The on-board carbon dioxide concentrations and ventilation performance in passenger cabins of US domestic flights

Abstract

Billions of people travel on airplanes every year, making the aircraft cabin a critical environment to understand with regard to public health. The main control over indoor environmental quality in the cabin is ventilation; therefore, maintaining sufficient ventilation rates on-board is essential for creating healthy and comfortable cabin environments. We measured real-time CO2 concentrations, an indicator of ventilation rates, and cabin pressure in the passenger cabins of 179 US domestic flights from boarding through deplaning. The average CO2 concentrations were 1353 ± 290 ppmv (mean ± SD) and the estimated outside airflow rates were 5.77 ± 2.09 L/s/p across all flights. The results indicated that 96% of observations met the minimum recommended outside airflow rates for acceptable air quality (3.5 L/s/p), but only 73% met the rate required in FAA design requirements (4.7 L/s/p), during flying phases. The CO2 levels on all flights were well below the occupational exposure limit of 5000 ppmv. Statistical analysis indicated that the ventilation rates during boarding phases were significantly lower than the others. The findings are of particular interest because low ventilation in other settings has been associated with increased rates of disease transmission, increased upper respiratory symptoms, and worse performance on cognitive function tests. Verification of ventilation performance rather than reliance on design estimates for determining compliance with ventilation standards is recommended.

Full text

Original Paper
I
Indoor
and
and B
uilt
uilt
Environment
The on-board carbon dioxide
concentrations and ventilation
performance in passenger cabins
of US domestic flights
Xiaodong Cao
1
, Christopher D. Zevitas
2
,
John D. Spengler
1
, Brent Coull
3
, Eileen McNeely
1
,
Byron Jones
4
, Sin Ming Loo
5
, Piers MacNaughton
1
and
Joseph G. Allen
1
Abstract
Billions of people travel on airplanes every year, making the aircraft cabin a critical environment to
understand with regard to public health. The main control over indoor environmental quality in the
cabin is ventilation; therefore, maintaining sufficient ventilation rates on-board is essential for creating
healthy and comfortable cabin environments. We measured real-time CO
2
concentrations, an indicator
of ventilation rates, and cabin pressure in the passenger cabins of 179 US domestic flights from board-
ing through deplaning. The average CO
2
concentrations were 1353 290 ppmv (mean SD) and the
estimated outside airflow rates were 5.77 2.09 L/s/p across all flights. The results indicated that 96% of
observations met the minimum recommended outside airflow rates for acceptable air quality (3.5 L/s/p),
but only 73% met the rate required in FAA design requirements (4.7 L/s/p), during flying phases. The CO
2
levels on all flights were well below the occupational exposure limit of 5000 ppmv. Statistical analysis
indicated that the ventilation rates during boarding phases were significantly lower than the others. The
findings are of particular interest because low ventilation in other settings has been associated with
increased rates of disease transmission, increased upper respiratory symptoms, and worse perfor-
mance on cognitive function tests. Verification of ventilation performance rather than reliance on
design estimates for determining compliance with ventilation standards is recommended.
Keywords
Aircraft cabin, Ventilation, Carbon dioxide, Airflow rate, Air quality
Accepted 20 July 2018
Introduction
The International Air Transport Association (IATA)
expects 7.2 billion passengers to travel in 2035, a near
doubling of the 3.8 billion air travellers in 2016, with a
3.7% annual Compound Average Growth Rate.
1
Transformational changes are underway, especially in
Asia and the Middle East where aviation growth is
expected to be even greater. Airplane cabin air quality
has received considerable attention. Several studies
have highlighted the in-flight transmission of infectious
disease, such as SARS, tuberculosis, common cold and
norovirus.
2–5
Other studies have focused on chemical
1
Department of Environmental Health, Harvard T.H. Chan
School of Public Health, Boston, MA, USA
2
U.S. Department of Transportation, Volpe National
Transportation Systems Center, Cambridge, MA, USA
3
Department of Biostatistics, Harvard T.H. Chan School of
Public Health, Boston, MA, USA
4
College of Engineering, Kansas State University, Manhattan,
KS, USA
5
Department of Electrical and Computer Engineering, Boise
State University, Boise, ID, USA
Corresponding author:
Joseph G. Allen, Department of Environmental Health, Harvard
T.H. Chan School of Public Health, 401 Park Drive, Boston
02215, MA, USA.
Email: jgallen@hsph.harvard.edu
Indoor and Built Environment
0(0) 1–11
!The Author(s) 2018
Article reuse guidelines:
sagepub.com/journals-
permissions
DOI: 10.1177/1420326X18793997
journals.sagepub.com/home/ibe

stressors inside the cabin, including flame retardants,
6,7
ozone,
8–10
VOCs
11,12
and particulate matter (PM).
13,14
The aircraft cabin is a unique environment that
relies on an aircraft environmental control system
(ECS) to move passengers and crew through an exter-
nal environment widely variable in temperature (80 to
>50C), pressure (10.1–101 kPa), and relative humidity
(near dry to saturation).
15
Air supplied to the aircraft
cabin is a combination of outside air brought in from
engines (bleed air) and air removed from the cabin,
which is filtered and recirculated. At the ECS terminal,
a mixing air distribution system regulates the cabin air
velocity/temperature, relative humidity and provides
sufficient fresh air to dilute gaseous and particulate
contaminants. Therefore, ventilation plays a key role
in maintaining a safe, healthy and comfort cabin envi-
ronment, by conditioning outdoor air, and controlling
in-cabin pollutants’ concentrations.
Ventilation standards for the aircraft cabin condi-
tions vary by country. In the USA, the airline industry
is regulated by the Federal Aviation Administration
(FAA). The FAA has established Federal Aviation
Regulations (FAR) to guide the operation of commer-
cial airliners. FAR 14 CFR 25.831
16
states that the
cabin ventilation system must provide at least 0.55 lb
(0.25 kg) of fresh air for each passenger per minute,
equivalent to 4.7 L/s/p at 8000 feet (2438 m) altitude
and cabin temperature of 22C (72F). This ventilation
rate is also specified by the joint design regulation
FAR/JAR Part 25
16
for crew members to perform
their duties without undue discomfort or fatigue and
to provide reasonable passenger comfort. It should be
noted that FAA only certifies performance based on
the design of aircraft ventilation system, not the
actual ventilation performance. US airlines are not
required to certify compliance through routine direct
monitoring or audits. As noted in a National Research
Council (NRC) report,
15
established FARs may be
inadequate to protect the health of the flight crew
and passengers.
