Chemical Substances and Biological Agents

Studies and
Research Projects
REPORT R-666

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria
under Real Conditions

Jacques Lavoie
Genevive Marchand
Yves Cloutier
Yves Beaudet
Jrme Lavou

Established in Qubec since 1980, the Institut de recherche

Robert-Sauv en sant et en scurit du travail (IRSST)
is a scientific research organization known for the quality
of its work and the expertise of its personnel.

OUR

RESEARCH
is working for you !

Mission
To contribute, through research, to the prevention of
industrial accidents and occupational diseases as well
as to the rehabilitation of affected workers.
To offer the laboratory services and expertise necessary
for the activities of the public occupational health and
safety prevention network.
To disseminate knowledge, and to act as scientific
benchmark and expert.
Funded by the Commission de la sant et de la scurit
du travail, the IRSST has a board of directors made up
of an equal number of employer and worker representatives.

To find out more

Visit our Web site for complete up-to-date information
about the IRSST. All our publications
can be downloaded at no charge.
www.irsst.qc.ca
To obtain the latest information on the research carried
out or funded by the IRSST, subscribe to Prvention au
travail, the free magazine published jointly by the
IRSST and the CSST.
Subscription: 1-877-221-7046

Legal Deposit
Bibliothque et Archives nationales du Qubec
2010
ISBN: 978-2-89631-502-4 (Print Version)
ISBN: 978-2-89631-503-1 (PDF)
ISSN: 0820-8395

IRSST Communications Division

505 De Maisonneuve Blvd. West
Montral, Qubec
H3A 3C2
Phone: 514 288-1551
Fax: 514 288-7636
publications@irsst.qc.ca
www.irsst.qc.ca
Institut de recherche Robert-Sauv
en sant et en scurit du travail,
september 2010

Chemical Substances and Biological Agents

Studies and
Research Projects
REPORT R-666

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria
under Real Conditions

Disclaimer
The IRSST makes no guarantee
regarding the accuracy, reliability
or completeness of the information
contained in this document. Under no
circumstances shall the IRSST be held
liable for any physical or
psychological injury or material
damage resulting from the use of
this information.
Note that the content of the documents is protected by Canadian
intellectual property legislation.

Jacques Lavoie,
Research Department, IRSST
Genevive Marchand and Yves Cloutier,
Research and Expertise Support Department, IRSST
Yves Beaudet,
Laboratory Services and Expertise, IRSST
Jrme Lavou,
CHUM Research Centre

This publication is available free

of charge on the Web site.

This study was financed by the IRSST. The conclusions and recommendations are those of the authors.

IN CONFORMITY WITH THE IRSSTS POLICIES

The results of the research work published
in this document have been peer-reviewed.

IRSST -

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

ACKNOWLEDGEMENTS

The authors wish to thank the non-industrial establishments that gave them access to their
systems and the HVAC cleaning companies that facilitated entry into these systems, as well as
Laurent Gratton of the IRSST and the members of the follow-up committee. Special thanks go to
the experts/evaluators: Ali Bahloul, Simon Aubin, Louis Lazure and Nicole Goyer, and
technicians of the IRSST, more specifically Carole Blanchard, Zlie Fortin and Claude
Ltourneau. Sincere thanks also go to Sylvie Bdard of the Association paritaire pour la sant et
la scurit au travail (joint sector-based occupational health and safety association) in the social
services sector, for having opened many doors to us.

