C
OUR DES
C
OMPTES
The costs of the
nuclear power sector
January 2012
Notice
Summary
of
the Public Thematic Report
T
his summary is designed to make the Cour des
Comptes report easier to read and use.
The Cour des Comptes is responsible only for the
content of the report.
Replies from government departments and entities
concerned are appended to the report.
Contents
3
Summary
of the Public Thematic Report by the
Cour des Comptes
Introduction
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5
1
Power generating costs spread
over a very long period
. . . . . . . . . . . . . . . . . . . . .
7
2
Expenditure funded by the public sector
. . . . . . .
19
3
Unresolved issues
. . . . . . . . . . . . . . . . . . . . . . . .
23
Introduction
5
Summary
of the Public Thematic Report by the
Cour des Comptes
I
n a letter dated 17 May 2011, the French Prime Minister asked the Cour des
Comptes to prepare, as part of its assignment to assist the French Government (now
defined under Article L.132-4-1 of the Financial Courts Code) a report on “the costs of
the nuclear power sector” to be submitted before 31 January 2012.
The resulting report brings together most of the factual data currently available on the
past, present and future costs of generating nuclear electricity in France. It does not take a
position as to the desirable level of this output or how it should be funded. The report can
be seen as a “database” prepared to inform the public and promote transparency.
For the most part, the study focuses on the running costs of existing nuclear power
plants as shown in operators’ accounts. These costs can be investment expenditure (past),
current expenditure (operating expenses) or future expenditure (dismantling and fuel and
waste management).
The Cour des Comptes wishes to measure these costs “for society”, not just for the oper-
ator. For this reason, it has also assessed expenditure funded by the public sector, which is
not generally taken into account in energy costs, for want of precise figures. For the first
time, an estimate was made of all French research expenditure relating to nuclear power
generation since 1957. The Cour des Comptes also measured security and safety expendi-
ture funded by the public sector in 2010.
Its analysis therefore excluded military nuclear expenditure and electricity transmission
and distribution costs, focusing solely on the cost of generating nuclear electricity, which only
accounts for some 40% of the price paid by the consumer.
The report does not compare the cost of different sources of energy or various energy
mix scenarios. It does not compare costs to tariffs. The Cour des Comptes does not judge
whether the public funds concerned are well or poorly managed. This is not an evaluation
report.
In order to carry out this study, the Cour des Comptes organized a very broad consul-
tation: a group of experts was set up to help steer this work and hearings were held not
only with managers in the nuclear power sector, but also with representatives of trade
unions and environmental protection NGOs. In addition, a written consultation of French
embassies in key “nuclear power” countries was undertaken.
7
Summary
of the Public Thematic Report by the
Cour des Comptes
Cour des Comptes
1
Power generating costs
spread over a very long
period
Generating nuclear electricity is a
highly capital-intensive activity where
costs are spread over a very long period
of time. For this reason, in order to cal-
culate the total generating cost of exist-
ing nuclear power plants from the oper-
ator’s point of view, past investment
costs, present operating costs and future
costs, which concern both investment
expenditure (for dismantling facilities)
and operating expenses (for spent fuel
and waste management) must all be
added together.
It is generally considered that
AREVA’s costs (investments and operat-
ing expenses, including future costs) are
incorporated in the fuel costs paid by
EDF with regard to AREVA’s involve-
ment in French nuclear power genera-
tion. To avoid accounting for the same
costs twice, calculations are thus based
on EDF accounts only.
Past expenditure is
reasonably well
identified
Heavy initial investment
In all, it cost €121 billion
2010
to
build the facilities required for nuclear
power
generation
(excluding
Superphénix). Of this, €96 billion
2010
represents the actual construction cost
of the 58 existing reactors. This amount
includes an “overnight”
(2)
cost of
€83
billion
2010
, which represents the
initial investments made between 1973
and 2002, plus interest during construc-
tion. The latter is due to the fact that it
takes several years to build nuclear
power plants and was estimated at
€13 billion
2010
by the Cour des
Comptes.
These 58 reactors have a combined
installed capacity of 62,510 MW.
_
__________
(2)The “overnight” cost represents the initial construction cost (€72.9 billion
2010
) plus
engineering expenses (€6.9 billion
2010
) plus pre-operating costs (€3.4 billion
2010
).
Power generating costs spread
over a very long period
8
Summary
of the Public Thematic Report by the Cour des Comptes
The construction cost per
megawatt (MW) has risen
over time
The initial construction cost, includ-
ing engineering (€79.751 million
2010
),
expressed in terms of reactor power,
has
risen
over
time,
from
€1.07 million
2010
/MW
in
1978
(Fessenheim) to €2.06 million
2010
in
2000 (Chooz 1 and 2) or €1.37 mil-
lion
2010
in 2002 (Civaux), with an aver-
age of €1.25 million
2010/
MW
for all 58
reactors. These rising costs can be large-
ly explained by safety baselines, which
have become increasingly stringent over
time.
