• Construction Noise
• Measurement
• Noise Barriers
  • Abatement
  • Acceptance Criteria
  • Design/Construction
  • Inventory
  • Research
• Noise Compatible Planning
• Regulation and Guidance
• Resources
• Tire Pavement Noise
• Traffic Noise Model
• Training

Noise Barrier Acceptance Criteria: Analysis

Chapter 3. Testing the Results for the Four Abatement Feasibility and Reasonableness Factors, Excluding Viewpoints

In this chapter, the results of the findings of Chapter 2, for the four studied feasibility and reasonableness factors, were tested by performing a sensitivity analysis. The purpose was to identify outcomes of possible combinations of factors and apply the combinations in a sampling of existing highway projects with abatement.

3.1 Sensitivity Analysis

As has been seen, there are a wide variety of combinations of the various factors in the reasonableness criteria. The previous chapter examined them and their relationships in a qualitative fashion. This section presents a more quantitative and analytical examination in the form of a sensitivity analysis.

3.1.1 Factors Studied

In order, four factors were studied:

1. Benefited noise reduction
2. NRDG in dB
3. NRDG quantity in terms of the number and percentage of benefited receptors
4. APBR

Ranges were defined for these factors, and the various resulting combinations were identified. Several scenarios were then developed and modeled with the FHWA Traffic Noise Model Version 2.5 (FHWA TNM) noise prediction program, which is specified in the regulation for use on federal and federal-aid highway projects.

The results of these modeling cases were then applied to the combinations of the factors to determine the resulting decision on reasonableness. For a barrier to be reasonable in this analysis,

• The NRDG had to be met for the specified number or percentage of benefited receptors, and
• The APBR had to be met.

The ranges in the factors were as follows:

• Benefited noise reduction: values from 5 dB up to the NRDG (for example, for a 7 dB NRDG, the regulation allows the benefited noise reduction to be 5, 6 or 7 dB)
• NRDG, ranging from 7 dB to 10 dB (7, 8, 9 and 10 dB), as permitted in the regulation
• NRDG quantity in terms of the number and percentage of benefited receptors
  • For the number of benefited receptors, values 1, 2 and 3 were used
  • For the percentage of benefited receptors, two situations were studied:
    • The percentage of first-row benefited receptors only, using 10%, 50%, 67% and 80%, and
    • The percentage of all of the benefited receptors, using 10%, 25%, 50%, 67% and 80%.
• APBR: A range of values from 500 to 2,800 SF/ benefited receptor (500, 1000, 1,500, 2,000 and 2,800)

The percentage ranges for both the first-row and all receptors were derived from the ranges in values in the SHA policies. The APBR values were also based on the ranges in the values used by the SHAs in their policies.

Use of the above values for these different factors resulted in 1,080 different criteria combinations to be analyzed.

3.1.2 Modeled Scenarios

In order to study these combinations for a variety of receptor scenarios with FHWA TNM, a four-lane divided highway model was created. The basic scenario consisted of two travel lanes in each direction. Each travel direction was modeled by a single FHWA TNM roadway with a width computed based on two travel lanes and paved shoulders, not by a FHWA TNM roadway for each travel lane. The difference was not felt to be significant for the purposes of this study. See Figure 7 for a plan view (from above) and a cross-sectional view of the roadway and noise barrier portion of the model.

Figure 7. Plan view (top) and cross-sectional view (bottom) of roadways and noise barrier portion of FHWA TNM model for sensitivity test cases.

Each FHWA TNM roadway carried 3,600 vehicles per hour per direction. The traffic was divided into 88% automobiles (3,168 veh/hr), and 12% trucks (two-thirds heavy trucks (288 veh/hr) and one-third medium trucks (144 veh/hr)). The travel speed was 65 miles per hour.

The noise barrier was located at the edge of the shoulder.^[8] "Lawn grass" ground cover and a flat site (elevation of 0 feet and no roadway grade) were assumed, with a grassy median between the two directions of travel. For each of the cases, a range of six barrier heights was examined: 6, 10, 14, 18, 22, and 26 feet.

Several receptor scenarios were modeled:

• Varying receiver density
• Varying number of rows of receptors
• Different distances back from the barrier to the first-row of receptors

Figure 8 shows that two different distances back from the barrier to the first-row of receptors were used. The assumption was that the barrier was located on the edge of the roadway shoulder and that there was 40 feet of right-of-way space from the barrier to the property line. The distances were:

• 90 feet (thus 50 feet back from the property line)
• 140 feet (thus 100 feet back from the property line)

Eleven FHWA TNM receiver points^[9] per row were placed at 200-foot spacing along a line parallel to the roadway over a distance of 2,000 feet. Figure 9 shows the receiver portion of the FHWA TNM model. In the analysis, each receiver was assigned to represent:

• One receptor, for a receptor spacing of 200 feet and a total of 11 receptors per row
• Two receptors, for a receptor spacing of 100 feet and a total of 22 receptors per row
• Four receptors, for a receptor spacing of 50 feet and a total of 44 receptors per row

In each case, levels were calculated at only eleven FHWA TNM receiver points per row.

The shielding provided by the rows of houses was modeled in FHWA TNM by placing a TNM "building row" object between the first and second and between the second and third rows of receptors. The building row blockage percentages for the three receptor spacings of 50, 100, and 200 feet were chosen as 40%, 60% and 80%, respectively.

