Bill Totten's Weblog

Thursday, December 16, 2004

The Energy Challenge 2004 - Natural Gas

by Murray Duffin (September 23 2004)


Natural gas is a more difficult subject to address than petroleum, because the data is much less complete and reliable, and because the USA situation appears much more precarious than the world situation. BP/Amoco statistics imply that at 1998 consumption rates, the world has about sixty years of resources remaining. However, known reserves are much lower, resource estimates are highly speculative, and the major resources (approximately 70%) are in the Middle East and former Soviet Union.

Natural gas can be readily transported by pipeline, but cannot be transported either in large quantities or economically by ship. Japan, Korea, and Taiwan have long-term contracts that lock up nearly all existing LNG shipping capacity. Europe may be able to depend on the Middle East and the former Soviet Union for several decades of natural gas supply. The USA does not have that luxury.

Because of transportation limitations, the USA has historically depended on North American natural gas. Mexico has long since reduced exports to zero and is now a small importer. Canada supplied about 16% of USA consumption in 2002, but has had very disappointing exploration results in recent years, and exports to the USA declined 8% averaged across 2003, reaching nearly a 14% year on year decline in the 4th quarter. Known USA reserves represent about eight years' supply at recent consumption rates, while demand has been projected to grow by more than 50% during the next twenty years. In 2003 the EIA revised their USA production projections from steady growth to essentially flat through 2020, and even that seems very doubtful in the light of recent trends. In spite of major increases in drilling, production in North America declined at least 3% in 2003 versus 2002, and is down another 3% YTD 2004.

Included in natural gas resource estimates are:

<> Associated resources - discovered along with oil fields, through drilling for oil.

<> Non-associated resources - free flowing natural gas discovered without petroleum.

<> Tight gases - natural gas in dense shale or sandstone deposits that requires extensive drilling and fracturing to recover.

<> Coal bed methane - gas released from coal deposits that again requires extensive drilling and fracturing to recover.

Estimates for total USA resources vary widely from about 300 to 1,400 trillion cubic feet (tcf), and methods of estimating are very imprecise and speculative. Background data is not freely available to the individual, but databases can be accessed at the cost of a few thousand dollars. It seems likely that the higher resource numbers result from arithmetic addition of low probability estimates, and may therefore be meaningless. A number near 1,100 trillion cubic feet or fifty years is widely used, but is a very risky multiple of proven reserves. The hard data we do have is not encouraging. What we do know is:

<> Drilling for natural gas in the five years from 1980 through 1984 was about double the average during the decade of the 1990s, but annual average discoveries were slightly less.

<> Because of the bad experience with wildcat drilling in the early 1980s, drilling in the 1990s tended to be concentrated near known large basins, extending their boundaries but not making major new finds.

<> 9,000 new gas fields were discovered from 1977-87, but only 2,500 from 1987-97.

<> With the application of new technology, especially hydraulic fracturing and horizontal drilling, initial production of new fields has been kept nearly constant for two decades, but depletion time has been shrinking rapidly. New wells average 56% depletion in the first year of production. Congressional testimony in 2004 stated that some tight sands wells deplete 50% in less than six months.

<> Wells drilled in 2000 were 60% above 1999, and early 2001 were about 50% above 2000. Production grew less than 2% in 2000, and less than 1.5% in 2001. After falling off in late 2001 and early 2002, drilling has increased steadily for the last two years while production continues to decline.

<> New finds are becoming progressively smaller.

<> Proved reserves of natural gas in the USA declined from a peak of 290 trillion cubic feet in 1967-70 to 167 trillion cubic feet in 1989, and, with some fluctuation, have been flat since, in spite of a major drilling peak in the early 1980s as noted above.

<> For twelve years through 2001, discovery just kept pace with production, and consumption growth was served by increasing imports.

<> Of 1999 EIA estimated resources of 1,280 trillion cubic feet, 890 trillion cubic feet were classified as "undiscovered", and 220 trillion cubic feet as expected reserve growth. (Most of the discovery in the 1990s was reserve growth. How much can be left?)

<> Natural gas production in the USA peaked in 1973.

<> Natural gas supply from the Gulf of Mexico (GOM) shelf is in decline.

<> Natural gas discovery in the deep Gulf of Mexico is much lower than expected, and NRG Associates in 2001 projected peak supply as 3 trillion cubic feet in 2007 versus the National Petroleum Council forecast of 4.5 trillion cubic feet in 2010.

<> Simmons has noted that rig count in the Gulf of Mexico grew 40% from April 1996 to April 2000, and 60% in Texas from January 1996 to October 2000, with production remaining flat.

