Electric fans: A potential stay-at-home cooling strategy during the COVID-19 pandemic this summer?

Current public health guidance designed to protect individuals against extreme heat and the ongoing COVID-19 pandemic is seemingly discordant, yet during the northern hemisphere summer, we are faced with the imminent threat of their simultaneous existence. Here we examine the environmental limits of electric fan-use in the context of the United States summer as a potential stay-at-home cooling strategy that aligns with existing efforts to mitigate the spread of SARS-COV-2.

Introduction

In the northern hemisphere summer, the United States faces two simultaneous threats to public health: heat waves and the COVID-19 pandemic. The intersection of these health threats is particularly salient given those most susceptible to the adverse outcomes of COVID-19 both clinically (e.g. older adults and those with cardiovascular and respiratory diseases) (Xie et al., 2020) and socioeconomically (i.e. poor and marginalized) (Bibbins-Domingo, 2020), mirror those most at risk during periods of extreme heat. Over the coming months, public health advice designed to protect against heatwaves must be compatible with the guidance issued to combat the further spread of COVID-19.

On average, 65,000 Americans visit the emergency room every summer for heat illness (Hess et al., 2014). As US healthcare continues to be stretched to capacity by the demands of COVID-19, the provision of effective interventions for preventing heat-related illness has never been more important. However, resources from Centers for Disease Control and Prevention (CDC) developed prior to the current pandemic (Centers for Disease Control and Prevention, 2017; Centers for Disease Control and Prevention, 2016), provide recommendations that could increase the risk of community transmission of SARS COV-2 — especially among those most at risk of its detrimental effects. For example, CDC urge people to “visit older adults or others at risk at least twice a day and closely watch them for signs of heat exhaustion or heat stroke” (Centers for Disease Control and Prevention, 2017; Centers for Disease Control and Prevention, 2016). Moreover, those without access to air-conditioning, or those who cannot afford to use it, are encouraged to congregate in local “cooling centers” (e.g. heat-relief shelters, shopping malls and public libraries) (Centers for Disease Control and Prevention, 2017; Centers for Disease Control and Prevention, 2016). The proportion of the population for which this applies is also likely to increase as COVID-19 exacerbates the financial strain experienced by many at a time of nationwide instability. These concerns have received some attention from mainstream media (Flavelle, 2020; Yuan et al., 2020) and in light of the current pandemic CDC have acknowledged that community cooling centers could in fact facilitate the spread of SARS-COV-2 among at-risk individuals, and have therefore issued interim guidance specifically aimed at reducing the risk of virus transmission in these centers (Centers for Disease Control and Prevention, 2020). This guidance emphasises the importance of physical distancing and personal hygiene at public cooling centers (Centers for Disease Control and Prevention, 2020), yet no alternative stay-at-home cooling solutions to air-conditioning are provided, that eliminate the need to seek refuge from the heat in public spaces and align with existing efforts to limit the spread of COVID-19 in the community.

A growing body of scientific evidence strongly supports the efficacy of several low-resource home-based cooling solutions. For example, skin-wetting has been shown to reduce physiological heat strain, dehydration, and thermal discomfort at temperatures up to 47 °C, irrespective of humidity (Morris et al., 2019a). Electric fans are another low-cost, low-energy demand (i.e. ~20–50-times less A/C) (Jay et al., 2019; Bachar et al., 2012) cooling strategy. However, their cooling effect during a heatwave is dependent on the prevailing combination of temperature and humidity (Morris et al., 2019b), which can vary greatly across the United States. Fans improve skin surface evaporation in humid conditions, but in low humidity conditions sweat evaporates readily, even without a fan, and therefore fans provide no additional benefit (Morris et al., 2019b). When air temperature exceeds skin temperature (~35 °C/95 °F) fans also accelerate dry heat transfer towards the body, via convection (Cramer and Jay, 2019). A recent clinical trial (Morris et al., 2019b) showed that increases in physiological heat strain and thermal discomfort were lower with fan use during an acute exposure to the peak conditions of the most deadly hot/humid heat wave in recent US history (Chicago, 1995; 40 °C/104 °F, 50%RH). While in very hot/arid heat wave conditions (e.g. Los Angeles, 2018; 47 °C/117 °F, 10%RH) fan-use was clearly detrimental, accelerating body heating and exacerbating cardiovascular strain and discomfort relative to a no-fan control condition. Other studies have shown that fans can provide physiological cooling up to air temperatures of 42 °C/108 °F with ~50%RH (Ravanelli et al., 2015). Heat loss may also be compromised by a reduced physiological capacity to secrete sweat, common in older adults (Gagnon et al., 2016; Inoue et al., 1991) and individuals taking certain medications (e.g. anticholinergics) (Cheshire and Fealey, 2008), effectively reducing the range of conditions under which a fan is beneficial (Gagnon et al., 2016). However, any potential decrements in sweating can be compensated by externally applying water directly to the skin with a spray bottle (Morris et al., 2019a). Therefore, our aim was assess the potential utility of electric fan-use with light water-spraying as a stay-at-home cooling solution across the United States this summer, by comparing the biophysically modelled humidity-dependent temperature limits for this strategy to peak historical summer weather conditions.

