Piping and Plant Design 

   
Academy   

 
 
Pump Suction Piping 
Rules 
 
 
 
Amir Razmi 
  2017   
Eccentric reducers and straight runs of pipe at pump suction | Hydrocarbon Processing | S... Page 1 of 4

COPYING AND DISTRIBUTING ARE PROHIBITED WITHOUT PERMISSION OF THE PUBLISHER

Eccentric reducers and straight runs of pipe at pump suction

09.01.2010 | Bloch, H. P., Hydrocarbon Processing Staff,

Keywords:

Questions relating to proper reducer application in centrifugal pump suction lines date back many decades. Until
his death (at age 84, in 1995), world-renowned pump expert Igor Karassik frequently corresponded with the writer
and other pump users on pump-related subjects. We rarely pass up an opportunity to highlight some of his
experience-based comments.

Once, a pump user referred to Fig. 1 and noted that this was quite typical of illustrations found in many textbooks.
In essence, Fig. 1 indicates that, with a suction line entering the pump in the horizontal plane, the eccentric reducer
is placed with the flat at the top. Available texts often give no indication as to whether the pumpage came from
above or below the pump.

Fig. 1. Illustration of eccentric reducer mounting from

Hydraulic Institute Standards.

Igor Karassik agreed that, if the supply source was from above the pump, the eccentric reducer should be installed
with the flat (horizontal) surface at the bottom. Entrained vapor bubbles could then migrate back into the source
instead of staying near the pump suction. If the pump suction piping entered after a long horizontal run or from
below the pump, the flat of the eccentric reducer should be at the top.1

Still, in many older texts it has been assumed that the pumpage source originated at a level below the pump suction
nozzle. Karassik reminded us that older Hydraulic Institute Standards commented on the suction pipe slope:

“...Any high point in the suction pipe will become filled with air and thus prevent proper operation of the pump. A
straight taper reducer should not be used in a horizontal suction line as an air pocket is formed in the top of the
reducer and the pipe. An eccentric reducer should be used instead.”

This instruction applies regardless of where the pumpage originates. Depending on the particulars of an installation,
trapped vapors can reduce the effective suction line cross-sectional area. Should that be the case, flow velocities
would tend to be higher than anticipated. Higher friction losses would occur and pump performance would be
adversely affected.

http://www.hydrocarbonprocessing.com/Article/2663961/Eccentric-reducers-and-straight-r... 3/22/2013
Eccentric reducers and straight runs of pipe at pump suction | Hydrocarbon Processing | S... Page 2 of 4

In the case of a liquid source above the pump suction, and particularly where the suction line consists of an eccentric
reducer followed by an elbow turned vertically upward and a vertical pipe length—all assembled in that sequence
from the pump suction flange upstream—it will be mandatory for the eccentric reducer flat side to be at the bottom.
That said, Fig. 2 should clarify what reliability-focused users need to implement.

Fig. 2. Suggested modifications for eccentric reducer

mountings.

Also, whenever vapors must be vented against the flow direction, the line size upstream of any low point must be
governed by an important criterion. The line must be a diameter that will limit the pumpage velocity to values below
those where bubbles will rise through the liquid.

In general, it can be stated that wherever a low point exists in a suction line, the horizontal piping run at that point
should be kept as short as possible. In a proper installation, the reducer flange will thus be located at the pump
suction nozzle and there is usually no straight piping between reducer outlet and pump nozzle. Straight pipe lengths
are, however, connected to the eccentric reducer inlet flange. On most pumps, one usually gets away with five
diameters of straight length next to the reducer. In the case of certain unspecified velocities and other interacting
variables (e.g., viscosity, NPSH margin, pump style, etc.), it might be wise to install as many as 10 diameters of
straight length next to the reducer inlet flange. The two different rules-of-thumb explain seeming inconsistencies in
the literature, where both the 5 and 10-D rules can be found. HP

LITERATURE CITED

1
Karassik, Igor J., Centrifugal Pump Clinic, 2nd Ed., Marcel Dekker, Inc., 1989.

The author

Heinz P. Bloch is HP’s Equipment/Reliability Editor. The author of 17 textbooks and over 470 papers or articles,
he advises process plants worldwide on reliability improvement and maintenance cost reduction opportunities. His
coauthored Bloch/Budris text, Pump User’s Handbook, is comprehensive and very widely used. Find the 2nd
edition under ISBN 0-88173-517-5. He can be contacted at HB@HydrocarbonProcessing.com.

