Existing Filter Standards Filter manufacturers test their products to different standards. Three of these in common usages are: Weight Arrestance Weight arrestance is tested using the arrestance test cycle method, carried out according to the ASHRAE Standard 52.1. This test measures how much dust a filter removes, by weight. Since this test assesses how well a filter captures large particles, it can not be used to rate filters for their ability to remove respirable particles from the airstream. For example, the arrestance value can be quite high (90%) for a filter that has a mediocre performance (<10%) in removing respirable particles. Dust Spot Efficiency Dust spot efficiency is tested using the atmospheric dust spot test method, carried out according to ASHRAE Standard 52.1. This test is useful for rating how well filters screen out some respirable particles, but it is not a good indicator of filter efficiency in the submicron (<1µm) range. DOP (Di-octyl Phthalate) Efficiency The DOP efficiency test measures the ability of the filter to remove small particles. This makes it particularly suitable for rating high-efficiency filters. For instance, to be certified as HEPA, a filter has to capture more than 99.97% of 0.3-µm particles during a DOP test.   In December 1999, ASHRAE published a new filter standard, 52.2-1999, which specifies filter performance over various particle sizes. Testing to this standard produces a minimum efficiency reporting value (MERV). It is likely that most higher quality filters will soon be tested to ASHRAE 52.2.

Figure 1. Only the commercial-grade 4-inch pleated filter, labeled 95MED, reduced flows significantly.

Table 2. Activity Effect on Particle Reduction Filter Particle Reduction Activity No Activity 4-in MED 9% 13% E.PAD 9% 29% HEPA 23% 38% 1-in MED 21% 57% ESP 31% 71%

Ozone in the Air Our study found that the ESP was the most consistently effective filter system. Because these units are known to produce ozone incidentally, we tested 15 houses with ESPs to determine whether ozone exposures were excessive. The news is both bad and good. All the ESPs we tested did increase ozone in the houses, and there was no general reduction in ozone concentrations if the devices were properly cleaned. On the other hand, the ozone concentrations in the houses, although elevated by the presence of the ESP, never reached the ozone levels found in the air outside the houses during our testing. (Ozone levels in outdoor air are generally higher than levels in indoor air, because there are more external sources of ozone. Ozone is unstable, and some of it is retransformed to oxygen as it passes through the house envelope.) CMHC is continuing the research on ESPs and ozone to further clarify the situation. The current conclusion is that ESPs are good filters and, considering all factors, produce clean air at a lower cost than other filters. The incremental electrical costs of running the ESP were relatively small. The big increase in electrical costs stems from running the fan continuously. Ozone increases are measurable, but relatively small.

Figure 2 shows indoor particle concentration over 24 hours, with the most effective filter, the electrostatic precipitator, installed.

Is it worth the time and expense to upgrade a furnace filter? Will putting in a pleated filter, or an electronic filter, reduce visible dust in your house? Will an upgraded filter make it easier for asthmatics to breathe? Will it seriously block air flow and make the furnace run on the high-limit switch? Manufacturers' specification sheets on filter performance abound (see "Existing Filter Standards," p. 38), but little data exist on what actually happens when you put filters in real furnace systems in homes with real occupants.

Canada Mortgage and Housing Corporation (CMHC)--the department of the Canadian government that deals with mortgage insurance, assisted housing, and various forms of housing research--ran into this roadblock when it set out to publish guidelines for consumers on the selection of furnace filters. Some research results were available in the scientific literature from studies conducted in laboratories or hospitals. But would filters work as well in homes as they did in the lab? And what is the best method for measuring the effectiveness of filters? In previous CMHC research, consultants testing for airborne particles often returned with statements such as "Occupant activity disturbed the instruments and invalidated the particulate testing." OK, but testing only in unoccupied houses does not answer any of the above questions.

