Safety and Operations Analyses in Wilmington

Re Safety and Operations Analyses at Selected Boston Region MPO Intersections, FFY 2012: Main Street (Route 38/129) at Church Street/Burlington Avenue (Route 62) in Wilmington

Introduction

This intersection ranked 84th on the Highway Division’s 2007–2009 Statewide Top 200 Intersection Crash list. Based on the MassDOT crash database, in that three-year period, this intersection had 52 crashes, 20 of which caused personal injuries. There were three crashes that involved a pedestrian and one that involved with a cyclist.

In addition to the high number of crashes, the intersection was selected for its congested conditions during peak hours, its regional significance due to its being the intersection of two major state routes (Routes 38/129 and Route 62) , and its transit significance because of its being adjacent to a major commuter rail station.

This memorandum summarizes safety and operations analyses and proposes improvement strategies at the intersection and at its adjacent intersection on Main Street at the MBTA (Massachusetts Bay Transportation Authority) commuter rail station driveway. It contains the following sections:

The memorandum also includes technical appendices that contain methods and data that were applied in the study and detailed reports of the intersection capacity analyses.

Existing Conditions

The intersection is under MassDOT Highway Division District 4’s jurisdiction. It is located in the central area of Wilmington, where two major state roadways meet. State Route 38/129, locally called Main Street, connects Interstate 95/Route 128 in the south with Interstate 495 and Route 3 in the north. Route 62, locally called Burlington Avenue/Church Street, links Interstate 93 in the east and Route 3 in the west. In addition to carrying regional traffic, the two roadways function as urban major and minor arterials that serve a number of major destinations in the town. During the morning and evening peak periods, the intersection is usually congested.

The intersection is adjacent to a major railroad, and its traffic operation is affected by train preemptions at the railroad’s at-grade crossing on Main Street (known as Wilmington Junction) about 1,000 feet north of the intersection. The railroad, which connects the city of Boston with locations in Maine and New Hampshire, serves both MBTA (Lowell Line and Haverhill Line) and Amtrak (Downeaster Line) trains. Near the intersection, it runs parallel and just west of Main Street from about a quarter mile south of this intersection to the Wilmington Junction, where the Lowell and Haverhill tracks split and the Haverhill Line crosses Main Street.

The intersection is signalized, and its traffic signal is coordinated with the signal at the MBTA driveway, which is located about 750 feet north of this intersection. Both signals are interconnected to the railroad crossing signal at Wilmington Junction, which is located about 250 feet north of the MBTA driveway. When a train is crossing Wilmington Junction, both signals operate under train preemption mode, which immediately clears the southbound approach and stops the northbound approach at both intersections. Then, it allows only right turns from the MBTA driveway and allows only through traffic from Burlington Avenue and Church Street and traffic turning onto Main Street Southbound at this intersection Figure 1 shows the locations of the three intersections, the connected roadways, the railroad, and the surrounding area.

Due to its proximity to the railroad junction, the intersection is elevated to allow trains to cross under Burlington Avenue. Thus, three of the approaches to the intersection, except Burlington Avenue, are on an uphill slope. Near the intersection, Burlington Avenue is relatively level, as its elevation rises gradually from a point west of the bridge over the railroad bed. The Main Street northbound approach is somewhat steeper than the other approaches and is located on a horizontal curve. Drivers on this approach cannot see the signal heads until about 200 feet from the intersection. They do not have a clear view of the entire intersection and vehicles from the opposite approach until arriving at the intersection.

In addition, the intersection layout is somewhat skewed, as the two roadways do not intersect each other perpendicularly. The skewed layout makes turning movements at the intersection difficult to maneuver, and most of the turns require larger turning radii than at a regular intersection, especially turns from Burlington Avenue and Church Street. The intersection carries a relatively high percentage of heavy trucks, including some semi-trailer trucks, which require even larger turning radii than autos for changing direction within the intersection.

All the streets connected to the intersection are two-lane roadways, except Main Street north of the intersection. It has two southbound lanes and one northbound lane. Approaching the intersection, one of the two southbound lanes is designated as a left- turn/through lane and one as a through/right-turn lane. The other three approaches are wider—two lanes instead of one—in the sections near the intersection. The northbound and westbound approaches each have one left-turn-only lane and one through/right-turn lane. The eastbound approach has a left-turn/through lane and one right-turn-only lane.

Figure 2 shows the existing layout of the intersection and its surroundings. Based on the existing configuration, all of the approaches require only one receiving lane, except the southbound approach (Main Street). The southbound receiving lane is wide enough for two lanes of traffic in the beginning, but within about 100 feet, its width decreases to only about 12 feet. Therefore, the southbound vehicles are forced to merge immediately after they pass through the intersection. They frequently have to slow down or even stop to avoid collisions. Sometimes they back up into the intersection and prevent the vehicles behind them from passing through the intersection. In addition, because the roadway slopes downhill, the lane reduction is not entirely visible to southbound drivers before they enter the intersection. Although there is a “Lane Ends” warning sign in place just south of the intersection, drivers still have to react to it quickly due to its location within an unexpected short distance.

