–Annotate the article–Research into human experience of this kind of place–Relevant organizational philosophy, goals, procedures, and policies related to your building type–Guidelines, standards or codes for design of this building type–Current design issues and trends, as reflected in the news, popular press, and professional journals of this building type.
Unformatted Attachment Preview
Building and Environment 112 (2017) 359e366
Contents lists available at ScienceDirect
Building and Environment
journal homepage: www.elsevier.com/locate/buildenv
Ten questions concerning thermal and indoor air quality effects on the
performance of ofﬁce work and schoolwork
Pawel Wargocki*, David P. Wyon
DTU-ICIEE, Technical University of Denmark, Denmark
a r t i c l e i n f o
a b s t r a c t
Received 7 October 2016
Received in revised form
8 November 2016
Accepted 10 November 2016
Available online 14 November 2016
Energy conservation in buildings as a way to reduce the emission of greenhouse gases is forcing an
urgent re-examination of how closely thermal and air quality conditions should be controlled in
buildings. Allowing conditions to drift outside the optimum range would conserve very large amounts of
energy and would in most cases have only marginal effects on health or subjective comfort. The question
that then arises is whether occupant performance would be negatively affected and if so, by how much.
This information is required for cost-beneﬁt analyses. The answers in this paper are based on laboratory
and ﬁeld experiments that have been carried out since the massive increase in energy costs that took
place in the 1970s. Although only a few of the mechanisms by which indoor environmental effects occur
have been identiﬁed, it is already clear that any economies achieved by energy conservation will be
greatly exceeded by the costs incurred due to decreased performance. Reducing emissions by allowing
indoor environmental conditions to deteriorate would thus be so expensive that it would justify greatly
increased investment in more efﬁcient use of energy in buildings in which conditions are not allowed to
deteriorate. Labour costs in buildings exceed energy costs by two orders of magnitude, and as even the
thermal and air quality conditions that the majority of building occupants currently accept can be shown
to reduce performance by 5e10% for adults and by 15e30% for children, we cannot afford to allow them
to deteriorate still further.
© 2016 Elsevier Ltd. All rights reserved.
Indoor air quality
1. Do thermal conditions affect ofﬁce work?
It seems that the thermal state of the body, however achieved, is
what affects arousal and thus performance: no difference in the
performance of a whole battery of different tasks was found between two conditions of thermal neutrality with very different
clothing insulation and air temperature: 0.6 clo at 23 C and 1.15 clo
at 19 C . In other words, thin clothing and warm air is equivalent to warm clothing and cool air, in terms of the resulting effect
on both performance and thermal comfort.
Under moderately warm conditions, above neutrality, it is
possible to avoid sweating by reducing metabolic heat production.
This leads to a lowering of arousal, as people relax and generally try
less hard to work fast. This is often a completely unconscious
response to warmth. Aspects of mental performance with a low
optimal level of arousal, such as memory  and creative thinking
, are improved by exposure to a few degrees above thermal
* Corresponding author.
E-mail address: email@example.com (P. Wargocki).
0360-1323/© 2016 Elsevier Ltd. All rights reserved.
neutrality, but they too are impaired at higher temperatures, closer
to and above the sweating threshold.
Tham and Willem  showed that increased accuracy in the
Tsai-Partington test indicates raised arousal, which improves concentration and would thus be expected to beneﬁt rule-based logical
thinking. They exposed heat acclimatised subjects living in the
Tropics to 20, 23 and 26 C and observed higher arousal at 20 C, as
indicated by measurements of stress biomarkers in saliva. Activation of the sympathetic nervous system and higher alertness are
beneﬁcial for tasks that require attention and the ability to sustain
prolonged mental effort.
Individual control of the thermal environment may be necessary
if optimal performance of ofﬁce work is to be achieved. Wyon 
showed that individual control equivalent to ±3 K would be expected to improve the performance of mental tasks requiring
concentration by 2.7%. A decrease of this magnitude (2.8%) in the
rate of claims-processing in an insurance ofﬁce had been demonstrated by Kroner et al.  when individual microclimate control
devices in an insurance ofﬁce were temporarily disabled. Wyon
 showed that this degree of individual control (±3 K) would be
P. Wargocki, D.P. Wyon / Building and Environment 112 (2017) 359e366
expected to improve the group mean performance of routine ofﬁce
tasks by 7%.
