SPE 94326
Wave Equation Simulation of Fluid Pound and Gas Interference
F. Yavuz, SPE, J.F. Lea, SPE, D. Garg, SPE, T. Oetama, SPE, and J. Cox, SPE, Texas Tech U., and H. Nickens, SPE, BP
Copyright 2005, Society of Petroleum Engineers
This paper was prepared for presentation at the SPE Production and Operations Symposium
held in Oklahoma City, Oklahoma, U.S.A., 1719 April 2005
This paper was selected for presentation by an SPE Program Committee following review of
information contained in an abstract submitted by the author(s). Contents of the paper, as
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to
correction by the author(s). The material, as presented, does not necessarily reflect any
position of the Society of Petroleum Engineers, its officers, or members. Papers presented at
SPE meetings are subject to publication review by Editorial Committees of the Society of
Petroleum Engineers. Electronic reproduction, distribution, or storage of any part of this paper
for commercial purposes without the written consent of the Society of Petroleum Engineers is
prohibited. Permission to reproduce in print is restricted to an abstract of not more than 300
words; illustrations may not be copied. The abstract must contain conspicuous
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O.
Box 833836, Richardson, TX 75083-3836 U.S.A., fax 01-972-952-9435.
Abstract
Using adiabatic compression of gas in the pump, a model is
shown that can be used to calculate the down-hole pump
dynamometer card for various degrees of pump fill of liquids
and gas at various degrees of pressure. The load release
portion of the card is emphasized. It is shown how the lower
pressure gas in the pump promotes what is commonly termed
fluid pound or slap and higher pressure gas in the pump
promotes what is termed gas interference.
The equations that are needed to model these effects
for inclusion into wave equation pump models are presented
and example calculated pump cards are shown, calculated
from wave equation simulations. Since even with so-called
fluid pound, gas is first compressed before the plunger
encounters the mostly incompressible fluid in the pump, the
traveling valve always encounters compressed gas sufficient to
open the valve before the plunger "hits" the fluid. This is true
as long as gas is in the barrel when incomplete fluid fill
occurs. Given this fact, the time step in wave equation solution
is examined as the load release in the so called "fluid pound"
can happen over a very short time and longer time steps in the
wave equation simulation could mask forces from short term
forces.
The effects of rapid load release on forces in the rods
are studied vs. what the forces might be due to the commonly
supposed forces from impact with the fluid when "fluid
pound" is modeled. Other parameters such as pump fillage,
intake pressure and sinker bars are examined for their effects
on calculated rod compression
Introduction
Fluid pound is associated with incomplete fillage of the
downhole pump. There is considerable opinion that the pound
and associated possible damage is related to the pump
traveling through gas and then striking or pounding the fluid
when it is encountered by the pump traveling downward.
However in this discussion it is shown that severe rod
compression can be encountered when no impact force is
simulated of the pump striking the fluid level. Instead it
appears that severe rod compression as a result of pump
incomplete fillage at low pressures is a result the pump
slowing as gas is compressed. This results in internal
compression of the rods as one section slows and the section
above slows etc. leading to compression above the pump.
Some examples of what can mediate this effect are presented
as well.
The dynamic loads during fluid pounding can
cause several detrimental effects on the downhole
equipment.: The rod string can experience buckling
that leads to rod breaks,; rod-to-tubing wear is
increased; shock loads contribute to coupling failure
due to unscrewing; and pump parts can be damaged
(as well as tubing), if unanchored. On the surface,
shock loads can damage pumping unit bearings, and
can lead to instantaneous torques that overloads the
speed reducer. (Takacs23)
Fluid pound is usually said to be the result of the pump
traveling through the gas filled portion of the pump and then
hitting the fluid. This is said to put compression into the
rods and to cause damage to the rod string, etc. However if
you view Figure 1, the first type of so-called pump card is the
type of card that is normally seen for fluid pound. In the first
type of card shown in Figure 1, there is not compression
shown. This is the typical pump card that is seen for the
situation of so-call fluid pound or rapid load release when the
intake pressure is low.
The second type of card shown in Figure 1, shows
that some compression is evident from presumably fluid
pound. This type of bottom hole card is not typical. Is it
possible that fluid pound does not in general lead to
compression or buckling in the lower rods?
Even in fluid pound, there is gas compression.
Assume 100 bottomhole stroke. Assume of the bottomhole
stroke is filled with fluid and the other half with gas at 100 psi.
Assume 2000 psi over the TV. Then:
P1/P2 = 100/2000 = V2/V1 = (X x A) / (50 x A) =
Solve for X and its equal to 2.5 inches or the TV will
open when the TV is 2 from the fluid.
2 SPE 94326
If the initial pressure in the barrel is 50 psi, then the
TV opens when the TV is 1 from the fluid. For lower
intake pressures it is seen that the traveling valve opens when
the pump is very close to the fluid level in the pump barrel.
So the TV always opens due to gas compression
