Future Electricity Generation in Belgium,
the Nuclear Phase Out
and its GHG Consequences
William D’haeseleer
University of Leuven Energy Institute, K.U.Leuven,
Celestijnenlaan 300 A, B-3001 Leuven
Reprint from article published in
Physicalia Magazine, 25 (2003), pp 63-75
- Introduction
Because of the threat of anthropogenic climate change [1], there is considerable pressure on society to reduce its CO2 emissions so as to mitigate the expected temperature changes. Although it is clear to most experts that the challenges ahead are ‘considerable’, and that all reasonable means in all sectors of society should be used to reduce the CO2 emissions, it is strange to notice the different signals from different countries regarding the role of nuclear power.
Especially under political pressure of the green movement, some countries have opted for a nuclear phase out. Germany and Belgium are prime examples. Sweden has recently ‘adjusted’ (say, softened) its phase out according to the German model, in which a maximal total number of TWh’s for the whole nuclear generating system is imposed, leaving it up to the operators to choose which plants they close first. On the other hand, the parliament of Finland has given permission to go ahead with the construction of a new nuclear power plant, and the Swiss population has decided by formal referendum in May 2003 that they want to stick with nuclear power; both a proposed phase out and a moratorium for the construction of new plants were turned down. Also in the USA, there are signals after the California ‘electricity debacle’ that there will be a need for expansion of nuclear capacity.
The European Green Paper on Energy [2], that projects a scaring energy dependence for Europe, and the quite unstable geo-political situation, should encourage us to follow a prudent energy policy based on the use of all reasonable options, according to the “and-and” principle, rather than to work with exclusives in the style of “either-or”.
Regrettably, in recent years, the nuclear issue in Belgium has become part of political negotiation tactics, whereby political opportunism and lack of objective knowledge, or a deliberate falsification of the data by some circles, lead to the nuclear phase-out decision. This decision, if executed, will cost the future Belgian economy a considerable amount of money, with its consequences for the welfare of all, but especially of the economically weaker classes. It is therefore time to present here some rational elements such as security of supply and energetic-technical, economic, safety-related and ecological arguments.
It should be clear that the author of this paper accepts that a democratically elected parliament has the last word on all issues of society, including electricity generation. However, rather than intervening in the implementation of electricity generation, the authorities should provide the framework, set some (perhaps strict) guidelines and define the boundary conditions, which subsequently should be taken into account by the ‘market’. On the other hand, as will be clear below, the author does not agree with the phase-out decision (although he will have to undergo it) and therefore aims to point out the consequences of this ‘unjustified’ decision. Regardless of politics, the nuclear phase-out decision in Belgium was a grave mistake.
- The Belgian Electricity System
Presently, about 58% of the Belgian electricity generation is of nuclear origin, 27% is based on gas while about 12% comes from coal-fired power plants. The renewable contribution amounts to less than 1%.
Future expansion of the generation system will mostly consist of a combination of gas fired combined cycle (CC) plants, consisting of a cascade of a gas turbine and a steam turbine, and of combined heat and power (CHP) units. With an efficiency of up to 60%, a very low investment cost and short construction time, the CC plants are currently considered as the prime choice for expansion of the centralized plant system. Because they can give rise to a considerable primary energy saving, of the order of 10-30 % with respect to the separate generation of heat and electric power, CHP units (of the gas-turbine or gas-engine type) represent the main choice for decentralized expansion investments. Because of the cheap coal prices, and because a properly functioning electricity system needs a certain fraction of storable fuel, coal-fired plants should not be ruled out. However, their CO2-emission per kWh of electricity is about the double of that produced by a CC, and therefore, with the Kyoto protocol[1] in mind, investors are somewhat hesitant to order that type of plants. From the year 2010 on, however, it may be necessary to invest in coal-fired plants anyway, which might potentially lead to a conflict with the post-Kyoto requirements.
Investment in renewable energy-conversion technologies is to a large extent socially driven. Because of their generally accepted environmentally friendliness, but notwithstanding their relatively high cost, they are considerably subsidized, which gives rise to a noticeable and well-publicized investment volume, at least in terms of number of installations. As will be dealt with below, however, the realistically feasible potential for Belgium (certainly with a horizon of 2020) is very limited.
There are no short-term plans to build new nuclear power plants. On the contrary, early 2003, a federal law was formally approved by the parliament which requires that the Belgian nuclear power plants be phased out after forty years of operation, and that no new nuclear plants for electricity generation be built. According to the phase-out law, the units Doel 1 & 2 and Tihange 1 are to be shut down in 2015, the units Doel 3 and Tihange 2 are supposed to be taken of the grid in 2023, while the last units, Doel 4 and Tihange 3 will have to stop producing electricity in 2025.
