Total world reserves (at end of 2002):
bituminous coal + anthracite	479 billion tons
subbituminous coal	272 billion tons
lignite	158 billion tons
Each coal class has a different energy content:
anthracite	30 MJ/kg
bituminous coal	18.8-29.3 MJ/kg
subbitiminous coal	8.3-25 MJ/kg
lignite	5.5-14.3 MJ/kg
wood	12 MJ/kg
coal	14-32.5 Mj/kg
oil	41.9 Mj/kg
natural gas	53.6 MJ/kg

On one hand, as more coal is discovered, as the price goes up, or as new mining machines are developed, coal reserves expand. On the other hand, as we extract and use enormous amounts of coal each year, we draw down those reserves.

One might expect that overall reserves figures would change fairly slowly and in a predictable fashion. In fact, as we will see, reserves figures for several nations have collapsed in recent years; and, over the past few decades, centuries’ worth of coal has disappeared from global reserves. Given that the world’s economy depends so heavily on coal, this trend is hardly reassuring. If we wish to understand how and why such downward reserves revisions are happening, it is essential that we look more deeply into the rather specialized, technical process of estimating coal reserves.

How Are Coal Reserves Estimated?

The estimation of coal reserves has evolved through the decades, and it now constitutes a sophisticated process entailing the work of thousands of trained and experienced coal geologists around the world.

The first step is to identify prospective areas. This is accomplished by means of old-fashioned, painstaking geological fieldwork, carried out with map, compass, and pick. Geologists typically look for coal outcroppings in rock strata exposed by streambeds or by ancient earth movements. Once a prospective area has been identified, cores are drilled to determine the thickness and depth of coal seams, as well as the quality and characteristics of the coal itself. These cores are carefully analyzed and logged to yield a three-dimensional map of the region. Then, using such maps, field sizes are estimated. Finally, reserves for entire regions are estimated by totaling field-by-field estimates.

No matter how carefully this process is pursued, it inevitably incorporates many judgment calls. Remember: reserves are defined not as the total amount of coal present (that’s the resource); rather, they consist of the portion of the resource that can be expected to be extractable at a profit using existing technology. Not only are reserves limited by resource quality, seam thickness, depth, and location, but analysts must also take into account the fact that the mining process will inevitably leave some of the resource behind. This is especially true in the case of underground mining, where in some instances a majority of the coal originally in place is left behind.⁵ Historically, practical recovery percentages for underground mining average about 50 percent of the coal that meets all economic criteria for minability; for surface mining, it is 85 percent.⁶

In the ideal case, all of these variables will have been taken into account when a final reserves number for a region or a nation is produced and published. However, ideal cases are rare.

The task of reserves analysts is made difficult, for example, by the fact that private coal companies often keep their data proprietary. Thus, when a public agency sets out to compile national reserves statistics, it may find significant gaps in available data. Moreover, some nations simply don’t have the personnel or funding needed in order to properly compile and update records.

Additionally, there is no single internationally recognized, uniform method for assessing and reporting reserves as a fraction of resources. In the United States, coal geologists work with the following carefully defined categories:

original resources
remaining resources
identified resources
inferred resources
measured resources
reserve base
inferred reserves
indicated reserves
measured reserves
marginal reserves, and
sub-economic resources.⁷

But other countries have their own sets of categories, with varying definitions. Assembling national reserves figures into a composite global picture is therefore a task of enormous complexity. One might expect that this would be the work of teams of data analysts working for the International Energy Agency (IEA) or some well-funded, prestigious institute. Surprisingly, the task is actually carried out by a two-person team — Alan Clarke and Judy Trinnaman, whose company, Energy Data Associates, is headquartered in Dorset, England. Clarke and Trinnaman send a questionnaire every three years to every coal-producing nation in the world. According to Clarke, about two-thirds of nations reply, but only about 50 of these replies typically are useful. Some reported data must simply be disregarded as unrealistic. No effort is made to verify reported national reserves figures through independent geological surveys.⁸

The figures from Energy Data Associates are then taken up in the triennial report of the World Energy Council, and are subsequently republished by the IEA, US Geological Survey, BP, etc.

Clarke and Trinnaman no doubt do an excellent service with the information available to them, but given the nature of this data the results can hardly be regarded with a high level of confidence.