Sidewalks are installed on both sides of the streets at the intersection, except on Main Street south of the intersection. That segment has a sidewalk only on the east side. North of the intersection, sidewalks continue on both sides on Main Street until Wilmington Junction. Away from the intersection, sidewalks continue on only the north side of Church Street, and only on the south side of Burlington Avenue.

The surrounding areas have mixed land uses. There are various commercial developments on Main Street. Located mainly on the east side, they include a limousine company, a small strip mall with off-street parking, a gas station/garage, and a tire store to the south of the intersection, and a number of storefronts, an office building, a bank, a realtor, and a flower store to the north of the intersection. Two-hour on-street parking is allowed in front the storefronts. About 10 cars can park on a stretch of wide shoulder area with no clear marking of parking spaces. Field observations indicate that one or two vehicles frequently were parked too close to the intersection and obstructed turning vehicles from Burlington Avenue and Church Street.

On the west side of Main Street adjacent to the intersection are a sandwich shop (Big Joe’s) and a small, vacant convenience store (previously a Quick Mart store), with its own driveway and parking lot. Further north, the MBTA’s Wilmington commuter rail station and its parking lot occupy most of the area between Main Street and the railroad intersection, drivers still have to react to it quickly because it is located within an unexpectedly short distance from the lane drop.

Crosswalks are installed on all approaches of the intersection, except the westbound approach. It appears reasonable not to have a crosswalk at the approach because the approach is relatively wide, with no shops or houses on either side of the approach.

Pedestrians are able to reach their destinations by using the existing three crosswalks. Pedestrian signals with push buttons are provided for the crosswalks. An exclusive pedestrian signal phase of 22 seconds is included in the traffic signal timing plan for pedestrian crossing calls. The pedestrian signal is not equipped with a countdown or accessible (audible) function.

North of the railroad junction, Main Street is a recently developed commercial area that has a number of major retail stores, specialty stores, restaurants, and gas stations, and a popular supermarket.

On Church Street, there are a few shops and small office buildings near the intersection. Further east on Church Street the land use is mainly residential, with a few institutional land uses. Wilmington High School is located about a mile east of the intersection. On Burlington Avenue, there are access roads leading to the warehouses, office buildings, and an industrial park that occupy the area west of the railroad. The land use farther west is mostly residential, with single-family homes in a wooded area.

Approaching the intersection, Church Street has a speed limit of 35 mile per hour (mph). Burlington Avenue, approaching in a horizontal curve and then on an elevated bridge, has a speed limit of 25 mph. Neither of the Main Street approaches has speed limit signs posted. Presumably, the speed limit is 35 mph. The crash data analyses (detailed in a later section) showed that there were a relatively high number of rear-end crashes on Main Street from both approaches. The northbound approach has several commercial driveways and is located on a horizontal curve. The southbound approach has a commercial driveway near the intersection and is relative straight. During peak traffic hours, drivers tend to change lanes when they are blocked by vehicles waiting to turn left.

Because there are industrial land uses in the area, the intersection carries a relatively high percentage of heavy-vehicle traffic. The Main Street section in the vicinity of the intersection is a state-designated truck route. CTPS traffic counts from April 2012 indicated that heavy vehicles comprise over 5% of the total entry traffic in the morning peak hour (7:30–8:30 AM) and about 2% of the total entry traffic in the evening peak hour (5:00–6:00 PM).

Currently, there are no bike lanes on any of the approaches. However, cyclists use the intersection to reach the MBTA train station and other destinations. The April traffic counts indicated that five bicycles used the intersection in the morning peak hour and two during the evening peak hour. Presumably, the number is much higher during warmer days, for example, in the summer. The MBTA station is equipped with bike racks that can accommodate about 20 bikes. A field visit on a July 2012 weekday morning indicated that over 80% of the bicycle parking spaces were occupied.

The adjacent MBTA parking lot has 161 spaces for motor vehicles, and is usually over 85% occupied during weekdays.1 In addition to the park-and ride entries and exists, there are kiss-and-ride drop-offs and pickups at the station driveway during the morning and evening commuting hours. The April traffic counts indicated that there were about 120 entering and 70 exiting vehicles at the driveway in the morning peak hour, and about 50 entry and 140 exit vehicles in the evening peak hour. The counts include a few transit vehicles, such as buses on Route 12, which is operated hourly by the Lowell Regional Transit Authority, and passenger vans run by private companies for their employees and customers.

Issues and Concerns

During peak hours, traffic is heavy on almost all of the approaches at the intersection. In the morning peak hour, traffic is especially heavy on Main Street southbound and relatively heavy on Church Street. In the evening peak hour, traffic is especially congested on Main Street northbound and on Burlington Avenue.

Meanwhile, the intersection’s traffic operation is affected by the train preemption operations at Wilmington Junction, especially during the PM peak period, when the northbound traffic is heavy. There are two train preemptions in the PM peak hour (5:00–6:00 PM). Each of the preemptions lasts about one minute to one and a quarter minutes. During this period, the northbound traffic is blocked. It usually takes two to three cycles for the northbound queue to dissipate. At times, the northbound queue extends close to the upstream intersection at Lowell Street and affects its traffic operations.

