Process Engineering Equipment Handbook Episode 3 Part 3 potx

Specific speed is a correlation of pump capacity, head, and speed at optimum efficiency, which classifies pump impellers with respect to their geometric similarity corresponding to the c

P-224 Steam production for different scenarios (Source: Ahlroth and Svedberg.)

FIG P-223 Power output for different scenarios (Source: Ahlroth and Svedberg.)

TABLE P-25 Simulation Results for the Hybrid System and the Tomlinson Boiler

Power Output Thermal Efficiency

Gas turbine Steam turbine

Black liquor gasifier malfunction 13.5 8.1 45.6

Steam Production Saturated Steam Produced in t/h

Black liquor gasifier malfunction 1.4

Production Minus Demand (t/h) “+” is surplus and “-” is deficit

Biomass gasifier malfunction -1.8 -11.9 Balance Black liquor gasifier malfunction Balance 1

1 Steam demand is lower in this case.

fuel used is difficult to refine to such a degree that it can be used in a gas turbine andthat it has a low heating value Secondly, comparing the efficiency of the Tomlinsonboiler and the steam turbine cycle to the efficiency of the hybrid-energy system, thepotential of the gas turbine in the pulp and paper industry is demonstrated For a

12 percent increase in fuel input to the pulp mill, the conventional Tomlinson boilerwith a steam cycle can deliver only 6.8 MW more power, whereas the hybrid-energysystem, thanks to its gas turbine, can deliver 13.3 MW more power This will makegas turbines an interesting option for pulp mills when biomass and black liquorgasification reaches full-scale commercial breakthrough

The hybrid energy system is restricted in many ways, since it has to fit in anexisting energy system with certain demands The decision not to use black liquorsynthesis gas in the gas turbine also limits the power output of the system Still,

it performs well In order to significantly increase the power output of a pulp mill,gasification of all the black liquor in an IGCC is needed For a pulp mill with equalflow of black liquor, 108 t DS/h, a power output of 96.6 MW is reported This doesnot include gasification of biomass, as in this study The total power output for such

a combined system would be 122.2 MW

Reference and Additional Reading

1 Bloch, H., and Soares, C M., Process Plant Machinery, 2d ed., Butterworth-Heinemann, 1998.

Pulsation Dampeners

Flow irregularities in fluid flow can cause pulsation in flow that is audible andcauses vibration in process machinery These irregularities can be corrected withpulsation dampeners, which essentially are vessels that provide the fluid (usuallythis problem is most common with gases) with enough residence time to steady theflow

A description of typical pulsation dampeners follows.*

Preferable manufacturing facilities are ASME-certified and are registered withthe National Board of Boiler and Pressure Vessel Inspectors, with fully accredited

“U” and “R” stamps Specialty services include fabrication with a wide range ofmaterials to pressures in excess of 10,000 psi Multicylinder dampeners can be builtwith tolerances of ±1/32in

See Table P-26 and Figs P-225 through P-232

Pumps

Pumps are the most common piece of equipment in any plant or process They varyconsiderably in operating parameter ranges and scope Manufacturers cancustomize a design for any operation with specific requirements, but, for the most

* Source: Peerless, USA.

TABLE P-26 Improved Efficiency Means Lower Costs

Pulsation/ Waves of compression (pulsation) set in motion  Reduce or eliminate equipment vibration by the periodic intake and discharge of gas or downtime/failure.

liquid from reciprocating compressors or  Increase safety: Reduce/ pumps can couple with resonant equipment eliminate possibility of rupture frequencies and produce damaging vibrations, caused by equipment vibration resulting in safety hazards, shortened fatigue.

equipment life, and increased downtime and  Lengthen equipment life by maintenance/replacement costs.

Pulsation eliminating rapidly cycling

pressures.

Solution The installation of pulsation dampeners Dampeners  Lower compressor piping costs.

or combination units eliminates these  Increase the accuracy of in-line problems Pulsation dampeners measurement equipment attenuate harmful pulsations and reduce without vibrations.

damaging vibrations by introducing a series  Save fuel: Less horsepower

of choking and expansion volumes required to run nonvibrating

systems.

