Content uploaded by C. P.G. Roelofsen
Author content
All content in this area was uploaded by C. P.G. Roelofsen on Jan 15, 2020
Content may be subject to copyright.
Performance loss in open-plan
offices due to noise by speech
Paul Roelofsen
GrontmijjTechnical Management, Amersfoort, The Netherlands
Abstract
Purpose – The purpose of this paper is to show how the performance of the personnel can be
negatively affected by conversations, adjacent to the working space, in an open-plan office
environment.
Design/methodology/approach – Using two scientific mathematical models it is possible to
quantify the performance losses, as a result of adjacent conversations, with various desk layouts in an
open-plan office.
Findings – The results obtained from the underlying Finnish study can be improved by applying a
regression analysis to the recorded research results. The modified deviation formula not only
corresponds more closely to the research results, but also produces a better translation of the speech
intelligibility index to the intelligibility qualifications, as shown in the guideline NPR 3438.
Originality/value – It is reasonable to conclude that performance loss, as a result of a poor acoustic
situation, can be related to the speech intelligibility in a space. The relationship between the speech
intelligibility index and performance loss makes it possible to design on the basis of productivity
improvement, resulting in a comfortable acoustic working environment and a consistent financial
advantage for the organisation.
Keywords Employee productivity, Performance levels, Noise control, Acoustics, Open plan offices,
Finland
Paper type Research paper
Introduction
More and more people are currently employed in the services sector and spend most of
their working days inside a building. A working environment which is as healthy and
comfortable as possible is therefore not unimportant, if people are to function
undisturbed and at their best within an organisation.
Noise, especially in an enclosed space, is an important aspect which affects both the
feeling of “well-being” and concentration in regard to functional performance.
Personnel costs are of a significantly higher factor than the costs for organisational
accommodation, therefore any productivity losses, expressed in financial terms, clearly
represents large amounts of money. Noise is often defined as undesired sound. Any
sound, however, that is disturbing is, in principle, noise. This can often mean that it is
not the particular sound or the magnitude of the sound which is defined as noise, but
the perception of the sound by the subject listener to which extent the sound is
experienced as noise. This is also relative to the working circumstances at that moment
in time. Everyday noise or noise within a noisy environment can be described as noise
when it disturbs or affects the capacity of the person to perform the working function.
Noise causes personnel to become irritated and interrupted and to perform less well,
especially in situations requiring creativity and thought, and may cause, occasionally,
short-term memory loss. Noise has a negative and often a slowing down effect upon
performance and memory. Noise in offices is a problem which is, currently and
The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1472-5967.htm
JFM
6,3
202
Journal of Facilities Management
Vol. 6 No. 3, 2008
pp. 202-211
qEmerald Group Publishing Limited
1472-5967
DOI 10.1108/14725960810885970
unfortunately, rarely recognised by the management in a corporate organisation
(Russel, 1999).
Two major reasons for the increase in objectionafble noise in the work place can be
blamed upon the introduction of open-plan offices and the transformation of formerly
closed office work areas into transparent and open character work areas. Magnification
of the problem has been caused by the intensification of the occupancy levels and the
use of speaker enabled telephone and computer communications.
This paper will show, by means of two scientific models, how the performance of
the personnel can be negatively affected by conversations, adjacent to the working
space, in an open-plan office environment.
The open-plan office
Office spaces can be roughly divided into two categories. Open office spaces with high
concentration levels of work areas (Plate 1), subdivided, or not, by screens and filing
cabinets and, conventional individual compartmentalised offices separated from each
other (Figure 1).
Open-plan offices are often preferred, partly because the working areas within the
office space are 50 per cent smaller than in conventional, individual, office
accommodations.
Alterations to the layout within open offices are simpler to carry out. An open-plan
office is also easier to rent. Furthermore, the economic advantages of open-plan offices
can be improved by the following:
(1) The shorter working distances.
(2) The promotion of better:
.communications;
.information routes;
Plate 1.
Open office space
Performance loss
in open-plan
offices
203
.colleague relations;
.work involvement;
.transparency and openness; and
.fresh and modern architecture.
Undesirable acoustic consequences – due to, for example, lack of speech privacy and
reduced concentration as an outcome of overhearing other conversations – are often
ignored, or not taken seriously, during strategic accommodation decisions, because of
the apparent advantages of the open-plan office economics and organisational benefits;
since it is expected that the personnel will, in due course, adjust to the new office noise
situation.
