Seattle City Light serves over half a million customers across a 2,100-mile distribution network in the Pacific Northwest. Major storms in 2006, 2019, and 2021 caused widespread outages, and the pattern was clear: extreme weather events were intensifying. But the utility's investment planning still relied on historical outage data—a rearview mirror approach that couldn't account for how climate change would reshape future grid vulnerabilities.

‍

Following recommendations in their Climate Change Vulnerability Assessment and Adaptation Plan, Seattle City Light sought to explore whether advanced analytics could help them understand the relationship between future weather patterns and grid performance. They partnered with Rhizome to pilot an approach that would correlate distribution grid outage and weather data to look forward in time under climate projections.

Rhizome applied the Aspen Platform to analyze 25 years of outage data (1997–2022) across 43 feeders in Seattle City Light's Duwamish region, an area known to experience persistent outages during storm events.

The analysis proceeded in two stages. First, Rhizome correlated historical weather conditions with grid performance. By examining how extreme weather events between 1997 and 2022—specifically wind, precipitation, and temperature—affected feeder-specific outages measured in customer-minutes interrupted, Rhizome established strong relationships between weather conditions and grid impacts.

In the second stage, the analysis applied these historical correlations to future climate projections. Rhizome modeled future projections of extreme weather using four dynamically downscaled climate models calibrated for the Pacific Northwest. These projections extended through 2100 under different emissions scenarios, allowing the team to estimate how outage patterns might evolve as extreme weather events change in frequency and intensity.

The pilot tested whether this two-stage approach was feasible with available utility data, and worked to identify what would be needed to make these insights actionable for investment planning.

The pilot validated that strong correlations exist between extreme weather and grid performance. The analysis found that extreme wind events cause 225x more outages than normal operating conditions, while combined wind and rain events lead to 20x more customer-minutes interrupted. Under a higher emissions scenario (SSP3-7.0), 2-year wind events are projected to increase 40% by 2070, with wind-related outage frequency potentially doubling in the Duwamish area.

The analysis also revealed specific failure patterns tied to weather conditions and asset types. On days with extreme wind, the system experienced over 200 additional tree-related outages and 2 additional elbow-related outages compared to normal conditions. Days with extreme heat saw 4.5 additional transformer-related outages, potentially linked to increased load on the system during hot weather.

The pilot also revealed the significant data and modeling challenges that must be addressed before utilities can confidently incorporate climate projections into investment decisions. The analysis highlighted gaps in asset-level data granularity, the need for better information on historical mitigation measures and their effectiveness, and the complexity of modeling system-level resilience impacts—particularly around outage duration, which is critical for resilience planning.

Seattle City Light and Rhizome demonstrated that linking extreme weather to grid performance is possible, and that climate projections can reveal how future conditions may differ from historical patterns. This pilot represents an important first step in integrating forward-looking information to plan ahead.

The work also clarified what's needed to make this approach decision-ready: more granular asset data, better documentation of infrastructure improvements and their impacts, and enhanced modeling of cascading system effects.

These insights are helping shape both Seattle City Light's data strategy and Rhizome's continued platform development, building toward a future where utilities can confidently plan for the climate they'll face, not the one they've known.