- Email: [email protected]
Acoustic evaluation and adjustment of an open-plan office through architectural design and noise control
Applied Ergonomics 43 (2012) 1066e1071
Contents lists available at SciVerse ScienceDirect
Applied Ergonomics journal homepage: www.elsevier.com/locate/apergo
Acoustic evaluation and adjustment of an open-plan office through architectural design and noise control Carolina Reich Marcon Passero, Paulo Henrique Trombetta Zannin* Laboratory of Environmental and Industrial Acoustics and Acoustic Comfort e LAAICA, Federal University of Paraná, PR, Brazil
a r t i c l e i n f o
a b s t r a c t
Article history: Received 9 August 2011 Accepted 26 March 2012
Arranging office space into a single open room offers advantages in terms of easy exchange of information and interaction among coworkers, but reduces privacy and acoustic comfort. Thus, the purpose of this work was to evaluate the acoustic quality of a real open-plan office and to propose changes in the room to improve the acoustic conditioning of this office. The computational model of the office under study was calibrated based on RT and STI measurements. Predictions were made of the RT and STI, which generated the radius of distraction rD, and the rate of spatial decay of sound pressure levels per distance doubling DL2 in the real conditions of the office and after modifications of the room. The insertion of dividers between work stations and an increase in the ceiling’s sound absorption improved the acoustic conditions in the office under study. Ó 2012 Elsevier Ltd and The Ergonomics Society. All rights reserved.
Keywords: Speech intelligibility Reverberation time Sound pressure level
1. Introduction Studies of the acoustic quality and suitability of rooms designed for intensively intellectual and cognitive activities, such as educational and work environments, have been the focus of scientific research (e.g., Ollerhead, 1973; Kleeman, 1982; Japuntich, 2001; Zannin and Marcon, 2007; Yildirima et al., 2007; Astolfi et al., 2008; Zannin and Zwirtes, 2009; Haka et al., 2009; Klatte and Hellbrück, 2010; Liebl et al., 2012). Since the 1950s, open-plan offices have been popular among design professionals (Durval et al., 2002) in the attempt to meet the communication needs and intense productivity. According to Duffy (1980), the main characteristic of open-plan offices is their free arrangement in large open plans. In the opinion of Van Der Voordt (2004), office designs should be aimed at stimulating new ways of working, improve productivity and reduce costs without reducing employee satisfaction. Based on this rationale, the priority of companies is to achieve better worker performance at low costs, while for the worker the most important factor is a pleasurable work environment. Most intellectual production is carried out in offices. The relative number of open-plan offices is still increasing and work that requires cognitive effort is performed in these spaces. As a consequence of the development of open-plan offices during the 1950s
* Corresponding author. Tel.: þ55 41 3361 3433. E-mail address: [email protected] (P.H.T. Zannin).
and 60s, studies were conducted about user perception concerning this type of environment (Malcolm, 1972; Sundstrom et al., 1982). Noise, especially human speech, is one of the main sources of annoyance among the occupants of open-plan offices (Hongisto et al., 2004; Beaman, 2005; Wang and Bowden, 2006; Helenius et al., 2007; Rashid and Zimring, 2008). Roy (2007) states that 71% of the distraction at office work stations is attributed to noise intrusion. Studies by Liebl et al. (2012) have shown that background speech of high intelligibility impedes cognitive performance. Hence, according to Helenius et al. (2007), increasing efforts should focus on the acoustic design of open-plan offices. Optimizing the acoustic design of these offices can be a difficult task, since many parameters must be considered. The International Organization for Standardization (ISO) recently published Part 3 of the ISO 3382 standard (ISO, 2012), which specifies methods for measuring acoustic parameters in open-plan offices. According to Parsons (1995), standards provide a useful starting point for successful design, but will not determine workplace design. Therefore, the objective of this work was to make an acoustic evaluation of a real open-plan office and improve its acoustic conditions by means of its architectural design and by controlling its ambient noise. 2. Materials and method The data for this study were obtained from measurements in the real office and from computer simulations. The purpose of these two work phases was to evaluate the acoustic quality of a real open-
0003-6870/$ e see front matter Ó 2012 Elsevier Ltd and The Ergonomics Society. All rights reserved. doi:10.1016/j.apergo.2012.03.007
C.R.M. Passero, P.H.T. Zannin / Applied Ergonomics 43 (2012) 1066e1071
plan office and the influence of some constructive elements and ambient noise on the acoustic conditioning of this office. The object of this study was an open-plan office located in Curitiba, the capital of the state of Paraná, Brazil. Table 1 presents the main characteristics of the office of this study (Fig. 1). For this research, acoustic measurements were taken of the sound pressure level (SPL), reverberation time (RT) and speech transmission index (STI) in the aforementioned open-plan office. Studies performed by Haka et al. (2009) presented the use of STI as an essential room acoustic design measure in open-plan offices. The measurement of these parameters was divided into two stages. In the first stage, measurements were taken during normal working hours to determine the SPL. The second stage involved taking measurements of the RT and STI after working hours or on weekends. The SPL was measured to obtain ambient noise data of the office during normal working hours. The measured parameters were: a) non-weighted sound pressure level in octave bands from 63 to 8000 Hz; and b) A-weighted equivalent sound level (LAeq). The sound pressure level was measured using a Brüel and Kjaer 2260 acoustic analyzer. The ANSI S1.4-1983 standard classifies this analyzer as Class 1. The measuring time at each point was 5 min and the points were selected so as to cover the entire expanse of the office. In the laboratory, the measurement data were transferred to a computer and the spatial average of all the measured points was calculated for each of the parameters, using the Brüel and Kjaer Evaluator Type 7820 computer program. The mean values of the sound pressure level, non-weighted and in octave bands, were used as background noise to calculate the STI. The mean values of LAeq were used for comparison with pertinent legislation. All the measurements of the sound pressure level were taken as recommended by the Brazilian NBR 10151 standard (ABNT, 2000). The reverberation time was measured using Dirac 3.1 software (Brüel and Kjaer 7841). This software calculates the acoustic parameters based on the room’s impulse response to a noise generated inside it. Studies by Passero and Zannin (2010) and Astolfi et al. (2008) used a similar method to evaluate the RT. Fig. 2 presents the measurement scheme and the equipment employed. Logarithmic scan type noise was chosen to excite the room, since this signal allows one to isolate the impulse response of the room’s components caused by nonlinearities (Masiero and Iazzetta, 2005). The position and number of source and receiver points were those established by the ISO 3382-2 (2008) standard. After taking the measurements, the spatial averages of the measured points were calculated. This spatial average, a single number for each octave-band frequency from 125 to 4000 Hz, was determined from the arithmetic mean of the individual reverberation times of all the sound source and microphone positions (ISO 3382-2, 2008).
Table 1 Principal characteristics of the office under study. Characteristics of the evaluated office Area (m2) Volume (m3) No. of work stations Predominant floor-to-ceiling height (m) Ceiling materiala Flooring materiala Height of dividers between stations (m) Material of dividers between stationsa a
Average a between the octave-band frequencies of 125e4000 Hz.
