80Scientific American March 1998
The End of Cheap Oil
logical breakthroughs justified the addi-tion of a staggering 287 Gbo. That in-crease is more than all the oil ever dis-covered in the U.S.—plus 40 percent.Non-OPEC countries, of course, are notabove fudging their numbers either: 59nations stated in 1997 that their reserveswere unchanged from 1996. Because re-serves naturally drop as old fields aredrained and jump when new fields arediscovered, perfectly stable numbers yearafter year are implausible.
nother source of systematic errorin the commonly accepted statisticsis that the definition of reserves varieswidely from region to region. In the U.S.,the Securities and Exchange Commissionallows companies to call reserves“proved” only if the oil lies near a pro-ducing well and there is “reasonable cer-tainty” that it can be recovered profitablyat current oil prices, using existing tech-nology. So a proved reserve estimate inthe U.S. is roughly equal to a P90 esti-mate.Regulators in most other countries donot enforce particular oil-reserve defini-tions. For many years, the former Sovietcountries have routinely released wildlyoptimistic figures—essentially P10 re-serves. Yet analysts have often misinter-preted these as estimates of “proved” re-serves.
reckoned reserves inthe former Soviet Union amounted to 190Gbo in 1996, whereas the
Oil and Gas Journal
put the number at 57 Gbo. Thislarge discrepancy shows just how elasticthese numbers can be.Using only P90 estimates is not theanswer, because adding what is 90 per-cent likely for each field, as is done in theU.S., does not in fact yield what is 90 per-cent likely for a country or the entireplanet. On the contrary, summing manyP90 reserve estimates always understatesthe amount of proved oil in a region. Theonly correct way to total up reserve num-bers is to add the mean, or average, esti-mates of oil in each field. In practice, themedian estimate, often called “proved andprobable,” or P50 reserves, is morewidely used and is good enough. The P50value is the number of barrels of oil thatare as likely as not to come out of a wellduring its lifetime, assuming prices re-main within a limited range. Errors in P50estimates tend to cancel one another out.We were able to work around many of the problems plaguing estimates of con-ventional reserves by using a large bodyof statistics maintained byPetroconsultants in Geneva. This infor-mation, assembled over 40 years frommyriad sources, covers some 18,000 oilfields worldwide. It, too, contains somedubious reports, but we did our best tocorrect these sporadic errors.According to our calculations, theworld had at the end of 1996 approxi-mately 850 Gbo of conventional oil in P50reserves—substantially less than the1,019 Gbo reported in the
Oil and Gas Journal
and the 1,160 Gbo estimated by
. The difference is actuallygreater than it appears because our valuerepresents the amount most likely to comeout of known oil fields, whereas the largernumber is supposedly a cautious estimateof proved reserves.For the purposes of calculating whenoil production will crest, even more criti-cal than the size of the world’s reserves isthe size of ultimate recovery—all thecheap oil there is to be had. In order toestimate that, we need to know whether,and how fast, reserves are moving up ordown. It is here that the official statisticsbecome dangerously misleading.
ccording to most accounts, worldoil reserves have marched steadilyupward over the past 20 years. Extend-ing that apparent trend into the future, onecould easily conclude, as the U.S. EnergyInformation Administration has, that oilproduction will continue to rise unhin-dered for decades to come, increasing al-most two thirds by 2020.Such growth is an illusion. About 80percent of the oil produced today flowsfrom fields that were found before 1973,and the great majority of them are declin-ing. In the 1990s oil companies have dis-covered an average of seven Gbo a year;last year they drained more than threetimes as much. Yet official figures indi-cated that proved reserves did not fall by16 Gbo, as one would expect rather theyexpanded by 11 Gbo. One reason is thatseveral dozen governments opted not toreport declines in their reserves, perhapsto enhance their political cachet and theirability to obtain loans. A more importantcause of the expansion lies in revisions:oil companies replaced earlier estimatesof the reserves left in many fields withhigher numbers. For most purposes, suchamendments are harmless, but they seri-ously distort forecasts extrapolated frompublished reports.To judge accurately how much oil ex-plorers will uncover in the future, one hasto backdate every revision to the year inwhich the field was first discovered—not
FLOW OF OIL starts to fall fromany large region when about half the crude is gone. Adding theoutput of fields of various sizes andages (
green curves at right
) usuallyyields a bell-shaped productioncurve for the region as a whole. M.King Hubbert (
), a geologistwith Shell Oil, exploited this factin 1956 to predict correctly that oilfrom the lower 48 American stateswould peak around 1969.