Development of a discrete infrastructure optimization model for economic assessment on dairy farms (DIOMOND)

•A tool to maximise return on investment in dairy farm infrastructure was proposed.•Best combination of technology, management practices and electricity tariff found.•Level of improvement in return on investment was 26.31% over a base case.•Five optimization algorithms were applied to the problem and compared.•Genetic algorithm found to be most suitable optimization algorithm for the problem. The objective of this paper was to develop a discrete infrastructure optimization model for economic assessment on dairy farms (DIOMOND) to maximize return on investment (ROI) in dairy farm infrastructure over a specified time horizon. Optimizing ROI in dairy farm infrastructure is essential since the choice of technology, electricity tariff and management practices affects the economic performance of the farm. The abolition of European Union milk quotas has started to increase investments in on-farm equipment, necessitating a means of financial decision support for farmers. The optimization of ROI in dairy farm infrastructure within DIOMOND was accomplished by obtaining the optimal selection of equipment, management processes and electricity tariff which affect dairy farm electricity and capital costs and consequently ROI for new infrastructure. Optimization algorithms were applied to previously published dairy farm electricity consumption and ROI models to obtain the combination of decision variables pertaining to dairy farm technology, management practices and electricity tariff which yielded the maximum ROI for a dairy farm over a specified time horizon, compared with a base investment scenario. Furthermore, an assessment was carried out to find the most suitable optimization algorithm for use in maximizing ROI in dairy farm infrastructure. Five optimization algorithms, namely Genetic Algorithm, Particle Swarm Optimization, Simulated Annealing, Tabu Search and Dynamic Programming, were assessed and their optimization performance compared. The Genetic Algorithm was selected due to its superior efficiency and performance. For a test case involving a 195 cow farm the optimal combination of farm technology, management and electricity tariff was found, resulting in a level of improvement in ROI of 26.31% over a ten year time horizon compared to a base investment scenario. The applicability of DIOMOND to a real world scenario was demonstrated successfully. Therefore it is anticipated that DIOMOND will provide a comprehensive tool to deliver substantial monetary savings