There is nothing in the known facts to support an optimistic resource estimate. Clearly the natural gas industry has to rapidly accelerate drilling, just to keep production flat. A large increase in wildcat drilling in the early 1980s didn't help and seems to be not helping much again.

Is Alaska going to help? Resources are projected by the EIA as 237 trillion cubic feet, but proven reserves are only 10 trillion cubic feet. (Does that make you wonder?) A three-foot -diameter pipeline, moving gas at 2,200 feet per second <1> would deliver only 0.5 trillion cubic feet per year, less than 2% of 2020 needs. The energy to move the gas increases with the square of the velocity, and, at this velocity, would require more than 2% of the gas moved just to drive the compressors. It may not be economical to build a 2,000-mile pipeline. (Maybe the natural gas can be converted to liquid syn fuel in situ and shipped via the existing oil pipeline?)

The National Petroleum Council has forecast natural gas demand as 29 trillion cubic feet in 2010, and the EIA as well as the NEPDG projected demand of 40 trillion cubic feet by 2020. Rising prices have already severely dampened such demand growth, with at least 25% of 2001 industrial demand already having been destroyed through closures or moves offshore. Unfortunately, because of low prices and high availability in the late 1990s, and to meet emission restrictions, almost all new electricity generation capacity built in the last decade has been gas fired, with the bulk of it coming on stream in 2002/3. Only unusually benign weather during the last fifteen or so months has saved us from severe natural gas shortages up to now.

To make matters worse, if a curve of USA discovery is superimposed on production with a shift of 28 years, the two curves match very well up to now. However, discovery went into a sharp decline about 28 years ago, so we can expect a similar decline in production soon. (We can't produce what hasn't been discovered). Production is likely to "fall off a cliff" and be down by half before 2015. See, Figures 78 and 89.

The Current Situation

While natural gas supply tends to be quite flat year round, demand is distinctly seasonal, with highest demand for winter heating load. Demand varies widely even week-to-week dependent on "temperature degree days" (TDDS), either "heating degree days" (HDDs) in winter or "cooling degree days" (CDDs) in summer. Winter demand is higher than production, requiring a build up of storage during spring, summer and fall to supply peak demand in winter. This splits the year into two main seasons, about thirty weeks from near 1 April to near 30 October, the "injection" (to storage) season, and about 22 weeks, 1 November to 31 March, the "withdrawal" (from storage) season. To get a good comparison from year to year one needs the "gas weighted" HDDs and CDDs.

Historically injection has varied from a low of 1600 billion cubic feet to a high of about 2450 billion cubic feet, mainly depending on how low storage got during the previous withdrawal season. The critical minimum storage level is from 700 to 900 billion cubic feet. Above 900 billion cubic feet the system operates fairly smoothly. Below 700 billion cubic feet there is a high probability of severe price spikes, and inability to deliver sufficient gas to users, due to low system pressure. Withdrawal also varies from a low of about 1800 billion cubic feet to a high of 2400 billion cubic feet, with the last ten-year average about 2000 billion cubic feet. Target end of injection season storage is above 3000 billion cubic feet.

In spring of 2003 storage got down to 720 billion cubic feet, and in February, when severe shortages were forecast, prices spiked to $18.00 per million Btu interday and $28.00 per million Btu intraday, versus a base level near $5.00 per million Btu. Because of this extremely low storage, Alan Greenspan began warning of an natural gas crisis in spring 2003. Then several factors worked together to provide record 2003 injections so that the 2003/2004 withdrawal season started with ample storage (3200 billion cubic feet). These factors were:

<> a substantial degree of industrial demand destruction in 2002 and 2003 due to high natural gas prices

<> a sharp increase in liquefied natural gas (LNG) imports, from about 200 billion cubic feet per year to 540 billion cubic feet per year

<> the decision by suppliers to leave natural gas liquids (NGL) in the supply, supplying wet gas instead of dry gas

<> an unusually mild summer with low CDDs until late August, that sharply reduced demand for peak electricity generation.

These factors were enough to offset declining production and rebuild adequate inventory for the withdrawal season.

During 2002 the USA used about 23.6 trillion cubic feet of natural gas, with about 16% imported from Canada by pipeline and a little under 1% supplied as imported LNG. During 2003 domestic production dropped sequentially by quarter by about 1%, 2%, 3% and 4%, for an average just below 3%. Imports from Canada dropped about 3%, 6%, 9% and 12%, for an average of near 8%. With supplies holding flat in 2004 at 4th Quarter 2003 levels, we can expect domestic production to be down 3%, 2%, 1% and 0% for an average of 1 to 1.5%, and imports from Canada to be down 9%, 6%, 3% and 0% for an average of 4%. Offsetting these declines, LNG imports should be up about 200 billion cubic feet, limited by regassification terminal capacity. Net, total supply was down about 1.5% in 2003 versus 2002, and 2004 will probably be down another 1.5% versus 2003. Demand has grown in both years due to high housing additions (80% of new houses are heated by natural gas), increased electricity generation fired by natural gas, and increased economic activity. These increases were fairly modest in 2003, mainly due to weather, but are quite sharp in 2004. During the last three weeks in May 2004 total electricity generation was up 5% year on year after adjustment for TDDs and holidays, and manufacturing activity is up sharply since late 2003.

