| NET POSITIVE SUCTION HEAD AD-19 (ISSUE A) Few subjects relating to the applica: tion of positive displacement (p.d.) pumps are more discussed and less lunderstood than Net Positive Suction Head — NPSH, Untila few years ago it was felt that NPSH was a term re- served for centrifugal pumps handling water and that it was not a subject of concern when applying slow speed ad. pumps to simple low pressure transfer jobs. This is no longer a valid assumption, TODAY — with sophisticated proc- ‘esses using p.d. pumps for drawing lig ulds from vessels under high vacuum (oil purtying and recycling} with users wanting more capacity from smaller pumps (higher speeds); with more cus: tomers installing reserve fuel supply tanks far beneath concreted parking lots (long suction lines with a ity, and with @ growing interest in installations that are quiet (do not exceed OSHA noise levels), that are highly efficient (conserve energy) and that do not wear ‘out (are inexpensive to maintain) — yes, today we are concerned about NPSH since a lack of consideration for NPSH could spell TROUBLE in any of the situations just mentioned. Another reason for concern is that more and more spec sheets and requests for quotes are asking for or are giving NPSH values. Net Positive Suction Head — must be indicated "available" or “required” to be meaningful — is the pressure in feet of liquid absolute measured at the ump suction port, less the vapor pressure. Seems simple enough but the available” and “required” terms and the “absolute” pressure cause some problems. For pd. rotary pumps instead of using NPSH expressed in feet of iquid absolute, the Hydraulic In- stitute (Viking Is a’ member) uses ‘Available Net Inlet Pressure (ANIP) and Required Net Inlet Pressure (RNIP) ex- pressed in PSIA. NPSH and NIP are the ‘same thing but expressed in differant units, To avoid confusion and addition. al problems; we will stay with the more frequently used NPSH and feet of liquid absolute throughout this article. NPSH available is 2 function of everything in the system on the suction Side of the pump up to the suction port, everything includes the pressure on the surface of the liquid in the supply tank, the difference between the liquid level and the centerline of the pump port, line losses, velocity head and vapor pressure, NPSH required is bbased on everything from the suction ‘bor to the point in the pump where the pressure starts to increase: here, everything includes the entrance losses and the friction losses or pres- sure drops getting into the pumping elements Since NPSHa is the absolute pres- sure available less the vapor pressure Of the liquid, i is logical to reason that the NPSHa ‘should always be greater than the NPSHr. I this were not the case, there would be vapor formed in the suction area of the pump. This reasoning is sound, Thus for a pump 10 operate properly the NPSHa must be treater than the NPSHr. With NPSHa less than NPSHr, the pressure at some Point in the pump suction area will be legs than the vapor pressure of the liq- Lid and CAVITATION will take place in the pump. With the pressure in the pump below the vapor pressure, bubbles (pockets ‘or cavities of vapor) start 1 form in the liquid. They are carried along with the liquid til they get to a higher pressure region in the pump where they, the bubbles, collapse. This phenomenon is known as cavitation. It is the violent coollapse of the bubbies of vapor with the resulting shock that causes many of the damaging effects associated with cavitation — noise, vibration, eroded parts, short service life. These, plus reduced capacity and efficiency, possible pulsations and an unhappy User, all make it mandatory that NPSHa be greater than (>) NPSHT. See Figure 1 FIGURE 1. Parts from a Viking In- ternal Gear Pump delib pe: ance Is typical of that caused by severe cavitation. PAGE 1NPSH Not Exact Remember!!! All values must be MULAis applied. Values for vapor pres- When a value for NPSH — available expressedin he same units; feet of i sure and line losses are approximate or required — is given, there is the im- uid pumped. and are meant for istration ony. pression thatthe value is exact. Unfor- The part that velocity head (Hy) has tunately this is not always the case. to play in NPSHa calculations and Such items as the following all tend to measurements is somewhat controver make stated values of NPSH possibly _sial. Hv= YE where V equals velocity of less exact than we would Ike. 4. entrained alt or gases in the iq. the liquid at the suction port in feet per wi second and g equals the acceleration 2, he ably fs pump, partoulery The vaueor Hy is normal qu sal ot the pd. type, to handle some vapor™ , Shh ite apatus uiteet 2 see Table 1, From Table 2 we see that the value of Hy is 1.3 feet of liquid or 3. the fact that pumps handling dit- Jess for all but two Viking Heavy-Duty ferent liquids particulary some hydro- (658 10" St but wo Niking Heavy Duty carbons will operate satisfactorily with fess NPSha hen woud be requ tor (0 fo#al Ving, pur applications water or other test liquids. This situa Nesta Yatues determined by caleula. tion is true for centrifugal pumps perin- thong, formation in the Hydraulic Institute Standards and has been observed as INSTALLATION 1 being true for internal gear rotary _ Wewillcalculate the NPSHa for a typi pumps. Viking will be doing some cal installation to show how THE FOR: testing in this area in the near future, NPSH Available eet ra caleuating Note wnien we shoud Velocity-Velocity Head (Hv) review and understand betorecontinu- ‘olocity of Liquid gercegeemmennet [mame 1s]. Lelolaluls 4s wo would Ik, its the bost ava bia, short of actal test data, and pro. | Velolly Head vides a basis for comparing systoms | (Hy). Feet of 4nd selecting pumps and can give an Levis 25 | 20 | 56| 76| 10 |1.25]155] 187] 224] 350 Sort to a potentially troublesome in- Stallion. As we have indicated before, TABLE 2. Ne Ooeae iepnionebieeen Velocity (V) of Liquid in Feet/Second and