Bubbles in the computation of carbon emission caps
April 30th, 2007 by Yves Smeers, University of LouvainThe Directive 2003/87 that establishes the European emission trading scheme (EU-ETS) is in many respects remarkable. It tackles one of the most serious problems of our time. It shows that Europeans can still pass critical pieces of legislation on difficult matters without endless discussions. Last, it introduces an economically sound policy … up to some point. Its implementation indeed reveals some bizarre practice of the Commission in the review of national allocation of CO2 allowances.
The CO2 emission market covers the whole of Europe but the number of total allowances and their allocation to individual installations are decided nationally in the so-called National Allocation Plans (NAPs). The Commission assesses the NAPs against twelve criteria listed in Annex III of the Directive. For instance, Austria proposed to allocate 32.8 million allowances but the Commission reduced this cap by 6.4%.
As of today, the Commission published 19 decisions on the NAP proposals of the second phase of the EU-ETS. The Commission finds a violation of criterion 3 of Annex III of the Directive 2003/87/EC in 14 of them. According to Criterion 3 “Quantities of allowances to be allocated shall be consistent with the potential, including the technological potential, of activities covered by the scheme to reduce emission”. The Commission interprets the “potential … to reduce emission” in terms of an “allowed annual average total quantity from 2008 to 2012” (hereafter, allowed annual allocation) and finds an excess with respect to this allowed annual allocation in these 14 NAPs. We argue that the computation of the allowed annual allocation and hence the implementation of criterion 3 is flawed.
The allowed annual allocation is computed as a product of three factors. The “verified emissions” in 2005 is the starting point. It is multiplied by “GDP growth between 2005 and 2010”. A multiplication by “Carbon intensity improvement with additional improvement of 2.5%” comes next. We argue that these factors are uncertain but treated as certain. This misrepresents the “potential…to reduce emission”. The Commission therefore does not prove that the allowed annual allocation is a good proxy of the “potential … to reduce emission” of criterion 3. We successively consider the three factors of the computation.
As all forecasts, the “GDP growth between 2005 and 2010” is uncertain. The April 2007 increase of German GDP forecast for 2007 and 2008 from 1.7% to 2.4% by the five leading economic institutes illustrates this uncertainty. Note also that uncertainty remain high even for very short term forecast. Selecting the most recent forecast, as done by the Commission is thus no remedy.
“Verified emissions” are facts, not forecasts and hence should not be affected by uncertainty. The Commission refers to 2005 emissions and argues as follows. 2005 is the first year for which emissions have been “independently and consistently verified”. It is a “normal” year since the Commission “does not have sufficient indications that a clear majority of exceptional circumstances manifestly pointed in one direction in 2005 and that therefore 2005 verified emission figures cannot be regarded as representative”. 2005 is thus a good reference since it is “normal” and “only the long term average can constitute a valid reference point”. Is this reasoning correct?
2005 emissions were independently and consistently verified. What does this imply? First was 2005 really a normal year? Gas prices were certainly not normal in 2005 (at least compared to the “long-term average”), with important consequences on coal and gas substitution and hence on emissions. Was it a normal hydro year? The Commission says so for Sweden but without quoting its sources. Sweden claims the opposite. Contradictions about facts suggest consulting open sources such as Nordel’s annual reports (Nordel is the association for electricity coordination in the Nordic countries). They reveal that Swedish hydro production was high in 2005 and exceptionally high if it is to be used as a benchmark for a five year period This implies that coal generation and hence CO2 reference emissions could have been exceptionally low. Does this mean that the Commission should have searched another “long-term average ” reference. In fact the long-term average may not matter as we explain.
Suppose we were not talking about emissions but about temperature. Suppose that 2005 was the first year for which average temperature would have been “independently and consistently verified” and suppose that 2005 was a normal year perfectly aligned with the long-run average of average annual temperature. Would this imply that 2005 average annual temperature could be used to foresee the average annual temperature 2008-2012? No! It could only be so if the average annual temperature does not exhibit too much variation. Transpose the reasoning to gas prices and hydro generation. Gas prices exhibit quite a lot of variation and this is important for a thermal electricity system. Nordel annual reports show that Swedish hydro generation is also quite variable, which is important for a hydro/thermal electricity system. Specifically the standard deviation (a common measure of variability) of hydro generation over a five-year period (that is after taking into account that different factors “generally balance each other”) is roughly equivalent (at 1 ton of CO2/1 Mwh) to the reduction of emissions imposed on Sweden! In fact, averages do not matter if there is a lot of variability. But long-term average may not matter, even with low variability, if successive years continue to be recorded as among the hottest ever observed. The past long-term trend becomes irrelevant if the short-term trend is changing. What happens when one takes exceptionally low emissions as reference and there is a high variability in emissions? One underestimates the “allowed emissions” and overestimates the “potential … for reducing emissions”.
