Beam Pump Balancing Based on

Motor Power Utilization

J.N. MCCOY, W.A. DRAKE

R.E. OTT

Echometer Company

Consultant

F. COLLIER

A.L. PODIO

Mobil Exploration & Production U.S., Inc.

University of Texas at Austin

Abstract
With more and more emphasis on reducing the operating
costs for beam units, operators are examining several approaches to cut costs. One of the biggest costs in beam unit operations
is the electrical power cost.
Controlling pumping units to adjust to the optimum counterbalance will reduce the electrical bill. Several pumping untits
have been checked with a PC software program (POWER),
which allows the operator to determine how far out of balance
the units are and what it will take to properly balance them.
Properly balanced pumping units will result in electrical savings
both in demand and in consumption, thus reducing total electrical costs.
Examples from several fields are provided illustrating actual
power costs and the reduction in expenses that occurred.

Introduction
The first approach in determining whether the pumping unit
requires re-balancing or motor downsizing is to conduct a power
survey of the field. When the POWER software first became
available, it was used on several individual wells to determine if
the unit had an out of balance problem or the motor was too large.
The POWER program requires actual data. Two current and
three voltage probes are connected to the incoming electrical
lines. During two strokes of the pumping unit, power and current
data is stored in computer memory for interpretation. The power
data is converted to gear box torque and if out of balance condition is observed the program recommends the counterweight
adjustment required to achieve a better balance. The actual time to
measure these values is small, but all the data is necessary to
properly interpret the condition of the well. If the well is operated
with a pump-off controller, one must determine the condition of
the well for the majority of the run time, since pumping conditions
may vary widely from first coming on to just before pump-off.
The first small field surveyed consisted of 17 producing wells
in the Waddell Field, near Crane, Texas. Production depths ranged
from 2,635 m to 2,937 m (8,700 to 9,700 ft.) and all wells were
equipped with beam pump units. The Waddell field has a history
of high gas-oil ratios, which has caused some lifting problems in
the past. Incomplete pump fillage is not uncommon since, in some
cases, the tubing intake is located above the perforations, due to
the presence of liners and open-hole completions.
Direction of pumping unit rotation was checked to determine
the direction that resulted in lower electrical power consumption.
In some cases, changing the direction of rotation did help reduce
Special Edition 1999, Volume 38, No. 13

the electrical power costs and the rotation was permanently

changed to take advantage of the savings. Other work was performed to determine which tap on a high-slip motor is best, and
whether direction of rotation makes a difference in operations.
Preliminary recommendations are to operate in the high-torque
mode and rotate counterclockwise, unless rod overloading
becomes a problem.
Table 1 lists the wells in the Waddell area and data gathered
from the field survey. From the data it was recommended that 12
of the 17 wells required counterbalance weight adjustment. The
estimated cost savings amounted to $168 per month.
One advantage of the POWER software is that the measurements are made at the motor end of the pumping unit rather than
at the polished rod. Other balancing programs require a
dynamometer card, the card interpreted, and then entering the
resultant data to software programs to determine the best counterbalance for a particular well. Another advantage of the POWER
software is that it does not have to have a particular counterbalance weight library. Since only the weight and not the position of
the counterweight is required, the dimensions of counterweight
can be used to calculate volume and its weight. The POWER software can be run, and the weight of the counterweight is input to
compute the distance it must be moved further in or out as computed by the program.
To confirm the proposed work on the wells in the Waddell
area, a dynamometer card was taken on each well prior to moving
the counterbalance weights. This data confirms polished rod data
as well as determining how complete the pump fillage was on
each well during its pumping cycle. All of the wells were
equipped with pump-off controllers (POCs). The POWER software was also run again prior to moving any counterbalance
weights to confirm that each well did require changing the position of the weights.
The POWER software allows the user to check the intermediate results while moving the counterbalance weights, since the
program only requires data from two strokes of the pumping unit.
This gives the operator much more confidence in moving the
counterbalance weights since the results are immediate and indicate if a better balance is advised. Table 1 presents the data from
one of the wells in which the counterbalance weights were moved
and POWER was run between each move.
After the counterbalance weights were moved to their final
positions, a confirming power measurement was made to verify
the initial survey. Table 2 presents the final results compared to
the initial survey data. Measurement of the power on the upstroke
and downstroke did confirm that the wells were better balanced
after counterweight movement. Seven wells showed decreased
power costs.
Actual billed electrical costs to operate the beam units prior to

PAPER: 95-25

Journal of Canadian Petroleum Technology

TABLE 1: University waddell wells requiring counterweight adjustment.

