So many studies, so little time. Just in the past couple of weeks analyses from DOE’s Energy Information Administration, Bloomberg New Energy Finance, British Petroleum and the International Renewable Energy Agency have hit my inbox (thank goodness we have moved beyond the old hardcopy stage, just those reports alone would have contributed to the world’s ongoing deforestation problem), and having the time to study them all has been difficult. But muddling through them does provide some fascinating glimpses of where the energy industry is today, and where it might be headed in the years to come.
EIA’s 2016 Annual Energy Outlook, released in abbreviated form last month with its full rollout slated for early July, includes more sobering news for electric utility executives: Sales growth really is gone, and it isn’t coming back. In its analysis, EIA estimates that overall electricity sales will grow at an average rate of 0.7 percent from 2015-2040, essentially unchanged from the 0.6 percent growth rate posted from 2000-2015. But a closer look at the numbers shows even that relatively anemic growth estimate may be optimistic.
For example, EIA estimates that electric sales in the residential sector will rise by an average of just 0.3 percent a year from 2015-2040—well under even the paltry 1.1 percent annual growth recorded from 2000-2015. According to EIA, the slow growth can be attributed to rising energy efficiency, especially in the lighting sector, and the broad adoption of distributed photovoltaics (PV). But what is most intriguing about EIA’s estimate is that virtually all of the growth occurs in the out-years (see chart below): From 2015 through 2030 there is essentially zero growth in residential sales. Specifically, EIA puts 2015 sales in the sector at 1,402 billion kilowatt-hours (kwh) and projects that sales in 2030 will rise to just 1,416 billion kwh—an increase, if you can call it that, of 0.1 percent annually. Rather than calling this growth it would be more appropriate to write it off as a rounding error. It also represents the continuation of a longer-term trend: Residential electric sales in 2007, just before the onset of the Great Recession, totaled 1,392 billion kwh. Measured from that starting point, sales are expected to climb just 24 billion kwh in 23 years, a miserly 0.07 percent annual increase.
From 2030-2040, EIA expects growth to pick up, with sales climbing roughly 7.5 percent during the decade. No reason is given for the stronger growth, perhaps it is simply a matter of the growth in population finally overwhelming the efficiency gains currently baked into the system. However, further efficiencies are just as likely, meaning it is entirely possible this expected growth won’t materialize—extending the flat line or residential sales essentially forever.
The outlook is similar in the commercial sector: Overall, EIA expects demand growth to climb by 0.8 percent annually through 2040, but here again the growth is weighted toward the out-years, raising the question of whether new, more efficient technologies could stifle that expected pickup in demand as well. Through 2025, for example, EIA estimates that commercial electricity sales will rise by less than 0.5 percent per year and then increase gradually after that. Stretching the estimate out to 2030, EIA projects growth will average a little more than 0.6 percent annually (from the 2015 starting point). But then from 2030-2040, EIA estimates that commercial sales will rise by more than 1 percent a year. It could happen, but to my mind that’s a sucker’s bet given the wave of new efficiencies sweeping through the sector. (See this related story here on my blog’s new technology page.)
And even if that projected growth pans out, the near-term still looks bleak for sales-starved electric utilities.
Looking past EIA’s admittedly slow, but probably still over-optimistic sales forecasts, it also seems clear that the DOE forecasting arm continues to understate the absolute upheaval under way in the power generation sector. It has been difficult for most analysts to grasp the speed of the ongoing transition, but given the importance of EIA’s annual outlook I think it is fair to ask if they are accurately projecting what their data are showing.
For example, the 2016 outlook projects that U.S. coal production will bounce around its 2015 level of 870 million tons through 2020. During this same period, coal-fired electric generation is expected to increase slightly as rising natural gas prices make coal more attractive. After 2020, EIA says, coal production is expected to start falling, and dip below the 800 million ton mark in 2023 or 2024—before falling all the way to 664 million tons in 2030. But a closer look at current statistics shows that far from leveling off, the freefall in coal output (and coal-fired electric generating) is continuing.