before it hits the fluid. It would never hit the fluid without gas
having enough pressure to open the TV first. Still the plunger
hits the fluid with the TV open and goes from travel in the gas
to impacting the fluid and traveling downward through the
fluid column.
Also what about when the pump is full of fluid? The
load release happens quickly but no compression or fluid
pound is reported. This is presumed to be because the pump is
traveling very slowly compared to mid-stroke.
In Figure 2, fluid pound is usually shown in
cartoon fashion as showing no unusual compression as a
result of fluid pound, although usually in the text describing
fluid pound, it usually says rod compression is a result of fluid
pound.
The cartoon in the Figure 2 is showing what is typical
of pumps cards in the industry to identify fluid pound. But
again where is the compression? If compression is in the rods,
it is not in the rods just above the pump according to these
types of cards as in Figure 2.
However in the Figure 3, compression is shown for a
fluid pound card. This would show compression due to fluid
pound, although this type of bottom hole card is usually not
shown with a fluid pound situation. Usually the downward
spike, indicating compression to the rods over the pump, is not
present. Also in this pump card there is a downward spike
earlier in the stroke and it is possible both spikes are due to
something other than fluid pound, such as solids in the pump
Analysis
Before considering the rod boundary condition for the rods
above the pump, equations for the complete wave equation
analysis are presented below. These equations are somewhat
unique as it shows a way to establish boundary conditions at
rod junctures where tapered rods are connected without having
to average the rod properties and sizes in the equations of
motion.
Equations are shown for both the so-called predictive
and diagnostic type of wave equation beam pump computer
models.
All necessary equations are shown in Appendix A
Results
Before simulations are shown, the forces required to buckle
rods are shown in Figure 5. Note that the loads are small that
cause buckling and for most rods the load is below 100 lbs.
Simulations of fluid pound situations are then
presented. Figure 6a and 6b show the base case of 60% pump
fill, intake pressure of 30 psig, fluid gravity of 1.0, pump size
of 1 , spm=8, 144 surface stroke, conventional unit
pumping clockwise. The rod string is an 86 tapered grade D
rod string.
In Figure 6a, the predicted surface and pump cards
are presented. In Figure 6b, the rod loads vs. depth are
presented. The effective loads being negative are an indication
of buckling if the load exceeds the buckling limit. A
discussion of true forces and effective forces is presented in
Reference 21. For this simulation a section of negative loading
is shown in the rod string. This is the dotted line on the left. In
fact very serious loading is indicated showing compression
many times what would be required to buckle the rods. This
shows that fluid pound (Figures 6a and 6b are for low intake
pressure and the load release is rapid showing a so-called fluid
pound situation) has apparently caused rod compression but
there is no compression right above the pump. Most programs
input some default value of pump resistance such as 100 lbfs
to indicate flow through the TV and friction of the plunger
through the barrel. However here it is not done to emphasize
the other loading shown.
This downstroke compression is NOT due to the
plunger hitting the fluid level. And again, no compression in
the rods is shown above the pump. So why is there
compression in the rod string in the fluid pound situation? It
seems to be that when the load is release rapidly with fluid
pound and the pump is moving downward and gradually slows
as compression occurs in the pump. This results in slowing of
the pump and the section above slows more etc. resulting in
compression in the rods above the pump. Again no upward
compressive forces are input into this simulation.
To further investigate, in Figures 7, the intake
pressure is varied from 50 psig to 110 psig. Note that the rod
compression gradually diminishes to zero as the intake
pressure is simulated to increase.
To investigate further, in Figures 8, the effects of
pump fill are examined from 50 to 90% pump fillage with all
at 7 spm. Note that the effects of fillage for these simulations
have little effect. Normally it is said that the maximum
velocity is near the center of the downstroke and if the fluid
level is at this location, then more damage can result. However
the simulations show that the effects of fillage are random and
rod compression seems somewhat independent of the pump
fillage. This may be because the pump velocity is not
sinusoidal but is more erratic and the maximum in velocity at
the pump is not in general at the center of the downstroke at
the pump. Figure 9 shows the erratic nature of the pump
velocity possible in a pump cycle.
Figure 10 shows the effects on the calculated
compression force with adding sinker bars. Note that, at least
in this case, the addition of sinker bars had little effect on the
compression in the rod string.
Conclusions
Rod compression from fluid pound may not be due
from the plunger hitting the fluid level. Substantial
compression is simulated without this effect. The
hitting effect may or may not add substantial
compression in the rods.
The effect of the intake pressure on the rod
compression is substantial.
The effect of the % pump fillage on rod compression
seems to be smaller.
Sinker bars may not eliminate the compression from
so-called fluid pound.
You cannot see the level of rod compression in the
rod string from examining the pump card.
SPE 94326 3