Thus, a total of 108 scenarios were studied (3 receptor spacings x 3 building row cases x 2 distances back from the barrier x 6 barrier heights). The result of examining the 1,080 combinations of reasonableness parameters for these 108 cases yielded a total of 116,640 decisions on barrier reasonableness.

Figure 8. Sketches of receptor scenarios used in modeling the sensitivity test cases for first‑row receptors 90 feet (top) and 140 feet (bottom) from barrier.

Figure 9. TNM plan view plots of receptor scenarios in FHWA TNM model for sensitivity test cases: 140 feet back (top) and 90 feet back (bottom).

3.1.3 Outcomes of Possible Combinations of Factors on the Likelihood of Abatement

These results of being reasonable or not being reasonable were arrayed in a color-coded spreadsheet of 1,080 rows by 108 columns, capable of being sorted by different factors. The goal was to be able to use this array to see the trends that emerge in the patterns of the decisions. It was a challenge to display this entire array in report format in a meaningful way, although the resulting series of charts should be useful. The spreadsheet developed for this analysis is available as a deliverable for this task. Some examples of the contents and how they can be used are illustrated below, followed by presentation of the array and discussion of the broader findings.

Figure 10 shows a very small portion of the upper left corner of the spreadsheet. The first five columns represent the reasonableness factors being varied. The rows are sorted first by NRDG, then by benefited noise reduction, then by NRDG type, then by value for the NRDG type, and finally by APBR. For all of the rows shown in this example, the NRDG is 7 dB and the benefited noise reduction is 5 dB. The NRDG type has one of three codes:

• N (in green) to indicate that NRDG number/percent will be the needed number of benefited receptors.
• P1 (in yellow) to indicate that NRDG number/percent will be the needed percentage of first-row benefited receptors, or
• PA (in red) to indicate that NRDG number/percent will be the needed percentage of all benefited receptors

The top rows indicate the scenarios being studied:

• A sequential case number
• The receptor spacing (50, 100 or 200-foot); in this sample all are at 50 feet
• The distance back to the first-row (50 or 100 feet back from the right-of-way line, representing 90 or 140 feet from the barrier); in this example all are at 50 feet
• The number of building rows (1, 2 or 3); in this sample the first six columns are for one row and the last two are for two rows, and
• The barrier height (6 to 26 feet in 4-foot increments)

Then, the green and yellow cells at the intersection of a row (combination of criterion factors) and a column (the scenario being studied) indicate the reasonableness decision of Yes or No, respectively. These answers are further coded by the deciding factor: i.e., number or percentage of receptors (-n, -p1, -pa) or APBR (-c).

Even in this small portion of the array, trends start to emerge. A 6-foot high barrier was shown to be reasonable for none of the combinations of factors (no receptors had a noise reduction at or above the 7 dB NRDG). Also, in none of the cases shown was a barrier reasonable for an APBR of 500 SF/benefited receptor, regardless of the barrier height (there were too few benefited receptors for these six one-row cases and two two-row cases, even at 50-foot spacing to achieve an APBR that low).

Figure 10. Upper left portion of reasonableness decision array for the sensitivity test cases: NRDG of 7 dB, Benefited Noise Reduction of 5 dB, 50-ft receptor spacing, 50-ft distance back to first-row (90 ft from barrier), and mostly one-row cases.

As a second example, Figure 11 shows the result of zooming out farther for the upper left corner of the decision array. The characters are not meant to be readable in this example, but the colored patterns will be discussed. The figure shows six major blocks of columns of decisions; these are all for the 50-foot receptor spacing. The first three blocks of columns are for the 50-foot distance back for cases of one, two, and three rows of houses, respectively. Then, the second three blocks of columns are for the 100-foot distance back for one, two, and three rows of houses, respectively. Within each block are columns for each of the six barrier heights.

Figure 11. Decision patterns indicated by color bands in portion of reasonableness array for 50-ft receptor spacing, NRDG of 7 dB and Benefited Noise Reductions of 5 dB (upper half) and 6 dB (lower half).

Not shown, the array has another set of six blocks of columns to the right for the 100-foot receptor spacing, and then six more blocks for the 200-foot receptor spacing.

The rows shown in this figure are for all of the combinations with an NRDG of 7 dB and benefited noise reductions of first 5 dB and then 6 dB. Results for NRDG of 7 dB and benefited noise reduction of 7 dB would be below these results, followed by the combinations for NRDG values of 8, 9 and 10 dB.

Trends can, again, be seen:

• For none of the combinations shown was the 6-foot high barrier reasonable^[10] (vertical yellow pattern; again, the NRDG of 7 dB was not achieved for this low height).
• More cases are reasonable at the taller barrier heights for two rows compared to one row, and more for three rows compared to two rows (more green and less yellow; the taller barriers are picking up second-row benefits in sufficient numbers to drive the APBR below the criterion values).
• For those cases with an APBR of 500 SF/benefited receptor, the "not reasonable" results for a one-row case (yellow horizontal bands) becomes reasonable (green) for the two-row and three-row cases, as more second row benefits are picked up, driving down the APBR.

