The Petroleum Truth Report

My goal is to offer clear and direct explanations of energy reality. These posts are data-driven interpretations of oil and gas topics that often challenge conventional thinking.

A Long Recovery for Natural Gas Price: revisiting the Haynesville Shale

Natural gas prices increased 39% from a 6 1/2 -year low of $3.19/MMBtu on April 27 to $4.42 on May 13, 2009. Some think that the worst of the price collapse that began in July 2008 is over, and that gas prices will return to normal. I do not believe that is the case, though I certainly hope that I am wrong. Chesapeake Energy proclaimed in a recent investor presentation that “the fix is under way”, and that natural gas prices will soon return to $7-8/Mcf. Chesapeake and other companies make the case that prices will rebound because of the drastic decrease in drilling. The gas-directed rig count has fallen from 1,606 to 728 since September 2008 and, because about 1,100 rigs are needed to maintain supply, we are creating a deficit that should cause the price to rise.

The argument is logical and may prove true in the long term, but it is difficult to support based on current events. During the same two-and-a-half week period of rising prices, working gas in storage has been well above the five-year average (23% above the 2004-2008 average), indicating that supply is strong (Figure 1). While proponents of increased gas price may find some support in short-term price fluctuations that are based on sentiment, gas storage is what drives traders, and traders determine price. In other words, until storage levels decrease to 5-year averages or lower, I doubt that there can be any sustained gas-price increase. The recent rally is probably related to rising crude oil price, a weaker US dollar, and short-selling of gas futures contracts rather than a change in market fundamentals. At this writing, gas prices have already lost most of their recent gain.

US gas supply includes two external sources: pipeline imports from Canada and LNG cargoes from all over the world. Imports from Canada are down, but LNG deliveries are way up. April LNG imports averaged 2 Bcfd, compared to half that amount in March, according to Jeffries & Company (1). Pritchard Capital Partners expects LNG deliveries to average 3.5 Bcfd for 2009, and to be as high as 5 Bcfd (1). Global liquefaction capacity has increased more than 5 Bcfd this year as several large projects came on line. Also, LNG price has decreased due to lower global demand, reduced tanker costs, and because contract prices are tied to a trailing index of crude oil and other commodity prices that have fallen.

Oversupply of gas may continue for longer than some expect. Average US gas production increased from about 62 Bcfd during the first half of 2007 to almost 65 Bcfd in the second half of 2008 (Figure 2). In addition, there are large volumes of gas available from wells that are not yet connected to sales because of limited pipeline capacity and low netback cost. The Rocky Mountain Express Pipeline will open considerable volumes of gas that have not been previously available (3.2 Bcfd by June 2009, and an additional 1.8 Bcfd in November), and the Mid-Continent Express Pipeline recently added 0.64 Bcfd of capacity. Also, in the Gulf of Mexico Independence Hub, Thunderhorse and Tahiti platforms, as well as initial production at Perdido, will increase gas production in 2009-2010.

Reduced demand because of the global economic crisis may contribute to a prolonged slump. Demand in February 2009 fell 14.5 Bcfd (16%) compared to January, and 7.4 Bcfd (9%) compared to February 2008. In Februrary, all sectors of gas usage fell.

While the gas-directed rig count is down, drilling activity is strong in the Haynesville Shale play, where high-volume initial production rates work at cross-purposes to offset the over-supply of gas. There are at least 75 horizontal wells that are currently drilling, completing, or shut-in pending pipeline connection. This could increase Haynesville daily production to more than 300 MMcfd.

I want to thank readers and operators for their willingness to share information with me in response to my earlier column on the Haynesville Shale (World Oil, April 2009). There is no doubt that the Haynesville is different from other shale plays, mainly because it is overpressured (~0.85 psi/ft). Overpressure and corresponding microfracturing combined with high shale porosity result in average initial production rates of more than 12 MMcfd and per-well EURs as high as 9.0 Bcf.