Fig. 1 shows the magnitude of the effects of temperature on
performance that have been documented in the literature. Fig. 2
shows an experimental relationship between thermal discomfort
and the performance of ofﬁce work.
2. Do thermal conditions affect schoolwork?
A comprehensive set of experiments on the effects of classroom
temperatures on the performance of schoolwork was carried out in
50 years ago Sweden . In these experiments, three parallel
classes of 9e10-year-old children were exposed for 2 h to each of
three classroom temperatures: 20, 27 and 30 C, encountered in
balanced order, and four classes of 11e12-year-old children were
similarly exposed to 20 and 30 C in the morning and the afternoon
in a 2 2 design, again in balanced order of presentation of conditions . The temperatures were artiﬁcially raised in these experiments. The children performed a number of school exercises,
including numerical tasks (addition, multiplication, numberchecking) and language-based tasks (reading and comprehension,
supplying synonyms and antonyms) so that their rate of working
and the number of errors they made could be quantiﬁed. The
children’s performance of both types of task was signiﬁcantly lower
at 27 and 30 C in comparison with 20 C. In the numerical tasks,
the effect was on rate of working, but both reading comprehension
and reading speed were reduced by raised temperatures. Performance tended to be lower, though not signiﬁcantly lower, at 27
than at 30 C, and the negative effects of raised classroom temperatures were signiﬁcant in the afternoon, when the children were
fatigued, but not in the morning. The magnitude of the negative
effect of temperature on performance was for some tasks as great as
Field intervention experiments were carried out in elementary
school classrooms in late summer by the present authors [47,48]. In
these experiments, the air temperature was reduced to about 20 C
by operating split cooling units that had been installed in the
classrooms for the purpose of the experiments. In the ﬁrst experiment, the average air temperature in the classrooms was about
20 C when cooling was provided and 23.6 C in the uncontrolled
reference condition, so that the difference was 3.6 K. In the second
experiment, with different children in the same two classrooms the
following summer, the air temperature in the classrooms was
21.6 C when cooling was provided and 24.9 C in the reference
condition, i.e. a 3.3 K difference. The classrooms in which the experiments were conducted were all mechanically ventilated with
100% outdoor air. The interventions were implemented in a
Fig. 1. The relationship between temperature and the performance of ofﬁce work that
was derived by Seppanen et al.  from a review of the literature.
Fig. 2. The experimental relationship between thermal discomfort due to heat and the
performance of ofﬁce work according to .
crossover design that was balanced for order of presentation: each
experiment was carried out in two parallel classrooms at a time and
each condition lasted for a week. In alternate weeks, the improved
condition was imposed in one classroom, the other classroom
acting as the reference condition during that week, and the conditions were then switched between the classrooms for the
following week. The children acted as their own controls, so any
observed differences in performance between conditions cannot
have been due to differences between groups of children. Both
teachers and pupils were blind to interventions. During the experiments, the teachers and pupils were allowed to open the
windows and doors as usual and the teaching environment and
daily routines were as normal. During each week the children
performed tasks resembling typical schoolwork: addition, multiplication, subtraction, number comparison, logical thinking (i.e.
grammatical reasoning), reading and comprehension, proofreading against dictation. The duration of the tasks was short
enough to ensure that children could not normally complete them
in the time available. Up to 10 min was allocated for each task.
When the temperature was reduced to 20 C, the performance of 2
arithmetical and 2 language-based tests improved signiﬁcantly, as
had been reported by Holmberg and Wyon . This time the
improvement was always in terms of speed of performance, not
accuracy. The average magnitude of the effects as a function of
Fig. 3. The experimental relationship between classroom temperature and the performance of schoolwork that was found by Wargocki and Wyon .
P. Wargocki, D.P. Wyon / Building and Environment 112 (2017) 359e366
temperature is shown in Fig. 3.
