From the pages of: World Energy, v5n3

The Case for a Coming Gas Shortage

by Matthew R. Simmons
Chairman and CEO
Simmons & Company International

As the world's largest energy consumer, America is fortunate to be virtually self-sufficient in each key energy source, with the exception of oil. And of the key sources, natural gas is assuming greater importance, both because it is available domestically and because it is the cleanest energy source used on any significant scale. Yet our natural gas supply could come up short as soon as this winter.

The case begins with the 1999 National Petroleum Council (NPC) report, "Natural Gas – Meeting the Challenges of the Nation's Growing Natural Gas Demand," which predicted U.S. gas demand would grow from 22 TCF per year to almost 30 TCF by 2010 based on the large number of new gas-fired power generation plants needed to meet growth in electricity demand.

There were three task forces that compiled this study. I served as the chairman of the Demand Task Force and a member of the integrating team of this study. When released, the report's demand assumptions generated considerable skepticism, principally because most readers did not realize the number of gas-fired power plants actually on order. In the 30 months following the NPC report's release, power generator orders far exceeded the total number of gas-fired power plants the NPC predicted by the end of 2010.

Although many power plant orders have been canceled, some 130,000 MW of new generating capacity will be complete by year's end, with another 50,000 to 70,000 MW of new power plants likely to come online in 2003. In fact, the number of new gas-fired plants will be double what the NPC report envisioned by 2015. Worse, the NPC study assumed that most new gas-fired plants would be dual-fuel plants, but instead, almost all are pure gas or cogenerated plants.

The NPC report said meeting a demand level of 30 TCF would be difficult but possible, as long as proper access to our gas reserves occurred in a timely manner and rig efficiency gains continued. The NPC report's model also suggested that this gas growth could happen while gas prices remained under $3 through 2015, though it acknowledged there would be price spikes from time to time.

The Last Drilling Boom

Within weeks after the report's publication, gas prices burst through the $3/mcf ceiling, rising steadily before peaking at $10. This price explosion led to an unprecedented gas well drilling boom. In 2000 and 2001, over 36,000 new gas wells were completed, twice the annual rates of the previous decade, exceeding the NPC report's model for new gas wells in this same period by 41 percent.

In 2001, the U.S. set a new record for gas wells completed, topping the 1981 record by over 2,000 wells (Figure 1). At the height of the 2001 boom, about 85 percent of the rigs at work were drilling for gas. Almost all these gas wells were development wells, which typically come on stream far faster than exploration wells since the delivery infrastructure is usually in place before the well is drilled.

By late spring 2001, gas prices began to soften as gas storage began to rapidly refill. Yet 2001 production figures show that in spite of the record-shattering drilling boom, daily gas production stayed flat for the eighth year in a row. Actually, gas storage was rising as a result of falling demand.

This weakness in demand arose from several causes. The $10 prices destroyed some key industrial gas markets. Some users shut down and others went to alternatives. Worse, the weather has remained mild, both summer and winter, ever since February 2000. Our stunned economy after 9/11 added impetus to the reduction in demand. So, after the mild 2001/2002 winter, storage levels had gone from "comfortable" to extremely high. By the first anniversary of $10 gas, prices had fallen to $1.85.

The drilling boom peaked at the end of July 2001, with over 1,070 gas rigs drilling for gas – a record. For the next few months, gas drilling declined but still stayed at very robust levels. Gas drilling really plummeted from November through its low point in early April 2002. By the bottom, gas drilling had dropped by 45 percent (Figure 2). Gas supplies, though, stayed relatively flat.

Supplies versus Drilling

Through the first quarter of 2002, reported gas well completions had fallen, but so far, the extent of the decline in supply is less than half the drop in drilling. All of this creates a two-part natural gas riddle: (a) Why did gas supplies fail to respond to the drilling boom? And (b) What impact will the drilling collapse have on gas supply? The answers could indicate whether the U.S. can ever deliver a 30 TCF market, or anything even close to it, absent calling on a massive amount of imported LNG and significant volumes of arctic gas.