IRSST -

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

iii

ABSTRACT
Dust accumulation in the components of HVAC systems is a potential source of contaminants.
To date, very little information is available on recognized methods for assessing dust build-up in
these systems. The few existing methods are either objective in nature, involving numerical
values, or subjective in nature, based on experts judgments. It is therefore difficult for building
managers to assess the proposals made by specialized cleaning companies.
An earlier project aimed at assessing different methods of sampling dust in ducts was carried out
in the IRSSTs laboratories. The goal was to reproduce different levels of dust accumulation in
non-porous metal ducts in the laboratory, to compare the different methods of sampling surface
dusts cited in the literature, to compare numerical evaluation methods with the visual inspection
method, and to establish application procedures. This laboratory study showed that all the
sampling methods were practicable, provided that a specific surface-dust cleaning initiation
criterion was used for each method. However, these conclusions were reached on the basis of
ideal conditions in a laboratory using a reference dust from the American Society of Heating,
Refrigerating and Air-Conditioning Engineers, Inc. (ASHRAE). The conclusions reached
therefore required validation under real conditions.
The objective of this study was to validate these laboratory results in the field. To this end, the
laboratory sampling templates were replicated in real ducts and the three sampling methods (the
IRSST method, the method of the American organization National Air Duct Cleaner Association
(NADCA) and that of the French organization Association pour la prvention et ltude de la
contamination (ASPEC)) were used simultaneously in a statistically representative number of
systems. The air return and supply ducts were also compared.
The cleaning initiation criteria under real conditions were found to be 6.0 mg/100 cm using the
IRSST method, 2.0 mg/100 cm using the NADCA method, and 23 mg/100 cm using the
ASPEC method. In the laboratory study, the criteria using the same methods were 6.0 for the
IRSST method, 2.0 for the NADCA method, and 3.0 for the ASPEC method. The laboratory
criteria for the IRSST and NADCA methods were therefore validated in the field. The ASPEC
criterion was the only one to change. Moreover, no statistical differences were found between the
return and supply ducts.
The ASPEC method therefore allows for the most accurate evaluation of dust accumulation in
HVAC system components. It is also the method that cleaning companies instinctively prefer as
it most closely resembles customary cleaning processes. We therefore recommend using the
latter method to objectively assess dust accumulation levels in HVAC systems.

IRSST -

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

TABLE OF CONTENTS

1. INTRODUCTION .............................................................................................1
2.

METHOD .............................................................................................................................. 5

3.

RESULTS AND DISCUSSION ........................................................................................... 9

3.1

Inter-expert agreement ................................................................................................... 10

3.2

Criteria for initiating cleaning ....................................................................................... 11

3.3

Return ducts .................................................................................................................... 12

4.

SCOPE AND LIMITATIONS ........................................................................................... 13

5.

CONCLUSION ................................................................................................................... 14

6.

BIBLIOGRAPHY ............................................................................................................... 15

vi

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

- IRSST

LIST OF TABLES
Table 1: Criteria for initiating cleaning of non-porous ducts (1) .............................. 2
Table 2: Advantages and disadvantages of the methods ........................................... 5
Table 3: Median of the seven experts ratings (ratings 1 to 3) and corresponding
weighings in mg/100cm........................................................... 9
Table 4: Dust concentrations (in mg/100 cm) in ducts to be cleaned.......... .. 11
LIST OF FIGURES
Figure 1 Return duct outfitted with sampling template............................... 6

IRSST -

1.

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

INTRODUCTION

Dust accumulation in the components of HVAC systems may be a potential source of

contaminants. Building occupants are becoming increasingly adamant that their systems be free
of dirt and dust. However, there is little information available on recognized or standardized
methods for assessing dust accumulation in these systems. It is therefore difficult for building
managers to evaluate the proposals made by specialized cleaning companies, as no
recommended objective methods exist for assessing dust concentrations on duct surfaces.
Ideally, HVAC systems should be maintained at optimum cleanliness levels. To do so, it is
important to have the necessary tools for measuring the amount of dust deposited in the
components of these systems. An objective diagnostic process helps prevent abuses and
unnecessary system cleaning (1).
In the United States and Canada, the initiation of air system cleaning is currently based on visual
inspection (2-4). However, these criteria are subjective and impractical when large-scale
cleaning work is involved (5). In 2006, the American organization National Air Duct Cleaner
Association (NADCA) published criteria for post-cleaning cleanliness acceptance. However,
these criteria are inadequate if someone wants to know when to initiate cleaning of the
components of HVAC systems (4).
Frances Association pour la prvention et ltude de la contamination (ASPEC, or association
for the prevention and study of contamination) has also published a guide on methods for
maintaining cleanliness in non-porous air systems for clean rooms and associated controlled
environments (1). This guide reports on the cleaning initiation criteria for buildings in the tertiary
sector (e.g. office buildings) and dust-sampling methods used in different countries. Table 1
presents these criteria.
In this table, the criteria cited refer to different dust sampling methods, thus making comparisons
difficult. According to ASPEC, these methods can only be applied in rigid, non-porous ducts that
are sufficiently large, i.e. round components that are larger than 30 cm in diameter; the ducts
must also be horizontal and the walls dry (1). Samples must be taken from a layer of dust
distributed on the bottom surface and not from dust piles (1). The sampling methods were found

