The Generation Professionals Group is for utility professionals who work in biomass, coal, gas/oil, hydro, natural gas, or nuclear power generation fields. 

14,441 Members


You need to be a member of Energy Central to access some features and content. Please or register to continue.


30-Year Global Energy Outlooks – Useful or Not So Much?

Source: CWB

30-Year Global Energy Outlooks – Useful or Not So Much?

It’s not like the outlooks matter—except for a few trillion dollars of investment, or the fate of the world.

This article presents a 30-year global energy outlook and compares it with the projections of other long-term global energy outlooks.

Who believes in energy outlooks, especially 30 years out? How can anyone see that far into the future? And most importantly, who cares? Actually, many stakeholders care:

  • Oil companies
  • Electric utilities
  • Corporations with a green conscience
  • Investors/funds
  • Countries/local governments
  • Humanity

Many of the above stakeholders compile long-term global energy outlooks. Why do these organizations get their projections so wrong and what can they do to improve their accuracy?

The easy answer to inaccurate outlooks is that the organizations behind the outlooks – US DOE (Energy Information Administration—EIA), the International Energy Agency (IEA), British Petroleum (BP), Shell, the International Renewable Energy Association (IRENA), and many others – bring inherent bias to their analyses, even this one. For EIA and IEA, both government organizations, the bias is political. For BP, Shell, ExxonMobil, and other traditional oil/energy companies, the historical bias is to fossil fuels. However, the transformation of these companies to alternative energy businesses may affect this historical bias. Organizations such as IRENA, and others heavily invested in promoting renewables, bring a green bias to their outlooks.

They can improve their accuracy by broadening their worldview, invite feedback from disparate perspectives, and take that feedback seriously. Will they? This remains to be seen.

Inherent bias could explain it, and does to a certain extent. For example, why else would EIA make such a non-asymptotic 30-year projection for US coal consumption? See the Coal section for a detailed analysis.

Is bias solely to blame?

Besides bias, the other part of the answer is: things change—fast. Things didn’t change that fast fifty years ago, or twenty, or even ten. Now they do. Natural gas was dying, then fracking and horizontal drilling happened, and now natural gas promises a stable price future – but for how long? An article titled “Interrogating Uncertainty in Energy Forecasts: The Case of the Shale Gas Boom”, Reed, etal., 2019, presents a range of issues that energy forecasts face, including Black Swan events, i.e., major disruptive events. The shale oil and shale gas revolution qualifies as a Black Swan event. In the matter of a few years, the economics of energy has changed radically and caused a cascading effect on all other energy sources:

  • Five years ago, even with the rumblings of a few coal plant retirements and problems with the nuclear industry, many organizations still projected the use of coal and nuclear out for fifty years, or even a hundred. Then, the economics of natural gas plants demonstrated, in solid market terms, that coal and nuclear were no longer viable.
  • Renewables’ slow cost decline over the years didn’t catch anyone’s attention, until, well suddenly, renewables went head-to-head with natural gas and sometimes won.


The economics of energy and power seem to change by the minute, which is a disruptive departure from ideologies of thirty or fifty years ago when electric utilities thought of plant lifetimes of forty years. The future projections about energy and power are all about quickly changing economics – $/MWh.

EIA has done very impressive work with establishing a well vetted Levelized Cost of Energy, or LCOE, methodology. As a commodity-centric metric, it accounts not only for power plant CAPEX costs, but also the OPEX costs over a capacity resource lifetime. With LCOE, it’s possible to compare the economics of a new coal plant versus that of a new combined cycle natural gas plant. However, you don’t only need LCOE for that. From a first principles perspective, the efficiency of a coal plant can be shown to be underperforming compared with a natural gas plant, as measured by the foundational metric of heat rate. Revisiting Thermodynamics, this is the amount of thermal energy required to produce a MWh of electrical energy, which for coal is about 25 percent higher than for natural gas. If you layer on carbon impacts and capture principles, the efficiency for coal generation gets even less appealing. Moreover, the present-day economics of extraordinarily low cost natural gas…well, you get the picture.


Look at coal. EIA has recorded the consumption of coal from 1950 to 2018, provided data for the first half of 2019, and has estimated 13GW of coal retirement by the end of 2020. EIA has also translated these data to heat content of electricity produced in quads (quadrillion British Thermal Units – BTUs), and the shape of the chart is nearly a direct overlay. This is because almost all coal consumed in the US is used to produce electricity. The downswing of this production over the last decade is dramatic.

