A critique of "A Clean Energy Future for Australia":

by M. Diesendorf and H. Sadler.

Ted Trainer.

This report is like many others in reinforcing the belief that increasingly efficient energy technology and use of renewable sources can solve the problems faced by consumer society without radical shift from its basic assumptions, institutions and values. (The study is concerned with "stationary" energy, especially electricity, and not liquid fuel.) In another source Diesendorf says of the report, "…existing renewable energy resources are capable of substituting for coal fired power stations."

I believe this general position is seriously mistaken and that it will not be possible to achieve a sustainable and just society without dramatic structural, lifestyle and value change.
Following is a brief indication of the challenges that can be made to the document.(For a lengthy discussion of the limits of renewable energy see http://socialwork.arts.unsw.edu.au/tsw/D74.RENEWABLE-ENERGY.html)

1. The energy supply target is set too low.

Many pages are devoted to arriving at the conclusion that by 2040 Australian stationary energy demand will be only 25% greater than it is now. This figure is derived by focusing on trends in the "energy intensity" of the Australian economy and its major sectors. "Energy intensity" refers to GDP or industry dollar output value divided by energy used. Following are some problems with this approach.

Many extraneous factors can influence the dollar value in the numerator of the index of an industry, or a country as a whole. These include shifting to producing and consuming more high value items without changing the amount of energy needed to produce each. Industries generally seek to move in this direction over time. Changes in the dollar value of inputs to production, especially energy inputs can also affect EI significantly.

Another important factor reducing apparent energy used in many industries is change in the "fuel mix", for instance change from use of coal to use of electricity which is the most flexible, transportable and convenient energy form, and therefore usually most capable of increasing dollar value of production achieved by a unit of energy. The study acknowledges that this is a major factor reducing energy intensities. (p. 43.) These are one-off changes and are likely to be taking place extensively now that energy costs have become important. Note that electricity use is the energy form undergoing most rapid growth in demand. Therefore changing the fuel mix can increase the value of an industry or national output while actual energy use per unit of output does not improve.

It is problematic to assume that trends evident in the 1994 to 2000 period will continue to 2040. Many major and potentially disruptive changes in the Australian economy are likely to occur in this period, including the peaking of petroleum supply and the impact of climate change which is likely to have significant energy implications. (e g., resort to energy-intensive water desalination.) The main reason for not extrapolating is that we are in the early stages of introducing energy conserving technologies, where the greatest gains can be made. We are "picking the low hanging fruit". Reducing energy use per unit of output is subject to diminishing returns. It is likely that the fall between 1994 and 2000 in energy needed to produce a kg of beef or a car will be significantly greater than that achievable in the decade 2030-2040, long after all the easy energy gains have been achieved.

The discussion makes no reference to the main challenge to the general dematerialisation thesis. Over time the economies of the rich countries have tended to move out of the energy and resource intensive industries, (more accurately their transnational corporations have relocated to "developing" countries.) Their consumption of the products of those industries increases, in the form of the products imported from the Third World, but these changes do not show up in their national energy accounting. The energy used to produce the goods they are importing is recorded as energy consumption in the Third World producing countries, not in the rich countries which consume them. This distorts "energy intensity" indices at any point in time, and trends in them over time. Trainer (2001) found that in US trade figures this tendency to relocate production and then import was greatest for the most resource and energy intensive industries. In other words, to focus on the energy used in a rich countries such as Australia will give a significantly distorted impression (underestimate) of the actual amount of energy going into providing its "living standards" and GDP.

Inview of the above points it would seem to be more satisfactory simply to focus on the existing, observable trends in energy growth, given that these automatically take into account the effect of technical advance in increasing EI, changing fuel mixes and changing prices. If there is good reason to expect that in future the rate of efficiency gains built into this extrapolation will or could change, then this can be factored in. (Dieslendorf and Sadler explore this in their Chapter 6 but see below on the double counting problem.)

