Peak Oil Demand Is A Risk – Why Nobody Knows If Or When

The lead story in a recent The Wall Street Journal Report on Innovations In Energy was titled “Get Ready for Peak Oil Demand.” We have been getting ready for the past couple of years. However, in dealing with such a critical issue overhanging the future of this important global industry, the unanswered question is when will oil demand peak and begin to decline. To answer the question one needs to know when those forces reducing oil consumption will become meaningful. Then, one has to know how much the demand will be reduced.

The assumptions underlying some of the most pessimistic forecasts for oil consumption treat demand-altering forces as switch-like – once thrown, oil use goes from light to dark, hot to cold, or rising to collapsing. It has a nearly immediate impact. However, it may be more appropriate to view the transition from expanding energy use to shrinking consumption as the equivalent to turning an ocean liner: it takes a long time and lots of distance, even when the ship is equipped with bow and stern thrusters. But as the WSJ article highlighted, there are a number of considerations that require unique judgement calls, and then all these individual calls must be rationalized before arriving at the final forecast.

The WSJ wrote: “New technologies that improve fuel efficiency are starting to push down the amount of gasoline and diesel that’s needed for transportation, and a consensus is growing that fuel demand for passenger cars could fall as carbon rules go into effect, electric vehicles grain traction and the internal combustion engine gets re-engineered to be dramatically more efficient. Western countries’ growth used to move in lockstep with their energy consumption, but that phenomenon is starting to decouple in advanced economies.” Based on their view, as these forces play out, energy consumption in meeting the global economy’s needs will stop rising and, over time, will most likely decline. Several questions come to mind: Is it possible that technology improvements will make fossil fuels significantly cleaner and more efficient, while keeping them cheaper than renewable power; If autonomous technology and artificial intelligence overcomes the historical decline in energy use by aging populations, might that offset some of the other energy-reducing trends; and while everyone assumes the adoption of these new technologies will match that of smart phones (five years to reach 60% penetration), but what if it is more like the pace of adoption of electricity (50 years)?

The role of long-term energy market forecasts was touched on in Steve Coll’s book, Private Empire – ExxonMobil and American Power. ExxonMobil’s (XOM-NYSE) forecasting changed in 2000 under CEO Lee Raymond. The company had been preparing 20-year energy forecasts starting in the 1940s. In 2000, Mr. Raymond, who was overseeing the first forecasts from the combined Exxon and Mobil staffs, asked them, “What did you say about 2000 in 1980?” The question was taken as criticism, but what Mr. Raymond was really asking was to assess what the forecasts had gotten right and what they had missed.

In this case, Exxon forecasters in 1980 had been half right and half wrong. They forecasted total global energy consumption in 2000 within 1% of actual use. Where Exxon was way off was in its crude oil price forecast. It had underestimated the pace of technological improvements that made finding and developing new oil supplies that pulled down oil prices, while also failing to anticipate geopolitical events that had a significant impact on oil prices. The conclusion reached by Mr. Raymond from this lookback was to stop forecasting oil prices. Had he spent time looking at numerous studies and reviews of past forecasts such as those conducted by the EIA, he would have seen how wildly wrong oil prices forecasts have been in the past, almost all being based on extrapolations of recent price trends. This is an excellent reason to not forecast oil prices, or at least make the forecasts as a broad range.

Exhibit 1. How Oil Price Forecasts Have Been So Wrong

Source: EIA

When it comes to forecasting oil demand, probably the most critical ingredient is the assumption about the pace of potential change in the transportation sector, principally the personal automobile. Every forecast must address when, and by what share of the automobile market, electric vehicles (EVs) will capture. Among oil companies, BP plc (BP-NYSE) expects that EVs will represent 100 million units by 2035, up from one million on the roads today. However, they believe the growth in EVs will only trim global oil consumption by 1.0-1.5 million barrels a day. On the other hand, oil consultant Wood Mackenzie expects EVs will shave 2%, or two million barrels a day off oil’s use. Norway’s Statoil is more pessimistic about oil demand due to the greater penetration of EVs into the global transportation fleet, likely driven by their experience of the rapid rate by which EVs have grown in the Scandinavian country. Of course, Norway has provided hefty EV subsidies that have helped stimulate sales, plus the country enjoys cheap power from its huge hydropower supplies.

