One of the major weapons in the wind energy proponents’ quiver is a report titled 20% Wind Energy by 2030. It was published by NREL, the National Renewable Energy Labs, which is part of the U.S. Dept. of Energy, in May 2008. It lays out a blueprint on how the U.S. could attain 20% of its electricity production by 2030. It has been widely used as an authoritative source by just about every industry body and many green-leaning politicians as well, all the way up to President Obama. Since the DOE is headed by a Nobel-prize-winning physicist (Dr. Steven Chu), and they’ve got lots of money and PhD-level people, you’d think such an important report would be unassailable, especially by a mere mortal from a small town in Ohio, working out of his garage (literally). You might be thinking that I’m just another anti-wind agitator who would always find something to quibble about in any otherwise solid piece of work. I hope that after reading this posting you’ll have some appreciation of just non-quibbly the problems are, and how truly stupid we are for using it to justify all the financial, environment and social costs of wind energy.
I wrote an earlier critique of how this report obtained its CO2 savings numbers, which still stands. But I’m constantly reviewing things I’ve written and as part of that process I chanced through this report again. In addition to the problems with the CO2 calculation I noticed a number of other problems, enough to warrant another posting – this time on a variety of topics contained in the Report.
First, a link to the main body of the report [backup link]. It’s a big report, 250 pages.
Second, a link to the 21-page executive summary [backup link].
Third, a link to a 35-slide presentation [backup link]. Some of my references are to here.
Third, a link to Hand, 2008 – an 11-page paper [backup link] that both I and the Report use.
The U.S. is projected to use 5,800,000,000 MW-H by 2030. This comes to an average generation of 662,000 MW. Wind energy’s 20% of this comes to 1,160,000,000 MW-H, or an average of 132,000 MW. So far, so good. The Report proposes installing 305,000 MW of wind capacity to meet this goal. Calculating the Capacity Factor: 132/305 comes to 43.3%. Whoops! Half of that would be more typical. Anyone with even a passing knowledge of wind energy should be rolling their eyes at this point. Surely all those clever people at NREL didn’t make this basic of a mistake, right? Well, let’s see.
If you look at their primary assumptions (slide 5), one of them is that the CF increases by 15% by 2030. 15% doesn’t sound like much, given the march of technology and all, but the devil is in the details. First off, most of us would think the 15% is an increase from the current base; in other words, if the current CF is 30%, the increase goes to 34.5%. However, that’s not what they meant. They’re going from 30% to 45%, which by most logic is a 50% gain in efficiency. They offer two lines of logic to justify this.
First, they speculate that turbines will be getting bigger and more importantly the towers higher, where the winds are better. The problem I see with this is that absent some unanticipated metallurgical breakthrough there’s no way to significantly increase the height without increasing the diameter, and there’s no way to transport anything much bigger on U.S. roads. They offer no clues as to how these larger towers will arrive at their projects – see p. 26 of the full report – and even acknowledge the size and height limitations of land-based turbines. Yet they continue to hold to the assumption. This is not scientific behavior.
Second, they mention that CF’s have increased over the last decade or so (see Table 2-5 on p. 27 of the full report), and will likely continue to do so, at the same rate. If you look at the EIA’s capacity and production numbers from 1998 to the very favorable 2006 (the last year they would have had access to) you might make a case. Here’s the full numbers, plus a chart.
If you take the trend line, it increases, for this period, by 4.5% over 9 years, or 0.5% per year. So in the period from 2000 to 2030, voila!, you get your 15%. Apparently the people at NREL came to age when solid-state advances were so large and apparently so easy they think large mechanical devices (arguably the largest rotating structures ever built by man) will also follow that kind of curve.
In another complication, as mentioned on p. 31 of the full report, current turbines extract about 50% of the wind energy present. Since the Betz limit is 59%, one wonders how the turbines will increase their efficiency by 50%.
Returning to the real world, how’s their projection doing? Not so well, as you can see in 2007 to 2009.
If you use a more reasonable 25% CF, you end up with 528,000 MW of capacity needed instead of 305,000, a 75% increase. This large an increase is not a quibble, and has major implications for the feasibility of their entire project, some of which I cover below.
At the risk of repeating my earlier critique, here’s some additional numbers. Just about every proponent (i.e. AWEA) quotes an annual emissions savings of 825 million tons. That number comes from this very Report – see slide 21. If you divide the savings by the wind production, you come up with a savings of 0.71MT/MW-H, or 1.57Lbs/KW-H. This number is consistent with my earlier critique, giving me confidence that my numbers are accurate. If you don’t want to follow the link to read it, I can sum it up by saying the 1.57 (it came to 1.52 there – close enough) came not from the government, but from AWEA (the box on p. 107 represents their sole basis), an industry lobbyist for Christ’s sake, which simply took an average emission rate of what they assumed wouldn’t be burned. The Bentek study put an end to that underlying assumption. There’s never been an empirical number published, and lots of reasons to think the emissions savings is for all practical purposes close to zero. This lack of any evidence for any savings is not a quibble.
