The Hydropower Signal Processor project will develop a data-driven methodology for classifying and comparing the water-to-energy and energy-to-water transfer functions that succinctly characterize the essential regulating and converting behavior of hydropower facilities. If one considers the time series of flow (and the specific energy conveyed by that flow) in a river as the signal, insight may be gained by examining how this inflow signal, with its myriad and periodic fluctuations, is lagged, filtered, and otherwise converted into an outflow signal by a hydropower facility, with a corresponding electric power output signal. By taking advantage of analytics from the signal processing and information flow domains, this effort will develop an efficient method for encapsulating the complex and facility-specific behavior of many hydropower facilities. The hypothesis of the project is that the transfer functions of facilities, derived from time series data, in the same archetype (run-of-river, ponding storage, and long-term storage for example) will exhibit similarities and features that can be used to classify facilities and model facilities more coherently and consistently in river and power system models, and understand which hydropower project archetypes warrant more detailed study and effort to improve their representation in models. These hydropower facility transfer functions and their facility-specific parameters derived from historical time series data will ultimately be intuitively and quantitatively linkable to hydropower parameters production cost modeling (e.g., modes of operation for hydropower facilities) and water balance modeling, routing, and scheduling. This research and proposed methodology is not intended to create yet another model for how hydropower participates in power systems; it will provide an analysis tool, lexicon, and set of concepts that enable river system and power system decision-makers and modelers to mutually convey the functionality and value of hydropower to electric power systems.
