Report_Traffic_2008.pdf

Government of The Virgin Islands Department of Public Works

Improvements to Veterans Drive (Route 30)

Contract No. PC-PWD-051-2006

Traffic Report

April 2008

Prepared by:

Parsons Brinckerhoff, Inc.

7650 Corporate Drive, Suite 300

Miami, Florida 33126

PB Americas, Inc. Traffic Report

Improvements to Veterans Drive (Route 30) Page i

Table of Contents

1. Introduction

1.1 Project Location

1.2 Project Description

2. Data Collection and Analysis

2.1 Existing Roadway Geometry

2.2 Existing Intersections

2.3 Daily Traffic Counts

2.4 Intersection Turning Movement Counts

2.5 Balanced Peak Hour Volumes

2.6 Crash Data

3. Traffic Simulation Model

3.1 Model Development

3.2 Model Calibration

4. Measures of Effectiveness

4.1 Level of Service Definition

4.2 Level of Service at Signalized Intersections

4.3 Level of Service at Unsignalized Intersections

4.4 Segments Level of Service

4.5 Network-wide Statistics

4.6 Link Speeds

5. Analysis of Existing Conditions

5.1 Intersection Level of Service

5.2 Segment Level of Service

6. Analysis of Future Conditions

6.1 Improvements Analyzed

6.2 Future Traffic Volumes

6.3 Future Signal Timing and Phasing

6.4 Future Intersection Level of Service Analysis

6.5 Future Segment Level of Service Analysis

6.6 Queue Length Analysis

6.7 Network-wide Comparison Analysis

7. Conclusion and Recommendations

Improvements to Veterans Drive (Route 30) Page ii

List of Figures

Figure 1.1 – Project Location Figure 2.1a – 2008 Existing AM Peak Hour Balanced Volumes Figure 2.1b – 2008 Existing AM Peak Hour Balanced Volumes Figure 2.2a – 2008 Existing PM Peak Hour Balanced Volumes Figure 2.2b – 2008 Existing PM Peak Hour Balanced Volumes Figure 3.1 – VISSIM Existing Condition Network Figure 3.2 – VISSIM Model Congestion along Veterans Drive, AM Peak Period Figure 3.3 – VISSIM Model Congestion along Veterans Drive, PM Peak Period Figure 4.1 – VISSIM Model Travel Time Section

List of Tables

Table 2.1 - Major Roads in the Study Area Table 2.2 – Intersections in the Study Area Table 2.3 – Crash Data Summary by Location Table 2.4 – Crash Data Summary by Crash Type Table 3.1 – AM Peak Period Calibration Results Table 3.2 – PM Peak Period Calibration Results Table 4.1 – Level of Service Definitions for Signalized Intersections Table 4.2 – Level of Service Definitions for Unsignalized Intersections Table 4.3 – Urban Street LOS Criteria (HCM Exhibit 15-2) Table 5.1 – Existing Levels of Service for Intersections along Veterans Drive Table 5.2 – Existing Levels of Service for Segments along Veterans Drive – AM Peak Hour- Eastbound Direction Table 5.3 – Existing Levels of Service for Segments along Veterans Drive – PM Peak Hour- Eastbound Direction Table 6.1 – Census Comparison Table 6.2 – Traffic Count Comparison Table 6.3 – Delay and Level of Service Summary – AM Peak Period Table 6.4 – Delay and Level of Service Summary – PM Peak Period Table 6.5 Segment Average Speed and Level of Service - Eastbound Direction (AM Peak Hour) Table 6.6 Segment Average Speed and Level of Service - Eastbound Direction (PM Peak Hour) Table 6.7 Storage Length – Build AM Table 6.8 Storage Length – Build PM Table 6.9 Recommended Storage Lengths Table 6.10 – Network-wide Performance

Appendices Appendix A – Intersection Approach Counts Appendix B – Turning Movement Counts Appendix C – Traffic Volume Diagrams Appendix D – Crash Data (Omitted) Appendix E – Detailed Network Diagrams Appendix F – VISSIM Simulation Output Tables

Improvements to Veterans Drive (Route 30) Page 1 of 42

1. Introduction

The purpose of this traffic study is to document the results of the analysis of the Veterans Drive Improvement Project in Saint Thomas, US Virgin Islands for the Department of Public Works. The analysis was prepared to assess existing conditions as well as future no-build and build conditions in the area.

A previous report was submitted in March 2007 documenting improvements to Veterans Drive between Kronprindsens Tvaer Gade and William G. Lewis Lane. The traffic data collected as part of this new study supersedes the data in the previous report

This report describes the data collection effort for this project, tools used to develop and calibrate the micro simulation model and discusses the results of the analyses. The report also evaluates traffic circulation and accident history within the study area with particular focus on Charlotte Amalie’s commercial district.

1.1 Project Location

Veterans Drive is located on the south side of the Saint Thomas and is currently the main thoroughfare used to link the west and east sides of the Island, via the Downtown area. In addition, as Veterans Drive runs through the heart of Charlotte Amalie’s waterfront, it takes the role of being the local distributor road, providing access to the commercial district, the Legislative building, and historic Fort Christian.

Veterans Drive also serves as the main link between Havensight, Charlotte Amalie, and the west side of the Island including the Cyril E. King International Airport. Although not located within the study area, Havensight is an important element of the transportation system on the Island as it is a commercial cluster and the docking area for most of the cruise ships coming to the Island throughout the year.

Figure 1.1 shows the location of the project in relation to the island. The network developed for the traffic analysis includes most of the Charlotte Amalie urban area. The limits of the study area are Veterans Drive from Kronprindsens Tvaer Gade to Tolbod Gade.

1.2 Project Description

As part of this project, reconstruction will add a raised median along Veterans Drive and add exclusive right turn lanes for the westbound approach at the following intersections:

• Veterans Drive and Kronprindsens Tvaer Gade

• Veterans Drive and General Gade

• Veterans Drive and Strand Gade

Improvements will also include the addition of one through lane on the eastbound approach at the intersection of Veterans Drive and Tolbod Gade.

