Most models used to predict peak oil do not address demand for oil, except to presume that demand will rise and fall as governed by an exogenously determined oil endowment. The underlying assumption is that the world will immediately use whatever oil can be pumped from the ground, and that supply is independent of demand—that is, oil exploration investments bear no relation to the current oil price or expectations of future demand.

Recent trends suggest that a demand-driven peak is increasingly plausible. First, demand for passenger travel may be saturating in industrialized countries. Second, recent global efficiency standards for passenger vehicles, international freight, and domestic freight present significant changes from decades of stagnation in efficiency. Third, price-competitive alternatives to conventional oil continue to expand.

...In this paper, we describe a transparent, user-operable, data-rich model to examine the possibility of whether “peak oil” may arise through falling demand for conventional oil. We consider three indicative scenarios that vary the rate of technological improvement and changes in fuel economy regulations. These three scenarios are not the only possibility: users can easily develop their own scenarios with different assumptions for economic and population growth; demand for passenger and freight travel; fuel economy; and the adoption rate of substitute liquid and non-liquid fuels.

—Brandt et al.

The basis for their analysis is their Interactive petroleum Demand EStimation (IDES) model—an interactive spreadsheet tool that can be accessed as Supporting Information (SI) for the paper. The model focuses primarily on oil demand from transport—the dominant consumer of oil and the most constrained in terms of substitution possibilities.

They projected conventional oil demand in five-year time-steps until 2100 under a variety of scenarios of population and economic growth, as well as varying rates of travel demand growth, efficiency improvement, and fuel substitution. Population and GDP projections are taken from IPCC SRES scenarios. The model performs country-level calculations that are summed in some instances to larger regional aggregations (e.g., UN regions).

In the paper, they consider three indicative scenarios: Historical; Efficiency Policy; and High Technology. The Historical scenario uses relationships derived only from historical trends. The Efficiency Policy scenario takes account of recent adopted or proposed fuel economy goals. The High Technology scenario adds more rapid penetration of alternative liquid fuels, as well as continuing the trend of aggressive fuel economy improvements seen in recent years.

Instead of modeling prices explicitly, they assumed that long-run prices for liquid fuels will remain in the ranges observed in the period of their underlying data sets (1970−2010, to above 140 $/bbl in constant 2010 dollars).

The major advantage of our approach is that projections of peak oil demand are inherently robust to a wide range of price scenarios. It is certainly possible that prices will rise beyond those driving our underlying data set. Under this condition, our model would underestimate electrification along with other shifts toward non-liquid fuels and the adoption of energy-saving technologies. Thus, demand would fall below our model’s estimates. On the other hand, if prices fall, then this suggests that there is no imminent scarcity of liquid hydrocarbons relative to demand (e.g., as in an economic contraction where demand and price decline). Thus, in scenarios where availability of conventional hydrocarbon supply is a concern, our model is more likely to overestimate demand than to underestimate it.

—Brandt et al.

Also unlike traditional economic models, they we adopted a flexible specification for demand for different end-use sectors, fitted to long-run historical relationships. In other words, they adopted different specifications for different end-use sectors, informed by both theory and historical data.

The IDES model defines conventional oil as hydrocarbons with density greater than 15° API and produced via primary and secondary recovery, and excluding natural gas liquids (NGLs). This is a common definition used by modelers and corresponds closely to what the public understands to be “oil.” Thus, in IDES, substitutes for conventional oil include NGLs; enhanced oil recovery (EOR); low-quality hydrocarbons such as bitumen and extra-heavy oil; coal-derived fuels (CTL); natural gas-derived fuels (GTL, CNG, LNG); biofuels; and electricity.

Historical scenario. (A) Demand for transport energy, including liquid fuels (dark tone) and non-liquid fuels (light tone). (B) Demand for refinery inputs. (C) Range in uncertainties surrounding the conventional oil path given uncertainties in transition speed to non-liquids (slow in wide dash, fast in short dash) and transitions to alternative liquids (fast in blue, slow in red); (D) GHG emissions from oil and oil substitutes in transport for central path, assuming medium (50% low carbon, 50% high carbon substitutes) and low (75%/25%) and high (25%/ 75%), respectively. Credit: ACS, Brandt et al. Click to enlarge.

