EXPERIMENT:Ramp meters in Minnesota


EXPERIMENT:Ramp meters in Minnesota

Originally published in issue 52 of Tollroads Newsletter, which came out in Nov 2000.

Page:26

Subjects:ramp meters

Agencies:Cambridge Systematics

Locations:Minneapolis-StPaul

Sources:Cutler

Minneapolis/St Paul have had perhaps the world’s most dense system of ramp meters: 450 of them covering almost the whole motorway network. They can work brilliantly for the mainline.

MinnStP maintains a freeflowing system with some amazing traffic volumes moved per lane – 2600 to 2800 veh/lane/hr are quite common. But this has been achieved at the price of subjecting users to extraordinary waits on the ramps. 10mins to 20mins is common. Motorist frustration eventually generated pressure to get a legislated switch-off of the ramp meters, and a before-&-after study which is being done by Cambridge Systematics in Boston.

The MinnStP ramp meters were turned off from Oct 16 to Dec 8. The results seem to have been a modest vindication of meters, but a reprimand to the state DOT for poor management of them. Cambridge Sys is a couple of months away from finalizing their report, but Marc C. Culter of the study kindly gave us his informal impressions on the phone. He says that without the meters there WAS some degradation of overall level of service, lower average mainline speeds and some overall increase in door-to-door travel times. It was uneven however. Quite a significant number of meters were working counter-productively – increasing wait times without improving mainline flow. So where those were turned off, things improved for everyone.

There was also an increase in accidents at the merges without meters, Cutler says. None of the accidents that occurred were serious, but over time you might get serious accidents. So the experiment did support the safety value of ramp meters as a means of aiding merges.

Managing volume problematic

Less vindicated was the idea of ramp meters as a mean of managing the volume of entering traffic. Cutler says the system was never equipped with the sensors to do this properly and he doubts the long ramp waits required are acceptable.

What seems to have emerged is a more modest-aiming ramp meter system. About 50 counterproductive ramp meters will be abandoned. The remaining ramp meters are being modified. They will operate shorter hours and their redtime will be capped at 13secs (previously often set to 20secs).

Dennis Eyler, a former MNDOT engineer now with SRF Consulting, says ramp meters in MinnStP have come to be used in a number of places as a mask to cover over serious capacity deficiencies. Unwarranted lane drops on the mainline have been protected by meters. The Ventura administration has taken this to heart and there is now a program for selective widening of the expressways at bottlenecks.

Eyler, too, says the ramp meter system has too few sensors ( ramp-end loops in this case) and management of the system has not been impressive. Still with the meters down completely he found his drive to the airport took 100mins vs 60mins with the meters operating.

‘Sprawlmakers”

Eyler says aggressive use of the meters undoubtedly influences the use of the system. It favors long distance and early joining drivers over those traveling shorter distances and joining midway or closer to the ends. This generates a sense of unfairness. It has also spawned the cry that ramp meters encourage the dread sprawl, so it has turned many enviros against them.

A market survey done by Cambridge Sys endorses the moderated ramp metering. Only 10% of the sample wanted to keep all the meters off, while 20% wanted the system restored to its pre-Oct 16 status. 70% favored a modified system.

The report seems likely to emphasize the need for better continuing traffic data from the system, more monitoring of operations and willingness to change them, and periodic independent review. The clearest benefits of the meters seem to be as an aid to merges, so they have a safety role. By reducing merge turbulence they can allow higher volumes in the mainlanes before flow breaks down, so they have some beneficial impact on capacity. But the benefits run out after a point as ramp waits get too protracted. Too much has been expected o the ramp meters in the past.

If a highway is overloaded there seems to be no alternative to working directly on supply, or demand, or both. A varied toll to alter demand, and/or more lanes to increase supply. The ramp meters have a useful spot role. (Contacts Marc Cutler, Camb Sys 617 354 0167, Dennis Eyler SRF 763 475 0010)