Over the past few years, the U.S. Energy Information Administration (EIA) has been widely criticized for consistently and significantly underestimating the growth of renewable energy, both in the U.S. and the world. For instance, see the December 2013 post here on Scaling Green, "EIA Renewable Energy Forecast Isn’t Just Wrong, It’s Wildly, Laughably Too Low." As we pointed out then:
*In 2005, EIA forecast that U.S. solar power capacity would hit about 1.2 GW in 2013. Where are we right now? According to Greentech Media, the U.S. is closing in (if it already hasn’t passed) the 10 GW mark in solar PV capacity right about now, and that’s not even counting solar thermal power generating capacity (according to this article, you can add another 1 GW or so of U.S. solar thermal power capacity). In sum, EIA forecast 1.2 GW of U.S. solar power capacity in 2013; the actual figure is around 11 GW – nearly 10 times higher than EIA forecast!
*In 2005, EIA forecast that U.S. wind power capacity would reach about 9 gigawatts (GW) in 2013. Where, in fact, are we right now? According to theAmerican Wind Energy Association (AWEA), installed U.S. wind power capacity at the end of 2012 was 60 GW. Quick math: EIA’s forecast of 9 GW compared to an actual 60 GW? That’s off by a factor of nearly 7!
More recently, see: DeSmogBlog's March 13, 2016 piece, "Renewable Energy Growth Blows EIA Forecasts Out of the Water, Again;" Climate Progress' April 2015 article, "This Federal Report Underestimates Renewables Every Year, And Energy Experts Have Had Enough;" Greentech Media's "Why EIA's Energy Outlook Misses the Real Value of Renewable Energy" (April 2015); CleanTechnica's post, "EIA 2040 Forecast Understates Renewables, Policy, Contingencies" (April 2015); and Get Energy Smart Now's take (May 2014). That's just a small sampling of the criticism EIA has received for its wildly low forecasts for clean energy scaling.
Now, EIA has fired back, with a new report, "Wind and Solar Data and Projections from the U.S. Energy Information Administration: Past Performance and Ongoing Enhancements." According to EIA, the criticisms it has received are unfair, as its reference cases "explicitly assume current laws and regulations in place at the time the projections are developed, consistent with EIA’s mandate to provide a neutral baseline for policy analysis." EIA adds, "When policies are assumed that support renewables, EIA’s projections show significantly higher penetration rates for both wind and solar than when such policies are not considered." EIA does admit that when it comes to solar, "some significant gaps between projected and realized outcomes are apparent, even when policy assumptions used in making the projections align with realized policies." EIA correctly points out that, "like many other industry trend watchers, [it] did not anticipate the sharp decline in solar PV costs seen over the past several years." Finally, EIA claims, without any evidence, that its "market projections for wind and solar electric generating capacity provide a reasonable basis for understanding the impacts of laws, policies, and market conditions on the growth of renewable energy markets."
With that in mind, here's a quick review of how EIA's done recently when it comes to wind and solar forecasts.
- In its 2008 Annual Energy Outlook (AEO), EIA's reference case forecast for U.S. wind power capacity in 2015 was 29.6 gigawatts (GW), or 152% lower than the 74.5 GW it turned out to be. That's not good, but it's great compared to EIA's 2008 AEO forecast for solar power capacity in 2015, which was off by a whopping 19,471% (0.14 GW forecast vs. 27.4 GW actual solar capacity in 2015).
- Similar results for EIA's reference case wind and solar forecasts were seen in its 2009 AEO.
- For its 2010 AEO, the reference case 2015 forecast for wind was a bit closer (only 16% low), but the solar forecast remained off by a huge margin (low by 19,471%).
- The 2012 AEO actually was even further off than the 2010 AEO for 2015 wind power capacity (31% low), while the solar forecast for 2015 improved markedly ("only" 122% low).
Perhaps an even bigger problem than badly missing clean energy forecasts for 2015 is that EIA's long-term forecasts, out to 2030 and beyond, consistently show extremely low growth rates for U.S. wind and solar power capacity. That includes the 2012 AEO, which shows only 8.6% annual growth for solar power capacity out to 2035 -- an absurdly low growth rate given what we're seeing actually happening. In fact, far from growing by only 8.6% per year, Greentech Media recently predicted U.S. solar power capacity growth of 119% in 2016, with continued rapid growth in the 2017-2021 time period as well.
Yet EIA's latest long-term forecast (AEO 2015) shows U.S. solar power capacity growing only 6.0% per year out to 2040, reaching 60.6 GW at that point (while wind grows at an even slower, 2.2% rate, hitting 109.7 GW in 2040). In stark contrast, Greentech Media believes that U.S. solar power capacity will exceed 100 GW by 2021, far past the 60.6-GW forecast by EIA for 2040. As for wind, the U.S. Department of Energy forecast in 2015 that U.S. wind power capacity could hit 224 GW by 2030 and 404 GW by 2050, again far beyond EIA's forecast of 109.7 GW by 2040.
The bottom line is that, yes, EIA is obligated to base its forecasts on current U.S. policy. However, even given that constraint, EIA's forecasts for wind and solar have been consistently and wildly low, both for 2015 and in the long term. The question is not just why that's the case, but more importantly what EIA's doing to correct it. Unfortunately, none of that has been cleared up by release of EIA's new report on its forecasting track record.
P.S. Also see Jigar Shah's response to EIA, which states: "Solar power is cost effective with only the 10% permanent tax credit, but EIA says that is absolutely not possible. It is like they don't know how to do math. You tell me whether their response admitted these fundamental facts or if they took a "holier than thou" tone to explain to us that they got it right the whole time -- we just looked for the data in all the wrong places."