Figure 3.1 Study Area
David E. Coate
Director, Noise and Vibration Control
Acentech Incorporated
33 Moulton Street
Cambridge, MA 02138
(617) 499-8019 email: dcoate@acentech.com
As a result of the Union Pacific/Southern Pacific (UP/SP) merger, freight rail operations will change along some of the existing corridors. Environmental impacts may occur in the communities along corridors where increases in freight rail operations are proposed. This paper presents the use of Esri's ArcView/ArcCAD to assess the noise impacts and mitigation options along the existing Union Pacific corridor through Wichita, Kansas.
Prepared for the Eighteenth Annual Esri International User Conference
July 27-31, 1998
San Diego, California
As a result of the Union Pacific/ Southern Pacific (UP/SP) merger, freight rail operations will change along some of the existing corridors. Environmental impacts may occur in the communities along corridors where increases in freight rail operations are proposed. This study presents an assessment and description of mitigation options to address the related concerns of community noise impacts along the existing Union Pacific corridor through Wichita, Kansas.
The Environmental Assessment (EA)(1) for the merger indicated that the rail line extending from Lost Springs to Wichita, Kansas would experience an increase in freight rail traffic of 10 trains per day. This increase is in excess of the Surface Transportation Board (STB) analysis threshold(2) of 8 trains per day. According to STB rules, environmental analyses are required for areas where this threshold is surpassed.
This study was initiated to 1) identify and quantify noise impacts, and 2) recommend appropriate mitigation measures where feasible and reasonable. A Geographic Information System (GIS) was employed to determine both noise impacts and mitigation measures.
The study area includes sensitive receptors (e.g., residences, schools, churches, and hospitals) along the existing Union Pacific freight rail corridor that extends from Haysville to Kechi, Kansas and is shown in Figure 3.1.
Figure 3.1 Study Area
Under the proposed merger, UP/SP would continue to operate through downtown Wichita on a north-south route that crosses 24 streets at grade between and including 21st street to the north and Pawnee Street to the south. The increased train movements on UP's Lost Springs to Wichita rail line would produce corresponding increases in horn, locomotive engine and exhaust noise, and wheel/rail noise.
In addition to through trains, other train activity in Wichita includes local and yard train movements. This study includes the noise effects of all of these train movements.
Environmental noise issues analyzed in this study include the effects of train horns, wheel/rail, and diesel locomotive noise. In the vicinity of grade crossings, the area of train horn noise impact for this project can extend as far as 400 feet from the track into the adjacent community, while wheel/rail noise impacts typically extend up to 80 feet from the track. These impact distances depend on a number of factors, including train speed, length, and the number of trains per day.
The basic criterion for evaluating a noise mitigation is the number of noise-sensitive receptors that would be no longer be impacted as a result of implementing the mitigation option. Section 8 presents various mitigation options and resulting number of affected noise sensitive receptors, as well as a discussion regarding the related costs.
Although the Union Pacific corridor in Wichita has been carrying freight rail traffic for many years, the proposed changes in train operations will affect the noise environment along the corridor. These changes in train operations were assessed based on the projected noise increase relative to pre-merger train operations.
Noise contours for both pre-merger and post-merger conditions were developed by use of a computer model designed for this project. The computer model was based on noise measurements conducted in Wichita, as described below in Section 7.3. Noise contours were incorporated into ArcView/ArcCAD to provide detailed information regarding noise-impacted land uses (e.g. residential, commercial, etc.).
The day-night average sound level (Ldn) is the noise descriptor that is used for this study. The Ldn is the time-average of the A-weighted noise levels obtained over a 24-hour period, with a 10 dB upward adjustment added to the nighttime levels (10 pm to 7 am). This adjustment is intended to account for the increased sensitivity to nighttime noise events. Appendix A, "About Noise", presents additional information regarding various measures of noise.
For this study, an adverse noise impact is deemed to occur when there is an increase to a noise level of 65 dBA Ldn or greater. These criteria apply to noise-sensitive receptors (e.g. residences, schools, libraries, retirement communities, and nursing homes) that are in the project area.
This section summarizes the results of train noise measurements conducted by Acentech Incorporated in Wichita, Kansas during the week of 17 March 1997. The purpose of these measurements was to verify and refine the assessment of pre-merger noise conditions, and to obtain a basis for modeling post-merger conditions. The noise measurements in this study take into account site-specific sound issues, such as actual train/horn equipment, shielding due to buildings, ground absorption, and the variability of train horn sounding sequences.
