Embedded Files
1 INTRODUCTION
This Memorandum documents the process generally used by UDOT to establish appropriate construction schedules to prevent excessive delays to traffic. UDOT is aggressive in protecting traffic during construction. Maintenance of Traffic (MOT) has a direct impact to the public, and therefore has a significant impact on the trust that UDOT has earned. A loss in that trust would be a loss in our ability to continue supporting Utah’s mobility goals.
Lane closures, flagging and full closures / diversions are evaluated based on construction needs and traffic demands for each specific project. MOT analyses are developed based on UDOT’s anonymous probe traffic data available statewide and anonymous 2,270 point traffic data stations collected through radar.
2 STEP 1: DETERMINE CONSTRUCTION SCHEDULE AND NEEDS
UDOT Traffic Operations works in partnership with the Resident Engineer, Project Manager and Structural Engineer (if applicable) to identify closures necessary to construct the project. For example, a restriping project may only need an outside lane closure while a bridge repair may need full closures and diversions.
This team also needs to identify the duration of those closures. Some work can be done in shifts which allow traffic to be closed and reopened during construction. Other work may require long-term closures that cover full days or even full weeks. These details are important when demand levels are considered.
Finally, the team also needs to identify when a project will be constructed. Some areas of Utah are highly susceptible to seasonal and holiday patterns. Projects constructed during peak seasons will have greater limitations. The month(s) and the year are both applicable as future years will require a forecast that accounts for typical growth in the area. These details are usually ironed out in meetings or over the phone, depending on the scope of the project. MOT requests for Regions 1, 3 and 4 are currently supported by the Traffic Management Division and can be submitted through the UDOT MOT Form. This is a support program and all final decisions to adopt recommendations are made by the Region Traffic Operations Engineer.
3 STEP 2: DETERMINE TRAFFIC DEMAND
The next step is to determine the traffic demand patterns that will be expected for the project. The goal is to identify a typical hourly demand for the construction season rather than trying to account for extraordinary travel days (such as holiday weekends) as construction limitations will usually prohibit closures during holidays and major events anyway.
3.1 Holidays
Holidays are identified for the project by compiling hourly volume data, recorded by UDOT’s radar data collection, for a full year. In general, expect to see demands spike during the following holidays:
New Years Day
MLK Day
President's Day
Memorial Day
Juneteenth
Independance Day
Poineer Day
First Day of Deer Hunt
Labor Day
Columbus Day Veterans Day
Thanksgiving
Christmas
Note that peak patterns will depend on the specific holiday. For example, Thanksgiving always falls on a Thursday, so outbound traffic (from Salt Lake, Davis and Utah County) will experience heavy demands on Wednesdays and inbound traffic (to Salt Lake, Davis and Utah County) will peak on the following Sunday. The location of the project will also play a significant role. Deer Hunt Season is typically a peak weekend for projects located in Region 4 whereas these patterns don’t typically show up in Salt Lake County. Some holidays, like Independence Day, will occur on various days so the outbound and inbound traffic will also shift accordingly. Holidays that always fall on a Monday, like Labor Day, will see a peak return on that Monday. It may be necessary to pull multiple years to identify the dominant pattern and relate that to the forecasted construction season.
3.2 Seasonal Pattern
Hourly volume data, recorded by UDOT’s radar data collection, for a full year will also reveal seasonal patterns for the project location. Note that Utah typically experiences (though many more exist) the following seasonal patterns depending on the project location:
☐ Urban and Rural commuter traffic (patterns tend to be static throughout the year)
☐ Winter recreational traffic (patterns tend to be heavier in January, February and March)
☐ Summer recreational traffic (patterns tend to be heavier in June, July and August)
MOT Schedules generally use the peak month for the construction window for the project. If severe variations between months exist, the project may consider multiple MOT schedules. However, this is not typically advisable as complicated MOT schedules often result in misinterpretations in the field and can potentially create more problems that it solves. An offer in the 555 Specification to “re-evaluate” the MOT schedule during construction is often more effective as it also accounts for diversions that are difficult to forecast. Monitoring construction delays, and the lack thereof, can also be done using UDOT’s Clearguide Construction Alerts System. This is a live monitoring of probe traffic travel times that sends and records times when travel times or speeds cross customizable thresholds.
