Over the past decade, regulators have pushed utilities to justify capital plans in terms of risk. In California, the Public Utilities Commission (CPUC) formalized this approach through its Risk Assessment Mitigation Process (RAMP), embedding risk-based planning into rate cases for investor-owned utilities. Similar approaches had emerged earlier in Europe, and today risk-informed planning is widely expected.
It wasn’t that long ago that infrastructure investment decisions were based on historical spend or engineering judgment alone. Requiring utilities to articulate expected outcomes has brought greater structure and transparency to capital planning, and that marks a huge shift for asset managers.
But this pressure has also stretched risk models beyond their limits. Utilities are now expected to show, with increasing rigor, how investment decisions affect customer outcomes. In many cases, the models used to quantify that impact were not designed to carry that weight.
When utilities submit rate applications, their goal is to demonstrate that planned spending is justified: the capital projects proposed are necessary to deliver safe, reliable, and affordable service. Increasingly, that justification is framed in terms of risk reduction: replace these assets to reduce this amount of risk. Yet how that risk is measured and how it maps to outcomes regulators care about, is usually unclear.
What planners, regulators, and customers ultimately focus on are outages (and the costs of reducing them). But estimating how likely an asset is to fail, and what happens if it does, is easier said than done. Not all asset failures lead to an outage, and even identical assets installed in different parts of the network can result in a huge range of outcomes if they fail. In practice, limited data and tools often reduce analysis to risk scores and abstract measures as “close-enough” proxies to what utilities are really trying to understand.
These measures fall short in many ways:
- Time variance: Over time, assets degrade, but the system also evolves. New loads, electrification, and climate change affect both failure probability and consequence. Modeling aging alone is no longer sufficient.
- Contingencies and mitigations: Just because an asset fails doesn’t mean there’s an outage. Spares, redundancy, and network configuration all shape the impact of asset failures, and the path to restoring power varies depending on how the grid was designed.
- System-level impacts: An asset failure impacts other assets too. A fault in a line may trigger a switch, load may be shed or redistributed, and increased load elsewhere raises the likelihood of additional failures.