The Texas
Highway Man

Home Search Miscellany Updates About me Contact Menu
Other San Antonio Area Roads
Restricted Crossing U‑turn (RCUT) Intersections

This page last updated February 19, 2024
RCUTs map

RCUTs map A "Restricted Crossing U‑turn" intersection (RCUT) — also known as a "superstreet" — is a divided roadway where the major intersections have been modified to eliminate left turns and straight-through traffic on the road that intersects the main roadway, and where minor intersections are converted to "right-in/right-out", i.e. traffic can only turn right onto or from those streets. Turnarounds are then provided to facilitate left turns and crossovers via a U‑turn instead.

This configuration improves traffic flow and safety. At the signalized intersections, doing this reduces the number of traffic signal phases required to move traffic through the intersections from typically five or more to just two, thereby allowing for longer green times for all movements without extending the overall cycle time, thus reducing wait times and the resulting congestion.

This is one of several "innovative" or "alternative" intersection types used to improve intersection throughput when grade-separation (overpasses or flyovers) is not possible, not warranted, or too expensive.

The San Antonio area has had four sections of RCUTs: one on US 281 North (four intersections, replaced when US 281 was upgraded to a freeway), one on Loop 1604 West (two intersections, also replaced by a freeway upgrade), one on Bandera Rd. (currently one intersection; construction underway on two more), and Loop 337 in New Braunfels (three intersections.) In the 281 and 1604 cases, the RCUTs were in a transition area between freeway and non-freeway sections of those roads, and both were intended to be short-term improvements while the upgrade of those segments to freeways was stalled due to various legal challenges and funding issues. The 281 and 1604 RCUTs were the first two superstreets in Texas.

An RCUT intersection was completed on Bandera Rd. at FM 1560 South in 2018. This was the first of several RCUT intersections planned for Bandera Rd. from Loop 1604 to Triana Parkway. Unlike those on US 281 and on Loop 1604, the RCUT plan on Bandera is considered to be a long-term solution for existing and expected congestion. It now appears that RCUTs will also be the plan for Bandera Rd. between Loop 1604 and Loop 410.

Finally, another RCUT section is planned on Loop 1604 between US 90 and Macdona Lacoste Rd., where it is also intended to be a long-term solution.


On this page:

Local RCUT history

US 281
In February 2009, with the proposed US 281 tollway stalled due to ongoing litigation that resulted in a new environmental study being required, engineers with Pape-Dawson Engineering made an unsolicited proposal to the Alamo Regional Mobility Authority (ARMA) to convert US 281 into a "superstreet" from Encino Rio to Marshall as a short-term fix while the lawsuits played out and the new environmental study for the long-delayed expansion project on 281 was completed. Modeling showed the superstreet plan could increase traffic throughput on 281 by 30%, increase average rush hour speeds by about 10 mph in both directions, and reduce travel time by about eight minutes between Loop 1604 and Marshall Rd.

The MPO approved funding for the project on March 23rd, 2009. Funding came from a combination of funds from the federal stimulus, Advanced Transportation District, and City of San Antonio. On January 14th, 2010, ARMA's board selected Ballenger Construction's bid of $5.2 million to build the project, which was nearly 20% below the final estimates and nearly 50% below initial projections. Ground was broken on the project on March 11th, 2010.

The first intersection (Encino Rio) was completed on August 29th, 2010. Weather delayed work on the Marshall Rd. intersection; it was finally completed on September 26th. The remaining intersections were completed on the weekends of October 2nd and 9th. As planned, the RCUT intersections were subsequently removed when the corridor was upgraded to a freeway in 2021.

Loop 1604
In October 2009, with plans for the US 281 RCUTs nearing completion, the Bexar County Commissioners Court approved funding to study a possible RCUT section in the Loop 1604 West corridor, specifically between Braun Rd. and SH 151, as a possible interim solution to ease congestion there while ARMA completed an environmental study for an expansion. The superstreet study, completed in early 2010, showed that a combination of conventional improvements (mainly additional through lanes at Braun) and RCUT intersections along that section of 1604 would provide substantial improvements.

