Abstract

In many US cities, indoor exposure to heat continues to be the underlying cause of a considerable fraction (up to 80% during extreme events) of heat-related mortality and morbidity, even in locations where most citizens have air conditioning (AC). Nevertheless, the existing literature on indoor exposure to heat often regards AC as a binary variable and assumes that its presence inevitably results in a safe thermal environment. This is also reflected in heat vulnerability assessments that assign a binary attribute to AC. In this study, we used thermal simulation of buildings to investigate overheating in residential buildings in three US cities (Houston, Phoenix, and Los Angeles) and focused on scenarios where an AC system is present; yet not fully functional. Moreover, we identified the role of key building characteristics and investigated the sensitivity of indoor environment to the ambient temperature. Our results show that energy poverty and/or faulty systems can expose a considerable fraction of AC-owning elderly in Phoenix and Houston to excess heat for more than 50% of summer. This highlights the need to reevaluate AC as the primary protective factor against heat and introduces several implications that need to be considered in heat vulnerability assessments.

1 Introduction

Acute and prolonged exposure to heat can cause a variety of health problems, ranging from fatal heat strokes to long-term cardiovascular implications and reduced sleep quality [14]. While several factors such as access to healthcare have resulted in an overall decline in the vulnerability of human population to heat [5], it continues to be one of the most important weather-related killers around the world. In the United States, where relatively reliable estimates are available, heat was identified as the underlying cause of death (61%) or a contributing factor (39%) in 3332 deaths between 2006 and 2010 [6]. Moreover, estimates suggest that heat-related hospitalizations are far more common than mortality [7,8] . Less severe (yet more common) health impacts of heat include heat exhaustion, syncope, cramps [9], reduced sleep quality [3], and reduced cognitive performance [10], as well as exacerbation of respiratory [11] and renal disease [12]). Prolonged exposure strains the cardiovascular system in elderly [13] and increases the risk of early mortality [14]. In addition, it has been identified as a contributing factor to congenital heart defects [15]. Notably, a significant portion of heat-related mortality and morbidity are associated with heat exposure inside a residence [16,17], with the majority of victims aged 65 years and older [1719]. This is, in part, explained by the significant portion of the day that population in developed countries (and specifically, vulnerable groups such as the elderly) spend indoors [20].

Underlying mechanisms of indoor exposure to health-implicating levels of heat are often categorized based on climate and air conditioning (AC) prevalence [21,22]. In colder climates with low AC prevalence, overheating occurs due to infrequent heat waves that render indoor thermal conditions unhealthy. In hot climates with extreme summers and high AC prevalence, thermal conditions inside homes can ideally be decoupled from the outdoor weather through mechanical cooling [23]. Therefore, in locations such as south and southwest United States—where AC presence in newly constructed residential buildings is approaching 100% [24,25]—indoor exposure is generally due to some form of AC malfunction and is not necessarily dependent on the intensity of heat events [18]. For instance, in Maricopa County, AZ, in all cases of death caused by indoor exposure to heat (61 of the total 154 heat-related deaths in 2017), the victims’ AC systems were dysfunctional [16]. Heat vulnerability surveys provide further evidence on this issue. For example, Hayden et al. [26] reported that while 87% of their respondents in Houston, TX (901 samples), had central air conditioning, 37% of them “felt too hot in their home,” and one-fifth of participants experienced heat-related symptoms. The same authors found very similar results in their survey of 362 households in Phoenix, AZ [27]. The causes of nonoperational or semi-operational AC include electricity cost, system repair cost, and a system with inadequate capacity (e.g., small window unit systems) [26,27]. In addition, even in the absence of these barriers, some elderly might not adequately cool their homes because of the limited ability to perceive heat [28].

Large-scale summertime power outages are the other potential cause of indoor overheating despite the presence of an AC system [21,23,2933]. An example is the large-scale power outage in the summer of 2012 in the US Midwest region that resulted in at least 20 heat-related deaths in a short period [34].

Despite these evidences, studies in the building science literature often consider AC as a binary variable. Scientific papers that look at the impacts of building characteristics on indoor heat exposure mostly focus on European cities with typically moderate summers, where most homes are not equipped with mechanical AC [3544]. In these locations, AC prevalence is low because of the underlying climate, existing building stock characteristics, and to some extent, socioeconomic and cultural factors. For example, in the United Kingdom (the most studied country based on our literature review), less than 1% of homes have air conditioning [45]. In contrast, in the United States, where more than 87% of homes have some type of AC [24], passive survivability of buildings is inadequately studied. We posit that the high prevalence of AC is the main cause of the reduced attention to the role that building characteristics can play as a protective factor. Meanwhile, the relatively low cost and availability of AC systems in the United States has led to construction practices that compensate the lack of climate-adaptive designs with mechanical cooling. The prevalence of lightweight wood frame construction, limited utilization of natural ventilation, energy codes that do not regulate passive survivability, and dominance of single-family detached buildings with no shared walls [24,25] cause the passive performance of an average residential building in the United States to be inferior to its European counterparts.

This disparity is also noticeable at the policy level. For example, while the European Environmental Agency and European Commission do not mention AC as an adaptive measure against indoor heat in their public guidelines, US government agencies such as CDC and EPA recommend “staying in an air-conditioned place” as the primary protective measure against heat. The other policy implication is the lack of building characteristics (excluding AC) in heat vulnerability assessments [46]. In part, this is due to the assumption that socioeconomic indicators of vulnerability that are often included in heat vulnerability assessments (in particular, income or poverty level) capture the aforementioned barriers to maintain a fully functional AC. However, the prevalence of these issues (in particular, a dysfunctional AC) is far beyond what can be explained by income [26,27,47]. This is even more consequential during city-wide power outages, where most people lose access to air conditioning.

In this study, we investigate overheating inside residential buildings under different AC functionality scenarios. Specifically, we shall (1) verify whether AC availability in a building or group of buildings guarantees safe and comfortable indoor thermal conditions and 2) identify the key factors underlying the indoor heat exposure. We have used a detailed building energy simulation software to assess indoor thermal conditions with (1) fully functional AC, (2) without any type of AC, (3) with inadequate cooling (see Sec. 2.3 for definition), and (4) under power outages. We selected Phoenix, AZ; Houston, TX; and Los Angeles, CA, as case studies because of their distinct climates and population demographics. We supplemented our simulations with a heat vulnerability survey that helped us estimate the fraction of population associated with each AC functionality scenario. Finally, we divided our simulations into subcategories to investigate the impacts from urban-induced warming and climate change, energy regulations, and certain passive strategies.

2 Methodology

We used whole-building simulations of archetypical (single family detached) residential buildings in Phoenix, Houston, and Los Angeles to estimate indoor heat exposure. Our simulation set included 27,648 permutations of the archetype in each city. In addition, based on our previously defined framework [48], we used a heat vulnerability survey to estimate the fraction of elderly population (>65 years old) associated with each AC functionality scenario. The reason for focusing on the elderly was their higher vulnerability to heat demonstrated by their disproportionate share of total heat-related mortalities in the past events [6,49].

2.1 Selected Locations and the Weather Data.

Phoenix has a hot/dry climate (average summer high = 40.3 °C). Therefore, nearly 90% of homes have central AC systems [25,50]. Houston has hot and humid summers (average summer high = 33.4 °C, average summer RH = 74.8%), with the same central AC prevalence as Phoenix [25]. In Los Angeles, because of the moderate climate, only half of the housing stock have central AC [25]. In addition to distinct climates and differences in AC prevalence, all three cities have a large population, with groups that are extremely vulnerable to heat [26,27,51], which make them interesting cities for our case study.