Well No.

Production
(O/W)

Monthly
Charge $

Out of Peak
Balance,
M in-lbs.

Torque
Power M in-lbs.

Avg.KW
POWER
Software

Move Weights

319
324
323
102
321
313
320
422
327
605
326
401

47/9
16/30
18/26
14/15
70/48
23/48
46/26
94/101
29/4
40/13
58/16
48/6

236.
106.
472.
146.
791.
348.
497.
443.
149.
176.
488.
227.

124.
82.
80.
75.
70.
66.
54.
48.
51.
40.
36.
38.

410
189.
389.
226.
379.
207.
327.
268.
210.
179.
450.
243.

17.5
2.9
25.2
8.3
20.9
9.7
13.8
13.3
8.6
10.4
13.6
15.5

Remove weight
Move all in 11.25
Move out 25.125
Move weights out
Move RJL in 38.25
Move wts in 13,375
Add heavier weights
Move in 8.5
Remove weights
Move out 2 weights
Move XJR in 9.5
Move 2-2ROs in

moving counterbalance weights were obtained. The Waddell Field

has one electric meter, from which power is supplied both to the
individual wells and to the water injection station. Costs associated with the water station were netted out to determine how much
electricity was being used by the pumping wells. Injection pump
run time and horsepower usage was determined and a table was
developed to reflect water injection electrical consumption. Table
3 presents the data from both before and after the work was done.
The average of the daily costs indicates a decrease of almost
$100/day which shows that the savings achieved from balancing
exceeded the predicted savings of $168/month. This was probably
due to the demand factor we were billed for as well as KWH consumption. The KW demand was not included in the initial cost
savings calculations. The KW demand factor was reduced about
10 per cent, and at $7.63 per KW demand amounted to another
$375 per month in savings. This, together with the reduced KWH
consumption provided a quicker payout than initially calculated.
In Table 3 is shown a three month decrease (May August) in
electrical costs and then an increase (August September). This
increase could not be accounted for, so a second survey of the
field with the POWER software was done. Any changes could
then be noted.
The resurvey of the Waddell area was quite a surprise. Several
wells were operating on Manual Control rather than with the

POC. Because of this, the electrical costs were higher, since the
wells do not require continuous pumping. Personnel changes
occurred during the increase in electrical costs and this probably
led to the lack of communication between the lease operators and
other field staff. There were also some POC electrical problems
that were not immediately repaired and, because of this, the
POCs were not effective in controlling the operation of the wells.
While this work was being done, questions arose as to whether
there was any Rule of Thumb for determining how far out of
balance a unit would have to be to require action being taken
based on potential cost savings. For API 456 units and smaller, if
the out of balance is less than 10 per cent of the unit rating, it is
probably not worth moving the weights. For units more than 10
per cent out of balance (torque values), it will be worthwhile to
reposition the weights. For API 640 units and larger, anything
over 50,000 in./lbs. is worth repositioning the weights. This Rule
of Thumb is not firm, but should give the operator some guidelines on when to start moving weights. This statement is based on
work done moving counterbalance weights when the units were
less than 10 per cent out of balance.

Second Phase
Because of the success of the work in the Waddell field, the

TABLE 2: University Waddell lease POWER software measurements.

POWER Software
Electrical Costs
$/month
Well #
319
324
323
102
321
313
320
422
327
605
326
401
-168.

TorqueValues

Before*

After

Change

Before*

After

236.
221.
770.
146.
753.
349.
595.
553.
123.
188.
488.
221.

194.
188.
853.
126.
781.
356.
605.
735.
87.
100.
262.
188.

-42.
-33.
+83.
-20.
+28.
+7.
+10.
+182.
-36.
-88.
-226.
-33.

301.
139.
388.
161.
316.
162.
344.
368.
178.
139.
404.
216.

306.
143.
429.
191.
297.
173.
349.
414.
180.
104
336.
196.

*The values reported here are from the resurvey of the field prior to moving weights
2

Journal of Canadian Petroleum Technology

TABLE 3: University Waddell electrical costs.

