Grid operations
Peak Is an Ops Problem
Peak demand is not a pricing slide. It is whether the co-op can move load before the wholesale window closes, with the people and devices it already has.
Most pitch decks treat peak as a market design problem. For a rural co-op or a small municipal utility, peak is an operations problem. Someone has to call the event, someone has to confirm the devices, and someone has to explain to the board why last Tuesday cost more than the forecast. Software that ignores those people is a science project.
The window is shorter than the model
Wholesale peak windows do not wait for a perfect forecast. The useful product is the one that gives a duty operator a short list of actions, the expected MW, and a clear abort path. At Lodegrid we design every peak screen as if the person using it has twelve other alarms and a radio call waiting. If the next step is not obvious in under a minute, we failed.
That is why we still do ride-alongs. A control-room afternoon will kill a feature idea faster than any user interview script. The operators who run roughly 210 MW of load across our customers have taught us more about peak than any wholesale textbook.
MW without people is vanity
Investors love megawatts. Operators love events that finish without a truck roll and without an angry member call. The product has to serve both, but the second one is the gate. Peak software that cannot survive a Tuesday afternoon with a short staffed shift is not infrastructure. It is a demo.
Peak is ops. Price signals matter. Markets matter. But for the co-ops and munis we serve, the product is the operating loop that makes the peak window survivable.