akshay
01.05.2013

Dear,
what benefits of eccentric reducer with 1",2" pipe at suction of C F pump? also which type of material service &
parameter required in pump (Viscocity,spe.gravity) are more suitable for eccentric reducer?

http://www.hydrocarbonprocessing.com/Article/2663961/Eccentric-reducers-and-straight-r... 3/22/2013
Eccentric reducers and straight runs of pipe at pump suction | Hydrocarbon Processing | S... Page 3 of 4

Mr.Piper007
08.30.2012

is there any possibility that concentric reducers be used on horizontal line piping?

mojtaba javan
06.11.2012

tanks for this paper.

www.mojtabajavan.ir

CHARLES
02.09.2012

Thanx for your info. Now i know.

Regards

Shankar
12.10.2011

Why should provide the eccentric reducer in the pump suction and why not provide concentric reducer????......if any
special purpose of providing the eccentric????...... pls give suggestion.....

02.28.2011

As to why the eccentric reducer should be mounted with flat bottom at bottom for pumpage Source above pump
suction is still very unclear to me..Kindly elaborate the same .

Rgds

Anirban

09.23.2010

Dear Sir,

Please also include some guide line for

(1) straight length requirement for various types of Pumps i.e. OH2 / BB1 / BB2 / Multistage Pumps and for type of
Suction i.e. single / double suction

(2) If due to Lay out constrain can we go for 3D stratght length for suction

Waitng for your reply

Thanks,
Kiran

09.21.2010

http://www.hydrocarbonprocessing.com/Article/2663961/Eccentric-reducers-and-straight-r... 3/22/2013
Eccentric reducers and straight runs of pipe at pump suction | Hydrocarbon Processing | S... Page 4 of 4

Thank God you clarified. There are innumerable references & books propagating the incorrect reducer connection
And there must be millions of incorrect installations. I as a project man (non Mech.) and going by basic principals
had argued with Pump specialists over this anomaly when a pump was misbehaving but was shown the text book.

09.20.2010

I had the priviledge to attend a pumping problems class led by Igor Karrasik. I still have the course notes and have
used them for over 30+ years.

http://www.hydrocarbonprocessing.com/Article/2663961/Eccentric-reducers-and-straight-r... 3/22/2013
 Innovative Mag-Drive, LLC
th
6911 West 59 Street
PUMP OUTLINE DRAWING Chicago, IL 60638
Advanced Sealless Pumps Tel: 773.586.6250 Fax:
773.586.6208

ANSI lb.
Pump Size D 2E1 2E2 F H O X Y CP SF DF
NO. (kg)
1.5 x 1 x 6 AA 5.25 6 0 7.25 0.625 11.75 6.5 4 8.7 1.5 1.0 75
(133) (152) 0 (184) (16) (298) (165) (102) (221) (38) (25) (34)
3 x 1.5 x 6 AB 5.25 6 0 7.25 0.625 11.75 6.5 4 8.7 3.0 1.5 84
(133) (152) 0 (184) (16) (298) (165) (102) (221) (76) (38) (38)
3 x 2 x 6A -- 5.25 6 0 7.25 0.625 11.75 6.5 4 11.3 3.0 2.0 85
(133) (152) 0 (184) (16) (298) (165) (102) (287) (76) (51) (39)
1.5 x 1 x 8 AA 5.25 6 0 7.25 0.625 11.75 6.5 4 11.3 1.5 1.0 129
(133) (152) 0 (184) (16) (298) (165) (102) (287) (38) (25) (59)
3 x 2 x 6B -- 5.25 6 0 7.25 0.625 11.75 6.5 4 11.3 3.0 2.0 139
(133) (152) 0 (184) (16) (298) (165) (102) (287) (76) (51) (63)
3 x 1.5 x 8 A50 8.25 9.75 7.25 12.5 0.625 16.75 8.5 4 11.3 3.0 1.5 142
(210) (248) (184) (318) (16) (425) (216) (102) (287) (76) (38) (64)
3x2x8 A60 8.25 9.75 7.25 12.5 0.625 17.75 9.5 4 11.3 3.0 2.0 148
(210) (248) (184) (318) (16) (451) (241) (102) (287) (76) (51) (67)
4x3x8 A70 8.25 9.75 7.25 12.5 0.625 19.25 11 4 11.3 4.0 3.0 182
(210) (248) (184) (318) (16) (489) (279) (102) (287) (102) (76) (83)
2 x 1 x 10 A05 8.25 9.75 7.25 12.5 0.625 16.75 8.5 4 11.3 2.0 1.0 205
(210) (248) (184) (318) (16) (425) (216) (102) (287) (51) (25) (93)
3 x 1.5 x 10 A50 8.25 9.75 7.25 12.5 0.625 16.75 8.5 4 11.3 3.0 1.5 211
(210) (248) (184) (318) (16) (425) (216) (102) (287) (76) (38) (96)
3 x 2 x 10 A60 8.25 9.75 7.25 12.5 0.625 17.75 9.5 4 11.3 3.0 2.0 223
(210) (248) (184) (318) (16) (451) (241) (102) (287) (76) (51) (101)
4 x 3 x 10 A70 8.25 9.75 7.25 12.5 0.625 19.25 11 4 11.3 4.0 3.0 235
(210) (248) (184) (318) (16) (489) (279) (102) (287) (102) (76) (107)
4 x 3 x 10H A70 8.25 9.75 7.25 12.5 0.625 19.25 11 4 11.3 4.0 3.0 248
(210) (248) (184) (318) (16) (489) (279) (102) (287) (102) (76) (113)
6 x 4 x 10 A80 10 9.75 7.25 12.5 0.625 23.5 13.5 4 11.3 6.0 4.0 255
(254) (248) (184) (318) (16) (597) (343) (102) (287) (152) (102) (116)
• Not to be used for construction
• Dimensions are: inches (mm)

