SPE-170588-MS

The Ultimate Sucker-Rod String Design Procedure

Gabor Takacs, PhD, and Mihaly Gajda, Jr, University of Miskolc. Hungary

Copyright 2014, Society of Petroleum Engineers

This paper was prepared for presentation at the SPE Annual Technical Conference and Exhibition held in Amsterdam, The Netherlands, 2729 October 2014.

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 have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material does not necessarily reflect
any position of the Society of Petroleum Engineers, its officers, or members. Electronic reproduction, distribution, or storage of any part of this paper 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 SPE copyright.

Abstract
A properly designed sucker-rod string should provide failure-free pumping operations for an extended
period. Improper design of rod tapers can lead to early mechanical failures (rod breaks) with a complete
termination of pumping action and an inevitable loss of production. Because of its prime importance in
sucker-rod pumping technology several rod string design procedures based on different assumptions were
developed in the past. Since most sucker rod breaks are fatigue failures the mechanical design of
sucker-rod string must consider the cyclic nature of rod loading and the appropriate fatigue endurance
limit of the rod material. This is why most string designs utilize the modified Goodman diagram for
calculating the lengths of rod tapers.
Available sucker-rod string design models calculate rod taper lengths that ensure proper operation
without premature fatigue failures. Their common design problems are (a) defining the principle of taper
length determination, and (b) calculating the true mechanical stresses along the string. The universally
accepted principle of taper length calculations is to provide the same level of safety against fatigue failures
in each taper section. Mechanical loads and stresses, on the other hand, are usually found in most of the
design procedures from very approximate formulas. These loads, therefore, can greatly deviate from the
true mechanical loads that would be measured in the rod string run in the well. The paper details the main
features of available rod string designs and discusses their main characteristics; it provides a thorough
comparison of designs involving the calculation of loads and stresses predicted from the solution of the
damped wave equation. Using a predictive analysis program rod stresses are calculated that, plotted on the
modified Goodman diagram, provide a proper comparison of the merits of the different rod string design
methods.
Rod strings designed by available design procedures usually do not have identical safety included in
the different tapers. This is due to the improper calculation of rod string loads that form the basis of
calculating rod taper percentages. Based on the evaluation of the different available rod string designs the
paper introduces a novel procedure that estimates rod loads from the predictive solution of the damped
wave equation during the design process. Since loads calculated that way simulate actual loads very
accurately, the most important limitation of previous rod string design procedures is eliminated. Strings
designed using the proposed model, therefore, will have identical safeties against fatigue failures in each
taper; the new design model provides the ultimate safety over previous designs.
2 SPE-170588-MS

Introduction
The sucker-rod string is a very peculiar piece of man-made structures because its maximum diameter
(about one inch) is absolutely negligible as compared to its length of several thousand feet making it an
absolute slender bar. The weight of the string is distributed along its length and any section has to carry
at least the weight of all the rods below it. This fact suggests that the ideal shape would be an inverted
cone, continuously tapering from top to bottom. Since such a rod string is impossible to manufacture one
tries to approach the ideal shape by designing tapered strings with sections of increasing diameters toward
the surface. For shallow wells, straight rod strings made up from one rod size only are also used but deeper
wells inevitably require the application of tapered strings.
The mechanical design of tapered rod strings is highly complicated because of the type of loading the
rods are subjected to. Investigation of the possible loads that occur during the pumping cycle shows that
the following distributed and concentrated loads act on the rod string:
Weight of rods in air; it is a distributed load along the string.
Buoyancy forces oppose the rod weight and are the result of the immersion of the rods into the
produced liquid.
Fluid load on the plunger of the downhole pump is a concentrated force acting during the upstroke
only.
Dynamic loads are the results of changes in acceleration of the moving masses (rods, fluid
column).
Frictional forces are: (1) fluid friction between the rods and the produced liquid, and (2)
mechanical friction between the rods and the tubing string.
After considering the variation of these forces during a complete pumping cycle, one can easily
conclude that the rod string is exposed to a cyclic mechanical loading. During the downstroke, the string
carries the buoyant weight of the rods minus dynamic and friction forces only, while on the upstroke it
also carries the load of the fluid lifted. Mechanical stresses follow the variation of rod loads and are cyclic,
too; they are typically tension stresses with the tension level considerably increasing during the upstroke
as compared to the downstroke; the loading of the rod string can be classified as pulsating tension.
The typical failure mechanism of sucker-rod strings is the consequence of the type of loading and is
mostly fatigue failure. Fatigue failures occur at much lower levels of mechanical stresses than the tensile
strength or even the yield point of the material and are caused by the extremely high number of repetitions
of the variable loads. This type of failure is absolutely different from tensile (overpull) failures and is the
root cause of the great majority of rod string breaks. Therefore, rod string design procedures must
inevitably take into account the cyclic nature of rod loading; this is why, in order to ensure a sufficiently
long service life, the string has to be designed for fatigue endurance.