For the weather data, we used Typical Meteorological Year (TMY3) (see the study by Hall et al. [52] for detailed description). While TMY is based on historic records and does not represent the future, it is a well-known and widely used representative climate for each city that can be relied upon with very few assumptions. In contrast, including future weather scenarios adds complexity and requires many assumptions (e.g., regarding representative concentration pathway (RCP) or the global circulation model (GCM)) that are beyond the scope of our work. The main question we are trying to answer is whether mechanical AC is an effective protective factor against indoor exposure to heat among the elderly. Therefore, running simulations using TMY3 is a fit-for-purpose analysis that provides a baseline for comparison among cities, building types, and AC functionality scenarios without adding complexity about many different scenarios for future weather. While this fits the scope of our analysis, we acknowledge that due to this limitation, the results presented here are inherently conservative. Table 1 presents the basic climate data of the three cities obtained from the TMY3 files.

2.2 Building Simulations.

Whole-building energy simulations apply physics-based models that account for all modes of heat and mass transfer to and from indoor spaces in response to outdoor weather, while accounting for internal loads and occupant behavior. Developed and validated by US Department of Energy [53], EnergyPlus is a state-of-the-art simulation engine that can output all physical variables needed for heat exposure analysis and is frequently used for this purpose [23,29,41,43,54,55]. The global energy balance of EnergyPlus has been validated in compliance with American Society of Heating Refrigeration and Air Conditioning Engineers (ASHRAE) standard 140. In addition, its ability to predict indoor temperature has been validated in a number of independent studies [22,56,57]. Based on our survey (see Sec. 2.5 for the description), the majority of residential units occupied by elderly (73% in Los Angeles, 76% in Houston, and 83% in Phoenix) in the three cities are detached single-family units with frame walls. Therefore, we limited our analysis to this building type and simulated numerous permutations (see the parameter tree in Fig. 1) of the archetype building to account for variations in the input. Each simulation consisted of a unique combination of the variables listed in this figure.

2.3 Modeling AC Nonfunctionality.

According to the findings from previous studies [26,27], AC functionality can be categorized based on these common scenarios:

  1. Fully functional AC

  2. Fully functional AC with high thermostat set point (TS) (e.g., due to electricity cost or decreased thermal sensation)

  3. Faulty AC system (e.g., due to refrigerant leakage) or system with inadequate cooling capacity (small window units)

  4. A combination of 2 and 3

  5. No AC (i.e., an AC system is not present)

  6. Temporary loss of AC due to power outage

Here, we put scenarios 1, 2, and 3 under the same category and refer to it as “inadequate cooling.” Since there is no direct way to model a faulty or undersized AC system in EnergyPlus, we used the input parameter “SizingFactor” as a proxy. The default sizing factor (SF) in EnergyPlus is 1, meaning that the system has enough cooling capacity to provide thermal comfort under all days in the TMY files. Accordingly, we replicated faulty/undersized AC systems by reducing the SF from 1 in 0.25 increments. We reviewed the existing literature on common AC faults in typical residential AC systems. Based on findings from Refs. [47,5868], common faults such as refrigeration leakage and air duct leakage can reduce cooling capacity by 50%. Therefore, without proper maintenance, an SF of 0.75–0.5 is a reasonable assumption for a system after several years of operation. Without reliable data on the prevalence of insufficient cooling capacity (i.e., homes that have a small window instead of central systems), we cannot put it in a distinct category. However, from a modeling perspective, this scenario is relatively similar to a dysfunctional central AC system and is captured by the same SF parameter.

We used a thermostat set point of 24 °C for the fully functional AC category based on Ref. [24]. To mimic energy poverty, we also modeled set points of 27 °C and 30 °C. The upper bound (30 °C) is based on results from Ref. [69] and the data from the study by Hondula [70].

Finally, we simulated a 24-h power outage scenario. The duration (24 h) was selected based on the historic data and is equal to the 75th percentile of outage duration in 1137 incidents that impacted more than half a million people in the past 17 years [71]. We selected 5:00 p.m. local time as the initiation time according to a sensitivity analysis done by Baniassadi et al. [72] who identified 24-h power outages initiated at this time as the “most intense” scenario. Second, this hour falls in the middle of the summertime peak demand during which there is a higher probability for power outages caused by an overloaded infrastructure. This was verified by analyzing the historic outage data from Ref. [71].

2.4 Overheating Metric and Threshold.

In this study, we used Wet Bulb Globe Temperature (WBGT) as the heat metric. It accounts for relative humidity, dry-bulb temperature, air velocity, and the long-wave radiation from adjacent surfaces. In addition, it is a widely recognized metric for studying indoor overheating [73].

The literature on reliable thresholds for health-implicating heat in residential buildings is very limited. This is in part due to the paucity of research on associations between indoor exposure to heat and different health outcomes. Kenny et al. [49] who conducted a literature review on this subject state that “to date there remains a paucity of research directed at defining “high risk” ambient conditions for the most vulnerable.” In another study, Anderson et al. [74] reviewed 96 peer-reviewed articles and concluded that “the data are sparse and inconclusive in terms of identifying evidence-based definitions for thresholds”. As a workaround, many researchers rely on thresholds developed for industrial settings [73] or available thermal comfort guidelines developed for designers and building engineers. The problem with the first approach is that most industrial (or army) thresholds were developed for young and healthy individuals and do not reflect the physiological response of more vulnerable groups. While the second approach is more applicable to people in residential environments, it overlooks the significant distinction between thermal comfort and health-implicating heat [73]. Moreover, in addition to excluding the impact of age, many of these thresholds overlook the gender differences in thermal perception and response to heat [75,76]. Despite all these limitations, having a threshold helps us interpret and communicate our results more clearly. Therefore, after reviewing the limited literature on this subject, we selected a WBGT of 23 °C as an overheating threshold for older adults. We note that the thresholds we observed varied between 21 °C (according to WHO guidelines at typical relative humidity values) and 28 °C (based on Ref. [73]). For reference, in a typical residential setting, a WBGT of 23 °C corresponds to dry-bulb temperatures of 29.1, 27.8, and 26.7 °C at RH levels of 40%, 50%, and 60%, respectively. We also acknowledge that we are using a general threshold for all elderly in our select cities, while factors such as sex, acclimation, and existence of chronic disease (cardiovascular problems or type 2 diabetes) can significantly affect individuals’ response to heat [49].

The current version of EnergyPlus does not directly output WBGT. However, it outputs the physical variables (dry-bulb indoor air temperature, relative humidity, and mean radiant temperature) that are required to calculate WBGT. Accordingly, we generated a matlab script that extracts raw data from EnergyPlus output files and performs the necessary postprocessing to calculate WBGT using the formulation described in Supplemental Material on the ASME Digital Collection (S2).

2.5 Heat Vulnerability Survey.

To arrive at an estimate of the fraction of population exposed to each of the AC functionality categories listed in Sec. 2.3, we relied on data from a previously conducted heat vulnerability survey that targeted residents aged 65 years and older in the three cities. The survey was conducted over the phone and had 306, 303, and 300 respondents in Los Angeles, Phoenix, and Houston, respectively. With overall sampling error of 5.6% in all cities (95% confidence level), the survey represents an estimated elderly (>65 years old) population of 437,224, 142,548, and 196,359, in Los Angeles, Phoenix, and Houston, respectively.