Recent Studies of Coal Reserves and Future Supplies

As underscored in the Introduction, questions regarding future world coal supplies are not just academic. The global economy is more reliant on coal today than at any time in the past as total production is at the highest level in history. Meanwhile, the questions of how and whether the world continues its coal consumption are crucial to the fate of the global climate. Accurate coal reserves figures and supply forecasts are therefore more important than ever, as the world plans its energy strategy for the remainder of this century.

The common assumption that the world has plenty of coal has been the subject of several recent studies. Five of the most important of these will be summarized below, and we will return to them in the next four chapters as we examine the status of coal reserves and production in the world’s main coal countries. These studies do not all reach the same conclusions, but they represent the best, most recent data and analysis available.

1

“Coal: Resources and Future Production” (Energy Watch Group). The organization Energy Watch Group (EWG) was founded by German parliamentarian Hans-Josef Fell, and is supported by the Ludwig-Bölkow-Foundation. Its mission is to assess future supplies of fossil and atomic energy resources, and develop scenarios for renewable energy sources and strategies for a long-term secure energy supply at affordable prices. The EWG report, “Coal: Resources and Future Production,” was released in March 2007.⁹ Its central conclusions are that minable global coal reserves are much smaller than is commonly thought, and that a peak in world coal production is likely within only ten to 15 years.

The report’s authors — Werner Zittel (Ludwig-Bölkow-Systemtechnik GmbH, Ottobrunn) and Jörg Schindler (Managing Director, Ludwig-Bölkow-Systemtechnik GmbH) — state their opinion that “the data quality [for coal reserves] is very unreliable,” especially for China, South Asia, and the former Soviet Union countries. Some nations (such as Vietnam) have not updated their “proved reserves” for decades, in some instances not since the 1960s. China’s last update was in 1992; since then, 20 percent of its reserves have been consumed, though this is not revealed in official figures.

Even more striking is the fact that since 1986, all nations with significant coal resources (excepting India and Australia) that have made efforts to update their reserves estimates have reported substantial downward revisions. (In its 2007 survey, the World Energy Council noted that India reduced its reserves from 92Gt to 56Gt; this survey was published after the EWG report.) Some countries — including Botswana, Germany, and the United Kingdom — have downgraded their reserves by more than 90 percent. Poland’s reserves are now 50 percent smaller than was the case 20 years ago. Each new assessment (again, except in the cases of India and Australia) has followed the general trend. These downgrades cannot be explained by the volumes of coal produced in this period. The best explanation, according to EWG, is that nations now have better data from more thorough surveys. If that is the case, then future downward revisions are likely from countries that continue to rely on decades-old reserves estimates.

Fig. 3

Fig. 4

The report concludes: “present and past experience does not support the common argument that reserves are increasing over time as new areas are explored and prices rise.” This conclusion is supported by the fact that even the world’s in-situ resources of coal have dwindled from 10 trillion tons of hard coal equivalent (hce) in 1980 to 4.2 trillion tons in 2005 — a 60 percent downward revision in 25 years.

The EWG group performed a peaking analysis of world coal, and arrived at the conclusion that world production will reach a maximum level around 2025, decline slowly for about two decades, and then fall off more rapidly beginning around 2050.

2

“A Supply-Driven Forecast for the Future Global Coal Production” (Höök, Zittel, Schindler, Aleklett; Uppsala Hydrocarbon Depletion Study Group). Two of the authors of this 2008 report were also responsible for the EWG study, so it might be expected that the conclusions of both would be similar — and this is indeed the case. However, the newer report is also somewhat more thorough. Future global coal production is forecasted using a logistic growth model, as well as experience from historical reserve and resource assessments. The result is the same:

Global coal production will be able to increase over the next 10 to 15 years by about 30%, mainly driven by China, India, Australia and South Africa. A plateau will be reached around 2020 and the global production will go into decline after 2050.¹⁰

The authors again show that throughout the past century both reserves and resources have been constantly downgraded in most nations. Competition from other energy sources and the introduction of various political restrictions were involved in these downward reductions, but a main driver of the trend has been better geological understanding of actual available coal amounts.