In addition to federal regulations for minimum ven-
tilation rates, ASHRAE issued the ANSI/ASHRAE
standard 161–2007
17
for specifying requirements for
air quality within commercial aircraft. This standard
specifies a minimum outside air ventilation rate of 3.5
L/s/p. Further, a total air supply (mixture of outside air
and recirculated air) rate of 7.1 L/s/p should be the
minimum, and a rate of 9.4 L/s/p is recommended.
When the supply air is from the on-board systems
(i.e. the auxiliary power unit (APU) or engine bleed
air), the standard requires a minimum of 3.5 L/s/p of
outdoor air. However, if the source of supply air is
from the airport terminal or ground carts, then the
ventilation rate of 9.4 L/s/p of total air is required.
CO
2
concentration has long been used as a proxy for
ventilation,
18
using CO
2
exhaled by passengers as a
tracer gas, and as an indicator of environments that
occupants will find unsatisfactory. ASHRAE standard
62.1–2013
19
references CO
2
with respect to odour per-
ception. It states that indoor CO
2
concentrations no
greater than 700 parts per million by volume (ppmv)
above outdoor CO
2
levels in typical office buildings is
an indicator that visitors entering the space will be sat-
isfied with respect to human bioeffluents (body odour).
In addition to its use as a proxy for ventilation and
satisfaction, recent evidence suggests that CO
2
may
also be a direct pollutant. Hypercapnia (CO
2
retention)
stimulates vasodilation of cerebral blood vessels,
increasing cerebral blood flow and elevating intracra-
nial pressure, which can cause dyspnea on exertion,
headaches, visual disturbance, impaired cognitive func-
tion, and other central nervous system symptoms.
20
A recent study
20
in the space station indicated that ele-
vated CO
2
can lead to symptoms such as headache and
lethargy. Other studies found that the exposure to CO
2
at concentrations typically encountered in indoor envi-
ronments (950–2500 ppm) is associated with decre-
ments in human cognitive function.
21–24
These
impacts were observed at concentrations well below
the 5000 ppmv exposure limit set by Occupational
Safety and Health Administration (OSHA).
25
Investigation of ventilation performance in aircraft
cabin can use simulated environment chamber. Cao
et al.
26–28
used a large-scale particle image velocimetry
(PIV) system to study the air distributions in a Boeing
737 cabin mockup occupied by 42 heating manikins.
They found that the cabin airflows were at low velocity
and high turbulence levels. The low-speed air motion
may lead to poor ventilation effectiveness in the pas-
senger zone, even at a total ventilation rate of 9.4 L/s/p,
as recommended by ASHRAE. Du et al.
29
experimen-
tally studied the jet flow characteristics from a ventila-
tion nozzle in an aircraft cabin mockup. They
concluded that the effect of nozzle jet on passenger’s
thermal comfort was mostly affected by air velocity
distribution. Wang et al.
30
studied the ventilation per-
formance in a Boeing 767 cabin mockup containing 35
manikins. The CO
2
released by manikins was used as a
tracer gas to evaluate the air exchange rate and venti-
lation effectiveness in the breathing zone of passengers.
They found that the local mean age of air decreased
almost linearly with an increase in fresh air supply
rates. Strøm-Tejsen et al.
31
used a 21-seat cabin
mockup with realistic pollution sources to investigate
the balance between fresh air and humidity during a
simulated 7-h flight. The study covered four conditions
of outside airflow rate ranged from 1.4 to 9.4 L/s per
person. Sixty-eight subjects acting as passengers were
exposed to the four simulated flight conditions.
2Indoor and Built Environment 0(0)

The results showed that the reduced outside air could
intensify the symptoms commonly associated with air
travel (headache, dizziness and claustrophobia). These
studies provided comprehensive evaluation of the ven-
tilation performance in controlled cabin environments
and simulated conditions.
Some researchers have further performed on-board
measurements of ventilation on actual flights. Spengler
et al.
32
conducted environmental monitoring in the pas-
senger cabin of 83 commercial flights, including six air-
craft types (2 Airbus and 4 Boeing), flying US domestic
and international routes. Various environmental
parameters were monitored including CO
2
, particles,
ozone, volatile organic compounds (VOCs) and noise.
They found that the ventilation rates on high-passenger
load flights could be below the required 3.5 L/s/p, as
indicated by high CO
2
concentrations. The concentra-
tions of some passenger-related VOCs were also higher
for flights with reduced ventilation rates. The air qual-
ity inside cabins of Boeing 747, A330 and A340 air-
crafts were investigated by Lee et al.
33
on 16 flights.
The measured CO
2
concentrations were well below
the OSHA exposure limit but exceeded the recom-
mended value by ASHRAE standard 62.1 (about
1000 ppmv), ranging from 2000 to 2500 ppmv.
Twenty one percent of the crew considered that the
air quality was poor during the flight. Similarly, a
study of 43 flights on commercial airlines showed that
the levels of CO
2
on most flights were higher than that
recommended by the ASHRAE Standard 62.
34
Li
et al.
13
and Guan et al.
11
conducted on-board measure-
ments in the passenger cabin of Boeing 737–800 aircraft
on a few domestic flights. CO
2
concentrations were also
recorded to calculate the outside air ventilation rates,
which were then used to identify contributions of air-
borne particles and VOCs from outside and inside the
cabin. The results showed that the outside airflows on
all tested flights were within relevant aircraft cabin ven-
tilation standards limits. For PM, contributions from
the bleed air and cabin interior were both important.
On the contrary, around 90% of VOCs came from
emissions inside the cabin.
Considering the vast numbers of flights that occur
annually on a global basis, the extent of actual meas-
urements on airplanes is quite limited. Further, com-
parison of on-board ventilation performance to
regulations and standards has not been fully explored.
Lastly, little is known about ventilation and the vari-
ability of carbon dioxide throughout the entire flight
experience, and whether there may be critical periods
with inadequate ventilation. Therefore, the overall
objectives of this study were to: (1) characterize the
variability of carbon dioxide throughout all phases of
flight, and (2) estimate ventilation during the whole
flight using CO
2
as a proxy for ventilation.
