This paper describes a tool for the interactive design of urban energy systems. The tool incorporates an optimisation model that can be used to select the resources, conversion processes and transport networks required to satisfy a spatially distributed and time varying pattern of demands across an urban area. The optimisation model can be formulated with specific goals and constraints such as minimising annualised costs while satisfying emission constraints. The optimisation model uses a temporal discretisation with two levels of time slices: major time periods for investment decisions, and minor time periods to represent the variation of demands within the major periods. The tool uses a Resource Technology Network to represent the potential technological choices for satisfying the demands. The tool provides a graphical interface for selecting the available conversion processes, resources and network links, and for specifying their operational parameters. This information is retrieved from a database implemented as a Protégé ontology and can be customised to reflect local conditions. The spatial grid and associated demands are specified as CSV files. Optimisation scenarios are constructed and submitted to a remote server. The results of the optimisation are retrieved and displayed within the tool. Resource flows and locations of conversion processes are visualised on a 2d representation of the spatial grid. Key performance indicators and a summary of results are displayed within tables and charts.
The tool described above is intended for use in interactive model development with an integrated database and results visualisation. The database and visualisation framework are implemented in Java, and the GAMS modelling language is used to formulate and solve the optimisation model. We are also developing an alternative implementation of the optimisation model written in the Pyomo modelling language. This is intended for use in integrated workflows where it could be connected with agent based models for estimating energy demands, and with other tools such as semantic models of urban areas. The Pyomo implementation could also be more convenient for use with tools for sensitivity analysis where the model has to be solved repeatedly.