Why the Best Path to a Low-Carbon Future is Not Wind or Solar Power

Figure A. source Economist Sun, wind and drain: Wind and solar power are even more expensive than is commonly thought

Figure B. source Charles Frank The Net Benefits of Low and No-Carbon Electricity Technologies

Figure A and B summarize some of the conclusions of the recent paper by economist and Brookings senior fellow Charles Frank. The paper might not have attracted much attention outside the usually wonkish energy policy circles. But The Economist wrote a full page review which quickly became a lightning-rod for much shouting by pro-renewables activists. There are three levels for you to study the results — in increasing order of difficulty:

  1. Economist: Sun, wind and drain: Wind and solar power are even more expensive than is commonly thought
  2. Brookings blog post by Charles Frank: Why the Best Path to a Low-Carbon Future is Not Wind or Solar Power
  3. Brookings paper by Charles Frank: The Net Benefits of Low and No-Carbon Electricity Technologies [PDF]

The Economist article will not be a favorite with Angela Merkel, as is nicely summarized in the last paragraph:

The implication of Mr Frank’s research is clear: governments should target emissions reductions from any source rather than focus on boosting certain kinds of renewable energy

I've read all 182 tedious comments, which I cannot recommend because the majority are non-referenced complaints from boosters. Approximately none of the Economist commenters had read the Frank paper. So my take is you can skip #1, read #2 for a good introduction, then work your way through #3.

Figure A is a nice graphic produced from Figure B which is the “money table” of the Frank paper. I've included Figure B so you can quickly grasp what the Cost vs Benefit bars mean in the graphic. There's a minor error in the graphic: the Wind cost/benefit bar is missing the mark for “net benefit” which is a negative $25k/MW not zero.

What Figure A and B claim to tell us is that in the USA new combined-cycle gas plants offer the greatest net benefit given a large set of assumptions. Dr. Frank's paper is a model of transparency — every assumption and parameter is referenced and further qualified by end-notes. Even though this is a simplified methodology for estimating net benefits, there are still a heap of assumptions that must be understood in order to assess where the results might be applicable. I'll summarize a few that I think are critical:

  • Net benefits are calculated on the assumption that new generation replaces on average 22 hours/day of coal non-peak generation and 2 hours/day of single-cycle gas peak generation
  • This is USA-centric, based upon EIA 2013 data
  • Therefore relatively very low methane (gas) prices
  • Therefore relatively high insolation, moderately high wind resource

For energy policy wonks I will highlight a few weak spots in the paper:

  • Most important is that Frank's Adjusted Capacity Cost does not fully reflect the negative reliability impact of VRE.
  • I will speculate that Dr. Frank chose to avoid the complexity of Capacity Credit to keep the presentation accessible. (Capacity Credit estimates the amount of firm, dispatchable generation that can be replaced by VRE without reducing reliability.)
  • Dr. Frank does not examine how Net Benefits vary with VRE penetration. Detailed modeling shows that increasing VRE has large effects on reliability.
  • Capacity Credit for VRE generation is inversely proportional to penetration. The more wind/solar you build the less marginal value you get.
  • The Frank paper is directed at a future powered by less coal (that's good) but not a zero-carbon future (which we must achieve).
  • If we build a strategy for the goal of Zero Emissions we will still likely build Gas CC in quantity because it is fast to build, relatively cheap and politically acceptable. But looking out a century to achieving Zero will help us focus on ramping up nuclear as fast as feasible and safe. We cannot wait 50 years to get started.

Why do I think the Frank paper is important? This is a serious effort to help policy-makers understand why subsidies supporting wind and solar are such an expensive and inefficient way to reduce carbon emissions. And Dr. Frank illustrates why traditional LCOE analysis overvalues wind and solar. And yes, the headline results are US-centric, but there is a serious effort to support generalizing the results by:

Sensitivity to Carbon Prices: In Tables 9A and 9B, the net benefits for both wind and solar are negative. However, if the carbon price is increased from $50 to $61.87 or above, then the net benefits of wind are positive (as shown in Table 11). Above $185.84, the net benefits of solar are also positive.

My interpretation of that result is that solar costs at least $185/ton CO2 avoided. For a society with finite resources, the cost/ton of CO2 abatement is a rather important number.

Sensitivity to Natural Gas Prices: The results in Tables 9A and 9B are highly sensitive to historically volatile natural gas prices. In the United States, the average annual cost of natural gas to electricity producers reached a high of $9.01 per million Btu in 2008. The average monthly cost reached a low of $2.68 in April 2012 (EIA, November 2013, Table 9.10.). The variation among countries, and the effect on net benefits, is illustrated in Table 12.