Observations from respondents who lived in multifamily units as well as those who reported leaving their home if they felt too hot were excluded. From the remaining samples, we matched the group who “felt too hot inside their homes” and/or had heat-related symptoms despite having mechanical (central or window units) or evaporative cooling systems to the “Inadequate Cooling” scenario (see Secs. 2 and 3 for definition). In addition, the respondents who did not “feel too hot” or experience heat-related symptoms and had mechanical cooling in their homes were assumed to have a fully functional AC. Finally, we assumed that the fraction of the population implicated by a power outage is equal to the fraction who remain in their home even when they “felt too hot” and/or experience symptoms of heat. The resulting estimated fractions associated with each AC functionality scenario are shown in Fig. 2. We acknowledge that our approach relies on subjective self-reports. However, until the empirical data are available from a large set of homes, these surveys are the most reliable sources for estimating these fractions.

3 Results and Discussion

By postprocessing the output from EnergyPlus simulations, we were able to assess different AC functionality scenarios and building characteristics, as well as the sensitivity to the ambient temperature.

3.1 Indoor Heat and Air Conditioning Functionality.

Figure 3 shows the distribution of indoor WBGT without AC (no AC) and with fully functional AC (operating at 100% capacity with a thermostat set point of 24 °C) during the summer months (June–August). The vertical axis shows the cumulative fraction of hours at or below the associated WBGT. The black curve is the average across all permutations, while the dashed lines show the 10th (light) and 90th (dark) percentiles. Therefore, the larger the difference between the dashed lines, the larger the difference between indoor thermal conditions in the best and worst permutations.

Our simulations show that a fully functional AC system with sufficient cooling capacity operating at a proper thermostat set point keeps indoor WBGT at comfortable levels during the whole summer. In addition, in this scenario, the variation in indoor WBGT across different permutations is small, meaning that building characteristics have a trivial impact. In contrast, without an AC, indoor WBGT is sensitive to climate and building characteristics. In Phoenix and Houston, indoor WBGT remains above the threshold for almost the entire summer. Moreover, building characteristics as well as active use of windows play a significant role under this scenario (note the difference between 10th and 90th percentile lines). Expectedly, only a small number of people do not have air conditioning in Phoenix and Houston (around 1% in both cities). In Los Angeles, while a significantly larger fraction (23%) of the respondents do not have any type of AC, on average, indoor WBGT is considerably lower than the other two cities.

Figure S1 in the Supplemental Material shows the same distribution under the inadequate cooling scenario, categorized by different combinations of SF and TS. Figure 4 is a simplified demonstration of the data and shows the effect of TS and SF on the average duration and intensity of indoor overheating across all permutations. These data show the sensitivity of indoor WBGT to AC operation variables (TS and SF), as well as the role of the underlying climate (note the difference in scales across the panels). In comparison with the “No AC” category, a significantly larger fraction of the population is exposed to inadequate cooling (17% in Houston, 18% in Phoenix, and 25% in Los Angeles). The role of the outdoor humidity level under this category is also important. In the dry climate of Phoenix, an AC system working with 50% of the initial cooling capacity can maintain the indoor WBGT at or below the threshold for more than 80% of the time with 27 °C thermostat set point. Under the same scenario, in Houston (which is more humid), average indoor WBGT remains above the threshold for more than 80% of the time.

3.2 Sensitivity to Outdoor Temperature.

Almost all projections regarding future climates of these cities point to an increase in frequency and intensity of heat waves as a result of climate change [77,78]. Therefore, the sensitivity of indoor thermal conditions to outdoor weather should be among the indicators of vulnerability to climate change. Figure 5 shows this sensitivity by comparing the indoor WBGT between 2 days, 50th and 90th percentiles of the outdoor dry-bulb3 temperature. In this figure, each box shows a pool of hourly indoor WBGT data across all permutations associated with a particular AC functionality scenario. While these results only show indoor WBGT during the 24-h cycle of interest, because we ran the models for the entire summer, and the impact from previous days (due to the thermal inertia) is also accounted for.

Expectedly, the sensitivity of indoor thermal conditions on the ambient temperature (i.e., the intensity of heat event) depends on AC functionality. In all three cities, a fully functional AC leads to comfortable conditions when ambient conditions are at the 50th and 90th percentile of outdoor dry-bulb temperature. In contrast, without AC, average daily indoor WBGT at the hotter day (90th percentile of daily mean ambient temperature) is 3–5 °C higher than that of the median day (50th percentile) in Phoenix and Los Angeles. These percentiles shift as cities become warmer. For example, in Phoenix, the current 90th percentile of daily mean temperature might be the 70th percentile in a typical mid-century year [79]. Therefore, the impact of climate change will be more pronounced on those who are not able to maintain a fully functional AC.

We also noted the variation among different AC functionality scenarios at the same day and city. In Phoenix and Houston, the difference between indoor WBGT with “fully functional AC” and “inadequate AC” is large enough to cause statistically significant health impacts based on exposure-health outcome associations reported by Kim et al. [80] and Laurent et al. [10]. The differences are less considerable in Los Angeles, meaning that AC dysfunction and energy poverty are more important in Phoenix and Houston. It is widely recognized that to some extent, both of these barriers are correlated with low socioeconomic status. Therefore, our physics-based simulations verify that climate change and/or urban-induced warming will be more consequential for indoor exposure among the most vulnerable groups in these cities.

Finally, as the data suggest, during a power outage, indoor heat exposure will be almost the same as buildings without AC. On the other hand, a conservative estimate (based on the survey results) of the fraction of elderly population that will stay at home during a city-wide outage is approximately 60% in all cities. Hence, despite their low probability, power outages coincident with heat waves pose a great threat to the public based on both the exposure intensity and the number of affected individuals.

3.3 Impacts From Building Characteristics.

As the next step, we compared the indoor WBGT data from different sets of permutations to assess how different building characteristics impact indoor heat. In particular, the impacts of building energy regulations are of interest, because they determine how the building stock is evolving.

The first comparison is between buildings compliant with 2003 (B) and 2018 (NC) versions of the IECC code. Moreover, to identify the mitigation potential of strategies that are not currently addressed by codes, we added a subset labeled “Passive Strategies” (PS), which included shade on exterior walls, a high albedo rooftop, and enhanced natural ventilation through operable windows. Figure 6 shows the parametric trees of the compared permutation sets.

In Fig. 7, we compared the indoor WBGT of these permutations sets. The format of this figure is similar to Fig. 5.

First, this comparison shows that building characteristics do not impact indoor WBGT when AC is fully functional and has sufficient capacity.4 As mentioned earlier, we posit that this is the key reason behind neglecting the roles that building characteristics (excluding AC) can play in mitigating indoor heat exposure in Sun Belt cities of the United States. Because most of the population of these cities fall in this category, less attention has been given to the effects that envelope properties, shade on exterior walls, thermal mass, ventilation and other building features have on indoor exposure to heat. This is reflected in the scientific literature on both building science and social vulnerability fronts, as well as policies (e.g., development of heat vulnerability maps or building energy regulations).

Without a fully functional AC, indoor WBGT can be highly sensitive to building characteristics (rows 3 and 4 in Phoenix and Houston, and rows 2–4 in Los Angeles). This highlights the problem with the aforementioned inattention to the role of buildings. For example, while the permutation set in Los Angeles with inadequate cooling shows a clear difference between indoor exposures to heat among different subsets, they might all be considered the same in heat vulnerability indices that only include AC ownership as a proxy for indoor exposure.

