Fossil or Nuclear Energy: A Comparison of Environmental Costs and Risks. Ari Rabl and Mona Dreicer, Centre d'Energétique, Ecole des Mines, 60 boul. St.-Michel, F-75272 Paris CEDEX 06, e-mail: RABL@CENERG.ENSMP.FR
In recent years several major studies of fuel cycle externalities have been carried out, based on a careful analysis of the impact pathways from emission of a pollutant to final impact and cost. Even though much uncertainty remains, a consensus is beginning to emerge about the dominant costs imposed on the public. The choice between nuclear and fossil power involves difficult considerations, such as the value of life and the right to impose costs on future generations who do not draw any direct benefit from our consumption of energy. In such a situation many people find that a simple comparison of discounted costs and benefits is unsatisfactory, even though it is the only measure that renders all commensurate. It is desirable to have, as much as possible, further criteria, for example a comparison of physical impacts and health risks. The present paper discusses several possible approaches:
No single comparison yields a satisfactory picture, but in
combination they can help clarify the issues. In the present
paper we consider only impacts on the general public (clearly
different from impacts on workers, in terms of internalization by
regulations or work contracts). Also we do not address the issue
of nuclear proliferation.
2.1. Methodology
By contrast to earlier studies, which tended to use a variety of different assumptions, a recent group [Ontario Hydro 1993, ORNL 1994, EC 1995, Curtiss et al 1995, Rowe et al 1995] is based on the methodology of impact pathway analysis, applied with a common set of clearly stated assumptions. One traces the impact pathway for each pollutant or other burden, from source to receptors, and evaluates the damage both in physical and in monetary units. The principal steps are:
The numbers are summed over all receptors that one wants to
include in the analysis. The detailed documentation can be found
in the above references.
2.2. Boundaries of the Assessment
The entire fuel cycles were analyzed, from the production of
the fuel to the disposal of the waste, including the
decommissioning of the power plant. The dispersion of pollutants
has been considered over a range sufficiently large to capture
essentially all of the impacts: global for pollutants with long
atmospheric residence time (such as C02), at least a
thousand km for others (particulates, SOx, NOx,
etc.).
A key assumption in the calculation of the response to
radiation doses is linearity (i.e., there is no threshold or, if
there is one, it is below the natural background). Future
population exposure patterns are assumed unchanged, with a world
population of 10 billion, and impacts are integrated over a time
horizon of 100,000 years. Due to the long half-life of some of
the radionuclides, low-level doses may exist very far into the
future; only 10% of the collective public dose occurs during the
first 100 years.
For the fossil fuel cycles the calculation of health impacts
has also been made with linear dose-response functions (as
suggested by recent epidemiological studies that find no
threshold at typical ambient concentrations). The major long term
effect for coal is global warming, with time scale of centuries.
2.3. Economic Valuation
The global warming cost estimates correspond to a GNP
reduction in the range of 1 to 2% per C02 doubling.
For mortality a value of statistical life of 3.4 M$ (2.6 MECU)
has been assumed [EC 1995c]. For the social discount rate ORNL
[1994] and EC [1995] have chosen 3% as central value, bracketed
by 0% and 10% to assess the sensitivity.
In the present paper we discount at 0% for the following
reasons. Using as criterion the preferences of future
generations, Rabl [1996] has shown that the appropriate discount
rate for inter-generational effects is significantly lower than
the conventional social discount rate, because it should include
only the growth of the economy (the pure time preference
component of the discount rate involves only redistribution
within the current generation and does not create wealth to
compensate future generations). Equally important is the rate at
which future costs will evolve; only the difference between this
rate and the discount rate matters. This difference
("effective discount rate") is likely to be positive
but small. To get a simple upper bound of costs and to facilitate
the conversion to impacts we set it equal to 0.
2.4. Site Dependence
The emission site does not matter for long-lived
globally-dispersing pollutants, such as C02 (global
warming), I129, and C14 (dominant long term
impact of the nuclear cycle). But for pollutants with regional
range there can be significant site dependence [Curtiss et al
1995, Curtiss and Rabl 1996], especially for particulates, NOx
and SOx whose health impacts are proportional to
population density in the region surrounding the power plant
(roughly a thousand km). Therefore we scale the mortality impacts
from coal according to the population density of the countries
where the plants are located.
For the fossil fuel cycles, Table 1, the major costs arise
from global warming and from health impacts (especially mortality
from particulates); depending on the assumptions, they could be
roughly on the order of ten percent of the price of electricity.
