Water Hammer Analysis by J. Parmakian

WATERHAMMER ANALYSIS by JOHN PARMAKIAN BIBLIOTECA DOVER PUBLICATIONS, INC. New YorkLay w PREFACE TO FIRST EDITION ‘Aum-Cuaucas Muorscronna Courar ‘Busowan-Lawa-Hascuron Comoran ‘Tam Janes Levent & Compan Nawrorr News Sumraonaixa axo Dat Doce Couraxe 3. Monga Surra Couranr In this connection I am especially indebted to G. Dugan Johnson, Chief Hydraulic Engineer of 8. Morgan Smith Company, for ini- tinting and obtaining this cooperative assistance. Much valuable assistance was given by my past and present colleagues at the Bureau of Reclamation. Of these I wish to thank F. E, Cornwell, C. C, Crawford, and W. E. Evans, whose suggestions have influenced much of the text material. The illustrations were prepared by R. H, Williams. ‘The solutions of the waterbammer problems at the end of the book were worked out independently by Benjamin Donsky and C. W. Lundgren. The assistance of Mr. Donsky and F. 0. Ruud in proofreading the text and illustrations is also acknowledged. Finally, I wish to thank L. N. McClellan, Assistant Commissioner and Chief Engi- neer of the Bureau of Reclamation, and Professor Warren Raeder, Head of the Department of Civil Engineering at the University ‘of Colorado, for the privilege and opportunity to teach the first ‘course of study on this subject at the University. Joun PARMAKIAN Denver, Colorado Preface to Dover Edition The Dover edition of Waterhammer Analysis is the same ‘as the original edition with the exception that several subscript errors and missing decimal points in a few of the figures have been corrected. To date no important exceptions to the material contained in the book have come to the author’s attention, Jorn Panmaxtan Denver, Colorado ve Water Column Theory. -- +. 5 + 1. Basic Equations. . 2. Uniform Gate Operation 3, Bmmple... 2... 4 Bauialn Leng of Pp i 5. Limitations of Theory . Elastic Water Column Theory ee 6. Basic Considerations . 7 Condition of Dynamic Equilibrium 8. Condition of Continuity : 9. Fundamental Waterhammer Equations. 10, Physidal Significance of Waterbammer Equations Velocity of Waterhammer Waves. . . - 1, Sioa, Cashion and TransitePpes 12, Wood-stave Pipes... - 13. Reinforced Concrete Pipes... M4. Circular Tunnels... = 15. StoeLlined Cirenlar Tunnels. Waterhammer Wave Reflection... . . . 16. Ata Reservoir. 1, Ata Dead End < | | Ata Change in Grossaectional Area. |. 10. Ata Junction of Three of More Pipes. - 20, Ata Partially Open Gate... .. 19 95Oo CONTENTS ‘Watethammer for Rapid Gate Movements. . - ‘21, Instantaneous Gate Movements... ss. . 30 22, Gata Movenenta Competed in Lam Than 21/6 Seconds errr 4 ‘Watethammer for Slow Gate Movements... - 23, Baslo Considerations. 6. ss ws eet 4, ‘Time Intervals Used in Computations - + 25, Example. ss... 30 ote a 28, Interpretation of Solution Theory of Graphical Waterhammer Analysis for Gate Operation... «+ 27, Conjugate Waterhammer Equations... 45 28, Gate Discharge Equation... . 5 48 29. Graphical Solution for Gate Closure a Watethammer Produced by Gate Closure. . « 30, At the Mid-length of Pipe. 31, At the Quarter Points ofthe Pipe Line 32. Pipe Line with « Stepwise Change in Diameter. 55 33. Partial Gate Closure 34. Gate Clonure with a Relel Valve... 2 88, Cont Gate ated Ary fom Outlet Bad of| cere vee OL 36. contol Gata Lacated ab Upper ea of Pipe oe OL Waterhammer Produced by Gate Opening . « ‘37, Fallin Presrure Produced by Gata Opening 63 £38, Pressure Riso Produced by Gate Opening 63 39, Rhythmic Opening and Closure of Control Gate. 65 Waterhammer for Uniform Gate Operation . . « 40, Uniform Gate Closure to Zero Gata 7 41, Uniform Gate Opening from Zero Gate « 30 37 45 52 63 70 xi. xi. Xill, XIV. XV. CONTENTS ‘Watethammer in Pump Discharge Lines Caused by Power Failue .. ++ - 42. Transient Conditions at Punp and Discharge Une. ss. es occ “ 43, Toertia Equation eee te 70) 4. Pump Characteristion se 78 45, Waterhammer Equations.) 2... ss a 46. Graphical Waterbammer Analysis... . 83 Wales Colomn Siparaton «s+ s ot Waterhammer Charts... Beet pra orn gas bocua charge Lines 50, Effect of Changes in Pumping Head... . 51, Pump Discharge Line with Check Valve... 32, Pump Discharge Lin with Long Suction Line 153, Effect of Control Valves - 34. Blarting and Stopping Pumps . esses ‘Waterhammer Analyses Including Effect of Hy- draulic Losses... . . - eee 55. Losses Concentrated at Intake . . 101 166. ‘Losses Concentrated at Pump Discharge Outlet . 103 ‘57, oases Concentrated at Several Points Along Pipe 105 Waterhammer Analysis for Compound Pipes . - bees s 106 . 0 Pipe Line with « Dead-Bod Branch Connection . 111 81. Compound Pipe System with Two Control Gan, 11 camer eer ne na Approximate Methods of Graphical Watetham- mer Analysis... se ee $3. Houvet Uniform Pipe Apposimation . - - 115 G4. Pipe Line with Surge Tank . . U6 93 101 106 115XVI. XVI. XVIIL, como? Discharge Characteristics of Gates and Valves . oe Soe cence de Sot iteawe Sooo Suge Tanks. 2 mean Air Chambers at Pumping Plants... 2... TL. Operation ee see BBL 72, Basie Considerations : 73. Graphical Solution ‘ 74, Air Chamber Charts... | eee 7. Eumple..... 1. ae ML 76. Water Level Controls 22) 2 oat References... ee Problems... ee ee Index ve ee ee 120 193 131 143 147 157 DQ Notation ———————— Description 1. Conduit Dimensions and General Symbols Cross-sectional area of pipe or conduit ‘Thickness of pipe wall Inside diameter of conduit, Inside radius of conduit Modulus of elasticity of pipe wall material ‘Modulus of rigidity of tunnel material Acceleration of gravity Bulk modulus of water ‘Total length of conduit, Length of any uniform section of conduit Specific weight of water Distance measured positive from lower end to point on conduit Distance measured in opposite direction from ‘Height of point on conduit messured above eater line of control gate Angle of slope of conduit Poisson's ratio for the pipe wall material ‘Longitudinal stress in pipe wall Circumferential stress in pipe wall 2. Pressure and Head Symbols for Gate Operation Pressure head for steady conditions (when gate is submerged, His taken as the ‘et head on gate) Presaure head for surge eondit Pressure rise above normal Pressure drop below normal Ratio H/He Friction head Ratio H;/He Unit pressure Symbol S Poms anmoen 8 Units fe ft fe ft Ib/te Ib/te ft/sec? Toye fe ft lye ft. fe ft byte Ib/te ft fe tt fe ft lotNOTATION Description Velocity and Time Symbols ‘Velocity of pressure wave Round trip wave travel time Number of 2L/a time intervals Initial steady flow in conduit, Time Time of gate travel ‘Velocity in conduit for final steady conditions Velocity in conduit for surge conditions ‘Velocity ratio ¥/Vs ot V/V, Difference between initial and final steady velocities J. Waterhammér Symbole Pressure wave travelling in direction of +2 Pressure wave travelling in direction of —2 Pipe line constant Reflection factor ‘Transmission factor . Gale-Opening Symbols Crose-sectional area through gate opening Gate opening factor as defined by equat (28) Coefficient of discharge through gate EBilective discharge area of gate Proportion of full gate effective opening . Pump Operation Symbols Ratio of pump speed at any time to rated Pump speed Ratio of pump input torque at any time to pump input torque at rated speed and head Pressure head for surge conditions measured above the pump intake water surface elevation ‘Symbol Units {ft/sec wee f8/e0 see ee. ft/sec ft/sec tt/see ft/e00 ft ft fe 10/200 te fe 3° NOTATION Description Pumping head for initial steady pumping conditions Rated pumping head Friction head Ratio of pressure head for surge conditions to rated pumping head; h= H/He Ratio of friction head to rated pumping head Defined by equation (50) Pump input torque Pump speed Pump efficiency at rated speed and head Angular velocity of pump and motor shaft Flywheel effect of rotating parts of motor, Pump and entrained water Mass moment of inertia of rotating parts 7, Surge Tank Symbots Cross-sectional area of pipe Cross-sectional area of surge tank orifice Dimensionless parameter Cross- 1NaisNvuL SONO93S 4B +719 SWOILIONOD anaisHYaL 80N0035. 9-8 <7 Av sworttoNO9 aNaiswvaL ) Onn w —— 5 Sang 0dy 80M Fes TE -suouy atom So sonoa2s 9- £+3.1y sowsianoo inaisuvus ——_swo035 $F +4 4¥ sworsiowoo saasHvua “h t — (LEI, etsy oa ee 044 anon ley ‘0p 08, : Sono0as $F+3.1¥ Swolsionod iNaiswveL ———-saw0035 34-243 4¥ sworLiaNoD LWaisMvUL {o) a — b ELITE. f Fe 44045 340m 40a) an0q. SONODS 9+ Juv SWOILIONOD LNSISNTEL ene eee yoy ano ory 0m. Ee, SONOD3S B+ 3.1" SWOILIONOD INBISNYEL -—-«$0NOD3S 9-2 «3.Av SWoHLIONOD ANaISWH¥L 0) LNs vain ne) 44034 2n0M sowooas 42.7 4¥ swoitionoo snaiswees Figure 19 o) quou anon fe = ae PEISK . unso79 ave snoanvinvisa yaiav soNoD3s-9 sovooss $41" swoisionoo svatswva ne nanan zs oO) =--yuoug anomy (9) — % = 3anssaa0 pasoasout yo uoibay |. 015 200 — — Goad em sm SaNODaS 9+ 2 +74¥ SWOILIONOD AN3ISNYYL —_ANINZKOM 3AvD O4 SMOIAIHE NOILIONOD 31V1S AGVILS hoy 310 2406 joo (0) . a %z [amavaad pasoarsap jo voibay | a n ae : Figure 19 (cont.)a 34 WATERHAMMER FOR RAPID GATE MOVEMENTS [29 NEAD CHANGES FOR RAPIO. ‘GATE MOVEMENTS Figure 20 22. » Gate movements completed in less than 2L/a seconds If the reservoir is at such a distance that a reflected wave cannot return to the gate before the gate motion is completed, the same maximum pressure change occurs at the gate as for an instantaneous gate movement. Hence the physical significance of Equation (31) is further enlarged and stated as follows: For $22] WATERHAMMER FOR RAPID GATE MOVEMENTS 35 (2) yen gote movement tine i gear than seconds ‘lee mon 4 secote (Maximum neod on 968 enum need on poe ‘or gate peng HEAD CHANGES FOR RAPIO GATE MOVEMENTS Figure 20 (cont.) an instantaneous gate movement or a gate movement which is ‘completed in less than 21,/a seconds, the maximum change in head at the gate is the same; that is, AH = —(a/g)AV. For any gate movement which takes place in less than 2L/a seconds, the maximum change in head extends from the gate to a certain limiting point along the pipe. This limiting point is determined as follows:* If zis the distance from the limiting point to the reservoir, the wave-travel time from the gate to the limiting point is (L — %)/a seconds. The time required for the pressure wave to ‘travel from the gate.to the reservoir and back to the limiting point is (L + 2,)/a seconds. If the total gate movement time is T, the elapsed time from the start of the gate movement to the instant of arrival of the final incremental pressure change at the * Bee Reference 30,rn 36 WATERHAMMER FOR RAPID GATE MOVEMENTS [22 limiting point is T + (L —)/a. Therefore the limiting point is located along the pipe where the direct wave is met by the reflected wave, that is, Ltn lp lan 2 a or ne me (2) A pictorial representation of the head changes along the pipe for gate movernent time equal to or less than 2L/a seconds is shown in Figure 20. In this figure AH = —*AV as defined by Equa tion (31). . vi Waterhammer for Slow Gate Movements 23. Basic considerations ‘When the gate movement time is equal to or less than 2/1 seconds, the maximum pressure change at the gate is doterminc«! from Equation (31) since no reflected waves return to the gate Soon enough. to alter the pressure at that location. However, when the gate movement time is greater than 2L/a seconds, this Stepwise gate movement BE sae ea Time m secoxos, Figure 21 isno longer true. For long gate movements the actual movement, is approximated hy a series of etccescive instantancous steyvins movements.! ‘These stepwise movements are selected vo that they have the exact valye of the actual gate opening at the time of completion of cach step as shown in Figure 21. ‘The pressure in- * See Reference 16, 37cr WATERHAMMER FOR SLOW GATE MOVEMENTS [$24 beatnt-oatn wot et Figure 22 crements produced at the gate by each of these stepwise movements are then displaced from one another by a distance aAt as shown in Figure 22, 24, Time intervals used in computations In order to reduce the computations to a minimum, the selected time interval should be as large as possible. After an instantaneous gate movement takes place at the lower end of the pipe, a reflected wave returns to the gate 2L/a seconds later. ‘Hence the maximum time interval which can be used in the com- utations to determine the waterhammer at the control gate is the complete round trip wave travel time of the pipe, that is 2L/a seconds. If the pressure-time history at any other point on the pipe line is desired, the largest time interval which can be used. in the computations is the shortest round trip wave travel time from the point on the pipe to a point of wave reflection. If more values are desired on the pressure-time-history curve for any particular point on the pipe, smaller time intervals are necessary. However, this will not result in a more accurate determination of the waterhammer for the points on the pressure-time history pre- viously determined with the larger time intervals. This is due to the fact that since the time interval used in the computations is not greater than 2L/a seconds, Equation (31) applics. ‘Therefore the head change at the end of a particular time interval depends only upon the velocity of water at the beginning and end of the time interval under consideration. In general, the time of occur- qd $25) WATERHAMMER FOR SLOW GATE MOVEMENTS 39 Tence of the maximum pressure change at the various points in the pipe line will not coincide with the instants of time selected for the stepwise representation of the actual gate movement. Hence, if the gate movement is very irregular, smaller time intervals must be used in the computations. 25. Example Consider a pipe line of uniform diameter and thickness as shown in Figure 23, At this installation the gate is closed in 6 seconds in a manner specified in Figure 24. imum tronsent pressure ‘ior haca seat ‘eeecntett fichead Hem seit Bi cult oer sec - “Agata Hoe. ie ae ah 3000 per see Figure 23 The following relations which were derived above are used in the waterhammer computations: Al any point on the pipe away from the gale H-H =F f. Al the gate +f. (15) (Cad B= CAD vay, (28) _ 2B, B fay Ve . v +e (2) + 4(it 4+ 22 +2/)s (2) Pa VW) +h (30) 2L 109 = ~#(: 22, ana 40 WATERHAMMER FOR SLOW GATE MOVEMENTS - [§ 25 A summary of the waterhammer computations is shown in Table 1. ‘The values given in the various columns in this table are obtained in the following manner: The time ¢ = 0 is taken at the start of the gate movement, and time intervals of 1 second as shown in column 1 are selected for the computations in order to determine the head rise at the mid-length of the pipe. Values seen on z eed ge! | 3% r 7 $ TIME IN Seconos C4.8g20-* GATE CLOSURE TINE RELATION Figure 24 of Cad, are read directly from Figure 24 and are recorded in column 2, Values of B are then computed for each 1-second time interval and recorded.in column 3. This completes the first three columns of the table. In the remaining computations it is necessary to compute successively the values of V, F, f and (F +f) for cach of the times listed in the first column. For example, for rasue a Tae ea Isteonds |e 4%] ® ran on [era aaOTE V0.1 mn i) 3 e2| Dy $25) WATERHAMMER FOR SLOW GATE MOVEMENTS. at t= 1 second the value of V is determined from Equation (20) using the value of B = 0.432 andf = 0. The computed value of V = 10.162 feet per second is recorded in column 4. ‘The magnitude of the direct wave F at ¢ = 1 second is then computed from Equation-(30). ‘This computed value of 53.2 feet is recorded in column 5. The magnitude of the f waves which reach the gate are then obtained from Equation (21). These values are recorded in column 6. At the time = 1 second no reflected wave has returned to the gate'so that f = 0. From Equation (15) the head rise at the gate is the sum of the F and f waves which are present at the gate. This value is recorded in column 7. For ¢ = 1 Taste 2 fr ' ern | second, the head rise is 53 feet. Similar computations are made for the terms V, F, and f for t = 2, 3, 4, ete., thus completing the tabulation. In Table 2 an account has heen kept of the waves which Teach the mid-length of the pipe line at various times. Since the first instantaneous movement of