Estimating Volumes forI-95 HOT Lanes in Virginia

Prepared for:2009 Planning Applications ConferenceHouston, TXMay 18, 2009

Prepared by:

Kenneth D. Kaltenbach, P.E.The Corradino Group, Inc

Jeremy Raw, P.E.Virginia Department of Transportation

Transportation & Mobility Planning Division

May 18, 2009

Summary

What are HOT lanes? I-95 HOT Lanes in

Fredericksburg How do you model HOT lanes? Structure of the finished model Model performance

What are HOT lanes?

High-Occupancy Toll (HOT) lanes are a class of managed lanes.

Designated lanes, usually on a freeway Separated from general purpose lanes Toll charged for low-occupancy vehicles No toll for high-occupancy vehicles Electronic toll collection

… Sell the excess HOV capacity

What are the benefits?

I-95 is over capacity, widening for more general purpose is not likely

Encourage three-occupant autos Sell excess capacity in HOV lanes Create a new revenue stream by

charging those willing to pay for premium service – 50 mph guaranteed

HOT Lanes Modeling in Fredericksburg

HOT Lanes already operating in Northern Virginia (Washington DC area)

HOT Lane extension to Fredericksburg is a new project

Fredericksburg Area MPO (FAMPO) needed to include the project in LRP and Air Quality Conformity.

Had to be implemented quickly, about 90 days.

Extent of the Virginia I-95HOT Lanes Project

Model Effort

Fredericksburg Only Daily 24 hours

No mode choice

No Transit

Daily assignment

Extend net into DC Estimate AM, PM

and off-peak periods Discrete auto

occupancy levels Virginia Railway

Express Period assignment

with toll and HOT lane allocation

“VERSION 2.0 MODEL”

“VERSION 3.0 MODEL”

Model Requirements

Important Factors for HOT Lane ModelPoint of entrance/exitVariable tolling to control lane congestionHOT lane effects outside model area held

constant Include Virginia Rail Express ridership

effects (DC-oriented commuter rail)

Current FAMPO Version 2.0 Model Network

Northern External is Prince William Co. Line

Updated Network

I-95 External Station split into distance stations

External skims (HOT vs. Free) computed from NCRTPB model

Use of I-95 outside model area held constant

Northern External is Potomac River

Modeling Assumptions/Sources Distribution of trips on I-95 corridor outside

Fredericksburg area borrowed from National Capital Region model (DC)Compiled via select link analysis

Auto occupancy rates disaggregated from FAMPO model

Time of Day directional traffic from VDOT traffic counts

Toll/free split borrowed from I-394 MnPASS Express Lane Stated Preference Survey

Modeling Assumptions

Length of trip outside area influences HOT lane use within area

Vehicle occupancy alters sensitivity to tollSOV, HOV-2, HOV-3

HOT versus free lane choice made in assignment

Model AM, PM and Off-Peak periods separately Value of time based on NCRTPB model, but

adjusted to 2008 values ($10.24 - $17.06 / hr)

Operating Assumptions

Goal: at least 50 mph in HOT lanes

3+ person carpools and buses: No Toll

2-person carpools and drive alone: Pay Toll

Directional operating plan for HOT lanes to conform with DC peak directions (AM-NB, PM-SB, Off-split)

Transponders -- no toll collection delay

Toll Diversion

TOLL DIVERSION CURVE

Toll($)/time saved (min) Fraction Toll

0 1.000

2 0.900

5 0.400

10 0.280

20 0.100

30 0.050

40 0.020

50 0.018

60 0.015

70 0.013

80 0.010

Toll Choice (from MnPass)

0.000

0.200

0.400

0.600

0.800

1.000

1.200

0 20 40 60 80 100

Toll $/minute saved

Frac

tion

Toll

Source: I-394 MnPASS Express Lane Stated Preference Survey

Toll Calculation

Function of the V/C on the managed lanes

Logit function Several toll

ranges explored Final toll rate

ranged between $.00 and $0.25 per mile

Toll Function

0.00

0.05

0.10

0.15

0.20

0.25

0.30

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1 1.2

Volume/Capacity

Toll $

per M

ile

Commuter RailVirginia Railway Express

VRE post-processing model Evaluates impact of HOT lanes on

VRE Incremental logit model AM peak period trips only Cube Voyager PT network

Model Performance and Sensitivity

Validation Sensitivity Toll Estimation HOT lane volumes

Validation

Volume Group Volume Range RMSE Number of Observations

1 1- 5,000 56.0% 480

2 5,000- 10,000 29.1% 128

3 10,000- 20,000 18.7% 102

4 20,000- 30,000 31.0% 30

5 30,000- 40,000 18.3% 36

6 40,000- 50,000 20.6% 6

7 50,000- 60,000 9.9% 18

8 60,000- 70,000 20.3% 4

10 80,000- 90,000 18.0% 2

Total 1-500,000 34.7% 806

Usual 4-step methods – for example …

Model Sensitivity

AM NORTHBOUND DIRECTION TRIPS LOADING AT FREDERICKSBURG, LEELAND RD., AND BROOKE STATIONS

AM NORTHBOUND

I-95 VEHICLES* YEAR ALTERNATIVE VRE BOARDINGS GP HOT TOTAL

VRE/ I-95

RATIO NOTE

2000 BASE - CALIBRATION 1,780 13,961 0 13,961 0.13 SURVEY/ BOARDING

COUNTS SAY 1,779 2015 BASE 3,703 20,104 0 20,104 0.18

2015 HOT 3,345 12,877 7,226 20,103 0.17 WITHOUT HOT LANES, VRE SHARE INCREASES

2025 BASE 5,293 24,148 0 24,148 0.22 AS CONGESTION INCREASES.

2025 HOT 4,195 14,976 9,267 24,243 0.17

2035 BASE 7,737 28,365 0 28,365 0.27 HOT LANES TAKE SOME VRE TRIPS, BUT

2035 HOT 4,972 18,164 10,203 28,367 0.18 VRE TRIPS STILL INCREASE.

* = At the Fredericksburg northern external station

Toll Estimation

AVERAGE SPEED (MPH) MINIMUM SPEED (MPH) Facility AM PM Off-Peak AM PM Off-Peak

HOT LANES 59.84 64.58 64.94 38.75 62.27 64.38

Other Freeways 55.71 43.76 55.41 30 15.41 30 Revenue $16,313 $33,492 $31,272

Acceptable Toll Range=$ 0.05- 0.25 per mile

Peak Period Off-Peak Period Drive Alone $3.52 Drive Alone $5.02

HOV-2 $4.39 HOV-2 $5.86 Daily revenue= $81,077

AM HOT vs. GP Traffic at Fredericksburg North I-95 Station, Northbound

HOT 9,279

GP 14,963

HOT Lane Volumes

HOT lanes coded explicitly

Multi-class assignment by vehicle occupancy

HOT Lanes

Band Width Plot

Next Steps

Model performed well! Despite tight Air Quality conformity

timeframe Additional improvements coming up

Recalibrate from new NHTS Time of Day Modeling (AM, PM, Off-Peak) Full Mode Choice & Vehicle Occupancy

Model Spatially improved network

Questions?

Thank You