key: cord-0869127-itn1x5gm authors: Aviv, Dorit; Wee Chen, Kian; Teitelbaum, Eric; Sheppard, Denon; Pantelic, Jovan; Rysanek, Adam; Meggers, Forrest title: A Fresh (Air) Look at Ventilation for COVID-19: Estimating the global energy savings potential of coupling natural ventilation with novel radiant cooling strategies date: 2021-03-22 journal: Appl Energy DOI: 10.1016/j.apenergy.2021.116848 sha: dc0c013dec7b91e02823b975db1b1938befc572a doc_id: 869127 cord_uid: itn1x5gm Radiant cooling-assisted natural ventilation is an innovative technical approach that combines new radiant cooling technology with natural ventilation to increase fresh air delivery into buildings year-round with minimal energy cost and improves air quality. Currently, the standard paradigm for HVAC (heating, ventilation and air conditioning) is based on central air systems that tie delivery of heating and cooling to delivery of fresh air. To prevent heat loss, the delivery of fresh air must be tightly controlled and often limited through recirculation of already heated air. Buildings are designed with air-tight envelopes, which do not allow for natural ventilation, and depend on energy-intensive central-air systems. As closed environments, buildings have become sites of rapid COVID-19 transmission. In this research, we demonstrate the energy cost of increasing outdoor air supply with standard systems per COVID-19 recommendations and introduce an alternative HVAC paradigm that maximizes the decoupling of ventilation and thermal control. We first consider a novel analysis of the energy costs of increasing the amount of conditioned fresh air using standard HVAC systems to address COVID-19 concerns. We then present an alternative that includes a novel membrane-assisted radiant system we have studied for cooling in humid climates, in place of an air conditioning system. The proposed system can work in conjunction with natural ventilation and thus decreases the risk of indoor spread of infectious diseases and significantly lowers energy consumption in buildings. Our results modeling HVAC energy in different climates show increased outdoor air in standard HVAC systems can double cooling costs, while increasing natural ventilation with radiant systems can halve HVAC costs. More specifically it is possible to add up to 100 days’ worth of natural ventilation while saving energy, when coupling natural ventilation and radiant systems. This combination decreases energy cost by 10-45% in 60 major cities globally, while increasing fresh air intake. In the fallout of the COVID-19 (novel coronavirus disease 2019) pandemic, it has become clear that our reliance on conventional building Heating, Ventilation and Air Conditioning (HVAC) systems need to be rethought. It has been established that enclosed areas with limited ventilation increase the risk of transmission [1] . Because of this, the building sector must rethink the current HVAC paradigm in buildings. Currently, the primary mechanism used to save energy in centrally ventilated buildings is to mix return air with fresh air intake, but this puts fresh air and energy efficiency at odds. In the face of today's health and environmental crises, if we aim to provide building environments that are both healthy and energy efficient, ventilation and air exchange must be decoupled from heating and cooling (Fig.1 ). Air should be primarily for breathing, not for heating and cooling. This study addresses the energy consequences of COVID-19 and building ventilation through two analyses: the first demonstrates a challenge, and the second demonstrates an opportunity. First, we analyze the energy consequences of increasing the fresh air delivered through typical commercial building recirculating ventilation systems, caused by the significant air-conditioning costs in the latent and sensible cooling derived from a real system model. Second, we analyze the potential of alternative radiant and convective systems to compensate using thermal comfort models combined with global weather data, which outputs the added natural ventilation potential and is correlated to previous work to estimate the energy savings. The first analysis relates directly to the challenge of COVID-19 ventilation energy costs for commercial buildings. The second analysis demonstrates how strict setpoints can be relaxed and shows an alternative pathway to energy savings for commercial buildings, as well as critical potential for residential buildings that often have no mechanical ventilation to increase natural ventilation while decreasing energy consumption. (a) (b) Figure 1 : (a) Current paradigm: airtight building with a centrally controlled air-based heating and cooling system with recirculation damper for energy savings; (b) Proposed paradigm: naturally ventilated building with radiant system for thermal comfort provision. The alternative systems considered in this analysis are convective fans, radiant systems and novel sub-dewpoint membrane radiant systems combined with natural ventilation to augment thermal comfort in lieu of air conditioning. To put this research in context, we will first provide a review of studies that have encouraged natural ventilation in buildings and recent developments in buildings' radiant systems that allow for the use of natural ventilation while maintaining thermal comfort. Based on the review, we will discuss the applicability of radiant systems in a naturally ventilated space through an analysis of climates around the world. This research will inform the design of naturally ventilated environments that maintain consistent thermal comfort using advanced radiant systems. The COVID-19 crisis has amplified and exposed the flaws of the current HVAC design paradigm, while at the same time, providing us an opportunity and the motivation to improve the health of our indoor environment. The building sector's share of U.S. annual electricity consumption has grown dramatically in the past five decades from 25% of in the 1950s, to 40% in the early 1970s [2] . By 2012, it's share increased to 76% with a significant fraction of it dedicated to heating and cooling systems. As a result of the oil crisis in 1973, energy conservation in buildings became an important priority. To conserve energy, buildings were designed to be tighter and less ventilated. Airtight spaces resulted in poor ventilation, and indoor environments ailed with "sick building syndrome" [3] [4] [5] . It has been clearly shown that inadequate outdoor air supply rates significantly impact air quality and cause the "sick building syndrome" [6] . Consequently, to mitigate "sick building syndrome," ASHRAE and other entities recommended increasing the outdoor air ratio (OAR: the amount of fresh clean outdoor air mixed with the return air from rooms) in the design of HVAC systems [7] . Past reviews have demonstrated the many ways by which design for increased thermal comfort and ventilation rates has directly led to increased energy demand [8] . The COVID-19 pandemic has shown that current ventilation design standards of most commercial and residential buildings may be insufficient in providing the necessary ventilation rates to avoid indoor contagion [9] . Past studies have shown that recirculated air in buildings can be a source for indoor disease transmission [10] . In order to avoid indoor SARS-Cov-2 transmission, REHVA guidelines published in April 2020 [11] recommended an increase in air supply with a higher air exchange rate, and to avoid recirculated air if possible. ASHRAE has also recently published independent peer