Author: Jan-Anders Larsson

Research

2014:56
Seismic design and analysis
of safety-related nuclear
structures in Sweden

Report number: 2014:56 ISSN: 2000-0456

Available at www.stralsakerhetsmyndigheten.se
SSM 2014:56
SSM perspective

Background
Following the severe accident in the Fukushima Dai-ichi nuclear power
plant in Japan on March 11, 2011, the European Council decided to request
stress tests to be performed on all European nuclear power plants. The Eu-
ropean Commission, the European Nuclear Safety Regulators Group (ENS-
REG) and the Western European Nuclear Regulators Association (WENRA)
were commissioned to develop the scope for the stress tests. It was decided
to focus the stress tests and the peer review on three main topics where na-
tural hazards including earthquake, tsunami and extreme weather was one
of these topics.

One step in the European stress tests was the international peer review of
each country’s activities. On the basis of the international peer review and
SSM’s own review of the Swedish nuclear power plants, SSM has specified
prioritized activities in “the Swedish action plan” with the intention to
handle all plant weaknesses identified by the European stress tests.

Objectives
The main objective of the project was to identify and evaluate different ap-
proaches for the design and analysis of safety-related nuclear structures in
Sweden with respect to severe earthquakes.

Results
The report presents the historical development of the seismic design for the
U.S., France and Sweden with special focus on issues related to severe earth-
quakes beyond the design basis as well as important aspects concerning the
design basis ground motions for the Swedish nuclear facilities.
The report provides recommendations on a revised model for seismic ha-
zard assessments, on minimum requirements for seismic analysis of safety-
related nuclear structures in Sweden as well as some recommendations for
new structural design or redesign of existing structures. It also provides
a proposal to address seismic margin assessments for severe earthquakes
beyond the design basis.

The detailed recommendations on minimum requirements and safety as-

sessments are focused on the safety-related building structures.

Need for further research

At present there is no need for further research in this area.

Project information
Contact person SSM: Kostas Xanthopoulos
Reference: SSM 2013/1651

SSM 2014:56
SSM 2014:56
Author: Jan-Anders Larsson
Scanscot Technology AB

2014:56
Seismic design and analysis
of safety-related nuclear
structures in Sweden

Date: December 2014

Report number: 2014:56 ISSN: 2000-0456
Available at www.stralsakerhetsmyndigheten.se
This report concerns a study which has been conducted for the
Swedish Radiation Safety Authority, SSM. The conclusions and view-
points presented in the report are those of the author/authors and
do not necessarily coincide with those of the SSM.

SSM 2014:56
Contents
Executive summary ...................................................................................... 5

Sammanfattning ............................................................................................ 7

1. Introduction ........................................................................................ 9

1.1 The European stress tests ...................................................... 9

1.1.1 General ...................................................................................... 9

1.1.2 European level recommendations ............................................. 9

1.1.3 The French action plan ............................................................ 11

1.1.4 The Swedish action plan .......................................................... 12

1.2 Scope of the report .................................................................15

2. Historical development of the seismic design basis .....................16

2.1 General ....................................................................................16

2.2 The United States ...................................................................16

2.2.1 Introduction .............................................................................. 16

2.2.2 Historical development of seismic design ground motions....... 17

2.2.3 Historical development of the structural modeling technique ... 20

2.2.4 Seismic design classification ................................................... 21

2.2.5 Codes and standards ............................................................... 22

2.2.6 Consequences of Fukushima Earthquake on Design Basis

Earthquake .............................................................................. 23

2.3 France ......................................................................................26

2.3.1 Introduction .............................................................................. 26

2.3.2 Historical development of seismic design ground motions....... 27

2.3.3 Evaluation of seismic safety margins ....................................... 30

1
SSM 2014:56
 2.4 Sweden ....................................................................................33

2.4.1 Introduction .............................................................................. 33

2.4.2 Historical development of the seismic design basis ................. 34

2.4.3 Seismic design classification ................................................... 37

2.4.4 Codes and Standards .............................................................. 38

3. Design basis ground motions ..........................................................40

3.1 General ....................................................................................40

3.2 Design basis considerations .................................................40

3.2.1 Basic requirements according to IAEA Safety Guides ............. 40

3.2.2 Aspects as regards an eventual Operating Basis Earthquake

(OBE) ....................................................................................... 40

3.2.3 Requirements as regards the Design Basis Earthquake (DBE) 41

3.2.4 Current design ground motions in SKI Technical Report 92:3 . 42

3.2.5 Recommendations on revised design ground motions ............ 44

4. Seismic analysis methods ................................................................46

4.1 General ....................................................................................46

4.2 Structural modeling ................................................................46

4.2.1 General requirements .............................................................. 46

4.2.2 Material properties ................................................................... 46

4.2.3 Modeling of stiffness of concrete elements .............................. 47

4.2.4 Modeling of mass distribution .................................................. 47

4.2.5 Modeling of damping................................................................ 47

4.2.6 Modeling of hydrodynamic effects ........................................... 48

4.3 Seismic analysis .....................................................................48

4.3.1 General .................................................................................... 48

2
SSM 2014:56
 4.3.2 Time history method ................................................................ 48

4.3.3 Response spectrum method .................................................... 49

4.3.4 Equivalent static method .......................................................... 49

4.3.5 Multiply-support systems ......................................................... 49

4.3.6 Combination of modal and component responses ................... 49

4.3.7 Soil-structure interaction .......................................................... 49

4.3.8 Input for subsystem seismic analysis ....................................... 50

5. Seismic design of nuclear structures..............................................51

5.1 General ....................................................................................51

5.2 Seismic design classification ................................................51

5.2.1 Seismic classification according to IAEA Safety Guides .......... 51

5.2.2 Seismic classification according to USNRC Regulatory Guides52

5.2.3 Seismic classification according to YVL Guides in Finland ...... 52

5.2.4 Seismic classification at the Swedish nuclear facilities ............ 53

5.3 Design Basis Earthquake .......................................................54

5.3.1 General .................................................................................... 54

5.3.2 Design Basis Earthquake ......................................................... 54

5.4 Design Extension Earthquake ...............................................55

5.4.1 General .................................................................................... 55

5.4.2 Design Extension Earthquake .................................................. 56

5.4.3 Recommendations on Design Extension Earthquake .............. 57

5.5 Seismic load combinations....................................................57

5.6 Seismic safety verification .....................................................60

5.6.1 General .................................................................................... 60

5.6.2 Codes and Standards .............................................................. 60

3
SSM 2014:56
 5.6.3 Seismic safety verification ........................................................ 61

6. Seismic evaluation of existing nuclear structures .........................63

6.1 General ....................................................................................63

6.1.1 General considerations ............................................................ 63

6.1.2 Objectives of the seismic safety evaluation ............................. 63

6.1.3 Selection of appropriate methodologies ................................... 64

6.2 Safety evaluation against the Design Basis Earthquake.....64

6.3 Seismic Margin Assessments ...............................................65

6.3.1 General .................................................................................... 65

6.3.2 Objectives of the seismic safety evaluation ............................. 66

6.3.3 Determine the Review Level Earthquake (RLE) ...................... 66

6.3.4 Selection of success paths and selected SSCs ....................... 67

6.3.5 Seismic response analysis ....................................................... 68

6.3.6 Capacity assessments of the selected SSCs ........................... 68

6.3.7 Determination of utilization ratios for the selected SSCs ......... 69

7. References .........................................................................................70

Appendix 1: List of figures ..........................................................................75

Appendix 2: List of tabels ...........................................................................77

Appendix 3: List of Acronyms ....................................................................79

4
SSM 2014:56
Executive summary
The severe earthquake and the subsequent tsunami that devastated the nuclear power plant at
Fukushima Dai-ichi in Japan on March 11, 2011 has resulted in extensive international discus-
sions and investigations as regards natural hazard assessments, and how to improve the exist-
ing safety evaluation methods for severe external events beyond the design basis.
In this report, the outcome of the stress tests of the European nuclear power plants is assessed,
with special focus on earthquake effects on building structures at nuclear facilities. The Swe-
dish action plan, which was developed after the stress tests, emphasize the needs to review and
update the seismic design basis as well as the conditions and methods for seismic analysis and
design. Additionally also, the methods for seismic margin assessments for ground motions
exceeding the Design Basis Earthquake (DBE) need to be improved.
The historical development of the seismic design practice is reported for the U.S., France and
Sweden. Important aspects regarding the design basis ground motions for the Swedish nuclear
facilities are addressed and recommendations on a revised model for seismic hazard assess-
ments are provided.
Seismic analysis methods and the seismic design process for new nuclear facilities as well as
safety evaluation procedures for existing facilities are covered at a general plant level for safe-
ty-related Structures, Systems and Components (SSCs). However, detailed recommendations
on minimum requirements and safety assessments are focused on the building structures.
A vast majority of the safety-related structures at the Swedish nuclear facilities consists of
concrete shear walls and slab systems of general heavy proportions. For steel framework struc-
tures, the effects of wind and snow loads normally govern the design. Hence, detailed require-
ments on material properties and procedures for structural analysis as well as determination of
failure modes and strength properties for seismic margin assessments are primarily addressed
for load-bearing concrete structures.
John D. Stevenson, Consulting Engineer and Jean-Pierre Touret, Scanscot Technology France
have provided essential input as regards the historical development of the seismic design basis
in section 2.2 for the U.S. and in section 2.3 for France. They have also reviewed the other
parts of the report.

5
SSM 2014:56
SSM 2014:56
Sammanfattning
Den svåra jordbävningen med den efterföljande tsunamin som ödelade kärnkraftverket i Fu-
kushima Dai-ichi i Japan den 11 mars 2011 har resulterat i omfattande internationella diskuss-
ioner och utredningar avseende riskbedömningar av naturkatastrofer, samt hur man kan för-
bättra de nuvarande metoderna för säkerhetsvärderingar av svåra yttre händelser utanför de-
sign.
I denna rapport genomförs en utvärdering av resultaten från stresstesterna av de europeiska
kärnkraftverken, med fokus på seismiska lasteffekter på byggnadskonstruktioner vid kärntek-
niska anläggningar. Den svenska handlingsplanen som togs fram efter stresstesterna betonar
behovet att se över och uppdatera de seismiska dimensioneringsförutsättningarna samt villko-
ren och metoderna för seismisk analys och design. Dessutom bör också metoderna för bedöm-
ningar av säkerhetsmarginaler för markrörelser som överstiger den dimensionerande jordbäv-
ningen (DBE) förbättras.
Den historiska utvecklingen av de seismiska dimensioneringsprinciperna redovisas för USA,
Frankrike och Sverige. Viktiga aspekter på de dimensionerande seismiska markrörelserna för
de svenska kärntekniska anläggningarna behandlas tillsammans med rekommendationer avse-
ende en reviderad modell för seismiska riskbedömningar.
Seismiska analysmetoder och den seismiska designprocessen för nya nukleära anläggningar
samt procedurerna för säkerhetsbedömningar av befintliga anläggningar, behandlas på en
övergripande anläggningsnivå för säkerhetsrelaterade byggnader, system och komponenter
(SSC). Emellertid fokuseras de detaljerade rekommendationerna avseende minimikrav och
säkerhetsbedömningar på byggnadskonstruktionerna.
Flertalet av de säkerhetsrelaterade byggnaderna vid de svenska nukleära anläggningarna består
av betongväggar och bjälklag med grova dimensioner. För byggnader med bärande stål-
stommar blir ofta effekter av vind- och snölaster dimensionerande. Därför redovisas detaljkrav
avseende materialparametrar och procedurer för strukturanalyserer samt bestämning av brott-
moder och hållfasthetsvärden vid seismiska säkerhetsutvärderingar primärt för bärande kon-
struktioner av betong.
John D. Stevenson, Consulting Engineer och Jean-Pierre Touret, Scanscot Technology France
har lämnat värdefulla bidrag avseende den historiska utvecklingen av de seismiska dimension-
eringsförutsättningarna i avsnitt 2.2 för USA och i avsnitt 2.3 för Frankrike. De har också
granskat de övriga delarna av rapporten.

7
SSM 2014:56
SSM 2014:56
1. Introduction
1.1 The European stress tests

1.1.1 General
Following the severe accidents which occured in the Fukushima Dai-ichi nuclear power plant,
the European Council in March 2011 requested stress tests to be performed on all European
nuclear power plants. The European Council invited ENSREG, the European Commision and
WENRA to develop the scope for the stress tests. It was decided to focus the stress tests and
the peer review on three main topics which were directly derived from the preliminary lessons
learned from the Fukushima disaster:
- Natural hazards, including earthquake, tsunami and extreme weather
- Loss of safety systems
- Severe accident management
The stress tests and the peer review assessed these topics in a three step process. The first step
required the operators to perform an assessment and set out proposals following the ENSREG
specifications. The second step was for the national regulators to perform an independent re-
view of the operators’ assessments and issue requirements whenever appropriate. The last step
was a peer review of the national reports submitted by regulators. The objectives of the peer
review were to assess the compliance of the stress tests with the ENSREG specification, to
check that no important issue has been overlooked and to identify strong features, weaknesses
and relevant proposals to increase plant robustness in light of the preliminary lessons learned
from the Fukushima accident. The operators submitted their final assessments in October 2011
and the regulators submitted their final reports in December 2011.
The peer review started in January 2012. The peer review was completed with a main report
that includes final conclusions and recommendations at European level as well as country re-
ports that included country-specific conclusions and recommendations. The report was ap-
proved by ENSREG and the European Council in April 2012. In a joint ENSREG/European
Council statement the stress test report was accepted and it was agreed that an ENSREG action
plan would be developed to track how well the recommendations were implemented. As part
of the ENSREG action plan each national regulator generated a country-specific action plan.

1.1.2 European level recommendations

The peer review was structured in accordance with the three topics of the stress tests; natural
hazards, loss of safety systems and severe accident management. The peer review identified
four main areas of improvement to be considered at the European level as compiled in [1] and
[2] and shown in Table 1.1.
Besides these four main recommendations, the peer review highlighted a number of other ob-
servations. For topic item no 1 (natural hazards), eight subtopics were highlighted, amongst
which five addressed various recommendations on seismic-related issues, as shown in Table
1.2.

9
SSM 2014:56
 Table 1.1 – European level recommendations according to [1] and [2]

No Area Recommendations
1 European guidance on asses- The peer review Board recommends that WENRA,
ment of natural hazards and involving best available expertise from Europe, devel-
margins op guidance on natural hazard assessments, including
earthquake, flooding and extreme weather conditions,
as well as corresponding guidance on assessment of
margins beyond the design basis and cliff-edge effects.
2 Periodic safety review The peer review Board recommends that ENSREG
underline the importance of periodic safety review. In
particular, ENSREG should highlight the necessity to
reevaluate natural hazards and relevant plant provi-
sions as often as appropriate but at least every 10 years.
3 Containment integrity Urgent implementation of the recognized measures to
protect containment integrity is a finding of the peer
review that national regulators should consider.
4 Prevention of accidents result- Necessary implementation of measures allowing pre-
ing from natural hazards and ventions of accidents and limitation of their conse-
limiting their consequences quencies in case of extreme natural hazards is a finding
of the peer review that national regulators should con-
sider.

Table 1.2 – Topic item no 1 (natural hazards) relating to seismic hazard according to [1]
and [2]

Subitem
Hazard frequency The use of a return frequency of 10-4 per annum (0.1g minimum
PGA for earthquakes) for plant review/back-fitting with respect to
external hazard safety cases.
Secondary effects of The possible secondary effects of seismic events, such as flood or
earthquakes fire arising as a result of the event, in future assessments.
Seismic monitoring The installation of seismic monitoring systems with related proce-
dures and training.
Qualified walkdowns The development of standards to address qualified plant walkdowns
with regard to earthquake, flooding and extreme weather, to provide
a more systematic search for non-conformities and correct them
(e.g. appropriate storage of equipment, particulary for temporary
and mobile plant and tools used to mitigate beyond design basis
(BDB) external events).
External hazard margins In conjunction with main recommendation 1 (European guidance on
assesment of natural hazards and margins), the formal assessment
of margins for all external hazards including, seismic, flooding and
severe weather, and identification of potential improvements.

10
SSM 2014:56
As regards the topic item “Hazard frequency”, the ENSREG peer review team states in [2]
that a good practice is that external events should be addressed by designing to the hazard level
consistens with a 10 000 year return period, i.e. an annual frequency equivalent to 10-4. How-
ever there are some countries where the acceleration levels consistent with the perceived 10 -4
yearly return frequency are very low. In these circumstances, IAEA guidance suggests that a
minimum 0.1g horizontal PGA should be adopted.
As regards the topic item “External hazard margins”, the peer review process noted that the
evaluation of margins beyond design basis (BDB) is not consistent in participating countries.
The majority have made only a general claim that margins exists and therefore there is no in-
formation on the basis of which to consider effective potential improvements. Very few coun-
tries have determined cliff-edge effects and the associated protection improvements in the
manner envisaged by ENSREG. There are well-established practices for assessing seismic
margins BDB, referred to as seismic margin assessment (SMA). This appears similar to a de-
terministic method, although the acceptance criteria are derived from probabilistic fragility
assessments. Alternatively, similar fragilities can be implemented in a seismic PSA. On the
basis of this outcome, the peer review team recommended that WENRA, involving the best
available expertise from Europe, should consider how to determine a consistent approach to
margin assessments for external events.

1.1.3 The French action plan

After having received and evaluated the complementary safety assessments from the French
operators, the French safety authority (ASN) concluded in the following general statements
[4]:
- The natural disaster which struck the Fukushima Dai-ichi NPP confirms that, whatever
the precautions taken in the design, construction and operation of nuclear facilities, an
accident can never be completely ruled out.
- The licensee has the overall responsibility for the safety of its facilities while, on behalf
of the State, ASN is responsible for regulating and monitoring nuclear safety, with the
technical support of IRSN and its Advisory Committees. Pursuant to the law, ASN en-
sures that the safety of French civil nuclear facilities shall be maintained continuously,
in particular through the periodic review process and the integration of experience
feedback.
ASN considers that the continued operation of the French nuclear facilities requires with high-
est priority their robustness to extreme situations to be increased beyond existing safety mar-
gins. ASN is thus requiring the licensees to adopt a range of measures to provide the facilities
with the means to deal with the following events:
- A combination of natural phenomena of an exceptional scale and exceeding those
adopted in the design or the periodic safety review of the facilities.
- Severe accident situations following the prolonged loss of electrical power or cooling
and liable to affect all the facilities at a given site.
Among these new provisions, ASN would in particular recommend the creation of a "hardened
safety core" of essential SSCs and organizational arrangements making it possible to manage
the fundamental safety functions in extreme situations, with the aim of preventing a severe
accident, limiting large-scale radioactive releases if the accident cannot be controlled and ena-
bling the licensee, even in extreme situations, to perform its emergency management duties.
This will for example involve setting up a "bunkerized" emergency management centre with
diesel electricity generator, and an ultimate backup water supply. The equipment to be includ-

11
SSM 2014:56
ed in this “hardened safety core” must be designed to withstand major events (earthquake,
flooding, etc.), of a scale far in excess of those used to determine the strength of the facilities,
even if not considered to be plausible.
In January 2014, the ASN adopted 19 resolutions [5] setting out additional requirements for
implementation of the post-Fukushima “hardened safety core” in EDF’s NPPs. These resolu-
tions specify the objectives and the contents of this “hardened safety core”, which shall com-
prise measures to:
- Prevent a severe accident affecting the core of the reactor or the spent fuel pool.
- Limit the consequences of an accident which could not be avoided, with the aim of pre-
serving the integrity of the containment without opening the venting system. This aim
of mitigating the consequences of an accident applies to all the phases of an accident.
- Enable the licensee to perform its emergency management duties.
This “hardened safety core” must be as independent as possible from the existing systems,
more specifically with regard to I&C and electrical power supplies. The ASN resolutions spec-
ify the design rules to be adopted for the “hardened safety core” equipment. These rules must
comply with the most demanding nuclear industry standards, used for the design and construc-
tion of installations requiring a high level of safety. Finally, they will lead EDF to determine
the maximum hazards to be considered for the “hardened safety core” equipment, in particular
for earthquake and flooding in order to ensure ultimate protection of the facilities. The hazard
level for the earthquake is still pending, it should probably be consistent with BDBE used in
the SMA. These resolutions will apply to all the NPPs in operation, as well as to the Flaman-
ville 3 EPR reactor currently under construction.
Over and above the “hardened safety core”, EDF is required to comply with the following pre-
scriptions:
- The “Nuclear rapid intervention force (FARN)”, fully operational no later than the end
of 2014. This force can provide assistance to a damaged site by providing specialized
teams to back up those of the plant concerned and mobile equipment to supply addi-
tional water and electricity. A number of modifications were therefore made to the re-
actors to make it easier to connect this equipment brought on-site by the FARN.
- Launch before 30th June 2013 the deployment of the 58 ultimate backup diesel genera-
tor sets for all the reactors, medium-power generator sets were added to each reactor.
- Additional training to its staff for intervention in the event of an earthquake and a se-
vere accident.
With regard to the basic safety requirements concerning the consideration of seismic hazard,
the prescriptions setting out additional requirements for deployment of the “hardened safety
core” significantly reinforce the ability of the NPPs to withstand this risk. Finally, together
with IRSN, ASN has begun to examine an update of the basic safety rule 2001-01 [40] con-
cerning the determination of the seismic risk.

1.1.4 The Swedish action plan

One step in the European stress tests was the international peer review of each country’s activi-
ties. The main observations from the peer review of the Swedish activities [3] as regards plants
assessments with respect to the Design basis Earthquake (DBE) can be summarized as follows:
- Licensing apply a DBE within a radius of twenty kilometers of a strength corresponding
to a magnitude of approximately 6.0 on the Richter scale and with a probability of
once per 100 000 years (10-5).