In May 2010, ARMA previewed the initial designs and expected benefits. Under the plan, the intersections of New Guilbeau Rd. and Shaenfield Rd. would be reconstructed as RCUT intersections. It was estimated that by doing so, travel speeds could be increased by 36% and delays reduced by 69% during the morning peak, and that average travel speeds could be increased by 65% and delays reduced by 79% during the evening peak period.

The project began construction in March 2011. The New Guilbeau intersection was completed August 18th, followed on September 15th by the Shaenfield intersection. As intended, the RCUT intersections were subsequently removed when the corridor was upgraded to a freeway in 2016.

In 2022, earlier plans for an expansion of Loop 1604 West from US 90 to Macdona Lacoste Rd. to a divided highway were updated to include a full RCUT configuration. Construction is expected to start in 2026.

Bandera Road
TxDOT has proposed two RCUT segments along Bandera Rd. One is through Helotes from Loop 1604 to Triana Parkway. Construction on the first intersection — at FM 1560 South — was completed in October 2018. This intersection was the first to be done because a project was already in the works to reconfigure the obsolete and congested dogleg intersection at Circle A Trail, so the RCUT design was incorporated into that project so that the new intersection wouldn't have to be torn-out and redone when the rest of the RCUT corridor is built in the future.

The next section — from Loop 1604 to FM 1560 — is currently under construction with completion set for late 2024. The remaining intersections from FM 1560 to Triana are currently unfunded and therefore there is currently no timeline for their completion.

The second proposed segment for Bandera Rd. was between Loop 410 and Loop 1604, but was shelved after Leon Valley declined to endorse it. However, an RCUT corridor is now the recommended alternative after other proposals were studied. Leon Valley has endorsed it this time around. There is currently no funding or timeline for its implementation.

Loop 337 (New Braunfels)
Work was completed in 2021 to expand a section of Loop 337 in New Braunfels from SH46 West to River Road to a four lane divided highway with RCUT intersections.


Please note that the intersection diagrams below are generalized for the types of intersections involved and not intended to reflect specific existing or future conditions. They are intended instead to illustrate the concepts being discussed. Thanks to "SPUI" for the base RCUT intersection illustration used in the intersection diagrams.

Conventional intersection operation

Before explaining how an RCUT works, it's important to understand how conventional intersections function, especially with regards to the traffic signal operation. Understanding the signal phasing is vital to grasping the improvements provided by an RCUT.

Question mark sign
What is a signal phase?

A signal phase is the green time assigned to a specified movement or collection of simultaneous movements in a traffic signal cycle. In other words, when the signal is green for a specific movement (straight through, left turn, etc.), that's a signal phase. When it changes to red and another movement gets a green signal, that's another phase. The complete rotation through of all of the phases is a cycle.

This is how a typical conventional intersection is configured; the road going left-to-right is the main road:


Typical conventional intersection
(Base graphic from Wikipedia modified by Brian Purcell)

The "split-phase" signal configuration often used at these intersections requires five signal phases (i.e. signal changes) per cycle (or three phases for "T" intersections) to allow for each movement through the intersection:

Phase 1


Phase 2


Phase 3


Phase 4


Phase 5


So, if you get stopped just as the light is turning red, you have to wait for three or four other directions to have a green light before you get to go. Some signal plans can even have as many as eight phases per cycle!

In a congested intersection, the number of phases and the required length of each phase due to heavy traffic volumes causes overall signal cycles to be long, thus causing long queues and congestion, especially during peak periods. Additionally, each signal phase adds up to 10 or so seconds of "lost" time due to yellow and all-way red clearance intervals, as well as the time it takes for stopped motorists to react to a green and to start moving.

Many times, people will say that the signal timings simply need to be adjusted in order to reduce the congestion. In some cases, signal timings can be further optimized to improve conditions. However, in most cases, signals are already optimized as much as they can be, and the laws of physics prevent wringing-out any further efficiency in the existing intersection. When this happens, the intersection must be physically changed.

How an RCUT intersection works

An RCUT reconfigures intersections like this:

Typical RCUT intersection diagram

Typical RCUT intersection
(Base graphic from Wikipedia modified by Brian Purcell)

At first glance, this looks quite complicated, perhaps overly so. But it actually makes sense once you understand how it works and why.

