SPE 90153

PCP Axial Load: Theory and Lab Results

F.J.S. Alhanati, SPE, C-FER Technologies and P. Skoczylas, SPE, C-FER Technologies

Copyright 2004, Society of Petroleum Engineers Inc.

Calculation of Pump Axial Load
This paper was prepared for presentation at the SPE Annual Technical Conference and The axial load generated at the pump is normally estimated
Exhibition held in Houston, Texas, U.S.A., 2629 September 2004.
based on the pressure forces acting on the pump rotor, which
This paper was selected for presentation by an SPE Program Committee following review of
information contained in a proposal submitted by the author(s). Contents of the paper, as
is connected to the rod string, as illustrated in Figure 1. Note
presented, have not been reviewed by the Society of Petroleum Engineers and are subject to that, for a downhole-driven PCP, the formulation would be
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 slightly different.
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
The pump axial load is therefore determined by
for commercial purposes without the written consent of the Society of Petroleum Engineers is multiplying the differential pressure across the pump by an
prohibited. Permission to reproduce in print is restricted to a proposal of not more than 300
words; illustrations may not be copied. The proposal must contain conspicuous effective pump cross sectional area, and applying a small
acknowledgment of where and by whom the paper was presented. Write Librarian, SPE, P.O. correction to account for the cross sectional area of the rod
Box 833836, Richardson, TX 75083-3836, U.S.A., fax 01-972-952-9435.
string:
Abstract F = P Aeff Pd Arod [Eq. 1]
Accurate estimation of pump axial load in Progressing Cavity where
Pump (PCP) applications is an important aspect of the
installation design. It affects the calculation of rod stresses, P = Pd Pi [Eq. 2]
and therefore the determination of the rod string torque
The geometry of a PCP pump is somehow complex; so, it
capacity. In deviated wells, it also directly affects the
is not immediately obvious what value for the cross-sectional
calculated rod-tubing contact loads, and therefore the
area (Aeff) area one should use. A few approaches have been
evaluation of wear related issues.
suggested in the literature1,2,3, and the formulas most
This paper discusses the current theory related to the
commonly used in the industry, for single-lobe pumps, are
calculation of axial load in PCP applications, in light of
listed below.
laboratory test results obtained with 25 different pump models.
The axial load is normally estimated by multiplying the
Stator Cavity Area. This is the cross-sectional area of the
differential pressure across the pump by an effective pump
stator cavity when the rotor is not present:
cross sectional area. However, given the complexity of the
pump geometry, it is not straightforward how to properly Aeff = 4 D 2 + 4eD [Eq. 3]
calculate this effective area. In the paper, it is shown that
intuitive choices, such as the circle defined by the rotor minor Rotor Minor Diameter Area. This is the cross-sectional area
diameter, the circle defined by the rotor major diameter, or the of the rotor:
pump cavity area, do not yield adequate results.
A new method is suggested that not only matches quite Aeff = 4 D 2 [Eq. 4]
well the extensive laboratory data for single-lobe pumps but
also matches reasonably well the limited data available for Rotor Major Diameter Area. This is the area of a circle
multi-lobe pumps. having a diameter equal to the major diameter of the rotor:

Aeff = 4 (D + 2e )
2
[Eq. 5]
Introduction
In a PCP application, the dominant load affecting stresses in Area Open to Flow. This is the area open to flow when the
the rod string is torque. Axial load does exist, however, both rotor is installed inside the stator:
due to rod weight and due to a load generated at the downhole
pump, and it can be significant in many cases. Other design Aeff = 4eD [Eq. 6]
considerations affected by the rod string axial load are pump
space-out and, in deviated wells, rod-tubing wear and rod Circumscribed Cavity Area. This is the area of the smallest
fatigue life. It is therefore quite important to be able to circle which can be drawn completely outside the entire stator
accurately calculate the axial load in many PCP applications. cavity:

Aeff = 4 (D + 4e )
2
[Eq. 7]
2 SPE 90153
Note that calculation of these areas requires knowledge of
the detailed geometry of the pump, namely the rotor minor
diameter (D) and the eccentricity (e), and these values are not
readily available from the manufacturers catalogs.
Later in the paper, it is shown that most of these intuitive
choices do not yield adequate results. In fact, only perhaps the
least intuitive one, the circumscribed cavity area, does.
New Method
When investigating this topic, the authors searched not only
for a formulation that would yield adequate results, but also
for a formulation that could be justified on a physical basis.
The proposed new formulation is:
Aeff = N S AFlow + ARotor [Eq. 8]
For a single lobe pump (Ns=2), this yields:
Aeff = 8eD + 4 D 2 [Eq. 9]
The physical justification for this formulation is that it is
the differential pressure across the cavities that results in the
pump axial load. In a PCP, the total differential pressure is
distributed through all the cavities, but there are as many trains
of cavities as the number of lobes in the stator. Therefore the
pump pressure differential would apply to the area open to
flow, but as many times as the number of lobes in the stator.
In addition, the pump differential pressure also applies to the
cross sectional area of the rotor.
Later in the paper, it is shown that this new formula
matches well laboratory data not only for single-lobe pumps,
but also for multi-lobe pumps.
Laboratory Test Program
Unlike beam pump aplications, a good measurement of axial
load cannot be easily obtained in a PCP application.
Therefore, to the authors knowledge, there has been little
verification of the formulations being used to calculate axial
loads to date.
As part of a Joint Industry Project (JIP), a laboratory study
was undertaken at C-FER Technologies laboratories in
Edmonton, Alberta, Canada, in 1995 and 1996, to study the
performance of PCPs.
Twenty-five pumps from different manufacturers were
installed in a test well, and run at several levels of speed,
intake and discharge pressure.
One of the many pump performance parameters measured
was axial load, using a load cell specifically designed and
constructed for this test program.
Figure 2 shows the test well, with the wellhead installed.
Also visible in the figure (at the top of the drivehead,
immediately below the polish rod clamp) is the load cell
which measured torque and axial load.
Table 1 lists the pumps which were tested in the study.
(Note that the test program took place in 1995-1996, and many
of these pumps may no longer be available.) Only one of
these pumps (#23) was a multilobe pump (with a 2:3 lobe
configuration); all the others were single-lobe pumps.
Table 2 lists the levels of speed, intake and discharge
pressure at which these pumps were tested. Note that not
every pump was tested at all combinations of intake and
discharge pressure, for three reasons: the rated differential
pressure of the pump was not to be exceeded; in some cases
(high speeds, high pressures, larger displacement pumps), the
power requirements exceeded the test systems limits; and in
some cases the pump was not able to produice any fluid at
higher differential pressures (zero efficiency). All tests were
conducted with water.
In the test well, the pump was in a vertical position and
just one tubing joint below the wellhead. Measured
independent parameters included flow rate, torque, and axial
load; however, only axial load results are reported here.
Test Results
Figure 3 shows the measured vs. calculated axial load for all
the tests conducted with single lobe pumps in water, and using
different effective area formulatins.
It is evident from the figure that the calculated values for
most pump effective area formulations are too small. Only the
Circumscribed Cavity Area and the New Area yield
acceptable results. Figure 4 plots the errors (calculated force
measured force) for each formulation.
Figure 5 is similar, except that it shows a relative error
(percent of measured). Note that some data points were
excluded from Figure 5 (while still present in Figures 3 and 4).
This corresponds to data points where the calculated force was
less than zero or the measured force was less than 5 kN. This
was done because small values of force can lead to very large
relative errors, even for very small absolute errors. This is
especially true for the Open Flow Area and Minor Diameter
Area, for which many points have a calculated force less than
zero (as can be seen in Figure 3.)
Figures 4 and 5 again show that only the the
Circumscribed Cavity Area and the New Area give reasonable
results, both in terms of having a small mean error and a small
standard deviation. Of these two, the Circumscribed Cavity
Area has a smaller mean error, but the New Area has a smaller
standard deviation (both in absolute and relative errors).
Multilobe Pumps.
As described above, the new method for calculating the pump
effective area was developed considering both single and
multi lobe pumps. Therefore, if the physical basis on which it
was developed was correct, it should also yield reasonable
results for for the 2:3 multilobe pumps. Figure 6 shows the
measured vs. calculated axial load for all the tests conducted
with this multilobe pump. The errors are plotted in Figure 7.
While the circumscribed cavity area provides a slightly
more accurate result for single lobe pumps, the New Area, or
Ns Flow Area + Rotor Area (Ns representing the number
of stator lobes) yielded better results for multilobe pumps, and
can perhaps be considered a better general equation.
Conclusion
1. The effective area for use in calculating the axial load on
the rotor of a progressing cavity pump subjected to a
differential pressure is not immediately obvious.
2. Various areas have been suggested in the past, but many
intuitive choices do not yield adequate results, when compared
to lab test data.
SPE 90153 3

3. A new method is suggested, that not only matches quite

well extensive laboratory data for single lobe pumps, but also
matches reasonably well limited data available for multi-lobe
pumps.