1324.40 4922.02 242 3.60 0.02 0.18 1.30 0.16
1067
The STI was measured using practically the same equipment as that used to measure the RT. Only the dodecahedron loudspeaker was substituted by a mouth simulator (Brüel and Kjaer 4227) and an octave-band frequency sound pressure equalizer (Behringer FBQ800) was included. Fig. 3 shows the measurement scheme of the STI and the equipment utilized. These modifications in the equipment were made according to the recommendations of the IEC 60268-16 (2003) standard. Prior to measuring the STI, the signal was equalized, after which the sound level was calibrated. After completing these calibrations, the actual measurements of the STI were performed. With regard to the position of the source and receiver, the IEC 60268-16 standard establishes only that they should be in the position of the speaker and listener. Since all the people working in open-plan offices are both listeners and speakers, several receiver positions were evaluated for each sound source position. This enabled us to evaluate the interference of the speaker on the listeners in his proximity. All the other speakers were considered as background noise and inserted in the STI simulation by inserting the ambient noise measured during working hours. The computer simulations were performed using Odeon 9.0 software. This software uses the hybrid method, in which the early reflections are calculated using a combination of the image source method and early scattering rays. After a given order of transmission, the reflections are calculated using the ray tracing method. The simulations were divided into three stages: a) calibration of the model; b) simulation of the real situation of the office; and c) simulation of the office with architectural and ambient noise modifications. Stages “b” and “c” involve simulating the RT, STI, and the rate of spatial decay of sound pressure levels per distance doubling (DL2). According to the ISO 14257 (2001) standard, the DL2 is the decay, in decibels per double the distance, of the spatial distribution curve of sound within a given range of distances. According to Ondet and Suer (1995), this parameter is strongly dependent on the reverberation of the room: a value of 0 dB corresponds to the case of a highly reverberant room; while a value of 6 dB corresponds to a room treated ideally, corresponding to the open field. To obtain an accurate simulation it was very important to use adequate calculation parameters. The computer model was calibrated using the procedure of altering the surface scattering and absorption coefficients of the materials and comparing the simulated values with the values of the same measured parameters. According to Katz and Wetherill (2007), the calibration procedure in simulations means the refinement of the acoustic properties of the materials. To calibrate the computer model, the real situation of the office studied here was simulated and the values of RT and STI were compared with the measured values of the same parameters. The precision of the model was checked using the standard deviation and the spatial average of the measured RT values. The simulated data should fall within the interval of the mean standard deviation (M s). Similar comparisons have been made in studies by Vorländer (1995). With regard to the measured and simulated STI values, their comparison served to confirm the calibration of the model. The STI is strongly dependent on the location of the point of measurement or simulation, and varies considerably around the room. Therefore, any minor imprecision at the moment of transcribing the location of the measured point to the computer model would result in a major variation of the STI. During the calibration of the model based on the STI values, care was taken to ensure that most of the simulated points showed a difference below the subjective perception threshold of 0.05 when compared to the measured values. When this was achieved, the calibration was considered to be concluded. Thus, the importance of this step for the precision of the data
1068
C.R.M. Passero, P.H.T. Zannin / Applied Ergonomics 43 (2012) 1066e1071
Fig. 1. Office under study.
simulated subsequently was confirmed after the computer model was calibrated. Calibration is an indispensable phase of this work, which requires in-depth knowledge of current literature and a certain amount of experience and sensitivity on the part of the researcher (Bork, 2000). During the calibrations, the source and receiver positions were located at the same points as the measurements. After the calibrations, still in the real condition of the room, simulations were performed with receiver grids. The purpose of these simulations was to evaluate the entire office, and to allow for a better comparison of the values obtained with these simulations against those obtained with the earlier simulations, with modifications in the room. For the simulations of RT with receiver grids, the source was located in two positions (ISO 3382-2, 2008) and the receivers on a quadrangular grid of 0.50 m 0.50 m covering the entire area of the office. After this simulation, the Odeon 9.0 program calculated the average of all the simulated points in octave-band frequencies of 125e4000 Hz. For the STI simulations with receiver grids, the source was located in a position, the work station of a speaker, and the receivers on the same 0.50 m 0.50 m grid. After the simulation, the radius of distraction (rD) was determined based on the calculated grid. Hongisto et al. (2007) defined the rD as the distance from the speaker where the STI is lower than 0.5. Based on this parameter, it is possible to obtain a single value to characterize the speech intelligibility in the room under study, since the rD is independent of the position of the receiver in the room, unlike the STI.
Computer with DIRAC 3.1
audio interface
amplifier
acoustic analyzer Fig. 2. Measurement scheme of reverberation time.
dodecahedron loudspeaker
The third simulated parameter was the DL2. The Odeon 9.0 program calculates the DL2 for each frequency band from 63 Hz to 8 kHz and the DL2,Co, which is the mean A-weighted decay rate, for the frequencies from 125 to 4000 Hz. In this study, the data were analyzed using only DL2,Co, since it presents the results of noise reduction over distance by means of a single number. The DL2 simulations were performed according to the ISO 14257 (2001) standard. The source used for the simulation of RT and DL2 is the omnidirectional pointwise type (ISO 14257, 2001; ISO 3382-2, 2008). The STI was simulated using a source with directivity similar to that of the human mouth (IEC 60268-16, 2003). The sound used for simulating the STI is similar to that of the human voice, frequency weighted as established by the IEC 6026816 (2003) standard. After simulating the STI with receiver grids, the value of the sound pressure level measured in the office was inserted into the simulation. After simulating the real situation of the office, modifications were made in the model, as new simulations were performed. The modifications involved the surface materials of the ceiling and floor, the presence or absence of divider panels, and changes in the level of ambient noise, the latter valid only for the STI simulations. Six simulations of the RT and DL2 and twelve simulations of the STI were performed, since the STI was simulated with two levels of ambient noise for each physical condition of the room. These noise levels are equivalent to the average sound pressure levels found in real offices (Passero, 2009).
Computer with DIRAC 3.1
audio interface
acoustic analyzer
equalizer
amplifier
mouth simulator
Fig. 3. Measurement scheme of the speech transmission index.
C.R.M. Passero, P.H.T. Zannin / Applied Ergonomics 43 (2012) 1066e1071
1069
Table 2 Characteristics of the physical changes and sound pressure level used in the simulations. Characteristics Ceiling materiala Flooring materiala Height of the divider panels between work stations (m) Material of the divider panels between work stationsa Sound pressure level (dB(A)) a
Situation 111
Situation 011
Situation 010
Situation 000
0.81 0.18 1.30
0.81 0.01 1.30
0.02 0.01 1.30
0.81 0.18 e
0.81 0.01 e
0.02 0.01 e
0.10
0.10
0.10
e
e
e
64 55
64 55
64 55
64 55
Situation 110
64 55
Situation 100
64 55
Average a between the octave-band frequencies of 125e4000 Hz.
Table 2 shows the simulated modifications in the open-plan office. Thus, as can be seen in Table 2, situations identified with the first digit “1” (1_ _) were characterized by the presence of divider panels between work stations. The divider panels were 1.30 m tall, and were made of melamine high pressure laminate board. This type of office partition is widely available on the market and maintains the characteristics of an open-plan office environment. Conversely, situations identified with the first digit “0” (0_ _) were characterized by the absence of divider panels between work stations. In situations identified with the second digit “1” (_1_), the ceiling was finished with acoustic material which is highly sound absorbing, especially at high frequencies. Conversely, in situations identified with the second digit “0” (_0_), the ceiling finishing material was removed and the concrete ceiling slab was left bare. In situations identified with the third digit “1” (_ _1), the floor was covered with high density carpeting. Lastly, in situations identified with the third digit “0” (_ _0), the flooring was replaced with ceramic tile. It was decided to change the ceiling and floor finishing materials because they are the largest surfaces in the office, and thus, according to the literature, cause the greatest impact on the sound environment of this type of construction. The STI was simulated with two background noise levels, 64 dB(A) and 55 dB(A), in real offices for the six physical situations of the room. These SPL values were recorded by Passero (2009). The combinations of materials, situation of divider panels and sound pressure level are commonly found in real offices. 3. Results Fig. 4 depicts the mean measured and simulated values of RT, as well as the M s of the measured values. In Fig. 4, note that the simulated RT values fall within the interval of the M s of the measured RT values at all the frequencies evaluated.