If the unprecedented 2000/2001 drilling boom could not increase production, it becomes hard to imagine the industry growing supplies by over 30 percent by 2010. Even worse, if supply suffers a serious reduction as a result of the drilling collapse, it then becomes problematic that production could climb back even to 52-53 bcf per day. This would certainly create a serious energy crisis. Unlike oil, where foreign imports can readily substitute for a fall in domestic production, imported natural gas can only come from Canada and four relatively small LNG terminals, two of which are still not even useable.

We might be facing an era where high prices stimulate more domestic drilling, but ever-rising decline rates in many North American producing basins keep supplies flat. In fact, few gas analysts now think gas supplies can grow until drilling recovers to boom conditions. On the other hand, these same supply forecasters also assume that any decline in gas production, as a result of the drilling drop, will be small.

The Marginal Well Theory

Those who assume that a 45 percent drop in drilling will only result in a modest change in daily gas supply believe a theory that also explains why supply never rose during the historic drilling boom. It's called the "marginal well" theory. It presumes that the drilling boom created by ultra-high gas prices induced the industry into drilling a pyramid of steadily more marginal wells.

Simplistically, the marginal well supply models assume that the first 10,000 wells drilled at "normal drilling rates" produce wells that average about 1 mmcf per day in their first year of production. As the rate of drilling increased during the boom, the next 5,000 wells' initial production was only a third of "normal" wells. The final wells brought on at boom's peak, the last 6,000 well completions of 2001, this theory suggests, had daily production levels of about 15 percent of the first 10,000 wells.

If these numbers are correct, and the base pre-2000/2001 production decline stayed around 26-28 percent, this perfectly explains why more than doubling 1999 well completions still failed to raise gas supply.

Some of us at Simmons & Company were troubled by this theory. It seemed unlikely that the well productivity could suddenly drop as much as 65 to 85 percent from the "norm" in such a short period. If the wells completed in 2001 were fairly similar to the wells completed in prior years, why was production flat? Absent "marginal wells" occurring, the only other answer is that the base decline rate has accelerated sharply from the past decade.

It could also turn out that the opposite of the marginal well theory happened. Not only were more "normal" wells drilled, but high gas prices could have created an environment where far more money was spent to maximize the volumes initially produced from each new well completed. This behavior could have led to a new generation of wells actually having higher peak rates of production, but this higher production would also result in far higher decline rates.

Testing the Theory

If marginal wells are not the answer to the riddle, a meaningful drop in gas supplies is in store. To investigate the marginal well issue, we at Simmons & Company made use of the remarkable new online database created by the Texas Railroad Commission called the ACTI. It contains monthly production history of every oil and gas well in Texas, from January 1993 onward (Figure 3).

We selected 53 Texas counties to represent a wide profile of production by type of well, ranging from counties with the largest number of wells to counties with the lowest well productivity to counties with the highest production per well (Figure 4). These 53 counties had total January 2002 gas production of 8.3 bcf per day. This amounts to 66 percent of Texas' daily gas-well gas production (as distinct from "casing head" gas, a by-product of oil production).

Since Texas represents the single largest gas-producing state in the U.S., these counties also accounted for 16 percent of the entire country's total daily gas supply. Our survey looked at January 2002 production for all the wells still producing that had been completed over the past four years, separating them into "vintages" by year of completion. We analyzed the current production from 10,648 wells completed in these four years, of which 7,024 were still producing in January 2002.

We also analyzed all the "giant gas wells" (GGW) in production that month, defined as any well still producing in excess of 3 mmcf per day. In addition to approximately 1,800 wells completed between 1998 and 2001, another 104 wells completed prior to the beginning of 1998 still produced in excess of 3 mmcf per day in January 2002.

Key Conclusions

A variety of important conclusions and endless interesting observations sprang from this extensive well-by-well analysis. The most important observation was the lack of any strong evidence in support of the "marginal well thesis" throughout the 53 counties surveyed.

The production per well from these 53 counties varied enormously. But even in the low well productivity counties, the new wells completed in 2000 and 2001 seemed to have well volumes close to those seen in prior years. In some cases, there was evidence that current peak production is higher than in the past, bringing with it higher declines, too.

In fact, the survey highlighted how the state of Texas' gas supply is now anchored by a small number of highly prolific wells. These 53 counties had 393 giant gas wells, which amounted to only 1 percent of the total gas wells in the survey, yet accounted for 28 percent of total gas produced.

