Wednesday, December 11, 2013

Turboprop update, and a Hot Air Oscar nomination

ATR600 - image from atraircraft.com My attention was recently drawn to an impressively fuel-efficient turboprop aircraft, the ATR72-600, which is claimed to be about one third more energy efficient than Bombardier's Q400 turboprop, which I featured on page 35 of SEWTHA.

HotAirOscar-ATR72 I've consequently written an update on turboprops, celebrating this achievement, but in the interests of balance I feel I should also nominate the advertisers of the ATR72-600 for this year's Hot Air Oscar for the most misleading "green" infographic, for this astonishing picture [at left] showing the difference between the fuel consumption of the ATR 72 and the Q400 on a 250-nautical-mile journey. As the numbers in the picture show, the ATR 72's fuel consumption is 70% of the Q400's, but the volume of the three-dimensional blue barrel shown is 30% of the volume of the orange barrel — a 2.3-fold exaggeration!
blue barrel:orange barrel
ratio of diameters91:134=0.68:1
ratio of heights 118:179=0.66:1
ratio of volumes
(as depicted)
=0.68×0.68×0.66
=0.30:1
true ratio of volumes
(735:1043)
=0.70:1

Do UK wind farms decline "very dramatically" with age?

In December 2013, Christopher Booker in the Telegraph discusses a study by Gordon Hughes, published by the Renewable Energy Foundation in December 2012, which is said to show, due to wear and tear on their mechanisms and blades, the amount of electricity generated by wind turbines "very dramatically falls over the years". Booker asserts that "Hughes showed his research to David MacKay, the chief scientific adviser to the Department of Energy and Climate Change, who could not dispute his findings." This is not true.
In fact, I doubted Hughes's assertions from the moment I first read his study, since they were so grossly at variance with the data.
Figure 1: Actual load factors of UK wind farms at ages 10, 11, and 15.
a) Histogram of average annual load factors of wind farms at age 10 years. For comparison, the blue vertical line indicates the assertion from the Renewable Energy Foundation's study that "the normalised load factor is 15% at age 10."
b) Histogram of average annual load factors of wind farms at age 11 years.
c) Histogram of average annual load factors of wind farms at age 15 years. For comparison, the red vertical line indicates the assertion from the Renewable Energy Foundation's study that "the normalised load factor is 11% at age 15." At all three ages shown above, the histogram of load factors has a mean and standard deviation of 24% ± 7%.

Moreover, by January 2013 I had figured out an explanation of the underlying reason for Hughes's spurious results. I immediately wrote a technical report about this flaw in Hughes's work, and sent it to the Renewable Energy Foundation, recommending that they should retract the study.
I would like to emphasize that I believe the Renewable Energy Foundation and Gordon Hughes have performed a valuable service by collating, visualizing, and making accessible a large database containing the performance of wind farms. This data, when properly analysed in conjunction with detailed wind data, will allow the decline in performance of wind turbines to be better understood. Iain Staffell and Richard Green, of Imperial College, have carried out such an analysis (in press), and it indicates that the performance of windfarms does decline, but at a much smaller rate than the "dramatic" rates claimed by Hughes. The evidence of decline is strongest for the oldest windfarms, for which there is more data. For newer windfarms, the error bars on the decline rates are larger, but Staffell and Green's analysis indicates that the decline rates may be even smaller.
I will finish this post with a graphical explanation of the flaw that I identified (as described in detail in my technical report) and that I believe underlies Hughes's spurious results.
The study by Hughes modelled a large number of energy-production measurements from 3000 onshore turbines, in terms of three parameterized functions: an age-performance function "f(a)", which describes how the performance of a typical wind-farm declines with its age; a wind-farm-dependent parameter "ui" describing how each windfarm compares to its peers; and a time-dependent parameter "vt" that captures national wind conditions as a function of time. The modelling method of Hughes is based on an underlying statistical model that is non-identifiable: the underlying model can fit the data in an infinite number of ways, by adjusting rising or falling trends in two of the three parametric functions to compensate for any choice of rising or falling trend in the third. Thus the underlying model could fit the data with a steeply dropping age-performance function, a steeply rising trend in national wind conditions, and a steep downward trend in the effectiveness of wind farms as a function of their commissioning date (three features seen in Hughes's fits). But all these trends are arbitrary, in the sense that the same underlying model could fit the data exactly as well, for example, by a less steep age-performance function, a flat trend (long-term) in national wind conditions, and a flat trend in the effectiveness of wind farms as a function of their commissioning date.
The animation above illustrates this non-identifiability. The truth, for a cartoon world, is shown on the left. On the bottom-left, the data from three farms (born in 87, 91, and 93) are shown in yellow, magenta, and grey; they are the sum of a age-dependent performance function f(a) [top left] and a wind variable v_t [middle left]. (The true site 'fixed effects 'variables u1, u2, u3 are all identical, for simplicity.) On the right, these identical data can be produced by adding the orange curve f(a) to the site-dependent 'fixed effects' variables u1, u2, u3 (shown in green), thus obtaining the orange curves shown bottom right, then adding the wind variable [middle right] shown in blue (v_t).