Figure 1.1 – Project Location

Page 2 of 42Improvements to Veterans Drive (Route 30)

Study Area

Improvements to Veterans Drive (Route 30) Page 3 of 42

2. Data Collection and Analysis

Collection of existing data within the study area is a crucial element in setting a base to be used for analysis of future conditions and to also have a good understanding of the existing conditions on the roadways to be analyzed. Extensive data collection was therefore performed including an inventory of existing roadways within the study area, daily traffic counts, peak hour intersection turning movements, crash data history, and license plate surveys. The following sections describe the data collection effort and the analysis of the collected data.

2.1 Existing Roadway Geometry

The layout of the roadway network, the Island’s geography and the land use development configuration of the urban area in Saint Thomas are set up in a way that the major traffic movement occurs in the east-west direction. Veterans Drive is the major arterial and is responsible for providing east-west mobility through the downtown area.

Veterans Drive is a two-way four-lane roadway west of Tolbod Gade. East of Tolbod Gade, in the heart of Charlotte Amalie, the eastbound direction is reduced to one lane while the westbound direction is still two lanes. Table 2.1 provides a description of the major roads found in the study area, their direction and their respective number of lanes.

Table 2.1 - Major Roads in the Study Area

Roadway Type Traffic

Direction Number of Lanes

Veterans Drive Two Way E-W 4 lanes from west of Kronprindsens Tvaer Gade to Tolbod Gade. 3 Lanes east of Tolbod Gade 1

(EB) and 2 (WB).

Kronprindsens Tvaer

Gade One Way N 2

NYE Tvaer Gade One Way S 2

General Gade Two Way N-S 2

Strand Gade One Way N 1

Guttets Gade One Way N 1

Store Tvaer Gade One Way S 1

Raadets Gade One Way S 1

Tolbod Gade One Way N 2

Improvements to Veterans Drive (Route 30) Page 4 of 42

2.2 Existing Intersections

As a general rule, intersections are critical on any roadway network and their operational characteristics have an important impact on the quality of mobility in specific areas. In the field it was found that even though there are loop detectors at the approaches along Veterans Drive, these are not operating at the moment and the intersections along the corridor have a Pre-timed operation. During peak hours there are policemen guiding traffic at all the signalized intersections. Table 2.2 shows a list of the intersections on the study area and the intersection type (signalized or non-signalized)

Table 2.2 – Intersections in the Study Area

Intersection Intersection Type

Veterans Drive and Kronprindsens Tvaer Gade Signalized

Veterans Drive and NYE Tvaer Gade Signalized

Veterans Drive and General Gade Signalized

Veterans Drive and Strand Gade Unsignalized

Veterans Drive and Guttets Gade Signalized

Veterans Drive and Store Tvaer Gade Unsignalized

Veterans Drive and Raadets Gade Signalized

Veterans Drive and Tolbod Gade Signalized

2.3 Daily Traffic Counts

Daily traffic counts are collected to provide an understanding of the traffic patterns over a 24-hour period along specific sections of roadways. The daily counts allow for calculation of the percentage of traffic occurring at each hour of the day and to determine the hours where the highest volume of traffic (peak hours) occurring on a particular roadway segment. Daily counts are also used to validate intersection turning movement volumes at each approach to the intersection where daily counts are taken.

In order to reduce fluctuations between days, daily counts are collected over a 48-hour period and averaged out to have a representative volume for a typical day.

Forty eight hour counts were collected at the following locations within the study area from January 9, 2008 to January 10, 2008.

o Veterans Drive west of Kronprindsens Tvaer Gade o Veterans Drive between Strand Gade and Guttets Gade o Veterans Drive east of Tolbod Gade

Improvements to Veterans Drive (Route 30) Page 5 of 42

The 48-hour counts included in Appendix A, show that between 26,400 and 31,850 vehicles travel along Veterans Drive on a daily basis. Based on these counts, the morning and afternoon peak periods occur between 7:00 AM and 8:00 AM and 5:00 PM and 6:00 PM, with approximately eight percent of the daily traffic occurring during the AM peak hour and seven percent in the PM peak hour.

As tourism is Saint Thomas’ main industry, with most of it associated with the cruise ship industry, traffic is impacted by the presence of ships on the island and by the seasonality of the industry. More than two million tourists arrived by cruise ship in 2004 in Saint Thomas. The ships only stay for a day, arriving in the morning and departing in the evening. Most of the cruise ship passengers head to Charlotte Amalie via taxi or public transport. Based on the cruise ship activity log obtained from the West Indian Company Ltd., the company who runs the port, last year’s cruise ship season peaked during the month of February. The data collected as part of this report was acquired in the middle of the peak tourist season, representing average conditions throughout the year.

In addition to personal vehicles, traffic in the Downtown area is composed of several other modes. Public transportation on the Island mainly consists of privately run Safari Taxis and traditional taxi services.

2.4 Intersection Turning Movement Counts

Intersection turning movements are used as input into the traffic operational software for analysis of existing conditions. Turning movements are also used to develop future traffic volumes at intersections for analysis of future conditions. While the daily counts were being collected, turning movement counts were also collected at the following intersections between 7:00 AM to 9:00 AM and from 4:00 PM to 6:00 PM:

o Veterans Drive and Kronprindsens Tvaer Gade o Veterans Drive and Nye Tvaer Gade o Veterans Drive and General Gade o Veterans Drive and Strand Gade o Veterans Drive and Guttets Gade o Veterans Drive and Store Tvaer Gade o Veterans Drive and Raadets Gade o Veterans Drive and Tolbod Gade

The actual turning movements by intersection and by 15-minute intervals are included in Appendix B.

Improvements to Veterans Drive (Route 30) Page 6 of 42

2.5 Balanced Peak Hour Volumes

Peak hour volumes from the turning movement counts were checked for reasonableness against the link volumes from the daily counts. Once the traffic counts were reviewed for consistency, a balancing process was conducted. The process of balancing traffic volumes refers to the reconciliation of volumes between intersections to minimize the difference in the volume of vehicles leaving one intersection and arriving to queue at an adjacent intersection. This is a standard process used to account for all volumes entering and exiting the study area based on the traffic counts collected.

The procedure followed to balance the volumes was to first establish the tolerance level for the difference between two points at approximately plus or minus ten percent. In locations where the differences were outside of the tolerance levels, the highest volume was used as the control volume and differences carried over to the next intersection. In some instances, significant building and/or land-uses were found between intersections.

In these cases, the differences in volumes were maintained between the intersections to account for traffic entering or exiting major land-uses.

