Utilizing Artificial Intelligence (AI) for the optimal design of geothermal cogeneration systems in zero energy building

• Developed an AI-based multi-objective optimization framework using Artificial Neural Networks to enhance geothermal cogeneration systems' efficiency and reduce costs. • Demonstrated the feasibility of meeting energy demands for a residential building in Copenhagen using a modified Organic Rankine Cycle (ORC) system. • Analyzed the performance of the system under Copenhagen's unique climatic conditions, providing insights into its scalability for other regions. • Achieved optimal exergy efficiency (63.79 %) and cost-effectiveness ($57.82/hour), contributing to sustainable energy practices and reducing environmental impacts. This research investigates the design and optimization of energy consumption for a five-story residential building in Copenhagen, Denmark, comprising ten units of 100 square meters each. The primary objective is to achieve the Zero Energy Building (ZEB) standard. Utilizing BEopt software for simulation and optimization, the study found that the building's annual electricity consumption is approximately 2789,416 kWh.A crucial aspect of energy analysis is the building's orientation. As energy optimization becomes increasingly vital, this research not only focuses on enhancing the building's energy system but also evaluates the optimal directional placement of the structure. The analysis revealed that the south-facing orientation results in the lowest cooling energy consumption at 90,511 kWh, while the east side records the least heating energy consumption at approximately 2133,357 kWh.To fulfill the building's energy requirements, a renewable geothermal energy system was proposed, capable of generating electricity, cooling, and heating. This proposed co-generation system incorporates a modified Organic Rankine Cycle (ORC) equipped with an ejector and preheater for electricity generation, utilizing waste heat for cooling and heating purposes. Modeling was conducted using the widely recognized EES software, while system optimization employed a combination of neural networks and intelligent optimization algorithms. The optimized configuration achieved an exergy efficiency of 63.79 % and a cost rate of $57.82 per hour. Economic analysis indicated that the heat recovery steam generator (HRVG) incurs the highest cost rate at $20.06 per hour. Additionally, a feasibility study incorporating Copenhagen's climate data demonstrated that the system could produce 10,465,920 kWh of electricity, 6340,320 kWh of cooling, and 7160,992 kWh of heating annually. The findings confirm that the geothermal system can adequately meet the energy demands of the five-story residential building throughout in one year.