Page 1 of 1 Pump Outline Drawing-082001

Pump Suction Conditions | Cavitation & Design | Empowering Pumps Page 1 of 6

Name

Email Address

Company Profiles
Events & Training
Latest Products

Search Empowering Pumps

About | Pump Topics | News | Papers | Industry Pros | Events & Training | Directories | Industry Support | Career Center |

Chemical & Abrasives   |   HVAC   |   Mining & Construction   |   Oil & Gas   |   Power & Energy   |   Water & Wastewater   |   Sanitary   |   Cavitation
& Design   |   Education & Training   |   Energy Efficiency   |   Life Cycle Costs   |   Maintenance & Reliability   |   Materials & Coatings   |   Pump
Control   |   NPSH   |   Sealing   |   Standards   |   Valves   |   Vibration, Balancing & Alignment

ARTICLE

Back to Articles

Pump Suction
Conditions

If a wide receiver has the right

speed and good hands, all
that's needed from the quarter-
back is to throw the ball
accurately and they'll probably
finish up with some
good yardage and maybe even
a touch-down. Believe it or not, What is the purpose of
much the same is true of a minimum flow? Read More...
pump and it's suction What questions
conditions. If the pump has the about bolted flange
connections do you
right speed and is the right
most f Read More...
size, all that's required from
What is NPIP? Read More...
the "quarter-back" is to deliver the liquid at the right pressure and with an even laminar flow into the eye of the impeller. If
the quarter-back's pass is off target, badly timed, or the balls turning end over end in the air, the receiver may not be able
to hold on to it and there's no gain on the play.

When that happens, the quarter-back knows he didn't throw it properly and doesn't blame the receiver. Unfortunately, that's
where the comparison ends. The engineering "quarter-backs" tend to blame the pump even when its their delivery that's
bad !!

Just as there are specific techniques a quarter-back must learn in order to throw accurately, so are there certain rules which
will ensure the liquid arrives at the impeller eye with the right pressure and the even laminar flow necessary for reliable
operation.

Rule #1. Provide Sufficient NPSH

Without getting too complicated on a subject about which complete books have been written, let's just accept the basic
premise that every impeller requires a minimum amount of pressure energy in the liquid being supplied in order to provide
adequate performance without cavitation difficulties. This pressure energy is referred to as Net Positive Suction Head.

The NPSH is supplied from the system and is solely a function of the system design on the suction side of the pump.
Consequently it's controlled by the system designer.
Follow PumpingTweets
To avoid Cavitation, the NPSH available from the system must be greater than the NPSH required by the pump, and the
biggest mistake that can be made by a system designer is to succumb to the temptation to provide only the minimum
required at the rated design point. This leaves no margin for error on the part of the designer, or the pump, or the system
and has proved to be a costly mistake on many occasions.

Figure 1

http://www.empoweringpumps.com/articles.php?parent_page_id=INDUSTRY_TOPICS&... 3/22/2013
Pump Suction Conditions | Cavitation & Design | Empowering Pumps Page 2 of 6

In the simple system as shown in Fig. 1., the NPSH Available can be calculated as follows:-

Fig. 2

NPSHA = Ha + Hs - Hvp - Hf

where, Ha = The head on the surface of the liquid in the tank. In an open system like this, it will be atmospheric pressure.