Methods
Flight information
On-board monitoring was conducted on a total of 179
US domestic flights from 10 July 2007 to 13 September
2009. The flights covered 24 aircraft types, 7 aircraft
series, 10 airlines and 58 flight routes. The largest pro-
portion of aircraft types were the Boeing 757 and MD-
88 models, accounting for 37% and 30% of all tested
flights, respectively. Most of the investigated aircraft
models are short- to medium-range narrow-body
twin-engine jet airliners (see Table 1), which are com-
monly used for domestic airline services. In these air-
liners, the conditioned air was delivered by means of
diffusers placed high in the cabin, and returned
through grilles at bottom sides of the cabin. The
median sampling duration was 1 h 14 min.
On-board measurement
We measured CO
2
concentrations and cabin pressure
using a modular and portable sensing package devel-
oped by Boise State University, known as the
ASCENT 1000.
35
The ASCENT 1000 package was
fixed to the seatback pocket or was placed on the
tray table to measure CO
2
concentrations in re-
circulated air, which could represent the average CO
2
level in the cabin under the fully mixing condition. The
seat locations varied from flight to flight, mostly in
economy cabins (only two tests performed in first-
class cabins). The cabin pressure and CO
2
concentra-
tions were recorded at intervals of either 10, 30 or 60 s.
The ASCENT 1000 passed tests for electromagnetic
interference by the Boeing Company and was approved
for use on commercial flights. The ASCENT 1000
package was pre-calibrated and tested in six flights by
Loo et al.
35
Inside the package, pressure sensing
was accomplished by a VTI Technologies SCP1000
sensor module with a measuring range of 30–120 kPa
(200 Pa). CO
2
concentrations were determined using
a Telaire 6004 CO
2
OEM sensor (General Electricity
Company). The measured CO
2
concentrations were
automatically corrected to compensate for air pressure
changes. This adjustment was done in real-time by the
sensor once it received the pressure measurement. The
sensor had a calibrated detection range of 0 to 2000
ppmv (40 ppmv or 3% of readings, whichever is
greater). The narrower range could provide a better
accuracy for lower levels more likely to be encountered
in aircraft cabins. In this study, 98% of measured CO
2
concentrations were within this range. The sensor could
detect concentrations above 2000 ppmv but with lower
measurement accuracy.
Cao et al. 3

The whole flight was divided into five phases: board-
ing, ascent, cruise, descent and deplaning. The flight
phases could be clearly distinguished by changes of
cabin pressure. Boarding phase was defined as a
period of constant cabin pressure prior to take-off.
Measurement campaign typically began when most or
all the passengers have boarded. So, the boarding
phase included a fraction of boarding process and
whole taxiing process. Ascent phase was a period char-
acterized by a steady decline in cabin pressure after
take-off until a constant pressure was achieved at
cruise phase. Descent phase was a period of steady
increase in cabin pressure after the cruise phase. After
landing, the cabin pressure became constant again
during deplaning phase. The pressure and CO
2
data
were continuously monitored from ascent to descent
on most flights. However, the measurement could not
be performed during boarding or deplaning phases on
some flights due to operational restrictions.
Estimation of outside airflow rate
The transient in-cabin/ambient CO
2
concentrations,
outside airflow rate and generation rate of CO
2
per
person together follow equation (1)
11,13,36
under the
fully mixed approximation
VdCi;j
dt 103¼NQjCo;j106
NQjCi;j106þNG
3600
(1)
where Q
j
¼outside airflow rate at time j(L/s/p);
G¼average CO
2
generation rate by a person (L/hr);
C
i,j
¼CO
2
concentration inside the cabin at time
point j(ppmv); C
o,j
¼CO
2
concentration outside the
cabin at time point j(ppmv); N¼number of passen-
gers; V¼cabin volume (m
3
); t¼time (s).
Based on this formula, the outside air ventilation
rates can be estimated using measured CO
2
concentra-
tions and a steady-state approximation, also known as
the constant concentration method. This method
assumes that the in-cabin CO
2
concentration is steady
during each measuring period, and the dominant
source of CO
2
is the exhaled breath by passengers at
an average generation rate. The constant concentration
method has been successfully applied in several studies
on in-flight ventilation performance.
11,13,32,37
Because
of the high number of air changes per hour on
Table 1. Measured CO
2
concentrations and estimated outside airflow rates on all the flights.
Aircraft series
Aircraft
types
Aircraft
numbers
CO
2
concentration (ppmv) Outside ventilation rate (L/s/p)
Min Max Mean Median SD Min Max Mean Median SD
Airbus 319/320 A319 2 1003 1635 1228 1211 112 4.05 8.19 6.11 6.13 0.82
A320 3 682 2990 1175 1235 382 1.94 17.08 8.15 5.95 4.04
Boeing 737 B737-300 4 1032 1773 1479 1497 111 3.64 7.83 4.68 4.55 0.54
B737-400 1 1656 1937 1808 1804 65 3.26 3.98 3.56 3.57 0.16
B737-500 1 1000 1220 1124 1142 49 6.06 8.23 6.88 6.69 0.49
B737-700 7 850 2947 1261 1272 188 1.97 10.90 6.05 5.71 1.40
B737-800 6 661 2976 1288 1355 316 1.95 18.38 6.73 5.22 3.62
Boeing 757 B757 67 703 2992 1438 1421 284 1.94 15.95 5.19 4.88 1.52
Boeing 767 B767-300 5 756 1662 1234 1238 160 3.96 13.66 6.25 5.93 1.65
Bombardier CR-7 1 1094 2209 1903 1911 160 2.77 7.14 3.38 3.32 0.48
CR-9 5 658 2616 1451 1427 265 2.27 18.59 5.04 4.86 1.32
CRJ-100 2 620 1620 889 878 160 4.10 21.60 10.61 10.25 2.91
CRJ-140 1 1264 1566 1398 1384 79 4.28 5.76 5.02 5.07 0.39
CRJ-150 1 793 1233 969 976 127 5.97 12.42 9.07 8.58 1.88
CRJ-200 2 721 1442 1070 1049 173 4.79 15.09 7.90 7.63 2.14
Embraer E-135 1 1143 2077 1417 1381 185 2.99 6.68 5.04 5.08 0.79
E-145 1 682 2054 1201 1070 403 3.03 17.08 7.86 7.40 3.85
E-170 1 855 1352 1097 1083 160 5.23 10.78 7.49 7.25 1.73
E-175 8 686 2137 1162 1120 200 2.89 16.85 6.91 6.89 1.69
E-190 1 1392 2340 1771 1717 233 2.59 5.03 3.75 3.80 0.60
MD DC-9 3 934 1951 1340 1159 320 3.23 9.23 5.82 6.54 1.59
MD-80 1 659 1094 897 891 111 7.14 18.52 10.41 10.01 2.45
MD-88 53 514 2979 1321 1318 264 1.95 39.50 5.89 5.42 1.98
MD-90 2 656 2993 1251 1265 416 1.94 18.72 6.97 5.75 2.83
Summary 179 514 2993 1353 1333 290 1.94 39.50 5.77 5.34 2.09
4Indoor and Built Environment 0(0)

airplanes, steady-state conditions are reached more
rapidly than other indoor environments. Under the
steady-state condition, the left side of equation (1)
becomes 0, so the outside airflow rate can be estimated
by equation (2)
Qj¼106G
3600 Ci;jCo;j
ðÞ (2)
If the CO
2
concentration is changing rapidly, equa-
tion (2) will not be accurate due to the transient term in
equation (1). A simple order of magnitude assessment
showed that rates of change greater than about
1000 ppm/h may adversely impact the accuracy of the
ventilation rate calculated using the steady-state
approximation. This requirement typically will not be
met in the beginning of the boarding phase when the
concentration rises rapidly as large numbers of people
enter the aircraft. Even though equation (2) may not
give an accurate measure of the actual instantaneous
value of outside air being supplied to the cabin during
rapid transients in concentrations, the levels of CO
2
in
the air are indicative of contaminant levels in the air.