You should simulate pumping conditions with a wave 1. Lea, J. F.: Boundary Conditions used with Dynamic
equation predictive to see if you are in a damaging Models of Beam Pump Performance presented at the
pumping condition that may be causing rod damage Southwestern Petroleum Short Course, Lubbock, TX,
with a program that plots or indicate rod loading with April, 1988, p 251-263.
depths. Loading at only node points is not sufficient 2. Gilbert, W. E.: "An Oil Well Pump Dynagraph," API
to see this type of loading. Drilling and Production Practice, 1936, pp. 94-115.
3. Gibbs, S. G.: "Computing Gearbox Torque and
Nomenclature Motor Loading for Beam Units with Consideration of
Inertia Effects," Journal of Petroleum Technology
A = Rod cross section area, ft2 [m2] (September 1975), pp. 1153-1159.
4. Gibbs, S. G.: "A General Method for Predicting Rod
c = Input dampening factor, dimensionless Pumping System Performance," paper SPE 6850
c 1 1 presented at the 1977 Annual Technical Conference
C = Dampening factor grouping, , , and Exhibition, Denver, October 9-12.
2 L ft m 5. Gibbs, S. G. and Neely, A. B.: "Computer Diagnosis
lbf Pa of Downhole Conditions in Sucker Rod Pumping
E = Modulus of elasticity (of sucker rods), Wells," Journal of Petroleum Technology, (January
ft 2 m 2 1966), pp. 91-98; Transactions, AIME, 237.
F = Force internal to rods due to dynamic stretch, lbf [Pa] 6. Gibbs, S. G.: "A Method of Determining Sucker Rod
lbm ft kg m Performance: U.S. Patent No. 3,343,409 (September
G= 32.17 , 26, 1967).
lbf sec 2 N sec
2
7. Gibbs, S. G.: "Predicting the Behavior of Sucker Rod
L = Length of rod string, ft [m] Pumping Systems," Journal of Petroleum Technology
(July 1963), pp. 769-778; Transactions, AIME, 228.
S = Time, sec [S] 8. Gibbs, S. G.: "Computing Gearbox Torque and
Motor Loading for Ultra High Slip Prime Movers,"
t = Increment of time, sec [S] SPE 5149, presented at the 49th Annual SPE Conf. in
Houston, TX, Oct. 6-9, 1974.
U = Dynamic rod displacement as a function, x and time, t, ft 9. Gray. H. E.: "Kinematics of Oilwell Pumping Units,"
[m] paper presented Amarillo, Texas, at the API Mid-
continent District Meeting, March 27-29, 1963.
X = Downward positive location from the surface to a position 10. Svinos, J. G.: "Exact Kinematic Analysis of Pumping
on the rod string, ft [m] Units," paper SPE 12201 presented at the 1983 SPE
Annual Technical Conference and Exhibition, San
X = Increment of position, ft [m] Francisco, October 5-8.
lbm kg 11. "API Catalog of Analog Computer Dynamometer
= Density if rod materials, Cards," API Bulletin llL2, First Edition, Washington,
ft 3 m 3 D.C., (1969).
ft m 12. "Design Calculations for Sucker Rod Pumping
V = Velocity sound in rods, Systems," API Bulletin IIL, Third Edition,
sec S Washington, D.C., 1977.
lbf Pa 13. Knapp, R. M.: "A Dynamic Investigation of Sucker
, , = Groupings of terms, Equation (8), , Rod Pumping, "M.S. Thesis, University of Kansas,
ft m Topeka, Kansas (1969).
14. Doty, D. R. and Schmidt, Z.: "An Improved Model
Subscripts for Sucker Rod Pumping," SPE Engineers Journal,
February 1983, pp. 33-40.
a - Signifies rod properties for a rods 15. Nicol, T. H.: "Dynamic Analysis of Sucker Rod
Pumping," M.S. Thesis, University of Oklahoma,
b - Signifies rod properties for b rods Norman, OK (1982).
16. Chacin, J. E.: "Diagnostic and Design Techniques in
i - Index for distance, x Oil Well Pumping," PhD Thesis, University of
Oklahoma, Norman, Oklahoma, 1986.
j - Index for time, t 17. Schafer, Donald J.: "An Investigation of Analytical
and Numerical Sucker Rod Pumping Mathematical
s - Indicates sum of groupings Models," M.S. Thesis, Texas A&M University,
College Station, Texas, (1987).
Reference 18. Chacin, U. and Purcupile, J. C., "A New Model for
Studying Oilwell Pumping Installations, SPE 16918,
4 SPE 94326