As a third example, Figure 12 shows a very small portion of the lower right corner of the array, representing all of the combinations of factors for a high NRDG of 10 dB and a high benefited noise reduction of 10 dB. The three major blocks of columns of decisions are all for the 200-foot receptor spacing and the 100-foot distance back. The first block is for one row of houses, the second for two rows, and the third for three rows. Within each block are columns for each of the six barrier heights. For the 200-foot spacing with these high values of NRDG and benefited noise reduction, very few cases are reasonable:

• None for the cases for one-row of receptors (all yellow in the left block) - the 10 dB NRDG cannot be achieved at the lower heights, and there are not enough receptors at this low density that achieve the 10 dB reduction to be benefited to counter the high costs for the taller heights.
• Only those two-row cases at the highest APBR of 2,800 SF/ benefited receptor and barrier heights of 18 and 22 feet (green highlights in the center block) - these tall heights result in enough second-row benefits to meet the liberally high APBR despite the high cost (yet, the cost becomes too high at 26-feet because there are not enough additional benefited receptors to overcome the extra cost), and
• Some of the three-row cases with APBRs of 2,000 and 2,800 SF/ benefited receptor for barrier heights of 18, 22, and 26 feet (green highlights in the right block) - it appears that enough third-row receptors are benefited to drive down the APBR.

Figure 12. Lower right portion of reasonableness decision array for the sensitivity test cases: NRDG of 10 dB, Benefited Noise Reduction of 10 dB, 200-ft receptor spacing, 100-ft distance back to first-row (140 ft from barrier) and one-, two- and three-row cases.

The above three examples illustrate the content and use of the decision array in its original format, sorted by NRDG. There are many ways that the array can be manipulated and sorted. Following are the results of two sorts, each of which provides a broad picture of how the reasonableness decisions are affected by parameter changes. These two sets of figures could be used by an SHA to access potential effects of policy changes. The spreadsheet containing the array will be made available as one of the tools developed in this research, which are presented in a separate report.

3.1.4 Reasonableness Based Solely on NRDG

One desired outcome of this research was an understanding of the sensitivity to the NRDG value and the quantity (number or percentage) of benefited receptors needing to meet the NRDG. The array was re-run taking APBR out of the decision process, and then sorted by the NRDG quantity. The next three figures present the results. The data are not meant to be read; the patterns of the color shading are what are of importance: green means abatement is reasonable; yellow means it is not reasonable.

Figure 13 is for the needed number of benefited receptors (N) and is divided into three horizontal sections for values of 1, 2, and 3. Each horizontal section is further divided by NRDG, which is further divided by benefited noise reduction (and further by APBR, which is not varied but was not easily removed from the array).

Figure 14 is for the needed percentage of first-row benefited receptors, divided into four horizontal sections for values of 10%, 50%, 67%, and 80%, with the above subdivisions by NRDG and benefited noise reduction.

Figure 15 is for the needed percentage of all benefited receptors, divided into five horizontal sections for values of 10%, 25%, 50%, 67%, and 80%, with the above subdivisions by NRDG and benefited noise reduction.

As illustrated and discussed for the three previous examples, each figure is divided into three major vertical sections of decisions for: 50-foot receptor spacing, 100-foot receptor spacing, and 200-foot receptor spacing. Each vertical section is divided into two subsections: 50-foot distance and 100-foot distance. Each subsection is divided into three blocks of columns for the cases of one, two, and three rows of houses. Within each block are columns for each of the six barrier heights.

Figure 13, Figure 14, and the top two sections of Figure 15 show the insensitivity of the NRDG reasonableness criterion to the NRDG quantity. The narrow vertical yellow bands are for the 6-foot barrier heights. These bands widen to include the 10-foot barrier when the NRDG increases from 7 and 8 dB to 9 and 10 dB for each of the horizontal sections in Figure 13 (values of N of 1, 2 and 3), in Figure 14 (the four percentages of first-row benefits), and for the 10% and 25% bands for "percentage of all benefited receptors" in Figure 15. This finding holds true no matter the receptor spacing, distance back from the barrier, or number of rows of receptors in the study area, at least for the range of values in the array.

As one moves down into the third horizontal band in Figure 15, the required percentage of all benefited receptors increases to 50%. Some of the 10-foot height cases for NRDG of 7 dB and 14-foot cases for NRDG of 8 dB begin to become "not reasonable" for the two-row and three-row study areas. Moving down to the 67% and 80% bands for "percentage of all benefits," many more of the two-row and three-row cases become "not reasonable" while the one-row cases stay reasonable. These results confirm exactly what was illustrated earlier in Section 2.2.1.3. A barrier that would be reasonable for a one-row neighborhood suddenly becomes "not reasonable" if that neighborhood was instead a two-row or three-row community. Adding more houses, many of which benefit to some degree from the barrier changes the decision because it is based on "percentage of all benefited receptors" instead of "percentage of first-row benefits" or even "percentage of impacted receptors."

Figure 13. Reasonableness decision array for NRDG for number of benefited receptors of 1 (top), 2 (middle), or 3 (bottom).

Figure 14. Reasonableness decision array for NRDG for percentage of first-row benefited receptors, divided into four horizontal sections for values of 10%, 50%, 67%, and 80%.

Figure 15. Reasonableness decision array for NRDG for percentage of all benefited receptors divided into five horizontal sections for values of 10%, 25%, 50%, 67%, and 80%.

