engineered safety features, and status of primary and secondary containment
(like isolation/ bypass failure).
Accident Progression and Containment Analysis
The purpose of the accident progression and containment analysis is to track
the physical progression of the accident further from the PDSs until it is
concluded that no additional release of radioactive material from the
containment building will occur. The analysis tracks the impact of the accident
progression on the containment building structure, with particular focus on
the threat to containment integrity posed by pressure and temperature loadings
or other physical and chemical phenomena.
For example, for boiling water reactors (BWRs), the phenomena can be divided
into three stages: (i) phenomena within reactor pressure vessel (RPV) and
reactor coolant system (RCS), (ii) phenomena within reactor cavity/vault and
(iii) phenomena within containment building. For pressurized heavy water
reactors (PHWRs) these can be divided into four stages: (i) phenomena within
reactor coolant circuit, (ii) phenomena within calandria, (iii) phenomena within
the calandria vault and (iv) phenomena within containment building.
Some of the accident phenomena are: core-concrete interaction (CCI), high
pressure melt ejection (HPME), direct containment heating (DCH), steam
explosion, hydrogen generation, deflagration and detonation etc. These are
highly complex set of physical and chemical phenomena. In Level 2 PSA these
phenomena are not described fully as for many of these phenomena, complete
understanding is still going on as research activities. However, the phenomena
are placed within overall structure of Level 2 PSA to account for their potential
effects on the containment integrity. These approximations contribute to
uncertainty in Level 2 PSA results.
Development of Accident Progression Event Trees
The plant-specific analyses of the progression of severe accidents are
performed using Level 2 PSA computer codes or in-house developed computer
codes. In addition, if available, literature for similar plants and containments
could be used as a basis for establishing an adequate framework for the accident
progression event trees (APETs)/containment event trees (CETs). These are
similar to event tree models used in Level 1 PSA in principle. Once the APETs
are developed, the next step is to assign proper nodal probabilities to each
branch point in APETs. The determination of conditional probabilities (at
each branch point) is based on deterministic analyses and expert judgement.
The quality of this expert judgement is dependent on the analyst’s current
state of knowledge to a particular issue.
Source Term Analysis
A source term (ST) is defined as the quantity, timing, duration and
characteristics of the release of radioactive material to the environment following
a postulated severe reactor accident. The core of a power reactor contains
several million curies of radioactivity of hazardous nuclides built up during
equilibrium power operation. Several barriers (e.g. fuel matrix, fuel cladding,
reactor coolant system and reactor containment) must be breached before any
significant part of this radioactivity can be released to the environment.
Establishing the timing and nature of the breaching of these barriers is an
essential part of ST analysis.
The aim of the ST analysis is to find out what part of the activity originally
released from the core will be retained in different areas of the plant, and what
will escape. Early containment failures are usually associated with high source
terms. On the other hand, a delayed containment failure will ensure that a
good part of the radioactivity reaching the containment is retained therein. It
must be pointed out that the possibility of containment bypass must be
established, for that would result in high ST scenarios even when the reactor
has a strong containment and its integrity is not lost.
Estimation of Frequencies for Release Categories
Once the APETs are developed, reliability of containment-engineered features
is to be evaluated and integrated with APETs along with the other nodal
probabilities. The analyst with the help of computer codes, evaluate the
frequency of the accident sequences originated from all the PDSs. As in Level-
1 PSA, here also large number of end states would result, some of which are
identical in terms of key release attributes. Depending upon the similar
characteristics, these end states are grouped together in terms of different
release categories. For each of these release categories, the corresponding
frequency can be calculated and from these the risk measure of Level 2 PSA
and large early release frequency (LERF) can be evaluated.
6. ELEMENTS OF LEVEL 3 PSA
Level 3 PSA takes input from the Level 2 PSA results and estimate the risk to
the public by performing few more specific steps. The major steps involved in
Level 3 PSA are described below.
Interface with Level 2 PSA
The starting point of Level 3 PSA is the ‘Source Term’ information provided
by a Level 2 PSA. This information is provided for each of the representative
accidents to be assessed, obtained by grouping accidents with similar release
characteristics together. The important attributes of the source term such as
timing and duration of releases, height of release and thermal energy associated
with releases are the direct input for the second step of the Level 3 PSA.
Other characteristics of release (the physical form and chemical properties of
radionuclides) are assumed to be constant in all release cases. It is assumed
that they are released in oxide form as aerosol particle with 1 mm activity
median aerodynamic diameter (AMAD) or one can use available distribution
of aerosol size, except noble gases, which appear in elemental form, and iodine,
which may appear in elemental, organically bound and particulate forms.
Atmospheric Dispersion and Deposition
Material released to the atmosphere is transported downwind and dispersed
according to normal atmospheric mixing processes. The diffusion-transport
equation is commonly used for estimating dispersion in the atmosphere. For
this, meteorological data is required to be obtained. It is a normal practice to
use the meteorological data from the meteorological station nearest to the
release point. Data compiled at other stations may, however, be acceptable if
they are representative of the general condition experienced by the plume.
