The I-93/I-95 Interchange is at the center of a regional highway system serving Massachusetts and New England. It is also a facility of importance to the host communities of Woburn, Reading, Stoneham, and Wakefield and neighboring towns.
The I-93/I-95 interchange, built in 1972, is a "system interchange" linking two of New England's busiest interstate highways. By 2004, approximately 377,500 vehicles passed through the interchange on an average weekday. Delays are common during peak commuting times, and traffic is often backed up from the ramps onto the mainline highways, creating a dangerous situation. Traffic in 2004 is approximately 123 percent higher than in 1974, when the interchange was processing approximately 169,000 vehicles per day. Traffic levels have not changed significantly since 2004, as the interchange still processes around 375,000 vehicles per day, and remains the highest single location for traffic volume in the state.
The following fact summaries are from documents used in the 2007 planning study:
Congestion and Traffic Operations:
The Level of Service (LOS) of the I-93/I95 interchange ramps and weave sections is LOS F for all weaves, two of the diverge (exit) ramps and one of the slip ramp entrances, depending on the time of day. This is the result of very high volumes and substandard radii and weave distances. The weaves between the Washington Street interchange and the central interchange are also LOS F, and the Route 128 weaves within the Route 128/Route 28 interchange are LOS F as well (NB: The 128NB exit ramp to 93 South is considered part of the weaving section with Washington St on ramp and therefore not analyzed separately. Similarly, the entrance ramp from 93 SB to 128 SB is considered part of the Washington St weaving section).
The queuing along the loop ramps has a cascading effect in the interchange. The northeast loop (93NB to 128SB) backs up in the morning and afternoon peak periods. This leads to a queue of cars in the right lane, a safety hazard, and eventually blocks vehicles exiting the southeast loop (128NB to 93 NB) causing it to back up. This queue in turn extends far enough to interfere with vehicles leaving the southwest loop (93SB to 128NB). The queues often extend far enough to block entry to the slip ramps in these three quadrants as well.
Traffic speeds are significantly reduced in the interchange during morning and afternoon peak hours. More work remains to refine the speed data, but the overall slowdown is readily apparent.
There is morning peak-hour congestion on Route 128 SB, which extends to Route 2 in Waltham/Lincoln and beyond. This appears to be due to a combination of high traffic volumes in every segment of Route 128 over this stretch of highway, combined with short acceleration lanes at many of the interchanges. The southbound congestion reduces speeds and further congests movements within the I-93/I-95 and Washington Street interchange area. This situation reduces the problem of matching speeds of entering vehicles from the interchange on ramps to Rte 128, but the close spacing of vehicles compounds the weaving problems. More work is needed with the microsimulation model to determine the effect of specific improvements connecting to Route 128 SB in the congested hours.
There is afternoon peak-hour congestion on Route 128 NB, which appears to begin in the vicinity of the lane drop north of the I-93/I-95 Interchange. The congestion on Route 128 NB improves north of Route 28.
There are numerous congested intersections in the Woburn commercial area, particularly on Washington Street and Mishawum Road at the Route 128 ramps, and cut-through traffic on local streets such as Route 129, West Street, and Montvale Avenue.
Available crash data show that the I-93/I-95 Interchange has significant safety problems, regardless of the measure used (crash ranking, crash rate, site-specific data), and regardless of the level of comparison:
Crash Ranking: Between 1995 and 2001, the I-93/I-95 interchange ranked between 1st and 6th on the list of 1,000 worst crash sites (all intersections and interchanges) in Massachusetts; which provides a strong indicator that a safety problem exists. This historical annual crash data provides the information for general observations and comparisons to other locations, and confirms a consistently high level of accidents at this location.
Crash Rate: The I-93/I-95 Interchange had the highest or nearly the highest crash rates for two historical analysis periods (1997-99 and 1999-2001) when compared to other cloverleaf interchanges and all interchanges as well. Calculating a crash rate is the next logical step in measuring the severity of a safety problem. A crash rate - expressed as number of crashes occurring per a given number of vehicles using a location - eliminates the inflating effect of high traffic volumes on total crashes (all other factors being equal, locations with higher traffic volumes will have more crashes than less-traveled locations). The crash rates of 1.88 and 1.36 per million entering vehicles (MEV) for the two analysis periods were worst among all cloverleaf interchanges and nearly the worst among all interchanges. Crash rates are again a general part of the overall picture and help to give a sense of priority with other locations in considering improvements. This historical crash rate information confirms a high rate of crashes occurring at this location.
Spatial Analysis: A spatial analysis of 2004 raw crash data shows clusters of crashes at a number of locations within the interchange, which generally correspond to the geometric problems observed there (see Geometrics below). For example, the cluster of crashes that occur entering the ramp from Route 128 NB to I-93 SB confirms the sight distance limitations present at that location. Conducting a spatial analysis identifies patterns of crashes and is the next logical step in measuring the severity, extent, and details of a safety problem (crashes are examined by type, time of day, weather conditions, etc.). The spatial analysis is valuable in that provides the basis for detailed observations and conclusions to further define the problem.
The interchange was built in 1972, and its geometry is substandard when measured against both MassDOT and AASHTO current design standards:
Loop ramp radii are too tight, requiring rapid deceleration of entering vehicles and requiring exiting vehicles to merge at a lower speed than the mainline traffic. This is both a safety problem and an impediment to traffic flow. During off-peak hours, rollovers sometimes occur.
The cloverleaf design creates four weaves that are significantly shorter than the required weaving distance, creating many conflicts between vehicles leaving one loop ramp and those entering the adjacent loop, whose paths must cross.
The acceleration and deceleration lanes are much shorter than required to match entering and exiting traffic speeds with those of mainline traffic. This poses a safety problem.
The sight distances for exiting and entering traffic are limited in some places, particularly the exit ramp from Route 128 northbound (NB) to I-93 southbound (SB).
The spacing between the central interchange and the Route 128/Route 28 interchange is shorter than standards recommend, and the spacing to the Washington Street interchange is significantly shorter than the standard. This causes additional weaves and traffic conflicts within a shorter distance, introducing additional safety and traffic flow problems.
There is a reduction from four to three travel lanes on Route 128 NB just east of the interchange, compounding the traffic flow and lane change problems in that direction.