ABARE’s expected growth rates to 2050 diminish but more or less average 2% to 2040, meaning energy demand then would be about 110% above present demand, as distinct from the 25% figure arrived at in this study. The CSIRO study Future Dilemmas (Foran and Poldy, Ch. 5.13) aligns with the ABARE figure, estimating that stationary energy use will rise to just under 4000PJ, compared with the 2352PJ stated by Diesendorf and Sadler. It would seem that the ABARE expectations can be taken to indicate that if the anticipated improvement in energy intensity, due to increased technical efficiency and all other factors, did not occur, 2040 energy use would be about 2.8 times the present amount, but if they do it will be about 2 times as big.

It would seem therefore that a more realistic target would be in the region of double the present amount of stationary energy, not a 25% increase…again unless reasons can be advanced for thinking that the rate of decline built into the ABARE expectations is too low.

Is there double counting of energy efficiency gains?

In chapters 4 and 5 the discussion of energy intensity concludes that the 2040 energy demand will be 57% higher than at present, but in the next chapter this is further reduced to the 25% increase quoted above. Chapter 6 argues that in general technical advance could cut the amount of energy required across industries by 20%.

As has been indicated, it would seem that this second step is not valid because it involves a double counting of gains already taken into account at least in part in the discussion of falling "energy intensity" trends, and feeding into the declining energy growth rate ABARE gives. The technical advances discussed in Chapter 6 are the kinds of changes that the presently observable declining trend in energy use involve and will continue to involve. At least it is not shown that these reductions will not be included in those involved in the declining trend ABARE expects.

Renewable energy.

The most challengeable elements in the study concern the contributions renewable energy sources are claimed to be capable of making. The general claim is that these plus use of natural gas can meet demand in 2040 and there is no hint of any problem in renewable energy sources being able to meet demand in subsequent years. Indeed the above quote asserts without qualification that renewables can meet demand.

Except for biomass and to some extent wind renewable technologies are described briefly without considering quantitative limits or deriving conclusions about the quantities they can provide, and the conclusions re contributions to the 2040 energy mix are stated not argued.

Wind; The main problem re wind energy is to do with variability and therefore the fraction of demand it can meet. The study gives an inadequate and misleading analysis of this problem. It claims it is a myth that " Since wind is an intermittent source it cannot replace coal fired power unless it has expensive dedicated long term storage." This is said to have been refuted by theoretical studies carried out 20 years ago.

Obviously wind can replace some coal-fired capacity, but the focal question is how much. The study says "It may well be possible to operate an electricity grid with 40% or more of its energy generated from the wind." "Denmark generates 18% of its electricity from wind." Firstly Denmark’s electrical production is equal to about 18% of its consumption, but the proportion of Denmark’s electricity demand met by wind is only around 5%, and there are significant problems of integration. Denmark has been able to develop wind energy extensively because of its capacity to frequently sell excess energy to neighbouring countries (and buy electricity from them at high prices.) Australia will not be able to do this. The situation in Germany, with the world’s highest commitment to wind energy, seems worse, with wind contributing under 5% of national electricity use and serious integration problems being encountered. In the E.On Netz supply system, Germany’s largest wind company, annual wind capacity was a remarkable 16% in 2003, and 19% in 2004… a long way below the 35 or 40% typically assumed by wind enthusiasts. Denmark is usually regarded as the leading wind nation and is in one of the best wind regions in the world, yet the average infeed capacity of its wind energy sustem in 2003 was 17%. (Sharman, 2005.)
A number of sources indicate that wind begins to cause significant integration problems when it approaches 10-15% of demand, and that the limit even in good regions such as the UK is probably 20%.

The study discusses the capacity to back up wind power sources by gas generators when winds are low, but fails to recognise that wind seems to have a very low "capacity credit"; i.e., it does not eliminate much need for fossil fuel plant.

It is probable that where wind is a significant contributor little fossil fuel capacity could be retired. Wind varies greatly over time and sometimes there are days on end with little or no wind, even in good sites. This means that almost as much back up capacity as wind capacity must always be available. The E.On Netz reports stress the issue of "capacity credit" (as does Sharman re Denmark and the UK.) If we build a lot more windmills we must also build a lot more coal, gas or nuclear power stations to turn to when the winds are down (…and grid extensions, and grid reinforcements to move large surpluses of wind energy around.)