Dynamic Fluctuating flow due to acoustic resonance can  Simplify and reduce frequency pressure drop cause significant additional pressure drop and/ of maintenance.

Solution The use of proven acoustic control techniques

Pulsation condensate.

diminishes excessive pulsation by effectively

dampeners  Protect machinery from solids decoupling cylinder excitation from the

resonant response of the piping.

Solid/liquid Formation of condensate and the presence of

entrainment solid particles may cause exchanger fouling

and/or compressor cylinder damage Combination Solution The combination dampener/separator effectively

dampener/

removes both solids and liquids in a singgle

separators vessel, providing maximum protection

at minimum cost.

FIG P-225 Natural gas transmission compressor station is typical of facilities where pulsation dampening equipment provides reliable long-term service (Source: Peerless.)

FIG P-226 General finite-element software is used to analyze structures such as this crosshead guide (Source: Peerless.)

FIG P-227 Specialized finite-element software is used to perform mechanical analysis of manifolds and complex piping systems (Source: Peerless.)

FIG P-228 Welding certified to the highest standards to the use of specified materials is prudent (Source: Peerless.)

resulting in thousands of successful structural piping designs (Source: Peerless.)

FIG P-230 An acoustical and mechanical engineer performs a field survey using state-of-the-art signal analyzers The company is known for its field service and works closely with customers to identify, solve, and correct vibration problems in compressor and pump installations (Source: Peerless.)

FIG P-231 Pulsation dampeners in low-pressure hydrogen service These dampeners, installed in

a fluid catalytic cracker (FCC) unit, ensure low refinery maintenance costs (Source: Peerless.)

P-217

part, picking a pump out of a catalogue will be possible After that, one has to choosebetween manufacturers, based on previous experience with items such as after-sales service, longevity of components, time between overhauls, and so forth Thereare various types of pumps such as centrifugal, piston, reciprocating, rotary, gear,lobe, and several others.

To help illustrate how varied features and types can be, information on common pump types in the process sector is included The sources in this sectionare primarily three different manufacturers, so contrasting operational/designphilosophies can be observed

Pump Theory*

Centrifugal pumps

The centrifugal pump is one of the most versatile types of machinery for industry.Every plant has in operation a multitude of pumps of this type, and moderncivilization could not be visualized without this equipment

Compared with other types of pumps, e.g., reciprocating and rotary pumps,centrifugal pumps operate at relatively high speeds and consequently are smallerand lighter when designed for comparable capacity and head Required floor space,weight, initial cost, and building costs are therefore reduced

Owing to their relatively high speed, centrifugal pumps are usually connected to the driver, the majority being electric-motor driven Having noreciprocating parts, centrifugal pumps are inherently balanced There are nointernal rubbing parts, and because running clearances are relatively large, wear

direct-is minimized The liquid direct-is delivered in a steady stream so that no receiver direct-is needed

to even out pulsations

In contrast to positive-type displacement pumps, centrifugal pumps develop alimited head at constant speed over the operating range from zero to rated capacity,and excessively high pressures cannot occur They can therefore be started against

FIG P-232 The pulsation dampeners are configured for dual nozzle suction manifolds Installed in

a natural gas reciprocating compressor transmission station in the northeastern United States, they reduce damaging vibrations and provide many years of trouble-free service at reduced costs (Source: Peerless.)

* Source: Demag Delaval, USA.

a closed discharge valve but should be operated at this condition for a minimumperiod (see subsection “Minimum Flow-Through Pump”) Generally, the bearingsare located outside the casing, so that the liquid does not come in contact with thelubricating oil and is not contaminated by it.