Consequently, the ventilation rates calculated with
equation (2) during transient periods may be consid-
ered as equivalent steady-state ventilation rates during
each measuring period and should be viewed in that
perspective.
The CO
2
concentration in the atmosphere generally
ranges from 365 ppmv to 390 ppmv, summarized by Li
et al.
13
Here, we used a mean outdoor CO
2
concentra-
tion of 386 ppmv, averaged from the global monthly
mean atmospheric CO
2
concentration data reported by
National Oceanic and Atmospheric Administration
(NOAA),
38
during the whole test period. The genera-
tion rate of CO
2
per person can be derived by
equation (3)
39
G¼RQð0:00276ADMÞ
0:23RQ þ0:77
ðÞ (3)
where RQ ¼respiratory quotient: 0.83 for an adult of
average size and engaged in sedentary activities;
39
M¼metabolic rate per unit surface area (METs);
A
D
¼DuBois surface area (m
2
), which can be calculat-
ed by equation (4) as
AD¼0:203H0:725W0:425 (4)
where H¼height (m) and W¼weight (kg).
The generation rates will vary for different people
depending on inputs in equations (3) and (4). Thus, a
Monte Carlo simulation was used to determine the
distribution of the CO
2
generation rate per person
from 100 air travellers.
32
The metabolic rate for each
passenger was selected using a uniform distribution in
the range of 0.9–1.3 METs, representing different phys-
ical active types from sleeping to sit/watch TV.
40
The
average generation rate of CO
2
per person from 10,000
simulation runs was 18.2 L/h.
Statistical analysis
Descriptive analysis was performed to show the mea-
sured CO
2
concentrations and estimated outside air-
flow rates by flight phases and aircraft types. Box
plots and probability distributions were used to com-
pare the data distributions by flight phases. Post-hoc
multiple comparisons using the Fisher’s least signifi-
cant difference (LSD) method
41
were performed to
test the influence of flight phases/duration on the ven-
tilation performance. A p-value <0.05 indicates a sta-
tistically significant difference for the mean value.
Results
Figure 1 depicts an example of the measured cabin
pressure and CO
2
concentrations during a typical
flight. As shown in Figure 1, the flight phases could
be clearly identified by changes in cabin pressure. The
FAR 14 CFR 25.831
16
states that the minimum cabin
pressure under normal operating conditions should not
be less than the pressure found at an altitude of 2438 m
(75.3 kPa), and all the tested flights met this standard.
This example highlights the typical variations in CO
2
concentration over the course of a flight. During the
boarding phase, the CO
2
concentrations increased rap-
idly, indicating low outside airflow rates. CO
2
concen-
trations decline steadily after taking-off and are lower,
but above levels typically acceptable in buildings (1000
ppmv) during the cruise and descent phases. Finally,
the CO
2
concentration increased again after landing.
The descriptive statistics of all measured CO
2
con-
centrations and corresponding outside air ventilation
rates by aircraft types are listed in Table 1. The CO
2
levels on all flights were well below the occupational
exposure limit of 5000 ppmv. Overall, across all aircraft
types and flight phases, the CO
2
concentrations varied
from 514 to 2993 ppmv, with an average value of 1353
290 ppmv (mean SD) and a median of 1333 ppmv.
Correspondingly, the outside airflow rates ranged from
1.94 to 39.50 L/s/p, and the average and median values
were 5.77 2.09 L/s/p and 5.34 L/s/p, respectively.
Figure 2 shows distributions of CO
2
concentrations
and corresponding outside airflow rates calculated
using equation (2) during each flight phase. Figure 2
(a) shows that CO
2
concentrations were the highest
during the boarding phase, with a median of 1539
Cao et al. 5

ppmv and an interquartile range (IQR) of 407 ppmv.
CO
2
levels during climbing were lower compared to
that during the boarding phase. The median CO
2
con-
centration was 1304 ppmv during cruise, the lowest
among all phases. During descent to deplaning, a grad-
ual increase in CO
2
levels was observed. The high CO
2
levels during the boarding phase indicated inadequate
ground ventilation by APU or gate-based ventilation
systems. As shown in Figure 2 (b), calculated ventila-
tion rate was significantly lower in the boarding phase
compared to all the other flight phases (p<0.0001),
with a median of 4.38 L/s/p (below the FAA regula-
tions) and an IQR of 1.84 L/s/p. The ventilation per-
formance gradually improved during ascent, and
reached a peak ventilation at the cruising phase, with
a median of 5.51 L/s/p and an IQR of 1.91 L/s/p.
