Skip to main content
eScholarship
Open Access Publications from the University of California

UC Berkeley

UC Berkeley Electronic Theses and Dissertations bannerUC Berkeley

A multi-method investigation into design and control of radiant cooling and heating systems

  • Author(s): Woolley, Jonathan
  • Advisor(s): Schiavon, Stefano
  • et al.
Abstract

Radiant cooling and heating offers compelling building energy performance opportunities compared to conventional cooling and heating systems. First, radiant systems operate with more moderate supply water temperatures, and – if designed and controlled strategically – can achieve better cooling and heating plant efficiency than conventional systems. In many scenarios, the supply water temperature needed for cooling is warm enough that an evaporative fluid cooler (water-side economizer) can be used in lieu of vapor-compression equipment. Second, whereas the timing and magnitude of cooling or heating plant loads for conventional systems are basically equivalent to the moment-to-moment needs for cooling or heating indoor spaces, high thermal mass radiant systems naturally decouple space heat transfer rates from plant heat transfer rates. As a result – if designed and controlled strategically – high thermal mass radiant systems can: utilize smaller plant equipment; operate more often at part capacity; and/or shift the timing of plant operation to periods with lower electricity prices or more favorable outdoor conditions.

However, these benefits are not unreservedly guaranteed. In fact, industry standard design procedures and the common tools for estimating cooling and heating loads obscure these opportunities and can lead designers to select systems that: operate with less favorable supply water temperatures, require larger equipment than necessary, operate more often at full load, and concentrate plant operation during periods with higher electricity prices and less favorable outdoor conditions.

For this dissertation we used several methods to investigate the design, control, and energy performance of radiant systems. First, we reviewed literature from simulations and case studies of radiant buildings which have built a strong case for the possible energy performance benefits of the technology. Then, by analyzing energy use intensity data from a statistically relevant sample of buildings we found that in practice, large office buildings with radiant cooling in mild climates are using 31% less energy on average than comparable existing building stock. This difference is almost certainly due to a variety of factors, but radiant cooling is very common among those buildings with the lowest energy use intensity.

Then, we conducted structured interviews with design professionals experienced with high thermal mass radiant cooling systems and documented the design and control strategies they typically employ in practice. We discovered that although there are many similarities among the strategies used, there is not consensus on best practices, and there is a lack of standards, tools, and guidelines to support design and control of high thermal mass radiant systems. Our interviews also revealed many opportunities to improve design and control that could reduce cost and improve energy performance of high thermal mass radiant. systems.

Next, we performed experiments to compare the space cooling loads for radiant and all-air systems. We conducted a series of very accurate, realistically scaled, multi-day, side-by-side tests that measured the space cooling rates required for each system to maintain equal indoor comfort conditions. These experiments proved that the peak space cooling load for a radiant system is larger than that of an all-air system (2–21%), and that in some cases the cumulative cooling load can be much larger than that of an all-air system (2–40%). These differences occur because a portion of the heat that would be absorbed by non-active surfaces in a building with all-air cooling, is instead extracted by radiation heat transfer with the internally cooled surfaces. Then, since less heat is stored in non-active thermal masses, less heat can be released passively to the environment when there is an opportunity to do so. Consequently, the differences are largest in scenarios with an opportunity for passive cooling overnight – such as with natural ventilation night pre-cooling. Importantly, these findings prove that the magnitude and timing of cooling loads for a space depend on the type of system and control strategy that is used to provide cooling – factors that are conspicuously absent from standard cooling load calculations and the associated system design procedure.

In view of these findings, we developed a thorough critique of the standard definition of “space cooling or heating loads” and the associated system design procedure, then we developed a new definition and an improved system design procedure. We argue that in addition to omitting important heat transfer fundamentals, the standard design procedure fails to account for the impact of system controls, does not facilitate many important design objectives, and imposes simple constraints that overlook fundamentals about thermal comfort. To resolve these issues, our improved approach shifts the focused objective of the system design procedure away from satisfying a singular – “ideal” – space cooling or heating load, and instead orients the designer toward selecting and sizing components and their controls that best satisfy performance objectives such as thermal comfort, indoor air quality, or life cycle cost.

Finally, to demonstrate the practical consequences, we simulated a high thermal mass radiant system designed with the standard procedure and compared it to several example design alternatives developed with the revised procedure. This comparison showed that the improved design procedure: reduced the size of cooling plant equipment by as much as 50%, increased median cooling supply water temperature by as much as 5.2 °C (9.4 °F), reduced chilled water consumption during periods with high electricity prices by as much as 100%, and reduced annual occupied discomfort hours by as much as 55%.

Main Content
Current View