Classification. There are three general classes of pumps, depending on theconfiguration of the pump impellers:

 Centrifugal or radial-flow pump

 Mixed-flow pump

 Axial-flow pump

Those classes can be subclassified according to

 Number of stages: single-stage pump, multistage pump

 Arrangement of liquid inlet: single-suction pump, double-suction pump

 Position of shaft: horizontal pump, vertical pump (dry-pit type), vertical pump(submerged type)

Specific speed. Specific speed is a correlation of pump capacity, head, and speed at

optimum efficiency, which classifies pump impellers with respect to their geometric

similarity corresponding to the classification mentioned above (see also Fig P-233).Specific speed is a number usually expressed as*

where N s= specific speed

N= rotative speed, rpm

Q= flow, gal/min, at or near optimum efficiency

H= head, ft per stage

The specific speed of an impeller is defined as the revolutions per minute at which

a geometrically similar impeller would run if it were of such a size as to dischargeone gallon per minute against one foot head

Specific speed N or

N Q

N QH H

1 4

FIG P-233 Profile of several pump-impeller designs, ranging from the low-specific-speed radial flow on the left to the high-specific-speed impeller design on the right, placed according to where each design fits on the specific-speed scale (Source: Hydraulic Institute.)

* The value of H3/4 may be found in Table P-28.

Specific speed is indicative of the shape and characteristics of an impeller, and ithas been found that the ratios of major dimensions vary uniformly with specificspeed Specific speed is useful to the designer in predicting required proportionsand to the application engineer in checking the suction limitation of pumps.Impeller form and proportions vary with specific speed, as shown in Fig P-233.Pumps are traditionally divided into three classes: the centrifugal or radial-flow,the mixed-flow, and the axial-flow, but it can be seen from Fig P-233 that there is acontinuous change from the radial-flow impeller, which develops pressure principally

by the action of centrifugal force, to the axial-flow impeller, which develops most ofits head by the propelling or lifting action of the vanes on the liquid

In the specific-speed range of approximately 1000 to 4000, double-suctionimpellers are used as frequently as single-suction impellers

Figure P-233 gives values of H3/4 for the accurate determination of the specific

speed N s.Figure P-234 may be used to find specific speed with sufficient accuracy forpractical purposes without calculating the head to the three-fourths power Thepoint located by plotting the total head and capacity in gallons per minute at the

FIG P-234 Diagram for determination of specific speed Using the diagram, plot the head-capacity point; move from this point parallel to heavy lines to correct speed; from there move horizontally

to the left and read specific speed Example (dashed lines): H = 1,000 ft; Q = 10,000 gal/min; N =

3575 r pm; N s= 2015 (Source: Demag Delaval.)

design point is moved parallel to the sloping lines to the correct speed in revolutionsper minute The specific speed is read at the left of the diagram The procedure isillustrated by the heavy dashed lines.

For double-suction impellers, total flow is used in the calculation, althoughhistorically there has been considerable use of half of the flow in the equation.For multistage pumps, the head per stage is used in the specific-speed equation.Generally, this is the total head of the pump divided by the number of stages.Impeller performance curves are intimately related to their types or specificspeeds Higher-specific-speed impellers operating at partial loads have higherheads, require more horsepower, and have lower efficiency This is illustrated inFig P-235 on a percentage basis for the three general types mentioned above

Hydraulics

Definition of static, pressure, and velocity heads. One of the most useful relationships of

hydraulics is the continuity equation, which is based upon the principle that after

steady conditions in any system have been established, the weight flow of fluid perunit of time passing any point is constant Since most liquids are practically

incompressible, this may be put in equation form as Q = AV, where Q = flow, ft3/s;

A= cross-sectional area, ft2; and V = velocity, ft/s This equation may be rewritten

in the form V = 0.321Q/a, where V = velocity, ft/s; Q = volume flow, gal/min; and

a= area of pipe, in2 This equation is of importance in determining the velocity ofthe fluid at various points in either the piping or the pump itself

FIG P-235 Comparison of performance curves for various types of impellers: (1) propeller; (2) mixed-flow; (3) Francis; (4) radial (Source: Demag Delaval.)

lb/ft If an open manometer tube is set perpendicular to the flow, the fluid in it will

rise to a height equal to P/ g This is referred to as static-pressure head.

