Insitu attenuation of speech in offices with mixed activities according

to ISO 22955

Morten Roar Berg 1 Ecophon AB Box 500 | SE-265 03 Hyllinge | Sweden Alexander Andreas Hasselström 2 Technical University of Denmark Ørsteds Plads, Bygn. 348, 2800 Kgs. Lyngby | Denmark Cheol-ho Jeong 3 Technical University of Denmark Ørsteds Plads, Bygn. 348, 2800 Kgs. Lyngby | Denmark

ABSTRACT In the new international standard “ISO 22955: Acoustics – Acoustic quality of open offices” six different space types are described in terms of acoustic characteristics and target values. Space type 6 describes open offices for mixed activities which encompass the majority of open office spaces. The standard generally prescribes acoustic target values depending on activity for each space type. Space type 6 (table 6) instead prescribes target values for the descriptor insitu acoustic attenuation of speech (D A,S ) between different types of activities in the same space. Because this approach of prescribing acoustic target values based on difference in activity is novel, further investigation is warranted to explore and support this approach. The present paper details measurements of D A,S values in three offices with varying degrees of acoustic treatment and either activity-based layout or division of activity types in the same office space. Only one of the six the measured adjacencies lived up to the standardized target values. Acoustic design measures and findings related to the application of ISO 22955’s table 6 are discussed. It is concluded that certain principles can be useful when evaluating open offices through ISO 22955’s Table 6. 1. INTRODUCTION

Features of offices relating to noise and acoustics are consistently rated as the second and third most problematic aspects of workplaces in the global employee workplace experience survey, The Leesman Index [1]. Irrelevant speech (i.e. lack of speech privacy) is the most common reason for 1 morten.berg@ecophon.se 2 hasselstrom.alexander@gmail.com 3 'Cheol-Ho Jeong'

Jai. inter noise 21-24 AUGUST SCOTTISH EVENT CAMPUS ? O ? . GLASGOW

distractions in offices [2]. Irrelevant speech has been demonstrated to have several negative effects on knowledge workers, including:

 Reduced short and long term memory recall [3].

 Reduced wellbeing [4].

 Reduced arithmetic accuracy [5].

 Increased stress [6].

In 2012 measurement techniques for acoustic descriptors indicating degrees of speech privacy were standardised in ISO 3382-3 [7]. These techniques were verified t hrough an international round robin test in 2021 [8] and studies have documented that sound-absorbing acoustic materials can reliably improve technical descriptors of speech privacy [9] and reduce perceived disturbance and stress in office workers [10]. ISO 3382-3 is gradually gaining in usage in the global markets but is still not widely applied. Its contents are generally not described in specialist reports on existing spaces nor as building performance target values contained in design briefs and building programs in the global markets. A revised version of 3382-3 was published in 2022. However, through investigation of 21 workspaces Haapakangas [11] indicated lacking correlation between the acoustic descriptors detailed in ISO 3382-3 and noise annoyance, except for the descriptor “distraction distance” which seemed better suited to predict perceived disturbance by noise in open-plan offices. In the 2019 paper: “ISO 3382-3: Necessary, but not sufficient – A new approach to acoustic design for activity-based-working offices,” Harvie-Clark [12] builds on these findings. It is argued that t he parameters from ISO 3382-3 are determined in an unoccupied office and therefore cannot include relevant factors for acoustic satisfaction such as type of work, culture, occupation density, etc. It is further argued that target values for acoustic quality should be determined based on the needs of the users of the office in question. Based on this reasoning, a new approach for evaluating acoustic quality of offices with mixed activities is introduced. This approach became the theoretical background of Table 6 in ISO 22955 [13], published in 2021. In short, ISO 22955 specifies guidelines for what constitutes acoustic quality in open offices in two primary ways: Firstly, by clearly defining the most important principles of room acoustic design for the overall office space and secondly, by describing six space types for specific work activities as well as their acoustic characteristics and target values. The new descriptor insitu attenuation of speech between workstations ( D A,S ), which logically denotes dampening of speech in decibels between two work activities in an office, was introduced. Of the space types detailed in the standard space type 6 is especially relevant as it deals with spaces encompassing mixed activities. Most open office spaces fall into this category. To support design of spaces for various activities, table 6 provides a list of potential target values of D A,S between different activity types. As this approach is novel in terms of prescribing acoustic target values this, study aims to explore and substantiate this approach.