3.1.5 Reasonableness Based on NRDG and APBR, Sorted by APBR

This next decision array includes both the NRDG criterion and the APBR criterion. It is divided into five figures based on APBR:

• Figure 16: APBR = 500 SF/benefited receptor
• Figure 17: APBR = 1,000 SF/benefited receptor
• Figure 18: APBR = 1,500 SF/benefited receptor
• Figure 19: APBR = 2,000 SF/benefited receptor
• Figure 20: APBR = 2,800 SF/benefited receptor

Again, the vertical sections and subsections are the same as for the previous figures. The horizontal bands are sorted first by NRDG, then benefited noise reduction, then NRDG type, and finally quantity.

The following trends can be observed while reviewing each figure.

For an APBR of 500 SF/ benefited receptor:

• None of the one-row, two-row, or three-row cases were reasonable for the 100-foot and 200-foot receptor spacings.
• None of the one-row cases were reasonable for the 50-foot receptor spacing
• Several barrier heights were reasonable for 50-foot spacing, but as the NRDG increased (moving down the horizontal bands in the figure), fewer heights were reasonable; the increase in barrier area grew proportionately faster than the number of benefited receptors meeting the NRDG.
• Combinations with high percentages of all benefits for the NRDG quantity were more likely to be "not reasonable"; these cases are evidenced by the yellow intrusions into the vertical green bands.
• For the 10 dB NRDG with the 10 dB benefited noise reduction, none of the cases for any of the receptor spacing at any height were reasonable.

For an APBR of 1,000 SF/ benefited receptor:

• None of the one-row, two-row, or three-row cases were reasonable for the 200-foot receptor spacing.
• None of the one-row cases were reasonable for the 100-foot receptor spacing. The two-row cases became reasonable for the 14- to 18-foot heights, and the three-row cases were reasonable for the 14- to 26-foot heights.
• Many more cases were reasonable for the 50-foot spacing, including for the high NRDG with high benefited noise reduction, although use of the high percentages of all benefits for the NRDG quantity continued to result in "not reasonable" decisions.

For an APBR of 1,500 SF/benefited receptor:

• For the 200-foot receptor spacing, some of the three-row cases at the middle heights of 14 to 18 feet were reasonable. The taller walls were unreasonable based on cost.
• For the 100-foot receptor spacing, the 10-foot height became reasonable for the one-row case for an NRDG of 7 to 8 dB, but no heights were reasonable when the NRDG increased to 9 to 10 dB.
• Many more of the two-row and three-row cases became reasonable for the 100-foot receptor spacing: for the two-row cases, heights from 10 to 26 feet were reasonable for an NRDG of 7 to 8 dB. However, for the NRDG of 10 dB, only the tallest heights for both the two-row and the three-row cases were reasonable; some unreasonable results for criteria combinations required 67% or 80% of all benefited receptors.
• For the 50-foot receptor spacing, many more cases became reasonable at the higher NRDG values and at the taller barrier heights.

For an APBR of 2,000 SF/ benefited receptor:

• All of one-row cases were still not reasonable for the 200-foot receptor spacing. Most of the two-row cases in the 14- to 18-foot range were reasonable. For the three-row cases, most of the heights between 14 and 26 feet were reasonable
• For the 100-foot spacing, the one-row cases were reasonable for the 10- to 18-foot range for the NRDG of 7 to 8 dB, but were only reasonable in the 14- to 18-foot range for the NRDG increased to 9 to 10 dB. For the two-row case, heights from 10 to 26 feet were reasonable for all NRDG values. For the three-row case, most of the heights between 10 and 26 feet were reasonable although some started to become unreasonable for the percentage of all rows equal to 67% and 80%
• For the 50-foot receptor spacing, more cases became reasonable at the higher NRDG values and at the taller barrier heights, even for the one-row case. The high values for "percentage of all benefited receptors" continue to result in "not reasonable" decisions.

For an APBR of 2800 SF/benefited receptor:

• For the 200-foot receptor spacing, the one-row cases became reasonable a height of 10 feet for a NRDG of 7 to 8 dB but remained not reasonable for NRDG values of 9 to 10 dB.
• For the 50-foot and 100-foot spacings, more cases at more heights became reasonable, except when the higher percentage of all benefited receptors was applied - the high percentages continued to return negative decisions, even for cases where all of the other parameter values led to barriers being reasonable.

For most of the criteria combinations, the decisions seemed very insensitive to the "distance back" parameter. The extra distance away from the barrier was only 50 feet. It is expected that a larger increase would start to show more differences in the two situations.

Figure 16. Reasonableness Decision Array, APBR = 500 SF/benefited receptor.

Horizontal blocks, from top to bottom, are for NRDG of 7, 8, 9, and 10 dB

Figure 17. Reasonableness Decision Array, APBR = 1,000 SF/benefited receptor.

Horizontal blocks, from top to bottom, are for NRDG of 7, 8, 9, and 10 dB

Figure 18. Reasonableness Decision Array, APBR = 1,500 SF/benefited receptor.

Horizontal blocks, from top to bottom, are for NRDG of 7, 8, 9, and 10 dB

Figure 19. Reasonableness Decision Array, APBR = 2,000 SF/benefited receptor.

Horizontal blocks, from top to bottom, are for NRDG of 7, 8, 9, and 10 dB

Figure 20. Reasonableness Decision Array, APBR = 2,800 SF/benefited receptor.