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

- IRSST

to have certain deficiencies, including primarily the absorption of moisture from the air by the
cellulose ester membranes and dust adhesion to the walls of the cassettes and sampling tubes.
Table 1: Criteria for initiating cleaning of non-porous ducts
Country

United
States
(NADCA
2006)

Cleaning initiation
criterion based on
surface density (g/m2)

Cleaning initiation
criterion based on
thickness (m)

Postcleaning
acceptance
criterion
(g/m2)

Sampling
method

Surface sampling
on membrane at
15 L/min (open
cassette)

0.075

Great
Britain
(1998)

Blowing: 1
Exhaust: 6

Blowing: 60
Exhaust: 180

0.1

Surface sampling
on membrane at
15 L/min

Finland
(2001)

Blowing: 2
Exhaust: 5

0.1

Surface sampling
on membrane at
10L/min
(with sampling
tube)

France
ASPEC
(2004)

Blowing: 0.4
Exhaust: 6

0.1

Surface sampling
on membrane at
15L/min
(with sampling
tube)

One method designed to reduce these problems involves weighing the entire cassette system and
using as the sampling head the pre-weighed IOM cassette with a 25-mm diameter (SKC Inc.
Eighty Four, PA, USA) equipped with a polyvinyl chloride (PVC) membrane with a pore size of
0.8 m. This sampling method was compared in the laboratory to the methods cited in the
literature in order to determine which was the most accurate (5). The objectives of the earlier
project (see IRSST Report R-525) were to reproduce different levels of dust accumulation in a
non-porous metal duct in the laboratory, to compare different surface-dust sampling methods, to

IRSST -

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

compare numerical evaluation methods to visual inspection methods, and to determine

application procedures.
All sampling methods were shown to be practical, provided that the corresponding surface-dust
cleaning initiation criterion is used. Using the visual inspection method, for an average experts
rating of 2 on a scale of 3, the corresponding values were 2.0 mg/100 cm using the NADCA
method, 3.0 mg/100 cm using the ASPEC method, and 6.0 mg/100 cm using the IRSST method
(5). Level 2 signified a uniform layer and localized dust accumulations in the ducts and therefore
that cleaning was required. Level 1 corresponded to a relatively clean duct, and level 3 to an
extremely dirty duct. These methods yielded surface sampling values that differed significantly
from one method to the other (p0.05). The dust used was the standard dust recommended by
ASHRAE (6).
The main objective of this actual project was to validate, under real conditions in the field, the
results obtained in the laboratory study regarding the criteria for initiating cleaning of nonindustrial HVAC systems (IRSST Report R-525) (office buildings, schools and hospitals), using
the dust deposited in the components of the HVAC systems of such buildings following normal
system use. To do so, we reproduced the laboratory sampling templates in real ducts and
repeated the three sampling procedures simultaneously (the IRSST, ASPEC, and NADCA
methods) in a statistically representative number of systems. We chose the method yielding the
most accurate results and the smallest standard deviation. The air supply and return ducts were
also compared.

IRSST -

2.

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

METHOD

The advantages and disadvantages of the three methods evaluated are reported in Table 2.
Contrary to the laboratory study, in this field study the IOM cassette (IRSST method) was made
of stainless steel rather than plastic, which eliminated the weight fluctuations attributable to the
moisture absorption associated with plastic.