Consider now the “Carbon intensity improvement with additional improvement of 2.5%”. As GDP, this is a forecast and hence by nature uncertain. But we have much less experience forecasting carbon intensity improvement than GDP growth. Carbon intensity can drastically vary through years: it dropped from 2.3% in 1995-2000 to 0.5% in 2000-2005 for EUR15 according to European statistics. The analysis of the power sector confirms this variability: changes of gas prices can have dramatic effects on the operation of coal and gas plants and hence on emissions. And gas price evolution is typically uncertain. Even so, the Commission forecast carbon intensity improvement with six-digit accuracy! The first remark is that this is not serious. Nobody would trust forecasts of German GDP growth of 2.4543%. One only wants to know with a sufficient confidence that growth will be 2.4% and not 1.7%. The remaining 0.0543% is spurious accuracy. But neglect this and consider the forecast of” Carbon intensity improvement with additional improvement of 2.5%”. It combines two terms namely the 2.5% improvement and a carbon intensity improvement computed by the model PRIMES. Those reading the decisions will quickly realize that the “additional improvement of 2.5%” comes from the air. The PRIMES component looks more serious. The Commission uses PRIMES extensively in different reports but usually notes that it “does not accept responsibility for any use made thereof (the results of these reports)”. Other public and private bodies also rely on analysis from PRIMES. Still common wisdom does not mean universal truth, especially when most of the documented information suggests that we do not know the truth (in this case how to assess carbon intensity improvement with the precision required by the modifications of the NAPs). The Commission refers the reader to a web site for further details on PRIMES. But this site only contains general information (compare with the US Energy Information Administration site on NEMS) and is essentially useless for those interested in understanding how the Commission arrives at its carbon intensity improvement and eventually at the amount of allowances that they will receive.
To sum up, the Climate Change Committee installed by the Directive “called on the Commission to assess all national allocations plans on a consistent, coherent and robust basis”. It also “stressed the crucial importance of transparent and credible baseline data and projected emission”. Replacing uncertain factors by a forecast or a long-term average is not robust. Relying on opaque modelling is not “transparent”.
Mitigating climate change in the non-cooperative environment of Kyoto will be costly but should not be unduly costly. What counts in global warming is the accumulation of greenhouse gas. Setting fixed targets by 5 years periods is unnecessarily restrictive if it is an accumulation process over 20 to 50 years that causes the damage. Setting these targets on the basis of murky computation adds to the damage. The author is not in a position to propose an alternative to the Commission but he can at least suggest something: look at the long-term gas contracts. They deal with huge quantities over horizons of 20 to 30 years. They do not try to present as certain what is inherently uncertain but instead offer contingency measures to adapt to what we do not know well and to some of what we ignore. These contracts built the European gas industry. Could sequences of rigid 5 years contracts have done the same?
N.B. A more extensive discussion of the EU-ETS will soon be published by Institut Français des Relations Internationales (IFRI) www.ifri.org.
April 24th, 2008 at 6:59 pm
The author remarks on the bizarre practice of the Commission in the review of national allocation of CO2 allowances reveals the fact that the NAPs for the first trading period were too complex and not sufficiently transparent. Complexity makes it hard for companies and other market actors to understand a NAP and thereby creates uncertainty. Also, a lack of transparency makes it very difficult for stakeholders to understand and form a view on plans. The first allocation process, for the 2005-2007 periods, yielded many important lessons which have been reflected in the Commission’s December 2005 guidance for the second trading period.
The allowed annual allocation for the existing allocations is computed based on the historic emissions, the expected growth per sector, the efficiency factor and a correction factor. For the determination of their historic emissions existing installations can now choose themselves three years from the reference period 2001 to 2005. This period has been chosen because it is the most recent period about which details are known and that is in principle verifiable before the point at which the allocation plan is sent to the Commission. It is also in accordance with criterion 10 of Annex III to the Directive which stipulates that a member state must take a decision in advance (before the planning period begins) about the absolute quantities of greenhouse gas allowances that will be allocated in total and to the operator of each installation.