Fig. 4. Graph of measured and simulated reverberation time, calibration of the computer model. “M” is the mean and “M s” is the mean the standard deviation.
Fig. 5 presents the measured and simulated STI values. Table 3 lists the results obtained in the simulations of the real situation and with modifications. It also presents the SPL measured in the office during the working day and the values of SPL used in the simulation of the STI and rD with modifications of the room. 4. Discussion With regard to the SPL measured in the office, the Brazilian NBR 10152 (1987) standard indicates that, for acoustic comfort in offices, the sound pressure level should be up to 45 dB(A), with 65 dB(A) indicated as the maximum acceptable limit. Hence, according to this standard, the value of 59 dB(A) measured in the office evaluated here is considered acceptable, although it is above the range of acoustic comfort. For comparison of the values of RT obtained in this office against the ideal values established by standards, the Brazilian Association of Technical Standards (ABNT) has the NBR 12179 (1992) standard, which establishes ideal values of RT according to the activity performed in the room and the volume of the room. However, this standard does not mention offices or working environments. International standards were therefore consulted for a comparison of the RT values. Chigot (2007) cites the following standards that present ideal values of RT specifically for open-plan offices: the Norwegian NS 8175 (2005); the Finnish SFS 5907 (2004); the Swedish SS 025268 (2001); and the Australian/New Zealand standard AS/NZS 2107 (2000). These standards establish ideal values for offices larger than 300 m3, since the NF S31-080 e the French standard also cited by Chigot (2007) e indicates ideal values of RT for offices larger than 250 m3. In addition to the above, Fischer et al. (2002) cite the German standard VDI 2569 (1990), which does not establish a minimum volume for the room under evaluation. A comparison was also made of the values of RT obtained in the simulations of the office in the various situations against those established by international standards, in terms of meeting or not meeting these standards. Table 4 presents the results of this comparison.
Fig. 5. Graph of the measured and simulated speech transmission index (STI) in the office, calibration of the computer model.
1070
C.R.M. Passero, P.H.T. Zannin / Applied Ergonomics 43 (2012) 1066e1071
Table 3 Acoustic parameters obtained in the real situation of the office and with modifications of the room. Acoustic parameter
Situation real
Situation 111
Situation 110
Situation 100
Situation 011
Situation 010
Situation 000
SPL (dB(A))
59
64 55
64 55
64 55
64 55
64 55
64 55
RT (s)a
1.08
0.74
0.75
1.42
0.79
0.80
1.45
rD
1.50
1.50 1.50
1.50 1.50
1.50 1.50
1.00 2.00
1.00 2.00
1.00 2.50
DL2,Co
3.61
3.97
4.02
3.31
3.76
3.79
2.75
Note: The first line of rD values presents the results obtained in the SPL simulations at 64 dB(A). The second line presents the values obtained in the SPL simulations at 55 dB(A). a Mean value at the octave-band frequencies of 500e2000 Hz.
Table 4 Comparison of the simulated RT values and the ideal values stipulated by international standards. Country of origin of the standard
Comparison of the simulated RT with data prescribed by the standards Situation real
Situation 111
Situation 110
Situation 100
Situation 011
Situation 010
Situation 000
Norway Finland Sweden Australia/New Zealand Germany
No No No No No
Yes No No No No
Yes No No No No
No No No No No
No No No No No
No No No No No
No No No No No
France Normal perform. Efficient perform. High perform.
Yes Yes Yes
Yes Yes Yes
No Yes Yes
No No No
No Yes No
Yes Yes No
No No No
No ¼ does not meet the standard; Yes ¼ meets the standard.
An analysis of Table 4 indicates that only the French and Norwegian standards are met in some of the evaluated situations. In the real situation, the reverberation times are adequate only if compared to the values established by the French standard, considering a normal performance of the workers. In situations 011 and 010, the recorded reverberation times are slightly above the maximum limit established by the Norwegian and French standards for high performance. However, in situations 111 and 110, the reverberation times found are slightly lower than these values, meeting the Norwegian and French standards in all the required ranges of performance. The other standards establish much lower maximum values for RT than those recorded in the open-plan office of this study, so none of those standards were met in any of the simulated situations. The values of RT found in situations 000 and 100 failed to meet any of the aforementioned standards. Upon analyzing the values of rD listed in Table 3, one finds that the inclusion of divider panels increases the values of this parameter when the noise is high. However, when the lowest SPL is used, the insertion of divider panels reduces this parameter. This can be explained by the fact that, with a high noise level, the speech intelligibility in the office is very low, so the insertion of divider panels improves the speech intelligibility at each work station, i.e., the sound is reflected by the divider panels and returns to the speaker, intensifying the sound in the work stations behind and diagonally from the speaker. With a reduced noise level, the speech intelligibility in the room is greater. Therefore, the insertion of divider panels served to block direct sound to the work stations in front of the speaker, reducing the rD. This finding is in agreement with Kang’s (2002) statement. According to this author, barriers can be designed not only to prevent the direct sound of a speaker from reaching other conversation groups, but also to increase the level of the signal, because the barrier keeps the energy of the speaker’s speech within his conversation group. In all the situations with divider panels, including the real situation, the rD remained unaltered, with no influence of the SPL or the ceiling and flooring materials on the values of rD in these situations.
As for the DL2, the French standard NF S31-080 specifies values of DL2 according to employee performance requirements: normal performance, DL2 > 2 dB(A); efficient performance, DL2 > 3 dB(A); high performance, DL2 > 4 dB(A) (Chigot, 2007). Hence, from Table 3, it can be concluded that the value of DL2 is adequate for the efficient performance of the workers in the real situation, according to this standard. In the other simulated situations, this parameter was found to be strongly influenced by the type of ceiling finishing material. Moreover, the insertion of divider panels resulted in an increase of this parameter. Thus, the highest value of DL2 obtained was in the simulation of situation 110, in other words, with divider panels, and acoustic absorbing material only on the ceiling. Only in situation 110 is the value of DL2 considered adequate for the development of activities that require high performance of the employees, according to the French standard. Contrary to situation 110, situation 000 presented the lowest value of DL2, making it suitable, according to the French standard, only for activities that require normal performance of the workers. This situation is characterized by the absence of sound-absorbing material on the ceiling and floor, and by the absence of divider panels between work stations. 5. Conclusions In this study, objective acoustic evaluations were made of a real open-plan office. These evaluations were based on acoustic measurements and simulations of several parameters. In addition, acoustic simulations were performed with modifications of the architecture and ambient noise in the office under study. With regard to the values of rD simulated in the situations with modifications in the room and the noise level, these values were influenced by the variation in ambient noise when the office was simulated in the situations without divider panels between work stations. When the office was simulated with divider panels, changing the ceiling and flooring materials or varying the SPL did not modify the values of rD.