Another policy-related observation is that at both days, overheating is more severe in samples that are compliant to the newer energy codes. This was previously observed in energy codes in Australia and United Kingdom and is in-line with findings in Refs. [38,40,44,81]. The problem is caused by stringent insulation and air tightness requirements, which may impede the ability of buildings to lose heat during the night in the absence of proper ventilation. We also tested this hypothesis using our simulation results by comparing the indoor WBGT between the baseline and new code sets during day (10:00 a.m.–6:00 p.m.) and night (10:00 p.m.–6:00 a.m.). We noted that in almost all cases, the negative impact of the newer code was larger at night. This can also be observed in Fig. 7 (especially in Los Angeles). Under the “No AC” and power outage scenarios, the interquartile range of indoor WBGT is smaller in the new code set, because the 25th percentile increases more than the 75th percentile. Put simply, this negative impact from building energy codes is more pronounced during night and at lower percentiles of indoor WBGT. On the other hand, there is evidence in the literature suggesting that minimum nighttime temperature is a better predictor of health outcomes than daily maximum or daily average [8284] . This is in part due to the fact that most people spend the night at home. Moreover, nighttime heat impedes body recovery from daytime exposure to heat. Despite these findings, it is neither possible nor beneficial to generalize this impact because these negative consequences are only observed under specific conditions [21,22,29,85]. For example, the findings of the study by Baniassadi et al. [21] suggest conflicting impacts between codes for high-rise residential buildings in Phoenix and Albuquerque. Regardless of the discrepancies in findings of different studies that are due to the differences in contexts and underlying climate, the existing evidence converges on overheating as an unintended consequence of more stringent codes under some scenarios and should be enough to drive policy change in code developments. One possibility is to include thermal resiliency in building energy codes. For example, compliancy might require a building to maintain habitable conditions for a certain number of hours in response to a summertime power outage.

Our results also show that a combination of relatively simple interventions (change of rooftop albedo, benefiting from ventilation through windows, and shade on exterior walls) can substantially mitigate indoor heat exposure. Notably, none of these strategies are currently governed by codes in single-family residential units. However, when added to the codes (the “NC + PS” subset in Fig. 6), they can compensate for their negative unintended consequences. Nevertheless, in extreme cases (power loss or lack of AC in Houston and Phoenix), buildings remain overheated even with these interventions. This highlights a limit in passive survivability of typical housing (excluding uncommon high-end buildings) in these two cities, meaning that interventions on the building side might not be enough to safeguard the citizens against heat in the absence of AC.

4 Limitations and Suggested Areas for Future Research

First, because of the limited literature on the epidemiology of indoor exposure to heat, our 23 °C WBGT threshold should not be looked at as a deterministic value. Second, variations in ambient temperature across different parts of a city can be significant. While inclusion of such variations is outside the scope of this paper, it is a worthwhile topic for future investigations. Moreover, while our residential building archetype (single-family detached house) was the most common housing type in the three studied cities, it may not be the most common housing type of the vulnerable population in different US cities. For example, in cities such as Philadelphia, Chicago, and Boston, the most common housing type for many vulnerable groups (based on demographic factors) is multifamily residential building [25]. Therefore, there are limitations in extrapolating our findings to a national level.

On the basis of abovementioned limitations, our general recommendation for future research is to generate data that can serve as more reliable input for similar assessments. This includes research on the epidemiology front to provide more reliable thresholds for indoor heat, generating future climate data for building energy simulations with an emphasis on heat events at fine spatial resolution and household surveys and measurement campaigns that result in reliable estimates of AC functionality parameters.

5 Conclusion and Policy Recommendations

In this study, we investigated overheating inside residential buildings in three US cities under different AC availability scenarios. We also matched the simulation outputs to fractions of population estimated based a heat vulnerability survey.

First, we noted that while there is a high potential for indoor overheating in Phoenix and Houston in the absence of AC, only a small fraction of the population did not have any type of AC. Therefore, the indoor exposure to heat in the absence of AC is the most intense, yet the least common type of exposure in these cities. While the indoor WBGT was generally lower in Los Angeles, significantly larger portion of the population (23%) was exposed. Second, our results show that with inadequate cooling (e.g., due to energy poverty or failure to upkeep a system), residents can potentially be exposed to health-implicating levels of indoor heat. More importantly, we estimated that a considerable fraction of elderly in our sample cities fall in this category. We identified large-scale power outages coincident with periods of hot weather as the most consequential scenario (especially in Phoenix and Houston), because they expose a large fraction of the elderly population (at least 60% in all cities) to intense levels of indoor heat. In addition, we investigated the sensitivity of indoor conditions to the ambient temperature. On average, we observed 0.6–0.8 increase in indoor WBGT per 1 °C increase in outdoor air temperature in buildings with dysfunctional AC. This highlights the vulnerability of people with the limited resource to impacts of climate change and urban-induced warming. Finally, we noted that passive survivability of typical residential buildings in the studied cities might be negatively impacted by more stringent energy codes. Low-cost strategies such as reflective rooftops (currently not included in codes) can mitigate indoor overheating in Los Angeles. However, in Phoenix and Houston, typical interventions on the building-side might not be enough to completely protect elderly from indoor heat in the absence of a fully functional AC system.

Heat is a multifaceted problem with complex exposure pathways, and our research implicates housing as a critical determinant of heat vulnerability and health. On the basis of our findings, and given that climate change is not a “future problem” anymore, we believe that short-term policies to mitigate citizens’ indoor exposure to heat are imperative and propose the following to policymakers and stakeholders.

  1. Heat vulnerability indices used by governments to identify the most vulnerable populations in a city must include housing characteristics in addition to well-known demographic indicators of vulnerability.

  2. There should be policies in place to improve the passive survivability of the housing stock. For new constructions, this can be achieved by including mandates in energy codes. For existing buildings, a short-term solution is to promote climate-specific interventions through incentives and public government programs (similar to many existing energy efficiency programs).

  3. There should be an emphasize on “passive” performance. Many cities expend considerable resources on utility assistant programs for low-income families. While we acknowledge the importance of such programs, we believe that they promote active cooling, which is vulnerable to infrastructure/technical failures. In contrast, helping citizens to improve passive performance of their homes (e.g., installing white reflective or shading the exterior of the building) results in a more resilient housing stock.

  4. Focusing on public housing is one of the most cost-effective strategies because they accommodate groups that are more vulnerable to heat. Moreover, compared with other types of construction projects, they are more feasible targets for top-down interventions.

  5. Cities need to develop response plans for large-scale power outages during heat waves. It is imperative that these plans include specific measures to address indoor exposure (e.g., providing care for people with mobility limitations).

  6. Utility companies need to be involved in the process on two fronts. There should be a mechanism in place to make sure that failure to pay the electricity bill does not result in a loss of power for an individual during hot weather. Moreover, during unavoidable summertime brownouts, given the importance of indoor exposure, neighborhoods with vulnerable populations and substandard housing should be prioritized.

Footnotes

3

We compared the percentiles with those based on outdoor Heat Index (a metric that also includes relative humidity) and noted that the ranking of days around the median and 90th percentile is very similar across the two sets.

4

However, they have a significant impact on energy demand of the AC system, which should not be overlooked.