Findings across all three scenarios were:

• Total transport energy demand increasing approximately linearly through ∼2060, but then with slowing growth slows toward the end of the century.

• Non-liquid fuels making significant inroads after ∼2030 in private passenger land transport, but of little importance in other sectors until the end of the century. This is due to the assumption that penetration will be slower for road freight, air, and water transport because of the presently low energy densities of non-liquid substitutes such as electricity.

• The rates of growth and decline in demand for transportation energy, liquid fuels, and conventional oil are much less rapid than those in most peak oil models, where a logistic functional form is assumed.

• Land passenger transport becomes less important, due to saturation in wealthy countries and projected efficiency improvements. In contrast, air travel and freight become more significant. Decades of observed data and theory of air travel demand show no near-term saturation, so air travel continues to grow with wealth. Similar relationships hold for freight, with decoupling occurring only at high levels of economic output.

• Non-transport uses of oil continue their decades-long decline in relative importance.

Under the Historical scenario, peak demand for conventional oil occurs within 25 years. Total demand for liquid fuels continues to increase, albeit much more slowly, until ∼2070. But after ∼2020, this demand is increasingly satisfied by alternative liquid fuels.

Although demand for transport services is likely to increase greatly over the coming decades, Historical trends show a growing divergence between transport activity and oil use. A simple continuation of the relationships between income and transport activity and the rates of efficiency improvement observed over the past 40 years would be sufficient to cause the transition away from conventional oil. The transition rates noted here are within historical envelopes of observed behavior. Concerted effort to reduce demand or introduce alternatives would only strengthen this conclusion. The important question surrounding peak oil therefore lies not with physical resource limitations, but with avoiding challenges that would impede a demand shift trajectory similar to historical rates of change.

...In particular, the speed of transition to alternative liquids is of key importance. The substitution effect, dependent on policy and the prices at which alternative fuels and vehicles become competitive, is perhaps the most uncertain, but excessive pessimism is unwarranted. Recent examples such as wind power installation growth (rapidly declining cost leading to a decade of growth rates of 50% per year) or shifts in natural gas production and consumption in North America show the possibilities for rapid shifts in purchases and investment. Indeed, market changes can be rapid and profound once shifts in relative cost and convenience occur. For example, improvements in battery technology, coupled with increasing divergence between oil and electricity prices, could lead to a situation where small electric “city” vehicles become more convenient, cheaper, quieter, and generally more desirable for significant portions of the world’s rapidly urbanizing population. At this point, oil substitution becomes the logical choice for an increasing number of consumers, and market changes could be rapid.

...Our model is agnostic about which combination of alternative fuels will triumph. From the perspective of oil scarcity, the eventual winner is immaterial. For climate change, however, the nature of the substitutes will have a major impact. If the transition shifts from conventional oil to coal-to-liquids, oil sands, and palm oil (with associated emissions from land-use change in the latter case), peak output of conventional oil will be a significant net negative for greenhouse gas emissions. If the transition is to natural gas and electricity generated from renewable sources, emissions from transport may well fall toward the middle of the century. This highlights the need for policies to direct the transition toward lower-emission substitutes for conventional oil, as is already happening in countries such as Argentina. Either way, policy makers should not rely on peak oil to constrain emissions, to constrain future emissions growth, or to stimulate policy actions for energy conservation.

Rather than the current focus on scarcity of conventional oil resources, we believe that more attention should be focused on understanding and anticipating the economic, environmental, and social consequences of adopting the various alternatives to conventional oil...These challenge—not petroleum scarcity—should be our key concerns surrounding the inevitable transition away from conventional oil.

—Brandt et al.

Resources

• Adam R. Brandt, Adam Millard-Ball, Matthew Ganser, and Steven M. Gorelick (2013) Peak Oil Demand: The Role of Fuel Efficiency and Alternative Fuels in a Global Oil Production Decline. Environmental Science & Technology doi: 10.1021/es401419t