As you can see on the schematic, an RCUT intersection eliminates the ability of traffic on the cross street to continue straight through the intersection or to turn left (that's the "restricted crossing" part of the name.) Instead, all traffic on the cross street must turn right onto the main road, then use a downstream turnaround (the "U‑turn" part of the name) to either go the opposite direction or to return to the cross street to continue on it. Those turnarounds are typically located approximately 1000 feet (about 210th of a mile) from the intersection they serve to allow sufficient room for merging and storage.

So how does this help? By eliminating the through-traffic and left-turns from the cross streets, the signal phases required for those movements can also be eliminated (or, more specifically, they're combined with the phase for traffic turning left from the main road onto the cross street.) It is by eliminating those signal phases that the benefit is derived (keep reading.)

While RCUTs in areas with low traffic volumes can operate fine with stop or yield signs, most RCUTs (including all those in San Antonio to date) are signalized, and, in fact, it actually doubles the number of signals in each direction. Yes, it seems counter-intuitive that adding signals can help improve traffic flow. But it's not the number of signals that causes problems; it's the time required to service each movement at the intersection that causes backups. By eliminating the straight-through and left turn movements on the cross street, the number of signal phases (green time for a movement) is reduced from the previous five phases to just two (or from three to two at "T" intersections):

Phase 1


Phase 2


That's it — just two signal phases! Notice how many movements are accomplished in the second phase. This essentially takes phases 4 and 5 and the left turn movements of phases 1 and 3 of the conventional intersection discussed above and combines them all into one, eliminating all the time required for those individual phases.

Ah ha!

If all of that sounds like technical gibberish, here's another way of explaining it:

At the conventional intersection, drivers who arrive as the light turns red have to wait for three or four other directions to get a green light before their light turns green again. At an RCUT intersection, drivers only have to wait for one signal change before they get a green again. Drivers then needing to use the turnaround may have to subsequently wait for one more signal change, but that's still less than before.

Furthermore, significant additional time savings is realized because every signal phase inherently includes some "wasted" time in the form of the yellow and all-way red clearance intervals, as well as the time it takes for drivers to start moving when they get a green light. All of that can add 5-10 seconds of lost time per phase.

All of these time savings allow the green time for through traffic on the main road to be increased without having to correspondingly reduce the green time for the cross street or lengthen the total cycle time.

Note on schematic simplification
It should be noted at this point that the diagrams above for the RCUT signal phases are simplified to show the corresponding phases on both sides of the main road operating simultaneously in order to demonstrate the overall concept. But in practice, with an RCUT, the signals on each side of the main road can operate independently of those on the other side, so the phases are usually staggered to better optimize the operation of each signal based on the traffic volumes at that time and location and to coordinate with other signals in the corridor.

RCUT "secret sauce" further explained
If the explanation above is still a bit muddy, here's another way of looking at it.

Imagine in the diagrams above that you're on the main road headed from left to right. You reach the conventional intersection just as the light turns red (phase 3.) You then have to wait through phase 3 for 20 seconds, phase 4 for 30 seconds, and phase 5 also for 30 seconds before you get a green light again on phase 1, for a total wait time of 80 seconds.

With the RCUT, all of the movements that took 80 seconds before can now move at once and therefore be accomplished in a single 30 second phase, which means that through traffic on the main road is now only stopped for 30 seconds per cycle instead of 80 seconds. The 50 seconds of time that's been freed-up by eliminating two phases can now be added to the green time for the main road. So if through traffic previously was allocated 40 seconds of green time, it can get 90 seconds with the RCUT without any change to the overall two minute cycle time. This an obvious and substantial improvement and is where the benefit of a RCUT is derived. See the box below for a graphic visualization of these timings. (Note that the timings I provide here are for illustrative purposes only and do not necessarily reflect any specific actual current or previous phase and cycle times.)

Example signal timings

Below are simplified timelines showing example signal phase timings for each movement in an RCUT intersection from the perspective of through traffic on the main road moving left to right in the diagrams above. Note that these timings are simplified for illustrative purposes only; they do not represent actual timings, and the actual phasing is a bit more complex and can vary throughout the day. Be sure to read the explanation above.

