Global Energy Outlook

Date October 4, 2016
Speaker Adam SIEMINSKI (Administrator, U.S. Energy Information Administration)
Moderator ISHII Hidehiko (Deputy Director, International Affairs Division, Agency for Natural Resources and Energy, METI)



Adam SIEMINSKI's Photo


The Energy Information Administration (EIA) is an independent organization within the Department of Energy (DOE). We work to produce quality data and accurate energy outlook analyses and projections for the United States and for the world based on existing law and regulation in both the United States and other countries. This contrasts with other forecasts, which are based on carbon emissions reduction policies, etc. We look at alternatives, but we start with the existing laws and regulations.

Key findings

We see global energy use growing by about 1.4% per year. That implies about a 50% increase in energy consumption around the world. Much of the growth will occur in non-Organisation for Economic Co-operation and Development (OECD) countries such as China and India, which will account for more than half of the increase. The industrial sector accounts for the largest share of delivered energy consumption. Economic growth in developing countries tends to be driven by industrial output.

Renewable energy is the fastest growing energy source, increasing by 2.6% per year. Nuclear is also growing rapidly along with natural gas. However, the total amount delivered will still not match the output from oil, coal and natural gas.

With the existing laws, fossil fuels will continue to be the dominant source of energy even by as late as 2040. The EIA is not the only forecasting agency that comes to this conclusion; the Institute of Energy Economics, Japan (IEEJ) and others show a similar pattern.

We see coal use flattening, and falling perhaps even in China. 40% of global electricity production was coal-generated as recently as 2012. We see it falling below 30% by 2040, but it is still a significant portion, especially in countries such as India and China. Annual carbon dioxide emissions will still rise from about 32-36 billion tons in 2010 to about 39 billion tons in 2030. The United Nations (UN) predicts that in 2030, assuming full implementation of the Paris Agreement, annual carbon dioxide emissions related to energy will be from 33-39 billion metric tons. Our 39-billion-ton prediction is at the upper end of the UN's range. I think some of the difference is related to our belief that some of the promised contributions at Paris might not be implemented. As an example, one country pledged to build five nuclear power plants, but it has no experience in nuclear power.

Long-term projections are almost guaranteed to be wrong because something will be missed: the prices will be wrong or some change in technology will be overlooked. As a consequence, it may be useful to consider how our projections could change. Implementing a global carbon tax would be the biggest difference. This would cause markets to reduce coal consumption significantly, maybe by as much as one-third, and would slow growth in petroleum and perhaps natural gas and strengthen growth in renewables and nuclear power.

For the United States, the EIA studied a hypothetical scenario in which a $25/ton fee was imposed on carbon, growing by about 5% per year. We found that by 2040, coal use would be nearly zero in the United States and renewables and nuclear would grow. Interestingly, we found that nuclear would grow by more than renewables in such a carbon-fee world.

We also see interesting developments in the global oil markets. As we get past 2025, demand for the Organization of the Petroleum Exporting Countries (OPEC) oil again begins to increase. This is one of the themes in our outlook: we think that oil prices will recover over time. Although the oil shale revolution in the United States has been very successful, we have not seen much of it outside the country. Even in the United States, the oil shale resources have some limits. With regard to natural gas, the resource base domestically and globally is very strong. We may see many natural gas developments worldwide.

Much of the oil consumption is happening in non-OECD states. It is likely to decline in the United States, Japan, and Europe, but oil growth will be strong elsewhere, including the Middle East, Africa and Latin America. Much of this will be a combination of transportation and industrial use. Oil is still a very good fuel for transportation. Until we have a breakthrough in battery technology, gasoline and diesel engines are still very good with regard to range, safety, and affordability.

We see liquefied natural gas (LNG) growing very strongly over the next few years. Australia and the United States will be adding significant LNG over the next few years and in fact we see the LNG markets as oversupplied, probably until the 2020s. This could be good news for Japan in terms of the affordability of LNG for the electric power industry.

We see these markets rebalancing eventually but probably not until the middle of the next decade. We see growth in renewables and natural gas and coal contributing to growth in electricity consumption out to 2040. The population is growing, and two billion people don't have electricity. We see this as an opportunity to improve the quality of life in many countries.