2. BACKGROUND

ISO 22955’s Table 6 lists typical office spaces based on activity type along with the adequate acoustic dampening of speech between them. Target values are traditionally prescribed for an entire room. If on room contains spaces with different acoustic needs this approach is not always functional. ISO 22955’s Table 6 introduces a new method of prescribing target values that takes this issue into account.

Table 1: ISO 22955’s Table 6. Activities listed from noise-generating to noise-sensitive from the bottom to the top in the first column and left to right in the top row.

Table 6 - Potential D A,S ratings between different types of spaces

Collaborative

Non- collaborative

Source/receiver:

Informal meetings

Outside of

Focused

Focused individual

-------------------

the room communication

(open

phone

Space type

(phone)

work

plan)

Social and welfare 15 15 18 24 27 32

15 12 15 21 24 29

Informal meetings

(open plan)

12 18 21 29

Outside of the

room Communication

(phone)

Collaborative 18 21 26

Non- collaborative 18 23

Focused phone 21 26

NOTE 1: In order to keep the noise level within the social and welfare space under control and avoid Lombard effect, a certain amount of absorption is needed. It is recommended to have an absorption area of at least 90 % of the floor surface.

A / S Floor = 0,9.

NOTE 2: These values are derived based on assumptions regarding background sound levels, source vocal effort, and proposed signal to noise ratios. These values may vary depending on the context.

As both the title and NOTE 2 of the table puts forward, these values are not generalizable and depend on several factors pertaining to the office in question. ISO 22955’s Table 6 references the formerly mentioned 2019 paper by Harvie-Clark [12] providing professionals with guidance on how to generate suiting D A,S target values for specific offices. To ensure comprehensiveness of the present paper a rough summary of the method by which the target values of ISO 22955’s Table 6 were generated is provided below. For a more exhaustive argumentation and theoretical justification for this approach, please see the original source [12]. The 5 step method of estimating target values in ISO 22955’s Table 6. 1. Characterizing acoustic conditions by liveliness.

2. Selecting activity type voice level.

3. Determining suitable workstation noise levels.

4. Evaluating suitable signal-to-noise ratios (SNR).

5. Based on step 1-4, estimate required D A,S target values between different types of activites.

In principal step 1, 2 and 3 require pre-existing or assumed data about the acoustic environment to support estimation of target values given the specific circumstances. The values in ISO 22955’ Table 6 is based on presumed values for steps 1-3 as described below.

2.1 Liveliness (Step 1)

Liveliness [14] is an acoustic measure that indicates how much sound levels of an environment fluctuate over time. Liveliness ratings are given in Table 2 below.

Table 2: Liveliness ratings assumed for target values of ISO 22955’s Table 6.

Non -

Informal meetings (open plan)

Outside of

Focused

Focused individual

phone

the room communication

Collaborative

collaborative

work

(phone)

7 6 6 5 4 3

Liveliness

rating

2.2 Speech level (Step 2)

An average A-weighted source speech level of 57 dB at 1 m is assumed for target values of ISO 22955’s Table 6.

2.3 Workstation noise levels (Step 3)

Workstation noise levels can vary greatly due to many factors. The assumed levels for the proposed target values is listed in Table 6. Table 3: ISO 22955’s table 5 lists noise levels assumed for target values of ISO 22955’s Table 6.

Workstation noise levels assumed for different types of activity

Informal meetings (open plan)

Non-

Outside of

Focused

Focused individual

phone

the room communication

Collaborative

collaborative

work

(phone)

48 48 45 42 42 40

Workstation

noise level

(dBA)

2.4 Determining (Step 4)

By the liveliness ratings of Table 2 a signal to noise ratio ( SNR ) can be calculated by the following equation (Equation 1):

𝑆𝑁𝑅= (𝑆𝑜𝑢𝑟𝑐𝑒𝐿𝑅−𝑅𝑒𝑐𝑖𝑒𝑣𝑒𝑟𝐿𝑅) ∗3 𝑑𝐵 (1)

2.5 Defining required D A,S target values (Step 5)

Target D A,S values for an open office with mixed activities can be estimated with the following equation (Equation 2):

𝐷 ஺,ௌ = 𝐿 ௦,஺ −𝑆𝑁𝑅−𝐿 ௕ (2)

SNR: Equation 1. L b : Background noise level at workstation Table 2. L s,A : Speech source level at 1 meter assumed at 57 dB.

