*3.2.1.1. Category 1 combinations of consequential hazards*

As already mentioned, for screening of causally related event combinations of natural hazards with other hazards, combinations of man-made or internal hazards with consequential natural hazards can be excluded.

For the site being investigated, only few individual natural hazards of the classes A, B, C, E, and F remain after the qualitative screening. For these, physically possible category 1 combinations have to be identified and screened out. As an exemplary result from [3], the following combinations with hydrological (Class B) hazards remain for quantitative screening (**Figure 2**):

#### *3.2.1.2. Category 2 combinations of correlated hazards*

For the qualitative screening of event correlations in the first step, the common causes have to be identified. These have to be systematically correlated to the different consequential events by phenomena. Typical correlations are possible between different hydrological hazards induced by precipitation (C1) as common cause. Examples of results from [3] are provided hereafter in **Figure 3**.

The final result of the qualitative screening of related hazards for the reference site analyzed is the following:

The longer duration external flooding hazards B3 and B4 resulting from snow melt or extreme precipitation can induce internal flooding hazards I2. All those external hydrological hazards with flooding potential not screened out individually besides B6a can in principle result in biological flotsam at the reference site.

**3.2. Hazard combinations screening**

136 Probabilistic Modeling in System Engineering

For a comprehensive Hazards probabilistic risk assessment according to the state-of-the-art, hazard combinations have to be included in the analyses. Since the number of hazards in *Lgen* and the resulting combinations is much too high to consider all combinations from the beginning in a generic manner, the screening of the hazard combinations starts from those hazards (*L***total**,**individual**) which cannot be qualitatively excluded at the nuclear power plant site being

In order to limit the analytical effort, at least for related hazards in a first step only first order combinations are qualitatively as well as quantitatively screened. For those combinations not screened out, potential second order combinations are identified and screened out. If there are still combinations remaining after screening, this process is repeated for the next order of

From the list of individual hazards which may occur at the plant under investigation, different types of hazard combinations (consequential, correlated, and unrelated ones) involving the remaining individual hazards after qualitative screening for rough or detailed analysis (stored in *L***rough,individual** and *L***detail,individual**) are identified. For these site- and plant-specific hazard combinations identified, the qualitative and quantitative hazards screening steps have again

As already mentioned, for screening of causally related event combinations of natural hazards with other hazards, combinations of man-made or internal hazards with consequential

For the site being investigated, only few individual natural hazards of the classes A, B, C, E, and F remain after the qualitative screening. For these, physically possible category 1 combinations have to be identified and screened out. As an exemplary result from [3], the following combinations with hydrological (Class B) hazards remain for quantitative screening (**Figure 2**):

For the qualitative screening of event correlations in the first step, the common causes have to be identified. These have to be systematically correlated to the different consequential events by phenomena. Typical correlations are possible between different hydrological hazards induced by precipitation (C1) as common cause. Examples of results from [3] are provided hereafter in **Figure 3**.

The final result of the qualitative screening of related hazards for the reference site analyzed

analyzed, and this in turn significantly reduces the hazards' screening effort.

combinations as long as there are combinations not yet screened out.

*3.2.1. Qualitative screening of hazard combinations*

*3.2.1.1. Category 1 combinations of consequential hazards*

*3.2.1.2. Category 2 combinations of correlated hazards*

to be carried out (cf. **Figure 1**).

natural hazards can be excluded.

is the following:

• External flooding hazards from various origins and with differing duration B2, B3, and B4 can occur as correlated events from the same root cause or together with the meteorological hazard C1 (precipitation) correlated, for example, by extreme weather conditions. In addition, flash flood B2 and heavy rainwater flooding B4 can occur correlated to F1 (subaerial slope instability).


**Figure 2.** Result of the qualitative screening for combinations of those hydrological hazards not screened out at the site being investigated with other hazards.


In the first step, all those individual hazards not qualitatively screened out can be considered for this third category of combinations. This results in a relatively long list of category 3 com-

Natural Hazards: Systematic Assessment of Their Contribution to Risk and Their Consequences

http://dx.doi.org/10.5772/intechopen.76503

139

In the example of a German nuclear power plant site, given cutoff values from the German regulation [4] have been applied to the occurrence frequency and to the damage frequency. The qualitative screening for category 1 combinations provides the following result for causally related combinations at the reference site with the hydrological hazards B2, B3, B4, B6a, B8, and B9a: Combinations of these hazards with E6 (biological flotsam) and F1 (subaerial slope instability) have been screened out quantitatively. Therefore, no category 1 combinations remain after the quantitative screening; higher order combinations are also not to be

The following category 2 combinations that remain after qualitative screening for the reference plant site have been analyzed: A meteorite fall (D2) can cause correlations of I2 (internal flooding) with B6a, B8, or B9a. I2 can also occur together with B6a, B8, or B9a as consequence of man-made explosions (Z1, Z4, Z6, Z11, or Z13). Resulting from a common cause such as a thunderstorm precipitation (C1) or F1 (subaerial slope instability) can be observed correlated with B2 or B4. All these correlations have been excluded quantitatively for the reference

For category 3 combinations B2, B3, B4, B6a, B8, or B9a have to be assumed to occur independent from other hazards. Such combinations have only to be analyzed, if their occurrence frequencies exceed a given cut-off value under consideration of the durations of the individual hazards. According to this argumentation, in the example of screening for hydrological hazards for a given German site, only combinations of B2 with B3 or B4 finally remain for further

It could be demonstrated that the remaining number of hazard combinations is significantly

The plant model for risk assessment of the facility under consideration needs to be extended by taking into account those hazards and hazard combinations remaining after screening. It has to be analyzed, which structural elements, plant operational components, or even complete systems maybe impaired in their required function (so-called initiating events, IEs). That also requires to extend the original list of risk-relevant functional unavailabilities, the so-called basic events (BEs) in the plant model by those ones related to the hazards and hazard combinations to be considered as well as by the corresponding failure dependencies. This

lower after qualitative and quantitative screening of hazard combinations.

binations, for which qualitative screening is necessary.

*3.2.2. Quantitative screening of hazard combinations*

assumed.

facility.

detailed risk analysis.

**4. Detailed analyses**

requires another two analytical steps:

**Figure 3.** Correlations of the hydrological hazards B2, B3, B4, B6a, B8, and B9a not screened out qualitatively with other hazards due to a common cause, from [3].

#### *3.2.1.3. Category 3 combinations of unrelated hazards*

Hazards that occur independently of each other have no common cause and are unrelated. The simultaneous occurrence is in general highly unlikely and is therefore investigated on an international level mainly for hazards of longer duration. In the example of the hydrological hazards not screened out qualitatively (B2, B3, B4, B6a, B8, B9a) a broad majority of combinations with unrelated events is not possible or very unlikely.

In the first step, all those individual hazards not qualitatively screened out can be considered for this third category of combinations. This results in a relatively long list of category 3 combinations, for which qualitative screening is necessary.
