An agent based approach to Irish dairy farm electricity consumption modeling

Dairy farming is an energy-intensive agricultural activity, and understanding electricity consumption patterns is essential for improving farm efficiency and reducing greenhouse gas emissions. This paper proposes an agent-based model (ABM) to estimate the electricity consumption of Irish dairy farms and the associated carbon dioxide equivalent ( C O 2 − e q ) and methane ( C H 4 ) emissions. The model represents major farm equipment as autonomous agents and incorporates key farm characteristics, including herd size, number of milking machines, seasonality, and system configuration. It produces hourly, daily, and annual electricity consumption profiles and corresponding emission estimates. The proposed model is validated using both the Decision Support System for Energy Use in Dairy Production (DSSED) and real farm electricity consumption data from Ireland. In addition, its performance is compared with three machine learning approaches: random forest regression, support vector machine regression, and neural networks. Results show that the agent-based model generates realistic electricity consumption profiles with competitive accuracy while maintaining full transparency and interpretability. Unlike data-driven methods, the model requires limited input data and enables equipment-level analysis, making it suitable for practical decision support in dairy farm energy management. • Simulates electricity consumption on Irish dairy farms using agent-based modeling. • Identifies major electricity consumers and key C O 2 − e q emission sources on farms. • Provides accurate and transparent results with minimal input data requirements. • Validated using the DSSED tool and real-world farm electricity consumption data. • Models hourly electricity consumption as well as C H 4 and C O 2 − e q emission profiles for detailed farm-level analysis.