OPERATIONAL PROCEDURES AND METHODOLOGY FOR IMPROVING LWD AND WIRELINE DEPTH CONTROL, KRISTIN FIELD, NORWEGIAN SEA

摘要

Kristin is a HP/HT gas/condensate field in the Norwegian Sea. Deep, hot deviated wells have contributed to depth differences between wireline (WL) and Logging While Drilling (LWD) depths of up to 20m, as well as significant variations between different bit runs. Such depth differences introduce unacceptable depth uncertainty for both reservoir modeling and well operations. Special procedures were implemented to help understand observed depth differences. Radioactive markers were installed in casing strings and logged routinely with both LWD and wireline. Intervals were re-logged on subsequent LWD runs to allow comparison, including logging in upwards direction. Primary depth control procedures were adopted for all WL descents. GR correlation logs were recorded during all relevant runs/passes to allow depth comparisons. Initially wireline depth was believed to be superior, supported by good consistency between runs and between down- and uplogs, unlike LWD depths. Depth differences, which increased systematically with depth, were believed to result primarily from stretch in drill pipe due to temperature and suspended pipe weight, causing LWD depths to be shallow compared to wireline depths. LWD depths, being based on drill pipe tally, were also shallow compared to casing tally, as observed using radioactive markers. Differences between LWD and wireline depths increased for later wells with higher inclination, and the validity of downlog wireline depths were questioned. Formation tops based on wireline depths resulted in questionable local corrections to seismic time-depth maps, with significant implications for mapped reservoir volumes. Given our distrust in both LWD and wireline depths, an alternative approach was sought. Simplified models available in literature were adopted to estimate pipe stretch/compression due to temperature, string weight and drilling parameters. Results obtained could explain only about 1/2 of earlier shifts applied to LWD logs. A similar methodology was developed to correct wireline depths for stretch, with estimated corrections far exceeding standard assumptions. Following run/passspecific correction of both LWD and WL logs, depth uncertainty has been significantly reduced. Similar corrections are recommended to LWD and wireline logs for future wells with large depth uncertainty. We propose to maintain today’s “Driller’s depth” (LWD depth) and “Logger’s depth” (WL depth) as wellsite depth references to avoid operational mistakes, and to establish “Corrected depth” based on corrected LWD- and/or WL depth for use in subsurface models.