Virginia Cooperative Organization Sponsored Jointly by the Virginia. Highway & Transportation Research Council. University of Virginia) Virginia

Transcription

1 K. Haviland J. Research Engineer Faculty F. Noble D. Scientist Research opinions, findings, and conclusions expressed in this (The are those of the authors and not necessarily those of report Highway & Transportation Research Council Virginia Cooperative Organization Sponsored Jointly by the Virginia (A of Highways & Transportation and Department University of Virginia) the 1979 July 80-RI VHTRC EVALUATION OF NOISE BARRIERS by and the sponsoring agencies) Charlottesville, Virginia

2 E. T. ROBB, Chairman, Asst. Environmental Quality Engineer, VDH&T R. E. BLASER, University Distinguished Professor, VPI & SU R. E. CAMPBELL, Assistant Transportation Planning Engineer, VDH&T A. R. CLINE, District Environmental Coordinator, VDH&T R. L. HUNDLEY, Environmental Quality Engineer, VDH&T A. B. JOHNSON, Assistant Construction Engineer, VDHST TOM STEPHENS, Associate Professor of Civil Engineering, VPI & SU R. G. WARNER, Assistant Construction Engineer, VDH&T ROBERT WELTON, Environmental Coordinator, FHWA J. E. YEATTS, Assistant Location and Design Engineer, VDH&T ENVIRONMENTAL RESEARCH ADVISORY COMMITTEE M. C. ANDAY, Assistant Head, VH&TRC L. E. BRETT, JR., District Engineer, VDH&T R. V. FIELDING, Materials Engineer, VDH&T D. D. MCGEEHAN, Research Analyst, VH&TRC W. P. TUCKER, Right-of-Way Engineer, VDH&T ii

3 measurements were taken at six barrier sites- Noise wooden, two metal, and one concrete barrier were studied; two sixth site had no barrier and was studied to determine the ground effect. The approach was to determine insertion the by taking simultaneous measurements behind the barrier losses different elevations over the same point. In this procedure, at the uppermost-microphone is clearly in the line of sight when the traffic, the difference between the level for the of For the measurements, three microphones were microphone. at different heights on a 9.l-m (30-ft) pole and a positioned microphone (with its own support) was placed 1.5 m fourth ft) above ground level. Unfortunately, with the microphones (5 arranged, the uppermost microphone was in line of sight so only the measurement sites close to the barriers. Thus of of the values derived from the analysis of the data must many viewed as differential insertion losses. be predicted and measured noise levels behind the Both were compared, and the results led to the conclusion barriers the barriers were performing as they should be expected that The principal recommendation that could be made, considering to. rather limited scope of this study, was that no policy the should be made that would eliminate the use of any decisions or construction technique on the basis of its material performance. SUMMARY microphone and the level for one of the lower uppermost is the insertion loss at the height of the lower microphones iii

4

5 K. Haviland J. Research Engineer Faculty F. Noble D. Scientist Research noise impact of those facilities when that impact exceeds the current noise standards. The only methods by which that the regulation can be satisfied are as follows- By zoning to prevent the juxtaposition of 2. and residential communities and highways By choosing a location for the facility so as to 3. any impact; as a long-term solution, this avoid would require that the area be zoned so method to prevent residential development after as method over which highway departments have the greatest The and on which they have looked most favorably is the control use of barriers located between the highway (source) The the community (receiver) to lessen the impact of highway and is increasing throughout many of the states. This trend noise especially evident in the urban-suburban areas of Virginia. is EVALUATION OF NOISE BARRIERS by and INTRODUCT ION to Title 23 of the United States Code, According responsible for transportation facilities must lessen agencies 1. By controlling noise emissions of vehicles. other noise sensitive activities. construction of the facility. 4. By constructing a noise barrier. construction of a noise barrier on the right-of-way.