Period
27Dec93 25Jan94
25Jan94 23Feb94
23Feb94 25Mar94
25Mar94 26Apr94
26Apr94 26May94

Days

KW
Demand

KWH
Used

Downtime
Days

Downtime
KWH

KWH w/o
Pump

Electric
Bill, $

$w/o
Pump

$/Day
w/o pump

28
28
28
31
30

549
576
559
547
543

314488
310092
347692
321419
280191

0
0
0
7
21

0
0
0
18799
56398

239291
234895
272495
256964
256021

15688.59
16046.21
16728.04
15849.78
14065.58

12068.43
13898.37
13165.93
12646.84
12351.67

431.02
496.37
470.21
407.96
411.72

Avg. 443.46
26May94 27Jun94
27Jun94 27Jul94
27Jul94 26Aug94
26Aug94 27Sep94
27Sep94 27Oct94
27Oct94 29Nov94

31
30
30
32
30
33

497
532
485
496
517
508

323456
268726
278387
303930
308138
331089

1.2
0
0
0
0
0

3134
0
0
0
0
0

243336
188158
197819
217991
227570
269049

14528.61
13659.94
13222.59
14031.25
14412.61
14846.52

11016.31
9921.84
9611.40
10276.22
10941.32
12985.96

355.36
330.73
320.38
321.13
354.71
393.51

Avg. 345.97
Note: Work was done in Mid-May 1994, to re-balance the counterweights

TABLE 4: PARKS FIELD wells requiring counter balance weight adjustment.

Well No.

Required
Counterweight
Adjustment

Power,
KWH

Peak Upstroke
Torque M-in lbs.

Peak Downstroke
Torque M-in lbs.

Balanced
M-in lbs.

18-26
25-5
19-14
19-17
18-21
23-15
18-14
19-18
23-11
25-3
MVB-1
AB-10
18-27
ABH-4
14-18
CBH-2
HSC-3
23-21
26-18
ABH-3
RPA-1
MVB-2
14-19
18-13
RPC-2
26-A9*

241 Out
212 Out
168 Out
125 Out
123 Out
114In
109In
108 Out
100In
99Out
87Out
86Out
73Out
70In
69Out
65In
60Out
59In
47Out
47Out
41Out
41Out
38Out
35Out
33Out
18In

10.1
6.6
6.4
11.3
8.7
17.2
7.5
16.7
7.5
11.0
9.5
14.2
7.0
11.4
7.5
3.7
10.1
14.5
13.8
8.5
4.9
3.4
9.3
7.4
7.7
6.0

324.
321.
384.
393.
459.
197.
213.
531.
154.
243.
360.
349.
171.
188.
227.
154.
192.
103.
240.
257.
238.
188.
98.
342.
212.
158.

20.
58.
132.
187.
216.
420.
391.
372.
315.
73.
202.
188.
79.

177.
185.
224.
282.
341.
307.
290.
437.
256.
160.
277.
270.
123.
118.
181.
97.
138.
160.
192.
2 18
200.
155.
135.
312.
184.
175.

117.

214.
148.
176.
131.
156.
277.
151.
189.

*Change Rotation Only

Parks field was surveyed. The Parks Field contains about 90 producing wells, most of which are on a beam pump. This field also
has a high gas-oil ratio. Table 4 presents the results of survey data
through listing only wells requiring some re-balancing.
Counterbalance weights were moved to reduce electrical costs.
At the north end of the Parks field there are a few wells either
with individual meters or with a small number of wells tied to a
single meter. The wells were re-balanced according to recommendations from the POWER software program which was used in
conjunction with a conventional dynamometer. Data was collected
for the electrical bills in these smaller installations as well as the
larger single field meter. The water injection station was also
included in the larger master meter, so it was deducted from the
base meter rate to try to determine any changes in power use in
the field.
Special Edition 1999, Volume 38, No. 13

Table 5 presents the data for the main area in the Parks field.
An average saving of $50/day was obtained as a result of the rebalancing project. Table 6 presents the single well data. The single
well data is not as conclusive, because changes at the well head
(i.e., run time, downtime, pump-off-control resetting) affect the
electrical readings for the month. Individual well problems are
explained in the notes at the bottom of Table 6.
One area of interest in this work is the measurement of the
power factor for each well. Low power factors usually suggest
room for improvement in the electrical design/installation and this
was confirmed by the data. However, trying to obtain large
improvements in the power factor is not easy. Each well was optimized with the best electrical installation available (without completely changing out equipment) and large improvements in the
power factor were not always achieved. Decreasing motor sizes
3

TABLE 5: Parks Field electrical costs.

