1.0 Introduction 

A conventional HEPA filter consists of a continuous sheet of a special paper-like, glass-fibre filter medium that is pleated into a 'vee' configuration with corrugated aluminium separators between the pleats. This forms the filter element, that is then bonded into a rigid frame using a special polyurethane compound. Standard filters are produced in a range of face dimensions and in two standard depths of nominally 150 mm and 300 mm. Another type of construction is used for 'minipleat' filters, that are produced in depths down to 50 mm. They have very close pleating of the filter medium, and manufacturers use various separation techniques to create minimal spacing between the pleats. 

High-quality filters are tested for efficiency against sub-micron particles. Each filter should be factory-tested and certified for efficiency and airflow resistance at the test flow rate. The two most common methods for determining efficiency are the British BS3928 Sodium Flame method and the US Hot DOP method. For laminar air flow (LAF), cleanroom or containment applications, filters should also be factory-tested for integrity (freedom from pin-hole leaks) using Cold DOP. 

Filters of special construction are required for corrosive or high temperature applications. Australian Standard AS4260 specifies requirements for the construction and performance of HEPA filters. 

2.0 Definition

The HEPA acronym is from 'High Efficiency Particulate Air Filter', used in earlier US military specifications. The precise definition of a HEPA varies somewhat from country-to-country, and from specifier-to-specifier in respect of the efficiency for a particular test method. A HEPA is normally defined as a high efficiency filter with pleated glass-fibre filter medium, that has a minimum efficiency of ³ 99.97% to a nominally-monodisperse 0.3µm test aerosol such as Hot DOP. 

This filter would not be suitable for a critical application, e.g. a LAF system or a safety cabinet, and would be at the bottom of today's classification of filters. HEPA filters for critical applications should be of ³ 99.99% efficiency. When considering filter efficiency to a given test method, the critical value is not so much the efficiency value, but the compliment of this value, which is the penetration of the test aerosol. For example a 99.97% filter, with 0.03% penetration, allows three (3) times more penetration of the test aerosol than a 99.99% filter with 0.01% penetration. When we consider the relative penetration allowed by a so-called MEPA or NEPA (or, 'semi-Absolute') filter compared to a HEPA, the penetration ratio is in the order of hundreds or thousands. 

The ULPA filter is an 'Ultra Low Penetration Air' filter that was developed for microelectronics cleanrooms. This filter uses a higher-efficiency medium and has higher pressure drop than HEPAs. ULPAs are not normally required outside of applications such as microelectronics and high magnification/ infrared optics.

Clyde-Apac supplies a wide range of HEPA and ULPA filters that are individually tested and certified. Australian-made Microseal HEPA filters are efficiency-tested to BS3928 Sodium Flame and are additionally scanned for leaks with Cold DOP.

3.0 Test methods

3.1 Efficiency

A close comparison of efficiency values for HEPAs tested to different test methods is invalid. For example, the Hot DOP method is size-selective @ 0.3µm, whereas the Sodium Flame test, notwithstanding that the geometric mean particle diameter is also about 0.3µm (actually 0.28µm), is not size-selective, and is looking for the penetration of particles both smaller and larger than this size. By number of particles, this method is challenging the filter with a greater proportion of smaller (than 0.3µm) particles than Hot DOP.

The Sodium Flame test is the only standardised test in Europe, and is in the EUROVENT specification. In the USA, Hot DOP is used. The above methods are the most common, and are specified in AS4260. Other methods include Paraffin Oil Mist (DIN 24 184, Test Aerosol No 1) and Uranine (AFNOR X 44011). These are used by some filter manufacturers in Europe.

3.2 Integrity and on-site testing

Filters for critical applications are also factory-tested for integrity (freedom from pin-hole leaks). The most-common method is the Cold DOP test described in AS1807.6 and 1807.7. This test is also called 'scanning', and is a challenge to the entire face of the filter.

A filter may pass the efficiency test, but could have a pinhole leak that , although not large enough to cause the filter to fail on efficiency, could allow the transmission of contaminant particles that would be unacceptable in an application such as work with HIV or sterile pharmaceutical products. Cold DOP test is also used as an on-site test for filters and their installations. 

