The Australian Labor Party intends taking to the federal election a promise to reduce Australia’s emissions by 45 per cent—well above the target Australia adopted in the Paris Agreement. As one means of reaching this target, Labor has promised to ensure that 50 per cent of Australia’s generation will come from renewables by 2030. This is nonsense on stilts—and very expensive nonsense at that.
A recent analysis by Brian Fisher, former head of the Australian Bureau of Agricultural and Resource Economics, showed just how costly this policy would be: a cumulative cost of $472 billion to 2030, compared with $69 billion for the Coalition’s 26-to-28 per cent reduction target.
Labor has shrugged this off, but the analysis was accepted as being in the right ballpark by Warwick McKibbin, probably Australia’s leading academic economist in the area of climate economics, and one with an impressive international reputation. McKibbin stated publicly that it agreed with his own recent analysis. Labor, however, has been using a 2015 report McKibbin produced for DFAT, based on assumptions then-current, to dismiss Fisher’s analysis. (McKibbin in 2015 used a carbon price of $5/tonne, rising to $10 by 2030; Fisher estimated it would reach $62 by 2030; the current EU price is around $42).
McKibbin had shown the Labor target knocked only a further 0.5 per cent (above the cost of Coalition’s policy) off Australia’s GDP. A cumulative cost of 0.5 per cent annually by 2030 amounts to $85 billion over ten years. (Australia’s GDP in 2017 was $1.7 trillion, so 0.5 per cent per year is $85 billion over ten years; more as the economy grows). The actual total cost of Labor’s target would therefore be around $17 billion annually, or $170 billion over ten years, or much more on current emission permit prices.
Fisher’s analysis also resonates with the best international research, informed by experience such as that in Germany, where its Energiewende program since 2000 has led to greatly increased costs for no recent reduction in greenhouse gas emissions. Energiewende has cost billions of euros in subsidies and, having dug an enormous hole, German policy-makers have chosen to dig deeper rather than admit they are not going to strike climate policy paydirt.
The reason why there has been no reduction in greenhouse gas emissions is that the system must be made reliable, and with limited hydro-electric resources (historically, about 3.6 per cent of generation) and batteries only being able to provide voltage and frequency stabilisation, rather than back-up over days, months or years, this reliability has to come by means of (inefficient but flexible) open-cycle gas turbines or by underloading combined-cycle gas turbines or coal-fired thermal generators. These generators typically require ten hours to start up from cold, so they frequently sit fired up, emitting but not generating, or at sub-optimal loads, producing more greenhouse gas emissions per kWh generated. France, which has an extensive nuclear program, has increased its greenhouse gas emissions for this reason: encouraged to install substantial wind capacity, it now needs more gas turbine generation to regulate the system because the nuclear plants that dominate it cannot cope readily with the fluctuations associated with large amounts of wind.
Add to that the costs of transmission. Average German capacity utilisation rates are only around 17 per cent for wind and 8 per cent for solar. Australia has better conditions, but the same problem remains: wind and solar both have low density. Average insolation at the top of the atmosphere, for example, is only 343 watts per square metre, with a lesser amount reaching the surface (depending on cloud cover and particulates), so the land area required is substantial. Renewables are therefore remote from sources of demand, and require transmission lines that can carry 100 per cent of output, but might only average 25 per cent of that load.
There are also transmission losses to consider—around five per cent in Australia, and dependent on length and load. Indeed, the Australian Energy Market Operator in 2018 adjusted the “marginal loss factor” (which reflects transmission losses) for renewables by up to 22 per cent after finding that the contribution of solar and wind to the market was less than expected, and some have been reduced by 20 per cent, so far, in 2019. Renewable generation is low-density, so must be located where land is cheap, usually remote from demand. Moreover, when the wind is blowing and the sun shining everywhere, there are problems with managing congestion.
Some people assume that 100 per cent renewables is possible, but it is not—or at least not at any sensible price (a point made by some leading climate scientists). For the sake of simplifying to give an example, let’s assume a 25 per cent capacity factor for renewables (likely slightly worse than achievable in Australia). A system of 100 per cent renewables then requires capacity four times the average demand to generate average demand. But, of course, this needs storage, and storage is both expensive and inefficient, only 70 to 80 per cent efficient for pumped storage hydro (let’s say 75 per cent), so in actuality even more capacity is required to supply a 100 per cent renewables system backed up by pumped storage. And similar transmission capacity is needed, but it is utilised to deliver energy only up to 20 per cent of the time, including to-and-from storage. (Batteries are perhaps 85 per cent to 87 per cent efficient, but are expensive and far from viable, except for ensuring short-term stability.)