Figure 3 depicts the percentages of in-cabin outside
airflow rates that met the two aircraft ventilation reg-
ulations. The flights met ASHRAE standard (3.5 L/s/p)
for 80% during boarding and the vast majority of time
Figure 2. Box plots of (a) CO
2
concentrations and (b) outside airflow rates by flight phases.
Figure 1. Measured CO
2
concentration and pressure during a typical flight.
6Indoor and Built Environment 0(0)

(96%) during flying, but this is a minimum for accept-
able air quality. They only met the FAR design require-
ment of 4.7 L/s/p, 42% of time during boarding and
73% of time during flying. Figure 4 shows the proba-
bility distributions of outside airflow rates during
boarding and cruising with an interval of 0.5 L/s/p.
The values higher than 20 L/s/p during boarding
(0.2%) are not shown for a clear comparison. The
black curve shown in each plot represents the corre-
sponding normal distribution with the same mean
value and standard deviation. As compared to the
black curve, the distribution of outside airflow rates
during boarding is obviously more positively skewed
(skewness ¼3.119) to the low end than during cruising
(skewness ¼1.964). Therefore, the boarding phase was
characterized by worse ventilation performance com-
pared to the cruising phase. A part of the very low
ventilation rates calculated for the boarding phase
may be not accurate due to the fact that the high
CO
2
concentration was changing rapidly in the begin-
ning of boarding. The effect of increased metabolic
activity during boarding may also account, in part,
for the increased CO
2
concentrations. Nevertheless,
the high CO
2
concentrations seen during boarding are
indicative of contaminants not being removed effective-
ly from the air by the ventilation.
Discussion
Cabin outside air ventilation rates were lower than
those typically found in other indoor environments.
Low ventilation and high physical activity levels possi-
bly increase the risk for airborne infectious disease
transmission, discomfort and other complaints. Low
ventilation rates in buildings are strongly associated
with higher risks of infectious airborne disease trans-
mission.
42
For instance, a recent study indicated that
crowded dormitories with low outdoor airflow rates
were associated with more respiratory infections
among college students.
43
Similarly, risk of disease
transmission within the aircraft cabin also seems to
be affected by ventilation.
2
A study of in-flight tuber-
culosis revealed that doubling ventilation rate within
the cabin could halve the infection risk.
44
On the
tested flights, passengers spent 20% of their boarding
time with equivalent ventilation rates below 3.5 L/s/p
and 58% of their time with rates below 4.7 L/s/p. Our
results showing low ventilation rates during boarding
also suggest that characterizing in-cabin exposures
based only on cruising conditions may overlook the
critical exposures likely occurring during ground oper-
ations. A recent Airport Cooperative Research
Program (ACRP) report
45
has highlighted the increas-
ing risks of transmission of infectious aerosols while the
aircraft is parked at the gate and APUs are shutdown.
Figure 3. Percentages of in-cabin outside airflow rates
met the aircraft regulations.
Figure 4. Probability distributions of outside airflow rates during boarding and cruise.
Cao et al. 7

The elevated CO
2
concentrations in the airplane
cabin are also of interest in light of recent studies on
impacts of CO
2
and ventilation on cognitive function
of occupants that saw effects at levels seen during all
flight phases. In a study of a controlled office environ-
ment,
21
participants were exposed to three CO
2
concen-
trations for 8 h each while they completed their normal
work activities. At the end of each day, they completed
1.5-h cognitive tests. Participants scored 15% and 50%
lower on cognitive tests on days when they were
exposed to 950 ppmv and 1400 ppmv of CO
2
, respec-
tively, compared to the reference condition of 550
ppmv. Similar findings were shown by Satish et al.
24
during 2.5-h exposure windows. Detrimental impacts
on cognitive function may influence the performance
of flight crew. In particular, the poor ventilation per-
formance during boarding warrant more attention. The
flight deck door is usually open during boarding, which
means CO
2
concentrations on the flight deck could be
influenced by the high CO
2
concentrations in the cabin.
Elevated CO
2
concentrations on the flight deck may
impact pilot performance during the subsequent taxiing
and take-off. A study by Allen et al. found that pilots
had 1.52 times higher odds of passing flight maneuvers
at 1500 ppm and 1.69 times higher odds at 700 ppm
compared to 2500 ppm, as rated by a FAA Designated
Pilot Examiner during simulated flight sessions. Pilots
were particularly impacted during maneuvers following
boarding; passing rates were twice as high at 700 ppm
and 1500 ppm during a takeoff with an engine fire than
at 2500 ppm, and four times as high at 700 ppm during
a rejected takeoff than at 2500.
46
In addition to cog-
nitive function, studies have found impacts on acute
health symptoms and physiological indicators such as
heart rate variability and peripheral blood circulation
from high indoor CO
2
levels.
47,48
Indoor CO
2
levels
higher than 800 ppm were associated with an increase
in workers’ SBS symptoms, such as eye irritation and
upper respiratory symptoms.
49
The measured ranges of CO
2
concentrations/outside
airflow rates were also comparable to other studies.
Spengler et al.
32
studied the cabin air quality during
the cruise phase of 83 flights flying US domestic and
international routes. In their study, the CO
2
concentra-
tions were 1404 297 ppmv with a range of 863–2065
ppmv, correspondingly the outside airflow rate was 5.5
1.8 L/s/p with a range of 3.0 to 10.9 L/s/p. Giaconia
et al.
37
measured the CO
2
concentrations on 14 short-
haul flights with A319 in Italy. The concentrations of
CO
2
varied between 734 and 2252 ppmv from boarding
to disembarking, which corresponded to outside air-
flows of 14.60 and 2.76 L/s/p, respectively. The mean
CO
2
concentrations on each flight ranged from 925 to
1449 ppmv, with the corresponding outside airflow
rates from 5.18 to 10.56 L/s/p. Similarly, Li et al.
13
reported that concentrations of CO
2
varied from 792
to 2253 ppmv during five domestic flights with Boeing
737–800 in China. The mean CO
2
concentration on
each flight ranged from 976 to 1135 ppmv, with the
corresponding outside airflow rates from 7.18 to 9.79
L/s/p. Hocking
50
calculated the ventilation capacity for
several representative aircraft types (Airbus, Boeing,
MD, USA) using their designed air exchange rates,
and reported that the outside air ventilation capacity
was within the range of 3.2 to 11.2 L/s/p. Our mean
outside ventilation rates were well within this range for
all 24 aircraft types.