Turning to EIA’s 30-year US coal projection from their 2019 Annual Energy Outlook, the chart in the EIA report shows a gradual decrease in coal percent of total energy usage from 28% in 2018 to 17% in 2050. However, when translated to units of million short tons consumed, the chart morphs, which presents a challenging-to-believe story:

Why would coal consumption suddenly exhibit a sharp right turn in 2020 when the fundamental economics of coal power production are driving an ever-increasing number of coal plants into early retirement? For example, PacificCorp has announced it will close 16 of its 24 coal plants by 2030 and another 4 by 2038, for a total retirement of 4.5GW. A more plausible 30-year projection for US coal consumption, for which this outlook uses, might be viewed as:

With this 30-year projection of US coal consumption, the percentage of coal use translates to zero in 2050. If an investor wanted to place a large bet on whether coal would or would not be the fuel of the future, which of the two projections might be the one to pick?

Of course, US coal consumption isn’t the whole story, or even the big story. The big story is all about India, China, and countries with emerging economies. India and China once built coal-powered electricity generation plants at a frantic rate. While they still build coal-powered plants, the rate of construction has decreased dramatically. This is where economics and investment bring context to the big story with something called Final Investment Decisions—FIDs. According to IEA, FIDs for new construction of coal plants globally have fallen from 88GW in 2015 to 22GW in 2018. The big twist to the big story is that given the retirement rate of existing coal plants, we may well see the peak in global coal plant capacity as early as 2021.


Relative to LCOE discussion, the LCOE value for new nuclear capacity is much higher than renewables, natural gas, and even coal—not a good thing for nuclear. The global outlook for nuclear power is dim. The last conventional nuclear power plant constructed in the US might be Vogtle 3 & 4, if even that happens. The new construction economics for nuclear power are mired with investment woes—escalating and unchecked costs, under-controlled schedules, and vexing questions on component and materials quality. You can decommission a nuclear plant using funds from the Decommissioning Trust Fund (DTF), but what of the costs and destination for the waste? If, as is present convention, it has to be stored on site, those costs seemingly go on forever. Who pays?

As if that weren’t enough—re-enter the newly resurrected and highly optioned low-cost commodity winner: natural gas. The new low cost of natural gas has all but destroyed the market for nuclear power investments. When baseload capacity from nuclear plant can’t clear a present day capacity market auction, it’s in deep trouble. Recently, this has begun happening with disturbing frequency. The retirement of nuclear plants isn’t quite as steep as coal plants, but it is still alarming to those who view nuclear power as low-carbon baseload generation. EIA projects the percentage of nuclear in the US electricity mix at 12% by 2050 (down from 19% in 2018). However, that may be on the high side considering retirement rates. This outlook uses a more plausible value of 5%. Globally, the nuclear outlook is only marginally better than the US.


The world is currently awash in oil—primarily due to tight oil production from the US’s Permian Basin. The global supply, which many projected several years ago to not meet future demand, now easily meets existing demand and looks to do so for the foreseeable future, notwithstanding low supplies from Libya, Iran, and other politically unstable areas. This has kept the price of oil well below its 2008 high, though price volatility is also a constant. Who knows when the next major disruption to a key producer such as Saudi Arabia might occur? Transportation is the primary demand for oil, though petrochemical production is also a factor.

If supply is sufficient, and price is reasonable, why wouldn’t 30-year projections say that oil demand would increase?

For transportation, a “reasonable today” price for oil won’t work when the competition is the electrification of transportation. The total cost of ownership of electrified transportation is already cheaper than that for traditional oil-powered vehicles, due to lower fuel and maintenance costs, despite the higher initial capital cost. Those purchase prices, which have predominantly been driven by the price of batteries, has dropped significantly over the years and is projected to drop to the point where it is equivalent to the cost of oil-powered vehicles. Transportation electrification will be paired with investments to modernize and vitalize an aging global electric system. This transformation is well underway and stands to benefit from distributed grid assets in all modes of electrified transportation.

The most significant outward sign of this transportation electrification revolution is that all major auto manufacturers (OEMs) now offer hybrid, plug-in hybrid, and pure electric vehicles for the light duty markets, which is the most cost-sensitive category of vehicles. Many OEMs indicate they will offer electrified versions of all models. Some are publicly declaring no further ICE development. The uptake in electrified transportation will take many years to penetrate new vehicle sales and many more years to displace existing oil-powered vehicles.