Acknowledgment

This research was supported in part by Assistance Agreement No. 83575402 awarded by the U.S. Environmental Protection Agency. It has not been formally reviewed by the EPA. The views expressed in this document are solely those of the authors and do not necessarily reflect those of the Agency. EPA does not endorse any products or commercial services mentioned in this publication. Parts of the data presented here resulted from the PhD thesis work of the first author [86].

References

References
1.
Basu
,
R.
, and
Samet
,
J. M.
,
2002
, “
Relation Between Elevated Ambient Temperature and Mortality: A Review of the Epidemiologic Evidence
,”
Epidemiol. Rev.
,
24
(
2
), pp.
190
202
. 10.1093/epirev/mxf007
2.
Basu
,
R.
,
Pearson
,
D.
,
Malig
,
B.
,
Broadwin
,
R.
, and
Green
,
R.
,
2012
, “
The Effect of High Ambient Temperature on Emergency Room Visits
,”
Epidemiology
,
23
(
6
), pp.
813
820
. 10.1097/EDE.0b013e31826b7f97
3.
Obradovich
,
N.
,
Migliorini
,
R.
,
Mednick
,
S. C.
, and
Fowler
,
J. H.
,
2017
, “
Nighttime Temperature and Human Sleep Loss in a Changing Climate
,”
Sci. Adv.
,
3
(
5
), p.
e1601555
. 10.1126/sciadv.1601555
4.
Winquist
,
A.
,
Grundstein
,
A.
,
Chang
,
H. H.
,
Hess
,
J.
, and
Sarnat
,
S. E.
,
2016
, “
Warm Season Temperatures and Emergency Department Visits in Atlanta, Georgia
,”
Environ. Res.
,
147
, pp.
314
323
. 10.1016/j.envres.2016.02.022
5.
Sheridan
,
S. C.
, and
Allen
,
M. J.
,
2018
, “
Temporal Trends in Human Vulnerability to Excessive Heat
,”
Environ. Res. Lett.
,
13
(
4
), p.
043001
. 10.1088/1748-9326/aab214
6.
Berko
,
J.
,
Ingram
,
D. D.
,
Saha
,
S.
, and
Parker
,
J. D.
,
2014
, “
Deaths Attributed to Heat, Cold, and Other Weather Events in the United States, 2006–2010
,”
Centers for Disease Control and Prevention National Center for Health Statistics
,
Hyattsville, MD
.
7.
Vaidyanathan
,
A.
,
Saha
,
S.
,
Vicedo-Cabrera
,
A. M.
,
Gasparrini
,
A.
,
Abdurehman
,
N.
,
Jordan
,
R.
,
Hawkins
,
M.
,
Hess
,
J.
, and
Elixhauser
,
A.
,
2019
, “
Assessment of Extreme Heat and Hospitalizations to Inform Early Warning Systems
,”
Proc. Natl. Acad. Sci. USA
,
116
(
12
), pp.
5420
5427
. 10.1073/pnas.1806393116
8.
CDC
,
2017
, “
Picture of America Heat-Related Illness Fact Sheet. Centers for Diseases Control and Prevention
,” https://www.cdc.gov/pictureofamerica/index.html, Accessed October 2019.
9.
Kilbourne
,
K. M.
,
1997
,
Heat Waves and Hot Environments, The Public Health Consequences of Disasters
,
Oxford University Press
,
New York
, pp.
245
269
.
10.
Laurent
,
J. G. C.
,
Williams
,
A.
,
Oulhote
,
Y.
,
Zanobetti
,
A.
,
Allen
,
J. G.
, and
Spengler
,
J. D.
,
2018
, “
Reduced Cognitive Function During a Heat Wave Among Residents of Non-Air-Conditioned Buildings: An Observational Study of Young Adults in the Summer of 2016
,”
PLoS Medicine
,
15
(
7
), p.
e1002605
. 10.1371/journal.pmed.1002605
11.
Strauss
,
R. H.
,
McFadden
,
E.
,
Ingram
,
R.
,
Deal
,
E.
, and
Jaeger
,
J.
,
1978
, “
Influence of Heat and Humidity on the AirwayObstruction Induced by Exercise in Asthma
,”
J. Clin. Invest.
,
61
, pp.
433
440
.
12.
Remigio
,
R. V.
,
Jiang
,
C.
,
Raimann
,
J.
,
Kotanko
,
P.
,
Usvyat
,
L.
,
Maddux
,
F. W.
,
Kinney
,
P.
, and
Sapkota
,
A.
,
2019
, “
Association of Extreme Heat Events With Hospital Admission or Mortality Among Patients With End-Stage Renal Disease
,”
JAMA Network Open
,
2
(
8
), pp.
e198904
e198904
.
13.
Kenney
,
W. L.
,
Craighead
,
D. H.
, and
Alexander
,
L. M.
,
2014
, “
Heat Waves, Aging, and Human Cardiovascular Health
,”
Med. Sci. Sports Exer.
,
46
, p.
1891
.
14.
Wallace
,
R. F.
,
Kriebel
,
D.
,
Punnett
,
L.
,
Wegman
,
D. H.
, and
Amoroso
,
P. J.
,
2007
, “
Prior Heat Illness Hospitalization and Risk of Early Death
,”
Envir. Res.
,
104
, pp.
290
295
.
15.
Zhang
,
W.
,
Spero
,
T. L.
,
Nolte
,
C. G.
,
Garcia
,
V. C.
,
Lin
,
Z.
,
Romitti
,
P. A.
,
Shaw
,
G. M.
,
Sheridan
,
S. C.
,
Feldkamp
,
M. L.
,
Woomert
,
A.
, and
Hwang
,
S. A.
,
2019
, “
Projected Changes in Maternal Heat Exposure During Early Pregnancy and the Associated Congenital Heart Defect Burden in the United States
,”
J. Amer. Heart Assoc.
,
8
(
3
), p.
e010995
.
16.
MCDPH
,
2019
,
Heat-associated deaths in Maricopa County, AZ: Final report for 2018. Maricopa County Department of Public Health
, https://www.maricopa.gov/1858/Heat-Surveillance, Accessed October 2019.
17.
NWS
,
2016
, “
Weather Fatalities 2017
,” https://www.weather.gov/hazstat/, Accessed January 2018.
18.
Putnam
,
H.
,
Hondula
,
D. M.
,
Urban
,
A.
,
Berisha
,
V.
,
Iñiguez
,
P.
, and
Roach
,
M.
,
2018
, “
It’s Not the Heat, It’s the Vulnerability: Attribution of the 2016 Spike in Heat-Associated Deaths in Maricopa County, Arizona
,”
Environ. Res. Lett.
,
13
(
9
), p.
094022
. 10.1088/1748-9326/aadb44
19.
Åström
,
D. O.
,
Bertil
,
F.
, and
Joacim
,
R.
,
2011
, “
Heat Wave Impact on Morbidity and Mortality in the Elderly Population: A Review of Recent Studies
,”
Maturitas
,
69
(
2
), pp.
99
105
. 10.1016/j.maturitas.2011.03.008
20.
Klepeis
,
N. E.
,
Nelson
,
W. C.
,
Ott
,
W. R.
,
Robinson
,
J. P.
,
Tsang
,
A. M.
, and
Switzer
,
P.
,
2001
, “
