Guidelines for Airport Sound Insulation Programs (2013)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

52 This chapter provides guidance for meeting FAA acoustical requirements for SIPs, an over- view of the acoustical engineer’s role on the SIP team, and a general background to the acoustical concepts, metrics, and nomenclature used for airport noise, which is important to understanding how noise is perceived, measured, and depicted. Noise is defined as unwanted sound and is therefore a relative term that depends on the disposition of the individual lis- tener as well as the sound source. The FAA employs the DNL noise exposure metric, established by the Environmental Protec- tion Agency (EPA) in 1974, for assessing all community noise impact in the United States.1 The FAA threshold of significance, based on extensive psychoacoustic research, is that degree of noise exposure at which the community becomes “highly annoyed.” This exposure has been estab- lished as a DNL of 65 dB.2 Thus the FAA has established an objective numerical noise exposure level based on subjective response research results. A lower local standard (e.g., DNL 60 dB) may be used for Part 150 purposes if the standard is formally adopted by the local jurisdiction for land-use compatibility and the airport sponsor has incorporated it (although the interior noise level standard of 45 dB does not change). Where a lower local noise standard is adopted outside of the Part 150 process, 49 USC 47141 requires that the land use compatibility plan be developed cooperatively by the airport sponsor and local jurisdiction to be eligible for a grant. Addi- tional information on these requirements is addressed in Paragraph 810b. Noise Exposure Maps used for Noise Insulation Programs must be current.3 4.1 Introduction and Acoustic Fundamentals Airborne sound is a rapid fluctuation of air pressure and local air velocity. Sound has proper- ties of both fluids and waves. It propagates outward from its source at high speed, bends around interposing structures, is partially reflected and partially absorbed by incident surfaces, and radi- ates through structures, which attenuate (i.e., reduce) the transmitted sound. Improved interior NLR (the difference in sound level from exterior to interior) is the objective of SIPs, and this NLR is achieved by retrofitting structures with building elements having higher sound transmission loss (TL) properties. C H A P T E R 4 Acoustical Engineering 1 The U.S. EPA Office of Noise Abatement and Control, Information on Levels of Environmental Noise Requisite to Protect Public Health and Welfare with an Adequate Margin of Safety, EPA/ONAC 550/9-74-004, March 1974. 2 The Federal Interagency Committee on Urban Noise issued its report entitled Guidelines for Considering Noise in Land Use Planning and Control in June 1980. This report established the federal government’s DNL 65 dB standard and related guidelines. 3 U.S. DOT, FAA, PGL 12-09, August 17, 2012, Attachment 1, §812 (b)(2), p. 1-2.

Acoustical Engineering 53 Three aspects of noise are important in determining subjective response: 1. Level (i.e., magnitude or loudness) of the sound. 2. The frequency composition or spectrum of the sound. 3. The variation in sound level with time. 4.1.1 Sound Perception and Combination of Sound Levels Sound is perceived in a very complex fashion by the human ear as we detect and assimilate sound level, sound frequency, and sound variation over time. Sound levels are measured and expressed in decibels, with 0 dB roughly equal to that level at the threshold of hearing. Sound is a measure of the pressure fluctuations per second, measured in units of hertz (Hz). Most sounds do not consist of a single frequency but are composed of a broad band of frequencies differing in level. The characterization of sound level magnitude with respect to frequency is the sound spectrum. Figure 4.1 depicts the audible range of sound spectra for various types of sounds. Changes in sound level and combinations of sound levels are nonlinear and do not behave as most other physical phenomena. Because the level and frequency of sound are perceived in a nonlinear way, the decibel scale is used to describe sound levels; the frequency scale is also mea- sured in logarithmic increments. Decibels, measuring sound energy, combine logarithmically. Courtesy of Charles M. Salter Associates, Inc. Figure 4.1. Range of sound spectra.

54 Guidelines for Airport Sound Insulation Programs A doubling of sound energy (for instance, from two identical automobiles passing simultane- ously) creates a 3-dB increase; the resultant sound level is the sound level from a single passing automobile plus 3 dB. It would take 10 identical cars passing by simultaneously to be judged as twice as loud as the single car pass-by, though this would be a tenfold, or 10-dB, increase in sound level. The rules and examples for decibel addition used in community noise prediction are given in Table 4.1. 4.1.2 Subjective Response to Noise The effects of noise on people can be classified into three general categories: 1. Interference with activities such as speech, sleep, and learning. 2. Physiological effects such as anxiety or hearing loss. 3. Subjective effects of annoyance, nuisance, and dissatisfaction. No universal measure for the subjective effects of noise has been developed, nor does a mea- sure exist for human reactions from noise annoyance. This is primarily due to the wide variation of individual attitude regarding noise sources. For aircraft noise, typical reactions vary from annoyance to anxiety to fear. An important factor in assessing a person’s subjective reaction is to compare the new noise environment to the prior noise environment. In general, the more a new noise exceeds the prior noise, the less acceptable it is. Therefore, a new noise source will be judged more annoying in a quiet area than it would be in a noisier location. There are two types of noise impact: 1. Absolute impacts, whereby noise level or noise exposure exceeds a specified numerical stan- dard, and 2. Relative impacts, whereby noise level or noise exposure increases by a specified value. Changes in the noise environment cause a relative impact; the magnitude of a noise environ- ment causes an absolute impact. Most people acclimate somewhat to their noise environment. 4.1.3 Frequency Weighting Many rating methods exist to analyze sounds of different spectra. The simplest method, A-weighting, is generally used so that measurements can be made and noise impacts readily assessed using basic acoustical instrumentation. This method evaluates audible frequencies by using a single weighting filter that progressively de-emphasizes frequency components below 1000 Hz and above 5000 Hz. This frequency bias reflects the relative decreased human sensitivity to low frequencies and to extreme high frequencies. A-weighting is applied by an electrical filter in all U.S. and international standard sound level meters. Figure 4.2 shows the A-weighted network. Rule Example Difference in Two Sound Levels Sum of Sound Levels Level A Level B Level A + B 0–1 dB Highest + 3 dB 86 dB 87 dB 90 dB 2–4 dB Highest + 2 dB 84 dB 87 dB 89 dB 5–9 dB Highest + 1 dB 80 dB 87 dB 88 dB >9 dB Highest 77 dB 87 dB 87 dB Table 4.1. Decibel addition used in community noise prediction.

