Improved Test Methods for Electronic Air Cleaners

(Editor’s Note: For many years, the filter industry thought that electronic air cleaners – aka electrostatic precipitators ESP’s – lost efficiency due to the cells loading with dust. While this is the case with some applications, the following paper presents new evidence concerning ESP efficiency loss in a residential facility with potential implications in other applications. We present this paper in its entirety with permission and our thanks to Hal Levin, Indoor Air 2002.)


The objective of this project was to develop a fractional filtration efficiency test protocol for residential ESPs that avoids the limitations of the ASHRAE 52.2 method. Specifically, the objectives were to a) determine the change in efficiency that residential ESPs undergo in real life and b) develop accelerated laboratory methods that reasonably reproduce the real life changes. This work was conducted as part of the Environmental Protection Agency’s (EPA) Environmental Technology Verification (ETV) program for Indoor Air Products.

Results from the real life study revealed that the observed decrease in ESP filtration efficiency was due to reduced corona discharge from the ionizing wires, as opposed to dust collection on the precipitator plates. Analysis of the wires showed that they had become coated with a silicon compound, presumably silicon dioxide formed by a corona-enhanced vapor deposition process. A laboratory method to reproduce and accelerate the vapor deposition process was developed.

INDEX OF TERMS: Electronic air cleaner, Filtration efficiency, Silicon dioxide, Verification, Chemical vapor deposition


ASHRAE recently issued ASHRAE Standard 52.2, a fractional efficiency test method for ventilation filters (ASHRAE, 1999). The method measures the initial and dust-loaded efficiency of air cleaners over the particle diameter size range of 0.3 – 10 :m. Unfortunately, due to the nature of the ASHRAE loading dust and the requirement to load the air cleaners to a prescribed pressure drop end point, the method is not fully applicable to residential electrostatic precipitators (ESPs).

Under cooperative agreement with EPA, RTI has conducted several tests on 2-stage electrostatic precipitators (ESP) (Hanley 2001). The overall objective of the project is to further the development of laboratory fractional efficiency test methods for ESPs that reflect their real life performance. This work continued earlier studies on this program (Hanley, Ensor, Smith et al., 1993; Hanley, Smith, Ensor et al., 1990).

The project included in-home exposures (one, two, and three month durations) to obtain baseline real life performance data on ESPs, lab exposure to loading dust, and lab exposure to silicone oil vapor. Each exposure scenario was followed by a filtration efficiency measurement.

At the onset of the project, we were searching for an appropriate loading dust for ESPs. However, as the work progressed, it became apparent that the process controlling the performance of the residential ESP was not dust loading, but was the formation of silicon dioxide deposits on the ionizing wires. The mechanism responsible for these formations is believed to be corona-enhanced chemical vapor deposition (Davidson and McKinney, 1998).


For the real-life exposure, in home exposures of one, two and three months were performed. Due to HVAC cycling and seasonal heating/cooling demands, the two-month exposure likely had the least run time for the ESP (Table 1). Figure 1 summarizes the real-life changes in filtration efficiency that occurred. The filtration efficiency was seen to decrease significantly with use.

Table 1. Summary of in-home ESP exposure
[table "1" not found /]

* “Auto” and “On” refer to the thermostat setting for the blower.

Visual examination of the corona pattern and intensity of the ionizing wires showed that the corona was greatly diminished on the one-month and three-month cells. Figure 2 shows photographs of the corona from a clean cell (left) and from a 1-month exposed cell (right). In the photograph, both of the cells were installed in a single two-cell ESP housing. This allowed equal voltage to be applied to both cells and allows a valid relative comparison of the corona intensity for the two cells because the photographic exposure is identical for both cells. It is readily apparent that the one-month cell had significantly diminished corona relative to the new cell.


Laboratory tests with artificial loading dusts (ISO Fine) showed little or no change in efficiency. Furthermore, examination of the ESP cells showed that while the plates were becoming dust loaded, the ionizing wires remained very clean throughout the dust loading process and retained their full corona pattern and intensity. Similarly, 75 hours of exposure to a sub-micrometer salt aerosol at approximately 50,000/cc resulted in no drop in efficiency and no visible deposits on the ionizing wires.