reviewed articles with guidelines that recommend systems "open minimum outdoor air dampers, as high as 100%, thus eliminating recirculation (in the mild weather season, this need not affect thermal comfort or humidity, but clearly becomes more difficult in extreme weather)" [12] . This recommendation recognizes the challenge of the large energy demand in extreme weather. More recently the ASHRAE Journal published that "researchers concluded that there is now sufficient evidence of airborne transmission of COVID-19 to justify improving ventilation and filtration where this would reduce SARS-CoV-2 exposure, and ASHRAE has adopted that position" [13] . If we continue to conform with the current building HVAC paradigm, increasing ventilation rates would result in a huge energy expenditure to support the heating or cooling of incoming air to achieve comfort levels. This is especially true in humid climates, where often more than half of the cooling ventilation load is due to dehumidification, which can result in significant overcooling of supply air [14] . This begs the question: how can we provide healthy indoor environments without a significant increase in energy consumption? One option is to simply open the windows. The use of natural ventilation in buildings has the potential for providing indoor air quality without expending energy on mechanical ventilation. Furthermore, certain studies have shown that the use of natural ventilation in buildings can provide means for prevention of airborne transmission, especially in tropical climates [15, 16] . When an air-based mechanical system is used, infiltration of unconditioned outdoor air can compromise the system's efficiency. Moreover, when the outdoor air is outside of the comfort zone, most models assume that the windows would be closed. We investigate a scenario in which comfort is maintained by a radiant system independent of the air conditions. In this scenario, the windows can be kept open for longer periods of time to maximize outdoor air to meet the recommended COVID-19 ventilation guidelines. Thanks to the use of surface-based radiant systems in lieu of air-based heating and cooling, we can increase the range of allowable outdoor air temperature and thus provide fresh air through additional natural ventilation hours beyond the conventional HVAC systems. Natural ventilation is as old as the human dwelling, but our buildings have evolved into complex thermal machines. Currently, we have many technical solutions that offer far more control over fresh air and thermal comfort than a simple facade opening, but what natural ventilation lacks in control it makes up for in simplicity and efficiency. Previous studies have shown the significant energy saving potential of natural ventilation, especially in hot climates [17] [18] [19] . The limits to natural ventilation's delivery of thermal comfort are primarily dependent on the conditions of outdoor air. These limits are often narrowly defined as 20-26 °C air temperature [20] . Natural ventilation has been superseded by HVAC systems because they could deliver controlled and predictable amounts of air at specific setpoints for temperature and humidity. But this control has resulted in significant increases in energy demand. Because of this, new strategies to design with natural ventilation and mixed-mode systems have gained importance. [21, 22] . Still, recent work continues to focus on losses from infiltration and lack of air-tightness of facades [23] , which leads to complex ventilation systems, and negates the potential of natural ventilation. Early thermal comfort models had fixed comfort zones that explicitly excluded many outdoor conditions. The Adaptive Comfort model allows for the zone to adapt to outdoor conditions, allowing warmer indoor temperatures in a naturally ventilated space during summer and cooler in winter, expanding the potential of natural ventilation [24, 25] . We propose to further consider other mechanisms to expand the potential of natural ventilation through adaptive thermal comfort. Radiant heat transfer and higher air velocity can significantly alter the net heat exchange of occupants independently of air temperature and humidity. These impact the adaptive thermal comfort calculations. We have demonstrated tools in previous work that can illustrate those potentials [26] . In order to increase the yearly hours during which natural ventilation can be utilized, we propose a strategy that uses fans and radiant systems for added thermal control and maintaining natural ventilation through porous enclosures. Previous work has demonstrated how natural ventilation hours can be determined by using comfort models combined with weather data available for different regions. These studies have gone from using standard comfort ranges [20] to using adaptive thermal comfort models in four US cities recognizing variations in climate and occupant expectations [27] . Another study has been carried out globally [28] using a large weather file dataset, but using comfort limits of 17 °C dewpoint for cooling and assuming internal heating gains of buildings to compensate for temperatures as low as 12.8 °C. This proposal builds on this most recent global analysis by adding the potential of radiant systems as a comfort mitigation strategy for an even wider range of conditions. It is important to note that local outdoor air quality is another major challenge to natural ventilation [29] . Outdoor pollution in many places makes natural ventilation not just challenging, but unhealthy, and has been shown to affect energy savings potential [45] . Yet, we also see the potential for centralized government policies and actions to mitigate causes of poor outdoor air quality at rates faster than similar efforts to reduce carbon emissions and energy consumption [30, 31] . The circumstances of the pandemic itself have had a significant impact on reducing air pollution in certain locations [32] , so this variable is subject to relatively rapid change. However, it is also possible to use state-of-the-art technologies for window-based solutions for filtering particulate matter from the air where air pollution is currently present, [33, 34] thus still allow for large quantities of air to enter the indoor space. The air velocity loss due to the filter can be overcome by using low-energy fans in the interior as we shall describe in the Methods section. When assessing the suitability of a location for natural ventilation we limited our assessment to climatic variables, as we believe it is viable to consider the theoretical potential of natural ventilation. The premise of this paper is to support discussions on the co-benefits of improving outdoor air quality alongside reducing global carbon emissions. Radiant heating and cooling systems rely on infrared radiation exchange between occupants and surrounding surfaces. This is a low-energy alternative to energy intensive air-based systems for thermal comfort delivery [35] . Under room conditions with similar air and surface temperatures, the heat transfer exchanged by radiation to surfaces and convection to air is roughly equivalent [36] . That is, half of one's thermal perception is driven by surface temperatures. Therefore, by manipulating surface temperatures to be colder or hotter, comfort can be maintained at a greater range of air temperatures. It is by this mechanism that increased potential for natural ventilation hours can be achieved. In addition, for commercial buildings using radiant systems, the setpoints for air supply systems can be