LIE HOV Clean Pass Vehicle Operational Performance ... · The Long Island Expressway Clean Pass Vehicle program allows eligible, low-emission, energy- efficient vehicles to use the

LIE HOV Clean Pass Vehicle Operational Performance Scenario Analysis

Final Report

Prepared by: WSP | Parsons Brinckerhoff &

EA Harper Consulting

Submitted on: December 8, 2015

Table of Contents

Executive Summary ........................................................................................................................ 1

I. Introduction & Overview ........................................................................................................ 1

Purpose of the Study ................................................................................................................... 1

Study Area Description ............................................................................................................... 4

General Approach to the Study ................................................................................................... 4

CPVs and Mode Definitions ....................................................................................................... 5

Scenarios for Evaluation ............................................................................................................. 5

II. Modeling Approach and Methodology ................................................................................... 7

Approach to Modeling with Adapted NYBPM 2010 Update ..................................................... 7

Methods for Base Year 2014 Calibration ................................................................................... 8

III. Calibration and Validation of 2014 No-Build Scenario .................................................... 12

2010 Initial Calibration ............................................................................................................. 12

2014 Calibration........................................................................................................................ 12

IV. Performance Measures for Scenario Evaluation ................................................................ 17

Peak Hour Volumes .................................................................................................................. 17

Average Speed .......................................................................................................................... 18

Demand Over Capacity Ratio ................................................................................................... 18

Lane Miles of Congestion ......................................................................................................... 19

Vehicle Hours of Peak Hour Delay .......................................................................................... 19

Vehicle Miles of Peak Hour Travel .......................................................................................... 19

Person Trips per Mile................................................................................................................ 20

V. Future Year Forecasting Methods ......................................................................................... 21

Standard Forecast Procedures ................................................................................................... 21

Unchecked (No Cap) Forecast Procedures ............................................................................... 22

VI. Discussion of Results: Baseline and Forecast .................................................................... 23

Baseline Conditions - 2014 ....................................................................................................... 23

Volumes and Speeds ............................................................................................................. 23

Volume Capacity Ratios ....................................................................................................... 26

Aggregate Measures – Congestion, Hours of Delay and Vehicle Miles of Travel ............... 27

Person Trips .......................................................................................................................... 27

Comparison of Policy Options and Forecasts ........................................................................... 28

Volumes and Speeds ............................................................................................................. 28

Volume Capacity Ratios ....................................................................................................... 36

Aggregate Measures – Congestion, Hours of Delay and Vehicle Miles of Travel ............... 42

Person Trips .......................................................................................................................... 42

VII. Conclusion ......................................................................................................................... 44

List of Acronyms .......................................................................................................................... 49

Technical Appendix A .................................................................................................................. 51

List of Tables

Table 1: Base 2010 Scenario Daily Validation ............................................................................. 14

Table 2: Base 2014 Scenario Peak Period Validation .................................................................. 15 Table 3: Level of Service (LOS) Definition ................................................................................. 18 Table 4: Occupancy Assumed by Mode in the NYBPM .............................................................. 20 Table 5: 2014 AM West Bound Average Speeds ......................................................................... 25 Table 6: 2014 AM West Bound – Difference in Averages Speeds (HOV-GPL) ......................... 26

Table 7: Aggregate Measures of Congestion – 2014 AM West Bound ....................................... 27 Table 8: 2014 AM West Bound Peak Hour - Person Trips by Facility Type .............................. 27 Table 9: 2014 AM Peak Hour - Share of Person Trips on the HOV Facility ............................... 28 Table 10: Forecasted AM West Bound Peak Hour Volumes and Speeds at Exit 49.................... 30

Table 11: Forecasted AM West Bound Peak Hour Volumes and Speeds at Exits 53-55 ............. 30 Table 12: Changes in AM Peak Hour Volumes on Parallel Corridors – At the County Boundary

....................................................................................................................................................... 31 Table 13: LIE Corridor Forecasted Aggregate Measures of Congestion – AM West Bound Peak

Hour .............................................................................................................................................. 42 Table 14: Forecasted AM Peak Hour West Bound - Person Trips at Exit 49 .............................. 43 Table 15: Forecasted AM Peak Hour West Bound - Person Trips at Exits 53-55 ........................ 43

List of Figures

Figure 1: Hybrid Market Share of US New Vehicle Sales ............................................................. 2

Figure 2: CPV Program Enrollment 2006 - 20121 .......................................................................... 2

Figure 3: CPV Share of LIE HOV Lane Traffic in Peak Hours ..................................................... 3 Figure 4: Study Limits and Parallel Routes .................................................................................... 4

Figure 5: HOV Lane Operational Scenarios ................................................................................... 6 Figure 6: Framework for Modeling .............................................................................................. 10

Figure 7: LIE HOV Lane Shed ..................................................................................................... 11 Figure 8: Locations for Comparison of MOEs ............................................................................. 17 Figure 9: 2014 AM West Bound Peak Hour - Total Volumes ..................................................... 23 Figure 10: 2014 AM West Bound Peak Hour – HOV-Eligible Vehicles by Mode ...................... 24

Figure 11: 2014 AM West Bound Peak Hour - HOV-Eligible Vehicles by Facility ................... 24 Figure 12: 2014 AM West Bound Peak Hour - Volume/Capacity Ratio ..................................... 26 Figure 13: Base Year 2014 AM Peak Hour West Bound Volumes and Speeds at Exit 49 .......... 32

Figure 14: Forecasted AM Peak Hour West Bound Volumes and Speeds at Exit 49 .................. 33 Figure 15: Base Year 2014 AM Peak Hour West Bound Volumes and Speeds at Exits 53-55 ... 34 Figure 16: Forecasted AM Peak Hour West Bound Volumes and Speeds at Exits 53-55 ........... 35 Figure 17: HOV Lane Composition at Exit 49-50: Carpools vs. CPVs ....................................... 44

Figure 18: Summary of Scenarios, in Throughput (Person Trips), AM WB ................................ 46 Figure 19: Summary of Scenarios, in Throughput (Person Trips), PM EB .................................. 46 Figure 20: Summary of Scenarios, in Vehicle Hours of Delay, AM WB .................................... 47 Figure 21: Summary of Scenarios, in Vehicle Hours of Delay, PM EB ...................................... 47

Executive Summary

LIE CPV Modeling – Final Report

Executive Summary

In recent years, increasing volumes on the Long Island Expressway (LIE) have led to worsening

congestion in both the general purpose lanes (GPL) and the High Occupancy Vehicle (HOV)

lanes. In 2014, both the GPL and HOV lanes were near their maximum capacity, with the

benefits of using the HOV lane diminishing as it has filled up. A large and growing portion of

the HOV lane traffic is composed of Clean Pass Vehicles (CPVs), which are allowed to use the

lane with only a single occupant. This study analyzes the performance of four alternative

operational scenarios for the HOV lane. The scenarios represent alternative eligibility criteria for

using the HOV lane, in a 2014 base year as well as 2016 and 2035 forecast years.

The operational scenarios are defined in terms of HOV lane use permissions by occupancy and

vehicle type. The goal of this study is to evaluate the capability for each of these scenarios to

move people more efficiently through the LIE corridor. One of the most effective ways to reduce

congestion is to increase the average occupancy of vehicles on the LIE. The scenarios used in

this study evaluate the effectiveness of using the HOV lane as both a carpooling incentive and as

an incentive to buy more fuel-efficient vehicles. The five scenarios evaluated in this study are

listed below:

(1) No Build / Base (Current 2015 HOV CPV policy: HOV2+, Single Occupant CPVs)

(2) HOV 2+ and discontinue Clean Pass Eligibility

(3) HOV2+, discontinue Clean Pass Eligibility but allow Single Occupant Plug-In Hybrid Electric Vehicles (PHEV) and Single Occupant 100% Electric Vehicles (EV)

(4) HOV 3+ and Clean Pass Eligibility Unchanged

(5) HOV 3+, CPV 2+, Single Occupant Plug-In Hybrid Electric Vehicles (PHEV), and Single Occupant 100% Electric Vehicles (EV)

The New York Metropolitan Transportation Council (NYMTC)’s regional travel demand model,

the New York Best Practices Model (NYBPM) 2010 update version, was used to evaluate these

operational scenarios. A set of measures of effectiveness was calculated to analyze the

congestion levels and passenger and vehicle throughput for each scenario and each year,

discussed in more detail in the body and appendix of this report. For example, one measure of

effectiveness is the number of person-trips passing through a certain point on the LIE. This

statistic provides a measure of throughput in terms of people rather than vehicles, reflecting the

effectiveness of the incentive to carpool. This person-throughput for each scenario is

summarized in the following figure.

