Thanks to Andrew Ferguson for this paper.
“ZERO CARBON BRITAIN 2030″ from the Centre for Alternative Technology.
Chapter 8, Renewables, reviewed by Andrew Ferguson
Abstract. The plan set out by the Centre for Alternative Technology (CAT), aiming at a goal of Britain achieving no net emissions of carbon dioxide by 2030, is, unfortunately, not based on reality. This is mainly due to it relying on a mistaken belief that an electrical system can accommodate an almost unlimited amount of uncontrollable input, and partly due to using land that is needed for growing food and fibres for growing biomass. Allowing only a plausible amount of uncontrollable input suggests that to be sustainable the population of Britain needs to reduce to about 20 million.
I have always admired the Centre for Alternative Technology (CAT) for its efforts to find out more about how humans could live in ways that are less ecologically damaging, and for its magazine Clean Slate, which provides information about those activities. However, I was much troubled when CAT engaged itself in a project to which it gave the title ZeroCarbonBritain2030 (ZCB2030), which boldly proclaimed the possibility of ensuring that Britain could so change its energy technology that it could achieve no net emissions of carbon dioxide by 2030. This seemed to me a fatuous hope, and I felt no more inclined to study it than I would a treatise on a perpetual motion machine ! But when I conveyed my thoughts to CAT, they sent me a copy of the 368 page book describing the scheme, “ZeroCarbonBritain2030 (the second report).” I decided that I should at least take the time to study the chapter on renewables to see where the authors – none specified but fifteen participants listed – thought that the required renewable energy could be obtained, for satisfying Britain’s approximate current 60 million inhabitants.
It did not take long to find out that the major failing of the analysis was the same as that which weakens many similar plans for a future based on renewable energy, including David MacKay’s otherwise excellent book, Sustainable Energy – without the hot air (2008), and to a lesser extent Patrick Moriarty and Damon Honnery’s Rise and Fall of the Carbon Civilisation (2010). These two books were reviewed in the April 2011 issue of the OPT Journal (http://tinyurl.com/optjournal). The common fault of such plans is a failure to recognize that uncontrollables can only contribute 30%, at most, to the total amount of electricity supplied via the grid. I have given the reasons for this in many previous issues of the OPT Journal, but they are germane to this review and I will briefly reiterate them.
Any system delivering electricity to a grid in an uncontrollable manner (i.e. delivering its energy whenever it is available, not when it is demanded), which has peaks of output interspersed with periods of much lower output (as do all uncontrollables), can only be fitted into an electricity system up to the size where the peaks do not exceed the demand for electricity. It is sometimes thought that wasting some of the electricity delivered at the peaks of output would not matter much. This is a difficult matter to get a clear handle on, but whether it would be practical is certainly dubious. The subject was dealt with in fair detail in the article The Problem with Uncontrollables (OPTJ 10/1, pp21-24), but mainly with reference to the United Kingdom. I look at it again in a more general way at Appendix A, but here let it suffice to say that the matter is sufficiently dubious that it should not be assumed to be an available solution until it can be demonstrated empirically that it is a feasible way of tackling the problem.
Lowest demand for electricity is around 66% of the average demand. So uncontrollables can only fit into the system provided that the peak of their output does not exceed 66% of the average electricity demand. Let us suppose that the performance of wind turbines in Britain improves, so that instead of the 2005 annual load factor (also known as capacity factor) of 28%, they achieve an average of 35%. With this capacity factor, the wind turbines could contribute 0.35 x 0.66 = 23% of the electricity delivered by the system; but the remaining 77% would need to be delivered by controllable systems.
Renewable energy enthusiasts put much store by the possibility of flattening demand, even to the extent that there is no ‘lowest demand’ – i.e. demand always stays near the average. That is a fairly incredible goal, but to probe the ultimate limits of that assumption let us imagine that it is achieved. With a 35% load factor the wind turbines could then satisfy 35% of the average electrical demand, i.e. supply 35% of the total electricity.
Not only is a complete flattening of demand unlikely, but there are other factors which reduce the size of the uncontrollable that can be fitted into the system. First, if Combined Heat and Power (CHP) plays a part in the system, when there is a strong demand for heat there are usually practical limits as to how much the electrical output of the system can be turned down and the heat output turned up. Secondly, if there is nuclear power in the system, there is no likelihood that the nuclear plant can be almost turned off to accommodate peaks of output from the uncontrollables. Both these factors further limit the size of the uncontrollable that can be fitted in. Thus while we may satisfy more than 23% of electrical demand, we are unlikely to get anywhere near to 35%. To assume that uncontrollables can satisfy 30% of the total demand for electricity would be optimistic.
