new frontiers for nuclear power plants safety
TRANSCRIPT
NEW FRONTIERS FOR NUCLEAR POWER PLANTS SAFETY
dr. Roberto Adinolfi
Chief Executive Officer, Ansaldo Nucleare SpA
Genova, 30 ottobre 2013
Genova, 30 0ttobre 2013
CPEXPO – COMMUNITY PROTECTION
NEW FRONTIERS FOR NUCLEAR POWER PLANTS SAFETY
How is it possible to protect community against man-induced risks…
…e.g. from Nuclear Power Plants?
• The key principles of nuclear safety
• Lessons learned from the Fukushima accident
• What is “acceptable” as man-induced risk?
• New plants: where are we going
• Is it possible to do better?
Genova, 30 0ttobre 2013
CPEXPO – COMMUNITY PROTECTION
THE KEY PRINCIPLES OF NUCLEAR SAFETY
• The development of a civil nuclear industry introduced a new
dimension in industrial safety:
– Not only protection of the investment
– Not only protection of the operators
– …but also protection of the general public
• A nuclear power plant shall not release radioactivity to the general public, even in accident condition, for more than a fraction of natural sources
• Proper barriers against the release of radioactivity shall be deployed, eg:
– The fuel pin
– The pressure boundary of the primary cooling system
– The containment, which envelopes all the radioactive systems of the plant
• The “defence in depth” principle: the barriers shall be designed in such a
way not to be impaired at the same time following an accident
Genova, 30 0ttobre 2013
CPEXPO – COMMUNITY PROTECTION
THE KEY PRINCIPLES OF NUCLEAR SAFETY (cont’d)
• According to the above principles, emergency plans for evacuation are
required by licensing authorities to limit the residual risk for the general
public.
• After Three Miles Island, the deterministic analysis is integrated and
supported by probabilistic risk assessment
• Protection against external events:
– Definition of “worst case” (“Design Basis”) event, on the basis of statistics and /or probabilistic analysis, with frequency of occurence in the range 10-3 – 10-4 ev/yr
– Design of structures and components to withstand, on deterministic bases, the worst case event
• Limited recourse to probabilistic analyses for external events
Genova, 30 0ttobre 2013
CPEXPO – COMMUNITY PROTECTION
LESSONS LEARNED FROM FUKUSHIMA ACCIDENT
(A thorough description of the Fukushima sequence of events is presented in the
memory of Mr. Alessandroni in a separate session)
• The actual Tsunami was largely over the Design
Basis event (the actual Eartquake* was only
marginally higher that the Design Basis one)
• The deterministic defenses against Tsunami
(e.g.a seaside wall of 6 m) were therefore totally
inadequate towards waves of 13-15 m
(according to present understanding, seismic
input didn’t create problems to
structures/systems integrity, but to the offsite
power supply)
• The flooding caused by the tsunami was a
common mode failure of (almost) all the electric
power onsite supply systems
(*) magnitude 9, the fourth largest
ever recorded worldwide
Slipping area 500x200 km
Genova, 30 0ttobre 2013
CPEXPO – COMMUNITY PROTECTION
LESSONS LEARNED FROM FUKUSHIMA ACCIDENT (cont.’d)
• Recovery actions to reinstate core and
spent fuel cooling were made very difficult
from the extensive damages and the
psycologycal situation
• Containment systems in Units 1, 2, 3 and 4
were subsequently impaired by explosions
caused by hydrogen generated from core
melting processes
Defense in depth didn’t work properly
• Evacuation of people (113.000 persons)
from the 20 km radius area was effective in
limiting the early exposure of the public
On-site testimony: “As tremendous aftershocks occurred, with our full face masks still on, we frantically
headed off to the upper ground.”
“While laying down cables at night, entailing the
search of penetrations and terminal treatment
work, we were terrified that we might be electrocuted due to the outside water
puddles.”