I now think that the Haynesville Shale reserve estimates that I presented previously were too low. I have evaluated 43 horizontally drilled wells with some production history, and 14 wells with initial production rates only (Figures 3a and 3b). The most-likely average EUR for all operators is 3.6 Bcf per well within a probabilistic range of 2.8-3.6-4.4 Bcf/well (RBC Capital is more pessimistic, projecting an average 2.5 Bcf EUR for all horizontal wells (2). The average EUR for key operators in the play varies: Petrohawk Energy Corporation has higher average EURs (3.9-5.1-6.2 Bcf) while Chesapeake Energy Corporation’s EURs are lower (2.2-2.8-3.3 Bcf). The average for other operators is 2.8-3.7-4.5 Bcf/well.

I have not changed my conclusion that the Haynesville Shale play is marginally commercial. Drilling and completion costs vary from $7.5 to $10.5 million per well. The marginal cost for operators to find and develop natural gas reserves is $7 to 8/Mcf, and current netback prices in the play are less than $3/Mcf. The threshold netback gas price for a better-than-average 5.5 Bcf well to break even is $7/Mcf at NPV10 (Bodell and Pittinger, in press). For companies that have favorable hedge positions, realized gas prices for 2009 will be as high as $6.50/Mcf and $6.00/Mcf for 2010. This means that the play is marginally commercial today for operators with favorable hedge positions, but not commercial based on cost and price fundamentals.

While many believe that natural gas prices will increase to $7-$8/Mcf by the end of this year, I am more pessimistic. Increased LNG imports and strong current gas supply, expanded pipeline capacity and ongoing gas-directed drilling contribute to strong gas supply, while the recession is reducing demand. This leads me to conclude that prices may not increase until the second quarter of 2010. I am also skeptical that price will recover beyond approximately $5.50/Mcf, the average inflation-adjusted gas price since 1995 (Figure 4). Shale plays have increased the marginal cost of production by approximately $2/Mcf, and I doubt that the market will reward that inefficiency. It seems more likely that LNG and conventional gas will play an increasingly important role in US gas supply in the future because of cost.

(1) Platt’s Inside FERC’s Gas Market Report, May 1, 2009

(2) RBC Capital Markets, Weekly Haynesville Shale Report: May 13, 2009

Haynesville Sizzle or Fizzle: Let’s be fair!

When I read some of the comments posted on this web log to a friend yesterday, he said, “Anything that gets this much flak, must be close to the mark.”

I have received dozens of e-mails and a half-dozen posted comments on this web log about the Haynesville Shale. Many of the e-mail authors strongly disagree with my opinions about the Haynesville, but are respectful and professional. In contrast, the authors of many web log postings who disagree with me are often disdainful, caustic and even vulgar, not only about my opinions but also about my professional qualifications. Some indignantly demand that I disclose the wells (be patient inquisitors—a list of wells follows as Figs. 1 and 2) that I used in my evaluation—I wonder if these same people issue similar demands to the companies that make pronouncements that the Haynesville Shale has 250 Tcf of reserves when there are fewer than 50 wells with any production so far, or that an average well will produce 6.0 Bcf when none have yet produced more than about 3.0 Bcf and most, considerably less.

To be fair from my side, readers have sent me data on Haynesville Shale production that was not available to me when I did my research and published my World Oil column and the previous web posting. That new information modifies my view of the Haynesville play somewhat, and requires an update to my observations and conclusions. Also, there is now a month or two of additional production history since I did the research for that work.

Based on this information, approximately 59% of Haynesville wells may have ultimately recoverable reserves of 0.5-2.0 Bcf (16 wells), while 41% may produce 2.0 Bcf or greater (11 wells), according to my analysis. The mode of 27 wells is 1.5 Bcf and the mean is 2.2 Bcf. These reserve projections are approximations, and are only intended to provide a range of possible outcomes for wells with too little production history to accurately project.

For those reservoir engineers who disparage my qualifications to pick a trend line through data points on a graph (something that apparently is beyond the capability of those with advanced degrees in science unless their degree is in reservoir engineering), I hope that you have never picked a top unless your degree is in geology.