Monday, November 18, 2013

Enormous solar power stations

Three spectacularly large solar power stations have recently been in the news: Ivanpah, located in California, but within spitting distance of Las Vegas, is a concentrating solar power station in which 300,000 flat mirrors focus sunshine onto three power-towers. Solana, located in Gila Bend, Arizona, has a collecting field of about 3200 parabolic-trough mirrors, each about 25 feet wide, 500 feet long and 10 feet high, and it can generate electricity at night thanks to its ability to store high-temperature heat in vast molten salt stores. Kagoshima, near the Southern tip of Japan, has 290,000 solar photovoltaic panels.
All three are enormous, and must be amazing to visit: Ivanpah occupies about 14 km2; Solana, 12.6 km2, and Kagoshima, 1 km2.

Now, I'm always interested in powers per unit area of energy-generating and energy-converting facilities, so I worked out the average power per unit area of all three of these, using the estimated outputs available on the internet. Interestingly, all three power stations are expected to generate about 8.7 W/m2, on average. This is at the low end of the range of powers per unit area of concentrating solar power stations that I discussed in Chapter 25 of Sustainable Energy - without the hot air; Andasol, the older cousin of Solana in Spain, is expected to produce about 10 W/m2, for example.

I published a paper on Solar energy in the context of energy use, energy transportation, and energy storage in the Phil Trans R Soc A Journal earlier this year, and these three new data points lie firmly in the middle of the other data that I showed in that paper's figure 8 (original figures are available here). .

These data should be useful to people who like to say "to power all of (some region) all we need is a solar farm the size of (so many football fields, or Greater Londons, or Waleses), if they want to get their facts right. For example, Softbank Corporation President Masayoshi Son recently alleged that "turning just 20% of Japan’s unused rice paddies into solar farms would replace all 50 million kilowatts of energy generated by the Tokyo Electric Power Company". Unfortunately, this is wishful thinking, as it is wrong by a factor about 5. The area of unused rice paddies is, according to Softbank, 1.3 million acres (a little more than 1% of the land area of Japan). If 20% of that unused-rice-paddies area were to deliver 8.7 W/m2 on average, the average output would be about 9 GW. To genuinely replace TEPCO, one would need to generate roughly five times as much electricity, and one would have to deliver it when the consumers want it.

Maybe a better way to put it (rather than in terms of TEPCO) is in national terms or in personal terms: to deliver Japan's total average electricity consumption (about 1000 TWh/y) would require 13,000 km2 of solar power stations (3.4% of Japan's land area), and systems to match solar production to customer demand; to deliver a Japanese person's average electricity consumption of 21 kWh per day, each person would need a 100 m2 share of a solar farm (that's the land area, not the panel area or mirror area). And, as always, don't forget that electricity is only about one third or one fifth of all energy consumption (depending how you do the accounting). So if you want to get a country like Japan or the UK off fossil fuels, you need to not only do something about the current electricity demand but also deal with transport, heating, and other industrial energy use.


Sources: NREL; abengoa.com; NREL; solarserver.com; and google planimeter.

Monday, October 14, 2013

Chinese translation of Sustainable Energy - without the hot air

The Chinese translation of Sustainable Energy - without the hot air is now available on amazon.cn
I am very grateful to the Chinese Academy of Sciences and President Li Jinhai for arranging both the translation and its publication. Thank you!