Available rod string design models

Fatigue endurance of rod materials
The maximum stress allowed in sucker-rod materials a.k.a. the fatigue endurance limit that ensures a
failure-free operation for a sufficiently high number of cycles (usually 10 million) under pulsating tension
loads typical for pumping operations is calculated from the modified Goodman formula [1]. This
formula, in a generalized form, is valid for different available rod materials and shows that the fatigue
endurance limit varies with the minimum stress that occurs in the rod and the tensile strength of the steel
material.
1
SPE-170588-MS 3

The constants A and B featuring in the Goodman Table 1Material parameters used in the Goodman formula.

formula were empirically determined for different Rod Type T A B

steel materials using experiments on material sam- psi - -
ples. Their values are listed, along with the mini-
API Grade K 90,000 4 0.5625
mum tensile strengths of available rods in Table 1. API Grade C 90,000 4 0.5625
The variable SF represents the effects of the envi- API Grade D 115,000 4 0.5625
ronment where the rod string is operating and can be High Strength (HS) rods 140,000 2.8 0.3750
considered as the inverse of a safety factor; it is Tenaris rods 125,000 2.3 0.3750
mainly used to allow for the corrosivity of the well
fluid and is usually held constant in a given field.
Overview of string design procedures
Early rod string design methods utilized the simplifying assumption that the string was exposed to a static
tension loading. Their goal was to keep the maximum rod stresses below a percentage of the tensile
strength or the yield point of the rod material. Several different procedures were developed to design
tapered rod strings; the one proposed in the Bethlehem Handbook [2] and later adopted by the API in
the earlier editions of RP 11L [3] gained wide acceptance. This is called a balanced design based on
static loading conditions and sets the maximum stresses at the top of each section equal.
The simple design procedures gave reasonable rod life in shallower wells, but as rod stresses became
higher in deeper wells it became inadequate and overloading of the lower rod sections was experienced,
as observed by Eickmeier [4] as early as in 1967. Rod breaks were occurring at operating stresses well
below the tensile strength or the yield strength of the rod material. Such breaks are typical fatigue failures;
this finding caused designers to realize that a proper rod string design must account for the cyclic nature
of rod loading. This is why the string has to be designed for fatigue endurance, as done in most of the
present-day procedures.
The first tapered rod string design procedure specifically developed for fatigue loading was proposed
by West [5, 6]. West developed a taper design that maintains the same amount of safety for every taper
section. The objective of the design is to have the same ratio of maximum stress to allowable stress
(represented by the Service Factor used) in each taper. Rod strings designed this way will have the same
safety factor included in every taper and will not have any weak points.
The design model proposed by Neely [7] introduced the concept of modified stress and aimed at
reaching the same modified stress values at the top of each taper section. His design method was adopted
by the American Petroleum Institute in 1976 and taper percentages calculated from his procedure were
published in the later editions of API RP 11L [3]. The use of these tables eliminates the time necessary
for detailed design calculations but, as proved by Gault [8] and Takacs [9], the taper percentages
presented for different API taper combinations are the sole function of the pump size with the other
operating parameters being assumed but not shown in the publication. Therefore, rod designs for operating
conditions different from those assumed in the RP 11L tables will inevitably result in significantly
different taper percentages; the use of the original Neely procedure is recommended instead.
The latest design procedure proposed by Gault-Takacs [10] is the only one including the effects of the
reflections of force waves in the rod string when calculating rod loads. These dynamic loads are
distributed along the string in proportion to the mass being moved. The goal of the design is to have the
same degree of safety in every taper section. Service factors for all tapers will be identical, and rods will
be subjected to a uniform level of fatigue loading all along the string.
Evaluation of the design procedures
Basic Features
Sucker-rod string design faces two basic problems when trying to find the lengths of the tapers in the
string:
4 SPE-170588-MS