The crucial issue about the Haynesville Shale play, however, is not rates and reserves, but cost. As I explained in “Haynesville Sizzle could fizzle”, threshold economics for the Haynesville Shale require netback gas prices of $8.50/Mcf, and minimum reserves of 2.5 Bcf/well. This is because drilling and completion costs are from $7.5-9.5 million. It is simple algebra once the costs are known.

I have studied the 10-K SEC filings by the major players in the Haynesville play. These are public documents prepared by the operators. With most operating costs between $2.50 and $3.50 per Mcf, rates and reserves simply do not matter at current gas prices of $2.50 netback in the Haynesville. When capital expenditures are added, it costs most operators about $7.50/Mcf to find, develop and operate in the play. While some operators are currently hedged at higher prices, this is a short-term situation, and no one will take the other side of a hedge at more than $7.50/Mcf today or at any time in the foreseeable future.

If you don’t believe me, you should read reports by Credit Suisse, “The True Cost of Shale Gas” (April 2009), and by Bernstein Research, “Why the Haynesville Won’t Work…at $4, $5, or $6/Mcf gas” (April 2009).

I am more optimistic now, based on new information, that the Haynesville Shale may be different from most other shale plays. If operators can substantially reduce cost, and if gas prices improve to levels during the first half of 2008 (average $10/Mcf Henry Hub), some percentage—perhaps 25-50%–of wells in this play may become commercial, but it’s really not about EUR as much as it is cost and gas price.

I have been fair in admitting that new information has modified my view of the Haynesville Shale play. I acknowledge that rates are extremely impressive for several wells, and that some wells have already produced more than 1.0 Bcfg.

For those who disagree with my views on this play, I ask that you be fair too. Look at costs, and not just rates and reserves. If the marginal cost to produce gas is more than $7.50/Mcf (which all operators admit and many state in public presentations–for example, Range Resources’ “IPAA 2009 Oil & Gas Investment Symposium”), then no one is making money on this play today regardless of impressive rates and strong reserves. Unless prices rise above levels they have reached during only 15 months over the past 10 years (or 20 years, for that matter), none of the wells in the Haynesville Shale play is likely to be commercial (Fig. 3).

I am not a gladiator. I don’t perform in my columns and web log waiting for thumbs up or down from readers to validate my methods or conclusions. I put my work in a public forum to share what I observe, and to generate a dialogue that may help us all move closer to the truth. I return every e-mail message that I get, because the people who write them want to engage in the conversation, and deserve my time and respect. For those who prefer to comment anonymously on this web log, I welcome your views also. I encourage you to join the conversation as peers and not as blood-sport spectators looking for entertainment at the Coliseum.

Haynesville Sizzle Might Fizzle

Despite lower natural gas prices, the Haynesville Shale is the hottest onshore play in North America. Production is more than 150 MMcfd from recently drilled horizontal wells, and single-well Initial Production (IP) rates are as high as 24 MMcfd.

I used standard rate-versus-time methods to determine estimated ultimately recoverable reserves (EUR) for 14 horizontally drilled wells that had sufficient production history to project a decline rate. Production was extrapolated using a hyperbolic decline, and an economic limit of 1.0 MMcf/month. The wells had an average EUR of 1.5 Bcf, and 67% (10 wells) had reserves less than 1.5 Bcf. This is an early evaluation, and does not include several recently completed wells because of insufficient production data. Reserves were, with one exception (5.3 Bcf), considerably lower than the 6.5 Bcfe most-likely per well reserves, and 4.5-8.5 Bcfe range, claimed by leading operators in the play Chesapeake Energy Corporations and Petrohawk Energy Corporation.

Problems with the Haynesville Shale include high decline rates and costs. Average monthly decline for the wells that I analyzed is 20–30%, and projected annual decline rates average 80−90%. Rapid decline makes IP rates unreliable indicators of well productivity. The average production history of wells used in this analysis is less than five months; current production rates already average only 48% of IP.

Drilling and completion (D&C) costs are about $7.5 million per well, although Petrohawk recently revised its D&C costs upward to $8.5–9.5 million. Average true vertical depth of wells in this study is 11,500 ft, and average measured depth is 15,250 ft. Five- to ten-stage hydraulic fracturing is typical with 600–750 lb sand/lateral foot in horizontal boreholes, which average 4,500 ft long. Leasing costs in active areas during 2008 were $10,000–30,000/acre, increasing capital expenditures for an 80-acre spacing unit $0.8-2.4 million above D&C costs.