presented at the 62nd Annual SPE Conf., Sept. 27-30, U 2U (X b ) 2

U i +1 = U i X b + 2 .......................................( A 5b)
1987. x ib x ib 2
19. Schafer, D. J., and Jennings, J. W., "An Investigation
of Analytical and Numerical Sucker Rod Pumping Solving Eq. A-5a and Eq. A-5b for the force at i gives:
Mathematical Models," SPE 16919, presented at the
62nd Annual SPE Conf., Sept. 27-30, 1987. U 2U X a U U i 1
20. Carnahan, B., Luther, H. A., and Wilkes, J. O., Fia = ( AE ) a = 2 ( AE )a + i ( AE )a .....( A 6a)

x ia x ia 2 X a
Applied Numerical Methods, Wiley, 1969; p. 463.
21. Newman, K., Bhalla, K., The Effective Force,
U 2U X b U Ui
CTES L.C., Conroe, TX. Technical Nole, Jan 13, Fib = ( AE )b = 2 ( AE )b + i +1 ( AE )b ......( A 6b)
1999 x ib x ib 2 X b
22. Eickmeier, J. R., Applications of the Delta II
Dynamometer Technique, 17th Annual Technical All Us are at time j in Eq. A-5a, Eq. A-5b, Eq. A-6a, and Eq.
Meeting. The Petroleum Society of C.I.M., Calgary, A-6b. Representing the time derivatives with finite difference
May, 1994 expressions for rods a and rods b in Eq. A-1 gives:
23. Takacs, G., Sucker-Rod Pumping Manual, Pennwell,
2003
24. Rourk, R. J., and Young, W. C., Formulas for Stress 2U U 2U i , j + U i , j 1 c
(U i , j +1 U i , j )...................( A 7 a)
2 = i , j +1 +
and Strain, McGraw-Hill, NY,NY, 1982, P. 539 x ia (Va t )2 Va t

Appendix A 2U 2U i , j + U i , j 1 c
(U i , j +1 U i , j )....................( A 7b)
U
The equation of motion, Eq. A-1, for the constant property 2 = i , j +1 +
x ib (Vb t )2 Vb t
sucker rod string is presented in several references (1, 4,5,13,
etc.) which includes viscous dampening.
Where j and i index time and distance. Equating Eq. A-6a and
Eq. A-6b and substituting for the second derivative of U with
U Eg U c
2 2
Eg U
= ...........................................................( A 1) respect to x from Eq. A-7 gives:
t 2 x 2 2 L t
a (U i , j +1 2U i , j + U i , j 1 ) + a (U i, j +1 U i, j ) + a (U i, j U i 1, j )
For changing rod properties with depth (tapered string) the
= b (U i , j +1 2U i , j + U i , j 1 ) b (U i , j +1 U i , j ) + b (U i +1, j U i , j )...........( A 8)
above equation can be written as Eq. A-2:

2U g U U
= EA CV ........................................................( A 2)
t 2 A x X t where:
XAE XAEC AE
= ,......... ..... = ,......... ...... =
where: 2 (V t ) 2 2V t X
c
C= In Eq. A-8 all the groupings are either groups of a or b
2L
and V is the speed of sound in the rod material. properties. No averaging of properties is required as the
If the following development, only Eq. A-1 is used calculation proceeds from a to b. Also the x increments can be
but a method of analysis for a tapered rod string is still equal or unequal across the boundary (i.e. as from a one rod to
developed below. The below force equation is or can be at a a different rod).
rod taper where properties and cross section change. There can Solving for Ui,j+1 gives an algorithm for the design
or predictive type of analysis:
be a x in the rod a, and a x in the rod at b of different
lengths. U i , j (2 s + s b a ) (U i , j 1 )( s ) + U i +1, j ( b ) + U i 1, j ( a )
U i , j +1 = .....( A 9)
( s + s )
Fa = Fb .......... .......... .......... .......... .......... .......... .......... .......... .......( A 3)
where;

or: s = a +b
s = a + b
( AE )a U U
= ( AE )b .............................................................( A 4)
x ia x ib
The technique of solution of the predictive problem is covered
Following a procedure in Reference 20 and expanding Ui+1 and elsewhere 4, 7, 13, 14.
Ui-1 about Ui gives in Eq. A-5a and Eq. A-5b: The solution involves evaluating the rod wave
equation over the complete length of the rod string for each
U 2U (X a )2 time in a complete pump cycle. The surface motion is supplied
U i 1 = U i X a + 2 ......................................( A 5a ) at each time by a geometric formula 9, 10. At each time, a pump
x ia x ia 2
boundary condition that specifies pump loading is used. The
SPE 94326 5

solution requires repetitive complete cycles until the solution

has converged usually requiring at least 2-4 cycles.
The stability of the solution is affected by the choice
of the time increment. For a constant x increment, the solution
is stable13 for t x/V where t is the smallest value for any
segment.
Another type of rod pump analysis is to input a
dynamometer card as points of load and position at the surface
and calculate down the rod string to just above the pump. This
is referred to as the diagnostic problem. Instead of using a
Fourier series solution4 to develop the diagnostic type of beam
pump analysis, Eq. A-8 is solved for Ui+1,j.