Table 2Basic features of available rod string design procedures.

Model Year Min. Load Max. Load Dyn. Loads Design Goal

Bethlehem 1953 - Fluid load plus rod weight in air - Equal max. stresses
West 1973 Rod weight in air Fluid load plus rod weight in air Mills acceleration factor SF const.
plus dynamic loads
Neely 1976 Buoyant rod Weight Fluid load plus buoyant rod weight Special formula Equal modified stresses
plus dynamic loads
Gault-Takacs 1990 Buoyant rod Weight Fluid load plus buoyant rod weight From RP 11L SF const.
plus dynamic loads

how to calculate rod loads during the pumping cycle, and

what principle to use for the determination of taper lengths.
The main features of available rod string design procedures are collected in Table 2 that contains
information on the different ways authors estimate rod loads and on the basic design goal of each model.
The early design proposed in the Bethlehem Handbook disregards minimum and dynamic loads and
does not consider buoyancy effects. West, too, ignores buoyancy forces to compensate for friction forces
that are usually unknown but tend to act opposite the buoyant force. Although he considers dynamic loads
but uses the obsolete Mills acceleration factor method which gives reasonable predicted loads for small
pumps and medium pumping depths only. Neelys approach to load calculations includes buoyancy as
well as dynamic loads for which he recommends an empirical correlation. The Gault-Takacs model is the
only one that tries to include the effect of the force wave reflections in the rod string when calculating rod
loads and finds the surface dynamic loads from RP 11L (now API TL 11L) calculations.
Investigation of the design goals of the different models reveals huge differences in basic principles.
The Bethlehem model designs strings with the same maximum stresses at the top of each taper. This
means that the Smin - Smax points plotted on the Modified Goodman diagram [1] for the different taper
sections will fall on a horizontal line which must inevitably cross several SF constant lines. Lower
tapers (with lower minimum stresses) have a higher service factor and a consequently reduced safety than
the tapers higher up the string. The fatigue loading on lower tapers, therefore, is higher and these sections
are more likely to experience premature failure.
Setting the service factors (SFs) equal in each taper, as done by West and Gault-Takacs, ensures the
same amount of safety for every taper section. Rod strings designed this way have the same safety factor
included in every taper and do not have any weak points.
Finally, Neely defines a modified stress and forces it to be equal at the top of each taper:
2

Setting the modified stresses equal in each taper means that Smin - Smax points belonging to the
different tapers, when plotted on the modified Goodman diagram (MGD), will lie on a parallel to the SF
1 line. This line, however, inevitably crosses the lines corresponding to any service factor other than
unity. Therefore, the design generates different safety factors for each taper; upper tapers are relatively
more loaded than lower ones. This situation is just the opposite of early design methods where the lower
tapers were under-designed.

Sample designs
In order to evaluate the available rod string design models an example case of a three-taper rod string is
presented. Well parameters are given here; calculated taper lengths found from each design model are
given in Table 3.
SPE-170588-MS 5

Table 3Sample rod string designs.