Operating costs average $2.25/Mcf, based on US SEC 10-K filings and annual reports. After gathering and transportation costs, netback gas prices for early March 2009 were less than $2.50/Mcf (RBC Richardson Barr). Net revenue interest, after royalties, is typically 75%, and Louisiana severance tax is $0.27/Mcf (included in operating cost) . While current prices are the lowest in many years, and hedging has helped careful operators, it cost many operators a $7.25/Mcf or more to produce gas during the fourth quarter of 2008.

Clearly, most Haynesville Shale wells will not approach a commercial threshold until both gas prices and per-well reserves increase. To quantify that reserve and price threshold, I ran a basic NPV10 model using the cost information already mentioned. I used decline rates from the Barnett Shale (65%—Year1, 40%—Year 2, 30%—Year 3, 25%—Year 4, and 20% thereafter) instead of the higher decline rates projected from Haynesville production to date.

The break-even (NPV10= 0), minimum per-well reserve volume is 2.5 Bcf with a netback gas price of $8/Mcf (~$9/MMBtu Henry Hub spot). This means that the play would have been marginally commercial in 2009 dollars during only 15 months (12.5%) over the past decade—and over the past 20 years since the advent of the natural gas commodity market in 1989—if an average well had reserves of 2.5 Bcf instead of only 1.5 Bcf. At 1.5 Bcf/well, $12/Mcf netback gas price is needed to break even.

Chesapeake CE O Aubrey McClendon recently said, “We only need gas prices to be ‘good’ for three to six months out of every two-year period.” (Houston Chronicle, February 11, 2009). If ‘good’ means to break even in the Haynesville Shale, it looks like he will meet costs no more than 12.5% of the time, and lose money the other 87.5%, assuming that per-well reserves can be doubled. That business model is difficult to understand, although successful hedging might change those percentages. But that’s not the entire business model.

“We believe in volatility…You can sell volatility. Volatility has value,” McClendon continued. “Our company makes additional money when we sell those calls.” What McClendon means is that his company can make money by selling deals to other companies that fear they will be left behind during brief periods of rising prices. For example, in 2008 Chesapeake sold interests in its shale plays to Plains, BP and StatoilHydro. Chesapeake made $10.3 billion on those transactions.

Why do I reach different conclusions about the Haynesville and other shale plays than some industry analysts? First, they are not industry insiders and, therefore, many do not incorporate true operational costs including interest expense for debt service, or netback gas prices into their evaluations. Second, investment company analysts are marketing a product and make a commission on stock that they sell to clients—their analyses cannot be truly objective. Third, they do little investigative research, and generally accept information on rates, reserves, and declines provided by the companies that promote these plays. They cannot have done independent decline analysis on the Haynesville Shale or they would have recognized the obvious reserve discrepancy (1.5 vs. 6.5 Bcf/well).

I expect shale plays to be part of the natural gas landscape for awhile, despite the fact that they are marginally commercial at best. Most companies in these plays have a lot of debt, and the only way to service the debt is to generate cash by drilling wells to produce gas.

The Haynesville Shale play appeared at a time when gas prices were rising. Companies rushed to pay great sums to obtain positions based on the irrational belief that prices would continue to rise. This is the same thinking that brought us the global financial crisis. The magnitude of capital expenditure for leasing and drilling illustrates a profound breakdown of due diligence by the financial and E&P industries.

It is difficult to imagine that the Haynesville Shale can become commercial when per-well reserves are similar to those of the Barnett Shale at more than twice the cost. Maybe the most recently completed wells will tell a different story; otherwise the Haynesville Shale play will likely be replaced by other shale plays that lose less money.