6 such as aesthetics, maintenance, structural Factors weight, and the need to be responsive to the characteristics, wishes create a need for public's variety of materials from a barriers can be constructed. Thus earth berms; metal, which and concrete walls; and combinations of these types of wood are used to attenuate noise. barriers barriers are designed with a computerized The model that uses the various factors that affect mathematical traffic noise as input and calculates highway predicted a level. At least ten barriers have been constructed in noise and many more are in the planning stage. Up to the Virginia the costs for constructing barriers have been very present, and the projected cost for the barriers in the planning high is in the millions of dollars. In addition, very few stage measurements have been taken to determine post-construction effectiveness of the barriers that have been erected. the the investment represented by the Department's Because.of to build barriers, the Department's designers wanted commitment relative to the attenuation achieved by information a group of barriers that already had been representative specific purpose of this study was to determine The effectiveness of some of the noise barriers constructed by the Department and to compare the measured attenuation with the the (design) attenuation for those barriers. predicted the effectiveness of a noise barrier in Determining a life situation is a more difficult task than might at first real assumed. So many factors have be effect on the results of an measurements that it is impossible to plan noise measurement a free of assumptions or corrections because of methodology that occur in some of the factors between measurements. variations constructed. PURPOSE MEASUREMENT RATIONALE

7 most frequently suggested methods for determining The effectiveness of noise barriers are as follows- the Taking measurements on a before and after basis 1. before measurements have the advantage of The have the disadvantage of requiring measurements or corrections concerning variations assumptions traffic conditions, changes in terrain that in as an unplanned result of the barrier occurred (removal of trees, different ground construction different atmospheric condit.ions, etc. cover), Choosing two sites along the highway that are 2. close to identical as possible, one with a as and one without- Because the presence barrier the barrier should be the only difference, of effectiveness could easily be determined. its finding such a situation in a state However, Virginia is very difficult; at the very best, like differences in tree and ground cover would minor making corrections. The traffic necessitate would have to be assumed to be conditions or would have to be sensed at each site identical corrections made to account for any differences and total count, percentage of trucks, or speed. in difficulty would be experienced in determining Great adjusting for differences in the ambient noise and of the sites. levels Taking simultaneous measurements from an array 3. microphones variously spaced on a vertical of Some, at least the uppermost, of the pole would be in line of sight with the microphones and would come as-close as possible to traffic the noise as if the barrier did not sensing The other microphones would be at exist. elevations so that the barrier would be lower them and the traffic. The bottom between would be at 1.5 m (5 ft.) above microphone of sight microphone would be so high above line ground that it might not be affected by the the the evaluations reported here, the decision to In the performance of the barriers investigate made after they was had been installed, which eliminated Method 1 above. Further, describing the noise environment accurately the barrier. However, the after without level so as to approximate ear level. ground possible disadvantage would be that the One surface effect as the 1.5-m (5-ft) level ground would be, if there were no barrier. microphone

8 of the barriers were installed in locations which would none Method 2 applicable. Therefore, Method 3 had to be used. make of the measurements made are given in Particulars i. This table includes the distances from the barrier Table which the microphones were placed and indicates the numbers at channels of recording made on both the NAGRA analog recorder of a digital recorder. The tape number shown is that of the and digital tape, while the figure number refers to the permanent summarized in Appendix A and the experimental parameters are presented in Appendix B. In addition to the five barriers are the measured noise values. Also, results of another study, on in references 3 and 4 (NCHRP 144 and 173), were reported simultaneous microphone readings were required Since several vertical locations, a 9-m (30-ft) pole was assembled at outfitted to accommodate microphones at three vertically and locations. adjustable data acquisition system used was that described The reference 2, except that it was expanded to four channels in recorder can record eight channels). Two additional sound (the meters, together with 131 m (430 ft) of cable and level an NAGRA tape recorder, were borrowed so that recordings additional be made on four channels simultaneously. Three 12.7 could mm microphones were mounted at the adjustable locations (i/12-in) the pole, while the fourth, 1.5 m (5 ft) above the ground, on part of the all-weather microphone system. The outputs was the four microphones were A-weighted in the recording from van four B&K 2204 or 2209 sound level meters. The DC outputs using fed to an A-D converter for recording on digital tape, were the AC outputs were fed to the four input channels of two while recorders. NAGRA in this report in which the geometry of the site is figure and in which results are summarized. Also, the results given a plain site (No. 7) was included as a reference, evaluated, to evaluate the effects of distance above ground especially included for comparison. INSTRUMENTATION

