After fracking came of age in the Barnett and Fayetteville Shale, the discovery of more prolific source rock in the Haynesville and then the Marcellus/Utica and Eagle Ford has put these older plays out to pasture. The Barnett, grandfather of shale gas plays, has seen its rig fleet emigrate, declining from the 100's at its peak to a meager14 horizontal gas rigs as of last week's count:
When the rigs first began to vanish, analysts debated the impact on production in the field. The well count was high, and average age of wells was also high. The flat slope of the shale gas decline curve in the 'late innings' suggested that production would gently trail off for many years.
Furthermore, an influential study of the Barnett Shale by the Bureau of Economic Geology at the University of Texas was recently published (Reviewed in Oil & Gas Journal in Jan 2014). The study was of excellent pedigree and purported to be extremely thorough and detailed. I guess the world is divided into those that see the source rock half-full and half-empty. And these authors were surprisingly optimistic about the future of the Barnett. Here is the study's 'base case' production forecast, which was predicated on HH gas prices of $4.00.
Now I would be the first to acknowledge that if I were planning publish something wildly bullish about natural gas, I would double check the market and maybe postpone publication by a day if gas were falling through the floor. No sense launching directly into a crashing wave. But look at the graph above. It helps to have a straightedge since vertical bars are unhelpfully absent. It shows production peaking in 2014, and down maybe 50 MMCFD in 2015. It falls below 5BCFD sometime in 2017 or 2018, for shall-we-say a 10% decline in 4-5 years, from a 5.6BCFD plateau in 2013-14.
Now Texas is noted for its latency in production data reporting. So allowances can be made for that. But let's start with the current reported production for the Barnett. First we have a graph directly from the TX Railroad Commission website (found here):
This is based on preliminary data, but shows a 2012 peak and a decline of 7.7% in 2013. The Jan '14 production number is likely well below the actual because recent data is typically supplemented by late production reports. Furthermore, there were significant temporary weather impacts on production in January.
But a more granular look at production, drilling down (as it were) to county and monthly statistics, gives a clearer picture. Here is data from the same source as above (TX RRC Query) that shows reported production in the four core counties of the Barnett, by month:
This shows a peak as far back as 2011, and a discernible decline curve for 2012 and 2013. Depending on how high production data are revised for late 2013, this is indicative of a far steeper decline. Here are the counties themselves:
A few complicating factors come to mind:
- Shut in volumes could well be a factor, as prices dipped painfully low in 2012 and many legacy wells in the Barnett would be considered stripper wells that would only be marginally economic.
- Cold weather could have heavily impacted production in Jan '14.
- Non-core counties are an important part of the mix, as the core counties above account for just over 80% of total production in 2012.
- Small but meaningful areas outside the core include the combo oil/gas play in Montague county:
But the overall impression from the production data indicate that production may not hold up as well as the models suggest. Or the University of Texas has some kind of Texas sized bias. I don't know, but I know how to draw a line through a series of points. I think Tarrant County is the easy place to look for indications of where the trend is headed. And only 3-6 rigs have been working there in the last year. I don't think that is enough to replace production of 2 BCFD.