Pump Setting Depth 6,000 ft Plunger Size 2 in
PR Stroke Length 120 in Pumping Speed 6 SPM Model taper 7/8 taper 1 taper
Rod Grade API D Service Factor 0.9
Bethlehem 2,364 ft 1,853 ft 1,783 ft
API Rod Code 86 Liquid Sp. Gr. 1.0
West 1,469 ft 2,400 ft 2,131 ft
Neely 1,543 ft 2,254 ft 2,203 ft
Gault Takacs 1,774 ft 2,239 ft 1,987 ft
Comparison of the four rod string designs is
done on the modified Goodman diagram in Fig. 1
where rod stresses estimated according to each mo-
dels basic assumptions are plotted. Fig. 1 presents a non-dimensional form of the modified Goodman
diagram (MGD) and contains Smin - Smax points for each taper. As seen, all four design methods produced
strings that can safely handle the estimated well loads as indicated by the fact that all maximum stresses
are below the fatigue endurance limits of the rod material for SF 0.9. The fatigue loading of the
individual tapers, however, is different for the four designs. The models West and Gault-Takacs have
their three points belonging to the three tapers located on SF constant lines; these strings are, therefore,
uniformly loaded and have tapers designed with the same safety against fatigue failure. The string
designed by Neelys procedure is not uniformly loaded because the three tapers have different service
factors. The top taper (represented by the rightmost point) is relatively more loaded and is bound to fail
first. The Bethlehem model results in even greater differences in the fatigue loading of the different
tapers. As shown in Fig. 1 maximum rod stresses are about the same and the three points belonging to the
tapers lie nearly on a horizontal. This means that the bottom taper (corresponding to the leftmost point)
is much more loaded than the upper ones. This fact was proved in the oil fields when in the late 1960s
rod strings designed according to this model failed almost invariably at the bottom taper. [4]
So far all loads and stresses were calculated according to each design procedures basic assumptions,
as detailed in Table 2. These loads, however, can only be considered as estimates for the given pumping
conditions. In order to exactly compare the features of the different rod string designs, one would have
to actually measure the loads occurring in the different strings just designed. Since this would be close to
impossible to do, this is, of course, not a viable approach. The best possible solution of the problem
involves the calculation of rod loads from the solution of the damped wave equation written on the rod
string. This approach is justified by experience gained since the introduction of the wave equation in the
late 1960s by Gibbs [11] proving that the predictive solution of the damped wave equation gives loads
that very closely match measurements. Since mechanical stresses in the rods must be calculated from
true loads the solution must include the effects of buoyant forces that occur on the different tapers. [12]
Therefore, the rod strings designed by different procedures will be compared in the following on the basis
of loads and stresses calculated using predictive analysis techniques.
The rod strings designed for the sample case were used to find the distribution of rod loads during the
up-, and the downstroke using the predictive solution of the damped wave equation. For the calculations
we assumed a conventional pumping unit, a vertical well, average damping factors, Grade D rods, and
pumping water in a pump-off condition. Minimum and maximum pumping loads were determined at the
top of each rod taper that allowed the calculation of mechanical stresses at the top sections of the tapers.
These stresses, plotted on the modified Goodman diagram (MGD) define the fatigue loading of the strings
and allow one to derive important conclusions on the merits of the different procedures.
Fig. 2 contains a comparison of the four design procedures for the example case based on the predicted
rod stresses and plotted on the modified Goodman diagram. Rod stresses, compared to those shown in Fig.
1, have considerably changed; the general trend is that minimum stresses became lower while maximum
stresses became higher. This is a clear indication that predicted dynamic loads are higher than those
assumed by the design procedures because dynamic loads tend to alter rod loads that way. This effect is
6 SPE-170588-MS