Sarbu and Pacurar  examined the performance of 18 university students in a classroom over a period of 36 days. The
classroom temperatures were manipulated by operating or idling
the air cooling and changing the air cooling set-point; they
remained between 22 and 29 C. The students repeatedly performed two attention tests: the concentrated attention test (Kraepelin test) and the distributive attention test (Prague test). These
tests took 10 min and 7 min to complete, respectively. The results
showed that the performance of students on both tests followed an
inverted-U curve, similarly to what was observed for the effects of
temperature on the performance of ofﬁce work by adults [20,32]
and in earlier experiments of Wyon . The temperatures for
optimal performance in each test were different: higher (about
27e28 C) for the Prague test and lower (about 25e26 C) for the
Kraepelin test, exactly as would be expected if the change in performance followed the psychological theory of arousal – moderately
higher temperatures reduce the arousal/stress level and are thus
beneﬁcial for tests requiring high cue-utilisation, as is the case for
the Prague test.
Haverinen-Shaghnessy and Shaughnessy  measured temperatures in 140 ﬁfth-grade classrooms in 70 elementary schools
for a week-long period and correlated them with a state-wide
assessment of learning. Average indoor temperature was 23 C
and it varied between about 20 C and 25 C. Modelling showed
that there was a 12e13% increase in mathematics score for each 1
decrease of temperature (0.5%/degree). Park  analysed more
than 4.5 million school-leaving examination results from New York
City high schools and found that taking the examination on a day
when the ambient temperature was 32 C increased the risk of
failing to pass by 12.3% compared with the results obtained when
the examination was taken on a day when the ambient temperature was 22 C. He showed additionally that an increase in temperature of 1 C will reduce the examination score by 0.4% which is
quite similar to was observed by Haverinen-Shaughnessy and
Shaughnessy . There was also an effect on learning, as an
increased number of hot days (above 27 C) during the school-year
prior to taking the school-leaving examination had a negative effect
on the results obtained in the examination.
3. How do thermal conditions affect performance?
Thermal conditions can affect the performance of ofﬁce work by
at least 6 different mechanisms : (i) Thermal discomfort
distracts attention; (ii) Warmth lowers arousal (the state of activation of an individual) [29,41], exacerbates and increases the
prevalence of building-related symptoms (SBS), causing distraction
and thus further negative effects on performance, and has a negative effect on cognition [19,52]; (iii) Cold conditions decrease ﬁnger
temperatures and thus have a negative effect on manual dexterity
; (iv) Rapid temperature swings have the same effects on ofﬁce
work as slightly raised room temperatures ; (v) Vertical thermal gradients reduce perceived air quality at head height and so
lead to a reduction in room temperature that then causes complaints of cold at ﬂoor level that are due to vasoconstriction, not to
the low temperatures at ﬂoor level ; and (vi) Raised temperatures can result in increased carbon dioxide (CO2) concentration in
the blood, which may cause headaches .
In interpreting experiments on thermal and air quality effects on
ofﬁce work and schoolwork, it should be assumed that thermal
discomfort sensations due to cold or heat stress distract attention,
and that the physiological responses to heat stress reduce arousal
and therefore motivation to exert effort. More generally, the effects
of thermal conditions appear to be mediated directly by the
physiological changes that take place, including vasoconstriction in
the cold, which reduces manual dexterity, and an increase in the
blood gas level of CO2 in response to heat, which causes headache
and increased difﬁculty in thinking clearly. These suggested
mechanisms are depicted as a diagram in Fig. 4.
Hancock and Vasmatzidis  suggested that a zone of psychological and physiological adaptability exists, in which people
can tolerate thermal stress and within which there will be no effects on cognitive performance. However, as indicated by Lan et al.
: increasing temperatures and the physiological responses that
then occur, including an increase in the blood gas concentration of
CO2, result in a range of negative health symptoms such as headaches or difﬁculty in concentrating and thinking clearly that can
reasonably be expected to have direct negative effects on cognitive
performance unless more effort is exerted to counteract them.Tanabe et al.  showed that exerting more effort in this way is only
possible for a limited time, so in stipulating the thermal conditions
in which adults or children will have to work all day and every day,
it should not be assumed that extra effort will always be exerted to
counteract the physiologically adaptive decrease in arousal and
motivation that occurs in warm conditions (as a way of decreasing
metabolic rate and thus the heat that must be lost to the indoor
environment). The Adaptive Thermal Comfort model  assumes
that this behavioural and physiological response to warmth will
Fig. 4. The mechanisms by which mental work is affected by the thermal environment.
