General Interest Oil&Gas Journal April 30, 2001

IEA, OPEC oil supply forecasts challenged

AM Samsam Bakhtiari   Fariba Shahbudaghlou

In the global petroleum industry, many major organizations tackle the job of forecasting future oil supplies.

Because the present standard forecast horizons are for 2010 and 2020, forecasters are taking little risk of being proven wrong, especially in a world that ticks ever faster. By those dates, who will remember forecasts made now?

The Organization of Petroleum Exporting Countries and the Paris-based International Energy Agency-the oil-sharing mechanism for members of the Organization for Economic Cooperation and Development-are among the organizations that release such forecasts.

For oil supply forecasts, a decade sounds like eternity. Even a 2-year span can bring dramatic changes in forecasts, as proven by a comparison of two IEA World Energy Outlook (WEO) forecasts for 2020 that were made in 1998 and 2000, respectively. There is an abyss between the two, as illustrated in Table 1.

The WEO's "unidentified unconventional oil" of 19 million b/d in 1998 looked much like a "mission impossible" prospect. It was rearranged 2 years later into a combination of non-OPEC oil and "identified unconventional oil," with non-OPEC carrying the main burden, as it was increased dramatically by more than 19 million b/d-seemingly yet another "mission impossible."

Furthermore, the IEA had added another 3.2 million b/d to the total of its 2020 oil supply forecast.

Forecast comparisons

Both WEO and OPEC's World Energy Model (OWEM) forecasts chose 1997 as their base year. The two sets of initial oil supply data used are compared in Table 2 along with the actual worldwide production figures issued by BP PLC predecessor BP Amoco Corp. in the report BP Amoco Statistical Review of World Energy-June 2000.

With the exception of WEO's "unconventional oil" item and its placement of Mexico in the "others" category, rather than OECD, as OWEM did-Mexico is a member of OECD-the two sets of data seem to match nicely, and the BP Amoco production statistics confirm the figures selected for the base cases.

Table 3 shows that WEO and OWEM forecasts for 2010 agree on a non-OPEC supply of 46-47 million b/d. However, the WEO's projected OPEC output is higher by a margin of 4.5 million b/d, which, added to its "unconventional oil" item, processing gains, and non-OPEC total, yields an overall higher WEO supply projection of 8 million b/d. This results in a discrepancy of roughly 10% between the two forecasts, beginning a significant divergence.

Table 4 compares the respective forecasts for 2020. Here again, both predictions seem to agree on a non-OPEC output of 46 million b/d but to disagree on OPEC supply (by a large margin of 10 million b/d) and the "unconventional oil" item (by 4 million b/d).

Consequently, for the 2020 horizon, the WEO forecast is 16 million b/d higher than that of OWEM, roughly doubling the 8 million b/d divergence noted in the 2010 forecasts.

Forecast evaluation

In general, both WEO and OWEM offer rather optimistic forecasts for both 2010 and 2020, with OWEM apparently taking a more cautious approach to future global oil supply.

The two forecasts again are generally in agreement over non-OPEC oil, but they diverge significantly once again over projected OPEC supply and the WEO's "unconventional oil" item. The three major subdivisions of these forecasts need closer examination:

Non-OPEC oil supply.

With Alaskan crude output in irreversible decline and the North Sea about to peak, it is difficult to explain the expected non-OPEC output of 46-47 million b/d by 2010-about the same as the present production level.

Non-OPEC supply is shown as remaining almost flat during 2010-20. In this respect, both WEO and OWEM apparently are pinning their hopes on the former Soviet Union-most probably the Caspian Sea area-and "Others," possibly the ultradeep waters of the Gulf of Mexico, Brazil, and West Africa.

However, these hopes encompass a very large amount of oil supply. The future non-OPEC oil reserves would have to make up for the decline of presently producing oil fields (from a current base of 44-45 million b/d). That may be difficult, with no new Alaska or North Sea province in hand. Interestingly, however, both the IEA and OPEC seem to be in accord over these rosy non-OPEC prospects.

OPEC oil supply.

Even more optimism is reflected in both forecasts for future OPEC oil supply. If the 2010 figures are still within the realm of the possible, those for 2020 defy the laws of reason, especially the WEO 61 million b/d projection.

It should be noted that, at present, an OPEC producing nearly 30 million b/d already seems near the end of its tether, and some major OPEC members have difficulty even meeting their output quotas today.

Moreover, in order to double its output during 2000-20 (as forecast by the WEO), OPEC would have to find an entirely new set of supergiant oil fields. Saudi Arabia, for example, would need another Ghawar field and Safaniyyah field and Iran would need new fields comparable to its Ahwaz, Marun, Gachsaran, and Aqajari fields in order to make the doubling possible at all.

As no new supergiant oil field has been found worldwide since 1986-and many leading upstream oil analysts seriously doubt undiscovered supergiants still exist-an OPEC producing 61 million b/d by 2020 or at any time in the future must be classified as within the "wishful thinking" category.

Unconventional oil.

For this special WEO item, a base of 1.3 million b/d in 1997 already seems excessive. Presently there are, at most, some 200,000 b/d of heavy and extra-heavy crude available from Venezuela and roughly 300,000 b/d from Canada's oil sands-adding up to only 500,000 b/d at best.

As for the 2010 forecast of 2.7 million b/d, it would require a quintupling of the present output, or an addition of roughly 2.2 million b/d within a decade. That's the equivalent of a large-size grassroots project in the Orinoco belt or in Canada's Alberta province every single year, with each project having a design capacity of 220,000 b/d and an attached price tag of about $4 billion.

During 2010-20, roughly the same scenario would have to be repeated, as some 1.5 million b/d of fresh capacity would be required under the forecast, and each yearly project would have a capacity of 150,000 b/d, for a related investment of roughly $3 billion.

Obviously the IEA's WEO and OPEC's OWEM forecasts for 2010 and 2020 are too optimistic, given the present status of global oil reserves and actual production capacities. And, although OWEM's results look somewhat more realistic than the WEO's, neither of the forecasts should be taken as a potential scenario for worldwide oil supplies over the next 2 decades.

Future higher prices could trigger massive investments in promising oil-rich provinces such as the Persian Gulf region, the Caspian Sea area, and even in new frontier provinces such as the ultradeep waters of the Atlantic Basin.

Such an eventuality would require close cooperation between potential reserve-holders and the major oil companies. In addition, a thorough revision of existing "future oil supply models" such as WEO and OWEM is advisable.

In the fresh approach, the basic econometric model could be linked to a crude oil dynamic model, based on the models developed by Prof. Jay W. Forrester of Massachusetts Institute of Technology, to include the above eventuality-and the whole possibly linked with a technological progress model.

In any event, a thorough revision of both WEO and OWEM forecasts is suggested in view of the ever more intricate nature of the future global crude oil supply picture.

The authors

Ali Morteza Samsam Bakhtiari is a senior expert in the corporate planning division of the National Iranian Oil Co. (NIOC), Tehran. He specializes in questions related to the global oil, gas, and petrochemical industries, with special emphasis on the Persian Gulf and OPEC. Formerly, he lectured on design and economics at the chemical engineering department of Tehran University's Technical Faculty. He holds a PhD in chemical engineering from the Swiss Federal Institute of Technology at Zurich.

Fariba Shahbudaghlou is a computer analyst with the corporate planning division of NIOC. She specializes in database compilation and computer graphics. She holds a BSc in physics from the Sciences University, Tehran.