In my December 1, 2009 issue I identified the Haynesville shale in Louisiana and Texas as one of the few areas in North America likely to grow natural gas production over the following 19 months. While the Haynesville may continue to grow from its current level of production of approximately 1.80 billion cubic feet per day (bcf/d) (this figure is the estimated combined production from the Haynesville in both Louisiana and Texas), it now appears that production growth from the play has stalled. While still a very large field by nearly every metric, to many of the vocal proponents of shale gas, the Haynesville has become a major disappointment. Some early estimates for production from the field were well north of 6 bcf/d. The peaking of production from the Haynesville shale has significant implications for North American natural gas supplies and will soon be recognized as a contributing factor to the end of shale gas production growth in the U.S.
As I have discussed in previous issues, many of the techniques implemented in the Haynesville were first developed by the operators in the Barnett shale more than five years ago. The two largest operators in the Haynesville, Chesapeake Energy and PetroHawk Energy, were able to drill their most productive wells right out of the gate due to a much better understanding of shale geology compared to when the Barnett was first explored. The combination of outstanding initial production rates from early wells in the Haynesville (some wells had initial production rates of over 20 million cubic feet per day (mmcf/d)) and a vaguely defined core area allowed the imaginations of some early promoters of shale gas and many analysts, armed with their magic extrapolation machine (sometimes called an Excel spreadsheet) to make some fairly bold predictions on the potential of the Haynesville. For example, below is a quote from Aubrey McClendon, CEO of Chesapeake Energy, one of the most active operators in the play, on the potential of the Haynesville at the Cambridge Energy Research Associates conference in Houston in February 2009:
“We think in time it will become the largest gas field in the world at 1.5 quadrillion cubic feet.” - Rigzone, February 11, 2009 https://www.rigzone.com/news/article.asp?a_id=72839
To put into context how ridiculous this comment is, consider the following. By early 2009, the Haynesville had yet to produce even 1 trillion cubic feet (tcf) of gas and 1,500 tcf of gas would represent over 65 years U.S. consumption. While it is unclear from the Rigzone article whether Mr. McClendon was referring to recoverable gas or original gas in place, he also stated that the Haynesville “will produce at least 500 Tcf over time and then recover around 700 Tcf before potentially growing even larger.” I have been very critical of Mr. McClendon in this publication since such hyperbole from an industry leader about the potential of a play in its infancy is inexcusable. Not only did his comments hurt Chesapeake’s long-term shareholders when reality did not live up to expectations, his comments and those of other shale promoters impacted other energy investors. With so little production history available about this play, “shale-hype” commentary is largely responsible for the current conventional wisdom that natural gas prices will remain low for years. On the other hand, shale hypsters have created a fantastic point of entry for investors who understand the true fundamentals of natural gas (such as readers of this publication) and go long natural gas or gas-weighted equities.
While a portion of the Haynesville extends into east Texas, wells in east Texas have been nearly all sub-economic to date and based on publically available information, do not seem to be improving materially. At the present time, it appears that over 90% of all Haynesville production is on the Louisiana side of the play and therefore for the purposes of this article, I will focus on the Louisiana side of the play.
One of the best sources of information on the Haynesville comes from the Louisiana Department of Natural Resources (DNR) website where it publishes a monthly status report on activity in the Haynesville. While the DNR does not publicly publish Haynesville production figures, we can draw several conclusions based on the publicly available information. Below is a chart reflecting the most recent drilling status, released on June 7, 2010.
The most important number in the DNR chart is the number of producing wells. Based on the 505 producing wells online at the end of May 2010 and a generous average production per well of 3.2 mmcf/d, we can estimate that the top end of production from the Louisiana portion of the Haynesville is approximately 1.6 bcf/d. Given the early stage nature of many of the wells that are currently producing and the incredibly steep decline curves of Haynesville wells, I expect productivity per well to fall in half by the end of Q1 2011. The declining well productivity in the Barnett shale, where the average well is only producing approximately 250 thousand cubic feet of gas per day (mcf/d) and falling, gives us a good example of just how fast the treadmill runs in shale plays that are drilled horizontally with multi-stage fracture stimulations.