Example timing timelines

Note that the 30 seconds for the cross street in the RCUT could then be increased to further reduce congestion on those approaches if necessary and still leave significantly more green time for the main road than it would have in a conventional intersection. For example, the cross street could be allocated 40 seconds of green time, leaving 80 seconds for the main road, which is still twice as much as it had before. This demonstrates the additional flexibility afforded by this design.

If that's still not making sense, below is yet another way of looking at it.

Example Bandera Rd. - Hausman Rd./Leslie Rd. signal timings

  • The whole pie below represents the time for a full signal cycle (~150 seconds.)
  • The pieces of the pie show the proportion of green time each movement gets during each cycle.
  • Blue sections represent Bandera Rd. traffic.
  • With an RCUT, all the movements except the through movements on Bandera get combined, and, in this example, are combined into the biggest piece of the pie from the conventional intersection. After doing so, that slice can (and probably will) be increased as needed to better accommodate Hausman/Leslie traffic, and even after doing so, everyone still gets more green time than they had in the conventional intersection.
  • Current timing splits are approximate as they change during the day, but are typical.
  • Timings were obtained by yours truly by monitoring the intersection for three half-hour periods during a Wednesday in April 2022.
  • Finally, in an RCUT, the signals on each side of Bandera Rd. can be timed independently of the other, and thus would be two separate pies, so the RCUT chart below combines both directions for simplicity.
Current signal phasing pie chart
RCUT signal phasing pie chart

Blue sections represent Bandera Rd. traffic
The other sections represent the other intersection movements relative to Bandera Rd.

Better synchronization
In addition to reducing signal phases, RCUTs allow for much better synchronization of the signals. With conventional intersections, signals along a corridor in both directions are inherently inter-dependent — in other words, changes cannot be made to one direction without affecting the other. This dependency makes signal coordination and signal progression (or "synchronization") tricky, especially two-way progression.

But with an RCUT, all the movements on one side of the intersection don't affect the other side at all, so the signals for traffic going in one direction on the main roadway are completely independent of those for the other direction. In effect, each side of the main road functions like an independent one-way street. This provides the ability to coordinate the signals much more easily and reliably without regards to the number of signals or their spacing, and the signals can react better to the traffic conditions on their side of the road. Essentially, it provides the optimal environment for signal coordination.

Traffic flow through RCUT intersections
While some of the movements through an RCUT intersection are the same as a conventional intersection, some are quite different and a bit unintuitive. Below are diagrams depicting the various routes through an RCUT intersection.

Straight-through traffic on the main road and traffic turning right onto the cross street works just like a conventional intersection:


Traffic turning left from the main road to the cross street also works just like a conventional intersection:


And traffic turning right onto the main road from the cross street works just like a conventional intersection:


However, traffic wanting to turn left from the cross street onto the main road instead turns right, then uses the downstream turnaround to head the other direction:


Similarly, traffic wanting to continue straight on the cross street across the main road instead has to turn right, use the downstream turnaround, then turn right again to get back onto the cross street:


Right on red prohibition
Although not required to make an RCUT function, RCUTs in San Antonio typically employ a full or partial right-on-red prohibition from the cross street. The reason why is two-fold and has to do with the fact that, at an RCUT intersection, many of the vehicles turning right will then subsequently be making a U‑turn. Doing this requires them to completely cross the through-traffic lanes to reach the turnaround in the center of the main road. Since a vehicle that makes a right turn on red will be traveling substantially slower than the through traffic and doesn't have a lot of space to get up to speed before merging over, those drivers can create a significant safety hazard as they cross the through-traffic lanes. Even if there isn't a collision, vehicles in the through lanes must slow down when encountering someone cutting across to the turnaround; in heavy traffic conditions, this can be the genesis of congestion.

Also, restricting when vehicles can cross helps to create a "platoon" of vehicles when the light turns green. When signals are coordinated, they are timed to more efficiently move platoons of vehicles, so vehicles crossing over individually while the light is red often won't get that benefit (i.e. still have to wait for the signal at the turnaround anyway.) Yes, this does inconvenience drivers who just want to turn right and not make the U‑turn, so where possible, the far right lane is allowed to make a right turn on red since drivers in that lane are very unlikely to be headed to the turnaround. This was done at a couple of intersections on US 281 and will be done at two intersections in the Helotes RCUT project. Right on red was also allowed for the Bandera/Circle A Trail intersections in Helotes due to the low traffic volumes there.