Acoustical Engineering 55 4.1.4 Noise Exposure Noise exposure refers to a measure of noise over a period of time, whereas noise level is a value at an instant in time. Although a single sound level may adequately describe the noise at any instant in time, airport and other community noise levels vary continuously. Most community noise is produced by many noise sources, which create a relatively steady background noise that has no identifiable source. These sources change gradually throughout the day and include traf- fic, wind through foliage, and distant industrial activities. Superimposed on this slowly varying background is a succession of identifiable noise events of brief duration. These include nearby activities, such as single vehicle pass-bys or aircraft flyovers, which cause the community noise level to vary from instant to instant. This fluctuating series of noise levels combines to form the noise exposure profile of a community. For purposes of quantifying noise that varies over a period of time, a standard term, equivalent sound level, has been adopted in the United States and internationally.4,5 Equivalent sound level is a single number and is typically referenced by the symbol Leq. Equivalent sound level is an energy average that takes varying sound levels of a time period and describes them as one constant noise level (i.e., the total sound energy divided by the duration). It is a construction of that constant sound level containing the same acoustic energy as the varying sound level during the same time period. Discrete, short-duration transient noise events, such as aircraft flyovers, may be described by their maximum A-weighted noise level or by their sound exposure level (SEL).6 The SEL value is preferred over maximum noise levels in defining individual events because measured results Courtesy of Charles M. Salter Associates, Inc. Figure 4.2. A-weighted network. 4 American National Standards Institute, ANSI S 1.8, American National Standard Reference Quantities for Acoustical Levels. 5 International Organization for Standardization, ISO 1996-1:2003, Acoustics – Description, measurement and assessment of environmental noise – Part 1: Basic quantities and assessment procedures. 6 See note 1.

56 Guidelines for Airport Sound Insulation Programs may be more reliably repeated and because the duration of the transient event is incorporated into the measure (thereby better relating to subjective response). Maximum levels of transient events vary with instantaneous propagation, measurement system time constant, and receiver conditions, while a total energy measure, like SEL, is more stable. The SEL of a transient event is a measure of the acoustic energy normalized to a constant duration of 1 second. The SEL dif- fers from the Leq in that SEL is the constant sound level containing the same acoustic energy as a 1-second event, whereas the Leq is the constant sound level containing the same acoustic energy over the entire measurement period. The SEL may be considered identical to the California standard single event noise exposure level (SENEL).7 Figure 4.3 depicts how SEL is computed. SEL values may be summed on an energy basis to compute Leq values over any period of time. This is useful for modeling noise in areas exposed to numerous transient noise events, such as communities around airports. Hourly Leq values are called hourly noise levels (HNLs). In determining the daily measure of community noise, it is important to account for the difference in human response to daytime and nighttime noise. During the night, people are more often at home and exterior background noise levels are generally lower than during the day, which causes exterior noise intrusions to become more noticeable. For these reasons, most people are more sensitive to noise at night than during the day. To account for human sensitivity to nighttime noise, the DNL (symbol Ldn) 8 descriptor is a U.S. and international standard adopted by the EPA in 1974 that describes community noise exposure from all sources. The DNL represents the 24-hour, A-weighted equivalent sound level with a 10-dB penalty added for nighttime noise between 10:00 p.m. and 7:00 a.m. The FAA has officially used DNL as its standard since 1981. In California, the CNEL9 has been the adopted standard since 1972. DNL and CNEL are typi- cally computed by energy summation of HNL values, with the proper adjustment applied for the period of evening or night. CNEL is computed identically to DNL but with a tripling of the Courtesy of Charles M. Salter Associates, Inc. Figure 4.3. Sound exposure level. 7 California Code of Regulations, 21 CCR Title 21. 8 See note 1. 9 See note 5.

Acoustical Engineering 57 evening (i.e., 7:00 p.m. to 10:00 p.m.) noise.10 The CNEL value is typically less than 1 dB above the DNL value. Since DNL and CNEL are so similar, all regulations are applied the same to both; only DNL will be referenced throughout these guidelines, with the understanding that CNEL is accepted identically in California. Noise exposure measures such as Leq, SEL, HNL, DNL, and CNEL are all A-weighted with units expressed in decibels. Figure 4.4 depicts the adjustments made to DNL and CNEL for the various periods of the day. 4.2 Basics of FAA-Sponsored Sound Insulation Understanding the acoustical impact environment created by transportation and applying the principles of noise reduction to affected buildings are the work of sound insulation programs. In order to achieve conformity in efforts across the country in varying climates and construction typologies, that work has to meet a consistent design standard. 4.2.1 FAA Acoustical Design Objectives There are two design objectives for eligible homes in sound insulation programs: 1. A target post-retrofit interior DNL of 45 dB in all habitable rooms. 2. A minimum NLR improvement of 5 dB. A. Basis for Design Objectives: Subjective Response The reason for the 5-dB NLR improvement criterion is to ensure a noticeable improvement in interior noise. A series of perception tests11 revealed that changes in noise exposure were per- ceived as follows: 1. Except under special conditions, a change in sound level of 1 dB cannot be perceived. 2. Outside of the laboratory, a 3-dB change is considered to be a just-noticeable difference. Courtesy of Charles M. Salter Associates, Inc. Figure 4.4. Hourly noise levels and annual metrics. 10 5 dB is often shown for simplicity, but the accurate formula is 10*log 3 = 4.8 dB. 11 See S. S. Stevens, “On the Psychophysical Law,” Psychological Review, 64: 153–181; S. S. Stevens, “Perceived Level of Noise by Mark VII and Decibels (E),” Journal of the Acoustical Society of America, 51: 575–601; E. Zwicker and B. Scharf, “A Model of Loudness Summation,” Psychological Review, 72: 3–26.

58 Guidelines for Airport Sound Insulation Programs 3. An increase or decrease in level of at least 5 dB is required before any noticeable change in response would be expected. 4. A 10-dB increase is subjectively heard as an approximate doubling in loudness. B. Perceived Airport Noise Exposure, More Complex However, this perception summary represents special test conditions. These perception tests were made by playing certain fixed broadband sounds to juries of listeners and then immedi- ately changing those levels and noting the juries’ average responses to sound level increases and decreases. The changes in perceived airport noise exposure, however, are considerably more complex for two reasons: 1. The spectra of individual aircraft flyovers change substantially between individual events, so the comparison in loudness is complicated by comparing dissimilar sounds. Issues of annoy- ance, nuisance, dissatisfaction, speech interference, sleep interference, learning impairment, anxiety, and hearing loss all affect subjective response to changes in noise level. 2. More significantly, noise exposure is an integrated measure of sounds over a period of time, whereas sound level is simpler and immediate. Subjective comparison of any sensory values is greatly affected by latency, or the time between events. So comparisons of long-term noise exposure measures, such as DNL, will yield much more varied responses than will the imme- diate changes in noise level used for this perception test. The FAA, in establishing performance criteria for SIPs, ensured that sound insulation pro- gram treatments would provide an audible improvement in affected buildings. 4.2.2 The Building Envelope Taking the design objectives into the field and applying them to structures needs further artic- ulation. While a building may have a consistent construction, it is possible for each room to pro- vide a different existing NLR depending on (1) the noise reduction properties of the individual façade building materials exposed to incident aircraft noise, and (2) the area of each individual building material exposed (i.e., ratio of wall to window). This is important in developing and implementing NLR design criteria for SIPs. Design approaches include: 1. Individually design each room to achieve a ≥5-dB ∆NLR and a ≤45-dB DNL. 2. Apply a uniform noise reduction treatment standard to create a homogeneous building envelope using consistent treatments. 3. Provide a hybrid approach where consistent treatments are applied across the building envelope but rooms that need additional treatments to achieve the ≤45-dB DNL will receive additional design attention. PGL 12-09 is not specific as to which of these design objectives meet FAA noise reduction goals. Program sponsors and consultants are advised to consult with their local ADO for fur- ther clarification regarding this issue. Depending on a building’s location in the contour and its existing noise reduction capabili- ties, treatments may need to achieve greater than 5 dB of reduction to meet the interior 45-dB DNL criteria. In the first approach, each room could have a different existing NLR depending on the ratio and composition of building materials. Achieving a uniform minimum 5-dB treatment could require different treatments for each room to achieve the required NLR. The second approach allows for use of uniform building materials and construction proce- dures throughout the program. This provides considerable cost savings in both material and labor.