Hs = The vertical distance of the free surface of the liquid above the centre-line of the pump impeller. If the liquid is
below the pump, this becomes a negative value.

Hvp = The vapour pressure of the liquid at the pumping temperature expressed in feet of head.

Hf = The friction losses in the suction piping.

The NPSH Available may also be determined with the following equation:-

NPSHA = Ha + Hg + V2/2g - Hvp

where, Ha = Atmospheric pressure in feet of head.

Hg = The guage pressure at the suction flange in feet of head.

V2

---- = The velocity head at the point of measurement of Hg.(Guage readings do not include velocity head.)

2g

Rule #2. Reduce the Friction Losses

When a pump is taking it's suction from a tank, the pump should be located as close to that tank as possible in order to
reduce the effect of friction losses on the NPSH available. This is usually accomplished by using a larger diameter line to
limit the linear velocity to a level appropriate to the particular liquid being pumped. Many industries work with a maximum
velocity of approximately 5 ft./sec., but this is not always acceptable.

When considering the proximity of the pump to the tank it is also imperative that it be far enough away to ensure that
correct piping practice can be followed.

Rule #3. No Elbows on the Suction Flange

Much discussion has taken place over the acceptable configuration of an elbow on the suction flange of a pump. Let's
simplify it. There isn't one !

http://www.empoweringpumps.com/articles.php?parent_page_id=INDUSTRY_TOPICS&... 3/22/2013
Pump Suction Conditions | Cavitation & Design | Empowering Pumps Page 3 of 6

Fig. 3

There is always an uneven flow in an elbow and, when one is

installed on the suction of any pump, it introduces that uneven flow
into the eye of the impeller. This can create turbulence and
air entrainment which can result in impeller damage and vibration.

When the elbow is installed in a horizontal plane on the inlet of a double suction pump, uneven flows are introduced into
the opposing eyes of the impeller and essentially destroy the hydraulic balance of the rotating element. Under these
conditions the overloaded bearing will fail prematurely and regularly if the pump is packed. If the pump is fitted with
mechanical seals, the seal will usually fail instead of the bearing, but just as regularly and often more frequently.

Fig. 4

The only thing worse than one elbow on

the suction of a pump is two elbows on the
suction of a pump - particularly if they are
positioned in planes at right angles to each other.
This creates a spinning effect in the liquid which
is carried into the impeller and causes
turbulence, inefficiency and vibration.

A well established and effective method of ensuring a laminar flow to the eye of the impeller is to provide the suction of the
pump with a straight run of pipe in a length equivalent to 5 to 10 times the diameter of that pipe. The smaller multiplier
would be used on the larger pipe diameters and vice versa.

Fig. 5

Rule #4. Stop Air or Vapour Entering the Suction Line

Any high spot in the suction line can become filled with air or vapour which, if transported into the eye of the impeller, will
create an effect similar to cavitation and with the same results. Services which are particularly susceptible to this situation
are those where the pumpage contains a significant amount of entrained air or vapour, as well as those operating on a
suction lift, where it can also cause the pump to lose it's prime.

http://www.empoweringpumps.com/articles.php?parent_page_id=INDUSTRY_TOPICS&... 3/22/2013
Pump Suction Conditions | Cavitation & Design | Empowering Pumps Page 4 of 6

Fig. 6

A similar effect can be caused by a concentric reducer. The suction

of a pump should be fitted with an eccentric reducer positioned with
the flat side uppermost as shown in Fig. 6.

If a pump is taking its suction from a sump or tank, the formation of vortices can draw air into the suction line. This can
usually be prevented by providing sufficient submergence of liquid over the suction opening, and a bell-mouth design on
the opening will reduce the amount of submergence required. This submergence is completely independent of the NPSH
required by the pump. It is worthwhile noting that these vortices are more difficult to trouble-shoot in a closed tank simply
because they can't be seen as easily.Great care should be taken in the design of a sump to ensure that any liquid
emptying into the sump does so in such a manner that air entrained in the inflow does not pass into the suction opening.
Any problem of this nature may require a change in the relative positions of the inflow and outlet if the sump is large
enough, or the use of baffles.