If the application demands a HEPA filter, the requirement for filter testing in situ should be axiomatic.This is not always expressed in project specifications.

4.0 Types

4.1 Conventional

The original and most common type is the conventional 'hard spacer' filter. There are two (2) common nominal filter depths: 

5 7/8" or 149 mm 

11 1/2" or 292 mm

The 149 mm-size is typically used in LAF and cleanroom applications. The 292 mm size is used in HVAC applications, where the lower pressure drop is attractive to specifiers. Australian standards for cleanrooms (AS1386) and integrity testing (AS1807) specify that the filter face velocity should be not > 0.6 m/s, which limits the application of the 292 mm size where higher airflows may be considered .

4.2 Minipleat

These filters are of compact, low-depth construction and use very close pleating and separation of media pleats by means such as ribbons of filter medium, fibreglass threads, dimpled medium or raised of hot-melt glue. Typically, a minipleat filter of a given face size with nominal medium depth of 50 mm has similar airflow/ pressure drop characteristics to a conventional 150 mm /6" filter. This approximate 3:1/ 1:3 ratio applies to other depths.

Minipleats are limited in their application in some LAF units because they do not attenuate air and fan noise as effectively as a 150 mm filter. This type of lighter, more-compact and aesthetically-superior filter is desirable in most non-cabinet applications - particularly large LAF cleanroom systems with remote fans, and in terminal filter modules. 

5.0 ULPA filters 

Australian Standard AS 4260 specifies ³ 99.999% on ³ 0.12µm for ULPA filters. These are designated as Grade 4. The definition varies between overseas manufacturers. Some say that efficiency is ³ 99.999% on ³ 0.3µm; some say ³ 99.9999% on ³ 0.3µm; others say theirs are the above efficiency on ³ 0.12µm.

5.2 Applications

This type of filter has typically been specified only for ultra-clean manufacturing applications such as microelectronics and high-magnification optics. In a limited number of cases, ULPA filters have been specified for LAF cabinet applications. 

5.3 Media

This type of filter needs a more-efficient medium, that has significantly-higher pressure drop. This means that, to get pressure drop/flow characteristics not greatly different from a HEPA (say, not > 30% higher), much more medium must be used. This really pushes the design to the limits of the conventional type (limited in effective medium-area capacity), and into the minipleat type.

6.0 Fluid seal filters

7.0 MEPA and NEPA filters

8.0 Installations

HEPA filter installations should be designed and constructed to provide the following:

(i) An effective filter clamping and sealing system. Double-gasket 'negative-pressure' seals are required in some applications.

(ii) A means of introducing the filter test aerosol upstream of the installation.

(iii) A means of obtaining a sample of the upstream aerosol concentration and of obtaining a manometer reading.

(iv) Access to the downstream side for testing and leak repair.

(v) Prefiltration of not < 90% to AS1132.4.

(vi) Effective dehumidification - particularly in Northern Australia and Asia.

8.2 Proprietary cleanroom modules

Significant limitations of a HEPA filter installation in respect of sealing or testing procedures may not be apparent until attempts are made to commission the system. Installation of filters into proprietary cleanroom modules, such as Clyde-Apac's TFP Series eliminates many of the problems that arise with other designs. These modules are available in a range of sizes and types, as filter-only units, or in fan-assisted configuration.

8.3 Locations

HEPAs should be installed at the point of discharge into the controlled zone. As biological and particulate contamination can accumulate downstream of remotely-located filters, the installation of HEPA filters in air handling units or plant rooms is bad practice. Additionally, if these fragile filters were located with other elements of an air-handling system, they may be exposed to inadvertent mechanical damage by unskilled service personnel.

8.4 'Commissioning-mode' protection

Before a ducted HEPA filter system is operated with the HEPA filters installed, the residual dust in the system should be purged by running the fans at the highest possible flow rate with the filters removed and/ or fitting temporary 'commissioning' prefilters immediately upstream of the HEPAs.