shifted to enable higher fractions of outdoor air and to allow 100% outdoor air (equivalent to enabled natural ventilation) for a larger number of hours during the year. There are two main challenges to radiant systems that successfully enable more natural ventilation and reduce energy use as a consequence of increased outdoor air. First, for natural ventilation in hot humid climates, the dew point temperature is often higher than the necessary surface temperatures for radiant cooling, which means condensation would create water problems inside. Second, when using radiant systems, large surface areas are generally needed for adequate view factors to occupants. Therefore the convective exchange with indoor air is also significant so even if setbacks to warmer air temperatures in summer or cooler in winter are desirable, the temperatures will tend to follow the radiant panel temperature [37, 38] . For natural ventilation cases, this also implies a larger fraction of losses as the high amount of air exchange would cost more convective losses. Still, radiant systems offer an alternative path to maintaining comfort through a wider range of air temperature conditions, and thus environmental conditions for natural ventilation. We will consider the potential of a typical radiant system to increase natural ventilation hours and consider the limitations of operation during humid or excessively cool conditions. We will also leverage our recent development of membrane-assisted radiant panels that mitigate condensation and convective-loss risk to further expand potential natural ventilation hours. We have developed a radiant panel with a membrane transparent to infrared emissions. This protects it from risk of condensation and minimizes convective exchange with the surroundings [39] . We built a pavilion in Singapore that demonstrated its operation under hot and humid conditions without condensation. Our thermal comfort study on the pavilion validated that people could be made to feel "cool" with panels operating below the dew point without condensation [40] . The infrared-transparent membrane allows water to be supplied below the ambient dewpoint without condensation, increasing the potential for radiant cooling to augment cooling in naturally ventilated spaces. The membrane convectively gains heat from the ambient air to remain warmer than the dewpoint, a relationship characterized in the Teitelbaum, Chen, et al. study [28] . The allowable water supply temperature depends (Twater in Eq. 1) on the ambient air and dewpoint temperatures and is given in equation 1. By inspection, as the air temperature increases, cooler water can be supplied to the panel, an elegant feature that allows more cooling to be supplied when more cooling is demanded. The system operated successfully without condensation in Singapore with supply temperatures between 13 and 17 o C, despite an outdoor dewpoint of 23 o C. The study described in Teitelbaum et al. [40] was limited to outdoor thermal comfort. However, this study demonstrates that the operation and control parameters from that study can be directly implemented to an indoor environment. This paper will refer to the combination of membrane-assisted radiant cooling panels with natural ventilation as 'radiant cooling-assisted natural ventilation'. This scope of this study includes the following: • Analysis of the energy cost of increasing outdoor air intake to a maximum with a conventional HVAC system • Analysis of an alternative HVAC solution which couples a membrane-assisted novel radiant system we developed with natural ventilation for ensuring increased fresh air intake while maintaining thermal comfort indoors without an energy uptick • Regression applied to the results of global natural ventilation potential study by Chen et al. [28] in order to generate a predictive relationship model between natural ventilation hours and energy savings • Estimates of the energy savings that can be achieved in 60 major cities globally by through the use of radiant cooling-assisted natural ventilation • Comparison of the cooling energy consumption of three cooling scenarios in 4 major US cities situated in different climatic zones • Analysis of the potential additional hours per climate zone globally that would be comfortable with radiant cooling-assisted natural ventilation using the adaptive comfort model We have developed a method that calculates and illustrates how much additional energy would be consumed by conventional air conditioning systems to meet increased fresh (outdoor) air recommendations for COVID-19. Conventional HVAC systems are reconfigured to introduce additional outdoor air, thereby increasing the system's outdoor air ratio (OAR). OAR is the ratio of outdoor air supplied into a recirculating central air conditioning system as it relates to the total air supplied to the building. A window air conditioner has 0% outdoor air and a dedicated outdoor air system (DOAS) has 100% outdoor air. But the majority of commercial buildings have ventilation systems with air handlers that use dampers to adjust the amount of outdoor air brought into the supply stream to be conditioned by heating or cooling coils. The OAR is thus adjusted for the needed ventilation. This is the control variable that facilities operators are being asked to increase to reduce recirculated air due to CoViD-19. We analyzed the increase in cooling energy as it has both sensible and latent components with varying influence on system performance and energy demand. For the heating case, the load is generally proportional to the increased OAR because it is dependent only on the sensible heat based on the dry bulb temperature differences. The cooling sensible and latent energy depend on both the outdoor temperature and humidity relative to the supply conditions, as well as the return air conditions that are influenced by internal gains of both heat and humidity. The relative amount of sensible cooling to the total cooling (sensible + latent) is called the sensible heat ratio (SHR), and it is determined by a combination of the aforementioned temperatures and humidities, along with the OAR. The cooling loads and SHR are critical factors in estimating the energy consumption for a typical chiller used to cool a building. We aimed to demonstrate how OAR influences it and the subsequent performance across a series of typical cities. To calculate cooling loads, we first needed to determine the annual cooling demand generated by a reference building envelope. We selected the standard small office building provided by the Department of Energy for use in EnergyPlus simulations [41] . The building is 511 m 2 (5,500 ft 2 ), and a single climate case for this building was selected. The building was simulated in EnergyPlus to generate a generic envelope load with an indoor thermostat setpoint of 24 o C (75 o F) during occupied hours (6am-7pm) with 10% OAR and no internal gains. This output is a baseline hourly envelope heat gain and cooling demand. The hourly envelope loads were inputted into the system model that was built in Engineering Equation Solver (EES). In this model additional parameters were tuned, which allowed for more flexible analysis and sizing of the unit for latent and sensible loads. The overall governing energy balance is shown in Equation 2 that relates energy change in the air to the change in sensible temperature (cooling) and the change is specific humidity (dehumidification). A12.5-ton Lennox air conditioner was modeled using the compressor performance curves from the