Table 2: Advantages and disadvantages of the methods

NADCA Vacuum test
(0.8 m cellulose ester
membrane in an open cassette
37 mm in diameter; 15 L/min.;
vacuumed duct surface of 100
cm)

Advantage

- Method recognized in industrial

hygiene
- Dust deposited on the inside
walls of the cassette

Disadvantages

- Cellulose ester membrane

fragile and sensitive to moisture

(2006)
French method

- Bevelled tube facilitates

sampling

ASPEC
(0.8 m cellulose ester
membrane in a closed cassette
37 mm in diameter; connected
to a bevelled tube; 15 L/min.;
vacuumed duct surface of
100 cm)
(2004)

Advantages

Disadvantages

(2009)

- Cellulose ester membrane

sensitive to moisture
- Adhesion to the inside walls of
the sampling tube of the cassette
- Membrane made of PVC (less
sensitive to moisture)

Method being developed by

the IRSST
(0.8 m PVC membrane in an
IOM stainless steel cassette 25
mm in diameter; 15 L/min.;
vacuumed duct surface of
33.82 cm)

- The fact of being able to stick

the sampling tube on the duct
surface makes it easier to aspirate
all the dust

Advantages

Disadvantages

- No loss of dust through

adhesion to the inside walls
because the whole cassette is
weighed
- The template needs to be
adapted to the circumference of
the cassette sampling head

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

- IRSST

The earlier results obtained in the laboratory study established that the minimum number of
samplings required to demonstrate a difference between two geometric means was 5 (n = 5) for
an acceptable error of 10% and a confidence level of 95% (t = 1.96) (5,7). First, some 30 HVAC
systems that had no interior insulation, and a maximum of three HVAC systems per building,
were identified. This step was facilitated by the involvement of members of the follow-up
committee (cleaning companies and building managers).
Three simultaneous sampling procedures were performed for each of the systems, using the three
previously evaluated methods described in Table 2 (5). Insofar as possible, the samples were
taken from the supply diffusers in the rooms, at a reasonable distance from the last elbow and at
the furthest accessible point in the air duct. Weighings were performed using the IRSSTs
standard method (8). Figure 1 shows a duct undergoing evaluation using a template.

Figure 1: Return duct outfitted with sampling template

Samplings were also taken from the return ducts in each building. The literature mentions the
need to clean such ducts when surface dust concentrations exceed 50 to 60 mg/cm of surface
(1,5). The final step involved subjective assessment through the visual inspection of cleanliness
levels by the same sub-committee of experts that took part in Project R-525. This evaluation
consisted of the visualization of the deposits shown in photographs that were taken from angles

IRSST -

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

both parallel to and perpendicular to the ducts evaluated. A three-level rating scale was used, in
which level 1 was normal, meaning clean ducts or those with a thin, uniform layer of dust; level
2, or above normal, corresponded to a uniform layer of dust and localized accumulations; and
level 3, or serious, corresponded to significant dust accumulations (5,9). The sub-committee of
experts had seven members. Given the discrete nature of the ratings, the median value was used
to calculate an average rating for each system, with a value of 2 corresponding to the initiation of
cleaning.
Forty-four different sites were evaluated, giving a total of 132 samples (44 sites X 3 methods
each). Eleven of these sites involved return ducts. Two-factor (ducts and methods) and threelevel (surface dust concentrations obtained using three methods) variance analyses (ANOVA)
were performed on the value logarithms to determine whether there were statistically significant
differences (p0.05) (10). Regressions were then performed, and the degree of correlation was
calculated in order to assess the relationship among the three sampling methods. Similarly, interexpert agreement in terms of their visual inspections was quantified using Krippendorffs alpha
index (11). An index of 1 indicated that the experts came up with the same rating for each system
evaluated. An index of 0 indicated the absence of concurrence among their evaluations.

IRSST -

3.