most pronounced for the Bethlehem model where

the bottom tapers fatigue loading has extremely
increased.
Comparison of Fig. 1 and Fig. 2 makes it clear
that the loads and stresses estimated by the different
design models do not represent actual pumping con-
ditions. This means that the mechanical stresses in
any string, when run in the well will be different
from those computed during the design process.
Since actual stresses at the top of rod tapers are
different from their designed values the fatigue
loading represented by the relevant service factors
will also be different. As a consequence, the basic
objective of the design procedure, i.e. having iden-
Figure 1Comparison of sucker rod string designs using estimated
tical service factors at the top of each taper, will not stresses.
be met. This is a general problem with all known
rod string designs that stems from the fact that rod
loads required in the design process are estimated
from approximate formulas only, as proved by
Takacs-Gajda [13]. The use of predicted loads
received from the solution of the damped wave
equation, on the other hand, ensures the highest
possible approximation of measured loads and can
thus give a reliable foundation to further investiga-
tions.
Development of the ultimate rod
string design procedure
Basic considerations
As discussed previously, one faces two basic prob-
lems when designing a sucker-rod string. The first is Figure 2Comparison of sucker rod string designs using predicted
the principle to be used for the determination of stresses.
taper lengths; as detailed before, the proper objec-
tive is to select tapers that have the same level of safety against fatigue failure. To achieve this goal, one
has to use the modified Goodman diagram and select taper lengths so that they have the same service
factor (SF) values. This is the final objective of the design procedure developed in this paper.
The second problem involves the calculation of rod loads during the pumping cycle. If loads in the
different tapers just designed could be measured during the design process then the final rod string would
surely meet the objectives of the design and would have tapers with identical safety against fatigue
failures. In reality, however, this approach is impossible to follow and some other solution must be found
to estimate the loading conditions of the rod string. A feasible answer to this problem is provided by the
calculation of rod loads from the solution of the one-dimensional damped wave equation introduced by
Gibbs [11]. The rod string design procedure developed in this paper, therefore, relies on true
mechanical loads including fluid buoyancy that are predicted from the solution of the wave equation.

Design Fundamentals
The solution of the problem i.e. finding a taper combination with identical service factors is surely an
iterative process. To reach the final rod string design one has to calculate rod loads and stresses in many
SPE-170588-MS 7

cases of assumed or calculated taper combinations.

Therefore it is important to have an idea on the
possible distribution of rod loads and stresses along
a rod string. Fig. 3 presents the distribution of rod
loads with well depth as calculated from the solu-
tion of the wave equation for a 6,070 ft deep well
using an API 86 rod string, 120 in polished rod
stroke and different pumping speeds. The two fam-
ilies of curves represent downstroke (the left-hand
curves) and upstroke (the right-hand curves) loads;
their difference is basically equal to the fluid load
on the plunger. Based on this and similar cases one
can conclude that the variation of maximum and
minimum loads in rod tapers can be fitted in func-
tion of taper length, l, with straight lines as follows:
3

For a given taper in the string, one can deter-

mine, based on the modified Goodman diagram, the Figure 3Distribution of loads along the rod string for a sample case.
safety of the taper against fatigue failure. In princi-
ple, the safety of any taper section against fatigue
failure is defined as the ratio of the actual maximum stress and the allowable stress. Therefore, expressing
the service factor (SF) valid at the top of any taper from the formula describing the modified Goodman
diagram, i.e. Eq. 1, provides a way to indicate the tapers safety level. Introducing rod loads instead of
mechanical stresses into Eq. 1 and solving the formula for the service factor, we receive:
5

Substituting Eqs. 2 and 3 in this formula the following equation is found:

6

This formula represents the relationship between the length of the taper, l, and the service factor, SFi,
which can be calculated from the maximum and minimum stresses along the taper. It will be used in the
design procedure to investigate the effect of changing the length of the taper for a fixed service factor. For
that reason, lets solve Eq. 6 for the taper length, denoted Li:
7

