Whether you believe some of its wilder aspects or not, the spectre of peak oil theory haunts the oil market. It has been a constant shadow encouraging long-only investment strategies and has contributed in no small part to the perception that there is only one direction for oil prices, and that is up. But the theory applies equally to natural gas, and in more than one way.

Natural gas too is a finite resource and as the market expands growth in supply can only be met by new production exceeding the loss of output from depleted fields. The larger the overall market, the greater the amount of depletion to be replaced, and the more resources that have to be committed to maintain growth. In mature basins, the peak already appears to have passed. UK gas output reached 3.826 Tcf in 2000 and has been on a downward trend since, dropping below 3 Tcf in 2006.

Formerly net producing nations shifting towards import dependency is an early sign of peak gas. And in parallel with oil, the 'easy gas' has been exploited first. Investing billions in the giant Shtokman field in the Barents Sea is no easy gas play, but that is where even Russia, with the world s largest natural gas reserves by some margin, is putting its money. This means that the marginal cost of production for natural gas will rise. Easy gas, by definition, was cheaper to exploit.

But peak theory looks set to hit the gas market sooner than might be expected, in the more immediate form of peak oil. LNG producers are aggressively seeking oil parity pricing for long-term LNG contracts and in the renewal of existing contracts. The UK s experiment with gas-to-gas competition appears likely to be annulled by its shift to net importer. If gas linked to oil from continental Europe sets the marginal price in the UK, then gas-to-gas competition on the UK local market is arguably a corpse.

Equally, if US dependence on LNG imports grows as forecast, and these are also priced on an oil parity basis, then gas-to-gas competition in the US may also cease. While market power remains in producers' hands, the slow evolution of a true international market in gas looks likely to be characterized by greater linkage to oil.

But there is a counter trend that adds a huge element of uncertainty. Reality, in the form of the US unconventional gas industry, appears to have broken ranks with theory. A look at domestic US gas production over the years suggests a twin peak or at least an extended plateau. Having fallen to about 16 Tcf in 1986, US domestic production of dry gas hit 19.6 Tcf in 2001 and remains well above the 1986 trough.

The trend is even better illustrated by US proved reserves, which at the start of 2008 stood at 211 Tcf, 10 Tcf higher than 1980. Almost half of US reserves now consist of non-conventional gas plays. The US can do this first because of its natural resource, but second because of its massive deployment of capital. Of the 1,778 gas rigs drilling the world in April, a staggering 1,471 were in the US. The ratio of gas to oil rigs in the US is 3.86 to 1; in the rest of the world it is 0.33 to 1.

New unconventional gas reserves are being discovered at a rate of knots. And while the US will eventually reach a new peak in unconventional gas production, the capacity of the rest of the world to change the shape of the post-peak gas curve has barely been scratched. Paradoxically, a strong link to oil is likely to provide the financial incentive to start scratching with a bit more vigor.

Ross McCracken, Editor of Energy Economist ross_mccracken@platts.com