C.R.M. Passero, P.H.T. Zannin / Applied Ergonomics 43 (2012) 1066e1071
As for the DL2, according to Asselineau (2007), the higher the values of this parameter, the better the acoustic conditions in offices, because noise will be more attenuated. The DL2 parameter varied significantly with the insertion/removal of divider panels. Situations with divider panels presented higher values of DL2. This finding is in agreement with the results reported by Hongisto et al. (2007). Moreover, the insertion of acoustic absorbing material on the surface of the ceiling increased the DL2 when compared with the simulated situation using material with low acoustic absorption on this surface. Therefore, it can be concluded from this study that for an openplan office to have good acoustic conditions, it must have: 1) divider panels between the work stations, and 2) the surface of the ceiling should be finished with high sound-absorbing material. References Asselineau, M., 2007. Integration of furnishing in open plan office design: case studies. In: International Congress on Acoustics, September, Madrid, Spain. Associação Brasileira de Normas Técnicas, 1987. NBR 10152: Níveis de ruído para conforto acústico: procedimento. Rio de Janeiro (in Portuguese). Associação Brasileira de Normas Técnicas, 1992. NBR 12179: Tratamento Acústico em Recintos Fechados. Rio de Janeiro (in Portuguese). Associação Brasileira de Normas Técnicas, 2000. NBR 10151: avaliação do ruído em áreas habitadas, visando o conforto da comunidade: procedimento. Rio de Janeiro (in Portuguese). Astolfi, A., Corrado, V., Griginis, A., 2008. Comparison between measured and calculated parameters for the acoustical characterization of small classrooms. Applied Acoustics 69, 966e976. Beaman, C.P., 2005. Auditory distraction from low-intensity noise: a review of the consequences for learning and workplace environments. Applied Cognitive Psychology 19, 1041e1064. Bork, I.A., 2000. Comparison of room simulation software e the 2nd round robin on room acoustical computer simulation. Acta Acustica 86, 943e956. Chigot, P., 2007. Alternative room acoustic descriptors for open offices e progresses in standardization. In: International Congress on Acoustics, September, Madrid, Spain. Duffy, F., 1980. Oficinas. H. Blume Ediciones, Madrid. Durval, C.L., Charles, K.E., Veitch, J.A., 2002. Open-Plan Office Density and Environmental Satisfaction. National Research Council Canada, Canada. IRC Research Report RR-150. Fischer, M., Jenisch, R., Klopfer, H., Freymuth, H., Richter, E., Petzold, K., Stohrer, M., 2002. Lehrbuch der Bauphysik, fifth ed. B.G. Teubner, Stuttgart. Haka, M., Haapakangas, A., Keränen, J., Hakala, J., Keskinen, E., Hongisto, V., 2009. Performance effects and subjective disturbance of speech in acoustically different office types e a laboratory experiment. Indoor Air 19, 454e467. Helenius, R., Keskinen, E., Haapakangas, A., Hongisto, V., 2007. Acoustic environment in Finnish offices e the summary of questionnaire studies. In: International Congress on Acoustics, September, Madrid, Spain. Hongisto, V., Keranen, J., Larm, P., 2004. Simple model for acoustical design of openplan offices. Acta Acustica united with Acustica 90, 481e495. Hongisto, V., Virjonen, P., Keränen, J., 2007. Determination of acoustic conditions in open offices and suggestions for acoustic classification. In: International Congress on Acoustics, September, Madrid, Spain. International Electrotechnical Commission, 2003. IEC 60268-16: sound system equipment e Part 16: Objective rating of speech intelligibility by speech transmission index. Switzerland.
1071
International Organization for Standardization, 2001. ISO 14257: acoustics e measurement and parametric description of spatial sound distribution curves in workrooms for evaluation of their acoustical performance. Switzerland. International Organization for Standardization, 2008. ISO 3382-2: acoustics e measurement of room acoustic parameters. Part 2: Reverberation time in ordinary rooms. Switzerland. International Organization for Standardization, 2012. ISO 3382-3: acoustics e measurement of room acoustic parameters. Part 3: Open plan offices. Switzerland. Japuntich, D.A., 2001. Polarized task lighting to reduce reflective glare in open-plan office cubicles. Applied Ergonomics 32 (5), 485e499. Kang, J., 2002. Numerical modeling of speech intelligibility in dining spaces. Applied Acoustics 63, 1315e1333. Katz, B.F.G., Wetherill, E.A., 2007. The fall and rise of the Fogg Art Museum Lecture Hall: a forensic study. Acoustics Today 3 (3), 10e16. Klatte, M., Hellbrück, J., 2010. Effects of classroom acoustics on performance and wellbeing in elementary school children: a field study. In: Inter Noise, June, Lisbon, Portugal. Kleeman Jr., W.B., 1982. The need for ergonomics expertise in office design. Applied Ergonomics 13 (2), 125e127. Liebl, A., Haller, J., Jödicke, B., Baumgartner, H., Schlittmeier, S., Hellbrück, J., 2012. Combined effects of acoustic and visual distraction on cognitive performance and well-being. Applied Ergonomics 43, 424e434. Malcolm, J.B., 1972. Office landscape: does it work? Applied Ergonomics 3 (4), 224e236. Masiero, B.S., Iazzetta, F., 2005. Estudo e implementação de métodos de medição de resposta impulsiva. Acústica e Vibrações 36, 34e41 (in Portuguese). Ollerhead, J.B., 1973. Noise: how can the nuisance be controlled? Applied Ergonomics 4 (3), 130e138. Ondet, A.M., Suer, J., 1995. Development and validation of a criterion for assessing the acoustic performance of industrial rooms. Journal of the Acoustical Society of America 97 (3), 1727e1731. Parsons, K.C., 1995. Ergonomics of the physical environment e international ergonomics standards concerning speech communication, danger signals, lighting, vibration and surface temperatures. Applied Ergonomics 4, 281e292. Passero, C.R.M., 2009. Avaliação e Adequação Acústica de Escritórios Panorâmicos através de Medições e Simulações. Tese de Doutorado, PhD thesis, Curitiba (in Portuguese). Passero, C.R.M., Zannin, P.H.T., 2010. Statistical comparison of reverberation times measured by the integrated impulse response and interrupted noise methods, computationally simulated with ODEON software, and calculated by Sabine, Eyring and Arau-Puchades’ formulas. Applied Acoustics 71, 1204e1210. Rashid, M., Zimring, C., 2008. A review of the empirical literature on the relationships between indoor environment and stress in health care and office settings e problems and prospects of sharing evidence. Environment and Behavior 40 (2), 151e190. Roy, K., 2007. Acoustic design in the new architectural trend of “exposed structures”. In: 19th International Congress on Acoustics, Madrid, 2e7 September. Sundstrom, E., Herbert, R.K., Brown, D.W., 1982. Privacy and communication in an open-plan office e a case study. Environment and Behavior 14 (3), 379e392. Van Der Voordt, T.J.M., 2004. Productivity and employee satisfaction in flexible workplaces. Journal of Corporate Real Estate 6 (2), 133e148. Vorländer, M., 1995. International round robin on room acoustical computer simulations. In: International Congress on Acoustics. Institute of Acoustics, United Kingdom, pp. 689e692. Wang, L.M., Bowden, E.E., 2006. Finding connections among indoor noise criteria, human perception, and worker performance. Acoustics Today 2 (1), 33e37. Yildirima, K., Akalin-Baskayab, A., Celebia, M., 2007. The effects of window proximity, partition height, and gender on perceptions of open-plan offices. Journal of Environmental Psychology 27, 154e165. Zannin, P.H.T., Marcon, C.R., 2007. Objective and subjective evaluation of the acoustic comfort in classrooms. Applied Ergonomics 38, 675e680. Zannin, P.H.T., Zwirtes, D.P.Z., 2009. Evaluation of the acoustic performance of classrooms in public schools. Applied Acoustics 70, 626e635.