To give some framework for future reference and comparison, the following numbers may be helpful. The total installed electric power capacity amounts to about 16 GW. The peak load occurring in wintertime is of the order of 12 to 13 GW, while the minimal demand taking place in the summer vacation period, is somewhat like 7 GW. The total electricity demand in 2002 was roughly 84 TWh (or 109 kWh) [3]. Although it is never easy to predict the future, the roughly expected electricity demand by the year 2020 may be very likely of the order of 100 TWh.
- The Belgian Electricity System and CO2 Emissions
Currently, the total annual amount of greenhouse-gas emissions in Belgium is of the order of 150 Mton. Of that, about 125 Mton is due to CO2, and 115 Mton is due to combustion-related CO2. The Belgian electricity sector emits about 20 Mton of CO2, which amounts to less than 20% of the CO2 due to combustion. Compared to the European average of about 30%, the Belgian electricity sector is doing quite well. This observation is confirmed by Table 1, which shows the average CO2 emissions per kWh produced in several countries.
Belgium307 g/kWhe
France 56 g/kWhe
Sweden 42 g/kWhe
Norway 5 g/kWhe
Germany588 g/kWhe
The Netherlands603 g/kWhe
UK521 g/kWhe
Spain471 g/kWhe
Denmark791 g/kWhe
Italy521 g/kWhe
European Union399 g/kWhe
USA610 g/kWhe
Japan350 g/kWhe
World (1994)544 g/kWhe
Table 1. Overview of the specific CO2 emissions due to electricity generation in a variety of countries (g CO2/kWhe)
The low specific CO2 emission due to electricity generation in Belgium is related to the considerable amount of nuclear generation. Figure 1 shows that the correlation between the nuclear electricity production and the CO2 emissions is unmistakable.
Figure 1. Belgian evolution of the annual CO2 emission due to electricity generation and the fraction of the annual nuclear electricity production over the period 1980 – 1997.
A further interesting figure on the Belgian CO2 emissions related to electricity generation is presented in Figure 2.
Figure 2. Real and fictitious evolution of CO2 emissions, depending of the presence of nuclear power generation: what would have been the emissions if we had never had nuclear power generation.
The bottom curve of Figure 2 shows the actual evolution of CO2 emissions due to electricity generation in Belgium. The other three curves are fictitious in the sense that they portray the CO2 emissions that would have occurred if we had never had any nuclear power plants (NPP) in Belgium. The top curve gives the emissions if, instead of NPP’s, we would have built coal fired plants. The third curve from above shows the same in case of only gas-fired plants, and the middle curve of the three shows the CO2 emissions in case we had instead invested in a mix of gas and coal that is proportional to their historical fraction in the mix. The figure shows that, if one were to shut down the NPP’s and replace them by gas-fired units, the CO2 emissions would increase by 16.5 Mton.
- The Lifetime of a Nuclear Power Plant
First it should be made clear that there is no such thing as a predetermined life of a nuclear power plant. A component has a certain lifetime, but a system does not. Indeed, every component is in principle replaceable. As prime examples, one could think of commercial airplanes that are more than ten years old or of the tourist class of racing cars (even the new ones), in which a large number of components has been replaced. As long as the NPP system can operate according to strict safety-technical standards, there is no problem and it can continue to operate. At some point in time, however, there would be a need to replace so many parts and components to keep up the safety level, that the operator may decide that it is no longer economically justified to replace them. At this moment, which is thus effectively an economical decision (but subject to the boundary condition of strict safety rules), the power plant will be shut down. The lifetime of 40 years that the nuclear phase-out law [4] imposes has no scientific-technical, and even economical basis. This type of lifetime is merely a political lifetime. A more detailed account of the issue of the lifetime of an NPP is given in [5].
- The Nuclear Phase Out and the AMPERE Commission
5.1 The Real Conclusions of the AMPERE Commission
A remarkable observation is that the Belgian government 1999-2003, from the time of its conception in July 1999, has totally ignored the then ongoing activities of the AMPERE Commission.(AMPERE stands for “Analysis of the Means for the Production of Electricity and a Reorientation of the Energy vectors”). The full report of this Commission is available at [6]. This Commission, consisting of 16 members, of which 13 Belgian university professors with energy as their expertise, was appointed in April 1999 by a previous government with the aim to analyze the possible electricity-generation means with a horizon 2020-2030. Disregarding the ongoing activities of the ampere Commission, and thus before one had a global view on the future needs for electricity generation, the coalition agreement already stipulated that Belgium would go ahead with a nuclear-phase out scenario. Politically, the nuclear phase-out law is nothing more than the formal ‘translation’ of this coalition agreement and the inaugural declaration of government. Scientifically and socio-economically, it is appropriate to question the rationale of this phase out.