The finding of higher CO
2
concentrations during
boarding was consistent with previous studies. In a
study of 16 flights, Lee et al.
33
found substantially
higher CO
2
concentrations during boarding and
deplaning than during cruising. They noted that CO
2
concentrations were higher during boarding and de-
boarding with typical levels of 2000 to 2500 ppmv.
Likewise, in a study of 26 intercontinental flights with
Boeing 767–300, Lindgren and Norb€
ack
51
found higher
CO
2
concentrations during non-cruise conditions due
to low air exchange rates. The CO
2
concentrations were
1656 ppmv under non-cruise conditions compared to
734 ppmv during cruise.
According to previous studies, the average bleed air
fraction of the air supplied the cabin is typically about
0.5 during the whole flight.
11,13,52
Thus, the mean total
ventilation rate may be approximately twice the mean
outside airflow rate (11.54 L/s/p). We used sampling
duration as a proxy to test the influence of flight dura-
tion. The tested flights included mostly short-haul
flights (<2 h) and the rest were medium-haul flights
(2–6 h). No significant difference for the mean outside
airflow rate can be found between the short-haul flights
with duration t<1 h and 1 h <t<2h (p¼0.944). The
ventilation rates were slightly higher on the medium-
haul flights (p<0.0001). The results of this study may
not be representative of average conditions on long-
haul international flights which have a longer cruise
phase and where high activity during boarding and
deplaning constitute a shorter fraction of the overall
flight experience.
The estimation of outside airflow rate has some
uncertainties. Through Monte Carlo simulations, we
estimated that the range of CO
2
generation rate by a
passenger is 17–19.3 L/h,
32
thus resulting a maximum
deviation of 6% for the estimated outside ventilation
rates due to uncertainty in metabolic rate. Our gener-
ation rate is consistent with that used by ASHRAE
19
(18 L/h), which corresponds to an average size adult
engaged in typical office work (1.2 METs). However,
other values have been suggested depending on sub-
jects. For example, Lee and Siconolfi
53
measured CO
2
generation rate of 23 astronauts in a seated position
8Indoor and Built Environment 0(0)

and obtained a value of 13.2 1.2 L/h. A recent study
54
suggested that CO
2
generation rates of young Chinese
people were overestimated by the empirical equation
used in our research, which should be corrected by
factors of 0.75 and 0.85 for females and males under
sitting/standing conditions, respectively. The genera-
tion rates of passengers may be increased during a frac-
tion of boarding phase when physical activity levels
were elevated. However, it is worth noting that a pas-
senger is seated during most of the boarding and taxi-
ing process before taking-off. For a sensitive test, we
assumed that on average the highest 10% of CO
2
con-
centrations was recorded during a passenger walking
inside the cabin and lifting up the luggage with a met-
abolic rate of 3 METs (implied walking, putting away
household items, moderate effort).
40
Under this
assumption, the median of ventilation rates during
boarding would increase to 4.77 L/s/p with an IQR
of 1.49 L/s/p, still significantly lower compared to
other phases (p<0.0001).
The application of constant concentration method
was based on the assumption that the in-cabin air
was well mixed by ventilation and the only source of
CO
2
was the respiration by passengers. The well-mixed
assumption has been validated by several experimental
studies.
30,55,56
However, no information was available
concerning whether there was any other source of CO
2
on tested flights (e.g. dry ice used for chilling food and
beverages in the galley systems). As discussed previous-
ly, rapid changes in CO
2
concentration will introduce
errors in the calculated rate. However, these errors
should have minimal effect on the average rates calcu-
lated over the period of a flight phase. These uncertain-
ties may lead to over- or underestimating ventilation
rates, but do not affect findings related to the absolute
CO
2
levels.
Conclusions
In this study of 179 US domestic flights, we found that
96% of the observations met the minimum recom-
mended outside airflow rates for acceptable air quality
during the flying phases (3.5 L/s/p), but only 73% met
the rate required by the FAR (4.7 L/s/p). The CO
2
levels on all flights were well below the occupational
exposure limit of 5000 ppmv. The ground operation
phases, especially the boarding phase, had higher
CO
2
concentrations than the flying phases, likely due
to the inadequate ground ventilation on many flights
and higher metabolic rates during boarding.
Ventilation rates were highest during the cruising
phase. Current ventilation requirements are often not
met, particularly during the boarding phase, which is of
concern because low ventilation in other settings has
been associated with increased rates of disease
transmission, increased upper respiratory symptoms,
and worse performance on cognitive function tests.
Verification of ventilation performance rather than reli-
ance on design estimates for determining compliance
with regulatory mandated minimum outdoor air venti-
lation rates is recommended.
Authors’ contribution
All authors contributed to the data analysis and manuscript
preparation. BJ and SML led the field sampling effort. CZ,
JGA, JDS, BC, and EM were involved in the initial data
analysis and manuscript preparation. XC and PM led the
final data analysis and manuscript preparation in coordina-
tion with JGA, BC and JDS. All authors contributed to and
approved the final manuscript.
Acknowledgements
Although the FAA has sponsored this project, it neither
endorses nor rejects the findings of this research. We thank
the anonymous peer-reviewers for their excellent comments
that have enhanced the manuscript.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with
respect to the research, authorship, and/or publication of
this article.
Funding
The author(s) disclosed receipt of the following financial sup-
port for the research, authorship, and/or publication of this
article: This study was funded partially by the US Federal
Aviation Administration (FAA) Office of Aerospace
Medicine through the National Air Transportation Center
of Excellence for Airliner Cabin Environment Research
(ACER)/Research in the Intermodal Transport
Environment (RITE), Cooperative Agreements 10-C-RITE-
HU, 07-C-RITE-HU and 04-C-ACE-HU.
References
1. IATA. IATA forecasts passenger demand to double over
20 years, http://www.iata.org/pressroom/pr/Pages/2016-
10-18-02.aspx (2016, accessed 15 June 2017).
2. Mangili A and Gendreau MA. Transmission of infectious
diseases during commercial air travel. Lancet 2005;
365: 989–996.