Measurements were made of the ambient noise (i.e. the noise environment without trains) and of noise associated with train passages. Train passage measurements were made at six locations at two distances along a line extending perpendicularly from the tracks in order to characterize the site-specific sound issues described above. The locations were chosen to be representative of suburban (with some building shielding) grade-crossing (horn noise) and non-grade crossing (no horn noise) areas. Train noise data for these two conditions were deemed to be sufficient to characterize the entire study area and to quantify how train noise decreases with distance, shielding, and ground effects.
The train noise measurement locations are identified below in Table 7.3.1:
| Location | Type | Nominal Measurement Distance from Tracks (feet) |
| 1. 71st Street, Haysville | Suburban grade crossing-with some shielding | 100, 200 |
| 2. 59th and Southern | No horns | 50, 100 |
| 3. 53rd Street | No horns | 75, 150 |
| 4. Pawnee Street | Suburban grade crossing-with some shielding | 100, 200 |
| 5. Osie Street | Urban/industrial grade crossing-with some shielding | 100, 200 |
| 6. 61st Street, Kechi | Suburban grade crossing-no shielding | 100, 200 |
Single-event Sound Exposure Level (SEL) data for each train noise event were used to determine how train noise decreases with distance for each of the measurement locations. The SEL permits comparison of the noise from events of different durations and is useful here since it takes the propagation of sound from the train to the measurement position for the entire train noise event, not just for the loudest portion of the noise event.
The rates of decrease of noise with distance calculated for each location identified in Table 7.3.1 for every measured train noise event were used to determine the distance from the tracks to the 65 dB Ldn contour.
Using the results of the noise measurements and analysis presented in Section 7.3, noise projections were made for both pre-merger and post-merger conditions along the entire rail corridor. Since the number and type of rail operations for both the pre-merger and the post-merger vary considerably along the corridor, the resulting noise impacts vary along the corridor. A table showing the distance from the track to the pre-merger and post-merger 65 dBA Ldn noise contour at each grade crossing was developed and incorporated into the GIS.
The results of the table described in Section 7.5 were incorporated into ArcCAD using interpolated buffer distances at small increments of mileposts along the track. In this fashion, numerous alternatives such as pre- and post-merger, with and without grade separations (i.e., under or overpasses), etc. could be evaluated in terms of noise impacts. These noise contours were overlaid on parcel-based GIS data provided by Sedgwick County. The GIS data includes actual property boundaries and detailed land-use information. Polygon intersection analyses with parcel data and noise contours were performed to determine which parcels were affected by each alternative. This technique allowed a detailed and quantitative method of determining noise impacts and associated mitigation costs. Figure 7.6 shows an example of this analysis. The expanded noise contours at grade crossings are due to locomotive horn soundings. The pink portion is the 65 dBA Ldn noise contour associated with pre-merger conditions, while the red band is the incremental increase associated with post-merger conditions
Figure 7.6 Example Train Noise Contours
Table 7.7 shows the results of an analysis that was done to determine the extent of the un-migitigated noise impacts and improvements derived from speeding up the trains and from constructing grade-separations. This particular combination of mitigation options was selected for analysis primarily to address traffic congestion issues.
Table 7.7 Number of Impacted Noise-Sensitive Receptors
| Condition | Number of Impacted Noise-Sensitive Receptors |
| Pre-merger | 313 |
| Post-merger (un-mitigated) | 391 |
| Speeding up trains | 427 |
| Pawnee grade separation and speeding up trains | 400 |
| Pawnee & Central grade separations and speeding up trains | 398 |
Speeding up trains has no effect on train horn noise impacts, but does affect wheel/rail noise impacts, since wheel/rail noise increases with increased train speed. As can be seen from Table 7.7, the Pawnee grade separation reduces the number of impacts by 27 because horn noise would be eliminated at that location, but also adding the Central grade separation only reduces the number of impacts by two more because there are 2 noise-sensitive locations in the vicinity.
Mitigation options include grade separations, building sound insulation treatments, local grade crossing warning devices (i.e. a horn sound source located at grade crossings), source noise controls (to reduce wheel/rail and diesel engine noise), noise barriers, four-quadrant gates and quiet zones, median barriers and quiet zones, and train horn modifications. The following is a brief description of several of these mitigation options and of the feasibility for this project.
