This portion of the study, 23 CFR 772 Streamlining, Analysis, and Outreach, examines the feasibility and reasonableness factors in the FHWA noise regulation in Title 23 CFR Part 772. Those factors are:
As a starting point, the focus was on the range of factors, individually and in combination as contained in the SHA noise policies in use throughout the country.
Then, a sensitivity analysis was performed to test over 1,000 combinations of the factors' values for over 100 cases of one-row, two-row, and three-row communities with receptors spaced at 50, 100, and 200 feet from each other. Two series of reasonableness decision arrays are presented - one considering only the NRDG criterion and one that considers both NRDG and the CE criterion in terms of the APBR.
Finally, the findings were tested on four actual highway project study areas previously evaluated for noise abatement. Seven sets of feasibility factor values were tested, as were 84 combinations of reasonableness criteria values.
All SHAs use 5 dB for the feasibility noise reduction. Feasibility is insensitive to the type of the quantity (number of impacted receptors, percentage of first-row impacted receptors and percentage of all impacted receptors) and its value. Positive decisions on feasibility were returned on all of the values considered, ranging from one impacted receptor to 75% of all impacted receptors.
While not directly studied in this task of this research, one concern of SHAs is the need to analyze abatement for cases of isolated single residences even though such cases will almost always result in a finding of "not reasonable" on the basis of CE. If one establishes in the policy that a noise barrier must achieve a minimum noise reduction of 5 dB at two or more impacted receptors, then any case of an isolated single residence would automatically return a decision of "not feasible" since there are not two or more receptors to be impacted.
In the first phase of this Study the 4 report titled Evaluation of 23 CFR 772 for Opportunities to Streamline and Establish Programmatic Agreements addresses this approach as a possible "streamlining" technique. It should be pointed out that such an approach would not work for an SHA that uses the feasibility criterion as a design mechanism instead of a screening mechanism. However, if the NRDG was changed to be based on impacted receptors instead of benefited receptors, as discussed in the next section, then a feasibility criterion based on a number of receptors could be used for isolated receptor screening for those SHAs. The counter-argument for screening out an isolated receptor - and not doing a reasonableness analysis - is that there can be the relatively rare case where a barrier for a receptor very close to the road will be reasonable in terms of both the NRDG and CE criteria
There are wide ranges of values used by the SHAs for each of the reasonableness factors studied. These ranges lead to very different decisions regarding abatement reasonableness for identical noise study area scenarios.
The decisions are not very sensitive to the NRDG and the NRDG quantity - up to a point. Meeting the NRDG for even a high percentage of first-row benefited receptors or a low-medium percentage of all benefited receptor was generally achievable. However, moving up to a requirement of over 50% of all benefited receptors results in "not reasonable" decisions for many cases that would otherwise be reasonable.
Having the NRDG quantity in terms of impacted receptors instead of benefited receptors would eliminate the problem of the number of benefited receptors changing in the design process as barrier height and/or length changes. Changing the number of benefited receptors changes the target number needed to meet a criterion based on percentage of all benefited receptors, complicating the design process. As an example of designing for impacted receptors, WSDOT uses its feasibility design goal, which needs to be achieved at greater than 50% of first row receptors, as a design criterion as well as a feasibility determinant.
Also, basing the NRDG decision on percentage of all benefited receptors can lead to counter-intuitive - and unfair - decision on reasonableness in cases where there are several rows of benefited receptors. A large number of benefited receptors coupled with a high percentage requirement for meeting the NRDG can lead to "not reasonable" decisions based on either or both of the NRDG and CE criteria. Yet, if those "extra" rows of houses were removed, then the total number of benefited receptors could not increase once all of the first-row receptors met the benefited noise reduction value, permitting the NRDG and CE criteria to be met. Either restricting the NRDG quantity to be a number or a percentage of first-row benefited receptors only or basing the NRDG on the number or percentage of impacted receptors will eliminate this problem.
Most SHAs use a CE criterion based on cost (in dollars) per benefited receptor (CPBR) and barrier surface area in square feet per benefited receptor (APBR). Several SHAs use sliding scales for determining the allowable values for either of these parameters. Some others use cost per benefited receptor per decibel of noise reduction.
The CE value in terms of CPBR for a project design is very dependent on the unit cost used by the SHA in developing that criterion and in computing the cost of a proposed abatement measure. As a result, the SHA must be very diligent in ensuring that any change in the unit cost is proportionately reflected in an adjustment to the allowable cost per benefited receptor (CPBR). The APBR approach factors cost out of the equation, as well as the ongoing need to keep unit costs and the CE criterion updated.
