A validated forecasting model for optimizing dispatch schedules of PV - battery system in grid-connected apartment buildings

The escalating emissions of greenhouse gases have emerged as the primary driver of global warming. Cities have been found responsible for 70 % of global CO2 emissions. This brings high potential for lowering greenhouse gas emission by introducing innovative technologies in urban environments. One promising technology is Photovoltaics (PV) coupled with energy storage systems (ESS). In this paper, the deployment, integration, and operation of ESS coupled PV is being investigated. This involves installing a real-life PV-ESS system in the KTH Live-In-Lab. This paper focuses on the optimization of battery operations and prepares the framework for an automated operation of the real-life battery installed in the KTH Live-in-Lab. The focus centers on achieving two separate objective functions: minimizing costs and maximizing self-consumption. These goals are pursued through the application of linear optimization, day-ahead forecasting, and real-time control. The results compare the performance of change in solar power self-consumption and cost for each year. Where the ‘Base case’ scenario served as a benchmark with a yearly cost of 210,078 SEK in 2022 and self-consumption of 71.86 %. The cost optimization considers two different scenarios where monthly peak demand billing is considered or disregarded. Without considering the monthly peak power billing, costs drop by 32.2 % and self-consumption increases significantly to 96 %. Considering power charges reduce costs by 14 % and self-consumption was slightly improved to 98 % compared to the base-case. The optimization for maximizing self-consumption shows an improvement in costs of 3 % and achieves 100 % self-consumption. The forecasting-optimization framework introduced in this study is a valuable decision support tool, aiding stakeholders in making informed choices about balancing costs and self-consumption for PV systems integrated with battery energy storage systems in grid-connected apartment buildings. The tool is adaptable and can be trained for different sites and is capable of handling different cost structures to evaluate the full energy trading landscape. The findings show that the battery dispatch strategies must be dynamic regardless of the season. This is due to the interplay between energy availability, market prices, and grid interactions to optimize performance and cost. They also highlight that for larger grid connections (above 80 A) a more flexible power billing structure would improve the financial viability of PV-BESS systems for the end user. • A robust, reusable day-ahead optimized dispatch schedule is developed and validated using the real-life setup of KTH Live-in Lab. • A comprehensive techno-economic analysis of the system's potential is conducted over the lifespan of the coupled PV system. • The potential and limitations of the Battery Energy Storage System (BESS) are analyzed over the PV installation's lifecycle. • Discussion on the need for further study on the interaction between peak power billing and flexible RTP energy pricing.