As the total final cost of the EPR is
not yet known, no precise comparison
can be made. Nevertheless, the Cour des
Comptes observes that the construction
cost per megawatt continues to rise with
this new generation of reactors which
must meet extremely strict safety
requirements, right from the construc-
tion stage. With an estimated construc-
tion
cost
of
€6 billion
for
the
Flamanville EPR (a first of a kind) and
1,630 MW capacity, the cost per MW is
€3.7 million. With an estimated cost of
€5 billion for the standard EPR, the cost
per MW works out at €3.1 million.
Clearly defined
current operating
expenses
EDF’s annual operating expenses
amounted to €8.9 billion
2010
for
407.9 TWh power output in 2010.
These expenses are clearly identified and
there is no particular difficulty involved
in their calculation.
Operating expenses in current €
In €million
2010
Trend
2008-2010
Nuclear fuel
2 135
+ 5%
Staff costs
2 676
+ 5%
External charges
2 095
+ 19%
Taxes and duties
1 176
+ 15%
Central functions
872
+ 30%
Total operating expenses
8 954
+ 11%
Source: Cour des Comptes
Power generating costs spread
over a very long period
9
Summary
of the Public Thematic Report by the
Cour des Comptes
Operating expenses represented a
cost of €22/MWh in 2010. They rose
by 11% between 2008 and 2010 (in cur-
rent €), largely as a result of increased
routine maintenance programmes and
taxes and duties.
Staff costs look set to rise over the
coming years as the recommendations
of the Autorité de sûreté nucléaire
(ASN,
French Nuclear Safety Authority
) fol-
lowing the Fukushima disaster are
implemented (with, in particular, the
creation of a nuclear rapid response
force) and because of the need to renew
the work force while maintaining opera-
tor skill levels.
Future costs are
uncertain by their
very nature
Precise estimates of dis-
mantling costs cannot be
obtained for want of gen-
uinely comparable nation-
al or international experi-
ence
Dismantling costs, i.e. the “end-of-
life” costs of nuclear power plants, are
currently estimated at €18.4 billion
2010
in terms of gross expenditure for the 58
existing reactors.
Dismantling costs are calculated on
an arbitrary historical cost basis,
although the results are corroborated by
much more elaborate methods. The
technical parameters of these methods
must nevertheless be validated by
experts from outside the company.
Current estimated costs should not
be taken at face value as experience in
the field, whether for EDF with its first-
generation plants, or the Commissariat à
l’énergie atomique et énergies alterna-
tives (CEA,
French Alternative Energies and
Atomic Energy Commission
) or AREVA,
shows that estimates have a general ten-
dency to rise as operations take shape.
Furthermore, international comparisons
produce results that are generally higher
than EDF estimations. Nevertheless,
the results of such international com-
parisons vary widely, thus highlighting
the uncertainty in this area.
The Cour des Comptes has
therefore made two recommenda-
tions regarding dismantling costs:
it wishes EDF to use the
Dampierre 2009 method as a basis for
calculating its dismantling provisions,
rather than the historical method which
does not enable sufficiently close moni-
toring of changes in provisions;
it confirms the need to have tech-
nical audits performed by outside
experts and firms to validate the techni-
cal parameters of the Dampierre 2009
method, as planned by the Direction
générale de l’énergie et du climat
(DGEC,
French General Directorate for
Energy and Climate
).
Power generating costs spread
over a very long period
10
Summary
of the Public Thematic Report by the Cour des Comptes
The estimated cost of
long-term radioactive
waste management is yet
to be stabilized
EDF’s gross expenditure relating to
long-term management of waste from
nuclear power facilities currently totals
€23 billion but is not yet stabilized.
Since 2005, the share attributable to the
management of high-level and interme-
diate-level long-lived waste has been
based on an estimate (€16.5 billion
2010
)
of ANDRA’s deep geological repository
project, but this was subject to an in-
depth review in 2009. The new figure of
€36 billion
2010
is twice that of the initial
estimate and is disputed by producers.
The official estimation must be deter-
mined by a French ministerial order by
2015, based on which EDF, AREVA
and the CEA may have to adjust their
provisions.
As there is no management solution
currently available for recycling the
quantities of spent MOX and ERU
(enriched recycled uranium) produced
by nuclear power plants, EDF calculates
its provisions for the long-term manage-
ment of this material as if it were waste
destined for the deep geological reposi-
tory under the same conditions as high-
level waste (HLW) and intermediate-
level long-lived waste (ILW-LL). This
method is consistent with the account-
ing conservatism principle but calls for a
well “calibrated” provision, which is not
the case today. In addition to the costing
aspect, it would be more reassuring if
this assumption could be properly
demonstrated and, in the long term,
possibly
developed,
should
the
Generation IV programme run into
difficulties.
The Cour des Comptes has
therefore made two recommenda-
tions regarding long-term waste
management:
the new estimate regarding the
cost of the deep geological repository
should be prepared as rapidly and as
realistically as possible and in accor-
dance with the decisions made by the
ASN, which is the only authority in a
position to give a verdict on the safety
level of this facility;
this new estimate should show the
cost of direct disposal of MOX and
ERU waste produced each year and this
assumption should be taken into consid-
eration in future design work on the
deep geological repository.