The National Human Activity Pattern Survey (Nhaps): A Resource for Assessing Exposure to Environmental Pollutants
,”
J. Exposure Anal. Environ. Epidemiol.
,
11
(
3
), pp.
231
252
. 10.1038/sj.jea.7500165
21.
Baniassadi
,
A.
,
Heusinger
,
J.
, and
Sailor
,
D. J.
,
2018
, “
Energy Efficiency vs Resiliency to Extreme Heat and Power Outages: The Role of Evolving Building Energy Codes
,”
Build. Environ.
,
139
, pp.
86
94
. 10.1016/j.buildenv.2018.05.024
22.
Sailor
,
D. J.
,
Baniassadi
,
A.
,
O'Lenick
,
C. R.
, and
Wilhelmi
,
O. V.
,
2019
, “
The Growing Threat of Heat Disasters
,”
Environ. Res. Lett.
,
14
(
5
), p.
054006
. 10.1088/1748-9326/ab0bb9
23.
Sailor
,
D. J.
,
2014
, “
Risks of Summertime Extreme Thermal Conditions in Buildings as a Result of Climate Change and Exacerbation of Urban Heat Islands
,”
Build. Environ.
,
78
, pp.
81
88
. 10.1016/j.buildenv.2014.04.012
24.
EIA
,
2015
, “
Residential Energy Consumption Survey
,”
U.S Department of Energy
.
25.
USCB
,
2017
, “
American Housing Survey
,”
United States Census Bureau
.
26.
Hayden
,
M. H.
,
Wilhelmi
,
O. V.
,
Banerjee
,
D.
,
Greasby
,
T.
,
Cavanaugh
,
J. L.
, and
Nepal
,
V.
,
2017
, “
Adaptive Capacity to Extreme Heat: Results From a Household Survey in Houston, Texas
,”
Weather Clim. Soc.
,
9
(
4
), pp.
787
799
. 10.1175/WCAS-D-16-0125.1
27.
Hayden
,
M. H.
,
Brenkert-Smith
,
H.
, and
Wilhelmi
,
O. V.
,
2011
, “
Differential Adaptive Capacity to Extreme Heat: A Phoenix, Arizona, Case Study
,”
Weather Clim. Soc.
,
3
(
4
), pp.
269
280
. 10.1175/WCAS-D-11-00010.1
28.
Dufour
,
A.
, and
Candas
,
V.
,
2007
, “
Ageing and Thermal Responses During Passive Heat Exposure: Sweating and Sensory Aspects
,”
Eur. J. Appl. Physiol.
,
100
(
1
), pp.
19
26
. 10.1007/s00421-007-0396-9
29.
Baniassadi
,
A.
, and
Sailor
,
D. J.
,
2018
, “
Synergies and Trade-Offs Between Energy Efficiency and Resiliency to Extreme Heat—A Case Study
,”
Build. Environ.
,
132
, pp.
263
272
. 10.1016/j.buildenv.2018.01.037
30.
O'Brien
,
W.
, and
Bennet
,
I.
,
2016
, “
Simulation-Based Evaluation of High-Rise Residential Building Thermal Resilience
,”
ASHRAE Transactions
,
122
(
1
), p.
455
.
31.
Nahlik
,
M. J.
,
Chester
,
M. V.
,
Pincetl
,
S. S.
,
Eisenman
,
D.
,
Sivaraman
,
D.
, and
English
,
P.
,
2016
, “
Building Thermal Performance, Extreme Heat, and Climate Change
,”
J. Infrastruct. Syst.
,
23
(
3
), p.
04016043
.
32.
Samuelson
,
H.
,
Claussnitzer
,
S.
,
Goyal
,
A.
,
Chen
,
Y.
, and
Romo-Castillo
,
A.
,
2016
, “
Parametric Energy Simulation in Early Design: High-Rise Residential Buildings in Urban Contexts
,”
Build. Environ.
,
101
, pp.
19
31
. 10.1016/j.buildenv.2016.02.018
33.
Sailor
,
D. J.
,
Baniassadi
,
A.
,
O'Lenick
,
C. R.
, and
Wilhelmi
,
O. V.
,
2019
, “
The Growing Threat of Heat Disasters
,”
Env. Res. Lett.
,
14
(
5
), p.
054006
.
34.
Fowler
,
D. R.
,
Mitchell
,
C. S.
,
Brown
,
A.
,
Pollock
,
T.
,
Bratka
,
L. A.
, and
Paulson
,
J.
,
2013
, “
Heat-Related Deaths After an Extreme Heat Event—Four States, 2012, and United States, 1999–2009
,”
MMWR Morbidity and Mortality Weekly Report
,
62
, pp.
433
436
.
35.
Dodoo
,
A.
, and
Gustavsson
,
L.
,
2016
, “
Energy Use and Overheating Risk of Swedish Multi-Story Residential Buildings Under Different Climate Scenarios
,”
Energy
,
97
, pp.
534
548
. 10.1016/j.energy.2015.12.086
36.
Lomas
,
K. J.
, and
Kane
,
T.
,
2013
, “
Summertime Temperatures and Thermal Comfort in UK Homes
,”
Build. Res. Inf.
,
41
(
3
), pp.
259
280
. 10.1080/09613218.2013.757886
37.
Mavrogianni
,
A.
,
Wilkinson
,
P.
,
Davies
,
M.
,
Biddulph
,
P.
, and
Oikonomou
,
E.
,
2012
, “
Building Characteristics as Determinants of Propensity to High Indoor Summer Temperatures in London Dwellings
,”
Build. Environ.
,
55
, pp.
117
130
. 10.1016/j.buildenv.2011.12.003
38.
McLeod
,
R. S.
,
Hopfe
,
C. J.
, and
Kwan
,
A.
,
2013
, “
An Investigation Into Future Performance and Overheating Risks in Passivhaus Dwellings
,”
Build. Environ.
,
70
, pp.
189
209
. 10.1016/j.buildenv.2013.08.024
39.
Mlakar
,
J.
, and
Strancar
,
J.
,
2011
, “
Overheating in Residential Passive House: Solution Strategies Revealed and Confirmed Through Data Analysis and Simulations
,”
Energy Build.
,
43
(
6
), pp.
1443
1451
. 10.1016/j.enbuild.2011.02.008
40.
Mulville
,
M.
, and
Stravoravdis
,
S.
,
2016
, “
The Impact of Regulations on Overheating Risk in Dwellings
,”
Build. Res. Inf.
,
44
(
5–6
), pp.
520
534
. 10.1080/09613218.2016.1153355
41.
Oikonomou
,
E.
,
Davies
,
M.
,
Mavrogianni
,
A.
,
Biddulph
,
P.
,
Wilkinson
,
P.
, and
Kolokotroni
,
M.
,
2012
, “
Modelling the Relative Importance of the Urban Heat Island and the Thermal Quality of Dwellings for Overheating in London
,”
Build. Environ.
,
57
, pp.
223
238
. 10.1016/j.buildenv.2012.04.002
42.
Pathan
,
A.
,
Mavrogianni
,
A.
,
Summerfield
,
A.
,
Oreszczyn
,
T.
, and
Davies
,
M.
,
2017
, “
Monitoring Summer Indoor Overheating in the London Housing Stock
,”
Energy Build.
,
141
, pp.
361
378
. 10.1016/j.enbuild.2017.02.049
43.
Porritt
,
S. M.
,
Cropper
,
P. C.
,
Shao
,
L.
, and
Goodier
,
C. I.
,
2012
, “
Ranking of Interventions to Reduce Dwelling Overheating During Heat Waves
,”
Energy Build.
,
55
, pp.
16
27