The intersection has little room for expansion or adjustment due to its proximity to the railroad and the surrounding built-up conditions. Drivers approaching the intersection have a limited sight distance from almost all the approaches. Largely due to the congested conditions, the intersection has a much higher crash rate than other signalized intersections in the area—about three times the average rate for signalized intersections in MassDOT Highway District 4 (see the next section for detailed analyses).

Based on field observations and the available crash and traffic data, the issues and concerns related to this intersection can be summarized as:

Crash Data Analysis

The crash analysis was based on two sources of data: crash reports obtained from the Wilmington Police Department (WPD) and the MassDOT Registry of Motor Vehicles Division crash data, for the years 2007 to 2011. For 2007, most of the data were derived from the MassDOT crash database; for 2008 and 2009, most of the data were from both the MassDOT crash database and the crash reports; and for 2010 and 2011, the crash reports were the sole source of data. The crash diagrams were created from the WPD crash reports.

Table 1 shows that, on average, about 23 crashes occurred at the intersection of Main Street at Burlington Avenue each year. About 32% of the crashes resulted in personal injuries, and about 64% of the crashes involved property damage only. None of the crashes caused a fatality. The crash types consist of 59% rear-end collisions, 20% angle collisions, 11% single-vehicle collisions, 5% sideswipe collisions, and 2% head-on collisions. Three crashes involved pedestrians, and two crashes involved a bicyclist. About 35% of the crashes occurred during peak periods. About 22% of the crashes occurred when the roadway pavement was wet or icy. Approximately 15% of the crashes occurred in dark conditions (dawn, dusk, and nighttime).

Crash rates are another effective metric for examining the relative safety of a particular location.2 Based on the 2007 to 2011 crash data and the recently collected traffic volume data, the crash rate for this intersection is 2.33 crashes per million entering vehicles (see Appendix A for the calculation). This crash rate is about three times the average rate for signalized intersections in MassDOT Highway Division District 4, reported by MassDOT to be 0.68 crashes per million entering vehicles.3

Using the WPD crash reports, staff constructed a collision diagram for the intersections (see Figure B-1 in Appendix B). Crashes in this area stretched from the driveway of Wilmington Station to a quarter mile south of the intersection. Crashes that occurred at or near the station driveway are shown on a separate crash diagram (Figure B-2 in Appendix B).

In the five-year period 2007 through 2011, 14 rear-end crashes occurred on the stretch of road between the Wilmington Station driveway and the driveway for Big Joe’s just before the intersection. Crash reports attributed most of these crashes to driver inattention. They also might have been related to the congested conditions during the peak periods, as nearly half of them occurred during a weekday peak period. Five collisions occurred at the driveway for the sandwich shop and vacant convenience store). The driveway is close to the intersection. It is difficult to exit the driveway when southbound traffic on Main Street is heavy.

Within the intersection, five crashes occurred between a vehicle traveling southbound on Main Street and a vehicle turning left onto Burlington Avenue. All of the crashes occurred during a period when the turning vehicle had a permitted green signal (rather than a dedicated left-turn signal). Two crashes occurred between vehicles traveling northbound on Main Street and a vehicle traveling eastbound on Burlington Avenue. One crash was an exceptional situation involving a towed trailer; the other was the result of a driver likely entering the intersection at the very end of a yellow phase. One crash occurred between a vehicle traveling southbound on Main Street and a vehicle traveling westbound on Church Street. The driver on Main Street did not see a red light and crashed into the westbound vehicle.

Near the intersection, the northbound, southbound, and westbound approaches each had five rear-end crashes, while the eastbound approach had only one rear-end crash. The insufficient sight distance of the intersection for drivers on the three approaches and/or congested conditions likely contributed to the rear-end crashes, while drivers on the westbound approach had a better view approaching the intersection. There were a noticeable number of out-of-control single-vehicle crashes at or near the intersection. They were also likely related to the intersection topography.

South of the intersection, a number of crashes are related to vehicles turning into and out of the driveways on Main Street. Some of the vehicles exiting these driveways caused rear-end collisions on the northbound approach. However, while many crashes were related to driveways, six rear-end crashes occurred on the southbound approach just past the intersection, where there are no driveways. The crash reports indicated that most of those crashes were due to the lane reduction of the approach.

Three crashes involved cars parked on Main Street in the allowable parking area near the sidewalk curb in front of the storefronts. The crash locations were all near the intersection. There were no clear marked parking spaces. Frequently cars were parked near the northeast corner of the intersection and obstructed the paths of vehicles turning from other approaches.

There were three crashes that involved a pedestrian or a cyclist. Two pedestrians were struck crossing the street at a location without a crosswalk. Each of them individually crossed approximately 100 feet north of the intersection, approximately where people leaving a commuter rail train might want to cross. One cyclist was sideswiped by a car turning into the Bank of America driveway.