Rule #5. Correct Piping Alignment

Piping flanges must be accurately aligned before the bolts are tightened and all piping, valves and associated fittings should
be independently supported without any strain being imposed on the pump. Any stresses imposed on the pump casing by
the piping reduces the probability of satisfactory performance. 

Under certain conditions the pump manufacturer may identify some maximum levels of forces and moments which may be
acceptable on the pump flanges.

In high temperature applications, some piping misalignment is inevitable owing to thermal growth during the operating
cycle. Under these conditions, thermal expansion joints are often introduced to avoid transmitting any piping strains to the
pump. However, if the end of the expansion joint closest to the pump is not anchored securely, the object of the excercise is
defeated as the piping strains are simply passed through to the pump.

Rule #6. When Rules 1 to 5 have been ignored, follow Rules 1 to 5.

Piping design is one area where the basic principles involved are regularly ignored resulting in problems such as hydraulic
instabilities in the impeller which translate into additional shaft loading, higher vibration levels and premature failure of the
seal or bearings. As there are many other reasons why pumps could vibrate, and why seals and bearings fail, the trouble is
rarely traced to incorrect piping.

It has been argued that, because many pumps are piped incorrectly and most of them are operating quite satisfactorily,
piping procedure is not important. Unfortunately, satisfactory operation is a relative term and what may be acceptable in
one plant, may be totally inappropriate in another. It should also be noted that the piping system is rarely placed under
scrutiny as a problem source when trouble-shooting a pump failure.

Even when 'satisfactory' pump operation is obtained, that doesn't automatically make a questionable piping practice correct,
it merely makes it lucky.

The suction side of a pump is much more important than the piping on the discharge. If any mistakes are made on the
discharge side, they can usually be accommodated by increasing the performance capability from the pump. Problems on
the suction side however, can be the source of ongoing and expensive difficulties which may never be traced back to that
area.

In other words, if your "receivers" aren't performing well, is it their fault ?

Or does the "quarter-back" need more training ?

http://www.empoweringpumps.com/articles.php?parent_page_id=INDUSTRY_TOPICS&... 3/22/2013
Pump Suction Conditions | Cavitation & Design | Empowering Pumps Page 5 of 6

Ross Mackay is an internationally renowned expert in pumping reliability. He specializes in helping companies increase their
pump asset reliability and reduce operating and maintenance costs through pump training programs. He is the author of
“The Practical Pumping Handbook”, and “The Mackay Self-Directed Pump Reliability Training System”. He can be reached at
1-800-465-6260 or by visiting

www.practicalpumping.com

Written by: Ross Mackay - Internationally renowned expert in pumping reliability

More Articles from this Author

1 comments for this entry

David Muhs
January 13th, 2012 on 4:51 PM

We have an impeller in our InstaPrime™ that has never shown any sign of cavitation in more than 14 years of use.
We pump flows from 2 GPM continuously to a max of 3500 GPM at various suction lifts up to 30' and heads from 5' to shut
off head with now problems. We are told it can't be, but it is

Leave a comment

Viewer’s Full Name:

E-mail:

Comment:

Security   (Type the security code)

C d
Cannot Read? Click Here to generate a new one.

Post Comment

INDUSTRY LINKS:
Pump Systems Matter
FSA
SMRP

http://www.empoweringpumps.com/articles.php?parent_page_id=INDUSTRY_TOPICS&... 3/22/2013
Pump Suction Conditions | Cavitation & Design | Empowering Pumps Page 6 of 6

Contact Info Resources Topics of Interest

Charli K. Matthews Subscribe Cavitation & Design NPSH

Founder & President Advertise Education & Training Controls & Monitoring
P.O. Box 2313 Submit an Article Energy Efficiency
Tuscaloosa, AL 35403 Networking Partners Life Cycle Costs
205-391-8422 PumpScout Maintenance & Reliability
charli@empoweringpumps.com

About | Pump Topics | News | Papers | Industry Pros | Events & Training | Directories | Industry Support | Career Center

© Empowering Pumps. All Rights Reserved. Designed and Hosted by

http://www.empoweringpumps.com/articles.php?parent_page_id=INDUSTRY_TOPICS&... 3/22/2013
 QUICK
RESPONSE
Saving life and property through effective licensing, plan review,
and inspection of fire protection systems.
June 2008