Some European nations and several US states propose the ban of oil-powered light duty vehicles by 2040 or even 2035. That, along with many other far-reaching regulatory requirements from all over the world, could translate to a major drop in oil demand by 2050. While electrification of light duty vehicles is low hanging fruit, what about medium- and heavy-duty vehicles? China, the US, and other world markets have begun a mass migration to electrified public transit and school buses. Two major auto manufacturers, Ford and GM, have announced programs to electrify pickup trucks to catch up with Tesla. Tesla has also led the pack in developing electrified heavy-duty trucks. Even Amazon has entered the electrified delivery truck business with its $440M investment in Rivian.

Natural gas

Tight oil has not only enabled the US to become the largest global oil producer and put supply concerns to rest worldwide, but tight gas has also revolutionized the energy market in the US. Tight gas has spurred natural gas to surpass coal as the number one producer of electricity in the US and continues its upward spiral. Globally, natural gas is also a big player, and set to become the largest producer of electricity in many markets.

We can blame the demise of coal and nuclear on the resurgence of natural gas. But why won’t natural gas forever remain the fuel that powers the planet?

We can blame economics. The LCOE range for natural gas is currently on the high side of solar and land-based wind and projected to become even more unfavorable.

Natural gas plants in the US currently bid in for unit commitment prices higher than solar, solar plus storage, wind, and wind with storage in many markets, even with very low natural gas prices. With renewable purchase power agreement and market bid prices declining, economies of scale are demonstrating compelling financial business cases for renewables versus natural gas. DOE’s Orange Button pro-forma methodology shows renewable project financing is becoming ever more defensible.

Over the last few years, new generation additions to the US grid and many other parts of the world have solely comprised natural gas, wind, and solar. EIA projects the percentage of US electricity production from natural gas will climb from 34% in 2018 to 38% in 2050. As with coal and nuclear, this projection appears optimistic. Based on renewables technology improvements and cost reductions, a more realistic projection might be on the order of a 20% share of the US electricity market for natural gas by 2050. A recent report by Rocky Mountain Institute analyzes the competitiveness of natural gas plants versus renewables, which currently bid in many markets at lower $/MWh rates than natural gas plants. The report concludes that as renewables increase their competitive position, investors will worry that natural gas plants will become stranded assets, beginning as early as 2025, and make the same non-FIDs move as with coal. For example, if an investor group funded construction of a natural gas plant in 2025, but operations became non-viable by 2035, that plant would become a stranded asset with over half its useful life lying fallow.

This downward economic trend is even more persuasive in other parts of the world where natural gas isn’t as plentiful or cheap as it is in the US.


EIA projects the percentage of US electricity production from renewables will climb from 18% in 2018 to 31% by 2050. However, with the probable more drastic declines of coal and nuclear than they predict in their 30-year outlook, and the likely decline of natural gas plant contribution, renewables will need to fill the gap.

Is this likely? Economics are on the side of renewables. The perennial knock against renewables is non-dispatchability and intermittency. However, with market bids approaching $20 USD/MWh, renewables are currently making a very strong showing, let alone five, ten, or twenty years from now. The addition of cost-effective energy storage greatly weakens the naysayers’ “when the wind doesn’t blow” or the “sun doesn’t shine” objections.

The argument that renewables have low capacity factors is also fading. The Hywind 30-MW floating wind farm (Scotland) has demonstrated capacity factors exceeding 60%, which is higher than many current coal and natural gas plants. Even without energy storage, this technology can provide a measure of baseload, dispatchable power.

The US is not the leader in renewable investment. China and other countries invest much more heavily in renewables as a percent of GDP. As renewables costs continue to fall and modernization continues, it follows that global investment and capacity ratio for renewables and storage will increase.

An increasingly common stewardship target for many of the world’s leading commercial and industrial organizations is to shift to and sustain utilization of 100 percent renewable energy for their operations. On a localized basis, accomplishing this by the 2035 to 2040 time-frame is easily facilitated by contracting for renewable power purchase agreements (PPAs). Presently, 80 percent of US renewable PPAs are virtual (i.e., VPPAs), which means the organization doesn’t actually own the electrons. The following are examples of recent corporate renewable PPA announcements:

  • Google’s signing of 1.6GW additional renewable power output to their portfolio is an example of this corporate movement. The addition brings Google’s worldwide renewables portfolio to 5.5GW
  • Microsoft signed up for an additional 230MW of wind and solar, bringing their portfolio to 1.9GW
  • AT&T recent signing of 960MW of wind and solar VPPAs brings their total to 1.5GW.
  • Amazon pledged 100 percent use of renewable energy by 2030 and has a current renewables portfolio of 1.3GW
  • Honda has committed to renewable VPPAs on the order of 320MW, which will cover approximately 60 percent of their North America energy use

Renewable PPAs, whether physical or virtual, represent a nascent market that encourages contracts for renewable capacity and shows excellent growth potential. Beyond Commercial and Industrial renewable PPAs, Duke Energy, the US’s largest electric utility, has committed to net zero carbon use by 2050. These commitments reveal many opportunities for innovation and profitability.