manufacturer. These performance curves relate system COP (coefficient of performance = cooling/electricity) to the hourly SHR, airflow rate, setpoints, and load. The load was a combination of the envelope gains, internal gains calculated in EES using daily occupancy for 80 individuals, and the air condition gains based on the EES model of outdoor air, return air influenced by envelope and internal gains, and supply setpoint that was aimed at 50% OAR and 24 o C (75°F). The supply setpoint and gains determined return air conditions which was where the impact of OAR was determined and was fed back to the SHR for system performance. This feedback was why EES interactive solving was ideal for the problem. From this we were able to determine the hourly energy consumption of the chiller and thus the cumulative cooling demand for the season as in equation 3. (2) = + ℎ where Q air is the cooling demand (J) m is mass (kg), c p is heat capacity of air (J/kg/K), ∆h cond is the heat of condensation (kg/J), ∆q is the specific humidity change due to dehumidification(kg/kg), the COP is the coefficient of performance (J cooling /J elec ), is the temperature difference of the outdoor air to the supplied air and Echiller is the electrical demand of the air conditioning system. This simulation was carried out using weather data files for Miami, Phoenix, Duluth, Chicago, Washington D.C., and San Francisco with a baseline OAR of 10% for each city, and then increasing the OAR to 25%, 50%, 75% and 100% respectively for each city. When the OAR was increased there was an additional step to check if the 12.5-ton unit was sufficiently sized to meet the peak demand as increasing OAR increased peak demand. This was done based on the EnergyPlus envelope peak load along with the design day conditions for 0.4% max sensible and coincident dew point temperature, meaning that 99.6% of the time it would be able to achieve the setpoint. The performance curves of the reference 12.5-ton system were scaled proportionally to meet the demand. We also noted where the 12.5-ton system failed as OAR is increased. We also developed a simple analytical solution for the linear relationship between raw air conditioning cooling energy and OAR based on outdoor air (OA) conditions, return air (RA) conditions, and supply air (SA) setpoints expanding equation (2) for outdoor and and return air energy fractions. Rearrange and substitute for : Recognize the components of outdoor air and indoor air: Where A is the specific energy (J/kg) of outdoor air and B is the specific energy for return air, gives the following linear relationship between OAR and air cooling energy: Therefore the energy to condition the supply air per unit supply air linearly increases with OAR at a rate determined by the difference in specific energy of outdoor air and return air, and with an intercept at zero OAR defined by the specific energy of the return air when no outside air is supplied. As this is normalized by total air supply it is clear that both air supply amount and the amount of fresh air both directly increase the energy needed to condition that air for indoors. We can also observe how the slope of the linear relationship describes the rate of energy increase per increase in %OAR: Finally, based on the definition of SHR we can again use substitution for OAR for the mass of outdoor air, return air, and supply air to develop the relationship: We built a surrogate regression model of the energy savings calculated by Chen et al. [28] . They provided estimates of the percent global energy savings, , of widespread adoption of (%) natural ventilation in the form of mixed-mode natural ventilation (MMNV) systems. Their savings compared annual energy use of a mixed-mode system versus a 100% mechanically-ventilated central air-conditioning (central AC) system. Energy use estimates were generated using a standard Department of Energy commercial building model, using the EnergyPlus simulation software, of an archetypal three-story office building. They simulated the performance of this archetypal building across 60 different cities around the globe. Chen et al. also developed a methodology for estimating the number of occupied hours for which natural ventilation is permissible (NV%) which could be calculated from the outputs of the energy simulation process. In their study, they defined the comfort zone for acceptable natural ventilation as having: A. A fixed lower temperature threshold of 12. The adaptive comfort model is described in equation (11) . where T cmfrange is the band of operative temperatures that an occupant can feel comfortable ( o C), T a,out is the mean outdoor temperature ( o C) and T accept is the width of the temperature band ( o C). T accept is 2.5 o C and 3.5 o C for 90% and 80% thermal acceptability respectively. The upper threshold can be further widened through elevated air movement. An elevated air velocity of 0.6-0.9 m/s, 0.9-1.2m/s and above 1.2 m/s will increase the upper threshold by 1.2 o C, 1.8 o C and 2.2 o C respectively. We generated a surrogate regression model, , of Chen et al.'s analysis to forecast the energy savings of radiant cooling-assisted natural ventilation, specifically establishing a relationship between ES% and applicable natural ventilation hours: Where: = Annual number of natural ventilation-applicable hours as a fraction of 8760 hours ℎ = Annual number of hours where it is too hot outside for natural ventilation; fraction ℎ 2 out of 8760 hours Figure 3 provides a plot of vs as per Chen et al, along with the established regression model. Though the model is probabilistic, with illustrated confidence bounds on the prediction, for the purposes of this study we will only use a deterministic model of the median, which can estimate within a standard error of 10% of its unit value. This deterministic model is represented by a quadratic equation 14 (see Fig. 2 plot) . (14) (ℎ 1 ,ℎ 2 ) = 23.18 ℎ 1 2 +29.04 ℎ 1 +24.14 ℎ 2 2 +18.74 ℎ 2 We model the potential extension of natural ventilation using membrane-assisted radiant cooling panels [40] along with increased air movement [57] . For the case of predicting energy savings of radiant cooling-assisted natural ventilation, we apply the following method: The adaptive comfort model is only applicable to conditions having T a,out (running average outdoor air temperature) between 10 o C to 33.5 o C and internal operative temperature between 17 °C and 32 °C. When relying on a radiant system for comfort as proposed in this paper, the acceptable air temperature range expands to as high as 40 o C. The adaptive comfort model is not shifted, only the approach for creating acceptable operative temperatures through radiant-only systems [26, 40] . where are the surface view factors between a point in the room and all the surrounding → surfaces, used to weigh the surface temperatures (K) at each surface. The MRT is scaled for a standing person [44] . The MRT is then used with the air temperature and air velocity to determine operative temperature and the comfort of an occupant in the adaptive comfort model. The radiant cooling NV hours, h-RCNV , and the too hot hours, h 2H,RCNV , are calculated below using the adaptive comfort ranges described in equations below: , , = + ( 10 ) where T cmfrange,lwr,i and T cmfrange,upr,i are lower and upper threshold of the acceptable comfort range of the adaptive comfort model as calculated in eqn(11) of a particular hour, i, in a year. T op,i,RV is the operative temperature of a particular hour, i, in a year. The T op,i,RV is calculated based on eqn (21) with MRT calculated using eqn (13) . v is the indoor air velocity (m/s). Once we have calculated the additional hours of natural ventilation, we can apply the quartile regression model, , to estimate the additional energy savings. But we must also incorporate the energy cost of running the membrane-assisted radiant panels, which we have shown operate only with sensible cooling and no latent cooling demand [40] . In order to estimate the energy savings, , we considered several precedents to estimate the energy required to operate the panel system. In a study by Niu et al. [53] , conducted in a hot and humid climate, the comparison between all-air system and radiant ceiling system yielded major reduction in sensible load in the radiant system case, thereby halving overall energy demand, when dehumidification is excluded. We reviewed climate data and found SHR (sensible heat ratio) varies from 50% to 90%. While the local climate is not the only factor that affects latent cooling ratios, latent cooling demand is found to be in the order of 20% to 50% of total building end-use cooling loads. [54, 55, 56] . We adopt an estimate of 35% , i.e., that the energy savings of using the radiant panel with = natural ventilation in comparison with a central A/C system, equals 35% of the savings of using natural ventilation alone. It is important to note that while the elimination of latent load can lead up to 50% savings based on precedents, there are significant additional savings on the efficiency of the sensible delivery. By eliminating the Chen et al [28] dew point limit, the new comfort hours also occur most prevalently in high latent cooling hours. Therefore a conservative 35% reduction provides a reasonable lower bound to the savings. While re-evaluating the natural ventilation hours for radiant panels we adjusted the boundary conditions used for Chen et al. [28] , eliminating the dew point limit, which is based on typical building system limits, not actual thermal comfort. We also recognized that much of the natural ventilation hours analysis in their study was influenced by building properties such as internal gains and thermal mass that generate the assumption that 12.8 °C air temperature is warm enough for natural ventilation (which is below the adaptive comfort range). Likewise, it is building air conditioning system properties that generate the assumption that there can't be natural ventilation above a 17 °C dew point. For a global analysis of NV hours, it is difficult to systematically aggregate diverse building properties. Therefore, we re-analyzed the global natural ventilation hours using the adaptive comfort model assuming no building influence by simply letting indoor conditions follow outdoor conditions without any assumed building influence. This would only be valid for an infinitely lowmass building, or one that is indicative of current common outdoor heating and cooling strategies. Our analysis used the typical meteorological year data of 2590 locations obtained from the EnergyPlus website, which is derived from 20 weather data sources [42] .The analysis is used to determine the number of potential naturally ventilated hours for these scenarios: 1. Base case scenario with no installation of mechanical systems. The indoor air velocity due to natural ventilation is calculated based on equation (21). where u in,max is the maximum indoor air velocity, C 1 is the wind speed coefficient, u out is the outdoor air velocity, C 2 is the buoyancy coefficient (ms -2 K -1 ), h is the vertical height of the opening (m), △T max is the maximum temperature difference between indoor and outdoor, C 3 is the turbulence coefficient (m 2 s -2 ). The values of C 1 , C 2, and C 3 are 0.001, 0.0035 ms -2 K -1 and 0.01 m 2 s -2 respectively [28] . The 90% acceptability temperature range of 5 o C is used for △T max . The vertical variation of air velocity due to height is not considered, thus h is assumed to be 1.5m. (Figure 2c ). 4. Membrane-assisted radiant scenario uses the membrane-assisted radiant panels for cooling. This scenario is similar to the third scenario with the exception that radiant cooling is allowed even when the dew point is above 18 o C (Figure 2b ). The effective panel temperature is determined using equation (1) . The effective panel temperature is 3 o C higher than T water . Our analysis demonstrated that there is an increase in cooling energy demand with increasing OAR (Fig. 4) . The baseline OAR for a central air system is 10%, which is the minimum allowable OAR intake per ASHRAE standard (10 L/s/person outdoor air for 20 people) for the office building with a 2,000 L/s AHU modeled. At 25% OAR the increases range from 6% increase in energy to a 26% increase and at 80-100% OAR they diverge more, and range from 29% to greater than 215%. The largest increase is for Miami across the board in both percentage and absolute values. Washington, DC is the next largest absolute increase in cooling energy. The percentage increase allows us to compare the form of the OAR-energy relationship across all cities, dictated by their climate. As expected, San Francisco with its moderate climate and low cooling demand experienced the least cooling increase even with 100% outdoor air. Miami and Washington DC simulations were capped at less than a 100% OAR since high outdoor air humidity made further increases require impractical physical changes to the HVAC control algorithms. (a) (b) Figure 4 : (a) The percent increase in annual cooling energy for 6 cities for increasing OAR from 10% to 100%; (b) the absolute cooling energy increase. We considered the analytical OAR and SHR relationships to interpret the potential influences on the form of the modeling results shown in Fig. 4. Fig. 5 shows a linear relationship between percentage energy increase and OAR, and also explains how SHR relates to OAR for fixed supply conditions with two different outdoor air and return air conditions that are representative of a warm humid climate and of a hot dry climate. We have observed that the rate of energy increase per percentage increase in OAR (the slope in Equation 9 ) is 25% when going to a more humid condition. This appears to be the case for the drier San Francisco having a lower slope than the other cities, as shown in Fig. 4 . Despite being in the desert, Phoenix actually has some relatively high latent loads to contend with and both high sensible and high latent loads contribute to increasing the slope. The OAR-SHR relationship, analytically plotted in Fig. 5 , shows how higher OAR ratio significantly decreases SHR, while drier conditions are not as sensitive to OAR. It also shows how completely dry conditions where dehumidification is not necessary will be solely dependent on managing internal humidity gains and thus SHR will increase with higher OAR. This is important to note as the EES model of the compressor in the reference air conditioner is dependent on SHR at each dry bulb and wet bulb air condition. Therefore, the cooling energy increase for Miami at the high OAR shown in Fig. 4 is likely due to the less efficient system performance at lower SHR. For Miami, we plotted the range of air conditioning electricity output from the hourly model for Miami. Our analysis summed for each month across the year for 25%, 50%, and 75% OAR (Fig. 6 ). It is clear how the values diverge in the summer months during peak cooling. The significant increase in demand is also clearly depicted. Even with a larger chiller, the inescapable coupling of dehumidification with air conditioning equipment