Pipeline Risk Assessment—Risk Profiling
Traditional QRA relies heavily on historical incident rates. When applied to pipelines, this means that a pipeline—a population of pipe segments in varying environments—is modeled to behave as point estimates of populations of other, supposedly comparable pipelines—also collections of individual segments. As the foundation for a risk assessment, this is potentially very inaccurate. Let’s examine some of the implied assumptions embedded in this approach:
- Comparison population (collections of pipe segments from similar pipelines) is accurately represented by a single value—ignoring extremes is appropriate; ie, the fact that some segments of pipelines may carry much less or much more risk, is not germane.
- Subject pipeline, considered as a whole, is fairly represented by the comparison population
- All parts of subject pipeline behave similarly—the sum of the parts equals a value represented by the comparison
Obviously, such assumptions will very often be very incorrect. It is a classic error of using the behavior of a population to estimate the behavior of an individual. We are trying to understand the individual—ie, what are the risk issues for this particular segment of pipeline? Risk management occurs on the segment level.
The production of a profile forces the consideration of changing factors along the route—pipe properties (wall thickness, age, coatings, etc); operational parameters (pressure, temperature, etc); and the many environmental changes (soil types, population density, nearby structures, etc). Some of these factors will change every few feet along the pipeline. So, a key first step is to divide the pipeline into segments appropriate for risk analysis.
Segmentation
The full risk assessment solution to any variation in any risk variable is to ‘dynamically segment’ on that variation. This means that a new segment should be created for any feature or length of pipe that has different characteristics from its neighbors. Every change in any aspect creates a new segment, reflecting a different crack potential, corrosion potential, ability to resist external force, consequence, or any of dozens of other factors. This will generate many segments.
A potential criticism to this high-resolution approach is that ‘management of such a high count of segments is problematic’. The response is direct and intuitive—these segments are currently already being ‘managed’ in the real world. Each segment really does have failure issues distinct from the adjacent pipe and must be managed accordingly. The risk assessment should acknowledge this reality. Furthermore, with today’s computers, high segment counts causes no real efficiency issues.
Comparing Two Pipeline Risk Profiles; Risk (vertical axis) vs Length (horizontal axis),
Risk Management
The end game in risk assessment is of course risk management. In risk management practice as with risk assessment, a potentially high segment count emerging from full risk assessment should not be worrisome. The count will not be burdensome to processes that rely on the assessment results since profiles can be readily summarized. Proper aggregation allows a ‘summary’ risk value for any stretch of pipe, regardless of the number of changes in risk properties along that stretch.
Once a pipe section—a collection of segments—becomes a candidate for risk management, the initial drill-down quickly reveals the cause(s) of all risk issues. Risk mitigation plans are made accordingly. Proper aggregation is important since any improper approach can lead to masking of real issues. This will be fully discussed in a subsequent article.
Prior to any summarization, however, the full profile itself reveals some risk management challenges. Note the sample figure above. The two pipelines have different lengths and widely different risk profiles. Even if the risk estimates underlying these profiles is perfect, risk management solutions are not immediately obvious—which pipeline warrants more immediate attention? Longer lengths of ‘medium risk’ pipeline may be just as troubling as ‘risk spikes’. But without the profiles, understanding of the risk is incomplete.
Proper risk management cannot even begin until the risk profile is understood. That is why the profile is an essential element of pipeline risk assessment.