The development of object oriented Bayesian networks to evaluate the social, economic and environmental impacts of solar PV Philip A. Leicester 2134/21657 https://repository.lboro.ac.uk/articles/thesis/The_development_of_object_oriented_Bayesian_networks_to_evaluate_the_social_economic_and_environmental_impacts_of_solar_PV/9539540 Domestic and community low carbon technologies are widely heralded as valuable means for delivering sustainability outcomes in the form of social, economic and environmental (SEE) policy objectives. To accelerate their diffusion they have benefited from a significant number and variety of subsidies worldwide. Considerable aleatory and epistemic uncertainties exist, however, both with regard to their net energy contribution and their SEE impacts. Furthermore the socio-economic contexts themselves exhibit enormous variability, and commensurate uncertainties in their parameterisation. This represents a significant risk for policy makers and technology adopters. This work describes an approach to these problems using Bayesian Network models. These are utilised to integrate extant knowledge from a variety of disciplines to quantify SEE impacts and endogenise uncertainties. A large-scale Object Oriented Bayesian network has been developed to model the specific case of solar photovoltaics (PV) installed on UK domestic roofs. Three specific model components have been developed. The PV component characterises the yield of UK systems, the building energy component characterises the energy consumption of the dwellings and their occupants and a third component characterises the building stock in four English urban communities. Three representative SEE indicators, fuel affordability, carbon emission reduction and discounted cash flow are integrated and used to test the model s ability to yield meaningful outputs in response to varying inputs. The variability in the percentage of the three indicators is highly responsive to the dwellings built form, age and orientation, but is not just due to building and solar physics but also to socio-economic factors. The model can accept observations or evidence in order to create scenarios which facilitate deliberative decision making. The BN methodology contributes to the synthesis of new knowledge from extant knowledge located between disciplines . As well as insights into the impacts of high PV penetration, an epistemic contribution has been made to transdisciplinary building energy modelling which can be replicated with a variety of low carbon interventions. 2016-06-15 11:46:29 Solar PV Bayesian networks Impacts Mechanical Engineering not elsewhere classified