Maintenance investments
set to increase
Maintenance investments are aimed
at guaranteeing reactor performance in
terms of power generation, gradually
enhancing security and safety and,
Power generating costs spread
over a very long period
11
Summary
of the Public Thematic Report by the
Cour des Comptes
where desirable, extending plant service
life.
Maintenance investments have been
on a downward trend since 2000. This
situation has drawn attention to their
importance as the drop has had a nega-
tive impact on the nuclear fleet’s unit
capability factor and, consequently, led
to a fall in power output. The need to
improve safety at nuclear power plants is
even greater than before, with the high
standard set by the EPR and in light of
the accident that occurred at Fukushima
in March 2011.
The EDF maintenance investment
programme for 2011-2025, which was
prepared
in
2010,
amounted
to
€50 billion, or an average of €3.3 billion
per year. This is nearly twice the amount
invested in 2010 (€1.7 billion), a figure
that was already up on previous years.
The investments needed to meet ASN
requirements under the complementary
safety assessment initiative following the
Fukushima disaster are currently esti-
mated at around €10 billion, of which
half was already accounted for in the
initial
€50
billion
programme.
Consequently, maintenance investments
should reach a yearly average of
€3.7 billion, for a programme in the
region of €55 billion
2010
covering the
period from 2011 to 2025.
In €billion 2010
Annual amount
Average 2008-2010
€1.5 billion
In 2010
€1.75 billion
Average with a programme of €55 billion between
now and 2025, including the impact of Fukushima
€3.7 billion
Investments and maintenance
Source: Cour des Comptes
Power generating costs spread
over a very long period
12
Summary
of the Public Thematic Report by the Cour des Comptes
Overall generating
costs to increase
A highly significant capital
cost that may give rise to
various estimations
according to objective
Generating electricity from nuclear
power is a long-cycle, capital-intensive
industry in which the cost of capital is a
variable – and one which has a highly
significant impact when calculating the
overall cost involved.
It is impossible to reconstruct the
funding of these nuclear investments
accurately for want of information and
lack of a specific funding method for
EDF activities. Furthermore, it is hard
to determine the economic value of the
existing fleet of nuclear power reactors
in the absence of a second-hand plant
market that is sufficiently “liquid” to cal-
culate the market value of EDF’s histor-
ical fleet. Stock market ratios are inoper-
ative in that there are no purely nuclear
listed operators and reactor fleets vary
from a structural point of view. Lastly, a
discounted cash flow approach would
be handicapped by the considerable
uncertainty regarding the future selling
price of electricity and the number of
years that the historical fleet will remain
in service. For all the above reasons,
capital cost is calculated on the basis of
conventions.
Various approaches are therefore
adopted to calculate capital cost and its
share in the overall generating cost. This
can be explained by the fact that several
parameters depend on the variable being
measured, and consequently the amount
of capital whose cost is to be calculated,
or even the distribution of this cost over
time (constant or degressive annual
cost). These approaches include:
-
full cost accounting for a given year
: this
consists in considering the depreciation
amount as the only factor in measuring
the share of investments and capital in
nuclear power generating costs. This
method does not, however, take into
account the cost of capital, i.e. the
earned rate of return. Another point is
that at the end of the period, total
depreciation can only be used to recon-
struct the amount of capital invested in
the fleet at its initial value, making no
allowance for inflation or changes in
technological and safety baselines con-
cerning nuclear reactors.
-
the Champsaur commission approach
:
this approach is aimed at calculating the
generating cost of the existing French
nuclear reactor fleet over the next 15
years, taking into account the fact that
the fleet was already 75% amortized in
2010, even though it had an average age
of 25 years for a total amortization term
of 40 years. This approach thus seeks to
calculate a tariff at a given moment in
time, taking into consideration the his-
tory of the fleet, focusing in particular
on how it was financed and on past
depreciation; the “rent”, which is sup-
Power generating costs spread
over a very long period
13
Summary
of the Public Thematic Report by the
Cour des Comptes
posed to measure capital cost, is calcu-
lated by applying an earned rate of
return to the net book value of the fleet
as of the calculation date (in this case,
2010).
-
the current economic cost (CEC):
this
method is used to calculate the average
overall generating cost of the nuclear
reactor fleet throughout its service life,
from the operators’ point of view. It
serves mainly to compare the generating
cost of different energy sources. It
measures the cost of return on and
reconstitution of the capital invested
through an economic rent paid in con-
stant annual instalments throughout the
service life of the reactor fleet. This rent
is calculated to reimburse and pay an
investor for his investment based on its
readjusted value at maturity. This
method makes no allowance for the his-
torical conditions under which the con-
struction of the fleet was financed and
gives an indication of what it would cost
to build the same fleet today.
In both the CEC and Champsaur
commission methods, the selected rate
of return on capital has an impact on
results. The service life of the fleet,
however, has relatively little impact on
the calculated amount.
The approaches above do not have
the same objective, so great care must
be taken to ensure that identical calcula-
tion methods are employed when com-
paring the generating costs of different
energy sources.