Velocity Head (Hv) in the operating conditions and the liquid Feet of Liquid at the Suction Port for Viking Heavy Duty Internal ee Fe a ape sta pean Gear Rotary Pumps at Nominal Rated Conditions can be calculated from THE FORMU- Pont aTeD ‘woulo vevociry= \eLociry wenn, th NeSHa = Has He Hi Hy ce | ER pegs | atimctanyoer | reap utas wis | im | 15 24 09 tus | iz | 30 a7 34 solute pressure on the surface Kis | 2 & 57 50 of the auld in the supply tank ex Keias | 2 8 76 20 Dressed in fest of iqud pumped us | 2 138 28 28 jrtical distance in feet trom the vars | 2% | 135 90 13 ‘surface of tho hiquld in the Supply tas | 3 140 64 58 tank to the centrine of the pump uss | 3 200 87 12 eteton part adie Gaon cote ais | 4 300 78 20 tne, He's negative. asi | 3 0 56 2 Ht= triton losses in suction piping © 3 108 1 pressed in feet of hauid pumped csi05 | 10 a7 21 Ha veloclty heed atthe suction pot | musios | 1% | 20 a2 16 in toot of found pumped. Huses | i | 3 ar a Hyp = absolute vapor pressure of liquid hea ae rs ee oe ‘tpumping tompereture expresses | AKATGB | 2 so aa 8 in feet of laud pumped " 3 7 “trere 1 m0 common agrment among Gea =] = = apuely at shud bo use 1 celorrne ras | 8 +100 7A 73 esata whan ty ya re ses Banal sa pean Salo PAGE 2INSTALLATION 1 Hat H2—HI—Hvp ‘absolute pressure on liquid ‘atmospheric pressure in feet of liquid 14,7 PSIA*)2.31" HOPS) 088 SG. of Fuel Oil 38.8 feet of fuel oil tical distance trom liquid level to centerline of suction port (use maximum or worst ccondition—emply tank). Lig- Lid is below pumps, so Hz is negative. (0 feet of fuel oil Ht =fiction Joss is suction pipe size and length and includes elbows, valves, other fittings {as equivalent length Hi =289 feet of fuel oil see pres- ssure loss charts in Section 510 ol Viking General Catalog) Hyp =vapor pressure of fuel oll at ‘75°F. in feet of fue} ail ab- solute, Hyp _=1 foot of fuel oil (maximum). Ha (38.8) —Hz (10)—H1 (2.9) Hp (1) 24,9 feet of fuel ol “ithe altitude atthe instalation was 2000 oo instead of sea fvel, the atmospheric pressure would have beon 13.6 PSIA (see Table 3) and Ha would have been 36.1 fe of hei ol. TABLE 3. Effect of Altitude on ‘Atmospheric Pressure and Barometer Readings In practice, an installation similar to this would have had a lower NPSHa because of additional losses (greater Hi through a shut off valve and possi- bly a foot valve and/or a strainer. Bo- cause there is offen entrained air in No. 2 fuel oll, the vacuum reading at the pump should not exceed 15" Hg. under the worst conditions, even though the NPSH available might in- dicate higher lifts could be handled satistactorly. 11 the liquid being pumped in Insta tion 1 was changed fram No. 2 fuel ol to gasoline, the NPSHa would be quite different. With gasoline, the specific gravity would change to .71 and the vapor pressure 108.5 psia. The formula for calculating NPSHa remains the (14.7) (2.31 7 47.8 feet of gasoline 10" of gasolne 29' of gasoline 8.5 psia winter gasoline at 75°F.) =85x251 7 27.6 feet of gasoine Ha (47 @)—He 10) Ht (2), ~ Hyp (27.6) 7.3 feet of gasoline. Note the big diference the vapor pressure has made in the NPSH availa bie rom the same system, 249 foe! of fuel ol vs. 73 feet of gasoline The NPSHa ot 7.3" of gasoline would normally be less by an additional fric- tin losses through a shut of valvo, foot valve and strainer It would also be affected by changes In tan iovel tom. perature and source Of Supp) FIGURE 2. Comparison Be- tween Absolute and Gage Pressure Seales. NOTE: To convert pressure i ps to foet of Iauld mutioy by 2°31 and divde by specitic ‘grauity of the Hguid. To convert inehas of mercury to feet of liquid multiply by 1.133 and divide by specific gravity ofthe lqui. *From a practical standpoint, the normal it ‘should not exceed 6° when handling ‘asolne ‘Since it s mandatory that all calcula tions be made in or converted to the same pressure unit, an understanding of the more frequently used units is necessary before doing any work rolat Ing to NPSH, The need for converting becomes ‘obvious when itis remembered that at ‘mospheric pressure is often given in inches of mercury absolute ("Ha. abs) (or PSIA, vapor pressure in mm Hg. ab: solute (mm Hg. abs.) or PSIA, ton in feet of iquid and tine loss in PSI or feet of liquid. Figure 2 compares the two basic pressure systems — Ab- solute and Gage. Always keep in mind that the gage pressure system zero point (atmospheric pressure) varies ac- ‘cording to the actual elevation above ‘sea level of the pumping site and that the atmospheric pressure at any given’ site can vary +1” Ha. from an average value. Thus the standard atmospheric pressure of 14.7 PSIA or 29.9" Hg. abs, is good as a reference but actually would seldom be the pressure ala par: ticular site at any given time. Table 4 shows the factors to use for converting trom one pressure unit to another. Those shown are the most fre- quently used, other units are occa- sionally encountered. Conversions for these other units can be found in Engineering Handbooks and similar references. As the SI or modem metric system of measurement becomes more widely used in the United States, we can expect to see pressures ex: pressed in pascals (Pa) or kilopascals (Pa), Figure 3 gives a visual comparison between the more frequently used pres- sure units, The conversions are not quite as exact as the chart would ind cate, s0 actual conversion factors should be used when doing calculations for a particular system, VAPOR PRESSURE In the FORMULA used on page 2 to calculate the NPSH available in a sys- tem we included the vapor pressure factor and used typical values for the liquids involved. At that time we did not define or discuss vapor pressure. We will do that here since an understand- ing of vapor pressure is as vital to prop- erly comprehending NPSH as is a good understanding of the relationship be- tween the various pressure units. Vapor pressure is one of the physi cal properties of liquid. By definition it is the pressure exerted