Of the 393 giant gas wells in the 53-county sample, 167 were completed in 2001, 43 percent of the total. These 167 wells accounted for 14.5 percent of the total 53-county production, or half of all production contributed by almost 2,300 2001 gas wells still producing in January 2002. The growth in total gas produced by these 167 wells is illustrated in Figure 5. Total production appears to have peaked in December and was already in decline in the first two months of 2002. The average life of these wells from the date of first production is seven months. The average life from the date of peak production is five months.

The production coming from gas wells completed over the last four years, plus 104 other giant gas wells from prior periods of time, amounts to 65 percent of all gas produced in these 53 counties. The other 49,000 gas wells now account for only 35 percent of this 53-county gas supply. This has created a remarkable gas pyramid.

A pyramid like this works well as long as new wells are constantly added. If only giant gas wells are added, supply can also possibly stay flat with even fewer wells added. But such a pyramid creates a high supply risk if any significant drilling decline occurs. To keep supply flat, an exponential growth in new gas wells is imperative.

Once the 53-county survey was finished, we also examined the annual gas production throughout Texas on a county-by-county basis. This led to another surprising finding – that Texas maintained eight years of flat gas production despite steep declines occurring in many key-producing counties, and even during a gas drilling boom.

Flat gas supplies were maintained by a handful of counties, often bolstered by the small number of giant gas wells that were rapidly growing their production fast enough to offset substantial production declines occurring in many other parts of the state.

Texas has 192 counties that produce natural gas. Between January 2001 and January 2002, 144 of these counties suffered a fall in production from 10.3 bcf/day to 8.8 bcf/day, a drop of almost 15 percent despite a record-setting drilling boom.

By contrast, 48 counties increased production from 4.3 bcf per day to almost 5.3 bcf per day over the same 12 months, which resulted in the impression that Texas gas production was staying flat. Of the 48 counties with higher production, five key counties made up almost half of the daily increase.

This county-by-county analysis proves that gas supplies can fall fast, even during a drilling boom. The 91 counties whose combined gas production totaled 34 percent of the state's total supply suffered a 26 percent decline over the course of 12 months, even during Texas' greatest drilling boom.

It is hard to imagine how much further supply would have fallen had the drilling boom not occurred. It also suggests that since this boom ended a year ago, there is no way Texas gas supply can grow. The big question is how fast supply might decline once the full impact of the drilling collapse is felt.

Dire Predictions

A handful of observers now worry that gas supplies might fall by 5 to as much as 7 percent by the end of 2002. But most analysts dismiss these dire forecasts as being overly bearish.

Through mid-April this year, the best support for a modest supply decline was the fact that supply had only dropped a little, despite the rig count peaking a year ago.

Then came a series of quite large gas production drops for many of the best-in-class public E&P companies. A close examination of Texas rigs drilling, compared to the gas wells reported to the RRC as completed, highlights how dangerous it is to disregard a falling rig count by citing the lack of evidence showing up so far.

Texas's rig count peaked at 509 rigs in July 2001. By April 2002, the Texas rig count had fallen to 293 rigs at work, a decline of over 40 percent.

The number of gas wells completed monthly throughout Texas previously averaged 250 to 300 new wells per month. In mid-2000, these well completions began to rise steadily. They reached a peak at an all-time high of 469 in March 2002. This peaking date reflects both the time delays in drilling many of these significant wells and also the lag in reporting the well completions.

Since the rig count has now fallen so much from its peak, six to nine months from now Texas' well completions should drop back into the low 300 wells per month. If this happens, it would seem almost impossible that gas supplies could only fall a percent or two.

Despite the drilling decline, reported well completions were at peak levels six to nine months after drilling peaked. While gas prices have now doubled from their recent lows, gas-related drilling could soon rebound, but through the end of September 2002, no sign of a big drilling recovery is underway.

However, even a quick drilling rebound will probably not reverse a supply drop because such a large proportion of the state's supplies are dependent on the steady growth of giant gas wells, and drilling for these wells has also declined.

Gas supplies in Texas will fall. Calculating the timing and magnitude of the decline is difficult. The risk that a supply drop could be significant is too serious to ignore. How big a drop is likely? The fact that 75 percent of Texas counties suffered declines of 15 percent during a major drilling boom illustrated that a statewide drop of this magnitude or more could be realistic. Handicapping the odds of such an event is impossible.