9 Site # Neck Rd. Great Beach) (Virginia Locn. # No. of Distance NAGRA From 4 : i i (Hampton) barrier Metal i i (Springfield) barrier Concrete earth berm 4! on North 1 29 Ch'ville) 2 (Near barrier 3 No 4 E CA) B (Milpitas, barrier C Masonary earth berm D on * * 26 i00 4 4* * O reported in Results 144 & 173 NCHRP Table 1 Summary of Recordings of Tape Figure Description No. # # Digital Channels Channels Barrier (feet) Denbigh Blvd. (Newport News barrier Metal barrier Wooden Churchland Br. (Portsmouth) barrier Wooden CONVERSION: m i ft. * Made in the laboratory.

10 two quartz-coated microphones that were The part of the a data acquisition system described in reference 2 were original calibrated at each site using the activators in their first covers, while the two additional microphones rain calibrated were of 12.2 db was selected because it corresponds to a figure input to the A-D converter, which is half of the 5-volt 10-volt input. Thus, there was maximum 6 db margin on a and a 4.7 db margin on the maximum sound level meter voltage this adjustment was made, the internal reference After of the sound level meters were turned on and their tones levels were read off the meters. Using this infor- equivalent the equivalent calibration level mation, determined for was channel. Later, when the internal reference tones each were as calibration signals, the equivalent calibration levels used the evaluation of the barriers, six sites were For as shown in Table I. At each of the six sites selected from 2 to 7), up to four locations (numbered selected for were recordings, taking into account the following criteria. making Distances behind the barrier should vary from a. 3 m (i0 ft) to about 61 m (200 ft). about Obstacles such as trees and houses should b. avoided as much as possible. be The top two microphones should be higher d. the barrier at their closest location than CALIBRATION PROCEDURES B&K calibrators and checked by with B&K piston_phone. a this procedure the meter was adjusted During that it would so 12.2 db (off the scale) at the estimated LI0* level. The read output at the estimated LI0 level. were used to convert the recorded data to decibel levels. RECORD ING PROCEDURES c. Nontraffic noise should be minimized. to the barrier. * See list of abbreviations on page 25.

11 recordings were made at each location, Fifteen-minute on the digital tape recorder and on the NAGRAs. The both recorder malfunctioned at site #7; however, the NAGRA digital were played back to obtain the digital tapes in the tapes after the digital recorder had been repaired. Table 1 laboratory the recordings made. summarizes the computer analysis programs reported in reference Using all of the available digital tapes were analyzed to obtain 2, (percentage exceedance) levels, L, LNP, NPL, TNI, and LE variances LSI t e and LEp S. Strip art recordings were made G the NAGRA tapes. It was immediately obvious that the of levels on the all-weather microphone were excessive, threshold had been feared during the measurements. Subsequent as in the laboratory have shown that the trouble investigations from a noisy heater power supply. (This trouble has since came rectified.) been for L l, L 10, L5 ' L9 Values data analysis 0 given in App ndix A. are MICNOISE l0 computer program was used to predict The LI0, and LEQ values for comparison with measured data, L50, these are also shown in Appendix A. For this purpose, the and of the program used by the Department was modified version Data output was modified to give decibels to 1. significant figures after the decimal. two LEQ was derived from total traffic noise (i.e., 2. the correct method) in addition to the present by from LI0 and L50 used in the Virginia calculation program. Truck stack height was included as a parameter. 3. Virginia program uses 4 m (13 1/2 ft). The Provisions were made to choose the NCHRP attenuation curve or the presently used barrier 117 curve. A comparison of these two NCHRP is given in Figure i. curves DATA ANALYSES and LEQ, obtained in the PREDICTIVE ANALYSES as follows.

