Identifying the Timing and Sources of Damage in Gas Storage Wells Using Smart Storage Technology

Abstract

Abstract Most underground gas storage (UGS) wells suffer loss of deliverability over time, often due to unknown causes. Consequently, deliverability maintenance is one of the primary concerns of UGs operators. The authors believe that a significant aid to identifying the potential causes of damage in UGs wells would be an increased knowledge of when the damage develops. specifically, does the damage develop during injection operations, during withdrawal operations, when a change in flow direction occurs, or continuously? Unfortunately, identifying trends in damage versus time using conventional buildup and falloff tests would be operationally challenging and prohibitively expensive. An increasing number of UGS operators are using electronic flow measurement (EFM) systems capable of collecting high-frequency (10 to 15 second intervals) pressure and flow rate data at the wellhead. Over the course of normal storage operations, wells will often be shut in for short periods of time. Wellhead EFM data collected during these shut-in periods can be analyzed as buildup or falloff tests. Consequently, UGs operators having wellhead EFM systems often have sufficient buildup and falloff test data to identify trends in damage over time. Tools have been developed to automatically 1) filter high-frequency EFM data from gas storage wells to identify shut-in periods that can be analyzed as buildup or falloff tests, 2) perform the appropriate pressure transient analysis of the test data so collected, 3) generate semilog and log-log analysis plots, and 4) generate composite graphs of apparent skin s' vs. flow rate and s' vs. date. These tools also allow the user to review and fine-tune the test analyses, and exclude results from any questionable test periods. The system has been used to analyze data from 20 wells in a major gas storage field for the period December 2003 through May 2005. since wellhead pressure and flowrate data are recorded for each well at 15-second intervals, the system had to be able to process large amounts of data. A total of 357 shut-in periods were identified, 207 of which were analyzable. Three wells had no interpretable tests, while eleven of the wells had 8 or more analyzable tests.