For the purposes of this study - in order to assess and characterize differences in the SHA policies - normalization for cost was performed (when abatement unit costs could be obtained from the SHAs) by dividing the CPBR by the unit cost to get an equivalent to the APBR.
The result is a very wide range in APBR criteria values, from the low end of the sliding scale at 250 SF/benefited receptor (for a case of in-fill development along a highway) to 2,750 SF/benefited receptor, a range of over 10:1.
The sensitivity analysis conducted in this research leads to the conclusion that the reasonableness decision is very dependent on the APBR. Values below 1,000 SF/benefited receptor for APBR resulted in "not reasonable" decisions even in many cases of dense receptor spacing. Conversely, high values of APBR resulted in very few "not reasonable" decisions, except when combined with a requirement for a high percentage of benefited receptors needing to meet the NRDG. APBR values no lower than 1,400-1,500 seem appropriate based on the results of this study.
This research has shown that there are a very large number of paths to the same decisions on reasonableness of abatement. There does not appear to a particular set of values for the various factors that is optimum. Instead, if extremes in the factors are avoided, then more uniform decisions from one SHA to the next and from one project to the next can be expected. Changing the feasibility quantity to two impacted receptors would eliminate the need to evaluate abatement unnecessarily in isolated receptor cases. Changing the NRDG quantity to be the percentage of first-row benefited receptors or impacted receptors would improve the design process and eliminate the problems with "not reasonable" decisions in study areas with multiple rows of houses. Not allowing low values for APBR or for normalized CPBR/unit cost will also lead to more consistent abatement decisions.
The charts in this report that display the results of the sensitivity analysis may help SHA analysts and policymakers understand the consequences of changes in the various factors covered by this study. The analysis tools resulting from this work, documented in a separate report, will allow the SHA analyst to examine the changes in decision-making for an individual project or a group of projects based on change in these factors.
As with the NRDG and CE/APBR reasonableness criteria, there are multiple factors and a wide range of values used by the SHAs when considering the viewpoints of the benefited property owners and residents. This variability can become more difficult to quantify when attempting to account for omissions, ambiguity, and nuances that exist in many of the policies. The unpredictability in voter response rate, which was demonstrated in Section 4.2 to be a determining factor, further adds to the complexity.
There are two main philosophies by which the barrier is judged to be reasonable:
Some SHAs base their decisions on the actual number of votes or responses that they receive. Others compute the percentage based on the total number of possible votes. Others are not specific about how the percentage is computed. Different decisions can easily be reached on reasonableness depending on whether the percentage is based on votes received or all possible votes. Regarding non-responses, some SHAs will make multiple contacts with the benefited receptors if they do not receive an initial response; others will not. The approach used can affect the decision that is reached.
Another important decision factor is whether a certain percentage of the benefited receptors are required to vote in order for the overall balloting to be considered valid. Some SHAs include such a requirement. In this case, it is more likely that a given barrier will not be built, especially if a strong effort is not made to get the eligible voters to respond.
The factor that has the most variation in values used by the SHAs is if and how the votes of the various stakeholders (typically, owner-occupant, non-resident owner and tenant) are weighted. Seventeen SHAs simply state that the views of the "benefited property owners and residents" will be obtained, without giving any indication of any weighting of the votes assigned to these two cohorts.
Others assign points or extra votes based on the stakeholder. Some policies give one vote to an owner-occupied dwelling, and one vote each to both a non-resident owner and the dwelling's tenant, resulting in a rental property having twice as many votes as an owner-occupied property. Giving rental units more votes over owner-occupied properties is not recommended by this research team.
For multifamily dwellings, most SHAs give the owner a vote for every rental unit in addition to a vote for every tenant; others try to give the tenants more say in the outcome.
Other weighting measures include additional votes or multipliers: 1) if a benefited receptor is in the "first" or "front" row adjacent to the highway; 2) if a benefited receptor is impacted in addition to being benefited; and in one case 3) based on the amount of noise reduction to be provided by the barrier.
Six SHAs were contacted for information on their experiences in implementing their Consideration of Viewpoints criterion. Implementation has been as learning process and several of the SHAs have refined their processes, while generally being satisfied that the voting results are leading to positive decisions on the abatement measure.
While the focus of this research was on residential land uses, it is noted that a third of the SHA policies contain language on the non-residential Activity Categories in 23 CFR 772. Examining these non-residential policy components might be a useful follow-up to this work.
Finally, to illustrate how differences in policies can yield different results for the same project, an exercise was conducted where nine different sets of Viewpoints factors were applied to six hypothetical voting "scenarios" for a relatively simple real-world study area. The results confirmed the overall conclusion of this research that the choice of factors and the values for those factors leads to different decisions.