Note that nuclear becomes the highest net-benefit policy when gas prices exceed about $9/MBtu. Current UK prices are above that level, which is where US prices were only six years ago.

My bottom line is: this paper is good starting point. Please keep in mind that the true cost of variability for wind and solar is significantly understated, as the value of VRE falls as penetration increases. Still, I appreciate that adding complete VRE analysis would have made this paper much more cumbersome.

Fortunately, there has been some very good work on VRE and System LCOE in the past couple of years. In a future post I will get into the research of Lion Hirth et al and the Potsdam Institute for Climate Impact Research. For the eager here are three good references for in-depth modeling studies of high penetration VRE:

  1. Hirth, Lion, The Optimal Share of Variable Renewables. How the Variability of Wind and Solar Power Affects Their Welfare-Optimal Deployment (November 8, 2013). FEEM Working Paper No. 90.2013. Available at SSRN: http://ssrn.com/abstract=2351754 or http://dx.doi.org/10.2139/ssrn.2351754
  2. Ueckerdt, Falko and Hirth, Lion and Luderer, Gunnar and Edenhofer, Ottmar, System LCOE: What are the Costs of Variable Renewables? (January 14, 2013). Available at SSRN: http://ssrn.com/abstract=2200572 or http://dx.doi.org/10.2139/ssrn.2200572
  3. Hirth, Lion and Ueckerdt, Falko and Edenhofer, Ottmar, Why Wind is Not Coal: On the Economics of Electricity (April 24, 2014). FEEM Working Paper No. 39.2014. Available at SSRN: http://ssrn.com/abstract=2428788 or http://dx.doi.org/10.2139/ssrn.2428788

 

 

Australia’s carbon tax – how much impact on global warming?

Regular Seekerblog readers know that we have been arguing for the efficacy of carbon taxes for a long time. In particular revenue-neutral carbon taxes, harmonized across all the major emitting countries. See e.g., Why revenue-neutral carbon taxes are superior to “cap and trade” schemes. Over these years of advocacy we’ve also learned to respect Roger Pielke’s “Iron Law“. In fact, the first Kindle book bought for our iPad was Pielke’s “The Climate Fix“.

We continue to favour energy policies that include harmonized revenue-neutral carbon taxes. But the Iron Law helps us appreciate that no effective carbon pricing is going to happen in the real world anytime soon — particularly where it matters: the developing countries such as China, India, Brazil. What we need to focus upon is promoting policies that directly accelerate conversion to sustainable low-carbon energy that is “cheaper than coal”. That is happily the focus of Google.org REC initiative. Sadly the “Google guys” are backing only “renewables” and completely ignore the only energy source able to contribute significantly to our goal of zero carbon electricity by 2050. That source is nuclear power, with an emphasis on the next generation of nuclear plants such as the IFR and LFTR.

Australian politicians are all in a lather over the Labour-proposed carbon tax. Their proposal is theoretically a not-bad design of a carbon tax. My view is that it will make no meaningful difference to global emissions — and critically, it distracts attention from what matters, and that is implementing low carbon energy sources, mainly nuclear power for electrical generation. Nothing else is going to get the job done, and the longer Australian politicians fail to adjust the more costly will be the ultimate conversion.

The reason I don’t believe the Australian carbon tax is effective is simple. First, if it ever should actually be enacted, it cannot survive politically because a unilateral economy can’t bear the competitive burden of meaningful carbon taxes (i.e., $25/ton+). Second, even if Australia enacted and sustained a “perfect” unilateral tax, the impact on global emissions is in the noise.

Here’s why: Australia’s CO2 emissions are about the same as Brazil or France, around 1.3% of global emissions (2007 UN Statistics). Number one emitter China was at 22.3% in 2007. On a “business as usual” track, if Australia’s total emissions grew at the same rate as China’s emissions, then by 2050-60 the Australian contribution would be the same proportion, around 1.3%. If Australia’s unilateral carbon tax scheme was a successful as we could hope, the tax effect might incentivize Australian emissions reductions of (optimistically) 25%. So the proposed carbon tax contribution to the global CO2 stock would be around 0.3%.

