In March 2008, word about the vast potential of the Haynesville shale play began to leak out with the announcements by Petrohawk Energy Corp. and Chesapeake Energy Corp. of a large lease play developing in northwestern Louisiana. Some stunning well results followed, and operators have since flocked to the play, which extends about 80 miles from west to east and spans the Texas-Louisiana border. (See “Haynesville 101,” Oil and Gas Investor, May 2008.)

The potential value of this shale was evident when the stock prices of key players surged in 2008 before the markets crashed. And in July of last year, Chesapeake unveiled a joint venture with Plains Exploration & Production Co. valued at an estimated $3.3 billion. Chesapeake sold half of its 80% working interest to Plains and will be carried for the costs of some 393 gross wells.

Interest in this new shale play has not abated, and the need for more information is critical as it unfolds. To reduce the time and capital E&P companies need to climb the learning curve, Houston-based Object Reservoir Inc., Dallas-based engineering firm DeGolyer and MacNaughton and a dozen operators have formed the Haynesville Shale Collaborative Exploitation Project (CEP).

This new working group will share best practices for well stimulation, completion techniques and spacing. As drilling moves increasingly to horizontal wells, operators have performed fracture stimulations with as many as 16 stages and drilled lateral sections of up to 7,000 feet.

Beyond these technical studies, a lot can be learned from mapping and graphing publicly available production data from the roughly 100 horizontal shale wells drilled in East Texas and North Louisiana. By reviewing data from Austin-based oil and gas production data provider HPDI LLC, as well as from operators, public filings, logs, industry technical papers and other public information, we have created peak flow-rate maps and plotted actual well performance against company-disclosed type curves.

What seemed like a dream during the leasing frenzy a year ago (leases at one time were selling for up to $30,000 per acre) could turn out to be a nightmare for some operators. Why? This play is highly variable, with some areas requiring $10-plus Nymex prices for operators to break even. Success depends on several factors, including the shale’s mineralogy, pressure gradient, net thickness, thermal maturity and operators’ expertise in dealing with extreme downhole conditions.

Wells in the emerging core have exceedingly high initial production (IP) rates, but are declining more steeply than many published type curves would suggest. However, some wells with more than six months of production history are beginning to exhibit the hyperbolic flattening expected of shale production.

Based on currently available data, we expect the first 30-day well rate (unrestricted) to have a 1:500 IP-to-30-year estimated ultimate recovery (EUR) ratio. The peak calendar-month rate-to-EUR, based on state data, would be closer to 1:600.

For example, Petrohawk expects a well that flows at 15 million cubic feet per day (30-day IP rate) to recover 7.5 Bcf, resulting in a 1:500 ratio. Chesapeake expects a well with an 8-million-a-day 30-day IP rate to recover 5.7 Bcf (6.5 over 50-plus years), yielding a 1:720 ratio. We believe Petrohawk’s ratio is a better fit.

By comparison, the Barnett shale has an average 30-day IP-to-EUR of 1:1000, effectively double the Haynesville figure. Admittedly, the data set is still very limited in the Haynesville.

Sweet spots

Like every shale play, there are good and bad areas and a wide distribution of results. The Haynesville is no different. Recently published petrophysical research shows distinct regional variation in the quality of the shale. Just because the Haynes­ville is present does not mean a company will have commercial results—Mother Nature loves a log-normal distribution.

Drilling activity and well performance clearly point to a core area in Louisiana, with most action taking place in southern Caddo, southern Bossier, De Soto and Red River parishes.

The Haynesville is deeper, hotter, more over-pressured and more mature in Louisiana, with better mineralogy and higher effective porosity than on the Texas side. Successful well results may extend the play’s core south and west towards Nacogdoches County, East Texas. The basin deepens in this direction, meaning both well costs and drilling risks are higher.

Our first stab at defining the Haynesville core encompasses about 750,000 acres, compared to two million acres in the Barnett shale’s core. We estimate that both plays can deliver about 50 trillion cubic feet out of their respective core areas. Chesapeake and Petrohawk have the highest acreage exposure to the Haynesville core.

Petrohawk’s southern Bossier Parish well results are unmatched in the play and have consistently exceeded every other operator’s results so far with some wells topping 20 million cubic feet per day (30-day IP). The company’s results in De Soto and Red River parishes are also superior to results of other operators drilling in the same areas.

Wells in Harrison County, across the border in East Texas, have much lower IP rates and contrary to some industry spectulation, are currently exhibiting similar decline profiles to those seen in Louisiana. Many of the nearly 20 horizontal wells drilled in East Texas will recover less than 2 Bcf, in our opinion. With time, experience, longer laterals and more frac stages (e.g. more money), all areas will show improved results. However, Harrison County will never be as good as southern Bossier Parish.

We caution that IP rates contained in press releases can sound more impressive than they should, due to a lack of standardized methodology. This is why we prefer to use monthly data as reported to the states. For example, one operator originally wrote in a press release that a group of its East Texas wells had an average IP of 9 million cubic feet a day, and subsequently reported that same group of wells had an average first-month rate of 5.8 million. This would be impressive if these wells were in the Barnett, but given steep declines, they may only deliver 3 Bcf each before royalties, at two or three times the cost of a Barnett core well.

We recommend investors and other interested parties always request a 30-day production rate. If one cannot be obtained, we suggest withholding judgment (good or bad) until this information is in hand.

The future

We expect some companies in noncore areas could come under pressure to write down 2008 proved-reserve assignments as their disclosed type curves have not materialized.

Based on three-year lease terms in Texas and Louisiana and assuming a maximum of 640 acres is held by a single well, we estimate that each 100,000 acres not held by production will require at least $1.4 billion of drilling expenditure over the next few years to prevent lease expirations. This will likely lead to campaigns for more farm-outs and joint ventures.

This article is adapted from two reports by Manuj Nikhanj, vice president and chartered financial analyst, and Trevor Sloan, vice president, finance, with Ross Smith Energy Group Ltd., a Calgary E&P research firm. Brock Murray, Jim Jarrell, Kerry Gregory and Salim Jamal also contributed to these reports.