Page 208 - Pipeline Risk Management Manual Ideas, Techniques, and Resources

Computer environments 811 85
Applications of risk management Oklahoma. A rank-ordered list for natural gas lines in
Oklahoma is generated.
A critical part of the role of risk assessment is of course its role 3. The corrosion control department wants to see a rank order-
in risk management. Some potential user applications are ing of all sections, ranked by corrosion indexes, lowest to
discussed in the following subsections. highest. All pipeline sections are ranked strictly by corro-
sion index score.
Application I: risk awareness 4. A pipeline company wants to compare risks for LPG
pipelines in Region 1 with crude oil pipelines in Region 2.
This is most likely the driving force behind performing risk Distributions of pertinent risk scores are generated. From
evaluations on a pipeline system. Owners andor operators the distributions, analysts see the relative average risks, the
want to know bow portions of their systems compare from a variability in risks, and the relative highest risks, between
risk standpoint. This comparison is perhaps best presented in the two pipeline types.
the form of a rank-ordered list. The rating or ranking list should
include some sort of reference point-a baseline or standard to Application 2: conzpliance
be used for comparisons. The reference point, or standard, gives
a sense of scale to the rank ordering of the company’s pipeline Another anticipated application of this program is a compari-
sections. son to determine compliance with local regulations or with
The standards may be based on: company policy. In this case, a standard is developed based on
the company’s interpretation of government regulations and on
the company policy for the operation ofpipelines (if that differs
Governing regulations, either from local government agen-
cies or from company policies. So, the standard is the risk from regulatory requirements). The computer program will
score of a hypothetical pipeline in some common environ- most likely be called on to search the database for instances of
ment that exactly meets minimum requirements of the regu- noncompliance with the standard(s).
To highlight these instances of noncompliance, the program
lations. must be able to make correct comparisons between standards
A pipeline or sections that are intuitively thought to be safer and sections evaluated. Liquid lines must be compared with liq-
than the other sections. uid regulations; Texas pipelines must be compared with Texas
A fictitious pipeline section-perhaps a low-pressure nitro- regulations, etc.
gen or water pipeline in an uninhabited area for a low-risk If the governing policies are performance based (“. . . corro-
score, perhaps a high-pressure hydrogen cyanide (very sion must be prevented . . .,” “. . . all design loadings anticipated
flammable and toxic) pipeline through a large metropolitan and allowed for. . .,” etc.), the standard may change with dif-
area for a high-risk score.
fering pipeline environments. It is a useful technique to pre-
define the pipeline company’s interpretations of regulatory
By including a standard, the user sees not only a rank- requirements and company policy. These definitions will be
ordered list of his facilities, he also sees how the whole list the prevention items in the risk evaluation. They can be used
compares to a reference point that he can understand. to have the computer program automatically create standards
Ideally, the software program to support Application 1 will for each section based on that specific section’s characteristics.
run something like this: Using the distinction between attributes and preventions, a
Data are input for the standard and for each section evalu- floating standard can be developed. In the floating standard, the
ated. The computer program calculates numerical values for standard changes with changing attributes. The program is
each index, the leak impact factor (product hazards and spill designed so that a pipeline section’s attributes are identified
scores), and the final risk rating for every section. Any of these and then preventions are assigned to those attributes based on
calculations may later be required for detailed comparisons to company policies. The computer can thus generate standards
standards or to other sections evaluated. Consequently, all data based on the attributes of the section and the level of preven-
and intermediate calculations must be preserved and available tions required according to company interpretations. The stan-
to search routines. The program will likely be called on to pro- dard changes, or floats, with changes in attributes or company
duce displays of pipeline sections in rank order. Sections may policy.
be grouped by product handled, by geographic area, by index,
by risk rating, etc. Example 8.1: Compliance
Examples o/risk data analyses There are countless ways in A company has decided that an appropriate level of public
which the risk picture may need to he presented. Four examples education is to be mailouts, advertisements, and speaking
of common applications are: engagements for urban areas, and mailouts with annual
landowneritenant visits for rural areas. With this definition, the
I. Pipeline company management wants to see the 20 pipeline computer program can assign a different level of preventions
sections that present the most risk to the community. A list is for the urban areas compared with rural areas. The program
generated, ranking all sections by their final relative risk generates these standards by simply identifying the population
number. A bar chart provides a graphic display of the 20 density value and assigning the points accordingly.
sections and their relative magnitude to each other.
2. Pipeline company management wants to see the 20 highest By having the appropriate level of preventions pre-assigned
risk pipeline sections in natural gas service in the state of into the computer, consistency is ensured. When policy is