Science of the Total Environment 381 (2007) 222 – 232

www.elsevier.com/locate/scitotenv

On the estimation of characteristic indoor air quality parameters

using analytical and numerical methods
Christos H. Halios, Costas G. Helmis ⁎
Department of Applied Physics, Faculty of Physics, University of Athens, Building Phys 5, University Campus, Athens 15784, Greece
Received 20 December 2005; received in revised form 6 March 2007; accepted 14 March 2007
Available online 26 April 2007

Abstract

Indoor exposure to air contaminants penetrating from the outdoor environment depends on a number of key processes and
parameters such as the ventilation rate, the geometric characteristics of the indoor environment, the outdoor concentration and the
indoor removal mechanisms. In this study two alternative methods are used, an analytical and a numerical one, in order to study the
time lag and the reduction of the variances of the indoor concentrations, and to estimate the deposition rate of the air contaminants
in the indoor environment employing both indoor and outdoor measurements of air contaminants.
The analytical method is based on a solution of the mass balance equation involving an outdoor concentration pulse which
varies sinusoidally with the time, while the numerical method involves the application of the MIAQ indoor air quality model
assuming a triangular pulse. The ratio of the fluctuation of the indoor concentrations to the outdoor ones and the time lag were
estimated for different values of the deposition velocity, the ventilation rate and the duration of the outdoor pulse.
Results have showed that the time lag between the indoor and outdoor concentrations is inversely proportional to the deposition
and ventilation rates, while is proportional to the duration of the outdoor pulse. The decrease of the ventilation and the deposition
rate results in a rapid decrement of the variance ratio of indoor to outdoor concentrations and to an increment of the variance ratio,
respectively.
The methods presented here can be applied for gaseous species as well as for particulate matter. The nomograms and theoretical
relationships that resulted from the simulation results and the analytical methods respectively were used in order to study indoor air
phenomena. In particular they were used for the estimation of SO2 deposition rate.
Implications of the studied parameters to exposure studies were estimated by calculating the ratio of the indoor exposure to the
exposure outdoors. Limitations of the methods were explored by testing various scenarios which are usually met in the indoor
environment. Strong indoor emissions, intense chemistry and varying ventilation rates (opening and closing of the windows) were
found to radically influence the time lag and fluctuation ratios.
© 2007 Elsevier B.V. All rights reserved.

Keywords: Mass balance equation; Analytical solution; Deposition rates; Ventilation rate; Exposure; Indoor chemistry; Urban Aerosol

1. Introduction

During recent years, the understanding of the mecha-

⁎ Corresponding author. Tel.: +30 210 7276927; fax: +30 210
7295285.
E-mail address: chelmis@phys.uoa.gr (C.G. Helmis).
0048-9697/$ - see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.scitotenv.2007.03.012
nisms that control indoor pollution is of increasing
interest among the scientific community, as a result of
the growing concern related to the effects of indoor
 C.H. Halios, C.G. Helmis / Science of the Total Environment 381 (2007) 222–232
exposure on human health (Jones, 1999). Indoor expo-
sure to air contaminants, especially to those that penetrate
from the outdoor environment, depends on a number of
key processes and parameters such as the ventilation rate,
the geometric characteristics of the indoor environment,
the outdoor concentration and the indoor removal
mechanisms. A number of theoretical models have
been developed for the assessment of indoor air quality
(Shair and Heitner, 1974; Nazaroff and Cass, 1987, 1989;
Switzer and Ott, 1992). The basis of these models is the
mass balance equation, which describes the relationship
between sources emitting pollutants into an enclosed
well-mixed compartment, the physical parameters of the
compartment, and the pollutant concentrations as a
function of time. Applications of theoretical models
have led to useful conclusions regarding indoor air
quality (Freijer and Bloemen, 2000).
One of the main indoor removal mechanisms results
from the interactions between the pollutants and the
fixed indoor surfaces (walls, floor), and is parameterized
with the deposition velocity (Nazaroff and Cass, 1987),
or the first order decay rate (Shair and Heitner, 1974).
Deposition of a specific pollutant indoors depends on
the material of the surfaces, indoor parameters related
mainly to humidity and temperature, and the ratio of the
indoor surfaces to the volume of the indoor environment
(A/V ratio). Even though technologies exist to directly
measure the deposition rate, they are often difficult to
implement due to the large number of the contributing
factors, or their cost. Thus many studies have focused on
the determination of the deposition rates using exper-
imental data of simultaneous indoor and outdoor mea-
surements that are often available in indoor air quality
studies and they are based on the application of the mass
balance equation under specific assumptions (Jamriska
and Morawska, 2003; Yang et al., 2003; Smolik et al.,
2005; He et al., 2005; Thatcher and Layton, 1995;
Thatcher et al., 2003).
On the other hand, a number of specific parameters
concerning indoor pollution still remain unclear. One of
the above parameters is the time lag between indoor
concentrations and outdoors when a rise of outdoor
concentrations is detected. A time lag of this type is
often observed in experimental data and has been reg-
ularly reported (Zhu et al., 2005; Hussein et al., 2005;
Morawska et al., 2001; Koponen et al., 2001; Jones
et al., 2000; Freijer and Bloemen, 2000; Ekberg, 1996)
but not studied in detail. Another aspect that has been
rarely studied is the “smoothing” of the indoor concen-
trations, i.e. the reduction of the fluctuation of the indoor
concentrations compared with the respective outdoor
values (Freijer and Bloemen, 2000). This particular
223
aspect of the indoor environment was employed in order
to theoretically deduce the ventilation rate between the
indoor and outdoor environment for a totally inert pol-
lutant (Switzer and Ott, 1992).
Time lag and indoor to outdoor fluctuation reduction
have been recognized and extensively studied in other
fields of building physics, with respect to thermal
comfort (Yam et al., 2003), but their influence on indoor
air quality with respect to indoor pollution has not been
explicitly described. The objective of this work was to
demonstrate two methods, an analytical and a numerical
one, in order to study the time lag and the fluctuation
reduction of the concentrations in the indoor environ-
ment. The differences between the two methods were
explored with the aid of statistical methods. A second
goal was to utilize the results of both methods in order to
create nomograms and theoretical relationships that can
be used in order to describe indoor air phenomena and
estimate the deposition rates in the indoor environment.
The methods were applied to a dataset obtained from the
Urban Aerosol project in order to estimate the indoor
deposition rate of sulphur dioxide in a one room apart-
ment. The significance of the studied parameters was
examined from the exposure assessment perspective.
2. Numerical approach
The numerical method involved an application of
the indoor air quality model MIAQ (Multichamber
Indoor Air Quality model), which has been used in
indoor air studies (Nazaroff and Cass, 1987, 1989;
Drakou et al., 1998). MIAQ is a general mathematical
model for both indoor aerosol dynamics and the con-
centrations of chemically reactive compounds in indoor
air. It accounts for the effects of ventilation, filtration,
direct emission, deposition onto surfaces, and coagula-
tion for particles. It also accounts for heterogeneous
removal, and photolytic and thermal chemical reactions
Fig. 1. Time evolution of the indoor concentrations, given by the
MIAQ model, along with the outdoor concentrations.
224 C.H. Halios, C.G. Helmis / Science of the Total Environment 381 (2007) 222–232