Costs that vary
significantly according
to the calculation
method used
Although they treat investments and
capital cost differently, the various
methods for calculating nuclear electric-
ity generating costs all take into account
operators’ past, present and future costs,
with future dismantling costs and spent
fuel and waste management costs being
discounted at a nominal discount rate of
5% (2.94% actual rate, without infla-
tion).
Calculations assuming a service life
of 40 years for the existing fleet of 58
reactors and based on the depreciation
and maintenance investment amounts
for 2010, give the following results for
the year 2010, when power output
reached 407.9 TWh:
Power generating costs spread
over a very long period
14
Summary
of the Public Thematic Report by the Cour des Comptes
This generating cost is not the cost
currently calculated in some internation-
al comparisons, such as those made by
the OECD Nuclear Energy Agency, or
compared to that of other energy
sources, such as in the reference costs of
the DGEC. In both the above cases, in
addition to the capital cost, which can
be calculated using still other methods,
the cost is calculated for an investor
arriving on the market today with new
nuclear power plants, such as the EPR
in the case of France. This calculation,
which simulates the fictitious cost of a
fictitious fleet, is very theoretical. At the
present time, the Cour des Comptes
only knows the estimated construction
cost of €6 billion for the Flamanville
EPR, which gives a generating cost of at
least €70 to €90/MWh, and bearing in
mind that this is not the cost of a “stan-
dard” EPR. The Cour des Comptes is
unable to validate these estimated costs
while the site is still under construction.
It is therefore far too early for it to cal-
culate and validate a generating cost for
an EPR fleet.
Results of the various calculations of generating cost per
MWh according to objective
In €
2010
- Cost using the full cost accounting method, which takes into
account the depreciation of the fleet but not the return on
capital
€33.4/MWh
- Cost using the Champsaur commission approach, which takes
into account the depreciation of the fleet and return on the
capital that is not amortized (with the objective of calculating a
tariff)
€33.1/MWh
- Current economic cost (CEC), which does not take into
account fleet depreciation and which pays the capital originally
invested, allowing for inflation (with the objective of calculating
an average generating cost with no historical reference).
€49.5/MWh
Source: Cour des Comptes
Power generating costs spread
over a very long period
15
Summary
of the Public Thematic Report by the
Cour des Comptes
Results are relatively
unaffected by provisions
for future costs
Calculations of future dismantling
and long-term waste management costs,
which will only need to be paid in a few
decades - or centuries in the case of
waste disposal facility monitoring - are
based on assumptions and often involve
significant degrees of uncertainty. It is
therefore important to measure the
impact of the most uncertain cost items
on the overall generating cost.
Simulations based on available
accounting data have been carried out
for 2010 conditions and using the cur-
rent economic cost method as calculat-
ed by the Cour des Comptes, in other
words on a total cost of €20 billion for
a power output of 407.9 TWh. These
simulations only measure the impact on
generating cost.
Impact of variations in the discount rate
As gross charges will not be paid off
until some point in the distant future,
they must be discounted so that they
can be booked in today’s accounts. This
is done by applying a 5% discount rate,
which includes an inflation rate of 2%
(actual rate is 2.94%). This reduces their
overall calculated amount by 48%,
bringing the total amount of gross
charges
for
all
operators
from
€79 billion
2010
to a €38 billion provi-
sion. This discount rate is roughly
equivalent to that used in other
countries.
Table of gross expenditure/provisions
Source: Cour des Comptes
In million € 2010
EDF
AREVA
CEA
Total
Gross expenditure
62.1
10.5
6.8
79.4
Provisions
28.3
5.6
4.5
38.4
Provisions/
gross expenditure
46%
54%
66%
48%
Power generating costs spread
over a very long period
16
Summary
of the Public Thematic Report by the Cour des Comptes
Reducing this rate by 1%, from 5%
to 4%, would induce a 21% increase in
EDF’s provisions (i.e. €6 billion more
than
the
current
provisions
of
€28.3 billion). As a result, the annual
nuclear power generating cost, calculat-
ed using the CEC method, would
increase by €162 million per year, i.e. an
increase of 0.8%.
Conversely, if the discount rate were
raised from 5 % to 6%, the annual cost
would fall by €131 million/year, which
represents a 0.6% drop.
Impact of variations in end-of-cycle costs
Regarding end-of-cycle costs, the
provisions calculated for spent fuel
management seem adequate. The waste
management
provision,
however,
should be reviewed rapidly.
The new
estimate produced by the Agence
nationale pour la gestion des déchets
radioactifs (ANDRA,
French National
Radioactive Waste Management Agency
), is
slightly more than twice the amount cur-
rently used for calculating provisions. It
would be worth measuring the impact
of doubling this provision. A more
accurate estimation of the impact of
spent MOX and enriched recycled ura-
nium waste disposal should also lead to
an increase in the provision.
Based on a simplified simulation and
ANDRA’s latest assumptions for its esti-
mate, the annual generating cost for
nuclear electricity would increase by
€200 million, which represents a 1%
increase in the generating cost per MWh
based on the CEC calculation.
