Large-Scale Deployment of a Closed-Loop Drilling Optimization System: Implementation and Field Results

Abstract

Summary On a land-based rig, the driller has many responsibilities and is rarely able to adjust the main process controls—weight on bit (WOB) and revolutions per minute (RPM)—on a continuous basis. The result is suboptimal drilling and longer spud-to-target depth (TD) time than necessary. Automatic, closed-loop optimization removes this burden and results in better and more consistent drilling performance. In this paper, a closed-loop drilling optimization system is presented with results from more than 1,700 wells in the field. The closed-loop system builds on industry-proven advisory technology and is designed uniquely for petroleum drilling. Because formation characteristics can change rapidly with depth, a fast optimization algorithm based on input signal dithering is used to continually adjust drilling parameters to search for the highest possible rate of penetration (ROP) and lowest possible mechanical-specific energy (MSE). Drilling dysfunctions, such as stick-slip and formation stringers, are treated as discrete time events and mitigated using software protocols triggered by accurate detection algorithms. Also, operation of the inner loop autodriller is of critical importance and controlled using a specialized set of autodriller management protocols. Over the past 2 years, the system has been deployed on more than 270 rigs for construction of more than 1,700 land wells in North America. One of the main deployment challenges was the need for full organizational buy-in from both rig and office personnel. Another major challenge was the need for parameter limit roadmaps that define reasonable, but nonconservative, optimization limits to be used during the drilling of each well. Drilling performance on approximately 90 wells was analyzed and compared with equivalent offset wells. The result was an ROP improvement of 7.2% and 8.0% (2-year average) and 9.7% and 17.3% (last year average) for the vertical and lateral sections, respectively. A case study is presented highlighting improved ROP and decreased nonproductive time (NPT) when the optimization system was used.