The laboratory exposure work then focused on how to reproduce the silicon oxide formations on the ionizing wires that were seen to develop during the real-life exposures. Expanding upon earlier work (Davidson and McKinney, 1998, Chen and Davidson, 1999), a sealed silicone oil exposure chamber (61 x 61 x 140 cm) was devised. The chamber held two ESP cells, a shallow pan (43 x 61 x 3 cm) of volatile silicone oil (Dow 244) and a small fan to keep the air mixed within the chamber. The power supply for the ESP cells was mounted outside of the chamber to minimize heat build up in the chamber; high voltage leads from the ESP power supply were run to each cell. Using this arrangement, cells from two different brands of ESP were exposed in the vapor chamber for 19 hours (an overnight exposure).

After the exposure, visual examination in the dark showed that the corona pattern was greatly disrupted for both brands of ESP with the corona being very weak (barely visible) for one ESP while the other had degraded into a large number of streamers. Examination of the ionizing wires from each cell showed the presence of silicon oxide deposits quite similar to those seen on the real-life exposure cells (Figure 3, right). The filtration efficiency of the cells was significantly reduced (Figures 4).

Real-life Laboratory Simulation

Work is continuing to determine the exposure duration of the cells in the silicone oil chamber needed to represent 1-month of full time in-home use of the ESP.


It is clear that the performance of a residential ESP can be significantly degraded by the formation of silicon dioxide deposits on the ionizing wires. Thus, laboratory tests attempting to reproduce real-life effects of residential ESPs should include controlled exposure to a silicone oil vapor; traditional dust and aerosol challenges do not appear to be the primary factors affecting the decrease in efficiency ESPs experienced during in-home use. This line of investigation is still evolving and several basic questions remain to be answered. For example: What are the significant sources and concentrations of silicone vapor in homes? How widespread is the silicone issue (just a few homes or the majority of homes)? Does the silicone deposition occur in office buildings or is it limited to residential scenarios? Does this apply to ESPs used in restaurants and bars where smoking is allowed, or will a different approach be needed for smoking applications?


The following conclusions were drawn from the test results:

  • In-home use of an ESP led to significant decreases in filtration efficiency relative the ESP’s initial efficiency.
  • Decreases in the in-home efficiency were due to a decrease in corona from the ionizing wires.
  • The decrease in corona was due to silicon dioxide deposits on the wires and was not due to traditional “loading dust” processes.
  • The silicon dioxide deposits appear to be the result of a chemical vapor deposition process as silicone vapor interacts with the corona.
  • A sealed silicone oil exposure chamber appears capable of reproducing the silicone oxide deposits and, by controlling the exposure duration, is expected to be able to reproduce real-life exposure efficiency changes.


The RTI authors thank the U.S. EPA for support of the program through Cooperative Agreements CR 822870-01 and CR 826394-01.


ASHRAE, 1999. ANSI/ASHRAE Standard 52.2-1999, Method of Testing General Ventilation Air-Cleaning Devices for Removal Efficiency by Particle Size, Atlanta

Chen H., and Davidson J.H. 1999. Effect of Silicone Concentration on Deposition of Silicon-Dioxide in the Corona Discharge of Electrostatic Air Cleaners, Proceedings of the 1999 Fall Topical Conference, pp 203-210, American Filtration and Separations Society.

Davidson J. H. and McKinney P. J. 1998. Chemical Vapor Deposition in the Corona Discharge of Electrostatic Air Cleaners. Aerosol Science and Technology, 29:2.

Hanley J. T. 2001. Development of Test Methods for Electronic Air Cleaners. Presented at the Electronic Air Cleaners Stakeholders Meeting, Environmental Technology Verification Program, held at Research Triangle Institute, October 25, 2001. (Meeting minutes available at

Hanley J.T., Ensor D.S., Smith D.D., and Sparks L.E. 1993. Fractional Aerosol Filtration Efficiency of In-Duct Ventilation Air Cleaners, Proceedings of the 6th International Conference on Indoor Air Quality and Climate – Indoor Air 93, Vol. 4, pp. 169-178, Helsinki: Indoor Air ‘93.

Hanley J.T., Smith D.D., Lawless P.A., Ensor D.S., and Sparks L.E., 1990. A Fundamental Evaluation of an Electronic Air Cleaner, Proceedings of the 5th International Conference on Indoor Air Quality and Climate – Indoor Air 90, Vol. 3, pp. 145-150, Toronto: Indoor Air ‘90.

Improved Test Methods for Electronic Air Cleaners; Fall 2002 issue of Air Media
Author(s): JT Hanley; et. al.

JT Hanley1*, DL Franke1, MK Owen1, DS Ensor1, and LE Sparks2

1Research Triangle Institute, Research Triangle Park, NC 27709

2National Risk Management Research Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, NC 27711