The ubiquitous HEPA (High-Efficiency Particulate Air) filter has evolved since it was introduced 40 years ago: New materials can be used as alternatives to the glass-fiber filter medium, for example, and housings have improved for ease of testing and maintenance. But the way these filters work remains largely the same.
Four mechanisms are at play when air passes through a HEPA filter: straining, impaction, interception, and diffusion.
Straining deals with the largest particle sizes, five microns and larger. Those particles are best captured with a less-costly pre-filter, optimally located in the containment space below the HEPA filter, says Stan Klassen, manager of containment services at the Canadian Science Centre for Human & Animal Health in Winnipeg, Manitoba.
Impaction occurs when particles 1 to 5 microns in size directly hit one of the fibers in the filter medium. Particles from 0.4 to 1 microns are intercepted when they get stuck on one of the filter fibers. And diffusion captures particulates on the submicron level, which are effected by a principle called Brownian motion.
“On the submicron level, the laws of gravity get fuzzy and these particles float around and get bounced around by molecules or particles in the air,” explains Klassen. “Sooner or later, they hit one of the filter fibers.”
A HEPA filter is most efficient at trapping the smaller and larger particles, and less efficient at trapping those between 0.15 and 0.3 microns.
“Diffusion gets more efficient as the particle size gets smaller, and interception and impaction get more efficient as the particle size gets larger,” explains Klassen. “The overall result is that the efficiency goes up with larger and smaller particle sizes.”
A typical two-stage housing has five doors across the front: The first is a prefilter; the second is the first HEPA filter; the third door is the scan section, which allows testing of the first HEPA filter in place without touching it, so that the seal between the filter and the housing is also tested; the fourth door is the second HEPA filter, and the fifth is the second scan section.
One lesson Klassen learned was to install the prefilter down below, in the lab itself. Otherwise, the entire decontamination process would need to be carried out just to change the prefilter.
“Installing it below saves in initial cost because the unit is smaller, and in operating cost because maintenance requires less time,” he says.
Annual Testing is a Must
Since HEPA filters are the first line of defense in preventing air-borne contamination, it is critical to make sure they are operating properly. By definition, that means they are capturing 99.97 percent of 0.3-micron particles, but most protocols set the standard at 99.99 percent, says Klassen. The filters need to be certified annually, according to new guidelines in the BMBL 5th Edition (see accompanying article).
The first step is to take the filter off line. Long before that happens, the certifier must have communicated with the scientists who occupy the labs, as well as the maintenance staff who operate them, because the filters can be off line for as long as a week.
“Scientists work on six-month, ten-month, one-year projects,” says Klassen. “If I go up on a mechanical floor and just turn off and isolate the HEPA filter, there would be a lot of upset scientists. Let them know weeks in advance, months if you can, when their HEPA filters are scheduled to come down.”
In addition, taking even a single HEPA filter off line can have a cascading effect on the HVAC system.
“In a typical containment suite, all the rooms are referencing each other, and there is directional air flow to maintain,” he says.
Once the filter is off line, it must be decontaminated, along with the housing. Filters are tested in place because the seal between the HEPA filter and the housing must be tested as well as the filter itself.
“We use a Certek 1414RH Formaldehyde Generator in combination with a circulating fan to force the formaldehyde gas through the HEPA housing,” he says. “The Certek equipment is automated, so that we start it at the end of our work day, and the decontamination is complete the next morning when we get in.”
Formaldehyde can leave a white residue behind and is not that friendly to electronics and optical equipment that may be in the lab, says Klassen, so some certifiers might opt to use hydrogen peroxide or chlorine dioxide instead. Both are cleaner and more compatible with electronics, but the equipment is much more costly to buy and maintain.
“Some of the VHP equipment we have here costs upwards of $100,000 each,” says Klassen. “Formaldehyde decontamination can be carried out in a small containment suite with a $35 electric skillet and a few dollars worth of chemicals.”