The kinetic or velocity head is the energy contained in a unit weight of the

fluid due to its motion and is given by the familiar expression for kinetic energy

V2/2g, where V = velocity, ft/s, and g = acceleration due to gravity (32.17 ft/s2) Values of this velocity head for various velocities are given in Table P-27 The head may be determined by taking the difference between a reading obtained

by a pitot tube facing the flow and a tube at right angles to the flow at the samelocation

The total energy of the fluid is equal to the sum of the three heads, or

Since energy cannot be created or destoyed, H is constant at any point of a closed hydraulic system (losses are neglected) This equation is known as Bernoulli’s

theorem The various forms of head may vary in magnitude at different sections,

but if losses are neglected, their sum is always the same

When liquid flows through a pipe, there will be a drop in pressure or head due

to friction losses

The total head H developed by a centrifugal pump is the measure of energy

increase of the liquid imparted to it by the pump and is the difference between thetotal discharge head and the total suction head Expressed in equation form,

H = h d - h s where H = total pump head, ft

h d = total discharge head, ft, above atmospheric pressure at datum elevation

h s = total suction head, ft, above atmospheric pressure at datum elevation

Note: h d and h sare negative if the corresponding pressures at the datum elevationare below the atmospheric pressure The common datum is taken through the pumpcenterline for horizontal pumps and at the entrance eye of the suction impeller forvertical-shaft pumps

Suction lift h sexists when the total suction head is below atmospheric pressure.The total suction lift as determined on test is the static pressure (vacuum) as measured by a mercury column expressed in feet of the liquid being pumped lessthe velocity head at the point of gauge connection This is equivalent to the staticlift plus entrance and friction losses in the piping if the water supply level is belowthe centerline of the pump In the case of water supply level above the pumpcenterline and at atmospheric pressure, suction lift will exist if the entrance andfriction losses in the suction piping are greater than the static head

Suction head h sexists when the total suction head is above atmospheric pressure.Total suction head as determined on test is the gauge reading expressed in feet ofthe liquid being pumped plus the velocity head at the point of gauge connection

In the case of a discharge into a closed vessel under pressure, the total dischargehead is equivalent to the static head corresponding to the water level in the vessel,

plus the gauge pressure expressed in feet of liquid corrected to this water level, plusthe friction losses in the discharge piping.

If the pump discharges to a level below the pump centerline or into a vessel undervacuum, the static head and pressure are taken as negative

Determination of head. Figure P-236 shows the usual arrangement for determining the total head developed by a pump working with suction pressure belowatmospheric Suction head is measured by means of a mercury U tube that is

connected to the suction through an air-filled tube Discharge head is measured by means of a Bourdon gauge connected to the pump discharge through a water-filled

tube

The total head for the arrangement shown in Fig P-236 is derived from thefollowing expressions:

where H = total pump head, ft

P g = discharge gauge reading, lb/in2

h sg = suction gauge reading, ftHg

g = specific weight of liquid being pumped, lb/ft3

gm = specific weight of mercury, lb/ft3

V d = average velocity in discharge pipe, ft/s

V s = average velocity in suction pipe, ft/s

z d and z s = elevation, ftFor water at 68°F, 1 lb/in2

= 2.3107 ft, and (144/g)P g = 2.3107 P g The specific gravity

of mercury at 68°F is gm/g = 13.57 Therefore, for water and mercury at 68°F, theformula simplifies to

g

V g

s

d

g d d

g d d

g

ggg

gg

Example: Determine the head developed, given a pump with a 10-in suction

and 8-in discharge with a capacity of 2000 gal/min The suction U tube reads

11 inHg and the discharge pressure gauge 50 lb/in2 The distance from the pumpcenterline to the U-tube connection is 7 in and to the center of the discharge gauge,+18 in