the gate is completed at ¢ = 1 second and the wave travel time from the gate to the mid-length of the pipe is 0.5 second, the first direct F wave of magnitude 53.2 feet reaches the mid-length of the pipe at { = 1.5 seconds. This value is recorded in column 2. This F wave reaches tho res. ervoir at ¢ = 2 seconds and the refleeted wave arrives at the mid- length att = 2.5 seconds as an f wave of opposite sign as recorded in column 3. Ina similar manner the sum of all the F and f waves which reach the mid-length of the pipe are recorded in the tabulation,43 i 42 WATERHAMMER FOR SLOW GATE MOVEMENTS [$26 T | f HUF + t)eHead nse of the gate 200 i; f simon seconosl iL fe mavement -2ooh-stortset to TIME HISTORY OF THE PRESSURE CHANGES WTHE PIPELINE ROJACENT TOTHE GATE Figure 25 i 26. Interpretation of solution ‘The phenomenon of waterhammer in a uniform pipe line for ate closure is scen to consist of successive pressure waves of both the F and f type which surge up and down the pipe as changes in velocity occur at the gate. For the example shown, theso waves aro initially positive but upoa traveling up the pipe to the reservoir they are reflected as negative pressure waves which return toward the gate, where they are again reflected. The summation of these waves of alternating magnitude at the gate at any prr- ticular time gives the rise or fall in pressure produced by the gate movernent. In the above problem time intervals of 1 second were selected between successive stepwise gate movements to approximate the GFE SERVZAMMER FOR SLOW GATE MOVEMENTS a eoeft tI i 7 [7 sees a ~ 8 20 Heod rise a mithengh ofthe pipeline |_| Fess I~ I~ i e's closed ol ta 6 seconds yom ee \ TN ri 7 SL TIME HISTORY OF THE PRESSURE CRANGES AT THE MIDLENGTH OF THE PIPE LINE Figure 26 ectual gete movement, The gate opening at the end of each of these time intervals corresponds to the actual gate opening. TTence the tabulation given in Table 1 gives correct values at ench instant of time shown. To find the correct values for other instants of time would require a recomputation, using smaller time intervals. However, to aid in visualizing the changes in the mng- nitude of the pressure waves and the pressure rise caused by this wave ection, the values of each 1-second time interval are connected by smooth curves ns shown in Figures 25 and 26. Ia tho present problem such curves represent. very good approximations for the iend rise at any time, since the gate motion specified in Figure 2 is a smoothly varying function, ‘he maximum hend rise in the pipe adjacent to the gato is computed ns 272 font and tha mavimiim hen? twa at the rider 44 WATERHAMMER FOR SLOW GATE MOVEMENTS [$26 longth of the pipe is 143 foot. ‘These values for the maximum, head rise are plotted on the profile of the pipe curve is drawn through these points as shown in Figure 23. fa curve is useful in determining the required pipe thickness to withstand safely the maximum head. For example, the maxi- um head in the pipe adjacent to the gate is 772 feet, and tho maximum head at the mid-length of the pipe is 563 feet. noted that the maximum head rises at these two points in the pipe line do not oceur at, the same tim vil Theory of Graphical \Waterhammer Analysis for Gate Operation 27. Conjugate waterhammer equations! ‘The analytical waterhammer solution shown in the preceding chapter gives an indication of the extent of the computations in- volved in obtaining a solution by that method. In general, the necessity of calculating the term V accurately from Bquation (29) make method a tedious one. A considerable ease in computations also resulls when there are physical discontinuities in the pipe line. Tn order to obtain the walerhammer solution more a graphical method of analysis is used. ‘The method is derived as follows: The fundamental waterhammer equations are HI re ®) ty (e 2); (as) V-We ~8[r(e -3) AC +2)} (16) Subtracting Equation (16) from (15) gives H-Wy=2£(V — Vo) +2F a ) - :). (33) a ‘The addition of Equations (15) and (16) gives ~ I= -2(V —Vo + z 34 H-M gh Yas a(e+2) 4) Since Equation (33) contains only the F wave term, this equation dofines the velocity and head relations in the pipe line in terms of F waves. Similarly, Equation (34) defines the velocity and head relations in terms of f waves. "Soe Reference 3. 4546 WATERHAMMER ANALYSIS FOR GATE OPERATION [§ 27 Consider the pipe line shown in Figure 27. ‘The lower section of pipe is of uniform diameter and thickness and has a wave velocity of a; while the upper section of pipe has a wave velocity of % feet per second. Suppose that there is an F wave at By at time 4. This wave moves up the pipe with a velocity of a; feet ove wipro Area, Figure 27 per second and reaches C; at 4 = 1 + In/ai seconds. Similarly, suppose that there is an fa wave at Crt time, This wave moves down the pipe with a velocity of a; feet per second and reaches By at & = ts + Li/a; seconds. Equation (33) is now applied at By and C; while Equation (34) is applied at By and C; as follows: How — How = 9 Vou ~ Vou) + 2Fs3 Flew — Hom = & (Vou — Ven) + 2Fi; Hou — Hew = (Vow = Vou) + fei Vw Haw = ~F Vou — Vou) + Ye Alter eliminating F, and f, from these equations and using the relations Han = Hew How = How Vou =Vew Vou = View the following equations are obtained: gv e0 = Vou) (35) How — Hein = Hou — Hau = ~& Veu ~ Via). (36) ys $27] WATERHAMMER ANALYSIS FOR GATE OPERATION 47 In the derivation of Equations (35) and (36), B, and C; are arbitrary points on the lower section of uniform pipe, while By and C; are arbitrary points on the upper section of uniform pipe. Hence these equations are applicable. between any two points, including the end points, of any section of uniform pipe in the pipe line. These equations are rewritten in another form involving the following ratios: for gate closing movements; for gate opening movements, I steady velocities of water in the pipe sections prior to gate closure, and V, and V! are the final steady velocities of water after gate opening has been completed, Then Ian, — hres = 201(Yaw = Yow); (37) few — hou = —2prlveum — tea), (38) where for gate closure, Bate Equoin (38) fe 2 Zenon ; aces vs Car Figure 2848 WATERHAMMER ANALY. FOR GATE OPERATION [$28 and for gate opening, aw, as, a ole and p; 29H Equations (37) and (98) are the conjugate waterhammer equations, while p, and p; are the pipe line constants, On a rectangular coordinate system of h and v the conjugate equations are represented by straight lines which have a slope of -£2p and pass through known points as shown in Figure 28. 28. Gate discharge equations Each conjugate waterhammer equation defines the relation between A and v at a point in a uniform pipe line in terms of A and » at another point in the pipe under the action of the waterhammer wave phenomena. In order to determine h and v explicitly, second independent relation is required. For gate operations relation is determined from the discharge at the gate. ‘The ial discharge through the gate under steady conditions is Va, = BV Ha; and the discharge at any gate opening and head as defined by Equation (28) is Va = BVH. Then m= t Via, (39) where ph hun He and r= 2, Although Equation (39) has heen derived for gate closure, an identical equation results for gate opening by letting 7 = B/B,, where B, is the gate opening factor at the end of the gate movement. 29. Graphical solution for gate closure Consider the pipe line and the gate closure relation shown in Figures 29a and b. “This problein is identical to that solved by analytical methods in the previous chapter. In order to illustrate the solution in its simplest form, the head rise at Lhe gate is deter $99] “WATERHAMMER ANALYSIS FOR GATE OPERATION 49 mined for 2-second time intervals. Since the wave travel time from A to E is 1 second, the necessary conjugate waterhammer equations for the end points of the pipe line are as follows: fe = +200 — U6)5 (0) hay = —200e, — t4)§ (b) +204, — ted; (0) a hay = =2(0e, — 95d; (A) tha — hey = +204 825 (0) thy — hha = —2(0e, — Pad, (O 1 > ? 1 (40) In these equations the head and velocity ratios h and v at the gate are indicated at ¢ = 0, 2, 4, and 6 seconds. Hence only the «is- charge relation at the gate at these specific times will be required in the waterhammer solution. ‘These are as follows: My = 1.0 Wha (m) t= 0.7 Ving (0) 24, = 03 Vins (0) ay 24 = 05 ©) 04, = 0 when ¢ > 6 seconds. Equations (40) and (41) are now solved simultaneously graphical methods as follows: On a set of coordinate axes of and h, parabolas m, n, v, and p arc drawn for values of + = 1.0. 1 in Figure 29e. The parabola r = 0 0.7, 0.3, and 0, ag sho coincides with the vertical axis at v = 0. The point ds prior to the gate movement is located at the coordinates k= 1, 6 = 1 Since the effect of a gate movement does not reach the reservoir end of the pipe until after the wave travel time of 1 second, Ey is also located at the same point as de. Equation (40a) is then rep resented as a point on the diagram at k= 1,0 = 1. Equation (0b) is the equation of a straight line with a slope of —2e which passes thronzh the point £3. Since the point correspontling to ts ix concurrently located on the parabola n and the sloping line 6. it must be located at the intersection of theze two lines as shown in the figure, Equation (4Nte) is the equation of a straight line of slope 2p which passes through the point as. For a nt reservoir level #3 is located on the horizontal axis ath = 1. The50 WATERHAMMER ANALYSI. .OR GATE OPERATION [5 29 ees aan ome GATE CLOSURE TIME RELATION © a by ie = oa oF ae ae 38 10 © igure 29 points corresponding to Ay, Ex, As, Er, Ay, ete., are then determined in the manner shown in the figure. A comparison of the results of this solution with those of the analytical method given in the preceding chapter shows that the results obtained from the two methods are identical. ‘The graphical method of waterhammer analysis is sen. to consist essentially of the graphical solution of simultaneous equations. One group $29] " WATERHAMMER ANALYSIS FOR GATE OPERATION 51, of equations depends upon the values of r, the effective gate opening, which is specified. ‘These gate relations are represented by parabolas on the h-v diagram. A second group of equations is obtained from the conjugate waterhammer equations. These ‘equations are represented on the h-v diagram by straight lines of slope -t2p. The intersections of the sloping lines and parabolns determine the head and velocity ratios at the gate and the intersections of the sloping lines with the horizontal line h = 1 determines these ratios at the reservoir end of the pipe.a vill Waterhammer Produced by Gate Closure 30. At the mid-length of pipe The graphical method of waterhammer analysis is extended to determine the head changes at the mid-length of the pipe for the example used in the preceding section. ‘Time intervals of 1 second are used in the computations, and parabolas corresponding to values of r for these time intervals are shown on the h-v dia gram in Figure 30. Conjugate waterhammer equations at 1- oF IE Ty i CORNY Figure 30 ~ i . 5 $30] WATERHAMMER PRODUCED BY GATE CLOSURE 53 second time intervals are now required for the pipe sections AC and CE, Instead of writing these equations formally as in the preceding chapter, a tabular representation of the equations can be used as shown in Table 3. . In this table ¢ = 0, 1, 2, 3, 4, ete., are shown in the column directly below the gate at A. The times recorded in the center column at C, the mid-length of the pipe, are equal to times recorded at the gate plus one-half second, the TABLE 3 ee. The numbers in the table ideale the hein seconds offer the stort of gate chsure of the points A,c oF € wave travel time between the gate and the mid-length of the pipe. Similarly, the times recorded under £, at the reservoir end of the Pipe, are the times recorded at the mid-length plus one-half see- ond. The lines wi equations and the direetion of travel of the press From a consideration of the action of the rellected waves, de; Cos, and 2,9 are located at ath =1,v = 1, Then Are and Crs are locate tial diteet and cominon point the intorsee-34 WATERHAMMER PRODUCED BY GATE CLOSURE [$31 tion of the line of slope —2p and the parabola corresponding to 7 = 0.9, The graphical representation of the remaining conjugate equations is shown in Figure 30 with the segments of the ‘sloping lines marked according to the designation shown in Table 3. ‘The maximum head rise at the gate as obtained from these computations is 0.54H, and occurs at t = 3 seconds, while the ‘maximum head rise at the mid-length is 0.28H, and occurs at t= 3.5 seconds, 31. At the quarter points of the pipe line ‘The graphical method of analysis is now extended to determine the pressure changes at other points in the pipe line. In general it is necessary to select points on the pipe line which divide it into an integral number of pipe sections of equal wave travel time, The maximum time interval used in the computations is then the round trip wave travel time for any pipe section. For example, in order to determine the pressure changes at the quarter Points in the pipe line shown in Figure 29a, one-half second time TOE URINE 7 vin 2 a) $39] WATERHAMMER PRODUCED BY GATE CLOSURE 35 intervals are used. The solution then determines simultaneously the pressure changes at the gate, the one-quarter point, the mid- point, and the three-quarter point of the pipe line for one-half second time intervals as shown in Figure 31. 32. Pipe line with a stepwise change in diameter In using the conjugate equations in the waterhammer solution the factors which are considered are p, the pipe line constant, ‘and the wave travel time between two points in a uniform section of the pipe line. For a pipe line which has stepwise change in the cross-sectional area or wave velocity, these conjugate equa- a Sia L500 ft ake [menoosat — levrassert lmewoasent Gare crosune Time RELATION ©) ie] Bel PAR fob eee HAT fly I VT [0 a * be MEP ha wale | Blom So tea Figure 31 “ mi Gs G © Figure 3256 \WATERHAMMER PRODUCED BY GATE CLOSURE [32 tions are also applicable at the end points of each section of uniform pipe. On the h-v diagram the graphical representation of these equations is similar to that shown in the preceding section with the exception that one set of the conjugate equations is represented by lines of slope -+2p,, while the other set has a slope of 2m. For example, consider the pipe line shown in Figure 32a and the gate closure time relation shown in Figure 32b. The round trip wave travel time in pipe sections AB and BC is 1 sec- TABLE & 30 “0 so ‘0 ae, ond. Therefore the largest time interval which ean be used in the computations is 1 second. Paraholas corresponding to the values of for 1-second time intervals are shown on the hv diagram in Figure 32c. ‘The applicable conjugate equations for 1-second time intervalsfor the pipe sections AB and BC are indicated in Table 4, This tabulation differs from Table 3 only in the respect that the value of p is different in the two sections of pipe. ‘The graphical representation of these conjugate equations is shown in Figure 32e, ‘There is no particular difficulty in obtaining a waterhammer 5 $32] WATERHAMMER PRODUCED BY GATE CLOSURE 57 solution when the round trip wave travel times in each of the two sections are different as shown for the pipe line in Figure 33. The method is also applicable when there are more than two stepwise changes in the diameter or wave velocity in the pipe line. How- F yss000 toe iigranr per ee GATE CLOSURE TIME RELATION o cy Figure 33 ever, when there are many stepwise changes in the pipe line this method becomes too tedious. In such cases the approximate method given in Chapter XV is useful and of sufficient accuracy for most purposes.38 WATERHAMMER PRODUC! BYGATE CLOSURE [533 I gate closure ‘When the control gate is closed only part way and held there, the waterhammer solution is as shown in Figure 34. This solu, tion shows that although pipe line friction has been. neglected, the partial reflections which occur at the open gate cause the pressure and velocity in the pipe to return quickly to a steady value, oe GATE CLOSURE TIME RELATION CG} ke A e/| so | a ye | ait © Figure 34 $34] WATERHAMMER PRODUCED BY GATE CLOSURE 59 34. Gate closure with a relief valve! Consider a pipe line which has a relief valve near the control gate as shown in Figure 3a. The type of relief valve which is considered is often used in turbine penstock installations for the control of waterhammer, and is operated in such a manner that the relief valve opens as the turbine gates are closing. A short time later the relief valve closes at a much slower rate with a negligible rise in pressure. ‘The assumed combined effective gate Uniform closure in 60 seconds. ==feambned movement (ontrorgote=s cam SS oat 77 —* — ore m Seconos GATE MOVEMENT TIME RELATION (od & a wal Lx} IF “eo Figure 35 * Bee Reference 40,oe WATERHAMMER PRODUCED BY GATE CLOSURE [36 h = 1 to the parabola for each gate opening is equal to B’B’ below the line h = 1. The solution indicates that the maximum drop in head at the gate is 0.53H, and the maximum head rise is 0.22 {a 3000 ft ot se. Agpio3 ft per see GATE CLOSURE TIME RELATION Cc) f je és Ba, BF gat 0 OT «© Figure 37 Ix Waterhammer Produced by Gate Opening 37. Fall in pressure produced by gate opening! ‘When the flow of water in a pressure conduit is retarded, the kinetic energy of the water ig converted into potential energy as evidenced by an increase in the hydrostatic pressure. Conversels, when the flow of water in a closed conduit is increased, enerzy is supplied to set the water in motion, with # consequent reduction in the hydrostatic pressure. For example, consider the pipe line shown in Figure 38 and the effective gate opening shown in Figure 38b. Parabolas corresponding to the values of r for 1- second time intervals are shown in Figure 38. Th itial velocity of water in the pipe is zero and hence the starting point of ‘the graphical solution on the A-» diagram is at the point A = 1, v = 0. This locates the point for Ao, Bos, and Ci. The graphical waterhammer solution is then completed as shown in the figure. 