Last week, the U.S. Energy Information Administration (EIA) released its latest annual forecast of U.S. energy, the Annual Energy Outlook 2018 (AEO2018). This year’s edition goes out to 2050 and provides a look at where the agency thinks production, consumption, imports, and exports of all major energy sources in the United States are headed.
In addition to the baseline “reference” case—which does not include the Obama Administration’s Clean Power Plan—EIA also runs a number of different scenarios that explore the impacts of things like higher and lower oil prices, faster and slower economic growth, a bigger or smaller oil and natural gas resource, and Clean Power Plan implementation in various combinations. For our purposes here, we’ll focus primarily on the reference case.
Before we get going, it’s good to remember this caveat from EIA: “Projections in the [AEO2018] are not predictions of what will happen, but rather modeled projections of what may happen given certain assumptions and methodologies.” It’s also important to understand that EIA operates under constraints that require it to model current policies. So for example (and as we’ll see below), AEO2018 assumes the subsidies for solar and wind will phase down in accordance with current law. If the law changes and the subsidies are extended, as some may hope, that would obviously change the outlook for these technologies. But at least in the baseline forecast, EIA doesn’t, and isn’t supposed to, make assumptions about future policy beyond what’s on the books already.
So what that in mind, let’s take a look at some of the high-level results.
- Energy “Self-Sufficiency” Moves Closer: The Global Energy Institute was, in the 2015 edition of our U.S. Energy Security Risk Index (see the foreword to that report here), the first to point out that sometime in the 2020s the U.S. would for produce more total energy than it consumes, something the country hasn’t managed to do since the 1950s. (There are still some forms of energy, however, that the U.S. would have to import to supplement domestic production, e.g., crude oil.) As Figure 1 shows, as recently as 2010, EIA was projecting that net imports would supply 20% to 25% of U.S. demand for the foreseeable future. Today EIA expects that by 2022, the U.S. will be a net exporter of energy. That’s quite a change. Look no further than the shale revolution for the biggest cause of this unexpected good news, though relatively flat energy demand also gets a share of the credit.
- Hydrocarbons Will Remain Dominant: Hydrocarbon sources of energy still dominate energy supplies in 2050. From a share of about 82% today, they are expected to provide 79% of U.S. needs in 2050. There are, however, some interesting shifts among different fuels. Both petroleum and coal see their shares of the energy mix decline from 39% to 34% and from 15% to 12%, respectively, while natural gas sees its share increase from 29% to 33%. Renewables capture a growing share of the market, with non-hydro renewables going from less than 4% today to nearly 9% in 2050. The share of demand met by biomass also grows while the shares met by nuclear and hydropower decline.
- Domestic Energy Production Rises: Domestic energy output is expected to increase at a rate of 0.8% annually, and by 2050 will increase a total of 31%. Every energy category except nuclear and coal sees its production increase. The fastest rate of increase is for “other” renewables (mostly wind and solar) at 3.1% annually followed by natural gas (1.4%), natural gas plant liquids such as ethane, propane, butane, etc. (1.2%), and crude oil (0.6%). Nuclear output is estimated to fall an average of 0.7% a year while the trends for coal and hydro are essentially flat.
- Crude Oil Production Sets Records: Remember when folks we saying the shale revolution was an unsustainable bump in the road on the way to lower and lower domestic oil production? Looks like they were wrong. EIA projects that U.S. crude oil production, mostly tight oil, will in 2022 breech 11 million barrels per day and fluctuate between 11 and 12 million barrels per day out to 2050. Wow. Put another way, from 2020 to 2050—a period of 3 decades—EIA expects crude oil output will be sustained at a rate 2 million barrels per day higher than the previous U.S. record of 9.6 million barrels per day set in 1970 (Figure 2). The economic and geopolitical benefits of such a trend are massive.
- Natural Gas Continues Healthy Growth Trend: EIA expects production of this fuel to jump an astonishing 58% out to 2050, from 27 to 43 trillion cubic feet, which should keep the U.S. as the world’s largest producer (Figure 2). The pull from greater demand for gas in the power and industrial sectors drive the vast majority of the increase in production. Importantly, increased demand is not expected to materially impact prices. In fact, over the next decade, EIA projects natural gas prices will go no higher than $4.20 per million Btu—significantly below a corresponding projection of $4.75 in last year’s forecast.