Figures 2.1 and 2.2 show the resulting balanced volumes. Detailed spreadsheets for each period are included in Appendix C. The sheets also show turning movement percentages at each of the intersections for existing conditions.

195 26

50% 50% 78% 22% 100% 35 34 3% 107 108 76 74 6% 153 42 54 4% 56 26

1215 1252 1217 1181 97% 1181 1201 1201 100% 1181 1289 1316 1240 1215 94% 1257 1210 1156 96% 1201 1227 1177

14 1% 15 12 1% 13 1420 1217 1217 99% 1276 1291 1231 1124 100% 1159 1159 1124 971 99% 1036 1049 983 983 100% 969 969 942

K ronprindsends

North

N ye

TvaerG ade

G eneralG ade

Strand G ade

Legend

12%

Approach Volume

Turning Movement Volume

Turning Movement Percentage

Balanced Volumes

48-hr Volume Count

St. Thomas FIGURE 2.1a

2008 Existing Conditions Balanced AM-Peak Hour Volumes

64 43 114

97% 3% 100% 65% 35% 57 5% 59 62 2 4 0% 4 43 74 40 487 42% 521

1183 1126 95% 1167 1169 1135 1131 100% 1165 1208 1082 1082 100% 1208 1248 1166 679 58% 727 727 789

94 9% 98 39 4% 39 983 921 91% 964 1062 1015 1015 100% 1051 1051 1015 941 100% 985 985 941 941 96% 938 977 980

G uttets

G ade

Store TvaerG ade

R aadets

G ade

Tolbod G ade

Legend

12%

Approach Volume

Turning Movement Volume

Turning Movement Percentage

Balanced Volumes

48-hr Volume Count

St. Thomas FIGURE 2.1b

2008 Existing Conditions Balanced AM-Peak Hour Volumes

177 133 45

48% 52% 58% 42% 100% 1092 75 8% 88 85 92 153 15% 176 77 56 140 16% 161 45 1132 970 895 92% 1044 1044 903 903 100% 1044 1136 988 835 85% 960 1016 886 746 84% 856 901 908

10 1% 10 26 3% 27 1269 1063 1063 99% 1099 1109 1073 988 100% 1006 1006 988 911 97% 938 965 937 937 100% 885 885 919

K ronprindsends

North

N ye

TvaerG ade

G eneralG ade

Strand G ade

Legend

12%

Approach Volume

Turning Movement Volume

Turning Movement Percentage

Balanced Volumes

48-hr Volume Count

St. Thomas FIGURE 2.2a

2008 Existing Conditions Balanced PM-Peak Hour Volumes

30 70 84

67% 33% 100% 43% 57% 82 8% 72 20 10 7 1% 6 70 36 48 378 36% 346

1029 947 92% 829 839 929 922 99% 833 903 885 885 100% 903 951 1040 662 64% 605 605 641

151.4 14% 158 67 6% 70 937 917 86% 957 1115 1069 1069 100% 1063 1063 1069 1033 100% 1073 1073 1033 1033 94% 1074 1144 1100

G uttets

G ade

Store TvaerG ade

R aadets

G ade

Tolbod G ade

Legend

12%

Approach Volume

Turning Movement Volume

Turning Movement Percentage

Balanced Volumes

48-hr Volume Count

St. Thomas FIGURE 2.2b

2008 Existing Conditions Balanced PM-Peak Hour Volumes

Improvements to Veterans Drive (Route 30) Page 11 of 42

2.6 Crash Data

Crash data was obtained from the Records Division of the Virgin Islands Police Department for the four most recent years available. The data provided ranged from 2003 to 2006. 219 records were collected for Veterans Drive within the study area.

Table 2.3 shows the summary of the location and the number of records obtained at each location. Table 2.4 provides a summary of the most prevalent crash types by year.

Due to the relatively large fluctuations in the number of crashes by year, there is a certain level of uncertainty about whether all available crash data for the area under study was supplied. This uncertainty decreases the ability to provide accurate trends within the crash data, but the data still provides insight into the safety of traffic operations within the study area.

From the data, rear-end crashes occurred most often with 66 crashes, or 30 percent of the total. This crash type, is typical of facilities that experience heavy congestion with “stop and go” traffic flows. Additionally, the bottle neck situation created by the lane drop for eastbound traffic at Veterans Drive and Tolbod Gade is largely responsible for the relatively high number of sideswipe crashes (40, 18 percent) occurring within the study area. It is expected that the future increase in travel lanes east of Tolbod Gade will alleviate a large percentage of these two crash types.

With the future creation of new right-turn lanes for westbound traffic, and the included protected signal phasing for these movements, it is likely that a number of rear-end, angle, and right-turn crashes will be prevented.

Table 2.3 – Crash Data Summary by Location

Intersection 2003 2004 2005 2006 Total Kronprindsens Tvaer Gade 5 13 3 10 31 Nye Tvaer Gade 1 1 0 1 3 General Gade 3 7 6 12 28 Strand Gade 5 10 0 8 23 Guttets Gade 1 5 0 5 11 Store Tvaer Gade 4 9 6 4 23 Raadets Gade 1 12 4 11 28 Tolbod Gade 0 0 0 4 4 Other 9 26 11 22 68

Improvements to Veterans Drive (Route 30) Page 12 of 42

Table 2.4 – Crash Data Summary by Crash Type

Summaries of the collected crash data is included in Appendix D.

2003 2004 2005 2006 Total Percent Total Crashes 29 83 30 77 219 - Fatal Crashes 0 0 0 0 0 - Injury Crashes 5 16 5 5 31 -

Rear-End 10 22 9 25 66 30.1% Head-On 2 1 0 1 4 1.8% Angle 2 13 6 12 33 15.1% Left-Turn 3 8 0 4 15 6.8% Right-Turn 3 9 2 2 16 7.3% Sideswipe 2 16 7 15 40 18.3% Backed Into 2 4 1 2 9 4.1% Coll w/ Parked Car 3 1 1 4 9 4.1% Other 2 9 4 12 27 12.3%

Crash Type

Improvements to Veterans Drive (Route 30) Page 13 of 42

3. Traffic Simulation Model

A traffic simulation model was developed to analyze traffic operation within the study area. The simulation model allowed analysis of the entire project area and it demonstrates how traffic conditions at one location in the area affect adjacent streets and intersections.

