Modeling multi-criteria decision analysis in residential PV adoption

Multi-Criteria Decision Analysis (MCDA) is a sub-discipline of operations research that aims to solve multi-objective optimization problems by evaluating competing factors in decision-making. MCDA supports multidimensional decision-making processes through the analysis of diverse inputs at several levels of description, e.g. economic, technical, social, and environmental. The use of MCDA has been gaining momentum in the energy field, especially in endeavors that require evaluating the feasibility of different adoption scenarios for renewable energy technologies. The study presented in this paper investigates the combined use of multi-agent simulation, Bayesian modeling and sensitivity analysis for the development of a novel MCDA approach that supports the analysis of residential solar Photovoltaic (PV) adoption. The ensuing MCDA approach is evaluated alongside four popular MCDA methods (AHP, TOPSIS, SAW and ELECTREII) in a variety of tests aimed to assess overlap in criterion rankings and decision-making outcomes, covariation of criteria rankings in alternative scenarios, and the capacity to provide a model of correct decision-making. The comparative evaluation with AHP, TOPSIS, SAW and ELECTREII shows that overall ABM-BN-SA is well correlated with most of other MCDA methods and provides the best performing model of decision-making, with reference to the PV adoption use case under analysis. TOPSIS shows the closest fit with ABM-BN-SA, as expected since it used a ranking approach considerably closer to ABM-BN-SA as compared to the other MCDA treatments. ELECTREII yields the lowest degree of ranking overlap with ABM-BN-SA. All the reviewed methods have been illustrated and evaluated within our residential solar Photovoltaic (PV) adoption decision support system. In general, these methods enable a user to select an optimal solution out of a set of plausible alternatives according to multiple criteria and assist him in the design and exploration of the decision space. The ensuing decision-making methodology can be applied not only by Solar PV panel purchasers but also by stakeholders in other industries to logically and straightforwardly model and simulate the adoption decision-process of the public based on their individual preferences, behavioral rules, and interaction within a social network, with specific reference to a consumer utility function.