Impact of variations in dismantling costs
Like those of AREVA and the
CEA, EDF’s dismantling costs are regu-
larly calculated and monitored. This
work shows that estimates have a gener-
al tendency to increase over time despite
progress made in methodology, an
increase which can be explained by the
fact that this field is very new and by the
lack of operating feedback. Secondly,
operators periodically make allowance
for these increases in their accounts,
thus reducing any risk of significant
misjudgement.
The simplified simulations carried
out based on current economic cost and
the same discount rate of 5% show that
a 50% increase in the estimate of dis-
mantling costs would cause a €505 mil-
lion increase in the annual generating
cost. This would, however, only repre-
sent a 2.5% increase in the generating
cost per MWh.
The above tests carried out to meas-
ure the impact of variations in a number
of parameters on future charges show
that, while this impact should certainly
not be overlooked, it remains relatively
limited, bearing in mind the assumption
of a 40-year reactor service life consid-
ered for these calculations.
Power generating costs spread
over a very long period
17
Summary
of the Public Thematic Report by the
Cour des Comptes
Significant impact of
changes in maintenance
investments
While an increase in future disman-
tling and waste management costs have
a relatively limited impact, increased
maintenance investments have a much
more significant impact.
Previous calculations are based on
the maintenance investment amount for
2010, namely €1.747 billion. EDF has a
projected investment programme of
some €55 billion for the period 2011-
2025, an amount which appears to
include the investments induced by
ASN work on complementary safety
assessments. This means an average
annual investment of €3.7 billion over
the coming years.
This increase in investment expendi-
ture induces a 10 to 15% increase in the
generating cost per MWh, depending on
the assessment method used, but in all
cases it has a significant impact.
Impact of the €55 billion investment programme between now and 2025
on the generating cost per MWh
Maintenance investments
Full cost
accounting
Champsaur
approach
CEC
Value in 2010
€1,747 million
33.4
33.1
49.5
Average value of the €55 billion
programme
€3.7 billion
38.2
37.9
54.2
Percentage variation
+ 14.5%
+ 14.5%
+ 9.5%
Source: Cour des Comptes
Power generating costs spread
over a very long period
18
Summary
of the Public Thematic Report by the Cour des Comptes
Extending the service life
of nuclear power plants
affects their profitability
The effect of extending the service
life of nuclear power plants cannot be
measured by calculating its impact on
measured costs using the approaches
described above, except for the full cost
accounting approach. Calculations car-
ried out using the other two methods
only take into account the initial invest-
ment value, not reactor service life.
If, however, it is assumed that the
calculated costs are covered by revenue
(prices, tariffs, etc.), it is clear that the
longer the fleet remains in operation,
the greater the return on the initial
investment and the more profitable this
initial investment is for its owner.
Furthermore, extending the service
life of the reactor fleet postpones not
only the payment of future dismantling
costs, thereby reducing the amount of
provisions, but also fleet renewal invest-
ments, which will call for considerable
financial resources, especially since the
new generations of nuclear power
plants cost more to build than their
predecessors.
19
Summary
of the Public Thematic Report by the
Cour des Comptes
Cour des Comptes
2
Expenditure funded by
the public sector
In order to calculate the cost of gen-
erating nuclear electricity “for society”,
then expenditure funded by the public
sector must be added to the cost for the
operator, as it does not appear in electric
utility accounts. Expenditure funded by
the public sector can be divided into two
main categories: a) research and b)
efforts to promote security, safety and
transparency and information to the
public. Five conclusions can be drawn
from the figures calculated by the Cour
des Comptes on these subjects.
In 2010, recurring expen-
diture funded by the pub-
lic setor represented a
limited amount, close to
that of the tax on basic
nuclear installations.
In 2010, expenditure funded by the
public sector totalled an estimated
€644 million (€414 million for public
research and €230 million for securi-
ty/safety/transparency). The analysis by
the Cour des Comptes focuses on how
these costs are determined and does not
set out to judge whether the funds are
adequate or put to effective use.
This expenditure only accounts for
5 to 6% of annual operating expenses
(including provisions for spent fuel and
waste management).
It represents roughly the same
amount as the tax on basic nuclear
installations (INBs), a special tax paid by
plant
operators
(amounting
to
€580 million in 2010). Although it can
be considered that it covers the related
public expenditure, the INB tax and the
charges that preceded it were originally
designed to cover security and safety
spending.
The development of
nuclear energy relies on
heavy research invest-
ment, mostly funded by
the public sector
As part of its work on this report,
the Cour des Comptes conducted a
study of research trends from the mid-
1950s up to the present day. The study
shows that total research expenditure in
the field of nuclear power can be esti-
mated at €55 billion
2010
, or in the region
of €1 billion
2010
per year.
Of this amount, €38 billion
2010
, (or
an average of €690 million
2010
per year)
was funded by the public sector, i.e. 70%
of the total. This is a significantly high-
er percentage than that observed for
Expenditure funded by
the public sector
20
Summary
of the Public Thematic Report by the Cour des Comptes
2010 and, more generally, over the past
ten years, where the figure was only in
the region of 40%.
