Development and evaluation of an automatic steam radiator control system for retrofitting legacy heating systems in existing buildings

Abstract

Inefficient building systems and a lack of building controls in existing buildings with legacy systems lead to increased energy consumption and reduced thermal comfort. Recent developments in low-cost sensors and microcontrollers have enabled the development and deployment of smart building sensing and controls solutions, but there remains a need for adapting these technologies to retrofit existing buildings with legacy systems. Here we introduce an automatic radiator control system, designed to address an application for which no commercial solutions currently exist, that can be used to retrofit low pressure steam heating systems in existing buildings. It allows for zone-level heat output control, which is adjusted continuously in real-time using data from custom wireless sensors. We describe the design and development of the physical, electrical, mechanical, and software components of the radiator control and present an evaluation of its performance while deployed in an historic building on the campus of Illinois Institute of Technology in Chicago, IL USA, over two winter seasons of 2019 and 2020. The automatic radiator control was deployed in a total of 17 rooms with varying uses and occupancy patterns. Five different control strategies were evaluated over a total of 15 weeks (9 weeks in 2019 and 6 weeks in 2020): (i) manual control (baseline), (ii) enforced schedule, (iii) occupancy-based control, (iv) proportional-integral-derivative (PID) control, and (v) PID and occupancy-based control. Each control strategy was applied one week at a time during a total of three different weeks, randomized throughout the experiment. Results show that using any of the automatic control strategies reduced radiator runtime compared to manual control, which reduces energy output. Using the PID + occupancy control strategy, which automatically adjusts the radiator heat output based on the setpoint alongside an occupancy sensor in each zone, yielded the highest potential energy savings during the 3-week-long measurement periods: approximately 74% compared to manual control. Using schedule- or occupancy-based control strategies led to some decreases in perceived comfort compared to manual control during the time of observation, but we also show that thermal comfort can be maintained and even improved by using a PID control strategy with the custom controller. In existing buildings where replacing legacy space heating systems with modern ones is financially and practically unfeasible, retrofitting them with custom automatic controls as demonstrated herein has the potential to considerably reduce energy consumption while maintaining or even improving thermal comfort, which can extend the life of the building.