United States EPA-600 /R-93-177
Environmental Protection
Aa<">cy September 1993
4>EFA Research and
Development
CONTRIBUTION TO
INDOOR OZONE LEVELS
OF AN OZONE GENERATOR
Prepared for
Office of Air and Radiation
Prepared by
Air and Energy Engineering Research
Laboratory
Research Triangle Park NC 27711
-------
EPA-600/R-93-177
September 1993
CONTRIBUTION
TO INDOOR OZONE LEVELS
OF AN OZONE GENERATOR
Final Report
by
Raymond S. Steiber
U.S. Environmental Protection Agency
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711
Prepared for:
U.S. Environmental Protection Agency
Office of Research and Development
Washington, D.C. 20460
-------
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency, and
approved for publication. Approval does not signify that the contents necessarily
reflect the views and policy of the Agency, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
ABSTRACT
Ozone generators are claimed to provide a way to clean indoor air.
The amount of ozone generated by these devices is a major concern.
A study of a commonly used commercially available unit was
undertaken to determine the impact of the ozone generator on indoor
ozone levels. Experiments were conducted in a typical mechanically
ventilated office and in a test house. The ozone generated by the
unit and the in-room ozone concentrations were measured. The
results showed that, when the unit was operated at the
manufacturer's recommended setting, it generated little if any
ozone. The indoor concentrations in this case were not
significantly above natural background. When operated at the
maximum setting the generator produced large amounts of ozone. In
this situation the indoor ozone concentration exceeded 100 ppb in
well ventilated spaces and nearly 1 ppm in poorly ventilated
spaces. When the ozone generator was turned off, ozone levels
quickly returned to background. No measurements were made to
determine the effect of the device on other aspects of indoor air
guality, such as the elimination of volatile organic compounds.
-------
CONTENTS
Page
Abstract ii
Tables iii
Metric Conversions iv
Introduction 1
Test Group 1 2
Test Group 2 3
Test Group 3 4
Conclusions 5
References 6
Appendix. Quality Evaluation Report 7
TABLES
1 Test Conditions (Test Group 2) 3
2 Test Conditions (Test Group 3) 4
3 Ozone Generator: Settings vs Ozone Output 6
111
-------
METRIC CONVERSIONS
Nonmetric units are used in this report for the reader's convenience.
Readers more familiar with the metric system may use the following
factors to convert to those units.
Nonmetric Multiplied by Yields Metric
ft 0.3048 m
ft2 0.0929 m2
in. 2.54 cm
IV
-------
INTRODUCTION
Ozone generators are claimed to help purify indoor air and
remove odors and other pollutants. Large industrial scale ozone
generators have been found useful in ameliorating unpleasant odors,
particularly in fire-damaged buildings. Ozone generators are also
claimed to be effective in combatting molds and bacteria. There
are no published data supporting these claims.
Although manufacturers of ozone generators claim that their
units do not increase indoor levels of ozone, they offer no
supporting data other than testimonials. This study was undertaken
to determine the impact of a commercial ozone generator on indoor
ozone levels under typical conditions. No measurements were made
to determine the effect of the device on other aspects of indoor
air quality, such as the elimination of volatile organic compounds.
The unit selected for test was an ozone generator designed for
use in the home or office. It has two user-adjustable controls:
one that varies the speed of the circulating fan and another that
controls the output of the ozone.
The unit was tested under three different sets of conditions.
Test Group 1 was a 24-hour test under normal operating conditions
in a well-ventilated room. Test Group 2 also took place in a well-
ventilated room, but this time only the higher ozone generator
settings were used. Test Group 3 was conducted in a poorly
ventilated room. Here the emphasis was on what happened when the
ozone generator was left running overnight with no air circulation
but normal leakage around the door and through the walls. In
running these tests, our chief interest was in the amount of ozone
being contributed to room air and not the ozone levels at the face
of the instrument. However, these were also measured (see Table
3).