12
SSM 2014:56
 - The assessment of the DBE uses a probabilistic approach based on a so called “average
Fennoscandian seismicity function” accounting for site conditions of hard rock. Con-
sideration of site effects leads to compute peak ground accelerations for the DBE by
the reduction of the PGA related to the Swedish 10-5 earthquake by 15% to account for
the favourable site conditions as all plants are sited on solid rock.
- It appears that the values of the DBEs for the different sites are close to IAEA’s sug-
gested minimum values at the background of the active deformation of Fennoscandia,
which is proved by geodetic and paleoseismologic data.
- The full compliance of the reactors, originally not designed to withstand seismic loads,
is expected in 2013 after the implementation of modifications (e.g. anchoring of me-
chanical components, emergency power supply) in accordance with the requirements
on seismic safety, in force in 2005.
The main observations of the peer review group as regards assessment of plants beyond design
basis earthquakes are as follows:
- The integrity assessments of the reactor containment, scrubber building and spent fuel
pools (SFPs) are based on approximate calculation methods and engineering judge-
ment on a best estimate basis due to the limited time available for the study.
- The integrity of reactor containments, SFPs and other important buildings are estimated
to be preserved in case of the 10-7-earthquake. However, there is need for refined anal-
yses and further investigations before definite conclusions are possible. Such investi-
gations should emphasize on evaluating margins to reach safe shutdown conditions.
According to SSM [3], the Swedish earthquake is based on observations and historical ac-
counts of earthquakes in Fennoscandia for about 500 years, as well as comparisons with the
occurrence of earthquakes in other low seismic regions in the world. Based on these facts,
SSM estimated that fairly reliable predictions can be performed concerning the earthquakes
that are likely to occur in Scandinavia in a 500 year geological time scale. However, this was
questioned as a restriction by the peer review team, due to the fact that geodetic and paleo-
seismologic data which according to some researchers indicates continuous active uplift and
deformation of Fennoscandia. Also, IAEA in SSG-9 [6] explicitly suggests the use of such data
in low seismicity intraplate regions. SSM has agreed to consider the existing approach by tak-
ing into account the geodetic and paleoseismologic data.
On the basis of the international peer review and SSM’s own review of the Swedish NPP
plants, SSM has specified prioritized activities in “the Swedish action plan” [7], with the inten-
tion to handle all plant weaknesses identified by the EU stress tests. As regards earthquake
hazards, actions are defined as shown in Table 1.3. These actions shall be finalished latest
2015.

13
SSM 2014:56
 Table 1.3 – Swedish action plan [7] regarding earthquake hazards

No Item Action
T1.LA.1 Seismic plant analyses A return frequency of 10-5/year (with a minimum peak
ground acceleration of 0.1g) shall be used as a basis for
plant reviews/back-fitting. The following actions shall
be performed:
- Further studies regarding the structural integrity
of the reactor containments, scrubber buildings
and fuel storage pools shall be performed.
- The pipes between the reactor containment and
the MVSS that allows a controlled pressure re-
lief of the reactor containment shall be evalu-
ated further.
T1.LA.2 Investigation regarding Investigations regarding secondary effects of an earth-
secondary effects of an quake shall be performed. Fire analyses at Swedish
earthquake NPPs are in general performed according to SAR but
analyses of the effects of fire as a result of an earth-
quake have not been carried out at any of the Swedish
NPPs. A more detailed analysis of earthquake induced
flood, where for example leakage from broken water
storage tanks and cracks in the cooling water channels
are taken into account have to be included in the anal-
yses regarding secondary effects.
T1.LA.3 Review of seismic Seismic monitoring systems are installed at all Swedish
monitoring sites. The utilities shall review the procedures and
training program for seismic monitoring and imple-
ment them.
T1.LA.9 Investigations of ex- In conjunction with recommendation regarding flood-
ternal hazard margins ing margin assessments, a formal assessment of mar-
gins for all external hazards including seismic, flood-
ing and severe weather, and identification of potential
improvements shall be performed. Weaknesses in the
plants shall be identified.
Regarding the seismic margins an evaluation of struc-
tures, systems and components against ground motions
exceeding DBE shall be performed. Such evaluations
shall emphasize on margins.
T1.RA.1 Research project re- SSM will start up a research project regarding the in-
garding the influence fluence of paleoseimological data on the existing mod-
of paleoseismological el regarding frequency and strength of the ground re-
data sponse spectra constructed in the project SKI 92:3 [8] .

14
SSM 2014:56
1.2 Scope of the report
A general conclusion from the topic “Natural Hazards” within the scope of the European stress
tests, is the necessity for further activities on the Swedish nuclear facilities in order to develop
the following issues:
- The approach to determine the seismic design basis, as well as the conditions and meth-
ods for seismic design, analysis and safety verification.
- Methods to address cliff-edge1 effects and seismic margin assessments for ground mo-
tions exceeding the DBE.
In this report, the outcome of the stress tests of the European nuclear power plants is assessed,
with special focus on earthquake effects on building structures at nuclear facilities.
The historical development of the seismic design practice is reported for the U.S., France and
Sweden in section 2, with especial focus on issues related to severe earthquakes beyond the
design basis.
In section 3, important aspects as regards the design basis ground motions for the Swedish
nuclear facilities are addresses and recommendations on a revised model for seismic hazard
assessments are provided.
Recommendation on minimum requirements for seismic analysis of safety-related nuclear
structures in Sweden, in accordance with ASCE 4-98 [24] and IAEA SG-G-1.6 [50], are pre-
sented in section 4.
The different steps in the seismic design process are addressed at a general plant level in sec-
tion 5, together with some recommendations for new structural design or redesign of existing
structures.
Specific considerations regarding seismic safety evaluation of existing structures not designed
against earthquakes are reported in section 6, together with a proposal to address seismic mar-
gin assessments for severe earthquakes beyond the design basis.
Seismic analysis methods, the seismic design process for new facilities and safety evaluation
procedures for existing nuclear facilities are covered at a general plant level for safety-related
Structures, Systems and Components (SSCs). However, detailed recommendations on mini-
mum requirements and safety assessments are focused on the safety-related building structures.
A vast majority of the buildings at the Swedish nuclear facilities consists of concrete shear
walls and slab systems of general heavy proportions. For steel framework structures, the ef-
fects of wind and snow loads normally govern the design. Hence, detailed requirements on
material properties and procedures for structural analysis as well as determination of failure
modes and strength properties for seismic margin assessments are primarily addressed for con-
crete structures.

1
when a small deviation of a design plant parameter give rise to an abrupt worsened sit-
uation for the whole plant.

15
SSM 2014:56
2. Historical development of the seismic design
basis
2.1 General
The first generation of nuclear power facilities in the U.S., which were commissioned during
the 1950s and the beginning of the 1960s, included only some general seismic design recom-
mendations that would be applied to any building structure, without any detailed requirements.
The rapid expansion of the nuclear power industry during the 1960s and 1970s was in fact one
important reason for the development of the seismic design requirements applicable to safety-
related nuclear structures, distribution systems and components. The new knowledge and expe-
rience in seismic engineering were soon reflected in new standards for NPPs.
Standards and guidelines for seismic design and analysis of NPPs have to a large extent been
developed in the U.S. under the superintendence of the USNRC. Later on, these standards and
guidelines were also adopted for nuclear facilities in many other countries. In order to better
understand the principles of current design criteria for seismic analysis of safety-related nucle-
ar structures and how to improve the assessments as regards beyond design issues for severe
earthquakes, a historical retrospect of the development of the international design practice is
presented in this chapter. The focus is on the developments in the U.S. and France. Some im-
portant perspective from the Swedish horizon is also considered.

2.2 The United States

2.2.1 Introduction
There is a hierarchy of requirement in the U.S. in order to regulate the seismic design of NPPs.
These requirements are as follows:
- Federal Laws of the U.S. These are laws passed by the U.S. Congress. These laws pro-
vide the highest tier of requirement which are in broadly stated objectives and have the
force of law and are mandatory.
- Code of Federal Regulations (CFR). These are requirements prepared by the USNRC
intended to provide more specific requirements to implement the laws. The require-
ments also have the force of law and are mandatory.
- Regulatory Guide (RG) and Standard Review Plan (SRP) procedures. These are re-
quirements that if followed would satisfy the USNRC’s interpretation of Federal Laws
and CFR. These are not mandatory, but if RG or SRP are not used or departed from,
the designer must justify the difference to the satisfaction of the USNRC. It should be
noted that the requirements of a Federal Regulation and provisions of a RG or SRP
provisions typically become USNRC’s policy one to three years before they are for-
mally published.
- Codes & Standards for design and constructiorn. The most important standards for nu-
clear concrete structures are ASME B&PV Code, Sect III, Div 2 [9] for concrete reac-
tor containments and ACI 349 [10] for other safety-related nuclear structures.
The evaluation of the seismic effects on safety-related nuclear structures can be divided into
three basic design activities as follows:

16
SSM 2014:56
 - Define earthquake phenomena and resultant loads typically in the form of Peak Ground
Acceleration (PGA) and response spectral shape.
- Identify procedures to convert earthquake loads to energized forces, stresses and defor-
mations or strains in safety-related nuclear structures, distribution systems and compo-
nents (SSCs).
- Provide acceptance criteria associated with the resultant generalized forces, deformation
or strains.
These three activities will be discussed historically as they were developed for nuclear safety,
from the initial static and dynamic deterministic based criteria to the current risk informed
probabilistic criteria used in the U.S. today.

2.2.2 Historical development of seismic design ground motions

The earthquake design effort for NPPs in the U.S. was first contained in TID 7024 [11] pub-
lished in 1963. TID 7024 [11] was a departure from conventional codes for building structures,
mechanical and electrical distribution systems and components design. Prior to 1965 the U.S.
Uniform Building Code [12] was used. For instance Connecticut Yankee NPP, whose design
began in 1964, was initially designed for 0.03g static acceleration as specified by the then ap-
plicable Uniform Building Code [12]. Later the plant was evaluated for a 0.17g seismic PGA
and response spectra. The TID 7024 [11] publication was prepared for the then U.S. Atomic
Energy Commission by a group of industry experts.
The definition of seismic PGA was addressed by the publication of 10 CFR100 Appendix A
[13], which defines the SSE, formally in 1973 but which had been in use since 1965. The SSE
corresponds to the DBE for commercial NPPs and is defined as that earthquake which produc-
es the maximum vibratory ground motion for which certain SSCs are designed to ensure the
following:
- Integrity of the reactor coolant pressure boundary.
- Capability to shut down the reactor and maintain it in a safe shutdown condition, or
- Capability to prevent or mitigate the consequences of accidents, which would result in
potential off-site exposures.
In 10 CFR100 Appendix A [13], the definition of seismic PGA applicable to the ground sup-
port of safety related nuclear structures was developed as follows:
For tectonic provinces without known capable fault locations, the largest historically recorded
earthquake occurring within the tectonic province of the site was moved to the site. In general,
the PGA which defined the anchor of the Housner shaped ground response spectra [14] in TID
7024 [11] was used to generate the seismic acceleration input to the foundation of safety-
related nuclear structures (Seismic Category I) in the period before 1967.
In some instances, large historical earthquakes in the Eastern U.S. such as Charleston in 1886
and in the Central U.S. such as New Madrid in 1811-1812, governed over the local tectonic
province historical earthquakes moved to the site.
TID 7024 [11] provided suggested reductions or attenuation of seismic intensities and associ-
ated PGAs as a function of the distance to the site from the hypocenter of earthquake in tecton-
ic provinces other than the site province. These attenuation criteria were based on Southern
California earthquake data. Subsequent studies for the Central/Eastern/Southern regions of the
U.S. showed that the slope of the attenuation of seismic intensity with distance in California
was too severe by a factor of 2 or more in the Central/Eastern/Southern U.S. There is continu-

17
SSM 2014:56
ing research into the attenuation parameter which is expected to be complete by the end of
2013. The current relationships were developed by EPRI in 2006.
Earthquake loads on SSCs are determined by the PGA to which the design basis ground re-
sponse spectra is anchored, (acceleration in excess of at least 33Hz). The Housner shaped
ground response spectrum [14] is based on an average of four measured strong motion earth-
quake response spectra. These spectra were used for NPP design in the U.S. between 1965 and
1967 and were replaced by the original and modified Newmark NBK response spectra which
were used between 1968 and 1971. These spectra were then replaced by the Newmark Blume
and Kapoor spectra in 1971. In 1973 the USNRC published the RG 1.60 [15] response spectra
as shown in Figure 2.1 which formally replaced the Newmark NBK spectra, as discussed by
Stevenson and Conan in [16].

Figure 2.1 – Horizontal design response spectra (5% critical damping) according to RG
1.60 [15], scaled to 1g horizontal ground acceleration

The design response spectra as defined by Housner in [14] and by USNRC in RG 1.60 [15] are
examples of so called standard, generic or site-independent spectra. The term standard here
refers to response spectra that have been developed by statistical analysis of a set of strong
motion data obtained within a wide range of distances of relatively large magnitude earth-
quakes and without specific consideration of the tectonic environment or the local subsurface
conditions at the site being evaluated, see Figure 2.2.

18
SSM 2014:56
Figure 2.2 – Example of a standard or site-independent response spectra, developed from
a statistical data from recorded earthquakes, according to [17].

Starting in the late 1980s, it began to be realized that the standard type of spectra were less
appropriate for soft soil foundations characterized by soils having a low strain and shear wave
velocity, and for sites susceptible to high frequency motions (where significant spectral ampli-
fication occurs at frequencies of 33 Hz and beyond) if systems and components are sensitive to
such motion. For such conditions, site-specific response spectra based on a reference annual
probability of exceedance approach started to be developed.
Site-specific spectra have the advantage of incorporating specific considerations of the tectonic
environment and subsurface conditions at a site. The development of these spectra may be
based on applicable response-spectral attenuation relationship or a statistical analysis of a se-
lected set of strong motion data to be particularly applicable to the site, and/or on modeling and
analysis of the effect of physical factors (earthquake source characteristics, geologic travel
path, and local soil conditions) on ground motions at the site.
This recognition of the probabilistic nature of the seismic hazard is based on the need to define
such risks probabilistically in order to meet overall safety goals of the USNRC. These goals
were expressed probabilistically by the publication of 10 CFR100.23 [19] and 10 CFR 50 Ap-
pendix S [20] in 1996, later on resulted in the publication of the USNRC RG 1.165 [21] in
1997, in which a probabilistic basis for determining the DBE was provided.
It should be noted that the RG 1.165 [21] introduced several more details to the process of
developing the DBE (SSE) requirements as follows:
- The requirement to develop a response spectra specifically applicable to the site rather
than use of only a generic response spectra (i.e. RG 1.60 [15]).
- A distinction between a site spectral ground motion on a real or assumed site rock sta-
tum and that applied to plant structures.

19
SSM 2014:56
 - The establishment of the DBE response spectra at the median (50th percentile) 10-5/yr.
probability of exceedence level.2
- However, this RG 1.165 [21] continued to assume the dominate earthquake acceleration
occurs in the 2 to 10 Hz frequency range characteristic of a large earthquakes in re-
gions (as in California) where there are many capable faults identified.
It should be noted that the RG 1.165 [21] was withdrawn in 2010 and replaced by another RG
1.208 [22] published in 2007. RG 1.208 [22] is the current USNRC recommended procedures
for developing the DBE (SSE) requirements in the U.S.
RG 1.208 [22] uses a performance-based approach instead of the reference probability ap-
proach as in RG 1.165 [21], in order to ensure that NPPs can withstand the effects of earth-
quakes with a desired performance. Further, the method consists of establishing a site-specific
Uniform Hazard Response Spectra (UHRS) with spectra coordinates at each frequency having
the same probability of occurrence and is based on the procedures developed in Chapter 2 of
ASCE 43-05 [23].

2.2.3 Historical development of the structural modeling technique

Prior to 1965, equivalent static seismic loads on structures were defined on the basis of the
fundamental frequency of the structure and an assumed flexibility based on a √(k/m) relation-
ship with m being the mass of the structure and k its shear and bending moment stiffness. Mul-
ti-degrees of freedom models using Housner response spectra [14] as input, and as well includ-
ing springs representing SSI where a plant was founded on other than rock or very stiff soil,
enhanced the modeling capability after 1965. At the end of the 1960s computer programs be-
came available in the nuclear industry, permitting multi-degree of freedom stick- or beam-
modeling technique and response spectrum modal analysis as well as developing in-structure
response spectra using two- or three-dimensional shell or plate models in three-dimensional
space as the basis for modeling and analyzing safety subsystems.
By the mid 1970s two- or three-dimensional shell or plate models were recognized as im-
portant to consider for unsymmetrical buildings and components, resulting in enhancements of
the traditional stick- or beam-models to handle the summation of the co-directional responses
from all three orthogonal earthquake excitation directions.
The first SSI-technique, using simple one-dimensional linear springs gave way to develop the
soil impedance function method in the 1980s. Since the 1990s finite element modeling of the
foundation media, coupled with the building model, has been developed and implemented in
various computer software products.
In contradiction to static loadcases when load values are determined independent of the math-
ematic model of the structure, the magnitudes of the seismic loads are dependent on the dy-
namic properties of the structural system being modeled. This means that the requirements on
the structural model and analysis must be more rigorous when dealing with seismic analysis
compared to conventional static analysis. This simple fact, in combination with the develop-
ment of more sophisticated 3D finite element models during the 1980s and 1990s, boosted the
need for more consistent requirements on the seismic analysis technique.
The USNRC released its first versions of SRP 3.7.1 [25] and 3.7.2 [26] regarding seismic sys-
tem design parameters and seismic system analysis in 1972 with revisions in 1981, 1989 and
2007 and the first versions of RG 1.92 [27] in 1976 for combining modal responses and spatial

2
For typical seismic hazard curves the mean 10 -4//yr. equals the median 10-5//yr. proba-
bility of exceedence.

20
SSM 2014:56
components in seismic analysis, as well as RG 1.122 [28] in 1978 regarding acceptable proce-
dures for developing in-structure response spectra. The first version of ASCE 4-98 Standard
[24] was released in 1986 and included requirements for modeling and analysis of safety-
related nuclear structures subjected to earthquake motions.

2.2.4 Seismic design classification

Seismic design classification for NPPs is addressed in RG 1.29 [29]. Those SSCs which have
to be designed to withstand the effects of an SSE, are specified and designated as Seismic Cat-
egory 1.
In RG 1.143 [30], in reference to radioactive waste stored at an NPP site, there are two safety
classes identified; Safety Class RW-IIa, RW-IIb and a Non-Safety Class RW-IIc. These Safety
Classes are a function of the radio nuclides and their quantities stored to include their gasous
liquid or solid forms and are described in RG 1.143 [30]. The off-site radiological release crite-
ria are set to 5 millisieverts (mSv) per year or an on-site dose of 50 mSv per year resulting
from a postulated failure requiring a RW-IIa classification and any value of radiation release
less than these values requiring Safety Class RW-llb.
The seismic load applicable to RW-IIa SSC is one half of the SSE load specified for Seismic
Class I SSC as shown in Table 2 in RG 1.143 [30]. The seismic load specified for RW-llb and
RW-llc are found in the ASCE 7-10 [32] Standard Risk Categories III and II respectively.
In RG 1.29 [29] there are references to SSE for fire protection system design in RG 1.189 [31],
where the following statement is made on page 54: “The fire suppression systems should retain
their original design capability for (1) natural phenomena of less severity and greater frequency
than the most sever natural phenomena (approximately once in 10 years) such as tornadoes,
hurricanes, floods, ice storms, or small-intensity earthquakes that are characteristic of the geo-
graphic region.” This statement would result in an earthquake stated load smaller than speci-
fied for conventional SSCs by ASCE 7-10 [32].
The USNRC has recognized the need to be less prescriptive in the designation of Seismic Cat-
egory I and relates such categories to safety-related and safety-significant risk categorization as
contained in Figure 1 in RG 1.201 [33], as shown in Figure 2.3, as an alternative to the RG
1.29 [29] deterministic seismic classification of SSCs which did not consider the level of risk
associated with a particular SSC failure.
In 10 CFR 50.69 [34], the application of safety-related and safety-significant categorization to
NPPs is explained. To date this procedure for risk informed reclassification of SSC has been
attempted on a trial basis related to maintenance, testing and examination activities for SSCs in
a small number of existing NPPs. There is an active effort in the ASME B&PVC Section III
Nuclear Component Design Code Committees to extend risk informed categorization to design
activities. But to date, definitive design criteria based on risk informed assessment categoriza-
tion has not been developed and it is expected it will take several years more.

21
SSM 2014:56
Figure 2.3 –Figure depicting the current safety-related versus nonsafety-related SSC cat-
egorization scheme with an overlay of the new safety-significance categoriza-
tion, according to [33].

2.2.5 Codes and standards

ACI 318 [35] prescribes minimum requirements for all types of ordinary concrete buildings in
the U.S. In general, the structural form consists of moment resisting frames designed for an
essentially elastic response for all loads and load combinations except those associated with
strong earthquake motions. ACI 318 [35] permits a seismic design based on loads correspond-
ing to an inelastic response to earthquake ground motions. In order to secure that the structural
elements can exhibit inelastic behavior during the translational earthquake motions, ACI 318
[35], chapter 21 provides minimum requirements on the reinforcing steel detailing.
ACI 349 [10] provides requirements for design of safety-related nuclear concrete structures.
The predominant structural form is shear wall and slab construction of geneal heavy propor-
tions. The structural elements are designed for an elastic behavior for all loads (except impul-
sive and impactive loads) and load combinations applicable to structures, distribution systems
or containments in Limit State D according to ASCE 43-05 [23] or Seismic Category 1 accord-
ing to RG 1.29 [29] including those associated with the DBE. The main reason to the choice of
structural form and the elastic design principle is of course to ensure a robust design with large
safety margins for SSC which provide for reactor safety and shutdown and spent fuel storage.
It should be noted that SSC categorized in Limit State A, B or C according to ASCE 43-05
[23] or in Safety Class IIa and IIb according to RG 1.143 [30] are allowed to respond inelas-
tically. But, as indicated in section 1.1 as well as Table C1-1 in ASCE 43-05 [23], SSCs in
modern NPPs fall, almost exclusively, in the highest Seismic Design Category (SDC 5), hence
being associated with an “essential elastic behavior” for the DBE event.
Even though ACI 349 [10] requires safety-related nuclear structures to be designed essentially
elastic to earthquake loads, it provides minimum requirements for reinforcing steel detailing
according to the requirements of chapter 21 in ACI 318 [35]. Besides maintaining the maxi-
mum possible compatibility between ACI 349 [10] and ACI 318 [35], the main reason for this
approach is to provide additional assurance that structural integrity is maintained in the unlike-
ly event of an earthquake beyond the design basis event DBE.
For mechanical components and distribution systems, elastically computed stress are used
when the allowable stress is typical in the range of 1.6 and 2.0 times specified minimum yield
stress.

