Session: 07-03-02 Risk Management

Paper Number: 87309

87309 - Optimizing the Prioritization of First-Time Ilis Using Quantitative Risk and Machine Learning 

Inline inspections are one of the most effective and economical methods for managing the integrity of pipelines.
However, many older pipelines were not designed to accommodate the passage of inline inspection tools. Historically, many pipeline operators often prioritized which pipelines to make inspectable on a risk-basis. While this risk-based approach has many merits, it does not necessarily result in the maximum risk reduction for a given amount of budget as it does not
consider the amount of risk-reduction that is acheved by completing the inline inspection. Therefore, an optimized prioritization strategy should consider both the current (uninspected) risk as well as the amount of risk reduction.

Unfortunately, it is not possible to directly calculate the risk-reduction acheved from performing a first-time inline inspection as
the resultant risk after performing the inspection would be based on the quantity and severity of imperfections detected by the tool. To overcome this limitation, TC Energy conducted an exploratory analysis of numerous first-time Inline inspection results to identify key parameters and built a machine learning models that predict the expected risk impact from performing a first-time inline inspection. Several different machine learning algorithms (neural network, decision tree, random forest etc.) were trained in a supervised learning environment using TCE's quantitative risk assement results from both before and after performing an first-time inspection. The models were trained at a dynamic segment level and aggregated to an assessment path for model performance evaluation. The best-performing machine learning model was selected that accurately predicts the risk reduction achieved from performing a first-time Inline inspection on a route level.

These results demonstrate the risk-reduction of a first-time ILI can be accurately predicted before the
inspection is performed. By combining the traditional risk-based prioritization approach with the
predictive abilities to estimate risk-reduction will allow TCE to optimize the selection of first-time
inspections by maximizing the amount of risk reduction.

Presenting Author: Brenn Snider TC Energy

Presenting Author Biography: Brenn Snider has been a risk engineer with TC Energy's Innovation and Decision optimization team for the last 4 years and has been leading development on several analysis tool and engineering data analytics projects to expand the utility of TCE's quantitative risk assessment adding long term forecasting and estimating risk impacts of integrity work.