Therefore,

Substituting in the formula,

Determination of power required. The work required for pumping depends on the totalhead and the weight or volume of the liquid to be pumped in a given time This willgive a theoretical liquid horsepower (Lhp) as expressed in the following formula:

where w = lb liquid pumped per minute

H = total head, ft of liquid

When the liquid is water at 68°F weighing 62.318 lb/ft3

, the formula becomes

If a liquid other than water is pumped or water at a temperature other than 68°F, the formula must be corrected for the specific gravity of the liquid so that

where sg = specific gravity of the liquid referred to water at 68°F

If the total head is expressed as pounds per square inch, the specific gravity doesnot enter into the final equation, which is as follows:

The theoretical horsepower is less than the actual or brake horsepower because

of losses in the pump such as friction and leakage The efficiency of the pump is therefore the ratio of the liquid-horsepower output to the brake-horsepower (bhp)input, or

2

2

and the bhp required at the coupling will be

where h = pump efficiency, percent/100

If the entire unit is considered, it is necessary to include the efficiency of thedriver and any other associated equipment to arrive at the overall efficiency

Pump performance. When studying the flow of liquid through a pump impeller,three types of velocities must be considered The first is the velocity of a point on

the impeller and is designated by the symbol U The second is the velocity of the

fluid relative to the casing, known as the absolute velocity and designated by the

symbol C The third is the velocity of the liquid relative to the impeller, known as the relative velocity and designated by the symbol W The relative velocity W is found by taking the vector difference of the absolute velocities U and C Subscripts

are generally used with these symbols, the subscript 1 designating the velocities atthe inlet to the vanes and the subscript 2 at the outlet of the vanes The angle

between the vectors U and W is designated as b and is the angle that the vanemakes with a tangent to the impeller The angle a between vectors C and U

represents the angle at which the fluid enters or leaves the wheel This is illustrated

in Fig P-237 for a typical impeller

The ideal head developed by the wheel is given by the equation

where the first term represents the head due to the centrifugal action, the secondthat due to the change in the relative velocity, and the third that due to the change

in the absolute velocity The first two terms represent the pressure head that

is developed in the impeller, while the last is the velocity head developed in theimpeller and converted into pressure in the volute or diffuser This expression mayalso be written

The actual head developed by the pump will be less than the ideal owing to fluidfriction and shock losses and to a circulatory flow that takes place between thevanes.

The basic formula can be expressed in a simplified form for the actual headdeveloped by a pump impeller at rated design conditions as follows:

where H= actual head, ft

or the impeller diameter

The net effect of changes in the outside diameter of the impeller is similar to that

of varying the speed of the unit

The effect of changes in operating conditions may be summarized by the

following equations, where the subscript a refers to the condition while b refers to

the new

where n= speed, rpm

D= impeller outside diameter, in

These relations may be applied over the entire range of a pump characteristiccurve but should be used only for relatively small changes in speed or impellerdiameter

b a b a

b a b a

b a b a

=

D n

H

i = 1( 2 2cosa2- 1 1cosa1)

Example: Assume that a pump is delivering 2500 gal/min of water against a

head of 150 ft when running 1760 rpm with an efficiency of 81 percent The bhp isthen 117 The outside diameter of the impeller is 131

/2in What would theperformance be if the impeller diameter were reduced to 13 in and the pumpspeeded up to 1800 rpm?

Figure P-238 shows a typical performance curve for a centrifugal pump Itillustrates the variations of head, power, and efficiency as a function of capacity at

constant speed These curves are called characteristic curves of pump performance,

and they are determined experimentally for each pump type

Figure P-239 shows diagrammatically the characteristic curves of a pump at two

speeds N1and N2, where points of similar flow conditions are related according tothe formulas given above

Parallel and series operation. When pumping requirements are variable, it may bedesirable to install several small pumps in parallel rather than use a single largepump When the demand drops, one or more smaller pumps may be shut down,thus allowing the remainder to operate at or near peak efficiency If a single pump

is used with lowered demand, the discharge must be throttled (for constant speed),and it will operate at reduced efficiency Moreover, when smaller units are used,opportunity is provided during slack-demand periods for repairing and maintainingeach pump in turn, thus avoiding the plant shutdowns that would be necessary