To get some answers, we at CMHC hired research consultant Dara Bowser to undertake filter testing in houses. Bowser began with extensive preliminary work in his own home, testing filters in his furnace and ducting system, and perfecting means of making accurate flow and particle measurements. Bowser tested ten filters, ranging from a $2 throwaway fiberglass filter to $700 electrostatic precipitators (also known as ESP, plate and wire, or electronic filters), as well as unusual variants such as bypass filters using high efficiency particulate arrestor (HEPA) or other media.

When he was satisfied with the quality of his measurement devices, Bowser took the show on the road to five other houses. In each house, he installed each of the filters in the furnace ductwork and measured its effects on indoor air quality (IAQ), furnace fan electric consumption, and other factors. Because researchers had identified the production of ozone by ESPs as a potential hazard, he took ozone measurements in 15 houses with ESPs, both before and after cleaning the filter elements (see "Ozone in the Air," p. 40).

All this testing took place during a Canadian winter. Because most people wisely close their doors and windows during this season, the entrance of outdoor particles is at a minimum and a filter's impact on IAQ is at a maximum. When considering how these results apply to your house and your climate area, remember that performance may vary, as they say, depending upon conditions outside and inside your house. Most importantly, these were all new filters, and the test periods consisted of days, not months. The effects of filter loading, which will improve efficiency but degrade air flow, cannot be determined using the data generated in this study.

Bowser collected particle data with a laser particle counter connected by sampling lines to five monitoring points. This device measured particle counts in the living or family room, in one bedroom, and outside the house, as well as at two sites in the furnace ductwork--immediately upstream and downstream from the filter. The data recording was comprehensive and revealing: For instance, any activity that prompted high concentrations of resuspended particles would be reflected in a concentration peak shortly after in the duct system.

The particle counter also broke down the particles collected into various sizes, so filter efficiency can be calculated for different particle diameters. Respirable particles, those smaller than 2.5 µm (one millionth of a meter), can cause respiratory problems. Particles between 2.5 and 10 µm are usually trapped in the nose or throat, generally preventing their entry into the lungs. Particles larger than 10 µm are typically not considered a health threat, as they rarely enter the lungs. Visible dust particles are usually 10 µm or larger; the very particles that you can see--and complain about--are therefore not likely to be a health problem. Also, furnace filters do not effectively remove visible dust, because it is not generally suspended long enough to be entrained. However, the presence of visible dust may signal high concentrations of respirable particles, since visible dust can create or promote the presence of suspended respirable particles.

Spare the Expense

As project manager, I was one of the first to review Bowser's findings, and a few of the results immediately leapt out at me. First, it is clear that, unless a client is willing to run the furnace fan continuously, there is little point in upgrading a furnace filter. Suppose you replace a furnace filter that removes 5% of the particles with one that takes out 50%--a tenfold improvement. A furnace operating only on a call for heat from the thermostat will have the fan running about 20% of the time during the heating season. The particle removal efficiency is the product of the filter efficiency and the fan run time. In this example, the 5% efficient filter at 20% run time drops to 1% particle collection over the heating season. The 50% efficient filter, which costs significantly more, has only a 10% collection efficiency over the heating season. Running the furnace fan continuously ensures that the filter is continuously working to remove particles, but noise, comfort, and cost must be considered as well. In many homes, leaving the fan on will add a couple of hundred dollars to the electrical bill over the course of the year. There may be people for whom it is reasonable to spend more than $200 to upgrade their filter and continuously run their furnace fan--those with asthma, allergies, or other health problems, for instance--but many people will decide that the air quality improvements are not worth the added expense.

Upgrading a furnace fan is an attractive option if a homeowner is considering continuous fan use. Efficient furnace fans, such as electronically commutated motors (ECMs), are being sold, but very few are being retrofitted to existing furnaces, because the costs are excessive. In Canada such a retrofit approaches CAN$1,000, although there is no real reason for this high a premium over ordinary motors. CMHC surveys of standard furnace fans show power consumption in the order of 400-800 watts, with only a small decrease on low-speed operation. A good fan motor should reduce consumption to a more reasonable 100-200 watts.