by the vapors ‘fa confined liquid. It varies with tem- perature. PAGE 3Liquids such as water and fuel oil putting some of the liquid above the that can be stored in open containers mercury in a barometer and noting the ‘at ambient temperatures have a rela- depression of the mercury column. For tively low vapor pressure (well below LP-Gas the vapor pressure can be de- almaspheric pressure). For liquids of termined by reading a gage attached this type the vapor pressure is normally to the vapor section ofa partially fled, expressed in mm Hg. abs. or PSIA. Lig- enclosed container. The temperature uids such 28 LP-Gas, Freons and am- should always be noted whenever such monia which must be stored in closed vapor pressure determinations are containers have a relatively high vapor made. pressure (considerably above atmos- _To help visualize another way that Bheric pressure), For such liquids the vapor pressure can be measured refer, vapor pressure is often expressed in to Figure 4. The procedure for finding Psic. the vapor pressure of aliquid using the For fuel oils and similar liquids the ®auioment is vapor pressure can be measured by 1. Fill quid chamber; close valve vi TABLE 4. Conversion Table. ‘wowes | KLoGRans rou Or | pensounne |" OF | gy | FERS, pr, | CENTIMETERS, | wenCURY, ‘ach, grea" | “rm ip 7s 3390 | 2990 | 1033 | 760 | 1013 | 14.70 1 | 08826 | 0030s | 22.41 | 290 |o4s86 1133 | 1 | 003483 | 254 | 329 | so12 281 | 2896 | 1 | 736.0 | e807 | 14.22 lo.o446 | 0.0304 | 0.00196 | 1 {0.1336 lo.oreas: jossas [0295 | ooro2 [74938 | 1 [01450 2307 | 2036 | 0.0703 | sis7 Jesos | 1 To convert {om a pressure unit nthe lft hand column mulply the numerical value {times the factors n the vertical coluran which shows the unt yau are converting to. For ‘example to conver 60 feel of water to PS! ao horizontally tothe right from the "Feet of ‘Water unit to the bor under PSI, the factor is 0.493. The 50 feet of water is then ‘multipied by 0.433 to get 216 PS) |when werking with liquids other than wat the Spectic Gravity of your liquid. {00KPa =" Bar 'Se0 Eng. Sect. 510.22 Conworsion Factors, and Catalog Sect. 141.17, Round off Conversion Factors to 4 significant places & no mare than fe places after cima. ‘multiply the conversion factors shown by 59.5 mm Ho x0.01935 = 14.696 = 14.70 PS! | H/0x0.4308 = 14.699: PAGE 4 2, Pull full vacuum on vapor cham: ber; close valve V2 8. Allow temperature to stabilize: ‘open valve V1 4. Read vapor pressure on the gage for manometer; record tempera: ture 5. Repeat this procedure at difer cent temperatures until enough points have been recorded to draw a curve ‘The Reid vapor pressure for gaso line and other volatile petroleum prod: Ucts Is determined using the same basic test vessel arrangement as shown in Figure 4. The procedure is, somewhat different from that outined ‘but the results are forall practical pur: poses the same. Details of the proce: ‘dure for determining Reid vapor pres: ‘sure are spelled out in ASTM standard 10328. The Reid vapor pressure is nor mally understood to be an indication of ‘the vapor pressure of the liquid at 10°F. The Reid vapor pressure isan impor- tant factor when selecting a gasoline blend for use during different seasons and for different locales to assure proper starting and performance of au fomobile engines. The Reid vapor pres sure for summer gasoline 's approxi: mately 9.5 PSIA while for winter gaso- line itis around 13.5 PSIA. The lower vapor pressure for summer gasoline tends 10 keep it from vapor locking in the summer heat. The higher vapor pressure for winter gasoline helps it vaporize in the carburetor at winter: time temperatures. Remember that the Reid vapor pressure is taken at 100°F. For comparison with gasoline the vapor pressure at 100°F. of several solvents and LP-Gases is shown below: Xylene — 16 mm Hg, abs, Water — 47 mm Hg. abs. Toluene — 54 mm He, abs. ‘Acetone — 391 mm Hg. abs. Bulane — 37 PSIG Propane — 170 PSIG Figures §, 6 and 7 show the Vapor Pressure — Temperature Curves for a umber of liquids frequently handled by Viking pumps, The torm volatile is sometimes en- ‘countered when working with vapor pressures and petroleum products. tis, fa rather general term which gives ‘some indication of the vapor pressure of @ liquid or more specifically, an in ication of how rapidly the liquid will vaporize. When comparing two liquids at the same temperature, the one with the higher vapor pressure would be considered the more volatile and would vaporize more rapidly. For example, ‘gasoline has a relatively high vaporPSIA Wg mma Psic abe ‘abe 20 a8 1000 5 atmospHer ATMOSPHERIC 14.7134. 760: {Sea Level) “Ho VACUUM 10 28——+ 800 + 1+ 280 ABSOLUTE 0: 0. 0. 2092: ZERO TPSIA. gO) mmHg Abs "abs vacuum FIGURE 3. Pressure Unit Comparison Chart. pressure and is considered mote voia- file than #2 fuel oll which has a very low vapor pressure ‘As mentioned earlier, the factors connected with NPSH determinations are not always precise. While the vapor pressure of a pure liquid is con: sistent and predictable, itis possible to have entrained or dissolved alr or gas in the liquid withthe result thatthe fluid facts as though it had a higher vapor pressure than figures for the pure lig: Lie indicate. In this case as the pres- sure in the suction system is reduced, any entrained air in the liquid will tend to expand or any dissolved gas will tend to be released. As this occurs, there willbe less liquid coming into the pump, resulting in a lower volumetric efficiency and possibly erratic flow, Pia GAGE yO Ye APOR To vacuum PUMP rasewoueTE® F— wow FIGURE 4. This situation happens most frequently Ina system where the liquid receives considerable agitation as it is being transported or unloaded or where the liquid is being continuously recircu- lated. A good example of a typical pumping application involving these problems is one handling #2 fuel ol. Air is offen entrained as the oll is moved from a jobber or distributor to the user in his heating oil system. The