This equation forms the cornerstone of the rod string design procedure developed in this paper because
it allows one to calculate the required length Li of any taper based on the required or assumed service
factor (SF) value. As will be shown later, taper lengths during the iteration process will be changed
according to this formula, while considering the actual values of the variables involved.
8 SPE-170588-MS

Calculation procedure
Although the developed design can handle any
number of tapers, calculation steps are detailed for
the case of a three-taper rod string in the following.
The description of the design procedure is accom-
plished with reference to Fig. 4.
At the initial conditions (iteration number J 0)
the total rod string length is divided into three equal
parts and the following taper lengths are calculated:
8

Taper lengths being known, one can use the

predictive solution of the wave equation to find the
distribution of minimum and maximum loads along
the entire length of the string. The input parameters
to the wave equation are the usual operating param-
eters as plunger size, polished rod stroke length, Figure 4 Illustration of the taper design procedure.

pumping speed, fluid parameters, etc. Similarly to

other string design models, pump-off conditions and
pumping of water are assumed to increase the safety of the design. The main parameters of the actual
pumping unit are needed also for the solution of the wave equation.
Based on the calculated rod loads the minimum and maximum loads in the first taper are fitted with
straight lines, according to Eqs. 3, and 4 and the parameters of the best-fitting lines a1 . . . d1 are found.
Service factors at the top of each taper SF1,0, SF2,0, SF3,0 are calculated according to Eq. 5 and their
average, SF0 is determined.
In the first iteration step (J 1) the length of the bottom taper (i 1) is modified while keeping the
combined length of taper one and two constant. To find the bottom tapers length Eq. 7 is utilized using
the target service factor SF0 and the parameters a1 . . . d1; the expressions to find adjusted taper lengths
are the following:
9

10

11

Using the modified taper lengths, the wave equation is solved again and the distribution of minimum
and maximum loads in taper two is utilized to find the parameters of the best-fitting lines, a2 . . . d2.
The second iteration step (J 2) starts with the adjustment of the length of the second taper and
keeping the length of the first taper. The third taper length is adjusted also and the relevant formulas are:
12

13

14
SPE-170588-MS 9

Table 4 Rod strings designed using the proposed design procedure.

API Grade D HS Rods Tenaris

Rod Size Li SFi Li SFi Li SFi

in ft - ft - ft -

1 1,916 0.842 1,723 0.534 1,708 0.494

7/8 2,112 0.844 1,853 0.533 1,854 0.494
3/4 1,972 0.842 2,424 0.534 2,438 0.494

Now the wave equation is solved again using the

adjusted taper lengths; service factors at the top of
each taper SF1,2, SF2,2, SF3,2 are calculated accord-
ing to Eq. 5 and their average, SF2 is determined.
Since the aim of the design is to reach identical
service factors in each taper, the deviation from the
average, SF2, of the individual SF values is evalu-
ated from the formula:
15

If the error is greater than the required accuracy

previously set then the string received in the second
iteration step must be adjusted further using the
procedure described in iteration step one, and the Figure 5Stresses in rod tapers designed using the proposed design
whole process is repeated. Otherwise, if the error is procedure.
below the required accuracy, the process has con-
verged and the final string design is reached. At this point a final check is made where the final service
factor is compared to its value required in the given field. In case the converged SF is below the accepted
value then the design is final; otherwise the calculations must be repeated using different (a) rod materials,
(b) numbers of tapers, or (c) rod sizes.