It is pointed out that optimistic forecasts of future supply — see the BGR report, below — rely on the potential of new technology to turn resources (coal in the ground) into reserves (coal that we believe can be economically extracted). The authors note that, during the past century, the introduction of new technology for exploration and mining has had very little impact on the available coal reserves. They assume that this situation will continue:

[T]he world’s future coal supply likely is overestimated, as the entire concept of having resources upgraded to reserves needs to be reassessed, since it is something that has not happened throughout history on a significant scale. More careful studies of this need to be undertaken and all details in the mechanism of resource upgrading should be examined to find a suitable behaviour for future production forecasts.¹¹

The authors note that, “Throughout history resources have not mattered much, and unless something causes a significant deviation in the historical trend, they will not matter much in the future.” They call for “better data and a more transparent and reliable system for reserve evaluations” in order “to form a solid basis for long-term decisions and forecasts regarding the energy system.”

3

“The Future of Coal.” This study, by B. Kavalov and S.D. Peteves of the Institute for Energy (IFE), prepared for the European Commission Joint Research Centre and published February 2007, questions future supply, but does not attempt a peaking analysis.

While Kavalov and Peteves discuss future supply in terms of the familiar but misleading reserves-to-production (R/P) ratio, nevertheless, the IFE’s conclusions broadly confirm those of EWG.

The three primary take-away conclusions from this study are as follows:¹²

• “World proven reserves (i.e., the reserves that are economically recoverable at current economic and operating conditions) of coal are decreasing fast ... .”

• “The bulk of coal production and exports is getting concentrated within a few countries and market players, which creates the risk of market imperfections.”

• “Coal production costs are steadily rising all over the world, due to the need to develop new fields, increasingly difficult geological conditions and additional infrastructure costs associated with the exploitation of new fields.”

Early in the paper the authors ask, “Will coal be a fuel of the future?” Their disturbing conclusion, many pages later, is that “coal might not be so abundant, widely available and reliable as an energy source in the future.” Along the way, they state “the world could run out of economically recoverable (at current economic and operating conditions) reserves of coal much earlier than widely anticipated.” The authors also highlight problems noted in the EWG study having to do with differing grades of coal and the likelihood of supply problems arising first with the highest-grade ores.

All of this translates to higher coal prices in coming years. The conclusion is repeated throughout the IFE report:¹³ “[I]t is true that historically coal has been cheaper than oil and gas on an energy content basis. This may change, however. ... The regional and country overview in the preceding chapter has revealed that coal recovery in most countries will incur higher production costs in [the] future. Since international coal prices are still linked to production costs ... an increase in the global price levels of coal can be expected.”

As prices for coal rise, “the relative gap between coal prices and oil and gas prices will most likely narrow,” with the result that “the future world oil, gas, and coal markets will most likely become increasingly inter-related and the energy market will tend to develop into a global market of hydrocarbons.”

4

Hubbert linearization and curve-fitting (studies by David Rutledge,¹⁴ Jean Laherrère,¹⁵ et al.). In the early 1980s, geophysicist M. King Hubbert (1900-1989) — who is generally credited with having pioneered the scientific study of oil depletion — developed a mathematical technique for forecasting ultimately recoverable figures for oil and the timing of production peaks using only production statistics. There are some who think that this technique can also be used to forecast future coal supplies.

Hubbert introduced the methodology, now known as “Hubbert linearization” (HL), in a paper titled “Techniques of Prediction as Applied to the Production of Oil and Gas,” published in 1982.¹⁶ It was later explained in some detail by Kenneth Deffeyes in his book Beyond Oil: The View from Hubbert’s Peak.¹⁷

The assumption inherent in HL is that the ability to produce a non-renewable resource depends entirely, and linearly, upon the unproduced fraction of the recoverable resource at any point in time. Simply put, this is a mathematical way of modeling the fact that we tend to find and produce the most accessible portion of the resource first, so that production requires more effort over time.

Cumulative production is logged on the horizontal axis of a chart, while the ratio of annual production to cumulative production is plotted on the vertical axis (P=production, Q=cumulative), using the equation P = 008

where a is annual production expressed as a fraction of cumulative production. In early years, production is naturally low, but it is a high percentage of total cumulative production. As time goes on, the cumulative figure goes up, but each year’s production is a smaller percentage of the cumulative amount to that date. Thus an entire production history tends to assume a more-or-less straight, downward-trending line.

If production is constrained for part of that time, or, on the other hand, if it is temporarily stimulated, the line will diverge from its previous path. If the method is applied too early in production history, its results will be fairly useless because it takes some time for the linear trend to appear. Also, the technique works best when assessing a large region: in a small area, a single new discovery occurring late in the production cycle can skew the production trend considerably, rendering the earlier trend-line misleading.