22
SSM 2014:56
For electrical distribution systems and components generally follow the civil engineering ac-
ceptance criteria for building (i.e. specified minimum yield stresss).

2.2.6 Consequences of Fukushima Earthquake on Design Basis Earth-

quake

2.2.6.1 Introduction
Following the severe accidents at the Fukushima Dai-ichi NPP, the USNRC decided to make
additional improvements to its regulatory system in order to enhance the protection against
accidents resulting from natural phenomena, mitigating the consequencies of such accidents
and ensuring emergency preparedness. The USNRC’s review of insights from the Fukushima
Dai-ichi accident resulted in recommendations for enhancing the reactor safety as reported in
[36].
Even though current USNRC regulations and associated regulatory guidance provide a robust
regulatory approach for evaluation of site hazards asscociated with natural phenomena, this
framework has evolved over time as new information regarding site hazards and their potential
consequences has become available. As a result, the licensing bases, design, and level of pro-
tection from natural phenomena differ among existing operating reactors in the U.S., depend-
ing on when the plant was constructed and when the plant was licensed for operation. Over the
years the USNRC has initiated several efforts to evaluate risks and potential safety issues re-
sulting from these differences. However, the USNRC has not yet undertaken a comprehensive
reestablishment of the design basis for existing plants that would reflect the current state of
knowledge of current licensing criteria. As a result, significant differences may exist between
plants in the way they protect against design-basis natural phenomena and the safety margin
provided.
With regard to seismic hazards, available seismic data and models show increased seismic
hazard estimates for some operating nuclear power plant sites, as reported in [36]. The state of
knowledge of seismic hazards within the U.S. has evolved to the point that it would be appro-
priate for licensees to reevaluate the designs of existing nuclear reactors to ensure that SSCs
important to safety will withstand a seismic event without loss of capability to perform their
intended safety function. As seismic knowledge continues to increase, new seismic hazard data
and models will be produced. Thus, the need to evaluate the implications of updated seismic
hazards on operating reactors will recur and need to be reevaluated at appropriate intervals.
In order to ensure adequate protection from natural phenomena, consistent with the current
state of knowledge and analythical methods as above, the USNRC initiated a number of ac-
tions, as reported in [36] and presented in Table 2.1. The outcome of Seismic Recommendation
2.1 and 2.3 are reported in the following sections.

23
SSM 2014:56
Table 2.1 – USNRC’s Recommendation 2 to enhance the reactor and spent fuel safety in
the U.S [36].

Recommendation
2.1 Order licensees to reevaluate the seismic and flooding hazards at their
sites against current NRC requirements and guidance, and if necessary,
update the design basis and SSCs important to safety to protect against
the updated hazards.
2.2 Initiate rulemaking to require licensees to confirm seismic hazards and
flooding hazards every 10 years and address any new and significant in-
formation. If necessary, update the design basis for SSCs important to
safety to protect against the updated hazards.
2.3 Order licensees to perform seismic and flood protection walkdowns in
accordance with their licensing basis to identify and address plant-
specific vulnerabilities and verify the adequacy of monitoring and
maintenance for protection features such as watertight barriers and seals
in the interim period until longer term actions are completed to update the
design basis for external events.

2.2.6.2 Recommendation 2.1: Seismic

The activity entitled Recommendation 2.1: Seismic was required by the USNRC to seismically
re-evaluate and potentially increase the seismic resistance of existing NPP. This included de-
velopment of a revised probabilistically based seismic hazard at the mean 10-4/yr. probability
of exceedence level to be instituted for each U.S. NPP site. This hazard for the various plant
sites has involved a complete re-evaluation of the earthquake level to be considered as a func-
tion of the magnitude of the probabilistic defined earthquake hazard and its motion attenuation
with distance from its epicenter or focus to the NPP site.
This re-evaluation and development of the seismic hazard earthquake is termed the Review
Level Earthquake (RLE) and is currently underway. It is expected that both the shape of the
ground response spectra and the PGA will change significantly from the spectral shape and
PGA used as the plant seismic design licensing basis, i.e., higher acceleration at frequencies
above 10Hz. This is particularly true for the Central and Eastern U.S. sites.
Also associated with this Recommendation 2.1: Seismic program will be the development of
fragility curves for nuclear safety-related SSC. In the development of these fragility curves it is
understood that the SSCs that are ductile will be allowed to respond into the inelastic range
beyond yield of the SSC’s material and that revised capability may be used in developing the
individual SSC fragility curves.
Guidance for the development and performance of the Recommendation 2.1: Seismic program
is contained in [37]. As part of the Recommendation 2.1: Seismic program is the so called Flex
Program. This program provides for re-evaluation of existing installed SSC and for the pro-
curement of additional SSC which are intended to mitigate or reduce seismic risk.

2.2.6.3 Recommendation 2.3: Seismic

The near-term program activity Recommendation 2.3: Seismic consisted of a review of the
seismic design adequacy on a walkdown and walkby basis (evaluation based on the existing
licensing basis for the NPP plant) which was completed based on existing outage schedules by

24
SSM 2014:56
November 201213. This was accomplished by a physical walkdown and walkby of a sample of
nuclear safety related mechanical and electrical components (approximately 120 items) in-
stalled in the NPP. The Seismic Evaluation Guidance [38] used to implement this program
was developed by EPRI and endorsed by the USNRC.
It should be noted a similar review of the flooding hazards, from all causes not just seismic, at
each NPP site was also performed by review of current design basis flooding hazard levels,
flood protection procedures and a physical plant walkdown of flooding mitigation and preven-
tion structures systems and components.

2.2.6.4 Beyond Design Basis Earthquake

The USNRC has developed a requirement to evaluate a BDBE to assure that there is margin
available to resist seismic load without failure of the SSC to perform required safety functions.
The load selected is 1.67 times the DBE load. It should be understood that for most of North
America the slope of most seismic hazard curves currently under review has an increase in
acceleration of between 1.5 and 2.0 for each doubling of the return period in the 104 to 105
mean return period range in years. As a result of the NRC’s 1.67 multiplication factor the
BDBE probabilistically defined mean return period would be increased from 10 000 years to
approximately 20 000 years.
Associated with this increase in seismic acceleration would be an increase in the applied ac-
ceptance criteria into the inelastic range beyond yield for ductile type structures. However,
specific acceptance criteria into the inelastic range has not yet been published by the USNRC.
ASCE 43-05 [23] provides acceptance criteria into the inelastic range for ductile SSC in SDC 3
to SDC 5. SDC 1 and SDC 2 category SSC are equivalent to USNRC Safety Class RW Class
IIc and IIb respectively as defined in RG 1.143 [30] and are designed to the loading require-
ments of ASCE 7-10 [32] for natural hazard loads and ACI 318 [35] for concrete structures.
It should also be observed that Section 1.1 as well as Table C1-1 in ASCE 43-05 [23] indicates
that SSCs in modern NPPs fall, almost exclusively, in the highest Seismic Design Category
(SDC 5) and is associated with an “essential elastic behavior” for design basis acceptance crite-
ria.
From ASCE 43-05 [23] for elastic analyses and ductile structures in other than the limit state D
category for SDC 3 to SDC 5, the total limiting capacity for an element shall be the yield stress
or design code ultimate strength equal to or greater than the sum of non-seismic demand, DNS,
and seismic demand, Ds, per the following load combination, as appropriate:
For bending moment, in-plane shear, and axial load in diagonal bracing:
U ≥ DNS + DS/ FµS
For other axial loads, other shear loads, and torsion:
U ≥ DNS + DS/ 1.0
where
U = Ultimate strength or specified minimum yield stress
DNS = Non-seismic demand acting on an element. Non-seismic demand shall include the code
effects of dead, live, equipment, fluid, snow and at-rest lateral soil loads.

3
The Program extends beyond November 2012 for SSC which were not available for
walkdowns because outages were not scheduled between July and November 2012.

25
SSM 2014:56
Ds = Calculated seismic response to the DBE using an elastic analysis approach
Fµs = System inelastic energy absorption factor for structural elements according to Table 5-1
in ASCE 43-05 [23].
The USNRC has not yet provided any acceptance criteria for Beyond Design Basis Event.

2.3 France

2.3.1 Introduction
The regulatory hierarchy of the safety requirements for safety-related nuclear structures in
France can be described as follows:
- Laws of the French republic, to be passed by the French Parliament. The TSN (Trans-
parancy and Security in Nuclear field) Law of June 2006 unifies previous laws and de-
crees, complemented with several decrees providing application details. According to
the TSN Law, ASN is declared as an independent regulatory authority and responsible
for technical and regulatory decisions, licensing and control of nuclear facilities, public
information, management of emergency situations and advices to the French Govern-
ment.
- Decrees, departmental orders and ASN decisions. Some examples are the environmen-
tal RTGE decree, the Regulation about working conditions and protection of workers
health and the Decree regarding quality assurance in nuclear activities.
- RFS fundamental safety rules, ASN guides and Technical directives. The RFS and ASN
guides are issued by ASN and define technically acceptable practice.
- Codes & Standards for design and construction. Typical French Codes & Standards for
civil structures are the RCC-G covering the existing French NPPs and the ETC-C for
EPR NPP design.
Following an earthquake, the objective of the protection of a NPP is to ensure that the safety
functions needed to return and maintain the plant to a safe shutdown state are not unacceptably
affected.
The SSCs required to achieve the safety objectives must be subject to seismic classification.
SSCs necessary for the safety must be designed so that they are able to fulfil their functions,
maintain their integrity or remain stable under the conditions caused by the seismic ground
motions.
The basic steps in the earthquake design process of safety-related nuclear structures together
with the applicable regulations, guidelines and design codes in France can be categorized as
follows:
a. Determine the soil characteristics at the NPP site according to RFS-I.3.c (1984) [39].
b. Determine the seismic design ground motions according to RFS 2001-01 (2001) [40].
c. Seismic modeling and analysis:
o Existing facilities according to RFS V.2.g (1985) [41].
o New facilities (except EPR) according to ASN/Guide/2/01 (2006) [42].
o EPR according to ETC-C, Appendix A.1 (2010) [43].
d. Seismic design and safety evaluation:

26
SSM 2014:56
 o Existing facilities according to RCC-G, volume 1-Design (1981 and 1985) [44].
o EPR according to ETC-C, Part 1 Design (2010) [43].
The earliest requirements for evaluating the seismic hazard in France was published in 1981 in
RFS I.2.c [45]. These requirements were replaced by ASN in 1985 through RFS V.2.g [41]. A
comprehensive review of RFS V.2.g [41] in 2006, resulted in new guidelines in
ASN/Guide/2/01 [42]. These guidelines define the seismic design requirements and acceptable
methods for civil works. The requirements for determining the seismic design ground motions
are specified in RFS 2001-01 (2001) [40].
The development of the RCC-G [44] began in 1976 when it was decided to establish a working
group involving EDF, FRAMATOME-CEA and the French Ministry of Industry, piloted by
EDF, to examine the possibility of issuing detailed documents (initially called ''Codes and
Standards'' and then, from 1978, ''Rules of Design and Construction”) with the following ob-
jectives:
- To serve as a basis for contractual relations between licensees and suppliers
- To facilitate discussions with nuclear safety authorities
The ETC-C [43] is an evolutionary development of the RCC-G [44]. It was undertaken for the
design and construction of EPR safety-classified buildings. The reasons for developing ETC-C
[43] were as follows:
- It was necessary for the EPR to comply with requirements from both French and Ger-
man regulations and practices
- New load cases were required to represent severe accident and more severe hazard con-
ditions
- Changes were needed to take into account the Eurocodes in the design of structures
- Updated operational experience feedback as well as current updated safety analysis re-
quirements had to be taken into account
- Updated knowledge of material and structure behaviour from laboratory and mock-up
tests had to be incorporated
A previous edition of the ETC-C [43] was issued by EDF in April 2006 and serves as a refer-
ence document for the Flamanville 3 project. Since 2009, the ETC-C [43] development has
continued under the lead of AFCEN resulting in revised editions in 2010 and 2012.
In ETC-C [43], the safety requirements are achieved through various specifications as regards
analysis methods or criteria, such as: linear analysis, requirements to limit cracking in concrete
structures, limitation on strains in materials, etc.

2.3.2 Historical development of seismic design ground motions

2.3.2.1 Seismic design ground motions for existing plants

There are in total 58 nuclear power units in operation in France. All of these are of PWR-type
and were designed by Framatome. There are three major standard types of designs.
- CP0 and CPY design types (900 MWe). There are in total 34 units still in operation.
They were designed and constructed in the 1970s and beginning of the 1980s.

27
SSM 2014:56
 - P4 and P’4 design types (1300 MWe). There are 20 reactors of these design types in op-
eration in France. They were designed and constructed during the late 1970s, the 1980s
and beginning of the 1990s.
- N4 design type (1400 MWe). Of this type, there are 4 reactors in operation in France.
They were designed and constructed during the 1980s and the beginning of the 1990s.
Accordingly, the design of the existing French nuclear fleet were carried out mainly during the
1970s and 1980s, before any consistent framework of requirements as regards seismic ground
motions was established. Therefore the seismic design basis were different for the different
design types as follows:
- CP0 and CPY: For the design of the CP0 and CPY plant series, the spectral shape used
was that known as the "EDF spectrum", defined as the smoothed mean of eight accel-
erograms recorded during five earthquakes of Californian origin. The accelerations
were normalized according to the local seismicity PGA.
o CP0: Bugey EDF spectrum anchored at 0.1 g PGA and for Fessenheim at 0.2g
associated to local soil conditions.
o CPY: EDF spectrum anchored at 0.2 g PGA associated to a range of soil’s con-
ditions: 500 to 2000 MPa in terms of dynamic young modulus for the soil.
- P4 and P’4: The DBE for Paluel, the first P4 site, was changed during the course of its
construction. At the beginning of construction in the late 1970s, the spectral shape
used hitherto for the units was that of the "EDF spectrum". Later during construction, a
new spectral shape was taken from that established by the USNRC in RG 1.60 [15],
which was also adopted in France as the reference for the design of the 1300 MWe
plant series. For the following P4 and P’4 reactors, EDF adopted the RG 1.60 [15]
spectrum, normalized to 0.15 g ZPA as the standard DBE applicable to nuclear island
design, compatible with the sites chosen for the reactors in this plant series. For the
buildings, this led EDF to use the following in turn:
o For a transitional period, the EDF spectrum anchored at 0.2 g PGA associated
with a range of soil conditions: 500 to 15000 MPa in terms of dynamic young
modulus for the soil.
o The RG 1.60 [15] spectra anchored at 0.15 g PGA associated with a range of
soil conditions: 500 to 15000 MPa in terms of dynamic young modulus for the
soil.
- N4: For the DBE, the RG 1.60 [15] spectrum anchored at 0.15 g PGA was applied at
Civaux site and RG 1.60 [15] spectrum anchored at 0.12 g PGA was applied at Chooz
site associated with local soil conditions

2.3.2.2 Approach in the Basic safety rules

In 2001 more consistent requirements with regard to seismic design ground motions were es-
tablished in the Basic Safety Rules in RFS 2001-01 (2001) [40]. The main objectives with this
document were to:
- Ensure that safety-related functions being maintained during and after plausible earth-
quakes that could affect nuclear installations.
- Define acceptable methods for determining the vibratory ground motions to be consid-
ered in the seismic design basis.

28
SSM 2014:56
The basic principles of the approach in RFS 2001-01 (2001) [40] can be summarized as fol-
lows:
- The approach is basicly deterministic and assuming that earthquakes comparable to his-
torically known earthquakes are liable to occur in the future.
- The definition of the characteristics of “Maximum Historically Probable Earthquakes
(MHPE) considered to be the most damaging earthquakes liable to occur over a period
comparable to the historical period of approximately 1000 years.
- The definition of a SSE to account for uncertainties in MHPE, which may be comple-
mented by paleoseismological evidences.
For an envisaged site, an intensity I(MHPE) is determined. In order to take account of uncer-
tainties inherent in the determination of the MHPE characteristics, a fixed safety margin is
defined as follows. For each MHPE, a SSE is defined, deduced from the MHPE by the follow-
ing simple equation in terms of intensity (I)4 on the site:
I (SSE) = I(MHPE)+1
Except for the particular case when the site is located in the immediate vicinity of an active
fault with surface fractures, the SSE are considered as the most aggressive earthquake to be
included in the design basis. The SSE can be preceded or followed by earthquakes capable of
reaching the MHPE level.
Seismic motion is defined by the response spectra of the horizontal and vertical components of
the motion on the surface of the site ground.

2.3.2.3 Seismic design motions for future plants

The design and qualification of SSCs in future plants, such as for instance the EPR, shall con-
sider the EUR standard design spectra in Figure 4 in [46] as shown in Figure 2.4 scaled to 0.25
g horizontal ground acceleration. These standard design spectra can be used under condition
that the range of soil characteristics forming the basis of the EUR spectra envelope the specific
site soil conditions at the facility. Soil property data for soft, medium and hard soil conditions
can be found in section 2.4-6.4.2.1 in [46]. The input ground motion shall be represented by
either response spectra or artificial time histories based on a damping value of 5 %. The earth-
quake excitations shall be represented by two horizontal and one vertical input motions simul-
taneously. Additional requirements regarding seismic ground motions and seismic analysis are
described in Appendix A.1 of ETC-C [43].

4
Intensity scales measure the amount of shaking at a particular location

29
SSM 2014:56
Figure 2.4 – EUR design basis ground motion spectra (horizontal, 5% damping) [46]

2.3.3 Evaluation of seismic safety margins

2.3.3.1 General
The methodology used in France to assess seismic hazards is based essentially on a determinis-
tic approach. The most penalising historical event, MHPE, is considered, to which large mar-
gins are added as described in section 2.3.2.2. This approach is supplemented by probabilistic
safety assessments (PSA), based on a systematic investigation of the different accident scenar-
ios to determine the probability of unacceptable consequences.
PSA as regards external events have not yet been applied systematically at the French NPPs for
external events. Seismic Probabilistic Safety Assessment (SPSA) studies were performed by
EDF for the St Alban, Flamanville and Civaux NPPs. The SPSA of the St Alban site was de-
veloped within the scope of the third PSR of the 1300 MWe series (P4/P’4-series), and EDF’s
conclusion was that the PGA return period of the DBE obtained by EDF is equal to about
10 000 years.
In addition, EDF has incorporated some of the seismic safety operating experience feedback
from the July 2007 earthquake at the Japanese NPP at Kashiwasaki-Kariwa, with especial con-
sideration of defining the scope of seismic inspections and studying the consequences of a
transformer fire.
The external hazards are reassessed periodically in the period safety reviews performed every
10 years. Moreover, the external hazards, particularly earthquakes and flooding, were the sub-
ject of a targeted reassessment as part of the European stress tests in 2011.

2.3.3.2 Approch for seismic margin assessments

Within the scope of the European stress tests, EDF identified three margin sources for its over-
all margin study:
- Margins between MHPE and SSE and between SSE and DBE
- Margins due to the response of the structure
- Margins due to the design criteria for the structures and equipment

30
SSM 2014:56
A specific study was conducted for the Tricastin site (900 MWe), applying the SMA-method.
In addition to the SMA-approach, EDF performed a SPSA for the St Alban site (1300 MWe),
as mentioned in section 2.3.3.1. For the next PSR of the 900 MW plants it is considered to
perform a robustness analysis based on a seismic PSA or SMA.
On the request of ASN, EDF performed some complementary safety assessments on the basis
of the available information as follows:
- To give an evaluation of the level of earthquake beyond which the loss of fundamental
safety functions or fuel damage (in vessel or pool) are inevitable.
- To identify weak points and cliff-edge effects.
- To propose measures to prevent these cliff-edge effects and reinforce the robustness of
the facility.
EDF reviewed the seismic strength margins of the structures and equipment important to safe-
ty, in order to determine the level of acceleration for which, with a high level of confidence,
the facility has a very low probability of failure (HCLPF).
EDF supplemented its general study with studies of equipment for which there could be per-
formance discontinuities, together with proposed modifications or reinforcements.
Finally, EDF carried out the seismic inspection of a sample of the equipment needed to operate
the unit in the event of total loss of off-site and on-site power supplies, whether or not seismic-
classified, for all the NPPs in service.

2.3.3.3 Seismic margin assessment of existing plants

EDF used a plant series response spectrum as the design spectrum for all the reactors of the
same series. For each site, EDF proposed a table of margin factors between the reassessed site
SSE and the DBE, between 1 Hz and 6 or 10 Hz. EDF adopted margin values between 1 and
1.7 depending on the sites and the buildings considered.
EDF concluded that the seismic capacity of the containment and equipment which, in the event
of failure, would compromise the safety functions, is at least 1.5 times larger than the forces
and stresses resulting from the SSE. EDF considered that this level easily exceeds the seismic
context of the sites, up to hazard values that are hardly plausible or implausible for these sites.
A robustness study regarding significant damage of the fuel assemblies identified no failure of
the necessary safety systems for a hazard equivalent to 1.5 times the SSE.
Beyond the current safety requirements, EDF proposed additional measures to prevent the
serious consequences of extreme situations, on a deterministic basis, regardless of their plausi-
bility. EDF proposed defining a hard core of reinforced equipment aimed to prevent severe
accidents and to avoid significant radioactive releases into the environment, over and above the
current safety requirements, for the deterministic situations studied in the complementary safe-
ty assessments, as described in section 1.1.3.

2.3.3.4 Seismic margin assessment of new plants

The EPR was originally designed to withstand the DBE according to EUR [46]. This DBE is
defined by standard design acceleration and a set of seismic spectra reflecting the European
conditions.
- A set of three European DBE spectra is given for “hard”, “medium”and “soft”soil, as
shown in Figure 2.4.