Contents lists available at SciVerse ScienceDirect
Applied Ergonomics journal homepage: www.elsevier.com/locate/apergo
Acoustic evaluation and adjustment of an open-plan office through architectural design and noise control Carolina Reich Marcon Passero, Paulo Henrique Trombetta Zannin* Laboratory of Environmental and Industrial Acoustics and Acoustic Comfort e LAAICA, Federal University of Paraná, PR, Brazil
a r t i c l e i n f o
a b s t r a c t
Article history: Received 9 August 2011 Accepted 26 March 2012
Arranging office space into a single open room offers advantages in terms of easy exchange of information and interaction among coworkers, but reduces privacy and acoustic comfort. Thus, the purpose of this work was to evaluate the acoustic quality of a real open-plan office and to propose changes in the room to improve the acoustic conditioning of this office. The computational model of the office under study was calibrated based on RT and STI measurements. Predictions were made of the RT and STI, which generated the radius of distraction rD, and the rate of spatial decay of sound pressure levels per distance doubling DL2 in the real conditions of the office and after modifications of the room. The insertion of dividers between work stations and an increase in the ceiling’s sound absorption improved the acoustic conditions in the office under study. Ó 2012 Elsevier Ltd and The Ergonomics Society. All rights reserved.
Keywords: Speech intelligibility Reverberation time Sound pressure level
1. Introduction Studies of the acoustic quality and suitability of rooms designed for intensively intellectual and cognitive activities, such as educational and work environments, have been the focus of scientific research (e.g., Ollerhead, 1973; Kleeman, 1982; Japuntich, 2001; Zannin and Marcon, 2007; Yildirima et al., 2007; Astolfi et al., 2008; Zannin and Zwirtes, 2009; Haka et al., 2009; Klatte and Hellbrück, 2010; Liebl et al., 2012). Since the 1950s, open-plan offices have been popular among design professionals (Durval et al., 2002) in the attempt to meet the communication needs and intense productivity. According to Duffy (1980), the main characteristic of open-plan offices is their free arrangement in large open plans. In the opinion of Van Der Voordt (2004), office designs should be aimed at stimulating new ways of working, improve productivity and reduce costs without reducing employee satisfaction. Based on this rationale, the priority of companies is to achieve better worker performance at low costs, while for the worker the most important factor is a pleasurable work environment. Most intellectual production is carried out in offices. The relative number of open-plan offices is still increasing and work that requires cognitive effort is performed in these spaces. As a consequence of the development of open-plan offices during the 1950s
* Corresponding author. Tel.: þ55 41 3361 3433. E-mail address: [email protected] (P.H.T. Zannin).
and 60s, studies were conducted about user perception concerning this type of environment (Malcolm, 1972; Sundstrom et al., 1982). Noise, especially human speech, is one of the main sources of annoyance among the occupants of open-plan offices (Hongisto et al., 2004; Beaman, 2005; Wang and Bowden, 2006; Helenius et al., 2007; Rashid and Zimring, 2008). Roy (2007) states that 71% of the distraction at office work stations is attributed to noise intrusion. Studies by Liebl et al. (2012) have shown that background speech of high intelligibility impedes cognitive performance. Hence, according to Helenius et al. (2007), increasing efforts should focus on the acoustic design of open-plan offices. Optimizing the acoustic design of these offices can be a difficult task, since many parameters must be considered. The International Organization for Standardization (ISO) recently published Part 3 of the ISO 3382 standard (ISO, 2012), which specifies methods for measuring acoustic parameters in open-plan offices. According to Parsons (1995), standards provide a useful starting point for successful design, but will not determine workplace design. Therefore, the objective of this work was to make an acoustic evaluation of a real open-plan office and improve its acoustic conditions by means of its architectural design and by controlling its ambient noise. 2. Materials and method The data for this study were obtained from measurements in the real office and from computer simulations. The purpose of these two work phases was to evaluate the acoustic quality of a real open-
0003-6870/$ e see front matter Ó 2012 Elsevier Ltd and The Ergonomics Society. All rights reserved. doi:10.1016/j.apergo.2012.03.007
C.R.M. Passero, P.H.T. Zannin / Applied Ergonomics 43 (2012) 1066e1071
plan office and the influence of some constructive elements and ambient noise on the acoustic conditioning of this office. The object of this study was an open-plan office located in Curitiba, the capital of the state of Paraná, Brazil. Table 1 presents the main characteristics of the office of this study (Fig. 1). For this research, acoustic measurements were taken of the sound pressure level (SPL), reverberation time (RT) and speech transmission index (STI) in the aforementioned open-plan office. Studies performed by Haka et al. (2009) presented the use of STI as an essential room acoustic design measure in open-plan offices. The measurement of these parameters was divided into two stages. In the first stage, measurements were taken during normal working hours to determine the SPL. The second stage involved taking measurements of the RT and STI after working hours or on weekends. The SPL was measured to obtain ambient noise data of the office during normal working hours. The measured parameters were: a) non-weighted sound pressure level in octave bands from 63 to 8000 Hz; and b) A-weighted equivalent sound level (LAeq). The sound pressure level was measured using a Brüel and Kjaer 2260 acoustic analyzer. The ANSI S1.4-1983 standard classifies this analyzer as Class 1. The measuring time at each point was 5 min and the points were selected so as to cover the entire expanse of the office. In the laboratory, the measurement data were transferred to a computer and the spatial average of all the measured points was calculated for each of the parameters, using the Brüel and Kjaer Evaluator Type 7820 computer program. The mean values of the sound pressure level, non-weighted and in octave bands, were used as background noise to calculate the STI. The mean values of LAeq were used for comparison with pertinent legislation. All the measurements of the sound pressure level were taken as recommended by the Brazilian NBR 10151 standard (ABNT, 2000). The reverberation time was measured using Dirac 3.1 software (Brüel and Kjaer 7841). This software calculates the acoustic parameters based on the room’s impulse response to a noise generated inside it. Studies by Passero and Zannin (2010) and Astolfi et al. (2008) used a similar method to evaluate the RT. Fig. 2 presents the measurement scheme and the equipment employed. Logarithmic scan type noise was chosen to excite the room, since this signal allows one to isolate the impulse response of the room’s components caused by nonlinearities (Masiero and Iazzetta, 2005). The position and number of source and receiver points were those established by the ISO 3382-2 (2008) standard. After taking the measurements, the spatial averages of the measured points were calculated. This spatial average, a single number for each octave-band frequency from 125 to 4000 Hz, was determined from the arithmetic mean of the individual reverberation times of all the sound source and microphone positions (ISO 3382-2, 2008).
Table 1 Principal characteristics of the office under study. Characteristics of the evaluated office Area (m2) Volume (m3) No. of work stations Predominant floor-to-ceiling height (m) Ceiling materiala Flooring materiala Height of dividers between stations (m) Material of dividers between stationsa a
Average a between the octave-band frequencies of 125e4000 Hz.
