Two-tenths
of a cent is not a huge difference, but if you consider the annual cost
of electricity, the lower efficiency pump adds an additional $3,200 to
the total electric bill.

Figure 1. The wire-to-water information for two,

3,000-gallon-per-minute pumps that operate 24/7

Another factor that can increase the attractiveness of higher

efficiency is the cost of electricity. The energy cost of 10 cents per
kilowatt hour, used in Figure 1, is the average commercial rate for
2010. However, it can vary significantly by state.
In some states, the cost of electricity can be as low as 6 cents, and in
certain parts of Washington, cooperative rates can be as low as 4
cents. New England states costs, however, are in the high teens.
What is the worst case? Hawaiiwith Oahu at 23 cents and some
outer islands topping out near 40 cents.
As the cost per kilowatt hour increases, so will the savings due to
increased pump efficiency. Seeing what the average rates will be for
2011 will be interesting. I suspect that they have increased
substantially. Another factor to consider when selecting a pump that

will not run continuously is the actual flow rate required. Does the end
user really need 3,000 gallons per minute, or can the same results be
achieved by running a 2,000-gallon-per-minute pump longer?
If you have the same head and efficiency at these two flow rates, the
cost per thousand gallons pumped is the same for both. In most
cases, reducing flow by 1,000 gallons per minute will result in a
substantial decrease in friction head. Since BHP is directly
proportional to head, an end user could see a substantial reduction in
the cost per thousand gallons pumped with the lower volume pump.

Figure 2. The H/Q curve for a 4-inch, vortex wastewater pump

When Is Efficiency Not As Important?

Selecting the most efficient pumps and motors will always reduce the
cost of electrical power, but sometimes, the payback versus initial
cost does not pencil out. Examples include smaller pumps, pumps
that are used infrequently and those installed for back-up or
emergency use only.

In many industrial applications, efficiency will take a back seat to a

pumps ability to reliably perform a particular process. A good
example is a slurry pump, with which larger clearances increase
useful life. Another is the vortex pump, which is popular in both
industrial and wastewater applications.

Figure 2 shows the H/Q curve for a 4-inch, vortex wastewater pump.
At 800 gallons per minute, its hydraulic efficiency is just 48 percent. A
standard 4-inch, non-clog pump with similar performance would have
an efficiency of 68 to 75 percent20 to 27 points better. The reason
for the lower efficiency is that vortex action is a two-step process, and
the overall efficiency is the product of the two individual efficiencies.
However even though efficiency is much lower than normally desired,
there is an extremely positive side. Almost anything that enters the
suction of a vortex pump will exit its discharge.
This is because the vortex impeller is recessed and seldom contacts
any of the solids or other material in the pumpage. This can be
beneficial when smaller wastewater pumps are required. The more
efficient 4inch, non-clog pump can plug frequently when rags and
stringy material are present, and this often results in removing the
pump from service for cleaning on a weekly basis.
In these applications, a vortex pump can be far more reliable, and the
maintenance cost savings is much greater than the additional energy
costs due to lower efficiency. One of the seminars I present to
specifying engineers is titled How Lower Pump Efficiency Can
Reduce Costs. It usually gets their attention.
If a pump is run by a gasoline engine, the case could be made that
the pumps efficiency is not too important. Although an 80 percent
efficient pump should save quite a bit of energy over one that is 65
percent efficient, the gas engine (approximately 20 percent) brings
their totals down to 16 percent and 13 percent respectively.
It may be hard to justify a higher initial pump cost for such a small
energy savings, unless the pump is used frequently and for long
periods of time.
Finally, some application design points exist for which reasonable
efficiency cannot be attained, but a pump is still required. Suppose
some million-dollar process line cannot use a positive displacement
pump but, instead, requires a centrifugal pump that can deliver 20
gallons per minute at 3,000 feet of head.