31
SSM 2014:56
 - A standard design acceleration level of PGA = 0.25 g is required for these three spectra.
- It shall be demonstrated that the standard plant remains in a safe condition for the whole
range of parameters (soil conditions) for the DBE.
An Operating Basis Earthquake (OBE) has not been specified in the EUR [46].
The standard design of the EPR is not intended to envelop all possible combinations of nation-
al regulations and site conditions. If necessary, the designer may have to make modifications
or additional studies to ensure the standard design is satisfactory for particular sites. The SSE
is usually defined in accordance with international practice for an annual probability of ex-
ceedance of 10-4 (i.e. a return period of 10 000 years)
A site-specific seismic margin analysis of the structures and equipment shall be carried out, to
ensure that adequate safety margins exist in the seismic design of the main structures and com-
ponents beyond the design basis conditions.
Hence, it is necessary to choose a Review Level Earthquake (RLE). According to EUR [46],
the design shall withstand potential earthquakes with a margin of 40 % on the horizontal PGA
above the design SSE level. At present, a 60 % margin is targeted for the EPR. The objective
shall be to establish the seismic capability of a minimum set of plant structures and systems
needed to avoid core damage, then bring the plant to and maintain it in a safe shutdown state.
This demonstration shall be made following a best-estimate methodology. The assessment
shall identify the items without sufficient margins in the capacity of the design. For items
without sufficient margins, a comprehensive SMA program shall be established using analyses
and tests.
The SMA-RLE is defined for a lower annual probability of exceedance than the SSE, e.g.10-5.
By this approach it is demonstrated that there are no cliff-edge effects from earthquakes slight-
ly exceeding the design basis.

2.3.3.5 Seismic margin assessment for UK-EPR

The seismic margin of the UK-EPR is assessed by a PSA-based SMA, following a methodolo-
gy developed by the USNRC. This approach uses the PSA model to identify combinations of
seismic equipment failures which could result in core damage, as well as combinations of
seismic failures, random failures and human errors which contribute significantly to seismic
risk. By identifying which equipment items and structures are of critical importance in seismic
events, the analysis approach ensures that vulnerabilities in the design are identified allowing
them to be corrected if necessary, thus helping ensure that the seismic risk is ALARP.
The UK-EPR is designed to resist an enveloping seismic event which is bounding for NPPs
constructed in Western Europe, with a Design Basis Earthquake (DBE) corresponding to a
0.25 g peak ground acceleration in the horizontal direction, according to Figure 2.4. SSCs re-
quired for controlling and mitigating accidents are designed so that they are able to fulfil their
functions, maintain their integrity or remain stable under the conditions caused by the seismic
motion.
The detailed PSA-based SMA is performed for at-power states. A simplified approach is used
for shutdown states. Also, at this first stage, internal hazards that might be caused by a seismic
event, such as fire or flooding, are not analysed in detail and are not included in the PSA model
supporting the SMA.
The purpose of the SMA is to show that the SSCs critical to achieving a safe shutdown state
following an earthquake are designed with large safety margins so that they have a low proba-
bility of failure in the Review Level Earthquake (RLE), which has a PGA of 1.6 times that
assumed for the DBE. The first step in this process is to define the ground motion spectrum for

32
SSM 2014:56
calculating the seismic capacities (fragilities) of the SSCs. The ground motion spectrum is a
characteristic of the EPR site in question.
For the purposes of the SMA, the free field ground motion spectra used as input data for the
estimation of seismic capacities of equipment and structures, are a bounding ‘hard’ and ‘soft’
site spectra derived by enveloping Uniform Risk Spectra (URS) for prospective UK new build
sites.
To perform the SMA, it is necessary to produce a Seismic Equipment List (SEL) containing
the SSCs whose seismic capacities need to be evaluated for the SMA. The seismic fragility
analysis for these SSCs is then performed.
The SEL for the UK-EPR SMA is developed using expert judgements in combination with the
Level 1 PSA model. The use of the PSA model to identify critical combinations of component
failures serves to confirm the completeness of the SEL.
The fragility assessment of the SSC items in the SEL evaluates the PGA at which their re-
sponse will exceed a threshold of acceptability for the characteristic motion spectrum adopted.
As noted above, the motion spectrum depends on the ground conditions.
The fragility assessment of the SSCs considers the capacity to withstand ground motion of
each component and its associated uncertainties. The capacity is defined as the free field PGA
value for which the seismic response at the component location exceeds the component re-
sistance capacity, resulting in the probability of failure of the SSC, i.e. the probability that the
response exceeds a defined threshold. The PGA capacity of the SSCs is estimated using infor-
mation on the plant design and ground parameters, test data from SSC qualification and fragili-
ty tests, data from generic seismic tests, earthquake experience results, material property data,
etc. Where identical components occur in different redundant trains, the seismic capacity of all
the components is conservatively set to that of the most vulnerable component, taking no bene-
fit for the redundancy of the system. Similarly the seismic capacity of electrical cables is as-
sumed to be the capacity of the cable tray anchorages, which is conservatively set to the capac-
ity at the most seismically vulnerable anchorage location.
The SMA shows that the UK-EPR could tolerate a seismic event with 0.61g PGA without sig-
nificant risk of a severe accident and release of radioactivity from the plant. Therefore, it has
been demonstrated that the seismic capacity of the UK EPR is higher than the RLE defined as
1.6 times the Design Basis Earthquake (i.e. corresponding to 1.6·0.25g = 0.4g PGA). The SMA
has shown that there are no cliff edge effects for seismic events with magnitudes above that
assumed in the design basis.

2.4 Sweden

2.4.1 Introduction
The Swedish Radiation Safety Authority (SSM) superintend the Swedish nuclear facilities as
well as all other activities in the country within the areas of nuclear safety, radiation protection
and nuclear non-proliferation. The SSM reports directly to the Ministry of the Environment.
The hierarchy of requirement in Sweden in order to regulate the design and analysis of NPPs
can be categorized as follows:
- The Act on Nuclear Activities (Kärntekniklagen) and the Radiation Protection Act
(Strålskyddslagen). These are laws passed by the Swedish Parliament.

33
SSM 2014:56
 - SSM Regulations (SSMFS). Regulations issued by SSM. These regulations are manda-
tory with the intention to provide more specific requirements on how to implement the
Act on Nuclear Activities and the Radiation Protection Act.
- General recommendations to SSMFS. These general recommendations are issued by
SSM. They are not mandatory but provides recommendations on the authority’s view
on the compliance with the Laws and the Regulations (SSMFS).
- Codes & Standards for design and construction. The most important standard for nu-
clear structures is applicable parts of the Eurocodes [52].

2.4.2 Historical development of the seismic design basis

The seismic activity in Scandinavia is low. Historically there are only a few registered events,
which might have caused damage to an industrial facility [47]. Thus, earthquake load effects
on buildings in Sweden have been regarded as negligible compared to other loads to be ex-
pected during the lifetime of a building. Accordingly, the design criteria for the oldest Swedish
nuclear power facilities in the beginning of the 1970s did not include any requirements on
structural integrity or maintaining safety functions due to earthquake ground motions.
Concurrently with an increasing safety consciousness in the nuclear power industry, there was
an increased understanding that seismic effects must be regarded for the Swedish nuclear facil-
ities. Thus, the design criteria for the latest Swedish NPPs, Forsmark unit 3 and Oskarshamn
unit 3, designed at the end of the 1970s included seismic load requirements. Due to lack of
statistical data regarding larger earthquakes in Scandinavia, design response spectra anchored
at 0.15 g PGA horizontally and 0.10 g vertically according to USNRC RG 1.60 [15] were ap-
plied.
With the purpose of deriving ground motions to be used in the safety analysis of the Swedish
nuclear power facilities, a joint project was initiated in the mid 1980s between the then Swe-
dish nuclear safety authority (SKI) and the Swedish nuclear power licensees. The project re-
sults are presented in SKI Technical Report 92:3 [8]. In this report, envelope ground response
spectra corresponding to a certain annual probability of exceedance (10-5, 10-6 and 10-7) are
defined for typical Swedish hard rock sites.
SS-EN 1998-1:2004; Eurocode 8 [48] has been in force since 2009 and applies to design and
construction of buildings and civil engineering works in seismic regions in Europe. The main
purpose with this standard is according to SS-EN 1998 [48], section 1.1.1 to protect human
lives, to limit damages and to secure that structures important for civil protection remain opera-
tional.
It must be observed that SS-EN 1998 [48] only includes complementary reuirements in addi-
tion to the requirements of the other relevant Eurocodes, to be applied for the design of struc-
tures in seismic regions. In this respect, SS-EN 1998 [48] is a complement to the other Euro-
codes.
An important limitation with SS-EN 1998 [48] is that it, as for other parts of the Eurocodes,
formally does not apply to nuclear power plants, offshore structures and large dam structures.
According to SS-EN 1998 [48], structures in seismic regions are recommended to be designed
and constructed to withstand a design seismic action associated with a reference probability of
exceedance of 10% in 50 years or a reference return period of 475 years. The hazard is de-
scribed in terms of the value of the reference PGA, which may be derived from zonation maps
found in National Annex.

34
SSM 2014:56
According to the Swedish National Annex to SS-EN 1998 [48], the Swedish National Board of
Housing, Building and Planning (Boverket) has not found it necessary to issue any regulations
or recommendations with regard to seismic load effects, since SS-EN 1998 [48] would only be
used in very specific cases where special expertise is required. The Swedish Transport Admin-
istration (Trafikverket) specifies as well its position in the Swedish National Annex to SS-EN
1998 [48]. Trafikverket states that earthquake loads do not need to be considered in Sweden,
since the other parts of the Eurocodes normally ensure the strength and durability of the struc-
ture for those earthquake hazard levels that could arise in Sweden.
One fundamental question is what values of the reference PGA are applicable in Sweden? In
[53], a hazard map corresponding to a mean return period of 475 years, i.e. the recommended
reference period in SS-EN 1998 [48], is presented for Fennoscandia, as shown in Figure 2.5.
From this map the horizontal PGA for the different NPP sites in Sweden can be identified,
according to Table 2.2. As can be noticed in Table 2.2, the maximum horizontal PGA for a
mean return perod of 475 years is 0.15 m/s2 for Forsmark and Oskarshamn NPP and 0.25 m/s2
for Ringhals site. Maximum PGA in Sweden arises in the territory inbetween Lake Vänern and
the Norwegian border where it can reach values of 0.35 m/s2.
According to SS-EN 1998 [48], a low seismicity case is defined as a region having a design
PGA not greater than 0.78 m/s2 (0.08g) and a very low seismicity case a region having a design
PGA not greater than 0.49 m/s2 (0.05g). It is also stated that in cases of very low seismicity, the
requirements of SS-EN 1998 [48] need not to be observed. Hence, on the basis of Figure 2.5
and Table 2.2, it can be concluded that the whole Swedish territory can be classified as a very
low seismicity case and the requirements of SS-EN 1998 [48] can be disregarded, as also are
concluded by Boverket and Trafikverket in the National Annex to SS-EN 1998 [48].
However, the safety conditions for safety-related structures at NPP sites in Sweden differ from
the conditions for bridges, conventional structures and industrial facilities. Safety-related struc-
tures at nuclear facilities house important safety systems, which in case of failure could result
in severe and unacceptable consequences for the personnel, the off-site public or the environ-
ment. Hence, safety-related structures at nuclear facilities should be designed against external
and internal hazards with higher reference return period than conventional structures and facili-
ties. SSM has also in SSMFS 2008:17 [49] specifically mentioned earthquake as one of several
natural events that the Swedish nuclear power plants must prove resistance to.
Within the framework of the large modernization and power uprate projects at the Swedish
NPP sites during the recent 20 years, new installed SSCs have been designed and existing
SSCs been evaluated against a DBE corresponding to an annual exceedance probability of 10-5
according to SKI Technical Report 92:3 [8] and shown in Figure 2.6.
Within the scope of the European stress tests, the resistance of the Swedish NPPs were as-
sessed against hazardous natural events with lower annual exceedance frequencies than 10 -5.
Assessments were performed for earthquakes, flooding and extreme weather conditions, as
further described in section 1.1.4.

35
SSM 2014:56
Figure 2.5 – Map of 90% probability of nonexceedence of horizontal PGA (m/s2) in 50
years, corresponding to a mean return period of 475 years, from [53].

Table 2.2 – Maximum reference horizontal PGA (m/s2) corresponding to a mean return
period of 475 years for the Swedish NPP sites.
NPP Site Maximum horizontal PGA
Forsmark 0.15 (m/s2) = 0.015 g
Oskarshamn 0.15 (m/s2) = 0.015 g
Ringhals 0.25 (m/s2) = 0.025 g
Swedish territory 0.35 (m/s2) = 0.035 g

36
SSM 2014:56
Figure 2.6 - Envelope ground response spectra for a typical Swedish hard rock site, cor-
responding to an annual exceedance probability of 10-5.

2.4.3 Seismic design classification

At the Swedish nuclear facilities, those SSCs which are identified to possess safety functions
necessary to be maintained during and after an earthquake, and even SSCs which although not
maintaining safety functions but for which loss of function could jeopardize the function of a
safety-related equipment, shall be designated a Seismic Category. There are three Seismic Cat-
egories (1, P and N) depending on type of safety function, as shown in Table 2.3.

Table 2.3 – Seismic design classification for SSCs at Swedish NPP

Seismic Structures Piping system Pump/valve Electrical com-

Category ponents
1 Leak-tightness Passive function 1) Active function Active function
P Load-bearing Mechanical in- Mechanical in- -
function tegrity tegrity
N No demand 2) No demand 2) No demand 2) No demand 2)
1) Refers for instance to ensure free flow of water or steam.
2) No demand as regards leak-tightness, load-bearing function or mechanical integrity. But SSCs in
Seismic Category N should not challenge any safety function of SSCs in Seismic Category 1 or P.

37
SSM 2014:56
Examples of typically requirements for building elements in respective Seismic Category can
be found in Table 2.4.

Table 2.4 – Examples on various types of requirements on building elements

Seismic Safety Requirement (examples)
Category function
1 Leak- Leak-tightness over the steel liner in the containment vessel.
tightness
Leak-tightness over the steel containment lid (BWR).
Leak-tightness over equipment hatches and openings in the con-
tainment vessel.
Leak-tightness over casing tubes around penetrations in the con-
tainment vessel.
Leak-tightness between the primary and secondary compartment
in BWRs.
Leak-tightness over the steel liner in Spent Fuel Pools (SFP)
Leak-tightness over building elements for protection against leak-
age from vessels in waste buildings.
Leak-tightess of culverts, against leakage from enclosed piping
containing radiactive waste in fluid phase.
P Load- Structural integrity of the load-bearing system.
bearing
To provide support and to shield safety systems and components
function
attached to the structural system.
N No demand No demand as regards leak-tightness, load-bearing function or
mechanical integrity. But SSCs in Seismic Category N should not
challenge any safety function of SSCs in Seismic Category 1 or P.

2.4.4 Codes and Standards

Within the scope of the modernization and power uprate projects at the Swedish nuclear facili-
ties during the 2000s, conditions and evaluation criteria according to DRB:2001 [54] were
generally used. In DRB:2001 [54], conditions and requirements for new design and for evalua-
tion or analysis of existing safety-related structures at NPPs are formulated. The rules in
DRB:2001 [54] were based on the then general Swedish building code, BKR [55], together
with additional conditions and requirements specifically applicable for building structures at
NPPs.
However, since 2011 all building activities in Sweden must comply to the Eurocodes [52], and
the conditions and requirements in BKR [55] and DRB:2001 [54] are no longer applicable. The
statute book of SSM does not yet include specific requirements and adequate guidance on how
safety-related structures at NPPs should be handled in safety analyses of existing structures as
well as in case of the design of new constructions. Therefore, the SSM together with the Swe-
dish licensees commissioned Scanscot Technology AB to further develop DRB:2001 [54] to be
based on the Eurocodes [52] instead. The scope of work was accomplished during 2012 and
2013 and the first version of DNB [51] was released in January 2014.

38
SSM 2014:56
The scope of DNB [51] includes instructions regarding design and analysis of loadbearing
concrete structures, covering reactor containments as well as other safety-related structures.
The main aim with DNB [51] is to complement the requirements in the Eurocodes [52] for
application at NPPs in Sweden. Thus, DNB [51] is based on the partial factor method and the
principles of design in limit states, as specified in the Eurocodes [52] including the National
Determined Parameters chosen by Swedish Authorities.
DNB [51] covers conditions, requirements and acceptance criteria for seismic design and anal-
yses and can be applied for design of new nuclear structures as well as for evaluation of exist-
ing facilities. The scope and application of DNB [51] are described in more detail in section 4
to 6.

39
SSM 2014:56
3. Design basis ground motions
3.1 General
In section 2.4.2 a survey of the historical development of the seismic design basis in Sweden is
described. Recommendations regarding applicable hazard frequencies have been addressed in
the European stress tests of the European NPPs, as described in section 1.1.2 to 1.1.4.
On the basis of the outcome of the survey and the European stress tests as well as applicable
parts of the IAEA Safety Guides, some recommendations will be provided in the following as
regards a reasonable seismic design basis for the Swedish nuclear facilities.

3.2 Design basis considerations

3.2.1 Basic requirements according to IAEA Safety Guides

In section 2 in IAEA SG-G-1.6 [50], it is stated that two levels (SL-1 and SL-2) of ground
motion hazard should be evaluated for an NPP. SL-1 corresponds to a level with a mean annual
exceedance frequency of 1·10-2, while SL-2 corresponds to a mean annual exceedance fre-
quency in the range of 1·10-3 to 1·10-4 or a median annual exceedance frequency in the range of
1·10-4 to 1·10-5. Additionally also for the SL-2 level, the PGA should be at least 0.1 g regard-
less of the actual seismic hazard.
When probabilistic seismic hazard analysis is used, either a reference annual exceedance fre-
quency in the U.S. according to RG 1.165 [21] as discussed in section 2.2.2, or a performance
based approach according to RG 1.208 [22] is allowed according to section 9.2 in IAEA SSG-9
[6]. As stated in section 9.4 in IAEA SSG-9 [6], a uniform hazard response spectrum shall be
developed by determining response spectral ordinate values that correspond to annual exceed-
ance frequiencies on the basis of seismic hazard curves.
The SL-1 correspond to the Operating Basis Earthquake (OBE), which normally relates to
operational requirements in case this is required by safety authorities. However, in IAEA SG-
G-1.6 [50] the SL-1 is addressed only in relation to safety analysis and design considerations.
In this connection, it should be noted that in the U.S in seismic design of new standardize
NPPs, the use of the OBE as a design basis is only necessary if the licensee defines the OBE as
exceeding ⅓ of the SSE (IAEA SL-2), which has not been done for any new standard NPP
licensed in the U.S to date.

3.2.2 Aspects as regards an eventual Operating Basis Earthquake

(OBE)
According to IAEA SG-G-1.6 [50], SL-1 (OBE) corresponds to a seismic hazard level with a
mean annual exceedance frequency of 1·10-2 as stated in section 3.2.1. Curves of annual seis-
mic hazard at 16%, 37%, 50% (median), 63% and 84% fractiles and the mean for the region
with the highest hazard in Sweden, i.e. the region between Lake Vänern and the Norwegian
border, are presented in [53], as shown in Figure 3.1. From this figure, the mean PGA can be
determined to be approximately 0.15 m/s2 = 0.015 g. This very low PGA level would only
result in negligible effects on safety-related structures as well as on safety equipment housed in
the buildings.
Hence, it can be concluded that an OBE as defined in IAEA SG-G-1.6 [50] would result in
negligible effects on safety-related SSCs at nuclear facilities in Sweden and need therefore not

40
SSM 2014:56
be considered. Finland has similar seismic conditions as in Sweden and as a comparison, the
OBE is not either required by STUK in their provisions for internal and external hazards at a
nuclear facility (YVL B.7 [56]).

Figure 3.1 - Curves of annual seismic hazard (expressed in horizontal PGA) at 16%,
37%, 50% (median), 63% and 84% fractiles and the mean for the region
with the highest hazard in Sweden, according to [53].

3.2.3 Requirements as regards the Design Basis Earthquake (DBE)

SL-2 corresponds to an earthquake level denoted as a Safe Shutdown Earthquake (SSE) or
more generally designated as the Design Basis Earthquake (DBE).
As stated in section 3.2.1, level SL-2 corresponds to a mean annual exceedance frequency in
the range of 1·10-3 to 1·10-4 or a median annual exceedance frequency in the range of 1·10-4 to
1·10-5, according to the recommendations in IAEA SG-G-1.6 [50]. The minimum PGA should
be 0.1 g regardless of the actual seismic hazard at the site.
Within the scope of the European stress tests, the ENSREG peer review team recommended in
[2] that external events should be addressed by designing to a hazard level corresponding to an
annual exceedance frequency of 1·10-4. In some countries the acceleration levels at the target
annual exceedance level 1·10-4 are very low and therefore ENSREG also recommended a min-
imum horizontal PGA of 0.1 g, with reference to the IAEA Safety Guides. It is not obvious
from ENSREG’s conclusions in [2] if the recommended annual exceedance level 1·10-4 relates
to the median (50% fractile) or the mean hazard. But judging from the fact that IAEA recom-
mends the usage of mean values for probability level 1·10-4 in IAEA SG-G-1.6 [50] , it seems
reasonable that ENSREG refers to the mean curves as well.
In this connection it is important to distinguish between mean and median hazard values. As
shown in Figure 3.1, the mean hazard values approximately target the 63% fractile curve for
relatively large annual exceedance frequencies (>10-3), while the mean hazard curves approach

41
SSM 2014:56
even larger fractiles for annual exceedance fractiles less than 10-3. As stated in section 2.2.2,
the mean annual exceedance frequency 10-4 approximately equals the median annual exceed-
ance frequency 10-5, which also is reflected in the IAEA recommendations as stated in section
3.2.1.
The median fractile curves can be considered as best estimates, while the 84% fractile is a con-
servative estimate consistent with general engineering practice in case of design. Historically
in the U.S, there has been focus on generating design response spectra at the 84% fractile level.
For instance, the design response spectra in RG 1.60 [15] are specified at that level.