12 o_ n W > 0 n U_ n O _ W n 0

13 sets of values are given in Appendix A. The first Two based on the 4-m (13 i/2-ft) stack height and is the on 117 curve of the Virginia program. The second is based NCHRP on 2.4-m (8-ft) truck stack height and on the NCHRP 174 curve. a data inputs for the MICNOISE l0 program are Essential in Appendix B. The traffic counts shown given obtained were the site, where only the traffic of the near (_east- on lanes was counted. Traffic counts for the far lanes bound) were mostly inferred from near lane counts. The of percentages (TMIX) given were for tractor trailers. truck were also made for medium trucks; these are not shown, Counts were relatively high. but of both measurements and tests are shown in Results A and in Figures 2-6. Note that the LEQ values, Appendix each figure, the near traffic lane, the barrier, and In of the microphone locations are shown. In some cases, all and vertical scales are different, horizontal indicated. as four solid curves show measured LI, LI0 L50 and Lg0 The all cases, the Virginia program with the 4-m (13 I/2-ft) In stack heights and NCHRP 117 curves give the highest truck calculated in each case, whereas, in reality, the curves were complex, since the NCHRP 117 curve is discontinuous. are ideal situations, all of the curves shown in Under 2-6 should have the same shape. Also, the predicted Figures measured LI0 curves should coincide. Were this the case, and would be incontrovertible evidence that the barriers there were performing exactly as the design procedures say evaluated should. they direct counting during the 15-minute recording periods. by the traffic on the roadway was counted with the exception All PRESENTATION OF RESULTS tabulated in Appendix A, are not shown in the figures. although that the only predicted values plotted are for LI0. Also, while the broken lines bordering the shaded regions values, the predicted LI0 values obtained by the two methods. show The shapes shown for the curves obtained from predictions. Virginia program are estimated from the four points which the INTERPRETATION OF RESULTS

14 LLI > LLI 0 >.- O I..U : 83 0 uo l-- d -r" U_ D lo

15 UJ rn > 0 < 0 rn 0 - n" U 0 < co >- U 0 < L9 0 "r > U I-- W UJ, LL 0 "r 0 0 o LO _J _.J W L9 ll

16 W 0 0 I >- O >. W 0 n, 0 0 W LU D T 12

17 LU k- b- 0 rr 0 0 rn, W n W r, I- LL 35 T n- 0 n-!3

18 W n," W n," 0 "r _1 7- W n," co U.l oo CL W U. "r" o3 74

19 is apparent from an examination of the figures It the above conditions were not met, except in an approxim.te that influenced the results. These factors are discussed evidently below. Performance of all-weather microphone. From a cursory 1. of the curves, inspec{i' 'n one compares noise levels as by the four microphones he sees that the lower the measured the more its reading is attenuated by the microphone, However, the lowest microphone, i.e., the all- barrier. microphone, shows a sharp increase in most cases. weather is consistent with a higher electronic background noise This which has since been traced to a noisy heater power level, Overall Noise Level Prediction. Consider the comparison 2. the predicted and measured LI0 levels at the top between cases. This discrepancy seems to be partly due to most effect of light trucks. Analyses treating all trucks the heavy trucks (not shown) tended to lead to overprediction, as the present analyses, in which light trucks were whereas as automobiles, tended to underprediction. Althouqh treated seems to confirm the newer practice of treating light this as intermediate noise sources, it was not considered trucks proper topic to pursue in this investigation. a Effect of Aircraft Overflights. Aircraft overhead contributed 3. to some o he measured values, especially when significantly the first site, on Denbigh Blvd., only very light At was encountered so that aircraft contributed traffic Typical of this effect are the curves shown considerably. Figure 2 for the 30.5-m (100-ft) and 45.7-m (150-ft) in Here the measured LI0 curves are higher than locations. and all of the measured values tend to be aligned predicted, because the barrier provides no attenuation of vertically, from overhead. noise the second site, on Great Neck Road, there were many At aircraft during the reading at the 30.5-m (100-ft) low-flying Thus, in order to demonstrate that the barriers do way. one must take into account the various factors which function, Thus, in effect, readings of the lowest microphone supply. be noted with care. must microphones. Although the shapes of the curves three reasonably, overall level.predictions are off in compare the road traffic was relatively light. The effect was not only to make the L location. and LI0 1 of Figure 3 almost vertical, but also to make them curves indicating that an aircraft was overhead at least merge, 10% of the time. aircraft were overhead during the measurements Although 1-64, they were higher, and the relatively heavy road on traffic reduced their effect somewhat, as shown in Figure 4. 15