Is Texas a National Model?

It is hard to accurately gross up the results of this 53-county gas survey to approximate the state of Texas. It is beyond sound analysis to even try to extend this sample to all of the U.S., despite the fact that these 53 counties represent 16 percent of total U.S. gas supply.

However, the 53 counties were specifically selected to serve as a reasonable proxy for not only the entire array of Texas wells but also the wide variance seen in the gas well productivity of states around the U.S.

Crockett County in west Texas, for instance, is a good proxy for a state like Kansas, with many wells but tiny per-well production. Brooks County in south Texas is a good representation for the best gas wells completed in the Gulf of Mexico. Logically, it is hard to see why this survey of 16 percent of the U.S. gas supply would not be a rough proxy for what is likely to happen to gas supply for the entire country.

U.S. gas rigs drilling declined by 45 percent. Texas rigs are down 44 percent. Both rig counts peaked at the same time. During the boom, the only states with higher growth rates were Oklahoma and Wyoming. The only states that fell from peak rig counts further than Texas were Oklahoma and New Mexico. The collapse in drilling was relatively uniform throughout the U.S.

Despite the fact that we surveyed production from almost 8,000 producing gas wells, there is no way to precisely guess the magnitude of the coming decline in gas supply, even from these 53 counties. It is even harder to use this data to precisely quantify what is likely to happen to total U.S. gas supply.

However, a decline in gas supply as little as 1 percent to 3 percent now seems almost impossible, once the full impact of a drilling collapse is finally felt. I think the U.S. will be fortunate if the decline is only 10 percent. It could be far higher.

Regardless of how much gas supply will ultimately fall, based on sampling individual gas wells in this 53-county survey, the full impact of the possible gas supply decline is unlikely to be felt until the fourth quarter of 2002 or 1st quarter of 2003.

By then, a robust drilling boom could be underway, assuming the industry has a sufficient inventory of drilling prospects. But the lag effect and a far higher decline rate now challenges the industry more than ever (and there is no sign of this renewed boom occurring any time soon).

Market Reactions

Each time prices soar and subsequently crash, it further erodes the confidence of oil and gas operators to begin drilling more wells when prices begin to rise.

If a big drop does occur, the dynamics of supply could make it hard, if not impossible, for the industry to build supply back to the levels we enjoyed for the past eight years, let alone ever grow supplies to meet a 30 TCF market by 2010 or even 2015.

If supply drops more than a marginal degree, there is also a risk that gas prices will face another violent and unhealthy upward explosion.

But early indications should be but a trickle compared to the production loss that should show up when a decline in completing the giant wells finally appears. The stunning drop in 75 percent of the Texas counties that produce the USA's biggest supply of gas, even during a drilling boom, attests to how fast gas supplies can drop.

The biggest risk embedded in a supply drop is that a new drilling boom might merely stabilize gas supply at the new lower supply level. Finding a way to return gas supplies to the 52 bcf per day base that the U.S. enjoyed for the past eight years might take years. If the industry's drilling boom did, in fact, use up many drill sites that were planned for 2002 and 2003, the industry will have a hard time quickly responding to a high gas price scenario, regardless of how attractive the economics might become.

How Much Could Supply Drop?

The unanswered questions from this intensive analysis are how far U.S. daily gas supply could drop, given the 45 percent drop in drilling new gas wells, how fast the drop will occur and when it will bottom out. The top 30 U.S. oil and gas producers' production results for the first quarter of 2002 showed a drop in gas supply from the fourth quarter 2001 that was a surprise even to most of the reporting companies.

When all the acquisitions and divestitures are removed, total gas production of these 30 companies fell by almost 3 percent in just one quarter. But half of these companies, producing 75 percent of the total gas, experienced average declines of 3 to 10 percent. The best half, accounting for 25 percent of total production, actually grew their gas supplies.

While these 30 companies represent over half of total U.S. production, they are probably not representative of all gas operators. The bottom 50 percent of U.S. gas supply comes primarily from smaller companies with no access to external capital. These companies were likely forced to curtail drilling earlier than the 30 large public companies did. It would be strange if these smaller companies did not suffer a more severe drop than the best-in-class larger public companies.

