Broner, N. The Missing 16 Hz – Can We Live With It?
19th INTERNATIONAL CONGRESS ON ACOUSTICS MADRID, 2-7 SEPTEMBER 2007
THE MISSING 16 HZ – CAN WE LIVE WITH IT?
Sinclair Knight Merz, 590 Orrong Road, Armadale, Victoria, Australia 3143;
As the need for power increases, power utilities are resorting to the use of peaking plants incorporating Open Cycle Gas Turbines. OCGT manufacturers generally supply noise data for these down to the 31.5 Hz octave band. However, most of these units also generate significant energy in the 16 Hz octave band. Both of these bands need to be considered when assessing potential noise impact on neighbouring residential communities. However, Regulatory Authorities generally have not considered this in their approach which is generally based on achieving a nominated A-weighted noise level criterion. This paper discusses criteria for low frequency noise including the 16 Hz octave band and the implications for the siting of peaking plants near residential or commercial/industrial locations.
The fact that low frequency noise can cause greater annoyance than the A-weighted level would suggest has been known for over 30 years eg Bryan , Broner [2, 3], Broner and Leventhall , Leventhall .
In particular, it is known that gas turbines and boilers can produce low frequency noise which can result in feelings of annoyance due to vibration induced rattle, nausea, headache and uneasiness.
In Australia over the last few years, there has been a drought which has recently culminated in the reduction of hydro power production. In order to meet the increasing power demand, utilities have turned to the use of peaking plants which utilise open Cycle Gas Turbines (OCGT’s) eg Figure 1. Generally, these plants are located away from residential areas but there have been recent instances where these plants have been placed or are planned to be close to residential locations. The question that then arises is how close to existing residential (or commercial) areas can these plants be placed without causing any acoustical (or other e.g. air quality) impact?
Figure 1.- Typical peaking power plant
There has been some concern raised about low frequency noise emission from such plants. Hessler  has described the low frequency noise problems that have occurred in the USA due to incorrect siting of such power plants close to residential areas. Typically, neighbours express complaints of low frequency rumble noise, vibration rattle, nausea and headaches. At low frequencies, apart from the spectral imbalance issue, a major factor in causing annoyance is the significant temporal level fluctuations that occur. Unfortunately, the Regulatory Authorities generally only require compliance with an A-weighted noise level criterion based on either amenity or intrusiveness considerations and also depending on the existing or planned zoning of the land. Where there is some consideration given to potential low frequency noise, this may be in the form of a penalty to be added to the base level if the low frequency noise exceeds a given threshold eg the New South Wales Industrial Noise Policy  considers that a 5 dBA penalty is to be incurred if the difference between the C- and A-weighted levels over the same time period is 15 dB or more. A similar approach is taken by the Alberta EUB  when the C- minus A-weighted difference exceeds 20 dB. However, this simplified approach is not likely to resolve potential situations where significant low frequency energy is causing annoyance and complaints.
DEGREE OF THE PROBLEM
When investigating the siting of a power station, it is necessary to obtain noise source data either from the manufacturers or by conducting site noise measurements at an existing installation. Currently, when seeking manufacturers data, the data range usually covers the range from either the 31.5 Hz or 63 Hz octave band to the 8 KHz band. This range is quite acceptable for noise sources with little low frequency noise contribution. However, Open Cycle Gas Turbines potentially generate significant low frequency noise, so the 16 Hz band must also be considered as well. Overlooking the 16 Hz band energy may result in an unacceptable noise environment in which annoyance due to vibration rattle, feelings of pressure, nausea or headaches may occur. Figure 2 shows manufacturers data for OCGT’s down to the 31.5 Hz octave band as well as data measured down to and including the 16 Hz octave band. It can be seen that the energy at the 16 Hz octave band is at least as important as that for the 31.5 Hz octave band.
See downloadable document at bottom of page for Figures
Figure 2.- Typical Sound Power Level Data
So, not including this significant energy can underestimate the degree of low frequency energy and the potential impact on neighbouring residences. The 16 Hz octave band sound pressure level must be considered!
HOW MUCH IS TOO MUCH?