= = 81 percent

hb ha

n n

D D

b a b a

= ÊË ˆ¯ ÊË ˆ¯ = 117 1 800

1 760

3 3

,,

b a b a

with single units Similarly, multiple pumps in series may be used when liquid must

be delivered at high heads

In planning such installations, a head-capacity curve for the system must first

be drawn The head required by the system is the sum of the static head (difference

in elevation and/or its pressure equivalent) plus the variable head (friction andshock losses in the pipes, heaters, etc.) The former is usually constant for a givensystem, whereas the latter increase approximately with the square of the flow The

resulting curve is represented as line AB in Figs P-240 and P-241.

Connecting two pumps in parallel to be driven by one motor is not a very commonpractice, and offhand such an arrangement may appear more expensive than a

FIG.P-239 Characteristic curves at two speeds N1and N2 (Source: Demag Delaval.)

FIG P-240 Head-capacity curves of pumps operating in parallel (Source: Demag Delaval.)

single pump However, it should be remembered that in most cases it is possible tooperate such a unit at about 40 percent higher speed, which may reduce the cost

of the motor materially Thus, the cost of two high-speed pumps may not be muchgreater than that of a single slow-speed pump

For units to operate satisfactorily in parallel, they must be working on the portion

of the characteristic curve that drops off with increased capacity in order to secure an even flow distribution Consider the action of two pumps operating in

parallel The system head-capacity curve AB shown in Fig P-240 starts at H static when the flow is zero and rises parabolically with increased flow Curve CD represents the characteristic curve of pump A operating alone; the similar curve for pump B is represented by EF Pump B will not start delivery until the discharge pressure of pump A falls below that of the shutoff head of B (point E) The combined

delivery for a given head is equal to the sum of the individual capacities of the twopumps at that head For a given combined delivery head, the capacity is divided

between the pumps and designated as Q A and Q B The combined characteristic curveshown on the figure is found by plotting these summations The combined bhp curve

can be found by adding the bhp of pump A corresponding to Q A to that of pump

B corresponding to Q B and by plotting this at the combined flow The efficiency curve of the combination may be determined by dividing the combined power

g(Q A + Q B )H/550 by the corresponding combined bhp (Q taken as ft3/s)

If two pumps are operated in series, the combined head for any flow is equal to

the sum of the individual heads as shown in Fig P-241 The combined bhp curvemay be found by adding the horsepowers given by the curves for the individualpumps Points on the combined efficiency curve are found by dividing the combined

fluid horsepower (H A + H B )Q g /550 by the combined bhp; Q is again in ft3/s

Pump performance for viscous liquids. As the viscosity of the liquid being handled

by a centrifugal pump is increased, the effect on the performance is a markedincrease in the bhp, a reduction in the head, and some reduction in the capacity

FIG P-241 Head-capacity curves of pumps operating in series (Source: Demag Delaval.)

Figure P-242 is taken from the Hydraulic Institute Standards and may be used

to estimate the magnitude of these effects for a particular liquid The diagramshould be used only for the conventional radial-type impeller and should not beextrapolated or used for nonuniform liquids such as gels, slurries, or paper stock

or when the net positive suction head is inadequate

The procedure for selecting a pump for a given head-capacity-viscosityrequirement when the desired capacity and head of the viscous liquid and theviscosity and specific gravity at the pumping temperature are known is as follows

Enter the diagram at the base with the desired viscous capacity Qvisand proceed

upward to the desired viscous head Hvisin feet of liquid For multistage pumps usethe head per stage Proceed horizontally (either left or right) to the fluid viscosity,

and then go upward to the correction curves Divide the viscous capacity Qvisby the

capacity correction factor C Q to get the approximate equivalent water capacity Q w

Divide the viscous head Hvisby the head correction factor C hfrom the curve labeled

“1.0 ¥ Q N ” to get the approximate equivalent water head H w Having this newequivalent water head-capacity point, select a pump in the usual manner Theviscous efficiency and the viscous bhp may then be calculated

FIG P-242 Performance correction diagram (Source: Hydraulic Institute.)

