Performing Risk Assessments

The coal of risk assessment is to evaluate the risks to each asset using information gathered from previous steps. Risk assessment is the main part of the risk- management process. Here we need to integrate the cost incurred and the probabil­ity that an asset will be destroyed or damaged. We know the vulnerabilities of the understudied system, but integrating the probability and impact of these vulnerabil­ities on the system is the outcome of this process. The probability of occurrence is a matter of contemplation because it has many components. Just because there is a vulnerability does not mean an attacker will absolutely exploit it, so this is associ­ated with a probability. Furthermore, this attempt can be successful or not. Finally, the degree of success (or cost incurred in each state) itself has a random distribu­tion that makes probability assessment much more difficult. On the other hand, the impact of each vulnerability dictates how much the company should invest to miti­gate possible losses [11). For instance, damage resulting from an earthquake in a department store is a risk that can strongly affect its operation, but the impact of this vulnerability may be more severe on a plant that produces petrochemical mate­rials. This company should make more efforts to control and reduce the conse­quences of this phenomenon should it occur.

Although managers tend to use prescriptive measures to assess the risk, a great number of quantitative risk assessment (QRA) methods have been widely pre­sented: Cagno et al. [28]; Krueger and Smith [29]; Metropolo and Brown [30]; Jo and Ahn [31]; Sklavounos and Rigas [32]; Jo and Crowl [33]; Brito and De Almeida [34]; Suardin et al. [35]; Brito et al. [36]; and Han and Weng [37], among others.

Jo and Ahn [31] presented a QRA approach applicable in the planning and building phases of new pipelines or modifying existing ones. By using the informa­tion of pipeline geometry and population density from geographic information sys­tems, they estimated the parameters of fatal length and cumulative fatal length. The former is used to determine individual risk (the probability of loss of life at any special location because of all unwanted events) and the pipeline failure rate. The 'atter plus the failure rate can be used to estimate risk (the relationship between the frequency of an event and the number of its casualties).


Han and Weng [37] present an integrated QRA method that is composed of the Probability assessment of accidents, consequence analysis, and risk evaluation. This method analyzes consequences, including those outside and inside gas pipe­lines. Individual and societal risk caused by different accidents are related to the °L'tside risk of pipelines and economic risk derived from pressure redistribution
related to outside risk of pipelines. In their method, using the FTA, event tree anal ysis, and historical data or modified empirical formula, the expected failure rate per unit pipeline is calculated as follows:


 

(20.1)

In the above formula, p shows the expected failure rate per unit pipeline (1 /year km), pk demonstrates the basic failure rate per unit length of pipeline (1/year km)

and Kk shows the correction function for any failure cause, aha2,____________ These are the

arguments of the correction function, and the subscript к indicates each failure cause such as corrosion, construction defects, external interference, or ground movement.

The main sources of harm to gas pipelines from outside are toxic gas diffusion, jet flames, fireballs, and unconfined vapor cloud explosions. They measure all of these adverse effects by quantitative criteria and finally calculate the fatal probabil­ity for destruction.

According to the dose—effect relationship between the dose of the concrete harm­ful load as toxicity, heat or pressure and such recipient categories as death or inju­ries, the function of fatality probability unit Pr is defined to quantitatively describe the harmful load. Fatality probability unit can be used for the measurement of the damage from an accident and that is the critical basis of the calculation of death probability percentage, which is the final result of the accident consequence. [37J

(20.2)

The main source of harm from outside gas pipelines is economic loss. For exam­ple, it is possible to estimate PT from an accident using the following formula:

PT = a + b InIf

where empirical constant a represents the hazard only related to a studied harmlul load and b (also an empirical constant) represents the vulnerability of recipients to the load. If, for a given exposure time, is a dose of the load. Also note that a rela­tionship exists between the death probability percentage and the fatality probability unit, by which we can calculate death probabilities.

(20.3)

They finally do the risk evaluation process in which risk is defined as a function of the probability of an accident and its consequences. For example, they calculate economic risk as follows (considering the assumption that economic production is in direct proportion to the gas supply pressure):

E{R) — * (К * (Pnode Pnode, now))

where E(R) is the financial risk, K'* is the expected failure rate of the nodes (calcu­lated in the first step), Pnode is gas supply pressure of the nodes in a normal situa­tion, and Pnotje novv is the gas supply pressure after a disruption. Note that the term in second parenthesis represents financial loss.

Brito et al. [36] tried to design a multiattribute model for investigating risk in natural gas pipelines. An evolutionary version of Brito and Almeida [34], this paper used the ELECTRE TRI method integrated with utility theory to do so. Identifying the hazard scenarios, they divided the pipeline into a definite number of sections (each section has specific threats and vulnerabilities), and then the impact of each scenario on each section is calculated. After estimation of payoff sets (H, E, F) in which H stands for human, E for environment, and F for financial, the utility func­tion of each is elicited (using the utility theory). Data from consequence probabili­ties on human environment and financial and utility functions are used to calculate human, environment, and financial losses, and these data are combined with hazard scenario probabilities to calculate human, environmental, and financial risks for each section. Finally, a number of risk categories are defined and all of the identi­fied risks are put into these categories (using ELECTRE TRI).

For the consequence analysis of the outside pipelines, heat and overpressure are considered to calculate the individual risk and societal risk. The economic risk of the gas pipeline network is used to study the results of the inside gas pipelines. A sample of an urban gas pipeline network is used to show the presented integrated quantitative risk-analysis method.

For more information on guidelines for QRA, readers are referred to the TNO Purple Book [38].