3.3 Typical Hourly Demand (Daily Profiles)
Once, the hourly volume data is narrowed down to the peak month during the construction window, a plot of this data, by hour reveals daily patterns to consider. It is important to clean the data set of anomalies before compiling a profile, starting with holidays unless these need to be specifically accommodated for long duration closures.
UDOT’s probe data for the target month is also used to weed out days that may have experienced significant delays due to incidents or weather. This is important because unaccounted for congestion could result in lower-than-typical demand. UDOT’s Clearguide Probe Traffic Data Tool is used to produce the following speed contour plot for the project location. Days that experience unusual delays are eliminated from the analysis.
Note that additional patterns may also be relevant based on project location. These patterns often create the need for separate MOT schedules, by direction of travel, as the days and hourly demands may differ substantially. Patterns that may contribute to this include:
Urban commuter traffic (patterns that, although may follow the same trajectory, peak differently by hour)
Rural commuter traffic (patterns that tend to be directional in the AM and PM such as I-80 to and from Tooele)
Heavy truck traffic (patterns that tend to reflect long haul trucking that route through Utah and back)
Destination traffic (patterns that tend to reflect destinations like ski resorts’ and national parks’ opening and closing hours)
Once a clean data set for the target month has been compiled, an average is taken of the hourly demand over a typical week. This is used to determine the demand that needs to be accommodated during construction.
3.4 Forecast Growth
Sometimes, a project is planned more than one year in advance. In these cases, the typical week is forecasted to consider the general growth of the area using historical data from UDOT’s radar. These forecasts do not use trip generation or Travel Demand Modeling, as would be typical for a full traffic analysis. They are a simplistic approach, using general growth over the years for the same typical week.
4 STEP 3: DETERMINE REDUCED ROADWAY CAPACITY
Once a final demand profile has been developed, the next step is to identify the capacity that can be accommodated for each MOT configuration. The maximum capacity is the optimum level of flow BEFORE speeds begin to drop (the tip of the speed flow curve)
4.1 General Enemies of Capacity
Every project location will experience different levels of capacity and the impacts to that capacity will change based on the severity of the closure. Some of the most common things to look out for include:
MOT Closure Type (in order of severity: Full Closure, Flagging, Lane Closure, Shoulder Closure)
Number of existing lanes (roadways with 5-7 lanes will suffer less impact to capacity than roadways that are 2-4 lanes)
Percent of Heavy Trucks (heavy trucks often have a passenger car equivalent of 4 in Utah and have a huge impact on capacity)
Grades (this becomes especially important for projects with a greater % of heavy trucks)
Vehicle Speeds (although high speeds are not a goal through a construction zone, they do improve capacity)
Business Accesses (left turns will have far more impact than right turns)
Width of lanes and shoulders (again, this is more impactful for locations with high heavy truck %)
Availability of a pilot car (pilot cars improve efficiency, particularly for metering and flagging operations)
4.2 Typical Capacity Levels – Urban Congested Locations
Capacity for urban locations are easier to determine as the roadway regularly experiences demands that exceed capacity.
The typical capacity that the roadway operates at is determined using the same hourly radar data used to determine traffic patterns. For simple urban locations, the max capacity for the roadway can be taken from the typical day when speeds start to drop as volumes peak as shown below.
Once the existing capacity of the project is determined, the enemies of capacity for the project are identified. As those enemies add up in the MOT configuration, the capacity drops.
4.3 Typical Capacity Levels – Rural uncongested locations
Projects in rural locations are much more difficult to determine capacities for as the demand will rarely reach capacity under non-construction conditions.
The first option is to look for holiday demands which may show that same drop in speed as volumes peak condition that is used for urban projects. The example below shows the Friday of Memorial Day Weekend with nearly double the typical demand. Occasionally, peaks like these will correlate to drops in speed, just as they do in urban areas, and a capacity can be determined the same way as section 4.1 demonstrates.
