Applying an analytics capability for a more proactive approach to decision-making

Northern Gas Networks (NGN) operates and manages one of the eight gas distribution networks (GDNs) in Great Britain. They distribute gas to 2.7 million homes and businesses in the North of the country.

NGN approached Business Modelling Applications (BMA) to work on a collaborative project to explore and develop a more data-driven method of analysis for their maintainable assets, such as domestic governors and pressure reduction system assets. Up until this point, the method for determining probability of failure and deterioration for these assets was subject matter expert elicitation. However, this did not fully account for the effect of maintenance regimes on the lifecycles of the assets, nor did it account for the interdependency of risk and resilience within the systems many of these assets operated in.

BMA’s role in the project was to:

• Collaboratively redefine failure to move to more data-driven methods.
• Build systems models to quantify the interplay between maintenance and capital investment.
• Explore if investment decision making from a systems rather than an asset perspective could produce additional value and/or cost efficiencies.

Delivering an energy capability that meets the United Kingdom’s long-term objectives on safety, sustainability, resilience and affordability is a challenging yet critical task. NGN are meeting this challenge by developing an analytics capability that supports proactive approaches based on data-driven decision making.

NGN were working with the other gas distribution networks (GDNs) on a cross-network monetised risk model (NOMs). At this time, subject matter expert (SME) elicitation was being used to determine failure and deterioration curves for assets such as domestic governors and Pressure Reduction Stations (PRS) assets.

“These approaches and tools remove blind spots which are endemic to current bottom-up asset investment strategy and planning that do not capture the end-to-end system and the internal and external inter-dependencies between interventions”

Ian Gray

Associate Director, Arup

Elicitation was undertaken as there was little failure data available to build robust statistical models. This was due to the high reliability of these assets resulting from the maintenance regimes carried out on them.

However, there are a number of challenges with elicitation for maintainable assets, even though this calls on significant subject matter expert knowledge and experience. Multiple interdependent factors are at play within integrated systems, combined with successive interventions being carried out on the assets over time.