20 Ground Effect. It had been expected that some of the 4. at the lower microphones might be attributable attenuation ground effects. For this reason, measurements were to at the seventh site on U.S. 29 north of Charlottesville made there is no barrier. The results, shown in Figure 7, where almost no effect; thus, most of the attenuation indicate in the other curves must be attributed to the barriers. seen lack of height effect is confirmed by an analysis This in reference 4. It appears that lush vegetation reported increase attenuation due to distance, but that the same does with distance is seen to an appreciable height attenuation ground. above Effect of Analysis Parameters. The figures show calculations 5. LI0 made with tw " x r me sets of asstumptions. The for 4 m (13 1/2 ft) and the NCHRP 117 (ref. 6) curves are at The lower values are based on the standard 2.4-m used. stack heights and on the lower NCHRP 174 (ref. 5) (8-ft) curves. attenuation the newer barrier attenuation curves. In fact, it was and a result of similar measurements reported in NCHRP 144 as 3) that the NCHRP 173 attenuation curves were (ref. developed. most completely documented investigatio n of barriers The to be that reported in NCHRP 144 (ref. 3). However, the appears but at the same location in the present study, they were heights simultaneously at different locations but at the same taken in order to support whatever comparison can However, made, one set of curves from NCHRP 144 is given in Figure 8. be range of L50 attenuation values is shown in shaded blocks, The the broken curves show predicted attenuations. Thus these while differ from those shown in Figures 2-7 in that the latter curves absolute values whilst Figure 8 shows relative levels. show predicted attenuations of Figure 8 are based on the NCHRP The methods. i17 values follow the analysis method presently used by higher Department. In these, truck stacks are assumed to be the into account some of the other factors mentioned, Taking measured data do appear to favor the lower stack height the difference between the present study and that in NCHRP 144 major that whereas readings were taken simultaneously at different is in the NCHRP 144 study. Thus the present study gives height barrier attenuation effects directly, whereas NCHRP 144 the distance effects directly and barrier attenuation only gives a statistical analysis. through 16

21 W n,' r7 W W W..J W > - W 0 _J >- W O W o n7 _J W 0 O _J _J 0 C W > W O < >- W W n W 17

22 W n

23 a further aid in interpreting these results, As 2 shows a comparison between measured and calculated Table effects. The latter were obtained by the Department barrier with truck stacks at 4.m (13.5 ft). These effects are method in terms of the differences between the.li0 levels expressed the upper and those at the second from lowest m±crophones. at of sight, the values given would be the insertion losses. line such conditions were met in only a few cases, as noted However, the table, so that the remaining values must be viewed as in insertion losses. Comparing the measured and differential values given, it is seen that differences are mostly computed 2 db, and that neither measured nor computed values are within One exception is the reading at 30.5 m (i00 ft) from favored. barrier at Great Neck Road, where aircraft flyovers reduced the measured differential loss to zero. Another is the the value at 3.1 m (i0 ft) from the barrier on calculated on a comparison between predicted and measured Based levels behind the barrier, and taking into account the noise already mentioned, the weight of evidence indicates factors the barriers are performing as they should be expected to. that is to say that if a barrier is designed according to present That and if one of the methods of construction (wood criteria, metal panels, or concrete) used on the sample barriers is panels, to, then the barrier will meet the designed objectives. adhered the lowest microphone because of difficulties already (Not If the upper microphone were always clearly in the mentioned.) the NCHRP 174 method predicted a differential of 5.7 db, Here, is much closer to the measured 6.7 db value. which CONCLUSIONS 19