FIRE PUMPS – SUCTION PIPING ARRANGMENT

Horizontal turns into a horizontal fire pump are problematic. It is important that the water entering a
horizontal fire pump load the fire pump impeller evenly. As water goes around a turn, momentum
Water pressure increases here causing
pushes the water to one side. If the water entering the
a greater flow to one side of the impeller Impeller suction flange does not even out, more water will push
to one side of the impeller. The extra load of that water
will cause the impeller to spin out of balance and
Suction
elbow
damage the pump. The figure to the left is a plan view
section showing the unbalanced loading of a double
suction impeller due to uneven flow through an elbow
adjacent to the pump. It is extremely important that
turbulence and changes in water flow direction are
Water
flow

Suction flange
carefully controlled close to the pump suction flange.
Top View

The ideal arrangement is when the water flows directly into the pump suction flange. This direct entry
minimizes the turbulence of the water and allows the impeller to load evenly. Unfortunately, it is not
uncommon that the water supply does not line up directly with the pump suction flange. This requires
alternative piping arrangements.

Vertical changes in direction do create some turbulence. However

the region of instability is in the vertical plane thus having no
effect on the loading of the fire pump impeller. As the water goes
into the impeller it crosses evenly across the horizontal axis of the
impeller. NFPA 20 - Standard for the Installation of Stationary
Pumps for Fire Protection, allows vertical direction changes Suction Discharge
directly on the suction flange of a horizontial fire pump.

As stated eailer, horizontal turns into a horizontal fire pump

Suction are problematic. To allow the water flow to straighten out,
NFPA 20 section 5.14.6.3.2 requires a straight run of pipe
prior to the suction flange. The length of this straight run
shall be greater than 10 times the diameter of the suction
pipe. This is measured from the end of the fitting to the fire
pump suction flange. The suction control valve and
Greater than 10 times
Discharge
eccentric reducer, if installed, are allowed to be included as
the suction pipe diameter
part of the straight run measurement.

Quick Response is presented monthly by the

Minnesota State Fire Marshal – Fire Protection Section
www.fire.state.mn.us
 RULES TO FOLLOW TO AVOID PUMP PROBLEMS
(Why Can’t Small Pumps Get Any Respect?)

By Flint Evans, President, Lang Engineering Equipment

Most pump problems are due to suction issues. In all the years I have been in the pump
business I have only found one instance of a pump problem that was related to the
discharge, other than of course pumps that have had a discharge valve shut while the
pump was running. Shutting a discharge valve off on a pump causes the fluid remaining
in the pump to get very hot and damages the housing, bushings, seals, etc. Hence, the
focus of this article will be about proper pump installations in regards to pump suction
conditions. I will first cover six basic rules and then some additional thoughts or
approaches to insure low maintenance and low cost operation of your pumps.

Before discussing the first rule about providing sufficient NPSH for the pump, we need to
discuss the term the concept of Feet of Head. Pumps don’t suck, rather the pump pushes
or throws the fluid out of the pump leaving a partial vacuum. Atmospheric pressure
(usually) then pushes the fluid into the pump. For centrifugals, this force is measured in
Feet of Head. Atmospheric pressure at sea level is 14.7 psia. At sea level, this can also
be stated in terms of 29.94 inches of mercury (barometers) or 33.9 feet of water. Hence,
at sea level we can say that a tank has 14.7 psia of pressure on it from atmospheric
pressure or we can say that it has 33.9 Feet of Head. The convention with centrifugals is
measure pressure in Feet of Head. A quick formula to convert between Feet of Head and
psi:

Pressure (lbs. per sq. in.) = Head in Feet X Specific Gravity of fluid
2.31

Where does the 2.31 come from? Divide 33.9 feet of head by 14.7 psia. Water has a
specific gravity of 1.0. So the formula always works. If the specific gravity is known for
the fluid that is being pumped, pressure gauge readings can be converted to Head in Feet,
which is useful for determining where the centrifugal pump is operating on its pump
curve. A final note before discussing NPSH. Pressure for centrifugal pumps (inlet and
outlet) is measured in Feet of Head and pressure for positive displacement pumps is
typically measured in psi. One of the exceptions is for air operated diaphragm (AOD)
pumps, which is a positive displacement pump where discharge pressure is measured in
Feet of Head.

Rule #1. PROVIDE SUFFICIENT NPSH

Simply put a pump will not operate properly without sufficient inlet pressure, the pump
will cavitate. Cavitation is caused by the rapid formation of vapor pockets (bubbles) in a
Page 2 of 5

flowing liquid in regions of very low pressure and collapsing in higher pressure regions,
often a frequent cause of structural damage to the propellers or other parts of the pump.
NPSHR or Net Positive Suction Head Required is the technical term used to determine
what pressure energy (in psia or feet of head) is needed to fill the pump inlet and not have
the pump cavitate. NPSHR is based on pump design. It is a characteristic which varies
primarily with pump speed and the viscosity of the fluid.

