Establishing a Metropolitan Network of CongestionFree Freeways Eliminating Freeway Bottlenecks in th

1
Establishing a Metropolitan Network of
Congestion-Free FreewaysEliminating Freeway
Bottlenecks in the Short-Term Without Major
Construction
Patrick DeCorla-Souza Federal Highway Administrat
ion TRB Transportation Applications Conference
May, 2007

2
Overview

• The freeway congestion dilemma
• Addressing traffic flow breakdown with a new
  approach to transportation system operation
  
• Costs, revenues and benefits

3
As Much as Half of Freeway Capacity is Wasted in
Rush Hours

• SR 91 Express Toll Lanes
• Vehicle throughput is double in the peak hour
• Free-flow speed vs. 15-20 mph in free lanes

4
Eastbound I-66, Northern VA Monday Morning,
March 5, 2007
Outside Capital Beltway
5
Travel Time on Eastbound I-66 7am to 10 am
6
Travel Time with No Flow Breakdown(for the same
19,000 vehicles)
Previous freeway traffic only. Recovered
capacity of 5,000 vehicles will attract
additional drivers
7
Annual Benefits of Free Flow on I-66
Additional benefits anticipated 9 10 am due to
recovered capacity (5,000 vehicles), including
(1) benefits to 5,000 vehicles (2) congestion
reduction on alternative routes.
At 12 per hour
8
The Problem is Solvable

• Keeping a small number of vehicles from arriving
  at critical times can prevent flow breakdown
  
• Evidence
• Washington DC in August
• State holiday in California
• Jewish holiday in Boston
• In each case less than 10 fewer vehicles

9
Can a Few Vehicles Shift Arrival Time?

• National Household travel Survey
• 20 of vehicle trips made during peak travel
  periods are made SOLELY to shop
  
• Total non-work travel constitutes
• 49 of vehicle trips in AM peak period
• 75 of vehicle trips in PM peak period

Source USDOT, FHWA National Household Travel
Survey 2001 Includes only trips not made durin
g a commute tour
10
Can a Few Vehicles Shift Arrival Time?

• Options for work trips
• Shift to transit
• Vanpool or carpool
• Use flextime, or telecommute
• Key is to get a few vehicles off the freeway
  during the critical period
  

11
Can We Prevent Flow Breakdown?

• The three pull strategies (transit, ridesharing
  and flextime/telecommuting) are necessary but not
  sufficient
  
• Any reduction in delays due to such shifts will
  cause replacement of reduced traffic by those
  who had previously been deterred by congestion
  
• Must ensure that user-borne cost of driving stays
  about the same as before
  
• Two ways (1) ramp metering (2) congestion
  pricing

12
(1) Ramp Metering

• Controls merging traffic at freeway entrance
  ramps, using time delay as a price to enter the
  freeway
  
• Main issues
• Queues can back up onto arterials and cause
  congestion
  
• Drivers frustrated by long waits
• Those living further out have advantage over
  those living in central city
  
• Motorists may choose to divert to alternate
  routes
  

13
(2) Congestion Pricing

• Variable toll price to manage demand for use of
  freeway
  
• Unlike time price in ramp metering, user-borne
  toll cost is not a wasted resource revenue
  can generate additional benefits
  
• Main issues
• Motorist perceives that free service will be
  taken away
  
• Some drivers may be made worse off

14
(3) Super HOT Congestion Pricing

• Toll-bypass lane plus modal choices, gives
  options
  
• Wait in toll-bypass lane and pay with time
• Go to transfer area and take fast, discounted
  transit
  
• Ride toll-free in a certified vanpool or carpool
• Pay variable toll and get guaranteed premium
  service
  
• Pay lower tolls or no toll by traveling earlier
  or later
  

15
Super HOT Pricing

• Dynamic pricing
• Can manage flow for optimum throughput in real
  time
  
• Can respond immediately when spare capacity is
  available by reducing toll rates, benefiting the
  motorist
  
• ALL lanes priced
• When only one or two lanes are priced, tolls must
  be high enough to price off a majority of
  users
  
• When all lanes are priced, only about 10 of
  users must be priced off at the critical
  breakdown time so tolls are a lot lower, and
  90 are willing to pay them

16
Super HOT Incident Management

• Overhead lane controls at short intervals
• During incident, one clear lane restricted for
  buses and HOVs
  
• Any spare capacity in the restricted lane managed
  for free-flow using an additional variable toll
  for SOVs willing to pay for premium service
  
• All other vehicles credited with amount of toll
  paid, if guaranteed speed is not experienced
  

17
Toll-Bypass Lanes

• With 10-15 of travelers shifted to other modes
  or times-of-day, all remaining vehicles could be
  accommodated, so toll-bypass lanes would have
  zero delays to begin with
• This would encourage more drivers to use the
  lane, until the queue is long enough that
  queue-delay cost, as perceived by those with
  the lowest value of time, would be equal to the
  toll rate
• Assuming a 5 min. queue delay, and a queue
  discharge rate of 15 vehicles per min., the
  toll-bypass lane would need to accommodate 75
  vehicles

18
Super HOT Concept Summary

• Manage traffic flow on all lanes of a freeway
• Variable tolls
• At bottleneck locations only
• During peak periods only
• Complementary strategies
• Operations ramp metering, toll collection, lane
  controls
  
• Discounted transit services
• Certified vanpools and carpools toll-free
• Telecommuting and flextime
• Multi-modal traveler information

19
Advantages

• Reduces congestion now on whole freeway network
• Creates an HOV network and a fixed guideway
  transit network practically overnight
  
• Advantages relative to priced lane networks
• Costs lower less right-of-way, pavement, easier
  freeway-to-freeway connections
  
• Capacity per lane higher
• Safety less weaving
• User cost - lower toll for premium service
• Entire facility congestion-free

20
Feasibility of a Super HOT Highway Network

• Toll rates
• Costs to implement
• Revenues
• Social benefits

21
Prototypical Urban Network

• 100-mile, severely congested freeway network,
  avg. 6 lanes
  
• Price all lanes peak periods only
• New express bus services with heavily discounted
  transit fares
  
• Park-and-ride at transit transfer sites
• Toll-bypass lanes at toll gantries at about
  5-mile intervals
  

22
Average Peak Period Toll Rate for 10-mile Freeway
Segment
Based on average value of time of 12.00, 50 of
travelers value their time equal to or higher
than 12.00 per hour
23
Annual Highway Costs for 100-mile Network
(million )
Based on total capital cost of 127 million,
covering operations technology and toll-bypass
lanes
24
Annual Transit Costs for 100-Mile Network
(million )
Annualized cost, including annualized capital
cost plus annual maintenance costs
25
Annual Costs for 100-mile Multimodal Network
(million )
26
Annual Benefits for 100-mile Multimodal Network
(million )
At 12 per hour Freeway only does not includ
e potential reductions due to reduced arterial
congestion
27
What This Means

• Use of congestion delay to allocate scarce
  roadway space during rush hours costs 370
  million a year
  
• Using a Super HOT pricing system instead would
  cost only 87 million annually, and would in
  addition provide other economic and quality of
  life benefits

28
Costs vs. Revenues and Benefits (million )
29
What This Means

• Super HOT highway network can be financially
  self-sufficient
  
• Benefits will exceed costs by a 4 to1 ratio
  (conservatively estimated).
  
• Potential to get public acceptance since no one
  is actually made worse off although some could
  still perceive differently