Acoustical Engineering 59 However, each room receives a slightly different NLR improvement, but each room receives a similar interior noise environment after retrofit. In the third approach, the exterior envelope of the whole building is reviewed for consistency of construction and building elements, and then individual rooms are verified for specific per- formance issues. This allows for use of uniform building materials and construction procedures for the majority of the treatments and acknowledges that more retrofit may be needed in limited cases to provide a continuous STC 40 building envelope. The majority of residential SIPs employ some form of the second or third approaches for treatment design; few programs attempt to achieve specific NLR performance for each room. Consequently, when applying a uniform sound transmission class (STC) performance enve- lope across a program, older homes and those having poorer pre-retrofit NLR performance will realize a greater NLR improvement (on average 7 dB to 8 dB) than newer and well-maintained homes, which may only realize a 4-dB to 5-dB improvement from the same treatments. In addition to the NLR properties of basic building elements, another significant noise path is the presence of acoustical leaks, termed flanking paths. These are typically cracks or poor seals where air and sound may infiltrate. Flanking may significantly degrade sound insulation perfor- mance and requires treatment in every instance. Sound has the property of always infiltrating the weakest spot. It is not feasible to apply excessive acoustical treatment in one location while allowing for flanking in another; therefore, attention will need to be paid to the building enve- lope beyond the major fenestration openings. 4.2.3 Achieving a Noise Level Reduction of at Least 5 dB Prior to sound insulation treatments, a structure will typically provide various degrees of noise reduction in various rooms. Corner rooms with three exposed façades, containing large non-sound–rated windows that make up a large percentage of the exposed exterior, will be much noisier inside than a contained first-floor room with a single exposed façade. The solid wall in the basic façade structure, typically stucco or wood siding, has much better sound attenuation properties than non-sound–rated windows. Certain uninsulated façades and lightweight façades (such as those of aluminum and lightweight vinyl) fall short of the standard sound transmission loss level; in these cases the windows provide more noise reduction than the façade. Heavy brick and stucco façades, on the other hand, typically provide more than the standard sound trans- mission loss level; in these cases the windows provide less noise reduction than does the façade. Given differences in room exposure, there has been considerable discussion as to how the minimum 5-dB NLR improvement criterion is to be applied in sound insulation programs. Alternative interpretations include: 1. Every room treated must achieve a minimum 5-dB NLR improvement. 2. The average NLR improvement for all tested rooms in a single dwelling must be at least 5 dB. 3. The average NLR for all dwellings in a single project or a single program must be at least 5 dB. PGL 12-09 is not specific as to which of these three treatment outcomes is consistent with FAA noise reduction goals; however, it does state, “The measurement of interior noise levels is an average for all habitable spaces in a particular residential unit.”12 This is consistent with the consensus of programs and the FAA offices that the second interpretation is the prevailing objec- tive for programs to meet. Sponsors and consultants should consult with their local ADO for further clarification regarding this issue. 12 U.S. DOT, FAA, PGL 12-09, August 17, 2012, Attachment 1, §812 (b)(1), Table 1 p. 1-3.

60 Guidelines for Airport Sound Insulation Programs Designing for a 5-dB NLR improvement in a room requires the application of technical calcu- lations containing information on the room area, the amount of wall openings, and the projected performance of the new treatments. Some examples of these calculations can be found in the 2005 version of the SIP guidelines.13 Rooms with varying amounts of wall openings will neces- sarily achieve different NLR improvements from a standardized program approach to window and door replacement. Therefore, the acoustical design criterion used in the majority of pro- grams is to achieve a uniform level of noise reduction for the entire building envelope. Thus the non-sound–rated windows are replaced with sound-rated windows whose noise level reduction properties are similar to that of the basic façade. It would be possible to achieve a full 5-dB NLR improvement for a contained first-floor room with a small window, but this might require retrofit treatment of the façade wall as well as the small window to balance the small ratio of window to wall. Consequently, this would render the small room much quieter than other rooms treated throughout the structure. In this case, meeting the ≤45-dB DNL interior may be sufficient. Most programs test the pre- and post-construction in approximately three rooms in any house. It is reasonable to focus on the major rooms of the house with normal to significant exterior exposure and let the consistent envelope treatment apply to rooms with smaller amounts of noise exposure. This allows for a balanced and uniform noise reduction envelope throughout the structure, using standard and similar building materi- als and elements throughout the program. This achieves the minimum 5-dB NLR improvement objective for nearly all rooms. In light of PGL 12-09, which requires an average of all treated space, programs will need to make policy decisions about what expense they are willing to incur to go above uniform treatments. 4.2.4 Interior Eligibility Standard PGL 12-09 has placed renewed emphasis on interior noise levels as a per-building qualifi- cation threshold for sound insulation treatment eligibility. This requires an assessment of the existing interior DNL for each residence or structure prior to its acceptance into or rejection from the SIP. This assessment will be conducted by either (1) composite transmission loss (CTL) computation of the existing NLR performance of the residence, or (2) measurement of the actual NLR performance. PGL 12-09 states: In 1992, FAA adopted guidance on testing frequency, sampling and other statistical measures that can be applied to a neighborhood to estimate the interior noise levels in the residences that are in the 65 dB contour. This information is compiled into the Acoustical Testing Plan. Long standing agency policy is that an airport sponsor must use the 1992 guidance to establish the existing interior noise levels to deter- mine whether or not the building qualifies for sound insulation using AIP.14 The FAA requires any testing beyond 30% to be justified to the FAA Planning and Environ- mental Division – Airport Planning and Programming (APP-400).15 Some programs currently conduct 100% testing for community confidence when excluding properties from treatment. PGL 12-09 clarifies that an average interior noise level of DNL 45 dB in all habitable rooms is an eligibility criterion for determining whether a structure qualifies for sound insulation. Two potential issues regarding implementation of this eligibility criterion are: 1. The testing methodology needs to accurately measure interior noise levels while taking into account the margin of error that is inherent in the various testing methodologies. The PGL 13 Department of the Navy, Naval Facilities Engineering Command, Guidelines for the Sound Insulation of Residences Exposed to Aircraft Operations, April 2005. 14 U.S. DOT, FAA, PGL 12-09, August 17, 2012, Attachment 1, §812 (c)(2), Table 2, p. 1-6. 15 U.S. DOT, FAA, PGL 12-09, August 17, 2012, Attachment 1, §812 (c)(2), Table 2, p. 1-7.