There are a number of low frequency noise criteria that can be used for planning situations. Most are based on empirical evidence and data. An example is ANSI B133.8 -1977  which recognises that for installations where frame structures are occupied by people near to gas turbine installations, the A-weighted sound level alone does not adequately define permissible low frequency sound emissions. Indeed, Appendix B recommends the selection of a maximum C-weighted level outside the nearest occupied framed structure and suggests the upper limit should be selected not to exceed 75 – 80 dBC. The range of values was given due to uncertainty as to the sound level required to induce a structural vibration in a frame structure.
Hessler  also considered that experience since 1971 has shown that the recommendation of ANSI B133.8 is woefully inadequate and that the problem continues to occur for combustion turbine open cycle plants. He therefore proposed C-weighted levels supplementary to the A- weighted site criteria as follows:
Figure 3.- Maximum Allowable dB(C ) Levels at Residential Areas to Minimise Infrasound Noise and Vibration Problems
These levels contained no factor of safety or margin of error and Hessler cautioned that these levels should be considered the maximum allowable.
In discussing low frequency gas turbine noise, Newman and McEwan  quote a British Gas Corporation criterion for specifying noise control for gas turbines viz. 60 dB in the 31.5 Hz octave band at the nearest dwelling. This value was said to have been determined by review of the noise levels which complainants found satisfactory.
Similarly, Annex D of ANSI S12.9 – 2005/Part 4  deals with sounds with strong low frequency content and for essentially continuous sound where the C-weighted sound level exceeds the A-weighted sound level by at least 10 dB. Annex D provides a means for calculating an adjustment to the sound exposure level based on the summation of the time – mean – square sound pressures in the 16, 31.5 and 63 Hz octave bands. ANSI recognises that generally, annoyance is minimal when octave band sound pressure levels are less than 65 dB at these octave bands and that to prevent the likelihood of noise-induced rattles, the low frequency sound pressure level should be less than 70 dB.
The Oregon State Noise Control Regulations  for industrial and commercial noise sources also quote low frequency allowable octave band sound pressure levels for the 31.5 Hz and 63 Hz octave bands as 65 dB and 62 dB respectively for the night time period 10 PM – 7 AM (the limits are 68 dB and 65 dB for the daytime period 7AM – 10 PM respectively).
Further evidence regarding the role of low frequency noise including the 16 Hz octave band comes from the work of Blazier . He investigated about 200 background indoor noise environments and considerations of sound quality, speech communication masking criteria and of noise-induced vibration at low frequencies led to the development of the Room Criterion (RC) curves for indoor assessment of HVAC noise. Due to the incidence of audible vibration rattle or feelable vibration in lightweight wall and ceiling constructions, the curves also included two regions where vibration rattle could occur viz. those where the vibration would be marginally or clearly noticeable. For marginally noticeable, the sound pressure levels were 65dB, 65 dB and 69 dB for the 16Hz, 31.5Hz and 63Hz octave bands respectively (for the clearly noticeable limits, the levels are 75dB, 75dB and 79 dB respectively). The RC curves were adopted by ASHRAE  as the preferred criteria for room noise assessment in recognition of mounting low frequency noise problems indoors.
As a result of ASHRAE sponsored research into rumble noise due to HVAC systems, Broner  determined that there was a greater sensitivity to energy below 31.5 Hz than had been allowed for in the RC curves and therefore recommended a modification to the 1981 RC curves which he called the Low Frequency Room Criterion curves, LFRC (similar to the LFNR curves developed earlier by Broner and Leventhall ). Also as a result of this research, Blazier , amongst other improvements, recommended the modification of the shape of the RC reference curves in the octave -band centred at 16 Hz. The refined methodology was identified as the RC Mark II procedure (see Figure 4) and this method was adopted by ASHRAE in the latest revision of their Handbook. Thus further evidence that energy at 16 Hz must be considered in dealing with potential low frequency noise problems.
Figure 4.- RC Mark II Curves with Increased Sensitivity at 16 Hz 
WHERE SHOULD POWER PLANTS BE LOCATED?
It seems that in order to prevent low frequency noise complaints due to OCGT’s, it is necessary to consider the 16 Hz octave band noise level and to limit the noise outside the nearest residences to the order of 60 – 65 dBC. What does this mean in terms of the siting of power plants “near” to residential areas?