Liquefied Natural Gas (LNG)

Most optimists and many analysts believe that LNG imports will save our bacon. That is not the case. Several countries with stranded natural gas are building liquefaction facilities so there will be plenty of LNG supply. The USA has expanded regasification facilities rapidly during 2002, 2003, and 2004 and has enough capacity for near term needs. New projects are approved or under construction to provide enough capacity to meet EIA import projections of 2.2 trillion cubic feet for 2010. The first problem is that at a decline rate for North American availability of 2% per year (remember 2003/4 declines have been much worse than this) we would need 2.6 incremental trillion cubic feet of LNG in 2010 to reach 2004 total natural gas supply level, and 3.7 trillion cubic feet to meet the 2002 level of consumption, not the 2.2 projected by the EIA. It is not clear that there will be enough regasification capacity to meet these numbers (there are a lot of NIMBY problems), and these numbers do not provide for any growth.

A much larger limitation is shipping capacity. The world LNG tanker fleet in Q1 2004 was 156 vessels, with 62 more on order for delivery through 2008. World shipbuilding capacity for LNG tankers is 20 ships per year. If this capacity is booked full another 50 or so vessels can be delivered by end 2008. All of the existing fleet is already under long-term contract, and not more than 18 of the vessels under order are available for shipments to the USA. Up to now about 40% of the USA LNG supply has come from Trinidad, 40% from North Africa, and the rest from the Middle East and Indonesia. Trinidad supply is now maxed out, but they will have some more capacity coming on stream in 2006. North Sea production is now in decline so Europe must become a much larger importer from North Africa. This means that most of the incremental USA supply will have to come from the Middle East, which means only about ten deliveries per ship per year. One modern ship has a capacity of about 2.6 to 2.8 billion cubic feet of regasified natural gas, but because of losses during transport, could only deliver about 2.3 billion cubic feet per trip from Qatar to the USA. At ten trips per year we would need an incremental 100 ships by the end of 2009 to meet 2010 demand equal to 2004 consumption, or 160 to meet 2002 consumption. Even if shipbuilding capacity is doubled by the end of 2006, (and the order book right now is not large enough to get that process started), and all of the incremental capacity went to serve USA demand, we would still be fifty ships short of minimum 2010 needs.

NB None of the above even considers North American production "falling off a cliff"! We can be confident that total USA NG supply in 2010 will be at least 1 trillion cubic feet less than the peak year of 2002, and there is a high probability that it will be less than half.

Other Sources

Users of natural gas have been urging the government to open up restricted areas for drilling. That will not solve the problem either. At least 40% of government lands in the Rockies are already open, but there is no great rush to drill them. Most of the land in question, where it contains gas, contains tight sands gas. It doesn't matter how much is available, it simply can't be produced fast enough to offset decline in the present major fields. The eastern Gulf of Mexico holds some promise, but deepwater Gulf of Mexico has been largely disappointing relative to early expectations so far. Coal bed methane has the producibility problem of tight sands plus a major environmental problem of water pollution.

If there were major attractive prospects in any of the areas in question the natural gas producers would be beating down the doors in Washington. That isn't happening, even with as energy corporation friendly government as we have now.


Natural gas is somewhere between a limit to growth and a disaster waiting to happen right now, and no one is doing anything about it. Only a few months of inclement weather will cause severe shortages and rocketing price spikes. There is a high risk of major availability declines with unimaginable economic impact, and there is no supply side solution. We urgently need a government driven, demand side oriented "Apollo Program" for energy. The wake-up call is already sounding, at least for natural gas, but the powers that be just aren't paying attention, and the energy industry, both users and suppliers of fossil fuel, are asleep at the switch or in denial. The bad news is that it looks like we will have to experience a prolonged crisis before there is any reaction to the danger. The good news is that the crisis is likely to happen real soon.


<1> The Fort St John, British Columbia to Chicago pipeline, completed in 1999, meets this specification. See Petroleum Review, November 2000, London, page 13.

Copyright 2004 CyberTech, Inc.
To post a comment, or to read comments posted by others,
please click on the word "comment" highlighted at the end of
the version of this essay posted at:

Bill Totten