From age-based maintenance to risk-based decisions: the case for health and criticality indices in power grids

As grids become more complex and capital-intensive, traditional maintenance approaches are no longer sufficient. This article explores how utilities can combine asset health and criticality indices to move toward risk-based decision-making—and why this shift is becoming essential.

Most utilities still maintain their networks using a familiar playbook: inspect assets periodically, intervene based on age or condition thresholds, and replace equipment when it reaches end-of-life.

This approach has worked reasonably well in stable systems. But today’s grid is anything but stable.

Rising demand from electrification, increasing penetration of variable renewables, and growing expectations around reliability are putting pressure on asset fleets that were not designed for this level of complexity. At the same time, capital constraints mean utilities can no longer afford to “over-maintain” or replace assets conservatively.

The result is a fundamental shift: maintenance is no longer a technical activity—it is a capital allocation problem under uncertainty.

Why traditional approaches fall short

Time-based maintenance assumes that assets degrade in predictable ways. In reality:

• Identical assets can age very differently depending on loading, environment, and operational history

• Failure modes are often non-linear and influenced by multiple interacting factors

• Replacing assets too early destroys value, while replacing too late increases system risk

Condition-based maintenance improves this by incorporating inspection data, but still struggles to answer the most important question:

Which assets matter most, and where should we intervene first?

Introducing a risk-based framework

A more robust approach combines two dimensions:

Asset health: how likely is failure?

Health indices aggregate information such as:

• Age and design characteristics

• Inspection results and condition assessments

• Operating history and loading patterns

• Sensor data and diagnostics

The goal is not perfect prediction, but a consistent, comparable measure of failure likelihood across the asset fleet.

Asset criticality: what happens if it fails?

Not all assets are equal.

Criticality captures:

• Customer impact (outages, industrial load disruption)

• System stability (network constraints, redundancy)

• Safety and environmental consequences

• Regulatory and reputational implications

This creates a structured view of consequence, which is often underrepresented in traditional maintenance planning.

From indices to decisions

When health and criticality are combined, utilities can quantify risk in a meaningful way.

This enables a fundamentally different set of decisions:

Prioritizing interventions based on risk, not age

Reallocating maintenance budgets toward high-impact assets

Optimizing inspection frequency rather than applying uniform schedules

Testing trade-offs between reliability targets and cost constraints

In practice, this often leads to counterintuitive insights:

• Some older assets can safely remain in operation longer

• Some relatively new assets may require early intervention due to operating conditions

• A small subset of assets typically drives a disproportionate share of system risk

The real challenge is not the model—it is the system

Many utilities underestimate this: building a health index is not the hard part.

The challenge lies in:

Standardizing data across fragmented systems

Embedding models into operational workflows

Aligning engineering, finance, and regulatory perspectives

Building trust in model outputs

Without this, even the most sophisticated models remain unused.

Conclusion

Health and criticality indices are not just analytical tools—they are a language for decision-making.

Utilities that adopt them successfully move from:

• Maintenance as routine → maintenance as optimization

• Asset-by-asset decisions → portfolio-level thinking

• Engineering judgment → transparent, defensible prioritization

In an environment where reliability expectations are rising and capital is constrained, this shift is no longer optional—it is foundational.