After considering various issues regarding exposure to PM, the workshop next explored several approaches to mitigating exposure as a means of reducing risk from exposure to indoor PM. William Fisk of the Lawrence Berkeley National Laboratory addressed the use of filtration to remove airborne PM. Sergey Grinshpun of the University of Cincinnati College of Medicine described several methods of controlling viable bioaerosol particles in indoor air. Brett Singer of Lawrence Berkeley National Laboratory then discussed the challenges of mitigating PM exposure in low-socioeconomic households. An open discussion moderated by Tiina Reponen followed the three presentations.
Filtration can be effective in reducing indoor levels of PM, William Fisk said, but current filtration practices are relatively ineffective even though the cost of doing better using existing technology is not prohibitive. Indeed, Fisk said, the filtration of incoming outdoor air and recirculated indoor air should be the first approach taken to mitigate individual exposure to PM. He added that there are also techniques, such as using ion generators and increasing air movement, to enhance particle deposition on indoor surfaces.
Many particle filtration technologies exist today, with the use of fibrous
1 This section is based on the presentation by William Fisk, a senior scientist at Lawrence Berkeley National Laboratory, and the statements are not endorsed or verified by the National Academies of Sciences, Engineering, and Medicine.
filters or stretched membranes predominating. Other technologies include electrostatic devices that charge particles and collect them on charged and grounded plates, combination technologies such as ion generation and filtration through fibrous materials, and combining a tight building envelope with exhaust ventilation. While Fisk limited his remarks to fibrous filtration, he said that these other technologies have potential for wider use.
Several factors determine the performance of a filter: its rate of particle removal, its energy use, the cost of filtration, the filter’s reliability, and inadvertent pollutant production. The factors that affect particle removal include the rate and duration of air flow, the particle removal efficiency as a function of particle size, and the location of the filter relative to pollutant sources and to the location of a building’s occupants. Factors affecting energy and cost include airflow resistance, pressure drop, fan and motor efficiency, and the particle-holding capacity as it relates to the filter’s lifetime. Fisk noted that there are three systems used in the United States to rate filters:
- Minimum efficiency reporting value (MERV), developed the American Society of Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE), which rates filters on their minimum efficiency within a set of particle size bins
- Micro-Particle Performance Rating (MPR), developed by 3M, which rates performance on removal of PM in the 0.3- to 1-micron size range
- Home Depot’s Filter Performance Rating system (FPR), which uses a scale of 1 to 10 and a color code to rate filters based on large particle removal, small particle removal, and particle-holding capacity.
Fisk said that while filtration can be highly effective, the effectiveness of filtration systems varies widely. One recent informative study of particle filtration in nine southern California classrooms (Polidori et al., 2013) resulted in the data displayed in Figure 5-1. This study compared filtration in a baseline scenario, in which the HVAC system fitted with a MERV 7 filter runs continuously, to several other filtration alternatives. Adding a standalone filtration unit with a MERV 16 filter produced a large increase in removal effectiveness, Fisk said, though adding a MERV 16 filter in the HVAC system itself was even more effective. None of the configurations tested were able to maintain the indoor concentration of PM10 below 60 percent of the outdoor air PM10 concentration, probably because of high indoor PM10 generation rates in classrooms. Many of the configurations were quite effective at reducing the indoor concentration of PM2.5 and UFPs relative to the baseline filtration system.
One limit to the impact that filtration currently has on reducing PM exposure, Fisk said, is that the current ASHRAE standards for both residences and commercial building only require MERV 6 filters except in areas of the country that are not in compliance with PM2.5 regulations (in which case the standard for commercial buildings is MERV 11 filters). However, MERV 6 filters remove less than 20 percent of the particles of most sizes (see Figure 5-2). “So the filters that we commonly use have a low efficiency for particles in the most interesting size range,” Fisk said. In homes, this deficiency is compounded by the fact that HVAC systems run intermittently. Given that approximately 55 percent of homes have filters with a rating of MERV 6 or lower (El Orch et al., 2014), which will remove about 7 percent of PM2.5 (Azimi et al., 2014), and that the HVAC in a typical home runs approximately 20 percent of the time (Cetin and Novoselac, 2015) with an air flow rate of approximately 4.4 air exchanges per hour (Jump et al., 1996; Stephens et al., 2011), Fisk calculated that the total removal rate is less than 10 percent of the indoor particle load per hour. “Those filters are not bringing us much benefit, but that is what we use today,” he said.