Nevertheless, if the region is large and if enough time has passed to enable the data to show a clear trend, it is possible to project that trend-line to the bottom horizontal axis to forecast the ultimately recoverable amount of the resource.

The method has worked well in forecasting ultimately recoverable amounts of oil in many producing nations such as the United States, the United Kingdom, Mexico, and Oman. There is no obvious reason it should fail to apply also to other non-renewable resources such as coal.

David Rutledge, the Tomiyasu Professor of Electrical Engineering and Chair of the Division of Engineering and Applied Sciences at the California Institute of Technology (Caltech), in a presentation at Caltech in October 2007, used the technique to estimate future global coal production, breaking the world up into eight regions — Australia, South Asia, East Asia, former Soviet Union, Africa, Europe, South America, and North America — and studying production statistics for each separately. Where no trend became apparent from the application of HL, Rutledge used reserves figures from the World Energy Council (WEC).

Fig. 5: Hubbert linearization plot for US oil production from the lower 48 states, showing an ultimate recovery of 225 billion barrels.

Region	Reserves (Gt)	Trends
North America	255	135
East Asia	190	70
Australia and New Zealand	79	50
Europe	55	21
Africa	30	10
Former Soviet Union	223	18
South Asia	111
Central and South America	20
World (at 3.6 boe/ton)	963 (3.5Tboe)	435 (1.6Tboe)
boe = barrels of oil equivalent Tboe = trillion barrels of oil equivalent ¹⁸

Rutledge found about half the reserves officially accepted by the WEC are likely actually to be produced, assuming the HL method holds.

In subsequent communication I’ve had with Rutledge, he noted that he does not consider HL useful for estimating the timing of the global coal production peak. However, as noted above, the method is used routinely by others for forecasting peaking dates for oil production within individual countries and the world as a whole.

Once about half the reserves are gone, it becomes progressively more difficult to maintain the same or a growing rate of production. The straight-line HL graph can be converted into a logistic curve, with the quantity beneath the curve equal to the ultimately recoverable amount forecast by the first graph. All that is necessary is to take a reciprocal of both sides of the equation:

Thus, in principle it is possible to obtain both an ultimately recoverable estimate and a peak production year forecast if there exists a sufficiently robust data set for production over time.

Fig. 6

Veteran petroleum geologist Jean Laherrère, who served as deputy exploration manager for TOTAL and has written several reports for Petroconsultants and Petroleum Economist on the world’s oil and gas potential and future production, examined the data for world coal reserves and production and produced the HL graph of future coal production reproduced here, showing a peak before 2050.¹⁹

It is worth noting parenthetically that Laherrère prefers to use round numbers in his calculations. He argues that reserves estimates even to one decimal place give a false impression of accuracy, where in fact the numbers are fluid, arguable, and imprecise. This attitude strikes me as being refreshingly realistic.

5

“Lignite and Hard Coal: Energy Suppliers for World Needs until the Year 2100 — An Outlook.” It would be wrong to give the impression that all recent studies have yielded pessimistic results regarding world coal reserves.

Thomas Thielemann, Sandro Schmidt, and J. Peter Gerling of The German Federal Institute for Geosciences and Natural Resources (BGR) have published in the International Journal of Coal Geology a report that forecasts no foreseeable bottleneck in coal supplies and a large potential for coal-to-liquids (CTL). The article’s abstract states:

For three years, international hard coal prices have been at rather expensive levels. Some argue that these higher prices might indicate the threat of a physical scarcity of [coal] — similar to the situation with oil and gas. This is not true. The supply situations with lignite and hard coal appear to be largely not critical. Adjusted to the rise in global coal consumption, which is expected until 2100, nature by and large can meet the world’s coal demand. ...The only area of potential concern is Asia (especially China). But today’s and coming eager efforts in China to convert coal resources into reserves will most likely deliver the coal needed for the Chinese market.²⁰

The conclusion of the report is unequivocal: “Up to the year 2100, and from a geoscientific point of view, there will be no bottleneck in coal supplies on this planet.”

The national coal reserves data used in this report are the same set used by the Energy Watch Group.

The BGR authors also assume a slowdown in world demand for coal, from the current six percent annual growth (largely resulting from consumption patterns in China) to two or three percent. They admit that their demand forecast looks “rather conservative,” and that stabilization of current high consumption growth is “uncertain.”