A multi-criteria logical based approach for optimal planning and assessment of rule-based hybrid load following micro combined heat and power systems

This study presents a logical based approach with constraint programming technique for applying rule-based hybrid strategies in the optimal planning of the micro cogeneration systems. Moreover, in this paper, a novel operation strategy that relied on the comparison of loss of supply probabilities for reducing the dependency on auxiliary systems and minimizing the production of redundant energy in micro cogeneration systems is also introduced. The presented approach is applied to evaluate the effectiveness of five operation strategies namely two basic, two common hybrid and the proposed hybrid strategy under three single optimization criteria for the total annual cost (TAC), primary energy consumption (PEC) and greenhouse gas generation (GHG) as well as one integrated criterion based on Utopia tracking method for the combination of these criteria. Three energy matching parameters are exploited to assess the performance of each strategy in matching the cogenerated energy carriers with the building demand requirements. The analysis of the obtained results for four types of residential buildings indicates that the common hybrid strategy which maximizes the exploitation of cogeneration system, and the proposed strategy as the most suitable strategies for integrated criterion have been able to achieve the best compromise solutions with over 55%, 14% and 9% reductions in TAC, PEC and GHG respectively. While the proposed strategy can achieve higher reductions for TAC (up to 4.6%) as well as higher overall performance for energy matching (up to 2.5%) with the same limitations bounds for GHG and PEC in comparison to this common hybrid strategy. By analyzing the model uncertainties, the feed-in tariff for power generation with near 60% decrement in TAC is found as the most dominant economic parameter for sensitivity analysis while the capacity of the micro cogeneration unit with over 3% increment is detected as the most affected outcome for stochastic analysis.