An organization can commit to use of 100 percent renewables by 2050 or sooner, and it’s possible to replace fossil fuel electricity generation in many markets by 2050. Primary energy, especially on the global scale, is a different thing. Primary energy, in addition to power generation, also includes burning oil to fuel our transportation needs, which could take much longer to replace with renewables via electrification. Getting to 100 percent renewable energy production and use for primary energy will not happen for most countries by 2040, or even 2050. The integrated picture of global energy is complex with many quickly evolving facets.

The Integrated Picture

The 30-Year Global Energy Picture – The CWB Outlook

The figure below presents an integrated global energy forecast over the next 30 years, according to the CWB outlook. The y-axis is billions of tonnes oil equivalent (toe). The CWB outlook uses the global primary energy from IEA’s and BP’s global outlook use to 2040 (17.5 Btoe) and extends this to 19 Btoe based on population growth and the 25-30 percent global per capita growth of energy use from 2020 to 2050. The CWB outlook to 2050 shows an approximate 70 percent penetration of renewables.

A brief list of 20- and 30-year outlooks, with the renewable percentage noted, are:

2040 (20-year outlooks)

  • BP (2019, Rapid Transition) – 36%
  • IEA (2018, NPS scenario) – 20%
  • ExxonMobil (2019) – 20%

2050 (30-year outlooks)

  • CWB (2019) – 70%
  • IRENA (2018) – 65%
  • BNEF (2019) – 62%
  • Statkraft (2019, low emissions scenario) – 47%
  • DNV GL (2019) – 45%
  • University of Utah (2018) – 40%
  • Shell (2018, Sky) – 33%
  • EIA (2019 International Energy Outlook, reference case) – 28%
  • Equinor (2019, Reform) – 22%

Who cares? Why Does the Percentage of Renewables Matter?

None of the major outlooks above show that the world will be at 100 percent renewables for primary energy by 2050, or even close. Is this necessary to save the world? Maybe not. The major push for renewables, aside from their economic dynamism, is to mitigate the effects of climate change from fossil fuel combustion.

As with long-term energy outlooks, modeling for climate change varies widely. The 2018 IPCC report recommends a CO2 reduction of 45 percent by 2030 to keep us on the right path. Even the 2040 outlooks listed above don’t show that level of reduction, let alone doing so by 2030. The report goes on to recommend a net zero level of CO2 by 2050, which would require 70-85 percent of global electricity production by renewables. The IRENA REmap case predicts a greater than 90 percent renewables penetration for electricity production from China, India, and the EU, while the US lags at 78 percent by 2050. This could mean that a greater than 60 percent primary energy use for renewables by 2050 may be enough to meet the current IPCC recommendations without requiring excessive carbon capture measures.

Do 30-year global energy outlooks matter? Only the fate of the world hangs in the balance.


Charles Botsford's picture

Thank Charles for the Post!

Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.


Matt Chester's picture
Matt Chester on Oct 24, 2019 9:19 pm GMT

Appreciate you throwing your 30 year outlook into the ring, as well. As you point out, each different authoritative outlook has its own inherent biases, mistakes, and shortcomings. But putting them together and looking in aggregate there's always a lot to learn. 

I often wonder why these 30 year outlooks don't put more effort in pointing out their mistakes or adjustments, even showing how they adjust year over year over year since those trends would probably tell us the truth about where things are actually headed. 

Charles Botsford's picture
Charles Botsford on Oct 24, 2019 9:56 pm GMT

Thanks Matt. I put together my first long-term global energy projection back in 2012 (Chemical Engineering Online), and I was off, really off, in hindsight of course. So I thought I would do a 2019 update and discovered that all the other global outlooks were wrong, too, by a lot. Turns out I was too conservative on renewables. Who knew wind and solar costs would drop so precipitously? You can see in my post that someone in the 2019 class has to be wrong, probably all of them--again, even me. I don't feel too bad, though, considering the horsepower these organizations have to throw at their efforts.