Sample design
The developed rod string design procedure was applied to the example problem presented previously
assuming a C-456D-256-120 pumping unit and three different rod materials. Calculated taper lengths
along with the service factors valid in tapers are presented in Table 4. As shown, Grade D rods are quite
heavily loaded since the average value of the service factors is SFavg 0.843. Use of stronger materials
(high-strength rods and rods with Tenaris premium connections) resulted in much lightly loaded strings
with average service factors of SFavg 0.534 and SFavg 0.494, respectively.
The merits of the string design procedure introduced in this paper are easily seen in Fig. 5, where
minimum and maximum rod stresses in the tapers are plotted in a dimensionless form of the modified
Goodman diagram. For every kind of material used the three points belonging to the three tapers fall on
lines representing the average service factors in the string. This proves that all tapers have the same service
factor, consequently the same safety included in their design. In comparison to other designs available
today, rod loads during the whole design procedure were found from the solution of the wave equation.
Since the wave equation predicts rod loads with the highest possible accuracy the design developed in this
paper can be considered as the ultimate tool for designing sucker rod strings.
10 SPE-170588-MS

Conclusions
Based on the detailed evaluation of the available procedures for rod string design and the development of
a new design model the following conclusions were drawn:
Available rod string design procedures estimate rod loads from approximate formulas which result
in loads that are not a true measure of the actual conditions. Since estimated loads are used in the
design procedures, calculated strings do not meet the basic requirement of reaching identical safety
in each taper.
The ultimate rod string design model presented here calculates rod loads from the solution of the
one-dimensional wave equation and, therefore, results in strings with a much enhanced safety
against fatigue failures as compared to previous designs.

Nomenclature
ai . . . di parameters of the best-fitting lines in Eqs. 3 and 4
l taper length, ft
Ai metal area of the rod in the ith taper, sq in
A, B empirical constants specific for the rod material, -
Fmin i minimum load in the ith taper, lbs
Fmax i maximum load in the ith taper, lbs
Li,j length of the ith taper in the jth iteration, ft
Ltotal total required length of the string, ft
Sa fatigue endurance limit (allowable stress), psi
Smax maximum rod stress, psi
Smin minimum rod stress, psi
Smod modified stress, psi
SF service factor, -
T minimum tensile strength of the rod material, psi

Acknowledgement
The authors used data from their previous publications at the 60th and 61st Annual Southwestern
Petroleum Short Course when preparing this paper; permission from SWPSC is gratefully acknowledged.

References
1. Recommended Practice for Care and Handling of Sucker Rods. API RP 11BR, 9th Ed.,
American Petroleum Institute, Dallas, Texas 2008.
2. Sucker Rod Handbook. Bethlehem Steel Company, Bethlehem, Pennsylvania 1953.
3. Design Calculations for Sucker-Rod Pumping Systems (Conventional Units). API TL 11L 5th
Ed. American Petroleum Institute, Washington, D.C., 2008.
4. Eickmeier, J. R.: Diagnostic Analysis of Dynamometer Cards. JPT January 1967, 97106.
5. West, P. A.: Improved Method of Sucker Rod String Design. Proc. 20th Southwestern
Petroleum Short Course, 1973, 15763.
6. West, P. A.: Improving Sucker Rod String Design. PE July 1973, 68 77.
7. Neely, A.B.: Sucker Rod String Design. PE March 1976, 58 66.
8. Gault, R. H.: Rod Stresses from RP 11L Calculations. Proc. 37th Southwestern Petroleum Short
Course, 1990 292301.
9. Takacs, G.: MODERN SUCKER-ROD PUMPING. PennWell Books, Tulsa, Oklahoma, 1993.
SPE-170588-MS 11

10. Gault, R. H. - Takacs, G.: Improved Rod String Taper Design. Paper SPE 20676 presented at
the 65th Annual Technical Conference and Exhibition of the SPE in New Orleans, Louisiana.
September 23-26, 1990.
11. Gibbs, S. G.: Method of Determining Sucker Rod Pump Performance. US Patent 3,343,409
(1967).
12. Lea, J. F. Pattillo, P. D. Studenmund, W. R.: Interpretation of Calculated Forces on Sucker
Rods. SPE PF February 1995, 4145.
13. Takacs, G. Gajda, M.: Critical Evaluation of Sucker Rod String Design Procedures. Proc.
60th Annual Southwestern Petroleum Short Course, 2013, 213222.