As long ago as October 1956, in a survey published by the American Petroleum Institute, the late Dr Marion King Hubbert forecast that crude oil production in the ‘lower forty-eight’ states of the United States would pass its peak by the year 1970, if not shortly before. Fourteen years later it did, almost precisely on time. Hubbert had suggested that a smoothed curve of production, ironing out year-to-year fluctuations, would peak before the end of the 1960s. It actually did so in 1968. He reckoned that the ‘single-year spike’ could occur any time between 1965 and 1975. The Texan geophysicist, whose forecast was to bring him fame (albeit for a time much derision), did not estimate volume in the ‘spike’ year of 1970 quite as precisely. It was 3.5 billion barrels (9.6 MBD) in 1970, against the 3 billion (8.8 MBD) that his growth curves, sketched fourteen years before had implied. But the ‘smoothed’ five-year average of production fitted Hubbert's curve rather more closely; and his timing of the peak turned out to be impeccable (Figure 10.1).

American crude oil production in that year has never been matched since, even though the US total has at times since included more than 2 MBD from Alaska. His estimates derived from the historical performance of the US oil industry in its first ninety years, the rates at which it had discovered oil and then brought it into production. For Alaska, at that time, there was no significant history to interpret, though small-scale production was developed there in the early sixties in the southern Kenai Peninsula and the Gulf of Alaska.