Table 2 shows the crash data for the intersection of Main Street at the MBTA driveway. All of the data for these crashes are derived from the WPD crash reports. There were nine reported crashes at or near the driveway intersection. Most of them were property-damage-only crashes and none caused fatalities. The predominant crash type was rear-end collisions. Few crashes occurred in peak traffic periods, on wet or icy roads, or in dark conditions. No pedestrians or cyclists were involved in those crashes.

Based on the WPD crash reports, staff constructed a collision diagram for the intersections (see Figure B-2 in Appendix B). Most of the crashes occurred on Main Street north of the MBTA driveway. Five rear-end crashes occurred on Main Street southbound after the railroad crossing, and two occurred on Main Street northbound before the railroad crossing. One rear-end crash occurred on Main Street southbound before the railroad crossing because a train blowing its horn caused a driver to stop abruptly before the crossing. The train was actually headed to Lowell and did not travel over this crossing. On the southbound approach, traffic signals at the crossing and at this intersection are synchronized in order to ensure that vehicles will not back up from the intersection onto the railroad track. It is a necessary precaution to prevent vehicles from sitting on the train track when a train is coming. The operation appears to be appropriate, as the number of crashes in that section is not alarmingly high and there were no crashes on the train track there.

The crash rate at the MBTA driveway intersection was calculated as 0.29 crashes per million entering vehicles (see Appendix A for the calculation), which is significantly lower than the MassDOT District 4 average. It should be noted that there were only a few hundred vehicles going in and out of the driveway, mainly during early morning and early evening hours. Most of the time, the signal remains green on the Main Street approaches. Once there is a call from cars exiting the driveway or from a pedestrian wanting to cross Main Street, the signal would switch to red for them to cross Main Street. However, they usually have to wait about a minute for their turn. Although there were no angle crashes reported in the study period between vehicles from the MBTA driveway and from Main Street, during a field visit in the early evening in April 2012, Staff observed two cars, at two different times, turning left from the driveway during a red light. The drivers probably were impatient about the long wait for the traffic light to change.

Intersection Capacity Analysis

Staff collected turning-movement counts at the study intersection and the intersection at the MBTA driveway for the commuter rail station on three individual midweek days in April 2012. The data were recorded in 15-minute intervals for peak traffic periods in the morning, from 7:00 to 9:00, and in the evening, from 4:00 to 6:00. Meanwhile, 24-hour automatic traffic recorder (ATR) counts at locations near the two intersections for three consecutive midweek days were collected by the MassDOT Highway Division in the week beginning March 12, 2012 (see Appendix C for the ATR counts summarized by hours of the day). Based on the 24-hour traffic counts, the turning-movement counts at the two intersections were adjusted and balanced.

Table 3 shows that the study intersection carried a total of nearly 2,500 vehicles in both the morning peak hour, from 7:30 to 8:30, and the evening peak hour, from 5:00 to 6:00. During the AM peak hour, 6 pedestrians crossed the intersection, and during the PM peak hour, 10 crossed. Table 4 shows that the MBTA driveway intersection carried nearly 1,700 vehicles in both the morning peak hour and in the evening peak hour, respectively. About five pedestrians crossed the intersection in either of the peak hours.

Five cyclists were observed crossing the study intersection in the morning peak hour, and two in the evening peak hour (not shown in the table). Two cyclists observed at the MBTA driveway in both the morning peak hour and in the evening peak hour. It should be noted that pedestrians and cyclists in April are usually fewer than in other month from May to October. At an additional site visit in July, staff observed about twice number of pedestrians and cyclists in both the morning and evening peak hours as there were during the April observations.

Based on the adjusted turning-movement counts, staff performed capacity analyses for the two intersections using Synchro software.4 The signals at the two intersections were modeled as coordinated signals, with the signal at the MBTA driveway as the master intersection. Signal timing input data were based on the timing plans provided to staff by MassDOT Highway Division District 4 and on manual measurements at the site.

Currently, the signal at the study intersection operates in the following sequence: (1) leading northbound movements with left turns protected, (2) northbound/southbound movements with left turns permitted, (3) lagging southbound movements with left turns protected, (4) westbound split phase, (5) eastbound split phase, (6) exclusive pedestrian phase if actuated. The northbound leading phase overlaps with the eastbound right-turn movement. Field observations indicated that the signal operated at a cycle length of about 155 seconds, including the exclusive pedestrian phase of about 22 seconds.

Table 5 summarizes the capacity analyses by lane for each approach of the study intersection in the AM and PM peak hours. The coordinated Main Street approaches were estimated to experience less delay and have a better level of services (LOS) than the side street approaches, which were evaluated to operate at LOS C or D, while the eastbound left-turn/through and the westbound right-turn/through movements were evaluated as LOS F, with extensive delays.

However, the analyses estimated that the southbound approach at the study intersection could experience a 95th-percentile queue5 of nearly 900 feet in the AM peak hour, and the northbound approach could experience a 95th-percentile queue of the same length in the PM peak hour. It also indicated that the eastbound Burlington Avenue approach could have a 95th-percentile queue of about 675 feet in the PM peak hour. Overall, the intersection was estimated to operate at LOS D in the AM peak hour with an average delay of nearly 55 seconds per vehicle (nearly LOS E), and to operate at LOS E in the PM peak hour with an average delay of nearly 70 seconds per vehicle. Details of the capacity analysis for both the AM and PM peak hours are in Appendix D.