But Australia will become an even smaller future carbon contributor, because of the relative emissions growth path of China and the other developing countries. The U.S. Energy Information Administration / International Energy Outlook 2010 forecasts primary energy consumption to 2035. Taking the EIA “High Economic Growth” case for 2007 vs 2035, we find for China the emissions growth ratio is 200 quads vs 70 quads = 285%. For Australia and NZ combined the ratio is 10 quads vs. 7.2 quads = 138%. So through 2035 the unilateral Australian carbon tax might contribute 1/3% x 138/285% = about 0.16% to our zero-carbon future. By 2050 I’m guessing that contribution will be halved again by the relative emissions growth rates of Australia vs. the developing economies. So my estimate for Australia’s best-case 2050 carbon-tax contribution is an insignificant 0.10%.

Instead of wasting political energy on the carbon tax, we want to see Australia focus on “cheaper than coal”; i.e., decarbonization of energy supply. There are dozens of articles on Seekerblog on this topic. For more background, try “How to Get Climate Policy Back on Course“. And please enjoy Bill Gates’ excellent 2010 TED presentation – we thought he explained the key ideas very clearly in Bill Gates wishes for more than “cheaper than coal.

The best single source for science-based energy policy is BraveNewClimate.com. In particular, I recommend a review of Scenarios for 2060.

In the supporting essays Barry Brook estimates that we need to build a daily average of about 680 MWe of new generation to achieve the low-carbon capacity required to meet 10 TWe of global demand in the 2050 timeframe. Achieving this goal gets harder and more costly every day that politicians squander the future of our children.

For reference: This 2010 Energy paper by Martin Nicholsona, Tom Bieglerb and Barry W. Brook is probably the best up to date source of levelized cost of electricity (LCOE), LCA emission intensity and the relative impacts of carbon pricing on baseload energy options.

Why revenue-neutral carbon taxes are superior to "cap and trade" schemes

To stabilize carbon emissions we need to choose effective policies to incent investors and consumers to make low-carbon choices immediately — not twenty years in the future. The most critical investment decisions are those of the power generation utilities — who are rapidly constructing inexpensive coal-fired plants around the world. To stop that impending train wreck the electric utility boards of directors need certainty of the future costs of carbon emissions. I.e., we need to establish a “price on carbon” that produces immediate results.

Fortunately, the requirements for efficient taxation and regulatory policies have been thoroughly studied — we do not need to create new “Bi-partisan Commissions” to recommend structural options. The following summarizes some of the relevant research demonstrating that a revenue-neutral carbon tax scheme is far superior to the politically popular “cap and trade” schemes such as Kyoto or the European “Emissions Trading Scheme” (ETS).

A revenue-neutral carbon tax scheme can be implemented with nearly zero administrative cost — existing tax collection and auditing channels are almost all that is required to implement a new tax schedule. The price mechanism allows each investor to confidently plan for her future cost of emissions. The carbon tax option gives us certainty of future costs, transparency, and low transaction costs associated with the carbon pricing.

The opposite of certain is what we get with the quantity mechanism of cap-and-trade. The investor faces a volatile, uncertain profile of future emissions costs. The ETS experience has demonstrated great volatility. Volatility is the enemy of what we urgently need — that is fast decisions to stop building dirty coal power generation, substituting the larger capital investments required for low- or zero- carbon plants.

The cap and trade scheme creates a new administrative monster that will be impossible to kill off once it gets going. Envision an apparatus ten times bigger than the Dept of Agriculture. Do not forget that cap and trade requires measuring and auditing reported quantities of emissions. Cap and trade will create a rich growth medium for rent seeking — which leads inevitably to corruption.

In “Bootleggers, Baptists, and Global Warming” Bruce Yandle looks at the post-Kyoto negotiations in the light of the “bootleggers and Baptists” theory of regulation. Yandle points out that in the South, Sunday closing laws make it illegal to sell alcohol on Sunday. These laws are maintained by an inadvertent coalition of bootleggers and Baptists. The Baptists (and other religious denominations) provide the public outcry against liquor on Sunday, while the bootleggers (who actually sell liquor on Sunday) quietly persuade legislatures and town councils to maintain the closing laws. In this paper, Yandle explains that something similar is happening with the treaty negotiations over climate change. Baptists are the environmental groups, and bootleggers are the companies, trade associations, and nations that are seeking favors through the global warming negotiations.

In the previous post Life After Kyoto: Alternative Approaches to Global Warming Policies I reviewed the 2005 Nordhaus study of the same name — which I think very compactly demonstrates the real-world advantages of harmonized carbon taxes vs. cap-and-trade.

