Page 209 - Pipeline Risk Management Manual Ideas, Techniques, and Resources

81186 Data Management and Analyses
changed, the standards can be easily updated. All comparisons ter evaluated using the cumulative risk techniques discussed in
between actual pipeline sections and standards will be instantly Chapter 15. This could then be judged against the effects of
updated an4 hence, based on the most current company policy. spending the same amount of money on, say, close interval sur-
It is reasonable to assume that whenever an instance of veys or new operator training programs.
noncompliance is found, a detailed explanation will be The costbenefit analyses will not initially produce absolute
required. The program can be designed to retrieve the whole values because this risk assessment program yields only rela-
record and highlight the specific item(s) that caused the tive answers. For a given pipeline system, relative answers are
noncompliance. usually adequate. The program should help the user decide
As policies and regulations change, it will be necessary to where his dollar spent has the greatest impact on risk reduction.
change the standards. Routines that allow easy changes will be Where absolute levels of spending are to be calculated tech-
useful. niques described in Chapters 14 and 15 will be needed.
Application 3: what-iftrials Application 5: detailed comparisons
A useful feature in the computer program will undoubtedly be In some of the above applications and as a stand-alone applica-
the ability to perform "what-if" trials. Here, the user can tion, comparisons among records will almost always be
change items within each index to see the effect on the risk pic- requested. A user may wish to make a detailed comparison
ture. For example, if air patrol frequency is increased, how between a standard and a specific record. She may wish to see
much risk reduction is obtained? What if an internal inspection all risk variables that exceed the standard or all variables that
device is run in this section? If we change our public education are less than their corresponding standard value.
program to include door-to-door visits, how does that influence Groups of records may also need to be compared. For exam-
the risk of third-party damage? ple, the threat of damaging land movements of all Texas
It will be important to preserve the original data during the pipelines could be compared with all Louisiana pipelines or the
what-if trial. The trial will most likely need to be done outside internal corrosion potential of natural gas pipelines could be
the current database. A secondary database of proposed actions compared with those for crude oil pipelines. Graphics would
and the resulting risk ratings could be built and saved using the enhance the presentation of the comparisons.
what-if trails. This second database might be seen as a target or
goal database, and it could be used for planning purposes. Additional applications
The program should allow specific records to be retrieved as
well as general groups of records. The whole record or group of Embedded or implied in some of the above applications are the
records will need to be easily modified while preserving the following tasks, which may also need to be supported by risk
original data. Comparisons or before-and-after studies will management sohare:
probably be desirable. Graphic displays will enhance these
comparisons. Due diligence-investigation and analysis of assets that
might be acquired
Application 4: spendingprioritization As an offshoot to the Project uppmvuls-as part of a regulatory process or com-
ranking list for relative risk assessment, it will most likely be pany internal, an examination of the levels of risk related to a
desirable to create rank-order lists for prioritizing spending on proposed project and the judgment of the acceptability of
pipeline maintenance and upgrades. The list of lowest scored those risks
sections from a corrosion risk standpoint should receive the Alternative route analyses-a comparison, on a risk basis, of
largest share of the corrosion control budget, for instance. The alternative routes of a proposed pipeline.
spending priority lists will most likely be driven by the rank- Budget setting-a determination of the value and optimum
orderedrelative risk lists, but there may be a need for some flex- timing of a potential project or group of projects from a risk
ibility. Spending priority lists for only natural gas pipelines may perspective
be needed, for example. The program could allow for the Risk communications-presenting risk results to a number
rearrangement ofrecords to facilitate ths. of different audiences with different interests and levels of
A special column, or field in the database, may be added to technical abilities.
tabulate the projected and actual costs associated with each
upgrade. Costs associated with a certain level of maintenance
(prevention) activities could also be placed into this field. This Properties of the software program
will help establish values of certain activities to further assist in
decision making. The risk assessment processes are often very dynamic. They
The user may want to analyze spending for projects on spe- must continuously respond to changing information if they are
cific pipeline sections. Alternatively, she may wish to perform to play a significant role in all planning and decision making.
costhenefit analyses on the effects of certain programs across The degree of use of this risk assessment is often directly
the whole pipeline system. For instance, if the third-party dam- related to the user friendliness and robustness of the software
age index is to be improved, the user may study the effects of that supports it.
increasing the patrol frequency across the whole system. The Properties of a complete risk assessment model are discussed
costs of the increased patrol could be weighed against the in Chapter 2 along with some simple tests that can be used as
aggregate risk reduction, perhaps expressed as a percentage measures of completeness and utility. Those same tests can be
reduction in the sum or the average of all risk values, but is bet- used to assess the environment that houses the risk model.