Methods

Employing a standard conceptual human heat balance model, we assessed the difference between the increased convective heat transfer towards the body and increased evaporative heat loss potential away from the body, with and without an electric fan (Morris et al., 2019b). A detailed description of the model methodology, including partitional calorimetry equations used with reference to the supporting evidence base can be found in Appendix A. Assessments were performed at 25–50 °C with a relative humidity of 0–60%. These ranges were chosen to capture all peak combinations of temperature and humidity naturally occurring across all regions of the mainland United States. The air velocity of 4.5 m·s−1 was equivalent to an 18″ diameter fan at maximum speed, at 1.0 m distance positioned at waist height. A metabolic rate of 65 W·m−2, body surface area of 1.8 m2, mean skin temperature of 35.5 °C, and the dry insulation and evaporative resistance of clothing for a summer ensemble were used (Morris et al., 2019b).

The model yielded threshold combinations of air temperature and relative humidity (RH) at which the increase in convective heat load with fan-use exceeded the increase in evaporative potential assuming the maximum sweat rate of an older adult (440 mL·h−1) (Inoue et al., 1991) spraying an additional 115 mL (~0.5 cup) of water onto the skin every hour (Morris et al., 2019a). Based on our pilot laboratory work this is equivalent to an individual that is wearing shorts and a sleeveless shirt spraying their exposed limbs (in 8 sweeping sprays) every 5 min. These temperature/humidity thresholds were subsequently compared to the highest hourly ambient temperature and associated humidity from May-to-July recorded over the last 20 years (2000–2019, inclusive) for 105 separate metropolis areas (100 most populous metropolises + top metropolis in each conterminous state not represented) across the United States.

Results

The modelled air temperature limit for fan-use was 37.2 °C at the lowest RH (0%), rising to 42.3 °C at 20%RH and then reducing to 39.9 °C at 50%RH (Fig. 1 ). Historically, fan-use would have been detrimental (i.e. exerted a heating effect) on 0% of summer days in 80 of 105 of the metropolises examined (home in 2018 to ~176 million individuals) (Fig. 2C). These metropolis areas are predominantly located in the Northeast, Southeast and Midwest regions of the United States, as well as the West Coast (Fig. 2A). According to our model, electric fan use would have been detrimental in a further 15 metropolises, mainly in the South (e.g. Austin, TX), on <0.5% of summer days; equivalent to 1–10 days in 20 years (Fig. 2B). In the remaining 10 metropolises, mostly in the hot-arid interior of the Southwest, fan use would have exerted a heating effect on up to 43.5% of summer days (Phoenix, AZ; Fig. 2B).

Fig. 1

Biophysically modelled humidity-dependent threshold temperatures at which electric fan-use and modest skin-wetting becomes detrimental; above the line fans exert a heating effect, below the line fans exert a cooling effect. Individual data points represent the highest 1-h ambient temperature and associated relative humidity recorded between 2000 and 2019 (inclusive) for each of the 105 metropolises analyzed. Data points are colour-coded to identify percentage of summer days in last 20 years on which environmental limits were exceeded where green: 0%; yellow: >0% but ≤0.5%; red: >0.5%. (For interpretation of the references to colour in this figure legend, the reader is directed to the web version of this article.)