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

RESULTS AND DISCUSSION

Table 3 shows the results obtained by the seven experts and the weighings obtained in mg/100
cm for the 44 ducts. The weighing detection limit was 0.02 mg, i.e. the minimum value reported
for the method, divided by the square root of 2 (12,13).
Table 3: Median of the seven experts ratings (ratings 1 to 3) and corresponding weighings
in mg/100cm
IRSST
9.11
1.00
0.92
1.51
0.21
0.24
0.56
0.74
3.58
2.84
1.18
55.38
0.89
0.13
23.99
13.99
1.42
19.94
10.27
1.92
14.44
7.23
2.37
51.2
67.27
110.63
0.56
281.38
3.98
22.78
0.71
13.45
8.58

ASPEC
7.16
2.95
2.00
4.04
27.9
0.39
8.45
1.09
2.23
5.20
0.90
155.25
6.74
1.62
17.89
28.25
29.08
36.16
46.55
6.57
7.56
13.61
42.13
62.31
67.09
218.33
7.12
333.77
1.97
2.70
7.60
8.47
14.56

NADCA
2.69
0.76
0.42
0.63
1.68
0.018
1.09
1.35
1.35
0.51
0.17
155.75
0.26
0.07
1.23
1.67
4.66
2.26
24.54
3.32
0.66
2.28
2.92
2.28
15.55
57.98
0.02
112.69
1.31
0.70
0.05
3.22
1.49

Median of the
experts ratings
2
1
1
1
2
2
2
1
2
1
1
2
1
1
1
1
2
2
2
1
1
1
3
1
1
2
2
3
1
2
1
1
2

Duct no.
1r
2
3
4
5
6
7
8
9r
10
11
12r
13
14
15
16
17r
18
19r
20
21
22
23
24
25r
26
27
28r
29r
30
31
32
33

10

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

15.95
23.29
7.13
0.14
0.12
2.80
0.68
5.00
0.95
3.43
1.54
58.75
87.73
159.18
169.41
159.97
68.16
237.97
1.6
29.57
16.07
32.75
r
= air return duct

3.1

1
0.02
0.12
0.34
0.14
2.61
32.64
53.37
21.25
1.12
1.02

1
1
1
1
2
2
2
2
2
2
2

- IRSST

34
35
36r
37
38r
39
40r
41
42
43
44

Inter-expert agreement

One of the objectives of this project was to replace the subjective evaluation of the cleanliness
levels of HVAC systems (the rating method) by a more objective method based on weighings.
To demonstrate the degree of subjectivity of the ratings, we compared the seven experts ratings
of the same ducts.
Taking the ratings as a whole, some experts ratings differed in a statistically significant manner
from the ratings given by other experts (p<0.001; F = 6.2) (log-ANOVA test) (10). F
represents Fishers test statistic. The higher F was, the greater the difference in the experts
ratings.
Based on a Tukey-Kramer multiple comparison test, Expert One cast gave significantly higher
ratings (1) (p0.05) than the six other experts, while Expert Five gave significantly lower ratings
than both experts One and Two (p0.05) (10).
Another criterion used in our study to measure the reliability of the different experts ratings was
Krippendorffs alpha index (11). The index was calculated at 0.37. At the very most, this
corresponds to moderate agreement among the experts and also shows the subjective nature of
the ratings.

IRSST -

3.2

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

11

Criteria for initiating cleaning

Table 3 shows that 20 out of 44 ducts obtained a median rating of 2, hence a recommendation
that cleaning be carried out. Table 4 shows the geometric means and geometric standard
deviations of these values.
A log-ANOVA test on these values showed statistically significant differences among the three
sampling methods (p<0. 001; F = 9.5).
Table 4: Dust concentrations (in mg/100 cm) in the ducts to be cleaned
IRSST
9.11
0.21
0.24
0.56
3.58
55.38
1.42
19.94
10.27
110.63
0.56
22.78
8.58
0.95
1.54
87.73
169.41
68.16
1.6
16.07

ASPEC
7.16
27.9
0.39
8.45
2.23
155.25
29.08
36.16
46.55
218.33
7.12
2.7
14.56
3.43
58.75
159.18
159.97
237.97
29.57
32.75

NADCA
2.69
1.68
0.018
1.09
1.35
155.75
4.66
2.26
24.54
57.98
0.02
0.7
1.49
0.14
2.61
32.64
53.37
21.25
1.12
1.02

6.48
7.91

22.56
5.48

2.39
10.93

Median of
the ratings
Duct no.
2
1
2
5
2
6
2
7
2
9
2
12
2
17
2
18
2
19
2
26
2
27
2
30
2
33
2
38
2
39
2
40
2
41
2
42
2
43
2
44
Geometric mean
(cleaning initiation criterion)
Geometric standard deviation