In case this logic and arithmetic is less than convincing, I will mention that Lenzen (2009, p19) stated that, “the main barrier to widespread wind power deployment is wind variability, which poses limits to grid integration at penetration rates above 20%.” And in an impressive analysis, Jim Oswald (2007) showed that the United Kingdom would have difficulty in handling more that 25 GW of wind capacity even if the wind turbines were placed all over Britain to try to limit the peaks. At a 35% load factor, that would contribute about (25 x 0.35 ) / 44 [GW average demand] = 20% of electrical demand.
After that lengthy introduction, we can now set about analysing a fairly realistic contribution from renewables were we to attempt to follow the ZCB2030 proposals – that is after making them fairly realistic by using only a realistic amount of uncontrollables. What would be fairly realistic? As just explained, an upper limit for uncontrollables is for them to provide 30% of the whole of the electricity supply. Another way of putting that is that uncontrollables can supplement controllables to the extent of increasing the total to 1 / 0.70 = 143%; that is they can add 43% to the output of the controllables.
Table 1 sums up the data taken from Figure 8.9 on electricity generation, page 261 of ZCB2030, , but adjusted to show only a realistic contribution from uncontrollables. Note that all the figures shown in bold in Table 1 are taken directly from Fig. 8.9. Now let us look in more detail at some of the lines in the table.
The first three lines under the section headed “Electrical output from controllables” refer to sources that are somewhat limited in the extent to which their output can be controlled: (1) Hydro is often limited in summer because the water in the reservoir needs to be retained for other purposes than providing electricity when wanted; (2) CHP is limited to the extent that if heat is required there is not likely to be the flexibility in the system to ensure that the CHP system delivers only heat and no electricity; (3) “Fixed tidal” is also somewhat limited; for instance, if one releases water when there is only a small head of water, then little electricity is generated. Only biogas and biochar are fully controllable. Thus the subtotal shown for the controllables contribution is supremely optimistic.
It should also be noted that to provide the amount of electricity shown from CHP, about 2.3 million hectares of good quality land would be needed to grow biomass crops. In this chapter of ZCB2030 I see no mention of the fact that the UK only supplies about 70% of the food it consumes and 15% of the wood it uses; thus far from having any land to spare, the UK is already lacking land to provide for its ecological demands. However, we will skip that omission for the present analysis. One explanation for the omission that would be offered by the authors is that by reducing our meat intake we would need less land for food, but it would need to be demonstrated that the land freed up would be sufficient to provide food, fibres and the proposed amount of biomass.
Next in Table 1 under the heading “Electrical output from uncontrollables,” the first line is the vital one – and the one which illustrates a huge difference from ZCB2030. It gives the figure for the above mentioned maximum 43% addition to the controllables as about 1.98 kilowatt hours per day per person (kWh /d /p), rather than the surmised contribution of ‘renewable uncontrollables’ (wind, wave, tidal stream and solar PV) suggested by the ZCB2030 plan, which amounts to 33.5 kWh /d /p (that is as much as 87% of the 38.5 kWh /d /p total electricity planned by ZCB2030 – far above the 30% limit).
Table 1 shows nuclear as an uncontrollable. This is not strictly true, as nuclear power stations can accept some variation in their output albeit at an economic penalty. However, as mentioned, we have been optimistic about the controllability of hydro, CHP, and fixed tidal, so being somewhat pessimistic about nuclear merely balances things out a bit.
When low grade heat is required for heating, it is better to use the electricity to drive heat pumps rather than use its electrical energy directly. Heat pumps vary in their Coefficient of Efficiency (COP), but the average is probably in the region of COP 4 (MacKay, 2008, p154), thus I use COP 4, which means that the heat provided will be four times the energy in the electricity itself. Following the plans in ZCB2030 closely (producing 6.93 kWh /d /p from the heat pumps, as shown, instead of 6.76 kWh /d /p), it is assumed that 25% of the electricity is used to provide heat via heat pumps. Note that the electricity used to produce heat must not be counted twice, so appropriate adjustments are made in Table 1.
After this subtraction, the amount of electricity available for other purposes is 5.2 kWh /d /p, while the total heat output from CHP, heat pumps, and solar hot water is 10.4 kWh /d /p. Adding the two we get a total of 15.6 kWh /d /p.