. 10.1016/j.enbuild.2012.01.043
44.
Taylor
,
J.
,
Symonds
,
P.
,
Wilkinson
,
P.
,
Heaviside
,
C.
,
Macintyre
,
H.
, and
Davies
,
M.
,
2018
, “
Estimating the Influence of Housing Energy Efficiency and Overheating Adaptations on Heat-Related Mortality in the West Midlands, UK
,”
Atmosphere
,
9
(
5
), p.
190
. 10.3390/atmos9050190
45.
Barford
,
V.
,
2013
, “
10 Ways the UK Is Ill-Prepared for a Heatwave
,”
BBC News Magazine.
46.
Bao
,
J.
,
Li
,
X.
, and
Yu
,
C.
,
2015
, “
The Construction and Validation of the Heat Vulnerability Index, a Review
,”
Int. J. Environ. Res. Public Health
,
12
(
7
), pp.
7220
7234
. 10.3390/ijerph120707220
47.
Stephens
,
B.
,
Siegel
,
J. A.
, and
Novoselac
,
A.
,
2011
, “
Operational Characteristics of Residential and Light-Commercial Air-Conditioning Systems in a Hot and Humid Climate Zone
,”
Build. Environ.
,
46
(
10
), pp.
1972
1983
. 10.1016/j.buildenv.2011.04.005
48.
O'Lenick
,
C. R.
,
Wilhelmi
,
O. V.
,
Michael
,
R.
,
Hayden
,
M. H.
,
Baniassadi
,
A.
,
Wiedinmyer
,
C.
,
Monaghan
,
A. J.
,
Crank
,
P. J.
, and
Sailor
,
D. J.
,
2019
, “
Urban Heat and Air Pollution: A Framework for Integrating Population Vulnerability and Indoor Exposure in Health Risk Analyses
,”
Sci. Total Environ.
,
660
, pp.
715
723
. 10.1016/j.scitotenv.2019.01.002
49.
Kenny
,
G. P.
,
Flouris
,
A. D.
,
Yagouti
,
A.
, and
Notley
,
S. R.
,
2019
, “
Towards Establishing Evidence-Based Guidelines on Maximum Indoor Temperatures During Hot Weather in Temperate Continental Climates
,”
Temperature
,
6
(
1
), pp.
11
36
. 10.1080/23328940.2018.1456257
50.
Fraser
,
A. M.
,
Chester
,
M. V.
,
Eisenman
,
D.
,
Hondula
,
D. M.
,
Pincetl
,
S. S.
, and
English
,
P.
,
2016
, “
Household Accessibility to Heat Refuges: Residential Air Conditioning, Public Cooled Space, and Walkability
,”
Environ. Planning B Planning Design
,
44
(
9
), pp.
1036
1055
.
51.
Reid
,
C. E.
,
O’neill
,
M. S.
,
Gronlund
,
C. J.
,
Brines
,
S. J.
,
Brown
,
D. G.
,
Diez-Roux
,
A. V.
, and
Schwartz
,
J.
,
2009
, “
Mapping Community Determinants of Heat Vulnerability
,”
Environ. Health Perspect.
,
117
(
11
), pp.
1730
1736
. 10.1289/ehp.0900683
52.
Hall
,
I. J.
,
Prairie
,
R.
,
Anderson
,
H.
, and
Boes
,
E.
,
1978
, “
Generation of a Typical Meteorological Year
,”
Sandia Labs.
,
Albuquerque, NM
.
53.
Crawley
,
D. B.
,
Lawrie
,
L. K.
,
Winkelmann
,
F. C.
,
Buhl
,
W. F.
,
Huang
,
Y. J.
, and
Pedersen
,
C. O.
,
2001
, “
Energyplus: Creating a New-Generation Building Energy Simulation Program
,”
Energy Build.
,
33
(
4
), pp.
319
331
. 10.1016/S0378-7788(00)00114-6
54.
Alam
,
M.
,
Sanjayan
,
J.
,
Zou
,
P. X.
,
Stewart
,
M. G.
, and
Wilson
,
J.
,
2016
, “
Modelling the Correlation Between Building Energy Ratings and Heat-Related Mortality and Morbidity
,”
Sustainable Cities Soc.
,
22
, pp.
29
39
. 10.1016/j.scs.2016.01.006
55.
Ramakrishnan
,
S.
,
Wang
,
X.
,
Sanjayan
,
J.
, and
Wilson
,
J.
,
2016
, “
Thermal Performance of Buildings Integrated With Phase Change Materials to Reduce Heat Stress Risks During Extreme Heatwave Events
,”
Appl. Energy
,
194
, pp.
410
421
. 10.1016/j.apenergy.2016.04.084
56.
Jamil
,
H.
,
Alam
,
M.
,
Sanjayan
,
J.
, and
Wilson
,
J.
,
2016
, “
Investigation of pcm as Retrofitting Option to Enhance Occupant Thermal Comfort in a Modern Residential Building
,”
Energy Build.
,
133
, pp.
217
229
. 10.1016/j.enbuild.2016.09.064
57.
Zhuang
,
C.
,
Deng
,
A.
,
Chen
,
Y.
,
Li
,
S.
,
Zhang
,
H.
, and
Fan
,
G.
,
2010
, “
Validation of Veracity on Simulating the Indoor Temperature in pcm Light Weight Building by EnergyPlus
,”
In: Life system modeling and intelligent computing: Springer
,
6328
, pp.
486
496
. 10.1007/978-3-642-15621-2_53
58.
Chen
,
B.
, and
Braun
,
J.
,
2000
, “
Simple Fault Detection and Diagnosis Methods for Packaged Air Conditioners
,”
International Refrigeration and Air Conditioning Conference
,
West Lafayette, IN
,
July 10–14
, pp.
321
328
.
59.
Palani
,
M.
,
O'Neal
,
D.
, and
Haberl
,
J.
,
1992
, “
Monitoring the Performance of a Residential Central Air Conditioner Under Degraded Conditions on a Test Bench
,”
Energy Systems Laboratory, Texas A&M University; Department of Mechanical Engineering, Texas A&M University
,
College Station, TX
.
60.
Kim
,
M.
,
Payne
,
W. V.
,
Domanski
,
P. A.
,
Yoon
,
S. H.
, and
Hermes
,
C. J.
,
2009
, “
Performance of a Residential Heat Pump Operating in the Cooling Mode With Single Faults Imposed
,”
Appl. Therm. Eng.
,
29
(
4
), pp.
770
778
. 10.1016/j.applthermaleng.2008.04.009
61.
Mehrabi
,
M.
, and
Yuill
,
D.
,
2017
, “
Generalized Effects of Refrigerant Charge on Normalized Performance Variables of Air Conditioners and Heat Pumps
,”
Int. J. Refrig.
,
76
, pp.
367
384
. 10.1016/j.ijrefrig.2017.02.014
62.
Siegel
,
J. A.
,
2002
, “
An Evaluation of Superheat-Based Refrigerant Charge Diagnostics for Residential Cooling Systems
,”
Annual Meeting, 108. LBNL Report No. LBNL-47476
, https://escholarship.org/uc/item/9322d1fk, Accessed Oct 2019.
63.
Cheung
,
H.
, and
Braun
,
J. E.
,
2017
, “
An Empirical Model for Simulating the Effects of Refrigerant Charge Faults on Air Conditioner Performance
,”
Sci. Technol. Built Environ.
,
23
(
5
), pp.
776
786