- Natural Gas Exports Grow: In 2017, the U.S. became a net exporter of natural gas for the first time since 1957. By 2027, net exports are expected to exceed 5 trillion cubic feet from just 0.09 trillion cubic feet in 2017. EIA notes that “U.S. natural gas trade, which historically was shipments by pipeline from Canada and to Mexico, is projected to be increasingly dominated by liquefied natural gas exports to more distant destinations.” By the mid-2020s, LNG shipments should be greater that pipeline shipments.
- Oil Prices Rise Steadily: There are a lot of factors that go into the price of a barrel of crude oil, but such strong U.S. production is sure to put downward pressure on prices going forward. Indeed, prices in the current forecast are lower than in the AEOs of just a couple of years ago. Crude oil prices rise gradually in EIA’s current projection and don’t top $100 per barrel until 2036, before climbing to about $113 per barrel in 2050. A lot can happen between now and then—some good, some not so good—and experience shows that the price for crude oil rarely travels such a smooth path. Don’t be surprised if the ride turns a bit bumpy.
- Energy Consumption: U.S. energy demand will increase about 13% between 2017 and 2050, an average of about 0.4% per year. Industrial energy use is forecast to jump 34% by 2050, commercial energy use to rise 12%, residential energy use to remain flat, and transportation energy use to decline 5%. Energy demand forecasts are much lower than in past years because of greater energy efficiency and an economic shift away from energy-intensive industries and towards service industries. The shale revolution and comparatively low energy prices, however, give U.S. industry a competitive edge that is buoying energy consumption in this sector. For example, EIA indicates about an 85% increase in petrochemical feedstocks.
- Transportation: The decline in energy use in this sector is driven largely by declining gasoline usage. It is important to note that EIA assumes implementation of the Obama Administration’s CAFE standards, which the Trump Administration is reviewing and may revise. This assumption, contributes to a very large drop in gasoline consumption (28%) between 2017 and 2050, from about 9.0 to 6.4 million barrels per day. Even if the CAFE rules are modified, however, it is more than likely that gasoline demand will continue to decline over the long run, both from greater auto efficiency and growth in electric vehicle sales, which increases from well less than 100,000 in 2017 to almost 2.3 million in 2050. Gasoline-powered cars and light trucks still dominate sales in the Reference case, however, accounting for 71% of all light-duty vehicles sold in 2050 (about 12.9 million vehicles).
- Electricity Demand: EIA projects that sales of electricity will grow 0.9% per year to 5.1 trillion kilowatt hours in 2050 from 3.8 trillion kilowatt hours in 2017. All economic sectors—residential, commercial, industrial, and transportation—see annual demand growth of 0.4% to 0.8%, but at an average of 8.5% a year, demand from electrified transportation grows the fastest. By 2050, the rise in electric and plug-in hybrid vehicle sales will drive electricity demand in the transportation sector to 3.3% of total demand versus the 0.3% it accounts for today.
- Electricity Generation Mix: The mix of technologies used to generate electricity changes considerably out to 2050 in EIA’s forecast. By 2050, the share of our electricity generated from coal is expected to decline 10 percentage points to 21%, while electricity generated from natural gas is expected to rise to 35% from 31%.The most dramatic change, however, is in the share of U.S. electricity coming from solar power. EIA forecasts that from less than 2% today, solar power (mostly photovoltaics) will account for more than 14% of our electricity by 2050.
Changes in Generating Technology Mix: The shifts in generation described above are reflected in Figure 3, which shows net generating capacity additions/retirements each year out to 2050 by technology. Bars above “0” indicate an increase in capacity and below “0” a decrease. Keep in mind that generating capacity does not reflect actual generation. EIA expects solar facilities, for example, to operate at about 21% of capacity. Nuclear, coal, and natural gas plants operate at much higher rates, which should be taken into account when comparing capacity figures (e.g., although solar capacity additions exceeded natural gas additions by a large amount, the natural gas plants will be able to produce considerably more electricity). Also, the chart presents “net” changes, not both additions and retirements. It is also important to keep in mind that the changes forecast below are very sensitive to changes in policies and fuel prices, especially the price of natural gas. With all of that said, here are some things of interest:
- Natural Gas: Many plants can use both oil and natural gas. Even though natural gas demand in the power sector has been rising and new natural gas-fired plants continue to be built, EIA shows a net decline in oil and gas capacity for a few years in the early 2020s. This is almost entirely because of the closure of many oil and gas thermal plants, which are much less efficient that natural gas plants. Once these plants are retired, natural gas capacity heads into positive territory and remains there through 2050.