will greatly overcool the air supplied to a room, as it will bring in fresh air, but create uncomfortable conditions. This is particularly true when the SHR is low because high capacity dehumidification is needed to cause excess sensible cooling. This is particularly true with high humidity climates during lower sensible temperature times when high OARs are deployed. Fig. 7 shows that at a thermostat setpoint of 24 o C (75 o F) air and with 50% OAR, there are hundreds of summer hours in the reference office building in Miami that have air that is well below this setpoint, due to the need to adequately dehumidify the air which results in overcooling. Our analysis estimated the potential impact of using the membrane-assisted radiant panels that enable increased natural ventilation for a building. Table 1shows results of the 60 locations examined in this study, the same cities as studied by Chen et al. [28] . The description of each column is as follows: • City: Name of the location. • Natural ventilation comfort hours: Number of hours comfortable without the operation of any mechanical systems, calculated as per Chen et al. [28] , and defined as hMMNV in section 3. With membrane-assisted cooling, the 60 cities achieved a range of 80-100% of comfortable hours with natural ventilation. In some climate contexts, particularly the hot and humid tropics, this is a significant increase in applicable NV hours compared to natural ventilation only as per Chen et al. [28] . This increase in applicable natural ventilation hours does also provide additional energy savings, though modest. In Fig. 8 , the assumed baseline, or backup, central AC system is a mixed fresh air / recirculated air system with an outdoor air ratio (OAR) of 10%, supplying 10 L/s/person of fresh air. With such a system in tropical cities like Miami, the additional energy savings offered by radiant cooling-assisted natural ventilation is in the order of 10-12% above natural ventilation only. In temperate cities such as Washington, the additional energy savings are less than 5%. These cities' total annual building energy use is increasingly dominated by annual heating energy demand. The estimated energy savings are altogether not significant as we recall that radiant cooling-assisted natural ventilation reduces latent cooling requirements, and associated fan power costs, but not the sensible cooling energy demand of a building -which can be substantial. The relationship between natural ventilation hours and energy savings becomes more pronounced when we consider a baseline scenario of a mechanical, central AC system that aims to provide a higher rate of fresh air. In Fig. 9 , we compare the four cities highlighted in figure 8 , but against different regimes of OAR. If, for example, one compares radiant cooling-assisted natural ventilation to a baseline building that seeks to provide occupants with a 50% OAR, the estimated annual energy savings increases to between 25 to 45%, highlighting the need to approach increasing outdoor air flow rates when energy use is constrained. Figure 9 : Ratio of change in cooling energy expended for different cooling scenarios:(left) All cooling and ventilation is provided through a central-air mechanical system; (middle) Mixed mode ventilation in which natural ventilation is allowed when outdoor air falls within the comfort zone and a mechanical system sets off otherwise; (right) A radiant system provides heating and cooling with natural ventilation providing fresh air at all times in which this combination falls within the adaptive comfort boundaries and a A/C used as backup otherwise. For all cases, the projected increase for 50% OAR is included. In section 3.2.2 and 3.3 we highlighted a difference in how Chen et al. [28] established criteria for estimating applicable natural ventilation hours, versus ours. Here we examine the relative changes in forecasting baseline applicable natural ventilation hours when we apply the latter method. Four maps were generated for visualizing the extended natural ventilation hours analysis. (Fig. 10) . The global map was divided into about 300km x 300km grids. Each grid shows the average results of the locations. The map in Fig. 10a shows the base case, i.e., the number of hours in a year that a location can achieve comfort, according to the adaptive comfort model, without running any mechanical system and with the windows open. We can see that due to the consistent temperatures of the tropical region, there is high potential for the region to achieve comfort without any use of mechanical system. Fig. 10b shows the improvement in the number of comfort hours by 480-960 hours (5-11%) if we elevate the air velocity to 0.8 m/s using mechanical fans. This strategy is only effective for the tropical region. By coupling elevated air velocity with radiant panels, we can further increase NV hours by 480-960 (5-11%) and maintain comfort (Fig. 10c) . The condensation risk of radiant panels limits the increase in comfort hours for regions with high humidity. When the condensation risk is mitigated by the membrane-assisted panels, and mechanical fans are coupled with membrane-assisted radiant panels, we can achieve 480-1440 hours (5-22%) of extra comfort according to Fig. 10d . The results of the absolute natural ventilation hours mapped correlates well with the previous studies for places that fit in the respective comfort bands, analysis of Singapore in the tropics [49] and cities in the United States [29] . Compared to the Chen et al. natural ventilation potential [28] , we find an expected shift in comfort limits for the bounds of areas with natural ventilation potential. Mapping both studies in Fig. 12 , we observe that because Chen et al.'s study uses a temperature range down to 12.8 °C independent of the adaptive comfort model, that study's results show more NV hours for colder regions. These would be deemed too cold by the adaptive comfort threshold. One could argue that their approach fails to recognize potential cold air infiltration in lightweight buildings, while our approach neglects any potential benefit of internal gains or thermal mass. Though less prominently displayed on the map, the tropical potential is also significantly expanded in our approach in comparison to the Chen et al. study by eliminating the dewpoint limit. We made the last assumption in light of our work on the more novel membrane-radiant systems being demonstrated in a fully outdoor environment with no significant internal gains to the air condition [40] . There we have shown that high humidity alone does not cause discomfort so long as the necessary metabolic heat dissipation is managed via convective and radiant heat transfer without requiring additional sweating/evaporative cooling. In comparing our analysis of the energy use penalty of increasing outdoor air (OAR), we were unable to find a paper or other published research that explicitly analyzed its physical relationship as an energy penalty for specific system operation. Numerous papers have explored ways to increase the performance of HVAC delivery of fresh air. Several of them were reviewed based on thermal comfort setbacks [8] , which show savings from 29% in Hong Kong to 69% in Las Vegas for standard systems. This confirms the sensitivity of energy demand to the difference from supply temperature to source, which was also confirmed in previous EnergyPlus modeling of setpoint setbacks [45] . Another review looked at energy conservation in ventilation strategies [46] and demonstrated systems that can reduce the penalty