Horizontal blocks, from top to bottom, are for NRDG of 7, 8, 9, and 10 dB

3.1.6 Findings of Sensitivity Analysis

One desired outcome of this analysis was to identify combinations of factors that are very likely to result in: (1) inclusion of noise abatement or (2) exclusion of noise abatement.

The first finding is that the NRDG reasonableness criterion is insensitive to the NRDG quantity for values of 1, 2, and 3 benefited receptors, for the tested percentages of first-row benefited receptors, and for 10% and 25% of all benefited receptors. As the required percentage of all benefited receptors increases to 50%, some cases begin to become "not reasonable" for the two-row and three-row study areas. As the required percentage of all benefited receptors increases to 67% and 80%, many more of the two-row and three-row cases become "not reasonable" while the one-row cases stay reasonable. These results illustrate how a barrier that would be reasonable for a one-row neighborhood suddenly becomes "not reasonable" if that neighborhood was, instead, a two-row or three-row community if the criterion is based on "percentage of all benefited receptors."

Next, the overall reasonableness is very dependent on the APBR.

For an APBR of 500 SF/ benefited receptor, none of the one-row, two-row, or three-row cases were reasonable for the 100-foot and 200-foot receptor spacings, and none of the one-row cases were reasonable for the 50-foot receptor spacing. Combinations with high percentages of all benefits for the NRDG quantity were more likely to be "not reasonable." For the 10 dB NRDG with the 10 dB benefited noise reduction, none of the cases for any of the receptor spacing at any height were reasonable.

As APBR increases, cases become increasingly reasonable, although even at the high values of APBR, the requirements for the higher percentages of all benefited receptors continued to result in negative decisions, even for cases where all of the other parameter values led to barriers being reasonable.

Finally, for most of the criteria combinations, reasonableness was insensitive to the "distance back" parameter, although the difference in distances back was only 50 feet

3.2 applying the combinations to a sampling of existing highway projects with abatement

The selected combinations of feasibility and reasonableness factors were applied to study areas from actual projects to assess how changes in values would affect noise abatement decisions.

3.2.1 Study Areas

Four study areas were selected for testing. The selection process involved reviewing numerous previously-studied areas along actual projects to identify unique but typical cases.

Land use type was a primary consideration in case selection. Comparison of test results would be significantly complicated for study areas with different land use types. Therefore, only areas of single-family residential development were considered.

Study areas were selected to provide different residential development characteristics such as density, distance to first-row residences, and first-row sound levels. Roadway characteristics were also considered including the number of travel lanes, road elevation relative to the adjacent residences (i.e. cut, fill), and design hour volume (DHV).

Table 19 summarizes the key characteristics of Study Areas A through D.

* From edge of near travel lane

Density correlates directly to the number of benefited residences and is often the most important factor in the reasonableness determination. Therefore, the study areas include a range of residential densities. Study Areas A and B are high density (small lots); Study Area C is medium density, and Study Area D is low density (large lot). Figures showing aerial photographs of each of the study areas are in the sections describing the results for each area.

The distances to first-row residences also vary. First-row residences in Study Areas A and C are as close as 100 feet of the edge of the near travel lane while residences in Study Areas B and D are a minimum of 160 feet and 240 feet, respectively, from the edge of the near travel lane.

The adjacent roadway has ten travel lanes adjacent to Study Areas A and C, eight lanes adjacent to Study Area D, and six lanes adjacent to Study Area B. The DHVs for Study Areas A and C are also higher than for Study Areas B and D. Truck percentages range from 6% to 15%

Finally, there is a wide range of predicted sound levels for first-row residences in each study area. First-row sound levels in Study Area C are up to 79 dB(A) due to the close proximity of residences, a very high DHV, and the fact the road is in cut. First-row sound levels are comparable for Study Areas A and B. Study Area D has the lowest sound levels due to the higher distances between first-row residences and the edge of the near travel lane.

3.2.2 Study Factors

The factors that were studied were developed by examining the range in values used in the SHA policies and the expected likelihood or unlikelihood of a positive decision on a barrier design. These factors include:

• Feasibility Noise Reduction
  • 5 dB (a "static" value, per the RFP)
• Feasibility Quantity
  • Number - one impacted receptor
  • Percentage of first-row impacted receptors - 25%, 50%, 75%
  • Percentage of all impacted receptors - 25%, 50%, 75%
• NRDG and benefited noise reduction (four combinations)
  • 7 dB and 5 dB (most common combination used by SHAs - lowest possible NRDG, with lowest possible benefited value)
  • 7 dB and 7 dB (lowest possible NRDG, with highest possible benefited value)
  • 9 dB and 9 dB (highest combination of equal values used by SHAs - none used 10 dB and 10 dB)
  • 10 dB and 5 dB (second most common combination used by SHAs - highest possible NRDG with lowest possible benefited value)
• NRDG Quantity
  • Number - one benefited receptor (no SHA policies used more than one)
  • Percentage of first-row benefited receptors - 25%, 50%, 75%
  • Percentage of all benefited receptors - 25%, 50%, 75%
• CE in terms of APBR
  • 800 SF/benefited receptor
  • 1,250 SF/ benefited receptor
  • 2,600 SF/ benefited receptor

Feasibility was examined separately from reasonableness, with seven individual cases studied for feasibility. Reasonableness was studied for combinations of values, together resulting in 84 cases: [4 pairs of NRDG and benefited noise reduction] x [7 NRDG quantities (1 numeric value + 3 first-row percentages + 3 all-benefits percentages)] x [3 CE values].