It was not possible, however, to cal-
culate past expenditure on securi-
ty/safety/transparency, although it is
probable that, unlike publicly funded
research expenditure, spending in this
area has tended to increase slightly over
time with the creation and gradual rein-
forcement of the ASN and the Institut
de radioprotection et de sûreté nucléaire
(IRSN,
French National Institute for
Radiological Protection and Nuclear Safety
)
the organizations that account for most
of the costs in this area.
Even without this information, it
can be considered that a new situation
has arisen with INB tax and expenditure
funded by the public sector coming to
roughly the same amount in 2010. This
situation has been brought about by two
opposite trends: a) the gradual drop in
research expenditure funded by the pub-
lic sector and b) the very significant
increase in tax income, which was multi-
plied by 4.6 between the years 2000 and
2010 (in current euros).
In constant euros, actual income
from the tax amounted to €3.3 billion
over the past decade (2000 to 2010),
while research expenditure funded by
the public sector over the same period
totalled €5.5 billion. Comparison of
these two figures clearly shows that the
previous situation was far less balanced
than in 2010.
The State will have to
fund the CEA’s provisions
As at the end of 2010, the CEA
faced future costs of €6.8 billion
2010
, or
a provision of €4.5 billion
2010
after dis-
count.
This
breaks
down
into
€2.9 billion for dismantling, €1.2 billion
for long-term waste management and
€0.3 billion for spent fuel management.
It is considered that €3.1 billion of
this provision is covered by dedicated
assets consisting for the most part of
receivables from the State or AREVA
securities. It is planned that the CEA
can sell these to the State as needed.
The State therefore directly or indi-
rectly finances these future costs, the
amount of which remains uncertain
based on dependable calculation meth-
ods. This is evidenced by the often con-
siderable readjustments made to esti-
mates of these future costs over the past
ten years. The State will have to cover
these costs when they fall due by means
of budget appropriations.
The Generation IV
programme significantly
increases future research
expenditure funded by the
public sector
The “future nuclear” programme
included in investments for the future
provides
funds
of
€650
million
(between 2011 and 2017) for the
detailed design of ASTRID, a demon-
strator designed for the development of
Expenditure funded by
the public sector
21
Summary
of the Public Thematic Report by the
Cour des Comptes
Generation IV sodium-cooled fast reac-
tors. If detailed design results encourage
France to continue along this path,
other forms of funding (probably pub-
lic for the most part) will need to be
sought to cover the cost of a) building
ASTRID and b) financing other facili-
ties, as the demonstrator will still be far
from reaching industrial maturity. All
these costs are unknown at the present
time.
The State insures some of
the “civil liability” risk in
the event of a nuclear
accident
The nuclear sector is in a very spe-
cial situation as far as insurance is con-
cerned. Although there is very little like-
lihood of a risk becoming reality, the
consequences of a severe accident can
be disastrous. The probability of an
accident and the seriousness of its con-
sequences are hard to estimate and are
the subject of heated debate. It is never-
theless certain that in the event of a
major accident, the upper ceiling for
operator civil liability for nuclear dam-
age, defined by international conven-
tions, would be rapidly reached and
probably exceeded.
According to current arrangements
for civil liability for nuclear damage, the
State could be obliged, in the event of a
nuclear accident, no matter how unlike-
ly, to pay damages above and beyond the
liability ceilings defined in the legislation
currently in force and to bear any eco-
nomic consequences not covered by
compensation mechanisms. Operators
currently have this guarantee free of
charge. The Cour des Comptes has
demonstrated that the cost of this guar-
antee is very low in comparison with
overall nuclear power generating costs.
But costs in the event of a severe acci-
dent could be huge and weigh very
heavily on the State’s resources, bearing
in mind that whatever happens, the
State is ultimately liable for the cost of
repairing nuclear damage and all the
consequences.
The Cour des Comptes has
therefore made two recommenda-
tions on this point:
it recommends that France should
make every effort to implement as rap-
idly as possible the amending protocols
to the Paris and Brussels Conventions,
signed in 2004, as they significantly raise
the ceiling for operator liability, even if
this ceiling remains limited;
it also emphasizes the need for the
provisions of current French substan-
tive law to be rigorously applied, espe-
cially with regard to fixing the financial
guarantee imposed on operators, which
implies applying regulations in full.
23
Summary
of the Public Thematic Report by the
Cour des Comptes
Cour des Comptes
3
Unresolved issues
Over and beyond the uncertainties
identified above, and for which the Cour
des Comptes has attempted to measure
impacts on nuclear power generating
costs, several issues which could have
significant consequences require partic-
ularly close attention.
While costs are important,
the positive and negative
externalities of different
energy sources must not
be overlooked
The impact of nuclear power gener-
ation, particularly on health, the envi-
ronment, the balance of payments and
the economy is generally very difficult, if
not impossible, to measure with cur-
rently available knowledge. These exter-
nalities should be taken into considera-
tion when comparing different sources
of energy.
The Cour des Comptes has never-
theless estimated the (low) cost of CO
2
emissions from nuclear power genera-
tion in 2010. The total cost was €90 mil-
lion for an average cost of €15/teq CO
2
(average cost of CO
2
over the recent
period) and €190 million for a cost of
€32 €/tCO
2
(Quinet report reference).