38. Pressure rise produced by gate opening After the initial drop in pressure due to a gate opening, pressure rise above normal occurs because of the reflected wat hammer waves which return from the reservoir. For example, consider the pipe line shown in Figure 30a and the gate opening shown in Figure 39b. ‘The waterhammer solution for the pressure changes at the gate is shown in Figure 39¢ from which it is seen ‘that at ¢ = 6 seconds the pressure rise is 0.10H ‘The maximum head rise due to a gate opening occurs when an initially closed gate is opened to its final position in 21,/a sec- "See Reforenees 1 and 25. 63a 64 WATERHAMMER PRODUCED BY GATE OPENING. [§38 onds or less and held there. To illustrate this, consider the pipe line shown in Figure 40a and the gate movement shown in Figure 40b, for which the graphical waterhammer solution is shown in Figure 40c. ‘The maximum head rise at the gate due to the return of the reflected waves from the reservoir is 0.23Hy and occurs at as GATE OPENING TIME RELATION oocle ERIE] FY, ENTREE Seney Figure 38 t= 4 seconds. For this type of gate opening the head rise depends only upon the pipe line constant p.. Hence the results of ‘waterhiammer solutions of this type can be indicated on a chart as shown in Figure 41. From this figure it is seen that the maxi mum head rise produced by a gate opening from zero gate is 23.1. 5 $39] WATERHAMMER PRODUCED BY GATE OPENING 65 SN ven3.22 ft per sec. Minne o rr GATE OPENING TIME RELATION ) © Figure 39 39. Rhythmic opening and closure of control gate* Under certain conditions of gate operation, it is possible to obtain a rhythmic opening and closure of the regulating gate. Such conditions have been observed, for example, in hydraulic turbine installations due to governor action. If the gate motion isin phase with the motion of the waterhammer waves in the pipe line an appreciable head rise ean occur. For example, consicler the pipe line shown in Figure 42a and the gate movement time history shown in Figure 42b, for which the waterhammer solution is shown in Figure 42c. This solution indicates a maximum head rise of 1.24» at the gate when a complete cycle of gate operation #80 References 1 and 21an 60 WATERHAMMER PRODUCED BY GATE CLOSURE [534 opening of the control gate and reliof valve is shown in Figure 35b. ' Since the waterhammer effects depend only upon the changes in velocity of water in the penstock, the waterhammer solution is obtained from the combined effective gate opening as shown in Figure 35¢. fag 3000 ft per see, 942250 ft per sec. e073 Peper see. WetO 73 ft per see o ipeors Sor fovon, ied ees sane eee Time m Seeonos GATE CLOSURE TIME RELATION ) 4q $35) WATERHAMMER PRODUCED BY GATE CLOSURE 61 35. Control gate located away from outlet end of pipe Consider the pipe line shown in Figure 36a and the gate closure shown in Figure 36b. Since the largest common round trip ‘wave travel time which occurs in both sections of the pipe is 1 second, this time interval is used in the waterhammer compu tions. ‘The discharge through the gate varies with VRaxy — hn ‘As the gata is closed, pressure waves form at B’ and B' move toward A and C. In the waterhammer solution By’ and C1 are located at h = 1,» = 1, while Bgand Asare located ath = 0, v = Las shown in Figure 86c. A line is now drawn from (this is the same point as C,) which has a slope of —2p: = —2on which Bi’ is to be located. Similarly, a line is drawn from 4p which has a slope of +2: = 1.5 on which By is to be located. Since the flow on both sides of the control valve are equal, ii’ and Bi are located on the same vertical line. Moreover, the-e two points are located in such a manner that the vertical distance from the point Bi’ to the parabola corresponding to r = 0.9is the same as the vertical distance from the horizontal line k = 0 to Bio that is, BiN = Bi'M, The reason for this is that at ¢ = 1.0) second, the flow through the gate is 0.9V hg, = Tae A simple method for locating the points Bi. and Bi’ is illustrated by the dotted lines in Figure 36c. Starting from By’ draw 1 dotted line of slope —2(o + p:) which intersects the parabola corresponding to r = 0.9 at the point M. ‘Then Bi» and Bi’, are located on the vertical line which passes through the point Mf. Other points which are required in the solution are determined in a similar manner as shown in the figure. Ye, = Yer, 36. Control gate located at upper end of pipe Consider the pipe line installation shown in Figure 37a and the gate closure time relation in Figure 37b. For this installation har = 1.0. . Hence, the waterhammer solution shown in Figure 87¢ is obtained by-locating the points %, Bi, By, ete., so that on the same vertical line the distance B’M above the horizontal line66 a WATERHAMMER PRODUCED BY GATE OPENING [539 322tt per see. p=03, i devort © rea ieee tree ee TIME 1m Seeonos. GATE OPENING TIME RELATION to) ; aD a a a @ Figure 40 5 $39] WATERHAMMER PRODUCED BY GATE OPENING 67 ia performed in 4L/a seconds. Since the head rise due to this type of gate movement depends only upon the parameter p,, the results of all possible solutions are indicated on a chart as shown, in Figure 43. From this figure it is seen that as p, increases, the maximum head rise approaches 2H as a limit. a FI mae aes i 1 eee arn a cE SUNTS | Ses b Ly T “CCH a a 1 e a 7%: = ‘i v oF cy Figure 4168 WATERHAMMER PRODUCE BY GATE OPENING [$39 $39) ies oer te "05 ooan ° 19 0 Se ee Se Time im seconos GATE, MOVEMENT TIME RELATION @ ay ee I ee ub S oY | to) KONG vl « lot or KON 4 os| | 7 V Nhe 0 % REGS OF BOE aE () Figure 42 ny WATERHAMMER PRODUCED BY GATE OPENING gfe 2g % INE constant - 3 cae | Peay 05 Tera ie wloty fet se send’ ‘ie Stoty wey of water the pipe im Tet per Second for the rmosimen gote opening, gsAczlerotion of growty in Feet er second, and Hy *Net Need in fet of water ‘ot the gore for Me itil stebdy flow conditions. MAXIMUM HEAD RISE AT GATE DUE TO RHYTHMIC GATE OPERATION Figure 43 Cs)x Waterhammer for Uniform Gate Operation 40. Uniform gate closure to rero gate ‘The graphical method of waterhammer analysis given in the preceding chapters provides a method for determining the head changes at various points in a pipe lino for any type of gate movement. When the effective gate opening varies uniformly respect (o time, the gate motion is called uniform gate operation. For this typo of gato motion it is not necessary to make a graphical waterhammer analysis for each pipe line and gate elosure, hecaueo the results of a large number of watethammer solutions can be shown on charts. Figure 44 shows one of these charts for un form gate closure to zero gate.' Tn order to determine the mas mum head rise at the gate, it is necessary to calculate the pipe ine characteristic p and the time of gate closure in terms of the number of 2//a intervals, that is N -3S. (42) ‘The value of I is then road from the chart and the head rise determined from the relation AH = 2pK Ho. (43) When s 1, thatis, whon the complete gate closure is performed during a time equal to or loss than 2/,/n seconds, K — 1 and AH = 2plJo, Yor such rapid gate closures the head rise is also equal to a AV as defined by Equation (31). 1 Seo References 1 and 26. 70 40] WATERHAMMER FOR UNIFORM GATE OPERATION: ed aT 4 : i NEA 5 Lies a she rmsr vaLues oF « de He srezpe RAY S. QUICK’S WATERHAMMER CHART FOR UNIFORM GATE OPERATION Figure 44 nch WATERHAMMER FOR UNIFORM GATE OPERATION = [§ 41 dewteet ‘= 3000 ft per second Qs (full gate discharge) 843cu.tt per sac. Full gate closure in 6 see. ,,,,. Gote closurein & seconds. tee oL_| stotieneten a a encentsoe oF wii are Oren HEAD RISE AT GATE OUE TOUNIFORM CLOSURE FROM VARIOUS INITIAL GATE OPENING Figure 45 If the gate closure can be accomplished from any initial gate position at the same uniform rate, the maximum head rise at the gate occurs when the uniform gate closure takes place in exactly 2L/a seconds. ‘This is illustrated for a typical case in Figure 45. 41. Uniform gate opening from zero gate” When the effective gate opening from the zero gate position is uniform with respect to time, the minimum head at the gate is found to oceur at exactly 2L/a seconds after the start of the gato movement, If the gate opening starts from the zero gate position, han = (-K + VE $1), (44) where K = LV,/gH.T. In this equation V, isthe final steady velocity of water in the pipe after the gate opening has heen com- 41] WATERHAMMER FOR UNIFORM GATE OPERATION 73 ‘there Hains minimum neod ot the gate in feet, Ta intial steady Neag at the gate infect, agin of pipe infect, ime for opening gate othe desired position in seconds (TS 24), Ve tinal steady velocity the pie in eet per ‘second ofter gate opening, = acceleration of gravity in feet per second® a. ai 3 4 ‘osoe bao = a MINIMUM HEAQ AT GATE DUE TO UNIFORM GATE OPENING FROM THE CLOSED POSITION Figure 46xi Waterhammer in Pump Discharge Lines Caused by Power Failure 43, Transient conditions at pump and discharge line! ‘There are many operating conditions at large motor-driven centrifugal pump installations which are eapable of producing substantial pressure changes in the discharge lines. Of these con ditions one of the most important is the rapid deceleration of the Pump motors because of a power failure. Figure 47 shows the time history of the pressure, flow, and speed changes at pump installation produced by power failure at the pump motors. When the power supply to the pump motor is suddenly cut off, the only energy that is let to drive the pump in the forward direc. tion is the kinetic energy of the rotating elements of the motor and pump and the entrained water in the pump. Since this energy is usually small when compared with that required to maintain the flow against the discharge head, the reduction in pump speed is quite rapid. As the pump speed reduces, the flow of water in the discharge line adjacent to the pump is also reduced. As a result of these rapid flow changes, waterhammer waves of subnormal pressure are formed in the discharge line at the pump, ‘These subnormal pressure waves move rapidly up the discharge line to the discharge outlet, where a wave reflection occurs. Soon the speed of the pump is reduced to a point where no water can bbe delivered against the existing head. If there is no control valve resent at tho pump, tho flow through the pump reverses, al. though the pump may be still rotating in the forward direction. The speed of the pump now drops more rapidly and passes: through zero speed. A short time later the pump, acting as a "See Reference 3, 33,34, and 35, 4 $49] WATERHAMMER IN PUMP DISCHARGE LINES 5 turbine, reaches runaway speed in reverse. As the pump approaches runaway speed, the revere flow through the pump reduces rapidly, and this reduction in the flow produces « pressure rise at the pump and along the length of the discharge line. In order to determine the transient hydraulic conditions at the pump and discharge line subsequent to 8 power failure at the ead of pump 38 ad enh 0 aseharge ne + cuvallar pian Ba \ org — Pome saved Zone of turbine operation] PERCENT OF noma BPtt0, neAO ano OSCUOREE Poae fare oeere of 170 ~ i ree Secows AFTER POWER Fane aia TRANSIENT CONDITIONS FOLLOWING POWER FAILURE Figure 47 ‘motor, three effects must be considered; namely, the pump tnd motor inert the pump characteristics, and the waterhammer wave phenomena in the discharge line. The effect of the pump aud motor inertia is obtained from the inertia equation. This equation defines the relation between the pump speed and torque at a given instant of time in terms of the kinetic enerzy of the rotating system. ‘The pump characteristics are obtained from a complete pump characteristic diagram. This diagram de-16 ‘WATERHAMMER IN PUMP DISCHARGE LINES ($43 fines the manner in which the pump torque and speed vary with the head and discharge throughout the range of operation as a pump, energy dissipator, and turbine. Finally, the waterhammer effects are obtained from the waterhammer equations. ‘These ‘equations define the relations between the head and flow in the discharge line during the transient flow conditions under the ao- tion of the waterhammer waves. 