The other very important number in the chart is the number of wells that were producing in June 2009. According to the DNR, there were 115 Haynesville wells producing a year ago. Simple math indicates that only 390 wells were put online during the past 12 months. Unless new wells come online at an accelerating rate or the new wells now coming online are significantly more productive than those currently online, Haynesville production will head into decline at some point over the next twelve months.
Based on the independent research I have done, I fully agree with Ben Dell of BernsteinResearch, one of the most vocal critics of the hype surrounding shale plays, who believes that operational results from the Haynesville have now peaked and are set to decline. Here is a quote from a recent research piece he put out:
“While other shale plays ramped up over the years, we believe that the Haynesville Shale is on an accelerated schedule and that operational results have now peaked. Multi-stage fracing was already well understood when aggressive development of the Haynesville began in 2008 and operators moved quickly to the current range of 10-12 frac stages per well. As a result, operators are now unable to increase the frac count to drive improvement in initial production rates.”
With operational results declining in the Haynesville, there is virtually no other way for the play to grow other than having operators put more wells into production at a faster rate. This brings us to the final important number in the DNR’s chart, the growing number of “Other Operations” wells. These are typically wells that have been drilled but are waiting on fracture stimulations or connections to the pipeline gathering system. According to the DNR there are a whopping 375 “Other Operations” wells in the Haynesville. While operators will give you different reasons for having drilled wells without putting them into production, the biggest impediment in turning wells to sales is lack of available frac crews. According comments made by Petrohawk President Dick Stoneburner in the company’s Q1 2010 conference call, there are approximately 20 frac fleets available in the Haynesville. Due to the high demand for very high horsepower pressure pumping from other unconventional resource plays such as the Bakken and the reluctance of service companies to increase the number of frac fleets due to the volatility of gas prices, service companies have been very slow to add capacity in the shale gas areas such as the Haynesville. Instead, service companies have mostly chosen to raise prices for their services. Increased service costs and limited capacity are certain to limit the number of wells moved from the “Other Operations” category to the “Producing” category, regardless of how much natural gas prices increase.
A final reason Haynesville production is unlikely to grow materially from today’s level is that some operators are now dropping drilling rigs in the play. Aubrey McClendon of CHK mentioned during his company’s Q1 2010 conference call that his company expects to drop 6 rigs in the Haynesville. HK also has plans to reduce its previously announced Haynesville capital program by 20% this year. While much of the reason for the reductions can be attributed to a reduced need to drill wells to hold leases, a reduction in drilling indicates operators do not expect to achieve acceptable rates of return from drilling. If operators are unwilling to drill at an increasing rate from today’s level, there is virtually no chance for production growth from the Haynesville.
In fact, based on data I received from a confidential and reliable source (this source was able to verify the data’s accuracy with the Louisiana Department of Natural Resources), it appears Haynesville production growth stalled out towards the end of last year. The below graph depicts Haynesville production growth in 2008 and 2009 in Louisiana’s five parishes responsible for virtually all of the State’s Haynesville production:
After a massive ramp up in production in 2008 and the first nine months 2009, production in the three largest producing parishes stalled in the final quarter of last year. The 200 wells that went into production from the end of 2009 thru May 2010 are unlikely to have increased production significantly due to the huge declines of older wells.
So what does a peak in Haynesville production mean for the United States’ natural gas supply? With production falling in Texas, the Gulf of Mexico, Wyoming, Colorado, Utah, New Mexico, California and Kansas (Source: Energy Information Agency and reporting by the individual states), U.S. natural gas production is headed for a significant fall. There simply are not enough gas wells being drilled to offset the rapid declines of existing wells and there is no way that growth from the Marcellus shale, the only area that will grow significantly this year, will be able to offset declines from all of these areas.