by W. Kent Muhlbauer, WKM Consulting, Austin, Texas, US
Simply, leak detection can reduce risk by reducing consequences – it has no effect on failure prevention. So, of the two parts of risk – probability and consequence – it plays a role only in potentially reducing the level of damage after a spill or release has occurred.
A few years ago, this column discussed a guiding equation for understanding consequence potential (“Risk is PoF x CoF – Where should the focus be?” September 2016). That article noted that this equation gives guidance on options to reduce consequence potential. Here is an excerpt:
Consequence of Failure (CoF) associated with any pipeline release can be efficiently understood as being comprised of four parts acting in a dependent relationship: CoF = P × V × D × R
Where
P = product hazard (toxicity, flammability, etc)
V = release quantity (quantity of the liquid or vapor release)
D = dispersion (spread or range of the release, including early- and late-ignition scenarios)
R = receptors (all things that could be damaged by the release).
The dependent relationship is illustrated in the use of the multiplier in this equation. Each factor can have a dramatic impact on total CoF. Any directional changes – higher or lower – in any of these four variables will generally forecast the change in consequence potential.
To reduce overall consequence potential, any single component can be reduced. If any goes to zero, then there are zero consequences. This helps us to understand the risk management options that focus on CoF. Consistent with this guiding equation, we can reduce CoF and, hence, risk, by actions targeting any of these four, such as:
changing the product
reducing product pressure or flowrate
limiting dispersion (e.g. secondary containment, boom deployment, etc.)
reducing spill quantities (e.g. leak detection, remotely operated equipment, etc.)
relocating people, property, environment.
Of course, these have varying levels of practicality. Even the more practical opportunities may be of limited benefit. Their ability to reliably reduce CoF are highly location- and scenario-specific. In some instances, they play a significant and valuable role; in others, much less so.
LDS capability analysis.
Of the four key determinants of consequence potential, leak detection can reduce the spill quantity and, in some scenarios, the dispersion. But is it a good choice for efficient risk reduction?
US regulations give much latitude in what risk reduction actions an operator employs. However, leak detection is specifically mandated as a potential risk management option that must be evaluated. Regulatory auditors can and do insist on reviewing these evaluations. Some operators have difficulty assessing their current capabilities.
Then, as related regulatory mandate, a formal decision process determining the sufficiency of that capability is also required. So, additional pipeline operators are required to assess leak detection capabilities and have a process to consistently judge when that capability should be enhanced. Let’s examine each of these facets.
A leak detection capability analysis must recognise two important aspects. First, there are almost always multiple ‘leak detection systems’ in place and, second, each has varying abilities to find leaks of varying sizes. So, step one is to identify all the systems. System types often include:
SCADA based systems such as monitoring via alarms (pressure, flowrate, temperature, etc.), transient models, mass balances, etc.
Field based systems such as staffing, patrol, sensors, ground water monitoring, and even passerby reporting.
Each system is sensitive to either leak rate or spilled volume. Many can find high leak rates. The noise, smell, vapor clouds, pressure drops, flows over ground surface, etc, from high rates are readily detectable.
At the other extreme, some small leak rates are undetectable until a certain volume has been released. Only a puddle, a sheen on water, ground water contamination or other visible evidence allows detection.
The ability to detect various leak rates can be plotted as a curve for each system and a composite curve can then be built that shows the combined capabilities of all systems at each leak rate. If plotted on a graph of leak rate versus time, the area under the composite curve is the volume released before detection.
Having analysed the family of curves representing current leak detection capabilities, an important input into the sufficiency determination emerges. The area under the composite curve provides insight into the amount of consequence that could theoretically impacted by improved leak detection capabilities. That’s the beginning of a cost/benefit analysis – the most defensible way to decide sufficiency.
Proposed leak detection enhancements will generate additional curves. Any proposed improvement to leak detection capabilities will generally focus on a specific part of the leak-rate vs time-to-detect curve. The difference between the current composite curve and the potential composite curve shows the amount of product loss that is avoided by the enhancement.
The volume reduction must be monetised to complete the cost/benefit analysis and for some products, a cost savings is readily assigned to this avoided volume loss. The savings realised may be simply the value of the lost product itself and cleanup or remediation expenses avoided, while for other products, scenarios involving ignition, fire, and thermal damages must be factored into potential consequence reduction.
A good risk assessment should be able to quantify the change in risk associated with any potential leak detection improvement. Ideally, this will be expressed in terms of expected loss in $/km-year.
Finally, the costs of the improvement in leak detection must be factored in. That cost must consider initial and ongoing expenses and apply those to the miles of pipeline and the time period for which the improvement provides benefits. So, total costs are expressed in the same $/km-year units as avoided loss (risk).
Recognising the extent of the leak detection improvement, over the lengths and time periods reveals some interesting things. Even a very expensive enhancement, such as a full, SCADA-based transient model, can be cost effective.
If such a computational system covers many kilometres of pipeline for many years, the per km-year cost could be an efficient way to reduce risk. On the other hand, a seemingly inexpensive solution applied very narrowly, in terms of lengths and time periods, may be hard to justify.
As with many issues in risk management, performing the calculations often results in new and interesting insights. This is, of course, the central intent of formal risk management –revealing the nuances that can optimise decision-making.
This article was featured in the July 2020 edition of Pipelines International. To view the magazine on your PC, Mac, tablet or mobile device, click here.
For more information visit the Pipeline Risk website.
If you have news you would like featured in Pipelines International contact Journalist Sophie Venz at svenz@gs-press.com.au