Posters

Abstract

Methods. Eni has developed a general-purpose optimization tool to support the optimal sizing and design of any so-called Multi Energy System, that is an energy generation facility that incorporates several closely integrated technologies. Among the possible applications is the design of hybrid plants combining photovoltaic power (PV), wind power and battery storage (BESS). When non-dispatchable and/or storage technologies are introduced, a complex interplay of different time scales arises when evaluating the economics of the project throughout its lifetime. For example, grid prices and other key economic assumptions might vary over a time scale of years. Simultaneously, accurate assessment of the optimal installed capacity requires to consider the hourly dynamics of generation and storage. Optimal design of MES is thus a multi-scale problem in terms of time resolution, and its optimization requires a specific class of mathematical models that is not readily available in commercial software. Another key challenge is the capability to provide a fair assessment of the economics in the face of uncertainty. For example, by means of extensive sensitivity analyses on the most impactful parameters. Most importantly, the projected hourly availability of non-dispatchable resources. Innovative contributions. The optimization tool available at Eni is able to overcome the challenges listed in the previous section, by implementing a multi-year mixed-integer linear programming formulation that can realistically take into account: i) complex plant topologies with multiple destination networks; ii) evolution of key economic parameters including hourly energy prices and resource availability over the plant lifetime; iii) hourly time resolution using typical days to keep the computational cost manageable; iv) realistic treatment of CAPEX and OPEX with an objective function closely aligned with standard lifetime net present value calculations. In addition, the capability to perform extensive sensitivity analyses up to 1000s of cases is enabled by the extensive HPC resources available in the company. The ability to couple short and long time scales in a holistic formulation is instrumental to provide a realistic assessment of challenging projects. This supersedes most conventional approaches, which are either based on monthly aggregates over the lifetime, or hourly calculations on a single reference year.  Results and conclusions. The optimization approach was tested on a real business case by Plenitude to properly design a multi GW hybrid wind/PV/BESS project. The energy produced shall be delivered across two different electrical grids in two different countries through an HVDC cable. More than 500 sensitivities were analysed changing key parameters as hourly selling prices, wind production distribution and costs, to obtain the most cost-efficient technology mix. This approach allows to design hybrid plants to maximise economic return instead of the classic method of maximising the installed capacity. One of the most interesting outcomes was the need to increase the wind share capacity since wind production occurred during the hours where energy has the higher cost, hence the maximum economic value.