for reactive gasses. Model results have been validated
against experimental data (Nazaroff and Cass, 1987,
1989; Drakou et al., 1998).
In the present study the outdoor concentration was
considered to have the form of a triangular pulse. This
approach is valid for some pollutants, for example NO,
for which the time evolution in the outdoor environment
is similar to a triangular pulse. The ratio of the fluc-
tuations of the indoor concentrations to the outdoor ones
and the time lag were estimated using MIAQ outputs for
several values of the deposition velocity (measured in
cm s− 1), the ventilation rate (measured in air changes
per hour — ACH) and duration of the outdoor con-
centration pulse.
In Fig. 1 a typical example, representative of the ones
obtained by the numerical approach, is presented, that
gives the time evolution of the indoor concentrations
(output of the model) and the outdoor concentrations
(set as input in the model experiment). In this example a
one room compartment with a 30 m3 volume and a total
of 60 m2 indoor surfaces was considered, with a venti-
lation rate of 1 ACH while the deposition velocity of the
pollutant was considered to be 2 × 10− 4 cms− 1. The
outdoor concentration is a triangular pulse that lasts for

Fig. 2. Nomograms that describe the relationship between the time lag, the deposition velocity and the ventilation rate for different duration of the
outdoor concentrations. a. and b.: the duration of the outdoor pulse is 7 h. c.: the duration of the outdoor pulse is 5 h. d.: the duration of the outdoor
pulse is 3 h e.: the duration of the outdoor pulse is 1 h. In a. isolines corresponds to 10 min. In b., c., d. and e isolines corresponds to 2 min.
 C.H. Halios, C.G. Helmis / Science of the Total Environment 381 (2007) 222–232 225