In a recent Today In Energy post on its web site, EIA noted that coal production in the first quarter had tumble to just 173 million short tons—the lowest quarterly output in 35 years. EIA attributed much of the decline to warmer-than-average winter temperatures and higher-than-normal stock buildups during the fourth quarter of 2015. All that may be true, and coal output may head upward again as the year continues, but it is worth noting that at least for the past five years, coal production has been surprisingly consistent. In fact, since the beginning of 2010, quarterly output has been almost exactly that, a quarter of total annual production (see the chart below).
In other words, output may tick upward somewhat, but a significant increase is unlikely given past production patterns. And that means it is entirely possible production this year will fall below 800 million short tons—seven or more years ahead of EIA’s outlook. While less likely, it is even possible that output will not top 700 million tons this year, a level not projected by EIA until the late 2020s. Should the quarterly production pattern hold true, 2016 output would come in at about 692 million tons; further bolstering this pessimistic scenario for coal is EIA’s weekly production data (found here), which shows output through June 11 (week 24 of the year) at just 289.6 million tons—do the math for the whole year and it’s not a very pretty picture.
Looking at EIA’s Monthly Electric Report reveals a similar situation. The 2016 Annual Outlook projects coal-fired electric generation this year will total 1,357 billion kwh, up ever-so-slightly from the 1,355 billion kwh produced in 2015, and remain roughly at that level through 2020 (see chart below). But unless there is a huge change in current generation patterns in the next several months there is no way that forecast is going to come to fruition—through the end of March total coal-fired generation was just 278.9 billion kwh, which equates to roughly 1,115 billion kwh for the year. Now, clearly a hotter-than-normal summer could help the coal-fired generation fleet inch closer to EIA’s projection, but a look at the 12-month rolling totals for the past two years makes it clear that it will be difficult for the sector to climb out of the hole it is in: For the 12 months ending March 2015, coal’s output totaled 1,513 billion kwh, while by March of 2016 the rolling 12-month total had dropped to 1,266 billion kwh.
Natural gas has been the largest reason behind the collapse in coal generation, but renewable are playing an increasingly important role as well, and here too EIA seems unwilling to recognize the speed of the ongoing transition, especially on the solar side of the equation. In its latest outlook, the data geeks estimate that installed PV capacity in the U.S. will struggle to top 50 gigawatts by 2020 and won’t hit the 100 GW mark until after 2025. By contrast, analysts at Greentech Media expect solar capacity to be nearing 50 GW by the end of the current year, and to shoot past 100 GW in 2021.
Looking at just the distributed side of the market (meaning residential and commercial installations), EIA’s estimates, while significant, are still dwarfed by those of other analysts. In its latest outlook, for example, EIA projects that roughly 36 percent of all the solar capacity installed through 2040 will be distributed PV; with an aggregate estimate of 250 GW, this means EIA expect about 90 GW of distributed to come online in the next 25 years. In contrast, in a recent study on distributed energy resources (DERs) in general, Navigant Research projected that 75 GW of solar powered DERs will be installed in the next 10 years alone. Those two estimates clearly can’t share much in common.
Finally, there has to be something wrong with any forecast that assumes essentially no windpower resources will be brought online after 2022 when the current production tax credit expires (see chart below). The PTC is worth $23 per megawatt-hour, clearly a significant benefit. But take it away (or add it to current purchased power agreements, which have been in the $25-$30 per mwh range, to make the comparison fair) and wind still comes in at less than $55 per mwh, clearly competitive with existing alternatives. (Price/production data was pulled from this article.) Given that, EIA’s flat-line for wind post 2022 simply can’t be believed.
It’s time for EIA to believe in its own data and forecast accordingly—an energy revolution is under way, but the agency may well miss it entirely unless it has the courage to follow its own numbers.