Shaping Energy at the Edge: How Hydro Ottawa leverages AI and Residential DERs for Localized Grid Flexibility
As rapid electrification accelerates, utilities face unprecedented demands on distribution assets—specifically at the feeder and secondary transformer level. Traditionally, these increases in peak load demand have driven asset constraints, forcing utilities to consider premature capital-intensive asset replacements and upgrades. This presentation explores GridFlex, a Hydro Ottawa initiative partially funded by Natural Resources Canada (NRCan) that sits squarely at the intersection of Energy at the Edge and advanced utility operations. GridFlex utilizes AI-driven short-term load forecasting and residential Distributed Energy Resources (DERs) to unlock flexible capacity and mitigate distribution system peaks at an ultra localized level.
To achieve granular, transformer-level visibility, the program employs a multimodal machine learning engine that ingests granular weather data and historical AMI meter data to generate high-resolution forecasts down to individual secondary transformers. However, technical accuracy is only effective if it is paired with a responsive network of participating customers who are willing to shift their load when the grid needs it most.
To bridge this gap, the presentation examines the project design of a frictionless customer journey paired with robust customer education and optimized incentives, turning the consumer into an active grid partner. The presentation details specific strategies used to recruit, onboard, and orchestrate a complex, multi-asset portfolio including Smart Thermostats, Electric Vehicles (EVs), and residential Battery Energy Storage Systems (BESS), alongside Behavioural Demand Response (BDR). It will also summarize how the project is exploring coordination with Ontario’s system operator and its DR program to manage these assets effectively. Managing this diverse mix of behind-the-meter devices requires a unified user experience that masks underlying technical complexity while simultaneously building vital “grid literacy” among participants.
Beyond the operational architecture, this session shares tangible operational impacts and data observed during the 2026 season. The presentation outlines the framework used to evaluate whether transforming residential edge consumers into reliable grid assets can provide a cost-effective Non-Wires Solution (NWS). This shift represents a significant departure from traditional utility-centric planning, moving toward an orchestrated, edge-optimized distribution ecosystem where customer-owned assets actively support grid stability.
Attendees will gain actionable insight into the immediate engagement barriers encountered during the 2026 season, from initial trust building to the ongoing education and financial signaling required to keep participants engaged in a dynamic, multi-asset program. The presentation provides a breakdown of preliminary findings regarding localized peak reduction and explains how this emerging data is being used to demonstrate the reliability of DERs to regulators. GridFlex provides an evolving blueprint for utilities navigating the technical, regulatory, and engagement hurdles of the modern grid edge.