At this point it might be well to insert a few words on how
room background levels were determined. Obviously background
levels could not be measured when a test was in progress, yet at
the same time results could be seriously affected by this number,
particularly at the lowest concentrations. In any indoor setting
where there are no secondary sources of ozone, such as actually
operating copiers, the chief source of ozone will be outside air
brought into the building through natural leakage or mechanical
ventilation. Weschler et al.1 have observed that there is roughly
a 2:1 correlation between outdoor and indoor or.one levels: that is
to say, if there is a 50 ppb concentration of ozone outside, indoor
concentrations will measure around 25 ppb. Insofar as these tests
are concerned, this is important not so much in terms of the
numbers themselves as it is in the manner in which indoor air
rigorously tracks outdoor air. Although the absolute
1
-------
concentrations may vary from day to day, the rise and fall of ozone
levels over the course of a day, given warm, sunny weather, is
absolutely predictable. In the early morning (before 8 am) ozone
concentrations will hover between 0 and 2-3 ppb. As the day
progresses, the levels will gradually rise, reaching their zenith
sometime between noon and 1:30 pm. On cloudy days the ozone levels
will remain at 0-2 ppb.
Now how does this relate to determining background indoor
ozone levels? If a background reading of 2 ppb is measured before
a test begins, and another background reading of 20 ppb is measured
after it ends, then a straight line increase from 2 to 20 ppb over
the course of the test can be inferred and subtracted sequentially
as background. On the other hand, if the pre-test reading in the
afternoon is 40 ppb and the after-test reading several hours later
is still 40 ppb, then it can be assumed that no significant
excursions have occurred. Imperfect as this system of background
correction may seem, it produces surprisingly consistent results,
particularly when checked against a standard ozone generator, such
as the TECO Model 565 ozone generator, the instrument used to
calibrate EPA's air network ozone detectors. As an additional
hedge against background excursions, many readings were taken
during these tests, instead of only a few, so that the overall
trend could be observed.
TEST GROUP I
Test Group 1 was run in a mechanically ventilated office
with continuously operating air ducts in the ceiling. The room
area was approximately 200 square feet* and there was a sealed
window at one end and a hallway door at the other. The unit was
placed on top of a 5-foot high filing cabinet located 3 feet from
the hallway wall. It was turned at a slight angle in order to give
coverage to the entire room. As in all these tests, measurements
were made with a Model 560 Thermo-Environmental Ozone Analyzer.
This instrument had recently been calibrated by personnel of the
Research Triangle Institute, located in Research Triangle Park,
North Carolina (see Appendix for details). The Model 560 was
placed on top of a 3-foot high table 6 feet 10 inches from the
ozone generator. Background ozone levels in the room
ranged from a few parts per billion (ppb) in the early morning
hours to over 40 ppb by mid-afternoon. By far the greatest
contributor to these background levels was the effect of sunlight
on atmospheric oxygen. When it was cloudy during any of the tests,
background ozone levels fell, when it was sunny they rose, and when
it was sunny with visible air pollution they rose dramatically.
Test Group 1 consisted of two experiments: one that lasted 29
*Readers more comfortable with the metric system may use the
factors listed on page iv to convert to those units.
2
-------
straight hours, during which the ozone generator unit ran
continuously, and a second in which the ozone generator unit was
shut off. For the first experiment both the fan knob and the
generator knob were moved to the medium setting (12 o'clock). This
is slightly above the manufacturer's recommended setting for normal
use (11 o'clock). The ozone concentration 14 inches from the face
of the unit is a continuous 5 ppb with the dial set at 12 o'clock.
Measurements were taken every 30-40 minutes except during the
nighttime hours when no one was present and the room was secured.
At no time during the 29-hour test did the ozone levels contributed
to room air by the ozone generator rise to the point where they
could be distinguished from the background; nor did room ozone
levels fall when the unit was shut off for the second test.
TEST GROUP 2
This group of experiments was also conducted in a
mechanically ventilated office with the same configuration and area
as the office used in Test Group 1. The test conditions for this
group of tests were the same as for the first except that much
higher settings were used on the ozone generator. Four separate
tests were run under the conditions listed in Table 1.