NPSHA or Net Positive Suction Head Available is based on the design of the system
around the pump inlet. The average pressure (in psia or feet of head) is measured at the
inlet port during operation, minus the vapor pressure of the liquid at operating
temperature. It indicates the amount of useful pressure energy available to fill the pump.
What we are asking is does the system provide enough pressure to fill the pump
completely and not cavitate (given the pump design, speed, fluid viscosity, etc.)? The
following is brief overview of NPSHA and how it is calculated.

NPSHA = Ha + Hs – Hvp – Hf
Where
Ha = Atmosphere Head is the head or pressure (pressure is measured in feet of head) on
the surface of the liquid in the tank that we are pumping out. In an open system
like this, it will be atmospheric pressure, 14.7 psi or 34 feet of water.

Hs = the vertical distance, measured in feet, between the free surface of the liquid to the
centerline of the pump impeller. If the liquid is below the pump, this becomes a
negative value.

Hvp = the vapor pressure of the liquid at the pumping temperature, expressed in feet of
head.

Hf = the friction losses in the suction piping, expressed in feet of head.

To put this formula in simpler terms think of NPSHA as being the result of atmospheric
head (pressure) pushing the fluid into the pump. The pump gains additional inlet head or
pressure if the liquid level is above the pump inlet or minus head if the liquid level is
below the pump. The fluid weight creates the pressure. The pump loses inlet head or
pressure from friction loss of the fluid moving through the suction pipe (small pipes or
long pipes have a lot of friction). And finally the inlet head or pressure is reduced by
vapor pressure. This is an issue if the fluid is evaporates easily or is very hot. So NPSHA
is atmosphere head plus or minus

One final note about NPSHR for a pump. Many pump manufacturers provide NPSHR
curves for their pumps. This curve is determined in labs using methodology as set forth
by Hydraulic Institute. The various points on this curve are determined by restricting the
inlet pressure with a valve. The restricted inlet pressure creates loss of flow or cavitation.
The NPSHR curve is drawn based upon the pump losing three percent of its rated flow.
At various flow points a vacuum reading is taken on the inlet of the pump. These points
are plotted below the pump curve showing the minimum inlet pressure the pump needs,
Page 3 of 5

but by definition this lost flow really is vapor bubbles and the pump is being damaged.
When installing a pump, insure that the inlet conditions are well above the NPSHR
requirements of the pump.

Rule #2. REDUCE THE FRICTION LOSSES

When a pump is taking its suction from a tank, it should be located as close to the tank as
possible. This reduces friction losses on the NPSH Available. However, the pump must
be far enough away that proper piping can supplied to the pump. Proper piping means
that a straight shot of pipe is supplied to the pump that is at least ten (10) diameters of the
pipe. We can this the 10D Rule. For example a minimum of 20” of straight pipe must be
immediately in front of the pump if the inlet pipe is 2” in diameter. Pipe friction is
reduced by using a larger diameter pipe. This limits the linear velocity, hence the friction
losses. Many industries use 5 to 7 feet/sec., but this is not always possible.

Rule #3. NO ELBOWS ON THE SUCTION INLET

It is never acceptable to install an elbow on a suction flange! There is always an uneven

flow in an elbow. When it is installed at the suction inlet of the pump, it introduces an
uneven flow into the eye of the impeller. This can introduce turbulence and air
entrainment, which may result in impeller damage and vibration. The only thing worse
than an elbow on inlet of a pump is two elbows. As mentioned above, the established
method of ensuring a laminar flow to the inlet of the pump is using the 10D rule, straight
pipe into the pump. This also means no valves, reducers, tees, etc.

Rule #4. STOP AIR OR VAPOR FROM ENTERING THE SUCTION LINE

Always check the suction line for leaks. As the pump operates it creates a partial
vacuum, which will suck air into the suction line. This will create an effect similar to
cavitation and with the same results. Another source of air in the suction line is the return
line in the tank if the pump is re-circulating the fluid through a system. If the return line
or supply line is above the tank liquid level, the liquid will become very become aerated.
This is a huge issue. Aerated tanks damage the pump just by creating cavitation like
conditions for the pump. The fix is to submerge the return or supply line. Return lines in
the tank can be to close to the outlet nozzle on the tank and can create the same issue.
The solution is relocating the return line or baffling the tank.