Billing Period
13Apr94 14Mar94
12May94 13Apr94
13Jun94 12May94
14Jul94 13Jun94

Days

KWH

Adjusted

$Adjusted

$/Day

30
29
32
31

439600.
425909.
447158.
459658.

246568.
239311.
241257.
260192.

13694.68
13259.23
13600.04
13781.35

7692.92
7442.57
7334.21
7805.76

256.43
256.64
229.19
251.80
Avg. 248.52

12Aug94 14Jul94
14Sep94 12Aug94
13Oct94 14Sep94
14Nov94 13Oct94
15Dec94 14Nov94

29
33
29
32
31

412300.
471520.
420234.
478343.
486449.

225702.
232653.
187004.
195258.
174825.*

12517.91
13957.08
12783.55
14301.63
14574.75

6861.34
6886.53
5684.92
5838.21
5244.75

236.60
208.68
196.03
182.44
169.19
Avg.198.59

* 1. Water Injection pumps (200 BHP) have been subtracted out

2. Five new wells were added to the field, beginning in August 1994 through November 1994
3. Work was done in late July to re-balance the counterweights

TABLE 6: North Parks Field area electrical costs.

Billing Period
13Apr94 14Mar94
12May94 13Apr94
13Jun94 12May94
14Jul94 13Jun94
12Aug94 14Jul94

Days

KWH

Adjusted

$/Kilowatt

$/Day

30
29
32
31
29

16172.
16290.
21542.
19007.
18891.

16172.
16290.
21542.
19007.
18891.

1404.74
1479.76
1549.88
1520.65
1402.63

0.0869
0.0908
0.0719
0.0800
0.0742

46.82
51.03
48.43
49.05
46.82
Avg. 48.74

14Sep94 12Aug94
13Oct94 14Sep94
14Nov94 13Oct94
15Dec94 14Nov94

33
29
32
31

20509.
18173.
19655.
17532.

20509.
18173.
19655.
19532.*

1420.49
1366.75
1465.78
1348.94

0.0693
0.0752
0.0746
0.0769

43.04
47.13
45.81
43.51
Avg. 44.87

* 1. The AB Harrington #2 well was shut in for two weeks in early December
2. Data summed is from wells where counterbalance weights were moved (five wells adjusted, seven were not)
3. Work was done in August 1994 to re-balance the counterweights

on wells with oversized motors can improve the power factor.

Changing from the 440 volt taps to the 762 volt taps, when
available, was attempted on some high slip motors to determine
potential savings by going to this particular winding rather than
the highest or lowest torque mode on the electric motor. The
results of this effort may explain part of the increase in the electrical bill at Waddell, since after the work was done, we noticed an
increased electric bill.

Future Work
Most of the wells in West Texas are sized for larger than actual
pumping conditions. When the waterflood hits we have to be
able to pump it off! was a common statement years ago when
selecting both motor size as well as pumping unit size.
Consequently, there are many pumping units in West Texas that
are over-designed both for lifting capacity and motor size.
As a result of this work, an evaluation of the economics of
downsizing motors has been initiated. The approach is to carefully
measure the pre-KWH consumption and KW demand prior to
downsizing the motor, make the motor change, and then carefully
measure the results of this work. A well was selected in the
Russell Ranch field to do this experiment. A multi-channel electrical meter was installed on the well prior to downsizing the motor,
to determine the KWH per month, KW demand, and power factor.
A 72 hp unit is working on the well. The POWER software indicated a 30 hp motor would be sufficient to produce the well.
Once the results of this well are known, the plan was initiated
to select a small group of candidate wells for motor downsizing
4

and expand the effort to determine economics of a single update

motor change versus a group motor change.
Another question resulting from this work is the frequency of
re-surveying the field with the POWER software. The Waddell
field was resurveyed approximately six months after the initial
survey. During this time, the water injection system was out of
service periodically. This resulted in lower bottomhole pressures
which would necessarily require additional power to lift the liquid
from a lower depth. This probably affected the data somewhat and
it cannot be concluded that the correct time to resurvey the field is
every six months.
A third area of further investigation is positioning of the counterbalance weights. Traditionally, most operators position the
counterbalance weights to be equally balanced on the crank arms.
However, on several wells, the counterweights were positioned
further out on the crank on the side of the crank which was leading. This would result in slinging. An attempt was also made to
smooth torque requirements during the stroke by having only
leading or lagging weights on the crank arm. This shifts the phase
of the weights in relationship to the load. Sufficient data was not
obtained to make recommendations about using lagging or leading
weights only but this procedure should be studied further in an
attempt to obtain more constant torque loadings throughout the
stroke. This may result in some potential savings.