TABLE 1. Test Conditions (Test Group 2)
TEST #
1
2
3
4
TIME
(hr)
2.0
22.5
24.0
43.0
FAN SETTING
(o'clock)
1
3
5*
Off
GENERATOR SETTING
(o'clock)
1
3
5*
Off
*Maximum setting
The average concentration of ozone contributed to room air during
Test 1 was 12.2 ppb and during Test 2 was 13.3 ppb. The
slightness of the difference in readings between Tests 1 and 2 is
probably indicative of the difficulty of taking measurements at
such low concentrations vis-a-vis the background and the non-
linearity of the controls on the ozone generator. During Test 3
ozone readings were much higher, averaging 100 ppb. All the above
averages have been corrected for the background.
Test 4 was an attempt to track room background across a
number of different meteorological conditions. Concentrations
during the period of its run ranged between 10 and 22 ppb.
3
-------
TEST GROUP 3
Test Group 3 was a repetition (with some modifications) of
all the previous tests, only in a closed room without mechanical
ventilation. This room had an area of 90 square feet and was
roughly the same size as a small bedroom. It had a single door at
one end and a closed window at the other. The background ozone
averaged 5.6 ppb in the daytime and 1.5 ppb in the early morning.
A total of six tests were run. The conditions are given in Table
2.
TABLE 2. Test Conditions (Test Group 3)
TEST #
1
2
3
4
5
6
TIME
(hr)
1.0
0.5
1.0
1.5
22.5
1.0
FAN SETTING
(o'clock)
12
5*
3
5*
5*
Off
GENERATOR SETTING
(o ' clock)
12
12
3
3
5*
Off
*Maximum setting
During Tests 1 and 2 no concentrations of ozone distinguishable
from the background were detected in the room. During Test 3 an
average of 20.1 ppb was detected. When the fan speed was increased
during Test 4, ozone concentrations more than doubled for an
average of 43.1 ppb. During Test 5 both fan and generator settings
were at their maximum. Of all the tests that were conducted, this
was the only one in which room concentrations outran the amount
being produced by the ozone generator; that is, ozone was being
produced at a rate higher than it could either be reacted or
dispersed by passive air flow. Starting at an initial
concentration of 150 ppb during the first few minutes of the test,
ozone levels rose to a high of 814 ppb, then leveled off between
600 and 700 ppb. As the level of these concentrations seemed to
vary according to the frequency with which the door was opened, it
was decided to let the unit run overnight with the door secured.
By 7 o'clock the next morning room concentrations had risen to
996.5 ppb. At those levels it was decidedly unpleasant for test
personnel to enter the room, and eye, throat, and olfactory
irritation reached the point where it was nearly impossible to
breathe.
The last test was an attempt to see how quickly high levels
4
-------
of ozone would dissipate from a closed room when their source was
eliminated. At the beginning of the test the ozone concentration
was 758.5 ppb. Within an hour it had dropped to 2.0 ppb with most
of the decay occurring in the first 10 minutes. However, for some
time afterward, a not unpleasant ozone-type odor persisted in the
room. This odor is best described as similar to that of sheets and
towels when they are fresh out of the drier, only much stronger.
Apparently this is not the odor of the ozone itself, but of some
typical reaction product.
For those who might be interested, Table 3 lists the ozone
output in parts per billion for six selected generator settings.
To obtain these readings, the intake nozzle of the analyzer was
held 14 inches from the face of the generator. The generator fan
was set at medium (12 o'clock on the dial), and all readings were
corrected for the background.
CONCLUSIONS
When operated at the recommended setting for normal use (dial
set at the 11 o'clock position), the ozone generator contributes
little or no ozone to the background air levels. At the highest
setting, however, it can contribute as much as 100 ppb to a well-
ventilated room and 10 times that much to a poorly ventilated one.
Our personal experience shows that at that level the air becomes
difficult to breathe and there is some irritation of the eyes and
the mucous membranes, perhaps due to excessive drying.