However, as shown below, the speeds for this example do not respond to the increased holiday demand. This means that the roadway is still operating under capacity which is common for rural locations.
The next option is to look for days when the roadway was under reduced conditions. If an appropriate example can be found for the right time period and the right MOT configuration, the resulting capacity can be used as the threshold directly. These can be found using the drops in speeds shown in the plot above, UDOT’s Clearguide Contour Plots discussed earlier and UDOT’s inventory of Data Quality Reviews. Use the radar data for the area to identify which lanes were closed and match those to the MOT configuration that is being considered.
Full Closure Data Example:
Below is an example of a full closure on August 18, 2024 and what that data looks like. Closures show up in Clearguide as black links and volumes in radar will typically drop below 50 vplph.
Note that the news article shows the roadway as completely closed.
Partial Closure Data Example:
Below is an example of a partial closure of I-215 and what that data looks like. Note that the day prior to, and the day after, August 15, 2024 both show the typical demand pattern for this section of roadway. The drops in speed on August 15 are unusual for the morning period here and correlate to substantially lower capacities.
Finally, if these do not produce results, the final option is to use experience and draw from general rules discussed in the next section.
Just as in urban congested locations, once the existing capacity of the project is determined, the enemies of capacity are identified. As those enemies add up in the MOT configuration, the capacity drops. This is very much dependent on experience as limited data is available. Section 4.3 covers the general ranges that UDOT roadways experience.
4.4 General Lane Closure Impacts on Capacity
As a general rule, capacity for the project is VERY project specific and dependent on the “enemies to capacity” that the project area must accommodate.
Although it is very tricky to set broad scale ranges for capacity, UDOT has generally found the following ranges to be expected:
The highest reported freeway capacity for flat, straight sections of I-15 with 4-5 lanes, 75 mph speed limits, and interchange spacing greater than 1 mile, is 1900 vehicles per lane per hour. Utah does not experience the 2000 vplph thresholds theorized in the HCM. Freeway sections through Davis, Salt Lake and Utah county often flow between 1300-1800 vehicles per lane per hour, with 1300 being typical during rush hour. Freeway sections in Southern Utah, typically operate around 1200-1600 vehicles per lane per hour.
Freeway sections 4-5 lanes wide tend to operate at an optimum level (1700-1800 vplph). As the number of lanes drops below 4-5, the max capacity per lane will substantially drop. Likewise, as the number of lanes increases above 4-5, the max capacity per lane will also dip. Sections of freeway with more than 5 lanes will actually achieve lower capacity per lane. This is because, as the number of lanes increases, the number of lane changes that vehicles need to make to get to those lanes, also increases. As lane changes increase, so do the gaps between vehicles as a matter of necessity for the maneuver. The increase in gaps reduces density. The drop in density ultimately decreases capacity. So, sections of freeway with 7 lanes may only see a capacity of 1500-1700 vplph.
The biggest impact to capacity is dropping a lane from a dual lane configuration as the ability to maneuver around slow-moving vehicles is eliminated. This means that the entire demand will move at a slower pace than the slowest moving vehicle. This is because a queue will often develop behind slow vehicles, and that queue often experiences shock waves as that slow vehicle speeds up and slows down. UDOT has seen capacities as low as 400 vplph on a single lane freeway MOT configuration.
The next biggest impact to capacity is dropping a lane from a 3-lane configuration as the ability to maneuver around slow moving vehicles is substantially hindered. One reason for this, often known as an elephant race, is when one heavy truck is passing another. During this time, vehicles can get trapped behind the passing truck, and as that queue develops, the amount of time for that queue to clear, also increases. Note that Southern Utah, specifically, has seen single lane reductions result in 800 vehicles per lane per hour capacity for the remaining 2 lanes. This is unusually severe and needs to be considered for projects in Region 4. For other locations this can drop capacity for the remaining 2 lanes down to 1200-1400 vplph.