Unfortunately, there are many variables that need to be considered when wanting to recommend a minimum distance away for the nearest residence. These include:-
• The power generation equipment itself – its size, package configuration and most importantly, the level of “standard” noise attenuation provided by the manufacturers. Most “standard” packages may reduce some low frequency noise but would not be aimed at achieving a significant noise reduction at 31.5 Hz and certainly not at 16 Hz, due to cost and size requirements.
• The local meteorological effects – e.g., temperature inversions may be a common occurrence in an area and can significantly increase the sound pressure level from the plant at the residence(s). Also, which way does the wind blow and for how long and at what speed?
• The background noise level in the area of the residence(s) – the background noise might help to mask noise from the plant.
As a rule of thumb based on case histories and Sound Power Level considerations, in practice, we would recommend that for OCGT plants with a total Sound Power Level in the range 115 – 120 dBA, the minimum distance to the nearest residential premises should be no closer than the order of 1500 – 2000 metres away.
With increased use of Open Cycle Gas Turbine power plants, there is a need to consider acoustic energy not only down to the 31.5 Hz octave band but also to the 16 Hz octave band. To prevent low frequency noise complaints, the noise level outside the nearest residences should be limited to the order of 60 – 65 dBC. In practice, this means that OCGT power plants with Sound Power Level of the order of 115 – 120 dBA should be sited so that they are at least of the order of 1500 – 2000 metres away from the nearest residential premises.
 M.E. Bryan: Low Frequency Noise Annoyance in Infrasound and Low Frequency Vibration edited by W. Tempest, Academic Press, London 65 – 96
 N. Broner: The Effects of Low Frequency Noise on People – A Review. Journal of Sound and Vibration 58(4) 1978 483-500
 N. Broner: A Criterion for Low Frequency Noise Annoyance. 10th ICA Sydney 1980, Paper C1-4.4
 N. Broner, H.G. Leventhall: Low Frequency Noise Annoyance Assessment by Low Frequency Noise Rating (LFNR) Curves. Journal of Low Frequency Noise and Vibration 2(1) 1983 20-28
 H.G. Leventhall: A Review of Published Research on Low Frequerncy Noise and Its Effects. Research Project Report (2003) http://www.defra.gov.uk/environment/noise/research/lowfrequency/pdf/lowfreqnoise.pdf
 G. F. Hessler Jr: Proposed Criteria for Low frequency industrial Noise in Residential Communities. Journal of Low Frequency Noise, Vibration and Active Control 24, No.2 (2005) 97-105
 New South Wales Industrial Noise Policy. Environmental Policy Branch, NSW Environment Protection Authority January 2000
 Alberta Energy and Utilities Board, Noise Control Directive 038 feb 16, 2007
 ANSI B133.8 – 1977 Gas Turbine Installation Sound Emissions Reaffirmed 1989 and 2001
 J.R. Newman, K.I. McEwan: Low Frequency Gas Turbine Noise. Transactions of the ASME 102 (1980) 476-481
 ANSI S12.9-2005/Part 4 Quantities and Procedures for Description and Measurement of Environmental Sound – Part 4: Noise Assessment and Prediction of Long-term Community Response
 Oregon Department of Environmental Quality, Noise Control Regulations for Industry and Commerce OAR 340-035-0035 http://www.deq.state.or.us/regulations/rules.htm
 W.E. Blazier: Revised Noise Criteria for Application in the Acoustical Design and Rating of HVAC Systems. Noise Control Engineering, 16 (2) (1981) 64-73
 ASHRAE Handbook: Sound and Vibration Control. Chapter 47 Atlanta (2003) American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc.
 N. Broner: Determination of the Relationship Between Low Frequency HVAC Noise and Comfort in Occupied Spaces – Subjective Phase. ASHRAE 879-RP (2004) Vipac Report 360120
 W.E. Blazier: RC Mark II: A Refined Procedure for Rating the Noise of Heating, Ventilating and Air-conditioning (HVAC) Systems in Buildings. Journal of Noise Control Engineering 45(6), (Nov – Dec 1997) 243-250