TH 7687.5 B73 HOE F1 Environment 1 (5D Ontario Copyright Provisions and Restrictions on Copying: This Ontario Ministry of the Environment work is protected by Crown copyright (unless otherwise indicated), which is held by the Queen's Printer for Ontario. It may be reproduced for non-commercial purposes if credit is given and Crown copyright is acknowledged. It may not be reproduced, in all or in part, for any commercial purpose except under a licence from the Queen's Printer for Ontario. For information on reproducing Government of Ontario works, please contact ServiceOntario Publications at copyrightc'u ontario.ca r ISBN 0-7729-9094-8 1 1 SURVEY OF OUTDOOR AIR CONDMONER NOISE FINAL REPORT - Revised February 27, 1991 RAC Report # 458G Report prepared for: Research and Technology Branch 1 Ontario Ministry of the Environment 1 Report prepared by: J.S. Bradley of the National Research Council Canada and the Heating, Refrigerating and Air Conditioning Institute of Canada 1 DECEMBER 1991 0 "" FUMB W Af6E Cette publication technique nest disponible qu'en anglais. Copyright: Queen's printer for Ontario, 1991 This publication may be reproduced for non-commercial purposes with appropriate attribution. PIBS 1761 1 1 1 A i3 l i I,- 7 '7- %-/61,3/ r- G 1 1 1 ACKNOWLEDGEMENT AND DISCLAIMER This report was prepared for the Ontario Ministry of the Environment. The views and ideas expressed in this report are those of the author and do not necessarily reflect the views and policies of the Ministry of the Environment, nor does mention of trade .names or 1 commercial products constitute endorsement or recommendation for use. The Ministry, however, encourages the distribution of information and strongly supports technology transfer and diffusion. Any person who wishes to republish part or all of this report shouted apply for permission to do so to the Research and Technology Branch, Ontario Ministry of the Environment, 135 St. Clair Avenue West, Toronto, Ontario, M4V 1P5, Canada. 1 1 1 Copyright: Queen's Printer for Ontario This publication may be reproduced for non-commercial purposes with appropriate attribution. 1 CR5954.2, Revised Page I of 52 Executive Summary This report presents the results of a survey of both subjective and objective ratings of outdoor domestic air conditioner noise. The survey was carried out in the Metro Toronto area, and both subjective and objective data were successfully obtained for 550 subjects. Subjects were selected to include the four combinations of low and high ambient noise level sites with owners and non- owners of air conditioners. In many cases, air conditioner noise levels were not substantially higher than existing background noise levels, making it difficult to obtain strong relationships between the air conditioner noise stimulus and subjective responses. Some highlights of the results include: (1) In spite of the inherent difficulties described above, adverse responses were significantly related to objective noise measures. While there were many statistically significant relationships between objective and subjective responses, the variance explained was often quite small and one should be careful not to assume these results to be more precise than is really justified. (2) Adverse responses were more strongly related to signal/noise ratios than to the measured noise levels. (3) Below 50 to 55 dB(A) (as measured at the property line), annoyance was not systematically influenced by air conditioner noise level. Similarly, below a signal/noise ratio of about +5 dB(A), annoyance was not related to measured noise levels. 1 (4) At sites with higher ambient noise, subjects were less annoyed by their neighbours' air conditioner noise by the amount expected due to the decreased signal/noise ratio. (5) Owners of air conditioners were less annoyed by air conditioner noise by an amount equivalent to 7 dB(A) lower air conditioner noise levels. (6) The ARI 275-84 prediction scheme was not found to be particularly accurate. Measured levels close to the air conditioner source were on average several decibels lower than predictions. Similarly, measured values at the facade of the neighbours' home were on average not predicted well. In most cases there was a quite large scatter in the comparison of predicted and measured values. CR5954.2, Revised Page 2 of 52 .................... ... Table of Contents 1. Introduction ... ... .. .. ........... .. .... ..., . .. 3 1 2. Methodology 4 . 