Most filters sold today contain embedded charged fibers, which can increase particle removal efficiency but only for a limited time (Raynor and Chae, 2004). Studies in several settings have shown that PM in cigarette smoke and diesel exhaust can quickly reduce the efficiency of charged fiber
A growing trend, he said, is to add “nanofibers,” a term the filtration industry uses for fibers with diameters of less than 0.5 microns (as compared with diameters of a few microns in the most common filters). Experimental data suggest that these nanofiber filters can produce a higher ratio of particle removal efficiency to air pressure drop (Ahn et al., 2006; Leung et al., 2009; Wang et al., 2008).
Fisk explained that the energy used to remove particles with filters varies dramatically depending on the system being used. His recent research showed, for example, that a standalone high-efficiency filter can remove particles for a fraction of the cost per gram compared to an HVAC system (Fisk and Chan, 2016) (see Table 5-1).
Fisk then discussed the common belief that better filters will substantially increase energy costs because they will increase airflow resistance. “For many situations,” he said, “the data do not bear that out” (Walker et al., 2013), and while better filters do cost more, he said he would argue that the increase in costs is not so high as to be prohibitive. Deeper filters with more pleating can reduce airflow resistance, which minimizes the effects on the energy consumed by the HVAC system fan, he explained. Calculations also show that for a system in a commercial building that has multiple
|HVAC + Low ε Filter||HVAC with High ε Filter||Standalone with High ε Filter|
|Flow rate (h–1)||4.3||4.3||1|
|House volume (m3)||433||433||433|
|Watt per m3 s–1||1090*||1090*||600*|
|PM2.5 removal efficiency||0.12*||0.27*||0.9|
|Time||1 year||1 year||1 year|
|Home PM2.5||20 µg m–3||20 µg m–3||20 µg m–3|
|PM removed (g)||39||88||68|
|Electricity cost ($)||$650||$650||$83|
|$ Elec. per gr. PM removed||$16.7||$7.4||$1.2|
Electricity Cost of operating a filtration system
Particle Mass Removed = flowrate efficiency time concentration
NOTE: Data marked with “*” are derived from Fisk and Chan, 2016.
SOURCE: Fisk slide 13.
filters and occupants, the increased life-cycle costs of a MERV 13 versus MERV 8 filter works out to at most $3 per person per month (Montgomery et al., 2012).
In 2013, Fisk reviewed the health benefits of filtration (Fisk, 2013) and came to two main conclusions. The first was that filtration has only a minor benefit with regard to reducing allergy and asthma outcomes. There is some evidence of benefit in homes with large sources of allergens, but only a fraction of health outcomes improved. The second conclusion was that the greatest potential comes from using better filtration to reduce indoor concentrations of outdoor PM, thus reducing the morbidity and mortality associated with outdoor air PM. “The health benefits are predicted to far exceed the costs for those interventions,” Fisk said. Other conclusions he drew in his review included
- Systems that delivered filtered air to the breathing zone when individuals are sleeping appear to be more effective in reducing allergy and asthma symptoms;
- Evidence of health benefits from filtration in homes, offices, and schools in subjects without allergies and asthma is limited; and
- The reductions in markers of future adverse coronary events with filtration support modeled health benefits of using filtration to reduce particles from outdoor air.