The diversity in forecasts about how quickly EVs and self-driving vehicles will impact fuel consumption is significant. Not only do the energy companies hold widely different views, but so do think-tanks and technology companies who are very active in developing autonomous vehicle technology. Potentially the most radical EV forecast we have seen is that of RethinkX, an independent think tank that focuses on technology-driven disruption and its implications across society. Its report says that the transition will be driven entirely by economics and will overcome the current desire for individual car ownership. This shift will begin in the largest metropolises on the West Coast and in the urban corridor extending from Boston to New York City and the immediately surrounding areas. The spread of this technology will then follow into the suburbs and eventually into the rest of the country.

Exhibit 2. Attributes Favoring On-demand Ride Services

Source: RethinkX

According to the report, the disruption caused by EVs and autonomous vehicles will have enormous implications across the transportation and oil industries, decimating entire portions of their value chains, causing oil demand and prices to plummet, and destroying trillions of dollars in investor value, not to mention the value of used cars. One can also point to the impact collapsing used car values will have on auto financing and the financial health of individual car owners, especially low income families.

How quickly will this transition occur? Tony Seba, an economist at Stanford University and a co-author of the report, says, “By 2030, within 10 years of regulatory approval of autonomous vehicles, 95 percent of US passenger miles traveled will be served by on-demand autonomous electric vehicles owned by fleets, not individuals, in a new business model we call ‘transport-as-a-service.’” He says that by 2030, 40% of the fleet will still be owned by people, but they will only account for 5% of passenger miles, as they opt for the on-demand autonomous EVs (AEVs).

Mr. Seba suggests autonomous cars will be used 10 times more than internal combustion vehicles were, and that they will last longer, possibly as much as one million miles. Importantly, this transition will provide U.S. consumers with upwards of $1 trillion of benefits by 2030. His forecast, he admits, is difficult to accept, but it is consistent with other major transitions that produced a 10x opportunity cost advantage. He says it happened with the printing press and with the first Model T car produced by Henry Ford. The Model T cost the same as a carriage and two horses, but offered 10x the horsepower. According to Mr. Seba, “Every time we have had a ten x change in technology, we had a disruption. This is going to be no different.” The reason it will be no different this time is that everything becomes cheaper.

Mr. Seba’s view is that in a few years the upfront costs of AEVs will be equal to those of internal combustion engine (ICE) cars, plus they will be owned by fleets and will “last a lifetime.” Maintenance costs will be significantly lower because these EVs will have only 20 moving parts in their powertrains compared to 2,000 for ICE vehicles. These factors will contribute to a much greater use of AEVs in on-demand services compared to their use of ICE cars. For on-demand service fleets, the longer vehicle life for AEVs means much lower depreciation expense, lower maintenance costs and significantly reduced insurance costs. This is why the report concludes that the cost for driving a paid-off car will cost 34-cents per mile compared to using an on-demand AEV service at an estimated 16-cents per mile. For Mr. Seba, what this means is that “cars powered by fossil fuels will no longer be made after 2024, as self-driving electric vehicles become vastly cheaper to use.”

Mr. Seba’s conclusion is shown by the report’s forecast for how U.S. auto sales will collapse starting in 2020, reaching zero for ICE vehicles in 2024, but only five million AEVs will be needed that year. Think of the impact this scenario has for automobile companies seeking to plan their future vehicle models and the necessary assembly plants. To some degree, this issue is being debated as it relates to the CEO change at Ford Motor Company (F-NYSE) last week.

Exhibit 3. A Questionable Forecast Of U.S. Vehicle Sales Source: RethinkX

One auto industry consultant, commenting on the chaos in the auto business from autonomous vehicles, EVs and on-demand car sharing services, pointed to the difference in business strategies between Ford and General Motors (GM-NYSE). Ford is talking about building a new assembly plant in Mexico, as it replaces its CEO, while GM is reducing its U.S. assembly capacity while also selling its car businesses in Europe, Russia, South Africa and Singapore.