2.1 Experimental Design 4 . . . . . . . . . . . . . . . . . . . . . . . . 2.2 The Questionnaire . . . . . . . . . . . . . . . . . . . . . . . . . .4 2.3 Noise Measurements. .. .. .. .. .. .. .. .. . ... . ... .. .. .. .. .. .. .. .. .. .. .. .. 5 3. The Survey Sample ... 7 .. .. .. 3.1 Acoustical Da.ta. .... .. .. ....... .... .. . . .. .. .. . ... .. .. .. .. . 7 3,2 Survey Data ............. .. .. ... .. .... 11 3.3 Quality of the Data .... ..... .... .13 .. 4. Initial Individual Subject Analyses ... .14 .. 5. Individual Data - Methods to. .Im..pr o.v.e .C.o.r re.l.a ti.o.n.s .. ... . 17 5.1 Composite Response Scales .... 17 . 5.2 Selecting Data According to Signal/Noise Ratio 18 . . . . 5.3 Signal/Noise Ratio as a Better Predictor 23 . . . . . . . . .. . . .. . . 6. Individual Data and Subject Selection Variables 27 . . . . 6.1 Success of Subject Selection P.ro.c.ed.u.re. ....... ...... 27 . . . . . . . . . . . . . 6.2 Analysis of Variance Results 27 6.3 Influence of Air Conditioder Ownership . .. ... . ... . . . . ... .. .. . 28 .. 6.4 Influence of Ambient Noise Conditions 30 . 7. Individual Data a.n.d. N.o.n.-N.o.is.e. P r.ed.i.c.to.r s .. .. .. .. . . .. . . .. . . . . 32 8. Grouped Data . . .35 8.1 Simple Relations with Noise Level Predic..to r.s. . . .. . . .. . . .. .. ....3 5 8.2 Influence of Subject Selection Vari.ab.l.es. . .. .. .. ...... 40 8.3 Influence of Non-Noise Predictors 42 9. Comparisons with the ARI Noise Level Prediction Procedure .43 9.1 Predictions from Manufacturers Ratings . . .. . . . .. . . .. ... . ... .. 4463 9.2 Predictions from Measured Source Levels . . 9.3 Summary . . . ... .. . .. .. . ... . . ... .. . ... . . . . . . . . . . . . . . 49 10. Conclusions . . . . . . . . . . . . . . . . .50 Appendix 1. The Questionnaire s Appendix 2. Statement of Work 1 I CR59542, Revised Page 3 of 52 1. Introduction This is the final report of a survey of both sound levels and subjective attitudes to residential air conditioner noise levels. The work was a joint project of the National Research Council (NRC), the Ontario Ministry of the Environment (MOE), and the Heating Refrigerating and Air Conditioning Institute of Canada (HRAI}. Support for the field measurements was also provided by the City of North York, the City of Toronto, Toronto Hydro, the Ontario Home Builders' Association, Ontario Hydro, and the Ontario Ministry of Housing. MOE was responsible for finding appropriate measurement sites and for all field measurements. NRC planned the experimental design and carried out all analyses of the data including the preparation of this report. 'The purpose of the survey was to consider the question of acceptable limits for outdoor residential air conditioner noise. The terms of agreement of the contract under which this work was performed are included in Appendix. 2. The principal objectives of the project were: (a) to quantify the relationship between subjective response to neighbours' air conditioner noise and the measured noise levels of these air conditioners; (b) to provide information to support the setting of acceptable limits for outdoor air conditioner noise levels; (c) to examine how local ambient noise and ownership of an air conditioner influence adverse reactions to air conditioner noise; (d) to consider the influence of other non-noise factors on adverse reactions to neighbours' air conditioner noise; (e) to compare measured air conditioner noise levels with predictions based on the ARI 275-84 procedure. It was initially planned that all measurements would be made during the summer of 1989 and that analyses of the data would be completed in the fall of 1989. The first measurement phase was not completely successful and a second phase of measurements was carried outduring the summer of 1990 at sites in the Metro Toronto area. The final data set included questionnaire responses along with air conditioner and ambient noise measurements for 550 subjects. This report describes the analysis of this data. CR5954.2, Revised Page 4 of 52 2. Methodology It had been suggested that residents in noisier areas might be more tolerant of noise from their neighbours' air 2.1 Experimental conditioner, and that subjects who owned air conditioners Design would be less disturbed by the noise from their neighbours' unit. The experimental design included a subject selection procedure that would make it possible to determine the effects of ambient noise levels and ownership of an air conditioner on adverse