By Jim Rosenthal, CAFS
Several years ago I wrote an article for this website comparing different commonly available filters for resistance and efficiency (at 0.3um). The article continues to be one of the most popular on our site. There just aren’t many places where one is able to compare filters from various manufacturers. Somehow it has filled a need. Hopefully, this update of the article will continue to do so.
Pictured above is our “test rig.” Actually, it is demonstration chamber that we built many years ago to show the resistance and efficiency of various types of filters. It has been used in seminars and presentations for such organizations as RSES, ACCA, BOMA, IAQA and a number of HVAC classes at Junior Colleges. It’s pretty simple – 8 feet long, filter slots for 1″, 2″ and 4″ filters, filter opening 24″X24″, fans capable of going up 750 fpm and a magnehelic gauge to measure “pressure drop” from the filters. Filter efficiency is measured with a Lighthouse Handheld 3016-IAQ with six particle sizes – 0.3 um, 0.5 um, 1 um, 2.5 um, 5 um and 10 um.
It wouldn’t pass for an ASHRAE 52.2 Test Duct. However, it can provide us with some very useful information. We tested a total of 26 different filters. The “retail” versions were obtained from a variety of outlets from hardware stores to supermarkets. In addition, a number of more “commercial” filters were included – such as 2″ filters from major manufacturers, polyester pads and ring panels.
All of the filters were tested with the same procedure. First, filters were tested at 300 fpm – both for resistance and efficiency. Then an ambient particle count was taken in the room. Then the speed was increased to 500 fpm and the filter resistance was measured. Efficiency was calculated as a percentage reduction from the ambient particle counts to the counts measured on the air exit side of the duct. The result was that all filters were treated in the same way, with the same conditions at the same time. Thus the results should provide a good comparison with each other.
(Resistance and efficiency were also compared with published data to confirm the relative accuracy of our numbers.)
Here they are:
| PD @300fpm | PD @500 fpm | Eff. 0.3um | Eff. 0.5um | Eff. 1um | Eff. 2.5um | Eff. 5um | Eff 10um | |
| 3M – MERV 13 – 1″ | 0.19″ wg | 0.29″ wg | 45% | 60% | 78% | 85% | 91% | 91% |
| 3M – MERV 5 Basic – 1″ | 0.20″ wg | 0.32″ wg | 1% | 10% | 26% | 54% | 53% | 62% |
| 3M Filtrete 1085 MERV 11 – 1″ | 0.20″ wg | 0.29″ wg | 32% | 40% | 52% | 69% | 75% | 75% |
| 3M Filtrete 1500 MERV 12 – 1″ | 0.20″ wg | 0.28″ wg | 48% | 55% | 75% | 95% | 94% | 91% |
| 3M Filtrete 1900 MERV 13 – 1″ | 0.21″ wg | 0.31″ wg | 45% | 53% | 66% | 86% | 90% | 84% |
| AAF Flanders Prepleat M8 – 2″ | 0.14″ wg | 0.21″ wg | 1% | 16% | 29% | 65% | 76% | 79% |
| AAF Perfect Pleat 2″ | 0.17″ wg | 0.31″ wg | 4% | 3% | 25% | 52% | 82% | 89% |
| ACE Hardware MERV 8 – 1″ | 0.15″ wg | 0.24″ wg | 7% | 8% | 27% | 36% | 40% | 74% |
| Aerostar MERV 8 – 2″ | 0.18″ wg | 0.24″ wg | 1% | 10% | 20% | 54% | 83% | 87% |
| Arm and Hammer 12000 – 1″ | 0.20″ wg | 0.33″ wg | 5% | 15% | 24% | 40% | 59% | 63% |
| Camfil 30/30 MERV 9 – 2″ | 0.19″ wg | 0.30″ wg | 1% | 5% | 14% | 42% | 77% | 75% |
| Fiberglass EZ- Flow – 1″ | 0.12″ wg | 0.19″ wg | 1% | 1% | 18% | 29% | 18% | 64% |
| Flanders Fiberglass 1″ | 0.10″ wg | 0.15″ wg | 1% | 9% | 11% | 27% | 50% | 50% |
| Honeywell FPR10 – 1″ | 0,27″ wg | 0.40″ wg | 41% | 59% | 76% | 79% | 86% | 87% |
| Honeywell FPR7 – 1″ | 0.26″ wg | 0.38″ wg | 23% | 39% | 63% | 62% | 92% | 86% |
| Polyester Pad – 2″ | 0.18″ wg | 0.26″ wg | 1% | 1% | 8% | 17% | 31% | 54% |
| Poly Pad – tackified MERV8 – 1″ | 0.21″ wg | 0.27″ wg | 4% | 6% | 25% | 46% | 66% | 78% |
| Polyester Ring Panel – 1″ | 0.19″ wg | 0.29″ wg | 1% | 4% | 19% | 26% | 49% | 76% |
| Precisionaire Poly – 2″ | 0.26″ wg | 0.32″ wg | 1% | 1% | 1% | 48% | 52% | 66% |
| Purafilter 2000 – 1″ | 0.19″ wg | 0.30″ wg | 1% | 11% | 15% | 59% | 75% | 80% |
| Tex-Air MERV 10 – 1″ | O.19″ wg | 0.27″ wg | 13% | 30% | 37% | 57% | 76% | 88% |
| Tex-Air MERV 10 – 2″ | 0.16″ wg | 0.20″ wg | 13% | 30% | 37% | 57% | 76% | 88% |
| Tex-Air MERV 10 – 4″ | 0.13″ wg | 0.19 ” wg | 13% | 30% | 37% | 57% | 76% | 88% |
| Tex-Air MERV 11 – 1″ | 0.21″ wg | 0.30″ wg | 26% | 44% | 55% | 62% | 79% | 89% |
| Tex-Air MERV 11 – 2″ | 0.16″ wg | 0.26″ wg | 26% | 44% | 55% | 62% | 79% | 89% |
| Tex-Air MERV 13 – 2″ | 0.17″ wg | 0.27″ wg | 46% | 50% | 56% | 71% | 89% | 91% |
Observations:
- It is not easy to get good particle counts above 5 um. The reason is pretty simple. There just are not very many of them – relatively speaking. In many cases particle counts at 0.3um/ft3 would be in the range of 400,000 while the 5um and 10um particle counts would be 600 or 400 per cubic foot. Small numbers can result in weird percentages. In these cases, I repeated the test when more larger particles were present in the room.