. 10.1080/23744731.2016.1260419
64.
Kim
,
W.
, and
Braun
,
J. E.
,
2012
, “
Evaluation of the Impacts of Refrigerant Charge on Air Conditioner and Heat Pump Performance
,”
Int. J. Refrig.
,
35
(
7
), pp.
1805
1814
. 10.1016/j.ijrefrig.2012.06.007
65.
Yoo
,
J. W.
,
Hong
,
S. B.
, and
Kim
,
M. S.
,
2017
, “
Refrigerant Leakage Detection in an EEV Installed Residential Air Conditioner With Limited Sensor Installations
,”
Int. J. Refrig.
,
78
, pp.
157
165
. 10.1016/j.ijrefrig.2017.03.001
66.
Yang
,
L.
,
Braun
,
J. E.
, and
Groll
,
E. A.
,
2007
, “
The Impact of Evaporator Fouling and Filtration on the Performance of Packaged Air Conditioners
,”
Int. J. Refrig.
,
30
(
3
), pp.
506
514
. 10.1016/j.ijrefrig.2006.08.010
67.
Yin
,
P.
, and
Sweeney
,
J. F.
,
2014
, “
The Impact of an ECM Blower on the System Performance of a 5-Ton Air Conditioner
,”
ASHRAE Transactions
,
120
, p.
1U
.
68.
Rodriguez
,
A. G.
,
O'Neal
,
D.
,
Davis
,
M.
, and
Kondepudi
,
S.
,
1996
, “
Effect of Reduced Evaporator Airflow on the High Temperature Performance of Air Conditioners
,”
Energy Build.
,
24
(
3
), pp.
195
201
. 10.1016/S0378-7788(96)00976-0
69.
White-Newsome
,
J. L.
,
Sánchez
,
B. N.
,
Jolliet
,
O.
,
Zhang
,
Z.
,
Parker
,
E. A.
, and
Dvonch
,
J. T.
,
2012
, “
Climate Change and Health: Indoor Heat Exposure in Vulnerable Populations
,”
Environ. Res.
,
112
, pp.
20
27
. 10.1016/j.envres.2011.10.008
70.
Hondula
,
D.
,
2018
, “
Summer Indoor Temperature Measurements in 46 Phoenix Homes
,”
Personal Communication.
71.
Mukherjee
,
S.
,
Nateghi
,
R.
, and
Hastak
,
M.
,
2018
, “
Data on Major Power Outage Events in the Continental U.S
,”
Data Brief
,
19
, pp.
2079
2083
. 10.1016/j.dib.2018.06.067
72.
Baniassadi
,
A.
,
Sailor
,
D. J.
, and
Bryan
,
H. J.
,
2019
, “
Effectiveness of Phase Change Materials for Improving the Resiliency of Residential Buildings to Extreme Thermal Conditions
,”
Sol. Energy
,
188
, pp.
190
199
. 10.1016/j.solener.2019.06.011
73.
Holmes
,
S. H.
,
Phillips
,
T.
, and
Wilson
,
A.
,
2016
, “
Overheating and Passive Habitability: Indoor Health and Heat Indices
,”
Build. Res. Inf.
,
44
(
1
), pp.
1
19
. 10.1080/09613218.2015.1033875
74.
Anderson
,
M.
,
Carmichael
,
C.
,
Murray
,
V.
,
Dengel
,
A.
, and
Swainson
,
M.
,
2013
, “
Defining Indoor Heat Thresholds for Health in the UK
,”
Perspect. Public Health
,
133
(
3
), pp.
158
164
. 10.1177/1757913912453411
75.
Kingma
,
B.
, and
van Marken Lichtenbelt
,
W.
,
2015
, “
Energy Consumption in Buildings and Female Thermal Demand
,”
Nature Climate Change
,
5
(
12
), p.
1054
1056
. 10.1038/nclimate2741
76.
Karjalainen
,
S.
,
2007
, “
Gender Differences in Thermal Comfort and Use of Thermostats in Everyday Thermal Environments
,”
Build. Environ.
,
42
(
4
), pp.
1594
1603
. 10.1016/j.buildenv.2006.01.009
77.
Ebi
,
K. L.
, and
Meehl
,
G. A.
,
2007
, “The Heat Is on: Climate Change and Heatwaves in the Midwest,”
Regional Impacts of Climate Change: Four Case Studies in the United States
,
National Center for Atmospheric Research
,
Boulder, CO
, pp.
8
21
.
78.
Gao
,
Y.
,
Fu
,
J. S.
,
Drake
,
J.
,
Liu
,
Y.
, and
Lamarque
,
J.-F.
,
2012
, “
Projected Changes of Extreme Weather Events in the Eastern United States Based on a High-Resolution Climate Modeling System
,”
Environ. Res. Lett.
,
7
(
4
), p.
044025
. 10.1088/1748-9326/7/4/044025
79.
Krayenhoff
,
E. S.
,
Moustaoui
,
M.
,
Broadbent
,
A. M.
,
Gupta
,
V.
, and
Georgescu
,
M.
,
2018
, “
Diurnal Interaction Between Urban Expansion, Climate Change and Adaptation in US Cities
,”
Nature Climate Change
,
8
(
12
), pp.
1097
1103
. 10.1038/s41558-018-0320-9
80.
Kim
,
Y.-M.
,
Kim
,
S.
,
Cheong
,
H.-K.
,
Ahn
,
B.
, and
Choi
,
K.
,
2012
, “
Effects of Heat Wave on Body Temperature and Blood Pressure in the Poor and Elderly
,”
Environ. Health Toxicol.
,
27
, pp.
e2012013
e2012013
.
81.
Ren
,
Z. G.
,
Wang
,
X. M.
, and
Chen
,
D.
,
2014
, “
Heat Stress Within Energy Efficient Dwellings in Australia
,”
Archit. Sci. Rev.
,
57
(
3
), pp.
227
236
. 10.1080/00038628.2014.903568
82.
Christidis
,
N.
,
Stott
,
P. A.
,
Brown
,
S.
,
Hegerl
,
G. C.
, and
Caesar
,
J.
,
2005
, “
Detection of Changes in Temperature Extremes During the Second Half of the 20th Century
,”
Geophys. Res. Lett.
,
32
(
20
), pp.
20716
20716
. 10.1029/2005GL023885
83.
Karl
,
T. R.
, and
Knight
,
R. W.
,
1997
, “
The 1995 Chicago Heat Wave: How Likely Is a Recurrence?
Bull. Am. Meteorol. Soc.
,
78
(
6
), pp.
1107
1119
. <1107:TCHWHL>2.0.CO;2
84.
Meehl
,
G. A.
, and
Tebaldi
,
C.
,
2004
, “
More Intense, More Frequent, and Longer Lasting Heat Waves in the 21st Century
,”
Science
,
305
(
5686
), pp.
994
997
. 10.1126/science.1098704
85.
Baniassadi
,
A.
,
Sailor
,
D. J.
,
Krayenhoff
,
E. S.
,
Broadbent
,
A. M.
, and
Georgescu
,
M.
,
2019
, “
Passive Survivability of Buildings Under Changing Urban Climates Across Eight US Cities
,”
Environ. Res. Lett.
,
14
(
7
), p.
074028
. 10.1088/1748-9326/ab28ba
86.
Baniassadi
,
A.
,
2019
, “
Vulnerability of U.S. Residential Building Stock to Heat: Status Quo, Trends, Mitigation Strategies, and the Role of Energy Efficiency
,”
Ph.D. dissertation
,
Arizona State University
,
Tempe, AZ
.