A case study of stuck drillpipe problems and development of statistical models to predict the probability of getting stuck and if stuck, the probability of getting free.

摘要

Stuck drillpipe is a major non-productive cost to the oil industry. As much as 35% of all drilling problems are caused by stuck pipe and the cost to industry exceeds {dollar}250 million annually. This work is an effort not only to understand the sticking mechanisms but to make recommendations to reduce their occurrence. Sticking is the single most important factor causing a serious economic impact on the drilling operations around the globe, specially in offshore drilling operations.; The potential cost savings and the need for further understanding are the driving forces behind this research. Stuck drillpipe incidents are broadly categorized as mechanically stuck and differentially stuck. These two categories have different causes; therefore, separate approaches are necessary for the treatment. This is the first such study which treats the two categories separately. In fact, we have introduced a third category of sticking, called "combined sticking."; We analyzed 143 wells drilled offshore Abu Dhabi (U.A.E) during the five-year period from 1989 to 1993 by ADMA-OPCO (Abu Dhabi Marine Operating Company). In this study efforts have been made to cure the stuck pipe problems, to analyze the operations, and to develop techniques to prevent the problems. Operational analysis was done side-by-side with statistical analysis to find out the parameters which are strongly related to differential and mechanical sticking. Based on the statistical analysis, three models have been developed: (i) to predict the probability of getting free, if stuck mechanically; (ii) to predict the probability of getting stuck mechanically; and (iii) to predict the probability of getting stuck differentially.; We developed a probability function called Sharif's Mechanically Stuck Freeing Factor (SMSFF) based on multivariate regression analysis of the data where pipe has already been stuck, that predicts only the chances of freeing mechanically stuck drillpipe, not the chances of getting stuck. A second model, Sharif's Mechanical Sticking Index (SMSI) predicts the probability of getting mechanically stuck to begin with. The third model, Sharif's Differential Sticking Index (SDSI) is based on the data where pipe were differentially stuck versus no-stuck situations 48 hours before and 24 hours after the differential sticking incidents. Use of these models will not only help to remove the uncertainty from the decision making process but will also help to manage the risk to a certain degree by providing degree of stickiness and time factor involved in freeing operations.