Fig. 2

A) Map of the conterminous United States identifying the 105 metropolises examined, with markers colour-coded to identify percentage of summer days between 2000 and 2019 (inclusive) that exceeded the biophysically-modelled environmental limits for fan use coupled with modest skin wetting (green: 0%; yellow: >0%, but ≤0.5%; red: >0.5% days); B) average number of days per summer month that fan-use would have exerted a body heating effect for each individual metropolis that exceeded the environmental limit for fan use on >0%, but ≤0.5% (top); and on >0.5% (bottom) of summer days; C) total population examined across 105 metropolises, based on 2018 census estimates, split to identify number of inhabitants for which fan-use exceeded environmental limits on 0% (green); >0%, but ≤0.5% (yellow); >0.5% (red) summer days in the last 20 years. (For interpretation of the references to colour in this figure legend, the reader is directed to the web version of this article.)

Discussion

The present analysis indicates that electric fan-use with light water-spraying potentially offers a feasible stay-at-home cooling strategy during heat extremes for large parts of the US historically experiencing hot-humid summer conditions. In comparison to existing experimental data that demonstrates fan-use providing physiological cooling up to air temperatures of 42 °C/108 °F with ~50%RH (Ravanelli et al., 2015) our modelled thresholds appear conservative. The aforementioned study (Ravanelli et al., 2015) and others (Morris et al., 2019b) were undertaken using young, healthy participants, but it is known that other factors such as age alter the environmental limits for fan-use (Gagnon et al., 2016) likely due to age-related decrements in sweating (Inoue et al., 1991), that limit the potential increase in evaporative heat loss a fan can provide. To establish environmental limits more generalizable to the American public that can be easily adopted in at-home settings we chose to incorporate several conservative components in the current model. Examples include the assumption of a low maximal sweat rate (440 mL·h−1), more representative of an older adult (Inoue et al., 1991) and considerably lower than sweat rates reported in previous experimental research examining fan effectiveness (i.e. average sweat rate in 47 °C/117 °F, 10%RH fan condition = 691 mL·h−1 (Morris et al., 2019b)) and a low volume of water used for skin-wetting (115 mL⸱h−1/0.5 cup⸱h−1), relative to previously reported self-dousing values (i.e. =698 mL⸱h−1 (Morris et al., 2019a)).

There are several important considerations when interpreting our model. Firstly, air temperatures used in the model were recorded from outdoor locations rather than from indoor locations, such as a home. Although, given the heterogeneity of building characteristics (i.e. thermal mass, ventilation rate, and insulation) known to influence the relationship between indoor and outdoor air temperatures (Nguyen and Dockery, 2016) we opted to use outdoor environmental conditions in the current model. Secondly, this cooling strategy is targeted as a stay-at-home cooling solution, rather than for use in public spaces. There is a lack of evidence suggesting fan-use may aid virus transmission, but given the suggested occurrences where air ventilation systems may have acted as a vector for virus transmission (Li et al., 2005; Yu et al., 2004) and our constantly developing knowledge of the nature of its spread (Van Doremalen et al., 2020; World Health Organization, 2020), it is possible that fan-use may accelerate the distribution of virus particles present in the home. There is indeed inherent transmission risk associated with co-habitation (Li et al., 2020; Wang et al., 2020). Importantly though, fan use during heat extremes in the home prevents people seeking cooling in public places among individuals whose virus status is less likely to be known than co-habitants, thus limiting personal risk of transmission and further spread in the community. Finally, our model identifies the point at which using a fan is better than not using a fan, and therefore does not quantify the amount of cooling, and whether it is sufficient to maintain body temperature within safe limits. Nevertheless, even with the conservative approach taken in developing the current model, for 200 million of the ~221.5 million residents (according to 2018 census estimates) in the 105 metropolis areas assessed, fan-use with light water-spraying would have been beneficial (i.e. exerted a cooling effect) relative to not using a fan on more than 199 of every 200 summer days in the past 20 years (Fig. 2C). It is therefore clear that public health officials should not advise people to turns fans off during heat waves as is current practice in a range of jurisdictions (New York State Department of Health, 2017; Department of Homeland Security, 2020).

