Condensation-free radiant cooling using infrared-transparent enclosures of chilled panels

Radiant cooling power in the humid climates is inherently limited by condensation. This research investigates a type of radiant cooling methodology whereby the cold temperature source is convectively and conductively isolated from the environment with a membrane transparent to visible radiation to allow supply temperatures to be decreased for radiant cooling systems in humid climates. We conduct an FTIR analysis on three candidate membrane materials and fabricate a prototype experimental test panel that allows for thermal performance evaluation at different panel orientation and depths. Our study shows that for a 5 °C chilled panel temperature, the exterior membrane surface temperature reaches 26 °C in a 32 °C / 70% RH environment resulting in an effective panel temperature of 15.8 °C. Such a panel construction would avoid condensation in many humid environments and allow for radiant cooling without any latent load handling. INTRODUCTION Radiant cooling environmental systems are a class of measures and technologies for space cooling in the built environment. They involve exposing building occupants to mechanicallycooled indoor enclosures, or parts of entire enclosures, allowing for a greater degree of heat to be rejected radiatively by the human body to the ambient environment than would otherwise occur. While thermal comfort models demonstrate the potential for radiant cooling systems to provide comfortable conditions in spaces with high indoor air temperatures (de Dear et al. 1996; Arens et al. 2006), in practice generating large air-to-panel temperature differences is hard to achieve without risking condensation occurring on chilled surfaces (Teitelbaum and Meggers, 2017; Feng, 2014). It is for this reason that radiant cooling systems are nearly always combined with mechanical ventilation systems that supply dehumidified air to interior spaces, ensuring indoor air dew point temperatures are sufficiently low to prevent condensation arising on cooled surfaces. An alternative solution to mitigating the risk of condensation can be found through a more focused investigation of the specific radiant heat transfer and convection processes occurring within and around radiant panel assemblies. In 1963, Morse (Morse, 1963) described a new type of radiant cooling panel for the tropical environments of Australia, whereby a membrane transparent to long wave infrared radiation is used to enclose, or isolate, the cold panel from the warm, humid ambient air as shown in figure 1a. Since the radiant panel and humans emit in the longwave regime, typically defined as wavelengths between 2.5 and 50 microns, their radiation is able to exchange proportionally to the transmissivity of the membrane. If the enclosure volume would be sufficiently large, and filled with dry air, internal convection would not be significant enough to lower the surface temperature of the membrane below the ambient dewpoint temperature of the interior space, thereby preventing condensation. Today, whilst there are some emerging commercial applications of Morse’s original idea (interpanel, 2018), there remains a lack of understanding of the spectral quality of potential membrane materials and how different material and geometric configurations of such panel assemblies affect overall radiant cooling flux and condensation risks. This paper presents an empirical study which expands on Morse’s original chilled panel design by carrying out: 1) Fourier Transfer Infrared (FTIR) Spectroscopy analysis of infrared transparent materials to select the most suitable, common building material for a future panel membrane; and 2) an experimental study of the radiant flux achieved with a prototype radiant cooling panel against varying geometric parameters such as the distance between the membrane and chilled panel, and the vertical/horizontal orientation of the panel itself. The objective of the overall study is to identify a potentially optimal radiant cooling panel design which would provide the greatest cooling flux in a very hot and humid environment without condensation occurring. MATERIALS AND METHODS Figure 1: (a) Original infrared transparent radiant cooling panel (Morse, 1963). (b) Test panel prior to black paint. (c) Finished panel with the PP membrane. (d) Visible image of the author holding a sample PP membrane sheet (background) in comparison to the equivalent infrared image of the scene (foreground) FTIR Analysis Many common household materials are transparent to longwave infrared radiation, such as high density polyethylene (HDPE) trash bags and low density polyethylene (LDPE) or polypropylene (PP) bottles. However, the comparatively large wavelengths for infrared radiation and correspondingly low frequencies contribute to faster extinction and absorption of the radiation in materials, so it becomes difficult to select these materials for a potential infrared transparent membrane without a detailed representation of their individual spectral properties. Three specific types of prototype LDPE, PP, and HDPE panels were procured for this research, respectively United States Plastic, 1/32” LDPE #42568; United States Plastic, 1⁄8” HDPE #42587 and a 50 micron-thick polypropylene panel produced proprietarily for interpanel GmbH, FTIR spectroscopy was conducted using a Nicolet i10 infrared spectrophotometer to measure the wavelength-based transmission spectra for each material between 2.5 and 15 microns. The FTIR transmission spectra was overlaid with a true black body emission curve to visualize the ability of each membrane to transmit radiation between the panel and a human. The resulting curve is a true spectral radiance diagram providing radiant power per steradian per micron. Integrating the curve numerically between the measured wavelengths provides a panel radiance value, in units W/m/sr. Assuming a Lambertian emission function over an arbitrary hemisphere about any point on the panel provides the integration constant for converting radiance to radiant exitance as π, providing a panel radiant power in W/m. This number is calculated for a panel of a known temperature through each candidate membrane. Dividing this number by the radiant exitance of a true black body provided the hemispherical transmissivity, τ.