Fisk concluded his presentation with a list of issues and challenges relating to filtration, which included
- Quantifying, demonstrating, and communicating benefits to motivate the use of better filters;
- Improving filtration effectiveness while reducing costs;
- Increasing minimum filtration efficiency requirements in standards;
- Limited expected effectiveness for locally resuspended coarse particles, such as some allergens;
- Pollutant generation by some electronic air cleaners;
- Soiled filters may emit pollutants and diminish perceived air quality;
- Many expensive and ineffective products are sold; and
- Empirical validation of predicted health benefits.
One of Fisk’s concerns relates to the trend of increasing the use of natural ventilation in commercial buildings, which will increase exposure to outdoor particles and ozone. He said that it will be important to identify approaches for mitigating those exposures in naturally ventilated commercial buildings. Fisk said he also believes that there is a need to better understand the relative health risks of outdoor PM versus PM that is generated indoors and also how filtration can be a tool to differentially affect exposures and risks.
One of the measurables that is specific to biological particles, Sergey Grinshpun said, is the percentage of airborne organisms that are viable. Viability is measured by counting colonies of microorganisms from collected particles that grow on agar plates. Other ways of analyzing biological materials include looking at antigens and allergens, quantifying molecules specific to the cell wall or membranes, and assessing the presence of fungal toxins, also known as mycotoxins. Biological particles appear in a wide
2 This section is based on the presentation by Sergey Grinshpun, a professor of environmental health at the University of Cincinnati College of Medicine, and the statements are not endorsed or verified by the National Academies of Sciences, Engineering, and Medicine.
range of sizes (see Figure 5-3). Viruses range in size from approximately 20 to 400 nanometers, the bacterial and fungal size range is approximately 0.5 to almost 10 micrometers, and pollen occupies the high end, from >10 to >100 micrometers in diameter. Air purification may not be an important issue for pollen because pollen grains settle out of the air rapidly.
Grinshpun said that there are two main approaches to mitigating exposure to biological particles. The first is to reduce the overall PM burden from all sources via, for example, filtration or electrostatic precipitation. While Fisk had already discussed filtration, Grinshpun pointed out that every type of filter has a characteristic particle size for which its removal efficiency is lowest. For the typical devices used for indoor air filtration, this size ranges between <0.1 and 1 micrometer, which corresponds to the sizes of larger viruses and many bacteria. Electrostatic precipitators have been found to be good at removing bacterial aerosol particles (Mainelis et al., 2002). While these precipitators are inexpensive and quiet to operate, they are not commonly used against bioaerosols.
Ozone generation devices are available but Grinshpun questioned their utility for removing indoor PM or reducing the viability of microorganisms in indoor air. Grinshpun and his colleagues have shown that commercially available ozone generators do not remove particles but instead create new ones (Grinshpun et al., 2010). Ion emission devices work by generating ions
that attach to particles, causing the particles to migrate toward and stick to indoor surfaces. Grinshpun and his colleagues evaluated this method in a test chamber and found that it does have the potential for removing particles, albeit with lower efficiency than filtration. The main problem is that most ion emitters also emit ozone, which, as Glenn Morrison had noted in his presentation, can trigger the generation of UFPs.
A number of different methods, including heat, UV light, and cold plasma, have been used to inactivate microorganisms. “With these methods, we do not care about concentration, we just want to kill viable microorganisms,” Grinshpun explained. He noted that some commercially available air cleaners use ion emission, ozone generation, and photocatalytic oxidation to inactivate microorganisms. The efficiency of these methods varies, he said.
Thermal inactivation has potential as a means of inactivating microorganisms, even stress-resistant bacterial spores such as anthrax spores. In one experiment, Grinshpun and his colleagues showed that only 0.1 percent of anthrax surrogate spores remained viable after passing through a thermal inactivation device at 315°C (Grinshpun et al., 2010). The major limitation to this technique, Grinshpun said, is being able to process enough air through such a device. A group of investigators in South Korea performed a study similar to what Grinshpun and his colleagues did with the anthrax surrogate spores (Jung et al., 2009) and showed that short-term exposure to high temperature changes the physical structure of aerosolized fungal spores. Whether this change is responsible for the ultimate inactivation of the spores is still unclear, Grinshpun said, “but the bottom line is that it is quite efficient.” Viruses, he added, are readily inactivated at air temperatures as low as 60°C.