+ s 2 + s b a
Ui +1, j = Ui, j +1 s + Ui, j 1 s Ui 1, j a Ui, j s



b b b b
.....................................................................(A 10)

For the rod boundary condition above the pump when the load
is released for the condition of partial pump fillage, the
nomenclature and equations used are identified in Figure 4.
Essentially as the pressure increases in the gas below the TV
as the pump goes downward, the equations adiabatic
compression are used. So the pressure at any location is found
from:

P 2 = {( P1 V 1 k ) / V 2k } 1 / k .......... .......... .......( A 11 )

where:
k is the ratio of specific heats for natural gas
V1 is the initial volume of gas in the pump before the
downstroke starts.
V2 is the compressed volume of gas at any downward location
of the pump
P2 is the pressure in the gas as it is compressed

The use of this equation interfacing with the stress in the rods
above the pump is, again, shown in Figure 4.
6 SPE 94326

We see this

Why not this???

Figure 1. Calculated (or measured) fluid pound dynamometer cards in the rods over the pump.

Figure 2. Cartoon representation of fluid pound pump cards.

SPE 94326 7

Ref: 7 Pmp Disp; 190 blpd

Fluid Disp: 98 blpd
Production: 81 bopd
Efficiency: 65%
Load

7300 well
7/8,3/4,5/8 rods
1 pump
11 SPM, 76 Stroke

Position

Figure 3. Measure/calculated surface/bottom hole cards showing some non-typical compression on due to fluid pound

Figure 4. Description of terminology of calculating rod loading above the pump for partially filled pump.
8 SPE 94326

Figure 5. Loads predicted to buckle sucker rods of various sizes24.

Maximum, True Minimum, True Maximum, Eff

Polished Rod Minimum, Eff

Effective Pump Load Depth (ft)

Load (1000 lbf) 0
Perm. Load - In Balance
26
SV/T V O pen
24 800

22

20 1600
Rod 2 Top
18

2400
16

14
CBE
3200
12

10 Top of Fluid in Pump Rod 3 Top

4000
8
SV
6
4800
4

2 5600

0
TV
6400
-2

-4
0 25 50 75 100 125 150
Pump
7200
Position from Bottom of Stroke (in) -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24

C:\PROGRAM FILES\PRODUCTION AND LIFT TECHNOLOGY, LLC\ACCUPUMP\EX1.ACP: 8.00 SPM, 144.000 in Strok e, Production = 199.7
Rod Load (1000 lbf)
C:\PROGRAM FILES\PRODUCTION AND LIFT TECHNOLOGY, LLC\ACCUPUMP\EX1.ACP: 8.00 SPM, 144.000 in Stroke, Production = 199.7

Figure 6 (a). Calculated Pump Cards for Figure 6 (b). Rod loads for 60% Pump Fillage.
Pump Fillage = 60%, Intake pressure= 30 psi.
SPE 94326 9

Polished Rod Maximum, True Minimum, True Maximum, Eff

Effective Pump Load Minimum, Eff

Loa d (10 0 0 lbf) Perm. Load - In Balance
De pth (ft)
26 0
SV/T V O pen
24
800
22 Case 1 Case 1
P um p, Intake Pressure 50 psi Pum p, I nt ake Pressure 50 psi

20
1600
Rod 2 Top
18

16 2400

14 CBE
3200
12

10 Top of Fluid in Pump Rod 3 Top

4000
8
SV
6 4800

4
5600
2

0
TV
6400
-2
Pump
-4 7200
0 25 50 75 100 125 150 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24
Position from Bottom of Strok e (in) Rod Loa d (1 0 0 0 lbf)
C: \P RO G RA M F ILE S \P RO DUCT I O N A ND LIF T T E CHNO LO G Y , LLC\A CCUP UM P \EX 1. ACP : 8.00 S P M, 144. 000 in S troke, Product ion = 199. 8 C: \ PRO G RAM F I LES\ PRO DUCT I O N A ND LI F T T ECHNO LO G Y , LLC\ ACCUPUM P\ EX1. ACP: 8. 00 SPM , 144. 000 in St roke, Product ion = 199. 8

Figure 7 (a). Intake Pressure = 50 psi. Figure 7 (b). Rod load vs. Depth.
Polished Rod