P. Wargocki, D.P. Wyon / Building and Environment 112 (2017) 359e366
take place if adjustments in clothing insulation and air velocity
prove inadequate, and that this will increase “thermal acceptability”. Although this is true, it has been shown above that it will
be at the cost of a substantial decrease in the performance of both
ofﬁce work and schoolwork.
4. Does subjective acceptance imply optimal performance?
As perceived thermal discomfort almost inevitably accompanies
the physiological changes that occur in response to thermal conditions that deviate from thermal neutrality, it has not yet been
possible to determine whether subjective acceptance of thermal
discomfort would be sufﬁcient to remove the direct effects of
physiological responses on performance, although this seems
5. Does indoor air quality affect ofﬁce work?
A series of experiments by Wargocki et al. [42,44,45], summarised by Wyon , examined the effects of indoor air quality on
the cognitive skills that are essential for ofﬁce work. In an intervention experiment , the indoor air quality in a normal ofﬁce
was altered while the health, comfort and performance of the occupants were measured. Indoor air quality was altered by
decreasing the pollution load, i.e. by physically removing a hidden
pollution source without informing the subjects, while maintaining
an outdoor air supply rate of 10 L/s per person. The major pollution
source was a carpet that had been used for 20 years in an ofﬁce and
was present behind a partition in a quantity corresponding to the
ﬂoor area of the ofﬁce in which the exposures took place, although
low background emissions and the bioefﬂuents emitted by the
subjects themselves were always present. Thirty subjects performed different cognitive tests and typing and addition tasks to
simulate typical ofﬁce work throughout 4.5 h exposures. The indoor
air quality caused 70% to be dissatisﬁed with the air quality when
the used carpet was present and 25% when it was absent. The
presence of the used carpet caused subjects to type 6.5% more
slowly, to make 18% more typing errors, and to experience more
Lagercrantz et al.  replicated the is study of Wargocki et al.
 with a different group of subjects and the same carpet. This
time the resulting indoor air quality caused 60% of the subjects to
report that they were dissatisﬁed with the air quality when the
carpet was present and 40% to be dissatisﬁed when the carpet was
absent. The subjects typed 1.5% more slowly and made 15% more
errors in an addition task. Meta-analysis of the two studies by
Wargocki et al.  conﬁrmed that their results are compatible.
The original study of Wargocki et al.  was repeated in the
same ofﬁce by Bako-Biro et al.  with a different pollution source.
Instead of carpet, there were 6 Personal Computers (PCs) with
Cathode Ray Tube Visual Display Units (CRT/VDU) that had been in
operation for about 500 h, corresponding to approximately 3
months of normal ofﬁce use. 40% of the subjects reported that they
were dissatisﬁed with the air quality when the PCs were present
behind the screen, while only 10% were dissatisﬁed when the PCs
were absent. Although the differences between conditions in terms
of typing speed were small, more subjects typed slowly and all
subjects made more typing errors when the PCs were present.
Proofreading was affected negatively but not signiﬁcantly.
Combining the observed effects on the speed and accuracy of
typing and the decrease in speed of proofreading, it could be shown
that overall text-processing would be performed 9% more slowly if
PCs were present, thus validating the ﬁndings of  with a very
different source of pollution.
Indoor air quality can also be modiﬁed by changing the
ventilation rate . The outdoor air supply rate was increased
from 3 to 10 or to 30 L/s per person with the original pollution
source, a used carpet, always present behind the partition. 60%
were dissatisﬁed with the resulting indoor air quality at the lowest
ventilation rate and 30% were dissatisﬁed at the highest rate. By
integrating speed and accuracy into an overall measure it was
possible to show that the performance of the text-typing task
improved by about 1% for every two-fold increase in the outdoor air
supply rate. The performance of the addition and proofreading
tasks followed the same trend, but the effects did not reach signiﬁcance. In an open-ended test of creative thinking, subjects
provided 10% more answers and more original answers at 10 L/s per
person than at 3 L/s per person. As stress (raised level of arousal) is
inimical to creative thinking (see Answer 1 above), this result is in
accordance with a more recent ﬁnding by Zhang et al.  that poor
air quality has the physiological effect of increasing arousal.
In another study in which the outdoor air supply …
Purchase answer to see full