3.2.3 Testing Protocol

The test method for each study area involved several steps. First, the study area was rerun in the FHWA TNM with each residence represented by a receiver point that was placed in close proximity to the residence, representing the area of frequent human use. All previous noise barrier designs for the study area were deleted. The barrier segments' height range was set from 10 feet to 26 feet in 2-foot increments. The traffic volumes and speeds for the original TNM run were not changed.

New barrier designs were then developed based on a protocol of designing to meet the NRDG. An important question had to be addressed in developing the design protocol for the study areas: Is achieving the NRDG the ultimate goal of the barrier design, or should the barrier be designed using other goals (i.e. protect all impacted first-row receptors) with the NRDG tested as a minimum threshold for reasonableness? This is an important distinction that is not specifically addressed in the regulation or guidance. However, the interpretation of NRDG has significant implications for barrier design and reasonableness determinations.

As discussed previously, the NRDG is met under 21 SHA policies if the barrier provides the NRDG at one or more benefited residences. Numerous different barrier designs could be developed that would meet this NRDG. Whether these designs provide adequate noise reduction for impacted receptors would depend on the SHAs design philosophy and not the NRDG which simply requires achieving the design goal at one or more benefited receptors.

For example, if the NRDG is 7 dB at only one benefited receptor, an analyst could design a barrier for a single receptor at the far end of the community or for only the residence with the highest sound level and then modify the design until enough benefits are achieved to meet the CE criterion. This design does not consider the number and locations of impacts and might provide very small noise reductions for impacted receptors. While this approach would be contrary to the intent of the regulation, one could argue that such design meets the "letter" of the regulation. This example also illustrates that the NRDG does not address how the barrier should be designed. One SHA (VDOT) reports this "mindset" on design-build projects. To ensure that the spirit of the regulation remains intact, VDOT added new language to its guidance manual for consideration in all projects. VDOT's noise specialist notes that, similar to other environmental mitigation features, the goal for noise abatement is to achieve 100% effectiveness, i.e., provide benefits to every impacted noise sensitive receptor. However, as the specialist notes, it is not always possible to benefit every impacted receptor for different reasons.

The noise barriers for the study areas were designed to provide the specified noise reduction for all of the first-row impacted receptors and not just the number or percentage specified by the NRDG. Additionally, the NRDG is based on the number of benefited receptors. However, SHAs do not typically design forbenefited receptors, which are the outcomes of design.

When a design was developed that met the NRDG criterion, the APBR criterion was tested. If the APBR criterion was not met, the barrier height was increased in an attempt to increase the benefits enough to lower the APBR to meet the criterion value.

This protocol assumes a philosophy of trying to provide abatement where impacts exist, which is in keeping with the stated intent in the regulation:

"Sec. 772.13 Analysis of noise abatement. (a) When traffic noise impacts are identified, noise abatement shall be considered and evaluated for feasibility and reasonableness. The highway agency shall determine and analyze alternative noise abatement measures to abate identified impacts by giving weight to the benefits and costs of abatement and the overall social, economic, and environmental effects by using feasible and reasonable noise abatement measures for decision-making." [emphasis added]

The focus is on the first-row because designing to meet the NRDG for more distant receptors generally increases barrier heights over a first-row design, raising the noise reductions for first-row receptors to well above the NRDG. While such a strategy might be employed for areas with very high first-row levels (e.g., worst-hour L_eq 75 or more dB(A)), it would not be commonly used for most typical situations (e.g., in the 66 to 74 dB(A) range).

3.2.4 Method for Presenting Results

Each study area and its associated results are described below. Two tables are provided for each study area: Feasibility Results (Tables 20, 22, 24, and 26) and Reasonableness Results (Tables 21, 23, 25, and 27).

Each row in each table represents a "case" for the study area (i.e., a particular combination of the criteria and their application to a particular barrier design) and includes the criteria used in the design and the results.

The Feasibility Results table includes the factor "Feas Type." This factor has three possible values (#, P1, and PA) that indicate whether feasibility is based on a number (#), percentage of first-row (P1), or percentage of total impacted receptors (PA). The results include the number of total and first-row impacted receptors (# Impacts ALL, # Impacts 1R), and, depending on Feas Type, the number or percentage of total or first-row impacted receptors that meet the feasibility noise reduction (Feas # 1R, Feas # ALL, Feas % 1R, Feas % ALL). The last column in the table shows the feasibility decision for each case.

Similarly, the Reasonableness Results table includes the factor "NRDG Type." This factor also has three possible values (#, P1, and PA) that indicate whether the NRDG criterion is based on a number (#), percentage of first-row (P1), or percentage of total benefited receptors (PA). The calculated barrier area (Area) is shown. The results include the number of total and first-row benefited receptors (# Ben 1R, # Ben ALL), and, as appropriate, the number or percentage of total or first-row benefited receptors that meet the NRDG criterion (# Ben NRDG 1R, # Ben NRDG ALL, Ben % 1R, Ben % ALL). The decision for NRDG is then shown. The last three columns of the table show the calculated APBR (APBR Calc), whether the calculated APBR is above (No) or below (Yes) the APBR criterion and the overall reasonableness decision (Yes, if both the NRDG and CE criteria are met).