-------
TABLE 3. Ozone Generator:
Settings vs Ozone Output
SETTING CORRECTED READOUT
(o'clock) (PPb ozone)
7 0.0
9 2.0
11* 3.0
12 5.0
1 8.0
3 69.0
5** 520.0
*Manufacturer's recommended setting
**Maximum setting
References;
1. Weschler, C.J., Shields, H.C., Naik, D.V. (1989), "Indoor Air
Exposures", APCA Journal 39:1562
-------
APPENDIX:
QUALITY EVALUATION REPORT
This study was designed to assess qualitatively the amount of
ozone generated by a commercially available product claimed by its
manufacturer to purify indoor air and remove odors and other
pollutants. Since validated test procedures and comparable data
were unavailable, planning for the study was restricted to
designing the test system and considering fundamental research
needs, and a quality assurance plan was not prepared. In
accordance with EPA's quality policy, data reported in this
document cannot be defended and should be considered for
qualitative purposes only.
However, quality control practices, such as routine good
laboratory practices, were applied. A specific example was the
calibration of the Model 560 Ozone Analyzer.
On April 11 and 12, 1990, the Model 560 Ozone Analyzer (Serial
No. 666864) was calibrated against an ozone photometer in
accordance with 40 CFR, Part 50, Appendix D. This was done by
Robert Murdoch and Richard Shores at the Center for Environmental
Quality Assurance (CEQA) at the Research Triangle Institute,
Research Triangle Park, North Carolina, 27709-5929. The results
were as follows:
CEQA Calibration
Concentrations in ppm
0.030
0.044
0.058
0.073
0.037
Analyzer Response in
millivolts (0.1 ppm scale)*
oToo
0.30
0.44
0.58
0.72
0.86
*CEQA Analyzer Response (ppm) = (Voltage) (0.100) - 0.000
The CEQA ozone photometer (Dasibi Model 1003AH, Ser. No, 2629) is
periodically checked against the EPA/EMSL QAD National Institute of
Standards and Technology (formerly Bureau of Standards) NIST-
traceable photometer.
-------
TEST GROUP 1
The room had continuously running mechanical sir ventilation. It
was approximately 200 square feet in area and had a sealed window
at one end and a door at the other.
Test # 1
Date: April 25-26, 1990 Settings: Fan, 12 o'clock
Weather: Sunny and hot with Generator, 12 o'clock
visible air pollution and
highs near 90 on the 25th;
hazy overcast till 11 am on
the 26th followed by hot and
sunny.
TIME UNCORRECTED READOUT CORRECTED READOUT
9:10 am 16 ppb *
10:30 am 16 ppb *
11:30 am 24 ppb *
1:00 pm 38 ppb *
2:3 0 pm 38 ppb *
4:0 0 pm 39 ppb *
8:00 am** 8 ppb *
8:30 am 5 ppb *
9:00 am 2 ppb *
9:40 am 3 ppb *
10:30 am 4 ppb *
11:10 am*** 18 ppb *
11:40 am 27 ppb *
12:35 pm 41 ppb *
1:15 pm 35 ppb *
1:55 pm 41 ppb *
*Corrected readout was indistinguishable from background.
**Unit was left running unattended overnight in secured room.
***At this point the sun broke through the haze, thereby
revealing the true source of the ozone readings.
-------
Test # 2
Date: April 26, 1990 Settings: Fan off
Weather: Hot, sunny, and Generator off
polluted.
TIME UNCORRECTED READOUT
2:25 pro 42 ppb
2:55 pm 42 ppb
3:25 pm 44 ppb
3:45 pm 42 ppb
3:50 pro Experiment Shut Down
-------
TEST GROUP 2
The room had a continuously running mechanical ventilation system
and was approximately 200 square feet in area. There was a single
sealed window and a single door.