three hours with a peak value of 70 ppb while the longer than 60 min. These findings are in agreement with
calculated indoor peak value was 48 ppb. The time step the results of other studies (Shair and Heitner, 1974;
for the model runs was set equal to 1 min. Freijer and Bloemen, 2000).
From this figure a prominent time lag can be es- In Fig. 3a, b, c and d the respective nomograms of
timated of the order of 37 min that was considered to be the ratio of the fluctuations in the ventilation rate-
the time interval between the outdoor and indoor peak deposition velocity plane are presented. Fig. 3a, b, c, and
values. It is also apparent that the indoor concentrations d correspond to 7 h, 5 h, 3 h and 1 h durations of the
are more “smoothed” compared to the outdoor ones, outdoor concentrations. The ratios of the indoor
indicating that the fluctuation of the outdoor concentra- fluctuations to the outdoor ones increase for an increase
tions is higher compared to that of the indoor air. The of the ventilation rate assuming constant values of the
fluctuations of the concentrations were estimated by deposition velocity and period of the outdoor concen-
means of the concentration range under different ven- trations. On the other hand, the deposition velocity has
tilation and deposition rates. Homogeneous chemistry the opposite effect: the decrease of the deposition
between the various pollutants was again considered to velocity values increases the fluctuation values as the
have limited importance when compared to other factors ventilation rate is kept constant. The fluctuation ratio
involved. gets higher as the outdoor pulse lasts for longer time
Results from the MIAQ model, analogous to the case intervals. Results agree with the ones found by Switzer
previously presented, were obtained for various values and Ott, 1992 even though their approach was applied to
of the ventilation rate and the deposition velocity. The a non reactive pollutant and thus no deposition was
deposition velocity was considered to have values in taken into account.
the range between 2 × 10− 4 cm s− 1 to 2 × 10− 2 cm s− 1 in
accordance with the values found by Cano-Ruiz et al.
(1993), while the ventilation rate ranged between 0.01
ACH to 40 ACH according to Said (1997), even though
ventilation rates of the order of 0.01 are not found in the
building environment (Maldonado, 1998). The simula-
tion results were then used to create nomograms that
give the dependence of the time lag from ventilation rate
and duration of the outdoor concentration pulse for
different deposition velocities.
In Fig. 2a, b, c, d and e nomograms of the time lag are
presented. These nomograms are presented as time lag
isopleths in the ventilation rate-deposition velocity
plane. Fig. 2a, b correspond to 7 h duration of the out-
door concentrations, while Fig. 2c, d and e correspond to
5 h, 3 h, and 1 h respectively. These nomograms allow
the approximate graphical computation of the time lag as
a function of the ventilation rate and the deposition
velocity. Time lag values decrease as the ventilation rate
increases, pointing to the fact that the outdoor originating
pollution enters more easily to the indoor environment.
On the other hand, time lag increases with the decrease of
the deposition velocity for the same ventilation rates. The
influence of the ventilation rate on the time lag is more
prominent for small values of the deposition velocity,
and the decrement of the deposition velocity enhances
more efficiently the time lag values for small ventilation
rates. It is apparent that the time lag increases as the
period of the outdoor pulse gets longer. Thus, for 7 h Fig. 3. Nomograms that describe the relationship between the ratio of
the indoor fluctuations to the outdoor ones, the deposition velocity and
duration of the outdoor concentrations, time lags of the the ventilation rate. a. Corresponds to 7 h duration of the outdoor
order of 200 min may be observed, while for 1 h duration concentrations and each isoline is 0.1. b., c., and d. Corresponds to 5, 3
of the outdoor concentrations the time lags can not be and 1 h respectively and each isoline is 0.05.
226 C.H. Halios, C.G. Helmis / Science of the Total Environment 381 (2007) 222–232
3. Analytical approach
The analytical approach is an extension of the ap-
proach proposed by Shair and Heitner, 1974 and
analogous to the one employed for the calculation of
the thermal mass into buildings (Yam et al., 2003). The
mass balance equation for a well-mixed indoor envi-
ronment (concentration gradients are neglected) of
volume V and surface A was considered. Air enters
with an airflow rate q0, P0 is the penetration rate of the
pollutant, Ci and C0 are the indoor and outdoor con-
centrations respectively. S is the net rate at which the
pollutant is added due to internal sources (direct emis-
sions) and Ho is the net production or loss rate due to
homogeneous chemical reactions. The pollutant's depo-
sition onto indoor surfaces is parameterised with the use
of the deposition velocity vd. Then the general form of
the mass balance equation is according to the following
relationship:
VdCi =dt ¼ qo Po Co  qo Ci  vd ACi þ S þ Ho
The above equation, taking P0 to be unity (Freijer and
Bloemen, 2000), in a more convenient, non-dimensional
form can be written as:
dCi =dt þ dCi ¼ eCo þ r
where:
d ¼ qo =V þ md A=V ; e ¼ qo =V and r ¼ S þ Ho=V :
The solution of the above equation is:
Z It is worthwhile to mention that the mathematical
t
Ci ¼ eexpðdt Þ expðdt ÞCo dt
0Z
t
þ expðdt Þ expðdt Þrdt þ Cio expðdt Þ
0
In the following, the outdoor concentration is taken
to vary sinusoidally with time following the equation
Co = ΔC sin (ωt), where ω is the angular velocity of
the pulse, an hypothesis which is valid for certain pol-
lutants, such as O3 (Shair and Heitner, 1974; Sabersky
et al., 1973). It should be noted that the previous
representation of C0 is valid only for values of ωt N 0.
Substituting Co = ΔC sin (ωt) in the above equation
then:
r e
Ci ¼ þ DC ðdsinðxt Þ  xcosðxt ÞÞ
d  x2 þ d2  indoor sources was considered. In Fig. 4 the combined
xe r effect of the ventilation rate, the deposition velocities
þ DC  expðdt Þ
x2 þ d2 d and the duration (period) of the outdoor pulse on the
The Eq. (4) is a global attractor (Yam et al., 2003) and
is the sum of three parts. The first part represents the