Executive Summary


The model results show that congestion on both the GPL and HOV lanes, already at a poor level

of service in the 2014 base year, are expected to increase substantially if the HOV lane eligibility

remains unchanged from its current configuration. However, the four alternative scenarios show

the potential to improve the level of service on the HOV lane in the 2016 and 2035 forecast

years. For example, relative to the current configuration, the 5th

scenario, which represents a

scheme of tiered occupancy requirements, is predicted to reduce the lane miles of congestion in

the HOV lane by 5% in the AM peak hour in 2035, without causing a notable increase in lane

miles of congestion on the GPL. The 2nd

scenario, in which Clean Pass eligibility is discontinued,

is predicted to result in a 33% decrease in lane miles of congestion in 2035, but also a 12%

increase in lane miles of congestion on the GPL, due to the single-occupant CPVs moving to the

GPL lanes as they are no longer allowed on the HOV lane. Meanwhile, the 3rd

scenario, in

which only EVs are allowed to use the HOV lane with a single occupant, produces smaller

improvements in HOV lane congestion (5% reduction in lane miles of congestion) with more

minimal impacts on the GPL (no difference in lane miles of congestion and only a 2% increase

in vehicle hours of delay). Interestingly, the 4th

scenario, in which traditional vehicles must have

at least three occupants but CPVs can have only a single occupant, is the only scenario that

results in a decrease in person-throughput on the HOV lane, because it does not change CPV

occupancy requirements at all, and vehicles with 3 or more vehicles make up a relatively small

portion of overall traffic. Additional tradeoffs between scenarios can be seen in a number of

different measures of effectiveness, presented in more detail in the body of this report and

Technical Appendix A. One clear finding is that, in the near term (i.e. before hybrid and electric

vehicles make up a more substantial portion of the fleet, as they are predicted to in 2035), the

“tiered occupancy” approach (scenario 5) has the greatest potential to alleviate congestion on the

HOV lane, and keep the lane as a meaningful incentive to purchase a more efficient vehicle, or

carpool.

I. Introduction & Overview

LIE CPV Modeling – Final Report 1


This report describes the model methodology and setup, measures of effectiveness, future year

forecasting methods, and analysis of model results for the operational performance scenario

analysis on the Long Island Expressway (LIE)’s High Occupancy Vehicle (HOV) Clean Pass

Vehicle (CPV) program.

Purpose of the Study The Long Island Expressway Clean Pass Vehicle program allows eligible, low-emission, energy-

efficient vehicles to use the 40-mile LIE-HOV lanes regardless of the number of occupants in the

vehicle. The program began in 2006 with the goal of promoting the use of low-emission energy-

efficient vehicles, saving gasoline and reducing emissions. Prior to 2006, the LIE-HOV lanes

were reserved for use by vehicles with two or more occupants, buses and motorcycles during

peak travel hours.

As more vehicles that meet the CPV criteria come on the market, congestion on the LIE-HOV

facility has increased. Over the past decade, hybrid vehicles (many of which are CPV-eligible)

have grown substantially as a proportion of new vehicle sales in the US, as shown in Figure 1.

At the same time, the rate of enrollment in NYSDOT’s CPV program has increased, particularly

in Suffolk and Nassau counties, as shown in Figure 2. In September of 2014, over 99% of the

14,000 CPV-eligible vehicles in Suffolk County were enrolled in the program1. Together, these

trends have contributed to CPVs accounting for a significant and growing share of vehicles in the

LIE HOV lane, as shown in Figure 3. Concurrently, the overall traffic in the HOV lane has been

increasing, and approaching capacity (shown by the horizontal line at 1650 vehicles per hour in

Figure 3). The CPV program has contributed to this growth in congestion, by allowing vehicles

with only a single occupant use the HOV lane, increasing the number of vehicles on the roadway

required to transport the same number of people. As it is today, the HOV lane is congested, in

part due to CPVs. With worsening congestion in the HOV lane making travel times similar to

those in the GPL, it no longer provides an incentive for either carpooling or purchasing more

efficient vehicles.

The purpose of this study is to analyze the operational performance of the LIE-HOV lane under

different eligibility criteria. Specifically, it considers the LIE between exits 32 and 64, in order to

understand the potential impacts of a range of potential HOV and CPV eligibility policies. This

study was initiated by the New York State Department of Transportation Policy & Planning

Division (NYSDOT). NYSDOT may revise the HOV CPV policy in the future, in response to

degrading conditions on the LIE HOV lanes.

1 Source: NYSDOT Data on trends relating to the CPV program, “Trends Feb15 Chart Info.xls”.



Figure 1: Hybrid Market Share of US New Vehicle Sales2

Figure 2: CPV Program Enrollment 2006 - 20121


2 Source: Monthly Hybrid Dashboard, hybridcars.com.

0.0%

0.5%

1.0%

1.5%

2.0%

2.5%

3.0%

3.5%

4.0%

4.5%

5.0%

Pe

rce

nta

ge o

f N

ew

Ve

hic

le S

ale

s Hybrid Market Share of all US New Vehicle Sales (%)

28%

3%

8%

16% 18%

64%

35%

11%

94%

9%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

Percentage of Clean Pass Eligible Vehicles Participating in the Program

2006

2009

2012



Figure 3: CPV Share of LIE HOV Lane Traffic in Peak Hours3


0

100

200

300

400

500

600

700

800

900

1000

1100

1200

1300

1400

1500

1600

1700

1800

1900

2000

Vo

lum

e

Westbound Peak Hour (7-8am) HOV Lane Volumes at Bagatelle Road (Exit 50)

CPV's

Carpools

Actual Projected

NOTE: Sept, Oct, and Nov 2011 - Data lost due to electrical problems related to major flooding. Data estimated on percent increase from 2010.

0100200300400500600700800900

10001100120013001400150016001700180019002000

Vo

lum

e

Eastbound Peak Hour (5-6pm) HOV Lane Volumes at Bagatelle Road (Exit 50)

CPV's

Carpools

Actual Projected

NOTE: Sept and Oct 2011 - Data lost due to electrical problems related to major flooding. Data estimated on percent increase from 2010.



Study Area Description The study area is defined as a 40-mile stretch of the LIE, also known as I-495, between Exit 32,

just east of the Cross Island Expressway in Little Neck to Exit 64, just east of State Route 112 in

Medford.

A 2014 count at Exit 49 in Melville, just east of State Route 110 shows that this portion of the

LIE carries 181,000 vehicles daily. Other major east-west facilities include:

Northern State Parkway (77,000 daily vehicles)

Southern State Parkway (160,000 daily vehicles)

Sunrise Highway (50,000 daily vehicles)

All of these east-west major corridors are experiencing congested conditions in the AM Peak

periods. And the LIE is experiencing highly congested conditions in both the general purpose

lanes (GPL) and on the HOV facility.

Figure 4: Study Limits and Parallel Routes

General Approach to the Study The general approach was to define a set of policy related options (or scenarios) for restricting

access to the LIE-HOV facility. The analysis makes use of the New York Metropolitan Planning

Council’s (NYMTC) travel forecasting tool, the New York Best Practices Model (NYBPM).

The NYBPM was customized to allow for a detailed analysis of LIE-HOV lane usage by

relevant HOV and CPV modes. In addition, the customized NYBPM was calibrated in the LIE

corridor to 2014; a year for which some limited counts by the relevant modes were available.

These modifications and the consequent calibration are fully documented and described in detail

in the next section. Model modifications also accommodated the reporting of a predefined set of

Measures of Effectiveness (MOE). The definitions of the scenarios, CPVs, and MOEs are

described below.



Testing of the scenarios was performed for the new base year (2014) and two forecast years

(2016 and 2035). No changes to the LIE are programmed between the calibration year and the

horizon year of 2035, so all operational changes observed are due to the policy scenarios tested

and segments of the LIE on which MOEs are reported are consistent for both forecast years.

CPVs and Mode Definitions The LIE CPVs are defined to be consistent with NYSDEC criteria. These criteria are:

Vehicles which are certified to the California Super Ultra Low Emission Vehicle (SULEV) standard and achieve a United States Environmental Protection Agency

(USEPA) highway fuel economy rating of 45 miles per gallon or more; or

Pre-model year 2005 hybrid vehicles which are certified to the California Ultra Low Emission Vehicle (ULEV) standard, and achieve a USEPA highway fuel economy rating

of 45 miles per gallon or more.

A list of eligible vehicles, by make and model and year, is provided on the NYSDOT website

(https://www.dot.ny.gov/programs/clean-pass). This list is modified on a regular basis as more

vehicles that meet the standards become available. An owner of any of the vehicles on the

official list can apply for a Clean Pass.

For this analysis electric vehicles and plug-in hybrid electric vehicles (PHEV/EV), which are a

subset of CPVs, are distinguished from CPV, provide another category of vehicle type that could

potentially be used for HOV lane usage policy.

Since motorcycles (MC) are permitted to use the HOV lanes the model was modified to allow

the modeling of MCs as a separate mode, and forecasted MC usage of the HOV facility is

reported. However, the incidence of MC usage is too small to constitute a vehicle type useful for

policy restrictions.

In summary, the CPV modes defined for analysis are:

SOV – Single Occupant Vehicles,

HOV2 – High Occupant Vehicles with at least two passengers,

HOV3+ - High Occupant Vehicles with three or more passengers,

CPV – Vehicles designated by NYSDOT as clean pass vehicles (excluding EVs),

PHEV/EV – Electric vehicles, including plug-in hybrid electric vehicles.

Scenarios for Evaluation The analysis required an assessment of policies affecting five modes in various combinations.