Numerous traffic simulation packages are available for analysis of traffic conditions in a study area. The traffic simulation model was built using Vissim 4.30. VISSIM is a microscopic, time-step and behavior-based simulation software that can model and analyze urban traffic and transit operations under different schemes of lane configuration, traffic composition, traffic signals and transit stops. The VISSIM model also allows analysis of left-hand driving as is the case in Saint Thomas.

The following sections describe the various elements used to develop the traffic simulation model and the calibration of such model to represent conditions in Saint Thomas within the study area.

3.1 Model Development

The following are the various elements used and considered in development of the traffic simulation model for the Veterans Drive Improvement Study. The model included all of the roadways within the study limits as shown on Figure 3.1, which is an actual screenshot from the model.

Traffic composition. Information collected from the 48-hour counts was used to create vehicle input points.

Links and connectors. Aerial photographs were used as background images to develop links and connectors at scale, according to the actual geometry of the network. Both the aerial photographs and the information collected in the field were used to confirm that the existing operational and geometrical characteristics were accurately coded. On each link and connector, the number of lanes and lane width were coded.

Routing decision points and associated routes. Routing decisions were used to define turning movements at the intersections. Turning movements were entered as percentages for each direction. The AM and PM peak hour volumes and turning movement percentages shown on the sheets included in Appendix C were used for the analysis.

Signal Controllers. Even though there are detection loops on the approaches at the intersections along Veterans Drive, the detection system is not in service and the intersections are operating as Fixed-Time Controllers. Furthermore, policemen direct traffic during the AM and the PM peak hours. For this reason it was decided to develop signal timings in Synchro based on the existing traffic volumes and utilize fixed-time controllers for the signalized intersections.

Improvements to Veterans Drive (Route 30) Page 14 of 42

Desired speed decisions. Desired speed decisions are locations in the network that were used to code posted speed limits across the network. The speed limit along Veterans Drive is 20 mph.

Turning Speeds. Vissim uses Reduced Speed Areas for replicating the reduction in speed that occurs when a vehicle approaches an intersection and turns. When approaching a reduced speed area in the model, vehicles progressively adjust to the lower turning speed and automatically get back to its original speed after leaving the reduced speed area. The reduced speed area for left turns follows a speed distribution with a minimum speed of 9.3 mph and a maximum of 12.4 mph, whereas the speed distribution for the right turns has a minimum value of 12.4 mph and a maximum value of

15.5 mph.

Priority rules. Priority rules were used to designate the right-of-way for conflicting movements at signalized and unsignalized intersections. Priority rules include defining stop line locations, minimum headways, minimum gap times, and conflict areas in order to model the travel behavior when vehicles go through intersections. Priority rules are used to simulate situations, such as permissive left turns, right turns on red and clear box areas.

Improvements to Veterans Drive (Route 30) Page 15 of 42

Figure 3.1 – VISSIM Existing Condition Network

Improvements to Veterans Drive (Route 30) Page 16 of 42

3.2 Model Calibration

An important part of the traffic simulation model development is making sure that the coded data replicates traffic conditions within the study area being analyzed. Because the main input into the model are the collected traffic counts, traffic volumes are the main measure of effectiveness used to determine whether the model accurately replicates traffic conditions. It is important to note that traffic conditions vary with time and are highly dependant on other factors such as individual travel behavior that cannot be accounted for in a traffic operation model. The model developed for the Veterans Drive Improvement project attempts to reflect average conditions typically encountered in the study area.

The volumes produced by the VISSIM model or processed volumes were compared to the balanced counts. The model was considered validated when for most of the roadway segments, the processed volumes were within 10 percent of the balanced counts. Comparisons between processed and balanced volumes are shown on Tables

3.1 and 3.2.

As seen from the tables, the overall difference between processed and balanced volumes is within the set threshold of 10 percent, which is an indication that the model is well calibrated.

Based on field observations, during both the AM and the PM peak hours, traffic in the eastbound direction along Veterans Drive experiences a considerable amount of delay due to the heavy volume and lane reduction occurring at the intersection with Tolbod Gade. Figures 3.2 and 3.3, show that the traffic simulation model in Vissim is also replicating this condition.

Improvements to Veterans Drive (Route 30) Page 17 of 42

Table 3.1 – AM Peak Period Calibration Results

B al an ce d To ta l B al an ce d

V ol um e

To ta l P ro ce ss ed

V ol um e

P ro ce ss ed C on tro l

V ol um e

B y

A pp ro ac h B y A pp ro ac h

V ol um e

D el ay [s

E B

T+ L

9%

D

W B

T+ R

-1

A

E

B T

9%

C

W B

T

-2

A

L

R

E B

T+ L

9%

C

W B

T+ R

-3

A

S

B L+

R

4%

C

E

B T+

L

D

T

R

S B

L+ R

-2

C

E

B T

E

W

B T

-4

A

L

R

L

T

T

R

B al an ce d To ta l B al an ce d

V ol um e

To ta l P ro ce um e

P ro ce ss ed S to pp ed

V ol um e

B y

A pp ro ac h B y A pp ro ac h

V ol um e

D el ay [s

E B

T+ L

D

W

B T

-3

A

S

B L

8%

C

E B

T+ L

D

W

B T

-4

A

S

B L

-9

B

S ig na liz ed In te rs ec tio ns

Ta bl e 3.