1324.40 4922.02 242 3.60 0.02 0.18 1.30 0.16
1067
The STI was measured using practically the same equipment as that used to measure the RT. Only the dodecahedron loudspeaker was substituted by a mouth simulator (Brüel and Kjaer 4227) and an octave-band frequency sound pressure equalizer (Behringer FBQ800) was included. Fig. 3 shows the measurement scheme of the STI and the equipment utilized. These modifications in the equipment were made according to the recommendations of the IEC 60268-16 (2003) standard. Prior to measuring the STI, the signal was equalized, after which the sound level was calibrated. After completing these calibrations, the actual measurements of the STI were performed. With regard to the position of the source and receiver, the IEC 60268-16 standard establishes only that they should be in the position of the speaker and listener. Since all the people working in open-plan offices are both listeners and speakers, several receiver positions were evaluated for each sound source position. This enabled us to evaluate the interference of the speaker on the listeners in his proximity. All the other speakers were considered as background noise and inserted in the STI simulation by inserting the ambient noise measured during working hours. The computer simulations were performed using Odeon 9.0 software. This software uses the hybrid method, in which the early reflections are calculated using a combination of the image source method and early scattering rays. After a given order of transmission, the reflections are calculated using the ray tracing method. The simulations were divided into three stages: a) calibration of the model; b) simulation of the real situation of the office; and c) simulation of the office with architectural and ambient noise modifications. Stages “b” and “c” involve simulating the RT, STI, and the rate of spatial decay of sound pressure levels per distance doubling (DL2). According to the ISO 14257 (2001) standard, the DL2 is the decay, in decibels per double the distance, of the spatial distribution curve of sound within a given range of distances. According to Ondet and Suer (1995), this parameter is strongly dependent on the reverberation of the room: a value of 0 dB corresponds to the case of a highly reverberant room; while a value of 6 dB corresponds to a room treated ideally, corresponding to the open field. To obtain an accurate simulation it was very important to use adequate calculation parameters. The computer model was calibrated using the procedure of altering the surface scattering and absorption coefficients of the materials and comparing the simulated values with the values of the same measured parameters. According to Katz and Wetherill (2007), the calibration procedure in simulations means the refinement of the acoustic properties of the materials. To calibrate the computer model, the real situation of the office studied here was simulated and the values of RT and STI were compared with the measured values of the same parameters. The precision of the model was checked using the standard deviation and the spatial average of the measured RT values. The simulated data should fall within the interval of the mean standard deviation (M s). Similar comparisons have been made in studies by Vorländer (1995). With regard to the measured and simulated STI values, their comparison served to confirm the calibration of the model. The STI is strongly dependent on the location of the point of measurement or simulation, and varies considerably around the room. Therefore, any minor imprecision at the moment of transcribing the location of the measured point to the computer model would result in a major variation of the STI. During the calibration of the model based on the STI values, care was taken to ensure that most of the simulated points showed a difference below the subjective perception threshold of 0.05 when compared to the measured values. When this was achieved, the calibration was considered to be concluded. Thus, the importance of this step for the precision of the data
1068
C.R.M. Passero, P.H.T. Zannin / Applied Ergonomics 43 (2012) 1066e1071
Fig. 1. Office under study.
simulated subsequently was confirmed after the computer model was calibrated. Calibration is an indispensable phase of this work, which requires in-depth knowledge of current literature and a certain amount of experience and sensitivity on the part of the researcher (Bork, 2000). During the calibrations, the source and receiver positions were located at the same points as the measurements. After the calibrations, still in the real condition of the room, simulations were performed with receiver grids. The purpose of these simulations was to evaluate the entire office, and to allow for a better comparison of the values obtained with these simulations against those obtained with the earlier simulations, with modifications in the room. For the simulations of RT with receiver grids, the source was located in two positions (ISO 3382-2, 2008) and the receivers on a quadrangular grid of 0.50 m 0.50 m covering the entire area of the office. After this simulation, the Odeon 9.0 program calculated the average of all the simulated points in octave-band frequencies of 125e4000 Hz. For the STI simulations with receiver grids, the source was located in a position, the work station of a speaker, and the receivers on the same 0.50 m 0.50 m grid. After the simulation, the radius of distraction (rD) was determined based on the calculated grid. Hongisto et al. (2007) defined the rD as the distance from the speaker where the STI is lower than 0.5. Based on this parameter, it is possible to obtain a single value to characterize the speech intelligibility in the room under study, since the rD is independent of the position of the receiver in the room, unlike the STI.
Computer with DIRAC 3.1
audio interface
amplifier
acoustic analyzer Fig. 2. Measurement scheme of reverberation time.
dodecahedron loudspeaker
The third simulated parameter was the DL2. The Odeon 9.0 program calculates the DL2 for each frequency band from 63 Hz to 8 kHz and the DL2,Co, which is the mean A-weighted decay rate, for the frequencies from 125 to 4000 Hz. In this study, the data were analyzed using only DL2,Co, since it presents the results of noise reduction over distance by means of a single number. The DL2 simulations were performed according to the ISO 14257 (2001) standard. The source used for the simulation of RT and DL2 is the omnidirectional pointwise type (ISO 14257, 2001; ISO 3382-2, 2008). The STI was simulated using a source with directivity similar to that of the human mouth (IEC 60268-16, 2003). The sound used for simulating the STI is similar to that of the human voice, frequency weighted as established by the IEC 6026816 (2003) standard. After simulating the STI with receiver grids, the value of the sound pressure level measured in the office was inserted into the simulation. After simulating the real situation of the office, modifications were made in the model, as new simulations were performed. The modifications involved the surface materials of the ceiling and floor, the presence or absence of divider panels, and changes in the level of ambient noise, the latter valid only for the STI simulations. Six simulations of the RT and DL2 and twelve simulations of the STI were performed, since the STI was simulated with two levels of ambient noise for each physical condition of the room. These noise levels are equivalent to the average sound pressure levels found in real offices (Passero, 2009).
Computer with DIRAC 3.1
audio interface
acoustic analyzer
equalizer
amplifier
mouth simulator
Fig. 3. Measurement scheme of the speech transmission index.
C.R.M. Passero, P.H.T. Zannin / Applied Ergonomics 43 (2012) 1066e1071
1069
Table 2 Characteristics of the physical changes and sound pressure level used in the simulations. Characteristics Ceiling materiala Flooring materiala Height of the divider panels between work stations (m) Material of the divider panels between work stationsa Sound pressure level (dB(A)) a
Situation 111
Situation 011
Situation 010
Situation 000
0.81 0.18 1.30
0.81 0.01 1.30
0.02 0.01 1.30
0.81 0.18 e
0.81 0.01 e
0.02 0.01 e
0.10
0.10
0.10
e
e
e
64 55
64 55
64 55
64 55
Situation 110
64 55
Situation 100
64 55
Average a between the octave-band frequencies of 125e4000 Hz.
Table 2 shows the simulated modifications in the open-plan office. Thus, as can be seen in Table 2, situations identified with the first digit “1” (1_ _) were characterized by the presence of divider panels between work stations. The divider panels were 1.30 m tall, and were made of melamine high pressure laminate board. This type of office partition is widely available on the market and maintains the characteristics of an open-plan office environment. Conversely, situations identified with the first digit “0” (0_ _) were characterized by the absence of divider panels between work stations. In situations identified with the second digit “1” (_1_), the ceiling was finished with acoustic material which is highly sound absorbing, especially at high frequencies. Conversely, in situations identified with the second digit “0” (_0_), the ceiling finishing material was removed and the concrete ceiling slab was left bare. In situations identified with the third digit “1” (_ _1), the floor was covered with high density carpeting. Lastly, in situations identified with the third digit “0” (_ _0), the flooring was replaced with ceramic tile. It was decided to change the ceiling and floor finishing materials because they are the largest surfaces in the office, and thus, according to the literature, cause the greatest impact on the sound environment of this type of construction. The STI was simulated with two background noise levels, 64 dB(A) and 55 dB(A), in real offices for the six physical situations of the room. These SPL values were recorded by Passero (2009). The combinations of materials, situation of divider panels and sound pressure level are commonly found in real offices. 3. Results Fig. 4 depicts the mean measured and simulated values of RT, as well as the M s of the measured values. In Fig. 4, note that the simulated RT values fall within the interval of the M s of the measured RT values at all the frequencies evaluated.