3. METHOD

To investigate the application of ISO 22955’s Table 6 an explorative approach to measurements were applied. Measurements were evaluated against both estimated and proposed target values from the standard. The estimated target values took the workplace noise levels of the specific offices into account but assumed the same Liveliness and speech source levels as the proposed target values. Both modified and proposed target values are valuated against the measurements. Through this approach, the difference between a ‘blind’ application of the target values and a more detailed approach is explored. Three very different office spaces and different activity adjacencies were examined. Measurements were conducted during the winter of 2019 (O3) as well as the winter and spring of 2022 (O1-2). Only offices with acoustic ceilings were examined. Acoustic aspects of the offices such as absorption area, floor plan and interior design etc. varied significantly. Two of the three offices had distinct activity- based layouts. The other office had clearly defined differences in work activities between workstations areas. Offices with activity-based layouts were an office collective in a sustainability innovation hub and a university administration department, respectively. The third office was an online supermarket administration office. Equipment from Brüel & Kjær and 01dB was utilized. Measurements were conducted according to the normative Annex A.2 of ISO 22955. Workstation noise level measurements were carried out during working hours and measurements of insitu attenuation of speech were carried out before or after normal working hours. An omnidirectional loudspeaker and measurement microphone were placed at specific workstations where they were most representative of the actual source and receiver positions of the offices. The heights of the measurement source and receiver were set to 1.2 meters to represent the approximate height of the ears and mouth of a seated worker. Measurement positions were selected to be a minimum of 0.5 meters from any reflective surface to represent the typical position that workers may occupy for the given activity type. Two source positions and three microphone positions were selected for measurements between every two workstations. Every measurement lasted a minimum of 6 seconds. The measurement source positions were chosen to represent the noisiest activity types and the receiver represented the most noise-sensitive activity types. No additional measurements were made due to results being close to the estimated or proposed target values ISO 22955’s Table 6.

4. RESULTS

In the following sections cursory descriptions of relevant acoustic aspects of the measured offices, plan drawings, measurement paths and measured D A,S values are presented. Proposed and modified target values and measurement techniques are evaluated.

4.1 Office 1 (O1) – Innovation Hub

The workspace layout is designed for activity-based working. An acoustic perforated metal ceiling (αw = 0.9) is suspended at 2.6 m. Cooling baffles are installed intermittently in the suspended ceiling, reducing the overall absorption area added by the ceiling. Only vertical absorption present was soft felt absorbers (~ 4 mm) in the informal meeting area and the remaining vertical surfaces are all triple insulated heavy glass. Wooden flooring covers the entire office space and interior design is estimated to add an average degree of diffusion to the room.

Table 4: O1 proposed and estimated target values

Proposed target

Estimated target value*

value in ISO 22955

15 dB 19 dB

Informal meetings -> Collaborative

Non- collaborative

18 dB

14 dB

-> Collaborative

* Based on method summarized in 2.1-6 and measured noise levels

Das = 18.4 dB i Vo Das=21.1dB |”

Figure 1: O1 plan drawing. Colour coded; green for informal meetings, yellow for collaborative,

blue for non-collaborative. Arrows indicate measurement paths.

Figure 2: O1 Picture - Collaborative area.

Measurement paths were chosen across the entire office. Measurement paths were therefore quite long as shown in Figure 1. For this reason, measured D A,S values complies with the proposed target values of ISO 22955’s Table 6 and the estimated target values for this space. It is highly unlikely this would be the case if adjacent workspaces had been measured.

4.2 Office 2 (O2) – Technical University of Denmark

The workspace layout is sectioned into different areas based on activity type. An acoustic perforated metal ceiling (αw ~ 0.8) is suspended in the ceiling at 2.6 m. Except for draperies for sectioning up the space visually, no vertical absorption is present. Other vertical surfaces are made of smooth vinyl tiles or triple insulated heavy windows. Stone tile flooring covers the entire space and the interior design is estimated to add a minimal degree of diffusion to the room.

