Speaker
Description
Heat pumps, specifically brine-to-water variants, provide an efficient alternative to traditional methods of heating and cooling production. These systems are versatile, accommodating various demands such as space and ventilation heating, cooling, and domestic hot water (DHW) production. However, the installation of brine-to-water heat pumps in existing buildings without hydronic distribution systems presents a challenge due to high implementation costs. Importantly, CO₂ heat pumps, which employ a natural refrigerant, possess thermodynamic properties that make them highly suitable for DHW production. Due to a global demand for reliable DHW production, the exploration of CO₂ as a working fluid in heat pumps is a necessity.
This study concerns the sustainable and efficient nature of CO₂ heat pump systems for DHW production. A state-of-the-art CO₂ heat pump system is analyzed, focusing on the impact of varying inlet temperatures on heat production and energy usage. The development and validation of an artificial neural network (ANN) model that enables efficient design and control of the CO₂ heat pump is presented. The study employs experimental data from an 8 kW CO₂ heat pump rig. The rig includes a heat pump with two gas coolers and a water-cooled evaporator, and a pump rig designed to generate system temperatures representative of various heat and DHW demands.
A comprehensive dataset was generated through systematic variation of inlet temperatures and production temperature setpoints. This data was used to train the ANN, with hyperparameters defined using Bayesian optimization. The resulting model, based on a multi-layer feedforward network with a back-propagation algorithm, predicts the outlet temperature, heat and cooling production at both gas coolers and evaporator, and electricity consumption based on inlet flows, temperatures, and production setpoints.
These predictions are integral to an optimization scheme that determines optimal operating schedules for heat production equipment, thermal storage, and heating set point temperature. Future research directions within the scope of this study are outlined, including data-driven demand prediction, development of production and storage models, geothermal system modeling, optimization strategies based on demand and electricity prices, and comparisons with alternative heating approaches such as direct electrical heating and non-optimized CO₂ systems. The implementation of the developed model could lead to significant improvements in both energy efficiency and profit generation compared to traditional electric heating and a rule-based baseline strategy for the CO₂ heat pump.
| Conference Topic Areas | Track9: Smart Energy Storage, Integration and Utilization |
|---|