43, Inertia equation ‘When the power to the pump motor is suddenly cut off, the deceleration of the pump at any instant depends upon the flywheel effect of the rotating parts of the pump and motor and the instantaneous torque exerted by the pump impeller. For a rotating system the accelerating torque is equal to the product of the mass moment of inertia of the rotating system and the angular acceleration. Following power failure at the pump motor, the decelerating torque on the rotating system corresponds to the pump torque. If the decelerating torque is taken as positive, WR da, %, (45) — 4, this equation is written da w= 1%. HE For a small time interval At = approximately as follows: Mi+My WR (wr = on) 2 7 at (46) ‘This equation is written with the ratios a = .N/Ne and B = M/Me as follows: mo = SHEE (6. + BAL (a7) ‘The decelerating torque at the rated head and pump speed is 60wH eQr. Ma- 2eNene « Thea = a =KiGi+ fds, 9) a K, = S50 gwHaQe _ 91,6001 xe, (50) * PW RyaNi WRN 2 8 T L__eos-ascrorce, Jpotea meee: Le FAs TYPICAL PUMP PERFORMANCE CURVES. Figure 48 Figure 4978 WATERHAMMER IN FUMP DISCHARGE LINES (5.44 44, Pump characteristics* = Typical pump performance data usually supplied by the pump manufacturer are shown in Figure 48 and include the head, brake horsepower, and efficiency plotted against the discharge. ‘These data are converted to a family of torque and speed curves on an fv disgram in the following manner: Referring to Figure Figure 50 49, the point corresponding to the rated head and discharge of the pump has the coordinates h = 1, v = 1, ‘The curves for a = 1 and 6 = 1 also pass through this point on the diagram, since these ratios are defined in terms of rated values. Other values of 6 on the curve for c = 1 are determined by computing values of h and v from the brake horsepower curve given in Figure 48, Points on * See Reference 39 > $44) WATERHAMMER IN PUMP DISCHARGE LINES 19 oss PO PUMP CHARACTERISTICS OIAGRAM ZONE OF NORMAL PUMP OPERATION Figure 512 the diagram for other values of « and f are then determined from tho laws of homologous pump operation, For example, for a given pump, If and Af vary with N2, while N and Vif vary with @. To use these relations on the A-v diagram a parabola of the type /s = Kye* is passed through a known point on the di such as that shown in Figure 50 ata = 1.0,8 = 0.8. The vertie: distance 4 = 1.10 at this point is then divided into cight equal parts, and the points corresponding to 8 = 0.7, 0.6, etc. are lo-80 WATERHAMMER IN PUMP DISCHARGE LINES (844 zee | SEye PUMP CHARACTERISTICS DIAGRAM ‘ZONE OF ENERGY DISSIPATION Figure 51 cated on the parabola at ordinate values of b = X(1.10), $(1.10), ete. Similarly, points corresponding to a = 0.9 and 0.8 are located on the same parabola at values of A = 1.10(0.9)* and 1.1008), A smooth eurve is then drawn throuth the points on the various paraholns corresponding to cach value of a and B. Figure 44) shows familics of Lorque and speed curves which were constructed in this manner. ‘Phe pump characteristics shown in Figure 49 are for the zone WATERHAMMER IN PUMP DISCHARGE LINES at Zs a ae ro BT | ea Be | hod [B+ ALT pes ol t "zpwE or tumaine OPERATION Figure 51¢ of normal purip operation. When power failure occurs at, the pump motor, these characteristies arc adequate for determining the minimum transient pressures at all points in the discharge line and for determining the likelihood of water column separation. However, after the flow reverses through the pump, additional pump characteristics are required for the zone of energy dissi tion in which the pump is running in the forward direction with the flow through the pump in reverse, and for the zone of turbine operation, in whicli the flow through the pump and the pump: rotation are both in reverse. Complete characteristics for a single82 WATERHAMMER IN PUMP DISCHARGE LINES (645, suction pump in each of the three zones of operation as determined from actual test data are shown in Figures 51a, b, and c. 45. Waterhammer equations The waterhammer equations for the pump discharge line shown in Figure 52 are as follows: fou = how = —2er(tan — toa)s (61) Few — hou = +2px(You — You) (51B) In these equations (f ~ t) is the wave travel time between By and C; for an F; type pressure wave, and (4 ~ 4) is the wave travel time between C; and Bs for an f; type pressure wave. For 8 uniform discharge line, p = aVe/2gH, the pipe line constant. Wore velocitys 0-7 Pipe line canstnt«zpe Wave velocitys 0,” Pige ine constonts2p, x= Figure 52 goretete Fae Sf Fema) 5 Beots'ap 5% ce yNeennan fa q 3 rr) Figure 53 * Bee Reference 29, 5 $46) WATERHAMMER IN PUMP DISCHARGE LINES 83 ‘These equations define the relations between the head and velocity ratios during the waterhammer wave phenomena. On a coordinate system of and », these equations are represented by straight lines which have a slope -&2p as shown in Figure 53. ‘Each waterhammer equation gives one relation between h and » at a point in a uniform section of the discharge line in terms of h and v which are known at another point in the line. In order to determine h and » explicitly, a second independent relation is required. This relation is obtained from the pump-inertia equation with the use of the pump characteristic diagram. 46." Graphical watethammer analysis Consider the pumping plant installation shown in Figure 54. ‘Ifa power failure occurs at all three pump motors, 2p = 2.31 and K, = 0.224. For a time interval At = L/4a, yan a= 262011 persee 1 per se (or 3 pms) 347 cut per see lor 3 pumps) Resatsa ft Hg= He 22ote Bump motor roting = 400 horsepower for each pump ond motor wil of rotating ports= 386910. t12 for eaen pump ené motor Pump speed = 760c0m Pump efficiency = 607 pres bensarsee. Figure 54 ntW~Ye Ye. 7) Anke te ke 84 WATERHAMMER IN PUMP DISCHARGE LINES Sa‘ Fo: ($46 ae « tole 22 03 06 08 08 OF OF 0510 co We. Ve "Qe ied ve $46) WATERHAMMER IN PUMP: DISCHARGE LINES 85 my — a = KiAt(Gs + B:) = 0.0782(8, + 6:). (52) ‘The simultaneous solution of the waterhammer equations and the inertia equations on the pump characteristics diagram by graphical methods is shown in Figure 55. ‘The complete pump characteristics used in this solution are those shown in Figures 5a, b, and c. The starting point in the solution corresponding to As is located on the h-v diagram at the coordinates h = I,v = 1. Au is then located on the line of slope 2p = +2.31 which passes TAGLE 5 PUMP SPEED AND TORQUE RATIOS a |a-a] a-a, Location | (ronaue| cseceo | (raom | crnon’ avio)|_eario) |_CuRvEsilequarion s2 1000 | 1000 | 0 ° 0.760 | 0863 | 0.137 | 0.138 esi [0.755 | 0108 | oor 0.520 | o.66@ | 0.087 | 0.088 0.440 | 0.892 | 076 | 0.075 0.sa0 | 0.826 | 0.066 | 0.088 fae feat po? leer fa ls? 0.328 | oarz | 0.056 | 0.055 J ATOS 05 =OF OS -05 -08 “05-08-01 0.205 | 0.425 | oer | 0.047 law for? fav? 0.250 | 0.383.| 0.042 | 0.042 o.70 | 0.350 | 0033 | 03 through As in the following manner: The location of Aus is first estimated. Values of ay and 8; are then read from the pump characteristics curves. ‘This value of 61 is used in Equation (52) to compute as. If the computed value of az does not agree with the value of a on the curves, the point for A. is shifted up or down on the sloping line 2p = +2.31 until these values of ay agree. This particular point is found to he located at a value of B = 0.760 and @ = 0.863. Other points for Aritu Astin Che, are determined in a similar manner. The values for several of these86 WATERHAMMER IN PUMP DISCHARGE LINES (47 points are shown in Table 5 and the graphical waterhammer solution is completed as shown in Figure 58, from which the following limiting values are read: Maximum drop in head at pump = 0.924, = 202 feet. ‘Maximum drop in head at mid-length of discharge line ~ 0.60H, = 152 feet. Maximum head rise at pump = 0.61H, = 134 feet. ‘Maximum head rise at mid-length of discharge line ~ 0.35Hy = 77 feet. A time history of the head, flow, and speed changes as obtained from the graphical solution is shown in Figure 47. 47. Water column separation The maximum positive and negative pressure changes obtained from the waterhammer solution are plotted on the discharge line profile in Figure 56 to show the limiting pressures for aig i pom ecteparaaee 300 Noma pumping gradient (tection neglected) euevarion, Feet Eman head on charge he ofr poner fore Figue 56 5 $48) WATERHAMMER IN PUMP DISCHARGE LINES a7 which the discharge line should be designed. When the minimum pressure at any point along the pump discharge line reaches the vapor pressure of water, the waterhammer solution is no longer valid. If this subatmospheric pressure condition inside the pipe persists for a sufficient period, the liquid water column parts and is separated by @ section of vapor. Water column separation sometimes occurs during the initial negative surge waves on long pump discharge lines ut high points which are near the hydraulic gradient, Wherever possible, this condition should be avoided by using either a surge tank, air chamber, or larger motor IR? because of the high pressure created when the two liquid water columns rejoin. When water column separation cannot he avoided, special means must be taken to minimize the violence of impact due to the rejoining of the water columns. This can be accomplished by positioning special control valves or other pro- tective devices which will either reduce the reverse velocity of the upper column or increase the reverse velocity of the lower water column, 48. Watethammer charts‘ In obtaining the graphical waterhammer solution for a pump installation subsequent to a power failure at the pump motors, two independent parameters are used with the complete pump characteristics. ‘These are p, the pipe line constant, and Ki(2L/a) constant which includes the effect of the pump and motor inertia and the waterhammer wave travel time of the discharge line. For a given set of pump characteristics, the results of a large number of waterhammer solutions can be shown on charts as indicated in Figures 57a to h. ‘These charts furnish & convenient method for obtaining the limiting transient conditions at the pump and discharge line when no control valves are present at the pump. Although the charts are theoretically applicable to one particular type of pump operating at the rated head prior to power failure, ‘they are very useful for obtaining the approximate waterhammer effects in any pump discharge line. ‘See References 12, 33, and 34.a _~ 88 WATERHAMMER IN PUMP DISCHARGE LINES (548 448) |WATERHAMMER IN’ PUMP DISCHARGE LINES 89 : I 3 I TTT . 3 ; Lo Te ! I | me E A hy # Wy Iz sy al 2 74 a rT re PT ann ! Sort HT i OLUD , fe 3 Md sa 7 z hs | “ i, l z 3" Vly i 3 2 3 i | io 7 vatues oF ws 2b 3 t a wawues oF «Be UPSURGE AT MIDLENGTH Figure 57¢,d OOWNSURGE aT mioLENGTH Figure 57a, boS WATERHAMMER IN PUMP DISCHARGE LINES : earrea 4 AS ee t MAXIMUM REVERSE SPEED 3 TIME OF FLOW REVERSAL AT PUMP Figure 576, f (48 $48) WATERHAMMER IN PUMP DISCHARGE LINES ” ! * Py * wy wwe Oo Tse sera ‘ime oF 2eR0 PUKE. SPEED x e220 32 ge™| az aie Bey Se ee vawues of x: 2 TIME OF MAXIMUM REVERSE PUMP SPEED Figure 579, hoe 2 |\WATERHAMMER IN PUMP DISCHARGE LINES (349 49. Conclusions Pressure surges in pump discharge lines subsequent '0 & power failure at the pump motors can ‘be computed accurately if power mplete characteristics of the pump are known. In most cases only the characteristics for the zone of normal pump opera~ tion are obtainable from the pump performance data supplied by te simp manufacturer. This permits en accurate determine tion of the waterhammer effects up to the point at which the flow toterses through the pump. When necessary, pump character eyes for the zones of energy dissipation and turbine operation van be estimated with sufficient accuracy for waterhammer Pur poses. Special W aterhammer Solutions for Pump Discharge Lines nn 50. Effect of changes in pum Consider the installation shown in same installation as that shown in ‘ter surface elevation 22 feet higher. i i98tt 3 oum0s ping head ‘Figure Figure 58. This is the ‘54 with the intake wa- pumping head, Figure 58 9394 SPECIAL SOLUTIONS FOR PUMP DISCHARGE LINES [551 Hy = 198 feet and te = He = 0.90, ‘The waterhammer solution for this installation due to a power failure at the pump motors is shown in Figure 58. This solution is obtained in a manner similar to that shown in Figure 55 with the exception that the starting point As, Czy is located at the intersection of the lines h = 0.90 and « = 1. 51. Pump For a pump discharge line with a swing check valve on the discharge side of the pump, the normal discharge of the pump charge line with check valve Figure 59 +See Reference 38, ‘sue sy ee $51] SPECIAL SOLUTIONS FOR PUMP DISCHARGE LINES 95 keeps the check valve open. However, when the flow through the pump reverses subsequent to a power failure, the swing check valve closes rapidly. ‘The waterhammer analysis for such an installation is shown in Figure 59. This is the same installation as that shown in Figure 54 with the addition of check valves on the discharge side of the pumps. The first part of the solution is the same as that shown in Figure 55 until zero discharge is reached. ‘At this moment the check valve ix assumed to close rapidly and aaa Li 5] rfl Figure 60 the waterhammer effects are then determined on the same basis asforadead end, ‘The maximum head rise at the check valve ix found to be 0.904, as compared with 0.612%, when the reverse flow passes through the pump. In the event that the check valve closure is delayed, the heal tise at the chock valva maay he hisgher. For example, in Figure sit is assumed that the check valve closes rapidly at 13/,/2a seconds after power failure. The maximutn head rise at the pump for this Uolayed valve closure is 2.541fy ws compared with 0.90/fy for the earlier closing check valveSPECIAL SOLUTIONS FOR PUMP DISCHARGE LINES Cy IGE LINE (0) sueTiON Une ( Figure 61 (st \ $52] SPECIAL SOLUTIONS FOR PUMP DISCHARGE LINES 7 52, Pump discharge line with long suction line Consider the installation shown in Figure 61a. This is the same installation as that shown in Figure 54 with the addition of ‘a long suction line. When the pumps operate with a pressure head of Hy on the discharge side and a head of H's on the suction side, the pump discharge depends upon the pumping head H ~ Hg — Is. Upon power failure at the pump motors, pressure waves of subnormal pressure form at B” on the discharge side of the pump, which move toward the discharge outlet at D, and positive pressure waves forma at B’ on the suction side of the pump and move toward the intake at A. In the graphical waterhammer solution of this problem a time interval of Li/a, = Lx/2a: seconds is used in the computations. Now By’ and Dry, are located at h = 1, v = 1, while By and Anja are located at h = 0, » = 1 a8 shown in Figures 61b and e. A line is now drawn from the point Dis Which has a slope of +-2p: = +2.31, on which the point corresponding to Bia, is located. Similarly a line is drawn from the point A,,j., which has a slope of —2p, = —2.31 on which the point corresponding to Biz is located. One of the requirements for locating the points Bim, and Bim is that since terran ™ Yrujay these points are located on the same vertical line. Another requirement is that at ¢ = Li/a, seconds, the change in pump speed must satisfy Equation (49) for a pumping head of (tariyay — Aa-sin)» A simple graphical method for locating the points Bi. and Bi, is illustrated by the dotted lines in Figures 61b and c. Starting from the point Diya. draw a dotted line whose tangent is +2(o + m). On this dotted line determine by trial the point on the pump curves which satisfies Equation (49). For example, the point corresponding to ¢ = In/ay is located at the point M, where a = 0.745. The points corresponding to Bei and Bi,,,.are then located on the vertical line whi passes through the point M. The remaining points for other time intervals are determined in a similat manner. 53. Effect of control valves ‘There are many types of pressure control valves in current ‘use at pumping plant installations. ‘The methods of waterham-8 SPECIAL SOLUTIONS FOR PUMP DISCHARGE LINES [§ 53. mer analysis which are given in the preceding chapters are readily applied to include the effect of any type of control valve which may be used. However, the time and manner in which the control valve acts must be known as well as the flow characteristics of the valve for the various degrees of opening. For the condition of power failure at the pump motors the graphical waterhammer solution is performed initially on the pump characteristics diagram, using the inertia and waterhammer equations with the CRE ig 7 fn a . TFeme tun? I a rE: Lagi oe $54] SPECIAL SOLUTIONS FOR PUMP DISCHARGE LINES 9 with the addition of the controlled valve closure after power failure. 