- E xi st in g

C on di tio ns

AM

P ea k

H ou r

(P ro ce ss ed

V ol um es

, C on tr ol

D el ay a nd In te rs ec tio n

LO

S)

IN

TE

R

SE

C

TI

O N

L O

S

0%

A pp ro ac h M ea su re

P oi nt

D el ay

S ec tio n

K ro np rin ds en s Tv ae r G ad e an d

Ve te ra ns

D riv e

In te rs ec tio n

D el ay T im e

(A vg

M ov em en t

D el ay T im e by

A pp ro ac h (A vg

VI

SS

IM

O

U

TP

U T

Ap pr oa ch

LO

S

C

F C

R aa de ts

G ad e an d Ve te ra ns D riv e

S B

W B

G en er al

G ad e an d Ve te ra ns D riv e

G ut te ts

G ad e an d Ve te ra ns D riv e

To lb od G ad e an d

Ve te ra ns

D riv e

N

YE

T va er G ad e an d

Ve te ra ns

D riv e

U ns ig na liz ed

In te rs ec tio ns

In te rs ec tio n

VI

SS

IM

O

U

TP

U T

IN

TE

R

SE

C

TI

O N

L O

S

M ea nt

D tio n

A pp ro ac h M ov em en t

B

D el ay T im ro ac h (A vg

C rit ic al

D

Ti m e Ap pr oa ch

LO

S

C

-3

D St ra nd

G ad e an d Ve te ra ns D riv e

St or e Tv ae r G ad e an d

Ve te ra ns

D riv e

D

A

C

D S

B

7%

E B

E C

A

W B

-4

Improvements to Veterans Drive (Route 30) Page 18 of 42

Table 3.2 – PM Peak Period Calibration Results

B al an ce d To ta l B al an ce d

V ol um e

To ta l P ro ce um e

P ro ce ss ed C on tro l

V ol um e

B y

A pp ro ac h B y A pp ro ac h

V ol um e

D el ay [s

E B

T+ L

7%

C

W B

T+ R

-5

A

E

B T

7%

C

W B

T

-4

A

L

R

E B

T+ L

5%

C

W B

T+ R

-5

A

S

B L+

R

-3

C

E B

T+ L

8%

D

T

R

S B

L+ R

-7

C

E

B T

E

W

B T

-4

A

L

R

L

T

T

R

B al an ce d To ta l B al an ce d

V ol um e

To ta l P ro ce um e

P ro ce ss ed S to pp ed

V ol um e

B y

A pp ro ac h B y A pp ro ac h

V ol um e

D el ay [s

E B

T+ L

7%

E

W B

T

-5

A

S B

L

4%

D

E

B T+

L

E

W B

T

-3

A

S B

L

6%

C

S ig na liz ed In te rs ec tio ns

Ta bl e 3.

- E xi st in g

C on di tio ns

PM

P ea k

H ou r

(P ro ce ss ed

V ol um es

, C on tr ol

D el ay a nd In te rs ec tio n

LO

S)

IN

TE

R

SE

C

TI

O N

L O

S

-1 1%

A pp ro ac h M ea nt

D tio n

K ro np rin ds en s Tv ae r G ad e an d

Ve te ra ns

D riv e

In te rs ec tio n

D el ay T im e

(A vg

M ov em en t

D el ay T im ro ac h (A vg

VI

SS

IM

O

U

TP

U T

Ap pr oa ch

LO

S

B

F C

R aa de ts

G ad e an d Ve te ra ns D riv e

S B

W B

G en er al

G ad e an d Ve te ra ns D riv e

G ut te ts

G ad e an d Ve te ra ns D riv e

To lb od G ad e an d

Ve te ra ns

D riv e

N

YE

T va er G ad e an d

Ve te ra ns

D riv e

U ns ig na liz ed

In te rs ec tio ns

In te rs ec tio n

VI

SS

IM

O

U

TP

U T

IN

TE

R

SE

C

TI

O N

L O

S

M ea nt

D tio n

A pp ro ac h M ov em en t

B

D el ay T im ro ac h (A vg

C rit ic al

D

Ti m e Ap pr oa ch

LO

S

C

-3

E St ra nd

G ad e an d Ve te ra ns D riv e

St or e Tv ae r G ad e an d

Ve te ra ns

D riv e

E

A

C C

S B

1%

E B

D B

A

W B

-4

Improvements to Veterans Drive (Route 30) Page 19 of 42

Figure 3.2 – VISSIM Model Congestion along Veterans Drive, AM Peak Period

Improvements to Veterans Drive (Route 30) Page 20 of 42

Figure 3.3 – VISSIM Model Congestion along Veterans Drive, PM Peak Period

Improvements to Veterans Drive (Route 30) Page 21 of 42

4. Measures of Effectiveness

In order to measure and describe the operational status of the local roadway network, measures of effectiveness are used. For the Veterans Drive Improvement Study, intersection level of service, segment level of service, queue lengths and overall network statistics were used to quantify existing and future traffic operations and the benefits of the build alternative. The following sections describe each of these measures of effectiveness.

4.1 Level of Service Definition

Traffic engineers use a grading system called Level of Service (LOS). The LOS grading system involves a rating scale from A, indicating free-flow traffic conditions with little or no delay at intersections; to E, representing unstable flow conditions with traffic volumes at or near design capacity, resulting in major delays for vehicles; and F, representing over-saturated conditions where traffic flows exceed design capacity resulting in long queues and delays.

4.2 Level of Service at Signalized Intersections

The LOS at signalized intersections for operational analysis is based on the weighted average control delay measured in seconds per vehicle. Peak hour traffic volumes, existing lane configurations, and signal timing/phasing are used as inputs in the level of service analysis. The 2000 Highway Capacity Manual (2000 HCM) operations methodology, Synchro 6 and VISSIM 4.3, were used for this analysis. Table 4.1 summarizes the relationship between the average control delay per vehicle and LOS for signalized intersections.

Table 4.1 – Level of Service Definitions for Signalized Intersections

Level of Service Description Avg. Control Delay

(Seconds)

A Operations with very low delay occurring with favorable progression and/or short cycle length. < 10.0

B Operations with low delay occurring with good progression and/or short cycle lengths. 10.1 to 20.0

C Operations with average delays resulting from fair progression and/or longer cycle lengths. Individual cycle failures begin to appear. 20.1 to 35.0

D Operations with longer delays due to a combination of unfavorable progression, long cycle lengths, or high V/C ratios. Many vehicles stop and individual cycle failures are noticeable.

35.1 to 55.0

E Operations with high delay values indicating poor progression, long cycle lengths, and high V/C ratios. Individual cycle failures are frequent occurrences. This is considered to be the limit of acceptable delay.

55.1 to 80.0

F Operation with delays unacceptable to most drivers occurring due to over saturation, poor progression, or very long cycle lengths. > 80.0

Source: Highway Capacity Manual (HCM) operation methodology, 2000.

Improvements to Veterans Drive (Route 30) Page 22 of 42

Figure 4.1 is an example of how travel time sections are determined in VISSIM. The travel time sections are used to determine the LOS for each intersection approach.