How an RCUT differs from a "Michigan Left"

People who have lived in Michigan and other parts of the country may be familiar with an intersection layout known as a "Michigan Left", also known as a "Median U‑turn or "MUT", and some have proffered that an RCUT is the same thing just with a different name. However, while a Michigan Left is similar to a RCUT, there are some significant differences between the two:

Below are simplified schematics that show the allowed maneuvers in each type of intersection:

RCUT ("Superstreet") traffic flow
MUT ("Michigan Left") traffic flow
When both roadways have heavy traffic volumes, U‑turns can also be provided on the cross street.
(Graphics by Brian Purcell)

RCUT ("Superstreet") traffic flow

MUT ("Michigan Left") traffic flow
When both roadways have heavy traffic volumes, U‑turns can also be provided on the cross street.
(Graphics by Brian Purcell)

Local RCUT results

US 281
To objectively measure whether the US 281 RCUT had the intended results, engineers did before-and-after measurements of travel times, average speeds, and traffic volumes. The study found that after the completion of the RCUT, travel times were notably reduced and average speeds increased even though overall traffic volumes increased in the corridor. The table below summarizes the findings. To ensure an "apples-to-apples" comparison, data was collected for Tuesday-Thursday periods when school was in session and when there were no abnormal incidents in the corridor (e.g. accidents, signal malfunctions, bad weather, etc.)

Metric Before RCUT After RCUT
Southbound travel time
(Bulverde to Loop 1604, morning rush hour)
23.3 minutes 19.2 minutes
Southbound average speed
(morning rush hour)
16 mph 19 mph
Northbound travel time
(Loop 1604 to Bulverde, evening rush hour)
19.2 minutes 12.7 minutes
Northbound average speed
(evening rush hour)
19 mph 29 mph
Traffic count (north of Evans) 60,100 vehicles/day 63,552 vehicles/day
Traffic count (south of Evans) 74,000 vehicles/day 81,526 vehicles/day

Anecdotal reports for the first few weeks from motorists in the corridor on social media, from media reports, and comments on various websites indicated that most of the time, the RCUTs provided some appreciable relief. Recurring traffic signal malfunctions for the first month or so after completion of the project did cause some initial problems; those were later attributed to the fact that the signal controllers used for the project were a new model, and the issues were resolved with assistance from the traffic signal equipment vendor.

Furthermore, because of their ability to better handle traffic, the RCUTs were kept during construction to upgrade US 281 to a freeway.

Loop 1604
Over on Loop 1604, a before-and-after study showed a reduction in peak period travel time of 35% southbound and 14% northbound. The primary reason for the significant discrepancy between northbound and southbound benefits was likely caused by an increase in traffic (induced or latent demand) drawn by the improved conditions of the RCUTs exceeding the capacity of the ancillary improvements made at the Braun intersection. This caused traffic to back up at the Braun intersection and cut into the travel time improvements of the RCUTs.

Anecdotally, the Loop 1604 superstreet had a more profound reduction in congestion than the one on US 281 due to the lower traffic volumes on Loop 1604. The most glaring example of the improvement was on weekends. Before the superstreet, the entire segment from Braun to Shaenfield was typically congested in both directions most of Saturday and Sunday afternoon; after the superstreet, there was virtually no weekend congestion on that segment.

Other sites of interest

Wikipedia - Superstreet
TxDOT - Virtual Public Meeting - SH 16 Helotes
Pape-Dawson "Superstreets in Texas" presentation
NC Department of Transportation presentation on superstreets
(Excellent explanation of all aspects of superstreets)
Utah DOT - ThrU-Turn Follow-Up
(Excellent video showing outcome of a Michigan Left project in Draper, Utah)
Federal Highway Administration - RCUT/MUT intersections
Federal Highway Administration - Restricted Crossing U‑turn Intersection
An Update on Superstreet Implementation and Research (Report)
An Update on Superstreet Implementation and Research (Presentation)
Virginia Department of Transportation - Restricted Crossing U‑turn