--John Kemp is a Reuters columnist. The views expressed are his own--
By John Kemp
LONDON (Reuters) - Like a hamster trapped on a wheel, the International Energy Agency (IEA)'s 2008 World Energy Outlook (WEO) paints a depressing picture of an oil industry having to run faster and faster just to keep pace with burgeoning oil demand over the next 20 years.
WEO2008 estimates the industry will need to find 64 million barrels per day (bpd) of new oil production capacity to meet the expected growth in demand by 21 million bpd by 2030 and offset 43 million bpd of expected declines from existing fields. The total cost is put at around $5 trillion at today's prices.
The gross capacity required is equivalent to more than three quarters of the world's current oil production (82 million bpd) and the capital expenditure exceeds the total annual output of Japan, Germany or China.
WEO2008 strongly implies herculean efforts will be required to bring all this new oil onstream, and the industry has no more than a moderate prospect of succeeding. Only a structural upward shift in prices can generate the incentives and resources to make this investment program possible. Oil consumers should accustom themselves to much more expensive oil throughout the forecasting horizon.
But projections over such a long period are notoriously sensitive to very small changes in the assumptions made. In particular, the WEO projections are highly dependent on assumptions made about the feedback between oil prices and long-term demand, and the cyclical relationships between prices, costs and supply.
Based on detailed data for the world's 580 largest oilfields and an extrapolation to the remaining 70,000 smaller fields, WEO2008 estimates output from existing and future fields will decline by 6.7-8.6 percent a year.
The field study reaches some interesting conclusions:
(a) Observed output declines are much slower for super-giant fields (>5 billion barrels of initial proven and probable reserves) than for giant fields (>500 million barrels) and large fields (>100 million barrels). Decline rates are just 3.4 percent per year for super-giants, rising to 6.5 percent for giants and an alarming 10.4 percent for large fields.
(b) Decline rates are much slower in the Middle East (2.7 percent) than for the world as whole (5.1 percent) and for OPEC (3.1 percent) than non-OPEC (7.1 percent).
(c) For the same field, decline rates accelerate from the post-peak period (5.1 percent) to the post-plateau period when production drops below 85 percent of peak (5.8 percent).
(d) Surprisingly, given accelerated declines over time within the same field, decline rates tend to be lower for the oldest pre-1970 fields (3.9 percent) than fields which entered production in the 1980s (7.9 percent) or the 2000s (12.6 percent).
(e) Finally, decline rates are slower for onshore fields (4.3 percent) and much faster for offshore shelf fields (6.6 percent) and super-accelerated for deepwater (13.3 percent).
MANAGEMENT AND GEOLOGY What explains these differences? Decline rates are field specific and depend on natural pressure loss as the field is produced and water penetration, as well as oil viscosity and the permeability of the reservoir.
They also depend on decisions about the production profile and development expenditure (infill drilling, well workovers, replacement of well casings, water flooding and gas injection).
All these techniques increase the proportion of oil eventually recovered from the reservoir either by tapping into isolated pockets of oil, sustaining pressure or pushing oil toward the wells.
Most of the supergiant and many giant fields are located in OPEC countries in the Middle East and were discovered between the 1940s and 1960s. They have highly favorable natural pressure and field characteristics; been produced by national oil companies at a moderate pace that has sustained pressure; and benefited from extensive development work (such as water flooding).
Most of the newer fields brought onstream since the 1980s have been much smaller, often located offshore and in deep water, making them much more expensive to develop on a per barrel basis. Private developers have pushed for accelerated production to recover huge capital outlays and maximize shareholder returns by front-loading cash flow and maximizing its net present value.
Based on an expected average decline of 6.7-8.6 percent per year, WEO2008 concludes output from existing fields will decline from 70 million bpd at present (plus 12 million bpd of natural gas liquids, unconventional oils and oil sands) to 51 million bpd in 2015 and 27 million bpd in 2030.
By 2030, the world will need 104 million bpd of production, according to the IEA. WEO2008 assumes the gap will be filled by 23 million bpd of new conventional oil production from known but not yet producing fields; 19 million bpd from fields that have yet to be found (including 11 million bpd from offshore); an extra 9.5 million bpd of natural gas liquids; 7.1 million bpd of non-conventional oils such as orimulsion; 4.7 million bpd from oil sands; and 6.4 million bpd from enhanced recovery techniques such as CO2 injection.
In contrast, Saudi Arabia currently produces just 8-9 million bpd. The scale of the new sources that need to be developed is vast. All this will cost $210 billion a year (in 2008 dollars) assuming costs level off at mid-2008 levels.
This vast expenditure will require substantially elevated prices to provide the incentives and resources for a massive investment program. But how realistic is the cost assumption?
Costs are strongly pro-cyclical. Upstream capital expenditure increased from $120 billion to $390 billion per year between 2000 and 2007. But WEO2008 estimates that two-thirds of investment increase was swallowed up by higher unit costs.
Drilling accounts for around half of upstream capital spending. As demand has surged and utilization rates for floating rigs and semisubmersibles have hit 100 percent, daily hire-rates for floaters have risen 350 percent since Jan 2004 and rates for semisubs are up 750 percent, according to ODS-Petrodata Consulting and Research (here).
Other costs (cement, steel, aluminum, seismic surveys and skilled labor) have all surged as the industry has experienced a cyclical boom.
In the United States, for example, the cost of offshore oil and gas exploration and development per foot drilled has soared almost 88 percent between 2002 and 2006 (here).
Given the long lead times involved, firms have so far been reluctant to cut exploration and development expenditure, despite the recent slump in oil prices. Rig rates have remained firm. ODS-Petrodata estimates that about 160 new offshore rigs will join the existing fleet of 730 between 2009 and 2011. Most of these are already fully committed.
But raw materials prices for steel, aluminum and cement have already fallen substantially. Labor and rig costs will be subject to downward cyclical pressure and in the longer term to structural factors as high rates and wages increase the supply of field services and workers.
It is not clear using the 2008 cyclical peak is a good way to project investment costs over the whole 2008-2030 period.
Long-term projections are terribly sensitive to these assumptions. In response to higher prices and evidence of both cyclical and structural demand destruction, IEA has already cut its estimate for 2030 production by a staggering 10 million bpd (9 percent) since WEO2007 (when it thought 113.7 million would be needed by 2030).
Both the amount of new oil production required by 2030 and the scale of the investment spending are unfeasibly large in WEO2008 (and worse in WEO2007). But that is more a confirmation that the demand-growth and cost-inflation rates witnessed in 2005-2008 were unsustainable rather than a good prediction for the future.
Market forces are already ensuring a correction is well underway, with both cyclical and structural declines in crude demand and the cost of supply.
There is no need to despair just yet.