Furnace Flow Unaffected

The second eye-catching result is that the furnace fan produced similar air flow rates, no matter which filter was installed. Granted, the filters were new and clean for the short duration of testing--circumstances that may not reflect common in-service conditions. Although the filter air-flow resistance was higher with certain filters installed, the furnace fan was able to overcome any increased resistance and produce very similar air-flow rates. For example, compare all flows in Figure 1 to the 1-inch fiberglass throwaway filter (labeled ORD in the graph). Only the commercial-grade 4-inch pleated filter, labeled 95MED, reduced flows significantly.

The bypass filters, either the turbulent flow precipitator (TFP) in this figure or the HEPA, may actually increase air flow in the duct. These filters are in a bypass duct loop, where air is taken from the return air duct, passed through the filter, and directed back to the return air. The amount of air taken from the return air duct will affect the apparent filter efficiency. For instance, a 100% efficient bypass filter that treats only 50% of the return air can achieve only a 50% efficiency rating on the furnace duct air flow. Bypass filters use their own supplementary fan. Bypass filters will never block furnace air flows, but the second fan increases the operating costs.

These test results lead to the conundrum that improved furnace filters do not make a big difference in particle exposure in most houses, even though the filters usually met their rated efficiencies. The cumulative results are shown in Table 1. Comparing Table 1 with Table 2, it can be seen that the improvement in house air particle concentrations is nowhere close to the high filter efficiencies. How is this contradiction possible?

Table 1 reveals that filters did indeed meet their stated efficiencies. Sampling particle concentrations before and after the filter in the duct system produced a continuous measurement of filter efficiency. The results shown in Table 1 are for PM10, or all particles less than or equal to 10 µm, and for PM1 (<1 µm). Efficiencies for other particle size cuts are available and show a roughly similar pattern.

The answer to this conundrum can be seen in Figure 2, a graph of particle concentration over 24 hours, in this case with the most effective filter, the ESP, in place. Look at the hours when there is activity in the house. It is clear that particle concentration will greatly increase during these hours, and that these concentrations will have the greatest effect on indoor particle exposure. A furnace filter many meters away down a duct will not make a difference to the "personal cloud" that occupants are creating just by getting out of bed, walking across the carpet, or making toast. These particle-generating activities overwhelm the capability of furnace filters to clean house air. If we look at a parallel graph for a period with no furnace filter, we see similar particle concentration patterns, but the peaks last a bit longer and the concentrations during inactive periods are significantly higher. If good filters cause lower particle concentrations at night, this could be seen as a big advantage. However, the inactive periods, at least in our study, have such low particle concentrations that they do not affect personal particle exposure to any great degree.

So where does this leave us? Filters can do a good job of cleaning duct air, but they cannot protect occupants from significant exposure to indoor particles. Furthermore, most homeowners connect particle exposure to observable surface dust, and most surface dust has little to do with respirable particles. The amount of surface dust present is more a function of dust sources than of filter efficiency. If homeowners expect that an improved filter will greatly reduce surface dust, they are likely to be disappointed. One way to explain this is to compare the mass or weight of a vacuum cleaner bag (surface dust) after a couple of months to the accumulated dust on a good filter. The dust in the vacuum bag will far exceed the collection of tiny "dust" particles on the furnace filter.

Upgrading a furnace filter and running the furnace fan continuously will cost a significant amount, generally $200 or more per year depending on the cost of the filter and the replacement schedule. One has to balance these increased costs against the actual improvement in particle reduction. Furthermore, if conditions in the client's house are much different from the test conditions (if, for example, the client likes to keep the windows open), the filter will have less of an effect on IAQ.