vapor pressure of #2 fuel oll at normal amb ‘ent temperatures is virtually nil, but be- ‘cause of possible problems resulting from expansion of alr entrained during hauling and recirculating itis seldom wise to operate a pump handling #2 fuel oll with @ vacuum greater than 15° Hg. Refer to Ad-22, Fuel Oli Systems. ‘Somewhat the same situation can ‘occur when handling volatile petrole- lm products since the liquid is proba: bly made up of many different trac: tions, each with a vapor pressure of its own, For such a liquid a slight vacuum might cause vaporizing of the lighter fractions, In summary, if the suction side of a system is so designed that the pump must develop or pull” a vacuum that results in an absolute pressure less than the vapor pressure of the liquid being handled, the liquid will tend to vaporize (form bubbles or boll"). This formation of vapor on the suction side ff the system reduces capacity and ‘can cause vapor lock. The collapse of the vapor bubbles on the discharge side of the pump causes noise, vibra- lion and rapid wear. This is the pheno- ‘menon known as cavitation, It's this phenomenon which we are trying to {avoid by developing an understanding (of NPSH, learning to calculate it and thus encourage system designs which provide an NPSH available greater than the NPSH requirement of the pump. NPSHA > NPSHR. NPSHR — Net Positive Suction Head Required — Is another way of in- dicating the pressure loss within the pump itself, Liquid cannot flow from fone point to another without a loss of pressure. For pipe this pressure loss or {drop has been calculated, checked by test and tabulated in charts or plotted in curves. The loss is expressed in PSI peer foo! of pipe length or in feet of head loss per 100 feet of pipe. Data is available for all standard pipe sizes and for a viscosity range from water thin to several hundred thousand SSU. A study of this informa: tion readily shows that the pressure loss increases as the rate of flow in- ‘creases for a given pipe size and vis: cosity and that the loss increases as the viscosity increases for the same pipe size and flow rate. To determine the pressure loss for tow through a pipe is relatively easy, Calculations have also been made and tests conducted to determine the pres- sure loss through pipe fitings, various ‘types of valves and other items found iia piping system. This data, as that for pipe, shows increasing loss within creasing tlow rate and viscosity. For low viscosity liquids with turbu- lent flow, the loss through a fitting oF valve Is expressed as the loss in so ‘many feet (equivalent length of straight pipe. For laminar flow the pressure Joss through the same fitting or valve may be expressed as a percentage of the equivalent length of pipe loss for turbulent flow, the percentage de- creasing as the viscosity Increases. ‘See Engineering Section 510.12. A study of the lass information for fittings ‘and valves indicates that it may not be fas accurate and easy to determine as that for straight pipe. This is logical when we consider that the liquid flow- ing through a fitting or valve may ex: perience a drastic change in direction and may go through sudden contr tions andior enlargements The point of this discussion is that the more complex the path of liquid flow, the more difficult itis to accurate ly determine the pressure loss be- PAGE 51% 2.3 i i [arom Pressure —resratarore comves| [25 macs Teuegstne coves Ps, a 1.5, Te) aa 5 se os : aa hn i | |e j 8% me gr ze . Ee i . a a [Es 238 5 Se. 28.5 2 pal is aT - i oe ae aa "i HE 28 " E as viKiNG ume TTT FIGURE 5. FIGURE 6. tween any two points short of conducting tests under actual TOT eta eoaton. oT aapsel TITIEE lee As there is a pressure loss between any two points in a pip- t | = ing eve, a0 rol poses bes oon belwoan ay t a4 two points within a pump. In addition to a short run of straight +. rtp ass How athe entrance tthe sucton por. tne fow witin mest me positive displacement pumps, depending upon principle, will t is Tau crostincen Slatgnct arosion Steve or areas a more than 80% may havea sudden enargment or corres: a i ton at he entanco toe sueten cavinr Tore may be 6 ; Ley picbed disruption to smooth liquid flow caused by the moving pump ing elements, the extent again depending on the pumping principle. AAs stated earlier, NPSH required is another way of stating the pressure loss in the suction area of a pump for a given set, of conditions. Figure 8 shows the internal gear pumping prin ciple with numbered points or zones of progression from the suction port to the point of minimum absolute pressure (max- imum vacuum). 1. Zone 1 iat the suction port of the pump. 2 Zone 2 is where the port throat opens into the suction ‘area of the casing, 3. Zone 3s in the area where the pumping elements (rotor and idler) are being filed with liquid, 4, Zone 4 is where the rotor and idler teeth are coming out ‘of mesh: the point of lowest absolute pressure (highest vacuum), ‘The curve in Figure 9 shows in a generalized manner the ‘gradual loss of pressure as liquid progresses from the suction ‘port (Zone 1) tothe point where the rotor and idler teeth come ‘out of mesh (Zone 4), Visualizing the pump in operation it is easy to see that as, the pump RPM increases, the pressure loss will also increase. PAGE 6 -berls eae nwo ape ef vee Pune if pipet lL TT oe Sp a eno a FIGURE 7.FIGURE 8. Internal gear pumping principle showing suction side pressure zones. ‘As there is a pressure loss in a pipe tance to flow) increases. See the ‘with increased flow, s0 there is in the dashed line in Figure 9, throat area (Zone 1} of the pump suc- any time a pump is operating there tion port. As the pressure loss in- ig & pressure loss within the pump: as creases with increased flow of liquid gs just been discussed and os shown through the zig zag path in a valve, soit jn Figure 9. Under normal conditions also increases as more liquid flows into Gn a well-designed system, even the faster moving pumping elements. though