Acoustical Engineering 61 does not address factoring in a margin of error. There may be homeowners who will be deemed ineligible by testing that has not accounted for a margin of error. 2. The averaging method will inevitably lead to situations where homes will be ineligible for treatment. If interior noise levels of the majority of rooms are above DNL 45 dB but one or more habitable rooms are unusually low, the overall average could result in the structure not being eligible for treatment. Program sponsors and consultants should refer to their local ADO for further information regarding these issues. 4.2.5 Acoustical Testing Plan Table 2 of PGL 12-09 specifies criteria that a sponsor must meet in order to be eligible for AIP grant funding. One of the requirements is that the sponsor prepare an acoustical testing plan to determine existing interior noise levels using FAA-adopted guidance as found in Guidelines for Sound Insulation of Residences Exposed to Aircraft Operations, October 1992. Table 2 of these guidelines describes a testing and treatment process that starts with the following steps in an initial testing phase: 1. Create an inventory of housing types and locations based on a windshield survey. 2. Develop packages of acoustical treatments specific to each housing type. 3. Test representative samples of each housing type.* 4. Develop acoustical treatments for each housing type that will meet FAA noise reduction goals. 5. Install treatments (although not specifically stated). 6. Conduct post-construction testing to determine if noise reduction goals have been met. 7. Submit a report (the acoustical testing plan) to the ADO describing pre- and post-construction test results along with (if required) any changes to sound insulation treatments for each housing type. 8. After approval of the acoustical testing plan by the ADO, begin the full sound insulation program using the acoustical treatment packages that have been tested and subsequently approved for each housing type. *Although not specifically stated in the process described by the PGL in item #3, it is assumed that if testing indicates that the average interior noise levels for all habitable rooms of a given housing type is less than a DNL of 45 dB, then this housing type will be deemed ineligible for AIP-funded sound insulation. As of the date of this writing, there has been some discussion regarding this process. Some sponsors and consultants suggest that this process may not accurately determine interior noise levels in homes or provide a defensible basis for determining a home’s eligibility or ineligibility for treatment. There are two main reasons for this concern: 1. As noted by the 1992 guidelines: Two houses may be very much alike and yet each will have unique features which require special treatment. While it is useful to discuss, in general terms, typical dwelling categories and classes of modi- fications, the actual site-specific design requires the services of an acoustics consultant or an acoustics- knowledgeable architect.16 2. Creating categories of homes in order to predict acoustical performance based solely on a windshield survey and exterior building characteristics does not account for factors such as exterior envelope air infiltration, homeowner modifications (e.g., pet doors, mail slots), and interior finishes that have a direct impact on interior noise levels. 16 U.S. DOT, FAA, Report No. DOT/FAA/PP-92-5, Guidelines for the Sound Insulation of Residences Exposed to Aircraft Opera- tions, October 1992, §3.6.1, page 3-64.

62 Guidelines for Airport Sound Insulation Programs Users of these updated guidelines are advised that further clarification from the FAA may occur. In response to these issues, a testing strategy is being formulated that uses a combination of typology sample testing with computer modeling. The core concept of this approach is to perform acoustical modeling of all potential program homes to determine the existing outdoor- to-indoor noise reduction (NLR) characteristics of all habitable rooms. Details about existing conditions in the homes that are needed for the modeling are obtained from assessment visits conducted to determine existing conditions and potential primary and secondary treatments at each residence. Thus, homes would not be qualified based on typology or windshield surveys but would be qualified based on actual inspection and modeling of every home. Ongoing research at the Georgia Institute of Technology is developing refinements to acoustic modeling software such as INSUL and IBANA-Calc17 that will take into account the spectrum of conditions in existing residential structures that the current PGL testing process does not include. Using acoustic modeling software will provide a higher level of confidence in calcula- tions of interior noise levels that will ultimately provide for a range of benefits, including: 1. Greater confidence from the homeowners that the particulars of their individual homes are being accounted for in determining eligibility and ineligibility. 2. Data that are empirically derived and that identify primary and secondary treatments and assurance that treatments will meet FAA noise reduction criteria. 3. More cost-efficient use of on-site acoustical testing to calibrate and add further information to the acoustical modeling database rather than using acoustical testing to determine the interior noise levels of every habitable space in every potentially eligible home. Users of these updated guidelines are advised that the use of acoustical modeling as a comple- mentary process to the process described in the PGL is not currently in widespread practice. It should be noted, however, that the 1992 guidelines (the FAA-adopted guidance per the PGL) confirm the value and validity of this approach: The current noise reduction capability of the dwelling can be determined in one of two ways. Perform- ing field measurements using the methods described in Section 3.3 gives a reliable value for the noise reduction. It may, however, be impractical to take measurements in each dwelling included in the project. Proprietary computerized models are an alternative, and equally valid, tool. In most home sound insula- tion projects the field measurements are primarily used to provide input data for calibrating the model and to validate the model predictions. The field measurements and model predictions usually agree to within 2 to 3 dB. In general, the more conservative noise reduction value should be used in setting the insulation goals and designing the modification package.18 Program sponsors and consultants should consult with the ADO responsible for review and approval of their program’s acoustical testing plan. 4.3 The Science Behind Evaluating Sound Insulating Building Elements, Materials, and Systems 4.3.1 Sound Transmission Loss Concepts Sound TL of individual building elements depends on their mass, resiliency, and acoustical decoupling properties, the spectra of the aircraft producing the noise environment, and the angle of incidence of all aircraft noise impinging on each building element on the structure’s façade. 17 Nathan Firesheets, Modeling the Transmission Loss of Typical Home Constructions Exposed to Aircraft Noise (Master’s Thesis), Georgia Institute of Technology, Atlanta, GA, 2012. 18 U.S. DOT, FAA, Report No. DOT/FAA/PP-92-5, Guidelines for the Sound Insulation of Residences Exposed to Aircraft Opera- tions, October 1992, §3.4.1, p. 3-18.

Acoustical Engineering 63 The most fundamental principle of sound TL is the mass law, which gives the TL at each fre- quency as a function of surface weight. According to the mass law, TL increases linearly with an octave or one-third octave band. This law works well only for limp monolithic (i.e., no composite structure) materials, but also forms the basis for TL properties for all materials and structures. Specifically, all materials and systems exhibit a general trend of increasing TL performance with increasing frequency. That is, higher frequencies are attenuated more effectively than low fre- quencies in all structures. Resiliency (or its inverse, stiffness) is an important property for TL in composite materi- als. Sound does not pass through materials but, rather, impinges on a material and reradiates from the other side at some reduced level. Materials attenuate sound energy through consuming mechanical vibration energy and converting it to small (almost immeasurable) quantities of heat. Thus, the more resilient a composite material, the more energy will be consumed and the greater the sound attenuation. The third principle in sound attenuation is decoupling, which is a property of composite materials to structurally and acoustically isolate parallel elements of the composite structure. One example of acoustical decoupling is the dual-glazed windows used in SIPs. Here sound impinges on the exterior glazing panel, which must then reradiate the sound through a sub- stantial air space (typically more than an inch) and then through a second layer of glass. This acoustical transmission inefficiency is very effective in reducing sound transmission through the assembly. The glazing panels are in resilient zipper gaskets, which minimize structural coupling through the framing system.19 High TL is most efficiently achieved by double-wall construction, allowing for greater TL with lighter-weight assemblies. Best results are achieved when the parallel panels are mechanically and acoustically isolated. Mechanical isolation is achieved by independent support of the paral- lel panels (no structural coupling), and acoustical isolation is achieved by increased air space between the panels. The net TL of two isolated panels may be computed from the individual TL properties of each.20 Several prediction methods may be used to compute the TL properties of building elements and assemblies. These models incorporate the mass, stiffness, geometry, mechanical isolation, and acoustic isolation properties of the building assembly. However, these models do not often yield precise results because of the difficulty in measuring the various properties, particularly stiffness and mechanical isolation. For this reason, laboratory TL testing is required for acousti- cal materials and assemblies used in SIPs. 4.3.2 Sound Transmission Class Rating While the TL characteristics of building materials and assemblies may generally be computed with reasonable accuracy and reliability or tested in sample field installations, the best method of ensuring TL performance is acoustical testing in an accredited laboratory and according to American Standard for Testing and Materials (ASTM) standards. Laboratory accreditation is by the National Voluntary Laboratory Accreditation Program (NVLAP) under the oversight of the National Institute of Standards and Technology. Architectural product manufacturers are generally required to submit such laboratory test results for all major building elements in order to obtain approval for use in SIPs. 19 See note 13. p. 2-6. 20 B. H. Sharp, “Prediction Methods for the Sound Transmission of Building Elements,” Noise Control Engineering, 11, 5533, 1978.