54, Starting and stopping pumps ‘The waterhammer effects due to the normal starting of pumps are usually relatively small. However, these effects can be further minimized by the use of proper control valves or start- us. Check volve mA a M 2.31 | A ne] ea V | 86 oF OF 08 08 18 thane REEL Figure 63 “he Figure 62 ing equipment, and by insuring that all air is removed from the pump characteristics in the manner described above until the flow is controlled by the control valve. The graphical solution is then completed with the waterhammer equations and parabolas corresponding to the flow through the control valve. Figuros 50 and 60 show the effects of rapidly closing gheck valves, When the valve closure is slower, the latter part of the graphical solution is performed with parabolas in the manner shown in Figure 62, ‘This figure shows the elfect of slow closing valves which control the flow at 11/2a seconds after power failure, This is the samo installation as that shown in Figure 54, discharge lines. If there is a shutoff valve on the discharge side of the pump, the pump is brought up to speed with the discharge valve closed. The discharge valve is then slowly opened with no appreciable waterhammer effect in the discharge line. In taking the pump out of service the waterhammer effect is also negligible if the discharge valve is closed slowly before shutting off the power to the pump motor. , If there are only swing check valves on the discharze side of the pumps, a nominal head rise occurs if the pumps are started rapidly. This effect is shown in Figure 63 for starting three pumps simultaneously. This is the same installation as that100 SPECIAL SOLUTIONS FOR PUMP DISCHARGE LINES [554 shown in Figure 54 with the addition of check valves. In this solution it is assumed that the starting up time of the pumps is less than the round trip wave travel time of the discharge line. ‘The starting point Ao, Dzie in the graphical solution is located at h=1,0= 0. Point Ar is then obtained by drawing a line of slope +2p from Dzj. until it intersects the curve a = 1.0. Other points in the graphical solution are obtained in the manner shown in Figure 63. xiii Waterhammer Analysis Including Effect of Hydraulic Losses 55. Losses concentrated at intake* In the derivation of the fundamental waterhammer equations for the elastic water column theory in Chapter II, it was assumed that the velocity head and hydraulic losses were nest gible when compared with the head changes. This assumption is justified for the majority of the pipe lines where rapid changes in flow are most likely to occur. Occasionally, however, it is necessary to include the effect of these hydraulic losses in the waterhammer computations. If the velocity head and the hydraulic losses are assumed to vary with V3, the effect of these losses ean be included in the waterhammer equations by the introduetion of a V? term in Equation (7), the equation of equilibrium for an element of water. However, the simultaneous solution of the resulting equation with Equation (8), the equation of continuity for an element, of water, is not possible.* For pipe lines involving gate movements, a first approximation to the effect of the hydraulic losses is obtained in the waterhammer solution by a hypothetical obstruction located at the upper end of the pipe line. This obstruction has the same total head loss as the entire pipe line, For example, consider the pipe line shown in Figure 64a and the gate closure time relation shown in Figure 64b. The entrance loss, pipe line friction, and velocity head in the pipe line for the initial flow is Hy = 13.4 fect, from which hy = . = 0.028, Th 1 See Reference 4 * See Reference 47. 101. 102 EFFECT OF HYDRAULIC LOSSES (555 sreuseti rgsumed pressure gradient for intial Flow eonitions State Bead | 500 oases eo GATE CLOSURE TINE RELATION ) Fai] a i ul cS ‘cy Ea esd oss porate ooo @ Figure 64 as shown by the head loss parabola in Figure G4e. In the graphical waterhammer solution shown in this figure the hypothetical obstruction is located at C. Hence, points for C; are located on the horizontal dotted line at h = 1,028 and points for C, are located on the head-loss parabola, ‘The graphical waterhammer solution using the head-loss parabola is completed as shown in the figure. $56) EFFECT OF HYDRAULIC LOSSES 103 56. Losses concentrated at pump discharge outlet For a pump discharge line the effect of the hydraulic losses in the entire line can be approximated in the waterhammer solution by @ hypothetical obstruction at the discharge outlet. For example, consider the installation shown in Figure 65. This is the Freee Toner L % sas Tae OF -BE OE OE-OS DEBIT | OF OF OF 03 OE OT OF UF 10 Figure 65 same installation as that shown in Figure 54. For this installa tion hy = 0.07et and hin = 1 — 0.07(1 ~ #) for normal pump flow and kp = 1 —0.07(1 +s) when the flow reverses throtzh the pump. These hoad-loss relations are shown in Figure 65 together with the graphical waterhammer analysis following a power failure at the pump motors.104 EFFECT OF HYDRAULIC LOSSES Assumed pressure 2 gradient fo iition Oazah 1 s000 tt per see, Vo" 537Ft per see. Pre + 7 i 7. ‘re im SeconDs GATE CLOSURE TIME RELATION @ Head. i056 parabola, Ohne , Oy oa es 80 ar os 8s (e) Figure 66 ($56 xr $57) EFFECT OF HYDRAULIC LOSSES 105 along pipe’ 57. Losses concentrated at several px If. closer approximation to the effect of the hydraulic losses is required, a number of hypothetical obstructions at various loca- tions slong the pipe line are used to replace the actual losses in the pipe line. For example, consider the pipe line shown in Figure 66a and the gate closure time relation in Figure 68b. For this installation fy = 0.1740", If Ube tolal head loss is divided evenly at three equidistant points along the pipe line, the loss at each obstruction is 0.058. ‘This is shown by the head-loss parabola in Figure 66¢ together with the graphical solution. In this solution a correction is made for the head loss at each obstruction. For example, B, and Bi are located by trial and error on the same vertical line with the distance betweén them equal to the head loss shown directly below on the head-loss parabola. This method of approximating the effect of the hydraulic losses at a number of obstructions along the pipe is also applicable for pump discharge lines, * Seo Referance 4,xIV Waterhammer Analysis for Compound Pipes 58. Basic equations! ‘The graphical method of waterhammer analysis used in the Preceding chapters was applied to. individual pipe lines without branch connections. The method was used for uniform pipe lines and also for pipe lines with stepwiso changes in the diameter or fengmn ut feength.u” Hy fhreo a . |iPressuce wave {Pressure move velocity 0” wave velocity along the length. By introducing the relations for the pressure ratios and the continuity of flow at the junction of a number of pipes, a method for determining the watetharamer effects for any system of compound piping is now developed. ‘Whenever threo pipes mect at a junetion as shown in Figure 67 and the velocity head and the hydraulic losses are neglected, hon = ho = hoon, (63) * Soe Reference 5. 106 $58] WATERHAMMER ANALYSIS FOR COMPOUND PIPES 107 ‘The requirement for the coritinuity of flow at the junction is AMV = A"V am + A'Vnw (4) If this equation is divided through by A”"Vs-o, AY | AV a "BV eo TA V re ‘The terms on the right-hand side of the equation are now defined as follows: Avy, AV Ban = Gott = bn = gy = ae This arbitrary definition of the velocity ratios Gav, and ip, sim- plifies the graphical representation of the waterhammer equations for compound pipes. The requirement for the continuity of flow at the junction is then et = He + Bee (85) ‘The conjugate waterhammer equation for an F type wave between B’ and E is Fan — Hm = (Vas ~ Ves where ‘This equation may be written in the following form: Fan — hey = 25! Gon — ten), (56) 5 = 208 where pled” ‘The equation for an f type wave between E and Bi is than — horn = —26'(Gen — tor). (56B) A’similar development from the conjugate waterhammer equation in pipe line B"C leads to a! _ avy pled” ~ ‘and for the pipe line B’”D, 1 08 i = A108 WATERHAMMER ANALYSIS FOR COMPOUND PIPES [ots n.sstt Seon peat. /esoas Ws a= 3000 per sec on os0see, Pa2o 25 iu mw seconos GATE CLOSURE TIME RELATION o ($59 aa rive Line 8c a) Figure 68 Pure Une OE 60=1--4 $03, -S0gt03 70 702 04 06 08 7 Jo Hs xa oS OF Par 3% ae lszgdns Sle a Be at 248 jszrh ae sigh 3 le a au og ls 5109! +. +3 —> +V PIPE LINE B'-E PIPE LINE BY x z 2 = e s a Figure 69om 110 WATERHAMMER ANALYSIS FOR COMPOUND PIPES [§ 59 ‘Itis noted that A’ and 5’” are not true pipe line constants for their respective pipes but are similar in form to the pipe line constants. ‘These equations are now used for several types of compound pipe systems. 59. Pipe line with a surge tank Consider the pipe system with a curge tank shown in Figure 68a and the gate closure time shown in Figure 68b. Dur ig the gate closure time the change in level of the water surface in the surge tank is small and is neglected. However, at a considerable time later the water level in the surge tank will rise slowly to a maximum value as the velocity of water in the Pipe line B'C is reduced to zero. ‘This slow surge phenomena is discussed in Chap- ter XVII. Prior to the gate movement no water is flowing into or out of the surge tank, Hence, B, Bi, By’, and Care located at h = 1, v = 1, while By” and Dy are located ath = 1, v = 0. Moreover, because of the respective wave travel times Bis, Bis, and C,4 are also located at h = 1,0 = 1; and By’, Da are located at h= 1, v= 0. The complete graphical solution ‘of the problem is shown in Figure 69, and the detailed construction is shown in Figures 68¢, d, and. The solution utilizes the conjugate waterhammer equations for the three pipe sections, the head ratios and the con tinuity of flow at the junction and the manner of gate closure. The solution consists of three h-v coordinate systems, one for each of the components of the pipe line with h = 1 on the same horizontal line, The grid system (e) on the right-hand side of Figure 68 applies for the pipe line B’E adjacent to the control ‘gate; and contains parabolas corresponding to the discharge through the gate. The grid system (d) at the center of the diagram applies for the pipe line B”C and the grid system (c) at the left-hand side of the diagram applies for the surge tank or pipe section B’’D, The detailed solution is obtained as follows: Line (1) having a slope of 26"” = 0.5 is set off from Dax on grid (e). Similarly, the line (2) with slope —2” = —1 is set off from Bi, on grid (d) and the line (3) with slope —2p’ = —4 is set off from Bos on grid (e). Lines (4, 5, and 6) are construction lines, Line (4) is parallel to line (1) and starts from A = 1,v = 0on grid (d) until it inter- $60] WATERHAMMER ANALYSIS FOR COMPOUND PIPES 111, sects with the line (2) at the.point M. Line (5) is the horizontal line passing through Mf until it intersects the vertical axis of grid (@) at point NV. Line (6) connects point N and By on grid (c). Since no pressure wave returns to the gate until 1 second has elapsed, line (3) is extended until it intersects the parabola at £),» corresponding to the gate opening of 1 second. The location of Eys is determined in a similar manner. Line (7) which originates at Ey, and has a slope of +25" is then drawn. The intersection of lines (6) and (7) locates Bi. Since hey = her = hen a horizontal line (8) is drawn through Bj. The intersection of (8) and (2) determines B;’ and that of (1) and (8) determines B{”. ‘This construction satisfies the condition of the continuity of flow at the junction as defined by Equation (55). From Bi” line (9) with slope —25’” = 0.5 is drawn to intersect the line h = 1 of grid (c), thus establishing D,.x. ‘The value A corresponding to the horizontal distance from Dy.» to Di. is next plotted on grid (d) along the line h = 1 and line (4’) is drawn from the point just estab- lished parallel to line (4). A similar procedure is followed for the remainder of the graphical solution as shown in Figure 69. 60. Pipe line with a dead-end branch connection Consider the pipe line installation shown in Figure 70a and the gate closure shown in Figure 70b. ‘The graphical waterhiam. mer analysis for this installation is shown in Figures 706, d, and e. This solution is similar to that shown in the preceding section with the exception that instead of the pressure ratio remaining, constant, the velocity ratio v = O remains constant at D, the dead end. 61. Compound pipe system with two control gates Consider the pipe line system shown in Figure 71a. Tn this system the control gates are assumed to start their closures simultaneously as shown in Figure 71. Two sots of parabolas are uscil to accommodate the variable flow at each control gate. ‘The starting point in the graphical solution is located from the initial flow ratios in the branch pipes, and the solution is completed as shown in the figure.113 @ vo ant aut x Sy, @ Nouwi3y anu aunsi fone am eg e0 0 14902: apy s2106 w pany 240s nO08 02,0 poo! Su sd" ogee “suolas 266 no 04) O10 WATERHAMMER ANALYSIS FOR COMPOUND PIPES ee $61] so ($61 Contal gate? 1335 @ 19 | o oa WATERHAMMER ANALYSIS FOR COMPOUND PIPES 112 Figure 74114 WATERHAMMER ANALYSIS FOR COMPOUND PIPES [§6. 62. Summary Tn the graphical waterhammer analysis for compound pipes one /-v grid is assigned to each pipe branch in the system. By reserving the center grid for the main or header pipe, itis possible ‘to combine the characteristics of one of the branch pipes with the main pipe and thus enter the third grid in such a way that the analysis of all components is carried out simultaneously. ‘This type of graphical analysis was made possible by the arbitrary definition of @ and is applicable for compound pipe system con- sisting of any number of pipes. The method is applicable for ‘compound pipe systems used in pumping plants. It can also be readily adapted to accommodate the effect of the hydraulic losses or restricting orifices which may be present at any point in the system. Revs, BIBLIOTECA XV Approximate Methods of Graphical \Waterhammer Analysis 63. Equivalent uniform pipe approximation’ ‘When there are numerous changes in the cross-sectional area or wave velocity along the pipe line, the number of sloping lines in the graphical solution becomes so numerous and so elase to each other that graphical waterhammer solution by the methods described above is impractical. For example, such a condition occurs in the waterhammer solution for a pipe line with numerous changes in the thickness along the length of the pipe as the profile of the line changes. In order to obtain a satisfactory waterham- ‘mer solution for pipe lines of this type with a minimum of effort, 2 useful approximation can be employed which reduces the actual pipe line to an approximately equivalent uniform pipe line. When the conjugate waterhammer cquations are utilized for the end points of each section of uniform pipe, the wave travel time for each section of pipe is used. In order to account for the speed of the waterhamimer waves in the actual pipe, the wave travel time of the equivalent uniforin pipe is taken as the sum of the wave travel times for the various pipe sections in the actual pipe line, that is, £i(4,4,h LV on ( 57) o-(2+3+2+..48)-2(2) ‘The changes in the cross-sectional area along the pipe are partially accounted for by using the relation given by Equation (6) in Chapter I for the equivalent uniform pipe based on the rigid water column theory';-that is, + See Reference 36, 115116 APPROXIMATE METHODS OF WATERHAMMER ANALYSIS [§64 Lo(uphyh De 2 (2s), a7 Gre ea 2) =(%) 8) It is noted that this approximation does not include the effect of ‘the partial wave reflections which occur at changes in the cross- sectional area or wave velocity. An approximate value for the pipe line constant for the equivalent uniform pipe line is then pe = OMA), (0) 2gHod (Ln/ an) To illustrate the use of an equivalent uniform pipe line approximation, consider the pipe line and gate closure shown in Figures 32a and b. The value of p, for the equivalent uniform pipe line which is used to replace the actual pipe line is ‘1800 , 1500’ ne 7008(735 +7300) _ ang (2)(32.2)(500)(1) “— This equivalent uniform pipe line is shown in Figure 72a for which the graphical waterhammer solution is shown in Figure 720. A comparison of the results using the equivalent pipe method with ‘the more accurate method shown in Figure 32 shows that the maximum head rises at the gate and mid-length of the pipe as computed by the equivalent pipe method are slightly higher. This would be expected since the approximation neglects the effect of the wave reflections at the change in section along the pipe line. In general, this approximation gives satisfactory results when the physical discontinuities are minor in nature, such as changes in the pipe shell thickness and small changes in the cross- sectional area. The method is of course applicable to pump discharge lines as well as pipe lines involving gate operation. 