The evidence is that wind creates little "capacity credit", perhaps 20% at best according to the E. On Netz reports, and much less according to others. The Clean Energy Futures report in effect recognises this later by saying that "…to maintain reliability … ( in a 2GW system)…up to 300MW of peak load gas turbines may have to be installed."..fossil fuel equal to half the wind system capacity (optimistically) expected.

Wind advocates often assume that dams can easily provide the necessary storage capacity to smooth out variability in wind generation. World hydro-electric capacity is only about 8% of world electricity use, so dams could not substitute for much wind capacity in a system with more than a small contribution from wind.

This means that, as has been pointed out by various people, wind is not a replacement for coal or nuclear capacity — it is an alternative. If we build a system with a lot of wind capacity we must also build almost as much coal, gas or nuclear capacity that will stand idle much of the time but have to be turned to when winds are down. One major implication is that the usually quoted capital cost of wind power, c $1000/kW is quite misleading. Firstly this is a peak figure and so in a system with a .16 capacity the figure per kW of electricity delivered is $7000. But then we should add the cost of the grid extensions and reinforcements, and of the additional perhaps .8 kW of fossil or nuclear plant that must also be built. So the real all-in capital cost of a wind system will be close to 10 times the figure wind enthusiasts usually state.

Easily overlooked is the need to reinforce existing grids so that when large amounts of surplus wind energy become available in one region they can be transmitted to others.

The variability problem cannot be overcome by improved forecasting. It is due to the fact that sometimes there is little or no wind, whether this is predicted or not. Davey and Coppin (2003, p. 11) found that in Eastern Australia the amount of a wind system’s capacity likely to be "reliably available" (i.e., with 95% certainty) is only a remarkable 4.6%…and this does not apply to the worst season of the year (Autumn.)
The important question regarding wind is therefore not to do with the areas getting good winds and the associated sheer quantity of electricity that could be generated, but the limits set by its integration into the supply system because of its variability… and serious problems are arising in large national systems where infeed is around 5%. Thus we should be quite cautious re the study’s assumption that wind can supply 20% of Australian demand.

Biomass.

The first problem with the study’s biomass energy discussion is the 35% and potentially 40% figures stated for electricity generation efficiency. It seems that the energy cost of biomass production, harvesting and delivery has not been taken into account. The all-inclusive net efficiency of the entire US biomass electricity generation system has been estimated at 22%.

The main problem re biomass is that in a renewable energy future it will be more than fully taken up in supplying liquid fuels and in effect none will be available for electricity generation. (For the reasons why hydrogen is not likely to solve the liquid fuel problem see RE. Ch.6.) The clearest conclusion in the entire renewable energy field is likely to be that biomass cannot come close to meeting present liquid plus gaseous fuel demand, let alone that generated by a world of 9 billion living as Australians do now (multiply demand by 10), let alone if they are all to live as we would by 2070 given our commitment to limitless growth in "living standards" (i.e., a world with 60 times present gross economic output.) (See RE. Ch.5. Liquid fuels.)

To meet present oil plus gas demand (via methanol at a net 5GJ/t; Foran and Mardon, 1999.) Australia would have to harvest 7 tonnes per ha of biomass from 70 million ha . (More recent estimates suggest more like 6.5 - 9GJ/t.)

Given that all our crop land is 20 million ha, all reasonable or better forest 40 million ha, and plantations are around 2 million ha, we are most unlikely to find the additional land capable of yielding anything like the 300-500 million tonnes required p. a.

Note that Diesendorf and Sadler refer to a predicted 4% pa rate of growth of transport fuel demand, which if it kept up to 2035 would mean 4 times present annual demand. Note too that Australia’s capacity for biomass production is far greater than almost all other rich countries, e.g., about twice US crop, pasture and forest land area per capita, and 7 times that of the UK.

In other words the study should not have made any provision for electricity generation from biomass, since in a renewable energy world its top priority use will be for transport fuel, and will fall far short of meeting that demand in Australia, let alone globally.

Cogeneration.