The ability of UV irradiation to inactivate viable microorganisms has been well studied, though not in aerosols. One study (Peccia et al., 2001) found that the rate of inactivation of aerosolized bacteria using UV light depended on the humidity. Other experiments have shown that the combination of UV light and heat is more effective at inactivating bacteria in indoor air environments and at lower temperatures than when heat alone is used. The one caveat to the use of UV light, Grinshpun said, is that UV lamps can generate ozone. Recently, Grinshpun and his colleagues have been studying the use of atmospheric-pressure cold plasma to inactive viable microorganisms, and they found that this method causes viruses to fragment (Wu et al., 2015). They have not yet studied the effects of cold plasma on other types of bioaerosol particles.
Viable microorganisms, Grinshpun said, can also be inactivated after they have been collected on filters. Biocidal chemicals, such as iodine (Eninger et al., 2008), have been shown to be effective at inactivating microorganisms on filters, as have microwave and infrared irradiation (Lee
et al., 2009; Ratnesar-Shumate et al., 2008; Zhang et al., 2010). Alumina nanofibers with a positive surface charge have been shown to strongly retain virus particles, which are often naturally negatively charged (Li et al., 2009).
As a final note, Grinshpun mentioned that filtering respirators can reduce exposure to biological aerosols (Eninger et al., 2008; Grinshpun et al., 2007) and are generally as effective as stationary air filters. Protection may not be efficient, however, because of leakage between the respirator and the user’s face. Grinshpun also expressed caution about the use of so-called antimicrobial respirator filters, given that the risks of inhaling biocidal agents or having them come in prolonged contact with skin are not known.
Low socioeconomic status is often equated with low income, Brett Singer noted, but it also is correlated with low education and, more importantly, low status and low access to information, all of which are important when it comes to thinking about changing behavior or practice. “When we talk about these fixes,” he said, “something that will work for a family that has it all together might not work for a family that is struggling just to get through the day.”
There are physical, economic, and sociological challenges to reducing PM in low-socioeconomic homes, Singer said (see Table 5-2), many of which Gary Adamkiewicz addressed in his earlier presentation. Singer pointed out two items in particular. “When we talk about low-cost remedies, what is low-cost for me is going to be different from what is low-cost to a family of four living on $30,000 a year that has no credit or very costly credit,” he said. “There’s also limited or no choice in their housing. They cannot move or readily change it.” He also pointed to the importance of low status, which often translates into a limited ability to demand repairs. “They do not complain because they are worried about being thrown out and they have no alternative place to live,” he said.
While the typical way to think about mitigation is to parse it into source control, ventilation, and filtration, Singer said he uses a different mental model, one that includes reducing PM from outdoor sources, reducing indoor sources, and accelerating the removal rate of indoor PM. The first step, he said is to reduce PM from the outdoors, given that a large
3 This section is based on the presentation by Brett Singer, a staff scientist in the Indoor Environment Group at Lawrence Berkeley National Laboratory, and the statements are not endorsed or verified by the National Academies of Sciences, Engineering, and Medicine.
SOURCES: Singer slides 6 and 7.
fraction of indoor PM originates outdoors (Allen et al., 2012; Meng et al., 2004). In a series of experiments conducted in a moderately tight, empty house located some 300 meters downwind of Interstate 80 in Sacramento, California, Singer and his colleagues studied the effects of a number of different combinations of ventilation and filtration on PM2.5 levels (see Figure 5-4) and UFP levels inside the house. They found that a relatively tight shell was very effective in reducing PM2.5 and UFP infiltration but was less effective in keeping out carbon black particles. Levels fluctuate, however, and the research observed times during the day when the indoor level of carbon particles was higher than that outdoors.