Even a target of 50 per cent renewables has similar problems, and one wonders why Labor thinks this is a sensible policy. More to the point, how has it managed to convince its affiliated trade unions to support this policy? True, unions have begun to support the Adani coal mine, but they seem so far to have accepted the 50 per cent target, which will almost certainly result in the transfer of the aluminium industry offshore, for example. When I was many years ago a member of the Tasmanian ALP Minerals and Energy Policy Committee, trying to develop a sensible energy policy after the Gordon-below-Franklin cancellation, the representatives of the ETU and the FEDFA were strong advocates for the interests of their members. Why the union silence now?
An important factor seems to be the prevalence of poor analysis that is insisting that renewables are cheaper than coal or gas. Last year, I pointed out that such claims by Professor Andrew Blakers and his colleagues at ANU rested on conflating the price renewable generators were bidding into the National Electricity Market with the cost of renewables. The price, of course, reflected the additional income the renewable generators realised from the sale of renewable energy certificates, the value of which themselves was about the cost of electricity from a new ultra-supercritical coal-fired power station ($81/MWh)—the kind that is being built in large numbers in Asia, and which can provide a 25 per cent reduction in carbon dioxide emissions over the existing black coal fleet, and around 40 per cent over brown coal.
Numbers like $50/MWh are frequently tossed around by spruikers of renewables, but this price is acceptable to investors only because they stand to double this income from the sale of renewable energy certificates. Fortunately, we have available some estimates of non-subsidised costs of wind and solar systems in Australia that are regularly updated by the company Lazard. Their most recent estimate (November 2018) is $US43 to 131/MWh for solar, or $A61.92 to 188.64/MWh converted at the most recent estimate for Purchasing Power Parity (PPP) of $A1.44 to $US1. The estimate for wind is $US34 to 73/MWh, or $A48.96 to 105.12. The spruikers of renewables are always promising us that costs will continue to come down, but Lazard’s Levelised Cost of Energy (LCOE) Analysis report warns that “over the past several years the rate of such LCOE declines have started to flatten”.
But, as noted above, income from generation plus sale of renewable energy certificates is only half the story, because this ignores the costs of integration into a reliable electricity system.
Analyses such as those from Blakers and his colleagues rely upon estimates of the LCOE from renewables, but such estimates ignore system costs that can double the cost of renewables. A more accurate estimate of cost—the System Levelised Cost of Energy (SLCOE)—is ignored by Blakers et al in their continuing attempts to convince us that we can have 100 per cent renewables at no net cost, and that the electricity sector alone can meet our economy-wide Paris target, and do so in a few short years.
Remarkably, that is the claim Blakers et al recently made. Extrapolating from a rapid growth in renewables installation over a couple of years, they noted that Australia’s growth in installations was the highest globally and all that was required was for government to get out of the way. This was a remarkable piece of analysis, to suggest that we would achieve a renewables nirvana that would meet all of Australia’s Paris commitments for the economy as a whole (and Labor’s 50 per cent renewables target) by 2024. However, they ignored the possibility that the recent level of investment might have been stimulated by something other than cost: a kind of gold rush in renewables investment to capitalise on the Renewable Energy Target scheme that was nearing its goal. They even acknowledged that “the target has now effectively been met, and new solar and wind farms can no longer expect significant subsidy support”. Renewable energy certificates will continue to be earned until 2030, but their value will be eroded by the addition of new capacity, unless propped up by a tightening of the target.