China, the United States, and India are currently the three biggest coal consumers. We expect consumption to be flat or to decline in the United States. Our official forecast for China is that consumption will continue to grow only for another 2-3 years. Some believe it may already have peaked. Economic growth from non-industrial activity will push it down. Coal consumption will continue to grow in India as electricity use grows. Coal is still the largest source of carbon dioxide on a global basis. Finding the technology to capture and sequester carbon dioxide (CCS) from coal would be very valuable. CCS technology is currently being heavily researched.

Thoughts about the U.S. energy outlook

The United States is still the world's second largest consumer of energy, with China leading. We are one of the largest producers of energy. The three biggest oil producers are Russia, the United States, and Saudi Arabia, each producing approximately 10 billion barrels per day. The United States is a big coal producer, though not nearly as big as China, and is the largest producer of natural gas. It is currently the leading producer of nuclear-generated electricity. The EIA just published an article stating that by 2040 China will be the biggest producer of nuclear electricity.

Gross domestic product (GDP) will increase. We have population and GDP growth in the United States, but the amount of energy used per unit of GDP is declining. Energy intensity is declining almost in line with economic and population growth, so total energy consumption is not growing very much. Electricity demand is growing very slowly, at less than 1% per year.

Market forces are driving up oil prices in our projections. We see oil getting back to $80 by 2030. We have oil at about $51 next year. Brent prices are just a little over $50 today compared to as low as $35 since the beginning of the year. We believe that the markets are rebalancing, but there is still a lot of inventory.

We see natural gas production increasing despite relatively low and stable prices because the resource base is so strong, and technological improvements will be the biggest driver of U.S. oil and gas consumption. We believe that shale oil and shale gas technology may continue to improve and that we may get a high resource case for oil and gas. Low prices encourage companies to innovate as they struggle to stay in business.

We see a lot of natural gas being exported from the United States. We also see considerable opportunity for pipeline exports to Mexico and LNG exports to Asia, and possibly also to Europe. Interestingly, the first tanker load of LNG from the Sabine Pass facility in the state of Louisiana went to Brazil. Brazil was having a drought and its hydroelectricity production was low. The first cargoes were expected to go to Asia.

I heard an interesting story last week in Washington DC. Chris Smith from the DOE said that the first shipment of LNG landed in Guangdong, China on August 22. The head of the National Energy Administration of China said, "Wonderful!" and he was smiling and the senior officials from China were clapping. Two years ago, they never would have thought that the United States would export to China. They were delighted to have a diversification of supply coming from the United States. On that same day, I was at another meeting. The first shipment went to England where it was met by people carrying "No LNG" and "No fracking" signs. What a contrast between Europe and China in terms of the attitude towards LNG.

Price scenarios

Oil will again approach $100 per barrel by 2030. I would be more comfortable with a number from $80 or $90. We ran an experiment on $100 oil and the result was a huge increase in oil production in the United States; I think the shale oil producers remain capable of producing a lot of oil. With regard to total primary energy consumption in the United States, coal comes down pretty sharply in our reference case, including clean power plants. The preliminary arguments were made just last week in the federal court for the District of Columbia on the clean power plan. The questions asked by the judges seemed to favor the Environmental Protection Agency's clean power plan rather than the current situation.

We see oil consumption remaining relatively flat and eventually declining. Renewables are growing very sharply. Natural gas is growing very sharply. Even without the clean power plan, consumption could continue to decline. The numerous air pollution laws and rules in the United States with regard to mercury and heavy metals will reduce the amount of coal used. In the reference case, from 2015 to 2040, U.S. net imports of oil are shrinking, but we are still importing oil, largely from Canada, Mexico, and possibly other countries in the Western Hemisphere. In the high resource case, which I don't think is impossible, the United States could become a net exporter of oil. Reporters here wanted to ask me about OPEC. The EIA doesn't have an opinion. We prefer market solutions. Even if we were an exporter, I don't think the United States would join OPEC.

We see the resource base for natural gas as enormous. Most of the side studies we have done for natural gas show higher rather than lower gas production. One of the interesting aspects of natural gas is the large geologic formation in the eastern part of the United States—the states of Pennsylvania, Ohio, and West Virginia—called the Marcellus Formation. It is very deep and has many different layers. It appears that the ability to move natural gas is going to be very strong. It's so strong that if we had the high resource case on the natural gas side, U.S. LNG exports could be even larger than what we are projecting in our reference case. If I had to say which, the reference case is our best guess with existing law, but I think that technology will continues to improve, making it a little more likely than our reference case.