contribution due to internal sources, or homogeneous
chemical reactions; the second part gives the influence
of the time lag development and the third part decays to
zero as time increases (Shair and Heitner, 1974). As the
third term decays to zero, the solution approaches the
following periodic relation:
r e
Ci ¼ þ 1=2 DCsinðxt  bÞ ð5Þ
d x2 þ d2
If the various pollutants do not undergo significant
chemical reactions and there are no indoor sources active,
then the first term is negligible and the time lag can be
expressed as
b ¼ 1=x arctan ðx=dÞ ð6Þ
and the indoor reduction of the concentrations fluctuation
ð1Þ can be expressed as:
e
 1=2 ð7Þ
x2 þ d2
Both parameters are functions of the frequency ω of
ð2Þ the outdoor concentration fluctuation, the parameter ε
that accounts for the ventilation of the indoor environ-
ment (measured in air changes per hour — ACH) and
the decay rate due to the deposition onto indoor surfaces
(expressed as the product of the deposition velocity with
the ratio of the indoor surfaces to the total indoor
volume).
expressions of the time lag and the ratios of the indoor to
outdoor fluctuations can be generated by the complex
number z = δ + iω in which the imaginary part accounts
ð3Þ
for the period of the outdoor concentration, while the
real part stands for intrinsic indoor characteristics. Time
lag, as expressed in Eq. (6), is the argument of z divided
by its imaginary part, while the ratio of the indoor to
outdoor fluctuations (Eq. (7)), is ε divided by the mod-
ulus of z. Thus, it is suggested that the complex number
z that accounts for the combined effects of both the
outdoor and indoor environment could be useful in
theoretical studies related to indoor air quality.
In the following paragraph, results obtained by the
analytical approach are presented. For simplification
reasons an ideal naturally ventilated compartment with a
surface to volume ratio equal to 2 (A/V = 2) and with no
ð4Þ
 C.H. Halios, C.G. Helmis / Science of the Total Environment 381 (2007) 222–232
Fig. 4. Time lag obtained from the analytical solution as a function of
the ventilation rate and the deposition velocities, for different periods
of the outdoor pulse.
time lag is presented. It is evident that the time lag is
negatively related to the ventilation rate when the depo-
sition velocity and the duration of the outdoor pulse are
kept constant and the same stands for the relation be-
tween the time lag and the duration pulse. The same is
the case for the relationship between the time lag and
deposition velocity for constant values of the ventilation
rate and the outdoor pulse duration, while the opposite is
true for the relationship between the time lag and the
duration of the outdoor pulse with constant the depo-
sition and ventilation rate. In Fig. 5 the ratio of the
fluctuations of the indoor concentrations versus the
outdoor values as a function of the ventilation rate and
deposition velocity for different periods of the outdoor
pulse are demonstrated. It is apparent that with the
Fig. 5. The ratio of the fluctuation of the indoor concentrations versus
the outdoor values obtained from the analytical solution as a function
of the ventilation rate and deposition velocity for different periods of
the outdoor pulse.
227
deposition velocity and the ventilation rate being
constant the fluctuation ratio is positively related to
the ventilation rate. The same stands for the relationship
between fluctuation ratio and deposition velocity with
constant ventilation rate and outdoor pulse duration. In
general, the plots of the analytical method qualitatively
and quantitatively agree with the respective results es-
timated using the numerical approach. The relationship
between the two approaches will be further examined
below.
4. Comparison between the two approaches
In order to examine the agreement between the
analytical and the numerical methods, a comparison was
made between the time lag and fluctuation calculations
using Eqs. (5) and (6) and the respective values that
were obtained using the numerical solution. Due to the
different form of the outdoor concentration pulse in the
analytical and the numerical methods (sinusoidal and
triangular respectively), the two pulses were considered
to have the same magnitude and the same volume of the
contaminants, i.e. equal area below the curves (dimen-
sions were concentration and time). The comparison
was established by means of the Pearson's correlation r
and the p-value of the Wilcoxon two-sample test of
subgroups of pairs of time lag and fluctuation ratios
estimated with the analytical and numerical method. The
coefficient r gives the quality of a least squares fitting to
the data, while the Wilcoxon method tests the statistical
significance of the means of the time lag (or fluctuation
ratio) values calculated with the two methods. In Table 1
the estimated values of the Pearson correlation coeffi-
cient r and the p-values of the Wilcoxon two-sample
test between subgroups of results obtained by the two
methods are presented. Each subgroup includes 30 pairs
of time lag and fluctuation ratios calculated with the two
methods, taking into consideration the same values
of the deposition velocity, the ventilation rate and the
Table 1
Pearson correlation coefficients (r) and Wilcoxon p-values between
subgroups of results obtained by the two methods
Deposition velocity (cm s− 1)
0 0.0002 0.002 0.02
Time lag r 0.996 0.996 0.998 0.999
Wilcoxon 0.00006 0.00003 0.00002 0.00003
Ratio of indoor r 0.989 0.990 0.996 0.999
to outdoor Wilcoxon 0.01401 0.01013 0.00016 0.00002
fluctuations
The duration of the outdoor concentration pulse was set equal to 7 h.
All r-values are statistically significant at p b 0.005.
228 C.H. Halios, C.G. Helmis / Science of the Total Environment 381 (2007) 222–232
duration of the outdoor pulse. Results clearly show that
there is an excellent correlation between results obtained
by the two methods (r values higher than 0.9 for all the
cases). Also, from the p-values obtained by the
Wilcoxon two-sample test it is apparent that there are
no significant differences between the two methods in
each subgroup, indicating that results obtained from the
two methods are from the same population.
5. Estimation of the indoor deposition velocity
The theoretical relationships and nomograms ob-
