Simulating systems with fast and slow dynamics: lessons from the electric power industry

Complex systems exhibit dynamics across multiple timescales. Strong couplings across timescales can be important, but common practice has been to ignore the couplings. This paper describes my experience modeling electric power systems when clients argued for models including both fast and slow dynamics. Their goals were realism, transparency for policymakers and fast simulations to enable policy design and sensitivity analysis in live workshop settings. I describe different methods developed to meet these needs using five modeling studies conducted over three decades. The studies focused on important policy issues, which are described for context. I summarize the strengths and weaknesses of each method, with emphasis on their applicability beyond the power industry. The paper concludes with an endorsement of a coupled system with a stand‐alone model of fast dynamics to support the design of the long‐term model. The appendices provide details to help those interested in applying the methods.