We have always argued (based chiefly on Vaclav Smil’s careful analysis) that it would be possible to maintain a civilized life on 48 kWh /d /p, even though that is far below the current total consumption of energy in the UK, which in terms of primary energy is 125 kWh /d /p. But much is lost in extraction of energy and in transformation to electricity, and the total energy supplied to consumers is 88 kWh /d /p. In order to get up to this minimum 48 kWh /d /p, using the energy shown in Table 1, it would be necessary to reduce the population from its current 60 million to 22 million (Table 1, 3rd line from the end). With such a reduced population there is a fair chance that we could feed ourselves, provide ourselves with fibres such as timber, and have sufficient land to spare to grow the biomass ZCB2030 plans for.
The last line of Table 1 shows that realistically only 18% of the electricity that CAT plans for is likely to be produced by the system That concludes the main analysis of this Chapter 8, but let us deal with some queries that may have arisen in the reader’s mind.
Even though the CAT plans indicate that transport is to be run on electricity, there will be a need for some liquid fuel, for instance when using tractors and harvesters. When dipping into another chapter I noted that ZCB2030 recognizes the need for some liquid fuels, but unless yet more biomass is to be grown on land that is not to spare, that would mean using some of the electricity to produce liquid fuels, and in the process of doing that more energy is lost, so the availability of even the above 15.6 kWh /d /p as a final supply is optimistic.
Notes to Table 1
a. Referencing MacLeay (2007), MacKay (2008, p56) writes that, in 2006, large-scale hydro produced 3515 GWh (0.14 kWh /d /p) from plant with 137 GW capacity, thus giving a capacity factor of 24%, and that small-scale hydro produced 212 GWh (0.01 kWh /d /p) from a capacity of 153 MW, giving a capacity factor of 16%.
b. The calculation was done using a biomass yield of 9 t/ha/y, of calorific value 18 GJ/t, giving power density of 5.1 kW/ha, and assuming that the electricity is produced with an efficiency of 30%.
It might be thought that although wind could only add 43% to the controllables, PV, for instance, could add to that 43%. There is some truth in that, but the extent is very small. The reason it could contribute some more is that the peak output of PV comes at midday, and although electricity demand is low at summer weekends, it is unlikely to be as low as in the middle of the night. Let us say, for example, that at such times demand drops to only 86% of average demand. Then while wind can be fitted in to the extent that its peaks reach 66% of electricity demand, PV can be deployed to fill the gap between 66% and 86%. However, the contribution would be tiny, because in the UK the capacity factor of PV is about 10%. Thus if a PV system is added, so that its peaks fill in the newly available 20% of electricity demand, its actual contribution to the electricity supply is one tenth of that, namely 2%. The mistake is often made by renewable energy enthusiasts of assuming that with lots of different sources of uncontrollable, there will be great benefits because the supply will flatten itself out; they conveniently ignore the possibility that the different sources might equally well experience their peaks at around the same time.
It is clear that the fundamental difference between this analysis and the ZCB2030 proposal is about the role of uncontrollables. My analysis, and that cited by Lenzen and the one made by Oswald are all theoretical analyses (although Oswald makes use of comprehensive data about wind speeds over the whole of Britain). It would be reassuring to back up those theoretical analyses with empirical examples. The conditions needed to provide good examples are twofold: (1) A specified area that is able to produce more electricity from wind turbines than it can practically make use of within its electrical grid; (2) the internal transmission lines are sufficiently good to be able to distribute the wind turbine electricity all over the specified area. With respect to item (1) Denmark is ideal, as the amount of wind turbine electricity produced over the year sometimes reaches about 25% of its consumption. However it fails on item (2), since its internal transmission system, taking electricity generated on the windy west coast to the densely populated east coast, is said to be inadequate. It is this lack of transmission capacity which is at least partly responsible for the fact that the amount of wind turbine electricity that is used directly in Denmark is only about 9% of its electricity consumption. The rest has to be exported to other nations, which it is relatively easy to do as there are good connectors to the Scandinavian countries to the north, whose electricity supply is largely from hydro (which can be turned off to allow wind to replace the supply). But other countries do not enjoy such advantages, and this 9% is at least suggestive of difficulties setting in quite early.
Further evidence comes from China. A report by the Worldwatch Institute, dated 28 February 2011, about wind turbines in China said, “Chinese industrial experts have warned that wind power should not exceed 10 percent of local grid capacity to avoid the risk of a grid collapse.” This again suggests that difficulties can arise at a lower wind penetration than the 20% cited by Lenzen. But not too much can be read into China’s experience as there may be special reasons for difficulties, such as poor flexibility of their controllable systems. At present the empirical evidence about the limits of wind penetration is weak, but the theoretical case is strong.
