336 PETROLEUM GEOLOGY/Reserves concerns have been raised again as the oil industry worldwide struggles to replace its reserves and achieve targeted rates of production growth Much of the published literature on the depletion debate falls into two camps On the one hand, there are those who warn of the imminent danger of a collapse in oil supply and the economic consequences of that occurring On the other hand, there are those who argue to varying degrees that there is no limit to oil supply in the near future What is common to both camps is the extensive reference to the work of King Hubbert Hubbert gained prominence in the American geological community because he anticipated the peak and subsequent fall of US oil and gas production At the heart of his prediction was the assumption that all resources are finite, and that eventually all resources will be depleted and exhausted over a period of time that is determined by the rate of production Hubbert argued that the complete cycle of exploitation must have the following characteristics Beginning at zero, the rate of production tends initially to increase exponentially Then, as difficulties of discovery and extraction increase, the production rate slows in its growth, passes one or more maxima and, as the resource is progressively depleted, declines eventually back to zero He argued that this cycle of production would be bell shaped when plotted against time The approximate rule here is that peak production will correspond to the mid-point of depletion and that this usually occurs about 20–25 years after the discovery of the mid-point Although some production curves have a shape that approximates such a ‘normal’ distribution, there is no inherent reason why production should follow such a pattern, and some production curves show very strong asymmetry For example, when production is dominated by a small number of large fields, peak output tends to precede the depletion mid-point Where the discovery pattern is more dispersed, or where offshore fields are significant, the production peak usually comes after the mid-point Hubbert’s basic assumption was that, if known past and prospective rates of production are combined with a reasonable estimate of the amount of fuel initially present, one can calculate the probable length of time that the fuel can be exploited In other words, the area under the complete production curve would be equal to the size of the resource However, estimating the amount of oil and gas that will ultimately be discovered and produced in a given area is full of uncertainty, as we have seen The Hubbert curve might have fitted US production, but, on a global scale, the modelled curve no longer fits the historical data (Figure 9) Up until the mid-1970s, the modelled production matched the actual production, but, after the supply shocks, which led to sharp oil price rises, annual increases in demand for oil fell from around 7% to just 2% per annum Figure World production and the Hubbert peak This chart shows a theoretical Hubbert curve based on production up to the oil price shocks of the 1970s The resulting high oil prices led to a decrease in demand for oil As a result, the slope of the oil supply curve departs from the predicted Hubbert curve after 1973 Assuming oil demand growth of 2% from 2003, and an ultimate reserve base of 2.25 trillion barrels, the peak in oil production could be reached in 2018 The chart highlights the impact that changes in demand can have on predictions of when global oil production will peak mmbbl, billion barrels; mmbbpd, million barrels per day