3.2.4 Current design ground motions in SKI Technical Report 92:3

As stated in section 2.4.2, it can be concluded that the whole Swedish territory can be classi-
fied as a very low seismicity case and therefore the seismic requirements of SS-EN 1998 [48]
can be disregarded, as also is concluded by Boverket and Trafikverket in the National Annex
to SS-EN 1998 [48]. But of course, safety-related structures at nuclear facilities must be de-
signed against external and internal hazards with much higher reference return period than
conventional structures and facilities. SSM has also in SSMFS 2008:17 [49] specifically men-
tioned earthquake as one of several natural events that the Swedish nuclear power plants must
prove resistance to.
The main purpose with project Seismic Safety in the beginning of the 1990’s was to provide
ground response spectra in accordance with Swedish geological and seismological conditions.
The result from this project is presented in SKI Technical Report 92:3 [8], in which ground
response spectra are presented for various annual exceedance probability levels (10-5, 10-6 and
10-7). Spectra were presented for typical hard rock conditions, but also specifically for the
Barsebäck NPP, which was decommissioned in 2005.
The hard rock spectra developed within the scope of Seismic Safety and presented in SKI
Technical Report 92:3 2 were based on a number of accelerograms from registred earth-
quakes in Japan. These Japanese spectra were transformed to Swedish spectra considering the
differences in fault mechanisms and geological conditions between these regions. However
variations in wave velocities along the wave paths due to variations of the hardness of the rock
were not considered. Later research showed that the Japanese recordings originated from sites
were the hardness variation, characterized by the ratio vso/vsb between the shear wave velocities
at the surface and the basement rock were considerable. Some of the Swedish licensees
claimed that the influence of this hardness variation on the results presented in SKI Technical
Report 2 could be considered by reducing the hard rock spectra values by 15%.
As stated in section 2.4.2, new installed SSCs have been designed and existing SSCs been
evaluated against a DBE corresponding to an annual exceedance probability of 1·10-5 accord-
ing to SKI Technical Report 92:3 [8] and shown in Figure 2.6. This design hazard level can be
compared with the corresponding hazard in south Finland, which also refers to an annual ex-
ceedance probability of 1·10-5 at the median confidence level and with a PGA of 0.1g accord-
ing to YVL B.7 [56], and shown in Figure 3.2.
In Figure 3.2, envelope ground response spectra corresponding to an annual exceedance proba-
bility of 1·10-5 with a horizontal PGA of 0.11 g (i.e. the current Swedish DBE), are in the same
figure compared to the following site-independent or site-dependent horizontal design response
spectra at 5 % critical damping:
- Finnish STUK YVL Guide B.7, PGA=0.1 g
- USNRC RG 1.60, PGA=0.11 g
- Eurocode 8 (SS-EN 1998-1:2004), Type 2, Ground type A (rock), PGA = 0.11 g

42
SSM 2014:56
 - EUR, hard soil, PGA=0.25 g
The EUR ground response spectra, which form the design basis for the EPR, as described in
section 2.3.2.3, have a PGA of 0.25 g and are thus enveloping all other design spectra with a
PGA of 0.1 or 0.11 g. As can be noticed, the Swedish DBE spectra will give lower acceleration
responses compared with all other design spectra in the important frequency range below 9 Hz,
where the fundamental eigenfrequencies of the reactor containments and auxiliary buildings
will arise. On the other hand, the Swedish DBE spectra will create higher responses in the fre-
quency range above 9 Hz (except for the EUR spectra), thus being more vulnerable to certain
safety installations primarily associated with electrical and instrumentation control devices in
the buildings.

Figure 3.2 - Comparison between the Swedish DBE ground response spectra (10-5) and
other international well-known DBE response spectra.

SKI Technical Report 92:3 [8] was reviewed by ENSREG peer review team within the scope
of the European stress tests. As stated in [3], ENSREG notified that the seismic hazard in SKI
Technical Report 92:3 [8] is based on observations and historical accounts for about 500 years
in Fennoscandia, together with transformed statistical data from earthquakes occurred in Japan.
ENSREG questioned this short geological time scale, due to the fact that geodetic and paleo-
seismologic data indicates continuous active uplift and deformation of Fennoscandia, accord-

43
SSM 2014:56
ing to some researchers. Besides plate techtonics, isostatic rebound may be a main contributor
in the seismogenic process in Fennoscandia, according to for instance [57]. One theory to the
cause of earthquakes in Fennoscandia is the release of stresses built up and propagated from
the North Atlantic Ridge versus stress adjustment connected to the postglacial land uplift.
According to section 3.25 in IAEA SSG-9 [6], all data on historical earthquakes, including also
events for which no instrumental recording were possible, should be collected as far back in
time as possible. Paleoseismic and archeological information on historical and prehistoric
earthquakes should also be taken into account. (The term “paleoseismic” refers to the evidence
of a prehistoric or historical earthquake displayed as for instance displacements on a fault or
secondary effects as ground deformations). Paleoseismic studies may be particularly useful in
areas for which historical earthquake records are lacking, like for instance in Fennoscandia. As
stated in the Swedish action plan [7], SSM will start up a research program regarding the influ-
ence of paleoseismological data on the existing Swedish seismic design basis.
The approach for determining the hard rock spectra in SKI Technical Report 92:3 [8] was
adopted in the 1980s and the applied statistical database, including the one from Japan, in-
cludes earthquake events until the 1970s. During the last 30 years the methodologies for ad-
dressing seismic hazard have been extensively developed and several European projects have
been active at global and regional scales. The development of the Eurocodes during the 1990s
and 2000s highlighted the need for development of a homogenous and unified probabilistic
seismic hazard assessment procedure for Europe. This development has resulted in the first
ever unified source model throughout the whole European-Mediterranean region and the gen-
eration of seismic hazard maps, expressing ground motion in different parameters, for different
soil conditions and probability levels, some examples for Fennoscandia are shown in Figure
2.5 and Figure 3.1. The seismic hazard assessments for buildings and civil works in the Euro-
codes [52] are based on these hazard maps.
Section 3.18 to 3.20 in IAEA SG-G-1.6 [50] defines how periodic safety reviews should be
addressed at NPPs. As described in recommendation 2.2 in [36] USNRC recommends licen-
sees to confirm seismic and flooding hazards every 10 years and address any new and signifi-
cant information. In France, the external hazards are as well reassessed periodically in the peri-
od safety reviews performed every 10 years, as stated in section 2.3.3.1.

3.2.5 Recommendations on revised design ground motions

With reference to the discussions, evaluations and conclusions stated in previous sections 3.2.1
to 3.2.4, some recommendations are presented regarding a revised seismic design basis for
defining the seismic input for safety analysis and design of nuclear facilities in Sweden.
As described in section 3.2.4 there are today a number of deficiencies in the characterization of
seismic ground motions in SKI Technical Report 92:3 [8]:
- The applied statictical database is from the 1970s and are thus lacking input from seis-
mic events from the last 30-40 years. According to international recommendations
from safety authorities the seismic hazard assessments should be confirmed every 10
years and address any new and significant information.
- The applied methodology for addressing the seismic hazard during project Seismic
Safety was probably in line with the scientific position at that time. But these methods
have been exensively developed and unified at an European level during recent years.
- Within the scope of the European stress tests, ENSREG has questioned the very short
geological time scale (500 years) for applied observations and historical accounts and
the lack of paleoseismic data.

44
SSM 2014:56
 On the basis of these shortcomings the following recommendations are proposed:
- Replace SKI Technical Report 92:3 [8] with a new model for seismic hazard assess-
ments on the basis of the last 20 years research and development within the subject
field.
- Establish annual seismic hazard curves for different fractiles (16%, 37%, 50%, 63% and
84%) and the mean, for different regions in Sweden and for annual probability of ex-
ceedance in the range 10-2 to 10-7.
- Develop new ground response spectra for various annual probability of exceedance, if
possible also for the different NPP sites in Sweden.
- Anchor the Design Basis Earthquake (DBE) at a minimum PGA of 0.1 g.
- Confirm the seismic hazard assesments every 10 years.
Additional recommendation:
- An eventual Operating Basis Earthquake (OBE) with a recommended annual probabil-
ity of exceedance according to IAEA SG-G-1.6 [50] would results in neglectible ef-
fects on SSCs and need not to be considered at nuclear facilities in Sweden.

45
SSM 2014:56
4. Seismic analysis methods
4.1 General
There are a number of different handbooks covering various aspects regarding modeling and
analysis of structural dynamic systems. ASCE 4-98 [24] is a standard which provides mini-
mum requirements and acceptable methods for seismic analysis of safety-related nuclear struc-
tures. This standard provides a comprehensive survey of the seismic analysis process, also
addressing requirements on seismic input and input for subsystem seismic analysis. ASCE 4-
98 [24] covers in principle all applicable requirements in Regulatory Guides and Standard Re-
view Plans issued by USNRC before 1998, for instance RG 1.61 [58], RG 1.92 [27], SRP 3.7.1
[59] and SRP 3.7.2 [60] and provide more extensive background information to the intentions
behind the requirements compared to the official USNRC documents.
ASCE 4-98 [24] provides much more stringent and robust requirements on structural analysis
methods, reflection the enhanced demands for nuclear facilities, compared to what is common
practice in standards for conventional buildings, as for instance in SS-EN 1998 [48]. ASCE 4-
98 [24] has undergone a major upgrading and has been approved for a revised publication in
2014 as an ASCE/ANSI Standard.
The recommendations in the following sections are based on the requirements of ASCE 4-98
[24] complemented with later provisions from USNRC issued after 1998. In order to comply
with the Eurocodes [52], complementary input as regards material parameters and design con-
siderations are also provided. These set of requirements are applicable in the design process of
new structures as well as for seismic evaluation of existing facilities in Sweden. All proposed
recommendations are in all essentials in accordance with applicable parts of IAEA SG-G-1.6
[50].
In this connection it is also worth mentioning that the technical recommendations in this sec-
tion may also be applicable in part to the design of a nuclear facility against vibrational phe-
nomena induced by sources other than erathquakes, as for instance explosions, aircraft crashes
or accidents with high speed rotating machinery, as stated in section 1 of IAEA SG-G-1.6 [50].

4.2 Structural modeling

4.2.1 General requirements

ASCE 4-98 [24], section 3.1.1 defines some general basic requirements on modeling of struc-
tures.

4.2.2 Material properties

In linear elastic analysis of concrete structures, for calculation of eigenfrequencies as well as
for determining sectional forces and moments in the structural elements, mean value of the
modulus of elasticity (Ecm), according to the principles in SS-EN 1992-1 [61], section 5.4 can
be used. The value of Ecm is then calculated according to SS-EN 1992-1 [61], Table 3.1.
Recommended value of the Poisson’s ratio (υ) is 0.2 for uncracked concrete and 0 for cracked
concrete according to SS-EN 1992-1 [61], section 3.1.3.
For eventual non-linear calculations, the general stress-strain diagram according to SS-EN
1992-1 [61] section 3.1.5 can be used.

46
SSM 2014:56
4.2.3 Modeling of stiffness of concrete elements
For determination of the load effects, linear elastic analysis may be carried out under assump-
tion of uncracked concrete cross-sections and mean value of the modulus of elasticity (Ecm).
That is, the structural model can be based on the nominal geometrical properties of the con-
crete elements.
However, if a linear elastic analysis indicates extensive cracking in concrete elements, the re-
duced stiffness must be considered. Qualified engineering assessments are needed to address
the stiffness reduction in an updated linear elastic calculation, whereby ASCE 4-98 [24], sec-
tion 3.1.3 can provide guidance. An acceptable approach to consider cracked concrete proper-
ties can be to reduce the stiffness of the uncracked members by a reduction factor as described
in ASCE 43-05 [23], section 3.4.1.

4.2.4 Modeling of mass distribution

The inertial mass properties in the load-bearing structures can be defined directly in the struc-
tural model through the geometrical properties of the structural elements and the density of the
material. In addition to the structural mass, mass equivalent to a distributed floor load of 250
kg/m2 could be included, to represent miscellaneous dead weights such as minor equipment,
piping and raceways, according to SRP 3.7.2 [26]. The mass of major permanent installed
equipment should be distributed over a representative floor area or included as concentrated
lumped masses at the equipment locations.
The structural model used for determining the seismic response shall also include the mass of
the quasipermanent part of the live load (ψ2Qk). Guidelines for applicable values on ψ2 for
different load types can be found in SS-EN 1990 [62]. Participating part of the mass of the live
load at floor slabs in nuclear facilities should be determined on a best estimate basis, but not
less than 25% (ψ2 ≥ 0.25) of the specified design live load, in accordance with ASCE 43-05
[23], section 3.4.2.

4.2.5 Modeling of damping

Damping represents the structural ability to absorb energy when responding to dynamic load-
ing. Damping is dependent on various factors such as type of connections between the struc-
tural elements, type of material and the stress levels during loading.
Applicable damping values to be used can be found in ASCE 4-98 [24], Table 3.1-1 for vari-
ous types of material. In ASCE 4-98 [24], section 3.1.2.2, the principles for determining damp-
ing values for design and structural evaluation of structures are described, as well as for deter-
mining in-structure response spectra to be used for subsystem seismic analysis.
In this connection, it must be observed that USNRC in their latest version of RG 1.61 [58]
from March 2007 has revised the damping values applicable when generating in-structure re-
sponse spectra to safety equipment at low stress levels. Hence, damping values for different
structural types can be determined according to Table 4.1. The principles for determining stress
level 1 and 2 respectively as described in ASCE 4-98 [24], section 3.1.2.2 are in all essentials
compatible with corresponding principles in RG 1.61 [58].
In practice, Stress Level 2 damping values may always be used in seismic design of structures,
while Stress Level 1 is most often used in development of in-structure response spectra.

47
SSM 2014:56
Table 4.1 – Modal damping ratios according to RG 1.61 [58], with stress level definitions
according to ASCE 4-98 [24].

Structure type Stress level 1 Stress level 2

Reinforced concrete 4% 7%
Prestressed concrete 3% 5%
Welded steel or bolted steel with friction connections 3% 4%
Bolted steel with bearing connections 5% 7%

4.2.6 Modeling of hydrodynamic effects

Hydrodynamic effects of large volumes of water in for instance fuel- and service pools and
condensation pools can be considered in accordance with ASCE 4-98 [24], section 3.1.6. In
this connection, the effects on the dynamical properties (eigenfrequencies) as well as the result-
ing load effects in the walls and floors of the pools and eventual separating walls between dif-
ferent pools should be considered.
ASCE 4-98 [24], section 3.1.6 provides acceptable methods for modeling hydrodynamic ef-
fects of water in pools. ASCE 4-98 [24], section 3.1.6.3 include examples on acceptable meth-
ods for determining convective and impulsive effects of water.

4.3 Seismic analysis

4.3.1 General
Following methods are acceptable to use when performing a seismic response analysis of safe-
ty-related structures at nuclear power plants:
1. The time history method
2. The response spectrum method
3. The equivalent static method
Minimum requirements for each method are described in the following.

4.3.2 Time history method

Time history analysis can be carried out using linear or non-linear analysis methods.
Modal dynamic time history analysis is the most common linear analysis method. The earth-
quake is then described in the form of acceleration-time histories. Requirements on the method
are described in ASCE 4-98 [24], section 3.2.2.2.1. It must be observed that USNRC does not
support ASCE 4-98 [24], section 3.2.2.2.1(f) regarding how many modes need to be included
in the modal superposition. USNRC states that ASCE 4-98 [24], section 3.2.2.2.1(f) is non-
conservative and recommend instead to apply RG 1.92 [27] for modal superposition and ad-
dressing that part of the mass not excited within the total modal mass (“missing mass”). Hence,
ASCE 4-98 [24], section 3.2.2.2.1(f) ought to be used with care and if not all mass is included
in the analysis, it need to be demonstrated that the effect of missing mass can be considered
negligible.
As an alternative to the modal dynamic analysis, the direct integration method can be used, see
ASCE 4-98 [24], section 3.2.2.2.2.

48
SSM 2014:56
In case that geometrical non-linearities, for instance gaps between structural elements, have a
significant impact on the response or where material non-linearities as for instance plasticity or
friction occur, non-linear time-history methods can be applied. Requirements on these methods
are described in ASCE 4-98 [24], section 3.2.2.3.

4.3.3 Response spectrum method

The response spectrum method enables calculation of the maximum response in the structure
when excited by an earthquake defined in the form of a ground response spectrum. The calcu-
lation of maximum values is carried out by combining the maximum responses for the partici-
pating modes. In ASCE 4-98 [24], section 3.2.3, requirements on how to apply the response
spectrum method are described. As regards the application of ASCE 4-98 [24], section
3.2.2.2.1(f), see section 4.3.2.

4.3.4 Equivalent static method

Equivalent static methods for determining seismic load effects in structures are allowed in
national standards for simple structure with symmetric and uniform geometry and mass distri-
bution. However, the method is inappropriate for structures with irregular shapes, likewise
there are restrictions for the method to be used at nuclear facilities. In general, the primary
usage of equivalent static methods are for simple estimates and feasibility assessments of re-
sults from more rigorous dynamic analysis for building structures and for component and dis-
tribution systems. The requirements on the use of equivalent static methods at nuclear facilities
are presented in ASCE 4-98 [24], section 3.2.5.

4.3.5 Multiply-support systems

For structures or safety systems supported on different structures or different structural ele-
ments within a building, the effect of different in-signals must be considered according to
ASCE 4-98 [24], section 3.2.6.

4.3.6 Combination of modal and component responses

Requirements on how modes and excitation directions shall be combined when the response
spectrum method is applied and how the excitation directions shall be considered when the
time history method is used are described in ASCE 4-98 [24], section 3.2.7. In this connection,
it is important to emphasize the requirement that the three directional components of earth-
quake motion in a time-history analysis must be statistically independent in order for the three
excitation directions to be applied simultaneously to the numerical model in one analysis. If the
directions have a statistical dependency, each excitation direction has to be applied separately
and the structural response should be adequately combined as described in ASCE 4-98 [24],
section 3.2.7.2.

4.3.7 Soil-structure interaction

In comparison with other dynamic loads, the earthquake load can be characterized in terms of
the ground motion rather than an applied external load. Ground response spectra or alternative-
ly synthetic developed time histories describe the ground motion in the free field without any
influence from the structure.
Depending on the characteristics of the earthquake, the foundation conditions and the dynamic
properties of the structure, the actual motion of the foundation will deviate from the ground
motion in the free field. For a light building with a flexible foundation slab founded on rock or

49
SSM 2014:56
on soil with high stiffness, the deviation will be negligible, since the building transfer only a
small amount of energy to the environment through the foundation. On the contrary, a heavy
building with a relatively stiffer foundation slab founded on softer soil conditions has a greater
ability to radiate energy to the environment, causing the ground motion in the foundation slab
to differ significantly from the motion in free field.
In case a significant difference can be expected between the motion in free field and the mo-
tion under influence from the structure, ASCE 4-98 [24], section 3.3 requires analysis to be
performed by considering the interaction between soil and structure, i.e. Soil-Structure Interac-
tion (SSI).
In ASCE 4-98 [24], section 3.3.1, it is required that SSI shall be considered for all structures
not founded on rock or rock-like soil foundation material. A fixed-base support may generally
be assumed when the structure is supported on rock or rock-like conditions, which approxi-
mately correspond to shear wave velocities > 1100 m/s. However, it should be verified that the
interaction frequency for a model with a completely stiff structure in combination with discrete
springs according to ASCE 4-98 [24], table 3.3-1 for a circular slab and table 3.3-3 respective-
ly for a rectangular slab, is at least twice the fixed-base frequency in a model with a flexible
structure. If the shear wave velocity > 2400 m/s, a fixed base assumption is accepted without
any further verification, according to SRP 3.7.2 [26].

4.3.8 Input for subsystem seismic analysis

The scope of seismic design of conventional buildings includes primarily calculation of load
effects and verification of sufficient capacity of the load-bearing structural elements. In addi-
tion for safety-related structures at nuclear facilities, the licensees also need to provide input
for seismic analysis of safety equipment in the building in the form of in-structure response
spectra or in-structure time histories at certain positions in the structure, normally at least at
each floor level. In general, in order to provide in-structure response spectra with sufficient
accuracy, the numerical model need to have higher geometrical resolution and more dense
mesh to catch higher local eigenfrequencies. In ASCE 4-98 [24], section 3.4, requirementss on
acceptable procedures for generating in-structure response spectra and time history motions are
provided.

50
SSM 2014:56
5. Seismic design of nuclear structures
5.1 General
In this section the seismic design process for nuclear facilities in Sweden is discussed. The
different steps in the design process are described at a general plant level, but the detailed as-
sessments and recommendations are focused on the safety-related structures.
The recommendations in this section are primarily applicable for new structural design or rede-
sign of existing structures, but can as well be applied for evaluation of existing structures. Spe-
cific assessments and recommendations for existing structures not designed against earth-
quakes are addressed in chapter 6.

5.2 Seismic design classification

5.2.1 Seismic classification according to IAEA Safety Guides

Seismic classification is the process by which each individual SSC at a nuclear facility should
be assigned to a specific seismic category, depending on its required performance during and
after an earthquake, in addition to other required safety classifications. SSCs should be as-
signed into a number of categories in accordance with the principles of section 2.11 to 2.26 in
IAEA SG NS-G-1.6 [50].
SSCs which should be designed to withstand the consequences of a SL-2 (DBE) earthquake
are classified to the highest safety category, Seismic Category 1 (SC1). In particular, following
items are classified as SC1:
- Items whose failure could directly or indirectly cause accident conditions as a conse-
quence of a SL-2 earthquake.
- Items necessary for shutting down and maintaining the reactor in a shutdown condition,
removing residual heat over a required period and monitoring parameters essential to
these functions.
- Items necessary to prevent or mitigate non-permissible radioactive releases, for any
postulated initiating events considered in the design basis, regardless of their probabil-
ity of occurrence, i.e. the defence in depth approach.
For any item in SC1, an appropriate acceptance criterion should be established in the form of
values of design parameters as regards leaktightness, structural integrity, passive or active
function, etc,.
Some examples of items in SC1 associated with the reactor are:
- The primary coolant system
- The main steam- and feedwater systems
- The primary heat removal system
- The control rod drive system
- The emergency power supply, including diesel generators, auxiliaries and distribution
systems.
- Instrumentation and control systems.