Table 6 summarizes the capacity analyses by lane in each approach at the MBTA driveway intersection. All of the approaches were found to operate at an acceptable LOS of C or better, with modest delays, except for the eastbound left-turn/through movements, which was estimated to have an LOS D with an average delay of about 40 to 45 seconds per vehicle in both the AM and PM peak hours. Overall, the intersection was estimated to operate at LOS B with an average delay of 11 to 12 seconds per vehicle in both the AM and PM peak hours. Details of the capacity analysis for both the AM and PM peak hours are in Appendix E.

The Synchro analyses generally coincide with the observations on the field. It should be noted that the northbound approach could experience somewhat greater delays in the PM peak hour during and after the train preemptions.

Improvement Alternatives

There appear to be a limited number of options for making the geometric modifications to improve traffic operations and safety at this intersection. However, traffic signal design modifications and various modifications to pavement markings may be feasible.

A major improvement project was proposed in the 1980s. It would grade-separate Route 38/129 and Route 62 by extending the elevated portion of Route 62 so that is also goes over Route 38/129. The improvement would make Route 38/129 at-grade and eliminate the vertical curves on Route 62. However, this design would require widening some sections of both roadways to connect various turning movements. The project was not favored by the Town of Wilmington. The intersection was reconstructed with no major geometric modifications when the railroad bridge on Burlington Avenue was replaced and widened in 2001.

In addition to the geometric limitations such as the lack of right-of-way for expansion, the surrounding built-up conditions, and the proximity to the railroad, signal operation improvements at this intersection are limited by the train preemption operation. The current traffic control settings of the two intersections’ coordinated traffic signals are dictated by the train-crossing signal, which is necessary for ensuring the safety of roadway users during train crossings, since crashes between a crossing train and vehicles on Main Street could be fatal. The train preemption sequence, which clears the southbound queues at both intersections, stops northbound traffic and allows only right turns from the MBTA driveway and traffic from either of the side streets at the study intersection, has to be preserved.

Staff tested four improvement alternatives that do not involve major geometric modifications. They are analyzed below, progressing from simple to more involved options. The four alternatives are:

Staff used Synchro to analyze the alternatives. Ideally, VISSIM6 or other sophisticated multimodal transportation models that can simulate train preemption operations would have been preferable for this study. However, those programs are much more expensive and were beyond this study’s budget. However, since field observations indicate that the intersection’s traffic operations generally are not significantly impacted by train preemptions in the AM and PM peak periods, except the preemption at around 5:20 PM, Synchro was considered to be adequate for this study.7

Table 7 summarizes the intersection capacity analyses for the four alternatives. Detailed analysis results for both the AM and PM peak hours at the study intersection for the alternatives are included in Appendices F to I. The analyses show that only Alternative 4 would have a noticeable operations improvement. Alternative 1 would improve the operation slightly. Alternative 2 operations would be worse than the existing operations in both the AM and PM peak hours. Alternative 3 would improve the operations moderately in the PM peak hour but could compromise the traffic safety on the side streets, making them worse than in the current split-phase operations.

The purpose of Alternative 1 was to examine the existing signal timing to determine if it could be adjusted using the most recent traffic counts. Iterations of Synchro cycle length and split optimizations indicated that the overall intersection delay could be reduced slightly, by about 5 seconds per vehicle, if the signal timing on the Main Street approaches is reduced by 10 seconds and the timing for other approaches remains the same. The adjustment would slightly reduce the average delay per vehicle and the 50th-percentile and 95th-percentile queues on almost all of the approaches. However, the analyses also estimated that in the PM peak hour, 50th-percentile and 95th-percentile queues on the northbound through/right-turn approach would increase by about one vehicle length (25 feet).

In Alternative 1, the same timing reduction was applied on the Main Street approaches at the MBTA driveway signal, as the two signals were modeled as coordinated intersections. The analyses indicated that all of the approaches of the driveway intersection would operate at a similar LOS, with comparable delays, except the eastbound left-turn approach. The adjustment would reduce the delay by about 12 seconds and 8 seconds per vehicle for the approach in the AM and PM peak hours, respectively. This potentially would reduce the red light running by vehicles turning left from the driveway. Detailed analysis results for both the AM and PM peak hours at the driveway intersection in Alternative 1 are included in Appendices J.8

Although Alternative 1 would moderately improve the operations for most of the approaches at the study intersection and noticeably improve the operations at the MBTA driveway, the potential increase of traffic queues on the northbound approach at the intersection is a concern. Presently, the approach endures an extensive queue during and after train preemption operations at around 5:20, during the PM peak hour.