The first scenario is the current, base condition and represents a policy option of no change.

Scenarios 2-5 are policy options with increasing levels of restriction on the types of vehicles that

https://www.dot.ny.gov/programs/clean-pass



are permitted to use the HOV facility. Motorcycles are permitted in the HOV lane in all

scenarios. The five scenarios, summarized in Figure 5, are:

(1) No Build / Base (Current HOV CPV use/policy)

(2) HOV 2+ and discontinue Clean Pass Eligibility

(3) HOV2+, discontinue Clean Pass Eligibility but allow Single Occupant Plug-In Hybrid Electric Vehicles (PHEV) and Single Occupant 100% Electric Vehicles (EV)

(4) HOV 3+ and Clean Pass Eligibility Unchanged

(5) HOV 3+, CPV 2+, Single Occupant Plug-In Hybrid Electric Vehicles (PHEV), and Single Occupant 100% Electric Vehicles (EV)

Each of the five operational scenarios was evaluated for a 2014 base year and 2016 and 2035

future year forecasts. The tool used to evaluate the alternatives is the New York Metropolitan

Transportation Council (NYMTC)’s regional travel demand model, the Best Practices Model

(NYBPM), 2010 version (the latest available at the time). The NYBPM was modified from its

native operating mode in order to accommodate the specific needs of this project (as described in

detail in section II), and highway traffic assignment results were used to calculate a set of key

measures of effectiveness (MOEs). These MOEs (described in detail in section IV) were then

compared and analyzed across operational scenarios and model years (see section VI).

Traditional Vehicle

Current CPV (excluding EV)

PHEV/ EV

Scenario SOV HOV2 HOV3+ SOV HOV2 HOV3+ SOV HOV2 HOV3+

1. Base Config.

2. No CPV

3. EV Only

4. HOV3+

5. Tiered Occupancy

= not permitted in HOV lane

= permitted in HOV lane

Figure 5: HOV Lane Operational Scenarios

II. Modeling Approach and Methodology



This section of the report describes the adaptation of the NYBPM 2010 for use in this custom

application to evaluate scenarios on the LIE HOV lane.

Approach to Modeling with Adapted NYBPM 2010 Update The first step in adapting the standard NYBPM 2010 version for use in this project was to code

the highway network to reflect the current policies of HOV 2+ and Clean Pass eligibility. Then,

the current seven-mode, multi-class highway assignment procedure was similarly modified, and

extended to incorporate Clean Pass Vehicle (CPV), Electric Vehicle (EV), and motorcycle (MC)

vehicle types. The standard set of NYBPM trip tables for the AM (6-10 am) and PM (4-8 pm)

peak periods were also extended to include estimated origin-destination (OD) flows of these

special user classes, with SOV and HOV tables distinguishing CPV, EV, and MC from non-CPV

autos. Existing vehicle class volume counts on the LIE for HOV2+, HOV3+ CPV and EV, as

well as vehicle registration information, were used to estimate 2014 base year CPV and EV trip

tables, and to calibrate the modified assignment procedures. Select link analyses for the different

user categories of the LIE project area segments were used to identify and calibrate the different

market shares and types of vehicles.

A supplementary source of data utilized for this project was the recent NYMTC Regional

Household Travel Survey (RHTS 2010). The RHTS 2010 contains detailed information about all

vehicles owned by households in the 28-county BPM modeled region, including which vehicles

are used for each trip. The RHTS data was analyzed to help identify CPV ownership and travel

use patterns by car sufficiency level (i.e. whether the number of cars in a household is zero, less

than the number of adults, or at least equal to the number of adults), household income (low,

defined as the bottom 15%; middle; or high, defined as the top 15%), as well as by trip purpose

(work vs. non-work).

The NYBPM commercial vehicle (heavy truck, medium truck, and commercial van) trip tables

were kept intact and not split into different sets of vehicle flows, as these categories are not

permitted on the LIE-HOV lanes.

The simulation output was analyzed and compared to the observed counts, both for overall peak

periods in 2010 and for peak 2-hour operations in 2014. The OD and path travel times for the

HOV lane, as well as for the LIE general lanes and other roadways in the corridor were extracted

for observation and evaluation. Based on the results of this analysis, the network capacities,

speeds, and assignment parameters were calibrated in order to adjust the trip tables so that the

model output closely matches the observed counts within an acceptable range of error (+10%) or

as per FHWA guidance.

This framework for modeling and the general approach to the preparation of the NYBPM 2010

Update regional model for use in this study is shown in Figure 6. This exhibit distinguishes two

levels of special modeling features to be developed and applied in this study as supplemental

functionality:

Main NYBPM modules with minor innovative adaptations for CPV and EV analysis.



New Utilities that will serve to process CPV and EV data between the execution of standard NYBPM modules.

Methods for Base Year 2014 Calibration The following specific steps were implemented to obtain a 2014 calibration of the NYBPM 2010

Update model with special features for CPV/EV LIE HOV lane modeling (described further in

section III). Note that for each of these calibration steps, the corresponding relevant stage or

model component of the study model is labeled in Figure 6. This figure is broken down by

NYBPM model component, including the Household-Auto-Journey generation step (HAJ),

Mode Destination and Stop Choice (MDSC), and the Pre-Assignment Processor (PAP).

1. Compare 2010 NYBPM Update results in the study corridor

a. 2010 Screenline traffic count database - full corridor and regional coverage

b. RHTS comparisons for HOV2, HOV3+ trip use patterns

2. Apply 2014 Socio-economic data (SED) and transportation network improvement projects to obtain a base year 2014 scenario modeling and reporting system.

a. Modify the NYBPM’s Pre-Assignment Processor (PAP) module to create 2 hour peak assignment periods for AM and PM

b. Refine link capacities and/or assignment parameters to obtain the best regular category fit for volumes (primary), and speeds or travel times (secondary).

3. Calculate CPV/EV and Motorcycle shares.

a. Using CPV/EV and Motorcycle registration information, calculate respective ratios of ownership by county in the study area (NY State only), as well as

county-specific rates of enrollment in the CPV program. This total number of

CPV trips will be allocated to specific auto trips using information from (b) and

(c) below.

b. Create an “HOV lane shed” (i.e. a group of TAZs likely to be using LIE HOV lane), using Select Link Analysis (SLA) along the LIE-HOV. SLA is a static

traffic assignment procedure that reports the origins, destinations, and link-by-link

volumes for vehicles which traverse a selected set of links (in this case, any of the

links on the LIE HOV lane). The HOV lane shed is a set of TAZs that are the

primary feeders of the HOV lane (determined by setting a threshold level of trip

ends). Auto trips in this “shed” will have a higher (and adjustable) probability of

being CPVs. Figure 7 presents the final HOV lane shed that was used, at the

cutoff threshold of 200 trip ends per TAZ during the PM peak period.

c. Calculate proportion of CPV/EV/MC vehicles by Income level, car sufficiency, and trip purpose, using RHTS percentages, weighting by HOV lane shed zones.

Use seeded Monte Carlo simulation to assign each trip to CPV, EV, MC, or non-

CPV using these calculated probabilities. Monte Carlo simulation is a method of

using repeated random draws to simulate an event. For example, in this context, a

random number is drawn from a uniform distribution between 0 and 1 for each

vehicle trip, and if the random number is less than the probability of that trip



being made by a CPV, the trip is assigned to the CPV mode. If the random

number is greater than the probability of that trip being made by a CPV, the

original mode (traditional SOV, traditional HOV2, or traditional HOV3+) is

retained.

4. Add bus preloads for the HOV lane, in order to ensure proper assignment and capacity usage on the lane

5. Calibrate CPV/EV/MC shares to counts, by adjusting or varying some of the probabilities:

a. First the split between HOV lane shed and the other areas

b. Then purpose, car sufficiency and income level shares, specifically the work trips and the split between middle and high income.



Figure 6: Framework for Modeling



Figure 7: LIE HOV Lane Shed

III. Calibration and Validation of 2014 No-Build Scenario



This section of the report describes process of calibrating and validating the NYBPM 2010 for its

use in evaluating the performance of the LIE HOV lane under the five scenarios. It discusses the

calibration methods used to improve the fit of modeled flows in the study corridor, particularly

for the new disaggregate vehicle classes used for this analysis, and presents the results of the

calibration procedure compared with traffic count data.

2010 Initial Calibration For the NYBPM’s native base year of 2010, the model has a set of screenline database traffic

counts that were used to perform a preliminary validation and calibration of the model’s

performance in modeling flows along the LIE corridor. Initial results showed that overall, flows

were slightly high on the primary highways in Nassau and Suffolk Counties (i.e. Long Island

Expressway [LIE], Northern State Parkway [NSP], Southern State Parkway [SSP], and Sunrise

Highway [SH]). Several adjustments to the core models’ county-to-county destination choice

calibration constants were tested, but ultimately it was decided that the county-to-county travel

patterns were in-line with observed data, and adjustments should be made at a link level. A 15%

speed reduction was implemented along the length of the LIE, both on the main line and the

HOV lane. An additional 5 mph in speed was added to the free-flow speeds of arterials (in the

BPM, Physical Link Type [PLT] 12-16) in area type 9 (defined as zones with a buffered

population density between 0 and 1800 per square mile and employment density between 4675

and 9600 per square mile, which includes a large number of zones in Nassau and Suffolk

Counties along the LIE corridor), in order to encourage more traffic to use arterials rather than

highways. The combination of these two changes improved the modeled flows along the LIE and

parallel routes along the study corridor.

Table 1 presents daily comparisons of County to County screenline results as well as specific

facilities along the Nassau/Suffolk demarcation (central to the study area). The simulated

volumes on the LIE, as well as the overall county-to-county flows, are well within the 10%

margin of error. A lower level of calibration was done for 2010, as the emphasis of this study is

on the 2014 base calibration for LIE specific volumes and correct HOV-CPV splits, as presented

in the next section.

2014 Calibration For 2014, a more limited set of count data was available for comparison with modeled flows.

These counts do, however, include a breakdown between the LIE mainline and HOV lanes, and

between CPV and non-CPV vehicles, as well as vehicle occupancy, in select locations along the

Expressway.

The calibration process for 2014 included both calibrating the BPM’s core models and highway

network settings to get the total traffic flows to match counts, as well as adjusting the factors

used to assign auto trips to CPV, EV, and motorcycle (MC) modes to meet those target shares. It

should be noted that with small target numbers, even small absolute deviations show large

percentage differences.

The targets used for the 2014 calibration come from several data items provided by NYSDOT.

Targets were needed for flows of each vehicle class (i.e. SOV, HOV2, HOV3+, CPV1, CPV2+,



EV, MC, Bus, Commercial) at several locations along the study corridor. A set of 2014 HOV

lane counts, broken down by HOV2, HOV3, HOV4, CPV, MC, and Bus, was available at Exits

46, 50, 59, and 624. These count data were used as the basis for the HOV lane counts at these

locations. The one location with both mainline and HOV lane counts available was at Exit 495.

October 2014 total volumes for the mainline and HOV lanes were used from this database to

create targets for Exit 49 flows. The total volumes in the HOV lane at Exit 49 were broken down

into classes using the proportions from the Exit 50 HOV lane occupancy data.

At Exits 46, 50, 59, and 62, the HOV lane data was further broken down into CPV occupancy

groups and EVs using additional data sets. Vehicle registration data was used to determine the

EV proportion of all CPV-eligible vehicles (6.3%), and this factor was used to calculate the EV

proportion of CPVs in the LIE HOV lane. The CPV occupancy data at Exit 506 was used to

create target proportions of CPV1 and CPV2+ out of the total CPV number (86% and 14%,

respectively), for all 5 count locations along the LIE. At Exit 49, where only a total HOV lane

count volume was available, the proportions of HOV2, HOV3+, CPV, MC, and Bus were used

from Exit 50 to estimate disaggregate counts. Table 2 presents the final calibrated results, while

a log file detailing the many simulation runs is presented in Appendix A. This log file contains

the details of the 40 model simulation runs that were needed to arrive to a satisfactory calibration

of the Base 2014 scenario.

The peak direction simulation results (shaded rows) are very close to the counts, within the

required 10% error range for most of the exits (only AM WB at 62 and PM_EB at 59 & 62 do

not meet this criteria) as are the splits between CPV and regular HOV (HOV2 & HOV3+) at the

same locations. The easternmost exits are more difficult to calibrate, as is the reverse peak

direction, for the simple reason that lower overall volumes lead to a larger proportion of HOV

and CPV to use the GPL lanes in reality, where changes between multiple lanes are possible,

while a static traffic assignment model does not see that difference and will have a tendency to

distribute volumes amongst all lanes in a more purely mathematical fashion. However, in the

areas where the network shows congestion, the behavior better reflects reality, with HOV and

CPV shares and totals all within the desired range.

4 from the file “HOV Occupancy Study 10_16_14.xlsx”, received from NYSDOT.

5 from the file “94-14h49_2014.xlsx”

6 from the file “CPVSpecialJune2014.pdf”



14

Table 1: Base 2010 Scenario Daily Validation

2010 Calibration Summary

COUNTY-TO-COUNTY SCREENLINE TOTALS MAJOR FACILITIES - AT EXIT 48-49:

All BPM modeled links, including arterials

Facility Direction Count Flow # Diff % Diff

Count Flow # Diff % Diff EB 90,364 90,317 -47 0%

EB 594,691 556,024 -38,666 -7% WB 90,581 87,479 -3,102 -3%

WB 538,049 512,879 -25,169 -5% EB 37,981 40,624 2,643 7%

EB 345,629 338,345 -7,284 -2% WB 39,073 40,326 1,253 3%

WB 372,703 358,732 -13,971 -4% EB 128,345 130,941 2,596 2%

WB 129,654 127,805 -1,849 -1%

EB 78,523 81,029 2,506 3%

WB 81,286 80,795 -491 -1%

EB 23,654 18,643 -5,011 -21%

WB 25,339 21,277 -4,062 -16%

EB 102,177 99,672 -2,505 -2%

WB 106,625 102,072 -4,553 -4%

EB 230,522 230,614 92 0%

WB 236,279 229,877 -6,402 -3%Grand Total

Queens / Nassau

Nassau / Suffolk

LIE Main + HOV

Northern State Parkway

Sub-Total: Northern

Southern State Parkway

Sunrise Highway

Sub-Total: Southern



15

Table 2: Base 2014 Scenario Peak Period Validation

At Exit 46:

AM Peak (7-9) PM Peak (4-6)

SOV HOV2 HOV3+ Taxi CPV1 CPV2+ EV CPV MC Total

Grand

Total SOV HOV2 HOV3+ Taxi CPV1 CPV2+ EV CPV MC Total

Grand

Total

LIE Mainline 9,340 290 84 4 145 39 11 195 8 10,701 WB 5,723 349 165 10 74 28 10 111 4 6,903

LIE HOV 0 1,221 481 30 935 197 87 1,220 34 3,012 13,712 0 1,242 499 46 155 72 23 250 10 2,064 8,967

LIE Mainline 4,516 211 81 15 46 16 6 68 2 5,415 EB 8,014 383 153 8 160 47 14 220 6 9,057

LIE HOV 0 955 257 33 87 24 14 125 8 1,387 6,803 0 1,316 525 26 605 143 63 811 31 2,730 11,787

HOV2 HOV3+ HOV CPV1 CPV2+ EV CPV MC Total HOV2 HOV3+ HOV CPV1 CPV2+ EV CPV MC Total

WB LIE HOV 1317 338 1655 1037 171 82 1290 8 2978 WB 1224 222 1446 159 26 13 198 11 1672

EB LIE HOV 581 182 763 169 28 13 210 10 992 EB 1368 355 1723 670 110 53 833 9 2586

WB LIE HOV -66 143 77 -102 26 5 -70 26 34 WB 64 277 341 -4 46 10 52 -1 392

EB LIE HOV 407 75 482 -82 -4 1 -85 -2 395 EB -26 170 145 -65 33 10 -22 22 144

WB LIE HOV -5% 42% 5% -10% 15% 6% -5% 330% 1% WB 5% 125% 24% -3% 176% 77% 26% -10% 23%

EB LIE HOV 70% 41% 63% -49% -15% 10% -40% -20% 40% EB -2% 48% 8% -10% 30% 19% -3% 241% 6%

At Exit 49:



Grand


Grand

Total

LIE Mainline 10,597 219 73 5 186 32 13 231 8 11,631 5,489 545 274 9 122 62 14 197 8 6,906

LIE HOV 0 1,181 479 29 965 193 88 1,246 38 3,024 14,655 0 988 411 39 143 62 21 226 5 1,681 8,587

LIE Mainline 4,729 271 120 2 46 12 10 68 2 5,586 10,451 250 111 4 165 38 18 222 5 11,207

LIE HOV 0 899 263 36 112 29 15 156 5 1,368 6,954 0 1,394 561 27 699 165 68 933 37 2,981 14,188


LIE Mainline 12,078 9,115

LIE HOV 1,447 296 1,743 984 162 78 1,223 18 3,026 15,104 919 276 1,195 96 16 8 120 7 1,336 10,451

LIE Mainline 8,240 10,572

LIE HOV 423 114 537 120 20 9 149 8 710 8,950 1,603 403 2,006 777 128 61 966 22 3,024 13,596

LIE Mainline -447 -2,209

LIE HOV -236 183 -53 -19 31 10 22 19 -2 -449 108 135 243 47 46 13 106 -2 345 -1,864

LIE Mainline -2,654 635

LIE HOV 512 149 662 -8 9 6 7 -3 658 -1,996 -182 158 -24 -78 37 7 -33 14 -43 592

LIE Mainline -4% -24%

LIE HOV -16% 62% -3% -2% 19% 13% 2% 107% 0% -3% 12% 49% 20% 49% 292% 175% 89% -30% 26% -18%

LIE Mainline -32% 6%

LIE HOV 121% 131% 123% -7% 46% 62% 5% -35% 93% -22% -11% 39% -1% -10% 29% 12% -3% 64% -1% 4%

WB

EB

WB

EB

WB

EB

WB

EB

EB

EB

WB

EB

Counts

# Diff

% Diff

WB

WB

EB

Modeled

Flows

Counts

# Diff

% Diff

Modeled

Flows

WB

EB

WB



16

At Exit 50:



Grand


Grand

Total

LIE Mainline 9,769 117 48 2 74 22 6 102 5 10,562 WB 5,633 240 120 4 83 31 12 125 3 6,485

LIE HOV 0 1,336 546 32 1,086 216 97 1,399 43 3,406 13,968 0 1,174 512 39 218 87 27 333 7 2,077 8,562

LIE Mainline 4,859 89 30 0 27 10 2 39 0 5,419 EB 9,832 208 93 7 107 31 12 150 2 10,505

LIE HOV 0 1,048 336 36 163 38 23 224 9 1,661 7,080 0 1,603 643 29 824 191 82 1,097 39 3,442 13,947


WB LIE HOV 1671 342 2013 1154 169 90 1413 21 3495 WB 906 272 1178 95 15 7 118 7 1317

EB LIE HOV 442 119 561 121 25 10 156 8 742 EB 1727 434 2161 824 151 66 1041 24 3258

WB LIE HOV -303 204 -99 -68 47 7 -14 22 -89 WB 307 240 547 123 72 20 215 0 760

EB LIE HOV 642 217 859 43 12 13 68 1 919 EB -95 209 114 0 40 16 56 15 184

WB LIE HOV -18% 60% -5% -6% 28% 8% -1% 103% -3% WB 34% 88% 46% 129% 476% 262% 182% 4% 58%

EB LIE HOV 145% 182% 153% 35% 48% 133% 44% 15% 124% EB -5% 48% 5% 0% 27% 24% 5% 64% 6%

At Exit 59:



Grand


Grand

Total

LIE Mainline 9,639 246 116 2 162 48 15 225 4 10,514 WB 5,719 792 427 7 226 122 25 373 6 7,512

LIE HOV 0 1,251 528 19 828 210 93 1,131 29 3,010 13,524 0 1,100 552 24 253 145 27 424 6 2,119 9,631

LIE Mainline 4,305 447 251 9 210 99 20 328 9 5,562 EB 9,809 543 260 7 221 83 23 327 6 11,062

LIE HOV 0 689 334 24 132 101 17 251 6 1,312 6,874 0 1,391 602 12 659 210 78 947 27 3,009 14,070


WB LIE HOV 1288 406 1694 926 152 73 1151 41 2951 WB 647 260 907 72 12 6 90 19 1021

EB LIE HOV 249 89 338 68 11 5 84 8 431 EB 1344 196 1540 507 84 40 631 37 2243

WB LIE HOV -17 122 105 -98 58 20 -20 -12 59 WB 478 292 770 181 133 21 334 -13 1,098

EB LIE HOV 464 245 709 64 90 12 167 -2 881 EB 59 406 464 152 126 38 316 -10 766

WB LIE HOV -1% 30% 6% -11% 38% 27% -2% -28% 2% WB 74% 112% 85% 252% 1104% 342% 371% -67% 108%

EB LIE HOV 186% 275% 210% 94% 817% 250% 198% -25% 204% EB 4% 207% 30% 30% 149% 95% 50% -26% 34%

At Exit 62:



Grand


Grand

Total

LIE Mainline 8,275 183 88 1 94 40 10 144 2 8,886 WB 3,919 489 249 5 146 82 15 243 6 5,023

LIE HOV 0 1,033 462 9 631 159 67 857 21 2,432 11,318 0 762 367 7 160 98 17 276 4 1,427 6,450

LIE Mainline 2,799 371 196 4 226 77 21 324 11 3,851 EB 7,767 415 135 4 145 61 11 217 3 8,623

LIE HOV 0 312 142 12 0 37 3 40 0 514 4,366 0 1,128 533 7 528 174 69 772 20 2,490 11,113


WB LIE HOV 872 205 1077 759 125 60 944 30 2092 WB 366 78 444 67 11 5 83 4 547

EB LIE HOV 183 50 233 41 7 3 51 4 295 EB 924 237 1161 446 73 35 554 33 1771

WB LIE HOV 170 257 426 -128 34 7 -87 -9 340 WB 403 289 692 93 87 12 193 0 880

EB LIE HOV 141 92 233 -41 30 0 -11 -4 219 EB 211 296 507 82 101 34 218 -13 719

WB LIE HOV 19% 125% 40% -17% 27% 12% -9% -31% 16% WB 110% 371% 156% 139% 790% 246% 232% -7% 161%

EB LIE HOV 77% 184% 100% -100% 429% 0% -22% -100% 74% EB 23% 125% 44% 18% 138% 98% 39% -40% 41%% Diff

Modeled

Flows

WB

EB

Counts

# Diff

% Diff

Modeled

Flows

WB

EB

Counts

# Diff

% Diff

Modeled

Flows

WB

EB

Counts

# Diff

IV. Performance Measures for Scenario Evaluation



A number of measures of effectiveness (MOE) were defined to evaluate the performance of the

facility under different conditions. These performance measures were calculated for each of the

five scenarios, in each of the three scenario years, to provide a variety of metrics for comparison.

The MOEs are:

Peak Hour Volumes

Average Speeds

Demand over Capacity Ratio

Lane Miles of Congestion Vehicle Hours of Peak Hour Delay

Vehicle Miles of Peak Hour Travel

Person Trips per Mile

MOEs are calculated and reported at seven locations along the corridor, as illustrated in Figure

8. The full set of MOEs for each scenario and model year are shown in Technical Appendix A.

Figure 8: Locations for Comparison of MOEs

Peak Hour Volumes This MOE measures the traffic volume per facility due to policy changes and the effect of future

traffic growth in relation to the future network changes.

In this study, the AM and PM peak periods are defined as 7-9am in the morning and 4–6pm in

the evening. These are considered to be the 2 busiest hours of the typical 4 hour morning and

evening peak periods used in the NYBPM. The peak hour within this flat peak period is obtained

by dividing the peak period volume by 1.96, for both AM and PM.

To report this performance measure, assigned volumes (expressed in vehicles) are reported for

both AM and PM peak hours by mode (i.e. SOV, HOV2, HOV3+, Taxi, CPV1, CPV2+, EV,

Motorcycle, medium truck, heavy truck, commercial van, and bus) by facility (i.e. GPL and

HOV) by direction along the LIE study corridor at selected locations.



Average Speed This MOE measures the mobility of a facility. It measures the average speed of vehicles between

two points, defined as the length of a given highway segment divided by the time required to

traverse the same segment at the end of the highway assignment. It is an indicator of quality of

travel. As congestion develops, the average speed decreases.

Average speed (expressed in miles per hour, or MPH) is reported for both AM and PM peak

hours by facility (i.e. GPL and HOV) by direction at selected locations.

Demand Over Capacity Ratio Demand over capacity ratio (also known as volume over capacity ratio, or V/C) measures the

capacity sufficiency ratio. It is calculated by dividing traffic volume (i.e. the travel demand) by

the total capacity of the roadway. It reflects mobility and quality of travel of a given segment of

road by comparing roadway demand (vehicle volumes) with roadway supply (i.e. carrying

capacity). In other words, congestion becomes severe when the demand over capacity ratio

becomes higher. Spill-back occurs when the ratio is higher than 1.0.

In this study, the demand over capacity ratio is defined as the demand over capacity ratio for the

2 hour peak periods (i.e. not the typical 4 hour AM and PM peak periods in NYBPM). Demand

over capacity ratio is a standard highway assignment output in NYBPM by link by direction, and

was modified to reflect the 2 hour peak periods.

This performance measure is reported for both AM and PM peak hours by facility (i.e. GPL and

HOV) by segment (e.g. between interchanges, cross street or highway exits) by direction.

The concept of Level of Service (LOS) uses a rating system ranging from A (best) to F (worst)

representing the V/C ratios to indicate congestion level, as presented in Table 3. The LOS results

are represented by mode by facility by segment by direction.

Table 3: Level of Service (LOS) Definition

Demand Over Capacity Ratio (V/C)

General Characteristics Level of Service (LOS)

V/C =< 0.6 Free flow A

0.6 < V/C =< 0.7 Stable flow with slight delays B

0.7 < V/C =< 0.8 Stable flow with delays C

0.8 < V/C =< 0.9 High density but stable flow D

0.9 < V/C =< 1.0 Unstable flow, approaching capacity E

V/C < 1.0 Forced flow, breakdown condition and over capacity

F



Lane Miles of Congestion Lane mileage is defined as the total distance, in miles, covered by links belonging to a specific

road. It measures the total length and lane count of given segments of roadway. It is calculated

by multiplying the length of a road by the number of lanes.

Similarly, lane miles of congestion quantifies congestion in terms of lane mileage. It measures

the number of peak hour lane miles that have congested travel. In this study, congestion is

defined when the demand over capacity ratio is greater than 0.8, or at LOS D or worse.