Test # 1
Date: May 7, 1990 Settings: Fan 1 o'clock
Weather: Sunny and Generator 1 o'clock
clear, high around 70
TIME UNCORRECTED READOUT CORRECTED READOUT
2:00 pm 34 ppb 11 ppb
2:35 pm 36 ppb 13 ppb
3:08 pm 34 ppb 11 ppb
3:40 pm 38 ppb 15 ppb
4:05 pm 34 ppb 11 ppb
Test # 2
Date: May 7-8, 1990 Settings: Fan 3 o'clock
Weather: Sunny and Generator 3 o'clock
clear both days with
highs in the 70s
TIME UNCORRECTED READOUT CORRECTED READOUT
4:10 pm 37 ppb 14 ppb
7:25 am* 12 ppb 12 ppb
8:05 am 11 ppb 11 ppb
8:40 am 32 ppb 10 ppb
9:20 am 30 ppb 8 ppb
10:00 am 35 ppb 13 ppb
10:30 am 32 ppb 10 ppb
11:10 am 34 ppb 12 ppb
11:40 am 38 ppb 16 ppb
12:30 pm 38 ppb 16 ppb
2:3 0 pm 46 ppb 2 4 ppb
*Unattended unit was allowed to run overnight in the secured
room.
10
-------
Test # 3
Date: May 8-9, 1990
Weather: Sunny and clear
with a high of 75 on the
8th; overcast with a high
of 70 on the 9th.
Settings: Fan 5 o'clock (max)
Generator 5 o'clock (max)
TIME
3:15 pm
4:08 pm
7:20 am*
8:10 am
8:38 am
10:20 am
12:40 pm
3:00 pm
UNCORRECTED READOUT
110 ppb
140 ppb
100 ppb
100 ppb
100 ppb
98 ppb
120 ppb
120 ppb
CORRECTED READOUT
88 ppb
118 ppb
100 ppb
100 ppb
100 ppb
98 ppb
98 ppb
98 ppb
*Unattended unit left running overnight in the secured room.
Room background varied from near zero in the forenoon to 22 ppb
from 12 o'clock on.
Test # 4
Date: May 9, 10, 11, 1990
Weather: Overcast and 70
on the 9th; rain and 70
on the 10th; sunny and 60
on the llth.
Settings: Fan off
Generator off
TIME
3:04 pm
3:50 pm
8:30 am
11:20 am
8:00 am
9:10 am
10:30 am
10:35 am
UNCORRECTED READOUT
90 ppb*
22 ppb
10 ppb
22 ppb
10 ppb
21 ppb
21 ppb
Experiment Shut Down
*The ozone generator unit had just been shut down when this reading
was taken, and it is corrected for background.
11
-------
TEST GROUP 3
The room had no mechanical ventilation and was approximately 90
square feet in area. It had a closed window and a single door.
Background ozone averaged 5.6 ppb during the principal hours of
daylight and 1.5 ppb in the early morning.
Test # 1
Date: May 21, 1990 Settings: Fan 12 o'clock
Weather: Partially sunny, Generator 12 o'clock
highs in the 80s
TIME UNCORRECTED READOUT CORRECTED READOUT
2:10 pm 4.8 ppb *
2:2 0 pm 4.6 ppb *
2:3 0 pm 5.5 ppb *
2:4 0 pm 6.4 ppb *
2:5 0 pm 5.4 ppb *
3:0 0 pm 6.1 ppb *
*Indistinguishable from background
Test # 2
Date: May 21, 1990 Settings: Fan 5 o'clock
Weather: As above. Generator 12 o'clock
TIME UNCORRECTED READOUT CORRECTED READOUT
3:10 pm 5.5 ppb *
3:20 pm 5.5 ppb *
*Indistinguishable from background
12
-------
Data/6
Test # 3
Date: May 21, 1990
Weather: Partially
sunny, high in the 80s
Settings: Fan 3 o'clock
Generator 3 o'clock
TIME
3:25 pm
3:35 pm
3:45 pro
3:55 pm
4:05 pro
4:17 pro
UNCORRECTED READOUT
23 ppb
24 ppb
27 ppb
27 ppb
26 ppb
27 ppb
CORRECTED READOUT
17.4 ppb
18.4 ppb
21.4 ppb
21.4 ppb
20.4 ppb
21.4 ppb
Test # 4
Date: May 22, 1990
Weather: Windy, rain,
high around 60
Settings: Fan 5 o'clock
Generator 5 o'clock
TIME
8:05 am
8:20 am
8:35 am
8:50 am
9:05 am
9:15 am
9:25 am
UNCORRECTED READOUT
54 ppb
50 ppb
34 ppb
46 ppb
46 ppb
42 ppb
40 ppb
CORRECTED READOUT
52
48
32
44
44
40
38
, 5 ppb
5 ppb
. 5 ppb
5 ppb
5 ppb
5 ppb
5 ppb
13
-------
Test # 5
Date: May 22, 1990 Settings: Fan 5 o'clock
Weather: Windy, rain, Generator 5 o'clock
high around 60