3. Thornley CN, Emslie NA, Sprott TW, Greening GE and
Rapana JP. Recurring norovirus transmission on an air-
plane. Clin Infect Dis 2011: 53(6): 515–520.
4. Yin S, Sze-To G and Chao CY. Retrospective analysis of
multi-drug resistant tuberculosis outbreak during a flight
using computational fluid dynamics and infection risk
assessment. Building and Environment. 2012; 47: 50–57.
5. Baker MG, Thornley CN, Mills C, Roberts S, Perera S,
Peters J, Kelso A, Barr I and Wilson N. Transmission of
Cao et al. 9

pandemic A/H1N1 2009 influenza on passenger aircraft:
retrospective cohort study. BMJ 2010; 340: c2424.
6. Allen JG, Sumner AL, Nishioka MG, Vallarino J, Turner
DJ, Saltman HK and Spengler JD. Air concentrations of
PBDEs on in-flight airplanes and assessment of flight
crew inhalation exposure. J Expo Sci Environ Epidemiol
2013; 23: 337–342.
7. Allen JG, Stapleton HM, Vallarino J, McNeely E,
McClean MD, Harrad SJ, Rauert CB and Spengler JD.
Exposure to flame retardant chemicals on commercial
airplanes. Environ Health 2013; 12: 17.
8. Bek€
o G, Allen JG, Weschler CJ, Vallarino J and Spengler
JD. Impact of cabin ozone concentrations on passenger
reported symptoms in commercial aircraft. PloS One
2015; 10: e0128454
9. Strøm-Tejsen P, Weschler CJ, Wargocki P, My
sk
ow D
and Zarzycka J. The influence of ozone on self-evaluation
of symptoms in a simulated aircraft cabin. J Expo Sci
Environ Epidemiol 2008; 18: 272–281.
10. Wisthaler A, Tama
´s G, Wyon DP, Strøm-Tejsen P, Space
D, Beauchamp J, Hansel A, M€
ark TD and Weschler CJ.
Products of ozone-initiated chemistry in a simulated air-
craft environment. Environ Sci Technol 2005;
39: 4823–4832.
11. Guan J, Li Z and Yang X. Net in-cabin emission rates of
VOCs and contributions from outside and inside the air-
craft cabin. Atmos Environ 2015; 111: 1–9.
12. Guan J, Gao K, Wang C, Yang X, Lin C-H, Lu C and
Gao P. Measurements of volatile organic compounds in
aircraft cabins. Part I: methodology and detected VOC
species in 107 commercial flights. Build Environ 2014;
72: 154–161.
13. Li Z, Guan J, Yang X and Lin C-H. Source apportion-
ment of airborne particles in commercial aircraft cabin
environment: contributions from outside and inside of
cabin. Atmos Environ 2014; 89: 119–128.
14. Ren J, Liu J, Cao X, Li F and Li J. Ultrafine particles in
the cabin of a waiting commercial airliner at Tianjin
International Airport, China. Indoor Built Environ 2017:
1420326X17713257 (in press).
15. NRC. The airliner cabin environment and the health of
passengers and crew. Washington, DC: National
Academies Press, 2002.
16. FAR 14 CFR 25.831. Part 25-Airworthiness standards:
transport category airplanes. Washington, DC: Federal
Aviation Administration (FAA), 2013.
17. ANSI/ASHRAE. Standard 161–2007, air quality within
commercial aircraft. Atlanta, GA: American Society of
Heating, Refrigerating and Air-Conditioning Engineers,
Inc, 2007.
18. Deng H-Y, Feng Z and Cao S-J. Influence of air change
rates on indoor CO2 stratification in terms of Richardson
number and vorticity. Build Environ 2018; 129: 74–84.
19. ANSI/ASHRAE. Standard 62.1-2013, ventilation for
acceptable indoor air quality. Atlanta, GA: American
Society of Heating, Refrigerating and Air-Conditioning
Engineers, Inc, 2013.
20. Law J, Watkins S and Alexander D. In-flight carbon
dioxide exposures and related symptoms: association,
susceptibility, and operational implications. NASA
Technical Paper 2010; 216126.
21. Allen JG, MacNaughton P, Satish U, Santanam S,
Vallarino J and Spengler JD. Associations of cognitive
function scores with carbon dioxide, ventilation, and
volatile organic compound exposures in office workers:
a controlled exposure study of green and conventional
office environments. Environ Health Perspect (Online)
2016; 124: 805.
22. MacNaughton P, Satish U, Laurent JGC, Flanigan S,
Vallarino J, Coull B, Spengler JD and Allen JG. The
impact of working in a green certified building on cogni-
tive function and health. Build Environ 2017;
114: 178–186.
23. MacNaughton P, Spengler J, Vallarino J, Santanam S,
Satish U and Allen J. Environmental perceptions and
health before and after relocation to a green building.
Build Environ 2016; 104: 138–144.
24. Satish U, Mendell MJ, Shekhar K, Hotchi T, Sullivan D,
Streufert S and Fisk WJ. Is CO
2
an indoor pollutant?
Direct effects of low-to-moderate CO
2
concentrations
on human decision-making performance. Environ
Health Perspect 2012; 120: 1671.
25. OSHA Carbon Dioxide, https://www.osha.gov/dts/chem
icalsampling/data/CH_225400.html (2017, accessed 15
June 2017).
26. Cao X, Li J, Liu J, and Yang W. 2D-PIV measurement of
isothermal air jets from a multi-slot diffuser in aircraft
cabin environment. Build Environ 2016; 99: 44–58.
27. Cao X, Liu J, Pei J, Zhang Y, Li J and Zhu X. 2D-PIV
measurement of aircraft cabin air distribution with a high
spatial resolution. Build Environ 2014; 82: 9–19.
28. Li J, Cao X, Liu J, Wang C and Zhang Y. Global airflow
field distribution in a cabin mock-up measured via large-
scale 2D-PIV. Build Environ 2015; 93: 234–244.
29. Du X, Li B, Liu H, Wu Y, Cheng T and Awbi HB. Effect
of nozzle air supply temperature and volume flowrate on
the jet flow from a typical ventilation nozzle in aircraft
cabins. Indoor Built Environ 2016; 27: 499–511.