Alternative 2 was developed with the intention of improving the traffic operations and safety on Main Street. It was expected that the LOS on Main Street would deteriorate, especially the southbound approach. The analyses showed that the southbound operations would deteriorate more than expected, especially in the AM peak hour. Meanwhile, all of the other approaches would also deteriorate significantly, as the underserved southbound demand would need the majority of the intersection’s capacity. The 50th-percentile and 95th-percentile queues on the southbound approach were estimated to be about twice the estimates of the existing conditions. The tests of this alternative indicate that no operations appear to be feasible for the Main Street approaches except the current operations.

Alternative 3 was developed to determine if the traffic operations on Burlington Avenue and Church Street could be improved by allowing concurrent through movements from both streets. The current split-phase operation is generally safer than other operations but demands more of a share of the intersection capacity. Because the intersection’s roadways do not intersect perpendicularly, the left-turn paths from the two streets tend to cross each other if they are have concurrent phases. As such, the lead/lag left-turn operation was considered for this alternative. The analyses showed that this alternative would moderately improve traffic operations on most of the approaches, but would cause traffic operations on Burlington Avenue to deteriorate, especially in the AM peak hour. Overall, the operations benefits of this alternative do not outweigh the safety benefits of the split-phase operation.

Alternative 4 was developed to utilize the relatively wide surface of the roadway section that goes over the railroad. Currently the approximately 48-foot-wide surface is divided evenly between the two In Alternative 1, the same timing reduction was applied on the Main Street approach at the MBTA driveway signal as at the signal at the study intersection, since the two signals were modeled as coordinated intersections. The analyses indicated that both of the approaches of the driveway intersection would operate at the same LOS, with comparable delays, except the eastbound left-turn approach. The adjustment would reduce the delay per vehicle by about 12 seconds in the AM peak hour, and by about 8 seconds in the PM peak hour. This would potentially reduce the frequency of red light directions. The eastbound side is further divided into two 12-foot lanes: one for right turns only and one shared by left turns and through movements. The westbound direction has a width of about 24 feet, and is designated as a single lane to receive traffic from all other streets. It appears that there is room for adding a lane on the eastbound approach by reconfiguring the lanes without widening the bridge.

Figure 3 shows the proposed lane reconfiguration for the intersection. The proposed configuration on Burlington Avenue would contain a 12-foot right-turn lane, an 11-foot through lane, and an 11-foot left-turn lane in the eastbound direction, and a 14-foot receiving lane in the westbound direction. The left-turn lane could store about three to four left-turning vehicles.

The capacity analyses indicated that Alternative 4 would improve traffic operations on all of the approaches at the intersection in both the AM and PM peak hours. Traffic queues would potentially be reduced by about 30% on the eastbound approach and by around 5% on most of the other approaches. The overall queue on the northbound approach would be slightly reduced in the AM peak hour, but would increase slightly in the PM peak hour. However, the queue in the northbound left-turn pocket would be reduced by about one vehicle length in the PM peak hour.

Note: Cell Values: Level of Service (A to F)/Average Delay (seconds per vehicle).

Alt. 1: Retime the traffic signal under the existing intersection layout and phasing sequence.

Alt. 2: Convert SB inside lane to a left-turn-only lane, and operate NB and SB left turns under protected/permissive phases.

Alt. 3: Convert EB inside lane to a left-turn-only lane and EB outside lane to a through/right-turn shared lane, and operate EB/WB left turns under leading/lagging protected phases.

Alt. 4: Add an exclusive left-turn lane and realign EB approach within the existing bridge width, and operate the traffic signal under the existing phasing sequence.

Currently, the left-turn pocket has a length of about 200 feet. The analyses estimate the 95th-percentile queue as about 150 feet for the existing conditions9 and about 130 feet for Alternative 4. Potentially, the northbound left-turn pocket could be reduced by about 50 feet, and the two-lane section of the southbound departure lane could be extended. This could be achieved by slightly adjusting the northbound center line northward (see Figure 3).

At this preliminary planning stage, Alternative 4 appears to have the potential to improve the intersection operations and is geometrically feasible. At the functional design stage, further engineering reviews should be performed to determine if the additional lane would allow trucks to turn left onto Main Street from Burlington Avenue and would not block the turning paths of trucks from the other approaches. If necessary, the stop line of the left-turn lane on Burlington Avenue or the stop line of the southbound Main Street left-turn lane could be set back somewhat.

Improvement Recommendations

This is a congested, high-crash intersection with a constrained geometric and operational environment. It is an intersection where two major state routes meet. It is adjacent to a major commuter railroad that has a train station and an at-grade railroad crossing. Its surrounding areas are mostly built up, with a number of stores along Main Street. Its traffic signal is coordinated with an adjacent signal at the MBTA station driveway and is interconnected with the signal at the railroad crossing. T carries multiple transportation modes and serves as a gateway to the train station.