Would it really matter if a single-stage pump had to be driven at

23,000 rpm and that its efficiency was less than 25 percent?
Probably not, and there are far more of these types of applications
than you might suspect.
There is definitely more than one side when it comes to pump
efficiency. Efficiency is a good thing, and we should always consider
a higher efficiency pump if the return on investment pencils out.
Often, a peak efficiency that adds one or two percentage points is not
that important since few pumps operate at their best efficiency points
(BEPs). The breadth of high efficiency, on either side of BEP, can be
far more beneficial.

About the Author

Joe Evans is responsible for customer and employee education at
PumpTech, Inc., a pump and packaged system manufacturer and
distributor with branches throughout the Pacific Northwest. He can
be reached via his website www.PumpEd101.com. If there are topics
that you would like to see discussed in future columns, drop him an
email.

Understanding System Efficiency in Motor-Driven Rotating Equipment

AUGUST 2015 ISSUE
THIS TOPIC SPONSORED BY
SHARE THIS PAGE
Users should consider system changes to comply with the new EISA standard.
by William Livoti

Energy efficiency has become a major focus for the U.S. government, municipalities,
power utilities and the industrial sector, with much of the attention falling on
components such as motors and pumps. For end users, understanding the difference
between component efficiency and system efficiency as applied to motor-driven

equipment is critical for evaluating a total system and making appropriate upgrades.
The Energy Independence and Security Act (EISA) is one standard that users must
understand and comply with to successfully improve system efficiency.

Efficiency Standards as Defined by EISA

For each general-purpose rating (Subtype 1) from 1 to 200 horsepower (HP) that was
previously covered by EPAct, the law specifies a nominal full-load efficiency level
based on National Electrical Manufacturers Association (NEMA) premium efficiency
as shown in NEMA MG 1, Table 12-12. All 230- or 460-volt (and 575-volt for
Canada) motors currently under EPAct that were manufactured after December 19,
2010, must meet or exceed this efficiency level.
General-purpose electric motors (Subtype II) not previously covered by EPAct will be
required to comply with energy efficiencies as defined by NEMA MG 1, Table 12-11.
The term general-purpose electric motor (Subtype II) refers to motors that incorporate
the design elements of a general-purpose electric motor (Subtype I) that are
configured as one of the following:

U-frame motor

Design C motor

Close-coupled pump motor

Footless motor vertical solid shaft normal thrust motor (as in a

horizontal configuration)
An 8-pole motor (900 rpm)
A poly-phase motor with voltage of not more than 600 volts (other
than 230 or 460 volts)

Motors that are 201 to 500 HP that were not previously covered by EPAct will be
required to comply with energy efficient efficiencies as defined by NEMA MG I,
Table 12-11.
This information and the Tables referenced above are readily available on the
Department of Energy (DOE) website.
ADVERTISEMENT

So, what does the new EISA Standard have to do with system efficiency? Many end
users believe that any system efficiency improvement is the result of an increase in
motor efficiency; however, that is not always the case. For example, consider a

centrifugal pump system operating at a fixed speed. The system requires variable flow
and is controlled by a motor-operated valve. One might believe that replacing the
standard-efficiency motor with the new EISA premium-efficient motor would lead to
an incremental gain in efficiency and a lower operating cost. This seems reasonable,
but more factors must be considered.
In order to meet the EISA standard, motor original equipment manufacturers (OEMs)
had to redesign their equipment to achieve the increased efficiency as mandated by
government regulations. To understand what is meant by "increased efficiency," users
must know the definition of a premium-efficiency motor and what affects that
efficiency.

Motor Losses
Losses in a motor include stray losses, rotors, stators, core losses and fan design
(windage).
To make a motor more efficient, a manufacturer must add more or better material.
These additions and adjustments could include more active material such as copper in
the winding, a longer stator, rotor cores and improved electrical steel (silicon steel is
used for the stator and rotor). A low-loss fan design could also be used to reduce
friction and windage losses. To reduce the stray load losses, manufacturing processes
are assured through International Organization for Standardization (ISO) 9001
procedures.
Some advantages of energy efficient motors are:
ADVERTISEMENT

Maximum Efficiency Energy-efficient motors operate at maximum

efficiency even when they are lightly loaded because of better
design.
Longer Life Energy-efficient motors dissipate less heat compared
with standard motors. Use of energy-efficient fans keeps the motor at
a lower temperature, which increases the life of the insulation and
windings as well as the overall life of the motor.
Lower Operating Cost The total energy cost of energy-efficient
motors during its life cycle is much lower when compared with
conventional motors.
Other Benefits Energy-efficient motors have better tolerance to
thermal and electrical stresses, the ability to operate at higher
temperatures, and the ability to withstand abnormal operating
conditions such as low voltage, high voltage or phase imbalance.

