Where I can test a grid technology claim

Ontario's first smart meters were bought largely to support time-of-use prices, and most distributors, mine included, used their data for billing and little else. I treat the replacement generation as a sensor network for the low-voltage grid. At Quillmere the new meters report fifteen-minute consumption, voltage at the customer's connection and a last-gasp message when supply fails. I use that data to find customers outside the normal voltage range, to see which service transformers are overloaded as heat pumps and vehicle chargers arrive, and to detect outages before anyone telephones. I sequence installation by feeder need rather than by the cheapest route, and I insist that a published statement of how we use interval data precedes the meter on the wall. I also work within the federal metrology rules on seal periods and the provincial arrangements for meter data, which shape timing more than technology does.

Freezing rain is the storm that defines reliability in our territory. Ice loads bring down limbs and conductors over wide areas at once, and crews can only work so many spans a day. My approach is to reduce how many customers each fault takes out and how long it takes to find it. That means automated reclosers and switches with fault location, isolation and service restoration on our longest feeders, covered conductor where ice loading and tree contact are worst, and undergrounding only where it is genuinely the cheapest answer per customer-minute saved. I measure resilience by customer-minutes lost per storm and by the share of customers restored automatically, not by kilometres of line rebuilt. I learnt in March 2025 that automation is only as resilient as its communications, and I now treat backup power for radio and cellular links as part of every switch.

A distributor's control room, substations and meter network were designed for availability, and many were connected to the outside world for convenience long before anyone thought about attackers. My work is to know every path into those systems, to close those that are always open and to make sure the grid can be run if the corporate network is lost. At Quillmere that has meant brokered, recorded and time-limited supplier access approved by two people, a one-way link for operational data leaving the control network, monitoring designed for industrial protocols and regular drills of isolated operation. As a distributor we sit largely outside the bulk-system reliability standards that bind transmitters, and our regulator's cyber security framework asks us to report our maturity each year. I treat that report as a floor. The evidence I give the board is how long we have actually run the grid alone.

Ontario distributors justify their capital spending to the regulator through a distribution system plan, and the most contested lines in it are usually asset replacements. I build asset-health models for poles, service transformers, underground cable and station equipment, but I judge them by one test: does the model change what we replace, and was it right when we checked? Every model we use is back-tested against the failures of the previous five years before it informs the plan, and we publish internally which assets it ranked badly. Meter data has made transformer models much better, because we now see real loading instead of estimates. Pole models remain weaker, because inspection records are uneven and wood decays below ground where nobody can see it. I tell our board and our regulator where each model is reliable and where engineers still override it, and how often they do.

Part of our service area includes, or borders, the lands of First Nations and Métis communities, and our newer projects cross territory where their rights and interests are long-standing. The legal duty to consult rests with the Crown, but much of the practical work falls to the proponent, and I think a technology organisation should prepare for partnership rather than treat it as permitting. My contribution is specific: early sharing of route and design information in a form a community's own advisers can test, agreements on how environmental and asset data collected on their lands will be held and shared, training places for community members in powerline and technology roles, and ownership structures in which a community's equity carries real information rights. I am careful not to speak for any community, and I defer to their leadership on what partnership should mean. Our current jointly owned substation project has taught me to slow down at the start.

I spent three years leading a carbon capture programme at an oil-sands producer, and I can read a capture proposal the way a sceptical engineer would. I ask whether the cost is quoted per tonne captured or per tonne avoided, after the energy used to run the capture plant. I ask how many hours a year the plant will really run, what the solvent will do in that particular flue gas, and who carries the transport and storage risk. I ask which part of the economics depends on a carbon credit price and which on a tax credit. I am not opposed to capture. Concentrated streams such as hydrogen production can be captured at sensible cost, and some heavy industry has few alternatives. I am opposed to capture figures that reach a board without the steam, the hours and the credit assumptions beside them.