Indoor PM levels were further reduced by filtration, Singer said, either when a MERV 16 filter was used to filter the air supply coming into the house or when a MERV 13 or better filter was used with a recirculating HVAC system. “You can get very effective outdoor particle reductions with a recirculating system,” Singer said. He noted that a simulation analysis using measured parameters conducted by Brent Stephens and his colleagues (Zhao et al., 2015) found that protecting the indoors from outdoor particles by sealing the envelope provides the biggest impact on reducing in-home exposure to outdoor PM, while filtration does more for reducing levels of indoor PM generated indoors. However, filtration was predicted to have a significant impact on indoor PM in old homes with significant infiltration of outdoor PM.
Singer said that the air tightness of homes has improved over the years (Chan et al., 2013) and that retrofitting measures do reduce air infiltration in existing homes. Weatherization assistance programs, he said, are achieving median reductions in PM levels of 30 percent for single-family homes and 28 percent for multifamily homes.
Indoor sources of PM2.5 and UFPs vary greatly by home and according to the time of day (Wallace et al., 2003). If there is habitual smoking in the house, that will typically be the most importance source, Singer said. He noted that efforts to get smokers to stop smoking indoors have been successful, particularly in low-socioeconomic homes, and that when smokers stop smoking indoors, it does produce meaningful reductions in indoor levels of PM2.5 (Semple et al., 2015; Wilson et al., 2012; Zhang et al., 2012). Portable filters placed in children’s bedrooms reduce PM levels in the homes of smokers and non-smokers alike (Batterman et al., 2012), though the researchers who conducted this study found that filter use waned over time.
Cooking is an important PM source in most homes, as are candles and incense in homes where these are used frequently. As had already been discussed, hot surfaces, resuspension, and cleaning can be important indoor sources of PM. With regard to cooking, range hoods can be effective at removing PM from indoor air, but only if they are used and only if they are installed correctly. Some range hoods, for example, simply run air through a charcoal filter and recirculate it back into the home. Singer
said that he and his colleagues at Lawrence Berkeley National Laboratory are working with the voluntary consensus standards organization ASTM International to develop a standard method of testing the efficiency with which range hoods capture cooking pollutants. As part of that work, he has found that the capture efficiency for cooking on the back burner is typically greater than 70 percent but that the efficiency is highly variable for cooking on the front burner (Delp and Singer, 2012). These measurements were made under controlled conditions in the laboratory, Singer cautioned, and the results were far worse when he went into the field and tested hoods in homes (Singer et al., 2012). A study using self-reports found that one-third of those responding to the survey used their hoods infrequently and that 10 percent reported they never used their hoods (Mullen and Singer, 2012). Some 20 percent of homes surveyed in California did not have an exhaust fan over their stoves.
With regard to costs, Singer said a quiet, energy-efficient, standalone high-efficiency particulate air/arrestance (HEPA) filter unit may cost $400 to $500 to cover 500 square feet of living space, with a less efficient, noisier unit costing perhaps half as much. The filters for the low-end units range from $20 to $30 and those for the high-efficiency units from $80 to $100. Low-income households are not likely to spend their limited funds on a filtering device, Singer said. Central HVAC filters are less expensive, but Singer said the landlord of a low-income housing unit is likely to think twice about spending $20 to $30 on a MERV 13 filter for every unit. A basic, noisy range hood can cost less than $50, but a quiet range hood with sufficient power to efficiently remove PM can cost $200 to $300.