Based on travel time sections VISSIM generates delay data for each one of the approaches at the intersections. The weighted average delay for all of the approaches determines the LOS of the intersection as a whole.

4.3 Level of Service at Unsignalized Intersections

At unsignalized intersections the LOS rating is based on the weighted total delay for the critical approach. Total delay is defined as the amount of time required for a driver to stop at the end of a queue of vehicles, wait in line, and depart the stop-controlled intersection. Table 4.2 summarizes the relationship between delay and LOS for unsignalized intersections.

Table 4.2 – Level of Service Definitions for Unsignalized Intersections

Level of Service Description Average Stopped Delay Per

Vehicle (Seconds)

A Little or no delays < 10

B Short traffic delays 10.1 to 15.0

C Average traffic delays 15.1 to 25.0

D Long traffic delays 25.1 to 35.0

E Very long traffic delays 35.1 to 50.0

F Extreme traffic delays with intersection capacity exceeded > 50.0

Source: Highway Capacity Manual (HCM) operation methodology, 2000.

4.4 Segments Level of Service

Based on the average travel time per segment collected by the travel time sections in the Vissim model, it was possible to calculate the average travel speed of the segments and calculate the LOS. The HCM 2000 Urban Street Methodology calculates the LOS of the segments based on the average through-vehicle travel speed, and the urban street functional and design classifications. Once the roadway segment has been classified, HCM Exhibit 15-2 is used to calculate the LOS for the roadway segment.

Improvements to Veterans Drive (Route 30) Page 23 of 42

Table 4.3 – Urban Street LOS Criteria (HCM Exhibit 15-2)

4.5 Network-wide Statistics

These measures are used to evaluate differences between alternatives over the entire study area. These statistics include total travel time, average speed, total network delay, and average delay time per vehicle.

4.6 Link Speeds

While the LOS measure of effectiveness provides an indication of the operation at each of the intersections, link speeds provide an indication of the level of mobility between intersections. Link speeds are obtained from the VISSIM model and compared to the posted speed limit along selected segments of the roadway system.

These measures of effectiveness are used to evaluate both existing and future conditions.

Improvements to Veterans Drive (Route 30) Page 24 of 42

Figure 4.1 – VISSIM Model Travel Time Section

Beginning Travel Time

Section

End Travel Time

Section

Improvements to Veterans Drive (Route 30) Page 25 of 42

5. Analysis of Existing Conditions

The results of the calibrated model were used to evaluate traffic operations at each intersection, for each approach and for all of the roadway segments within the study area.

5.1 Intersection Level of Service

Table 5.1 shows the resulting levels of service at each intersection along Veterans Drive.

Despite the level of congestion experienced in the eastbound direction, the level of service of the intersections along Veterans Drive is usually below saturated conditions.

The reason for this occurrence is that the level of service of an intersection is calculated based on the weighted average delay for all of the approaches. Since the westbound direction carries an important number of vehicles during both the AM and PM peak hours and drivers experience short delays on this direction, the average delay of the intersections as a whole are not critical as the delay in one particular direction.

Table 5.1 – Existing Levels of Service for Intersections along Veterans Drive

Intersection LOS (AM Peak Hour)

LOS

(PM Peak Hour)

Veterans Drive and Kronprindsens Tvaer Gade C B

Veterans Drive and Nye Tvaer Gade C C

Veterans Drive and General Gade B B

Veterans Drive and Strand Gade D E

Veterans Drive and Guttets Gade C C

Veterans Drive and Store Tvaer Gade D E

Veterans Drive and Raadets Gade D C

Veterans Drive and Tolbod Gade C B

As it is shown in Tables 1 and 2 in Appendix F, the level of service of the westbound approach for all the intersections along Veterans Drive during the AM and PM peak hours have minimum delays, which results in a LOS A. The level of service of the eastbound direction, on the other hand, ranges from LOS D and E as a result of higher delays caused for the most part by the lane reduction on the eastbound direction at the intersection of Tolbod Gade. The seemingly low delay on the eastbound direction, despite of the congestion experienced by drivers, results from the fact that the intersections are relatively close to each other and some of the approaches do not have sufficient capacity to store a large number of vehicles that would experience excessive delay. From the field observations, and as simulated by the model, when eastbound vehicles are given the green time, most of the vehicles clear the intersections.

*Note: Intersection Level of Service is calculated as a weighted average of the approaches delay

Improvements to Veterans Drive (Route 30) Page 26 of 42

Additionally, the congestion experienced by drivers does not occur throughout the peak hour. As it is shown in the simulation, at the beginning of the peak hour, vehicles are able to travel at relatively high speeds and experiencing low delays even in the eastbound direction.

Overall, the AM and PM peak hours show similar operational conditions. This is primarily due to the similarity between the traffic counts collected in the field for both periods and the fact that the bottleneck formed at the intersection of Tolbod Gade, produces congestion along the eastbound direction in the morning as well as in the afternoon.

5.2 Segment Level of Service

Another important measure of effectiveness is to verify the average through-vehicle travel speed in order to determine the level of service of the segments as it is discussed in the Highway Capacity Manual. Even though the intersection delay might not be remarkably high, as it was explained in the previous section, drivers experience heavy congestion along the eastbound direction of the corridor under study. Tables 5.2 and

5.3 show that the average speeds along the eastbound direction are in fact very low, and that most of the segments are experiencing over saturated conditions (LOS F) during both the AM and the PM peak hours.

Table 5.2 – Existing Levels of Service for Segments along Veterans Drive – AM Peak Hour- Eastbound Direction

Roadway From To Speed (mph)

LOS

Veterans Drive

West of Kronprindsens Kronprindsens 8.1 E

Kronprindsens NyeTvaer Gade 5.8 F

Nye Tvaer Gade General Gade 5.8 F

General Gade Strand Gade 5.2 F

Strand Gade Guttets Gade 3.9 F

Guttets Gade Store Tvaer Gade 3.8 F

Store Tvaer Gade Raadets Gade 3.4 F

Raadets Gade Tolbod Gade 3.3 F

Improvements to Veterans Drive (Route 30) Page 27 of 42

Table 5.3 – Existing Levels of Service for Segments along Veterans Drive – PM Peak Hour- Eastbound Direction

Roadway From To Speed (mph)

LOS

Veterans Drive

West of Kronprindsens Kronprindsens 11.9 D

Kronprindsens Nye Tvaer Gade 7.4 E

NyeTvaer Gade General Gade 7.4 E

General Gade Strand Gade 5.3 F

Strand Gade Guttets Gade 3.3 F

Guttets Gade Store Tvaer Gade 2.8 F

Store Tvaer Gade Raadets Gade 2.3 F

Raadets Gade Tolbod Gade 2.4 F

As it is shown in the previous tables the lowest average speeds are reached between Raadets Gade and Tolbod Gade right before the lane reduction at Tolbod Gade.