If oil prices remain low, exports of LNG from the United States will be constrained simply by economics, i.e., the cost of getting LNG to Asia or Europe to compete against Russian-supplied pipeline gas would limit the level of LNG exports.

In our reference case with and without the clean power plan, we still see growth in renewables. Nuclear is relatively flat. Natural gas grows faster in our reference case. Interestingly, with the clean power plan in effect, gas grows faster because coal drops very rapidly, and, although renewables are growing sharply, they can't grow fast enough to replace all of the coal retirements so there is more natural gas.

The United States will become a net exporter of natural gas next year. We will continue to be an importer of oil, but in the high resource case, it could approach zero. The United States turns out—surprisingly, to many people—to be a potential exporter of most fuels.


Q1. What are the parameters for the very low oil price scenario?

This would perhaps require a combination of the market and policy. For example, we may have stronger rules on auto fuel efficiency that would reduce gasoline and diesel fuel consumption. A carbon tax could be implemented, which would influence both coal and oil production. The level of the carbon tax is interesting. A rough rule we use at the EIA is that a $1 carbon fee is worth about one cent/gallon. If a carbon tax were applied equally across all fuels, a $25/ton tax on carbon escalating by 5% per year, an over-$50 tax by 2035 would reduce coal almost to zero in the United States but would only increase the price of gasoline by 50 cents/gallon from an average of maybe $3/gallon, not enough to really influence gasoline consumption. To influence gasoline consumption, if the carbon tax is going to be the same for all fuels, fuel efficiency standards would still be needed. That would be the policy side.

The second side would be the markets. What if we have the high technology and resource case for oil in the United States and it becomes a net exporter of oil in competition with OPEC? What if Iraq and Iran want to compete with Saudi Arabia for oil consumption revenue? The resource base is very strong in Iraq. It currently produces three billion barrels per day and could produce up to seven billion. Too much supply could be coming online, which could keep the price down. It has been very difficult for OPEC to agree to limit production, and the competition between countries such as Saudi Arabia, Iraq, and Iran for power and prestige in the Middle East and for oil production is still a very powerful incentive for higher oil production. From the International Energy Agency (IEA)'s publications, the monthly oil market report, and the work we have done at EIA, many countries now have very low oil production: Libya, Sudan, Nigeria and even countries such as Yemen and Syria. The list is very long of countries where oil production could be coming back into the market if the diplomats are successful in bringing some peace and order to those countries. So, the low oil price scenario could be driven simply by supply, production technology, and policy choices. One comment on supply: a question I get is what is the break-even price for shale in the United States. The EIA tried to look at that, but it is a very complicated calculation. Between work done in Paris and Washington and by consultants in different places, many people believe that shale oil production could begin to rise again at about $60-$70/barrel. We are already seeing even at $50 that the rig count for oil in the United States has flattened and is coming up.

Q2. I suspect that the reason for the sudden Saudi Aramco IPO is the assumption that low prices will continue. What other reason could there be?

Does Saudi Arabia share that view? I sense that after the Paris Agreement the view in Saudi Arabia was that it is very possible that they currently have peak oil demand and that producing the oil now might be a viable strategy. The reports of the last few weeks about Saudi Arabia and OPEC considering production freezes and limits are an interesting turn-around from that, but I have a feeling that oil demand in non-OECD countries suggests that demand for oil is going to continue to rise for some time. We are showing a world in which fossil fuels are still dominant in 2040. Many people believe that to be impossible because of the climate issues, which creates a natural tension in these forecasts that says something is going to have to change, but we are unsure.

Q3. It's very interesting to listen to the differences between the IEA and EIA forecasts. Did the IEA study the scenarios that you did not? In that sense, your analysis shows that under the Paris Agreement, two degrees is an illusion. If we just look at the global outlook, global emissions should be one billion tons negative. For that purpose, in 2080, global emissions should be negative. However, not necessarily only because of Paris but also other agreements of the same type, it is an illusion to have this kind of projection because I think globally, including the United States, we are doing some further measures and policies. Scenarios and projections which take this into account may be valuable. What do you think?

Right. Even the IEA says that the 450 parts per million (ppm) scenario looks impossible to achieve. The UN's own forecast is that the Paris Agreement is insufficient to hit the 2-degree target. Are you asking whether the EIA should show a scenario that achieves a 2-degree or maybe even a 1.5-degree target?