The best control, then, starts with a good building, Singer said in summary: one with an airtight envelope, a vented range hood that is also quiet (so that it will be used), a central forced air HVAC system with an efficient blower and a 2- to 5-inch filter slot, robust venting of combustion appliances, and limited use of carpeting, except perhaps in the case of housing for the elderly where slipping on uncarpeted surfaces can be hazardous. Singer also suggested a number of actions that individuals can take to reduce their exposure to PM, including closing windows to reduce the levels of outdoor PM, particularly when pollution is bad or likely to be bad; restricting smoking and burning candles and incense; using a range hood and cooking on back burners; using a HEPA vacuum cleaner and ventilating when cleaning; and investing in good filters and using filtration. As a final thought, he said that PM is just one element of the indoor environment and rather than worry about which elements of green housing are most important, the key point is that providing good housing for people will provide a great deal of benefits in many areas beyond reducing PM.
Tiina Reponen started the discussion by asking the three session speakers to list the key questions that need answering with regard to mitigating exposure to PM. Fisk responded that there is a need to identify appropriate techniques to deal with airborne allergens or particles that are inflammatory so as to help reduce the effects that these have on respiratory difficulties, including asthma. Fisk also said he thought that while filtration is predicted to produce large reductions in mortality and morbidity, empirical data are needed to support those predictions even though acquiring such data will be challenging. He did note that there are some data in the asthma mitigation literature indicating that a broad combination of approaches is more effective at improving health than any single approach. Also, a subset of studies that looked at the effects of using filtration systems to ventilate the breathing zone of asthmatic individuals when they were sleeping found that these systems produced more benefits more consistently than whole-house measures (Fisk, 2013).
Grinshpun said that given the rapidly growing U.S. and European markets for air purifiers, he would like to see more research to identify the various byproducts produced by some of the methods for removing bioaerosols from circulation and to determine the optimal condition under which a given method is most efficient at reducing exposure without doing any harm. He said in response to a question about ozone production from ion generators that no device that emits ozone should be deployed and that he has tested ion generators that do not emit any measurable ozone. However, he added, filter-based air purifiers are generally more efficient than ion generators at removing indoor PM. Singer said that one of the big questions for him is how to better communicate what is already known to the public about the effectiveness of and issues associated with various mitigation strategies so that the public, including building professionals, can make use of information on how to best mitigate exposure to PM.
Grinshpun, responding to a question about the mechanism by which cold plasma inactivates viruses, said that the mechanism is still not well understood. With regard to thermal inactivation, he said that mechanical disintegration of microorganisms may occur at 600 to 700°C, which is where 100 percent inactivation has been observed, but the mechanism by which lower temperatures produce moderate levels of inactivation is still not adequately characterized.
When asked about the difficulty of retrofitting an HVAC system to take deeper MERV 13 or MERV 16 filters, Fisk said that manufacturers are now making filters with higher than MERV 7 efficiency that fit in standard 1-inch filter slots and that there are systems that can be installed over the return grill instead of in the furnace system that are not hard to install. “In
many cases, you do not need to retrofit,” he said, “but having said that, a home owner or renter in a low-income environment is not likely to have the resources to make even modest improvements their highest priority.” Singer said that adding more efficient filters to some older HVAC systems could increase the pressure drop significantly, raising a legitimate concern about whether a retrofit on an existing system is a good idea. With regard to retrofitting kitchen exhaust hoods, Richard Corsi commented that many range hoods, at least in Texas where he lives, get vented into attics, which can lead to a buildup of chemically reactive unsaturated fatty acids on surfaces in that space.
Terry Brennan said that while tightening the building envelope can produce large reductions in the transport of outdoor PM to the indoors, that would also lead to increases in the levels of indoor PM from indoor sources. Singer replied that the judicious use of indoor ventilation and reducing the production of indoor PM through education have to go hand-in-hand with envelope tightening. He also responded to a question about air quality in net-zero energy homes by noting that such homes also need to make judicious use of effective ventilation. The one criticism Singer had of some of these homes is that they may forgo range hoods in the mistaken belief that cooking on electric stoves does not produce UFPs, which in fact it does. A possible solution, he said, would be to use high-efficiency filters in the forced-air HVAC systems that are installed in at least some of these homes.