The presence of an air pocket in the suction line is another example of a cause for pump
troubles, which should never happen. Any high point in the suction line can become
filled with air and interfere with proper operation of the pump. This is particularly true
when the liquid being pumped contains an appreciable amount of air in the solution or of
entrained air and the pump is handling a suction lift. Long suction lines are too
frequently installed with improper pitch or with humps and high spots, where air can
Page 4 of 5

accumulate. If the liquid supply is below the pump the suction line should run up to the
pump. Straight reducers are definitely a no-no. Use an eccentric reducer, mounted with
the flat portion on top and sloping portion on the bottom. Install the other way around if
source of supply is above pump.

Another common problem is pumping a tank to low or having a short tank that in general
has low liquid levels above the outlet nozzle of the tank. If a pump is taking its suction
from a tank with low liquid levels, the formation of vortices can draw air into the suction
line and hence the pump. Vorticing can be eliminated, by installing a low liquid level
sensor to turn off the pump. Alternatively, install a bell-mouth connection on the tank
opening to lower the velocity on the tank outlet nozzle, hence lowering the liquid level
requirements to keep the tank from vorticing. Or a vortex breaker can be installed on the
discharge nozzle of the tank. They look very similar to the drain stopper in a modern
bathroom sink, except the diameter of the top round disk on top is 1½ times the size of
the ID of the tank discharge nozzle. Placing the tank outlet nozzle near the wall of the
tank will also help break a vortex.

The following table shows the minimum submergence required over opening unless some
of the suggested solutions mentioned above are employed:

Minimum Submergence Velocity of Outlet

of Outlet Nozzle Nozzle
1 foot 2 ft./sec.
2 feet 3.5 ft./sec.
3 foot 5 ft./sec.
4 foot 6 ft./sec.
5 foot 6.5 ft./sec.
6 foot 7.5 ft./sec.
7 foot 8 ft./sec.
8 foot 8.6 ft./sec.
9 foot 9.5 ft./sec.
10 foot 10 ft./sec.

The Hydraulic Institute states that typically one foot submergence for each foot per
second of velocity at the suction pipe inlet is recommended, with a suggested maximum
inlet velocity of six feet per second.

Rule #5. CORRECT PIPING ALIGNMENT

Piping flanges must be accurately aligned before the bolts are tightened and all piping,
valves and associated fittings should be independently supported, so as to place no strain
on the pump housing. Magnetically coupled pumps can have very short lives due to this
issue. Plastic pumps will not take these forces and moments. Piping strains can affect
seal life and bearings as well. Stress imposed on the pump casing by the piping reduces
the probability of satisfactory performance and pump life.
Page 5 of 5

ADDITIONAL THINGS TO WATCH

Sometimes when an electrician hooks up the motor is wired backward, meaning the pump
may be spinning the wrong direction. The result is low flow and head. Before the pump
is installed on the motor, quickly turn the motor on and off or “bump” it and check the
direction of rotation and compare that to the direction marked on the pump casing. If the
direction is wrong, reverse the electrical leads.

Special pumps are available from many manufacturers to handle slurries, yet most pumps
are not designed to handle foreign material without damage to the pump. For this reason
many applications have strainers or filters installed in front of the pump. The major
problem with this that users fail to monitor the pressure drop that develops across the
strainer or filter as it loads up with foreign matter. The result is high friction losses,
which result in inadequate NPSHA and the pump cavitates. The solution is to install
differential pressure drop instrumentation or a vacuum gauge or better yet switch, which
can automatically alarm the operators. Sometimes the damage from insufficient NPSH is
worse than if no strainer or filter was installed.

SUMMARY

When any of the above rules have been ignored, follow rules 1 through 5.

Lang Engineering has found that basic pipe design in small pumps is routinely ignored.
This results in shorter life in seals or bearings. Just because the pump works does not
mean that the pump is piped correctly! Even when the pump is working satisfactorily it
doesn’t mean that it is piped correctly, it merely makes it lucky.

The suction side of the pump is much more important than the piping on the discharge. If
any mistakes are made on the discharge side, they can usually be compensated, by
increasing the performance capability of the chosen pump. Problems on the suction side,
however, can be the source of ongoing and expensive difficulties, which may never be
traced back to rules 1 to 5.

The solution then on problem pumps may not be the pump, but the piping, the tank or any
of the other issues discussed above. Good luck and happy pumping!

References:

1. ROSS C. MACKAY
2. IGOR J. KARASSIK