8.1 Grade Separations
"Grade separation" refers to an overpass or underpass. Train horn noise could be eliminated at grade crossings where this option is employed. The cost of constructing a grade separation is substantial (on the order of several million dollars). Consequently, construction of a grade separation would be substantially more costly than other noise mitigation options, such as building sound insulation. However, if a grade separation were constructed to alleviate traffic impacts, noise reduction benefits would follow. In order to identify which grade separations would provide the best "bang for the buck," the GIS was used to show which parcels would be impacted by noise with and without a grade separation. Table 8.1, below, shows a list of grade separations, starting at the top with the most-heavily impacted location and ending with the least impacted location.
Table 8.1 Noise Impacts at Grade Crossings
| Grade Crossing | Number of Impacted Receptors |
| Pawnee | 61 |
| 71st | 47 |
| Kincaid | 32 |
| 63rd | 31 |
| Boston | 26 |
| Skinner | 23 |
| Mt. Vernon | 23 |
| Osie | 21 |
| Fundstrom | 20 |
| 55th | 20 |
| Clark | 18 |
| 117th | 18 |
| Harry | 15 |
Assuming a cost of $5,800,000 for a grade separation at Pawnee, the cost per protected dwelling unit would be $95,000. This figure demonstrates that grade-separations are not cost-effective if considered solely for noise mitigation.
8.2 Building Sound Insulation
Building sound insulation refers to improving the noise attenuation characteristics of a building envelope in order to reduce the intrusion of outdoor noise into the building. Sound insulation treatments usually involve improving the noise insulation characteristics of windows, since windows are usually the weak acoustical link. Special acoustical windows or modifications to existing windows can provide up to 10 dBA increased noise reduction. Nominal sound insulation treatment costs are on the order of $10,000-$20,000 per dwelling unit, depending on air-conditioning costs. Unlike other mitigation options, this building sound insulation obviously only can reduce the noise inside buildings.
An analysis was performed to estimate the order-of-magnitude cost for sound insulating impacted buildings for this project. The GIS was used to create a detailed list of impacted properties and the estimated cost of sound insulation for each property. The total of sound insulation treatments for all of these properties amounts to approximately $8,000,000.
8.3 Local Grade Crossing Warning Devices
The Federal Railroad Administration (FRA) and the Union Pacific Railroad have been assessing the viability of alternative local grade crossing warning devices, such as locating a horn or loudspeaker at the grade crossing. The benefit of such a device would be to limit the extent of the impacted community. Currently, train horns are sounded 1/4 mile from a grade crossing, resulting in noise exposure to residences in a fairly large area. Since the sole purpose of the horn is to warn motorists and others at the crossing, a device that delivers horn noise only to the area at or near the crossing is preferable.
The FRA has tested a prototype automated horn system (AHS) designed to increase the warning effectiveness at grade crossing while minimizing community noise impact. The system consists of a single electronic horn placed directly at a grade crossing and directed along approaching roadways. Since the horn is located at the grade crossing, the community noise exposure due to horn noise on a moving train is eliminated. The directivity of the system results in sound levels that are higher directly in front of the horn and lower to the rear and the sides. Consequently, not only is the area of community impact reduced, but the horn is more effective because of its greater audibility to motorists further down the road.
The 65 dBA Ldn contour footprint for the AHS is much smaller compared with the contour for conventional locomotive horn sounding. In fact, the noise contour for the AHS easily fits within the conventional locomotive horn noise contour. The GIS was used to determine which parcels would be within the 65 dBA Ldn noise contour with and without the AHS at each grade crossing. Figure 8.3 shows an example of this analysis. The parcels highlighted in blue are those affected by the conventional moving locomotive horn; while the "figure eight" noise contour associated with the ahs touches only two parcels. If the AHS were used at all of the grade crossings listed in Table 8.1, the number of impacted receptors would be reduced to 152 properties.
The approximate cost of an AHS installation at a grade crossing is $12,000-$15,000 (3). The range of costs depends on whether or not the road is two-lanes or a divided highway; this affects the complexity of the installation. This cost assumes that the crossing is state-of-the-art with appropriate circuitry for the AHS.
Other considerations regarding the AHS include its effectiveness as a warning device in comparison with a conventional train horn.