The bottom line is that the reason for considering the viewpoints of the benefited property owners and residents is to provide abatement where it is wanted and not provide it where it is not wanted. It is concluded that SHAs have taken many different paths toward determining that intent. It is also concluded that these different paths will lead to different decisions on a given project with a given set of voting results simply based on the factors being considered and weights being assigned.
It is further concluded that many of the SHA policies are not specific enough in how votes are assigned, counted (including non-votes), and/or weighted. While ambiguity does offer flexibility, it also opens up the possibility of inconsistent application that could lead to challenges. Follow-up is recommended to see how the non-specific policies are actually applied by the SHAs. Then serious consideration should be given to modifying these policies to make them more specific. This report and the companion NAFRAT (Noise Abatement Feasibility and Reasonableness Analysis Tool) spreadsheet should be helpful in this regard, especially in the areas of weighting and counting votes.
It is also concluded that requiring an abatement measure to be voted down is a more difficult test than requiring a vote in favor of it simply because it is unusual to get 100% of the eligible voters to vote. Requiring a vote against an abatement measure is also based on the philosophy that abatement should be provided unless there is sufficient demonstrated opposition to it, a philosophy stated in several of the SHA policies.
Requiring a vote in favor to move forward is very much influenced by whether the percentage of positive votes is computed based on all of the possible votes or on only the actual votes received. Further, the decision is affected by how the non-votes are treated.
Should non-votes be considered as being in support of the measure, or are non-votes an indication that the non-voter either does not care or, less likely, is opposed? The preamble to the revised FHWA noise regulation in the July 13, 2010, Federal Register is clear regarding how non-votes should not be considered: "It is FHWA's position that the failure to respond to a survey may demonstrate lack interest in noise abatement, particularly when there is a low response rate from the community, but only explicit 'no' votes should be considered as 'no' votes." This position is restated and expanded in the Frequently Asked Questions (FAQ) on the Noise Regulations and Guidance page of the FHWA web site: "In order to obtain a majority of responses on which to base their decision, highway agencies should only consider votes that are submitted, and should not assume a 'no response' is a vote for or against the noise abatement" [italics added].
While none of the SHAs explicitly count a non-vote as a "no" vote, several do count a non-vote as a "yes" vote, which is counter to what is stated in the FAQ. It is concluded that these inconsistencies should be addressed.
As an example, if only four voters respond out of a possible 20 eligible voters and all four are in favor, should the sense of the community be taken as only 20% in favor (4/20) or as 100% in favor (4/4, or (4 + 16 non-votes taken as "yes")/20)?
Conversely if only four voters respond out of a possible 20 eligible voters and all four are opposed to the measure, should the sense of the community be taken to be as 100% opposed (4/4) or as 80% in favor (16 non-votes taken as "yes" divided by 20 total possible votes)?
In both cases, the reality is likely somewhere in between; yet "in-between" could be in favor or opposed. The wide swing in apparent support or opposition in this simple example illustrates the importance of how the viewpoints criterion is specified and then implemented.
Some SHAs attempt to address this ambiguity by having a desired response rate and will make multiple attempts to obtain a vote. The level of effort in trying to obtain a vote from each eligible voter can affect the outcome. Two SHAs gives eligible voters only one opportunity to vote and six give only two opportunities. It is concluded that a single attempt to obtain votes is insufficient if a low response means that abatement will not be provided.
It is further concluded that how the votes are weighted can play a critical role in the decision, such as giving an owner-occupant less votes than a non-resident owner and his or her tenant. Similarly, giving a multi-unit owner the same number of votes as the sum of the benefited tenants virtually assures that the owner's vote will always decide the matter regardless of what the tenants want. This is a philosophical issue of the whether those living in the noisy environment should have a real say in the decision to abate the noise they experience. One SHA addresses this inequity by allowing a positive vote by 75% of the tenants to override an owner's "no" vote.
A multi-unit owner can also have the deciding vote in a noise study area that is a mix of a multi-unit complex and owner-occupied single-family residences. Such an outcome could be viewed as unfair to the single-family owner-occupants who want abatement for both quality of life and value of property. In such a case, it is concluded that there should be some flexibility in the SHA policies to allow the noise study area to be broken into single-family and multi-unit areas that can be separately analyzed, even if it means that a barrier for the single-family area may have to be extended partially in front to the multi-family units for feasibility and reasonableness.
Finally, it is concluded that the use of weighting factors - such as giving more points to those benefited receptors who are impacted, are in the front row closest to the road, or will receive a large noise reduction - is an appropriate way to put the decision more into the hands of those most affected by the noise.
