Paper: Rod Pump Design Using Personal Computers

Design calculations for sucker rod pumping systems (conventional units) using the API RP 1lL Recommended Practice is now the "standard" for the oil industry. Prior to the RP 1lL publications, the Mills, Marsh, and Coberly formulas were used. It was obvious to many designers that a more accurate method was needed. In 1954 a group of users and manufacturers undertook a study of the complex problems associated with sucker rod pumping. The services of Midwest Research Institute were retained to study these problems. Based on their correlations and test data obtained from an electrical analog study of sucker rod pumping systems, the design procedures outlined in API RP .llL were developed. Since its development, API RP 1lL has become widely used. In the past two years, it has been programmed for use on the personal computers. In general, API RP 1lL gives reasonably close answers on an average for conventional units for the assumed conditions. However, the answer for a specific case can easily exceed 7 percent for loads and 10 percent for peak torques. RP 11L does a good job of predicting the percentage change by altering pump conditions. Improvements that are predicted by the calculations will often be achieved in practice. The API RP 11L method has a number of simplifications and assumptions that must be recognized. The work was based on an "average" conventional unit geometry, running with only medium slip, having complete pump fillage, and having no abnormal dampening or friction. In addition the assumptions were made that fluid acceleration was negligible and that there were no mechanical problems. For simplification, one average set of design curves were developed