Select a pump for a water capacity of 790 gal/min at 109 ft head The selectionshould be at or close to the maximum efficiency point for water performance If thepump selected has an efficiency with water of 81 percent at 790 gal/min, then theefficiency for the viscous liquid will be

Evis= C E ¥ E W= 0.635 ¥ 81 percent = 51.5 percentThe bhp for pumping the viscous liquid will be

The procedure to determine the pump performance on a viscous liquid when itsperformance with water is known is as follows From the efficiency curve locate the water capacity (1.0 ¥ Q N) at which maximum efficiency is obtained From this capacity determine the capacities 0.6 ¥ Q N, 0.8 ¥ Q N, and 1.2 ¥ Q N Enter thediagram at the bottom with the capacity at best efficiency (1.0 ¥ Q N), go upward to

the head developed (in one stage) H wat this capacity, then horizontally (either left

or right) to the desired viscosity, and then proceed upward to the various correction

curves Read the values of C E and C Q and of C Hfor all four capacities Multiply each

capacity by C Q to obtain the corrected capacities Multiply each head by itscorresponding head correction factor to obtain the corrected heads Multiply each

efficiency value by C E to obtain the corrected efficiency values, which apply at the corresponding corrected capacities The head at shutoff can be taken asapproximately the same as that for water Calculate the viscous bhp from theequation

Net positive suction head. If the pressure at any point inside a pump drops belowthe vapor pressure corresponding to the temperature of the liquid, the liquid willvaporize These bubbles of vapor will be carried along to a point of higher pressure

where they will suddenly collapse This phenomenon is known as cavitation It is

accompanied by removal of metal in the pump, reduced flow, loss in efficiency, andnoise and hence should be avoided It occurs around the pump suction and inletedge of the vanes when the absolute suction pressure is low

The net positive suction head (NPSH) of a pump is the equivalent total head atthe pump centerline corrected for vapor pressure It is found from the equation

open to the atmosphere or the absolute pressure in the closed tank or condenserfrom which the pump takes liquid In a deaerating heater, it is normally the

saturated pressure at the existing water temperature H zis the height in feet of thefluid surface above or below the pump centerline If above, it is considered to be

plus since the suction head is then increased; if below, it is minus H vpis the head

corresponding to the vapor pressure at the existing temperature of the liquid H fisthe head lost because of friction and turbulence between the surface of the liquidand the pump suction flange

In designing a pump installation and purchasing a pump, there are two types

of NPSH to be considered One is the available NPSH of the system, and the other

is the required NPSH of the pump to be placed in the system The former is

determined by the plant designer and is based upon the pump location, fluidtemperature, etc., while the latter is based upon suppression pump tests of themanufacturer To secure satisfactory operating conditions, the available NPSHmust be greater than the required NPSH In higher-energy pumps such as boiler-feed pumps, values of NPSHA should be 1.5 to 2.0 times larger than NPSHR (asnormally measured with a 3 percent drop in head)

The calculation of available NPSH will be illustrated by two examples The head

corresponding to a given pressure is given by the equation H p = 2.31 p/sg, where

p is the pressure in pounds per square inch and sg the specific gravity of the

liquid

Assume that water at 80°F is to be pumped from a sump The unit is located at

an altitude of 800 ft above sea level, and the suction lift (from water surface to pumpcenterline) is 7 ft The pipe losses amount to 1 ft head What is the available NPSH?The atmospheric pressure at an altitude of 800 ft is 14.27 psia The specificgravity of the water at 80°F is 0.9984, and the vapor pressure is 0.5069 psia