For the nuclear fuel cycle, Table 2, the dominant costs are
due to cancers and severe hereditary effects; their magnitude
could be roughly on the order of one percent of the price of
electricity -- although controversy continues about waste,
accident and proliferation. The evaluation of high-level waste
repositories is difficult due to lack of practical experience and
difficulty in predicting the far-off future. The assessment of a
severe accident depends on the assumptions of the probability of
occurrence of an accident, probability of a release, and the
consequences that would be expected to occur after radiation
protection actions were taken. The release fraction was estimated
to be about 1% of the core and the probability of a core melt
accident was taken to be 1.0E-5 per reactor-year, with a
conditional probability of a release occurring after the accident
of 0.19, based on engineering fault tree analysis. The resulting
total regional collective dose is about 58,000 man·Sv (compared
to the 560,000 man·Sv estimated for Chernobyl), and is
multiplied by 1.9E-6 for a risk of 0.11 man·Sv per
reactor·year.
Most probably current nuclear and coal technologies will
either evolve towards lower emissions or be replaced altogether
by cleaner sources (fusion, solar, ...). The full implementation
of cleaner technologies will be reached only gradually, probably
over a time scale on the order of many decades. For a simple
order of magnitude assessment of impacts we therefore assume that
these power plants (= currently best available technology) will
be in use for a century. We also need to estimate the total
amount of electricity that might be produced by nuclear or coal
during this time. Even though the demand for electricity is
growing world wide, general consumption trends of materials and
energy in industrialized countries suggest that saturation is
likely, perhaps not far above current levels of kWh/capita in
industrialized countries.
To get a rough idea of possible scenarios, we list in Table 3
data for population and current electricity production in France,
Japan, the USA, the European Union and the world [OECD 1995]. The
current world production by nuclear power is about 2000 TWh/yr.
This number could decrease somewhat if old plants are not
replaced, but in view of the commitment to nuclear power in
certain countries, in particular France and Japan, it seems that
a production equivalent to the current demand of France and Japan
is a plausible lower bound for the nuclear power output during
the next 100 years (= minimal scenario). For an upper bound one
might consider an output equal to the current electricity demand
of the world (= maximal scenario).
To keep the numbers simple and transparent we show, for each
country or region in Table 3, the public deaths if the entire
production of 100 years at current demand is supplied by nuclear
or by coal. The numbers for coal are comparable to or higher than
those for nuclear. However, the deaths from coal are fairly
immediate and certain, whereas those from nuclear tend to occur
in the distant future and are less certain (e.g. a cure for
cancer may be found). The fraction of nuclear deaths within the
100 yr period is less than 10% of the total.
The dose/yr decreases with time. To estimate how large the
dose from 100 yrs of nuclear power would be compared to the
average natural background, 2.4 mSv/yr, let us assume as rough
approximation that 10% of the total committed dose of the 100
year production occurs during the first 100 yrs. Table 3 shows
that the man made dose represents less than 1% of the natural
background for the minimal and 6.5 % for the maximal scenario.
Based on work supported in part by a grant from the European
Commission, DG XII, under contract JOUL2-CT-92-0236, ExternE
Program. We have benefited from discussions with our colleagues
in the ExternE Program, especially N. Eyre, J. Pellissier-Tanon
and J. Lochard.
References
Curtiss, P. S. and A. Rabl. 1996. "Impacts of Air
Pollution: General Relationships and Site Dependence". To be
published in Atmospheric Environment.
Curtiss, P. S., B. Hernandez, A. Pons, A. Rabl, M. Dreicer,
V. Tort, H. Margerie, G. Landrieu, B. Desaigues and D. Proult.
1995. "Environmental Impacts and Costs: the Nuclear and the
Fossil Fuel Cycles". Centre d'Energétique, Ecole des Mines,
60 boul. St.-Michel, 75272 Paris.
Dreicer, M., V. Tort and P. Manen. 1995. ExternE:
Externalities of Energy. The Nuclear Fuel Cycle. European
Commission, Directorate-General XII, Science Research and
Development.
EC 1995. ExternE: Externalities of Energy. European
Commission, Directorate-General XII, Science Research and
Development. JOULE programme.
OECD 1995. Environmental Data Compendium 1995. OECD, 2
rue André-Pascal, F75775 Paris. Ontario Hydro. 1993. Full
Cost Accounting for Decision Making. Toronto: Ontario Hydro,
December 1993.
ORNL 1994. Reports on the External Costs and Benefits of Fuel
Cycles. Oak Ridge National Laboratory, Oak Ridge, TN 37831.
Rabl, A. 1996. "Discounting of long term costs: what
would future generations prefer us to do?" Ecological
Economics, to be published (1996).
Rowe, R.D., C.M. Lang, L.G. Chestnut, D. Latimer, D. Rae,
S.M. Bernow, and D.White. 1995. The New York Electricity
Externality Study. Oceana Publications, Dobbs Ferry, New
York.