Figure 8.3 Noise Contour of AHS Fixed Horn versus Moving Locomotive Horn
8.4 Source Noise Control
Source noise control refers to reduction of noise at the source; in the case of freight trains, source noise controls apply to wheel/rail and diesel engine noise. Wheel/rail noise can be reduced by wheel truing and rail grinding. Diesel engine noise may be reduced by improved exhaust silencer technology or active noise control. Active noise control consists of adding loudspeaker-generated sound pressures to those resulting from engine noise so that these pressures cancel each other, resulting in greatly reduced noise. The Federal Railroad Administration (FRA) is currently conducting a study of the feasibility of active noise control for diesel locomotives.
Source noise controls could reduce the area of impact in regions where impact is not due to horn noise. However, since the impacts in these regions are very limited, source noise controls would only have a minor benefit for this project.
8.5 Noise Barriers
Noise barriers are effective for reducing wheel/rail noise that reaches the community. Because train horn noise is likely the dominant noise source, noise barriers would be useful only in those areas where horn noise is not present.
An analysis was conducted using the GIS to determine whether noise barriers would be effective for impacted locations in the study area. The three locations where noise barriers were found to be potentially useful are shown in Table 8.5, together with related cost estimates.
| Location | Approximate Barrier Length (ft) | Estimated Number of protected dwelling units | Approximate Cost | Approximate Cost/dwelling unit |
| Greenway St. North of Wassall St. | 900 | 6 | $135,000 | $23,000 |
| Between 44th and 43rd St. | 1400 | 8 | $210,000 | $26,000 |
| Turkle Avenue, Haysville | 2600 | 32 | $390,000 | $12,000 |
8.6 Four-Quadrant Gates and Quiet Zones, Median Barriers and Quiet Zones
The FRA currently is considering the use of four-quadrant gates or median barriers, both of which are designed to keep motorists from driving around the crossing gate arm as a train approaches. This approach could eliminate train horn noise (hence the term "quiet zone") at specific grade crossings as do grade separations, at a much lower cost; however it does not alleviate vehicular traffic congestion.
Quiet zones are not practical in the central part of Wichita, where horn soundings from nearby adjacent railroad tracks would substantially reduce the effectiveness of the quiet zone. Quiet zones might be feasible at grade crossings south of Wichita, such as at 71st Street in Haysville. The associated costs would include those for upgrading the crossing to a four-quadrant crossing gate or those costs associated with installing median barriers. Quantification of these costs was not part of this study.
8.7 Train Horn Modifications
Train horn types include the three-chime (i.e. three discrete tones) Leslie and five-chime Nathan. One of the Leslie horns, the RSL-3L-RF (used on the Union Pacific GE Dash-8 locomotive), is more efficient toward the sides and rear of the engine than the front. The sound level measured in front of the engine is approximately 6 dBA lower than that at the sides and 8 dBA lower than that measured at the rear. Therefore, in order to meet the FRA's requirement of 96 dBA measured at 100' in front of the engine, this particular horn produces 102 dBA at the sides and 104 dBA at the rear of the engine. Clearly, if the goal is to direct as much sound energy to the front of the engine, this particular horn design is not optimal. This example shows that substantial mitigation might be achieved by assessment and modification of the horn designs.
Reducing horn noise at the source could be cost-effective even if a large number of locomotives were involved. There are an estimated 23,000 locomotives (4) in the U.S.; a fraction of this number are Dash 8 locomotives. The cost associated with this mitigation option which would include the re-design and installation of the horn system, was not quantified as part of this study.
Table 8.7, below, lists the mitigation options described above, and indicates the total number of noise-sensitive locations impacted if the option is in place, and shows the total cost of the mitigation.
| Mitigation Option | Number of properties/buildings impacted | Order-of-Magnitude Cost |
| Pre-merger | 313 | NE |
| Post-merger (unmitigated) | 391 | NE |
| Speeding up trains | 427 | NE |
| Pawnee grade separation and speeding up trains | 400 | $5,800,000 |
| Pawnee and Central grade separation and speeding up trains | 398 | $12,000,000 |
| Building Sound Insulation | 0 | $8,000,000 |
| Local Grade Crossing Warning Device | 152 | $245,000 |
| Source Noise Controls (Improved silencers and active controls) | NE | NE |
| Noise Barriers (for wheel/rail noise) and speeding up trains | 381 | $735,000 |
| Four-quadrant gates or median barriers/quiet zones | NE | NE |
| Train horn modifications | NE | NE |
8.8 NOISE MITIGATION RECOMMENDATIONS
Noise mitigation for this project falls into two different categories: 1) mitigation aimed at other environmental impacts that happen to also reduce noise, and 2) mitigation aimed exclusively at noise. Examples of 1) include the construction of grade separations (underpasses or overpasses) which are aimed at reducing traffic impacts, but which also reduce noise due to eliminating the need for train horn sounding. This report presents the benefit (i.e., reduction in the number of impacted noise-sensitive receptors) of noise mitigation measures.