What would be the case globally for a $25 carbon tax? This is never going to be enough to achieve the 2-degree so just forget about that.

It's enough to eliminate coal in the United States but would it do so in China and India? Perhaps not. The EIA was conceived mainly as a domestic agency. We could try to do that, though we are better at internal forecasting and projection.

I initially thought that your question was leaning toward another question: why don't we show a low carbon scenario as our reference case? Maybe we are misleading people or we are showing a very bleak outlook with these fossil fuel levels. I like to think of the EIA that our base, reference case scenario is actually valuable because it says, "With what we know we are doing now, this is what we will get, and if you'd like to see a different world, the laws and regulations will have to change to reflect your goals." Some kind of consensus may be emerging on the idea of a carbon tax, but if we were to have, for example, a $25 or $50/ton global carbon tax, it would imply in some way that EIA was advocating that. We try to stay out of the business of advocacy and policy-setting. There are other scenarios. One is simply a cap-and-trade system. Another way to achieve lower carbon is to ban coal. There are many ways to achieve the desired outcome. It could be done. I think an interesting question is what is the lowest-cost way of achieving lower carbon? I am not really sure if that's an EIA or even an IEA question. That might be a question for the Environmental Protection Agency (EPA) or the Council of Economic Advisors or the UN. In my experience, the lowest-cost path is pricing.

Q4. How do you see the excess or additional capacity of non-OPEC shale oil? If it is a very high amount, for example in the United States, then it could undermine an agreement by OPEC to limit production.

The first thing to come to mind is that OPEC has been talking about freezing production. Libya then said maybe there would be a cut. Russia says it has some problems. Iran said it is not quite ready to agree. Economists would say that OPEC will have to be very careful if they do set a production limit that they don't push the price so high that they begin to get a supply response from short-cycle developments like shale oil in the United States.

Oil production in the United States dropped very rapidly. Year-on-year, it has declined by about a million barrels a day. However, the rig count is beginning to come back up. The decline in production is starting to slow. If prices move from $50 to $60/barrel, production may rise again. If OPEC tried to push it to $80 immediately, we may see a strong reaction.

Over the next few years, OPEC's control in the oil markets has some limits. Longer term, going through the resource base, if we don't see oil shale development outside the United States, then we will be back to relying on more conventional kinds of oil and the resource base for that still lies mostly in the Middle East and OPEC countries.

Another thing that we have done work on is the volatility of oil prices. To get a 95% confidence level on the oil price, say $50, the confidence level range is anywhere between $30 and $90 per barrel.

Q5. What methodology did you use to generate this outlook?

For the United States, we generally tend to use something called the National Energy Modeling System, which iterates fuels against prices but takes a lot of inputs externally so our economic input is external and so on. Much of our international forecast is subjective. We have international models, but they are not nearly as well-developed as our domestic models. Although the modeling is computer-driven and better than an Excel spreadsheet, it's not as good as the more complex models.

Q6. I have two questions concerning future oil price projections. First, how do you account for technological innovation over the longer term? Second, how do you account for future oil production potential and unconventional oil exploration?

We do our own global oil resource assessment. We generally tend to rely on outside consultants for that work, but we also look at the U.S. Geological Survey estimates. I think a number of companies have looked overall at the heavy hydrocarbon resource base, and it's probably two or three times the proven resources. The real issue is not so much the amount of hydrocarbons in the ground but the economics and extraction technology. What made the shale oil revolution in the United States so interesting was the formations. They were always known to exist but had been believed to be technically impossible to extract. Technology greatly improved the extraction economics beyond what engineers had thought was possible. I think we are learning that the resource base is there and that the constraints are either going to come from the demand side—laws that say those hydrocarbons can't be used for some reason—or the extraction technology side—the technology doesn't improve so much. If conventional resources in OPEC countries can be produced at $75/barrel, it might reduce investment in oil sands in Canada. The possibility of having oil shortages that can quickly drive prices up is still part of the existing system. One of the reasons we have such high and low forecasts out into the future is that the further out we go, the more there is of variability. We just don't know what technical, economic and policy choices are going to be made, but having ranges like that is a valuable tool to assess the world.

*This summary was compiled by RIETI Editorial staff.