In the mean time, and long before the coalition agreement was translated into law, the AMPERE Commission issued its final report in October of 2000. Its most important conclusions were subsequently confirmed by an international review panel, which was appointed by the government then in office. [7] Although the “Conclusions and Recommendations” of the AMPERE Commission [6] have been written in some kind of ‘diplomatic language’, whereby one has to read somewhat in between the lines, the document is quite clear that the nuclear option should be kept open for the future. Furthermore, those readers that make the effort to read the “synthesis report” of AMPERE, or even better the complete report, will come to the conclusion that even with a very opportunistic energy policy concerning the use of renewable sources (with even up to 1500 MWe of installed wind power) and Combined Heat and Power (with an extra of 1000 MWe), it will be almost impossible to satisfy the Belgian Kyoto commitment (at least as far as the electricity sector is concerned, and assuming that that sector also should decrease its emissions by 7.5%). Subsequently satisfying the expected post-Kyoto reductions will already be very challenging with the nuclear power plants present; with a nuclear phase out it will almost certainly be impossible.
Unless…. Unless one opts for dramatic electricity rationing with heavy governmental intervention, not unlike earlier plan economies, with strict obligations or bans on the choice for energy-conversion technologies (on the demand side as well as on the supply side). Unless one opts for a massive crash program in R&D for CO2 capture in exhaust gases and subsequent CO2 storage, such that an increase of fossil fuels (whereby the abundantly available, and hence cheap coal will become a serious candidate) is compatible with the constraints of post-Kyoto. Unless, one simply ignores the expected post-Kyoto requirements, or tries to “cheat”. Unless one pushes through the nuclear phase out and risks the security of supply of the country and one is ready to pay a considerable economic price.
5.2 Post-AMPERE MARKAL Simulations
It is interesting to look at the so-called post-AMPERE MARKAL simulations. [8] This analysis investigates what is the ‘economical optimal’ electricity mix by the year 2030, given a normal evolution of the fuel prices. As one of the scenario’s, a nuclear phase out, as stipulated in the Belgian law, and a reasonable post-Kyoto reduction of -15%, which implies a linear extrapolation of the current Kyoto reduction[2], has been run. The simulations have been performed with the MARKAL computer code. This code starts from a technology data base, consisting of all kinds of energy-conversion technologies, both on the supply side (including renewable sources) as on the demand side (including very efficient technologies), and tries to find the most economically efficient investment strategy under the given boundary conditions. Sometimes, MARKAL will decide to invest in more energy-efficient demand side technologies in order to save primary energy, at other moments along its time history between 2000 and 2030, MARKAL decides to invest in technologies on the supply side, e.g., to increase the electric generation capacity. It is important to note that MARKAL optimizes the whole Belgian energy sector and not only the electricity sector. The results of these simulations, reported in [8], are not very “favorable”. Under the imposed conditions of a simultaneous nuclear phase out and abiding by the post-Kyoto requirements, Belgium would rely for 85% on gas for its electricity generation. Of course, in itself there is little wrong with gas as a primary source, but putting all your eggs in the same basket has never been a good guideline. In case of (probably to be expected) gas-price hikes, Belgium would find itself in a very vulnerable position, with “heavy” socio-economic consequences. In addition, the price tag of such a scenario, even in the absence of gas-price fluctuations and thus with a “normal” evolution of the fuel prices, is substantial: the combined burden of a nuclear phase out and post-Kyoto CO2 reductions, would cost the country a bit more than 3% of the Gross National Product of the year 2000. Other investment strategies, which deviate from the optimal MARKAL mix, will be even more expensive. For completeness, we mention that a nuclear phase-out scenario does not cause any dramatic economic hardship in the absence of Kyoto or post-Kyoto limitations. In that case, MARKAL simply invests in coal-fired electricity generation plants. (See Ref. [8].)
From a realistic perspective (as was done by the AMPERE Commission), it follows that Belgium cannot afford to phase out its nuclear electricity generation if one takes the GHG issue seriously. Indeed, at the present time about 55-60% of our electricity is generated by nuclear means.
5.3 Non-Nuclear Means to Reduce Greenhouse Gasses
Because of our geographical location and the meteorological circumstances, there is in Belgium hardly any renewable potential to speak of. De “renewable experts” of the AMPERE Commission estimate that the maximal technical potential by the year 2020 would be about 8-10% of the electricity generation at that time. Of this technical potential, about 4-5% is considered as “realistic”. This estimate assumes about 500 MWe of wind turbines on land and about 1000 MWe off shore. Whether these off-shore wind farms will actually be realized and be successful remains to be seen. Presently, there are several attempts to go ahead with off-shore farms in the North Sea and elsewhere, but the somewhat hasty drive towards realization of massive farms (although it does not seem straightforward to get environmental or construction permits) risks to compromise the expected performance, because of corrosion issues, difficulty for maintenance and the resistance again the often very violent North Sea storms. Furthermore, the farther off shore one goes, the more expensive the foundations and the electric connection become. This off-shore route is certainly to be considered in the long run, but a somewhat more cautious demonstration-like attitude in subsequent phases would perhaps be more wise. One should avoid at any cost that the investors become disappointed with the results of the projects; because once the investment market has turned its back to such projects, it will be difficult to turn the tide.