The cleaning initiation criteria corresponding to the geometric mean obtained using each of the
three methods for the 20 ducts were therefore 6.0 mg/100 cm for the IRSST method (geometric
mean of 6.48), 2.0 mg/100 cm for the NADCA method (geometric mean of 2.39) and 23
mg/100 cm for the ASPEC method (geometric mean of 22.56). In the laboratory study, the
criteria for these same methods were 6.0 for the IRSST method, 2.0 for the NADCA method and

12

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

- IRSST

3.0 for the ASPEC method (5). The field criteria obtained were therefore similar to the
laboratory criteria with respect to both the IRSST and NADCA methods. Only the ASPEC
criterion changed. It must be recalled that the two other methods did not involve direct contact
with the dust. In fact, there was a distance of 0.381 mm between the cassette and the duct surface
(4,5), such that only the surface dust was aspirated and aspiration was incomplete. By contrast,
the ASPEC method allows virtually the entire dust deposit to be scraped and collected. All the
dust is therefore dislodged and aspirated. This method also yields the smallest geometric
standard deviation (Table 4) and thus, the most reliable results.
A statistically significant correlation exists among these three methods. The correlation
coefficient between the IRSST and NADCA methods is 0.74, with a value of F = 51.01 and a
p<0.001. The correlation coefficient between the ASPEC and NADCA methods is 0.83, with a
value of F = 90.89 and a p<0.001. Lastly, the correlation coefficient between the IRSST and
ASPEC methods is 0.67, with a value of F = 33.68 and a p<0.001. These correlation coefficients
confirm that any of the three methods can be used, provided that its specific initiation criterion is
applied.

3.3

Return ducts

Of the 44 ducts evaluated, 11 were return ducts (see Table 3). It should be recalled that the main
objective of this study was to validate the cleaning initiation criteria for HVAC systems.
However, further to a recommendation made by the projects follow-up committee, we also
tested the hypothesis that the air return ducts always have greater dust accumulation than the
supply ducts. We therefore compared the two types of ducts and obtained surprising results. A
log-ANOVA analysis did not establish any statistically significant difference between the dust
deposits measured in the return and supply ducts, using both the IRSST method (p = 0.23; F =
1.45) and the ASPEC method (p = 0.25; F = 1.37). However, using the NADCA method, the
difference was statistically significant (p = 0.01; F = 6.25). We were thus able to formulate the
hypothesis that the sensitivity of the different methods played a significant role in the differences
found between the two types of ducts. The more sensitive the method, as was the case of the
ASPEC method, the fewer differences there appeared to be. Although the number of ducts
involved was sufficient to make such a comparison, it would be relevant to push this comparison
further in a possible future study of broader scope.

IRSST -

4.

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

13

SCOPE AND LIMITATIONS

As stated in the introduction, the NADCA and ASPEC methods underestimated the real
concentrations of dust due to losses. In fact, the dust that adheres to the inside walls of the
cassettes, and more specifically, inside the sampling tube used in the ASPEC method, is not
weighed. Moreover, the NADCA and IRSST methods vacuum only the surface dust. By contrast,
the ASPEC method is the one that most closely resembles the cleaning processes used by HVAC
cleaning companies (12). It is also the one that gives the best geometric standard deviation. We
therefore recommend that the ASPEC method, with its specific cleaning initiation criterion of 23
mg/100 cm2, be used to objectively assess cleanliness levels of HVAC systems. The value
proposed by ASPEC in Table 1 for initiating duct cleaning is 4.0 mg/100 cm (0.4 g/m).
However, given that the methodology is not described in its entirety, we are not in a position to
discuss it. Regarding the other two methods, it is the first time that a cleaning initiation criterion
has been proposed.
Eventually, the ASPEC method could also be improved, for example, by using an Accucap
membrane in the sampling cassette. With this membrane, all the dust removed is weighed and
there are no losses. The aspiration tube should also be included in the weighing. In this case, the
losses would be virtually nil.
With regard to the return ducts, we simply compared them to the supply ducts. Even though there
was a sufficient number of ducts to allow for such a statistical comparison, it would be advisable
to do a systematic comparison in the context of another research project, by using, for example,
ducts in the same HVAC systems and the ASPEC method, in order to determine the presence of
any statistically significant differences, as well as the factor or number of times bigger the
concentrations are in the return ducts than those in the supply ducts, if indeed there are any
differences at all.

