2027 Event Agenda

For a large transmission system, the processing power required to reconstruct the thermal history of tens of thousands of conductors isn't a software problem. It's an engineering problem at a scale utilities haven't attempted. This is the story of how Dominion Energy and Data Society built toward that scale - and what it actually took to get there.

The journey started where most utility innovation does: a spreadsheet. A Dominion engineer built an Excel prototype to model conductor temperatures using IEEE 738 - proof the concept was sound and proof that it couldn't handle the amount of data a real transmission network produces. 

Phase 1 moved the work into a Python-based pipeline, but even programmatic computation couldn't handle 55,000+ conductors at 15-minute intervals across five years. The math is unforgiving: each conductor run isn't a single calculation but an optimization algorithm that takes tens or hundreds of iterations to converge. Phase 2 distributed and parallelized the pipeline across seven on-premises servers - and even then, the full computation took two months to complete.

And that was just the mechanics. The harder problems were in physics. Deriving a defensible definition of thermal cycling, reconciling competing assumptions about conductor behavior, and tracking down the source of unexplained temperature spikes - one of which turned out to be buried in an older version of the IEEE 738 standard itself and resolved only by implementing the updated specification - required as much engineering judgment as computational horsepower.

Phase 3, now actively deploying on OpenShift on-premises at Dominion - representing one of the first uses of that infrastructure for Dominion's electric transmission organization - addresses the remaining challenges: scale, computational efficiency, robustness, modularization, and the quality of computed temperatures.

The platform that emerged reconstructs conductor thermal history at 15-minute resolution, quantifies cumulative stress and thermal cycling, and gives asset managers a view of their fleet they've never had before. A key outcome: an ability for the engineers to see the entire forest as well as individual trees to avoid unnecessary equipment replacement while maintaining reliability objectives.

This session will walk through all three phases - what broke, what we learned, and what it took to build something that could actually run at transmission-system scale. Attendees will leave with a grounded understanding of what physics-based conductor modeling demands in practice at scale, not in theory.

Learning Objectives:

• Understand the computational realities of IEEE 738 conductor temperature modeling at transmission-system scale - and why standard approaches fail when you're modeling 55,000+ conductors at 15-minute intervals across multiple years of history.
• Learn how Dominion Energy and Data Society evolved from prototype to distributed system to container-based distributed enterprise solution (OpenShift) deployment across three phases of development.
• Explore how thermal cycling, cumulative stress analytics, and anomaly investigation change the way asset managers make replacement and maintenance decisions.
• Take away honest lessons from a multi-year, multi-phase engineering effort - including what broke, what surprised us, and what we'd do differently.