64 Guidelines for Airport Sound Insulation Programs The sound TL properties of building elements are tested and reported according to national standards in one-third octave bands, classified as an STC rating.21 Each building element, such as a particular window, may be expected to have a unique TL signature represented by 16 TL values from 125 Hz to 4000 Hz. As mentioned in the previous section discussing mass law, the nature of sound attenuation through structures is such that all TL tests have generally up-sloping properties from low frequency to high frequency, indicating generally increasing noise reduction in higher frequencies. Figure 4.5 depicts a typical TL test result. 4.3.3 Transmission Loss Metrics STC is the oldest and most established rating for the TL properties of building elements and systems. STC is computed by using the standard three-straight-line-segment curve in Fig- ure 4.5 and computing the TL deficiencies (differences in measured TL and curve value) in each of the 24 one-third octave bands. The STC rating is determined as the highest value of the curve at 500 Hz for which the sum of deficiencies does not exceed 32 and no single defi- ciency exceeds 8. This procedure was developed for two purposes: to consider the subjective response of the human ear at various frequencies with the shape of the segmented curve and to account for the annoyance effects of panel resonance and coincidence dips. These later effects are most prevalent with lightweight structures, where specific frequencies are reinforced and cause annoying buzzing tones. However, with most building elements of STC 40 or greater, these effects become imperceptible. Coincidence dip is a drop in the TL of a material or assem- bly at a certain frequency caused by resonance effects. Figure reprinted, with permission, from ASTM E413-10, Classification for Rating Sound Insulation, copyright ASTM International, 100 Barr Harbor Drive, West Conshohocken, PA 19428. A copy of the complete standard may be obtained from ASTM International, www.astm.org. Figure 4.5. STC test result. 21 ASTM International, ASTM E413, Classification for Rating Sound Insulation.

Acoustical Engineering 65 Another TL standard was adopted by ASTM, the outdoor–indoor transmission class (OITC).22 The OITC method is simpler and more easily understood than STC. It was developed specifically to assess the TL properties of materials and systems subjected to transportation noise. Specifi- cally it (1) employs a reference sound spectrum comprised as the average from railroad, freeway, and aircraft noise sources; (2) subtracts the 18 one-third octave band TL values from 80 Hz to 4k Hz; and (3) A-weights the resulting sound spectrum to produce the OITC value. Some SIPs have shown interest in OITC, and some have accepted OITC tests as an option to STC test results. The sound transmission class metric has withstood the test of time and remains the most commonly used metric for rating the sound transmission loss characteristics of various build- ing materials and systems for SIPs. Many programs are reluctant to require vendors of proven building materials and systems to spend the time and money to retest their established products for the new standard. 4.4 The Acoustical Design Process The acoustical consultant plays a key role throughout the acoustical design process, extending from the project planning phase through conceptual and detail design, construction consulting, acoustical testing, and project performance reporting to the FAA. In the planning phase of the project, the consultant works with the project team in selecting building materials and acousti- cal treatment protocols for use in the program, and reviews acoustical test reports from product manufacturers to ensure compliance with the TL standards for the program, which is typically specified using STC performance. Often the consultant will confer with vendors and suppliers to ensure that there have been no changes in the test reports for the design or fabrication of materials or systems. After selection of program structures for treatment, the acoustical consultant performs pre- retrofit acoustical testing of representative structures in the program. Established or ongoing programs typically test 10% of program structures. Pilot programs or programs with widely vari- ant housing types may test a higher percentage, from 25% to 100%, depending on the amount of data needed to make decisions on treatments. In addition, any special structures, such as histori- cal homes, will likely require specialized acoustical treatment and would be appropriate to test. Based on test results, the consultant prepares an acoustical conceptual design, which identifies the performance needs for various treatment elements. Acoustical consultant input is particularly important when reviewing exceptions (referred to as waivers in some regions) to project acoustical standards, such as homeowner requests to have special building elements (e.g., decorative or stained glass windows) remain rather than be replaced. Incorporating this review may prevent exceptions that would result in nonconformance with the NLR and DNL standards of the program. Following the review, the project architect and engineers develop drawings for each structure to be sound insu- lated, list specifications for all building materials and systems, and outline project details for on-site construction and installation of building elements and systems. Customized treatments for each residential structure are developed based on a standard set of treatments as well as program policies and procedures. Institutional or public buildings will likely require fully customized treatments. Following are typical acoustical treatments for building elements and systems: • Replacement of windows with sound-rated windows. • Replacement of exterior doors into habitable/occupied spaces with sound-rated entry doors, or addition of sound-rated storm doors over new or existing doors. Also, new perimeter gas- kets and threshold systems are installed. 22 ASTM, ASTM E1332, Standard Classification of Outdoor-Indoor Transmission Class.