Sunday, June 9, 2013

David MacKay's "Map of the World" - an update

I've updated my "Map of the World" which shows, country by country, how human power-consumption per unit area compares with the power-production per unit area of renewables. I originally published this graph on my blog in August 2009. I've made quite a few improvements to it since then, including the representation of country size by point size, and colour coding of continents in the style of Gapminder.
One interesting thing I figured out while working on this graph is that, while the average power consumption per unit land area of the world is 0.1 W/m2, 78% of the world's population lives in countries where the average power consumption per unit land area of the world is greater than 0.1 W/m2 — much as, in a town with some crowded buses and many empty buses, the average number of passengers per bus may be small, but the vast majority of passengers find themselves on crowded buses.
Please follow this "Map of the World" link to see multiple versions of the graph, and to download high-resolution originals, which everyone is welcome to use.
My "Map of the World" graphs are published this year in two journal papers, which I will blog about shortly.
David J C MacKay (2013a) Could energy-intensive industries be powered by carbon-free electricity? Phil Trans R Soc A 371: 20110560. http://dx.doi.org/10.1098/rsta.2011.0560 This paper also contains detailed information about the power per unit area of wind farms in the UK and USA, and of nuclear power facilities
David J C MacKay (2013b) Solar energy in the context of energy use, energy transportation and energy storage. Phil Trans R Soc A 371: 20110431. http://dx.doi.org/10.1098/rsta.2011.0431 This paper also contains detailed information about the power per unit area of solar farms

Monday, April 8, 2013

I've been unfair on Hydrogen

In Sustainable Energy - without the hot air I spent a couple of pages discussing hydrogen transportation, under the title "Hydrogen cars – blimp your ride". While I still think that some people have been overhyping hydrogen - even Nature magazine, who praised Governor Arnold for filling up a hydrogen-powered Hummer - some of the criticisms I wrote were incorrect and I wish to correct them.

On page 131 I wrote: ... hydrogen gradually leaks out of any practical container. If you park your hydrogen car at the railway station with a full tank and come back a week later, you should expect to find most of the hydrogen has gone. Both of these statements are incorrect.

First, while hydrogen is a very leaky little molecule, it is possible to make practical containers that contain compressed hydrogen gas for long durations. It's just necessary to have sufficient thickness of the right type of material; this material may be somewhat heavy, but practical solutions exist. The technical term used in the hydrogen community for this topic is "permeation", and it's especially discussed when ensuring that hydrogen vehicles will be safe when left in garages. Hydrogen containers are currently classed in four types, and the metallic containers and containers with metallic liners (Types 1, 2, and 3) have negligible permeation rate. However, hydrogen permeation is an issue for containers with non-metallic (polymer) liners (Type 4) which readily allow the permeation of hydrogen. [Source: P. Adams et al]

Second, when discussing the hydrogen vehicle that is left for 7 days, I incorrectly tarred all hydrogen vehicles with a hydrogen-loss brush that applies only to vehicles that store liquified hydrogen at cryogenic temperatures. There are in fact three types of hydrogen storage: Compressed gas (typically at 350 or 700 bar); Cryogenic (typically at less than 10 bar and at extremely low temperature) and Cryo-compressed (at low temperature and at pressures up to about 350 bar). The hydrogen community discuss the "loss-free dormancy time" and the "mean autonomy time" of a system, which are respectively the time after which the system starts to lose hydrogen, and the time after which the car has lost so much hydrogen it really needs refilling. In the US Department of Energy's hydrogen plans, the targets are for a loss-free dormancy time of 5 days and a mean autonomy time of 30 days. Cryogenic liquid-hydrogen systems (such as the one in the BMW Hydrogen 7, which I featured in my book) do not currently achieve either of these targets. (And the reason is not that the hydrogen is permeating out, it's that heat is permeating in, at a rate of 1 watt or so, which gradually boils the hydrogen; the boiled hydrogen is vented to keep the remaining liquid cold.) However, compressed-gas systems at 700 bar can achieve both of these targets, so what I wrote was unfair on hydrogen vehicles. [Source: EERE 2006 Cryo-Compressed Hydrogen Storage for Vehicular Applications]

I apologise to the hydrogen community for these errors.

I will add a correction to the errata imminently.