Nomenclature
Aeff = effective cross-sectional area for the pump (m)
Aflow = area of the stator cavity open to flow (m)
Arod = cross-sectional area of the rod at the top of the pump
rotor (m)
Arotor = rotor cross-sectional area (m)
D= rotor minor diameter (m)
e= rotor eccentricity (m)
F = axial load (N)
Pd = pump discharge pressure (Pa)
Pi = pump intake pressure (Pa)
P = differential pressure (Pa)

Acknowledgements
The authors would like to acknowledge the Participants of
C-FERs Progressing Cavity Pumping System Technology
Development JIP, for their contributions to the test program
reported in this paper.

References
1. Cholet, H.: Progressing Cavity Pumps, Editions Technip,
Paris (1997).
2. Presber, T.: Lifting Heavy Oil with Screw Pumps,
presented at the 2nd AOSTRA Can.-China Heavy Oil
Technical Symposium, Beijing, China, Oct. 29-Nov. 1,
1990.
3. Klein, S.T. et al: Well Optimization Package for
Progressive Cavity Pumping Systems, paper SPE 52162
presented at the 1999 SPE Mid-Continent Operations
Symposium, Oklahoma City, March 28-31.
4 SPE 90153

Pump Pump Pump Nominal Pressure

# Manufactur Model Displacement Rating
er (m/d/RPM) (kPa)
1 BMW Pump 28-1200 0.28 11760
2 Rotalift- 12-H-10 0.1 11775
Corod
3 Griffin- 40-B-095 0.151 11968
Legrand
4 Rotalift- 12-E-17 0.17 11772
Corod
5 BMW Pump 28-1200 0.28 11772
(HS)
6 Rotalift- 12-N-27 0.27 11772
Corod
7 Rotalift- 12-N-27 0.27 11772
Corod
8 Rotalift- 12-N-27 0.27 11772
Corod
9 Griffin- 40-A-195 0.31 11968
Legrand
10 Griffin- 40-A-195 0.31 11968
Legrand
11 National 14-1200 0.35 9663
Oilwell
12 National 14-1200 0.35 9663
Oilwell
13 BMW Pump 83-600 0.83 5886
14 BMW Pump 120-600 1.2 5886 Figure 1. Pressure Forces on PCP
15 Rotalift- 8-H-75 0.75 7848
Corod
16 Griffin- 30-A-400 0.636 8927
Legrand
17 National 18-1500 0.48 12429
Oilwell
18 Rotalift- 5-H-110 1.1 4905
Corod
19 Griffin- 30-600 0.95 8927
Legrand
20 National 14-2100 0.626 9663
Oilwell
21 Kudu 400TP900 0.806 8829
Industries
22 Kudu 600TP900 1.2 8829
Industries
23 Kudu 900TP1000 1.68 9810
Industries
24 Kudu 200TP1200 0.392 11772
Industries
25 Kudu 200TP1200 0.392 11772
Industries
Table 1. Pumps Tested in Study

Test Speeds
75, 100, 150, 200 RPM Pumps 1, 2, 3, 4, 6, 7, 8
100,200,300,400 RPM Pumps 5, 9, 10, 11, 12, 13, 14, 15, 16,
17, 18, 19, 20, 21, 22, 24, 25
50,100,150,200 RPM Pump 23
Intake Pressures
200, 2000, 4000, 8000 kPa
Discharge Pressures
200, 1000, 2000, 3000, 4000, 5000, 6000, 7000, 8000, 9000,
10000, 11000, 12000 kPa
Table 2. Test Conditions

Figure 2. Test Well

 SPE 90153 5

40

35

30

25
Measured Force (kN)

20

15

10

0
Cavity Area
Circumscribed Cavity Area
5 Minor Diameter Area
Major Diameter Area
Open Area
New Area
10
10 0 10 20 30 40 50
P x Area P x A (kN)
d rod
Figure 3. Measured vs. Calculated Axial Load

New Area New Area

Open Area Open Area

Effective Area Type

Effective Area Type

Major Diameter Area Major Diameter Area

Minor Diameter Area Minor Diameter Area

Circumscribed Cavity Area Circumscribed Cavity Area

Cavity Area Cavity Area

40 30 20 10 0 10 100 50 0 50
Error (kN) Percent Error
Figure 4. Error in Calculated Axial Load Figure 5. Percent Error in Calculated Axial Load
 6 SPE 90153

20

15
Measured Axial Load, kN

10

Circumscribed Cavity Area

3 x Flow Area + Rotor Area
5
10 5 0 5 10 15 20
Calculated Axial Load, kN
Figure 6. Measured vs. Calculated Axial Load for Multilobe PCP

3 x Flow Area + Rotor Area

Circumscribed Cavity Area

5 4 3 2 1 0 1
Error in Calculated Load, kN
Figure 7. Error in Calculated Load for Multilobe PCP