The atmospheric dispersion and dose calculation are repeated for a large
number of sequences of conditions selected from the meteorological data file
used to predict the full distribution of consequences, which may occur. Ideally
the calculation may be performed for every possible sequence of weather
conditions in the data file, in other words a weather sequence at each hour on
the file. It is neither practicable nor necessary to consider every such sequence.
Instead, the one or more year’s data is sampled in such a way that a truly
representative set of weather sequences is selected. The selection should be
made in such a way that the sequences chosen represent the complete set of
possible sequences, and yield the correct probability distribution of
Once these data are obtained, atmospheric dispersion and deposition of the
radionuclides are modeled. Several models have been developed for this
purpose using a variety of boundary conditions and simplifying assumptions.
Many simple theoretical formulations of dispersion predict that concentration
profile will have a Gaussian shape. Additionally, they assume that the downwind
transport goes along a straight line. Although the assumption of simple theories
does not hold for real atmosphere, the Gaussian shapes have been found
empirically to be approximately valid in many situations and it forms the basis
of the Gaussian plume model which has been, and still is, widely used in
Identification of Different Exposure Pathways
There are six principal pathways (i.e. external b and g irradiation from the
radioactive materials in the cloud, inhalation of radioactive materials in cloud,
external dose from radioactive material deposited on skin and clothing, external
g irradiation from deposited radionuclides on ground, inhalation of resuspended
material and ingestion dose) by which people can accumulate a radiation dose
after an accidental release of radioactive materials to the atmosphere. For
each pathway a dosimetric model is required to convert the concentration of
radionuclides in the atmosphere, on the ground, in foodstuffs, or on skin and
clothing to dose to humans.
Once the exposure pathways are identified, the doses received by the humans
are required to be calculated from each of these exposure pathways to find out
the risk. Different dose conversion factors such as attenuation factor, shielding
factor, re-suspension factor etc. are used for this purpose.
A variety of possible countermeasures or protective actions may be taken
following an accidental release to reduce the impact of the accident on the
environment and the public. For realistic estimate of the exposure of the
population, appropriate account of these countermeasures is taken in the risk
evaluation task of Level 3 PSA.
The various protective actions available fall broadly into two categories
depending upon the time at which they are implemented and the effects for
which they are designed to mitigate: short-term protective actions and long-
Short-term countermeasures include sheltering, evacuation, the issuing of
stable iodine tablets, and the decontamination of people. The primary objective
of such measures is to limit the exposure of the population to both internal and
external irradiation with the intention of preventing deterministic effects and
minimizing risks of stochastic effects.
Long-term countermeasures include changes to agricultural practices, deep
ploughing, alternate feed, cesium binders, alternative crops and alternate
production. Long-term countermeasures are designed to reduce chronic
exposure to radiation, both externally from deposited material and internally
from ingestion of contaminated food, with the intention of reducing the
incidence of late health effects.
Estimation of Health Effects and Other Risks
The exposure of individuals to ionizing radiation can lead to health effects,
which are generally classified as either ‘deterministic’ or ‘stochastic’.
Deterministic effects and stochastic effects are often referred to as ‘early’
effects and ‘late’ effects, respectively. Effects observed in exposed individuals,
i.e., deterministic effects and cancers are termed ‘somatic’ effects, while those
observed in their descendants are known as ‘hereditary’ (genetic) effects.
Different models are available for this purpose.
The most common risk measure of Level 3 PSA is presented in the form of
complementary cumulative distribution functions. An example is shown in
FIG. 2 : AN EXAMPLE OF CUMULATIVE DISTRIBUTION
The ordinate of these cumulative distribution function (CCDF) is the probability
of equaling or exceeding the consequence magnitude indicated by the curve.
The abscissa is the numerical value of the consequence, which may be any of
the effects, such as number of early fatalities or injuries, the number of latent
cancer fatalities, the size of the area contaminated to such a level that
decontamination is required, and so on. Logarithmic scales are employed on
both axes to accommodate the wide range of frequencies and consequences
involved. CCDFs are often used as a measure of public risk. In addition, the
expected (mean) value of the CCDF (which corresponds to the integral of the
CCDF) is frequently used as a summary measure of risk.
CONDITIONAL PROBABILITY OF ³³³³³ X
X MAGNITUDE OF CONSEQUENCE
7. COMPUTER CODES USED FOR PSA
A number of computer codes and software packages are currently used for
performing PSA. Typically, an integrated software package is used in the
Level 1 PSA analyses for the development and storage of system models,
sequence models, failure data, and sequence quantification. Level 2 PSA and
Level 3 PSA analyses will also require the use of large computer codes. Finally,
smaller pieces of software may be used for special analyses, conversion or
transport of data. Increasingly, integrated software packages are developed
and used, covering almost all levels and tasks of a PSA.