Matt Chester's picture
Matt Chester on Oct 25, 2019 1:14 pm GMT

Thanks Charles. Making these outlooks and putting your name behind them, knowing full well the reality is that it's really really hard to be right, is commendable. What do you think you learned the most from your missteps in the 2012 projection that you tried to adapt for in the 2019 version. Obviously you mention that renewable costs dropped more so the penetration was higher than anticipated, but what about from a broader perspective in how to best make these kind of forecasts?

Charles Botsford's picture
Charles Botsford on Oct 25, 2019 4:25 pm GMT

Hi Matt, In the broader perspective, one of the primary variables is world population. Not population growth itself, but per capita energy use projections. In the US, per capita tons of oil equivalent (toe) consumption has been dropping, which is good. Globally, however, it should substantially increase by 2050. To me, this is the difference in EIA's 2050 projection of primary global energy of ~23 Btoe versus DNV's projection of ~16.5 Btoe. BP, IEA, and I are in the middle at ~19 Btoe. 

This has large impacts on projected carbon equivalent emissions, especially if one outlook is 40% higher than another. Who do you believe, and if you're a government organization trying to make policy, how do you justify your actions?

Digging into the weeds, I think I've once again gone conservative on renewables penetration at 70%. My outlook and IRENA's are similar, and both are way out of touch with the mainstream outlooks, but at this level, global carbon emissions will still take us well above the IPCC 1.5C carbon budget. We need to have more renewables, faster, to hit the IPCC budget.

Matt Chester's picture
Matt Chester on Oct 25, 2019 5:16 pm GMT

Really great insights, and thanks for responding. Hope you don't mind if I continue to pester since I find it really interesting!

Digging into the weeds, I think I've once again gone conservative on renewables penetration at 70%. My outlook and IRENA's are similar, and both are way out of touch with the mainstream outlooks, but at this level, global carbon emissions will still take us well above the IPCC 1.5C carbon budget. We need to have more renewables, faster, to hit the IPCC budget.

Is what you're saying here basically coming down to: "to stay under 1.5degC threshold, we need to have greater than 70% renewables, therefore it would make more sense if the forecast predicted greater than 70% renewables"? I ask because if your forecast shows the technologically/economically more feasible number is 70% rather than a greater one, than I don't know how moved I would be by the need to get above 70% to be what dictated the forecast. Lots of well intentioned clean energy goals have fallen short in the past, and while it is an existential necessity we get there, part of me remains skeptical that all actors involved will do what needs to be done.

Charles Botsford's picture
Charles Botsford on Oct 25, 2019 7:03 pm GMT

Hi Matt, I believe in economics dictating the self-interests of the actors. EIA's LCOE and LACE metrics are powerful guides for those actors in doing the right thing -- make money. Fortunately for climate change mitigation, the economics of renewables in terms of LCOE are outstanding, and getting better. 

To answer your question, I struggled with the 70% renewables penetration level. That level is not enough to meet the IPCC carbon budget--too little, too late. I just couldn't justify a higher level, primarily because I couldn't see a steeper reduction in global oil consumption. I'm a proponent of EVs, but oil use for transportation--medium/heavy duty, ships, air, light duty--will take a while to electrify.

Gary Hilberg's picture
Gary Hilberg on Oct 28, 2019 2:29 pm GMT

Charles - I agree with Matt, each one of these should start with a comparison to their last outlook.  Natural Gas pricing in North America has driven dramatic change, just looking at presentations from a Rural Electric conference in 2006 - gas was $6.5/MMBTU and nuclear plants would be built for $4B/unit.  I appreciate your bringing up the point and the inherent bias of most of these studies.  One point that needs to be considered on Cost of Electricity is the expected run time, most fossil plants have massively reduced their operating hours and this drives their COE up dramatically with many coal plants in reserve, the national COE's will be high, but in areas where there are few other choices like Indiana and West Virginia - these plants have a much more competitive COE, so it is difficult to shut them them down and invest in renewables - particularly when the renewable profiles in these areas is not good.  Thanks for the analysis, I do like the 30 year outlooks but am very careful to see who puts them out. 

Vladimir Vinogradov's picture
Vladimir Vinogradov on Oct 29, 2019 1:27 pm GMT

Thanks, Charles - your text is very good. World energy is changing rapidly. The speed of these changes is increasing. We do not know in which world we will live tomorrow. 

Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »