LONG-TERM DEMAND FORECASTING OF MANAGED LANES

LONG-TERM DEMAND FORECASTING OF MANAGED LANES

Christopher Mwalwanda

13th TRB Transportation Planning Applications ConferenceMay 10, 2011

Challenges in Addressing Key Influential Risk Parameters

• More Complex than Traditional Forecasting– Competition Conditions are immediately apparent

• More Data for Operational Assessments– Public Behavioral Characteristics– Geometrical Consideration/Travel Speed

Deterioration Analysis– Time of Day Profiling

• Eligibility and Pricing Options– Operational Demand Management versus

Revenue Generation

MANAGED LANE FORECASTING 101

SR 167, Seattle, WA•2008

I-680, Alameda, CA•2010

SR 91, Orange, CA•1995

I-15, San Diego, CA•1998

Houston, TX•US 290 QuickRide 1998 •I-10 Katy Freeway Managed Lanes, 2009

I-95, Miami, FL•2008

Minneapolis, MN•I-394 , 2005•I-35W, 2009

I-15, Salt Lake, UT•2006

I-25, Denver, CO•2006

OPERATING MANAGED LANE PROJECTS

•I-580•SR 237•SR 85 & US 101

•IH-635 /LBJ•NTE

I-95 Section 100

Route 495 Lincoln Tunnel

Atlanta (Various)

I-405

US 36

US 290

Existing Managed Lanes Projects

Planned or Under ConstructionStudied

I-595

I-25 North

RECENT HOT/MANAGED LANE PROJECTS

MoPacLoop 1

• New and Innovative Demand Management Techniques– Dynamic Speed Limits/Dynamic Re-striping– Shoulder Lane Utilization– GPS/Dynamic Re-routing Procedures

• How does one develop a forecast?– Point forecasts for financial feasibility– Ranges for procurement assessment

FORECASTING CHALLENGES

MANAGED LANE POLICIES

• HOV’s• HOT’s• ETL’s• TOT’s

Facility Type

PricingTypeFacility Location Comments

Fixed Variable Rates

SR 91 Orange County, CA ETL's Preset Varies by day of week and hour of day

I-25 HOT Lanes Denver, CO ETL's (HOT) Preset HOV's free – reversible/Free Flow for Buses

I-95 Express Lanes Miami, FL ETL's (HOT)

Dynamic Pricing

I-15 Managed Lanes San Diego, CA ETL's (HOT) Dynamic Must keep free flow for HOV

I-394 MNPASS Minneapolis, MN ETL's (HOT) Dynamic Must keep free flow for HOV

SR 167 Seattle, WA ETL's (HOT) Dynamic Must keep free flow for HOV

IH 10 Toll Lanes Houston, TX

I-15 Managed Lanes Salt Lake City, UT ETL's (HOT) Dynamic Must keep free flow for HOVDynamic* Registered HOV

ETL's (HOT) Preset HOV's free during peak periods

VARIABLE PRICING EXAMPLES

Project Name Length

Lanes Daily VolumeAnnual Revenue (million)

Tolling PolicyML GP ML (000) GP (000)

SR 167_WA 9 2 4 2 - 2.3 112 - 115 $0.4 - $0.5 HOV2+ free

I-394_MN* 11 1/2 4 4 - 4.5 150 – 160 $1.4 - $1.6 HOV2+ free

I-25_CO* 7 2 8 4 – 5 220 – 230 $2.0 - $2.5 HOV2+ free

IH 10_TX 12 4 10 25 - 27 220 - 225 $6.0 - $7.0 HOV2+ free peak period

I-95_FL 6 2 8 50 – 55 210 - 250 $13 - $14.0 HOV3+ free, Registered

SR 91_CA 10 4 8 35 - 40 215 - 220 $35 -$40HOV3+ discount in PM, free all other times

* Reversible facilities

EXISTING ML OVERVIEW

Mod

erat

ely

Con

gest

ed

Pea

k P

erio

d C

onge

sted

Hyp

er-C

onge

sted

# of Years

RE

VE

NU

E

• Market Capture– Attracting User

Markets– Peak Period HOV

Discounting– HOV 2+ or 3+ Market

Segmentation– Already Relatively

Mature Corridors

• Maturation of Targeted Demand

― Captures Sufficient Targeted Daily Demand

• Management of Demand

― High Toll Rates

― Discourage excessive usage

EVOLUTION OF MANAGED LANES

• Hockey Stick Revenue Achievable?? It Depends and requires:– Detailed Assessment of the all key variables– Focus on Future Operational Performances (GP & ML)

• Key Risk Associated with Forecasts– Competing Facilities– Escalation of Toll Rates– Maximum Demand Capture Rates– Off-peak/Directionality Considerations– Local Corridor Characteristics– Future Geometrical and Network Connectivity

EVOLUTION IMPLICATIONS

Annual Revenue growth has been very strong: 9.6% AAGR (1998 - 2004) [Inflation ~ 2.9%]

16.9% AAGR (2004 - 2007) [Inflation ~ 4.0%] Recession effect: -4.8% AAGR (2007 - 2010)

Overall nominal growth:7.5% AAGR (1998 - 2010)[Inflation ~ 2.8%]

Real Growth ~ 4.7% AAGR

REVENUE GROWTH IMPLICAITON?

$0

$5,000

$10,000

$15,000

$20,000

$25,000

$30,000

$35,000

$40,000

$45,000

$50,000

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Ann

ual R

evne

ues

(000

's)

SR 91 - CA

0

5,000

10,000

15,000

20,000

25,000

30,000

SR 167_WA I-394 MN (Reversible)

I-25_CO (Reversible)

IH-10_TX I-95_FL SR 91_CA

Ge

ne

ral P

urp

os

e D

aily

AA

AD

T

pe

r la

ne

2010 Estimates from Available Data

GP Daily AADT/Lane

$0

$20,000

$40,000

$60,000

$80,000

$100,000

$120,000

$140,000

$160,000




Mo

nth

ly R

ev

ne

ue

s


Monthly Revenue/Mile/Direction

$0.00

$0.50

$1.00

$1.50

$2.00

$2.50

$3.00

$3.50




Re

ve

nu

e p

er

To

lled

Ve

hic

le


Average Revenue/Vehicle

REVENUE – POLICY IMPLICATIONS

• Corridor Demand (Peaking/ Directionality) • Market/ OD pattern (Diversification)• Weekend Traffic Profile

• Traffic Conditions/Operations • GP Lane Congestion, Queuing/Metering, Time Saving

• Traveler’s Characteristics • Willingness-to-pay, Value of Reliability, Safety

• Toll Rate Pricing Structures, ML Access etc.

MANAGED LANE TRAFFIC – KEY FACTORS

• Economic Growth– Long-term Cyclical Trends/ Diversification of Growth

• Traffic Growth Profiles– Seasonality/Weekly/Daily/Hourly Distributions

• Values of Time– Income Growth and Distributions

A good forecaster is not smarter than everyone else, they merely have their ignorance better organized Anonymous

LONG-TERM CONSIDERATIONS

• Mode Trends/Market Shifts– HOV/Commercial Vehicle Market Trends – Aging Population/Migration Patterns

• Inflationary Trends– Toll Rate Escalation and Disposable Income

• Additional Influential Factors– Incident Rates/ Fuel Prices– Geometric/Operational Impedances on Speeds

LONG-TERM CONSIDERATIONS

• Risk Ranges (Tend to be Situational)– Location Dependent (Mature vs Undeveloped/Corridor

vs Regional)– Economic Diversity – Dependency on Single Markets/Industries

• There are many ways to get to the same place– Concave versus Convex Growth

ECONOMIC GROWTH

The past does not repeat itself, but it rhymes. Mark Twain

Brazoria Co.

Galveston Co.

Harris Co.

Fort Bend Co.

“Forecasters tend to use historical data for support rather than illumination” Montgomery County

0

100,000

200,000

300,000

400,000

500,000

600,000

700,000

1960 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020

Census

1972 Projection

1986 Projection

1992 Projection

2005 Projection

Harris County

0

1,000,000

2,000,000

3,000,000

4,000,000

5,000,000

6,000,000

1960 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020

Census

1972 Projection

1986 Projection

1992 Projection

2005 Projection

Fort Bend County

0

100,000

200,000

300,000

400,000

500,000

600,000

700,000

1960 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020

Census

1972 Projection

1986 Projection

1992 Projection

2005 Projection

ECONOMIC GROWTH

• Key Factors:– Motorist value of time (varied and situational)– Anticipated time savings “Error of anticipation”

• Equilibrium Sensitivity to Market Capture Rates– Elasticity is 4.0 (not 0.4) i.e. A small 10% change

in Traffic can result in 40% change in Revenues– Major Revenue Declines with higher gas prices

• Short-term or Long-term?

DETERMINING OPTIMUM TOLLS RATES

• Does it Necessarily Fall in Line with CPI?– Traditional Toll Facilities have not kept up with inflationary trends – What about managed lanes?

TOLL RATE ESCALATION

-2.0%

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12

:00

AM

1:0

0 A

M

2:0

0 A

M

3:0

0 A

M

4:0

0 A

M

5:0

0 A

M

6:0

0 A

M

7:0

0 A

M

8:0

0 A

M

9:0

0 A

M

10

:00

AM

11

:00

AM

12

:00

PM

1:0

0 P

M

2:0

0 P

M

3:0

0 P

M

4:0

0 P

M

5:0

0 P

M

6:0

0 P

M

7:0

0 P

M

8:0

0 P

M

9:0

0 P

M

10

:00

PM

11

:00

PM

Ave

rag

e A

nn

ual

Gro

wth

(20

01-2

010)

SR 91 Toll Rate Trends

Westbound Eastbound Regional LA CPI

• Revenue Days/ Annualization Factors– Difference between 275 and 365 can yield significant

revenue changes

• Ramp-up Assumptions– Brownfield versus Greenfield– Duration of Ramp-up (typically short for MLs)

• Peak Spreading Characteristics– Composition of Demand (Work versus Non Work)– Radial versus Circumferential– Corridor Volume Capacity

MAJOR REVENUE DETERMINANTS

• Are the Capture Rates Expected to be similar in both directions?– Diversion to managed lanes is very situational…

SR 91 Sample Profiling Example

$0.00

$0.10

$0.20

$0.30

$0.40

$0.50

$0.60

$0.70

$0.80

$0.90

$1.00

12

:00

AM

1:0

0 A

M

2:0

0 A

M

3:0

0 A

M

4:0

0 A

M

5:0

0 A

M

6:0

0 A

M

7:0

0 A

M

8:0

0 A

M

9:0

0 A

M

10

:00

AM

11

:00

AM

12

:00

PM

1:0

0 P

M

2:0

0 P

M

3:0

0 P

M

4:0

0 P

M

5:0

0 P

M

6:0

0 P

M

7:0

0 P

M

8:0

0 P

M

9:0

0 P

M

10

:00

PM

11

:00

PM

To

ll R

ate

Pe

r M

ile

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

Vo

lum

e/C

ap

ac

ity

Ra

tio

EB Toll Rate per Mile WB Toll Rate per Mile EB V/C WB V/C

MARKET CAPTURE RATES

Note: Market Share reflects toll paying patronage only

MANAGED LANE MARKET SHARES

3%

10%12%

16%

29%

18%

2%

6%4%

10%

20%

14%

0%

5%

10%

15%

20%

25%

30%

35%

40%




Bi-

Dir

ec

tio

na

l M

LM

ark

et

Sh

are

(T

OT

AL

)(T

oll-

pa

yin

gM

ark

et)


Peak (6-10 & 3-7) Weekday

• Long-term Commercial Vehicle Trends– Global/Local Effects of Trade Policies– Just-in-Time Delivery– Supply Chain Strategies– Evolution in Truck Sizes– Vehicle Operating Costs

• Aviation and Intercity Rail Trends– Competing versus Complementary Modes– New Transportation Policies (fuel efficiency etc.)

MODAL UTILIZATION CONSIDERATIONS

• Defining Risk– Where is the Risk – How to Quantify– How Significant is the Risk– Discrete versus Ranges

• Dependent on Data Availability– Historical Profiling– Accuracy/Variability of Forecast Sources– Data Filtering– New Modeling Approaches– Value of Reliability

• Incorporate all the Key variables to create realistic ranges– Correlation Dependency– Unknown/Unforeseen Variability– Prioritization of Key Factors

RISK PROFILING

2015 2020 2025 2030 2035 2040 2045 2050

Year

Base Case EstimateEarly OccurrenceLate Occurrence

Moderately Congested

MANAGED LANE RISK PROPAGATION

To expect the unexpected shows a thoroughly modern intellect. Oscar Wilde

# of years

RE

VE

NU

E

f( Key Subset Variables)

BASELINE

f( Key Subset Variables)

f( Full Universe of Variables)

f( Full Universe of Variables)

UNCERTAINTY RANGES

RECENT MANAGED LANE FINANCINGS

Managed Lane Project

FinancingMethod

Miles (Ultimate)

Project Costs

Public Grant

/Subsidy TIFIAFinancial

Close

Capital Beltway (Washington D.C.) PPP/DB 14.0 $1.9B $409M* $589M

Dec2007

I-595 Express Lanes (Miami)

Availability/DBFOM 10.5 $1.8B $232M** $603M

Marc h2009

North Tarrant Express (Fort Worth) PPP/DBFOM 13.0 $2.0B $573M $650M

June2009

IH 635 LBJ (Dallas) PPP/DBFOM 13.0 $2.7B $489M $850M June 2010

* Commonwealth of Virginia grant ** FDOT qualifying development funds

$0.6$1.0

$0.1 $0.3

$1.2$1.8

$1.3$1.6

$2.1$2.6

$2.1 $2.1

$4.0

$7.2

$3.4

$4.5$3.9

$5.1

$2.9$3.3

$0.0

$1.0

$2.0

$3.0

$4.0

$5.0

$6.0

$7.0

$8.0

S.R. 91 (1999)

S.R. 91 (2010)

Existing Existing Lender Sponsor Lender Sponsor Lender Sponsor Lender Sponsor

S.R. 91 IH 10 I-95 Miami

LBJ NTE Segment 1 NTE Segment 2 Capital Beltway

Ann

ual R

even

ue P

er M

ile p

er La

ne (M

illio

ns R

eal $

201

0)

2010 2020 2050

GP

Dai

lyVe

hs/L

ane

22,4

00

29,0

00

23,0

00

24,3

00

31,1

06

10,4

83

13,4

20

26,0

00

21,6

52

27,7

17

26,2

50

33,6

02

MANAGED LANE REVENUE RISK

4.7%

4.8%

3.6%

1.5%*

2.3%

*Escalated from 2040 results

• Quantification may unintentionally create an aura of precision and confidence – Clear Understanding of the Assumptions is a MUST.

• Context of how will the ranges be utilized– Project Feasibility – Bonding/Capital Improvement Plans– Identification of Subsidy Requirements

• How to narrow the likely ranges?– Detailed data on current ranges– Assessment of Key Variables – Explore Alternative/New Influential Variables

INTERPRETATION AND CONCLUSIONS

THANK YOU

Christopher Mwalwanda

Vice President

Wilbur Smith Associates

[email protected]

Documents

LONG-TERM DEMAND FORECASTING OF MANAGED LANES