paid by increasing outdoor air including energy and heat recovery as well as geothermal pre-cooling. We did not consider the role of these devices and strategies in improving performance. Still, while these are also great options to increase fresh air, they are not yet prevalent in standard installations. Finally, we considered our analysis relative to research and analysis that consider air enthalpy instead of just sensible temperature as is done in degree-day calculations. Degree-day analysis includes the comparison of modeling accuracy [47] that describes the significant energy cost of humidity on the air-side of cooling systems and also shows an improvement in energy prediction using enthalpy cooling degree-days. Similarly it was shown in another study how expected climate change will increase the enthalpy gradients that buildings will have to address, thus exacerbating the energy penalty of increased outdoor air ventilation rates [48] . We can therefore confirm our results qualitatively from previous work that the intuitive nature of bringing in more unconditioned air, particularly if it is hot humid air, has very negative feedback that increases energy costs. The form of the simulation results illustrates potential negative feedback on machine performance from changing SHR, and the fundamental analytical relationships provide a more general backdrop to directly consider OAR increase penalties that should be considered in practice. Further, we have confirmed that this scenario has occurred in the field during the COVID crisis, as the facilities managers at Princeton University were instructed to open outdoor air dampers to their maximum setting to bring as much outdoor air as possible into buildings, which is the scenario that was modeled. The energy savings assessment for natural ventilation in our analysis is dependent on the regression of the cooling analysis done by Chen et al. [28] It is important to recognize that in their paper the energy savings assessment is still based on their EnergyPlus model of a standard system. If we consider the energy consumed by a ceiling fans, it is easily as low as 1 W/m 2 , and we therefore argue, it would be insignificant. But we do recognize the importance of considering the delivery of radiant cooling to the surfaces modeled. As the novel membrane-assisted radiant system does not have yet an in-depth energy model for calculating its cooling load, our analysis considers 35% savings for this system in comparison with a central A/C, based on radiant cooling precedents. We plan future analysis to more accurately model the energy savings of radiant cooling-assisted natural ventilation. We also recognize and advocate that building performance estimates presented in this paper, as per tables 1, figures 8 and 9, should be considered rough approximations. There are many assumptions which underlay our analysis, along with the complementary work of Chen et al. [28] . We have not examined different building design and typology to judge whether natural ventilation can be freely provided with windows open in all building spaces. We have not examined outdoor air quality and the effect of noise and pollution on natural ventilation viability. We have not considered the impact of climate change on these forecasts nor optimization of mechanical ventilation strategies. For example, implementation of heat recovery ventilators (HRVs) can reduce central AC energy use under higher OAR configurations. Likewise, the baseline backup system for a radiant cooling-assisted natural ventilation system may be a HRV-based air handling system, such as dedicated outdoor air system (DOAS). Finally, we recognize there are many other factors affecting access to air and airflow in buildings, such as the sensitivity of natural ventilation to outdoor air pollution. Yet, we suggest that studies which show that the common, "solution to pollution is dilution" cliché for increasing OAR and ACH in rooms as the best method of mitigating disease spread fail to recognize the non-uniform airflows and increased mixing that occurs in high velocity systems [51] . For COVID-19, the spread of the virus has been shown to be caused by directional airflows as well. Still, it remains clear that the vast majority of cases are transmitted indoors, and transmission would be reduced by having more fresh air exchange. Our analysis demonstrates significant potential for expanded use of mixed-mode systems to allow for more naturally ventilated hours. Also, in considering the cities and places that can achieve up to 100% natural ventilation, it is implied that building design in these locations can be freed from conventional ducting and air handling systems, which changes the design paradigm of walls and enclosures significantly. While teaching classes in outdoor open air during the 2020 pandemic to mitigate transmission risk, heat and humidity as well as cold temperatures were experienced during different days in temperate climates. Both could be addressed by the alternative radiant and fan paradigms we presented here. Future work should also consider how specific times of the year, such as semesters for schools, could be prioritized for analysis, and how occupied time of commercial buildings and residential buildings could be used to more accurately consider which hours in which buildings are most valuable to have natural ventilation potential increased. We recognize that our proposed interventions are unconventional, but with the amount of change happening in work settings and building operations, alternative paradigms need to be considered, especially those that improve health and safety of occupants and dramatically decrease energy consumption. In light of mounting evidence for the airborne transmission risks of the SARS-COV-2 virus, we must rethink the fundamentals of HVAC systems designed to provide adequate fresh air intake without compromising buildings' energy performance. Natural ventilation is an affordable and sustainable method for achieving this goal, but it presents challenges for thermal comfort, when the air temperature falls outside of the comfort zone. We have demonstrated that it is possible to significantly increase the potential for natural ventilation hours in buildings within multiple climatic zones by using radiant systems instead of air-based systems for temperature control. Radiant systems are proven to be more energy efficient than air-based systems, and at the same time provide a high degree of occupant satisfaction. Our models and calculations demonstrated that increasing to 100% fresh air (OAR) in standard building HVAC can significantly increase energy costs in humid climates. Alternatively, we have demonstrated that increasing natural ventilation hours by using a radiant cooling system significantly reduces the cooling energy usage. When considering a delivery 50% OAR scenario, 20-45% colling load reduction can be achieved with the use of a radiant system in comparison to a central A/C system. An even higher cooling load reduction if 100% OAR delivery is considered. Using the adaptive comfort model, we have shown that it is possible to increase natural ventilation hours by up to 100 extra days (2400 hours) per year using our proposed novel radiant system for indoor temperature control. Most of the regions with the highest potential to make this switch to natural ventilation are located in Southeast Asia and South Asia. While recognizing the need to address the challenges of air pollution and noise mitigation in many of the megacities of these regions, this