- Airflow velocity makes a huge difference in resistance or pressure drop (PD) of a filter. The higher the velocity, the greater the drop in pressure from behind the filter to in front of the filter. (that is what the Magnehelic measures).
- The construction of the filter affects resistance. Pleated filters have less resistance than flat panel filters because the velocity pressure of the airflow is distributed over a larger media area. So a 2″ pleated filter will have lower pressure drop than a 1″ pleated filter made of the same media. Likewise a 4″ filter will have less resistance than a 2″ filter. This is a very important concept for residential filters. Many residential HVAC installations do not have the proper ductwork or filter size to be able to function properly with a more restrictive filter. (Whenever possible going to a thicker pleated filter is a good idea.)
- Efficiency often affects resistance. Generally a higher MERV (explained here – https://www.texairfilters.com/what-is-a-merv-air-filter-testing-explained/) with a higher particle efficiency creates more pressure drop. Thus a MERV 13 filter is often more resistant than a MERV 8 or a MERV 10 filter. However, as we shall see in the next point, there are exceptions to this rule. Again, the HVAC system should be analyzed before moving up to a MERV 13 filter. If the increased resistance could be a problem for the operation of the system, go to the next highest MERV filters that are compatible with the HVAC system.
- One of the most surprising points from this analysis is the major differences in resistance between filters of different types and different manufacturers. Some pleated filters with the same or lesser MERV’s have higher pressure drops. In fact, some MERV 8 and MERV 9 filters have similar or higher pressure drops than the MERV 13 filters. All filters will increase in pressure drop in use. The “dirt” on the filters adds resistance. Some filters will increase in resistance faster than others – which brings us to our last point.
- A number of filters in this analysis have very low efficiencies in the 0.3um, 0.5um, 1um and 2.5um particle sizes. These would translate roughly into the E1 (0.3um to 1um) and E2 (1um to 3um) size ranges in the MERV system. Many of these filters are classified as MERV 8 or MERV 9. Yet, they are 2 to 7 times less efficient in these smaller sizes than other filters of a similar MERV. Here is the problem. Right now air filters are a critical component of our defense against the spread of Covid-19. Currently, there is widespread acceptance of the fact that Covid-19 is being transmitted through aerosols that can stay in the air for extended periods of time. The Covid-19 aerosol size range of greatest concern is 1um to 3um. (The largest number of particles with Covid-19 RNA is in the <1um range, but it is questionable if they are contagious.) In other words, these filters with very low efficiencies below 3um are doing very little to capture Covid-19 aerosols.
- The filters with higher efficiencies below 3um use what is known as electret media. Filters using this media have been shown to decrease in efficiency in use. This can be a significant factor for filters used for operating rooms, clean rooms and other environments where clean air is required. However, for filters like we are analyzing in this chart the decrease in efficiency is generally insignificant. (This article provides a more detailed explanation – What Happens to the Electret Charge on a MERV 11 Pleated Filter in Service? | Tex-Air Filters) The bottom line is that filters with electret media will outperform strictly mechanical filters in the critical size ranges for their entire life.
Notes:
- ASHRAE 52.2 – 2017 – Is the US standard for testing air filters. As shown above it could be improved by focusing on the needs of the user. The user wants to know how the filter will perform in their residential or commercial HVAC system. The MERV system (at least MERV 1-11) is oriented toward showing efficiency on larger particles. The problem is most of those particles do not make it to the filter. (They settle out of the air because of gravity.) See – Are Air Filter Tests Representative of Indoor Particle Counts? | Tex-Air Filters The system could be improved by providing more information on small particle (less than 3um) removal.
- The international air filter test standard – ISO16890 – is even worse. Most of the filters used in the US would be classified using a “Rural” particle distribution. This distribution is actually skewed toward larger particles – making the system even more inaccurate. This article explains the issue in greater detail and gives information on where the “rural” distribution originated – Analysis Shows ISO 16890 Uses Outdated and Inaccurate Particle Distribution Data | Tex-Air Filters
- This comparison also emphasizes the fallacy of the “anti-electret bias” in the filter industry. While this should be a concern with filters of MERV 14 and above that are used in critical areas. This should not be a concern for the types of pleated filters used by the vast majority of residential and commercial users. As one can see in the table – all electret filters are not the same. Properly made – they provide better airflow and higher efficiency on sub-3um particles. We see this in N95 masks. In order to achieve the necessary efficiency, N95 masks are generally made with electret media. Instead of forcing tests to neutralize the charge, we should be working toward tests that could help manufacturers refine the charged media technology to provide longer life.