I don’t recall heavy quant involvement. Certainly there were model-driven transactions, but the key models used in power (as I recall) are scheduling-oriented. Power management is perhaps the operations research professional’s highest calling. I used to think it was airline scheduling, but then I learned about the dynamics associated with LMP and I was blown away by their complexity.
Again, though, these modes are for scheduling more than pricing. I believe the Enron issues were more bald-faced fraud than model manipulation.