51
SSM 2014:56
 - Control rooms required for safe shutdown.
- The reactor containment.
- Structures and buildings which house or support systems for safe shutdown, power sys-
tems and instrumentation and control systems.
In IAEA SG NS-G-1.6 [50], items categorized to a Seismic Category 2 (SC2) mainly includes
items whose failure due to collapse, falling or displacement as a consequence of a SL-2 earth-
quake may jeopardize items in SC1 and SC3, or items that may influence safety functions of
items in SC1 and SC3 or safety related operator actions.
Collapse, falling or displacements caused by an earthquake may for example generate missiles
due to failure of rotating machinery, pressure waves due to bursting tanks, blocking of emer-
gency cooling lines, flooding or fire.
Some examples of SC2 items are:
- The turbine building
- The cooling water intake structures
- The emergency access roads
Seismic Category 3 (SC3) items should include all items that could pose a radiological hazard
but that are not related to the reactor. Some examples of SC3 items:
- The spent fuel building when the fuel is no longer active (i.e. requires forced cooling)
- The radioactive waste building
Seismic Category 4 (SC4) items should include all items that are not in SC1, SC2 or SC3.
Some examples on SC4 items are storage and workshop buildings and administrative build-
ings. Structures and buildings in SC4 could be designed in accordance with conventional non-
nuclear building standards, such as for instance the Eurocodes [52].

5.2.2 Seismic classification according to USNRC Regulatory Guides

The basic approach for seismic design classification in the U.S. is described in section 2.2.4.
The classification approach for NPPs follows RG 1.29 [29], while for radioactive waste at NPP
sites, RG 1.143 [30] is applied. Basically, RG 1.29 [29] only use one seismic category, Seismic
Category 1, which in general complies with SC1 according to IAEA SG NS-G-1.6 [50].
For radioactive waste management SSC in the U.S., RG 1.143 [30] applies. The safety classes
RW-IIa and RW-Ib in RG 1.143 [30] are a function of radiological releases and thus represent-
ing a somewhat more sophisticated classification approach compared to SC3 in IAEA SG NS-
G-1.6 [50] and a reduction of the SL-2 earthqauke motion is specified for design.
As also stated in section 2.2.4, the USNRC has recognized the need to be less prescriptive in
the designation of SC1 items, with the aim of implementing a more safety-related and safety-
significant risk categorization as described in RG 1.201 [33] as an alternative to the determinis-
tic seismic classification approach in RG 1.29 [29]. However, this procedure for risk informed
reclassification of SSCs has so far only been attempted on a trial basis at a small number of
existing NPPs.

5.2.3 Seismic classification according to YVL Guides in Finland

The seismic classification approach to be applied at the Finnish nuclear facilities is addressed
in section 3.4 in YVL Guide B.2 [63]. At the Finnish NPPs, each SSC shall be assigned to one

52
SSM 2014:56
of three seismic categories (S1, S2A and S2B) on the basis of the specified seismic resistance
requirements.
Seismic Category S1 items shall maintain their leak-tightness, integrity, functionality and
proper position in a loading situation caused by a DBE. Some items may be assigned only a
certain feature, for instance leak-tightness, that need to be maintained. The requirements for
Seismic Category S1 items are in all essentials in compliance with corresponding requirements
for SC1 items as defined in IAEA SG NS-G-1.6 [50].
Seismic Category S2A comprises SSCs for which operability and integrity are not essential,
but whose failure due to for instance collapse or falling or other reasons may jeopardize the
safety-related operation and integrity of S1 items. The definition of Seismic Category S2A
items complies with corresponding definition of SC2 items in IAEA SG NS-G-1.6 [50].
It is also mentioned in YVL Guide B.2 [63] that S1 and S2A items should have specifications
as regards acceptance criteria for e.g. operability, integrity and leak-tightness, as also are ad-
dressed in IAEA SG NS-G-1.6 [50].
Seismic Category S2B shall comprise all other SSCs of a nuclear facility, corresponding to the
definition of SC4 in IAEA SG NS-G-1.6 [50].
Seismic Category S1 in YVL Guide B.2 [63] does not distinguish between seismic categoriza-
tion of SSC items related to the reactor and other items that could pose a radiological hazard to
the site as do IAEA SG NS-G-1.6 [50] (SC1 and SC3 respectively).

5.2.4 Seismic classification at the Swedish nuclear facilities

The seismic classification approach in Sweden is described in section 2.4.3. There are three
seismic categories (1, P and N). The designation of the categories (1, P, N) differ from corre-
sponding in IAEA SG NS-G-1.6 [50] (SC1, SC2, SC3 and SC4) and in YVL Guide B.2 [63]
(S1, S2A and S2B). But in practice there are a good coherence between the Swedish practice
and international standards as discussed below.
IAEA SG NS-G-1.6 [50] strongly emphasize on that an appropriate acceptance criterion should
be established in accordance with the required safety function. As examples of such acceptance
criterion, IAEA SG NS-G-1.6 [50] specifically mention design parameters for functionality,
leaktightness or maximum distortion. YVL Guide B.2 [63] also emphasize on specification of
appropriate acceptance criteria for e.g. operability, integrity and leak-tightness. The Swedish
seismic classification approach further addressing these issues by separating Seismic Category
1 (SC1) in IAEA SG NS-G-1.6 [50] or S1 in YVL Guide B.2 [63] into two categories; 1 (leak-
tightness) and P (integrity) as shown in Table 2.3. Seismic Category 1 is the highest category
and includes integrity as well.
This is a strong feature, which facilitates the seismic categorization process by more specifical-
ly addressing the type of acceptance criterion for each different SSC as illustrated in Table 2.3.
Sesmic Category N corresponds to Seismic Category SC2 in IAEA SG NS-G-1.6 [50] and
Seismic Category S2A in YVL Guide B.2 [63].
In Sweden there are no specific category for SSC items whithout any specified safety function
during a DBE earthquake. (SC4 in IAEA SG NS-G-1.6 [50] and S2B in YVL Guide B.2 [63]).
As for YVL Guide B.2 [63], the Swedish seismic classification approach does not distinguish
between seismic categorization of safety items for reactor systems and items that could pose a
radiological hazard not related to the reactor.

53
SSM 2014:56
5.3 Design Basis Earthquake

5.3.1 General
In section 3, some basic conditions for determining design basis ground motions according to
international standards are described and summarized. Also, recommendations on revisions of
the characterizations of the seismic ground motions in SKI Technical Report 92:3 [8] are pro-
vided.
In this section, the main sources and references for establishing the Design Basis Earthquake
(DBE) for the Swedish NPPs are analyzed and some conclusions are provided.

5.3.2 Design Basis Earthquake

In section 3.2.1 it is stated with reference to IAEA SG NS-G-1.6 [50] that the SL-2 (DBE)
earthquake corresponds to a level with a mean annual exceedance frequency in the range of
1·10-3 to 1·10-4 or a median annual exceedance frequency in the range of 1·10-4 to 1·10-5. Addi-
tionally also for the SL-2 level, the PGA should be at least 0.1 g regardless of the actual seis-
mic hazard.
In section 9.2 in IAEA SSG-9 [6], it is stated that the SL-2 (DBE) should be defined by means
of appropriate spectral representations and time histories. The ground motions should be de-
fined for free field conditions, at the level of ground surface or key embedment depths.
Design ground response spectra can be defined either as a standard or a uniform hazard re-
sponse type of response spectrum as further described in section 2.2.2 or section 9.4 and 9.5 in
IAEA SSG-9 [6]. A standard spectrum is obtained from various response spectra derived on
the basis of earthquake records and engineering considerations, scaled to envelope the mean
ground motion levels at low and high frequencies. A uniform hazard response spectrum is de-
veloped by selecting the values of the response spectral ordinates that correspond to the annual
frequencies of exceedance of interest from the seismic hazard curves. As stated in section
2.2.2, the existing standard type of spectra are not appropriate for sites susceptible to high fre-
quency motions, i.e. the situation in Sweden with the dominating contribution from near-field
earthquakes. Even though it might be able to justify the usage of standard response spectra for
certain special cases, it is today an internationally establish common practice to consider the
probabilistic nature of seismic hazard and accordingly to apply site-specific uniform hazard
response spectra for seismic safety evaluations. Therefore, usage of standard response spectra
is not further discussed or evaluated.
According to RG 1.165 [21], a median based reference probability of 1·10-5 is recommended
for determining the DBE. In RG 1.208 [22], a performance-based approach according to
ANSI/ANS 2.26 [64] and ASCE 43-05 [23] is recommended instead. For the highest Seismic
Design Category class (SDC 5), it is recommended to use a mean annual exceedance frequency
of 1·10-4, for the minimum structural damage state, i.e. essentially elastic behavior. Essentially
elastic behavior means that localized inelasticity might occur at stress concentration points, but
the overall seismic response should be essentially elastic. As stated in section 2.2.2, for typical
seismic hazard curves the mean annual exceedance probability 1·10-4 equals the median annual
exceedance probability 1·10-5.
In section 1.5.3.5 in SS-EN 1990 [62], an accidental action is defined as an action, usually of
short duration but of significant magnitude, that is unlikely to occur on a given structure during
the design working life. Further it is stated that impact, snow, wind and seismic actions may be
classified as variable or accidental actions, depending on the available information on statisti-
cal distributions. In section 4.1.1 in in SS-EN 1990 [62], it is stated that the design value of an

54
SSM 2014:56
accidental action should be specified for individual projects and that the design value for seis-
mic actions should be assessed from the characteristic value or specified for individual pro-
jects.
Further as described in 2.4.2, non-nuclear structures in seismic regions in Europe are recom-
mended to be designed to withstand a design seismic action associated with a reference proba-
bility of exceedance of 10% in 50 years or a reference return period of 475 years. It is also
concluded that for hazards with such low reference return period, the whole Swedish territory
can be classified as a very low seismicity case and the requirements of SS-EN 1998 [48] can be
disregarded (i.e. earthquakes of such short return periods in Sweden are not damaging to con-
ventional structures). But due to the fact that nuclear facilities must be designed against inter-
nal and external hazards with much higher reference return period (in the range of 10 000 to
100 000 years), the resulting seismic action should be classified as an accidental action accord-
ing to the definitions in the Eurocodes [52], and could therefore be specified for individual
projects, without considering the characteristic values.
DNB [51] includes complementary rules to the requirements in the Eurocodes [52] for applica-
tion at NPPs in Sweden. The DBE load is in section 4.2.3 in DNB [51] defined as an accidental
load and denominated EDBE. Further in section 7.4.1 in DNB [51], EDBE is specified as a seis-
mic action corresponding to an annual exceedance frecuency of 1·10-5 and defined in the form
of envelope ground response spectra for a typical hard rock site in Sweden according to Ap-
pendix 1 in SKI Technical Report 92:3 [8]. As a comparison, the DBE in Finland is according
to section 4.1 in YVL Guide B.7 [56] defined as having an annual exceedance probability less
than 1·10-5 on the median confidence level. The Finnish design ground response spectra is
anchored at 0.1 g PGA and has a similar shape and spectra ordinate values as the spectra in
SKI Technical Report 92:3 [8] at the same exceedance probability, as illustrated in Figure 3.2.
Some conclusions:
- The basic definitions of the DBE (SL-2) earthquake can be found in IAEA SG NS-G-
1.6 [50] and IAEA SSG-9 [6].
- It is an internationally established common practice to consider the probabilistic nature
of seismic hazard and accordingly to apply site-specific uniform hazard response spec-
tra for seismic safety evaluations.
- The DBE earthquake corresponds to a level with a mean annual exceedance frequency
in the range of 1·10-3 to 1·10-4 or a median annual exceedance frequency in the range
of 1·10-4 to 1·10-5.
- According to the Eurocodes [52] and DNB [51], the DBE load is defined as an acci-
dental load and denominated EDBE.
- EDBE corresponds to ground response spectra for a typical hard rock site with an annual
exceedance probability of 1·10-5 according to SKI Technical Report 92:3 [8].
- Some important remarks and conclusions as regards the deficiencies in the characteriza-
tion of seismic ground motions in SKI Technical Report 92:3 [8] can be found in sec-
tion 3.2.5.

5.4 Design Extension Earthquake

5.4.1 General
In the aftermath of the Fukushima nuclear accident, there has been ongoing activities and dis-
cussons in the nuclear community about how to take into account very unlikely external events

55
SSM 2014:56
beyond the design basis, in order to assess the existing safety margins as well as to identify
potential improvements in the existing plants.
Issues related to severe earthquake effects beyond the design basis for new build safety-related
structures are discussed in this section. For existing structures not designed against earth-
quakes, specific issues are addressed in chapter 6.

5.4.2 Design Extension Earthquake

Already within the Seismic Safety project during the 1980s, discussions were initiated as how
to maintain certain very essential SSCs against seismic loads with extreme low probability of
exceedance levels (10-7), in addition to the requirements on the safe shutdown and cooling of
the reactor to withstand a seismic load corresponding to an annual exceedance frequency of
10-5. It was concluded that SSCs essential for the containment integrity, isolation and pressure
relief should have a sufficient ultimate capability to withstand a seismic load corresponding to
an annual exceedance frequency of 10-7. However, no formal recommendations were estab-
lished and the actions were limited to some minor pilot studies and assessments at some of the
Swedish NPP units.
As part of the European stress tests during 2011, the Swedish licensees carried out calculations
and engineering judgments for a probability level of 10-7 at a majority of the Swedish NPPs, in
order to assess the robustness for severe earthquakes beyond the design basis. The integrity
assessments were focused on the reactor containments, the scrubber buildings, the spent fuel
pools and certain structural walls and slab elements housing safety functions essential for the
containment and pressure relief function and the spent fuel pool integrity. However, these as-
sessments were based on rough up-scaling calculation methods and engineering judgments
owing the very limited time available, as also is notified by ENSREG in [3] and described in
section 1.1.4. Therefore, further refined analyses and investigations are necessary before defi-
nite conclusions can be established.
According to section 2.39 in IAEA SG NS-G-1.6 [50], the seismic design should be carried out
in accordance with the general recommendations in section 1 to 3 in [50], in order to provide
sufficient margins for seismic events beyond the design basis to avoid severely abnormal plant
behavior (“cliff edge effects”). A “cliff edge effect” has reference to when a small deviation of
a design plant parameter give rise to an abrupt worsened situation for the whole plant.
Further, it is also important to create sufficient margins in the design against for instance future
elevated seismic hazard levels, new modern standards for hazard assessments and more sop-
fisticated design and analysis methods.
As stated in section 1.1.3, the French safety authority (ASN) has recommended the French
licensee (EDF) the creation of a “hardened safety core” of essential SSCs and organizational
arrangements in order to manage the fundamental safety functions in extreme situations, aim-
ing on preventing a severe accident and limiting large-scale radioactive releases if the accident
becomes uncontrollable. This approach involves setting up a bunkerized emergency manage-
ment centre with diesel generator and an ultimate backup water supply. The equipment to be
included in the “hardened safety core” must be designed to withstand external events, includ-
ing earthquakes, far beyond the design basis of the plant. ASN has adopted a number of resolu-
tions, which specify the objectives of the “hardened safety core”:
- Prevent a severe accident to affect the core of the reactor or the spent fuel pool.
- Limit the consequencies of an accident which could not be avoided, with the aim of
preserving the integrity of the reactor containment without opening the venting system.
- Enable the licensee to perform its emergency management duties.

56
SSM 2014:56
ASN has specified the design rules for the “hardened safety core” equipment to comply with
the most demanding nuclear industry standards. The seismic hazard level is not yet determined
but should probably be consistent with the BDBE for SMA, that is approximately 1.6 times the
DBE level, which is applied for the EPR. As a comparison, USNRC has developed a require-
ment to evaluate essential SSCs to perform required functions for a seismic load of 1.67 times
the DBE load, according to section 2.2.6.4.
In the Swedish DNB [51], accidental loads corresponding to very unlikely events (Design Ex-
tension Conditions) are specified. One such load is the Design Extension Earthquake (DEE),
denominated EDEE, according to section 4.2.4 in DNB [51]. Further in section 7.4.1 in DNB
[51], it is stated that the seismic load level, EDEE should be specified by the Swedish nuclear
safety authority, SSM.

5.4.3 Recommendations on Design Extension Earthquake

DNB [51] provides recommendations to address consequences of earthquakes beyond design
basis through the seismic load EDEE under Design Extension Condition. Thus, it is now possi-
ble to design certain essential safety functions for a very unlikely earthquake, as was the inten-
tion already in the Seismic Safety project during the 1980s and further emphasized through the
intention of the French safety authorities as regards the “hardened safety core” equipment.
Some recommendations as regards the Design Extension Earthquake:
- Determine a reasonable seismic hazard level for a Design Extension Earthquake (EDEE).
A reasonable hazard level could be approximately 1.6 to 1.7 times the DBE in line
with ongoing discussions in the U.S. and France, as discussed in 2.2.6, 2.3.3 and 5.4.2.
Another possibility could be to apply an annual exceedance probability of 1·10-6 ac-
cording to SKI Technical Report 92:3 [8], with a PGA of 0.23g, i.e. approximately 2
times the DBE. In order to comply with the recommendations provided in the Seismic
Safety project and to create sufficient margins for future increased seismic hazard lev-
els due to identified deficiencies in the characterization of seismic ground motions in
SKI Technical Report 92:3 [8] as stated in section 3.2.5, it can be recommended to ap-
ply an annual exceedance probability of 1·10-7 according to SKI Technical Report 92:3
[8] for certain essential safety-critical SSCs.
- The essential SSCs which need to be maintained for a EDEE can be determined by means
of a seismic classification, with similar procedures as for the DBE load.
- As regards acceptance criteria for an EDEE , see section 5.6.

5.5 Seismic load combinations

General recommendations on how to combine the earthquake load with other loads can be
found in section 2.27 to 2.30 in IAEA SG NS-G-1.6 [50]. However, these paragraphs only
provide some overall recommendations for normal conditions, anticipated operational condi-
tions and accident conditions respectively, and for each Seismic Category, without any specific
recommendations.
In the Eurocodes [52], there are requirements for conventional buildings and industrial facili-
ties on how to combine permanent, variable and accidental loads. These requirements should
be followed for nuclear structures as well, together with additional requirements to consider
essential safety issues for nuclear facilities. DNB [51] specifically provides such supplemen-
tary requirements in addition to corresponding requirements for ordinary building structures as
covered in SS-EN 1990 [62].

57
SSM 2014:56
The general load combination rule for seismic design situations according to section 6.4.3.4 in
SS-EN 1990 [62] is:
ΣG + P + Σ(ψ2·Q) + EDBE where
ΣG = Sum of the different permanent loads
P = Prestressing load
Σ(ψ2·Q) = Sum of the different variable loads
EDBE = Seismic load
In Table 4.6 in DNB [51], this general expression is further decomposed to the different indi-
vidual loads, together with some specified nuclear-related loads:
D + Hgw + Hge + Pp + ψ2L + ψ2S + ψ2Wq + ψ2ΔT + ψ2Hqw + ψ2Hqe + Md + EDBE where
D = Dead load of the structures and permanently installed equipment.
Hgw = permanent groundwater pressure at the mean water level and hydrostatic pres-
sure from water in basins at normal operating conditions.
Hge = Permanent soil pressure.
Pp = Prestressing forces.
ψ2L = Quasi-permanent value of the live loads.
ψ2S = Quasi-permanent value of the snow load.
ψ2Wq = Quasi-permanent value of the wind load.
ψ2ΔT = Quasi-permanent value of climate-related temperature difference and tempera-
ture change.
ψ2Hqw = Quasi-permanent value of the difference between variable waterpressure and
Hgw.
ψ2Hqe = Quasi-permanent value of the difference between variable soil pressure due to
moving load and Hge.
Md = Process-related loads from pipe and process-system, differential pressures, tem-
perature differences and loads resulting from relief valve or other high energy deviced
actuation, during operation, outages or operational disturbances.
EDBE = Seismic load DBE. Determined as the inertia effects from all excited mass, cal-
culated from a dynamic analysis. Can be expressed as a/g (D + ψ2L + ψ2S), where a
represents the seismic acceleration of the mass.
The quasi-permanent or long-term value of a variable action (ψ2·Q) is determined so that the
total period of time for which it will be exceeded is a large fraction of the reference period.
Recommended values for some variable loads can be found in Table A1.1 in SS-EN 1990 [62]
as follows:
Snow load (S): ψ2 = 0.2
Wind load (Wq): ψ2 = 0
Climate-related temperature load (ΔT): ψ2 = 0

58
SSM 2014:56
Live loads (L) at slab structures in nuclear facilities: ψ2 > 0.25 5)
For load Hqw and Hqe,, in general at NPPs: ψ2 = 0
The seismic load combination as above can then be condensed to following expression:
D + Hgw + Hge + Pp + 0.25L + 0.2S + Md + EDBE
The US design standard ASME Section III, Div 2 [9] for reactor containments and ACI 349-06
[10] for other safety-related buildings are commonly used internationally as reference stand-
ards for design of safety-related structures at nuclear facilities. Hence, the seismic load combi-
nation rule in DNB [51] is here compared with corresponding combinations rules in ASME
Section III, Div 2 [9] for reactor containments in Table 5.1 and ACI 349-06 [10] for other safe-
ty-related structures in Table 5.2, with load symbols in accordance with DNB [51].

Table 5.1 – Comparison of load combination rules between DNB [51] and ASME Section
III, Div 2 [9] for reactor containments.
Code Permanent loads Variable loads Accidental loads
- D Hgw Hge Pp L S ΔT Hqw Hqe Md Pa ΔTa R EDBE
DNB 1.0 1.0 1.0 1.0 0.25 0.2 - - - 1.0 - - - 1.0
ASME 1) 1.0 1.0 1.0 1.0 1.0 - 1.0 1.0 1.0 1.0 - - - 1.0
ASME 2) 1.0 1.0 1.0 1.0 1.0 - - 1.0 1.0 - 1.0 1.0 1.0 1.0
1) Load combination “Extreme environmental” according to Table CC-3230-1 in ASME [9].
2) Load combination Abnormal/extreme environmental ” according to Table CC-3230-1 in ASME [9].