Effective Pump Load Maximum, T rue Minimum, T rue Maximum, Ef f

Loa d (1 0 0 0 lbf) Perm. Load - In Balance Minimum, Ef f

26 De pth (ft)
SV/TV O pen 0
24

22 Case 2
Pum p, I nt ake P ressure 70 psi 800
Case 2
P um p, I nt ake Pressure 70 psi
20

1600
18
R od 2 Top

16
2400
14 CBE

12 3200

10 Top of Fluid in Pump

R od 3 Top
4000
8
SV
6 4800

5600
2

0 TV 6400
-2
Pump
-4 7200
0 25 50 75 100 125 150 -2 0 2 4 6 8 10 12 14 16 18 20 22 24

P os ition from Bottom of S trok e (in) Rod Loa d (1 0 0 0 lbf)

C: \ PRO G RAM F I LES \P RO DUCT I O N AND LIF T T E CHNO LO G Y, LLC\ ACCUP UM P\ EX 1.A CP: 8.00 SP M , 144. 000 in St roke, Product ion = 199. 6 C: \ PRO G RAM F I LE S \ P RO DUCT I O N A ND LI F T T E CHNO LO G Y , LLC\ A CCUP UM P \ E X 1. A CP : 8. 00 S P M , 144. 000 in S t roke, P roduct ion = 199. 6

Figure 7 (c). Intake Pressure = 70 psi. Figure 7 (d). Rod load vs. Depth.
Maximum, T rue Minimum, T rue Maximum, Eff

Minimum, Eff
Polished Rod
De pth (ft)
Ef fective Pump Load 0
Loa d (10 0 0 lbf) Perm. Load - In Balance
26
SV/T V O pen
800
24 Case 3
P um p, I nt ake P ressure 90 psi

22 Cas e 3
P um p, I nt ake Pres sure 90 psi
1600
20 Rod 2 Top

18
2400
16

14 C BE 3200

12
Rod 3 Top
4000
10 Top of Fluid in Pump

8
SV 4800
6

4
5600
2

0 6400
TV

-2 Pump
7200
-4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24
0 25 50 75 100 125 150

P osition from Bottom of S trok e (in) Rod Loa d (1 0 0 0 lbf)

C: \ P RO G RA M F I LE S \ P RO DUCT I O N A ND LI F T T E CHNO LO G Y , LLC\ A CCUP UM P \ E X 1. A CP : 8. 00 S P M , 144. 000 in S t roke, P roduct ion = 199. 9
C: \ P RO G RA M F I LE S \ P RO DUCT I O N A ND LI F T T E CHNO LO G Y , LLC\ A CCUP UM P \ EX 1. ACP : 8. 00 S P M , 144. 000 in S t roke, Product ion = 199. 9

Figure 7 (e). Intake Pressure = 90 psi. Figure 7 (f). Rod load vs. Depth.
Polished Rod Maximum, T rue Minimum, T rue Maximum, Eff

Ef fective Pump Load

Minimum, Eff
Loa d (10 0 0 lbf) Perm. Load - In Balance De pth (ft)
26
0
SV/T V O pen
24

22 Cas e 4 800
P um p, I nt ake Pres sure 110 psi Case 4
P um p, I nt ake P ressure 110 psi

20
1600
18 Rod 2 Top

16
2400
14 C BE

12 3200

10 Top of Fluid in Pump

Rod 3 Top
8 4000
SV
6
4800
4

2
5600
0
TV

-2 6400

-4
0 25 50 75 100 125 150 Pump
P osition from Bottom of S trok e (in) 7200
-2 0 2 4 6 8 10 12 14 16 18 20 22 24

C: \ P RO G RA M F I LE S \ P RO DUCT I O N A ND LI F T T E CHNO LO G Y , LLC\ A CCUP UM P \ EX 1. ACP : 8. 00 S P M , 144. 000 in S t roke, Product ion = 199. 5 Rod Loa d (1 0 0 0 lbf)

C: \ P RO G RA M F I LE S \ P RO DUCT I O N A ND LI F T T E CHNO LO G Y , LLC\ A CCUP UM P \ E X 1. A CP : 8. 00 S P M , 144. 000 in S t roke, P roduct ion = 199. 5

Figure 7 (g). Vary Intake Pressure: 110 psi. Figure 7 (h). Rod load vs. Depth.

Figure 7 (a, b, c, d, e, f, g, h).