64, Pipe line with surge tank Consider the pipe lino with a surge tank shown in Figuro 73a and the gato closure shown in Figure 73b. This is the same installation as that shown in Figure 68. When a pressure wave from the control gate reaches the junction at B, the wave reflection factor and magnitude of the reflected wave which returns to the gate can be computed from Equation (27b). ‘The wave re- $64] APPROXIMATE METHODS OF WATERHAMMER ANALYSIS 117 Hoss00ft Fas re . ol “ I t GATE CLOSURE TIME RELATION118 APPROXIMATE METHODS OF WATERHAMMER ANALYSIS. [564 10H 367th ger sec (2030 Fo" £3000 per sec Spas Ocenia Tige_ scons GATE CLOSURE “TIME RELATION (o) Figure 73a, b flection factor is 0.88 for pipe line B’E when all three pipes at the junction are considered. If pipe line BC is entirely neglected at the junction, the wave reflection factor for pipe line B’E is 0.85. Thus a satisfactory approximation of tho waterhammer effects in pipe line B'H may be obtained by neglecting the presence of the pipe line B”’C and by considering the pipe line B’E to he directly connected to the surge tank. ‘The watorhammer solution for such an equivalent pipe line is shown in Figure 73e. A comparison between this approximate solution with the more accurate solution shown in Figure 68 indicates the approximation to be a very satisfactory one. aS $64] APPROXIMATE METHODS OF WATERHAMMER ANALYSIS 119 an Tor TF be Figure 73¢Discharge Characteristics of Gates and Valves ——————— 65. Control gates In the solution of waterhammer problems for pipe lines in= volving gate operation, it was assumed that the variation in the effective area of the gate as a function of time was known from other considerations, In order to determine this gate factor, the flow characteristics of various types of gates and valves in common use will now be considered. At the present time available test data on the flow through partially open control gates are very limited. In order to determine these flow characteristics, the open area of the gate normal to the axis of the pipe was first compute ‘and the effective atea of the gato was then determined by mul plying the open area by a suitable coefficient of discharge for the articular gate opening. Figure 74 shows the flow characteristics obtained in this manner for a typical butterfly valve, dise gate valve, ring-follower gate, and a plug valve. 66. Turbine wicket gates The discharge through the turbine wicket gates as determined from actual tests in the field are shown in Figuro 75 for a 10,500- and 150,000-horsepower unit. When the rate of closure of any of these control gates or valves is known; the variation in the effective area of the gate as a function of time can be determined. ‘This permits the completion of the waterhammor solution by the graphical methods described above. 120 $66] DISCHARGE CHARACTERISTICS OF GATES AND VALVES 121 Y Tae WE care levoseos PERCENT OF FUL GATE EFFECTIVE AME 4 scat fsa . fen acan orn on aaron RUE pen area avourrenruy vauve Open Poenore, ps | SL Rae Oo Crem C+RING FOLLOWER CATE O-SPHERE On PLUG vAL¥E Figure 74122 DISCHARGE CHARACTERISTICS OF GATES AND VALVES [$66 PERCENT OF FULL cate Errecrive arta + 83t0n8 he | F eens of c 5 se a 5 aS Figure 75 xVil Surge Tanks 67. Operation A surge tank is often used at a power or pumping plant to control the pressure changes resulting from rapid changes in the flow. For example, when the turbine gates are closed at a power plant which is supplied by a long penstock, the water surface in the surge tank rises slowly above the original running level as the kinetic energy of the rejected flow is converted into potential cnergy. Such a conversion of energy reduces the rate of change of flow and the waterhammer in the penstock between the forebay and surge tank. Similarly, upon an opening movement of the turbine gates, energy is provided by the surge tank for the im- medinte demand of the turbine. This action reduces the waterhammer effects in the long penstock and assists the turbine to pick up its increased load more rapidly. Ato pumping plant with a long discharge line, a surge tank can also be usod.to effectively control the pressure changes in the discharge line resulting from the shutdown or starting up of a pump. For example, following the sudden shutdown of a pump, the surge tank provides cnergy to reduce the rate of change of fiow and the waterhammer in the discharge line. Upon starting a pump, most of tho initial flow from the pump enters the surge tank and this action reduces the waterhammer effects in the long discharge line. In order to accomplish its mission most effectively, the surge tank dimensions and location are based on the following considerations: (@) Ata power plant where the turbine output is controlled by a governor, tho surge tank must have sufficient cross-sectional area to prevent unstable action. In the event the area of the tank is too small, a load change on the turbine will cause continuous 123134 SURGE TANKS (668 oscillations of the water level in tho surge tank, possibly with in- creasing amplitude. This problem of surge’ tank instability is outside the scope of this treatment.' In addition, the cross. sectional area of a surge tank at a power plant should be large enough that the magnitude of the surges will be small during normal load changes on the turbine. Otherwise, turbiné speed regulation will be difficult or impossible, (b) Tho surge tank shuld be located as close to the power or pumping plant as possible, {c) The surge tank should be of sufficient height to prevent overfiow for all conditions of operation unless an overfiow spill. way is provided, (d) The bottom of the surge tank should be low enough that during its operation the tank will not drain and admit air into the turbine penstock or pump discharge line, On high-head plants, where from other considerations it is necessary to place the surge tank at a considerable distance from the power or pumping plant, the farther the surge tank is away 68. Analysis neglecting hydraulic losses Consider the simple surge tank installation shown in Figure 76 where tho initial flow through the control gate is cut off rap. idly. Ttis desired to find the maximum upsurgo in the surge tank and the time at which this upsurge occurs. In order to present Figure 76 "See Reference 44, see $69) SURGE TANKS 195 the phenomena in its most elementary form, the hydraulic losses and the velocity head in the pipe line are initially neglected. ‘Moreover, the rigid water column theory of waterhammer is ut lized since the effect on the upsurge of the stretehing of the pipe wails and the compressibility of the water due to an increase in pressure is negligible, Prior to the gate closure, the mass of water which is moving in the penstock is LAw/g. ‘Upon gate closure the unbalanced force acting on this water column is wAS. From Newton's see- ond law of motion the deceleration of the water column in the penstock is -% 208, (Go) ad L From the condition of continuity of flow following complete gate closure, the flow of water into the surge tank is the same as that out of the penstock, that is, = 61 PS - ay. (61) ‘The simultaneous solution of Equations (60) and (61) is performed with the following boundary conditions: When ¢ = 0, S = and dS/dt = Qo/F. FE in fu 6 =& ‘Then s-% Qe fe, (63 Sane ve (63) and the time required to reach the maximum upsurge is JZ A For the installation shown in Figure 76 the maximum upsurge in the surge tank above the static level due to the gate closure is computed to be 51.8 fcet and the time required to reach this upsurge is 30.3 seconds. from which (Gs) 69. Analysis including hydraulic losses and throttling Consider the surge tank system shown in Figure 77 where the positive directions of fiow and surge are designated. The196 SURGE TANKS [bo inagnitude of the surge in’ the tank with the friction effects ine cluded will now be determined. The head tending to accelerate ra water in the pipe line in the direction of the positive velocity sis Hy = -8 & ois Vi, (65) ‘where the signs of the last two terms depend on the direction of Vrand Vs. In this equation ais a constant such that eM = Hy, Stic level ‘Hydraulic gradient —; Initial eating lve eDeott Panstock rotting device head loss = Hq Inti Figure 77 and represents the sum of the entrance oss, pipe line friction loss, and velocity head in the pipe line. ‘Then a is a constant such that «Vi » H,, and represents the throttling loss for the flow into or out of the surge tank, The following tabulation gives H. for four possible cases: (4) Case 1 (upsurge caused by turbine shutdown), V: > 0, V>0, Ha = ~S ~ V3 - av}. (65a) (6) Case 2 (downsurge caused by starting up turbine), U>0, <0. He = -S —aVi+aVi. (65B) 7, Lo) “#808 (downsurge caused by pump shutdown), V, < 0, 1 <0. He = -S + Vi + 0.V2 (650) y, AP Case 4 (upsurge caused by starting pump), V, <0, 1 >0. He = ~S+0Vt~ eV}. (65D) a) $69) SURGE TANKS 127 ‘The mass of fluid inthe pipe line being accelerated is wAL/o and its acceleration at any time is dV;/di. Then from Newton's second law of motion ag 66) a La : For continuity of flow vate ease a3 _ AVi-Q (68) a a F By substituting Equation (65) into (66) and using (67) and (58) to eliminate V; and Vs, a differential equation is obtained in and t, By suitable changes in variable this equation reduces to the following form: * op (a +2 (a) #3, 148, ‘at *3( In this equation 2 [Fo _ Hy [Fo ba ale Vaz” *@ VE/a Now S; is a function of S, and (1 is 8 function of t. For example, Case 1 for turbine shutdown reduces to the following differential ‘equation: eS. Fy AY/(4S* 9, Qo aS #8 4 Fo («a ay NE +205 Ag » +H (6 tah ‘The substitutions +8 = (69) (70) ALS, 2Fgler + eo(A7/Ai)] and t= vie. reduce Equation (70) to one form of Equation (69). ‘The soln- tions* of Equation (69) for the four special eases of turbine and See Reference 19.SURGE TANKS. 129 $69) ‘SURGE TANKS [$69 128 20 Mons 40 ONILUvIS SnoaNYANYISN! 04 30 xNeL 3DuNs AL JOURS WHMixER, a to 90 go 0 zm 40 00199 coo_ 09 a Tae aren EE uve aouns £0 Mord 40 DNIddO1s SnoaNVANYE: 01 30 XNYL 39uNS NI DUNS wow HH worn oxmvval is woisaa oaomvwe i Figure 79 Figure 78130 SURGE TANKS {870 Pump operation are shown in Figures 78 and 79, In these figures cither the initial or final flow through the pump or turbine is zero. ‘These figures are plotted in terms of the parameter by or b with the ratios Sa Ss Hy He of and Fh Figure 78 is used to determine the surge resulting from an instantancous stopping of the complete flow al either a turbine or pump installation, while Figure 79 is used to determine the surge caused by the instantaneous starting of the flow at either a turbine or pump installation, The broken curve is used for balanced design, where the maximum surge Sp or S4 equals the throttling loss H, 70. Example To illustrate the use of the charts consider the surge tank installation shown in Figure 77. At this installation a flow of 843, cubic feet per second is shut off rapidly. For the given throttling ratio of Hp/H, = 1.49 and by = 0.259, the upsuige S, above the running level as determined from the charts is (3.5)(13.4) = 46.9 feet. ‘The upsurge S above the static level as shown in Figure 77 is then 46.9 — 13.4 = 33.5 feot. Surge values can also be determined in a similar manner from the charts due to the starting of turbines at a power plant or for stopping or starting the pumps at a pumping plant, xVIII Air Chambers at Pumping Plants 11. Oper If 8 power failure occurs at the motors of the pumps which supply water to a long discharge line, the initial negative surge wave may cause water column separation to occur at the high points in the discharge line which are near the hydraulic gradient. ion -€L1985 | MSELISTO £12004 Normal hiro gradient auisz0- j A s1965 aft Wo" 309 per se. 2° 2960 ft pr sec. Figure 80 Whenever possible, this condition should be avoided because of the high pressures ereated when the water colurnns rejoin. .\ device which can effectively control the pressure surges in a lons pump discharge line is an air chamber as shown in Figure 80. ‘The lower portion of the chamber contains water, while the upper portion contains compressed air. When power failure occurs, the head developed by the pump and the flow decrease rapidly. ‘The ‘compressed air in the chamber then forces water out of the hot- tom of the chamber into the discharge line and minimizes the 131_— em “ 132 ‘AIR CHAMBERS AT PUMPING PLANTS (71 velocity changes and waterhammer effects in the discharge line. As the pump speed is reduced to the point where it cannot deliver water against the existing head, the check valve at the discharge side of the pump closes, and the pump comes slowly to a stop. ‘The water in the discharge line then comes to rest, reverses, and flows back into the chamber. As the reverse flow enters the cham- or, the air volume decrenses, and a head rise above the normal pumping head occurs in the discharge line, gure 81 Air chambers at pumping plants can usually be made smaller Gan open surge tanks which are designed for the same purpose. Moreover, they do not have to be placed in an upright position. However, in order to be effective, chambers must be kept supplied with small amounts of compressed air to replace that which dis- solvesin the water. In order for a chamber to he most effective, it is necessary to throttle the reverse flow of water from the dis charge line into the chamber, while very little throttling is provided for the flow out of the chumber. One device for necomplish- inng this is a differential oritice of the type shown in Figure 81.! When tested in the laboratory this particular orifice was found to xive2.5 timesas much head loss for return flow into the air chain- ber as for flow out of the chamber. * See Iteference 9, 4 $72) ‘AIR CHAMBERS AT PUMPING PLANTS: 133 72. Basic considerations Following a power interruption at the pump motor, the pressure head at the discharge side of the pump is initially maintained by the air chamber, while the pump speed and discharge fall off rapidly. ‘This causes the check valve to close very soon after power failure. Hence it can be assumed that the check valve closes simultaneously with power failure, and that thereafter all the flow from the discharge line is from or into the chamber. ‘This, assumption eliminates the pump characteristics from the waterhammer computations but introduces an abrupt pressure wave corresponding to the instantaneous head drop across the throttling, orifice at the base of the chamber. During the transient conditions the air in the chamber initially expands as the head drops in the discharge line and later compresses as the head in the discharge line increases. In the following analysis the pressure-volume changes for the air in the chamber are taken midway between adiabatic and isothermal ex- pansion, that is HC! = H3C}# = constant. a) This is written in the following form: Lest (72) where h* = H*/Hs and ¢ = C/Cs. The volume of air in the chamber at any time f, in terms of the volume at a previous time tis C= C,-A f “yan. (73) For small time intervals the variation in V may be assumed to bw: linear, and the volume of air in the chamber is thea n=O, — Aa Vat ws) where the time interval At = 4 ~ 4. In order to use Equation (7A) with the graphical waterhammer solution, it is necessary to choose At as a fraction of the wave travel time in the discharge line, that is, At = 8(/a). ‘Then,134 ‘AIR CHAMBERS AT PUMPING PLANTS (673 ‘Atmosphere. pressure need ir chamber Go 100 eutt Pune Check vale Throtting = * $01 for O20 eu ft per see 520.0 cu ft per see Vo 5.00 tt per sec = 3220 ft per sac 2p 2200 2650 oe Bat rm 0.0 ee-gengnee | os [a a a Ye era eg Figure 82a, banne eS $$$ s cn 9. 136 ‘AIR CHAMBERS AT PUMPING PLANTS (73 Gy _ AW, + V,)QL8, Ci” Ge ~ B0,Meda ag — ws, a fe BC a 7OML’ Re) Since the air volume changes in the chamber are defined in terms of absolute pressures, it is preferable to define the pipe line characteristic in terms of absolute pressure heads, that is, (76) ‘The head changes in the discharge line are then defined by the conjugate waterhammer Equations (51A) and (51B) where p is replaced by p*. 