I. Using this manual

II. Beginning risk management

III. Risk assessment models

IV. Designing a risk assessment model

V. Lessons learned in establishing a risk assessment program

VI. Examples of scoring algorithms

To get answers quick!

While the topic of pipeline risk management does fill the pages of this book, the process does not have to be highly complex or expensive. Portions of this book can be used as a “cookbook” to quickly implement a risk management system or simply provide ideas to pipeline evaluators. A fairly detailed pipeline risk assessment system can be set up and functioning in a relatively short time by just one evaluator.

A reader could adopt the risk assessment framework described in Chapters 3 through 7 to begin assessing risk immediately. An overview of the base model with suggested weightings of all risk variables is shown in Risk Assessment at a Glance, with each variable fully described in later chapters. A risk evaluator with little or no pipeline operating experience could most certainly adopt this approach, at least initially. Similarly, an evaluator who wants to assess pipelines covering a wide range of services, environments, and operators may wish to use this general approach, since that was the original purpose of the basic framework.

By using simple computer tools such as a spreadsheet or desktop database to hold risk data, and then establishing some administrative processes around the maintenance and use of the information, the quick-start applicator now has a system to support risk management. Experienced risk managers may balk at such a simplification of an often complex and time-consuming process. However, the point is that the process and underlying ideas are straightforward, and rapid establishment of a very useful decision support system is certainly possible. It may not be of sufficient rigor for a very detailed assessment, but the user will nonetheless have a more formal structure from which to better ensure decisions of consistency and completeness of information.

For pipeline operators

Whereas the approach described above is a way to get started quickly, this tool becomes even more powerful if the user customizes it, perhaps adding new dimensions to the process to better suit his or her particular needs. As with any engineered system (the risk assessment system described herein employs many engineering principles), a degree of due diligence is also warranted. The experienced pipeline operator should challenge the example point schedules: Do they match your operating experience? Read the reasoning behind the schedules: Do you agree with that reasoning? Invite (or require) input from employees at all levels. Most pipeline operators have a wealth of practical expertise that can be used to fine-tune this tool to their unique operating environment. Although customizing can create some new issues, problems can be avoided for the most part by carefully planning and controlling the process of model setup and maintenance.

The point here again is to build a useful tool—one that is regularly used to aid in everyday business and operating decision making, one that is accepted and used throughout the organization. Refer also to Chapter 1 for ideas on evaluating the measuring capability of the tool.

Step 1: Acquire a risk assessment model

A pipeline risk assessment model is a set of algorithms or “rules” that use available information and data relationships to measure levels of risk along a pipeline. A risk assessment model can be selected from some commercially available models, customized from existing models, or created “from scratch” depending on requirements.

Step 2: Collect and prepare data

Data preparation are the processes that result in data sets that are ready to be read into and used by the risk assessment model.

Step 3: Devise and implement a segmentation strategy

Because risks are rarely constant along a pipeline, it is advantageous to first segment the line into sections with constant risk characteristics (dynamic segmentation) or otherwise divide the pipeline into manageable pieces.

Step 4: Assess the risks

After a risk model has been selected and the data have been prepared, risks along the pipeline route can be assessed. This is the process of applying the algorithm—the rules—to the collected data. Each pipeline segment will get a unique risk score that reflects its current condition, environment, and the operating/maintenance activities. These relative risk numbers can later be converted into absolute risk numbers. Risk assessment will need to be repeated periodically to capture changing conditions.

Step 5: Manage the risks

This step consists of determining what actions are appropriate given the risk assessment results. This is discussed in Chapter 15.

Model design and data collection are often the most costly parts of the process. These steps can be time consuming not only in the hands-on aspects, but also in obtaining the necessary consensus from all key players. The initial consensus often makes the difference between a widely accepted and a partially resisted system. Time and resources spent in these steps can be viewed as initial investments in a successful risk management tool. Program management and maintenance are normally small relative to initial setup costs.

What is a model?

Armed with an understanding of the scenarios that compose the hazard (see Chapter 1 discussion of risk model building blocks), a risk assessment model can be constructed. The model is the set of rules by which we will predict the future performance of the pipeline from a risk perspective. The model will be the constructor’s representation of risk.

The goal of any risk assessment model is to quantify the risks, in either a relative or absolute sense. The risk assessment phase is the critical first step in practicing risk management. It is also the most difficult phase. Although we understand engineering concepts about corrosion and fluids flow, predicting failures beyond the laboratory in a complex “real” environment can prove impossible. No one can definitively state where or when an accidental pipeline failure will occur. However, the more likely failure mechanisms, locations, and frequencies can be estimated in order to focus risk efforts.

Some make a distinction between a model and a simulation, where a model is a simplification of the real process and a simulation is a direct replica. A model seeks to increase our understanding at the expense of realism, whereas a simulation attempts to duplicate reality, perhaps at the expense of understandability and usability. Neither is necessarily superior—either might be more appropriate for specific...