When the V/C exceeds 1.0, spillback will occur, but static traffic assignment does represent such

queues, nor on which upstream links additional queues would form if there is more than one

roadway connecting into the link which is over-capacity. In order to define the extent of

congestion, it is assumed that (for V/C >1.0) the congestion would spill back in proportion to the

V/C ratio in relation to the length of the congested segment. Therefore, lane miles of congestion

is calculated for segments of roadway where the V/C threshold is greater than 0.8. For 0.8 < V/C



Person Trips per Mile Similar to vehicle miles of travel, person trips per mile is a measure of the density of travelers on

the road, in units of person trips per mile. It is a function of traffic volume and vehicle

occupancy.

A person trip is a trip made by one person in any mode of transportation. If there is more than

one person in a vehicle, each person is considered as making one person trip. For the purpose of

this study, vehicle occupancy by mode is defined in Table 4 below.

Table 4: Occupancy Assumed by Mode in the NYBPM

Vehicle Mode (Non Commercial) Vehicle Occupancy

SOV, CPV1 1.0

HOV2, CPV2 2.0

HOV3+, CPV3+ 3.46

TAXI 1.4

MC 1.0

This performance measure is reported for both AM and PM peak hours by facility (i.e. GPL and

HOV) by segment (e.g. between interchanges, cross street or highway exits) by direction, for

private vehicles only.

V. Future Year Forecasting Methods



Future years forecast model runs use the standard set of NYBPM 2010 future year highway

projects and socioeconomic demographic (SED) data forecasts. These model components are

used to predict the travel demand and highway network conditions in any future year.

Standard Forecast Procedures In addition to these standard NYBPM forecasts, for the LIE CPV custom application, forecasting

methods were derived for the special vehicles classes added for this analysis (i.e. CPVs, EVs,

and motorcycles). In order to generate future year CPV, EV, and motorcycle forecasts, the

following key assumptions were made:

(1) CPV-Eligible Vehicles Constant – Clean Pass enrollment requirements will become more stringent over time (i.e. between 2014, 2016, and 2035) so that the total number of

CPV-eligible vehicles remains constant at present day levels. This assumption was made

because the HOV lane is already operating at congested conditions, near its maximum

capacity, in 2014. In order for the lane to provide a benefit to travelers, and thus to

incentivize carpooling and/or purchasing more efficient vehicles, the total number of

CPV eligible vehicles would need to be controlled. Additionally, as vehicles become

more fuel efficient over time, as the cost of hybrid and electric vehicles goes down, and

as a growing number of people own hybrid or electric vehicles, the current 45 miles per

gallon (mpg) minimum requirement will become obsolete. In fact, US fuel efficiency

standards will be increased to 54.5 mpg for cars and light-duty trucks by model year

20257.

This was implemented through a factor in the algorithm that determines the probability

that each auto trip is made by a CPV mode (see section II’s sub-section on Approach to

Modeling with Adapted NYBPM 2010 Update, step 3). This factor was adjusted

iteratively until the number of CPV-eligible vehicles matched the absolute value from the

2014 base scenario.

(2) Ratio of EVs to CPVs – The ratio of EVs to total CPV-eligible vehicles will grow from 6.3% in 2014 to 12.9% in 2016 and 44.6% in 2035. These numbers are based on PHEV

and EV sales projections consistent with 6 NYCRR Part 218-4 Zero Emissions Vehicle

Mandate - NYSDEC, 2014. For the 2016 forecast (as for the 2014 baseline scenario),

both battery electric vehicles (BEVs) and plug-in hybrid electric vehicles (PHEVs) are

considered part of the EV category, and the ratio of EVs is calculated as the sum of the

BEV and PHEV forecast numbers divided by the total number of CPV-eligible vehicles.

The absolute number of BEVs and PHEVs from the 2025 NYSDEC forecasts exceeds the

total number of CPV-eligible vehicles. For the 2035 scenario, the 2025 proportion of EVs

out of the total number of CPV-eligible vehicles was used. Because of the high forecasts

of EVs, for the 2035 scenario it was assumed that the criteria for non-EV CPVs would be

defined as PHEVs (being that both hybrid and electric vehicles are expected to constitute

a substantially larger proportion of vehicles on the road), and NYSDEC’s BEV projection

would constitute the EVs for the purposes of the LIE HOV lane scenarios. These

7 White House Press Release, August 28, 2012. Available at https://www.whitehouse.gov/the-press-

office/2012/08/28/obama-administration-finalizes-historic-545-mpg-fuel-efficiency-standard.

https://govt.westlaw.com/nycrr/Document/I4e8fc622cd1711dda432a117e6e0f345?viewType=FullText&originationContext=documenttoc&transitionType=CategoryPageItem&contextData=(sc.Default)https://govt.westlaw.com/nycrr/Document/I4e8fc622cd1711dda432a117e6e0f345?viewType=FullText&originationContext=documenttoc&transitionType=CategoryPageItem&contextData=(sc.Default)



assumptions result in 12.9% and 44.6% being used as the percentage of CPV vehicles

assumed to be EVs in the 2016 and 2035 future year trip tables, respectively.

(3) Motorcycle Proportion – The probability for any given vehicle trip to be taken by motorcycle does not change in future years.

Following these assumptions, the procedures outlined in section II’s sub-section on Approach to

Modeling with Adapted NYBPM 2010 Update, were followed for the future year scenarios to

produce future year trip tables and highway network assignment results.

Unchecked (No Cap) Forecast Procedures In addition to the full set of forecasts under the previously described conditions, which assume

increasingly more stringent requirements for Clean Pass eligibility, a scenario was run assuming

“uncontrolled” growth in the number of Clean Pass eligible vehicles. This scenario was only run

for 2035, assuming that the criteria of 45 mpg for the Clean Pass program remained unchanged.

This scenario was implemented in the modeling framework by setting assumptions for the

proportion of hybrid and electric vehicles on the road in 2035. Based on analysis conducted by

the National Highway Traffic Safety Administration (NHTSA) presented to the Society of

Automotive Engineers (SAE), most vehicle manufacturers could comply with the 2025 CAFE

(Corporate Average Fuel Economy) standards by producing a fleet composed of 12% hybrid

electric vehicles, 1% PHEVs or BEVs, and a combination of other vehicle types8. For this study,

it was assumed that, after the 10 years between the 2025 adoption of the new CAFE standard and

the 2035 model year, vehicles on the road would be comprised of approximately 13% CPV-

eligible vehicles, with 1% of vehicles being EVs and 12% non-EV CPVs. This represents

substantial growth from the 2014 conditions, in which hybrid vehicles account for 0.7% of NY

State registered vehicles, and EVs only 0.033%.

In the remainder of this report, the 2035 scenario which allows “uncontrolled” growth in CPVs

will be referred to as the 2035 “No Cap” scenario.

8 Tamm, J. (January 21, 2015) "Regulatory Analysis of Powertrain Technologies: One Pathway for Compliance with

CAFE and GHG Emissions Standards." [PowerPoint Slides] SAE 2015 Government/Industry Meeting, Washington DC. Available at http://www.nhtsa.gov/staticfiles/rulemaking/pdf/cafe/Tamm_J_SAE-GI-Light-Duty-Powertrain-Technologies-for-CAFE.pdf.

VI. Discussion of Results: Baseline and Forecast



This section describes the MOEs at critical links in the LIE corridor for the baseline (2014) and

then for the two forecast years (2016 and 2035). Congestion levels in the corridor are very

similar in the AM peak direction as in the PM peak direction, and therefore, to simplify this

discussion, all results are discussed in terms of the AM peak hour (7 – 8 AM) measures. The

complete set of MOEs for all years, policy scenarios, peak periods and directions are provided in

Technical Appendix A for reference and further examination.

Seven key segments were identified for reporting, these include locations between HOV access

and egress which are roughly comparable to adjacent segments on the GPL. The seven key

segments are identified on the map in Figure 8.

Baseline Conditions - 2014 The 2014 Base Configuration results provide a general understanding of the current conditions

on the LIE and the relationship between the GPL and HOV facilities as they operated during the

calibration base year of 2014. The MOEs for this baseline condition are discussed below.

Volumes and Speeds Beginning at Exit 63 and moving in the peak, westerly direction, 2014 peak hour volumes in the

corridor (GPL and HOV combined) begin at under 5,000 and increase to over 8,000 vehicles

between Exits 53 and 55, near the interchange with the Sagtikos Parkway (see Figure 9). This is

the most heavily used section in the study area. Total corridor volumes diminish to

approximately 6,500 vehicles west of the interchanges with the Northern State Parkway.

Figure 9: 2014 AM West Bound Peak Hour - Total Volumes

-

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

9,000

Exit 36 - 37 Exit 39 - 40 Exit 46 Exit 49 Exit 53 - 55 Exit 59 Exit 62

Total Volumes

East to West, AM Peak



The volumes of vehicles that are eligible for use of the HOV facility follow a similar pattern

except that after the drop-off in volumes that is seen west of Sagtikos Parkway, HOV-Eligible

Vehicles (i.e. vehicles eligible to use the LIE HOV lane) increase in volume all the way to the

westerly end of the study area (see Figure 10). The HOV facility remains at approximately

1,500 vehicles per lane with the excess vehicles spilling over into the GPL (see Figure 11).