there is @ pressure loss be- ‘And finally, since the time for the quid tween Zone 1 and Zone 4, the pump to fil the void at Zone 4is shortened as pressure at Zone 1 Is high enough so the pump RPM goes up, the liquid must that the pressure at Zone 4 does not move faster and to do so, requires drop below the vapor pressure of the more pressure (experiences @ larger liquid. As long as this is the case, pressure loss). Increasing pump RPM cavitation does not occur and there is does increase the pressure loss in the no. pumping problem. For a system pump. handling a viscous liquid, as long as As all pipe line loss charts show, the the pressure at Zone 1 is high enough hhigher the viscosity the greater the to assure a complete filling of the oss, other conditions remaining the pumping elements in Zone 3, there will same; so it is within a pump. Fore be no “starvation” (incomplete filing given’ RPM the pressure loss will in- of the pumping elements) and thus no Crease as the viscosity (aliquid’s resis-_problem. vacuum PRESSURE ZONES FIGURE 9. Absolute pressure vs. pump suction side pressure zones. It is when the pressure above the vapor pressure of the liquid (NPSHA — Net Positive Suction Head Available) at Zone 1 is not sufficient to overcome the pressure loss within the pump, that ‘cavitation or starvation” will occur. ‘An engineer or designer when laying ‘out a piping system should determine the amount of pressure (NPSHA) that will be available at the pump port. Its then important for him to know what the pressure loss (NPSHR) is for the various pumps available so that a proo- ‘selection can be made, Thus, a line (of pumps for which there is no NPSHR. information may not be given consider- ation when a pump selection is being made. Viking has accumulated a wealth of practical field experience over the past 75 years in applying pumps to applica- tions with limited NPSHA. During re- ‘cent years an aggressive program of RAD lab testing has been underway to accumulate NPSHA data for a wide range of speeds and viscosities on the ‘complete pump line, x FIGURE 10. Drawing of test Set-up for checking NPSHA. ‘The drawing in Figure 10 shows the arrangement of the various pieces of equipment actually used in conducting the NPSHA tests in the R&D Lab. The 400 gallon tank shown in Figure 11 is, the vacuum vessel used when testing ‘at 50 GPM and higher flow rates. In theory, the procedure being used to collect data for making NPSHR determinations would be as follows; in practice some revisions have been ‘made to the procedure to stay within the limitations of the equipment: 1. Use the vacuum pump to reduce the pressure in the vacuum ves sel 10 the vapor pressure of the PAGE 7liquid (when using fuel ols or lube oils as test liquids the pressure in =. the vacuum vessel approaches zer0 absolute (29.92" Hg. vacu- tum] since their vapor pressures are almost nil) 2, Raise the vacuum vessel above the pump. 3, Start the pump and set the speed. 4, Adjust the height of the vacuum vessel 10 a point that gives a vacuum gauge reading (approxi- mately 3.4° Hg. vacuum) that is ‘equivalent to an absolute pres: Sure-of 30 feet o quid (spectic. FIGURE 11. 400 galion araity of 1.0) Set dacharge pres vacuum vessel. ‘sure at 50 psig. Record capacity. 5, Lower the vacuum vessel to a until an absolve pressure of ap paint that gives a vacuum gauge proximately 2270 feet s reached feading equvalent oan absolute Capacity a his point wil sso be pressure of 25 fee! of qué. Re- 2010 Gord the capacity, As the vacuum vessel is lowered 6. Continue lowering the vacuum {here will be 2 point — normally vessel and recording capackty between 10 and 2 feet of abso aT 70) 640 RPM | 50 PSI NpsHR-7'~/ 60 2 { = 90) | & 4zo_rpm | 5 Psi 2 i 2 40 30 PS 5 NPSHR z 3 ws °[_280_RPM| 50 PSI MH 2 NPSHR~2.0' i Feo < z $ 19] old 30 20.15 Ose reSase10 NPSH — FEET OF LIQUID, SP. GR. 1.0 FIGURE 12. Curves for Viking Pump Model K-125. Plotting capacity vs. NPSHA when handling 38 SSU liquid. PAGE 8 lute pressure — where, depend: ing upon pump speed and liquid viscosity, the capacity wil start 0 drop off, indicating thatthe pump is cavitating or “starving”. Isat this point that more pressure is required to get the liquid into the jumping elements than is aval- able at the suction port and as a result, the elements are nol being completely filled with liquid. 7. Repeat the above procedure at Various speeds and viscosities. The curves shown in Figure 12 are plotted from actual test data taken ac- ‘cording to the just reviewed procedure for a Viking Model K125 pump handling 38 SSU liquid. The actual NPSHR value {or the various speeds is indicated by the "X""on the capacity curve. It is de termined as the point at which the pump capacity deviates from a straight line. NPSHA Calculations — The basic formula, NPSHA=Ha + Hz—Hi—Hyp, was discussed earlier The Hydraulics Institute is advoc: ing the use of the term Net inlet Pres: sure (NIP) in place of NPSH when working with positive displacement pumps. As more spec sheets are writ: ten specifically for positive displace: ment pumps, there will be a growing Use of the term Net inlet Pressure. Net Inlet Pressure is expressed in PSIA. ‘The Required Net Inlet Pressure (RNIP) Is the counterpart to NPSHA. The Available Net Inlet Pressure (ANIP) is: Used in place of NPSHA, Ai the present time information on Viking internal gear pumps willbe given as NPSHR ex- pressed in feet of liquid with a specific. gravity of 1.0. To convert this to RNIP in PSIA multiply the value by 0.433, ‘As before, Ha Is the absolute pres- sure on the gurtace of the liquid, Hz is the distance from the surtace ofthe lig id to the centerline ofthe suction port, Hf is the pipe friction or line loss. and Hip is the vapor pressure ofthe liquid, All terms to be expressed in the same Units, feet of liquid pumped, In the following examples the basic \NPSHA formula is applied to three typi cal pump installations. These three are representative of the types which most frequently