The next biggest impact to capacity is dropping a lane from a 4-5 lane configuration, which will usually drop capacity by that of the lanes next to it (1400-1800 vplph) as this configuration falls within the range optimum levels.
Dropping a lane from wider sections of freeway (6-7) will have the least amount of impact on capacity (usually dropping capacity by that of the lanes next to it (1300-1700 vplph) as this configuration has already passed optimum levels.
Signals with 45-55 mph speeds may process up to a maximum of 850 vehicles per lane per hour. Note that this is HIGHLY dependent on the signal timing and the demands on that signal from other movements. Use this number only as a reference when considering results of a MOT model done in VISSIM.
Unsignalized arterials with 45-55 mph speeds typically have 900-1200 vplph capacities but note that this is highly dependent on roadway geometry and accesses. Short sections of arterials are much more likely to reflect signal capacity than a 10 mile arterial absent of signals and accesses.
4.5 Capacity of Ramp-Surfing
One common method UDOT uses to accommodate a lane closure when demands normally be higher than the remaining lanes, is to surf traffic through an off ramp and back onto the on ramp through a diamond interchange. Note that this method does not work for SPUIs as the on and off ramps do not directly line up. For urban locations, this can often be accommodated through signal timing changes that prioritize the surfing movement. UDOT uses this methodology both for adjacent lanes and for split lanes (a closure of an outside lane and an inside open lane). The changes in signal timing will increase delays on the cross street so this is something to consider when developing the MOT plan.
For rural unsignalized diamonds, UDOT will sometimes bag the stop signs of the off ramp and install temporary stop signs on the cross street to protect the ramp surf movement. This improves efficiency of movement and increases capacity. However, the slower movement of the ramp surf will limit the capacity that can be gained through this method, especially for locations with heavy trucks.
Typically a ramp surf condition can accommodate between 700-1000 vplph. However, UDOT recently conducted a long term ramp surf condition for an urban interchange on SR-201 during bridge reconstruction. The interchange was reconfigured from a SPUI to a diamond and all of the mainline SR-201 lanes were surfed through concrete barrier-separated channelized lanes. The concrete barrier, widened 3-lane ramps, long tapers and long duration of the project all contributed to much higher ramp surf capacities that UDOT had ever experienced; maxing out at 1,300 vplph. More details on this study can be found in the SR-201 Ramp Surfing at 3200 W Traffic Performance Report.
4.6 Capacity of One-Way Flagging Operations
UDOT has begun to expand its MOT program and as part of that, specific efforts are being made to quantify thresholds more thoroughly. UDOT most recently conducted a study on the capacity of flagging operations and this is where we have the most data. Future research is planned to refine numbers for lane closures both in urban and rural locations specific to Utah. Those studies will be used to further guide estimates relating to enemies of capacity in both urban and rural areas. As those results become available, they will be incorporated into this guide.
The Evaluation of One-Way Flagging Operations research project, conducted specifically on Utah roads, examined the impacts of one-way flagging practices used during construction and maintenance on two-lane highways. This research aimed to develop reliable, data-informed standards for estimating the capacity of one-way flagging operations. The following document details key findings from the study Figure 1 illustrates the finding that the capacity of a one-way flagging operation declines as the length of the flagging zone increases. This is due to the increase in clearance time needed at longer flagging zones. Additionally, a higher percentage of trucks lowers the capacity at all lengths.
While this figure does not incorporate roadway grade as a variable, the research study found a 1% increase in grade would equal a 1.81% increase in truck percentage regarding its impact on capacity. This relationship can be used with the figure above to estimate the impact of grades on capacity. For example, a 6% grade would add an additional 10.9% trucks to the truck percentage.
