Nathan Mayre and Dylan Bucanek
Apergy Corporation
Sucker rod pump or “rod pump” is a common method of artificial lift for oil and gas wells in the United States. For decades well analysts and production engineers have looked at surface and downhole dynamometer cards to diagnose various downhole and surface equipment issues alike. In more recent years, helpful rod pump diagnostic tools have aided well analysts and production engineers in training and the analysis of downhole dynamometers utilizing generalized libraries with known behavior for downhole dynamometer cards. Unfortunately, the same generalized libraries do not exist for surface dynamometer cards limiting these tools to base their diagnostics solely on information captured in the downhole dynamometer card. Although a majority of data used for analytics and diagnostics can be found in the downhole dynamometer card, it has been known for years that still more helpful analysis can be done utilizing data and patterns found in the surface dynamometer card. Recently, strides have been made in software tools to analyze data and patterns not only found in downhole dynamometer cards, but also the surface dynamometer card. It has been well known within groups with expertise on dynamometer card analysis that pump tagging and shallow friction can be seen more obviously in the surface dynamometer card than the downhole dynamometer card. Mimicking the thought process of these experts, algorithms leveraging data science tools and statistical methods have been implemented in diagnostic software tools that can better detect both shallow friction and pump tagging problems that can be seen in the surface dynamometer card well before they are seen in the downhole dynamometer card, especially for deep wells. These new algorithms will be yet another tool in the continual aid of well analysts and production engineers to more quickly and effectively analyze dynamometer cards and optimize production for the sucker rod pumping system. Although current downhole analytical software provides great benefits to users, including these algorithms allows for a more robust and effective dynamometer card analysis and diagnostics software.