TIME UNCORRECTED READOUT CORRECTED READOUT
9:30 am 160 ppb 158.5 ppb
9:40 am 260 ppb 258.5 ppb
10:25 am 270 ppb 268.5 ppb
10:45 am 360 ppb 354.4 ppb
11:15 am 340 ppb 334.4 ppb
12:10 pm 600 ppb 594.4 ppb
1:45 pm 820 ppb 814.4 ppb
2:45 pm 700 ppb 694.4 ppb
3:30 pm 680 ppb 674.4 ppb
4:15 pm* 620 ppb 614.4 ppb
7:10 am 998 ppb 996.5 ppb
7:50 am 760 ppb 758.5 ppb
*At this point the room was secured and no more readings were taken
until the next morning. May 23rd dawned sunny and cool with
morning temperatures in the low 50s.
Test # 6 (Shutdown/Decay)
Date: May 23, 1990 Settings: Fan off
Weather: Sunny and cool, Generator off
temp, in low 50s
TIME UNCORRECTED READOUT CORRECTED READOUT
7:54 am Ozone Generator Shut Off 758.5 ppb
7:55 am 540 ppb 538.5 ppb
8:02 am 70 ppb 68.5 ppb
8:10 am 4 ppb 4.0 ppb
8 : 2 0 am 2 ppb 2 . 0 ppb
8:4 5 am 2 ppb 2.0 ppb
8:50 am Experiment Shut Down
14
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing'
1. REPORT NO.
EPA-600/R-93-177
2.
4. TITLE AND SUBTITLE
Contribution to Indoor Ozone Levels of an Ozone
Generator
3. Rl
5. REi «... ~~, _
September 1993
6. PERFORMING ORGANIZATION CODE
7, AUTHOR(S)
Raymond S. Steiber
8. PERFORMING ORGANIZATION REPORT NO
9. PERFORMING OROANIZAT1ON NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
See Block 12
11. CONTRACT/GRANT NO.
NA (Inhouse)
12, SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 4/90 - 12/92
14, SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES ^EERL project officer is Raymond S. Steiber, Mail Drop 54, 919 /
541-2288.
16. ABSTRACT
repOrt gives results of a study of a commonly used commercially avail-
able ozone generator, undertaken to determine its impact on indoor ozone levels. Ex-
periments were conducted in a typical mechanically ventilated office and in a test
house. The generated ozone and the in-room ozone concentrations were measured.
The results showed that, when the unit was operated at the manufacturer's recom-
mended setting, it generated little if any ozone. The indoor concentrations in this
case were not significantly above natural background. When operated at the meximum
setting, the generator produced large amounts of ozone: over 100 ppb in well ventila-
ted spaces, and nearly 1 ppm in poorly ventilated spaces. When the ozone generator
was turned off, ozone levels quickly returned to background. No measurements
were made to determine the effect of the device on other aspects of indoor air quality
such as the elimination of volatile organic compounds.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.tDENTIFiERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Ozone
Pollution Control
Stationary Sources
Ozone Generators
Indoor Air Quality
13 B
07B
3. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
19
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-i (9-73)
15
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Center for Environmental Research Information
Cincinnati. Ohio 45268
OFFICIAL BUSINESS
PENALTY FOR PRIVATE USE. S3OO
AN EQUAL OPPORTUNITY EMPLOYER
If your address is incorrect, please change on the above label
tear off; and return to the above address.
If you do not desire to continue receiving these technical
reports, CHECK HERED; tear off label, and return it to the
above address.
Publication No. EPA-eoo /R- 93-177
------- |