Staff performed a series of safety and operations analyses in order to identify geometric design and operational deficiencies at the intersection. In general, the analyses found that the congestion and most of the crashes were caused by heavy peak-period traffic, significant nearby commercial and commuting activities, and roadway grade changes due to the adjacent railroad tracks. The analyses indicated that the current signal design and coordination at this intersection and at the MBTA driveway intersection are appropriate under the existing conditions of high traffic demand, right-of-way and geometric limitations, and train-crossing safety considerations.10

In addition, staff tested four improvement alternatives that mainly focus on signal timing and phasing adjustments under the existing intersection right-of-way. The four alternatives are:

The analyses of the alternatives indicated that only Alternative 4 would have a noticeable operations improvement over the existing conditions. Alternative 1 would moderately improve the operations on the approaches, except for the northbound approach. Alternative 2 operations would be worse than the existing operations in both the AM and PM peak hours. Alternative 3 would somewhat improve the operation in the PM peak hour but could make traffic safety on the side streets worse than with the current split-phase operations.

In the short term, staff proposes the following measures to improve operations and safety at the intersection and the adjacent roadways. All but the last two items are low-cost measures that could be implemented in a relatively short time.

For the long term, the staff proposes the following additional measures that would potentially further improve operations and safety at the intersection and the adjacent roadways:

One additional long-term improvement option would be to add a left-turn bay on the southbound approach. It would significantly improve the operations and safety at the intersection, but it would require the removal of about four to five on-street parking spaces from the storefronts on Main Street.16 The parking spaces are much needed by the adjacent businesses and the Town does not favor removing them. However, this improvement option should be considered when the area of the storefronts is to be redeveloped or when the roadway north of the intersection is to be reconstructed as part of a major development in the future.

1 The number of parking spaces and occupancy rate are based on the counts performed in 2010 for the Boston Region MPO Congestion Management Process.

2 Crash rates are estimated based on crash frequency (crashes per year) and vehicle exposure (traffic volumes or miles traveled). Crash rates are expressed as “crashes per million entering vehicles” for intersection locations and as “crashes per million miles traveled” for roadway segments.

3 The average crash rates estimated by the MassDOT Highway Division are based on a database of intersection crash rates submitted to MassDOT as part of the review process for an Environmental Impact Report or Functional Design Report. The most recent average crash rates, which are updated on a nearly annual basis, are based on all entries in the database, not just those entries made within the past year. The average crash rate for District 6 was calculated on July 7, 2011.

4 Synchro is software used for intersection capacity analysis and traffic signal coordination that is developed and distributed by Trafficware Ltd. It can be combined with SimTraffic to perform traffic simulation for an individual intersection or a series of intersections.

5 The 95th-percentile queue is defined to be the queue length (25 feet per vehicle) that has only a 5 percent probability of being exceeded during the analysis time period. It is a useful parameter for determining the appropriate length of turn pockets, but it is not typical of what an average driver would experience. It can be regarded as the potential maximum queue length under the input traffic conditions.

6 VISSIM is a microscopic simulation program for multimodal traffic flow modeling. With a high level of detailed input data, it proclaims to be able to accurately simulate urban and highway traffic, including pedestrians, cyclists, and transit vehicles. The software is maintained and distributed by PTV America Inc. in Corvallis, Oregon.

7 The intersection operations are affected by three train preemptions in the morning peak period, at around 7:20, 7:50, and 8:05, and three in the evening peak period, at around 4:35, 5:20, and 6:00.

8 The timing and phasing settings at the driveway intersection in Alternative 1 were also applied to Alternatives 2 to 4. The appendix thus represents the analysis results for the intersection in Alternatives 1 to 4.

9 Field observations of the existing conditions are generally consistent with the estimation. Usually two to three vehicles and infrequently five to six vehicles queue in the lane during the peak hours.

10 Usually, a signalized roadway train preemption control scheme’s primary purpose is safety, and congestion mitigation is a secondary goal. With the advance of signal communication and traffic monitoring technologies, a number of researchers are currently seeking to find ways to achieve both objectives.

11 U.S. Department of Transportation, Federal Highway Administration, Chapter 2B, “Regulatory Signs, Barricades, and Gates,” in Manual for Uniform Traffic Control Devices, 2009 Edition, December 2009.

12 MUTCD, Chapter 2C, “Warning Sings and Object Markers,” 2009 Edition.

13 There are a number of commuters using bikes to access the station. Currently there is no shared-lane operation for bikes in that short section of roadway due to the limitation of the existing right-of-way. Preferably, bike travels should be separated from traffic in the busy Route 38/129 corridor. The “complete street” design concept should be considered for future corridor development. A possible solution would be adding a wide shoulder for bikes traveling in the corridor.

14 Traffic signal control technologies that can adapt to serve demand by adjusting the cycle lengths, splits, and/or offsets of traffic signals in a corridor based on volume or occupancy data collected in real time.

15 The corridor is a merged section of two major state routes, Route 38 (whose whole length is on Main Street in Wilmington) and Route 129 (diverging onto Lowell Street in the south and onto Shawsheen Avenue in the north). Essentially, roadway capacity is reduced from four lanes to two to three lanes, therefore making this section of the Route 38/129 more congested than other sections of Route 38 and of Route 129

16 The left-turn bay should have a length of at least 100 feet in order to meet the left-turn demand in peak hours.

TABLE 1
Summary of Crashes at the Intersection of Main Street at Burlington Avenue/Church Street, Wilmington: January 2007–December 2011

		2007	2008	2009	2010	2011	5-Yr. Total	Annual Avg.