Genova, 30 0ttobre 2013
CPEXPO – COMMUNITY PROTECTION
A FEW CONSIDERATIONS
• Fukushima has been a very dramatic natural disaster, affecting a large
number of infrastructures (civil buildings, roads and bridges, distribution
networks, industry plants, dams etc.). All this resulted in almost 20.000
immediate deaths:it is a fact that, out of them, none was related to the
nuclear accidents
• This was largely due to the defense in depth approach, which offered
protection at different levels, even if there were flaws on how it had been
implemented
These flaws were mitigated by the evacuation of a significant number of people, which after one year are still living out of their home, even if large part of the restricted area is no longer “at risk” from a purely scientific point of view.
Genova, 30 0ttobre 2013
CPEXPO – COMMUNITY PROTECTION
WHAT IS “ACCEPTABLE” AS MAN-INDUCED RISK?
• Preventing accidents is the first requirement for any risky industry or
infrastructure
• Protecting against the consequences of any credible accident, either
internally generated or due to an external event, has been generally
considered an adequate measure for plant safety
• Nuclear industry , with its defence-in-depth approach, went already further:
at least for internal events, the goal becomes to mitigate the consequences
(i.e. the releases to the environment) even in the case of beyond design
bases events. (Note that this is not so common in other risky industries…)
• Fukushima shows the need to consistently extend this approach also to
severe natural events with very low probability of occurence
This was the goal of the stress tests conducted in Europe
Genova, 30 0ttobre 2013
CPEXPO – COMMUNITY PROTECTION
WHAT IS “ACCEPTABLE” AS MAN-INDUCED RISK? (cont’d)
Is it enough to limit the direct consequences even of very unlikely events?
• Fukushima persisting evacuation (as it is also at Chernobyl) creates a
strong sentiment of “inadequate protection”
• The new demand seems to be to limit, or to avoid, socio-economic
consequences which add on the direct consequences of the initiating
event
• This is true also for other risks than nuclear. In Emilia, the structural failure
of industrial buildings was perceived as “unacceptable” not only in terms of
human losses but even in terms of loss of jobs.
• Since the recovery time to eliminate those consequences becomes a crucial
factor for acceptability, nuclear is particularly “weak” (“no threeshold”
approach to long term effects)
Genova, 30 0ttobre 2013
CPEXPO – COMMUNITY PROTECTION
NEW PLANTS: WHERE ARE WE GOING?
Generation III reactors
• The design of these advanced plants was based on the lessons learned
from Three Miles Island and Chernobyl accidents and is largely based on a
probabilistic approach to safety
• They are a consistent step forward in the technology, extensively tested and
reviewed by safety authorities through a ten years long process
• Two main types of plant design approach:
– Evolutionary approach: improvement of redundancy and independence
of «standard» safety features (EPR from AREVA)
– Alternate approach: extensive use of passive safety features to
prevent common mode failures and/or operator errors (AP1000 from
Westinghouse)
– Both types are currently under construction (4 EPR and 8 AP1000)
Genova, 30 0ttobre 2013
CPEXPO – COMMUNITY PROTECTION
NEW PLANTS: WHERE ARE WE GOING?
Generation III reactors: some significant steps forward
Improved Defense in Depth
the containment system is designed to withstand core melt scenarios through:
• retention & cooling of corium (in vessel or through core catcher)
• Hydrogen control (igniters and/or recombiners)
• Independent cooling systems (totally passive in the case of AP1000)
so ensuring independence towards other barriers
Lower risk of Total station Blackout
The Ap1000 has no need of diesels to withstand an accident: only batteries are
required, with a capability of at least…….
Lower dependance on operator action
The Ap1000 doesn’t require any operator action for 72 hours after an accident.
Genova, 30 0ttobre 2013
CPEXPO – COMMUNITY PROTECTION
IS IT POSSIBLE TO DO BETTER?
Generation III reactors are addressing many of the safety issues raIsed by
the Fukushima accident.
Acceptability requires reduced socio-economic consequences
There are at least two areas of improvement for the future:
1. Better protection against natural hazards
• Probalistic approach to be extended to external events
• «cliff edge» effects to be avoided, i.e. no sharp increase in
consequences for limited variations in the intensity of the phenomenon
2. Limited impact outside the fence
• Improved containment performances in terms of radiological releases
• Improved post accident monitoring/emergency management
• Recovery strategies for extreme situations
• Clear criteria for evacuation/sheltering measures
Thank you for your attention!