Energy Probe
January 17, 2002

“This EA review is being conducted under a process with outrageous shortcomings . . .”

Energy Probe’s Supplementary Comments on Atomic Energy of Canada Limited’s (AECL’s) Cernavoda Reactor 2
Environmental Assessment Summary

(Note: This version contains minor editorial changes from the original
location on Energy Probe’s web site, but retains the meaning of the
original copy.)
Introduction

Energy Probe has formally endorsed the elaborate and carefully
documented Comments by Non-Government Organizations on Atomic Energy of
Canada Limited’s (AECL) Cernavoda Reactor 2 Environmental Assessment
Summary (henceforth “NGO Comments”), largely prepared by David Martin
and the recently tragically deceased Irene Kock, both of Sierra Club of
Canada’s Nuclear Project. In addition, we hereby submit the following
supplementary comments on this Environmental Assessment (the EA).

The context of this Environmental Assessment, including “other
impacts”: This EA review is being conducted under a process with
outrageous shortcomings – detailed in the NGO Comments and in the
comments from our sister organization Probe International. Briefly,
these include incomplete disclosure, rushed time lines, and total lack
of accountability. Perhaps an even greater concern is the fact that
this EA review is apparently the only opportunity for the citizens and
taxpayers of Canada and Romania to participate in the decision of their
two governments to proceed with this project at taxpayers’ expense and
risk.

Obviously, any assessment of whether or not this publicly funded
project is worthy to proceed must properly include an Environmental
Assessment, but it must also properly include an assessment of the
project’s other impacts ‚Äì both its expected impacts (costs and
benefits) and its possible impacts, or “risks.” These other impacts are
primarily the commercial costs, benefits, and risks that ordinary,
private buyers and sellers routinely consider and weigh in deciding
whether or not to proceed in a business transaction. In the case of
government-agency transactions like the export of AECL’s CANDU reactors
to a foreign country like Romania, the citizens of both countries are
entitled to a full assessment of these costs and benefits and risks,
not just the predicted impact on the local water temperature, etc.
Although this EA review process is a deeply flawed assessment of the
project’s environmental costs, benefits, and risks, it does at least
address some important environmental issues. As an assessment of the
commercial impacts of this project, the EA is completely worthless. The
lack of an open assessment of these other impacts – either in Canada or
in Romania – renders this EA review absurd, especially given the
current unwillingness of private buyers and sellers anywhere in the
world to begin construction of nuclear generating stations, CANDU or
otherwise. That unwillingness suggests prima facie that the
non-environmental impacts of this project are likely strongly negative,
when fully integrated.

Put bluntly, the level of environmental impact that might be acceptable
in return for a least-cost supply of reliable electricity produced at
the risk of willing investors, is one thing; the level that might be
acceptable in return for a publicly subsidized, unreliable source of
electricity (like Pickering-A, Bruce-A, or Point Lepreau, for example)
at the continued risk of taxpayers as unwilling investors, is clearly
another.

It is important to note that the commercial risks and net costs of
CANDU reactors used to be theoretical, but are now painfully real: The
two Canadian utilities with significant nuclear investments relative to
their size – the former Ontario Hydro and New Brunswick Power – were
essentially bankrupted (brought to a state where their liabilities
exceeded their assets) by their CANDU reactor investments, despite
receiving extensive subsidization and monopoly privileges from their
governments. By not expressly addressing the significant likelihood
that Cernavoda-2 will also be a drain on public finances and an
unreliable source of electricity, this EA does not judge the project’s
environmental impacts against the proper standards. And it certainly
provides no rational basis for concluding that this project is worthy
to proceed.

The EA’s inadequate assessment of the need for Cernavoda-2: The needs
assessment for the reactor in the EA documents is a reflection of
classical central planning. There appears to be no consideration of
competition as an alternative means of meeting the electricity needs of
Romanians. There appears to be no consideration of whether the reactor
investment could stand on its own as a private venture – not surprising
given the obvious answer to this question. The EA repeatedly assures
the reader that the electricity the reactor will produce will be
inexpensive – without any supporting figures, assumptions or
calculations. Unfortunately, Canadians have no recourse against AECL
for similar assurances, now demonstrably false, concerning past reactor
projects, and Romanians are presumably no better protected in this
project. Such unaccountable assurances from a usually unreliable source
have no credibility, in our view. It is ironic that Canadian
jurisdictions are generally escaping the very form of central-planning
decision-making that AECL – a Canadian corporation – adopts without
question in this EA.

Past Environmental Assessments of Canadian nuclear projects: The EA
report claims that: “The CANDU 6 technology has undergone environmental
assessments, both in Canada and internationally, starting with the
Canadian panel review of Point Lepreau Unit 1, in New Brunswick in
1977, and including the Chinese environmental assessment of the Qinshan
Phase III Project. Each of these projects was approved to proceed.” But
the environmental assessment of the Qinshan Phase III Project was
secret and not independently adjudicated. The complete Cernavoda-2 EA
documents are secret and the EA will not be subject to independent
adjudication.

