FOCUS: EXPLORATION What do recent N. Sea unit cost changes mean?

M.A. Adelman
Massachusetts Institute of Technology
Cambridge, Mass.

Field Development Expenditures - Table 2 [11712 bytes]

Recent innovations have raised oil productivity and cut expenses, especially offshore. But how much have unit costs been reduced?

Estimates of "finding cost per barrel of oil equivalent" have become popular, but they make no sense. They are derived by dividing total finding plus development expenditures by total reserve additions, discoveries plus development, oil plus "oil equivalent" gas. This is wrong because no one knows how much was discovered last year or in any recent year. A field once found keeps expanding for many years.

But there is an indirect way to measure trends in finding costs, although not the level. If newly found fields are more costly to find because they are smaller, deeper, more faulted, etc., their development, too, will turn out more expensive.

Furthermore, if the oil industry sees discovery as increasingly expensive per unit found, it will turn increasingly to more intensive development, again driving up development costs. Thus, the change in development cost per unit, if known, is also an index of changes in discovery cost.

Looking now at development cost: there is no such animal as "oil equivalent." The U.S. Department of Energy estimates that a barrel of crude oil has about 5.6 times the heat value of one thousand cubic feet of gas.¹ A more popular number is six; others use seven or 10. But none of them is justified.

The prices and the in-ground values of gas and of oil are not in lock step. In the U.S. during 1984-94, the average oil price per barrel at the wellhead was nine times the price of gas per Mcf, with a standard deviation of 1.5. The ratios of in-ground values of oil reserves and gas reserves have changed, also.²

In the U.S., gas and gas liquids were 52% of "oil equivalent" reserve additions in 1984-89 but 66% in 1990-94.³ These changing proportions would change "cost per barrel of oil equivalent" even if neither oil nor gas costs changed. In the U.S., the industries are diverging; oil is shrinking, while gas is holding its own or expanding.

Most development projects are mixed oil and gas. But given three pieces of information for each project-capital expenditures, oil reserves developed, and gas reserves developed-one can, with a sufficient sample, estimate by least-squares regression the separate investment per barrel of oil and per Mcf of gas.

During the last 4 years, OGJ has published four compilations of forthcoming North Sea projects. For each group, the regression results are in Table 1 [37250 bytes].

For the 1992 group, estimated development cost was $5.46/bbl and 93¢/Mcf. But then costs dropped substantially, although they kept fluctuating. In 1992, the oil-to-gas cost ratio was actually close to 6:1. But then it went much higher, before dropping to under 5:1.

For gas-plus-oil, we can now make comparisons over time by making explicit allowance for changing gas-to-oil proportions. The 1992-96 quantities multiplied by the 1996-2003 unit costs would require 31% less investment.

Lower costs standing alone may be an ambiguous signal. An industry may lower costs by downsizing: keep the best projects and cut the rest. This has apparently happened in the U.S. in oil, not gas.⁴ But in the North Sea, producers in 1990-95 added some 50% net capacity (about 2 million b/d) plus a large amount of newly installed capacity needed to offset decline. More reserves created at lower unit costs spelled more plentiful supply. In economic jargon, the supply curve moved out to the right.

Group 3 consisted wholly of Norwegian projects, and a number of reported values had to be eliminated as impossibly low, which makes the whole group harder to use. There is no doubt of much lower costs, but much of the spectacular drop in gas investment requirements was temporary.

I suspect that many European gas deposits, mingled with oil or close by, had been left undeveloped because operators (and the Norwegian government) expected the price to rise soon. Some believed there was only a temporary surplus which rising demand would soon dry out. Or else "the fix was in," and governments and industry would cooperate to get European prices up. But around 1992, companies in Europe tired of waiting for the higher price and began developing the gas projects or developing gas horizons or reservoirs near an oil project. If this speculation is correct, as the inventory of hoarded gas projects shrank, gas development cost rose, though not to 1992 levels.

The Center for Global Energy Studies, London, in 1992 reviewed the data underlying Group 1 of Table 1 and expressed doubt that most of the projects would pay out at current prices.⁵ It was a reasonable doubt then. Few would state it now, even aside from the higher prices prevailing in 1996. Table 1 is a fragment, but helps us make sense of recent developments.

The method used here can easily be applied elsewhere. It should be possible for public or private groups to obtain the records of past projects from the companies, who would benefit from the exchange.

For example, "predict" the expenditures in recently reported fields by using the factors estimated for Table 1, Group 4: Andrew field, reserves of 120 million bbl of oil and 130 bcf of gas, capital expenditures calculated at $506 million, while actual spending, originally estimated at $675 million, came in at $435 million (OGJ, July 8, 1996, p. 23). Curlew field, reserves of 71 million bbl of oil and 244 bcf of gas, calculated $463 million, actual $450 million.

A good beginning would be with records from other offshore areas, such as the Gulf of Mexico. Given the intense current interest in improving productivity, the modest cost would be well worth paying.

References

1. U.S. Department of Energy, Monthly Energy Review, Tables A2, A4.

2. Adelman, M.A., and Watkins, G.C., "The value of U.S. oil and gas reserves," working paper MIT-CEEPR 96-004 WP, May 1996.

3. Energy Information Administration, U.S. crude oil, natural gas, and natural gas liquids reserves: 1994 annual report, Table 1, converting gas at 7.2 Mcf/bbl.

4. Adelman, M.A., "Why rising U.S. demand may not hike prices in the '90s," "OGJ, Apr. 13, 1992, p. 95; and Adelman, The genie out of the bottle: world oil since 1970, MIT Press, Cambridge, 1995, p. 22.

5. Centre for Global Energy Studies, Global Oil Report, Vol. 3, No. 1, 1992, p. 17.

The Author

M.A. Adelman is an economist and member of the Center for Energy & Environmental Policy Research at Massachusetts Institute of Technology. He is an MIT professor emeritus.