tained by the two methods can be applied to exper-
imental data in order to estimate the indoor deposition
rates. The time lag and the fluctuation reduction can be
easily estimated from concurrent indoor and outdoor
measurements. The nomograms presented at Figs. 2
and 3 can be used in order to approximately estimate the
deposition velocity in the cases that ventilation rates are
known and constant. For example, if a time lag of the
order of 30 min is observed in the time series of a
hypothetical pollutant and the ventilation rate is mea-
sured to be 0.95 ACH then from Fig. 7a the deposition
velocity can be estimated as 0.01 cm s− 1.
In the following, an application of the method is
described, in order to estimate the indoor deposition rate
of SO2 corresponding to a sinusoidal outdoor pulse. The
SO2 concentrations were measured in the indoor and
immediate outdoor environment of a one-room apart-
ment in Athens during the Urban Aerosol experimental
campaign, in the winter of 2002. The Urban Aerosol
campaign aimed at the characterization of air pollutants
and PM both indoors and outdoors and the associated
actual human exposure in selected European urban areas
(Lazaridis et al., 2001). Details about the experimental
Fig. 6. Time dependence plot of indoor and outdoor concentrations of SO2 for the time interval 18/3/2002 17:00 to 19/3/2002 0:00 (Urban Aerosol
experimental campaign).
procedure can be found elsewhere (Halios et al., 2003;
Halios and Helmis, 2005).
The time dependence plot of indoor and outdoor
concentrations of SO2 for the period 18/3/2002 17:00
LST to 19/3/2002 0:00 LST is presented in Fig. 6. The
indoor environment had surfaces that cover an area of
139.2 m2 with 66.6 m3 volume, thus the A/V ratio is
2.1 m− 1. The ventilation rate was measured to be 1 ACH.
The outdoor SO2 concentrations resembled a sinusoidal-
like pulse with a period of 12 h. The time lag was
estimated to be 17 min (outdoor and indoor peak values
occurred at 18/3/2002 19:49 LST and 20:06 LST re-
spectively) and the concentrations range were estimated
at 11 ppb and 39 ppb for the indoor and outdoor envi-
onment respectively. From relationships (6) and (7) the
deposition rates were calculated to be about 7 × 10− 4 s− 1.
The indoor environment was a one room apartment in
a 30 years old apartment block. The floor was wooden
and part of it was covered with a synthetic carpet.
Relative humidity ranged between 25% and 30% during
the time intervals presented here. The deposition velocity
value for SO2 was estimated to be 0.033 cm s− 1 and is
within the range found in the literature (Nazaroff and
Cass 1987; Grontoft and Raychaudhuri, 2004) for the
same values of relative humidity and similar materials of
the indoor surfaces.
6. Exposure assessment
From Fig. 1 it is apparent that a 3 h outdoor pulse
lasts for 7 h in the indoor environment. Consequently,
the exposure of the inhabitants to outdoor originating
pollutant lasts longer in the indoor than the outdoor
environment. According to Monn (2001), “integrated
exposure can be calculated by integrating the
 C.H. Halios, C.G. Helmis / Science of the Total Environment 381 (2007) 222–232
Fig. 7. The ratio of the exposures in the indoor environment to the outdoor ones, as a function of the ventilation rate for different periods of the outdoor
pulse and deposition velocities.
concentration over time”. In order to explore the
combined effect of the time lag and the fluctuation
reduction on the human exposure, in the following we
present the ratios of indoor exposure to the outdoor ones
for different values of the deposition velocity, ventila-
tion rates and durations of the outdoor pulse. Exposures
were calculated by integrating the indoor and outdoor
concentrations, like the ones presented in Fig. 1, over
time. The results are presented in Fig. 7. It is clear that
for the scenarios examined (pollution is introduced from
the outdoor environment indoors) indoor exposures are
always less than the outdoor ones. For ventilation rates
usually found in the indoor environment with closed
windows (less than 1 ACH) the exposure to pollutants
with small deposition velocities (less than 0.002 cm s− 1)
is greater than 80% of the respective exposure outdoors.
On the other hand, pollutants with high deposition
velocities (higher than 0.02 cm s− 1) cannot penetrate
Fig. 8. Time dependence plot of the second and third term of Eq. (4)
and their sum.
229
effectively in the indoor environment (exposure ratios
lower than 0.4). Thus, the indoor environment protects
the inhabitants from the exposure to pollutants of high
deposition velocities more effectively than from those of
low deposition velocities. It is apparent that the duration
of the outdoor pulse does not influence significantly the
indoor exposure.
7. Limitations and extension of the method
In the presentation of the analytical method, it was
stated that Eq. (5) is valid for values of the exponential
part of Eq. (4) very close to zero. In order to stress the
limits of this assumption, the time evolution of each of
the three parts (the periodic, the exponential and the sum
of them) for various values of the ventilation rate and the
deposition velocity was calculated. In Fig. 8 results for
ventilation rate with values 1 ACH and deposition
velocity with value of 2 × 10− 3 cms− 1 are presented. For
reasons of clarity, all parts were allowed to go negative
as well as positive. It can be observed that the third term,
that initially has large values, rapidly decays to zero,
allowing thus the left hand term of Eq. (4) to be a
function only of the second term (the periodic part). It
should be noted that the exponential term decays more
slowly as the deposition velocity and the ventilation
rates get smaller. Thus, it should be expected that this
approach gives reliable results for relatively high values
of both the deposition velocity and the ventilation rate.
Up to this point, only idealistic situations have been
considered and our methods account neither for indoor
sources nor for chemical reactions. Complete absence of
chemical reactions is a strong idealistic assumption
which in the majority of cases is not met (Weschler and
Shields, 1997). In the presence of significant indoor
homogeneous chemistry or direct indoor emissions, the
230 C.H. Halios, C.G. Helmis / Science of the Total Environment 381 (2007) 222–232