Non-economists will probably always be uncomfortable with using indirect instruments like prices, just as patients may wonder how little yellow pills can cure their disease. Nonetheless, the fact that prices are more indirect than quantity restraints should not prevent us from recognizing their superior power as a coordinator and motivator for global warming.

For an in-depth exposition of the superiority of the carbon tax strategy, see The Challenge of Global Warming- Economic Models and Environmental Policy by Yale’s William Nordhaus. You can read a clear discussion of the above issues by searching the PDF for this section The Many Advantages of Carbon Taxes [A. Prices versus Quantities for Global Public Goods].

Nordhaus has probably done the best work demonstrating the many efficiencies of price schemes over quantity schemes. As just one example, a key requirement of any scheme is that it be adaptive. In Nature, 8 May 2008: among the 4 letters responding to the Nature PWG commentary is this letter from Richels, Tol and Yohe, which makes the point on the adaptive requirement better than I have:

In their Commentary ‘Dangerous assumptions’ (Nature 452, 531-531; 2008), Pielke et al. show that the 2000 Special Report on Emissions Scenarios (SRES) reflects unrealistic progress on both the supply and demand sides of the energy sector. These unduly optimistic baselines cause serious underestimation of the costs of policy-induced mitigation required to achieve a given stabilization level.

This is well known among experts but perhaps not to the public, which may explain why some politicians overstate the impact of their (plans for) climate policy, and why others argue incorrectly that ‘available’ off-the-shelf technologies can reduce emissions at very little or no cost.

The numbers presented by Pielke et al. are revealing, but they divert attention from a more serious problem underlying the SRES approach to calculating mitigation costs: a failure to incorporate the dynamic nature of the decision problem into climate-policy analysis. Until we can keep adjusting the analysis by continually incorporating uncertainty, correction and learning, we shall continue to offer policy-makers an incomplete guide to decision-making.

The focus of policy analysis should not be on what to do over the next 100 years, but on what to do today in the face of many important long-term uncertainties. The minute details of any particular scenario for 2100 are then not that important. This can be achieved through an iterative risk management approach in which uncertain long-term goals are used to develop short-term emission targets. As new information arises, emission scenarios, long-term goals and short-term targets are adjusted as necessary. Analyses would be conducted periodically (every 5-10 years), making it easier to distinguish autonomous trends from policy-induced developments — a major concern of Pielke and colleagues. If actual emissions are carefully monitored and analysed, the true efficacy and costs of past policies would be revealed and estimates of the impact of future policy interventions would be less uncertain.

Such an approach would incorporate recent actions by developed and developing countries. In an ‘act then learn’ framework, climate policy is altered in response to how businesses change their behavior in reaction to existing climate policies and in anticipation of future ones. This differs from SRES-like analyses, which ignore the dynamic nature of the decision process and opportunities for mid-course corrections as they compare scenarios without policy with global, century-long plans.

For a recent study of the costly, messy world of trading permits and offsets see the April 2008 working paper [PDF] by reliable sources Michael Wara and David Victor. Excerpt:

This article reviews the actual experience in the world’s largest offset market–the Kyoto Protocol Clean Development Mechanism (CDM)–and finds an urgent need for reform. Well- designed offsets markets can play a role in engaging developing countries and encouraging sound investment in low-cost strategies for controlling emissions. However, in practice, much of the current CDM market does not reflect actual reductions in emissions, and that trend is poised to get worse. Nor are CDM-like offsets likely to be effective cost control mechanisms.

This is excellent work — and compelling results. I hope that all the politicians pushing “son of Kyoto” deals will read it and think carefully about what they are proposing.

The complexity of cap-and-trade vs. the simplicity of revenue-neutral carbon taxes

The previous post on the ecology of tree planting illustrates the mess that is about to be created by the cap-and-trade schemes. These require vast administrative bureaucracies to determine, measure, monitor both sides of every emissions “trade”. And it isn’t simple to determine whether an activity being sold as an “offset” is in fact. It may simply be scientific rubbish, such as tree planting above 20° latitude. Or as common today, it is an activity that was going to be undertaken anyway, but now is enormously profitable due to scalping an “offset fee” in addition to the already favorable project economics. I.e., tree planting.

A revenue-neutral carbon tax scheme has none of these problems, needing nearly zero administration — no complex determinations. And the taxes can be collected by already existing channels, such as corporate income tax.

Cap-and-trade schemes have already proven to be huge generators of new rent seeking special interest groups. Every single type of “emission offset” will gather around it a host of parties profiting from the activity and lobbying for more and more of it. Again none of this happens with carbon taxes.