Additional variables may impact capacity which are not in the figure; these include the following:
Driveways or intersections inside the one-way flagging zone
Trucks entering and exiting the construction zone requiring both directions to be stopped
Percent trucks travel through the one-way flagging zone
Vehicles speeds and travel times within the one-way flagging zone
Width of the lanes and shoulders within the one-way flagging zone
The need for a Pilot Car for vehicles to safely traverse the one-way flagging zone
Green Time Evaluation
A limited evaluation of optimal green time at one-way flagging operations was performed using calibrated Vissim simulation models. For this evaluation, scenarios used one mile of flagging zone distance and 10 percent trucks. Five different scenarios using split times (the time given to vehicles to enter the one-way flagging zone on an approach and time to clear the zone before the other direction is started) between 5 and 15 minutes were used to calculate capacity. Figure 2 shows that capacity increases as green time increases with diminishing improvement for longer green times.
While some benefit may be derived from longer one-way flagging split times, there is also an increase in stopped time for vehicles waiting to enter the one-way flagging zone.
Delay Calculations
The research study utilized equations for headway, clearance time, and maximum queue length to calculate total delay (stopped and clearance time) at one-way flagging operations. The calculated total delays (converted to minutes) for several scenarios are shown in Table 1. These scenarios included percent grades of 0 and 6 percent, percent trucks of 10 and 35 percent, flagging zone lengths of 0.5, 1.5, and 2.5 miles, and hourly approach volumes between 100 and 450 vehicles per hour.
The calculated delay times shown represent a scenario that is under the capacity of the one-way flagging operation, as it is based on the measured data from operations that were under the capacity of the operation. For traffic demand over the capacity of a one-way flagging operation the delay will grow exponentially.
Maximum Queue Length
Based on the collected data for the one-way flagging research the follow formula was developed to determine the maximum queue length expected for a one-way flagging operation. This formula is based on the distance of the one-way flagging zone, the hourly traffic volumes, and the percentage of trucks on the routes.
Maximum Queue Length (mi) = -0.13 + 0.042 × Distance + 0.002 × Hourly Volume + 0.003 × Percent Trucks(%)
However, it should be noted the data collected only included one-way flagging operations where the traffic demand was below the capacity of the operation. At locations where the demand exceeds capacity the queues will grow exponentially.
5 STEP 4: DETERMINE MOT SCHEDULE
Quite simply, the MOT schedule is set to restrict construction windows to time blocks when demand falls below the capacity provided by the MOT configuration. Note that slowing due to tapers is specifically ignored in these analyses so even with the MOT schedule, some slowing in reality is expected for any MOT configuration.
Generally, time blocks are only identified for usable windows. An hour window is usually not productive for the contractor as the time to set up barrels and take them down is more than the hour is worth. Work with the Resident Engineer and the Project Manager to identify projects that can benefit from these short windows, but as a general rule, short intervals do not provide as much value.
When considering flexibility of construction windows, it is important to note that failure of traffic operations is non-linear. Once speeds start to drop, traffic operations will start to fail. Once operations have failed, the path to return is different from that of failure. The demand must actually drop below capacity in order to recover. This means that it is far more important to protect the front of the peak period than it is the back. For projects that are looking to extend construction windows, it is much better to allow the contractor to start early in the construction window than it is to allow the contractor to end late. For example, construction for a project is restricted from 7 am to 10 am. It is better to allow the contractor to take lanes from 9:30 am to 10:00 am than it is to allow the contractor to continue work from 7 am to 7:30 am. That way additional traffic impacts do not compound, and recovery is much more likely because demand is actively dropping during the window encroachment. This is especially true for project locations that experience sharp increases in demand over short periods of time.
Below is an example of the MOT Schedules that are included in the 555 Specification. Utah Contractors are familiar with this format so the consistency of this table is important to maintain.
6 DETERMINE DISINCENTIVES TO ENFORCE MOT SCHEDULE
Finally, disincentives are used in UDOT’s 555 Specification to enforce MOT Schedules. These are very general numbers developed based on roughly 10% of estimated user costs. These numbers are updated every few years depending on stability of construction costs, stability of travel patterns and updates to AADT numbers. Updates are not automatic as the impact these numbers have on projects is carefully considered. Abrupt increases and decreases would make projects difficult to bid on. This process has been simplified to a form available online: UDOT User Cost Worksheet
Page updated
Report abuse