have NPSH problems. INSTALLATION 2 — Suction Lift ‘Transfer 50 GPM of Toluene from an 8' diameter horizontal buried tank 2° below grade. Normal tank temperature is 60°F, specific gravity is 0.87, vis Ccosity 's 0.8 Cps (about the same as ‘cold water), ana the vapor pressure is (0.36 PSIA at 60°F. Pump Is to be in-stalled at grade above the tank, line size is 2°, the job site isin the upper rmigwast. INSTALLATION 2 NPSHA=Ha Hz —Ht—Hvp Ha =27" Hg, abs, based on an elevation of 2000" and @ low barometer. Ha =(27{1.13)=35" Toluene, oar He =8'+2'4+1'=11' Toluene. This is the distance from the surface of the liquid to the Centerline of the pump. How: ever, always figure worst ‘condition with the tank emp: ty. The liquid is below the ump; therefore, Hz will be negative. HE =Use the line loss through 35° of pipe [12° actual (11° ver tical plus 1" horizontal) plus 23" equivalent for elbows, a gate valve and a check Valve]. Line loss in feet of Toluene per foot of pipe (02 x2.3x.87 =.046; from 87 information in Section 510 ot the General Catalog HE =35x0.086 Hi =1.6' of Toluene Hyp =0.96 PSIA at 60°F; use 1.7 SIA, the vapor pressure at 120°F which is possible summertime temperatures at the pump. Hyp 7(2.31)=45' Toluene 087 NPSHA=95'— 11'—1.6'-45" 17.9’ Toluene ANIP) =Hg+He—Hi—Hvp Ha 7" Hg. abs. x0.49 Ha 323 PSIA He of TolueneX$pG) 281 He ~ =(1K87)+231 Hz =4.14PSIA Hf PSIIFt.\(SpG) times (Total Equiv) Ht 02M 8712-423) ‘0174y35), 61 PSIA Hp =1.7 PSIA ANIP =1323—4.14—061~1.7 ANIP =6.78 PSIA The NPSHR for a KK-size Viking at 420 RPM is 3.3 Ft. (SG ~ 1.0). (3.3) =(87)=3.79 Ft. 17.9 Ft The required Net Inlet Pressure (RNIP) is equal to: (Ft Lig 10.433) = Sp NIP + 87 NIP ANIP ‘Any sel-priming pump with an NPSHR. less than 17.9" should work, Ht, in Installation 2 the tank was in: stalled above grade, the liquid level ‘would be above the pump suction port In this ease the NPSHA would be signif. icantly inoreased because the Hz fac- tor would be plus instead of minus. This is not to say that such an installation should not be reviewed closely when at the design stage or that it should not be checked it pump problems develop at startup. It is quite possible to fall into the trap of believing that for an installa- tion with a "flooded" suction (liquid level above the pump), there fs no need to check the suction side of the pump. Flooded’ suction means positive pressure at the pump suction port. Many times the suction port is not “flooded” because of long suction line, small piping, high viscosity, a fine mesh strainer, or other fittings which together restrict the flow of liquid and cause NPSH problems, NPSHA or ANIP should always be calculated on extreme conditions to prevent start up problems. You will note from the caloulations tor the NPSHA for Installation 2 that several of the terms were figured on & ‘conservative or extreme condition. As pump manufacturers, we feel that this is akvays the preferred approach. If problems develop after the system is Installed no amount of checking and veritying data or calculations will cor fect a situation that might not have developed if more conservative values had been used to start with INSTALLATION 3— High Vapor Pressure Transfer 30 GPM of LP Gas (Pro- pane) at 65°F from a large storage tank, Specific gravity 0.50, viscosity is 0.1 Gps, vapor pressure 100.7 PSIG. INSTALLATION 3 Has Hz—Hi—Hyp absolute pressure at surface Of liquid. atmospheric pressure + va- pot pressure, =14.7 + 100.7 =1154 PSIA, 118.442:31) 0.50 Ha__ =533' of LP Gas. Hz =dlstance trom liquid level to He port centerine "+4", liquid is above pump. ction line loss from pres sure drop charts, 5! of LP Gas, vapor pressure of LP Gas at 65"F, (00.7 PSIG + atmospheric pressure of 14.7 psia, = 115-4 psia Hyp _ =533' of LP Gas, NPSHA'=533'+ 4’ 1.5'—533" NPSHA=2.5' of LP Gas, ‘Note that Tora high vapor pressure liquid (@ liquid which nas o be stored ina closed ‘vessel fo keep it rom boling away) the Ha ‘and Hvp terms cancel out and that NPSHA becomes the difference detween the elevation head andthe ine los: He — Hf In addition to the LP gases, propane and butane, other high vapor pressure liquids that are frequently handled with Viking pumps are anhydrous ammonia and Freon. ‘Adequate NPSHA when handling the high vapor pressure liquids is obviously ‘a must. There is evidence and some test data that indicates that these li: vids, particularly the hydrocarbons, ‘can be pumped satisfactorily with less NPSHA than would be required for cold water or other low vapor pressure test liquids, Under some conditions and with certain liquids, pumps will operate satisfactory with 5 litle as 50 per ‘cent of the NPSHA that might be re: Quired if handling cold water. Some {uidelines for the NPSHA required for Viking LP Gas pumps are given in the General Catalog Section 440, and Technical Service Manual, TSM442 and 443, PAGE 9INSTALLATION 4 — Vacuum Vessel ‘Transfer 40 GPM of soybean oll at 240°F from a vessel under 24” Ha, vacuum. Viscosity 40 SSU, specific gravity 0.8, INSTALLATION 4 NPSHA=Ha + He —Ht— Hyp Ha =absolute pressure at surtace of liquid. Assume a 1000" tlevation installation with @ low barometer (use 27.8° Hg, abs. as barometric pres: sure), = (CH abs.)