20H 73. Graphical solution Consider the pumping system with an air chamber shown in Figure 82a. From the data given, 2p* = 2.00 and 2C,a/Q.L = 10.0. For 8 = }, Equation (75) reduces to iy = Cy = 0.050(0% ++ tn). (77) The pressure-volume relationship in the chamber as defined by Equation (72) is plotted in Figure 82b.. Referring now to Figure 82e, the starting point in the graphical waterhammer solution corresponding to as is located at the coordinates h* = 1, v = 1. ‘The assumption of instantaneous check valve closure causes an abrupt change of flow in the discharge line from Ay to Ay. At the same instant a head drop deveiops across the throttling orifice as represented by the vertical line Ajay. ‘The head changes at A from time zero to 2L/a seconds are lorated on the line of slope 2p* which passes through As. ‘The head changes inside the chamber at A’ from time zero to 2/a seconds are located on the throttling line PN where the vertical distance hetween A and A’ for agiven flow corrospondato the throttling lors, Now Ais located at the intersection of PN with the horizontal line A* = 1, and Ag is located directly below. The location of Axis. on the line of slope 2p* is then determined in the following manner: The approximate location is first assumed and value of vzjm is read from Figure 82c, Then cus is computed from Equation (77) and $73) |AIR CHAMBERS AT PUMPING PLANTS 107 as of os} a 4 | ep os a ) * 2699 Oat Nombers on curves ore SURGES IN PUMP DISCHARGE LINE, K + 0 Figure 83 ‘ims is read from Figure 82b. The assumed value of rise andl the computed value of M, define the loeation of sim which ‘must fall on line PN. If it does not, the trial process is repeate until this condition is met. ‘This computation is then repeated for times Z/a, 8L/2a, and 2L/a seconds. At time 2L/a seconds, the abrupt pressure wave which was assumed to have left the chamber at time zero returns from the discharge end of the line. ‘This return wave causes an instantaneous change in velority from tore to Abi as shown in Figure S2c. The complete solution of the air chamber problem is then completed as shown in the figure.eS 138 ‘AIR CHAMBERS AT PUMPING PLANTS. ($74 a ay a Zl ad SH? IB i 03} ad 7 7 | aa iv zo ures are wales of 26% a eatncet pump. a stig ‘SURGES IN PUMP DISCHARGE LINE, K #0.3, Figure 836 74, Airchamber charts From a study of the graphical waterhammer solution given in the preceding section it is seen that the basic parameters in. volved in determining the waterhammer effects in a pump dis. charge line with an air-chamber are 2o* and 2C,a/QsL. Hence the results of a large number of air chamber solutions can be shown on a chart as shown in Figure 83. In this figure the surges in the discharge line adjacent to the chamber and at the mid. * See References 2 and 16. $74] ‘AIR CHAMBERS AT PUMPING PLANTS 139 MAX pownsunce Tas ore a a oY 2C00 Tat Numbers on curves are values of 2p ‘slecert 10 pun. a atieitengt SURGES IN PUMP DISCHARGE LINE, K #0.8 Figure 83¢ length are plotted in terms of the basic parameters. The values shown in this figure are based on the following assumptions: (a) The air chamber is located near the pump. (b) The check vaive at the pump closes immediately upon Power failure. (©) The pressure volume relation for the compressed air in the air chamber is H1*C' = a constant. (d) The ratio of the total head loss for the same flow into and out of the air chamber is 2.5 to 1; KI/3 is the sum of the hydraulic losses in the discharge line and the throttling losses at the differ-a 140 ‘AIR CHAMBERS AT PUMPING PLANTS (574 os ¥ - 7% SURGES IN PUMP DISCHARGE LINE, K =0.7 Figure 83d ential orifice when a reverse flow equal to Qs is passing into the air chamber. To insure that air will not enter the discharge line when the maximum downsurge is attained, the total volume of the air chamber must be greater than C’, where C” is defined as follows: 1 og (ty © = (He) 3) where Hie = Hs minus maximum downsurge adjacent to pump. $75) ‘AIR CHAMBERS AT PUMPING PLANTS 141 75. Example Consider the pumping plant installation with an sir chamber shown in Figure 80. It is desired to determine a chamber size such that the maximum upsurge in the discharge line adjacent to the pump will not exceed 0.43H and the maximum downsurge at the mid-length will not exceed 0.2143. From the charts in Figure 83 it is found that these requirements are met by using the values K = 0.3 and 2Ca/QoL = 21 for which the maximum upsurge at the pump = 0.27H3, the maximum downsurge at the mid-length“= 0.2143, and the maximum downsurge adjacent to the pump = 0.32H. For this installation the pipe line friction loss for a fow Qyis about 3 feet. The differential orifice required at the chamber must then give a head loss for a flow of Q, into the chamber of 0.3 X 234 — 3 = G7 feet. With 2C,a/QuL known, the initial volume of compressed air in the chamber C, is 709 cubic feet and the minimum volume for the whole air chamber C’ as determined from Equation (78) is 1040 eubic feet. 76. Water level controls In the previous example, the calculations for downsurge and upsurge were based on the same initial volume of air in the chamber. In an actual installation some fluctuation in the water surface level in the chamber will occur since air is constantly going into solution and the air compressor cannot maintain a fixed water level. These changes in water level in the chamber can he used to actuate the controls in the pumping plant. Two sets of control levels are usually required, namely, the upper and lower emergency levels and the compressor ‘on’ and ‘off’ levels as shown in Figure 84. When the water level in the chamber reaches either of the emergency levels, the pumping plant is automatically shut down, ‘The volume of the air in the chamber is determined from the consideration that there must be adequate air in the chamber above the upper emergency level to control the surges in the dis charge line to desirable limits, and there must also be enough water in the chamber below the lower emergency level to prevent air from entering the discharge line. In order to obtain the total142 cr [AIR. CHAMBERS AT PUMPING PLANTS (876 Lower emergency ‘eve! AIR CHAMBER CONTROL LEVELS Figure 84 hese conditions, Co is computed fore and is then increased by an ‘mount equal to the volume of the chamber between the upper nd lower emergency levels. Anew value of His at the Oo ts ‘the chamber is then obtained from the charts, ‘The to oa tf the chamber C’ is then determined from Equation (78), ‘the new values of Hii. and Co i for t volume of the air chamber from the charts in Figure 83 as be References ( 1. Ausanmy, L,, Theory of Waterhammer, translated by E, E. Halmos, printed by Riceardo Garoni, Rome, Italy, 1925. 242, Atasemy, L., “Air Chambers for Discharge Pipes,” Transactions ASME, Vol. 59, Paper Hiyd-59-7, November 1937, pp. 651-659. 3, Anave, R. W., “Simple Graphical Solution for Pressure Rise in Pipes and Pump Discharge Lines,” The Journal of the Engineering Institue of Canada, February, 1935, pp. 72-81, Including Those Supplied by Cone ‘Bulletin 152, University of Toronto 5, Awous, R. W. “Waterhammer Pressures in Compound and Branched Pipes,” Proceedings ASCE, No. 202, January 1938, pp. 340-401, 38. Awavs, R. W., “Air Chambers and Valves in Relation to Waterhammer,"” Transactions ASME, Vol. 59, Paper Hyd-59-8, November, 1937, pp. 661- 68, 7. Benatnow, L., “Etude des variations de regime dans les conduites «l'eau: Solution graphique generale,” Renue gnerale de Vhydrautique, Paris, Vo. 1, 1036, pp. 12-60. 8, Bexornon, L., Du coup de belier en hydraulique au coup de foudre on elec- ‘tricie, Dunod, Paris, 1950, Discussion of L. Allici’s paper, “Air Chambers for Dis- Transactions ASME, Vol. G, July, 1999, pp. 441-445, 10, Bruaiwos, A. W. IC, Dooxix, 0. HL, Kware, F., and Sastos, A., Je, “‘High-Head Penstock Design,” Symposium on Walerhammer, ASHE- ASCE, 1938, pp. 20-61, 11, Catan, Joy and Gaver, D., Theorie des chambres dequilire, Cauthur- Villars, Paris, 1926, 12, Crawron, C.C, “Curves for the Quick Estimation of Tra Conditions Following Failure of Bleetrie Power to a Pura ss for M.S. Degree, University of Coloralo, IDS, 18. Dawson, F.M., and KAUNSKE, A. A., “Methous of Calculating hammer Pressures," Journal of the American Water Works Association, Vol 431, No. 11, November, 1989, pp. 1535-1864 14, Duran, W. F, Hydraulics of Pipe Lines, D. Van Nostrand Company, Ine., New York, 1921. fater- 143CO 144 REFERENCES 15, Evans, W. I. and Crawrono, C. C., “Design Charts for Air Chambers on Pump Lines,” Transactions ASCE, Vol. 119, No. 2710, 1954, pp. 1025-1046. 16. Ginsow, N. R., “Presaures in Penstocks Caused by the Gradual Closing of ‘Turbine Gates," Transactions ASCE, Vol, $3, 1919-1920, pp. 707-175. 17, Guoven, R. B,, “Computation of Waterhammor Pressures in Compound Pipes,” Symposium on Walerhammer, ASME—ASCE, 1983, pp. 64-60, 18, Hatatos, E. E, “Effects of Surge Tanks on the Magnitude of Waterammer in Pipe Lines,” Symposium on Waterhanmer, ASME—ASCE, 1983, pp. 72-80. 10. Jaconsox, R 8., “Charts for Analysis of Surge Tanks in Turbine or Pump estallations,” Special Report 104, Bureau of Reclamation, Denver, Cole. ado, February, 1952, 20, JazaeR, and Public Works Review, 1948, 21, Jazonn, C, “Theory of Resonanco in Pressure Conduits,” Transactions ASME, Vol. 61, February, 1989, pp, 109-15, 22, Jouwsox, R. D., “The Diferential Surge Tank,” Transactions ASCE, Vol. 78, 1915, pp. 760-805. 23, Jouxsox, R. Dy “The Surge Tank in Water Power Plants,” Transactions ASME, Vol. 30, No. 1204, 1908, pp. 448-501. 2%, Jovxowsxy, N., “Waterhammer,” Translated by Miss O. Simin, Proced= ‘ings AWWA, Vol. 24, 1008, pp, 341-424, 25. Kenn, 8. L., “Fallin Pressure in Iydraulie Turbine Peastocks Due to Ao- celeration of Flow," Power, Vol. 60, No. 7, August, 1925, pn. 206-268, 25. Kenn, $I “New Aspects of Maximum Pressure Rise in Closed Conduit,” Transactions ASMB, Vol. 51 (1), Paper Iyd-31-B, 1929, pp. 13-30. 21. Kenn, S, L., and Smowcen, B. B., "Resumé of Theory of Waterhammer in Simple Conduits,” Symposium on Waterhammer, aSME—ASCE, 1983, pp. 15-24, 28, Kessuen, L.H. “Speed of Waterhammer Pressure Waves in Traasite Pipe,” Transactions ASME, Vol. 61, January, 1999, pp. 11-15, 20, Kare, RT, "Complete Charncterstios of Centrifugal Pumps and Their se in the Prodietion of Transient Behavior,” Transactions ASALE, Vol. £59, Paper Hyd-50-11, November, 1937, yp. 683-640, 30. Knvse, 0. ¥., Discussion of N R. Gibson's paper, “Pressures in Penstocks Caused by the Gradual Closing of Turbine Gates,” Transactions ASCE, Vol. 3, 1019-1920, pp. 741-747, 31, Le Conte, J. N., "Experiments and Caleulutions on the Resurge Phase of Wotechammer,” Transactions ASME, Vol. 50, Paper Ilyd-50-12, Novers ber, 1097, pp. 691-694, REFERENCES 145 32. Moopr, L. F., "Simplified Derivation of Waterhammer Formula,” Sympo- sium on Waterhammer, ASME—ASCE, 1933, pp. 25-28. 83, Panuanian, J., “Pressure Surges at Large Pump Installations,” Trans- octions ASME, Vol. 75, August, 1953, pp. 995-1006, A, Pamuartan, J., “Pressure Surge Control at Tracy Pumping Plant,” Proceadings ASCE, Vol. 79, Separate No. 361, December, 1953, 35. Peanovr, R. M,, “Typical Analysis of Waterhammer in a Pumping Plant ‘of the Colorado River Aqueduot,” Transactions ASME, Vol. 61, February, 1939, pp. 117-124, 36, Qutex, R. 8., "Comparison and Limitations of Various Waterhammer Theories,” Transactions ASME, Vol. 49, No. 6a, May, 1927, pp. 524-530. 37. Rien, G. R., Hydraulic Transients, Ist ed, Engineering Societies Mono- sraphs, McGraw-Hill Book Co, Inc., New York, 1951. (Dover Reprint) 38, Scmerpen, O., “Comparisons Between Calculated and Test Results on Waterhammer in Pumping Plant,” Transactions ASHE, Vol. 59, Paper Hyd-59-13, November, 1937, pp. 695-700, 99. Sreranorr, A. J., Centrifugal and Azial Flow Pumps, John Wiley & Sons, Tne, New York, 1948, 40. Smoworn, E. B., “Relation of Relief-Valve and Turbine Characteristics im the Determination of Waterhammer,” Transactions ASME, Vol. 52, Paper Hyd-59-14, November, 1937, pp. 701-705. 41, Stnowars, E. B,,“Waterbammer Problems ia Connection withthe Design of Hydroelectric Plants,” Transactions ASME, Vol. 67, July, 1945, pp. 377- 392, 42, Smowarn, E. B., and Kens, S. L., “Speed Changes of Hydraulic Turbines for Sudden Changes of Load,” Transactions ASME, Vol.48, No. 2009, 1920, pp. 200-202, 43. Symposium on Waterhanmer, ASME—ASCE, 1933, 44, Tuows, D., Heitrage sur Theorie des Wasserachloescs bei selbliatig gereglien Turbinenanlagen, Oldenburg, Munchen, 1910, 45. Thwosmento, 8, Strength of Materials, 2d ed, Part 2, D. Van Nostrand Company, Ine., New York, 1941, 46, Wannen, M. M., “‘Penstock aad Surge Tank Problems,” Transactions ASCE, Vol. 79, 1915, pp. 238-208, 47. Wooo, F. MM. "The Application of Heavisides Operational Caleulus to the Solutions of Problems in Watorhammer,” Transactions ASME, Vol, 53, Paper Ilyd-59-15, November, 1987, pp. 707-713,Problems Chapter | 1 A pipe line has the following characteristics: L ~ 2600 fect, Hy = 160 feet, and D = 6 inches. (8) What i the maximum pressure at the control gate if the ‘low of 600 gallons per minute is cut off uniformly in 10 seconds? (b) What is the factor of safety ifthe bumting strength of the pipe is 375, pounds per square inch? Ans. (s) 92 pounds per aquare inch, (b) 4.1. steady P SCantrat ore Figure 85, 2, The fll gate discharge for the pie line shown in Figure 85 ia 300 cubic feet per second. ‘What is the maximum head at the control gate if one-half of the fall harge is eut off by a uniform gate closure in 4 seconds? (©) What isthe minimum head at the centerline ofthe pipe at B ifthe coutol gate is opened from 2aro gate to fll gate in § seconde? Ane. (a) 108 feet, (6) 15 fet. 3. (a) Show that the equivalent length of uniform pipe of diameter Ds which can be used to replace the tapered tunnel section show ia Figure 80 in the waterhammer computations it Z,Dy/Dy gate Figure 86 (b) What is the maximum head rise at the control gato if the initial low ‘of 250 eubie fet per second is cut off uniformly in 12 seconds? Ans. ()) GS feet. 147a 148 PROBLEMS: Control gate pi fiete n= 38th srt “Reducer sector Figure 87 4, Calculate the equivalent length of uniform pipe of 42-inch diameter for tha pipe line shown in Figure 87. Ana. 285 fect, Chepter Ill 5, What the watrbamer wave vl ina circular conduit with ig walls? Ans, 4720 feet per second. " ae 6. (0) Compute the wataranimer wave veloity in a tel pipe 42 ncaa in outside iametr ed eh thie the pie ne sacha at on end iy (b) What is the wave velocity if the water is replaced with erude oil whose sowie gravity i 030 and bulk modula a 220,00 pounds per quae ash? Ans, (a) 3300 feet per second, (b) 3400 feet per second. 3 inches thick, Assume that the pipe line is anchored at one end only. Ans. (a) 1720 feet per second, 8, Compute the waterhammer wave velocity in a 4-foot inside diameter rsinforced conerete pipe line which is anchored against longitudinal movement throughout its length. The pipe walls consist of eonerete G inches thick with J-inch diameter steel reinforcement bare spaced at 12-inch centers. Ans. 3190 {feet per second. 