The role of innovative econometric models in short-term hydropower optimization

The recent transformation of the electricity market has modified the hydropower production paradigm, especially for storage reservoir systems. In particular, the process of market liberation has led to a shift in hydropower management approaches. These have moved from strategies oriented to maximizing energy production to strategies aimed at revenue maximization. Indeed, hydropower producers bid their energy production scheduling in advance, attempting to align the operational plan for the ensuing day (i.e., allocating 1-day ahead the hourly time series of turbined water discharges) with hours where the expected electricity prices are higher. As a result, the accuracy of 1-day ahead electricity prices forecasts, as given by econometric models, has started to play a key role in the short-term optimization of storage reservoir systems. Though recognized, this aspect has so far received limited attention in the literature.

This work aims to contribute to the topic by presenting a comparative assessment of revenues provided by the solution of short-term hydropower optimization problems driven by two econometric models during an entire year of simulation. Both models are autoregressive time-adapting hourly forecasting models which exploit the information provided by past values of electricity prices. One model, referred as Autoarimax, can be considered as the state-of-the-art in electricity prices forecasting, the peculiarities of which are rooted in the use of time-varying exogenous variables related to electricity demand and production, while the other, referred to as the Benchmark, can be considered a standard autoregressive model.

The added value of using an innovative econometric model is exemplified in two selected hydropower systems with different storage capacities located in the south- eastern Alpine region. The enhanced accuracy of electricity prices forecasting is not constant across the year due to the large uncertainties characterizing the electricity market, the fluctuations of which are controlled by short-term and seasonal imbalances in factors affecting electricity demand and production. Our results also show that the adoption of this more accurate econometric model leads to larger revenues with respect to the use of a standard model. The increased revenues depend strongly on the hydropower system characteristics, such as reservoir capacity and the ratio between inflows and maximum turbined water discharge that can be conveyed to the plant. Specifically, we showed that, for the reservoir characterized by a larger storage capacity, the use of Autoarimax forecasts led to a revenue increase of up to 2.31% at monthly scale with respect to the case in which Benchmark forecasts are used in the optimizations. This revenue gain can reach up to a 31.06% increase if we consider the maximum daily deviations.