3.2.5 Study Area A Description and Results

Study Area A represents a proposed widening to a ten-lane highway. The road is on a slight fill and there are several rows of relatively closely spaced houses, as shown in Figure 21.

This project is one for which abatement should be feasible and reasonable because of the number and density of the receptors.

Figure 21. Study Area A: Ten-lane highway with dense development close to the highway.

Table 20 shows the feasibility results for Study Area A, and Table 21 shows the reasonableness results for Study Area A.

A barrier was feasible for all seven cases of the tested feasibility factors, an expected result given the density of houses and closeness to the road. Even for Case 7 - 75% of all impacts - enough non-first-row impacts received a noise reduction of at least 5 dB.

A barrier was reasonable for all of the cases for the 7 dB NRDG with a 5 dB benefited noise reduction (Cases 1 through 21), even for Case 21 for 75% of all benefited receptors. All of the APBR were under 800 SF/benefited receptor, a result of the large number of benefits. Note that while the APBR values are in a small range, the barrier area increases nearly 25% from 68,000 SF for the easiest case (Case 1) to nearly 84,000 SF for the most difficult case (Case 21). This increase in area is due mainly to the increase in height, which, on this project, resulted in adding enough benefits to offset the increase in area and keep the APBR in a tight range.

While not shown, the results of the increased height are first-row noise reductions well above the NRDG of 7 dB. It is a matter of design and abatement philosophy as to whether an SHA would push the barrier height up in this manner as long as benefits were still being accrued. Some SHAs might be reluctant to do so; instead they might design a barrier that just meets the 7 dB NRDG criterion, then test that design for CE and draw a conclusion on reasonableness. With such an approach, many of the above cases would result in a decision of "not reasonable." Choice of a philosophy also has critical cost implications, as evidenced by the 25% increase in barrier area from Case 1 to Case 21.

Cases 22 through 42 are for the 7 dB NRDG with a 7 dB benefited noise reduction. All of the cases with the 800 SF/benefited receptor APBR criterion were unreasonable; the cases with the higher APBR were reasonable. These barriers were all in the 74,000 to 84,000 SF range, and the APBR could not be reduced to below 932 SF/benefited receptor. The need to achieve 7 dB in order to count a receptor as benefited required too tall of a barrier to keep the barrier area low enough to reduce the APBR below the 800 SF/benefited receptor criterion.

Cases 43 through 63 are for the 9 dB NRDG with a 9 dB benefited noise reduction ("9 dB/9 dB"). As with Cases 22 through 42, all of the cases with the 800 SF/benefited receptor APBR criterion were unreasonable. While the 9 dB reduction could be achieved at 100% of the benefited receptors, the "penalty" in increase barrier area prevented the reaching of the 800 ABPR value and just barely achieved the 1,250 APBR criterion with a value of 1,249. The barrier height had to be increased substantially to achieve the 1,250 SF/benefited receptor criterion as evidenced by the necessary total area exceeding 100,000 SF. In the process of raising the height, however, more noise reduction was provided.

Cases 64 through 84 are for the 10 dB NRDG with a 5 dB benefited noise reduction ("10 dB/5 dB"). Contrasting with the 9 dB/9 dB cases above, two of these cases (Cases 64 and 67) met the 800 SF/benefited receptor APBR criterion. The NRDG quantities in these two cases were small - one benefited receptor and 25% of first-row benefits. With only a 5 dB benefited noise reduction criterion instead of 9 dB, enough extra benefits were realized to drive down the APBR for these cases. For the higher NRDG percentages, while there were enough benefited receptors to meet the NRDG criterion, there were not enough to meet the 800 APBR criterion. The minimum barrier area for a reasonable design for the 10 dB/5 dB criteria was 72,000 SF. For the higher NRDG percentages with the higher APBR, the barrier area jumped to 105,000 SF. Contrast this amount with the 68,000 SF needed to meet some of the reasonable 7 dB/5 dB designs.

Also, the last three cases for 10 dB/5 dB - where the NRDG criterion was 75% of all benefited receptor - were not reasonable because the NRDG criterion could not be met. In two of these cases, the APBR criterion could not be met either.

Based on experience, it was expected that a noise barrier would be reasonable for Study Area A . The lessons from Study Area A, with the densely developed residential community, include:

• Feasibility was easily demonstrated, even at 75% of all impacted receptors.
• A 7 dB NRDG with a 5 dB benefited noise reduction resulted in reasonable designs regardless of the NRDG quantity and with calculated APBR as low as 756 SF/benefited receptor. If one is satisfied with 7 dB noise reductions, then these criteria can be used successfully with APBRs as low as 800 SF/benefited receptor, even with high NRDG percentages.
• A 7 dB NRDG with a 7 dB benefited noise reduction resulted in "not reasonable" decisions for the 800 SF/benefited receptor APBR cases, regardless of the NRDG quantity. Use of 7 dB/7 dB suggests an APBR of at least 1,000 SF/benefited receptor, which would give reasonable designs even for high NRDG percentages.
• A 9 dB NRDG with a 9 dB benefited noise reduction would require an APBR of at least 1,250 SF/benefited receptor and potentially a bit higher to result in a "reasonable" abatement decision for this project. Moving up to 9 dB for the NRDG increases barrier cost because of the increased height but provides greater noise reduction at the most impacted receptors.
• The 10 dB NRDG with a 5 dB benefited noise reduction would require an APBR of at least 1,300 SF/benefited receptor for reasonable designs, but only if the NRDG quantity was kept to no more than 50% of all benefited receptors. This combination resulted in the greatest barrier areas needed in order to reach the 10 dB NRDG. A noise reduction of 10 dB also affords more protection to the impacted residents.