The Cour des Comptes recom-
mends encouraging and supporting
research and studies on these issues, for
other energy sources as well as nuclear
energy. Although a monetary value can-
not be calculated for many types of
impact - at least not at the moment -
knowledge of impacts is useful for com-
paring different energy sources.
Calculations concerning the
cost of complementary
safety assessments
performed in the wake of
the Fukushima disaster
must be completed and
refined
Following the Fukushima accident,
the ASN initiated at the French
Government’s request an in-depth secu-
rity and safety audit of existing nuclear
facilities. Its report and opinion on “pri-
ority facilities” were published on
3 January 2012. Although they do not as
yet allow a comprehensive and precise
estimation of all the lessons to be
learned from this accident, they do bring
some details to light, taking into account
the fact that EDF, AREVA and the
CEA are not in the same situation.
Unresolved issues
24
Summary
of the Public Thematic Report by the Cour des Comptes
EDF’s situation
EDF is the most concerned by these
issues. Two main types of cost can be
distinguished, considering only the
financial consequences:
- measures to “increase the resist-
ance of facilities to extreme situations”.
EDF currently estimates that in terms
of investments these measures will cost
around ten billion euros over a period of
a few years. Some of this amount is
already included in EDF’s provisional
investment programmes. The costs
induced by these measures will, howev-
er, also need to considered in terms of
staff, especially for setting up a “rapid
response force”. According to EDF, this
aspect represents a significant cost in
the region of €300 million per year;
- social, organizational and human
factors. These costs are even harder to
determine today, though they will have
an impact on staff numbers and the
resulting payroll costs, as well as on the
organization of subcontracting.
AREVA’s situation
AREVA’s facilities vary considerably
and specifications, which were originally
designed for power reactors, must now
be adapted to the specific characteristics
of each facility. As a result, AREVA’s
efforts are still in progress and the com-
pany must continue to work on safety
improvements. By mid-2012, it must
define concrete measures as part of its
interdisciplinary emergency response
studies.
As for EDF, the emphasis is on set-
ting up a hard core for each AREVA
“platform”, together with supplemen-
tary measures for more robust pool fill-
ing. The creation of a rapid response
force, however, seems less adapted to
the smaller number of sites and greater
diversity of activities. A more logical
solution would be to reinforce emer-
gency response organisation on each
“platform”.
AREVA’s investments are set out in
a five-year strategic plan and amount to
€2 billion for the period concerned. At
present, the company seems to consider
that investments relating to complemen-
tary safety assessments should represent
several hundred million euros of addi-
tional costs over the period, although
the Cour des Comptes has no way of
confirming these figures, especially
because ASN recommendations are far
from precise at the moment.
The CEA’s situation
The CEA is in a rather similar situa-
tion to AREVA in terms of the diversi-
ty of its facilities, although this situation
is made a little more exceptional in that
most of its facilities will be examined in
2012 and that three of the five facilities
examined during the first series of
Unresolved issues
25
Summary
of the Public Thematic Report by the
Cour des Comptes
assessments,
namely
Phénix,
the
Plutonium facility and Osiris, have now
been shut down and are currently being
dismantled. For this reason, it must
specified in each case what investments
are needed as risks continue to drop and
dismantling work progresses.
Current CEA estimates concerning
the possible financial impact of comple-
mentary safety assessments cover a rela-
tively
wide
range,
from
€50
to
€500 million over a period of three or
four years.
It is therefore generally too early to
calculate and verify the amount of
investments and the costs in human
terms induced by this first round of
complementary
safety
assessments.
Also, as the ASN mentions in its report,
“feedback from the Fukushima accident
could take some ten years to filter
through. It appeared necessary to assess
the resistance of facilities to extreme sit-
uations as rapidly as possible”, although
this is only the first step in what will be
a long, drawn-out process of analysis
and reflection.
The growing number of
departures from the Act of
2006 and the impact of
the financial crisis on ded-
icated assets management
should lead to a review of
the conditions for imple-
menting this mechanism
Out
of
a
total
provision
of
€27.8 billion for end-of-cycle operations
to be covered by dedicated assets,
€18.2 billion was covered by listed finan-
cial securities, at as 31 December 2010,
€2.7 billion was not supposed to be cov-
ered at that date and €6.9 billion consist-
ed of mutual hedges between operators
in the nuclear sector, including the State.
In all, the State can be considered direct-
ly or indirectly liable for €4.6 billion,
excluding RTE securities (€2.3 billion)
accounted for in EDF’s dedicated assets.
Several decisions have significantly
modified the application of the initial
Act of 2006 relative to Transparency
and Security in the Nuclear Field: the
date by which dedicated assets must
cover provisions in full has been pushed
back from June 2011 to June 2016; the
liquidity of these assets has been
impaired by the acceptance of securities
issued by operators’ subsidiaries and by
“inter-operator” receivables; and, lastly,
the very notion of dedicated assets has
been dropped for the CEA.
Unresolved issues
26
Summary
of the Public Thematic Report by the Cour des Comptes
At the same time, the financial crisis
has increased uncertainties as to the
medium- and long-term profitability of
assets making up the portfolios and,
therefore, raises doubts as to whether
they are sufficient to cover future costs.