Table 5: O2 proposed and estimated target values

Proposed target

Estimated target value*

value in ISO 22955

15 dB 1,3

Informal meetings -> Collaborative

21 dB 21,1

Informal meetings ->

Non- collaborative

* Based on method summarized in 2.1-6 and measured noise levels

Figure 3: O2 plan drawing. Colour coded; green for informal meetings, yellow for collaborative,

blue for non-collaborative. Arrows indicate measurement paths.

ly=55,5 dB — | Ly = 58,7 dB & Ly = 39,9 dB Das= 12.9 dB | cs — = Das = 12.9 dB

Figure 4: O2 Picture – Informal meeting area

Figure 5: O2 Picture - Collaborative area

Measurements showed higher workstation noise levels in the collaborative area than the informal meeting area. For this reason, measurement path direction should arguable be reversed +++ and estimated target values would be slightly higher. Due to higher noise levels than assumed in Table 3 one of the measurement paths complies with the estimated target values of ISO 22955’s Table 6. The most noise-sensitive activity was the non- collaborative and its closest and noisiest workplace adjacency was the collaborative area. A measurement path between these two areas would therefore have been more indicative of the acoustic quality of the office.

4.3 Office 3 (O3) – Online supermarket coordination office

The workspace layout is classically designed with rows of double desks with workers’ stations facing each other. The acoustic ceiling of perforated plaster (αw = 0.60) is suspended at a height of 2.4 m. Acoustic tiles are interspersed with lines of non-absorbent smooth tiles of plaster. Installations in the suspended ceiling are relatively sparse relative to modern ceilings. All walls are heavy concrete with heavy double-paned windows. There is no vertical absorption present. Modern hygienic carpet (αw ~ 0.10) covers the entire floor and low wooden shelving between every other desk is the only interior

design elements besides desks. The interior design of the open office space and interior design is estimated to add an average degree of diffusion to the room.

Table 6: O3 proposed and estimated target values

Proposed target

Estimated target value*

value in ISO 22955

15 dB 13

Collaborative

-> Non- collaborative

18 dB 13

Informal meetings ->

Non- collaborative

* Based on method summarized in 2.1-6 and measured noise levels

Das = 9.6 dB

Figure 6: O3 plan drawing. Colour coded; yellow for collaborative, purple for phone intensive, blue

for non-collaborative. Arrows indicate measurement paths.

Figure 7: O3 Picture – Concentration area

None of the measurement paths complies with the target values of ISO 22955’s Table 6. The most noise-sensitive activity was the non-collaborative area and its closest workplace adjacency was the collaborative area. However, the complaints of the users in the non-collaborative area mainly concerned noise from the "phone-focused" area. It seems the proposed target values or modified values of ISO 22955’s Table 6 would be very hard to reach with this type of office layout, even with a much higher degree of acoustic absorption.

5. DISCUSSION

Three offices of distinctly different designs were investigated through explorative measurement approaches. They suggest that certain principles can be useful to keep in mind when evaluating open offices through the estimated or proposed target values in ISO 22955’s Table 6. Because these target values are based on signal-to-noise ratios, certain data about the acoustic environment has to be available or assumed before target values can be estimated. The proposed target values in ISO 22955’s Table 6 assume lower noise levels than were measured in this investigation. The most critical adjacencies (noise generating to noise-sensitive workspaces) are most indicative of acoustic quality in open offices. Offices for activity-based working will typically be better equipped to comply with the method behind ISO 22955’s Table 6. Classically designed offices will typically have fewer square meters per workspace and a homogenous typology. Given the lack of vertical absorbers and free-standing acoustic screens, it is not surprising that only one measurement path in the investigated offices live up to the proposed or estimated target values of ISO 22955’s Table 6. Especially considering recent research by Keränen [9] into the effects of absorbing materials on speech privacy in offices which indicates that both screens and wall panels are necessary to ensure adequate speech privacy. Brockman & Klein’s [14] investigations into D A,S simulation similarly highlight this point. In relation to ISO 22955’s Table 6, this could indicate that typical offices generally do not live up to the given standards of acoustic quality, lacking vertical absorption both in the form of wall absorbers and acoustic screens. Conversely, the proposed target values in Table 6 might be relatively high due to assuming relatively low workstation noise levels. Further research of the application of the method of ISO 22955’s Table 6 is highly relevant to support the optimal implementation and future development of this part of the standard.