In order to ensure quality assurance (QA) for the PSA, all computer codes
used in the development of the PSA must be verified and validated, either in
the course of their development or by the PSA group. Computer codes that
are available commercially may be verified and validated by the code developer.
For software that is not commercially procured but, for example, written internally
in the PSA organization, a verification, validation and QA process should be
Computer Codes used for Level 1 PSA
Some of the computer codes, which are used for Level 1 PSA studies are:
Risk Spectrum PSA Professional
Computer codes used for Level 2 PSA
Some of the deterministic computer codes, which are used for assessing the
accident consequence for input in Level 2 PSA studies are:
Computer Codes Used for Level 3 PSA
Some of the computer codes which are used for Level 3 PSA studies are:
8. APPLICATIONS OF PSA
PSA can be used to explore the risk significance of various aspects of NPP
design and operation, the risk impact of changes in NPP design or modification
of operating procedures and for the evaluation of the abnormal events that
occur at NPP. To use PSA for such applications, PSA should be performed
with stat-of-the-art methodology and should be updated with respect to the
changes/modifications in plant configuration and reliability data obtained from
the plant experience. The following are the PSA applications:
PSA Applications for Design of NPPs
This is the most important application of PSA as it helps to identify design
deficiencies that can challenge plant safety during the operation phase. PSA
can be used at design stage of NPPs. However, it is to be understood that the
PSA for a new plant design would contain substantial uncertainties due to
incomplete information of design details, limited database, reliance on
preliminary procedures, preliminary thermal-hydraulic analyses, etc. Hence,
the PSA analysis at design stage should be supported with uncertainty and
sensitivity studies. Some PSA applications during NPP design are:
To Support NPP Design
A PSA provides a fully integrated model of the entire plant that can be used to
examine the risk from a variety of possible initiating events (e.g. transients,
LOCA, support system failures, etc.). The model combines front-line safety
systems and support systems in a manner that allows designers to identify
the risk significance of important inter-system dependencies. The PSA allows
designers to examine the significance of single failures and multiple failures,
and to determine the risk importance of ‘safety’, ‘safety related’ and ‘non-
safety’ systems. Consideration of only a limited set of design basis accidents
and application of traditional deterministic design criteria for individual safety
functions, systems, and components do not provide the same benefits as the
combination of traditional approaches and PSA.
To Support NPP Upgrade and Backfitting Activities and Plant Modifications
One of the major goals of PSA is to assess the level of safety of existing plants
and to identify design weaknesses that need to be corrected by plant
improvements (backfits). If the frequency of core damage or severe off-site
releases is largely dominated by a very limited number of accident sequences,
effective backfits may be proposed to prevent or to mitigate these scenarios.
Proposed backfits may involve changes to system designs and installation of
new hardware. They may also involve changes to operational procedures,
development of specific accident management procedures, etc. PSA
evaluations can also be used to demonstrate which modifications are
acceptable and to compare or suggest possible alternatives.
PSA Applications for Operation of NPPs
PSA can provide valuable insights for the NPP operation. It provides the
framework for risk-informed operational activities. Some of the applications of
PSA in NPP operation are as follows:
Tool as a Safety/Risk Monitor
A safety/risk monitor is a plant specific real-time analysis tool used to determine
the instantaneous risk based on the actual status of the systems and
components. At any given time, the safety monitor reflects the current plant
configuration in terms of the known status of the various systems and/or
components, e.g. whether there are any components out of service for
maintenance or tests. It is necessary to control the risk due to plant
configurations during power operation as well as during the shutdown state
of the plant.
There are two main tasks in the risk based configuration control, risk planning
and risk follow-up. Risk planning is a forward-looking application of PSA and
it consists of supporting the preparation, planning and scheduling of plant
activities and configurations. This application can be performed with an on-
line or off-line PSA model. Risk follow-up involves the online use of the PSA
by plant personnel in order to keep the risk due to actual configurations, plant
activities and unanticipated events, at an acceptable level.
Evaluation of Technical Specifications for Operation
Technical specifications for operation (TS) are operating rules for NPPs that
are approved by the regulatory authority. The technical specifications define
limits and conditions for operations, testing, and maintenance activities as a
way to assure that the plant is operated safely in a manner that is consistent
with the plant safety analyses. The TS define limiting conditions for operation
(LCOs) and surveillance requirements (SRs).
LCOs also define equipment operability requirements and allowed outage
times (AOTs). Surveillance requirements define the safety system and safety
related/supporting systems testing requirements and the surveillance test
intervals (STIs). PSAs can be used to develop quantitative bases for optimized
limits on equipment AOTs, STIs and testing strategies.
Periodic Safety Review
A safety assessment process consists of identifying safety issues, determining
their safety significance and making decisions on the need for corrective
measures. This has to be done continuously during the life of the plant. In
Documents you may be interested
Documents you may be interested