Fig. 4. Graph of measured and simulated reverberation time, calibration of the computer model. “M” is the mean and “M s” is the mean the standard deviation.
Fig. 5 presents the measured and simulated STI values. Table 3 lists the results obtained in the simulations of the real situation and with modifications. It also presents the SPL measured in the office during the working day and the values of SPL used in the simulation of the STI and rD with modifications of the room. 4. Discussion With regard to the SPL measured in the office, the Brazilian NBR 10152 (1987) standard indicates that, for acoustic comfort in offices, the sound pressure level should be up to 45 dB(A), with 65 dB(A) indicated as the maximum acceptable limit. Hence, according to this standard, the value of 59 dB(A) measured in the office evaluated here is considered acceptable, although it is above the range of acoustic comfort. For comparison of the values of RT obtained in this office against the ideal values established by standards, the Brazilian Association of Technical Standards (ABNT) has the NBR 12179 (1992) standard, which establishes ideal values of RT according to the activity performed in the room and the volume of the room. However, this standard does not mention offices or working environments. International standards were therefore consulted for a comparison of the RT values. Chigot (2007) cites the following standards that present ideal values of RT specifically for open-plan offices: the Norwegian NS 8175 (2005); the Finnish SFS 5907 (2004); the Swedish SS 025268 (2001); and the Australian/New Zealand standard AS/NZS 2107 (2000). These standards establish ideal values for offices larger than 300 m3, since the NF S31-080 e the French standard also cited by Chigot (2007) e indicates ideal values of RT for offices larger than 250 m3. In addition to the above, Fischer et al. (2002) cite the German standard VDI 2569 (1990), which does not establish a minimum volume for the room under evaluation. A comparison was also made of the values of RT obtained in the simulations of the office in the various situations against those established by international standards, in terms of meeting or not meeting these standards. Table 4 presents the results of this comparison.
Fig. 5. Graph of the measured and simulated speech transmission index (STI) in the office, calibration of the computer model.
1070
C.R.M. Passero, P.H.T. Zannin / Applied Ergonomics 43 (2012) 1066e1071
Table 3 Acoustic parameters obtained in the real situation of the office and with modifications of the room. Acoustic parameter
Situation real
Situation 111
Situation 110
Situation 100
Situation 011
Situation 010
Situation 000
SPL (dB(A))
59
64 55
64 55
64 55
64 55
64 55
64 55
RT (s)a
1.08
0.74
0.75
1.42
0.79
0.80
1.45
rD
1.50
1.50 1.50
1.50 1.50
1.50 1.50
1.00 2.00
1.00 2.00
1.00 2.50
DL2,Co
3.61
3.97
4.02
3.31
3.76
3.79
2.75
Note: The first line of rD values presents the results obtained in the SPL simulations at 64 dB(A). The second line presents the values obtained in the SPL simulations at 55 dB(A). a Mean value at the octave-band frequencies of 500e2000 Hz.
Table 4 Comparison of the simulated RT values and the ideal values stipulated by international standards. Country of origin of the standard
Comparison of the simulated RT with data prescribed by the standards Situation real
Situation 111
Situation 110
Situation 100
Situation 011
Situation 010
Situation 000
Norway Finland Sweden Australia/New Zealand Germany
No No No No No
Yes No No No No
Yes No No No No
No No No No No
No No No No No
No No No No No
No No No No No
France Normal perform. Efficient perform. High perform.
Yes Yes Yes
Yes Yes Yes
No Yes Yes
No No No
No Yes No
Yes Yes No
No No No
No ¼ does not meet the standard; Yes ¼ meets the standard.
An analysis of Table 4 indicates that only the French and Norwegian standards are met in some of the evaluated situations. In the real situation, the reverberation times are adequate only if compared to the values established by the French standard, considering a normal performance of the workers. In situations 011 and 010, the recorded reverberation times are slightly above the maximum limit established by the Norwegian and French standards for high performance. However, in situations 111 and 110, the reverberation times found are slightly lower than these values, meeting the Norwegian and French standards in all the required ranges of performance. The other standards establish much lower maximum values for RT than those recorded in the open-plan office of this study, so none of those standards were met in any of the simulated situations. The values of RT found in situations 000 and 100 failed to meet any of the aforementioned standards. Upon analyzing the values of rD listed in Table 3, one finds that the inclusion of divider panels increases the values of this parameter when the noise is high. However, when the lowest SPL is used, the insertion of divider panels reduces this parameter. This can be explained by the fact that, with a high noise level, the speech intelligibility in the office is very low, so the insertion of divider panels improves the speech intelligibility at each work station, i.e., the sound is reflected by the divider panels and returns to the speaker, intensifying the sound in the work stations behind and diagonally from the speaker. With a reduced noise level, the speech intelligibility in the room is greater. Therefore, the insertion of divider panels served to block direct sound to the work stations in front of the speaker, reducing the rD. This finding is in agreement with Kang’s (2002) statement. According to this author, barriers can be designed not only to prevent the direct sound of a speaker from reaching other conversation groups, but also to increase the level of the signal, because the barrier keeps the energy of the speaker’s speech within his conversation group. In all the situations with divider panels, including the real situation, the rD remained unaltered, with no influence of the SPL or the ceiling and flooring materials on the values of rD in these situations.
As for the DL2, the French standard NF S31-080 specifies values of DL2 according to employee performance requirements: normal performance, DL2 > 2 dB(A); efficient performance, DL2 > 3 dB(A); high performance, DL2 > 4 dB(A) (Chigot, 2007). Hence, from Table 3, it can be concluded that the value of DL2 is adequate for the efficient performance of the workers in the real situation, according to this standard. In the other simulated situations, this parameter was found to be strongly influenced by the type of ceiling finishing material. Moreover, the insertion of divider panels resulted in an increase of this parameter. Thus, the highest value of DL2 obtained was in the simulation of situation 110, in other words, with divider panels, and acoustic absorbing material only on the ceiling. Only in situation 110 is the value of DL2 considered adequate for the development of activities that require high performance of the employees, according to the French standard. Contrary to situation 110, situation 000 presented the lowest value of DL2, making it suitable, according to the French standard, only for activities that require normal performance of the workers. This situation is characterized by the absence of sound-absorbing material on the ceiling and floor, and by the absence of divider panels between work stations. 5. Conclusions In this study, objective acoustic evaluations were made of a real open-plan office. These evaluations were based on acoustic measurements and simulations of several parameters. In addition, acoustic simulations were performed with modifications of the architecture and ambient noise in the office under study. With regard to the values of rD simulated in the situations with modifications in the room and the noise level, these values were influenced by the variation in ambient noise when the office was simulated in the situations without divider panels between work stations. When the office was simulated with divider panels, changing the ceiling and flooring materials or varying the SPL did not modify the values of rD.