Table 5.2 – Comparison of load combination rules between DNB [51] and ACI 349-06
[10] for other safety-related buildings.
Code Permanent loads Variable loads Accidental loads
- D Hgw Hge Pp L S ΔT Hqw Hqe Md Pa ΔTa R EDBE
DNB 1.0 1.0 1.0 1.0 0.25 0.2 - - - 1.0 - - - 1.0
ACI 349 1) 1.0 1.0 1.0 1.0 0.8 - 1.0 1.0 1.0 1.0 - - - 1.0
ACI 349 2) 1.0 1.0 1.0 1.0 0.8 - - 1.0 1.0 - 1.0 1.0 1.0 1.0
1) Load combination (9-6) according to ACI 349-06 [10].
2) Load combination (9-9) according to ACI 349-06 [10].

After comparing the standards some differences can be distinguished:

- DNB/Eurocodes use an approach in which reduction factors are applied on the variable
loads, reflecting the long-term values of these loads, in equivalence with the approach
for determining the participating mass during the earthquake. In contrast, ASME [9]
uses the full value of the live loads, despite the fact that the participating mass is re-
duced, while the live load according to ACI 349-06 [10] includes a minor reduction
(20%) of the live load.
- It seems that neither ASME [9] nor ACI 349-06 [10] consider any snow load in combi-
nation with the seismic load.
- Climate-related temperature distributions are not considered in combination with seis-
mic loads in DNB/Eurocodes.

5)
To be determined on a best estimate basis, but not less than 0.25, see section 4.2.4.

59
SSM 2014:56
 - In contrast to ASME [9] and ACI 349-06 [10], effects of a pipe break accident need not
to be combined with seismic loads in DNB/Eurocodes.
Regarding the approach in DNB/Eurocodes to consider the long-term quasi-permanent part of
the variable loads in combination with the seismic loads is in line with the basic statistical
principles of the standard, while ASME [9] and ACI 349-06 [10] use a more conservatively
deterministic approach.
The same statistical design principles in DNB/Eurocodes are applied to disregard climate-
related temperatures in combination with the seismic load.
Effects of a pipe break as a consequence of the seismic DBE load have not been adopted in the
design basis for the Swedish NPP design. The design philosophy is instead to prove that the
high-energy piping system can withstand the DBE load. The same design principle is adopted
in Finland. In section 4.2.2 in YVL Guide B.7 [56], it is stated that the DBE loads need not to
be combined with other accidental loads if it can be proved that the consequences of the seis-
mic loads can be prevented by the SSC whose failure could initiate another accident (i.e. a pipe
break). In practice in the U.S. today, Leak Before Break (LBB) is used to eliminate postulated
pipe breaks in new NPPs.
The load combination principles for the DEE load are identical with the DBE load as shown in
Table 4.7 in DNB [51]. Thus, the symbol EDBE can simply be replaced by EDEE. As for the
EDBE, pipe break must be proved to be prevented.

5.6 Seismic safety verification

5.6.1 General
Verification of the resistance of SSCs at a nuclear facility against earthquake loads can be
executed by means of one of following methods or a combination of them:
- Experience based methods
- Testing
- Numerical simulations (dynamic analysis)
The experience-based methods consist mainly of assessments of existing facilities’ resistance
against actual strong motion earthquakes. These methods can be used on facilities not designed
against earthquake load effects or facilities designed for a certain earthquake hazard level, but
where the site need to be re-assessed for a more severe hazard level. The best known methods
are SMA and SPSA.
Testing of components is carried out on shaking tables according to specified routines and for
equipment that is difficult to evaluate by other methods. Most commonly, testing is done on
electrical instrumentation and control components and devices.
The predominant method for seismic safety verification of building structures is numerical
simulations by means of dynamic analyses.

5.6.2 Codes and Standards

As stated in section 2.4.4, all building activities in Sweden shall comply to the Eurocodes [52].
However, one important limitation with the Eurocodes [52] is that it formally does not apply to
nuclear power plants, offshore structures and large dam structures. Additionally also, the stat-
ute book of SSM does not yet cover specific requirements and guidance for safety-related
structures at nuclear facilities. Therefore, in order to remedy these deficiencies, SSM together

60
SSM 2014:56
with the Swedish licensees commissioned Scanscot Technology AB to develop DNB [51], as
further described in section 2.4.4.
As described in section 2.4.2, SS-EN 1998-1 [48] applies to design and construction of build-
ings and civil engineering works in seismic regions in Europe. However, the whole Swedish
territory can be classified as a very low seismicity region and the requirements of SS-EN 1998
[48] can be disregarded, as also are concluded by Boverket and Trafikverket in the National
Annex to SS-EN 1998 [48]. Besides this formal non-validity, Eurocode 8 [48] does not provide
sufficient stringent requirements to meet the high level of seismic safety requirements at nucle-
ar facilities.
ASCE 43-05 [23] uses a modern seismic design approach for safety-related SSCs at a broad
range of nuclear facilities. It uses a graded approach in developing the seismic design criteria.
When subjected to the DBE, significant damage are accepted at Limit State A, while no dam-
age and essentially elastic behavior are the goal at Limit State D. Performance goal are ex-
pressed as the mean annual probability of exceedance of the specified Limit State. In addition,
all buildings are classified in a Seismic Design Category (SDC) developed in ANSI/ANS 2.26
[64], in order to set the design earthquake levels. Five SDCs have been established. The ap-
proach for assigning an SSCs SDC is based on the severity of the consequences of its failure to
perform its safety functions as determined by safety analysis. Conventional buildings may be
assigned to SDC-1 while nuclear power plants associated with reactor and spent fuel safety
may be assigned SDC-5. In practice, all modern NPPs fall in SDC-5 with the Limit State D
acceptance criteria “essential elastic behavior”.
ACI 349 [10] provides requirements for design of safety-related nuclear structures As de-
scribed in section 2.2.5, ACI 349 [10] requires reinforcing steel detailing to provide ductility to
the structures in accordance with chapter 21 in the code. The main reason for this requirement
despite the “essential elastic behavior” criterion for the DBE load, is to provide additional as-
surance that the structural integrity is maintained in the unlikely event of an earthquake beyond
the DBE or other unforeseen circumstances.

5.6.3 Seismic safety verification

The seismic design process can in general be described in three steps:
1. Establish the design earthquake level, i.e. DBE or DEE.
2. For each earthquake level assign each SSCs to any of the Seismic Categories (1, P or
N).
3. Verify that the acceptance criteria for each SSCs are fulfilled, during and after the
earthquake.
The procedures for step 1 is described in section 5.3 for DBE and in section 5.4 for DEE. Step
2 procedures are discussed in section 5.2. Recommendations as regards seismic safety verifica-
tion of the safety-related structures are described below.
Minimum requirements for structural modeling and seismic analysis are described in section 4.
It can be concluded that modern nuclear facilities should be classified as SDC-5 structures as
defined in ASCE 43-05 [23] and thus be verified against the acceptance criteria “essential elas-
tic behavior”. The requirements in chapter 21 in ACI 349 [10] have the same purpose as corre-
sponding requirements in section 5 in Eurocode 8 [48], that is to ensure adequate ductility to
dissipate energy without substantial reduction of the overall resistance of the building. Howev-
er, Eurocode 8 [48] recommends concrete buildings to be designed for low dissipation capacity
and low ductility in cases of low seismicity regions (as in Sweden). This means that building
structures can be designed for the seismic design situations in principle with the same methods

61
SSM 2014:56
as for other accidental loading situations according to SS-EN 1992-1-1 [61], without any spe-
cific ductile reinforcement arrangement.
Design of safety-related structures in Seismic Category 1 and P for the DBE is carried out in
accordance with requirements in the ultimate limit state in DNB [51] for the seismic load com-
bination as described in section 5.5. In order to meet the “essential elastic behavior” criteria,
the idealization of the behavior in the structural analysis is limited to linear elastic behavior,
according to section 6.6.6.1 in DNB [51]. Idealization of the behavior in terms of “linear elas-
tic behavior with limited redistribution” or “plastic behavior” should be avoided for the DBE
load because of the cyclic character of the seismic load and the absence of ductile reinforce-
ment. Eventual comprehensive concrete cracking is considered linear elastically in accordance
with the priciples for stiffness reduction, as described in section 4.2.3.
Safety-related structures in Seismic Category N need not comply with any formal seismic safe-
ty requirements. But structures or structural elements in Seismic Category N should not jeop-
ardize SSCs in Seismic Category 1 or P.
Design of structures in Seismic Category 1 and P for the Design Extension Earthquake (DEE)
can be carried out in accordance with the same approach as for the DBE. The “essential elastic
behavior” acceptance criteria and the non-ductile design philosophy as for the DBE applies as
well.

62
SSM 2014:56
6. Seismic evaluation of existing nuclear struc-
tures
6.1 General

6.1.1 General considerations

It is important to distinguish between seismic design and seismic safety evaluation, in that
seismic design of SSCs is primarily related to the design stage of the installation prior to its
construction and operation, whereas seismic safety evaluation is applied after the installation
has been constructed. Seismic design can of course also be applied after its construction in case
of design of a new SSC or upgrading of an existing SSC.
The seismic design procedures for new structural design as described in section 5 can be ap-
plied to existing structures as well. However, when evaluating existing structures, alternative
methods are generally applied, as described in this section. The main reference is IAEA NS-G-
2.13 [69], but also section A1.0 in ASCE 4-98 [24].
It is usually recognized that nuclear facilities designed against seismic loads in accordance
with established international practice have an inherent capacity to resist larger earthquakes
than considered in the design basis, as discussed in section 5.6.3. This inherent robustness is
usually described as the “seismic design margin” and is a direct consequence of
a. the conservatism in the seismic design procedures and
b. the fact that other loads than the seismic load are governing for many of the SSCs.
For the Swedish existing nuclear facilities (except Oskarshamn unit 3 and Forsmark unit 3), the
first reason (a) is not applicable since seismic design requirements were not included in the
original design. In case of a seismic evaluation against earthquake loads, the second reason (2)
is often at hand at the Swedish facilities, especially for load-bearing steel structures, where
wind and snow loads usually are governing, or for the reactor containment vessels, for which
the inherent over-preussurization capacity normally envelope the relatively moderate seismic
load effects. Due to the fact that nuclear installations are designed for a wide range of internal
and external accidental and natural hazards other than seismic events, the seismic design mar-
gin varies a lot for different locations and different SSCs in a facility.

6.1.2 Objectives of the seismic safety evaluation

In accordance with international accepted practice, systematic safety reassessments of a NPP
shall be performed throughout its operational lifetime, taking operating experience and signifi-
cant new safety information into account. A seismic safety evaluation of an existing nuclear
facility should be performed in the event of any one of the following:
- Evidence that the seismic hazard at the site is significantly greater than the DBE, be-
cause of new seismological data or new methods for seismic hazard assessments.
- Regulatory requirements as regards for instance periodic safety reviews that take into
account the state of the knowledge and the actual condition of the plant.
- Inadequate seismic design, due to for instance old and obsolete design solutions.
- New technical findings, such as vulnerability of selected SSCs.
- New experience from the occurrence of actual earthquakes.

63
SSM 2014:56
 - The need to address the performance of the facility for beyond design basis earth-
quakes, in order to demonstrate that no significant failure would occur for a slightly
greater earthquake than the DBE, that is to prove that there is no “cliff-edge effect”.
- As part of a programme for long term operation of the plant.
It is important to clearly establish the purposes of the seismic evaluation before the process
starts-up, because there are significant differencies in the evaluation procedures and acceptance
criteria depending on the purpose of the evaluation. The main objectives of a seismic safety
evaluation may include one or more of the following:
- To demonstrate the seismic safety margin beyond the original design basis and to con-
firm that there are no cliff edge effects.
- To identify weak links in the facility and its operations with regard to seismic loads.
- To identify and prioritize possible upgrades.
- To assess plant capacity parameters, for instance HCLPF-values, against regulatory ex-
pectations.

6.1.3 Selection of appropriate methodologies

There are two main approaches for evaluating the seismic safety of existing facilities; the
Seismic Margin Assessments (SMA) and the seismic PSA (SPSA) methodology. The main
differencies between these two methods are related to the modeling approach and the capacity
evaluation. The SMA uses success paths for system modeling, while SPSA applies so called
event trees or fault trees. Capacity evaluations of SSCs are performed in terms of HCLPF val-
ues in the SMA method, whereas capacity evaluations are made by probabilistically defined
fragility functions in the SPSA method.
In general, the level of effort required is much higher for the SPSA-method than the SMA-
method because of the extended scope of the SPSA even when done at the same detail level.
The approach for seismic safety evaluation in this report is limited to safety-related nuclear
structures and structural elements. Hence, in section 6.3 and thereafter focus will be on appli-
cable parts of the SMA-methodology to address issues related to seismic safety evaluation of
safety-related structures and equipment-building interfaces at the existing Swedish nuclear
facilities.
However, in section 6.2 for the safety evaluation of existing structures against the DBE, a con-
ventional evaluation is recommended in accordance with the same procedures as for new struc-
tures, as described in section 5.6.3, for the DBE load.
Walkdown procedures and use of earthquake experience data are important parts of seismic
safety assessments, especially as regards selecting the success paths and determining the seis-
mic capacities of the selected SSCs. However, these activities are mainly related to the overall
plant level and are not dealt with in this report.

6.2 Safety evaluation against the Design Basis Earthquake

According to the Swedish action plan [3] developed within the framework of the European
stress tests, the existing Swedish NPPs should withstand the effects of a DBE with an annual
exceedance probability of 1·10-5 according to SKI Technical Report 92:3 [8]. For the DBE it is
reasonable to apply a conventional evaluation of the seismic safety of building structures, ac-
cording to the same procedures as for new structures as described in section 5.6.3.

64
SSM 2014:56
It should be noted that for seismic evaluation of existing NPP distribution systems and compo-
nents as described in the IAEA Safety Report No. 28 [65] development of success paths sam-
pling, earthquake experience data and engineering judgement play a more significant role.
In this connection it is important to notice that in principle all existing safety-related structures
at the Swedish NPPs were designed during the late 1960s to the beginning of the 1980s, in
accordance with the then applicable design standard “Bestämmelser för betongkonstruktioner”
(BfB) for concrete structures [66] and “Statens Planverks spännbetongnormer” SBN-
S25.21[67] for prestressed concrete structures. The steel structures were designed according to
Stålbyggnadsnormerna StBK-N1, N2 och N4 [68].
The design philosophy in BfB [66], SBN-S25.21[67] and StBK-N1, N2 och N4 [68], were
based on the principles of certain specified “allowable stresses” for concrete and reinforce-
ment. Load cases were defined as “ordinary load cases” for which low allowable stress values
were used, whereas 20% higher allowable values could be used for “exceptional load cases”.
Earthquake loads were not included in the design basis and consequently no specific ductile
reinforcement were required. In addition, requirements on shear reinforcement in BfB [66]
were not as stringent as in the Eurocodes [52], thus further limiting eventual inherent ductility
compared to the Eurocodes [52].
Considering the absence of ductile reinforcement in the existing Swedish nuclear facilities, it is
reasonable to apply the “essential elastic” acceptance criteria as described in section 5.6.3 for
new structures. It shall however be emphasized that the “essential elastic behavior” as stated in
RG 1.208 [22] can be somewhat mitigated in the sense that localized inelasticity are accepted
at stress concentration points, but the overall seismic response should be essentially elastic. In
addition in SRP 3.7.2 [26], it is stated that the SRP acceptance criteria address linear elastic
analysis with allowable stresses near elastic limits of the structures. However, for certain spe-
cial cases (e.g., stability analysis and evaluation of as-built structures), reliance on limited ine-
lastic/nonlinear behavior is acceptable when appropriate.

6.3 Seismic Margin Assessments

6.3.1 General
Besides evaluation of the seismic safety of existing structures against the DBE load, it might
also be necessary to consider effects of beyond the design basis, in order to avoid any “cliff-
edge effects”, as described in section 5.4.2. As described in section 3.2.5 a number of deficien-
cies in the characterization of the seismic ground motions in SKI Technical Report 92:3 [8] is
specified. For instance, the applied database is very old (from the 1970s) and has not been up-
dated despite international recommendations to confirm/update the seismic hazard assessments
every 10 years, and ENSREG has specifically also questioned the very short geological time
scale (500 years) for applied observations and historical account as well as the lack of paleo-
seismic data.
Thus, it is quite likely that a future reassessment of the seismic hazard would result in in-
creased hazard levels. Further as stated in section 5.4.2 it is also important to consider that new
modern standards for seismic hazard assessments and more sophisticated analysis methods
may indicate insufficient margins. Also, as will for instance be the case in France, there might
be new requirements on verifying specified essential safety equipment for a seismic event be-
yond the design basis, as discussed in section section 5.4.2, not only for new structures but also
for existing ones.
To sum up it might be necessary to provide safety evaluations and assessments for beyond
design basis seismic events for the existing Swedish NPPs.

65
SSM 2014:56
The main aim of a seismic safety evaluation of an existing nuclear facility is to determine the
true state of the evaluated SSCs in terms of their required safety function and their seismic
capacity and thus to assess the seismic safety margin. The approach used by the SMA method-
ology is to consider a reasonable higher seismic hazard level than the design basis and to asso-
ciate this with realistic and best estimate values for the as-is condition of the SSCs. Hence, the
inherent excess capacity of the SSCs can be accounted for.
The SMA approach is in general performed in six different steps:
- Define the objectives of the seismic safety evaluation.
- Determine the Review Level Earthquake (RLE).
- Selection of success paths and of selected SSCs.
- Seismic response analysis.
- Capacity assessments of the selected SSCs.
- Determination of utilization ratios for the selected SSCs
On the basis of the SMA approach with its different procedures and specifically also the Con-
servative Deterministic Failure Margin (CDFM) method as described in EPRI NP-6041-SLR1
[70], a proposal for a somewhat simplified deterministic method for seismic safety evaluation
of safety-related nuclear structures is presented in section 6.3.2 to 6.3.7.

6.3.2 Objectives of the seismic safety evaluation

The first important step is to establish the main purposes of the seismic safety evaluation. In
section 6.1.2, the four most common objectives are specified. The approach presented here is
limited to a safety evaluation of the building structures in line with the following main objec-
tives:
- To evaluate the seismic safety margin beyond the DBE and to confirm that there are no
cliff edge effects.
- To identify weak links as regards maintaining required functions in the structures dur-
ing and after an earthquake beyond the DBE.
In accordance with these objectives, it is important to determine the RLE level of the seismic
ground motion, the safety functions to be ensured and the acceptance criteria for the identified
critical failure modes to be prevented.
The methodology presented here is not dependent on any eventual SMA for the safety equip-
ment in the plant, but can of course be one acitivity in a larger SMA scope at plant level. Any-
way, the step to select the success paths and selection of essential SSCs ought to be determined
on a plant level, to be able to identify essential safety equipment and their locations in the
plant, as basis for defining the structural elements which must maintain supporting and/or
shielding functions of these SSCs.
One disadvantage with the approach described above is that only two of the Swedish NPP units
(Oskarshamn 3 and Forsmark 3) include seismic load in their design basis. This means that for
the other units, a successful seismic evaluation against a post-defined DBE must first be exe-
cuted in accordance with for instance the procedures in section 6.2.

6.3.3 Determine the Review Level Earthquake (RLE)

The Review Level Earthquake (RLE) defines a screening level in the evaluation process. The
RLE should be defined with a sufficient margin over the original DBE, in order to ensure plant

66
SSM 2014:56
safety and to identify any essential “weak links” that may limit the overall capacity of the plant
to withstand a seismic event beyond the original DBE.
Most of the procedures developed and implemented to date have defined two screening levels;
PGA of 0.3 g and PGA of 0.5g. These PGA levels were established on the basis of seismic
hazard values and earthquake experience data from the U.S.
According to section 2.2.6.4, the USNRC has recommended the US licensee to assure a safety
margin for a Beyond Design Basis Earthquake (BDBE) corresponding to a load of 1.67 times
the DBE. This BDBE correspond in probabilistic terms to a doubling of the mean return time
from 10 000 years (DBE) to approximately 20 000 years. For the new EPR NPP plant, seismic
margin assessments have been performed with a target margin of 1.6 times that assumed for
the DBE, as stated in section 2.3.3.4 and 2.3.3.5.
The DBE at the Swedish nuclear facilities is defined in terms of a uniform annual frequency of
exceedance response spectrum shape according to SKI Technical Report 92:3 [8], in contrast to
for instance the standard type of response spectra in RG 1.60 [15]. The CDFM method as de-
scribed in EPRI NP-6041-SLR1 [70] recommend to select the RLE in terms of a uniform haz-
ard spectra in case the DBE is defined similarly, in order to achieve uniformity in the frequen-
cy domain between the DBE and RLE and in the safety margin assessments. Then, statements
as regard seismic margins can be expressed in probabilistic terms of annual frequency of ex-
ceedance, as opposed to the deterministic PGA.
According to international practice, the RLE level often is determined to be approximately 1.6
to 2.0 times the DBE level. Thus, a first approach could be to apply a trial responce spectra
with an annual exceedance probability of 1·10-6 according to SKI Technical Report 92:3 [8],
which correspond to a PGA of 0.23 g, that is 2.0 times the PGA of the DBE (annual exceed-
ance probability of 1·10-5). If too many essential structural elements fail in the evaluation at
this level, it can be necessary to reduce the RLE level in a second step.
As stated in section 5.4.2, it was in the joint project Seismic Safety during the 1980s between
the then Swedish nuclear safety authority (SKI) and the Swedish licensees concluded that
SSCs essential for the containment integrity, isolation and pressure relief should have a suffi-
cient ultimate capability to withstand a seismic load corresponding to an annual exceedance
frequency of 10-7. Consequently, the Swedish licensees carried out calculations and engineer-
ing judgements for a selection of the most important of these safety functions at an annual ex-
ceedance probability of 10-7, within the scope of the European stress tests during 2011. Hence,
in order to comply with the recommendations in the project Seismic Safety, but also to specifi-
cally consider future increased seismic hazard levels due to identified deficiencies in the char-
acterization of seismic ground motions in SKI Technical Report 92:3 [8] as stated in section
3.2.5, it may be advisable to apply an annual exceedance probability of 1·10-7 according to SKI
Technical Report 92:3 [8] as RLE for certain essential safety-critical SSCs.