If filters can't remove dust and particulates, what can be done? CMHC recommends instituting a comprehensive program to reduce all types of particle and dust entry. CMHC has not yet measured the effectiveness of this strategy, but most of the following recommendations should reduce particle sources, because tracked-in dust, for instance, will include both the large, visible particles and smaller, respirable particles. In addition, large particulate can be crushed and resuspended as small particles. To confront the particulate problem, then, homeowners have to eliminate all dust sources by:

• removing footwear on entry;
• keeping major dust generators (smoking, pets, and so forth) out of the house;
• keeping dust collecting surfaces (open shelves, carpets, upholstered furniture) to a minimum;
• vacuuming diligently and frequently with an efficient vacuum cleaner (a HEPA vacuum or a central vacuum work best at not reblowing household dust back into the home);
• reducing the entry of particulate-laden outdoor air by closing windows, improving house airtightness, and installing an intake filter on the air supply; and
• using as effective a furnace filter as the homeowner's budget permits.

Based on our research, a furnace filter upgrade would have to be combined with these steps to significantly decrease indoor respirable particle exposure. CMHC is continuing to research the effects of furnace filters on IAQ. In this new research, we will be delving deeper into issues that arose during this first round of testing, such as particle resuspension by bare floors versus carpets.

Don Fugler is a mechanical engineer who manages research projects for the Canada Mortgage and Housing Corporation.

For more information: "Clearing the Air: Filters for Residential Forced-Air Systems," HE July/August '96, p. 14. The research report, Evaluation of Residential Furnace Filters, and the consumer publication, About Your House CE22: Your Furnace Filter, are available from CMHC by phoning (613)748-2367. The consumer publication is free. It can also be downloaded from the CMHC Web site at www.cmhc-schl.gc.ca/cmhc.html. The research report is CAN $12.95 plus shipping (about US $18 in total).

Table 1. Filter Performance and Cost
Code	Generic Description	Manufacturer Claimed Performance	Results for Test House #1: Upstream/Downstream Efficiency
			Reduction in PM1	Reduction in PM10	Cost (CAN$)
ORD	Ordinary furnace filter	Change monthly; made from recycled material	-2%	-3%	$2
1-in PLT	25-mm pleated media filter	Eliminates 92% of airborne allergens; 2-3 times more efficient than standard filters	4%	6%	$7
1-in MED	25-mm pleated media	99% of particles in air consist of high quality microparticles; 20 times better than ordinary filters; 7 times better than ordinary pleated filters	25%	32%	$22
PAS.E	25-mm passive electrostatic	93% average arrestance	6%	10%	$100
E.PAD	Electronic charged pad	Removes 98% of submicron particles on multipass basis; single pass @ 0.3-0.5 µm = 33-75%; @ 0.5-1.0 µm = 75-95%; multiple pass @ 0.3-0.5 µm = 97%; @ 0.5-1.0 µm = 98.6%	11%	15%	$150 purchase, $35/year in operating costs
4-in MED	100-mm pleated media	32% average dust spot efficiency; 92% average arrestance	19%	19%	$500 installed, $60/year in operating costs
95MED	95% dust-spot pleated media	95% average dust spot efficiency; 99% average arrestance	56%	58%	$400 installed, $200/year in operating costs
ESP	Electronic plate & wire type	75% average dust spot efficiency; 98% average arrestance	94%	95%	$700 installed; no annual costs if homeowner does the cleaning
TFP	Turbulent flow precipitator	0.5 µm = 84%; 0.7-0.9 µm = 87%; 1 µm = 92%; 2-3 µm 95%; 5+ µm = 99% (according to ASHRAE 52.1)	21%*	22%*	$1,000 installed, $40/year in operating costs
HEPA	HEPA	99.97% D.O.P. 5	1%*	51%*	$2,200 installed, $93/year in operating costs
*The TFP and HEPA filters are bypass filters handling 22% and 51% of the system air flow respectively. Efficiencies are measured in total air flow, so the implied efficiency directly though the bypass filter approaches 100% for both types.

Home Energy can be reached at: contact@homeenergy.org
Home Energy magazine -- Please read our Copyright Notice