66 Guidelines for Airport Sound Insulation Programs • Addition of attic insulation. Often fiberglass insulation is added by blowing in, although roll out insulation works equally well. • Addition of vent baffles for outsized gable vents in the attic. Eave vents are typically not treated due to their small size. • Addition of flex ducting to bath fans and dampers atop kitchen fan exhaust stacks. • Addition of chimney-top dampers or glass doors on fireplaces. • Patching and sealing of extraneous protrusions through the façade such as mail slots, pet doors, through-wall air conditioning, and various homeowner modifications. • Addition of or modification to heating, ventilating, and air conditioning (HVAC) systems to ensure air quality and comfort. • Addition of secondary glazing to skylights. • Addition of ceiling or wall materials where needed. On completion of construction, the acoustical consultant performs post-acoustical testing on the same structures originally tested. Testing locations, procedures, and conditions are replicated to the maximum extent possible in order to determine NLR improvement as accurately and reli- ably as possible. The final step in the acoustical design process is preparation of sections of the project’s final report that are specific to acoustical treatments. Topics and issues typically covered are structures treated, design criteria, treatments effected, special structures, the pre- and post-retrofit DNL and NLR, and assessment of compliance with the FAA acoustical objectives for the program. These reports are typically submitted to the airport sponsor, which submits the full report to the local FAA ADO. See Chapter 12 for more information about project reporting and closeout. 4.5 Noise Effects The FAA has accepted the DNL metric exclusively to assess and evaluate airport noise expo- sure in SIPs. This metric is also the EPA standard and is used almost exclusively in all community noise assessments throughout the United States (though the very similar CNEL is used through- out California). The DNL metric is typically accepted as the best single metric for describing airport noise impacts that may be interfering with speech and sleep and creating adverse learn- ing effects on school children. Certainly, a reduction in interior DNL provides a degree of relief from these impacts. 4.5.1 Improved Speech Communication The greatest single result from sound insulation is improved speech communication. As mentioned, the greatest noise reduction from sound insulation occurs in the higher frequencies where speech occurs, from 500 Hz to 2,000 Hz. Typical sound insulation results will be from 5 dB to 8 dB in terms of A-weighting, though the improvement in the higher speech frequencies is typically 10 dB or more. Homeowners surveyed after noise retrofit typically cite improved speech communication. They no longer need to interrupt dinner and phone conversations or turn up the TV volume during flyovers. However, there is less improvement in the low-frequency roar, typically about 2 dB, below 500 Hz. But this roar only minimally affects speech communi- cation because it is outside of the speech frequencies and does not mask, or drown out, speech. 4.5.2 Supplemental Noise Metrics DNL is by far the most widely accepted noise metric and is employed as the standard for all environmental impact assessments throughout the United States for transportation projects.

Acoustical Engineering 67 Other metrics for noise-related impacts such as speech interference and sleep interference may be useful to help evaluate specific use areas such as hospitals, schools, and places of worship. While the interior DNL ≤ 45 dB and ∆NLR ≥ 5 dB metrics are the measurement standards for SIP performance, the use of other metrics is only allowed by the FAA with specific approval in concert with the regional ADO. In part FAA AC 150/5020-1, Noise Control and Compatibility Planning for Airports, states, “it is recommended that additional analysis via single event maxi- mum sound level and/or sound pressure level versus frequency data be used to determine the necessity (and/or eligibility) for soundproofing.”23 Various noise metrics are available to assist in assessing various noise effects. For speech the most common are speech interference level24 (SIL, the average sound level in the 500-, 1,000-, and 2,000-Hz octave bands) and speech intelligibility index.25 Other metrics have been devel- oped for speech intelligibility, and some are supported by national and international standards, but many of these are overly complex, while others are applied to particular conditions such as assessment of audio systems or HVAC background noise. SIL is used most commonly for its simplicity, and often a simple A-weighted maximum level of 60 dB is used as a design criterion for speech communication in an aircraft noise environment. This is particularly expedient in designing treatments for educational environments where speech intelligibility is critical. As of 2002, the American National Standards Institute (ANSI) promulgated a national stan- dard for the noise environment in school classrooms: Acoustical Performance Criteria, Design Requirements, and Guidelines for Schools. This standard addresses background noise, room acoustics (the liveliness and acoustical reflectivity of the interior space), and sound insulation in terms of STC. The sound insulation standard depends in part on whether the interior noise from exterior sources exceeds 40 dB for more than 10% of the time during any hour of class. This assessment requires modeling or measurement of the aircraft noise environment. Specifically, the Integrated Noise Model (INM) computer program used to develop the DNL noise contour for the Part 150 study must be rerun using a time-above analysis. Alternatively, continuous mea- surements may be made in the classroom during the noisiest hour on a typical day. Sleep interference is another important issue often addressed for aircraft noise environments. A national standard, ANSI S12.9-2008, Quantities and Procedures for Description and Measure- ment of Environmental Sound—Part 6: Methods for Estimation of Awakenings Associated with Outdoor Noise Events Heard in Homes,26 was developed to assess awakenings and has been recommended by the Federal Interagency Committee on Aviation Noise (FICAN). This was developed from assimilating various sleep interference tests and considers the SEL values of individual events and the volume (quantity) of events. There has been considerable discussion about awakening evaluation because it is evident that many people in an aircraft noise-affected community significantly acclimate to the noise environment. Thus, those new to an aircraft noise environment are more likely awakened than those who have resided in the area for a substantial period. This is still a new area of study in regard to sound insulation. One program at an airport that serves as a hub for an express package service with nighttime operations allowed treatments in bedrooms that provided greater NLR to create a better sleeping environment. This was not part of a Part 150 program but rather part of a mitigation agreement based on an EIS for the addition of new nighttime flights. 23 U.S. Department of Transportation, FAA, AC 150/5020-1, Noise Control and Compatibility Planning for Airports, August 5, 1983. 24 David A. Bies and Colin H. Hansen, Engineering Noise Control, Theory and Practice, 3rd ed., Spoon Press, 2003, pp. 150–151. 25 ANSI, ANSI S 3.5, American National Standard Methods for Calculation of the Speech Intelligibility Index. 26 ANSI, ANSI S 12.9, Quantities and Procedures for Description and Measurement of Environmental Sound – Part 6: Meth- ods for Estimation of Awakenings Associated with Aircraft Noise Events Heard in Homes.

68 Guidelines for Airport Sound Insulation Programs 4.6 Acoustical Testing The program objectives are to meet the interior DNL ≤ 45 dB and ∆NLR ≥ 5 dB on the basis of the design year NEM (which, per PGL 05-04, must be currently valid). PGL 12-09 goes on to clarify this requirement: In general, NEMs less than 5 years old are considered current unless conditions have created a sig- nificant change that would affect noise contours. NEMs older than 5 years old must be certified by the sponsor and updated as required as discussed in the PGL.27 Therefore, the aircraft noise environment used for acoustical design and for acoustical testing of NLR should consider the aircraft fleet mix, flight tracks, and so forth for all flight operations from the NEM. An ideal acoustical test program for pre- and post-retrofit would accomplish the following for each structure tested: • Consider only aircraft noise, ignoring all other noise sources. • Integrate the sound spectra from all aircraft used in the NEM. • Integrate the flight tracks by aircraft type from all aircraft used in the NEM. • Integrate the sound incidence angles of all aircraft on all flight tracks, by aircraft type, from all aircraft used in the NEM. • Integrate the change in meteorological conditions with all other parameters used in the NEM. The only way all of these goals can be accomplished would be through continuous attended noise monitoring and simultaneous monitoring inside and outside each residence for the NEM scenario, while manually deleting non-aircraft acoustic events. Since this is infeasible, testing methods that allow for data generation that can be further analyzed by computer models are a benefit for an acoustical testing program. Two methods have been most commonly used in SIPs for field testing the TL of rooms within a structure: the aircraft flyover test and the artificial noise source test. A third method has also been employed: the indoor-outdoor speaker test. Each method has technical and logistical advantages and disadvantages. It should be noted that PGL 12-09 states, “Long standing agency policy is that an airport sponsor must use the 1992 guidance to establish the existing interior noise levels to determine whether or not the building qualifies for sound insulation using AIP.” Since the 1992 guidelines mention only overflight as a testing methodology, SIP sponsors and consultants have requested that the FAA clarify if the PGL is intended to preclude testing using artificial noise. Sponsors are advised to consult with their local ADOs for direction on allowable means and methods of testing. The most valid and precise acoustical testing for sound TL is laboratory testing of building systems and materials;28 field testing procedures29 are designed to parallel lab procedures as much as is practicable. Laboratory testing is performed by inserting the window, door, or other build- ing element into an opening between two large rooms in the testing facility, then generating a dif- fuse sound field on the source room side, recording the spatial average diffuse sound level in each room, and subtracting the receiver room noise level from the source room noise level to obtain the NLR. A small correction is then applied to account for the noise buildup and absorption effects of the receiving room. The diffuse sound field has sound waves traveling in all directions with equal probability. This is necessary because the incidence angle at which sound impinges on a material affects its TL properties. However, it is clearly impossible to field test existing buildings 27 See note 3. 28 ASTM, ASTM E90, Test Method for Laboratory Measurement of Airborne-Sound Transmission Loss of Partitions. 29 ASTM, ASTM E966, Guide for Field Measurement of Airborne Sound Insulation of Building Façades and Façade Elements.