Table 2 assumed that the indoor temperature was 297 K and

Statistical values of the simulation considered for the numerical the simulation location was set to Los Angeles. The
experiments with photochemistry and reaction rates calculated by the
model
maximum outdoor photolysis rate was calculated by the
model to be 0.53 min− 1. This value is calculated for the
Statistical values NO NO2 O3
simulation location during midday and under clear
Mean (ppb) 9 42 12 skies, and is in accordance with the values reported in
Standard 11 20 15
Mc Rae et al. (1982). The model calculates the indoor
deviation
(ppb) photolysis rate with the methodology described in
Min (ppb) 1 0 2 Nazaroff and Cass (1987). The same methodology was
Max (ppb) 50 70 60 recently incorporated by Carslaw (2007) for the devel-
Reaction rates NO + O3 → 23.68 ppm− 1 min− 1 opment of a detailed chemical model for indoor
NO2 + O2
pollution. For these calculations a light attenuation
NO2 + h.ν → Time dependent Mean value
NO + O 0.05 min− 1 factor is required as input to the model. This factor
max value corresponds to the attenuation of the outdoor light as it
0.13 min− 1 is transmitted indoors through windows. Following
min value Nazaroff and Cass (1987) and Drakou et al (1998), we
0 min− 1
assumed that 70% of the visible light and 25% of the
UV component is transmitted indoors. The model cal-
first term in Eq. (5) cannot be neglected and thus Eqs. (6) culated reactions rates are presented in Table 2. The
and (7) are not solutions of the equation. But, in cases outdoor and indoor concentrations of NO2 resulting
when chemistry or indoor emissions are small enough from this numerical simulation are presented in Fig. 9.
compared to the other term of the equation, fractions (6) Even though indoor concentrations are higher when
and (7) are solutions of Eq. (5). In the next paragraphs reactions are taken into consideration when comparing
different scenarios that are usually met in the indoor to the no chemistry (basic) scenario, the ratio of the
environment will be explored, in order to signify the fluctuations during the two simulations is very close
importance of these remarks. (0.68 for the basic simulations and 0.71 for the chem-
Consider that the example presented in Fig. 1 refers istry scenario). The time lags observed during the two
to the time evolution of the indoor and outdoor concen- simulations are equal. It should be stressed though
trations of NO2 and that mean indoor and outdoor NO that there are instances when indoor chemistry could
and O3 concentrations have the values that are presented radically alter the diurnal pattern of the concentrations
in Table 2. Numerical experiments were performed, leading thus to completely different time lags and
analogous to the ones described in the previous par- fluctuation ratios.
agraph, but in this scenario, chemical reactions among In Fig. 9, the src_1 and src_2 indoor concentrations
the three pollutants was taken into consideration. It was refer to results from numerical experiments during

