Correcting wildcat success

Statistically speaking, the past decade of petroleum exploration in the U.S. was better than originally thought.

The U.S. Energy Information Administration recently corrected major errors in its data for U.S. well completions. It turns out there were a lot more exploratory oil and gas wells drilled than the original numbers showed for 1985-95. For most of the period 1989-92, exploratory oil and gas well completions were more than double the totals originally reported. Errors for exploratory dry holes weren't nearly as great. On balance, therefore, the success rate for exploratory drilling jumped after the correction (see chart Exploratory Success Rates [31,476 bytes] ).

Improvement in exploratory success rates will surprise no one. But evidence until now has been anecdotal: individual companies reporting impressive results. Because the overall success rate in the uncorrected EIA data languished below 30% until 1994, however, the suspicion had to be that failures were concentrated within a segment of the industry not calling attention to its results.

As the new EIA data show, solid improvement began in the mid-1980s and has remained strong. The turn coincides with surges in collection and use of 3D seismic data.

Why didn't EIA's data capture the change until now? For one thing, counting well completions is never easy. It involves a reporting lag as great as 5 years. Estimating the lag's effect was especially difficult during the period under question. A major drilling slump was under way. And rapid technological development skewed traditional statistical tests; for example, completion totals stray from old correlations with the rig count when operators increasingly complete two or more lateral wells from a single vertical hole.

Turbulence in a data series already plagued by a large reporting lag thus camouflaged errors that invaded numbers for which EIA relies on a private vendor. As the agency explains it, "The data files since 1987 were missing some records and contained duplicates of others, updates to many records were not passed along, and records for recompletions-which weren't expected-were present in files for some months but not in others."

In addition to the exploratory success rate change, EIA points out that drilling responses to oil price movements of 1993-96 were not as great as the uncorrected data made them appear to be. Also, the increase in exploratory well completions in the corrected data will reduce the volumetric finding rate that EIA uses in its projections.

The reporting error's most important consequence, however, is its implicit overstatement of the dry-hole risks of exploratory drilling since the early 1980s. Coupled with the oil-price risks that became manifest after 1986, that error, to the extent it influenced investment decisions, may have cost U.S. exploration a lot of capital. Correction of the error and the exaggerated perception of risk that it conveyed should help operators raise funds for exploration. It will not hurt that a prominent feature of the corrected ex- ploratory success rate is that the upswing continues.

Copyright 1998 Oil & Gas Journal. All Rights Reserved.