A Worst-Case Scenario

Handicapping the odds of a severe supply drop is difficult. But the probability that it could happen is too high to ignore. Public policy energy planners and major users of gas now need to begin planning for the various unforeseen consequences that will occur if gas supply does drop.

Should a material drop occur, it seems unlikely that any drilling boom could grow daily supply enough to get back to a 52-53 bcf per day U.S. base in any reasonable period of time. The 2001 drilling boom, if carefully assessed, was unsustainable. All useable rigs were working. Rigs were drilling for gas, not oil. A large percentage of these rigs were applied to development instead of exploration wells.

If the U.S. has a sharp fall in gas production it creates an urgent need for a series of action programs, government-assisted or voluntary, to encourage a steady growth in drilling an ever-increasing supply of giant gas wells and to expand drilling in high impact areas like deep gas formations in the shallow waters or the Gulf and deep formation Rocky Mountain wells.

In order to mitigate against any further unforeseen drilling collapses, some form of price floor might be necessary. A price ceiling might also be needed to keep gas prices from rising to destructive levels.

If 1 percent of Texas wells could essentially overcome a decline in production throughout most of Texas' remaining supply, this also points to another action. If the industry could crank up its drilling efforts to begin completing hundreds of these highly prolific wells, this would go a long way toward reversing a production drop.

But giant gas wells are typically deep, take a long time to drill and are very expensive. In a highly volatile price scenario, it is unlikely that enough new wells with peak productivity to make a difference could materialize fast enough to make a difference. Solving the gas supply problem through a relentless growth in new giant gas wells also creates an imperative that such an effort should never slow down.

The decline curve for these wells is so steep that it takes exponential growth in these wells merely to keep production flat. There is, of course, also a finite number of prospects for such wells. A supply drop of even a modest degree also highlights the importance of creating the necessary infrastructure to bring Arctic gas to the lower 48 states as fast as possible, along with a rapid expansion of LNG terminals and LNG facilities to receive imported gas from overseas locations.

If supply falls by 10 percent or more, the concept of a single Arctic gas pipeline suddenly becomes barely adequate. Two Arctic lines become almost mandatory if natural gas is to remain a key energy source. The cost of two sizeable Arctic pipelines could exceed $40 billion and take a long time to build.

The 2002 Offshore Technology Conference saw announcements of projects using both compressed natural gas on vessels and a new generation of vessels that import LNG and discharge the gas into conventional pipelines instead of expensive and difficult-to-site offloading LNG terminals. Both become far more important alternatives if gas supplies start to drop by any meaningful degree.

If the pending supply drop is severe, it is time for America to abandon its paranoia about exploring and producing gas in our offshore basins outside the Western and Central portions of the Gulf of Mexico. Natural gas spills do not happen. No one has ever shown that developing offshore natural gas creates any environmental risk.

Placing an offshore platform or pipeline clearly impacts the environment. But there is no evidence it hurts the environment. A major supply drop will create a painful wakeup call for Americans to begin to distinguish between an event that impacts the environment and one that hurts the environment.

Since America is the largest energy consumer on earth, it makes no sense to ban trying to locate natural gas supplies in the eastern portion of the Gulf of Mexico or offshore New England, let alone the potentially gas-starved Pacific Coast states, simply because of environmental scare tactics.

If natural gas is not our energy future, as so many people have assumed for so long, then the only realistic way for the U.S. economy to continue to prosper is to embrace more coal-fired power plants and initiate a return to more nuclear plants. Coal gasification, the most basic way to process coal into its cleanest state, becomes an idea that was 25 years before its time when it became the poster child of the Carter Administration's energy solutions.

Renewable energy sources take on a far higher urgency, but the tough limits keeping sources like wind and solar at only a tiny piece of the energy mix are real barriers. For the foreseeable future, the U.S. will be highly dependent on natural gas, coal and nuclear to heat and cool homes, power industry and light up our offices and information highways.

As unpleasant as this may sound to environmentalists' ears, if natural gas cannot do the job, there is no real alternative to providing energy to the U.S. economy than tilting heavily to coal and nuclear. The industry and the government will simply have to find ways to make coal clean, resolve the "spent fuel" issue and replace the aging nuclear plants that now impede any bright prospect for nuclear energy's future.