(FtH,O per "H) 8G Ha =(27.8-24.0)(1.133) 0.88 Ha __ =4,9 feet of soybean oil Hz distance from liquid level to port centertine. Hz =6’of soybean oil this value is plus singe the liquid is above the pump, HI =suction line loss in feet of soybean oil. Ref. Sec. 610 of Viking General Catalog for line loss charts, (.02)(4),88)(2.31) Ht__ =0.2' of soybean ol Hyp =vapor pressure of soybean cll at 240°F. 2! estimate) of soybean oil NPSHA=4.9'+6'—0.2"~2 NPSHA=8.7" of soybean oll ‘As with Installation 2 itis well when. making calculations at the design stage for a system of this type to be conservative wherever possible. Installation 4 is also quite typical of applications involving transformer oil purilying, lube oil reclaiming and sol- vent recovery. One important point to remember on this type of application is tat the pump is a the low point in the PAGE 10 system and is, therefore, wetted at slarlup and able to develop a good in: ial vacuum. A pump with a lift often hhas to evacuate air or vapor from the suction line before it becomes wetted with the liquid pumped. ‘Another point to keep in mind is that ‘on some applications, the liquid is quite viscous and thal itis possible to have ‘more friction loss per toot of vertical pipe than there is gain because ot ele- \ation, Larger pipe size will correct this, situation, Possibly expressing the information in he formula by line chart would be of help in visualizing how the actual NPSHA figure is determined. Figure 13, shows the steps taken in arriving at the NPSHA for Installation 2 Determine Maximum Suction Lift— The NPSHA formula Is most often used for making calculations in con- junction with existing or completely de- ‘signed systems as we have just done for Installations 2, 8 and 4. It can also serve as an effective tool when used in the early stages of system design 10 determine limiting conditions One of the admonitions stated sev: eral times is that tohave an acceptable Installation, the NPSHA of the system NPSHAS +H, “Hyp oH, (LIET)* i) NPSHA 16-4.5') REFERENCE +179" INSTALLATION 2 SUCTION LIFT HANDLING TOLUENE] ‘ABSOLUTE ZERO FIGURE 13. Line chart showing sis takon in calcula Ing NPSHA “Ifthe liquid level was 11" above the pump ie would te plus and would be aaded 10 Instead of being subtracted trom Ha, The INPSHA would then be 38.9 ‘must be oreater than the NPSHR of the pump. If, n the early stages of system design, & pump is selected, it is then possible to determine the maximum al fowable lift by equating NPSHA 10 NNPSHA and soWving for Hz. AS an ex- ample, consider that for Installation 2, the question had been raised "What is the maximum vertical It possible?” From data available for the pump se- lected, determine the NPSHR: as an ex ‘ample consider that the NPSHR for the ump selected is equal to 3.5 of Tol. ‘ene; use 5.0" to be conservative. Then NPSHA=NPSHR=5.0'=Ha—H2— Hf-Hvp Values for the different fac- tors are taken from the cal ‘culations on Installation 2 Ha fz it 0,046" (Hz +24'**) Ht 0,046 H+ 1.1 Hyp =45" 50" =Ha—Hz—Hi—Hyp 50! =35'—Hz—(0.046 Hz+ 1.11) 45" $0! =35'—1.086 H2=1.1'-45' 1.046 Hz =35'—1.1'—45'—500" Hz =2333) A vertical lift of 23.3" is higher than ‘would normally be encountered but the exercise does illustrate an additional use for the NPSH formula. The same approach applied to other systems ‘would permit determination of 1) the minimum liquid leg required above a pump when handing a high vapor pres- sure liquid or when pulling from a vac- tuum vessel, 2) the proper line size (d- termined by the Hf factor) when the lift Or liquid leg is predetermined. Line loss i f00t of Toluene per foot of pipe. “One foot of horizontal pipe plus 23° of ‘ouivatent length equals 24° Determining NPSHA By Gauge— It is possible by using a pump suc- tion port gauge to determine the NPSHA of an operating system without making all of the calculations shown for Installations 2, 3 and 4 By definition from an earlier chapter, NPSH equals “the pressure in feet of liquid absolute measured at the pump suction port, less the vapor pressure” if we introduce a new term, Hi and de. fine it as the absolute pressure of the liquid at the pump suction port ex pressed in feet of liquid, we can set it fequal to the first three terms of the NPSHA formula — Hi=Has Hz —HfFestating the NPSHA formula using the Hi term it becomes — NPSHA=Hi- Hyp The value of Hi can be determined by converting a suction port gauge reading into feet of liquid absolute. When making the conversion use the local absolute barometric pressure, the specific gravity of the liquid and the appropriate factors. The final step is to subtract the vapor pressure of the liq: uid at pumping temperature to arrive at the NPSHA for the system. To illustrate refer to Installation 2 and assume the unit is operating under the following conditions: liquid temper- ature is 60°F, the tank is one-half full, the barometer is 27° Hg. absolute. The vacuum gauge reads 6” Hg., deter- mine the NPSHA of the system, ‘The suction port gauge reading of 6” Hg. vacuum subtracted from an abso- lute barometric presure reading of 27° Hg. gives a pressure reading at the pump of 21" Hg, absolute. The caloula- tions below convert the 21" Hg. abso- lute to 27.3 feet of Toluene absolute; this isthe value for the Hi term. Calcu- lation — 21.("H9) 1.18 (feet of HO) 1 Cho)” 087 (feet of Toluene) =27.3" of Toluene (leet of HO) The vapor pressure at 60°F is 0.38 PSIA or 1” of Toluene. NPSHA =Hi— Hy 26.3’ of Toluene This is considerably greater than the 17.9" determined by calculations for In- stallation 2. The difference, of course, relates to the fact that for calculation ‘purposes we used the extreme condi tions of having an emply tank and 120°F operating temperature with its resulting higher vapor pressure 27.8 Reducing NPSH Problems — ‘In the discussion following the cal: ‘culations for Installations 2, 3 and 4, the desirabity of using conservative values for the factors atfecting NPSH was mentioned several times. Some ‘specific comments and suggestions re- ‘garding each of the factors follows: Ha — the absolute pressure on the surface of a liquid — does not lend itself to much change or increase since itis a function of either atmos- pheric pressure or the process. Hz — an indication ofthe location of the surface of the liquid with respect to the pump suction port — is relatively easy to change at the design stage; it may be very difficult to change after an Installation Is in operation. Ht_—tine loss or pressure drop — is relatively easy to reduce when design- ing the system. For example, larger line sizes, fewer elbows, larger strainers, different type shutoff valve, etc. all tend to reduce the line loss. Such changes may result in added initial ex- pense, but if they are necessary to af- ford the difference between