9. (a) Compute the waterhammer wave velocity for o stecl-lined circular pressure tunnel at which D = 10 feet, e = 1 inch, H = 30 X 10° pounds per square inch and G = 2 X 10" pounds per square inch. (b) If the steal liner is removed what isthe waterbammer wave velocity in the tunuel? Ans. (a) 4440 fest per second, (b) 4400 feet per second. Chapter IV. 410, At the installation shown in Figure 88 an instantaneous closing move ment of the regulating gato at A produces a positive pressure wave of 25 feet at the gate. Dotormine the magnitude of the preseute wave which (a) Reflected back from the junction C toward the gate in pipe section CA, (b) ‘Transmitted to pipe sections CB, CE, and CG. (6) Reflected back from tho reservoir at 2, (@) Rellected back from the free water surface at PROBLEMS 149 ws pe20ft 9°2800 ftper see. Doe Closed sdoooft per ste. oes AN ‘0v3s00 ft per see. Regulating gate 3 Figure 08 (@) Reflected baek from the closed gate at B. Ans, (a) —23.3 feet, (b) $17 feet, (¢) =117 foot, (2) ~L.7 fet, (e) +1.7 feet. 11, Assume that at a given instant a positive pressure wave of 10-foot ‘magnitude traveling in the direction of flow inside the pipe has just reached an ‘open control gate, Determine the magnitude of the rellected pressure wave leaving the gate if I = 0.25, y= 676 foot, Ve = 0.5 feet per second, anil 4a = 2950 feet per socond. Ans. +3.9 feet. Chapter V 12, A pipe line 3000 foot long has a waterhammer wave velocity of 2007 {eet per soeond and sn initial water velocity of 4.5 feet per second. (@) What is the head rise at the control gate if the entre flow is cut off in 0.6 secon? (b) What isthe farthest point along the pipe line away from the control sgate at which this maximuen head rise is attained? Ans. (a) 405 fet. (b) 300 feet 13, Compute the head tise for ench foot per second of velocity suients ‘extinguished fora pipe line carrying ol of specific gravity 0.75 and bul: modulis fof 200,000 pounds per square inch. ‘The pipe ine has an inside diameter of 12 inches, a wall thickness of { inch and is anchored at one end only. tn, 110 feet. 14. Construct n pressure wave time history diagram similar to Figure 12 in the text for an instantaneous partial gate elosure. Chapter VI 15. As.a result of a closing movement of the control gate at the pie li gate varies in the following manner shown in Figure 8 the head rise at Ui LGseeomld, 0 025 O50 075 1 125 1.50 C+D, OF LIT 2% HW Bz What isthe total ue at Band Cat ¢= Land ¢ = 1.75 seconds? Ams, 1 foot and 147 foot at 1, 10 foet wend LS feet at C.150 (02800 Foe se Figure 89 16, Compute the minimum head at the gate and mid-leagth of the pipe line shown in Figure 23 of the text for the following gate opening: tGeccad), 01 2 3 4 5 6 7 8B Cady (square feet), 0 0.50 1.0 1.55 315 360 4.20 4.50 4.70 ‘Use one-second time intervals in computations. Ans, 285 feet and 260 feet, Chapter Vil 17. A pipe line has the following charncterinticn: L = 1360 feet, a = 3600 feet per second, Vo = 3.25 feet per second and /1s = 180 feet. Compute the ‘maximum head rise at the control gate for the following gute closure: "(econds), 0 075 1.50 225 3.00 375 450 5.25 6 7» 100 0.875 0.925 0.85 0.80 0.85 050 0.35 0 ‘Ant, 110 feet, Chapter VIIL "18. A pipe line has the following characterstis: Z = 2400 feet, a = 3200 fect per second, He = 325 feet and Ve = 6.5 feet per second, (s) Compute tho maximum head rise at the control gate and one-third points in the pipe line for uniform gate closure to the half open gate position in 3 seconds, (b) Construct a pressure time history diagram for the pressure changes at the control gate from ¢= 0 to ¢= 6 seconds, Ans, (a) 85 feet, 60 feet, and 30 feet 19. Compute the maximum head at the control gate and the junction B {or the pipe ine shown in Figure 00 for the following gate closure: Figure 90 151 Voesttersec a=ssooft persec Figure 91 er 4350 0 05 1p 15 20 25 20 35 40 weenie [ao O27 O@0 O40 O38 O28 020 O15 010 00s 0 Ana, 655 feet and 405 fect. 20, Compute the maximum head rise at B” forthe pipe line showm in Figure 91 fora uniform gate clogure in § seconds. Ans. 75 feet. i ‘and mid-length for the pipe 21, Compute the maximum head at the gate and mi! so line shown in Figure 92 forthe following control gate and relief valve operatic [asaisaitparaeeT ssort gs 861 per sec Figure 92 HGeconds), 0-15-80 60 + (ontrl gate), 100 058 0.30 ° (elit valve), 0-014 040 os Ans, 770 feet nnd 705 feet. 22, Compute the minimum head at B' for the pipe ine shown in Figure 93 for the following gate closure: Woe 2ett per see Figure 93a 152 PROBLEMS. fi HQeemd), 0 45 30 45 00 og * 100 090 070 050 030 6 Ans, 235 feet, Chapter IX Verattpersee, Figure 94 Compute the minimum head at the pointe A, B, and C for the pipe inte A, B, F the pi Tao in Figure 4 fora uniform gate opening in 8feaconds. Ans. OD ese Chapter X seconds Ant) 20 et () 35 ft, (210 foet and 8a Ok Chapter XI typeof eh instalation has the following characteristics; Number and Pe of pumps, two single suction sumps with sharacteristie sh Fi areal el ps with characteristics shown in Figure 5 PROBLEMS uz) WR? of each pump and motor, 325,000 pound fet, Pump speed, 327 revolutions per minute, Rated bead of pump, 186 feet, Discharge capacity ofeach pun at rated head, 205 cubic feet per second, Pump elicieney at rated head, 89.0 per cent, Tide diameter of discharge lin pipe, 11 feet, Length of discharge line, 4800 fest, Average wave velocity of discharge lie pipe, 4000 feet per second. Determine from a graphical watechammer analysis the maximum rte and drop i head at the pump and mid-length ofthe discharge in fora simultaneous power failute at both pump motors. Use time intervals of 0.8 second in the computations and neglect hydrauls losses. Ans. Head rise = 95 fert and 50 feet, Head drop = 150 feet and 90 feet. 27, Whatis the maximum rise and drop in heid atthe pump ond mid-length of the discharge line of Problem 28 hea a power failure occurs with only one pump in operation? Ane. Head rise = 60 feet and 30 feet. Head drop ~ 125 foot and 85 fet. Chapter XII 28, Re-olve Problom 26 assuming that there is a check valve on the dis: charge side of ench pump which closes immediately upon a flow reversal at the ump. Ans. Head rise = 150 feet and. 86 feet. Head drop = 150 feet and Si feat. 29, Resolve Problem 26 assuming that the 1VR* of the pump and motor in increased 100 per cent. Ans. Head rise = 60 feet and 30 feet. Head drop = 125 feet and 70 feet, 30, What is the maximum head rise in the discharge line of Problem 26 caused by the sudden starting of one pump or two pumps simultaneously with the discharge line full? Ans. 19 feet. 31, Resolve Problem 26, assuming that suction line of wave travel time Zn/ox = 06 second and having the same pipe line characteristic as the discharge line is added to the installation. Ans. Head rise = 80 feot and 45 feet Head drop = 125 foet and 70 feet. ‘32. Solve Problems 26 and 27 by using the waterhammer charts in Figure 57 of the text. Ans. For 2-unit shutdown: Head rise = 60 feet and 30 feet, ead drop = 140 fect and 85 fect. For I-unit shutdown: Head rise = 40 fet ‘and 25 foet, Head drop = 115 feet and 70 feet. Chapter XIll 33, A pipo line has the following characteristics: L = 30,000 feet, 1) = 2.5 feet, Ve = 8.37 fect per second, a = 3000 feet per second, static heal = GOS feet, and hydraulic losses = 108 feet. Compute the maximum head at the154 PROBLEMS ‘contol fate for @ uniform gate closure in one minute for th following case: (s) Hydraulic losses concentrated at the intak (©) Hydrauli loses concentrated at the midlength, (c) Hydraulic losses concentrated at the control gate, 1 ptinting eect of hydraulic lass, An. (a) 715 fet, (0) 729 fee, (©) 720 feet, (d) 740 feet, i asolve Problem 26 for an assumed pipeline head lots of 10 fet with the hydraulic lostes concentrated atthe dichargn line cutot cod, ances roe = 70 feet and 30 feet, ‘Head drop = 150 feet and 100 fest, Chapter XIV 35: Compute the maximum head at the control gate of the installation shown ia Figure 05 fora uniform gate clonur in 3 eeconde with va tater 0f 1200 eubio feet per second. Ans, 630 feet ws. 20° dia. surge tonk~ Figure 95 36. For the pipe line shown in Figure 71 of is initially closed. ‘Compute tho maximum head 1 initial Now of 150 cubic feet per gection is cut off u Ans, 105 feet, 80 feet, 75 fet. Chapter XV jay, elv® Problem 35 by the approvimate surge tank mothod, Ans 620 feet. canine pfobem 19 forthe maximum head a the conto gate by using ‘the equivalent uniform pipe approximation. Ane, 676 feet, Chapter XVI 20. Construct 8 gate cloure time curve for tho butterfly valve shown ia Figure 74 of the text where tho initial two-thirds of the angen closing move. PROBLEMS 155 ‘ment is performed in 10 seconds and the remaining one-third movement is ac- ‘complished in 20 seconds. {0,1 te extn of Figure 2 inthe tat ie placed ith «battery vale biplane ah ona pata na ‘sxta position in 12 seconds at a uniform angular rate? Ana. 165 fet Chapter XVIL 6) Compute te masinu water us eleva inthe srt ioe va Fig 8 ose fl oon 00 ea 16 de. suge tank overt Thoting ing 2o0ft Riser vine F005 Figure 96 — i riser pipe loss ie dat the turbine. ‘The surge-tank throttling ring and riser pip. {Oat fors tow o 0 cats fet per ecnd into ue he age tank () Compute the effect on the upsurge of removing the throttling effect at ‘the aurge tank. Ans. (a) 227 fet, (b) 242 fect. i 1 installation shown 42, Compute the downsurge in the surge tank forthe ia Figure, asning tat a pupa ih lore edi dlvering WO ea ct per second when power failure occurs. Ans. 50 feet. GpeScu Me per see F009 (0+ 3000 ft pe se. Figure 97156 PROBLEMS Chapter XVIII 49, Compute the minimum volume of n ir chamber required at the pump- a ee isn nd 0 te a ing pn er ng ead and we oot a oy ate 10 ec re ale atthe pomp motor, Use K = 0 "Ans. 1000 eubie feat. A ¢ Air chambers: Calne, Jy and Gaten, D., 143 ‘basic considerations, 133, Gast iron pipe, wave velocity, 19 caractorietio 134 Geotrfugal pumps: charts, 138 characteristics, 78 example, 11 ‘pple performance curves, 77 ical solution, 138 WRerbamimer upon power fllue, 74, ‘operation, 131 8 Throting device, 182 role required, 140 16, 28, 65, 18 rater level control 141 fused at pump discharge Allies Ly 143 ines, 94, 131 ‘Angus, RAW, 143 cingular pipes and tunnel, wave veloc ‘Approxioate analysis 19 Mguivaleat usiform pipe, 115 Closure of control gate fips lie with aurge tank, 116 complete, 4, $8 Figid water column theory, 5 fnstantaneous, 90 ‘Arithmetic anne’, 39 parti, 4, 58 ‘Assumptions: flow, 37 ‘ir chamber, 139 ‘Compound pipe laste water column theory, 8 108 field water eoluenn theory, 2 8 Balanced design of surge tank, 130 toro control entes, 11 ‘Basie considerations: Compreesibiity of water. 8 Conerete pipe, wave velocity, 22 Condition of continuity: “dastie water eoluma tory, 10 {unetion of three pipes, 107 ‘is chamber, 135 fompound pipes, 106 antic water eolumo theory, 8 figid water column theory, 1 dow gate movernents, 37 ‘copition of dynaraic equilibrium, 8 urge tanks, 123 Cehugnte waterbammer equations, 45, Dergston, Ln 149 ‘3, 82, 107 oul equation, 1 Control gates: ert Billings, AWC, and otbers, 143 ow characteritit, 120 ‘Branehed pipet: ooated at upper end of pipe, OL “aaalais 108 reeted away from outlet end of Pipe, seni ymgection and transmission, 28 OF ‘Balk modulus of watar, 14 Control valves used ia pump di Bitter val Tinea, 97 iachargt, 120 fuse at pumping plants, 98158 INDEX D Damon, F.M,, and Kalinke, A.A, 143 Dead end: jn eompound pipe ayatems, 111 wave reletion a, 26 Derivation of waterhammer equations: ‘compound pipe aystem, 108 clastic water column theory, 8 rigid water column theory, I Differential orice used with airbase bere, 132 Dine gate valve, 120 Disebarge: cbaracteratcs of gates and valves, 120, through control gata, 3, 29, 48, OL Durand, W.F., 143 Downsurge ia surge tanks, 128 Dynamic equilibrium, condition of, 8 Blastie water column theory: basic coosiderstions, 8 condition of continuity, 10 condition of dynamie equilibrium, 8 fundamental waterhammer equations, 16 physical significance of waterhammer equations, 16 Blastcty of pipe sbell, 12 Equation ‘Beraoul ‘conjugate, 45,83, 82, 107 continuity, 10 dynamic equilibrium, 8 fundamental waterhammer, 16 inertia for pump and motor, 76 surge tank, 127 Equivalent length of pipe tin, 6, 115 uivalont wave travel time, 115 Brana, W.E,, and Crawlord, C.C., 144 F Fallin presure dus to: fete opening, 62 power failure at pump ‘motor, 74 ‘uniform gate opening, 8, 72 Flow: ‘in compound pipe ayatems, 106 through control gntea and valves, 120 Fundamental waterbammer equations, 10 Friction loses in pipeline, 101 s Gat flow through, 3,29, 48, 61 located at upper end of pipe, 61 located away from outlet end of pipe, oy ‘ing-follower, 120 turbine wicket, 120 Gate cloure: compete, 4,48, 52 ste located at upper ead of pipe, 61 tate located away from outlet end of pipe, 81 partial, 4 58 elit valve, 59 theory of graphical analysis, 45 uniform, 4,70 with stepwise change in diameter, 65 Gate discharge equations, 3, 29, 8, 61 Gato movement completed in fem than 21/a seconds, Et instantaneous, 30 tow, 37 Gate opening: ‘sraphieal solution, 63 retour rise prodiueed by, 63 thythmie opening and clonure, 65 uniform, 4, 72 Gates, dicharge through, 120, Gibson, NR, 144 Glover, RAE, 44 Graphical analy tie chambers, 136 compound pipe, 108 ste operation, 45 ump discharge line, 83 theory, 45 4 Halmos, EXE, 144 eed, lect of changes In pumping, 13 Homologous relatis for pumps, 79 Hydraulic lowes concentrated at: Tntake, 101 ‘pump divcharge outlet, 103, everal pointe along pipe, 105 a INDEX 159 ' ° Incompresible Dui, 2 of contol gates, 5,63, 72 Inertia equation for pumps, 76 Operation of: Instability of surge tanks, 123 tr cbaunbers, 131 Tastantaneous gate movements, 30 surge taka, 123 Tnatantaneows flow changes with surge ‘tanks, 128, Intake, lowes concentrated st, 101 Iterretation of wateraminer elton, i ‘acobson, RS, 144 Jaeger, C, 148 Tohosoo, R-D., 144 oukowsky, N., 144 ‘Tanetion of three or more pipes, wave reflection, 28 Kerr, 164 Kenler, LH, 144 Koapp, RT, 146 Kruse, O.V., 144 L Le Conte, JN. 144 Length, equivalent of pie ine with step vwise changes, 6, 115 Level, coatrols in ait chamber, 141 Limitations of rigid water column the: oy,7 Limiting poit of maximum head change, 35 Lomes, effect of hyleaulie, 101 ™ ‘Maximum reverié speed of pumps, 99 “Maximum time interval use in compas tations, 38 Midlength of pipe, head changes at, 40, 82, 138 Moody, EaF., 145 Motors, power faire at ium, 74 N ‘Newton's secon law of nation; 8, 8 10, 125, 127 ‘Numerical waterbaminee solution, 39 Orifice, throttling, used for: sir chambers, 182 surge tanks, 125, P Parmatian, J, 145, Partial gate closure, §, 68 Partially open gate, wave refletions, Peabody, TA. 145, Physical'signicance of waterhammce ‘equations, 16 Pipe tne: ‘anchorage, 13 constant, 48, 82, 10, 196 with stepmise changes ia diameter, 6, 26, 55,115 Pipes: ‘ast ion, 19 rminforeed concrete, 22 steel, 19 ‘Transit, 19 roodeatave, 20 Presure-volume changes for ait cham ber, 13 Pressure wave velocitirs, 19 Problems, 147 Propagation of pressure waves, St Pump and motor inertia envatinn. 1 Panp discharge line with air chamber, 131 Dattertly valve, 98 chock valve, 04 Tong suction Tne, 97 ‘0 control valve, 74 surge tank, 126 Pumping head, effet of ehangrs in Pump: characteristics, 78 hhomologaus nperation, 79 performance eurves, 77 starting and stopping, wateehanmer charts 37 waterhanmer due to shulsown of, 7 125, 131= =——_——= 77 { CTT 1 wo y 100 noe o erin 128 ss plot pootoast De inn wy 110,18 Soares Tie ete 125 Sea or wim ue do, Quick's chart for uniform gate a ‘Symposium oa Waterhammer, 145. at t apd sat movements 0 ‘aba ermntaon of eatin, Tala anne pot 9 ‘Tape toa etn 68 Itences 1 ney ates prewure wave a ine ater lum, 8 hangin meetin 38 bil enya dead end, 26 free water eurace, 25 junction of three oF more pipes, 23 Dartinlly open gate, 28 reservoir, 25, Relief valve, 59 Reinforeed conerete pipe, 22 Restricted orifice urge tank, 125, Rhythmic opening and closure of control ante, 65 Rich, GR, 145 Rigid water colurin theo basic equa equivalent length of pipe line, 6 example, 5 ‘uniform gate operttion, 4 Ring-follower gate discharge, 120 ‘Round trip wave travel time, 38,56 Runaway speed of pump, 75,00 s Schayder,0,, 145 Slow gate movements, 37 Spend of pump after power failure, 90 Sphere or plug valve, 120 Starting and stopping pumps, 90 Steel pipe, wave velocity in, 19. Steritined circular tuncel, wave velocity i, 23 Stepanod, A, 145 Strowger, Eh, 15 Suction line, pump discharge line with tong, 97 Surge tank: ‘alaneed design, 190 charts, 128. example, 130 fnetabuity, 123, rigid water eoluma, 1 Thoma, D, 145 Time: flow reversal at pump, 90 Warren, M.M, 145 intervala used in waterbammer com: Water: Putations, 33, 56 column separation, 86, 131 compressibility, 1 ‘uprurg in surge tank, 125 “8 a Tevel controls ia ‘maximum reverse pump epecd, 91 wave travel, 17 ero pump speed, 9 Timoshenko, 8, 146 ‘Transient conditions at pump discharge Tines, 74 ‘Transite pipe, 19 ‘Transmission of pressure waves a ‘change in eroseseetional area, 28 junction of three of mote pipes, 23 ‘Tunnels, wave velocity ins circular, 2 Meet-lined cieeuler, 28 ‘Turbino wicket gates, discharge through, 120 chamber, Mt u Uniform gate closure: clastic water column theory, 70 rigid water column theory, 4 Uniform gate opening: clastic water column theory, 72 gid water column theory, 4 Uniform pipe npproximation, 6 115 Upourge in surge tank charts for determining, 128 imple wurge tan 126 v buttery, 98, 120 cheek, contre, 7 161 Waterhammer: ‘latio water columa theory, 8 Figid water column theory, 1 theory of graphical analyis, 45 Wave: propagation, 32 ‘election, 25 tranumision, 27 travel time, 88, 66, 115 veloeity, 19 Wicket gates, turbine, 120 Wood, FM, 145 Woodt-tave pipe, 20 z Zoot of pumnp operation: energy disipation, 80 ‘ormal, 79

Water Hammer Analysis by J. Parmakian

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Water Hammer Analysis by J. Parmakian

Uploaded by

Copyright:

Available Formats

You might also like