Energy efficiency improvements in a microbrewery in South Africa

Abstract

The brewing industry constitutes a significant portion of energy demand in the food and beverage sector, with microbreweries consuming approximately 0.65 kWh of electrical energy to produce 1 L of beer, which is almost 50% more than large-scale breweries. This equates to around 5.2% of total production costs, which highlights the critical need for energy efficiency improvements. Despite this substantial consumption, the implementation of optimised energy management strategies remains limited, particularly for thermal processes in craft beer production, which account for 8% of operational costs. More studies are needed to explore effective energy management schemes for breweries, with a particular focus on heating and cooling loads during critical production phases. This study shows that demand-side management, coupled with time-of-use (ToU) tariff optimisation, presents opportunities for significant cost savings in brewery operations. The research aimed to address key challenges in energy-intensive brewing processes, particularly in South Africa, where craft beer production has grown substantially in recent years. This work emphasises predictive modelling of brewery energy performance using machine learning, optimal control strategies, and comprehensive economic analysis to minimise energy costs while maintaining production quality. The study proposes an integrated approach that combines process optimisation, a hybrid thermal–electrical energy system incorporating solar thermal collectors, a heat pump, and TES, and an optimisation-based control methodology designed to leverage ToU tariff structures and reduce peak-period grid dependence. The investigation employed the Performance, Operation, Equipment, and Technology (POET) framework to systematically analyse energy usage patterns in brewery operations. A real-time monitoring system using a Power Quality Analyzer (PEL103) was installed in a Bloemfontein microbrewery to collect detailed load profile data. Subsequently, an artificial neural network (ANN) model was developed using shallow neural net fitting (nftool) to predict energy requirements across various production stages, including mashing, boiling, fermentation, and cold storage. The ANN model processed input variables such as process temperatures, flow rates, and equipment states, which achieved regression values between 0.90 and 0.98, with mean squared errors below 15%. The study also implemented optimisation strategies using model predictive control methods to evaluate hybrid energy system configurations that incorporated renewable sources and thermal energy storage (TES). Simulations demonstrated that shifting energy-intensive processes to off-peak periods could yield 33% cost savings, while the integration of photovoltaic systems with battery storage reduced grid dependence by 10% to 15%. The POET framework analysis projected potential energy savings ranging from 10% to 70%, depending on the level of implementation: conceptual improvements (50% to 60%), active controls (25% to 35%), technical upgrades (60% to 70%), and further engineering optimisations (10% to 20%). The results indicate that TES systems can achieve up to 69% energy savings compared to conventional grid-powered operations, with a breakeven point within 1.5 years for a system that costs approximately USD 50 000. The ANN model proved particularly effective in optimising batch scheduling to align with ToU tariff periods while maintaining product quality standards. Furthermore, the study found that comprehensive implementation of energy efficiency measures could reduce long-term energy costs by 35% over a 20-year period, assuming 5% annual inflation and 10% electricity price increases. The study demonstrates that the synergistic integration of ANN-based predictive modelling, an optimisation-based control strategy, and a hybrid thermal-electrical energy system significantly enhances energy efficiency in microbrewery operations. These integrated methods also improve operational costs by optimising thermal supply, reducing dependence on grid electricity, and increasing the effective use of renewable energy. The developed methodologies provide a replicable framework for energy management in food and beverage production, with particular relevance for small to medium enterprises in emerging markets. The findings underscore the importance of integrated energy solutions that combine technological improvements with operational optimisation to achieve both economic and environmental sustainability goals in industrial processes.