From various studies, it has been shown that speech (incidental and/or formal
meeting discussions, telephone conversations, etc.) form the most disturbing source of
noise in an open office accommodation. The disturbance is not, apparently, due to any
increase of speech volume, whether individually or due to a multitude of voices
speaking simultaneously but, and in particular, as a result of over-hearing a
conversation which is distracting. Studies reveal that once a person has been distracted
it will take ten minutes or more for that person to return to the same concentration level
they were at before the distraction. In practice, everyday, in an open-plan office, there
are virtually constant conversations which can be of a disturbing nature (Folsom and
Koenig, 2007). The reduction of the speech intelligibility disturbance between various
working areas in open-plan offices takes on great importance.
Figure 1.
Examples of work areas
within open office spaces
JFM
6,3
204
Speech intelligibility versus speech privacy
Speech intelligibility
The speech intelligibility in a space can be evaluated on the basis of the speech
transmission index (STI) as mentioned in NEN-EN-ISO-9921 (2003). The STI is a
measurable and calculable value for the speech intelligibility, whereby the values vary
between 0 (not intelligible) to 1 (perfectly intelligible). Speech is modulated by a test
signal which has particular speech characteristics; given that speech can be described
as a broad band white noise modulated with particular fluctuation frequencies. At the
receiver location, the modulation depth of the received signal, over various frequency
bands, is compared to the test signal transmitted. The speech intelligibility is related to
the reduction in the modulation depth.
Speech privacy
Speech privacy is the complete opposite of speech intelligibility and can best be shown
on the basis of the Privacy Index, as follows:
Privacy Index ¼12STI ð2Þ:
The relationship speech intelligibility index, speech intelligibility and speech privacy
is shown in Table I.
For a good speech privacy, a speech intelligibility index equal to or smaller than 0.2
is advised (Bradley et al., 2004). In order to evaluate the speech intelligibility between
two adjacent working areas, whether or not separated by a sound barrier, this study
uses the Wang and Bradley model (Wang et al., 2002).
Other influential aspects
When reading, the sounds of a conversation are only a disturbance if there is a
similarity between the text being read and the content of the conversation, and not by
the loudness of the speech. Conversations have a more negative influence than any
back ground noise, if the conversation concerns a recognisable subject.
Conversations that one cannot follow, for example in a foreign language, are much
less disturbing. Noise can also have a negative influence on the memory.
Conversational noise makes memory retention more difficult. Other sorts of noises
are disturbing but to a lesser extent. Performance studies carried out in noisy
situations but with good acoustic environments do not always reveal large differences.
This can be answered by understanding that one compensates and subdues daily and
familiar background noises by a higher level of concentration. However, this very often
results in tiredness, lack of comprehension and irritation. One performs less well and
cooperation is poorer with colleagues (Russel, 1999).
Speech intelligibility index (-) Speech intelligibility Speech privacy
0.00-0.30 Bad Good
0.30-0.45 Poor Reasonable
0.45-0.60 Reasonable Bad
0.60-0.75 Good Very bad
0.75-1.00 Excellent No
Table I.
Translation of the speech
intelligibility index to the
intelligibility qualities,
NPR 3438 (2006)
Performance loss
in open-plan
offices
205
Human beings have an almost instinctive reaction to sound. There is always a reaction
to a new sound and this causes a lack of focus in the activity being carried out. The
consequences can be that important details are overlooked and/or the creative thought
processes are detrimentally disturbed.
Mans sensory perceptions keep a persons active and subconsciously aware of what
is occurring around them, but only to a certain level. Above this level, any disturbance
is disruptive. The disturbance level is variable from person to person and is dependent
upon the work activity, the situation, other overriding sensory impressions, etc. Sound
can also result in performance improvement. Certain monotonous manual work, such
as painting and decorating and production line work can perceive sound as acceptable.
Conversely, more difficult tasks demand concentration. The activity level of an
individual is also influenced by personal circumstances, such as health, sleep, drugs,
medications, work pressures, etc.
Noise becomes a disturbance when one is deciding how to carry out a particular
activity. Unwelcome noise causes one to diminish the value of and ignore available
information. The result is that the activity is not always performed in the best manner.
Noise is an influence on strategic solutions to a problem. If one is affected by noise,
human nature chooses the simplest most favoured strategy, even when, subsequently,
circumstances change and the noise disappears.
Noise increases working pressure. A problem that couldbe solved by using a little more
effort may be carried out, but less effectively, despite every intentions of more effort.
Noise reduces performance capability, even when the noise has subsided and may
later result in tiredness, irritability and depression.
