Table 1. Filters selected for testing.

Filters were originally conceived to protect heating and cooling equipment, for example to prevent large particles from clogging the air passages of coils. The old, familiar fiberglass filters do a fair job protecting equipment, but do little to enhance IAQ.

Over the past several years, energy efficiency and green programs have begun to adopt requirements for filters that can remove the smaller particulates that cause allergic reactions and other health problems. The Energy Star Indoor Air Package, Department of Energy (DOE) Builders Challenge, Leadership in Energy and Environmental Design for Homes, and EarthCraft programs all call for a MERV rating of 8 or better.

MERV (Minimum Efficiency Reporting Value) is a measure of the efficiency with which filters remove particles of specific sizes. The test protocol for determining MERV ratings is described in the American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE) Standard 52.2-2007, Method of Testing General Ventilation Air-Cleaning Devices for Removal Efficiency by Particle Size.

A MERV 8 filter can remove particles down to 3 microns in size at an efficiency of 70 percent or greater (1 micron is about 0.04 thousandths of an inch). According to the standard’s application guidelines, particles in this size range include pollen, dust mites, mold spores, hair spray, powdered milk, and of course, snuff. Higher MERV ratings (13-16) are needed to remove bacteria and smoke particles. At the top of the scale are high-efficiency particulate air filters with MERV ratings from 17 to 20, which can filter out particles smaller than 0.30 microns, including some viruses.

It stands to reason that the ability to filter smaller particles would come with the penalty of increased resistance to airflow. Could the shift to better filters mean that they could cause problems with inadequate airflow or greater fan energy use? Are all high-MERV filters equal, or do some have less pressure drop than others? How much better are 2- and 4-inch-thick pleated filters than 1-inch filters? Are larger ducts required to offset the added pressure drop of the filter?

With a pat on the back from our Building America program sponsors, we decided to run some tests to answer these questions.


Thirteen filters (listed in Table 1) were selected for testing. The objective was to choose filters that homeowners would be likely to purchase to replace the filter provided by the builder, so we selected brands and models that are found in most big box and chain retail stores, and that represent a variety of MERV ratings and thicknesses.

One fiberglass filter (Ace 30 Day) was chosen as a reference. Its rating was not listed and it was assumed to be MERV 2. All filters except the Ace 30 Day and WEB Lifetime filters have pleated media. All but one of the filters (a 2-inch MERV 13 product) were purchased from chain retail stores.

We decided to limit our tests to filters having outside dimensions of 16 inches by 25 inches. This is a commonly available size that allowed us to compare a wide range of products. Standard 52.2 specifies that filters are to be tested at 492 feet per minute (fpm) face velocity, which equates to 1,367 cfm for the 400-square-inch face area. Applying the 400 cfm per ton rule of thumb, this size would be appropriate for 3- to 3.5-ton air conditioners.

Our testing evaluated filter pressure drop and blower motor energy over a range of airflow rates for each filter type. The test apparatus and measurement of standard airflow was based on ASHRAE Standard 41.22, and used the equipment diagramed in Figure 1. The apparatus included a calibrated nozzle box with integral pressure-balancing fan, pressure sensors for measuring airflow and filter pressure drop, temperature and rh sensors to normalize airflow to standard conditions, and power monitors for measuring blower energy. The filters were attached to a typical air handler that was coupled to the nozzle box. The air handler has a 12-inch-diameter by 10-inch-wide blower wheel.

Figure 1. Test apparatus.

We tested the filters using a standard ¾-hp permanent split capacitor (PSC) motor and a 1-hp GE ECM™ 2.3 variable-speed motor connected to the same blower. Most economy-priced furnaces use PSC motors. These respond to increasing flow restriction by moving less air with little change in power.

The airflow reduction that occurs with PSC motors can affect the amount of power consumed by the compressor. It can also result in a decrease in cooling capacity and extended run times, therefore greater energy use.

Many higher-priced furnaces and heat pumps use electronically commutated motors (ECMs). These respond to increasing flow restriction by maintaining a fairly constant airflow rate but at the expense of increased power. Testing with the ECM allowed us to dial in the airflow rate, whereas the flow rate for the PSC motor was whatever it could deliver at its three different tap settings.

Figure 2. Airflow rate vs. MERV rating, PSC motor tests.

We wanted to determine the incremental impact of the filter on airflow and fan power, beyond what the cooling coil and ductwork would contribute. California field studies indicate the median pressure drop for residential duct systems is 0.18 inch, and for cooling coils is 0.27 inch, resulting in a static pressure downstream from the filter of 0.45 inch. To simulate ducting, we adjusted the pressure-balance fan until the airflow and pressure drop fell on a typical flow vs. pressure curve, developed by assuming 1,400 cfm would produce 0.45 inch of static pressure at the discharge of the air-handling unit. Thus, at different measured airflows, the virtual duct and coil pressure drop followed this curve, just as it would for a normally installed system.

With the ECM installed, filters were evaluated at nominal airflows of 750, 1,000, 1,250, and 1,500 cfm. These airflows were set using a control that generates a pulse width modulation (PWM) signal that is proportional to the airflow the motor is programmed to deliver. The PSC motor was tested at each of its three tap settings. Second-order, polynomial, curve fits of each data set were used to interpolate filter pressure drop and power, at the standard velocity of 492 fpm (airflow rate of 1,367 cfm).

Figure 3a. Filter pressure drop at 492 fpm.


Q. Do high-MERV filters reduce airflow?

A. Though there are considerable differences in how different filters affect airflow, Figure 2 shows that there is a definite trend toward lower airflow with higher-MERV filters for systems using PSC motors. Over the range of filters tested, there was no such correlation between airflow and filter MERV rating for the ECM, which can maintain constant airflow over a large range of external static pressure.