Acoustical Engineering 69 using this laboratory-based method, even though an interior room approximates one room with a diffuse sound field. The sound field outside the structure constitutes both a free field, not influ- enced by local sound reflection, and a far field, not close to or influenced by the noise source size. PGL 12-09 specifies several testing procedures and protocols for AIP-funded programs. • Interior noise testing is to be conducted with windows and doors closed. This protocol applies without regard to the presence of ventilation systems.30 • The measurement of interior noise levels is an average for all habitable spaces in a particular residential unit.31 • FAA-accepted guidance on testing frequency, sampling, and other statistical measures is con- tained in the Guidelines for Sound Insulation of Residences Exposed to Aircraft Operations, pre- pared for the Department of the Navy by Wyle Laboratories in 1992.32 • The ADO must approve or disapprove a sponsor request for reimbursement for testing more than 10%, but not more than 30%, of the residences of a particular construction type.33 • For requests for reimbursement for more than 30% of the residences of a particular type, the ADO must receive APP-400 approval.34 • Occasionally residents may request that their residence be tested specifically. This may be because of the condition of the home or because the resident believes that the residence will test differently than others. These additional tests are generally allowable. However, if an addi- tional residence is tested, it must be tested both before and after any noise insulation work to ensure that the 5-dB NLR is achieved.35 4.6.1 Aircraft Flyover Method The aircraft flyover test is used throughout sound insulation programs. This method simulta- neously measures the exterior free-field incident sound of flyovers and the diffuse sound field in the test room inside the structure. The difference in the two A-weighted sound exposure levels (SEL values) is subtracted to yield the NLR of the room. In practice, synchronized digital pro- grammable sound level meters are positioned in the free field outside the home and in the room. They simultaneously record multiple events, allowing for computation of the NLR for each event and statistics for a series of flyover events. Typically, two interior rooms are measured simulta- neously. These measurements generally follow a national standard for field TL measurement.36 The flyover method is assumed to provide a reasonable approximation of the TL in each room, but does have the following limitations and sources of error: • The sound spectra of the aircraft flyover samples are assumed to be the average for all air- craft used in the NEM. This is unlikely since measurement on a single day will likely record aircraft during a single operation type (i.e., all landings or all takeoffs) and not reflect that of the annual mix. This method is best used at an active airport with a high number of daily operations. • Extraneous noise sources for occupant or neighborhood activity are recorded and may be included with the aircraft noise measurements. These sources include occupants, local vehi- cles, construction, recreation, and other neighborhood activities. 30 U.S. DOT, FAA, PGL 12-09, August 17, 2012, Attachment 1, §812 (c)(1), Table 1, p. 1-4. 31 U.S. DOT, FAA, PGL 12-09, August 17, 2012, Attachment 1, §812 (c)(1), Table 1, p. 1-3. 32 This document may be found on the FAA Airport Noise website at http://www.faa.gov/airports/resources/advisory_circulars/ index.cfm/go/document.information/documentNumber/150_5000-9A. 33 See note 15. 34 See note 15. 35 U.S. DOT, FAA, PGL 12-09, August 17, 2012, Attachment 1, §812 (c)(2), Table 2, p. 1-8. 36 See note 29.

70 Guidelines for Airport Sound Insulation Programs • A single stationary microphone in a room does not give a good measure of the diffuse sound field. Laboratory tests use large rooms of special dimensions, often with moving microphones or vanes, and nonparallel walls to minimize standing wave effects. Microphones in small rooms are significantly influenced by location, particularly with pre-retrofit testing where the location relative to a poor sound-attenuating window may have considerable effect. It is dif- ficult to obtain a satisfactory reverberant field in a small room. • The total NLR is primarily from the composite transmission loss characteristics of the structure but also from the room acoustics controlled by the size and absorptive proper- ties of the receiving room. Therefore, if there is a considerable change in room furnishing between the pre- and post-retrofit testing, a significant change in room absorption will affect the measured NLR. • Different aircraft under different operating conditions are recorded for the pre-construction and post-construction acoustical measurements. These different operating conditions result in different spectra and incidence, therefore posing another source of error. 4.6.2 Artificial Noise Source Method The testing method using a generated noise source is similar to the flyover test method and is also used throughout sound insulation programs. As with the flyover method, this method simultaneously measures the exterior free-field incident sound of flyovers and the diffuse sound field in the test room inside the structure. The difference in the two sound exposure levels (SEL values) is subtracted to yield the NLR of the room. This method uses an artificial aircraft sound source (loudspeaker), usually mounted on a tele- scoped boom from a truck and elevated to a position to approximate the incident angle of aircraft. However, the physical limitations for employing a boom truck may require that the loudspeaker simply be mounted on a tripod set at ground level. Free-field exterior and diffuse interior measure- ments are made by consultants who often record the noise to allow for spectral measurement. These measurements may also identify flanking paths and may use near-field measurements to identify the approximate TL of individual building elements. The interior microphone is circulated through- out the room by a consultant during the measurement to optimize reverberant field measurement. Typically, two interior rooms are measured sequentially. These measurements generally follow a national standard for field TL measurement.37 The artificial noise source method is assumed to provide a reasonable approximation of the TL in each room but does have the following limitations and sources of error: • The sound spectrum of the artificial noise source is assumed to be the annual energy average for all aircraft used in the NEM. This is unlikely since measurement on a single day will likely not reflect that of the mix of airplanes recorded in the NEM. • The free-field coverage of the loudspeaker over the structure is assumed to approximate that for all aircraft used in the NEM. In practice, loudspeaker coverage is considerably inferior to that from aircraft. • The sound incidence of the loudspeaker noise is assumed to be the average for all aircraft used in the NEM. This is unlikely since measurement on a single day will likely not reflect that of the annual mix. Angle of incidence is important since normal incidence (perpendicular to the surface) may provide as much as 5 dB more attenuation that tangential incidence (parallel to the surface) due to incitement of flexural waves. • The total NLR is primarily from the composite transmission loss characteristics of the structure but is also from the room acoustics controlled by the size and absorptive proper- 37 See note 29.