30. Wang A, Zhang Y, Sun Y and Wang X. Experimental
study of ventilation effectiveness and air velocity distri-
bution in an aircraft cabin mockup. Build Environ 2008;
43: 337–343.
31. Strøm-Tejsen P, Wyon DP, Lagercrantz L and Fang L.
Passenger evaluation of the optimum balance between
fresh air supply and humidity from 7h exposures in a
simulated aircraft cabin. Indoor Air 2007; 17: 92–108.
32. Spengler JD, Vallarino J, McNeely E, Estephan H and
Sumner AL. In-flight/onboard monitoring: ACER’s com-
ponent for ASHRAE 1262, Part 2. 2012. Boston, MA:
National Air Transportation Center of Excellence for
Research in the Intermodal Transport Environment
(RITE)-Airliner Cabin Environmental Research
(ACER) Program, 2012.
33. Lee SC, Poon CS, Li XD and Luk F. Indoor air quality
investigation on commercial aircraft. Indoor Air 1999;
9: 180–187.
34. Haghighat F, Allard F, Megri AC, Blondeau P and
Shimotakahara R. Measurement of thermal comfort
10 Indoor and Built Environment 0(0)

and indoor air quality aboard 43 flights on commercial
airlines. Indoor Built Environ 1999; 8: 58–66.
35. Loo SM, Owen M, Kiepert J and Planting A. Modular,
Portable, Reconfigurable, and Wireless Sensing System
for the Aircraft Cabin. J ASTM Int 2008; 5: 1–9.
36. You Y, Niu C, Zhou J, Liu Y, Bai Z, Zhang J, He F and
Zhang N. Measurement of air exchange rates in different
indoor environments using continuous CO
2
sensors.
J Environ Sci 2012; 24: 657–664.
37. Giaconia C, Orioli A and Di Gangi A. Air quality and
relative humidity in commercial aircrafts: an experimen-
tal investigation on short-haul domestic flights. Build
Environ 2013; 67: 69–81.
38. NOAA Trends in atmospheric carbon dioxide, Mauna
Loa CO
2
monthly mean data http://www.esrl.noaa.gov/
gmd/ccgg/trends/#mlo_full (2017, accessed 15
June 2017).
39. Persily AK. Evaluating building IAQ and ventilation
with indoor carbon dioxide. Transact-Am Soc Heat
Refrigerat Air Condition Eng 1997; 103: 193–204.
40. Ainsworth BE, Haskell WL, Herrmann SD, Meckes N,
Bassett DR Jr, Tudor-Locke C, Greer JL, Vezina J,
Whitt-Glover MC and Leon AS. 2011 Compendium of
physical activities: a second update of codes and MET
values. Med Sci Sports Exercise 2011; 43: 1575–1581.
41. Williams LJ and Abdi H. Fisher’s least significant differ-
ence (LSD) test. Encyclopedia Res Des 2010;
218: 840–853.
42. Li Y, Leung GM, Tang J, Yang X, Chao CYH, Lin JZ,
Lu J, Nielsen PV, Niu J and Qian H. Role of ventilation
in airborne transmission of infectious agents in the built
environment – a multidisciplinary systematic review.
Indoor Air 2007; 17: 2–18.
43. Sun Y, Wang Z, Zhang Y and Sundell J. In China, stu-
dents in crowded dormitories with a low ventilation rate
have more common colds: evidence for airborne trans-
mission. PloS One 2011; 6: e27140
44. Ko G, Thompson KM and Nardell EA. Estimation of
tuberculosis risk on a commercial airliner. Risk Analys
2004; 24: 379–388.
45. TRB. ACRP report 91: infectious disease mitigation in
airports and on aircraft. Washington, DC: National
Academy of Sciences, 2013.
46. Allen JG, MacNaughton P, Cedeno-Laurent JG, Cao X,
Flanigan S, Vallarino J, Rueda, F, Donnelly-McLay D,
and Spengler JD. Airplane pilot flight performance on 21
maneuvers in a flight simulator under varying carbon
dioxide concentrations. Journal of Exposure Science and
Environmental Epidemiology 2018; 28(5).
47. Vehvil€
ainen T, Lindholm H, Rintam€
aki H, P€
a€
akk€
onen
R, Hirvonen A, Niemi O and Vinha J. High indoor
CO
2
concentrations in an office environment increases
the transcutaneous CO
2
level and sleepiness during cog-
nitive work. J Occup Environ Hygiene 2016; 13: 19–29.
48. Wargocki P, Wyon DP, Sundell J, Clausen G and Fanger
P. The effects of outdoor air supply rate in an office on
perceived air quality, sick building syndrome (SBS)
symptoms and productivity. Indoor Air 2000;
10: 222–236.
49. Tsai D-H, Lin J-S and Chan C-C. Office workers’ sick
building syndrome and indoor carbon dioxide concentra-
tions. J Occup Environ Hygiene 2012; 9: 345–351.
50. Hocking MB. Passenger aircraft cabin air quality: trends,
effects, societal costs, proposals. Chemosphere 2000;
41: 603–615.
51. Lindgren T and Norb€
ack D. Cabin air quality: indoor
pollutants and climate during intercontinental flights
with and without tobacco smoking. Indoor Air 2002;
12: 263–272.
52. Rydock JP. Tracer study of proximity and recirculation
effects on exposure risk in an airliner cabin. Aviation,
Space Environ Med 2004; 75: 168–171.
53. Lee S and Siconolfi SF. Carbon dioxide and water vapor
production at rest and during exercise. A report on data
collection for the crew and thermal systems division.
Houston, TX: NASA Technical Reports, 1994.
54. Qi M, Li X, Weschler L and Sundell J. CO
2
generation
rate in Chinese people. Indoor Air 2014; 24: 559–566.
55. Patel JA, Jones BW, Hosni MH and Keshavarz A.
Experimental investigation of ventilation effectiveness
in an airliner cabin mockup. In: ASME 2016
international mechanical engineering congress and exposi-
tion, Phoenix, AZ, November 11–17, 2016, pp.
V007T009A012-V007T009A012. American Society of
Mechanical Engineers.
56. Zhang Z, Chen X, Mazumdar S, Zhang T and Chen Q.
Experimental and numerical investigation of airflow and
contaminant transport in an airliner cabin mockup. Build
Environ 2009; 44: 85–94.
Cao et al. 11