While public health officials strive to protect all citizens during the current pandemic, parallel efforts are also required to proactively prepare for the likely overlap of COVID-19 with extreme heat. Heatwave preparation plans are increasingly centered on building community resilience and protecting the most vulnerable members of society (Manangan et al., 2020; Abbinett et al., n.d.; Fox et al., 2019). While this approach must continue, we require adaptive, yet evidence-based, efforts to protect against the ill-effects of extreme heat that align with current public health recommendations, such as physical distancing and stay-at-home orders, that are crucial to mitigating the spread of SARS-COV-2 this summer. Collectively, this analysis highlights how a stay-at-home fan-use with skin-wetting approach could better enable the simultaneous mitigation of heat stress and the spread of SARS-COV-2 across much of the United States. The widespread utility of this method also provides evidence for this alternative, low-cost, low-energy cooling strategy in a post-COVID climate. While, in hot-arid regions, environmental conditions have regularly exceeded the threshold for a heating effect with fan use, so alternative at-home cooling strategies, e.g. water-dousing without fans and/or cold-water foot immersion (Morris et al., 2019a), could be considered.

CRediT authorship contribution statement

Lily Hospers: Conceptualization, Data curation, Visualization, Writing - original draft, Writing - review & editing. James W. Smallcombe: Conceptualization, Data curation, Writing - original draft, Writing - review & editing. Nathan B. Morris: Methodology, Data curation. Anthony Capon: Supervision, Writing - review & editing. Ollie Jay: Conceptualization, Methodology, Supervision, Writing - review & editing.

Declaration of competing interest

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Table A1

Weather analysis by city using summer (June–August) weather data (CustomWeather, Inc., 271 Miller Avenue Mill Valley, CA, USA 94941) from 2000 to 2019.