Determine the available NPSH of a condensate pump drawing water from acondenser in which a 28-in vacuum, referred to a 30-in barometer, is maintained.The friction and turbulence head loss in the piping is estimated to be 2 ft Theminimum height of water in the condenser above the pump centerline is 5 ft The absolute pressure in the condenser is 30 - 28 = 2 inHg, or 0.982 lb/in2 Thecorresponding specific gravity is 0.9945

A third example is that of a deaerating heater having a water level 180 ft abovethe pump centerline The water temperature is 350°F The pipe friction loss is

12 ft Since the water in the deareator is in a saturated condition, this absolutepressure on the surface liquid equals the vapor pressure at 350°F Then

The required NPSH must be determined in most cases by means of suppressiontests The Hydraulic Institute has prepared a series of diagrams to estimate thisrequired head (see Figs P-243 to P-246 inclusive) These diagrams are not to beconsidered as the highest values that can be obtained by careful design, but theymay be used for estimating as they represent average results of good present-daypractice

The use of the diagrams is simple and may be illustrated by an example A suction pump operating at 3600 rpm delivers 1000 gal/min against a total head of

double-200 ft What should the minimum NPSH be for satisfactory operation? The specificspeed as found from Fig P-234 is 2200 By referring to Fig P-243, the pointcorresponding to this specific speed for double-suction pumps and a total dynamichead of 200 ft gives a 12-ft suction lift as the safe maximum If the same conditions

were applied to a single-suction pump with the shaft through the eye of the impeller,the safe minimum suction condition would require at least a 1 ft positive head (i.e.,the suction head would have to be at least +1 ft rather than -12 ft; hence, therequired NPSH would be 35 ft instead of 22 ft).

These curves are based upon handling clear water at 85°F and sea-levelbarometric pressure If the water temperature is higher, the difference in headcorresponding to the difference in vapor pressures between 85°F and thetemperature of the water pumped should be subtracted from the suction lift oradded to the suction head Also, if the unit is to be located above sea level, thedifference in head corresponding to the difference in atmospheric pressures should

be subtracted from the suction lift or added to the suction head

Thus, in the above example, if the water temperature is 140°F and the plant islocated at an altitude of 2000 ft, the correction for vapor pressure will be 2.889 -0.596 = 2.293 lb/in2, and the correction for altitude will be 14.69 - 13.66 = 1.03 lb/in2.The corresponding head change will be 2.31 (2.293 + 1.03)/0.9850 = 7.8 ft For the double-suction pump the maximum suction lift would be 12.0 - 7.8 = 4.2 ft, and for the single-suction pump the positive suction head would have to be 1.0 +7.8 = 8.8 ft

FIG P-244 Upper limits of specific speeds: single-suction shaft through eye pumps handling clear water at 85°F at sea level (Source: Hydraulic Institute.)

A series of diagrams (Figs P-247 through P-249) have been prepared by theHydraulic Institute to determine NPSH on the basis of the flow, operating speed,and discharge pressure for hot-water and condensate pumps They may also be used

to find the maximum permissible flow for a given available NPSH

Hot water. Two curves, Figs P-247 and P-248, have been prepared for pumps handling hot water at temperatures of 212°F and above These curves show therecommended minimum NPSH in feet for different design capacities and speeds.Figure P-247 applies to single-suction pumps and Fig P-248 to double-suctionpumps These curves serve as guides in determining the NPSH for hot-water pumpsand do not necessarily represent absolute minimum values

Net positive suction head for condensate pumps. Figure P-249 indicates NPSH forcondensate pumps with the shaft passing through the eye of the impeller It applies

to pumps having a maximum of three stages, the lower scale representing suction pumps and the upper scale double-suction pumps or pumps with a double-suction first-stage impeller

single-For single-suction overhung impellers the curve may be used by dividing the specifiedcapacity, if 400 gal/min or less, by 1.2, and if greater than 400 gal/min, by 1.15

FIG P-245 Upper limits of specific speeds: single-suction overhung impeller pumps handling clear water at 85°F at sea level (Source: Hydraulic Institute.)