Figure 10: 2014 AM West Bound Peak Hour – HOV-Eligible Vehicles by Mode

Figure 11: 2014 AM West Bound Peak Hour - HOV-Eligible Vehicles by Facility

-

500

1,000

1,500

2,000

2,500


HOV-Eligible Vehicles

HOV2 HOV3+ TAXI CPV1 CPV2+ EV MC

-

500

1,000

1,500

2,000

2,500


2014 HOV-Eligible Vehicles by Facility

Vehicles in HOV Lane HOV-eligible in GPL



The approximate percentage of HOV-Eligible Vehicles that are served by the HOV facility

ranges from 65% to 85%. The theoretical excess demand for the currently defined HOV facility

of 15% to 35% may represent demand that cannot be served by an HOV facility that has reached

its capacity, and/or HOV-Eligible vehicles whose origin-destination needs are not served by the

access/egress locations of the HOV facility.

These volumes result in the average speeds on the LIE GPL and HOV facilities as shown in

Table 5. Consistent with these high levels of volumes, both facilities are showing average

speeds considerably lower than free flow. A key to the successful implementation of an HOV

facility is maintaining an average speed above that of the GPLs. Table 6 shows the comparison

of averages speeds on these facilities at the key locations. In most locations the average speed on

the HOV facility is only minimally above that on the GPL, and in some cases it is lower.

Table 5: 2014 AM West Bound Average Speeds

Facility Location Congested Speed (mph)

Free Flow Speed (mph)

Difference in Average Speeds

Percentage Difference

GPL Exit 36 - 37 32 55 -23 -42%

Exit 39 - 40 34 57 -23 -40%

Exit 46 21 49 -28 -57%

Exit 49 18 49 -31 -63%

Exit 53 - 55 19 54 -35 -65%

Exit 59 34 55 -21 -38%

Exit 62 44 54 -10 -19%

HOV Exit 36 - 37 38 55 -17 -31%

Exit 39 - 40 31 57 -26 -46%

Exit 46 23 49 -26 -53%

Exit 49 21 49 -28 -57%

Exit 53 - 55 33 54 -21 -39%

Exit 59 26 55 -29 -53%

Exit 62 40 54 -14 -26%



Table 6: 2014 AM West Bound – Difference in Averages Speeds (HOV-GPL)

Location GPL HOV Difference

Exit 36 - 37 32 38 6

Exit 39 - 40 34 31 -3

Exit 46 21 23 2

Exit 49 18 21 3

Exit 53 - 55 19 33 14

Exit 59 34 26 -8

Exit 62 44 40 -4

Volume Capacity Ratios The differences in average speeds shown above are consistent with the Volume Capacity ratios

(V/C) along the corridor, shown in Figure 12. Most MOE’s in this report are illustrated at the

seven key locations discussed above, but V/C is presented for every unique segment in order to

provide a general profile of congestion and flow in the corridor, and interactions between the

GPL and HOV lanes. With the exception of the segment between Exit 42 and 43, the V/C

conditions on the GPL and HOV lanes are similar, and hover generally between approximately

0.8 and 1.0.

Figure 12: 2014 AM West Bound Peak Hour - Volume/Capacity Ratio

The outlying segment, Exit 42-43, lies east of the interchange with Northern State Parkway at

Glen Cove Road. To access the Northern State Parkway from the LIE-HOV requires exiting the

HOV facility about three miles back near the interchange with Jericho Turnpike. Issues of

access/egress and desired travel patterns likely result in this anomaly.

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

Exit

33

- 3

6

Exit

36

- 3

9

Exit

39

- 4

0

Exit

40

Are

a

Exit

40

- 4

2

Exit

42

- 4

3

Exit

43

- 4

5

Exit

45

Are

a

Exit

45

- 5

0

Exit

50

- 5

1

Exit

51

- 5

3

Exit

53

Are

a

Exit

53

- 5

4

Exit

54

- 5

5

Exit

55

Are

a

Exit

55

- 5

6

Exit

56

- 5

7

Exit

57

- 5

8

Exit

58

- 6

0

Exit

60

Are

a

Exit

60

- 6

1

Exit

61

- 6

2

Exit

62

- 6

3

2014 Base Config. (1) HOV V/C GPL V/C



Aggregate Measures – Congestion, Hours of Delay and Vehicle Miles of Travel Three of the MOEs are aggregate measures of operations in a corridor. They are intended to

describe conditions on the corridor as a whole in terms that can be compared across scenarios.

These include Lane Miles of Congestions, Vehicle Hours of Peak Hour Delay, and Vehicle Miles

of Peak Hour Travel. The aggregate measures for this 40-mile LIE study area for the 2014

baseline are shown in Table 7.

Table 7: Aggregate Measures of Congestion – 2014 AM West Bound

Lane Miles of Congestion 38

Vehicles Hours of Peak Hour Delay 1,761

Vehicle Miles of Peak Hour Travel 281,379

Person Trips An important objective and benefit of HOV facilities is the ability to move more people through

the corridor with the same number of lanes and or vehicles. This is a direct result of an increase

in vehicle occupancy. The number of person trips at each of the key locations for the 2014

baseline condition is shown in Table 8. Approximately 35% of the person trips in the corridor

are currently served by the HOV facility (see Table 9).

Table 8: 2014 AM West Bound Peak Hour - Person Trips by Facility Type

Location SOV HOV2 HOV3+ TAXI CPV1 CPV2 CPV3+ EV1 EV2 TOTAL

GPL Exit 36 - 37 3,869 974 484 38 78 34 73 6 2 5,558

Exit 39 - 40 4,496 308 176 1 74 10 35 5 - 5,105

Exit 46 4,765 296 149 3 74 20 31 5 - 5,343

Exit 49 5,406 224 128 3 95 16 28 5 - 5,905

Exit 53 - 55 6,099 436 343 4 152 32 90 9 2 7,167

Exit 59 4,918 250 204 1 83 20 52 5 2 5,535

Exit 62 4,222 188 156 1 48 18 38 4 2 4,677

HOV Exit 36 - 37 - 1,372 817 64 282 68 156 22 4 2,785

Exit 39 - 40 - 1,542 893 64 338 82 180 28 4 3,131

Exit 46 - 1,246 848 21 477 94 183 38 4 2,911

Exit 49 - 1,206 844 21 492 92 183 39 4 2,882

Exit 53 - 55 - 1,154 848 15 476 82 183 36 6 2,800

Exit 59 - 1,276 934 14 423 78 232 40 8 3,005

Exit 62 - 1,054 817 7 322 58 180 29 4 2,470

TOTAL Exit 36 - 37 3,869 2,346 1,301 102 360 102 228 28 6 8,343

Exit 39 - 40 4,496 1,850 1,069 66 412 92 215 33 4 8,236

Exit 46 4,765 1,542 996 24 551 114 215 43 4 8,254

Exit 49 5,406 1,430 972 24 587 108 211 44 4 8,786

Exit 53 - 55 6,099 1,590 1,190 20 628 114 273 45 8 9,967

Exit 59 4,918 1,526 1,138 15 506 98 284 45 10 8,540

Exit 62 4,222 1,242 972 8 370 76 218 33 6 7,148



Table 9: 2014 AM Peak Hour - Share of Person Trips on the HOV Facility

Location % of Total HOV-Eligible

Vehicles in the HOV Lane % of All Vehicles Using the

HOV Lane % of All Person Trips in the

HOV Lane

Exit 36 - 37 64% 21% 33%

Exit 39 - 40 84% 23% 38%

Exit 46 84% 22% 35%

Exit 53 - 55 85% 21% 28%

Exit 59 73% 18% 35%

Exit 62 83% 23% 35%

Comparison of Policy Options and Forecasts The 2014 baseline conditions reflect the characteristics of the corridor during the calibration

year. Two additional years were analyzed – an initial implementation year (2016) and a long

range forecast year (2035). These forecasts are provided below. To simplify the report, only

two of the key sections of the LIE are discussed below. Exit 49 is the location for which full

classification counts were available for the calibration, and therefore is the location considered to

be most accurate. Exit 53-55 is the location with the highest volumes and congestion, and as

such presents the worst case condition. Also, with the baseline condition, only the AM Peak

West Bound direction is discussed here. All results for all scenarios, time periods, and directions

are provided in Technical Appendix A.

Volumes and Speeds Volumes and Speeds for 2016 and 2035 at Exit 49 are shown in Table 10. Table 11 contains

volumes and speeds for Exit 53-55. These data are further illustrated by the graphs in Figure 14,

Figure 17, Figure 16, and Figure 19. The center graphs in these figures show the number of

vehicles on the LIE eligible to use the HOV lane under each scenario’s criteria, separated by

those that are using the HOV lane and those that use the GPL despite being HOV-eligible. These

graphs also show a horizontal line at 1650 vehicles, assumed to be the hourly capacity of the

HOV lane. These graphs illustrate that the number of vehicles eligible to use the HOV lane

exceeds its capacity in 2014, 2016, and 2035 under the existing (or base) HOV eligibility

configuration.

