Earlier this week, I attended the Energy Information Administrations (EIA) annual energy conference in Washington, DC. This years conference was a two-day event marking the 30-year anniversary of the EIAs founding as the statistical and research arm of the US Dept of Energy.

One of the more high-profile debates in the crude oil market today and at the EIA conference is the concept of peak oil. Peak oil is simply the idea that global oil production is at or near a limit and wont be able to expand meaningfully in coming years. Even worse, many argue that global oil production will actually begin to decline; that decline rate could be rapid, given the age of some of the worlds largest fields.

If peak oil theorists are correct, crude oil at $100 per barrel would seem ridiculously cheap. Rapidly rising oil demand from the developing world coupled with shrinking supplies would likely produce an oil spike of epic proportions.

One of the main reasons peak oil has been so talked about in recent years is because of an outstanding, yet controversial, book by Matt Simmons entitled Twilight in the Desert, which was released in 2005. Simmons was at the conference, on a panel set up to debate peak oil with two EIA analysts and an analyst from Cambridge Energy Research Associates (CERA).

As you might expect, the panelists didnt see eye to eye on this issue. The EIAs estimates for global crude oil production out to 2030 are summarized in the chart below.

Source: EIA International Energy Outlook 2007
The chart shows that the EIA projects that global liquids production (mainly crude and natural gas liquids) will expand from the current 83 million barrels per day to around 118 million barrels per day in 2030, an overall increase in 42 percent.

Ive also broken down the chart by source of liquids production. To summarize, the EIA is looking for modest gains in non-OPEC (Organization of the Petroleum Exporting Countries) oil production out to 2030, totaling around 6 million barrels of oil per day.
The biggest increase will come from OPEC conventional crude oil production. The EIA has penciled in nearly 21 million barrels per day from OPEC. And Saudi Arabia alone makes up a sizeable chunk of the projected OPEC increase, ranging between approximately 6 million and 6.5 million barrels per day.

The final growth component of the EIAs forecast is unconventional fuels such as biofuels, heavy oils, oil sands and coal-to-liquids. In total, these sources add another 8 million barrels per day to global crude oil production.

These liquids growth estimates rest on a handful of key assumptions. One is that recovery factors will rise. Reserve estimates you often hear quoted in the news are for estimates of original oil in place (OOIP), the total amount of oil contained in the reservoir. But oil and gas arent found in giant underground caves or lakes but trapped in the pores of rocks.

Some of this oil is stranded in sections of the field where the rock is impermeable, and therefore, the oil cant reach the wellbore. And some of the oil will simply be left behind during production; theres no way to pump it out as if it were in storage.

Typically, a producer wont recover anything close to 100 percent of the OOIP even after many decades of production. Some reservoirs with lower permeability may only yield 15 percent of the OOIP. And even the best, most permeable and most heavily-developed fields in the world rarely reach a 70 percent recovery factor.

The assumption underlying much of the growth in OPEC and non-OPEC production is that better technologies will allow recovery factors to rise globally. And producers will squeeze more oil from existing fields.

There is truth to this argument. After all, as Ive written before in TEL, the use of newer techniques such as fracturing have allowed producers to show impressive growth in oil output from reserves such as North Dakotas Bakken formation. (See TEL, March 14, 2008, Oil Bull Market Uncovers Two New Opportunities, for details.)

But Simmons and other peak-oil theorists bring up valid points about rising recovery rates. One is that we really cant accurately assess the current state of production or recovery rates from most OPEC fields.

The EIA and other firms rely on third-party data providers for their information and as a basis for projections. That data is often vague and provides little real field-by-field production and reserves detail. In most cases, its also disseminated by the countries themselves, which have an interest in inflating their production capacity.

In addition, for non-OPEC countries, decline rates from existing fields have been faster than expected. In other words, production from mature fields is falling at a faster rate than most analysts had predicted. For example, in a recent conference call, oil services giant Schlumberger indicated that global decline rates are higher than most projections assume.