research shows that the potential for natural ventilation in hot humid climates should not be overlooked, as it presents significant energy savings and potential health benefits. The most significant general finding from this study is that increasing the outdoor air intake to 100% fresh air with existing mechanical ventilation systems will result in doubling space cooling energy demand in hot and humid climates, whereas switching to a 100% radiant cooling system coupled with natural ventilation will reduce net cooling energy demand and retain thermally comfortable conditions. Under the new proposed paradigm, air is supplied to building occupants for the purpose of breathing, not for heating and cooling. Thermal comfort can be instead controlled by surface temperature manipulation. By decoupling air supply from temperature control, it is possible to achieve both high levels of indoor air quality and energy savings in interior spaces at the same time. World Health Organization. Transmission of SARS-CoV-2: implications for infection prevention precautions: scientific brief Sick building syndrome and the problem of uncertainty: Environmental politics, technoscience, and women workers Reflections on the history of indoor air science, focusing on the last 50 years The Effect of Varying Levels of Outdoor-Air Supply on the Symptoms of Sick Building Syndrome The effects of outdoor air supply rate in an office on perceived air quality, sick building syndrome (SBS) symptoms and productivity Refrigerating, Air-Conditioning Engineers, American National Standards Institute. Ventilation for acceptable indoor air quality. American Society of Heating, Refrigerating and Air-Conditioning Engineers Thermal comfort and building energy consumption implications -A review How can airborne transmission of COVID-19 indoors be minimised? Airborne spread of measles in a suburban elementary school Guidance for Building Operations During the COVID-19 Pandemic Filling the Knowledge Gaps HVAC and COVID-19 Dehumidification and Cooling Loads From Ventilation Air Natural ventilation for the prevention of airborne contagion Roles of sunlight and natural ventilation for controlling infection: historical and current perspectives Energy saving potential of utilizing natural ventilation under warm conditions-A case study of Mexico Analysing natural ventilation to reduce the cooling energy consumption and the fuel poverty of social dwellings in coastal zones Naturally comfortable and sustainable: Informed design guidance and performance labeling for passive commercial buildings in hot climates A Method to Assess the Suitability of a Climate for Natural Ventilation of Commercial Buildings Mixed-mode cooling Principles of Hybrid Ventilation Weather correlations to calculate infiltration rates for U Developing an adaptive model of thermal comfort and preference Nudging the adaptive thermal comfort model Design with Comfort: Expanding the psychrometric chart with radiation and convection dimensions Impact of adaptive thermal comfort on climatic suitability of natural ventilation in office buildings Investigating natural ventilation potentials across the globe: Regional and climatic variations Impact of outdoor air quality on the natural ventilation usage of commercial buildings in the US Air pollution reduction in China: Recent success but great challenge for the future Air pollution success stories in the United States: The value of long-term observations Changes in US air pollution during the COVID-19 pandemic Direct blow-spinning of nanofibers on a window screen for highly efficient PM2. 5 removal Mass production of nanowire-nylon flexible transparent smart windows for PM2. 5 capture Thermal-comfort analysis and simulation for various low-energy cooling-technologies applied to an office building in a subtropical climate Subjective response to radiant and convective environments Comparison of mean radiant and air temperatures in mechanically-conditioned commercial buildings from over 200,000 field and laboratory measurements Optimizing Radiant Systems for Energy Efficiency and Revisiting radiant cooling: condensation-free heat rejection using infrared-transparent enclosures of chilled panels Membraneassisted radiant cooling for expanding thermal comfort zones globally without air conditioning Weather Data Sources Simulating Invisible Light: Adapting Lighting and Geometry Models for Radiant Heat Transfer Standardization IO for. ISO7726 Ergonomics of the thermal environment. Instruments for measuring physical quantities Extending air temperature setpoints: Simulated energy savings and design considerations for new and retrofit buildings Review of energy conservation technologies for fresh air supply in zero energy buildings Prediction of cooling energy use in buildings using an enthalpy-based cooling degree days method in a hot and humid climate Daily enthalpy gradients and the effects of climate change on the thermal energy demand of buildings in the United States Applying Natural Ventilation for Thermal Comfort in Residential Buildings in Singapore Energy saving potential of natural ventilation in China: The impact of ambient air pollution Adequacy of air change rate as the sole indicator of an air distribution system's effectiveness to mitigate airborne infectious disease transmission caused by a cough release in the room with overhead mixing ventilation: A case study High Efficiency 'Low-Lift'Vapour-Compression Chiller for High-Temperature Cooling Applications in Non-Residential Buildings in Hot-Humid Climates Energy Savings Potential of Chilled-Ceiling Combined with Desiccant Cooling in Hot and Humid Climates Matching the sensible heat ratio of air conditioning equipment with the building load shr Evaluating the use heat pipe for dedicated ventilation of office buildings in Hong Kong Energy savings potential of a hybrid desiccant dehumidification air conditioning system in Beirut Exploring membrane-assisted radiant cooling for designing comfortable naturally ventilated spaces in the tropics Writing -Review & Editing, Investigation, Visualization Methodology, Software, Writing -Original Draft, Writing -Review & Editing, Investigation, Visualization, Formal Analysis, Data Curation Eric Teitelbaum Conceptualization, Methodology, Writing -Original Draft, Writing -Review & Editing, Investigation, Formal Analysis Denon Sheppard Resources, Formal Analysis Jovan Pantelic Conceptualization Methodology, Writing -Original Draft, Writing -Review & Editing, Investigation, Visualization, Formal Analysis Methodology, Writing -Original Draft, Writing -Review & Editing, Investigation, Formal Analysis recommends increasing indoor fresh air supply to mitigate COVID-19 spread • Increased fresh air supply may double energy costs when operating a conventional HVAC system • We propose an HVAC solution which couples novel radiant systems with natural ventilation • We estimate this solution would address thermal comfort needs in all major climate zones • Where adopted, this solution could reduce building HVAC energy The authors would like to thank Miranda Mote for her help with the final proofs of this paper. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. In section 3.2.1, a method for utilizing a surrogate model derived from Chen et al. [28] was described. The method makes it possible to estimate the energy savings attributed to expanding applicable natural ventilation hours in a radiant-assisted natural ventilation regime. The results of this analysis, as applied to the 60 cities listed in table 1, is shown in Fig. 8 .