6.3.4 Selection of success paths and selected SSCs

The SMA uses a success path approach to determine which SSCs should be selected. A suc-
cess path means a path of SSCs (selected SSCs) that can successfully bring the plant to the
desired end state for the plant (e.g. safe shutdown). Other types of successful end states could
for instance include the defence in depth and system redundancy for the nuclear reactors, and
successful confinement of nuclear waste during and after an earthquake for a NPP and other
nuclear facilities.
The most fundamental safety functions to be ensured for the RLE level are:
- Control of reactivity.

67
SSM 2014:56
 - Removal of heat from the core.
- Confinement of radioactive materials, control of operational discharges as well as limi-
tation of accidental releases.
- Cooling of spent fuel.
In terms of structures and structural elements whose safety functions (confinement, support or
shielding) need to be preserved, the following elements can be specified as a minimum:
- The reactor containment.
- The scrubber building.
- The spent fuel pools.
- The structural elements housing or shielding essential safety equipment to preserve the
fundamental safety functions as above.

6.3.5 Seismic response analysis

For those SSCs selected as part of the safe shutdown success path, the seismic response should
be determined for:
a. Evaluation of the structural capacity on the basis of required function to be maintained
and the damage mode.
b. Generation of in-structure response spectra.
For each building and structure which has been defined as part of the selected SSCs, the safety
functions to be maintained, the damage modes for these functions and associated acceptance
criteria should be defined.
The evaluation procedure for determining the best estimate (median centred) of the seismic
response should be defined. When using best estimate parameters it is acceptable to use Re-
sponse Level 3 damping values according to Table 3.2 in ASCE 43-05, if a linear elastic anal-
yses is used and if the structural elements are near yield. Response Level 3 is limited to the
evaluation of the structural elements. For generation of in-structure response spectra Response
Level 2 is used.
The normal procedures for the response calculations are through dynamic structural analysis,
but in certain simple cases, scaling can also be accepted.
Evaluation of equipment-building interfaces (e.g. anchorages, anchor plates and anchor bolts)
should be included in the evaluation of selected equipment and distribution systems. All domi-
nant failure modes of these interfaces, such as failure of the anchorage or the substructure
should be identified and evaluated on the basis of as-is conditions. The expected behavior of
the supporting building structure elements should also account for local concrete cracking
which may reduce the capacity of expansion anchor bolts.

6.3.6 Capacity assessments of the selected SSCs

In a SMA, capacities of selected SSCs are defined as HCLPF capacities. In probabilistic terms,
the HCLPF capacity of an SSC is the earthquake level at which there is a high confidence
(95%) of a low (5%) probability of failure. Although defined with a probabilistic approach,
HCLPF values are almost always calculated by deterministic methods. Deterministic guide-
lines in the form of the CDFM-method as described in EPRI NP-6041-SLR1 [70], have been
developed and demonstrated to yield the approximate probabilistic definition.

68
SSM 2014:56
The required functions of the selected SSCs must be identified. For example, required func-
tions of structures can be confinement, support and/or protection of other SSCs. Realistic fail-
ure modes of the SSCs, that is the inability of the SSCs to perform its required safety function
due to inadequate seismic capacity, have to be determined. For structures with proven ductility
capacity against cyclic loads, some non-linear behavior can be accepted, but at lower levels
than for conventional industrial facilities. In seismic safety evaluations of SSCs, ageing degra-
dation should be considered if the ageing effects reduce the seismic capacity of SSCs.
The as-is concrete classes used for the construction of the structures should be verified on the
basis of existing plant specific tests and industry standards for concrete. Either destructive or
non-destructive methods may be used. The actual material properties of the reinforcement steel
should be used in the evaluation. Material properties should be available from existing test
data.
For Swedish conditions it can be reasonable to accept characteristic values on concrete and
reinforcement, i.e. to use fyk for reinforcement and fck and fctk for concrete without any reduc-
tion to reflect design, and including effects of concrete ageing if appropriate.
In order to assess the vulnerability due to eventual non-compliancy of certain structural ele-
ments when using characteristic values, it might be reasonable to instead apply median centred
strength property values in a second step.
The ductility of the Swedish nuclear structures to cyclic earthquake loads is limited because of
the absence of ductile reinforcement arrangements according to international standards. How-
ever, limited non-linear assessments may be applied for exceptional cases when appropriate.
Then, especial attention should be paid to ensuring that shear or other non-ductile failure
modes can be excluded with sufficient margins for the investigated elements.

6.3.7 Determination of utilization ratios for the selected SSCs

The calculated load effects from the seismic response analysis as determined according to sec-
tion 6.3.5 (i.e. demand) are compared to the capacities determined in section 6.3.6.
Thus, eventual weak links in the structures may be identified. Alternatively, if too many essen-
tial structural elements have failed, an updated SMA can be executed for a lower RLE level.
Finally, a seismic margin assessment for the building structures can be expressed in probabilis-
tic terms, by interpolating between applicable typical seismic hazard curves for the different
exceedance probability levels of the DBE and RLE respectively.

69
SSM 2014:56
7. References
[1] ENSREG: Post-Fukushima accident. Compilation of recommendations and suggestions
– Peer review of stress tests performed on European nuclear power plants. 2012.

[2] ENSREG: Post-Fukushima accident. Peer review report. Stress tests performed on Eu-
ropean nuclear power plants. 2012.

[3] ENSREG: Post-Fukushima accident. Peer review country report-Sweden. Stress tests
performed on European nuclear power plants. 2012.

[4] ASN: Nuclear Safety Authority (ASN) opinion no 2012-AV-0139 of 3rd January 2012
concerning the complementary safety assessments of the priority nuclear facilities in the
light of the accident that occurred on the nuclear power plant at Fukushima Daiichi,
2012.

[5] ASN Resolution 2014-DC-0403 of 21th January 2014 instructing Electricité de France.
2014.

[6] IAEA Safety Guide SSG-9. Seismic Hazards in Site Evaluation for Nuclear Installations.

[7] Strålsäkerhetsmyndigheten (SSM): Swedish action plan for nuclear power plants. Re-
sponse to ENSREG’s request. December 2012.

[8] SKI Technical Report 92:3: Project Seismic Safety. Characterization of seismic ground
motions for probabilistic safety analyses of nuclear facilities in Sweden. Summary Re-
port. April 1992.

[9] American Society of Mechanical Engineers: ASME Boiler and Pressure Vessel Code,
Section III, Division 2, Subsection CC, Code for Concrete Reactor Vessel and Contain-
ments.

[10] American Concrete Institute, Code Requirements for Nuclear Safety-Related Structural
Concrete Structures (ACI 349-06) with Commentary, september 2007

[11] TID 7024 Nuclear Reactors and Earthquakes: Prepared by the Lockheed Aircraft Corp.
et al for U.S. Atomic Energy Commission. August 1963.

[12] Uniform Building Code: Volume 1. International Conference of Building officials. Vol-
ume 1. 1963 Edition.

[13] US Atomic Energy Commission: US Code of Federal Regulations 10CFR Part 100, Ap-
pendix A: Seismic and Geologic Siting Criteria for Nuclear Power Plants. 1973.

[14] Housner G.W., Martel R.R. and Alford J.L; Spectrum Analysis of Strong Motion Earth-
quake. Bull. Seismo1. Soc.Am 43. 1953.

[15] U.S.NRC: Regulatory Guide 1.60. Design Response Spectra for Seismic Design of Nu-
clear Power Plants. December 1973.

70
SSM 2014:56
[16] Stevenson, J.D. and Conan, O; Design of Hazardous Mechanical Structures, Systems
and Components for Extreme Loads, ASME Press, 2006.

[17] Chopra A.K., Dynamics of Structures, 1995.

[18] EPRI, U.S. DOE and U.S. NRC: Technical Report: Central and Eastern United States
Seismic Source Characterization for Nuclear Facilities. 2006.

[19] U.S.NRC: US Code of Federal Regulations 10CFR Part 100.23, Geologic and Seismic
Siting Criteria. 1973.

[20] U.S.NRC: US Code of Federal Regulations 10CFR Part 50, Appendix S: Earthquake
Engineering Criteria for Nuclear Power Plants. 1996.

[21] U.S.NRC: Regulatory Guide 1.165. Identification and Characterization of Seismic

Sources and Determination of Safet Shutdown Earthquake Ground Motion. March
1997.

[22] U.S.NRC: Regulatory Guide 1.208. A Performance-Based Approach to Define the Site-
Specific Earthquake Ground Motion. March 2007.

[23] American Society of Civil Engineers: ASCE/SEI 43-05. Seismic Design Criteria for
Structures, Systems, and Components in Nuclear Facilities. 2005.

[24] American Society of Civil Engineers, ASCE 4-98 Seismic Analysis of Safety-Related
Nuclear Structures and Commentary. 1998.

[25] U.S. Nuclear Regulatory Commission, Standard Review Plan 3.7.1, Seismic Design Pa-
rameters, Rev 3, mars 2007.

[26] U.S. Nuclear Regulatory Commission, Standard Review Plan 3.7.2, Seismic System
Analysis, Rev 3 mars 2007.

[27] U.S. Nuclear Regulatory Commission, Regulatory Guide 1.92, Combining Modal Re-
sponses and Spatial Components in Seismic Response Analysis. Rev 2, juli 2006.

[28] U.S. Nuclear Regulatory Commission, Regulatory Guide 1.122; Development of Floor
Design Response Spectra för Seismic Design of Floor-supported Equipment or Compo-
nents. February 1978.

[29] U.S. Nuclear Regulatory Commission, Regulatory Guide 1.29, Seismic Design Classifi-
cation. Rev 4, March 2007.

[30] U.S. Nuclear Regulatory Commission, Regulatory Guide 1.143, Design Guidance for
Radioactive Waste Management Systems, Structures, And Components Installed in
Light-Water-C Nuclear Power Plants, Rev 2, November 2001.

[31] U.S. Nuclear Regulatory Commission, Regulatory Guide 1.189, Fire Protection for Nu-
clear Power Plants, Rev 2, October 2009.

[32] American Society of Civil Engineers, ASCE 7-10. Minimum Design Loads for Build-
ings and Other Structures. 2010.

71
SSM 2014:56
[33] U.S. Nuclear Regulatory Commission, Regulatory Guide 1.201. Guidelines for Catego-
rizing Structures, Systems and Components in Nuclear Power Plants According to Their
safety Significance. May 2006.

[34] U.S.NRC: US Code of Federal Regulations 10CFR Part 50.69, SSC Categorization
Guideline. 2005.

[35] American Concrete Institute, Building Code Requirements for Structural Concrete (ACI
318-11) with Commentary, augusti 2011

[36] USNRC: Recommendations for Enhancing Reactor Safety in the 21st Century, The near-
term Task Force Review of Insights from the Fukushima Dai-Ichi Accident. July 2011.

[37] EPRI: Seismic Evaluation Guidance. Screening, Prioritization and Implementation De-
tails (SPID) for the Resolution of Fukushima Near-Term Task Force Recommendation
2.1: Seismic. 2013.

[38] EPRI: Seismic Walkdown Guidance. For Resolution of Fukushima Near-Term Task
Force Recommendation 2.3: Seismic. 2012

[39] ASN RFS-I.3.c: Etudes géologiques et géotechniques du site ; détermination des caracté-
ristiques des sols et études du comportement des terrains (1er août 1985)

[40] ASN: Basic Safety Rule. Fundamental safety rule no 2001-01 concerning basic nuclear
installations. 2001.

[41] ASN RFS V.2.g: Règles générales concernant plusieurs systems, structures ou équipe-
ments. 1985.

[42] ASN/Guide/2/01: prise en eompte du risqué sismique a la conception des ouvrages de

genie civil d’installations nucléaires de base a l’exception des stockages a long terme
des déchets radioactifs. 1985.

[43] AFSEN ETC-C: EPR Technical Code for Civil Works. 2010

[44] EDF RCC-G: Design and construction rules for civil works of PWR nuclear islands,
July 1988.

[45] Institute for Nuclear Safety and Protection, Basic Safety Rule I.2.c. 1981.

[46] EUR – European Utility Requirements for LWR Nuclear power plants, Volume 2, Chap-
ter 4. 2012

[47] Seminar on NPP’s earthquake safety, SKI Technical Report 646/86, 1986.

[48] SS-EN 1998-1:2004; Eurocode 8: Design of structures for earthquake resistance-Part 1:

General rules, seismic actions and rules for buildings, edition 1, March 2009

[49] Strålsäkerhetsmyndighetens författningssamling, SSMFS 2008:17: Strålsäkerhetsmyn-

dighetens föreskrifter om konstruktion och utförande av kärnkraftsreaktorer.

72
SSM 2014:56
[50] IAEA Safety Guide NS-G-1.6. Seismic Design and Qualification for Nuclear Power
Plants. November 2003.

[51] Scanscot Technology: Dimensionering av Nukleära Byggnadskonstruktioner (DNB),

Januari 2014.

[52] Boverket, BFS 2011:10 - EKS 8 Boverkets föreskrifter och allmänna råd om tillämpning
av europeiska konstruktionsstandarder (eurokoder), april 2011.

[53] Wahlström, R., Grunthal, G: probabilistic seismic hazard assessment (horizontal PGA)
for Fennoscandia using the logic tree approach for regionalization and nonregionaliza-
tion models. 2001.

[54] Scanscot Technology AB, Dimensioneringsregler för byggnader (DRB:2001), March

2002.

[55] Boverket: Boverkets Konstruktionsregler, BKR, 2010.

[56] YVL Guide B.7. Provisions for internal and external hazards at a nuclear facility. 2013.

[57] Wahlström, R. Fennoscandia seismicity and its relation to the isostatic rebound. 1993.

[58] U.S.NRC: Regulatory Guide 1.61. Damping values for Seismic Design of Nuclear Pow-
er Plants, Rev. 1, March 2007.

[59] U.S.NRC: Standard Review Plan 3.7.1, Seismic Design Parameters. Rev. 3, March
2007.

[60] U.S.NRC: Standard Review Plan 3.7.2, Seismic System Analysis. Rev. 4, September
2013.

[61] SS-EN 1992-1-1: 2005; Eurocode 2: Design of concrete structures-Part 1-1: General
rules and rules for buildings. Edition 1. November 2008.

[62] SS-EN 1990; Eurocode 0: Basis of structural design. December 2010.

[63] STUK-YVL Guide B.2; Safety classification of systems, structures and components in
nuclear facilities. Draft 4, 21 september 2011.

[64] American Nuclear Society National Standard, ANSI/ANS 2.26-2004: Categorization of

nuclear facility structures, systems and components for seismic design, 2004.

[65] IAEA Safety Report No 28. Seismic Evaluation of Existing Nuclear Power Plants,
IAEA, 2003.

[66] Betongbestämmelser för betongkonstruktioner (B5, B6 and B7), Statens betongkom-

mitté, 1968.

[67] Statens Planverks spännbetongnormer.

[68] Stålbyggnadsnorm StBK- N1, -N2, -N4.

73
SSM 2014:56
[69] IAEA Safety Guide No. NS-G-2.13: Evaluation of Seismic Safety for Existing Nuclear
Installations. IAEA 2009

[70] EPRI: A Methodology for Assessment of Nuclear Power Plant Seismic Margin. Rep.
NP-6041-SL, Rev 1. 1991.

74
SSM 2014:56
Appendix 1: List of figures

Page
Figure 2.1 – Horizontal design response spectra (5% critical damping) according 18
to RG 1.60 [15], scaled to 1g horizontal ground acceleration
Figure 2.2 – Example of a standard or site-independent response spectra, devel- 19
oped from a statistical data from recorded earthquakes, according to [17].
Figure 2.3 –Figure depicting the current safety-related versus nonsafety-related 22
SSC categorization scheme with an overlay of the new safety-significance cate-
gorization, according to [33].
Figure 2.4 – EUR design basis ground motion spectra (horizontal, 5% damping) 30
Figure 2.5 – Map of 90% probability of nonexceedence of horizontal PGA 36
(m/s2) in 50 years, corresponding to a mean return period of 475 years, from
[53].
Figure 2.6 - Envelope ground response spectra for a typical Swedish hard rock 37
site, corresponding to an annual exceedance probability of 10-5.
Figure 3.1 - Curves of annual seismic hazard (expressed in horizontal PGA) at 41
16%, 37%, 50% (median), 63% and 84% fractiles and the mean for the region
with the highest hazard in Sweden, according to [53].
Figure 3.2 - Comparison between the Swedish DBE ground response spectra (10- 43
5) and other international well-known DBE response spectra.

75
SSM 2014:56
SSM 2014:56
Appendix 2: List of tabels

Page
Table 1.1 – European level recommendations according to [1] and [2] 10
Table 1.2 – Topic item no 1 (natural hazards) relating to seismic hazard accord- 10
ing to [1] and [2]
Table 1.3 – Swedish action plan [7] regarding earthquake hazards 14
Table 2.1 – USNRC’s Recommendation 2 to enhance the reactor and spent fuel 24
safety in the U.S [36].
Table 2.2 – Maximum reference horizontal PGA (m/s2) corresponding to a 36
mean return period of 475 years for the Swedish NPP sites.
Table 2.3 – Seismic design classification for SSCs at Swedish NPP 37
Table 2.4 – Examples on various types of requirements on building elements 38
Table 4.1 – Modal damping ratios according to RG 1.61 [58], with stress level 48
definitions according to ASCE 4-98 [24].
Table 5.1 – Comparison of load combination rules between DNB [51] and 59
ASME Section III, Div 2 [9] for reactor containments.
Table 5.2 – Comparison of load combination rules between DNB [51] and ACI 59
349-06 [10] for other safety-related buildings.

77
SSM 2014:56
SSM 2014:56
Appendix 3: List of Acronyms

ALARP As Low As Reasonably Practicable

ASN Autorité de Sureté Nucléaire
DRB:2001 Design Rules for Buildings
BFS 2011:10 – Boverkets föreskrifter och allmänna råd om tillämpning av europeiska
EKS 8 konstruktionsstandarder (The Eurocodes)
BDBE Beyond Design Basis Earthquake
BOVERKET The Swedish National Board of Housing, Building and Planning
BWR Boiling Water Reactor
CDFM Conservative Deterministic Failure Margin
CFR Code of Federal Regulations
DBE Design Basis Earthquake
DRS Design Response Spectra
DEC Design Extension Conditions
DEE Design Extension Earthquake
EKS see BFS 2011:10 – EKS 8
EC European Council
ENSREG European Nuclear Safety Regulators Group
EPR European Pressurized Reactor
EPRI Electric Power Resaerch Institute
ETC-C French Association for Design, Construction, and In-Service Inspection
Rules for Nuclear Island Components, afcen, ETC-C EPR Technical
Code for Civil Works
HCLPF High Conficence of a Low Probability of Failure
HD Hazard exceedance probability
DNB Dimensionering av Nukleära Byggnadskonstruktioner (Design of Nu-
clear Buildings)
IAEA International Atomic Energy Agency
MHPE Maximum Historically Probable Earthquake
MVSS Multi-Venturi Scrubber System
NA National Annex
NPP Nuclear Power Plants
OBE Operating Basis Earthquake

79
SSM 2014:56
PGA Peak Ground Acceleration
PSR Periodic Safety Review
PSA Probability Safety Assessment
PSAR Preliminary Safety Analysis Report
RFS Fundamental Safety Rule
RG Regulatory Guide
RLE Review Level Eartquake
SAR Safety Analysis Report
SC Seismic Category
SDC Seismic Design Category
SKI Statens Kärnkraftsinspektion (today SSM)
SFP Spent Fuel Pool
SMA Seismic Margin Assessment
SPSA Seismic Probabilistic Safety Assessment
SRP Standard Review Plan
SSC Structures, Systems and Components
SSE Safe Shutdown Earthquake
SS-EN Swedish Eurocodes in general
SSM The Swedish Radiation Safety Authority (Strålsäkerhetsmyndigheten)
SSMFS Safety Regulations issued by SSM
TRAFIKVERKET The Swedish Transport Administration
TSN Transparancy and Security in Nuclear field
UHRS Uniform Hazard Response Spectra
USNRC United States Nuclear Regulatory Commission
WENRA Western European Nuclear Regulators’ Association
YVL Guides Guides from the Finnish Radiation and Nuclear Safety Authority

80
SSM 2014:56
SSM 2014:56
 2014:56 The Swedish Radiation Safety Authority has a
comprehensive responsibility to ensure that
society is safe from the effects of radiation.
The Authority works to achieve radiation safety
in a number of areas: nuclear power, medical
care as well as commercial products and
services. The Authority also works to achieve
protection from natural radiation and to
increase the level of radiation safety
internationally.

The Swedish Radiation Safety Authority works

proactively and preventively to protect people
and the environment from the harmful effects
of radiation, now and in the future. The Authority
issues regulations and supervises compliance,
while also supporting research, providing
training and information, and issuing advice.
Often, activities involving radiation require
licences issued by the Authority. The Swedish
Radiation Safety Authority maintains emergency
preparedness around the clock with the aim of
limiting the aftermath of radiation accidents
and the unintentional spreading of radioactive
substances. The Authority participates in
international co-operation in order to promote
radiation safety and finances projects aiming
to raise the level of radiation safety in certain
Eastern European countries.

The Authority reports to the Ministry of the

Environment and has around 315 employees
with competencies in the fields of engineering,
natural and behavioural sciences, law, economics
and communications. We have received quality,
environmental and working environment
certification.

Strålsäkerhetsmyndigheten
Swedish Radiation Safety Authority

SE-171 16 Stockholm Tel: +46 8 799 40 00 E-mail: [email protected]

Solna strandväg 96 Fax: +46 8 799 40 10 Web: stralsakerhetsmyndigheten.se