In a highly congested corridor, such as the LIE, changing restrictions on the HOV lanes tends to

shift HOV-Eligible vehicles from to the GPLs and vehicles from the GPLs out of the corridor. In

this case, as observed at Exit 49, as HOV-eligibility becomes more restrictive we see the total

volumes, both on the HOV facility and in the GPLs remain constant. The exception to this

pattern occurs with the most restrictive policy scenario; the Tiered Occupancy scenario. In this

case the restrictions on the HOV-eligibility substantially decrease the HOV lane usage

sufficiently to increase the average speed on the HOV facility to 35 mph. Because the GPLs are

already operating over capacity, this reduction in volumes in the HOV lane at Exit 49 causes a

shift out of the corridor all together; reducing total volume by approximately 250 vehicles.

In 2035, at Exit 49 this same pattern occurs. As restrictions on the use of the HOV lane are

increased, congestion on the HOV facility as well as the GPLs remains constant until the most

restrictive policy (Tiered Occupancy) is implemented, at which point vehicles begin to move off



the LIE all together. However, the impact of restrictions on HOV-eligibility in 2035 is less

dramatic at Exit 49, because of the increased congestion on all facilities in the region. There is

less of a difference between the Tiered Occupancy policy scenario and the other scenarios, and

the improved average speed for the Tiered Occupancy scenario reaches only 24 mph in 2035 as

compared to 34 mph in 2016.

The conditions at Exits 53-55 are different from those at Exit 49. Here, as noted in the baseline

condition the volumes are the highest of any of the measured locations in the corridor. Total

corridor volumes are approximately 1,000 vehicles greater than at Exit 49. However the patterns

of HOV-Eligible vehicles and total volumes in the LIE corridor are similar to those evidenced at

Exit 49. For all the policy options except the Tiered Occupancy option, the congestion and

average speeds remain more or less constant. With the most restrictive policy scenario, the

Tiered Occupancy scenario, volumes in the HOV lane decrease relative to volumes in the GPL

sufficiently to allow an increase in speeds. In 2016 speeds in the HOV lane reach 37 mph, but in

the more congested 2035 speeds in the HOV lane in the Tiered Occupancy scenario, while faster

than the GPL by 11 mph, only reach an average speed of 27 mph.

At both Exit 49 and Exits 53-55 the total vehicles in the corridor remains more-or-less constant

except for the Tiered Occupancy scenario. In 2016 over 200 vehicles are lost from the corridor

(as compared to the Base Configuration) at both locations. In 2035 around 85 vehicles are lost

from the corridor. Also, in 2035, all scenarios reach or exceed capacity in the HOV lane at Exits

53-55, even the Tiered Occupancy scenario.

Table 12 shows changes in volumes on the parallel roadways near the county boundary between

Nassau and Suffolk Counties. This implies that the volumes lost from the corridor are generally

dispersed. However, these are relatively small numbers, and the travel patterns in the LIE

corridor suggest that many of the trips may be shorter distance trips. It is unlikely that these

additional trips on the parallel roadways are long distance trips that have been diverted from the

LIE.



Table 10: Forecasted AM West Bound Peak Hour Volumes and Speeds at Exit 49

No Cap Base Config.

No CPV EV Only HOV3+ Tiered

20

16

Non HOV-Eligible in GPL -- 5,662 5,770 5,805 5,860 5,962

HOV-Eligible in GPL -- 264 189 141 102 10

Total Veh. In GPL -- 5,926 5,959 5,946 5,962 5,972

Vehicles in HOV Lane -- 1,680 1,639 1,647 1,621 1,346

Total Veh. In Corridor -- 7,606 7,598 7,593 7,583 7,318

GPL Congested Speed -- 18 18 18 18 18

HOV Lane Congested Speed

-- 21 22 22 23 34

20

35

Non HOV-Eligible in GPL 5,947 5,866 5,838 5,962 6,018 6,150

HOV-Eligible in GPL 248 255 304 151 103 20

Total Veh. In GPL 6,195 6,121 6,142 6,113 6,121 6,170

Vehicles in HOV Lane 1,801 1,749 1,711 1,726 1,694 1,583

Total Veh. In Corridor 7,996 7,870 7,853 7,839 7,815 7,753

GPL Congested Speed 15 15 15 15 15 15

HOV Lane Congested Speed

18 18 20 19 20 24

Table 11: Forecasted AM West Bound Peak Hour Volumes and Speeds at Exits 53-55

No Cap Base Config. No CPV EV Only HOV3+ Tiered

20

16

Non HOV-Eligible in GPL -- 6,321 6,393 6,456 6,479 6,605

HOV-Eligible in GPL -- 552 537 478 475 422

Total Veh. In GPL -- 6,873 6,930 6,934 6,954 7,027

Vehicles in HOV Lane -- 1,798 1,748 1,754 1,704 1,409

Total Veh. In Corridor -- 8,671 8,678 8,688 8,658 8,436

GPL Congested Speed -- 19 18 18 18 18

HOV Lane Congested Speed -- 22 24 23 25 37

203

5

Non HOV-Eligible in GPL 6,569 6,422 6,322 6,494 6,551 6,589

HOV-Eligible in GPL 387 523 684 499 501 482

Total Veh. In GPL 6,956 6,945 7,006 6,993 7,052 7,071

Vehicles in HOV Lane 1,862 1,856 1,807 1,830 1,772 1,664

Total Veh. In Corridor 8,818 8,801 8,813 8,823 8,824 8,735

GPL Congested Speed 17 17 17 17 17 16

HOV Lane Congested Speed 22 20 22 21 23 27



Table 12: Changes in AM Peak Hour Volumes on Parallel Corridors – At the County Boundary

2016 2035

Base Config. Tiered Diff Base Config. Tiered Diff

Northern State Pkwy 8,315 8,398 84 8,795 8,840 45

Southern State Pkwy 12,364 12,416 52 12,837 12,837 1

Sunrise Highway 5,713 5,761 48 6,574 6,603 29

Total 26,392 26,576 184 28,206 28,281 75



2

01

4

Figure 13: Base Year 2014 AM Peak Hour West Bound Volumes and Speeds at Exit 49

7,050

7,250

7,450

7,650

7,850

8,050

8,250

8,450

8,650

8,850

Base Cnfg. No CPV EV Only HOV3+ Tiered

Total Veh. In Corridor

-

500

1,000

1,500

2,000

2,500


Vehicles in HOV Lane HOV Eligible in GPL

-

5

10

15

20

25

30

35

40


GPL Congested Speed HOV Lane Congested Speed



20

16

20

35

Figure 14: Forecasted AM Peak Hour West Bound Volumes and Speeds at Exit 49

7,050

7,250

7,450

7,650

7,850

8,050

8,250

8,450

8,650

8,850



-

500

1,000

1,500

2,000

2,500



-

5

10

15

20

25

30

35

40



7,050

7,250

7,450

7,650

7,850

8,050

8,250

8,450

8,650

8,850

No Cap Base Cnfg. No CPV EV Only HOV3+ Tiered


-

500

1,000

1,500

2,000

2,500

No Cap BaseCnfg.



-

5

10

15

20

25

30

35

40





20

14

Figure 15: Base Year 2014 AM Peak Hour West Bound Volumes and Speeds at Exits 53-55

7,050

7,250

7,450

7,650

7,850

8,050

8,250

8,450

8,650

8,850

Base Cnfg No CPV EV Only HOV3+ Tiered


-

500

1,000

1,500

2,000

2,500



-

5

10

15

20

25

30

35

40





20

16

20

35

Figure 16: Forecasted AM Peak Hour West Bound Volumes and Speeds at Exits 53-55

7,050

7,250

7,450

7,650

7,850

8,050

8,250

8,450

8,650

8,850



-

500

1,000

1,500

2,000

2,500



-

5

10

15

20

25

30

35

40



7,050

7,250

7,450

7,650

7,850

8,050

8,250

8,450

8,650

8,850



-

500

1,000

1,500

2,000

2,500

No Cap BaseCnfg.



-

5

10

15

20

25

30

35

40





Volume Capacity Ratios

In the 2016 Forecast Year

The five figures below show a comparison of the V/C ratios for the HOV lanes and the GPL for

all policy scenarios for 2016. The 2016 Base Configuration scenario is very similar to the 2014

Base Configuration scenario, as shown in Figure 12. The anomaly at Exits 42-43 discussed

above is apparent in all the policy scenarios for both the 2016 and 2035 forecasts. In 2016 the

V/C ratios show a similar pattern for the first three policy scenarios – Base Configuration, No

CPV, and EV Only. The HOV and GPLs are both highly congested to the extent that there is

little difference between the V/C ratios of the HOV and the GPL.

The HOV3+ policy scenario begins to show some improvement in the operations of the HOV

lanes at the ends of the corridor – before Exit 55 and after Exit 45. Between these exits the

highly congested conditions continue to exist for both the HOV and the GPL. With the

additional HOV usage restrictions introduced in the Tiered Occupancy policy scenario for 2016

we see an improvement in V/C ratios for the entire corridor. V/C ratios for the GPLs

LIE HOV Clean Pass Vehicle Operational Performance ... · The Long Island Expressway Clean Pass Vehicle program allows eligible, low-emission, energy- efficient vehicles to use the

Documents