The decontamination itself must be validated or proved with biological indicators. A test strip of harmless but hearty spores is placed on the clean side of the filter. After decontamination, the strip is incubated for anywhere from two to seven days to see if the spores survived.
“If we can kill those, we’ve proven that we’ve killed everything,” says Klassen.
It is always advisable to consult a biosafety officer to provide guidance regarding specific pathogens, and what type of decontamination would be most appropriate. For example, TSE (i.e., mad cow disease) is known to be resistant to formaldehyde decontaminations.
The decontamination should be done only by trained personnel wearing appropriate personal protection equipment, says Klassen. For example, current North American regulations do not permit the use of air purifying respirators for protection against hydrogen peroxide, so if hydrogen peroxide is used instead of formaldehyde, the certifier should wear a self-contained breathing apparatus. While the decontamination is under way, the area must be off limits to everyone but the certifier, and remain inaccessible until the air is tested.
When the decontamination is complete, the filter itself is tested, a process that takes only about 20 minutes.
“Generating 0.3-micron particles is the tricky part,” says Klassen.
An aerosol generator outfitted with a Laskin nozzle generates a fog of either Emery or DOP particles from 0.1 to 0.3 microns in size. A machine called an aerosol photometer scans the filter on the clean side to quantify any leaks.
“The leak rate is measured as a percent of the upstream concentration,” he says. “Anything higher than 0.01 percent is a failure.”
“Every square inch of the filter has to be scanned, at about two inches per second,” cautions Klassen. “So if you are hiring contractors, and they can scan a six-foot biosafety cabinet filter in 30 seconds, they are going a little quick. The seal between the filter and the housing, and the bond between the medium and the frame, also should be looked at.”
Protocols vary among industries, but generally, leaks can be repaired with silicon and then retested; no more than 1 percent of the filter surface should be patched. Gasket leaks and gel leaks can usually be repaired with silicone grease. The nuclear industry permits no repairs; if any leaks are detected, the filter must be replaced.
The filters must be decontaminated before they are disposed of, either by autoclave or incineration.
The only time bench testing of a filter is recommended is when it arrives from the manufacturer, before it is installed, says Klassen.
“Filters are most vulnerable between the manufacturer and when they get to your loading dock,” he says. “We have seen failure rates as high as 20 percent or more in a shipment.”
Klassen recommends bench testing a random sample of 10 to 20 percent of a shipment. Pay attention to the arrow on the boxes, he warns. HEPA filters should be stored with the pleats vertical. If they are stored horizontally, the pleats will sag and put pressure on the side where the media is joined to the filter. He also recommends never lifting the filter out of the box because of the likelihood of damaging the medium.
“Gently flip the box upside down and pull the box up off of the filter,” he says. “Do a quick visual examination. You can usually tell if a gasket is going to leak. If it’s got nicks in it, address that before you put it in.”
Filter Innovations and Options
The typical HEPA filter medium is made of glass fibers, as it has been for the past 40 years, which works well in most environments. But it can cause problems in the electronics industry, where some of the silica molecules come off of the filter medium and interfere with microelectronics. In that case, a Teflon medium works as a good alternative to glass.
“The medium manufacturers have progressed to the point where they can manufacture a medium to your specifications,” says Klassen. “And the technology is not standing still. There has been a lot of discussion about how HEPA filters will handle nanotechnology, with particles as small 100 nanometers or less.”
It’s not just the filter medium that is evolving, but also its configuration, says Chris Kiley, biocontainment mechanical engineer at Merrick & Company in Atlanta, Ga. Until recently, for example, most facilities used the standard HEPA filter rated at 250 feet per minute. A high-capacity filter has been developed with more filter medium compressed into the same space, providing almost twice the surface area. Another option is a “V-bed filter,” which is 1.1 inches deep with the medium in a V configuration that increases the surface area even more, bumping the rating to 2400 cubic feet per minute. Of course, higher capacity comes with a higher price: A standard HEPA filter costs about $400; the high-capacity filters are about $520 each, and the V-beds are $650 each.
By Lisa Wesel