Noise disturbs verbal communication and makes speech and hearing less
distinct. A noisy environment demands more effort to speak and listen. The difficulty
is increasedwhen the listener must also remain concentrated on theirwork (Russel, 1999).
Performance loss
It is reasonable to conclude that performance loss, as a result of a poor acoustic
situation, can be related to the speech intelligibility in a space. For this purpose, the
research results from the Institute of Occupational Health in Finland (Hongisto, 2005)
are used. A summary of the results from the Finnish study show that the performance
loss varies between 4 and 45 per cent, dependent upon the task. The best performance
was shown to occur when there was no speech present (STI ¼0) and the largest
performance loss occurred when speech was perfectly intelligible (STI ¼1). Between
these two extremes, the assumption is that, the performance loss, as a function of the
speech intelligibility, in principle, follows a similar curve to the subjective speech
intelligibility, as a function of the speech intelligibility index, conforming to IEC
60268-16 with a maximum of 7 per cent, which is the minimum performance loss,
within the parameters of the previously mentioned inventory for proofreading (see the
lower curve with squares in Figure 2).
In the Finnish study, the model was not influenced by a regression analysis of the
research results. Consequently, and for this presentation, it seems sensible to perform a
regression analysis upon the results in order to evaluate the extent to which this may
affect the eventual model. The results are graphically shown in Figure 2 and the upper
curve with circles illustrates the influence of the regression analysis.
JFM
6,3
206
By means of the Boltzmann sigmoid function, which is advised in the previously
mentioned study, the deviation in the prediction model due to regression analysis
conforms to the following formula (see upper curve with circles in Figure 2):
DP ¼8:03ð121=ð1þEXPððSTI 20:30Þ=0:06ÞÞÞ
where DP – performance loss (per cent); and STI – speech intelligibility index (-).
The maximum of the modified deviation formula appears to be somewhat higher
than the Finnish study. Larger differences are also apparent between the speech
intelligibility indexes 0.2 and 0.5. In our opinion, the application of this deviant formula
not only corresponds more closely to the research results but produces a better
translation of the speech intelligibility index to the intelligibility qualifications, as
shown in the NPR 3438 (2006). The model should be further developed by continuing
the studies of Ellermeier and Hellbru
¨ck and Venetjoki et al. as mentioned in the Finnish
study. For the record it is pointed out that culture may be of influence. This should be
further investigated too in future research.
Situation under consideration
To obtain an impression of the performance loss by an employee in an open-plan office
situation, a calculation study has been carried out. An open work area is considered, as
shown in Figure 3, whereby the screen height, the sound absorption coefficient of the
ceiling and the background noise levels are variables. In order to remain in line with
the previously mentioned Finnish study, their proposed prediction model is used
instead of the modified deviation function.
Calculation results
The calculated results are shown graphically in Figures 4-6.
Figure 2.
Curve fitting and the
research results of the
Institute of Occupational
Health in Finland
as a function of the speech intelligibility
Performance loss [%]
Speech intelligibility index [–]
14
12
10
8
6
4
2
00 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Hongisto Curve fitting Ellermeier & Hellbruck Venetjpki Kaarlela
Performance loss
Source: Hongisto (2005)
Performance loss
in open-plan
offices
207
Conclusions
Based upon of the observations, the following can be concluded:
.Using the previously described models, it is possible to quantify the performance
losses, as a result of adjacent conversations, with various desk layouts in an
open-plan office.
.The results obtained from the Finnish study can be improved by applying a
regression analysis to the recorded research results.
.The modified deviation formula, in our opinion, not only corresponds more
closely to the research results, shown in Figure 2, but produces a better
translation of the speech intelligibility index to the intelligibility qualifications,
as shown in the NPR 3438 (2006).
Figure 3.
Plan considered work
station diagram (top view):
two desks opposite of each
other; lights (open grill)
over screen
Screen height: 1.7m
3m 3m
3
m
Figure 4.
Summary of calculated
results
Performance loss
Performance loss [%]
Mean sound absorption ceiling
7
6
5
4
3
2
1
01 1.1
0.42 0.61 0.9 0.97
1.2 1.3 1.4 1.5 1.6 1.7 1.8
Screen height [m]
as a function of the speech intelligibility, the screen height and the
sound absorption of the ceiling
Note: Background noise level 41.9 dB(A)
JFM
6,3
208
.The maximum performance loss for office workers influenced by intelligible
adjacent conversations would be approximately 8 per cent, using the modified
deviation formula.
Figure 5.