C.R.M. Passero, P.H.T. Zannin / Applied Ergonomics 43 (2012) 1066e1071
As for the DL2, according to Asselineau (2007), the higher the values of this parameter, the better the acoustic conditions in offices, because noise will be more attenuated. The DL2 parameter varied significantly with the insertion/removal of divider panels. Situations with divider panels presented higher values of DL2. This finding is in agreement with the results reported by Hongisto et al. (2007). Moreover, the insertion of acoustic absorbing material on the surface of the ceiling increased the DL2 when compared with the simulated situation using material with low acoustic absorption on this surface. Therefore, it can be concluded from this study that for an openplan office to have good acoustic conditions, it must have: 1) divider panels between the work stations, and 2) the surface of the ceiling should be finished with high sound-absorbing material. References Asselineau, M., 2007. Integration of furnishing in open plan office design: case studies. In: International Congress on Acoustics, September, Madrid, Spain. Associação Brasileira de Normas Técnicas, 1987. NBR 10152: Níveis de ruído para conforto acústico: procedimento. Rio de Janeiro (in Portuguese). Associação Brasileira de Normas Técnicas, 1992. NBR 12179: Tratamento Acústico em Recintos Fechados. Rio de Janeiro (in Portuguese). Associação Brasileira de Normas Técnicas, 2000. NBR 10151: avaliação do ruído em áreas habitadas, visando o conforto da comunidade: procedimento. Rio de Janeiro (in Portuguese). Astolfi, A., Corrado, V., Griginis, A., 2008. Comparison between measured and calculated parameters for the acoustical characterization of small classrooms. Applied Acoustics 69, 966e976. Beaman, C.P., 2005. Auditory distraction from low-intensity noise: a review of the consequences for learning and workplace environments. Applied Cognitive Psychology 19, 1041e1064. Bork, I.A., 2000. Comparison of room simulation software e the 2nd round robin on room acoustical computer simulation. Acta Acustica 86, 943e956. Chigot, P., 2007. Alternative room acoustic descriptors for open offices e progresses in standardization. In: International Congress on Acoustics, September, Madrid, Spain. Duffy, F., 1980. Oficinas. H. Blume Ediciones, Madrid. Durval, C.L., Charles, K.E., Veitch, J.A., 2002. Open-Plan Office Density and Environmental Satisfaction. National Research Council Canada, Canada. IRC Research Report RR-150. Fischer, M., Jenisch, R., Klopfer, H., Freymuth, H., Richter, E., Petzold, K., Stohrer, M., 2002. Lehrbuch der Bauphysik, fifth ed. B.G. Teubner, Stuttgart. Haka, M., Haapakangas, A., Keränen, J., Hakala, J., Keskinen, E., Hongisto, V., 2009. Performance effects and subjective disturbance of speech in acoustically different office types e a laboratory experiment. Indoor Air 19, 454e467. Helenius, R., Keskinen, E., Haapakangas, A., Hongisto, V., 2007. Acoustic environment in Finnish offices e the summary of questionnaire studies. In: International Congress on Acoustics, September, Madrid, Spain. Hongisto, V., Keranen, J., Larm, P., 2004. Simple model for acoustical design of openplan offices. Acta Acustica united with Acustica 90, 481e495. Hongisto, V., Virjonen, P., Keränen, J., 2007. Determination of acoustic conditions in open offices and suggestions for acoustic classification. In: International Congress on Acoustics, September, Madrid, Spain. International Electrotechnical Commission, 2003. IEC 60268-16: sound system equipment e Part 16: Objective rating of speech intelligibility by speech transmission index. Switzerland.
1071
International Organization for Standardization, 2001. ISO 14257: acoustics e measurement and parametric description of spatial sound distribution curves in workrooms for evaluation of their acoustical performance. Switzerland. International Organization for Standardization, 2008. ISO 3382-2: acoustics e measurement of room acoustic parameters. Part 2: Reverberation time in ordinary rooms. Switzerland. International Organization for Standardization, 2012. ISO 3382-3: acoustics e measurement of room acoustic parameters. Part 3: Open plan offices. Switzerland. Japuntich, D.A., 2001. Polarized task lighting to reduce reflective glare in open-plan office cubicles. Applied Ergonomics 32 (5), 485e499. Kang, J., 2002. Numerical modeling of speech intelligibility in dining spaces. Applied Acoustics 63, 1315e1333. Katz, B.F.G., Wetherill, E.A., 2007. The fall and rise of the Fogg Art Museum Lecture Hall: a forensic study. Acoustics Today 3 (3), 10e16. Klatte, M., Hellbrück, J., 2010. Effects of classroom acoustics on performance and wellbeing in elementary school children: a field study. In: Inter Noise, June, Lisbon, Portugal. Kleeman Jr., W.B., 1982. The need for ergonomics expertise in office design. Applied Ergonomics 13 (2), 125e127. Liebl, A., Haller, J., Jödicke, B., Baumgartner, H., Schlittmeier, S., Hellbrück, J., 2012. Combined effects of acoustic and visual distraction on cognitive performance and well-being. Applied Ergonomics 43, 424e434. Malcolm, J.B., 1972. Office landscape: does it work? Applied Ergonomics 3 (4), 224e236. Masiero, B.S., Iazzetta, F., 2005. Estudo e implementação de métodos de medição de resposta impulsiva. Acústica e Vibrações 36, 34e41 (in Portuguese). Ollerhead, J.B., 1973. Noise: how can the nuisance be controlled? Applied Ergonomics 4 (3), 130e138. Ondet, A.M., Suer, J., 1995. Development and validation of a criterion for assessing the acoustic performance of industrial rooms. Journal of the Acoustical Society of America 97 (3), 1727e1731. Parsons, K.C., 1995. Ergonomics of the physical environment e international ergonomics standards concerning speech communication, danger signals, lighting, vibration and surface temperatures. Applied Ergonomics 4, 281e292. Passero, C.R.M., 2009. Avaliação e Adequação Acústica de Escritórios Panorâmicos através de Medições e Simulações. Tese de Doutorado, PhD thesis, Curitiba (in Portuguese). Passero, C.R.M., Zannin, P.H.T., 2010. Statistical comparison of reverberation times measured by the integrated impulse response and interrupted noise methods, computationally simulated with ODEON software, and calculated by Sabine, Eyring and Arau-Puchades’ formulas. Applied Acoustics 71, 1204e1210. Rashid, M., Zimring, C., 2008. A review of the empirical literature on the relationships between indoor environment and stress in health care and office settings e problems and prospects of sharing evidence. Environment and Behavior 40 (2), 151e190. Roy, K., 2007. Acoustic design in the new architectural trend of “exposed structures”. In: 19th International Congress on Acoustics, Madrid, 2e7 September. Sundstrom, E., Herbert, R.K., Brown, D.W., 1982. Privacy and communication in an open-plan office e a case study. Environment and Behavior 14 (3), 379e392. Van Der Voordt, T.J.M., 2004. Productivity and employee satisfaction in flexible workplaces. Journal of Corporate Real Estate 6 (2), 133e148. Vorländer, M., 1995. International round robin on room acoustical computer simulations. In: International Congress on Acoustics. Institute of Acoustics, United Kingdom, pp. 689e692. Wang, L.M., Bowden, E.E., 2006. Finding connections among indoor noise criteria, human perception, and worker performance. Acoustics Today 2 (1), 33e37. Yildirima, K., Akalin-Baskayab, A., Celebia, M., 2007. The effects of window proximity, partition height, and gender on perceptions of open-plan offices. Journal of Environmental Psychology 27, 154e165. Zannin, P.H.T., Marcon, C.R., 2007. Objective and subjective evaluation of the acoustic comfort in classrooms. Applied Ergonomics 38, 675e680. Zannin, P.H.T., Zwirtes, D.P.Z., 2009. Evaluation of the acoustic performance of classrooms in public schools. Applied Acoustics 70, 626e635.