Metropolitan statistical area	Peak temperature (ºC)	Peak relative humidity (%)	Total days fans detrimental	Total days in analysis	Average detrimental days per summer	Detrimental days (%)
Phoenix, AZ	48.3	6	801	1839	40.1	43.5
Las Vegas, NV	47.0	2	611	1747	30.6	35.0
Tucson, AZ	46.1	7	232	1839	11.6	12.6
Fresno, CA	44.4	19	74	1839	3.7	4.0
Bakersfield, CA	44.0	16	50	1839	2.5	2.7
El Paso, TX	43.0	10	38	1839	1.9	2.1
Riverside, CA	45.0	13	24	1839	1.2	1.3
Stockton, CA	46.0	16	19	1838	1.0	1.0
Boise City, ID	43.0	9	16	1839	0.8	0.9
Sacramento, CA	43.0	17	13	1839	0.7	0.7
Tulsa, OK	44.4	19	10	1839	0.5	0.5
Oklahoma City, OK	43.9	14	9	1839	0.5	0.5
Wichita, KS	43.3	19	8	1839	0.4	0.4
Dallas, TX	42.8	22	5	1839	0.3	0.3
Austin, TX	43.3	15	3	1839	0.2	0.2
Salt Lake City, UT	41.1	11	3	1839	0.2	0.2
Billings, MT	41.1	12	3	1839	0.2	0.2
Little Rock, AR	45.0	19	2	1839	0.1	0.1
Denver, CO	39.4	9	1	1839	0.1	0.1
St. Louis, MO	42.2	13	1	1839	0.1	0.1
San Antonio, TX	42.8	16	1	1839	0.1	0.1
Albuquerque, NM	40.0	5	1	1839	0.1	0.1
McAllen, TX	43.9	19	1	1839	0.1	0.1
Columbia, OH	42.2	17	1	1839	0.1	0.1
Provo, UT	41.0	9	1	1839	0.1	0.1
New York, NY	39.4	31	0	1812	0.0	0.0
Los Angeles, CA	34.0	26	0	1839	0.0	0.0
Chicago, IL	39.4	31	0	1793	0.0	0.0
Houston, TX	42.2	20	0	1747	0.0	0.0
Washington, DC	40.6	35	0	1839	0.0	0.0
Miami, FL	36.7	38	0	1839	0.0	0.0
Philadelphia, PA	39.0	24	0	1839	0.0	0.0
Atlanta, GA	40.6	25	0	1839	0.0	0.0
Boston, MA	38.9	30	0	1839	0.0	0.0
San Francisco, CA	39.4	19	0	1747	0.0	0.0
Detroit, MI	38.3	32	0	1839	0.0	0.0
Seattle, WA	39.4	20	0	1840	0.0	0.0
Minneapolis, MN	38.9	32	0	1839	0.0	0.0
San Diego, CA	36.0	39	0	1839	0.0	0.0
Tampa, FL	36.0	47	0	1839	0.0	0.0
Baltimore, MD	41.0	20	0	1839	0.0	0.0
Orlando, FL	37.2	39	0	1839	0.0	0.0
Charlotte, NC	39.4	27	0	1839	0.0	0.0
Portland, OR	41.0	20	0	1748	0.0	0.0
Pittsburgh, PA	36.1	43	0	1839	0.0	0.0
Cincinnati, OH	39.4	31	0	1839	0.0	0.0
Kansas City, MO	41.7	26	0	1839	0.0	0.0
Columbus, SC	37.2	38	0	1839	0.0	0.0
Cleveland, OH	36.7	35	0	1835	0.0	0.0
Indianapolis, IN	40.6	25	0	1839	0.0	0.0
San Jose, CA	41.0	23	0	1839	0.0	0.0
Nashville, TN	41.7	17	0	1839	0.0	0.0
Virginia Beach, VA	38.3	44	0	1836	0.0	0.0
Providence, RI	38.3	33	0	1839	0.0	0.0
Milwaukee, WI	38.3	29	0	1839	0.0	0.0
Jacksonville, FL	38.3	42	0	1839	0.0	0.0
Raleigh, NC	40.6	30	0	1839	0.0	0.0
Memphis, TN	41.1	25	0	1839	0.0	0.0
Richmond, VA	41.0	23	0	1839	0.0	0.0
Louisville, KY	40.0	22	0	1839	0.0	0.0
New Orleans, LA	38.0	37	0	1839	0.0	0.0
Hartford, CT	39.0	29	0	1839	0.0	0.0
Birmingham, AL	41.0	17	0	1839	0.0	0.0
Buffalo, NY	35.0	37	0	1839	0.0	0.0
Rochester, NY	36.1	46	0	1839	0.0	0.0
Grand Rapids, MI	40.0	30	0	1839	0.0	0.0
Worcester, MA	39.4	17	0	1826	0.0	0.0
Bridgeport, CT	38.3	38	0	1839	0.0	0.0
Omaha, NE	40.6	25	0	1839	0.0	0.0
Greenville, SC	40.6	30	0	1836	0.0	0.0
Knoxville, TN	40.6	22	0	1839	0.0	0.0
Albany, NY	37.2	41	0	1839	0.0	0.0
New Haven, CT	38.0	37	0	1838	0.0	0.0
Oxnard, CA	30.0	45	0	1839	0.0	0.0
Allentown, PA	39.4	36	0	1838	0.0	0.0
Baton Rouge, LA	39.4	30	0	1747	0.0	0.0
North Port, FL	37.2	42	0	1839	0.0	0.0
Dayton, OH	38.9	30	0	1839	0.0	0.0
Charleston, SC	38.3	33	0	1839	0.0	0.0
Greensboro, NC	38.3	33	0	1839	0.0	0.0
Cape Coral, FL	37.0	37	0	1839	0.0	0.0
Colorado Springs, CO	37.2	18	0	1839	0.0	0.0
Lakeland, FL	39.0	33	0	1839	0.0	0.0
Akron, OH	37.8	37	0	1839	0.0	0.0
Poughkeepsie, NY	39.0	29	0	1837	0.0	0.0
Ogden, UT	39.0	14	0	1834	0.0	0.0
Winston, NC	38.9	30	0	1838	0.0	0.0
Madison, WI	39.4	30	0	1839	0.0	0.0
Deltona, FL	36.7	34	0	1839	0.0	0.0
Des Moines, IA	41.1	26	0	1839	0.0	0.0
Syracuse, NY	38.0	27	0	1839	0.0	0.0
Springfield, MA	38.3	26	0	1838	0.0	0.0
Augusta, GA	42.0	25	0	1839	0.0	0.0
Toledo, OH	38.9	32	0	1839	0.0	0.0
Palm Bay, FL	37.0	47	0	1839	0.0	0.0
Jackson, MS	41.1	25	0	1839	0.0	0.0
Durham, NC	40.6	30	0	1839	0.0	0.0
Harrisburg, PA	38.9	40	0	1839	0.0	0.0
Spokane, WA	40.6	13	0	1839	0.0	0.0
Chattanooga, TN	41.1	21	0	1839	0.0	0.0
Portland, ME	37.8	30	0	1839	0.0	0.0
Fargo, ND	39.0	33	0	1839	0.0	0.0
Sioux Falls, SD	39.4	26	0	1839	0.0	0.0
Burlington, VT	36.7	32	0	1839	0.0	0.0
Cheyenne, WY	36.7	13	0	1839	0.0	0.0