Blakers et al were immediately criticised, even by those who supported policies to encourage renewables. Their ANU colleague, the economist Frank Jotzo, said it was “a very big assumption that renewables deployment would continue at present rates. And all it is is a straight-line extrapolation from one year’s renewables deployment.” Melbourne University’s Dylan McConnell tweeted that the analysis “seems not only internally inconsistent, but seriously flawed”. Blakers et al and many other analysts simply ignore integration costs, which are substantial. Even at 30 to 40 per cent wind market share, the integration costs are up to 50 per cent of generation costs—€25 to 35/MWh ($A49.45 to 69.23/MWh converted at Purchasing Power Parity). This is the estimate for Germany, which requires less storage back-up thanks to interconnections to other European countries with nuclear and hydro capacity. Lazards estimate the cost of solar plus storage at about 2.7 times the cost of solar alone.
Blakers et al are not alone in ignoring these costs. The analysis performed for Greenpeace by Reputex published in 2018 (which examines the economics of Labor’s 45 per cent target) similarly simply ignores transmission. If Labor has been encouraged by the Reputex analysis, it has been encouraged in its policy on research that simply ignores integration costs.
An important integration cost is the need to provide storage to ensure system reliability. This is less important at lower levels of renewables penetration, because the system can draw on large amounts of dispatchable generation. However, German economist Lion Hirth found that the value of wind power fell rapidly as wind penetration increased from zero to 30 per cent of total electricity consumption; for solar power, similarly low value levels were reached at 15 per cent penetration.
There are four kinds of storage necessary in a system with large amounts of renewables: short-term storage to maintain grid stability (frequency and voltage); daily storage to capture solar energy for when the sun goes down and the wind drops to zero (or is so strong turbines have to be shut down for safety); intraseasonal storage needed to cover intermittency of wind and solar, the output of which can fall to near zero for several days at a time; and interseasonal storage that could store surplus solar-generated electricity in the summer months for use in the depths of winter. (This last is less of a problem in Australia, with a summer peak.)
Batteries can cover short-term storage, but at a considerable cost. Jack Ponton, Emeritus Professor of Engineering at the University of Edinburgh, has estimated the cost of the “world’s largest battery” installed by Tesla in South Australia (a 129 MWh system believed to have cost around $US38 million, which can perform this function for the South Australian system for four minutes), as in excess of $400,000 per megawatt hour. (It is worth noting that Lazard sees the price of batteries possibly increasing because of plant constraints and rising lithium prices.) The costs of stability for 30 to 40 per cent wind penetration in Germany are less than $12; this exceeds the estimate by Blakers at al, who state: “The cost of hourly balancing of the Australian electricity grid is modest: about $5 per megawatt hour for a renewable energy fraction of 50 per cent, rising to $25 per megawatt hour for 100 per cent renewables.” (The source they give for this estimate is a self-reference to an earlier post of theirs on The Conversation.)
Pumped hydroelectricity can provide daily storage at around $60/MWh – bearing in mind that this is a net consumer of electricity—but there is currently no technology that can provide intraseasonal or interseasonal storage. Renewables advocates usually place their faith on interconnection and the hope that the wind will be blowing or the sun shining elsewhere, but Australia has the world’s longest transmission system and this entails losses exacerbated by distance (currently 5 per cent)—not to mention the impact of events such as dust storms on the output of solar installations, both domestic rooftop and grid. Blakers et al place enormous faith in solar and wind output “counter-correlating”, but there are many widespread calm nights, and this does not overcome low capacity factors for each that are not a problem when they are operating at the margins of a system dominated by dispatchable generation.
Blakers et al simply wish most of these issues away, stating: “Stabilising the electricity grid when it has 50–100 per cent renewable energy is straightforward using off-the-shelf techniques that are already widely used in Australia.” For them, these off-the-shelf techniques are storage (pumped hydro and batteries), demand management, and “strong interstate interconnection using high voltage transmission lines to smooth out the effect of local weather”. They don’t cost these techniques and we are being asked to believe that they will come at prices where they will simply walk off the shelves.
At low levels of penetration, renewables can be a useful addition to a modern electricity system—but we must be careful how we evaluate them, because they very quickly escalate the cost of the system as they achieve substantial penetration. The problem confronting Australia is that we have subsidised and regulated our way to higher system costs. As economist Paul Simshauser pointed out five years ago, we have gone from first to last in terms of electricity prices, and we have done so by focusing solely on LCOE of particular generation sources, ignoring what we were doing to the system—a mistake common to the work of both Blakers et al and Reputex. The situation has worsened since then, and Labor is promising to make it even worse, and it cannot simply wave away the Fisher analysis, because these realities tend very much to support it.