Shale Plays, Risk Analysis and Other Perils of Conventional Thinking: Haynesville Shale Sizzle Turns to Fizzle

In mid-July 2008, the United States somewhat unexpectedly discovered that it had an oversupply of natural gas, and prices fell sharply. Jen Snyder, head of Wood Mackenzie Ltd’s North American Gas Research Group, recently said that the development of shale gas plays has caused “a significant potential over-supply” (Oil and Gas Journal, December 1, 2008). Shale plays had become increasingly irresistable to the North American industry before prices fell this summer. Many traditional E&P companies, including some majors, decided to become shale players, and many are still considering the possiblity despite low gas prices. The global financial crisis has accentuated the aversion to risk that fueled shale plays to begin with, and it seems that no one now wants to pursue anything but shale.

In the first half of July, spot gas prices were more than $13.00 per million British thermal units (MMBtu). Six weeks later, the price had fallen below $8.00, and in March 2009, it is around $4.25/MMBtu. Some analysts predict that gas prices will average $4.00-6.00/MMBtu range at least through the end of 2010.

A total of 1,966 horizontally-drilled producing wells from the Barnett Shale were evaluated to determine commercial gas reserves using standard decline methods. Based on this analysis, only 30% of Barnett Shale wells will realize revenues that meet or exceed drilling, completion and operating costs in the most-likely case based on assumptions incorporated into a 10% net present value (NPV10) economic model. The economic model includes per-well drilling and completion costs of $3.25 million, a wellhead gas price of $6.25/MMbtu (the average spot sales price for 2007), 75% net revenue interest, 7.5% Texas severance tax, and $1.25/Mcfg lease operating and overhead cost. These assumptions are consistent with information published in 10-K U.S. Securities and Exchange Commission (SEC) filings by key Barnett Shale operators. The model requires per-well cumulative production of about 1,325 MMcfg over 10 years to reach an economic threshold.

An early analysis of 20 horizontally drilled wells in the Haynesville Shale play in Louisiana and parts of adjacent East Texas suggests a disappointing outcome because of extremely high decline rates. Average monthly decline rates are 24%, with 75% of wells declining 20-35% per month. The impressive initial production rates (IP) for these wells do not, therefore, necessarily translate into high reserves (actual daily production rates from the maximum 30-day period were, in fact, about 20% lower than reported IPs).

Fifteen Haynesville Shale wells had sufficient production history to analyze using standard rate-versus-time decline methods. Estimated ultimately recoverable reserves (EUR) averaged 1.5 Bcfg, and 67% of wells had reserves between 0.5 and 1.5 Bcf. These results indicate that Haynesville Shale reserves will be about the same as Barnett Shale wells at approximately twice the cost to lease, drill and complete.

I have struggled to understand the appeal of shale plays based on economic factors, and thought that low gas prices would greatly reduce activity. At $10.00/MMBtu, about half of horizontally drilled and fractue-stimulated Barnett Shale wells were commercial so, while prices were rising even higher, shale plays made some sense. At current prices, however, only about 11% of Barnett wells pay out, and all indications are that prices will fall lower or, at best, remain at current levels. While leasing has largely stopped, drilling continues*, and enthusiasm from both companies and analysts seems strong, at least for the Barnett, Haynesville and Fayetteville shales.

How can we understand what is happening with shale plays?

The diffusion model of innovation (Ryan and Gross, 1943; Rogers, 1962) shows that people adopt new ideas and technologies slowly, and that only about 5% of people make the decision to adopt based on information. The other 95% decide because of the the views of opinion leaders in the community, and on the eventual social momentum that develops—what Malcolm Gladwell called the “tipping point”. The 5% who base decisions on information in the diffusion model are critical thinkers; the rest are conventional thinkers.

What causes people to decide to abandon an idea that almost everyone previously accepted? It is reasonable that only critical thinkers make this decision based on information, and that conventional thinkers follow in what may become a stampede. Thomas Kuhn (1962) explained that scientists resist abandoning a ruling theory in favor of a new paradigm with a kind of orthodox fervor of conventional thinking, and often ostracize those critical thinkers who point out problems with the existing model. At some point, when opinion shifts to support a new paradigm, the previous theory is unceremoniously dropped, and its remaining supporters are criticized as dinosaurs.