a system that works and one that does not then, of course, the added expense is jus tied. To make changes in the piping sys tem after it is instaled to reduce the line loss value Is difficult and very ex- pensive. For an operating system it may be possible to reduce line loss by increasing the temperature and thus reducing the viscosity. This is okay it the reduction in jie loss is not more than offset by an increase in vapor pressure. It may also be possible 10 Teduce the line loss (Hf) by reducing the flow rate, For a continuous opera: tion this may not be practical. For a batch type operation or a transfer job reducing the flow rate may be a prac tical means of increasing NPSHA. Hyp — vapor pressure of the liquid at operating temperature — is tied to ambient conditions or to the process. In those few cases where it may be possible to reduce the vapor pressure by lowering the temperature it should only be done i the necessary reduc- tion in temperature does not cause a more than offsetting increase In the Hf (line Joss) term because of higher vis: cosity, It to reduce NPSH problems, changes to the system do not seem the most practical, the NPSHR of the pump Should be reviewed. For a given pump, ducing the speed (and thus lowering the capacity) will educe the NPSHR; it will also increase the NPSHA because the Hf (line loss) factor will be smaller. It itis not practical to reduce the ca- pacity because of system or process Fequirements, a larger pump running slower willbe able to deliver the same capacity with a lower NPSHA figure, Variables — ‘When an actual NPSHA increase of 1 foot of liquid can spel the difference between an acceptable installation and fone that is troublesome it seems worthwhile to pursue all avenues that might reduce potential problems. The following points or variables can prove to be a source of trouble if not given proper consideration, 1, Liquid level in the supply vessel. Ina storage tank this level goes up and down as the product is used and as the tank is refilled. If the NPSHA is ‘marginal problems may oycle with the level in the tank, Seldom is liquid in the bottom ‘quarter of a buried tank used, there- fore, the suction pipe does not ex: tend clear to the bottom of the tank. (On those occasions when the liquid level is unusually low, itis possible to get abit of swirling or vortexing at the inlet to the suction pipe with re- sulting air entrainment and reduced capacity ‘Many supply vessels and vacuum vessels use float actuated controls to maintain liquid level; sometimes there is more variation in level than was intended. Otten location of the start and stop levels can be ‘changed to helo an NPSH problem. 2. Ambient temperatures. Higher tom- peratures than anticipated may ‘cause the product to vaporize in the suction line or may raise the vapor pressure; lower temperatures may cause problems because of in- creased viscosity and line losses. 8. Suction gage location. A suction ¢gage on the tank side of a strainer does not give the proper picture of ‘conditions at the pump, although it ‘may only be a couple of feet from the suction port. 4, Light ends vaporizing. Some liquids, particulary fuel olls, may have light fractions which will vaporize under slight vacuum conditions and cause ‘cavitation and noise. The problem may come and go when handling heavy fuel oils, depending on the source of supply and whether the particular oll was a blend or a straight distilad product. Lowering the oll temperature a few degrees may help. 5. Air in the liquid. Dissolved or en trained air in a product may also ‘cause problems if allowance is not made for this possibility in the ‘design of the system. The extent of potential problem can best be de- termined by discussing the actual liquid with the user to get as much information about it as possible. Ifa liquid containing entrained air can be left undisturbed for several hours some of the air wil rise to the sur face and escape, Comments — ‘As was mentioned earlier, working with NPSH involves a number of factors and variables, often interrelated, which must be given consideration. Thus Viking, as @ pump manufacturer, tends to be’ somewhat conservative when ‘making recommendations for a system (or pump on an application which may PAGE 11involve potential NPSH problems. As necessary to understand the conver you can appreciate, if the sysiem sions between the various pressure doesn’t work, the first thing that is units, suspect isthe pump, which is he heart "The Viking R_& D Laboratory has of the system, completed the NPSHR testing of al the ¥ you have followed through the cal- internal gear purnp sizes currently cata cculations you can appreciate that it iS jogued. Its one more tool in the hands Net Positive Suction Head Required by Viking Pumps of the Viking marketing organization which will permit applying Viking pumps to the fullest extent of their capabilities, while giving assurance that the pump will perform satisfactory NPSHR — Feet of Liquid Sp. Gr. 1.0 UWP SPEED PW rune sre vou [as | 15 | woo | veo | om | ose | co | so | oe | ro | mo | vce | ve | re c | 17 | 19 | 22 | 24 F.FH | 18 | 19] 21 | 23 | 28] 34 6.66 18| 20| 22 | 26[ a1 | a0] se| 76 ‘HAH, HE 17|18| 19] 21| 24| 28 | 34| 45 | 62 | 95 [135 AS, AK, AL 16| 17] 18 | 20| 23 | 27/32] 39 | 55 17.0 [233 JAK KK 17| 12| 19 [2a | 29 | 20 | 39 63 Luts 17| 1a[ 20] 22[ 25 [30] 38] 50 Q.as 4a | an 27 | 33 | 42| 61 | a4 ™ 24 | 23 34 | 43 | 60 N 21 | 25 45 | 63 | 95 [150 |< CATALOG SPEED RATING a 27| a2| 42| sa | e2 [io NOTES: 1. NPSHA (Net postive suction head avaliable) must be greater than the NPSHA (Net postive sucton heed requires) gen in the above tbe. 2. VISCOSITY — Above chart apples fo vicostes up thu 750 SSU. Consult factor or Viking representative for visostios shove 750 5SU 2. Foriquis other than water, die by Sp. FIGURE 14. Viking Pump Inc., Unit of IDEX Corporation Cedar Falls, lowa 50613 * Telephone (319) 266-1741 * FAX (319) 273-8157 ipoaeM S94 Copyright 1987 PAGE 12 Priniedin USA.