The installation of this much wind generation represents a very large and potentially destabilizing change to the grid upon which our standard of living depends. NREL points to studies that show this much wind energy can be accommodated, but these studies are inevitably based on assumptions and computer models, created by people who have an interest in the results.
Slide 16 has a small picture of the change, while the full report contains two Figures from which a better picture can be seen, and which are reproduced below. First, Figure 5, from page 4 of Hand, shows the capacity changes that the Report is suggesting. I’ve added some horizontal lines and my guess at the numbers the bars represent.Adding the numbers up, the “normal” case has 1170GW of capacity. Recall that 622GW is the assumed average generation. The 20% case has a total capacity of 1435GW, of which 305 is wind and the remaining 1130 is conventional, representing a 21% penetration of wind into the grid. Never mind that Germany with a penetration of 19% is having significant grid balancing problems.
Thus the conventional (as in, reliable and dispatchable) capacity drops by 40GW. Apparently NREL thinks 305GW of wind can permanently replace 40GW of conventional generation. This implies what is called a Capacity Value of 13%. The full report goes into this discussion starting on p. 87, but I remain unconvinced that wind, which can be becalmed over large areas for long periods, has any CV at all. What happens if the 43% CF happens, one year, to be “just” 25%? The 132GW average drops to 76GW, and now you’ve lost 96GW of capacity – some 14.5% of your average generation.
The major changes are an increase in the use of “CT” gas turbines – aka open cycle, or simple cycle turbines, that are responsive enough to balance wind’s variability. In fact, the capacity of the CT’s matches the capacity of the wind. There’s also a lot less new coal – which emits the most emissions – being built.
Using my more realistic 25% CF, the resulting 528GW of wind now increases the total capacity from 1435GW to 1658GW and the wind penetration increases from 21% to 32%. But you’d also need more CT’s to match the wind, so now the capacity goes up to 1881GW. A likely 60% increase in capacity and all the associated infrastructure costs to meet your goal is not a quibble.
The next figure is the proposed generation breakdown, as shown in Fig. 6, from p. 5 of Hand, again with my additions.The numbers add up to about 5600 TW-H, compared to the 5800 TW-H mentioned at the start. Maybe there’s some “other” they left out. No matter, it’s close enough. The biggest item that pops off the chart is the low usage of the load-balancing CT’s. It represents a 1% CF. I doubt many investors would be willing to build at CT if it were projected to have a 1% CF. The promise of more subsidies and/or very high (and thus volatile) wholesale prices would be needed to get someone interested. The full Report has a short section A.4.1 on p. 151 on this topic, but it doesn’t get into enough detail to understand how they plan on getting anyone to build all those CT’s. One of the stated benefits to their plan is to “stabilize electric rates”. Exactly how this meshes with the low CF of the CT’s is not clear.
There’s also a problem with their balancing assumptions. Wind production varies continuously, and can vary by +/- 20% in a 5 minute period. Exactly how will 305GW of CT’s running at an average of 3GW balance305GW of wind that can swing multiple GW’s per minute? You could speculate about perfect forecasting or a super grid providing a potential fix, but neither of these may ever come to fruition, and the costs in neither case are mentioned.
They claim that the incremental cost of wind is a total of $43B (full report, p. 19), mostly because the actual capital cost is offset by the fuel savings. I have serious doubts about the fuel savings – like emissions savings, they’ve never been empirically demonstrated. I’d bet those CT’s, with their higher fuel consumption and emissions compared to the CC’s, will be cranking out a lot more than 3GW.
But even aside from their dubious assumed offsets, I cannot get their numbers to add up. From Section B.3.4, p. 181, a MW of wind capacity has an average capital cost $1,650,000/MW. If you build 305,000MW of capacity, I think that comes to a total of a little over $500B. Yet, on p. 19 they claim a capital cost of $197B. Maybe Washington considers $300B a quibble.
Chapter 5 of the full Report covers the environmental downsides to wind energy. The answers it provides could have come from AWEA’s web site. In fact, they probably did. Birds? Other things kill more. Never mind Wolfe Island. Bats? Barely mentioned, needs more study, but in the meantime, build away! Noise? No louder than a refrigerator. Never mind Rand and James. Land values? REPP and Hoen found no effect. Never mind McCann.
One of the things they studied was the availability of the raw materials needed to build all those turbines. P. 64 mentions saving copper (and electricity) by using permanent magnets, of which 90% come from China, and 95% of the rare-earth raw materials. Their statement that “…availability is not expected to be restricted,…” lasted roughly 18 months before China brought the hammer down.
I could go on and on about the problems this report has, but it’s been a long posting and I thank you for reading it. The NREL has a scientific-sounding name, but I fear it has been seized by a religious fervor. Somewhere from on high the 20% goal came down and from that point on all scientific skepticism was discarded, the assumptions were worked and reworked to produce a report that looked as authoritative as it could. Unfortunately, looking authoritative is not the same as being true. Instead of searching for the truth, NREL has now become a renewable advocacy organization, really tight with people like AWEA, whom they reference regularly.