It is useful to review some of the history of how our industry arrived in its present state. The collapse of oil prices in 1982-1986, and the ensuing 13 years of over-supply and low prices created an environment in the E&P business where cutting cost and reducing risk were paramount. Thousands of jobs were lost, and companies disappeared as layoffs, reorganizations, mergers and consolidation became the core business of oil and gas companies.

As oil prices slowly recovered in the late 1990s, risk analysis teams were formed to manage technical work. Executives abdicated their technical responsibilities to risk committees, and turned their attention to buiness models. With the help of consultants, they envisioned companies in which exploration and production would become a manufacturing operation, and risk was eliminated. Execution was paramount, standardization was essential, and new geological ideas were unnecessary. The new vision for the E&P business represented the victory of conventional over critical thinking.

Shale plays not only satisfied this model, but also solved the perennial E&P problem of being opportunity-constrained: because shale is practically ubiquitous, there are no limits to what can be spent pursuing new and existing opportunities. This shift was widely supported by the capital investment community because of the low perceived risk, and the fact that non-scientists could understand the play.

Returning to the present, myths about the current state of domestic E&P must be clarified in order to put shale plays in context. These plays are an important component of domestic natural gas production, but represent a relatively small—though growing—portion of the total gas supply. Even among unconventional gas resources, tight gas and coal-bed methane dominate production.

Second, these plays involve considerable risk. The fact that 75% of wells are commercial failures at current gas prices is a tangible risk. Great emphasis is placed on engineering ideas and technology, but it seems that concern for geological and geophysical understanding is uneven among shale players. All shale plays are different, and require unique approaches based on thermal maturity, structural factors, fracturability, and identification of sweet spots.

Third, economic models must be aligned with full-cycle PV10 industry standards. Wood MacKenzie’s Snyder says that established shale plays have “sufficient volumes available at a development break-even price of $5.50/MMbtu or below” (Oil and Gas Journal, December 1, 2008). I don’t believe that. I do not know any credible industry analysts who believe that shale plays are commercial below $7.50. The only way to arrive at the thresholds that Snyder mentions is to understate or ignore current levels of capital expenditure, as well as general and administrative, lease operation, midstream, and discounted capital costs, or to inflate rates and reserves beyond what can be supported by performance history.

Additionally, the over-supply of natural gas that analysts describe may be relative, and that would be positive for shale plays. Spot prices rose to $13.00/MMcf because of an imbalance between supply and demand. Prices fell when about 2 Bcfd of additional supply came online from the Independence Hub, Thunder Horse and Tahiti in the offshore Gulf of Mexico, in addition to increased unconventional gas production, including shale gas. Monthly natural gas production over the past year averaged approximately 1.75 Tcf. The additional 2-3 Bcfd that produced an over-supply is only 3.5-5.5% of total production. Many circumstances might quickly upset the supply-demand balance and result in higher prices. At the same time, the global financial crisis will probably reduce demand, and somewhat offset other factors that may favor rising price. The point, however, is that the difference between what the market perceives as over- and under-supply can be razor thin.

Finally, gas rig counts and rates have fallen sharply in recent months from more than 1,600 in September 2008 to 970 in late February 2009. Some predict that rig counts may fall to 800-900 in coming months. Unconventional wells have steep decline rates, and any decrease in drilling will quickly result in dramatically lower gas production from these plays. That, in turn, will affect supply, and prices could rise, but may also expose the ephemeral contribution of unconventional gas sources to total natural gas supply.

There is little doubt that Shale Plays are likely to be important for some time. I hope that operators will continue to learn how to reduce cost, optimize production, and better incorporate geology and geophysics into their play strategies. It is not certain that the U.S. has a long-term over-supply of natural gas, or that today’s surplus is chiefly because of shale gas production.

Shale plays represent a disturbing tendency in the E&P business away from critical thinking. The belief in reward without risk is irrational. Failure to acknowledge the marginal economics of the play is bewildering. Unless opinion leaders confront the underlying economic and geological risks of these plays, I fear that a financial crisis may develop that will discredit the E&P industry.

*Barnett Shale completions have dropped from about 250 wells/month in May 2008 to approximately 125 wells/month in late 2008.