Segments located in the west side of the corridor reach higher average speeds because the queue formed at Tolbod Gade grows as a function of time along the eastbound direction and has a shorter influence on the intersections located on that side of the network. In fact, at the beginning of the AM and PM peak hours the queue does not reach intersections on the west side of the corridor and vehicles are able to travel at normal speeds.

6. Analysis of Future Conditions

The analysis of future conditions within the study area was performed using the calibrated model developed for existing conditions as a base. The following sections describe the improvements evaluated, the methodology used to develop the future traffic volumes as well as the signal timings used for analysis of future conditions.

6.1 Improvements Analyzed

Two scenarios were analyzed for future conditions: No-Build and one Build Alternative.

No-Build Alternative – This alternative is based on the existing geometry assuming no other improvements would be made through the year 2025. The No-Build Alternative is used to identify locations where congestion is expected to significantly increase and as a base for evaluating the improvements proposed as part of the Veterans Drive Improvement project.

Improvements to Veterans Drive (Route 30) Page 28 of 42

Build Alternative – This alternative includes the proposed improvements under the Veterans Drive Improvement project. The improvements consist of the following:

• Veterans Drive east of Tolbod Gade was widened to 4 lanes.

• An exclusive right turn lane was added for vehicles traveling in the westbound direction at the following intersections:

Veterans Drive and Kronprindsens Tvaer Gade

Veterans Drive and General Gade

Veterans Drive and Strand Gade

6.2 Future Traffic Volumes

The first step in analyzing future conditions was to determine a growth rate that would represent future conditions in the study area by the project’s design year of 2025.

Various sources were identified where indication of future growth rates could be obtained from. These sources included previous traffic counts on Veterans Drive, vehicle registration history, and Census data. Based on available data, the Census data was selected as the tool that captured growth over a ten-year period and could be applied to existing counts to estimate traffic growth for the next 17 years. Comparison of population figures between 1990 and 2000 obtained from the Census data, resulted in an average growth rate of two percent per year (See Table 6.1). As an additional check for annual growth, the traffic counts collected were compared to counts collected for a previous study performed by Kimball Chase (Veterans Drive Data Collection Report) for this same area, in 1994. As seen from Table 6.2, the location compared for daily counts shows an annual growth of 2.48 percent.

Table 6.1 – Census Comparison

Characteristic

Census

Census

Annual Growth Rate

Population 48,166 51,181 0.70% Housing Units 18,433 24,030 3.37% Workers 22,058 23,721 0.84% Car, Truck, or Van 15,781 17,720 1.37% Public Transportation 2,645 3,202 2.34%

None 4,329 4,968 1.64% 1 6,899 8,913 3.24% 2+ 4,417 5,577 2.92%

2.05%

Household Vehicles

AVERAGE=

Improvements to Veterans Drive (Route 30) Page 29 of 42

Table 6.2 – Traffic Count Comparison

Location

Kimbal Chase

Report

Collected Counts

Annual Growth Rate

Veterans Drive ADT 22,000 28,000 2.48%

Veterans Drive west of Tolbod Gade (AM) 2,364 2,104 -1.00% Veterans Drive west of Tolbod Gade (PM) 2,229 1,565 -2.71% Veterans Drive west of Hospital Gade (AM) 1,460 1,523 0.39% Veterans Drive west of Hospital Gade (PM) 1,412 1,311 -0.65%

-0.41%AVERAGE=

Daily Counts

Peak Hour Counts

The growth rate was applied to existing volumes including turning movements at the intersections. Because the commercial district of Charlotte Amalie is a well established built-out area and major changes in land-use are not expected, traffic distribution at the intersections was assumed to remain as it is today.

2025 AM and PM peak hour volumes were developed for the No-Build as well as for the build alternative. While the No-Build Alternative volumes were developed by applying the growth factor to existing conditions, traffic volumes needed to be re-assigned for the Build Alternative since vehicles will be able to make a right turn at the intersection of Veterans Drive and Strand Gade.

Future volumes are shown on Figure 3 through Figure 6 in Appendix C for AM and PM peak conditions for No-Build and the Build Alternative, respectively.

6.3 Future Signal Timing and Phasing

Once the projected volumes were calculated for the AM and PM peak hours the next step was to modify the existing traffic signal timing and phasing to reflect the future traffic demand and changes in geometry. The new signal timing and phasing were developed using the Synchro software. This software allows for signal optimization based on traffic volumes at intersections.

Since vehicles in Saint Thomas travel on the left side of the road instead of the right side, the default setup in Synchro, some adjustments were made in order to model the traffic conditions at the intersections.

The most important consideration when simulating left-side traffic in Synchro is that even though a conversion to right-side traffic is needed, the operational characteristics of the conflicting movements must not be affected. To that end, right turns in Saint Thomas become left turns in the Synchro model. Figure 6.6 shows how the phasing configuration was converted from the left-side traffic to the right-side traffic for one of the intersections.

Improvements to Veterans Drive (Route 30) Page 30 of 42

Figure 6.1 – Turning Movement Conversion in Synchro

This conversion ensures that the delay calculated in both scenarios is not affected given that the number of vehicles in the conflicting movement, which is the right turn for Saint Thomas and the left turn for right-hand driving, does not change. Using this methodology, the cycle length and the splits for all of the pre-timed controllers were optimized. The optimization was performed according to the time of the day (AM or PM peak hour) and the geometric configuration based on the alternative being evaluated, No-Build or the Build Alternative. The resulting optimized signal timing and phasing at each intersection were coded into the VISSIM model for analysis of future traffic operations in the study area. A total of four scenarios were analyzed with future volumes: No Build AM and PM, Build AM and PM.