As quoted by the WSJ, “The peak-demand discussion is only at most a couple of years old,” according to Daniel Yergin, vice chairman of IHS Markit Ltd., an energy research firm and the sponsor of the CERAWeek energy conference. According to Mr. Yergin, as this debate has progressed and gained greater credence, American oil companies have pushed back on the idea. It is important to remember that it was barely a dozen years ago that Matt Simmons published his book, Twilight In The Dessert, discussing the failing oil reservoirs of Saudi Arabia that would propel the world into a peak oil supply scenario resulting in extremely high oil prices and global economic and energy deprivation. Average oil prices of $79 a barrel in 2010, however, were well below Mr. Simmons’ projection that they would average at least $200. For oil companies that pushed back against the peak oil supply scenario and won that battle, one can understand why they might push back equally hard against a peak demand scenario. Does history repeat or even rhyme?

At the heart of the peak oil demand argument is the question of how quickly new energy-reducing technologies will be adopted.

Exhibit 4. The History Of Technology Product Acceptance

Source: Mauldin Economics

When the pace of adoption of technologies is examined, there are a number of interesting questions that bear on the projections of how quickly EVs and AEVs, as well as on-demand ride services, will be accepted. Are they going to be adopted as consumer technology items or truly revolutionary technologies and labor-saving devices? As shown in Exhibit 10, proponents of rapid technology adoption point to the cellphone, which took about a decade to go from zero to 60% penetration. That was about the same time span as the internet, but maybe only slightly longer than the VCR. On the other hand, the telephone needed nearly 50 years, while electricity needed only about 25 years, to reach the 60% penetration level. However, maybe we should look at these vehicle technologies as akin to those that brought significant lifestyle changes such as the stove, the clothes washer and the dishwasher, which needed between 35 and 50 years to reach 60% of American homes.

Our best guess is that the adoption rate will be somewhere between the cellphone and electricity, 10 to 25 years, but with a bias toward the longer timeframe. Why do we say that? It is important to understand that vehicles play an important role in family evolutions, something that hasn’t changed over generations. The hyped concern about millennials not getting married, starting families and buying homes, which was very popular during the years immediately following the global financial crisis of 2008, is disappearing. We now see millennials coming out of their parents’ basements, getting married, starting families and buying homes – although maybe not of the same size or in the same locations as their parents. These millennials are, however, continuing the generational pattern of societal evolution, although they are taking longer than previous generations to take some of the steps down that road. Given the pace of this phenomenon’s development, it is important to remember that automobiles remain the second largest purchase after homes for families. These purchases are not made frequently, they usually require significant research and time to reach a decision, and the decisions are often based on economic considerations involving all aspects of families’ lives and not just social concerns, such as climate change.

Given the factors involved in new car purchases, those forecasting the demise of petroleum must explain how those with limited incomes and wealth will voluntarily give up their perfectly functioning fossil fuel vehicle for an expensive EV, which because of battery technology may not get anywhere close to the advertised performance due to the climate where they reside. Their lives will become more complex until electric charging stations are as ubiquitous as gasoline stations, since they may not be able to afford the wait for battery recharges nor the cost of an installed charger in their home, if that option even exists for them.

There is also the question of what happens to the economics of EVs versus ICE cars when the values of used ICE cars go essentially to zero? In that case, unless gasoline and diesel fuels are banned, which may be the next target of environmental activists, it will be much cheaper to own and operate ICE cars than EVs.

There is also the question of how quickly the fleet of American vehicles can be converted to EVs or AEVs. For the past several years, Americans have purchased 17 million or slightly more new vehicles each year. At that pace, it will take 15 1/3 years to completely replace the approximately 260 million vehicles currently on America’s roads. To reach the magic 60% penetration rate, Americans must buy 17 million new EVs every year for more than nine years. Despite the high number of EVs in the fleet, it still leaves 104 million ICE vehicles on the roads burning fossil fuels.

Without some critical technology breakthroughs, principally in battery and fast charging technology, that improve the time commitment related to traveling in EVs, the distance they can travel and their cost, EVs are likely to remain a niche market. EVs do become a viable solution for high urban density areas, but are probably less viable in lower population density regions. Just as ExxonMobil’s Mr. Raymond questioned what his forecasters got right and wrong in their forecasts, EVs, consumer attitudes toward car ownership and on-demand ride service developments are forces that need to be monitored to assess their possible impact on petroleum consumption. Energy company managements need to develop alternative scenarios of the pace of these trends impacting their businesses, but claiming that the Age of Petroleum is facing a near-term demise is premature.

oilpro.com