This has meant that non-OPEC production estimates have been consistently disappointing. Its difficult to assess if better technology and recovery rates can do more than simply offset some of the drop in production from natural decline.

Simmons also brought up data on the Yibal field in Oman. Shell, the main operator of the field, began using the most advanced horizontal drilling and production techniques in the 1990s. The result was a quick uptick in production, followed by a peak and precipitous decline in production starting in 1997. Simmons contends that technology didnt recover more oil from Yibal but just allowed the oil to be produced more quickly for a short period of time.

And then theres deepwater. Although Im convinced this is one of the only frontiers of oil production that will see growth in coming years, Simmons is certainly correct to bring up key risks to that outlook. Chief among those is that producers typically only drill a few appraisal wells in their deepwater fields; drilling such wells is extremely expensive in the deepwater. But with just a handful of wells to test a field, its hard to accurately gauge production potential, recovery factors or the size of the reserve.

When it comes to unconventional production, theres even more dissention. One panelist indicated that conventional crude oil production would indeed peak in 2017 to 2020 but that unconventional production and liquids such as biofuels would offset that decline. This would lead to a long period of steady or near-steady liquids production--a case Simmons called an undulating plateau.

But there have been serious delays and cost overruns to many unconventional liquids projects, including gas and coal-to-liquids plants. Moreover, the worlds increasing output of biofuels is having some serious implications in the agricultural markets. Chief among those are a rapid run-up in commodity prices and increasing water shortages globally.

Consider, for example, that nearly 70 percent of the worlds freshwater supplies are used by the agricultural industry. Rising demand for crop output has, over time, resulted in a depletion of water supplies. According to UK consultancy Bidwells, the groundwater levels in the North China Plain are dropping at an alarming rate of 1 to 3 meters (3 to 9 feet) per year. The loss of water supply is already having a deleterious effect on agricultural productivity.

Toward a Consensus
But these disagreements are less important from an investment standpoint than the consensus that did emerge. Growing oil production fast enough to meet the worlds rapidly growing demand will be difficult, expensive and subject to a great deal of risk.

For example, the CERA panelist noted that every year producers need to find an additional 3 million to 3.5 million barrels per day of oil production per day just to offset declines from older fields. Global liquids supply is running on a treadmill; drilling and field development must continue just to maintain current production rates.

According to CERA, there are 546 giant oilfields in the world that currently account for just more than half of production. Only 80 of those fields are currently undeveloped or not fully developed. And the inventory of giant fields is declining because few major discoveries have been made in recent years. Therefore, producers will increasingly target smaller fields in which production is more complex and expensive.

And costs overall in the industry are rising. CERA publishes an index of total capital costs in the oil industry. That index was set to 100 in the year 2000. Currently, the CERA capital costs index is closer to 200. Actual costs have, in the very least, doubled.

There are also above-ground risks. In other words, even though oilfield geology may allow for expanded production, there are other factors influencing production. For example, many of the most promising fields are located in countries with significant political instability such as parts of Africa.

Other promising plays are located in nations where production is controlled by state-owned national oil companies (NOC). NOCs have, in recent years, sought to exert more control over their resources. In many cases, this has meant that the big western producers with the best technology simply cant get access to reserves on favorable terms.

Then theres the simple problem of infrastructure. Much existing infrastructure in global oilfields is ageing and needs to be replaced. In addition, key assets such as deepwater rigs are in short supply, resulting in delays to key projects.

All told, global oil production estimates are an exercise in probabilities, not certainties. My view remains that some forecasts of future production are way too optimistic. There are just too many best-case assumptions in the EIA projections. I seriously doubt that the world will see 120 million barrels per day of liquids production in 2030.

Nonetheless, an immediate peak and decline would appear somewhat overly pessimistic. Higher energy prices have allowed producers to target fields once thought uneconomical, such as those in deepwater.

Whats absolutely clear is that future oil production growth isnt a certainty and will be expensive to bring online. This suggests that the era of low energy prices is over: Good whether oil production has already peaked. Or can it continue to grind higher for another 15 to 20 years?