Conclusions
The conclusion of these investigations are:
A. The utilization of the POWER software can help reduce
electrical costs.
Journal of Canadian Petroleum Technology

Proper balancing of the pumping units reduced the electrical

operating costs in the fields we examined. Better balancing
of the pumping unit adds to the longevity of the unit. The
smoother operation of the pumping unit also decreases the
possibility of damaging equipment because of overloading.
B. An additional saving is available when re-balancing counterweights due to decrease in KW demand.
The KW demand portion of the electric bill can be significant. This was a pleasant surprise to us when we observed
the 10% reduction in KW demand. In the Crane area, this
helped increase the cost savings of the monthly electrical
bill, far exceeding our original cost savings estimates. KW
demand should be closely examined to try to reduce it as
much as possible.
C. Direction of rotation should be checked for each well, since
there may be electrical savings available by changing
direction.
Changing direction of rotation on pumping units is simple,
and may result in cost savings available. In some cases a
reduction of peak torque is achieved by changing rotation.
The cost savings were not great, but the torque reduction in
one direction was reduced and would better load the unit for
torque in both directions. This will improve the life of the
pumping unit.
D. There will not be a huge saving in electrical costs, but savings of up to 10% have been observed on smaller fields.
The POWER software will help reduce electrical costs.
However, it will not be a 25% savings, but more likely savings of up to 10 12% are available. For a small lease these
savings may not be large. However, most small leases have
higher consumption charges and the wells would be worth
examining to note any potential savings with the POWER
software. For large fields, we have observed potential savings and are surveying all of Mobil operations with the
POWER software in West Texas.

Acknowledgement
We would like to thank Mobil E & P U.S. for permission to
publish this paper. Thanks also are extended to Larry Logan and
James Wolf of Mobil for conducting many of the surveys.

Authors Biographies
James McCoy graduated from the
University of Oklahoma with a B.S. in
petroleum engineering and from Penn State
University with a M.S. in petroleum and
natural gas engineering. He is president of
Echometer Company, is a member of the
Petroleum Society and the Society of
Petroleum Engineers, and was the 1994
recipient of the Southwestern Petroleum
Short Course Slonneger Award.
Bill Drake graduated from the University
of Texas at Arlington with a B.S. in electrical engineering. He is an engineer for
Echometer Company and a member of the
Society of Petroleum Engineers.

R.E. Ott photograph and biography unavailable.

F. Collier photograph and biography unavailable.
A.L. Podio is holder of the H.B. (Burt)
Harkins Jr. Professorship in Petroleum
Engineering at the University of Texas at
Austin. His teaching and research interests
are primarily in the areas of drilling and
production and include studies of multiphase flow in pipes, efficiency of beam
pumping multi-phase flow metering, well
pressure control in drilling and applications
of sonics and ultrasonics to the solution of
problems related to hydrocarbon production systems.

REFERENCES
1. MCCOY, J.N., COLLIER, F.B., and PODIO, A.L., Application of
Real-time Measurement of Motor Power to Determination of Beam
Pump Efficiency; Southwestern Petroleum Short Course, Lubbock,
TX, 1994.
2. MCCOY, J.N., PODIO, A.L., JENNINGS, J., and DRAKE B.,
Motor Power, Current and Torque Analysis to Improve Efficiency of
Beam Pumps; Southwestern Short Course, Lubbock, TX, 1993.

ProvenanceOriginal Petroleum Society manuscript, Beam

Pump Balancing Based on Motor Power Utilization, (95-25),
first presented at the 46th Annual Technical Meeting, May 14-17,
1995, in Banff, Alberta. Abstract submitted for review September
20, 1994; editorial comments sent to the author(s) May 25, 1998;
revised manuscript received May 7, 1998; paper approved for prepress June 22, 1999; final approval November 8, 1999.M

Special Edition 1999, Volume 38, No. 13