Summary of calculated
results
Performance loss
Performance loss [%]
Screen height [m]
6
5
4
3
2
1
01 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8
Note: Background noise level 44.9 dB(A)
Mean sound absorption ceiling
0.42 0.61 0.9 0.97
as a function of the speech intelligibility, the screen height and the
sound absorption of the ceiling
Figure 6.
Summary of calculated
results
Performance loss
5
4
3
2
1
01.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8
0.42 0.61 0.9 0.97
Performance loss [%]
Screen height [m]
Note: Background noise level 46.9 dB(A)
Mean sound absorption ceiling
as a function of the speech intelligibility, th screen height and the
sound absorption of the ceiling
Performance loss
in open-plan
offices
209
.By applying the modified deviation formula, the curves shown in Figures 4-6
would be drawn somewhat higher in the graphical representation.
.Especially, for the speech intelligibility index interval between 0.2 and 0.5, the
modified deviation formula produces a larger off-set compared with the Finnish
study.
.The speech intelligibility index should be lower than 0.5 to produce any positive
influence upon performance.
.In order to obtain acceptable conversation privacy, a speech intelligibility index
lower or equal to 0.2 is advised. (Bradley et al., 2004).
.An increase in background noise level causes a worsening of the speech
intelligibility of the participants and a decrease in performance loss.
An open-plan office requires a minimum background noise level of 45 dB(A),
where potentially disturbing conversations/discussions are carried on.
The background noise level may, however, never be higher than 48 dB(A)
(Bradley et al., 2004).
.Practically speaking, this means that ceilings with very good sound absorbing
qualities together with high separation screens should be used in order to mask
speech sounds. Without these measures, a good speech privacy is impossible in
an open-plan office environment.
.Other general measures to limit or prevent noisy activities will improve speech
privacy and increase performance.
.In addition, it is strongly recommended that during the design of an open-plan
office environment, the incorporation of smaller closed office spaces should be
planed, where employees may retreat for telephone conversations or to carry out
activities which require more creativity and/or accuracy.
References
Bradley, J.S. and Gover, B.N. (2004), “Criteria for acoustic comfort in open-plan offices”,
Inter-Noise 2004 – The 33rd International Congress and Exposition on Noise Control
Engineering Prague, Czech Republic, August 22, 2004, pp. 1-6.
Folsom, F.W. and Koenig, Th.A. (2007), “The impact of open plan speech noise on employee
productivity & satisfaction and the facility solutions”, Dynasound Inc., available at: www.
ifma-holland.nl/files/Congressen/The_Impact_of_Open_Plan_Speech_Noise_on.doc
(accessed 23 January 2007).
Hongisto, V. (2005), “A model predicting the effect of speech of varying intelligibility on work
performance”, Indoor Air, Vol. 15 No. 6, pp. 458-68.
NEN-EN-ISO-9921 (2003), “Ergonomie – Beoordeling van spraakverstaanbaarheid”, October,
NEN, Delft.
NPR 3438 (2006), Ergonomie – Geluidhinder op de arbeidsplaats – Bepaling van de mate van
verstoring van communicatie en concentratie, November, NEN, Delft.
Russel, P. (1999), Geluid en het moderne kantoor, Hilanders, Helsingborg.
Wang, C. and Bradley, J.S. (2002), “Sound propagation between two adjacent rectangular
workstations in an open plan office”, I Mathematical modelling: NRCC-46314, available at:
http://irc.nrc-cnrc.gc.ca/pubs/fulltext/nrcc46314/nrcc46314.pdf
JFM
6,3
210
About the author
Paul Roelofsen, CFM, MSc Eng., studied Architecture, Urban Planning and Housing at the
Eindhoven University of Technology specialising in the Physical Aspects of the Built
Environment. After graduation, he continued his studies through a post-Higher Professional
Course in Advanced Installation Techniques and an advanced course in Environmental Noise, as
well as a course in Facility Management and Quality Management. Paul is a visiting lecturer and
occupies various supplementary functions within both his working and tutoring activities. He is
also active with various contact groups, related to his expertise in building constructional
techniques, and performs several functions for, and within, these groups. The IFMA Award of
Excellence for outstanding achievement in Facility Management was presented to Paul in 2004
as well as an Award for Building Services Innovation in The Netherlands in 2007. Paul Roelofsen
can be contacted at: paul.roelofsen@grontmij.nl
Performance loss
in open-plan
offices
211
To purchase reprints of this article please e-mail: reprints@emeraldinsight.com
Or visit our web site for further details: www.emeraldinsight.com/reprints