Acoustical Engineering 71 ties of the receiving room. Therefore, if there is a considerable change in room furnishing between the pre- and post-retrofit testing, a significant change in room absorption will affect the measured NLR. 4.6.3 The Indoor–Outdoor Speaker Method The indoor–outdoor method has been used only a relatively short time, and as of the publica- tion of these guidelines, does not have the endorsement of the FAA. It is currently used by a single acoustical consulting firm, and only general information about the procedure has been made publicly available. This method reverses the location of the sound source and receiver from that of the aircraft and speaker methods; that is, the interior room is the sound source room and the exterior is the receiving area. A calibrated sound source is generated within the room to be tested generating a diffuse sound field, and the receiver noise levels are measured in near-field areas immediately outside of the structure. Since the structure is composed of building elements with varying TL performance (particularly for the pre-retrofit structure), it is necessary to measure the near field of various building components over all exterior façades, measure the area of each, and compute the composite transmission loss of the structure. Currently no national or inter- national standards for the procedure have been promulgated. 4.6.4 Comparison of Acoustical Test Methods As noted previously, each method has significant advantages and drawbacks. The flyover method has the logistical advantage of not requiring anyone present during the measurement. The artificial noise source method has the advantage of only requiring a single meter transferred between the two rooms; also, the measurement may be conducted in less time unless setup with the bucket truck is complicated. More important are the sources of error and unreliability associ- ated with each method. These are summarized in Table 4.2. It is difficult to determine a best test method. Each has significant benefits and limitations. The decision of which method the acoustical consultant selects is influenced by logistical and technical factors. Some SIP work is for future flight operations, making any flyover measurement program impossible. Many other locations are difficult to traverse and have many trees and util- ity wires, making it impossible to employ a crane or bucket truck for the artificial noise source methods. There appears to be no strong consensus among acoustical consultants on a preferred Flyover Artificial Noise Source Diffuse sound source Yes Marginal Repeatable measurement No Yes Adverse room acoustic effects Yes Minimal Statistical results Yes No Elimination of bad results Some Yes ID of weak building elements No Yes Flanking measurement Marginal Yes Courtesy of Freytag & Associates. Table 4.2. Comparison of acoustical test methods.

72 Guidelines for Airport Sound Insulation Programs test method or which provides the most accurate and reliable results. However, most consultants appear to agree on the following limitations and benefits of the two common methods: • It is advantageous to have flyovers for the sound source since they provide the most realis- tic acoustical environment for testing. It is also advantageous to have statistical results since these at least minimize random errors, though not consistent errors such as those from room acoustic effects. • The attended measurements for the artificial noise source method allow use of a moving microphone in the receiving room, significantly reducing room acoustic effects. This attended method also prevents contamination from extraneous noise events and allows for collection of approximate NLR information on individual building elements. Measurements of noise levels in field situations are subject to many variables, including the type and accuracy of instrumentation, the source of test noise (whether it be aircraft flyovers or loudspeaker), the shape and size of the interior room, and the type of furnishings in the room. ASTM E966-04, Standard Guide for Field Measurements of Airborne Sound Attenuation of Building Façades and Façade Elements, recognizes these uncertainties with the following state- ment regarding measurement precision: No body of experience in the use of this guide exists at present; however, it is estimated that the repeat- ability standard deviation of the test procedure is of the order of 2 to 4 dB, depending on frequency. That is, a number of teams of operators testing the same field situation using this procedure would produce a population of test results whose scatter about the arithmetic mean would corre- spond to a standard deviation of 2 dB to 4 dB. Even taking the smaller of these figures, a standard deviation of 2 dB, this means that there is a 16% chance (almost 1 in 6) that the real noise reduc- tion is at least 2 dB less than the measured value. To be confident that eligibility is determined fairly and equitably, a 3-dB allowance should be introduced to account for these uncertainties. No studies have been identified that provide comprehensive statistical assessment of the accuracy and reliability of the two common methods. Such a study would comprise compara- tive testing for the two methods on the same structures and an analytic CTL computation (see Section 4.2.4). The results of such a study would quantify the accuracy and repeatability of the two methods in terms of confidence interval and probability level (e.g., a confidence interval of ±2 dB @ 95% certainty level). This margin of error is a potential issue since it affects program eligibility. Some homeowners may challenge the eligibility testing, necessitating a testing methodology that is beyond dispute or a margin of error for the measurements. However, the FAA has not addressed the issue of mar- gin of error in measurements. Table 4.3 shows the potential effect on program eligibility from the perspective of a ±3-dB margin. One method to address the margin of error is available in the 1992 guidelines, which state, “The exterior levels are taken from mapped DNL contours which show current DNL levels in 5-dB increments. In determining the required noise reduction, the higher end of the noise zone range is always used.”38 This procedure could adjust for some of the margin of error that is possible when using 1-dB contour increments. The FAA’s 1992 AC 150-5000-9A, page 3-18, states that “the exterior levels are taken from mapped DNL contours which show current DNL levels in 5 dB increments. In determining the required noise reduction, the higher end of the noise zone range is always used.” The PGL does not mention the increments of the contour needed to use in calculating interior noise levels. Sponsors and consultants have requested that the FAA clarify whether the recommendation of using the higher end of the 5-dB contour still applies. 38 See note 18.

Acoustical Engineering 73 MEASURED DNL VALUES TRUE MEASUREMENT ERROR DNLin -3 -2 -1 0 +1 +2 +3 42 39 40 41 42 43 44 45 43 40 41 42 43 44 45 46 44 41 42 43 44 45 46 47 45 42 43 44 45 46 47 48 46 43 44 45 46 47 48 49 47 44 45 46 47 48 49 50 48 45 46 47 48 49 50 51 Correctly eligible — True positive (lower right) Incorrectly eligible — False positive (upper right) Correctly ineligible — True negative (upper left) Incorrectly ineligible — False negative (lower left) Courtesy of Freytag & Associates. Table 4.3. Margin of error effects in eligibility testing. 4.7 Best Practice Recommendations: Acoustical Engineering 1. Assist the project team at the outset of the program by participating in the public outreach program to explain program approach, FAA policy and guidelines, home- owner responsibilities, and expectations from the retrofit treatment. 2. The acoustical design criterion used in the majority of programs is to achieve a uni- form level of noise reduction for the entire building envelope. Focus on the major rooms of the house with normal to significant exterior exposure and let the consis- tent envelope treatment apply to rooms with smaller amounts of noise exposure to avoid over- and under-design of habitable rooms. 3. Provide technical input for policy decisions regarding what expenses the program may incur to go above uniform treatments, whether to treat residences where some but not all rooms currently meet an interior DNL of 45 dB, and how to account for measurement tolerances. 4. Design a pre- and post-construction measurement program to efficiently measure the difference in NLR and provide a reasonable estimate for the accuracy and reli- ability of measurement results. 5. To ensure that FAA eligibility requirements and acoustical criteria are met, deter- mine existing interior noise levels for each residence and design treatment protocols using acoustical modeling. Continuously calibrate the TL values used in the model- ing process by field testing typical and diverse housing conditions. Consult with the SIP’s FAA point of contact regarding any eligibility questions. 6. Consult with building material and building system vendors to review new products for use on the SIPs. Assist with material inspections in cases of new, unusual, or ques- tionable products or materials.