Pump Fillage= 60%, 7 spm.
10 SPE 94326

Polished Rod

Effective Pump Load

Maximum, T rue Minimum, T rue Maximum, Eff
Loa d (1 0 0 0 lbf) Perm. Load - In Balance
26 Minimum, Eff

SV/TV O pen De pth (ft)

24 0

22 Base Case

800
B ase Case
20

18 1600
R od 2 Top
16
C BE
2400
14

12 3200

10 Top of Fluid in Pump

R od 3 Top
4000
8
SV
6
4800

2 5600

0 TV
6400
-2
Pump
-4 7200
0 25 50 75 100 125 150 -2 0 2 4 6 8 10 12 14 16 18 20 22 24

Pos ition from Bottom of S trok e (in) Rod Loa d (1 0 0 0 lbf)

C: \ P RO G RA M F I LE S \ P RO DUCT I O N A ND LI F T T E CHNO LO G Y , LLC\ A CCUP UM P \ E RRO R2. A CP : 7. 00 S P M , 144. 000 in S t roke, P roduct ion = 127
C: \P RO G RAM F I LE S\ P RO DUCT I O N A ND LI F T T E CHNO LO G Y , LLC\ A CCUPUM P \ E RRO R2. A CP : 7. 00 S P M , 144. 000 in St roke, P roduct ion = 127

Figure 8 (a). Pump Fillage = 50%. Figure 8 (b). Rod load vs. Depth.
Maximum, True Minimum, T rue Maximum, Eff

Polished Rod Minimum, Eff

Ef fective Pump Load

De pth (ft)
0
Loa d (1 0 0 0 lbf) Perm. Load - In Balance
26
SV/T V O pen
24 800
Case 1
Pum p, F illage 60. 0 %
22 Cas e 1
P um p, F illage 60. 0 %

20 1600
R od 2 Top
18
2400
16
C BE
14
3200
12
Top of Fluid in Pump
10 R od 3 Top
4000
8
SV
6 4800

2 5600

0 TV
6400
-2

-4 Pump
0 25 50 75 100 125 150
7200
-2 0 2 4 6 8 10 12 14 16 18 20 22 24
P os ition from Bottom of S trok e (in)
Rod Loa d (1 0 0 0 lbf)
C: \ P RO G RA M F I LE S \ P RO DUCT I O N A ND LI F T T E CHNO LO G Y , LLC\ A CCUP UM P \ ERRO R2. ACP : 7. 00 S P M , 144. 000 in S t roke, P roduct ion = 162

C: \P RO G RAM F I LE S\ P RO DUCT I O N A ND LI F T T E CHNO LO G Y , LLC\ A CCUPUM P \ E RRO R2. A CP : 7. 00 S P M , 144. 000 in St roke, P roduct ion = 162

Figure 8 (c). Pump Fillage = 60%. Figure 8 (d). Rod load vs. Depth.
Maximum, True Minimum, T rue Maximum, Eff

Polished Rod Minimum, Eff

De pth (ft)
Effective Pump Load
0
Loa d (1 0 0 0 lbf) Perm. Load - In Balance
26
SV/T V O pen
24 800
Case 3
Pum p, F illage 75. 0 %
22 Case 2
P um p, F illage 70. 0 %

1600
20
Rod 2 Top
18
2400
16

14 CBE
3200
12

10 Top of Fluid in Pump Rod 3 Top

4000

8
SV
6 4800

4
5600
2

0 TV
6400
-2
Pump
-4
0 25 50 75 100 125 150 7200
-2 0 2 4 6 8 10 12 14 16 18 20 22 24
P os ition from Bottom of S trok e (in)
Rod Loa d (1 0 0 0 lbf)
C: \ P RO G RA M F I LE S \ P RO DUCT I O N A ND LI F T T E CHNO LO G Y , LLC\ A CCUP UM P \ E RRO R2. A CP : 7. 00 S P M , 144. 000 in S t roke, P roduct ion = 196
C: \P RO G RAM F I LE S\ P RO DUCT I O N A ND LI F T T E CHNO LO G Y , LLC\ A CCUPUM P \ E RRO R2. A CP : 7. 00 S P M , 144. 000 in St roke, P roduct ion = 213

Figure 8 (e). Pump Fillage = 70. Figure 8 (f). Rod load vs. Depth.
Maximum, True Minimum, True Maximum, Eff

Polished Rod Minimum, Eff

De pth (ft)
Effective Pump Load
0
Loa d (1 0 0 0 lbf) Perm. Load - In Balance
26
SV/T V O pen
24 800
Case 2
P um p, F illage 70. 0 %
22 Case 3
P um p, F illage 75. 0 %
1600
20 Rod 2 Top

18
2400
16

14 CBE 3200

12
Rod 3 Top
10 Top of Fluid in Pump 4000

8
SV
4800
6

4
5600
2

0 TV 6400

-2
Pump
-4 7200
0 25 50 75 100 125 150 -2 0 2 4 6 8 10 12 14 16 18 20 22 24

P os ition from Bottom of S trok e (in) Rod Loa d (1 0 0 0 lbf)

C: \ P RO G RA M F I LE S \ P RO DUCT I O N A ND LI F T T E CHNO LO G Y , LLC\ A CCUP UM P \ E RRO R2. A CP : 7. 00 S P M , 144. 000 in S t roke, P roduct ion = 213 C: \ P RO G RA M F I LE S \ P RO DUCT I O N A ND LI F T T E CHNO LO G Y , LLC\ A CCUP UM P \ ERRO R2. ACP : 7. 00 S P M , 144. 000 in S t roke, P roduct ion = 196