About 65% of the cogenerated electricity in the study’s Scenario 2 for 2040 comes from fossil fuels. In a renewable energy world there can be almost no use of fossil fuels.

The renewable energy goal is much lower than it would have to be.

Scenario 2 involves considerable use of gas, which will probably be becoming quite scarce by 2040, and it involves use of fossil fuels equal to 36% of the present amount.

The IPCC carbon emission analyses show that if we are to stabilise carbon in the atmosphere even at 450 ppm (far higher than at present when alarming effects are already becoming evident) we must almost entirely eliminate use of fossil fuels. The commonly quoted need to reduce our use by 60% is usually taken to mean that rich countries must make such a reduction; but this is incorrect. The figure applies to total world use, so rich world reductions must be to per capita use rates around 5% of present levels. In other words those who believe a consumer-capitalist society with a growth economy can go on must explain how this can be based almost entirely on renewable energy sources. Diesendorf and Sadler do not go one-tenth of the way to showing this; see below.

The study ends before the problems begin.

Something like the study’s "Scenario 2" for stationary energy might be achievable by 2040, but this would have been achieved by drawing down heavily on potential renewables and eliminating their liquid fuel contribution, using far more fossil fuel than is viable in the long term, and depending heavily on gas whose availability is likely to crash about 2040. No doubt is raised re the possibility of going on from 2040 towards a fully renewable future.

A more realistic outlook would recognise that by 2040 petroleum and gas are likely to be very scarce and depleting fast, and that we should be phasing fossil fuels out altogether.

The "Jeavons effect" is ignored.

It well known that technical advance which increases energy efficiency does not necessarily lead to reduced energy use. Sometimes it leads to increased use, as in the case of the US car fleet petrol consumption average in recent years. (Heinberg, 2003, p. 160.). Better technology can reduce costs and this can encourage greater use. This study does not discuss whether the technical efficiency gains they assume are likely to translate into reduced consumption.

Putting it all together

Something like Scenario 2 might be achievable by 2040, in view of its considerable dependence on fossil fuels, its heavy use of biomass, and its focus only on stationary energy. However a more realistic account of a renewable energy future would involve the following elements.

* A stationary energy demand target of around 3700PJ, not 2352PJ.

* No electricity generated from biomass or fossil fuels.

* Wind contribution maybe 10-15%.

* Cogeneration contribution quite small.

* Renewables contribution perhaps 68 PJ hydro, 38PJ solar, (taking their figures) and wind 15% of demand…perhaps a total of 335 PJ, or 9% of possible demand.

Conclusions

There is a very powerful tendency to assume that renewable energy sources can substitute for fossil fuels and enable continuation of capitalist-consumer society with its high "living standards" and economic growth. There has been almost no literature critical of this assumption. It has been almost impossible to get onto the public agenda the possibility that renewable sources can’t save us and that solutions to global problems cannot be achieved unless we abandon some of the basic commitments of our society, notably to high living standards and limitless economic growth. Clean Energy Futures is another "technical fix" argument, reinforcing the comforting belief that we do not have to think about radical change from affluent consumer-capitalist society. This is of course what just about everyone wants to hear.

If on the other hand we should be thinking very seriously about the possibility that unless we abandon consumer-capitalist society and move to far lower "living standards" in a zero-growth economy there will be catastrophic beakdowns…then reports such as A Clean Energy Future are reinforcing precisely the wrong ideas, values and assumptions. Especially important is the impression the study gives that renewable energy sources plus better conservation technology can solve the energy problem. No reference is made to any possibility that consumer-capitalist society needs to be fundamentally questioned.

ABARE www.abareconomics.com/data_services?energy_fig.html?period=13182
Davy, R. and P. Coppin, (2003), South East Australian Wind Power Study, Wind Energy Research Unit, CSIRO, Australia.
Heinberg, R., (2003), The Party’s Over, Gabriola Island, New Society.
Sharman, H., (2005), "The dash for wind; West Denmark’s experience and UK energy aspirations", www.glebemountaingroup.org/Articles/DanishLessons.pdf
Trainer, F. E. (T.), (2001), "The dematerialisation myth", Technology in Society, 23, 505-514.