6.4 Future Intersection Level of Service Analysis

Similar to the analysis of existing conditions, the VISSIM model was used to analyze the level of service for each of the intersections within the study area for the Build Alternative. Existing levels of service and delays for signalized intersections were included in the following table for comparison purposes. Delays shown are in seconds per vehicle. Tables 6.3 and 6.4 show the average intersection delay and level of service for the AM and PM peak periods, respectively.

Table 6.3 – Delay and Level of Service Summary – AM Peak Period

Intersection

Existing Conditions

No-Build

Build Alternative

Veterans Drive and Kronprindsens Gade 21/C 67/E 3/A

Veterans Drive and Nye Tvaer Gade 23/C 44/D 10/B

Veterans Drive and General Gade 17/B 26/C 11/B

Veterans Drive and Guttets Gade 26/C 28/C 8/A

Veterans Drive and Raadets Gade 39/D 36/D 8/A

Veterans Drive and Tolbod Gade 33/C 31/C 10/A

Actual Movements Synchro Model

Improvements to Veterans Drive (Route 30) Page 31 of 42

Table 6.4 – Delay and Level of Service Summary – PM Peak Period

Intersection

Existing Conditions

No-Build

Build Alternative

Veterans Drive and Kronprindsens Gade 10/B 86/F 1/A

Veterans Drive and Nye Tvaer Gade 24/C 61/E 8/A

Veterans Drive and General Gade 18/B 32/C 8/A

Veterans Drive and Guttets Gade 21/C 32/C 10/B

Veterans Drive and Raadets Gade 34/C 46/D 5/A

Tolbod Gade 18/B 28/C 7/A

No-Build Alternative – Based on the analysis, traffic conditions will worsen during both AM and PM peak periods for the No Build condition in 2025. As it was anticipated, the lane reduction at Tolbod Gade creates a critical condition for the eastbound direction and therefore the average delay and level of service worsen for most of the intersections when future volumes are introduced into the network. The operational conditions of the intersections could be even worse but since the westbound approach for all the intersections is still experiencing low delays, the overall weighted average delay does not have a dramatic increase.

As it is shown in Tables 3 and 4 in Appendix F the westbound approach for all the intersections during both the AM and the PM peak hours is operating at LOS A.

Conversely, the operational conditions for the eastbound approach for all the intersections within the corridor are close or over capacity. The delay and level of service could be even higher along the eastbound direction but as it is shown in these two tables, the eastbound approach at the intersection of Veterans Drive and Kronprindsens Tvaer Gade is not capable of processing the projected demand (the processed volumes are 42% and 48% short of the projected volumes during the AM and PM peak hours) and since this is one of the vehicle input points, traffic volumes are not able to be loaded into the network in that particular direction. This means that the inability of the traffic model to process volumes on the eastbound direction is the clearest indication that the current geometric conditions would be unable to accommodate future traffic volumes.

Build Alternative – The results shown in Tables 6.3 and 6.4 indicate that the improvements planned for the year 2025 under the build condition will have an important effect on the delay and the level of service of the intersections along the corridor under study. The weighted average delays of the signalized intersections for both the AM and PM peak hours are low and the level of service range between LOS A and B.

Tables 5 and 6 in Appendix F show that the delays obtained from the Vissim traffic simulation Model accurately reveal the conditions that are likely to occur in the future

Improvements to Veterans Drive (Route 30) Page 32 of 42 given that the percent difference between the volumes processed by the model and the projected volumes is low and therefore the model is calibrated.

6.5 Future Segment Level of Service Analysis

Tables 6.5 and 6.6 show the slow speeds and poor conditions (LOS F) experienced by drivers along the eastbound direction of Veterans Drive for both the AM and the PM peak hours. As it was the case for existing conditions, the bottleneck caused by the lane reduction at the intersection of Tolbod Gade creates a queue along the eastbound direction that spills back to the intersection of Kronprindsens Tvaer Gade for the No Build scenario. Since the queue is generated at Tolbod Gade, the segments close to this intersection experience congestion for most of the peak hour and therefore have lower speeds. Given that it takes some time for the queue to build up and reach the intersections on the west side of the corridor, the average speed of the segments located close to the intersection of Kronprindsens Tvaer Gade are not as low as the segments located downstream.

Table 6.5 Segment Average Speed and Level of Service - Eastbound Direction (AM Peak Hour)

From To

Existing AM No Build AM Build AM

Speed (mph) LOS Speed

(mph) LOS Speed (mph) LOS

West of Kronprindsens Kronprindsens 8.1 E 2.9 F 17.0 C

Kronprindsens Nye Tvaer Gade 5.8 F 3.3 F 10.7 D

Nye Tvaer Gade General Gade 5.8 F 3.7 F 11.0 D

General Gade Strand Gade 5.2 F 3.2 F 17.7 C

Strand Gade Guttets Gade 3.9 F 3.3 F 11.0 D

Guttets Gade Store Tvaer Gade 3.8 F 3.3 F 18.8 C

Store Tvaer Gade Raadets Gade 3.4 F 3.2 F 13.6 C

Raadets Gade Tolbod Gade 3.3 F 3.1 F 8.6 E

Improvements to Veterans Drive (Route 30) Page 33 of 42

Table 6.6 Segment Average Speed and Level of Service - Eastbound Direction (PM Peak Hour)

From To

Existing PM No Build PM Build PM

Speed (mph) LOS Speed

(mph) LOS Speed (mph) LOS

West of Kronprindsens Kronprindsens 11.9 D 2.3 F 18.5 C

Kronprindsens Nye Tvaer Gade 7.4 E 2.5 F 12.7 D

Nye Tvaer Gade General Gade 7.4 E 3.0 F 11.1 D

General Gade Strand Gade 5.3 F 2.3 F 18.1 C

Strand Gade Guttets Gade 3.3 F 2.3 F 8.1 E

Guttets Gade Store Tvaer Gade 2.8 F 2.1 F 19.5 B

Store Tvaer Gade Raadets Gade 2.3 F 2.3 F 17.5 C

Raadets Gade Tolbod Gade 2.4 F 2.3 F 10.3 D

Once the additional lane is implemented east of the intersection of Tolbod Gade and the bottleneck is dissipated, the speeds of the segments for the Build scenario are considerably faster than the No build and existing conditions.