We desperately need good policy analysis that focuses on the System LCOE of variable renewable energy, defined as the sum of their LCOE plus integration costs per unit of variable renewable energy generation. It is a measure that seeks to comprise the total economic costs of variable renewable energy. A large component of integration costs has already been felt, but rarely made explicit, in Australia: reduced utilisation of capital embodied in thermal plants, which has not been accounted for in most integration studies.
Labor would head us down a path where costs would increase still further. Our current system is cannibalising the dispatchable generators, and Labor would have us double down on this. It is also discouraging investment in new ultra-supercritical coal-fired plant that can reduce greenhouse gas emissions by 25 per cent over the existing black coal fleet and 40 per cent over the brown coal generators in Victoria. There is an enormous risk in all this: what happens in 2030, when many of the renewables generators will have repaid their capital but no investor will have any appetite for investment in thermal? Indeed, many of the early renewables will be ageing by then; what price will be needed to induce investment with no renewable energy target?
Our non-systems thinking is systematically driving us towards an impoverished energy future.
Aynsley Kellow is Professor Emeritus of Government at the University of Tasmania.
Blakers, Andrew, Matt Stocks, Bin Lu (2019) ‘Australia: the renewable energy superstar.’ http://re100.eng.anu.edu.au/
Heard, B. P., Brook, B. W., Wigley, T. M. L., & Bradshaw, C. J. A. (2017). Burden of proof: ‘A comprehensive review of the feasibility of 100 per cent renewable-electricity systems.’ Renewable and Sustainable Energy Reviews, 76, 1122-1133.
Hirth, L. (2013). ‘The market value of variable renewables: The effect of solar wind power variability on their relative price.’ Energy economics, 38, 218-236.
Hirth, L. (2015). ‘The optimal share of variable renewables: How the variability of wind and solar power affects their welfare-optimal deployment.’ The Energy Journal, 149-184.
Hirth, Lion, Falko Ueckerdt & Ottmar Edenhofer (2015): ‘Integration Costs Revisited – An economic framework of wind and solar variability.’ Renewable Energy 74, 925–939.
Jotzo, Frank ‘Australia is not on track to meet Paris emissions target – not without policy support’. 8 February 2019. https://reneweconomy.com.au/australia-is-not-on-track-to-meet-paris-emissions-target-not-without-policy-support-71969/
Kellow, Aynsley. (2018) ‘Why the Future is not Solar.’ Quadrant. (5 July) https://quadrant.org.au/writer/aynsley-kellow/
Lazard. (2018) ‘Levelized Cost of Energy and Levelized Cost of Storage 2018.’ <https://www.lazard.com/perspective/levelized-cost-of-energy-and-levelized-cost-of-storage-2018/>
Packham, Ben. (2019). ‘Coalition to add $1bn to climate fund.’ Australian. 22 February. https://www.theaustralian.com.au/national-affairs/climate/coalition-to-add-1bn-to-climate-fund/news-story/b8e3ac4065fd31869481ecd69c05d2e9
Parkinson, Giles ‘Australia could be 100 per cent renewables by 2032 at current rate of wind and solar installs’ Renew Economy 8 February 2019. <https://reneweconomy.com.au/australia-could-be-100-renewables-by-2032-at-current-rate-of-wind-and-solar-installs-20759/>
Parkinson, Giles (2019), ‘New solar, wind projects may stall in face of network “bloodbath”.’ Renew Economy March 12, https://reneweconomy.com.au/new-solar-wind-projects-may-stall-in-face-of-network-bloodbath-49402/
Ponton, Jack (2018) ‘Grid-scale storage. Can it solve the intermittency problem?’ London: The Global Warming Policy Foundation.
Reputex (2018) More expensive, more pollution: The impact of the NEG on carbon emissions and power prices. Report commissioned for Greenpeace. 20 July.
Simshauser, P. (2014). ‘From First Place to Last: The National Electricity Market’s Policy‐Induced “Energy Market Death Spiral”.’ Australian Economic Review, 47(4), 540-62.
Ueckerdt, F., Hirth, L., Luderer, G., & Edenhofer, O. (2013). ‘System LCOE: What are the costs of variable renewables?’ Energy, 63, 61-75.