Agena Company Logo
leftl
Latest Publication
Free Download
You can download a free evaluation
of AgenaRisk Desktop, which contains the complete functionality,
here.
AgenaRisk users include:
You are here: Home » Case Studies » Optimal infrastructure replacement strategies in the energy industry
Optimal infrastructure replacement strategies in the energy industry
Problem
- The infrastructure assets (like gas pipes, oil platforms
and vessels) of major energy companies are expected to last for many
decades
- Repairing and replacing such assets is a continually ongoing process
- Since such assets must inevitably be repaired and replaced
in an incremental manner, a massive challenge for the companies is to
determine an optimal replacement strategy
- Specifically, the challenge is to determine, at any given time, which components should be replaced first
- Two of Agena's clients faced this challenge: one in the gas
sector was looking for an optimal strategy of pipe replacement, the
other in the oil and gas sector was looking for an optimal strategy for
vessel repair and modernisation
- In the case of the gas company there was a special concern
about the effects of data quality on failure rates. Sampling uncovered
the fact that some materials were wrongly classified. The risk of
failure needed to be reassessed to take into account this reduction in
data quality
Solution
- In the case of the vessel update problem a risk map was
developed in AgenaRisk to provide predictions of most likely critical
failures based on a range of objective indicators and expert judgement
- For the pipeline problem AgenaRisk was used to encapsulate existing material failure prediction models and known failure rates
- The generated models were then enhanced using sampled material misclassification statistics to compute "true" failure rates
- AgenaRisk predicted failure more accurately
Benefits
- AgenaRisk enhances the accuracy of the materials replacement planning process
- AgenaRisk is used to satisfy regulators that best practice has been applied and that data quality risks have been accounted for