Conclusion. As with nearly all proposals by renewable energy enthusiasts, the major fault in the CAT project is to fail to take account of the limited possible contribution of uncontrollables to an electrical system. An additional fault is to fail to take account of the fact that Britain only manages to grow about 70% of the food it consumes and 15% of the fibres consumed in the form of wood and wood products, so there is no ecologically productive land available to grow significant amounts of biomass.
Oswald, J. 2007. The twenty-one page report 25 GW of Distributed Wind on the UK Electricity System. is available for download on the internet: http://www.ref.org.uk/images/pdfs/ref.wind.smoothing.08.12.06.pdf
Lenzen, M.. 2009. Current state of development of electricity-generating technologies – a literature review. Integrated Life Cycle Analysis. Dept. of Physics, University of Sydney.
MacKay, D.J.C. 2008. Sustainable Energy – without the hot air. UIT Cambridge. 372 pp. £20. ISBN 978-0-9544529-3-3 Available free online from www.withouthotair.com
Moriarty D. and Honnery D. 2010. Rise and Fall of the Carbon Civilisation. Springer. 2011. Hardcover (ISBN: 978-1-84996-482-1, Oct. 11, 2010) US$116; GBP90; Aus$216.
OPTJ 10/1. 2010. Optimum Population Trust Journal, Vol. 10, No 1, April 2010. Manchester (U.K.): Optimum Population Trust. 28 pp. Archived on the web at http://tinyurl.com/optjournal
Appendix A. Over sizing uncontrollables – an overview
Looking in a more general way at the numbers given in The Problem with Uncontrollables (OPTJ 10/1, pp21-24), the following possibilities and problems of overloading a system with uncontrollables become apparent.
To make an analysis with varying electrical demand is very difficult, but renewable energy enthusiasts often argue that demand can be flattened by such things as recharging batteries during off-peak hours, and being able to control demand by temporarily turning off non-essential electrical machinery. It seems improbable that completely flat demand could be achieved, but in order to simplify the analysis we will assume it can.
We will take as a first option the installation of wind turbines such that their capacity exceeds electrical demand by 25%. With turbines that achieve a capacity factor of 30%, this would mean that about (0.30 x 1.25) = 37.5% of electrical demand would be satisfied by the wind turbines. If we assume that, as in Germany, output from wind turbines as an entire group reaches no more than 80% of full capacity, then no electricity would be lost.
As a second option, let us assume that wind turbines are installed with a capacity that exceeds demand by 150%. The losses then become significant. Although the wind turbines now satisfy 66% of the electricity demand (a 76% increase over 37.5%), this has been achieved by doubling the installed wind capacity.
There are further introduced costs which are likely to prove more significant. First, with the wind turbines now satisfying 66% of demand, the controllable power plant – that has to be available to take over from the wind turbines during a lull – will have to reduce its capacity factor to 100 – 66 = 34%. It will be expensive to have plant and operators remaining available for the purpose of satisfying only the 34% of electrical demand that cannot be satisfied by the wind turbines. The consequence is that although this electricity may not be as expensive as peak electricity is today, it will certainly be expensive.
Perhaps a more terminal problem is that with wind satisfying 66% of electrical demand, whenever there is a rapid decrease in wind, a large amount of plant will have to be brought on line over a short period of time. This means that the plant that satisfies the remaining 34% of demand will have to be largely rapid-response plant, which means that it will be significantly less efficient than it might be. The very high efficiency of some gas turbines is achieved mainly by using combined cycle turbines, and these are not amenable to being brought rapidly on line. Indeed it may be impossible to achieve the required flexibility using biomass powered plant: in February 2011 it was reported by the Worldwatch Institute that “Chinese industrial experts have warned that wind power should not exceed 10 percent of local grid capacity to avoid the risk of a grid collapse.” The Chinese controllable plant would be fuelled by coal, and biomass would behave similarly.
There is also a problem with the storage of biomass to deal with say a ten day period when there is little wind. To have available natural gas or even coal stocks to deal with that would not be much of a problem, but to store the amount of biomass needed to deal with such eventualities may prove to be an insuperable problem. Biomass needs to be kept fairly dry if it is to burn efficiently.
The above is not a rigorous analysis, and indeed it is probably impossible to achieve that as a theoretical exercise. The important point is that those who are trying to cajole us into a sense of security about future energy supplies need to do so on the basis of what is probable, not what just might be; because the quality of life – and in many cases the lives – of billions of people will depend on having sufficient energy once fossil fuels become scarce. If it is uncertain that we will have sufficient energy, then there is one action which will definitely alleviate the scale of disaster should it occur, that is to have a smaller population. But that is a solution which always seems to take a back seat !