Q. Do some high-MERV filters have higher pressure drop than others? Is pressure drop lower for thicker pleated filters?

A. As shown in Figure 3a, there was a significant variation in pressure drop, particularly amongst the five MERV 8 filters tested, and not as close a correlation between pressure drop and MERV rating as we expected to see. The WEB Lifetime, a washable electrostatic filter, was the best high-MERV performer. Its MERV 8 rating is surprising given the relatively open appearance of the media compared to replaceable filters.

Figure 3b. Fan motor power for ECM motor at 492 fpm.

Q. Do the reduced airflow rates of high-MERV filters in PSC systems affect compressor energy use?

A. Air conditioner and heat pump rated performance is a function of airflow over the indoor coil. Lower airflow can decrease capacity and consequently the EER or HSPF of the system. We looked at representative manufacturers’ performance curves and found that, over the less than 200-cfm reduction in airflow observed between MERV 2 and MERV 13 filters (see Figure 2), the impact on compressor power is almost insignificant.

However, filter pressure drop can more than triple when the filters are fully loaded with dust particles, so loaded filters probably affect system performance. The deeper-pleated filters allow dirt to be spread over a larger surface area, reducing pressure drop and/or allowing less-frequent filter changes before system performance is significantly affected.

Q. What can be done to mitigate the added pressure drop of high-MERV filters?

A. Friction contributed by coils, ducting, and filters is additive, so reducing the friction of one component will offset the added friction of another. When completing ACCA Manual D for duct sizing, use a realistic filter pressure drop. If using a MERV 8 or higher filter, a pressure drop value of at least 0.5 inch should be entered for the filter. In one Manual D example we ran, we found that increasing the filter pressure drop assumption from 0.2 to 0.5 inch increased the required size of some of the ducts.

Oversizing the filter to decrease the face velocity can substantially reduce pressure drop, provided the equipment can accommodate larger filters. Figure 5 shows the pressure drop vs. face velocity curve measured for the Filtrete 600. At the standard 492-fpm face velocity, the static pressure is about 0.32 inch wc. Increasing the filter size by 25 percent decreases the static pressure to 0.20 inch, and doubling the size of the filter will reduce it to about 0.07 inch.

Figure 4. Fan power vs. filter pressure drop at 492 fpm.

Q. How do measured fan power values compare to industry standards and field data?

A. At the 1,367-cfm standard airflow used in these tests, PSC motor power ranged from 0.502 to 0.523 Watts/cfm; ECM power ranged from 0.347 to 0.390 Watts/cfm. In calculating air conditioner SEER ratings, AHRI assumes 0.365 Watts/cfm of fan energy (unless data are supplied by the manufacturer).

California Title 24 standards apply a more-realistic 0.58 Watts/cfm prescriptive fan energy value. Median values obtained from Lawrence Berkeley National Laboratories tests of 141 furnaces were 0.44 Watts/cfm, and test data from new California homes yielded a median value of 0.52 Watts/cfm. Irrespective of the measured impacts of high-MERV filters on system energy use, SEER ratings appear to significantly underestimate the contribution of fan power.

California field test results and Title 24 default Watt/cfm values are reasonably consistent with our laboratory tests of the PSC fan with clean filters, which are both considerably higher than the 0.365 Watt/cfm default value. As high-MERV filters become more prevalent, the DOE should consider establishing a default value that is at least consistent with clean MERV 8 filters. More research is needed to look at the energy and airflow impacts of dirt-laden high-MERV filters.

Figure 5. Pressure drop vs. filter face velocity, Filtrete 600.


Accurate measurement of the airflow of ECM-powered fan systems takes an immense amount of patience. If a pressure-balance fan is used to adjust the static pressure seen by the blower, in particular, the motor reacts dynamically to the changing pressure and interacts with the pressure-balance fan, making it difficult to zero in on a particular set of pressure conditions.

ECMs also have harmonics that might confuse some power monitors. We found fairly good agreement between a precision power monitor and an inexpensive meter. Interestingly, using multiple wraps around the current transformer to improve measurement resolution produced erroneously low power measurements.

Pressure drop and system performance are definitely worthy of consideration when applying high-MERV filters, and duct design and filter sizing should be considered in the design process. However, if no accommodations are made for the greater pressure drop of high-MERV filters, airflow and energy penalties are not likely to be severe, at least until the filter is loaded with dirt. More study is needed in this area.

Care must be taken to reduce register and grille, duct and plenum, and filter pressure drops to ensure a sustainably efficient system. Making filters accessible for easy replacement and providing controls that tell homeowners when a filter replacement is due will help eliminate problems such as clogging and filter collapse that are more likely to occur with more-efficient filters.

As we have shown, there is a wide variation in pressure drop performance, particularly amongst the MERV 8 filters; inadequate product labeling makes informed choices still more difficult. Some products do not list the MERV rating, and none list standard pressure drop on their labels.

Although ASHRAE Standard 52.2 does not address the energy implications of filtration, there is growing interest in the subject. A recent ASHRAE paper presented two methods for rating the energy performance of filters, and we may expect an expansion of Standard 52.2, or development of other standards that will incorporate filter energy ratings as emphasis on energy efficiency grows.

Incidental to the filter evaluations, the data reveal the significant energy-savings potential of properly designed systems with variable-speed ECMs. However, these motors will degrade system performance if there is excessive pressure loss in filters, coils, and ducting.

Some furnaces are beginning to incorporate a newer, nonvariable-speed, brushless permanent magnet motor, the GE/Regal Beloit ECM X13. This motor should use about the same power as the ECM 2.3 version tested, but since it does not regulate airflow, for good or bad, its power use will be relatively unaffected by filter pressure drop associated with a high-MERV rating or dirt accumulation.

Publication date:10/11/2010