14

5.

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

- IRSST

CONCLUSION

This study was carried out under real dust-accumulation conditions inside the components of
non-industrial HVAC systems (e.g. schools, office buildings, or hospitals). We established
HVAC-system cleaning initiation criteria for these conditions. These criteria were 23 mg/100
cm for the ASPEC method, 6.0 mg/100 cm for the IRSST method, and 2.0 mg/100 cm for the
NADCA method. These criteria were identical to those obtained in the laboratory for the IRSST
and NADCA methods. All three methods can therefore be used, provided that their specific
cleaning initiation criterion is applied. Using the ASPEC method under real conditions, although
the criterion differs from that determined in the laboratory, this method aspirates most of the
surface dust in the duct and is therefore a preferable option to the other two, which aspirate only
the surface dust. In conclusion, it is this method that yields the most accurate results and the
smallest standard deviation. It is also the one we recommend.

IRSST -

6.

Heating, Ventilation and Air Conditioning

Validation of System Cleaning Initiation Criteria under Real Conditions

15

BIBLIOGRAPHY
1. ASPEC. 2004. Maintien en propret des rseaux araulique pour salles propre et
environnements matriss apparents, Association pour la prvention et ltude de la
contamination, France, March. (In French only)
2. BROSSEAU, L.M., Vesley, D., Kuehn et al. 2000. Methods and Criteria for Cleaning
Contaminated Ducts and Air-Handling Equipment. ASHRAE Transaction 4335:88-199.
3. LAVOIE, J. and Lazure, L. 1994. Guide for the Prevention of Microbial Growth in
Ventilation Systems, Institut de recherche Robert-Sauv en sant et en scurit du travail
du Qubec, Technical Guide RG-089.
4. NADCA. 2006. Assesssment, Cleaning, and Restoration of HVAC Systems. ACR 2006.
National Air Duct Cleaner Association, Washington, DC, 44 p.
5. LAVOIE, J., Gravel, R., Cloutier, Y., Bahloul, A. 2007. Critres de dclenchement du
nettoyage des systmes de chauffage, de ventilation et de conditionnement dair
ddifices non industriels. Institut de recherche Robert-Sauv en sant et en scurit du
travail du Qubec, Studies and Research Projects, Report R-525, 20 pages. (In French
only)
6. ASHRAE. 1992. Gravimetric and dust-spot procedures for testing air cleaning devices
used in general ventilation for removing particulate matter. Atlanta, American Society of
Heating, Refrigerating, and Air-Conditioning Engineers, Inc. (ANSI/ASHRAE Standard
52.1-1992).
7. BOHS. 1993. Sampling Strategies for Airborne Contaminants in the Workplace. British
Occupational Hygiene Society Technology Committee Working Group, H and H
Scientific Consultants Ltd., Leeds, UK, 84 p.
8. IRSST. 1985. Mesure de concentrations pondrales en poussires respirables et totales.
Institut de recherche Robert-Sauv en sant et en scurit du travail, Notes et rapports
scientifiques et techniques, mthode 48-1, Montral. (In French only)
9. GOYER, N., Lavoie, J., Lazure, L., et al. 2005. La qualit de lair dans les
tablissements du rseau de la sant et des services sociaux. Institut de recherche RobertSauv en sant et en scurit du travail du Qubec, Guide RG-410, 148 pages. (In French
only)
10. NCSS 2007. Statistical and Power Analysis Software, Kaysville, Utah, USA.
11. KRIPPENDORFF, K. 2008. Content Analysis: An Introduction to Its Methodology. Sage
Publications, Thousand Oaks, CA, 481 p.
12. FINKELSTEIN, M.M. and Verma, D.K. 2001. Exposure estimation in the presence of
non-detectable values: another look. American Industrial Hygiene Association Journal
Vol. 62, (2) pp.195-198.