6. CONCLUSION

ISO 22955’s Table 6 presents a new way of evaluating acoustic quality in open office spaces with mixed activities by specifying acoustic needs based on activity types. To test and substantiate this approach three offices were measured and evaluated through an explorative approach. Findings indicate that certain principles should be upheld when applying this approach. The target values of ISO Table 6 should be modified to fit the acoustic environment in question. When evaluating workplaces through this approach main adjacencies or the most critical (noise generating to noise-sensitive) measurement paths should be evaluated.

7. REFERENCES

1. Leesman Index (2021) Leesman review, issue 31, p. 15, Review Global Workplace Survey,

https://www.leesmanindex.com/media/Leesman-Review-31-DPS.pdf 2. Roper, Kathy & Juneja, Parminder. (2008). Distractions in the workplace revisited. Journal of

Facilities Management. 6. 91-109. 10.1108/14725960810872622. 3. Zimmer, Karin. (2014). The psychoacoustics of the Irrelevant Sound Effect. A review. Acoustical

Science and Technology. 35. 10.1250/ast.35.10.

4. Haka, Matts & Haapakangas, Annu & Keränen, Jukka & Hakala, Jarkko & Keskinen, Esko &

Hongisto, Valtteri. (2009). Performance effects and subjective disturbance of speech in acoustically different office types - A laboratory experiment. Indoor air. 19. 454-67. 10.1111/j.1600-0668.2009.00608.x. 5. Banbury, S. and Berry, D.C. (1998). Disruption of office-related tasks by speech and office noise.

British Journal of Psychology, 89: 499-517. https://doi.org/10.1111/j.2044-8295.1998.tb02699.x 6. Radun, Jenni & Maula, Henna & Rajala, Ville & Scheinin, Mika & Hongisto, Valtteri. (2020).

Speech is Special. The Stress Effects of Speech, Noise, and Silence during Tasks Requiring Concentration. Indoor Air. 31. 10.1111/ina.12733 7. International Organization for Standardization. (2022). Acoustics – Measurement of room

acoustic parameters – Part 3: Open offices. (ISO Standard No. 77437). Retrieved from: https://www.iso.org/standard/77437.html 8. Hongisto, Valtteri & Keränen, Jukka & Labia, Laura & Alakoivu, Reijo. (2021). Precision of ISO

3382-2 and ISO 3382-3 – A Round-Robin test in an open-plan office. Applied Acoustics. 175. 107846. 10.1016/j.apacoust.2020.107846. 9. Keränen, Jukka & Hakala, Jarkko & Hongisto, Valtteri. (2020). Effect of sound absorption and

screen height on spatial decay of speech – Experimental study in an open-plan office. Applied Acoustics. 166. 107340. 10.1016/j.apacoust.2020.107340. 10. Seddigh, Aram & Berntson, Erik & Jönsson, Fredrik & Bodin Danielsson, Christina &

Westerlund, Hugo. (2015). Effect of variation in noise absorption in open-plan office: A field study with a cross-over design. Journal of Environmental Psychology. 44. 34-44. 10.1016/j.jenvp.2015.08.004. 11. Haapakangas A, Hongisto V. Distraction distance and perceived disturbance by noise — An

analysis of 21 open-plan offices, The Journal of the Acoustical Society of America 141, 127 (2017 ); https://doi.org/10.1121/1.4973690 . 12. Harvie-Clark J., Opsanger C., Larrieu F. A new approach to acoustic design for activity-based

working offices , Proceedings of the ICA 2019 Aachen (Germany), 2019, pp. 2407-2414. URL: http:// pub .dega -akustik .de/ ICA2019/ data/ articles/ 000840 .pdf 13. International Organization for Standardization. (2021). Acoustics - Acoustic quality of open

offices. (ISO Standard No. 74237). Retrieved from: https://www.iso.org/standard/74237.html 14. “ Zur Anwendung der ISO 22955 in einem simulierten Multi-Space“ Holger Brokmann, Achim

Klein, Rainer Machner, Jack Harvie-Clark, Presented at the DAGA 2022, Stuttgart