Total number of crashes		31	24	25	13	23	116	23.2


Severity	Property damage only	19	14	14	10	17	74	14.8
	Non-fatal injury	9	8	11	3	6	37	7.4
	Fatality	0	0	0	0	0	0	0
	Not reported/unknown	3	2	0	0	0	5	1


Collision type	Single vehicle	5	3	3	0	2	13	2.6
	Rear-end	14	17	14	11	12	68	13.6
	Angle	7	3	8	1	4	23	4.6
	Sideswipe, same direction	1	0	0	1	3	5	1
	Sideswipe, opposite direction	1	0	0	0	0	1	0.2
	Head-on	0	1	0	0	1	2	0.4
	Not reported/unknown	3	0	0	0	1	4	0.8


Involved pedestrian(s)		1	1	1	0	0	3	0.6
Involved cyclist(s)		1	0	0	0	1	2	0.4
Occurred during weekday peak periods		9	8	10	7	7	41	8.2
Wet or icy pavement conditions		5	9	6	0	6	26	5.2
Dark conditions (lit or unlit)		4	5	3	1	4	17	3.4


Source	MassDOT database	22	9	3	0	0	34
	Wilmington Police Department	1	0	1	13	23	39
	Both MassDOT and WPD	8	15	21	0	0	44


Note: The Wilmington Police Department provided crash reports for the period between mid-August 2007 and December 2011. Statistics for crashes before
August 2007 were obtained from the MassDOT crash database. Crashes that occurred before August 2007 are not shown on the crash diagram.

TABLE 2
Summary of Crashes at the Intersection of Main Street at MBTA Driveway, Wilmington: August 2007–December 2011

		2007	2008	2009	2010	2011	5-Yr. Total	Annual Avg.

Total number of crashes		0	3	2	1	3	9	1.8


Severity	Property damage only	0	3	2	1	1	7	0.8
	Non-fatal injury	0	0	0	0	2	2	0.4
	Fatality	0	0	0	0	0	0	0
	Not reported/unknown	0	0	0	0	0	0	0


Collision type	Single vehicle	0	0	0	0	0	0	0
	Rear-end	0	2	2	1	2	7	0.8
	Angle	0	0	0	0	1	1	0.2
	Sideswipe, same direction	0	0	0	0	0	0	0
	Sideswipe, opposite direction	0	1	0	0	0	1	0.2
	Head-on	0	0	0	0	0	0	0
	Not reported/unknown	0	0	0	0	0	0	0


Involved pedestrian(s)		0	0	0	0	0	0	0
Involved cyclist(s)		0	0	0	0	0	0	0
Occurred during weekday peak periods		0	0	1	0	0	1	0.2
Wet or icy pavement conditions		0	0	1	0	0	1	0.2
Dark conditions (lit or unlit)		0	1	1	0	0	2	0.4


Source: The summary is based on crash reports obtained from the Wilmington Police Department.

Main Street at Burlington Avenue/Church Street, Wilmington		Main St.						Burlington Ave.			Church St.			Total
		Northbound			Southbound			Eastbound			Westbound
		LT	TH	RT	LT	TH	RT	LT	TH	RT	LT	TH	RT
AM peak hour	Turning volume	118	345	68	40	968	59	53	194	214	121	224	42	2,447
	Approach volume	531			1,068			461			387
	Ped. crossings	2			3			0			1			2
PM peak hour	Turning volume	80	618	59	72	625	107	65	270	170	92	247	48	2,453
	Approach volume	757			804			505			387
	Ped. crossings	0			9			0			1			10

Main Street at the MBTA Driveway, Wilmington		Main St.							MBTA Driveway			Private Driveway			Total
		Northbound			Southbound				Eastbound			Westbound
		LT	TH	RT	LT	TH		RT	LT	TH	RT	LT	TH	RT
AM peak hour	Turning volume	15	424	1	6	1038		97	37	0	30	0	0	0	1,653
	Approach volume	440			1,146				67			0
	Ped. crossings	4			1				0			0			5
PM peak hour	Turning volume	9	714	8	6		750	41	88	0	48	6	1	0	1,671
	Approach volume	731			797				136			7
	Ped. crossings	3			0				2			0			5

		Main Street		Burlington Avenue	Church Street	Overall
		Northbound	Southbound	Eastbound	Westbound	Overall
AM Peak Hour	Existing	C/31	D/46	E/58	F/109	D/55
	Alternative 1	C/28	D/41	E/56	F/94	D/49
	Alternative 2	C/27	F/216	F/103	F/145	F/142
	Alternative 3	C/26	C/34	F/123	F/104	E/61
	Alternative 4	C/25	D/38	D/51	F/94	D/49
PM Peak Hour	Existing	D/35	D/54	F/108	F/124	E/70
	Alternative 1	C/30	D/44	F/99	F/125	E/64
	Alternative 2	D/44	D/52	F/99	F/125	E/71
	Alternative 3	C/26	D/41	F/114	E/76	E/57
	Alternative 4	C/30	D/44	E/69	F/106	D/54