The use of heat pumps for heating and cooling in buildings is increasing significantly. This is due to their crucial role in facilitating the integration of renewable energy sources in buildings. Previous research conducted on the optimization of a cascade of chillers considering the variation of the coefficient of performance (COP) in relation to the load distribution demonstrated energy savings. This work presents an analogous optimization strategy by drawing upon the similarity between the COP curves of chillers and heat pumps. A heuristic mathematical optimization is used to calculate the optimal threshold values for the heat pump cascade. A heating installation of a cascade of three heat pumps is simulated at virtual level, and the performance of the optimized cascade is compared to the baseline configuration. The results show a reduction in electric consumption of up to 12.51%.

The accurate forecasting of weather conditions and electricity demand is of great importance in smart building and distributed energy resource operations. It ensures efficient resource allocation, operational cost reduction, and mitigation of environmental impacts. Traditional forecast methods often fall short due to their reliance on simplistic statistical techniques and limited adaptability to complex, dynamic patterns. Machine learning has emerged as a powerful (but complex) approach for improving the accuracy of such predictions by leveraging advanced algorithms. In this context, we introduce a framework which hosts a publicly available library of traditional and state-of-the-art machine learning models tailored for weather and electricity demand forecasting in buildings. Models are dynamically selected and combined based on an internal optimization algorithm to autonomously operate without user interaction and to ensure optimal forecast performance over the lifetime of the installation.

This paper addresses the problem of cooling tower control in water-cooled chiller systems. These types of systems use significant amounts of energy and improvements in the control improve their energy efficiency. A detailed Modelica simulation of a chilled water system serving a data center is used as a testbed to demonstrate the potential of a new type of real-time optimization (RTO) for maximizing energy efficiency. The paper also highlights the utility of coupling a pre-built system model with Python-based code as a way to streamline testing and development of new controls. The RTO algorithm is model-free and does not require traditional temperature and humidity sensors and is applied to controlling the cooling tower fans. Results show that the algorithm can save 15% energy compared to a baseline during summer periods in two different climates.

Vintage building mechanical systems are prevalent throughout the country, with many steam radiator systems found in northern climates and urban centers. While these systems are no longer state of the art, they can be run efficiently by experienced operating engineers and modern building controls. One area of control and maintenance that is often overlooked is diagnostic testing for steam traps. However, such work is complicated, requires trained professionals with costly precision equipment, and can only be performed as frequently as budgets allow. This paper describes the development and testing of a low cost, continuous steam trap diagnostic monitoring device to address these limitations.

Photovoltaic (PV) and battery storage prescriptive requirements for nonresidential new construction in California’s energy code are specified based on the type of building and do not account for diversity in space floor areas that may be found in buildings of the same type. To improve the way energy codes specify PV and battery sizes, a new method was developed that provides a space function-based approach to PV and battery sizing. This new method uses three modules to (a) optimize PV and battery systems to meet export targets, (b) create a relationship map linking modeled systems to spaces, and (c) proportionally distribute building level PV and battery to modeled spaces. Collectively, these modules create a framework that can be used to specify energy code PV and battery storage sizing requirements based on space function and the relative floor area of various spaces in a specific building or project.