- Coal: Coal capacity declines by about 66 gigawatts by 2030 partly as a result of compliance with the U.S. Environmental Protection Agency’s Mercury and Air Toxics Standards, but after 2030 there is very little change in coal capacity. The coal plants that do remain are expected to operate at a higher capacity (70% versus about 55%). As a result, coal demand and production are expected to remain relative steady through 2050.
- Wind and Solar: The dip in both wind and solar capacity additions in the early 2020s corresponds to the expiration of the Investment Tax Credit for solar and wind and the Production Tax Credit for wind. After 2022, EIA forecasts very little growth in wind power. On the other hand, while solar power capacity additions shrink, by 2030 they recover and maintain a healthy pace through the projection period as solar PV costs continue to decrease.
- Average Electricity Prices Stay Low: Average electricity prices are anticipated to grow from 10.6 to 11.0 cents per kilowatt-hour between 2017 and 2050, a remarkably low 4% increase over 33 years. This is less than half of the forecast increase in the Reference case that included the Clean Power Plan. Rates rise fastest in the transportation sector, reflecting the increase in electricity demand for battery-powered and plug-in hybrid vehicles. The Commercial and industrial sectors see essentially flat electricity prices, maintaining a huge competitive edge for U.S. industries.
- Energy-Related Carbon Dioxide Emissions: Under the current suite of policies, EIA forecasts that energy-related carbon dioxide emissions decline thru the mid-2030s and then gradually rise to, in 2050, about the level they were at in 2017—a wash. Reductions in residential and transportation emissions are expected to be largely offset by increases in industrial and commercial emissions. Power sector emissions barely budge over the period. By 2025, EIA forecasts that total energy-related emissions will be about 15.2% below their 2005 level, well off the aim set by President Obama in his Paris pledge. (With the Clean Power Plan 2025 emissions would be 16.6% lower.)
- Kaya Factors and Emissions: Energy-related carbon dioxide emissions, which is what EIA measures, are the product of trends in four areas: (1) per capita GPD; (2) population; (3) energy intensity (i.e. the amount of energy is used to create a unit of GDP); and (4) the carbon intensity of the energy supply (i.e., the average amount of carbon dioxide emitted per unit of energy consumed). These combined factors are known as the “Kaya Identity.” Figure 4 shows the cumulative impacts of each of these Kaya factors on emissions out to 2050. It is no surprise that GDP growth is the biggest contributor to emissions, and population contributes, as well. Offsetting these increases are very large improvements in energy intensity with decarbonization of energy supplies really contributing a modest amount (except in the power sector). These offsetting trends produce the flat emissions trajectory in EIA’s forecast.
- GDP Growth: We mentioned that these Reference forecasts are sensitive to a lot of different variables. GDP growth is certainly a big one that often gets overlooked. The AEO2018 Reference case has an average annual rate of GDP growth of just 2%. EIA’s High Economic Growth scenario has an annual GDP growth rate of 2.6%. Back in 2007, before the Great Recession, EIA’s AEO2007 Reference forecast had an annual GDP growth rate of 2.9%—faster than the 2.6% rate in this year’s High Economic Growth forecast. My, how times have changed.
Seemingly differences in growth rates can lead to vastly different economic, energy, and emissions outcomes. Consider that this year’s EIA’s High Economic Growth case vis-à-vis the Reference case has:
- A cumulative $81 trillion (in 2015$) more in economic activity out to 2050;
- An annual energy consumption rate of 0.7% versus a 0.4% rate leading to 2050 demand 4.1% greater when compared to the reference case;
- An annual energy production rate of 1.0% versus a 0.8% rate leading to 2050 production 6.5% greater when compared to the reference case; and
- An annual emissions growth rate of 0.4% versus 0.1% leading to 2050 emissions being 9.2% greater when compared to the reference case.
If President Trump’s economic agenda can kick the U.S. economy into a higher gear—and if his “Energy Dominance” policies also gain traction—that could, as these examples show, have big impacts on the future U.S. energy economy.