The mechanism underwent a num-
ber of changes before the commission
that was supposed to structure its gover-
nance was set up. This is unfortunate.
The CNEF
(3
)
is now operational and in
a position to give an opinion on the
mechanism as it now stands and, if need
be, on how it could be adapted to the
present financial situation.
The Cour des Comptes recom-
mends
re-examining this issue. Each
new difficulty is accompanied by a new
dispensation, radically changing the
structure and initial logic of the mecha-
nism. This situation is unhealthy.
Power plant service life is
a strategic variable that
should be framed by
explicit guidelines
The ASN performs a ten-yearly
inspection to determine whether a
nuclear power plant can be granted a
licence to remain in service and under
what conditions. So far, only two reac-
tors (at the Tricastin and Fessenheim
plants) have been licensed to operate for
40 years, subject to significant work to
improve safety.
In accounting terms, however, EDF
nuclear power plants have been depreci-
ated over 40 years since 2003. Plant
service life has a significant impact on
the actual generating cost by allowing
investments to be depreciated over a
longer period of time. In addition, it
postpones dismantling costs and the
need to invest in new power facilities.
The Cour des Comptes observes
that ten years from now (by the end of
2020), 22 out of 58 reactors will reach
their
fortieth
year
of
operation.
Consequently, assuming that nuclear
reactors will have a 40-year service life
and that nuclear power output will
remain at its present level, eleven EPRs
will have to be built by the end of 2022.
This seems highly unlikely, if not impos-
sible, including for industrial reasons.
This implies one of two things: a)
either it is assumed that plants will oper-
ate for more than 40 years, as the
“multi-year investment programme”
(PPI) for electricity generation for 2009-
2012 would seem to indicate, since it
“favours the central scenario of extend-
ing the service life of nuclear power
___________
(3) National Committee for the Evaluation of Funding for the costs of dismantling
basic nuclear installations and spent fuel and radioactive waste management
Unresolved issues
27
Summary
of the Public Thematic Report by the
Cour des Comptes
plants beyond 40 years”; b) or the ener-
gy mix will move towards other energy
sources. However, no clear public deci-
sion has been made concerning these
major strategic issues, even though they
call for short-term action and major
investments.
Major investments
required in the short or
medium term will have a
significant impact on the
overall generating cost
In
order
to
extend
operating
licences up to 40 years, make allowance
for the consequences of safety assess-
ments
carried
out
following
the
Fukushima accident (ASN report) and
maintain plant unit capability factor at
an acceptable level (between 80 and
85%),
the current rate of maintenance
investments will have to be doubled.
This represents an increase in generat-
ing costs of around 10% in terms of
current economic cost.
In addition, if existing plants were
to be replaced by EPRs, which have a
much higher construction cost (at least
€5 billion for a “standard” EPR) and
assuming that the service life of existing
plants
is
extended
to
50
years,
€55 billion (covering the cost of eleven
EPRs) would need to be invested in the
next twenty years.
The Cour des Comptes notes that,
whatever the solutions found for these
issues in the future, significant invest-
ment expenditure will be required in the
short and medium term for both main-
tenance and new build purposes.
This
will be in addition to distribution grid
and research investments if the authori-
ties decide to move ahead with the
Generation IV reactor development
programme. This should entail a consid-
erably higher level of investment than
that currently observed in this area,
although the exact amounts involved
cannot be calculated as yet.
The strategic and financial impact of
this situation must be analysed and
taken into account in medium-term
energy policy guidelines that are made
public and accessible to all stakeholders
in the nuclear sector. In the field of
energy policy, it takes a long time for the
effects of a decision to be felt. This time
lag is particularly long in the nuclear sec-
tor, although it can be observed in all
other sectors, too, including for energy
saving initiatives. Not making a decision
on this issue is the same as choosing to
extend the service life of existing reac-
tors beyond 40 years.
Unresolved issues
28
Summary
of the Public Thematic Report by the Cour des Comptes
Cost calculations must be
transparent and the data
in this report regularly
updated
Given the complexity of the subject,
uncertainties concerning data and the
many assumptions used in calculating
the figures in the report, this work must
be regularly reviewed and improved as
part of a governance approach that
takes into account the strategic aspects
of the energy issue and the fact that
public opinion is highly sensitive to this
subject.
The Cour des Comptes therefore
recommends that this survey should
be regularly updated, adopting a
transparent and objective approach,
in order to:
gradually specify the assessment
methods required to evaluate decisions
from an economic point of view in the
event of uncertainty. In particular, the
costs and probabilities of accidents
should be investigated in greater depth;
keep track of future trends in the
various cost items analysed, drawing on
available feedback and focusing in par-
ticular on the financial impact of com-
plementary safety assessments following
the Fukushima accident;
capitalize on the contributions of
the different parties involved and
experts in the field.
The significance of externalities, the
impact of which, particularly on the
environment, health, employment and
the trade balance, cannot be calculated
except, perhaps, by comparison with
other solutions, highlights the fact that
costs are certainly not the only variables
to be considered in decisions relating to
nuclear power generation.