Figure 8 (g). Pump Fillage = 80%. Figure 8 (h). Rod load vs. Depth.
SPE 94326 11

Polished Rod Maximum, T rue Minimum, T rue Maximum, Ef f

Effective Pump Load Minimum, Ef f

Loa d (1 0 0 0 lbf) Perm. Load - I n Balance
De pth (ft)
26 0
SV/T V O pen
24
800
22 Case 4 Case 4
Pum p, F illage 90. 0 % P um p, F illage 90. 0 %

20
1600
18 R od 2 Top

16 2400

14 CBE
3200
12

10 Top of Fluid in Pump R od 3 Top

4000
8
SV
6 4800

4
5600
2

0 TV
6400
-2
Pump
-4 7200
0 25 50 75 100 125 150 -2 0 2 4 6 8 10 12 14 16 18 20 22 24
P os ition from Bottom of S trok e (in) Rod Loa d (1 0 0 0 lbf)
C: \ P RO G RAM F I LES\ P RO DUCT I O N A ND LI F T T E CHNO LO G Y, LLC\ A CCUPUM P\ ERRO R2. ACP: 7. 00 S PM , 144. 000 in S t roke, P roduct ion = 264 C: \ P RO G RA M F I LE S \ P RO DUCT I O N A ND LI F T T E CHNO LO G Y , LLC\ A CCUP UM P \ E RRO R2. A CP : 7. 00 S P M , 144. 000 in S t roke, P roduct ion = 264

Figure 8 (i). Pump Fillage = 90%. Figure 8 (j). Rod load vs. Depth.

Figure 8 (a, b, c, d, e, f, g, h, i, j).

Vary Intake Pressure= 50 psi 7 spm.

Polished Rod

Pump
Velocity (ft/sec)
8

-2

-4

-6

-8

-10

-12
0 1 2 3 4 5 6 7 8

Time in Stroke (sec)

EX1.ACP: 8.00 SPM, 144.000 in Stroke, Production = 199.7 bbl/d

Figure 9: Pump and surface velocity in a pump cycle.

12 SPE 94326

Polished Rod
Maximum, True Minimum, True Maximum, Eff

Effective Pump Load

Minimum, Eff
Loa d (1 0 0 0 lbf) Perm. Load - In Balance Depth (ft)
26
0
SV/T V O pen
24

22 800

20
1600
18
Rod 2 Top
16
2400
14 CBE

12
3200

10 Top of Fluid in Pump

Rod 3 Top
8 4000
SV
6
4800
4

2
5600
0
TV

-2 6400

-4
0 25 50 75 100 125 150 Pump
P os ition from Bottom of Strok e (in) 7200
-4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24
C: \P RO G RA M F ILE S \P RO DUCT I O N A ND LIF T T E CHNO LO G Y , LLC\ ACCUP UM P \E RRO R2.A CP: 8.00 S P M , 144.000 in S troke, P roduction = 202 Rod Load (1000 lbf)
C:\PROGRAM FILES\PRODUCTION AND LIFT TECHNOLOGY, LLC\ACCUPUMP\ERROR2.ACP: 8.00 SPM, 144.000 in Stroke, Production = 202

Figure 10 (a). Pump fillage = 60%, 8spm Figure 10 (b). Rod load vs. Depth.
86 rod string.

Polished Rod Maximum, True Minimum, True Maximum, Eff

Effective Pump Load Minimum, Eff

Load (1000 lbf) Perm. Load - In Balance
Depth (ft)
28 0
SV/TV O pen

800
24

1600
20 Rod 2 Top

2400
16

3200
CBE
12
Rod 3 Top
Top of Fluid in Pump 4000

8
SV 4800

4
5600

0
TV 6400
Rod 4 Top
Pump
-4 7200
0 25 50 75 100 125 150 -4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24

Position from Bottom of Stroke (in) Rod Load (1000 lbf)

C:\PROGRAM FILES\PRODUCTION AND LIFT TECHNOLOGY, LLC\ACCUPUMP\ERROR2.ACP: 8.00 SPM, 144.000 in Stroke, Production = 202 C:\PROGRAM FILES\PRODUCTION AND LIFT TECHNOLOGY, LLC\ACCUPUMP\ERROR2.ACP: 8.00 SPM, 144.000 in Stroke, Production = 202

Figure 10 (c) Pump fillage = 60%, 8spm Figure (d). Rod load vs. Depth.
86 rod strings with 300 feet
of 1 sinker bars.

Figure 10 (a, b, c, d) Effect of Sinker Bars.