Table 21. Reasonableness Results for Study Area A: Ten-Lane Highway with Dense Development Close to the Highway.

3.2.6 Study Area B Description and Results

Study Area B represents a proposed widening to a six-lane highway facility. The road is at-grade with the community. There are two rows of closely-spaced houses that angle away from the road, with two other houses closer to the road than those in the rows, as shown in Figure 22.

For this project, feasibility and reasonableness are both uncertain due the relatively limited number of receptors and the manner in which they angle away from the road.

Figure 22. Study Area B: Six-lane highway with two rows of closely-spaced houses angling away from the highway.

Table 22 shows the feasibility results for Study Area B, and Table 23 shows the reasonableness results for Study Area B.

Table 22. Feasibility Results for Study Area B: Six-Lane Highway with Two Rows of Closely-Spaced Houses Angling Away from Highway

A barrier was feasible for all seven combinations of the feasibility factors, even for Case 7 - 75% of all impacts - which required extra barrier height to reach the 75% value.

A barrier was not reasonable using the APBR criterion of 800 SF/benefited receptor for any of the combinations of NRDG and benefited noise reduction (Cases 1 through 84).

For the 7 dB NRDG and 5 dB benefited noise reduction in Cases 1 through 21, barriers were reasonable for the 1,250 and 2,600 APBR criteria for all options for the NRDG quantity. Area requirements were in the 21,000 to 27,000 SF range for these designs.

Cases 22 through 42 for 7 dB NRDG and 7 dB benefited noise reduction, and Cases 43 through 63 for 9 dB NRDG and 9 dB benefited noise reduction were also unreasonable using the 1,250 SF/benefited receptor APBR criterion. For the 7 dB/7 dB cases, 1,343 SF/ benefited receptor was the lowest APBR for any of the designs; and for the 9 dB/9 dB cases, 2,524 SF/ benefited receptor was the lowest value. However, all of the cases for those two combinations were reasonable for the 2,600 SF/benefited receptor APBR criterion. Square footage requirements were in the 27,000 to 28,000 SF range for the 7 dB/7 dB cases and approximately 33,000 for the 9 dB/9 dB cases. The extra area for the latter cases was necessary to achieve the higher 9 dB NRDG, although fewer receptors were actually counted as "benefited" due to the higher benefited noise reduction criterion.

For the 10 dB NRDG and 5 dB benefited noise reduction in Cases 64 through 84, most of the cases using the 1,250 and 2,600 SF/benefited receptor APBR criteria were reasonable. However, for Case 81, requiring 50% of all benefited receptors to meet the NRDG, only the 2,600 APBR criterion was met. Also, none of the cases requiring 75% of all benefited receptors to meet the NRDG were reasonable for any of the three tested APBR criteria (Cases 82 through 84). The barrier area for the reasonable case was as low as 30,000 SF when only one benefited receptor or only 25% of the first-row benefited receptors had to meet the 10 dB NRDG. The needed area increased to 37,000 when 50% of all benefited receptors were required to meet the NRDG. For the unreasonable "75% of all" cases even an area of nearly 55,000 SF was not enough to meet the NRDG at the needed number of receptors.

Study Area B was one where the expectation, based on experience, was that a noise barrier was probably necessary based on the high predicted no-barrier levels for the receptors closest to the road. The close spacing of the receptors supported a decision of "reasonable," while the fact that the rows of houses angled away from the road raised the question as to whether sufficient benefits could be obtained. The mixed results bore out these considerations. The lessons from Study Area B include:

• Feasibility was easily demonstrated for all cases.
• An APBR of 800 SF/benefited receptor was too low for reasonableness in any case, even for a 7 dB NRDG and 5 dB benefited noise reduction with only a single benefited receptor needing to meet the NRDG.
• An APBR of 1,400 SF/benefited receptor is probably the lowest that would result in positive decisions on reasonableness for an NRDG of 7 dB.
• An APBR of 2,600 SF/benefited receptor is probably the lowest that would result in positive decisions on reasonableness for an NRDG of 9 dB when the benefited noise reduction is also high.
• However, a lower APBR of 1,300 SF/benefited receptor would appear sufficient a high NRDG of 10 dB when the benefited noise reduction criterion is kept low.
• Requiring a high percentage of all benefited receptors to meet the NRDG is likely to result in a decision of "not reasonable" when the NRDG is high.
• Requiring only one benefited receptor or even up to 50% of first-row or all benefited receptors to meet the NRDG is likely to result in a decision of "reasonable" regardless of the NRDG.

3.2.7 Study Area C Description and Results

Study Area C represents a proposed widening to an eight-lane Interstate facility. Figure 23 shows the area. The road is in slight cut and there are groupings of houses on two curved residential streets with open area between them.

For this project, feasibility and reasonableness are somewhat uncertain due to the moderate residential density and the fact that the road is primarily in cut.

Figure 23. Study Area C: Ten-lane highway with medium-density houses curving away from the highway.

Table 24 shows the feasibility results for Study Area C, and Table 25 shows the reasonableness results Study Area C.