System Efficiency
Energy-efficient motors can also improve system efficiency, but end users must
consider the following factors:

Motors meeting higher efficiencies tend to run faster than their less
efficient counterparts.

Matching speeds to application need (such as pump flow) is

important to consider.

Drives may be required, which offers the opportunity to increase

system efficiency in applications with variable output requirements.
Variable frequency drives (VFDs) require further considerations for
optimum reliability and efficiency.

In some cases, mounting dimensions for motor into machinery may

be slightly different.

Case Study
The following case study graphically illustrates the impact of a premium-efficient
motor in a centrifugal pumping application.
Users should consider system changes to comply with the new EISA standard.
by William Livoti

Figure 1. Four separate scenarios for reducing energy consumption in a cooling

tower pumping system (Courtesy of WEG)

Figure 1 provides four separate scenarios for reducing energy consumption in

a cooling tower pumping system. The portrayed system is a typical closed
loop configuration where the discharge is being throttled over a range of
operation. The system in this example operates 24/7, 365 days per year. At

this particular load point, that means it operates 70 percent of the timeor
6,250 hours per year.
Columns 1 and 2 in Figure 1 indicate the various components factored into the
system efficiency calculation. Column A is the base condition where the
system operates 50 percent of the time. The component efficiencies for
the VFD and gearbox are at 100 percent because they were not used.
Under the base condition, the total power required is approximately 1,777 HP;
almost 356 HP is being lost (wasted) across a control valve. In addition, the
pump is operating back on the curve at 65 percent efficiency. Under these
conditions, the total system efficiency is 49 percent.
Column B provides the new operating conditions with the addition of a VFD.
The head required has been reduced to 150 feet because the loss across the
valve has been eliminated by reducing the speed of the pump to meet
required system demand. Motor efficiency remains the same, and a 2 percent
loss has been added as a result of heat generated across the drive. Note the
dramatic improvement in the overall system efficiency (81 percent) and the
total operating cost reduction from $414,306 to $187,360. The total cost
savings is $226,946 per year.
Column C addresses the impact on the system by improving the efficiency of
the pump. Nothing else in the system was changed.
Advertisement

The minimal improvement of the overall system efficiency (53 percent) results
from increasing the pump efficiency by 5 percent. The 50 feet of head loss
across the control valve remains, so the total power required is 1,650 HP. This
scenario does not present huge savings based on the cost of a new pump and
installation and potential piping changes. Factor in the ongoing reliability
issues, such as the pump operating back on the curve, and $29,593 would be
difficult to justify.
Column D identifies potential savings when motor efficiency is improved by 2
percent. Again, nothing has changed in the system with the exception of an
additional 5 feet of friction loss across the valve as a result of the reduced slip
in the premium-efficient motor (head increases to the square of the speed). In
this case, the system efficiency remains the same at 49 percent. Note that the

power required for the additional friction has increased to 330 HP. The total
power required was reduced to 1,650.2 HP (a reduction of 127 HP) with a total
savings of $518 per year.
References
1.
2.

EISA Standards Department of Energy

WEG Electric

Pages

first

previous

ABOUT THE AUTHOR

William Livoti is the power generation business development manager for WEG
Electric Corporation. Livoti may be reached at wlivoti@weg.net.

- See more at: http://www.pumpsandsystems.com/motors/august-2015understanding-system-efficiency-motor-driven-rotating-equipment?

page=2#sthash.gAz9p1nu.dpuf
- See more at: http://www.pumpsandsystems.com/motors/august-2015understanding-system-efficiency-motor-driven-rotatingequipment#sthash.ZpegQmMt.dpuf