Electric utilities in Europe are faced with squezeed margins in the marketing of its production units due to increased pressure on the electricity price levels assigned to renewable energy effects. As electricity markets gain dynamics and move closer to real-time redispatch and planning of the production facilites, partly also due to technical requirments of the transmission grid which maintains the energy balance in the system at all times, efficient and performant asset optimization and management software solutions become increasingly important. Well studied algorithms like the linear or mixed integer programming reach their limits in the ability to find the optimal dispatch in short time while taking volatility of the input parameters in consideration, such as water inflow or intraday prices. We are trying to tackle the challenges of the fast changing electricity market by developing models in different systems which satisfy our requirements for optimality and performance. We were happy to find out that an open-source programming language Julia with good performance reputation offers a great support for optimization solutions and algorithms, even more the trending ones like dynamic programming or stochastic dynamic programming. We were happy to find out that we dont have to ‚reinvent the wheel‘ and implement the algorithm ourselves but rather focus on the analysis part and analyse the results. We have built an optimized dispatch application of our pumped-storage hydro power plant in Julia and would like to share our experiences in working with SDDP Packages in Julia, while benchmarking it with the existing applications that we use in our everyday work.