Fig. 9. Time evolution of the indoor concentrations under different scenarios, given by the MIAQ model. chem corresponds to numerical simulations
when photochemistry was taken into consideration, src_1 to simulations with direct indoor emissions from four cigarette, src_2 to direct indoor
emissions from one cigarette and vent to simulations with varying simulation rates. Basic corresponds to numerical simulations with no chemistry or
direct emissions and with constant ventilation rates. Details are discussed in the text.
 C.H. Halios, C.G. Helmis / Science of the Total Environment 381 (2007) 222–232
which indoor sources of various strength were taken
into consideration. Indoor sources correspond to NO2
release during cigarette smoking and it was assumed
that the NO2 production was 0.79 mg per cigarette
(Moschandreas et al., 1985). During src_1 scenario it
was assumed that four cigarettes were smoked between
300 min and 320 min, while scenario src_2 refers to the
smoking of one cigarette from 300 min to 305 min. No
chemistry was taken into consideration. It is apparent
that the daily pattern of indoor NO2 concentrations is
distorted when compared to the daily pattern observed
during the basic scenario. This deviation is more pro-
nounced during the src_1 scenario when completely
different fluctuation ratios are observed (1.12 and 0.75
for src_1 and src_2 respectively). During src_1 sce-
nario there is no time lag observed. In the same figure
the time evolution of NO2 concentrations with varying
ventilation rate is presented (vent). It was assumed that
during 180 min–240 min the ventilation rate was 0.2
ACH, during the time interval 300 min–360 min 6
ACH and during the time intervals 240 min–300 min
and 420 min–500 min 10 ACH (values usually met
when the windows are open). Ventilation rate was 0.5
ACH during the rest of the day. It is clear that the
fluctuation ratios are substantially greater and the time
lag reduced to 5 min.
The methods presented here can be applied to
particulate matter penetrating from the outdoor envi-
ronment as well as to gaseous pollutants. Time lags have
been observed in experimental time series of aerosol
particles (Hussein et al., 2005), as well as of gaseous
pollutants (Freijer and Bloemen, 2000). Furthermore,
Eq. (1) has been modified to size specific mass balance
equation with respect to aerosol number or concentra-
tions in a number of studies (Hussein et al., 2005; Riley
et al., 2002; Zhu et al., 2005; Smolik et al., 2005).
Eq. (1) is then modified as follows:
VdNi ðnÞ=dt ¼ qo Po ðnÞNo ðnÞ  qo Ni ðnÞ
 vdi ANi ðnÞ þ Q  J ð8Þ
The index (n) indicates that the equation is valid for
the particular size range (n), V is again the volume and A
the surface of the indoor environment, q0 the airflow
rate, P0(n) the penetration rate at the specific size range,
Ni(n) and N0(n) are the indoor and outdoor concentra-
tion at the specific size range respectively. Q is the
source strength and J is the sink strength. The last two
terms account for direct emissions, resuspension, co-
agulation, growth, nucleation and chemical transforma-
tions. Then, d ¼ qo Po ðnÞ=V þ vd A=V ; r ¼ ðQ  J Þ=V .
Assuming that there are no indoor particle sources, that
231
particles are not resuspended, they are physically and
chemically stable, and there is no coagulation, the
source and sink term may be neglected and the so-
lutions to Eqs. (6) and (7) can be extended to include
particulates.
The methods can be applied to rooms with a com-
bination of natural and mechanical ventilation, as well
as recirculation of the indoor air, adding the respective
terms in the equations described above.
8. Conclusions
In this work two characteristic parameters of the
indoor environment, the time lag and the reduction of
the indoor fluctuations, were theoretically examined
with the aim of numerical and analytical solutions of the
mass balance equation. It was found that the time lag of
the indoor concentrations versus the outdoor ones is
inversely proportional to the deposition and ventilation
rates, and proportional to the duration of the outdoor
concentrations. On the other hand, a decrease in the
ventilation rate and duration of the outdoor concentra-
tions results in a rapid fall of the fluctuation ratio of
indoor to outdoor concentrations. The deposition rate
decrease results in an increment of the fluctuation ratio.
It was found that the complex number z = δ + iω that
accounts for the characteristic time scales of both the
indoor and outdoor environment could be useful in
indoor air quality theoretical studies.
The results obtained by the two methods were com-
pared and found to be in fairly good agreement. Sen-
sitivity tests of the analytical methods showed that
results obtained for very small values of the deposition
velocity and the ventilation rate should be examined
with caution. The nomograms and equations presented
here are valid for constant ventilation rates, weak indoor
emissions and negligible chemistry when compared to
the effect of deposition and ventilation. Under varying
ventilation (opening and closing of the windows), strong
indoor emissions and significant chemical reactions,
results from the methods presented here should be
considered with caution. In such cases, the time lag
values may vary with time, or not be observed at all.
When the methods are applied to particulates, the pen-
etration factor for each size bin should be evaluated
carefully.
From an exposure perspective it was found that the
indoor environment protects the inhabitants from the
exposure to pollutants of high deposition velocities
more effectively than from those of low deposition
velocities. For ventilation rates usually found in the
indoor environment with closed windows the exposure
232 C.H. Halios, C.G. Helmis / Science of the Total Environment 381 (2007) 222–232
to pollutants with small deposition velocities is about
80%.
The theoretical relationships and nomograms that