responses to air conditioner noise. The goal was to interview a total of 600 subjects from four equal groups. The groups were to be the four combinations of high and low ambient noise with owning and not owning an air conditioner. Thus, it was hoped to obtain 150 subjects in each of these four groups. Of course, each subject had to have at least one neighbour with a central air conditioner or heat pump. Subjects also had to be at least 18 years of age, and it was hoped to get an approximately equal number of males and females. Potential subjects first received an introductory letter from the Ontario Ministry of the Environment describing the survey as concerning particular aspects of the 1 neighbourhood environment, and encouraging them to participate. Shortly after receiving the introductory letter, subjects were approached by an interviewer, and an interview was carried out at the convenience of the subject. No mention was made that the survey concerned air conditioner noise until after the interview. However, after a few questions, subjects were asked if any of their immediate neighbours had a central air conditioner. If not, the subject was thanked and the interview terminated. After the interview was successfully completed, permission to make noise measurements was requested. These were made as soon as possible after the completion of interviews in that particular area. As a final gesture, each subject was given a thank you letter expressing appreciation for their help, revealing that the survey was focussed on air conditioner noise, and giving the names of the three main sponsoring groups. 2.2 The Questionnaire The questionnaire was administered directly to subjects in their homes by a trained interviewer. The interviewer wrote the coded responses on the questionnaire form. Most responses were in the form of seven-point response scales, and subjects were shown a response card from which to select a numerical response. The card included the integer numbers from 0 to 7; the value "0" was labeled CR5954.2, Revised Page 5 of 52 1 "Not Applicable", the value "1" was labelled "Not at all"; the value "4" was labelled "Medium", and the value "T' was labelled "Very". By not labeling intermediate values of the seven-point scale, it was hoped that subjects would linearly interpolate between the extremes of the end values and valid interval scale response data would be obtained. Where subjects refused to answer particular questions (e.g. concerning income), a %1" response was to be entered. A copy of the questionnaire is included in Appendix 1. The first two questions in the interview were open-ended questions about things subjects liked and didn't like about their"neighbourhood. This gave respondents a chance to spontaneously mention air conditioner noise without any prompting by more direct questions. This was followed by questions concerning how considerate their neighbours were thought to be, and whether an immediate neighbour had a central air conditioner or heat pump. This initial section was followed by the main body of the questionnaire that included direct questions concerning how often they heard or were annoyed by various environmental noises. These included: road traffic noise, aircraft noise, train noise, and residential air conditioner noise. This block of questions was repeated four times. The questions were asked for day- and night-time conditions as well as for the subject being indoors or outdoors. This was followed by questions concerning possible sleep disturbance by each type of environmental noise. The final section of the questionnaire gathered the usual socio-economic information about respondents. Questions concerned the total family income, the subject's years of formal education, the subject's age, the number of adults and the number of young people in each home. They were also asked- whether air conditioner noise was more disturbing in a particular room, whether they owned a central or window air conditioner, and how much extra they would pay if buying a $2,000.00 central air conditioner for a very quiet model. The interviewer also recorded whether the subject was male or female, whether they appeared to have any hearing problems. and whether their home was a row house, semi-detached, or a detached home. 2.3 Noise The noise measurement data were first entered onto Measurements sheets attached to the subjects questionnaire responses. Integrated A-weighted sound level measurements were