The Eye of the Hurricane

Two years ago, a convergence of events – cold winter, humid and hot summer and a surging economy – created a "perfect storm" which sent gas prices soaring to their $10 peak. Quickly, prices fell, finally bottoming at between $1.85 and $2.10. Many observers credited this fast correction to the market efficiencies of a deregulated market-driven energy business.

But this rapid change could merely have been "the perfect storm" passing into a typical "eye of the hurricane."

The real reason gas prices plummeted was that gas storage went from almost empty to safely full. The "savior" was 5 bcf per day of reduced demand. Half of this came from abnormally mild winter and summer weather. Most of the balance came from the destruction of key industrial markets through extremely high prices.

A tiny bit came when NGL processing margins turned negative and some NGLs were left in the gas stream, artificially boosting natural gas supply by 1.0 bcf per day during the first quarter of 2001.

The eye of this perfect storm now seems to be passing. How bad the back-end of the perfect storm might be will be dependent on the extent to which natural gas supply falls in the third and fourth quarters of 2002, how quickly gas drilling rebounds, how long the lag effect takes before a drilling upturn halts the supply drop and what happens to natural gas demand over this same period of time.

Conclusion

America's most precious energy supply, natural gas, is in for a major jolt. The country is now as exposed to a serious energy shock as during the 1973 Oil Embargo. But this time, geo-politics and OPEC are absent. This new threat is merely the by-product of a drilling slump.

Regardless of the extent of the pending supply drop, it now seems unlikely that conventional gas supplies can grow beyond the steady levels enjoyed over the past eight years through at least 2010. Whether supplies can even return to a 52 BCF per day level is also now a serious energy issue.

The unforeseen consequences are almost too numerous to list. A drop in natural gas supply of a severe magnitude is not a foregone conclusion, but it not a ridiculous exaggeration either.

For too many years, a looming number of concerns on the reliability of safe, low-cost energy in both oil and gas were dismissed as views of contrarians or inveterate energy bulls, and for too long, unplanned and unpredictable events like benign weather have kept serious energy concerns from being addressed.

Too many "flashing yellow" and subsequent "flashing red lights" were dismissed by many supposed energy experts as simply the normal cycles of energy. America has never been particularly good at predicting crises and warding them off. Our great strength has been how effectively we marshal our collective assets to cope with a great crisis.

If gas supplies suffer even a 10 percent decline by year-end, America will have an abrupt energy wake-up call. We will know the answer to this gas riddle by the fourth quarter of 2002. Hopefully, this article will prove to be more alarming than the real facts. We will know all of these answers soon.

Matthew R. Simmons is chairman and CEO of Simmons & Company International, a specialized energy investment banking firm. The firm has guided its broad client base to complete over 450 investment banking projects at a combined dollar value of approximately $56 billion.

Mr. Simmons was raised in Kaysville, Utah. He graduated cum laude from the University of Utah and received an M.B.A. with distinction from Harvard Business School. He served on the faculty of Harvard Business School as a research associate for two years and was a doctoral candidate.

Mr. Simmons founded Simmons & Company International in 1974. Over the past 28 years, the firm has played a leading role in assisting its energy client companies in executing a wide range of financial transactions, from mergers and acquisitions to private and public funding.

Today the firm has approximately 135 employees and enjoys a leading role as one of the largest energy investment banking groups in the world. Its offices are in Houston, Texas and Aberdeen, Scotland.

Mr. Simmons is a trustee of the Museum of Fine Arts, Houston and the Farnsworth Art Museum in Rockland, Maine. He serves on the board of directors of Kerr-McGee Corporation (Oklahoma City), the Atlantic Council of the United States (Washington, D.C.), the Initiative for a Competitive Inner City (Boston), Houston Technology Center (Houston) and the Center for Houston's Future (Houston). He is also on the University of Texas M. D. Anderson Cancer Center Foundation Board of Visitors (Houston) and is a charter member of the University of Houston National Advisory Council (Houston). In addition, he is past chairman of the National Ocean Industry Association. He serves on the board of directors of the Associates of Harvard Business School and is a past president of the Harvard Business School Alumni Association and a former member of the Visiting Committee of Harvard Business School. He is a member of the Council on Foreign Relations and the Advisory Council of the National Trust for Historic Preservation.