have been produced can be used for the estimation of
indoor deposition rates. The methods have been im-
plemented for indoor and outdoor concentrations of
sulfur dioxide at a one-room experimental site during
the Urban Aerosol project and the deposition rate was
calculated to be 7 10− 4 s− 1.
Acknowledgments
This work was supported by the General Secretariat
for Research and Technology of the Ministry of Devel-
opment, in the frame of the project ‘PENED 2001’.
References
Carslaw N. A new detailed chemical model for indoor air pollution.
Atmos Environ 2007;41(6):1164–79.
Cano-Ruiz JA, Kong D, Balas RB, Nazaroff WW. Removal of reactive
gasses at indoor surfaces: combining mass transport and surface
kinetics. Atmos Environ 1993;27A:2039–50.
Drakou G, Zerefos C, Ziomas I, Voyatzaki M. Measurements and
numerical simulations of indoor O-3 and NOx in two different
cases. Atmos Environ 1998;32(4):595–610.
Ekberg LE. Relationships between indoor and outdoor contaminants in
mechanically ventilated building. Indoor Air 1996;6(1):41–7.
Freijer JI, Bloemen JTh. Modeling relationships between indoor and
outdoor air quality. J Air Waste Manage Assoc 2000;50:292–300.
Grontoft T, Raychaudhuri MR. Compilation of tables of surface
deposition velocities for O3, NO2 and SO2 to a range of indoor
surfaces. Atmos Environ 2004;38:533–44.
Halios CH, Helmis CG. An indoor air quality study regarding outdoor
originating pollutants. IASME Trans 2005;4(2):646–50.
Halios CH, Helmis CG, Flocas HA, Assimakopoulos VD, Elefther-
iadis K. A statistical analysis of aerosol precursors and black
carbon in indoor and outdoor environment at a typical residential
apartment. J Aerosol Sci 2003;S34:385–6.
He CR, Morawska L, Gilbert D. Particle deposition rates in residential
houses. Atmos Environ 2005;39(21):3891–9.
Hussein T, Hameri K, Heikkinen MA, Kulmala M. Indoor and outdoor
particle size characterization at a family house in Espoo–Finland.
Atmos Environ 2005;39:3697–709.
Jamriska M, Morawska L. Quantitative assessment of the effect of
surface deposition and coagulation on the dynamics of submic-
rometer particles indoors. Aerosol Sci Tech 2003;37:425–36.
Jones AP. Indoor air quality and health. Atmos Environ 1999;33: 4535–64.
Jones NC, Thornton CA, Mark D, Harrison RM. Indoor/outdoor
relationships of particulate matter in domestic homes with roadside,
urban and rural locations. Atmos Environ 2000;34: 2603–12.
Koponen IK, Asmi A, Keronen P, Puhto K, Kulmala M. Indoor air
measurement campaign in Helsinki, Finland 1999: the effect of out-
door air pollution on indoor air. Atmos Environ 2001;35:1465–77.
Lazaridis M, Asimakopoulos DN, Branis M, Colbeck L, Drossinos I,
Dye C, et al. Characterization of urban air quality indoor/outdoor
particulate matter chemical characteristics and source-to-inhaled
dose relationships (Urban Aerosol). J Aerosol Sci 2001;32(S1):
1075–6.
Mc Rae GJ, Goodin WR, Seinfeld JH. Mathematical modelling of
photochemical air pollution. Final Report of Research under ARB
Contract Nos. A5–046–87 and A7–187–30; 1982.
Maldonado E. Naturally ventilated buildings. In: Allard F, editor.
Natural Ventilation in Buildings. A Design Handbook. UK: James
and James ltd.; 1998. p. 255–324.
Monn C. Exposure Assessment of air pollutants: a review of spatial
heterogeneity and indoor/outdoor/personal exposure to suspended
particulate matter, nitrogen dioxide and ozone. Atmos Environ
2001;35(1):1–32.
Morawska L, He C, Hitchins J, Gilbert D, Parappukkaran S. The
relationship between indoor and outdoor airborne particles in the
residential environment. Atmos Environ 2001;35:3463–73.
Moschandreas DJ, Relwani SM, O'Neill HJ, Cole JT, Elkins RH,
Macriss RA. Characterization of emission rates from indoor com-
bustion sources. Chicago, Illinois: Gas Research Institute; 1985
(Report No. GRI 85/0075).
Nazaroff WW, Cass GR. Mathematical modelling of chemically re-
active pollutants in indoor air. Environ Sci Technol 1987;20(9):
24–934.
Nazaroff WW, Cass GR. Mathematical modeling of indoor aerosol
dynamics. Environ Sci Technol 1989;24:66–7.
Riley WJ, Mckone TE, Lai ACK, Nazaroff WW. Indoor particulate
matter of outdoor origin: importance of size-dependent removal
mechanisms. Environ Sci Technol 2002;36:200–7.
Sabersky RH, Sinema DA, Shair FH. Concentrations, decay rates and
removal of ozone and their relation to establishing clean indoor air.
Environ Sci Technol 1973;7(4):347–53.
Said MNA. Measurenments of air change rates and air flow patterns in
large single cell buildings. Energ Build 1997;26:175–82.
Shair FH, Heitner KL. Theoretical model for relating indoor pollutant
concentrations to those outside. Environ Sci Technol 1974;8(5):
444–51.
Smolik J, Lazaridis M, Moravec P, Schwarz J, Zaripov S, Zdimal V.
Indoor aerosol particle deposition in an empty office. Water Air
Soil Pollut 2005;165(1–4):301–12.
Switzer P, Ott W. Derivation of an indoor air averaging time model from
the mass balance equation for the case of independent source inputs
and fixed air exchange rates. J Expo Anal Environ Epidemiol
1992;2(S2): 113–35.
Thatcher TL, Layton DW. Deposition, resuspension, and penetration of
particles within a residence. Atmos Environ 1995;29(13):1487–97.
Thatcher TL, Lunden MM, Revzan KL, Sextro RG, Brown NJ. A
concentration rebound method for measuring particle penetration and
deposition in the indoor environment. Aerosol Sci Tech 2003;37:
847–64.
Weschler C, Shields CH. Potential reaction among indoor pollutants.
Atmos Environ 1997;31(21):3487–95.
Yam J, Li Y, Zheng Z. Nonlinear coupling between thermal mass and
natural ventilation in buildings. Int J Heat Mass Transfer 2003;46:
1251–64.
Yang W, Lee K, Chung M. Characterization of indoor air quality using
multiple measurements of nitrogen dioxide. Indoor Air 2003;14:
105–11.
Zhu Y, Hinds WC, Krudysz M, Kuhn T, Froines J, Sioutas C.
Penetration of freeway ultrafine particles into indoor environ-
ments. J Aerosol Sci 2005;36:303–22.