I attended the Biophysical Economics 2nd International Conference in Syracuse on Friday and sat through a long day of very intensive, eye-opening presentations. The first part of the day covered ERO(E)I, with presentations by David Murphy and Bryan Sell.
Sell’s presentation, on the EROI of shale natural gas plays was most instructive. He studied and compared a conventional well region (Indian Cty, PA) with the longest-running shale play, the Barnett Shale (Wise Cty. TX). Both areas offer around 11,000 wells, and are mature, yielding good data sets.
The first thing to note was that the wells in the Barnett have a much greater initial yield than current conventional wells, about 10x the peak production, although lifetime volumes were not this high. Usually, this is the number the gas companies throw around- initial production, which comes online in the first year.
There is a linear ramp up to peak production, usually in the first year, followed by a very steep exponential decline. Most wells in the Barnett are done in 7 years or less. More disturbing than this is that the overall field has shown the same curve shape, in a fractal relation to wells and groups of wells. In 2000, production in the study area was peaking at 180 MCF / day and declining over 7 years. By 2007, production was peaking much higher at 300 MCF / day, but declining much more rapidly in 3-4 years. The EROI went from 84:1 in 2000 to 38:1 in 2007, and overall volume per well had also dropped to half over the same period. This trend suggests another halving in 7 years, a 10% decline rate. Despite initial positive EROI, Barnett will show lower EROI than the conventional PA play in about 10 years time.
Recent studies have shown only 28% of these wells have been profitable, and Sell showed costs per foot drilled in the Barnett at $150, three times conventional well costs. Shale plays also tend to be much deeper than conventional wells, driving up per-well cost. The Marcellus and Haynesville plays are more difficult and deeper than Barnett, and cost per foot drilled is double or more what it is for Barnett.
While Chesapeake, XTO and others have touted that the Barnett will yield 26 TCF, Sell calculates that the actual recoverable will not exceed 8.8 TCF.
The numbers don’t add up. The earlier profitable wells, and new initial peaks are being used to pay off debt and early entrants’ royalties. Even in the Barnett, this Ponzi approach will not last, and the 8.8 TCF may turn out to be optimistic as the economics start feeding back.
This need to keep ratcheting up production explains the tremendous pressure to open up the Marcellus, but it is clear that if Barnett doesn’t pay, none of the others will either.