It is understandable but unfortunate that the EA document does not
recount the most striking fact about environmental assessment reviews
of CANDU reactors in Canada: All recent public environmental assessment
panel reviews of planned nuclear expansions in Canada contributed to
the sensible and beneficial decisions not to construct those reactors.
The two most recent panel reviews of nuclear power reactors in Canada
are the federal assessment of the subsequently cancelled Point Lepreau
Unit 2 and the provincial (Ontario) environmental assessment of the
then Ontario Hydro’s so-called “Demand/Supply Plan,” which envisaged
the construction of 10 CANDU reactors, all of which were cancelled. It
is also worth noting that the most recent environmental assessment
panel review of any nuclear proposal from the author of this EA – AECL
‚Äì was the federal panel review of AECL’s concept for deep geological
disposal of radioactive waste (“spent fuel”). That review was charged
with deciding whether or not AECL’s concept was “safe” and
“acceptable,” and whether or not it should receive approval to proceed
to the next step – selection of an underground disposal site. After
nine years of deliberation, that independent panel:

‚Ä¢ unanimously concluded that AECL’s concept was not acceptable;

‚Ä¢ could not reach agreement on whether or not AECL’s concept was safe; and

• unanimously recommended that approval not be given to AECL to proceed to site selection.

In short, the record of the CANDU reactor and AECL before public,
independent EA panel reviews is at best mixed, and the recent record is
one of consistent rejection, either during the review or afterwards.
This fact sets the context for (a) the inadequacies of this EA and (b)
the Canadian government’s decision to leave the Cernavoda-2 decisions
in the hands of AECL and EDC, rather than submitting them to an
independent panel review.

Cernavoda-2 design flaws

The Cernavoda-2 design repeats many of the flaws of CANDU designs used in Canada. For example:

• Like all Canadian CANDUs, the station will have open loop
service water systems, a design that increases environmental insults
compared to closed loop systems. Radioactive waste getting into the
service water will be collected and discharged into the environment via
the Active Liquid Waste Treatment (ALWT) System. Based on the Canadian
experience, we assume that little or no treatment will be applied to
the waste before discharge, relying on the obsolete belief that “the
solution to pollution is dilution.”

• Similar to the Pickering Station, the cooling water intake
comes from a shipping canal, making the station particularly vulnerable
to marine security threats (as discussed below, security threats are
ignored in the EA).

• The Emergency Water Supply System is shared between C1 and
C2, a safety shortcut that is not permitted by U.S. nuclear regulators.
This oversight is also understandable, since the possibility of
two-unit accidents at Cernavoda 1 and 2 is dismissed in the EA (see
below).

The EA’s treatment of decommissioning

The EA’s claims regarding decommissioning are especially false and
misleading: “Canadian experience indicates that decommissioning nuclear
facilities can be carried out without significant adverse health and
environmental effects. The ICRP 60 occupational dose limit (Section
2.2.4) was not exceeded by any of the 130 workers decommissioning the
Gentilly 1 NPP over a two-year period, and the average dose was 0.35
mSv¬∑a¬π.” Nowhere in the documents does AECL admit that G1 was a failed
prototype reactor that operated for no more than a few hundred
full-power hours and therefore contained only a tiny fraction of the
radioactive contamination that would be experienced if C2 operates for
any significant period of time, nor the fact that the decommissioning
is not yet complete, and has not included dismantlement – the activity
that most people think of as “decommissioning” (as indicated in the NGO
Comments).

The EA’s failure to address sabotage, terrorism, human malice

The EA completely ignores the threat of malicious action. There is no
indication that even the minimal security enhancement required in
Canada after the events of Sept. 11, 2001, are being considered for C2
(or C1). The main summary document does not contain a single reference
to any grammatical forms of the words “sabotage,” terror,” or
“malice”(!).

Among the many passages displaying accidental or intentional ignorance
of the possibility and significance of malicious action are the
following:

1. The report states “For the Cernavoda NPP Site, these potential
events included meteorological events, hydrological events (including
flooding due to water level variations), earthquakes and human-induced
events (such as land, sea and air traffic accidents).” (Parentheses as
per original).

Comment: “Human-induced events” should extend beyond just accidents to include sabotage, terrorism and malice.

2. The definition of abnormal events used in the report is “within the
context of the environmental assessment, ‘abnormal events’ refer to
process system failures, component failures or operational occurrences
that result in a release of radioactive or non-radioactive substances.
However, abnormal events exclude events in which one of the safety
systems is required to limit the consequences of releases from
radioactive substances; events that result in fuel failure; and events
where the normal operations regulatory limits are exceeded.”

Comment: Those excluded categories include the results of effective malicious action.

3. Accidents are defined as “A failure of a process system or component
or a procedure which, in absence of a protective or mitigative action
either automatic or operator-initiated, could lead to a significant
release of radioactive material within and/or outside of the plant.”

Comment: The authors are assuming that the “protective or mitigative
action” always occurs. That assumption is unduly optimistic in general,
and especially so in the context of malicious action like a terrorist
attack.

4. The report states that “Cumulative or simultaneous abnormal events
would be unlikely because the two units [Cernavoda 1 and 2] operate
independently. . . . There could, however, be exceptional circumstances
where accidents could occur at both units with higher than normal
releases, doses and effects. Such occurrences, however, would be highly
unlikely given the extensive operational experience from
western-designed NPPs.”

Comment: As previously noted, the U.S. Nuclear Regulatory Commission
does not allow Emergency Water Supply Systems sharing because of the
potential implications of simultaneous multiple unit accidents. In the
U.S., similar rules apply to other systems such as emergency backup
power, a system that is not even mentioned in the EA. Given a variety
of credible multi-unit failure scenarios, such as the large but unknown
contribution of seismic events to total CANDU accident risk, such
scenarios should be reflected in the design basis for safety systems.

Conclusion:

Both before and after reviewing this EA summary, Energy Probe does not
believe that the costs and risks of this project are acceptable to
Canadians, Romanians, or the environment.