Real GDP growth has been less in recent years than in the last half of the 1990s but the recent rate of increase in CO2 emissions has been greater than in the earlier period. The inescapable implication is that the rate in improvement in CO2 intensity slowed noticeably around 2000. These trends are especially pronounced in the EU15, but they are also true for the new East European member states. CO2 intensity is still improving, but the rate of improvement has diminished sufficiently that CO2 emissions are increasing at a faster rate than they were in the late 1990s despite slower growth in real GDP. Growth has been set at 1.7% per year for all sectors for the period from 2006 to 2010, 2010 being the middle year of the planning period 2008–2012. This is a matter of CO2-related growth. It is based on the estimates of ECN.
The measure for the relative energy efficiency for the combustion emissions has been derived from the performance levels within the energy covenants. In this way account is taken of the use of (voluntary) energy-efficient technologies in the past. This gives effect to criterion 3 of Annex III to the Directive. At the same time justice is also done to criterion 7 of Annex III to the Directive in this way.
Environmentalists have criticised carbon intensity as tool for policy makers. Reductions in carbon intensity do not mean a reduction in actual carbon emissions. It has been labelled as “a book keeping trick which allows the administration to do nothing about global warming while unsafe levels of emissions continue to rise.”
To make the sum of the allowances allocated to each installation equal to the available CO2 capacity the same correction has been applied to each installation for the quantity of allowances for all installations. This correction factor will be applied to all the individual installations so that all the installations contribute to the respective MS climate objectives and in this way to lower CO2 production
Its true that the long term gas contracts have built the European gas industry but we have consider the fact in order to face the consequences of drastic global climate change and its adverse ill effects , its really necessary to push the countries with stringent short term goals for long term accumulation damage which will be act as a trend setter in the years ahead in our fight against global warming and climate mitigation. The first period gave us a good insight of various loop holes in our efforts to reduce emissions and various other factors to be considered to ascertain the way the allocations were made.
So , we could say the European Commission’s efforts in bringing together the MS to have a accord in their allocation plans strictly adhering to the Directive will have a positive impact in their emission reduction targets wherein some of MS may profit in their EU ETS, which will enhance healthy competition among the Member States to gain advantage in respecting their own goals which will come down as one of the great success of the European Union in the near future.
June 19th, 2008 at 10:35 am
The environment in which we’re living belongs to all of us, that’s the reason why we have to look after. The article “Bubbles in the computation of carbon emissions caps” follows one critical point concerning the implementation of carbon reductions. It criticises that uncertain factors are treated as certain. The year 2005 is the starting point of an annual allocation of carbon emissions. So the real emissions of the year 2005 were taken to suggest of following years. In my opinion it’s right to mention, that the average of just one year is not a serious point to represent the potential of a certain country to reduce emissions, but everything has to start at one point. I’d like to say, that also long-term averages can variegate in big dimensions. It’s not a question of time how significant a represent average is, it’s a question of the explanatory power of a certain period. Also a long-term period can become irrelevant, when short-term periods change in big dimensions. But in my opinion it’s not necessary to quarrel with the momentousness of durability of studies, the most important thing is, what the countries are able to achieve in future belonging reductions of carbon emissions. All directives and all commitments have to use basic principles. So what is the problem to take the year 2005 as starting point to assess potentials to reduce emissions? The article criticises also, that the gas price was unusually high in the year 2005. But isn’t it a fact of time that prices of raw materials are growing with an uncertain high rate? It is not only the energy sector who is moaning.
Further I’d like to mention the problem that not all countries have to make equal reductions. So every country has the legislation of a certain number of carbon allowances. But there is also the privilege to obtain more certificates. The decision is on national level. Is it just an ambition of an individual country to make reductions? Shouldn’t it just be the aim to reduce emissions in the European Union as one big object? The premises are not the same. Further I can see the problem of small business concerns. They risk becoming less important in competitions of the energy market. Furthermore countries variegate in a big dimension of the mode to produce energy. Also the technological potentials to make reductions are not the same. So is it really equitable? I would not say, that the allowed annual allocation is flawed, like it is mentioned in the article, but I think it is a hard question of comparability between different countries.