Estimating Coal Production Until 2100

Several different models have been devised to predict total remaining worldwide coal yield.  Projections of mineral resources are often made using the Hubbert linearization model, a method invented in 1976 that has been quite accurate in estimating oil production in the United States. Coal reserves have also been estimated using a method that sums estimates of current reserves with estimates of cumulative production.  S.H. Mohr and G.M. Evans devised a new methodology for estimating coal production that takes into account supply and demand interactions in different countries and for different types of coal (Mohr and Evans, 2009).  These three models are compared based on the predicted amounts of coal production by country, production levels by type of coal, and energy produced by the estimated amounts of coal.  The model devised by Mohr and Evans, the “Best Guess Method,” predicts that overall worldwide coal production will last much longer and provide more energy than the predictions of the Hubbert linearization model, but the model that sums reserves and cumulative production predicts higher coal production than the Best Guess and the Hubbert methods.  Based on the Best Guess methodology, the worldwide coal yield will between 700 and 1,243 gigatonnes, and worldwide coal production will peak in 2034 on a tonnage basis, and in 2026 on an energy basis.  The Best Guess model indicates that the notion that coal is widely abundant appears to be unjustified.— Caitrin O’Brien
Mohr, S., Evans, G., 2009. Forecasting coal production until 2100. Fuel 88, 2059-2067.
S.M. Mohr and G.M. Evans created a model for estimating coal production that takes into account more than 400 constants for 132 countries and many different coal types.  This new model includes data from external disruptions, such as wars and depressions, and can be used for any resource where production is derived from mining.  Mohr and Evans compared their new model to two more traditional models used to estimate coal production.  The Hubbert linearization method was devised in 1976 to plot production data for the US oil industry, and this method has been used to plot the depletion of other finite mineral resources.  This model assumes production to be a symmetric bell curve, and does not take into account estimated reserves of coal.  The second method that Mohr and Evans compare is denoted as the “R+C” model, which is the summation of current estimates of coal reserves and cumulative production worldwide.  The authors used these three methods to compare the coal production for each country or region of the world, in gigatonnes.  The three models were also compared based on their results for estimated production based on type of coal, and the amount of energy produced by each type of coal. 
The scientists found large discrepancies between the three models of coal production.  When comparing the scenarios based on coal production predictions, the R+C scenario predicted much higher overall coal production than the other two, and the Best Guess scenario is in between the Hubbert estimate and the R+C estimate.  All three of the models show China running out of coal around 2100, but the R+C scenario and the Best Guess scenario both predict a huge increase in coal production in the Former Soviet Union around 2110.  The Hubbert method shows very little production in any region past 2100, whereas the other two scenarios estimate that coal production will last beyond 2200, especially in the United States.  All of the scenarios show that Western Europe is the only continent where coal production has already peaked and is declining, and the study shows that European coal production peaked in 1988 and is declining at a rate of 3% per year.  All of the models indicate that world coal production in tones will peak between 2010 and 2048. 
Mohr and Evans compared worldwide coal production for each of the three models based on the amounts of coal estimated for each model.  Four different types of coal were considered- anthracite, bituminous, sub-bituminous, and lignite.  For all three scenarios, bituminous coal is the most prevalent, because bituminous coal is the most common coal in the world. Anthracite estimates are very low and approximately the same for each scenario. The Hubbert method estimates that the coal that is produced worldwide will be primarily bituminous, whereas the Best Guess and the R+C methods predict a huge spike in lignite extraction.  This is presumably because as bituminous coal begins to run out, the energy industry will increase production of less efficient types of coal such as lignite and sub-bituminous.
Finally, the authors compared the three models based on the energy values of the predicted coal extraction.  All models predicted that anthracite would produce high amounts of energy compared to its mass, although bituminous coal will produce the most overall energy because it is the world’s most prevalent coal.  The Hubbert method estimated that by the year 2100, less than 30 exajoules per year will be produced by coal worldwide, whereas the R+C scenario’s estimate for the same year is about 60 exajoules per year. The Best Guess methodology predicts that by 2100, over 90 exajoules per year will be produced by coal.  The author’s Best Guess scenario indicates that the amount of worldwide energy produced by coal will peak between 2011 and 2047.  Mohr and Evans conclude that the worldwide coal yield is between 700 and 1,243 gigatonnes, and that worldwide coal production will peak in 2034 on a tonnage basis and 2026 on an energy basis.  According to the research by Mohr and Evans, the world has almost reached its maximum levels of coal production, and by 2100 will have all but exhausted worldwide coal supplies.  

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s