24 0 2O

25 measurements reported here were made on the The that they would be treated as preliminary, and understanding if the methodology to be used proved feasible, then there that be an opportunity to rectify any apparent discrepancies and would make more accurate measurements. Therefore, these recommen- to address the question of whether more measurements should dations Based on the somewhat limited study reported here, it is 1. that no policy decisions be made at present recommended would eliminate the use of any material or construction which on the basis of performance. This assumes, of technique It is recommended that a second series of barrier 2. be planned to accomplish the following. measurements Support a recommendation on whether changes should be c. in Virginia's design methodology. made It is further recommended that the following discrepancies 3. available test equipment be rectified. in Build a pole that will satisfactorily hold all four b. so as to free the all-weather microphone microphones Build a calibration system to drive four activators c. simultaneously. Add a push-button cutoff to stop measurement while d. are overhead. aircraft RECOMMENDATI ONS made so as to provide additional input for the decision- be process. making that the proper design technique is used in each course, case. a. Add to the variety of barrier types tested. b. Provide verification of the performance of barriers. Procure three 12.7 mm (i/2-in) quartz.-coated microphones a. rain covers and dessicators. with for installation near the highway.

26

27 Haviland, J K and D F Noble, Tentative Working l Evaluation of Standing Noise Barriers", unpublished, Plan May Computer Programs in the Design of Noise Barriers Predictive Before and After Approach, Part I. The Data Acquisition A February "Highway Noise A Evaluation of Traffic Noise Reduction Methods", Field 144, Los Angeles, California, NCHRP B A., Commins, D. E., and Galloway, W. J., Kugler, Noise A Design Guide for Prediction and Control" "Highway J., Kugler, B. A., and Nelson, W. Design Guide for Highway Engineers A REFERENCES Haviland, J K., and D. F. Noble, "Effectiveness of Virginia Highway and Transportation Research System", Charlottesville, Virginia, VHTRC 78-R32, Council, Kugler, B. A., and Piersol, A.G., and "Highway Noise Generation and Bolt, Newman, Control" NCHRP Beranek, 173, Los Angeles, California, NCHRP 174, Los Angeles, California, C. G., Galloway, Gordon, L., "Highway Noise. D. NCHRP 117, Los Angeles, California,

28

29 percent exceedance sound level, the A-weighted E level equaled or exceeded E% of the time. sound deviation expressed in decibels, based on Standard that L assuming is derived from a normal distribution. E deviation expressed in decibels, based on Standard on a specific channel. readings LIST OF ABBREVIATIONS Equivalent A-weighted sound level over a 24-hour period. LEQ Noise Pollution Level computed from- LNp LNp L50 + (LI0- L90) + L90) (L10 60 NPL Noise pollution level. TNI Traffic noise index computed from 4(L10 L50) + L TNI LSIG LEPS 25

30

31 OF 3A.qRIER EVALUATION RESULTS Levels in dba) (Noise Dist. Measured Levels Caiculated for Location trucks 13.5' trucks and 8.0' 174 NCHRP APPENDIX A Calculated for barrier method.enbich O 5O 5O I00 i00!00 I Great Neck: !CO A-i

32 Si,a.No. 23 I O 9 APPm. D A (continued) for Calculated 13.5' trucks Grea._ :;eck !0 2O 17!0 33!0 A-2

33 Si e Local±on i{ [ Levels Calculated for.measured g trucks ! S!.,5,, ,,72.,9, trucks and 8.0' 174 NCHRP L LEQ 60.2.,z,, /70 75.! 54.3, 56.6 APPENDIX A (continued) Calculated for barrier method ,5,,,..-.,o.,,,4.4 S.o. ss. s. a , O ,:? 8.2 4_ ,66.,4 6_.i,.., ,.2 6,8.0!_6_!, ! = : : :,2 s.8 60 ],._ o ,0 6 4i6! I DDe ! ,-,! "J, ;', ;' I i00 14! O OO 2OO A-3

34

35

36 ' o o r- Lr n't - B-2

37 r.q o

38 o ill

39 L 0

40 o

41 1 LE) 0 (:D (:D r--! (:D C I r,- c C,,i -,-! r-t-] O B-7

42 ('Xl : i 0 r--i 0 C',4 CO,--I re) oe'} o") C l m.-i C:) 0 r-'i m-l! o") re) (',') r-i Xl 1 t n' u*) m-t r-i I C) t 0 0

43 O " Lr) 0 N Xl,--!,-4 (Y) (Y) 1 (N B-9

44 B-IO