Recommendations for measures aimed exclusively at noise mitigation follow:
Noise Mitigation
State agencies that routinely design and build noise mitigation typically limit the extent of the noise mitigation to $10,000-$25,000 per protected dwelling unit. This interpretation of "reasonable" mitigation arises primarily because a residential property typically can be bought out at prices not much higher than this range.
Applying this general rule of "reasonable" to the mitigation options listed in Table 8.7 reduces the list to: sound insulation, noise barriers, and local grade crossing warning devices.
Ranking these three options in terms of cost effectiveness (with Option 1 being most cost-effective) yields:
Option 1 Local Grade Crossing Warning Devices (Directional Horn)
Option 2 Noise Barriers
Option 3 Sound Insulation
Noise impacts for this project could be reduced substantially by employing Options 1 and 2, which would reduce the number of houses to be insulated in Option 3.
Noise is usually defined as sound that is undesirable because it interferes with speech communication and hearing, or is otherwise annoying (unwanted sound). The characteristics of sound include parameters such as amplitude, frequency, and duration.
Sound pressure levels can vary over an extremely large range of amplitudes. The decibel (dB) is the accepted standard unit for measuring the amplitude of sound because it accounts for these large variations in amplitude and reflects the way people perceive changes in sound amplitude.
Different sounds may have different frequency content. Frequency content of a sound refers to its tonal quality or pitch. When describing sound and its effect on humans, A-weighted (dBA) sound levels are typically used to account for the response of the human ear. The term "A-weighted" refers to a filtering of the noise signal to emphasize frequencies in the middle of the audible spectrum and to de-emphasize low and high frequencies in a manner corresponding to the way the human ear perceives sound. This filtering network has been established by the American National Standards Institute (ANSI). The A-weighted noise level has been found to correlate well with peoples' judgments of the noisiness of different sounds and has been used for many years as a measure of community noise.
Community noise levels usually change continuously during the day. However, community noise typically exhibits a daily, weekly, and yearly pattern. To compare noise levels over different time periods, several descriptors have been developed. One descriptor, the equivalent sound level (Leq), is the equivalent steady-state A-weighted sound level that would contain the same acoustical energy as the time-varying A-weighted sound level during the same time interval. The hourly Leq is often used to describe traffic noise.
Another descriptor for noise is the statistical A-weighted noise level exceeded in a given percentage of the time. For example, the L50 is the level exceeded 50 percent of the time and the L10 is the level exceeded 10 percent of the time.
The Single-event Sound Exposure Level (SEL) is a noise descriptor that normalizes all of the sound energy of a noise event to a one second duration. The SEL provides a meaningful way to compare noise levels of two different noise events of different durations. The SEL is useful for calculating the drop-off rate in the present case since it takes into account the propagation of sound from the train to the measurement position for the entire train noise event, not just for the loudest portion of the noise event. In addition, the SEL in conjunction with the number of daytime and nighttime train noise events can be used to directly calculate the Ldn (the day-night average noise level used for identification of impacts in this study).
The day-night average sound level (DNL or Ldn), was developed to evaluate the total daily community noise environment. The Ldn is the time average of all A-weighted levels for a 24-hour period with a 10 dB upward adjustment added to the nighttime levels (2200 to 0700). This adjustment is an effort to account for the increased sensitivity to nighttime noise events. The Ldn noise metric has been adopted by federal agencies including the Environmental Protection Agency (EPA), the Federal Transit Administration (FTA), the Federal Aviation Administration (FAA), and the Housing and Urban Development (HUD) as the accepted unit for quantifying human annoyance to environmental noise.
When high noise levels are experienced inside or outside people's homes, as may occur from the passage of motor vehicles or the operation of mechanical equipment, a feeling of annoyance may result. These noise levels may also interfere with the performance of various activities such as conversation, TV watching, sleeping, etc. The degree to which there is annoyance and/or activity interference depends on the magnitude of the intruding noise level, the frequency with which it occurs and the time of day of occurrence. In response to the Noise Control Act of 1977, which directed the EPA to establish a recommended measure to describe community noise, the day-night average sound level (Ldn) was selected as the unit of measurement to be used to predict annoyance from noise exposure.