Transportation Research Board

1
The Secrets to HCM Consistency Using Simulation
Models
Given by Ronald T. Milam, AICP Contributing
Analysts David Stanek, PE Chris Breiland Mike
Iswalt

• Transportation Research Board
• Planning Applications Conference, May 2007

2
Typical Simulation Examples
3
When to Use Simulation
4
When to Use Simulation
5
When vs. How?

• The HCM methodologies may not be appropriate in
  environments with multiple travel modes or
  multiple hours of congestion.
• the HCM methods are generally not
  appropriatefor the evaluation of queues that are
  building over both time and space. Page 9-1,
  HCM 2000
• Certain freeway traffic conditions cannot
  easily be analyzed by the methodology. Multiple
  overlapping bottlenecks are an example.
  Therefore, other tools may be more appropriate
  - Page 22-1, HCM 2000

6
Challenges/Issues

• Multiple software packages
• AIMSUN
• CORSIM
• Dynasim
• Paramics
• SimTraffic
• TransModeler
• VISSIM
• others
• Which ones are HCM consistent?
• How can a user or reviewer be sure?

7
Simulation Performance Measures

• Traffic simulation programs provide many
  performance measures
• Total delay
• Density
• Travel time
• Average speed
• Travel distance
• Number of stops
• However, the definition of the performance
  measures may differ from the HCM definition.

• Stopped delay
• Throughput
• Emissions
• Fuel consumption
• Queue length
• Others

8
HCM Performance Measures

• HCM methods determine LOS for facility types
  based on different performance measures

• Facility Type
• Intersections
• Urban Streets
• Two-Lane Highways
• Multi-lane Highways
• Freeways

• Performance Measure
• Control Delay
• Speed
• Percent Time Spent Following
• Density
• Density

9
Consistency Issues

• Roadway System
• Peak hour factor
• Intersections
• Defining the intersection approach
• Freeways
• Ramp influence areas

10
Peak Hour Factor

• HCM defines the LOS for the peak 15 minutes of
  the peak hour

11
Peak Hour Factor, continued

• Edit the periods file to create 15-minute
  interval demand periods.

Paramics Example
12
Peak Hour Factor, continued
Use the divisor parameter in the demands file to
adjust peak hour volumes. Hourly data with PHF
of 0.95 implies a divisor of 3.8.
Paramics Example

13
Intersection Approach Delay

• Include delay for all vehicles on approach
• Model link delay may not include entire
  intersection approach

Model Network
14
Intersection Approach Delay, continued

• Upstream unsignalized intersection divides the
  southbound approach into two links.

SimTraffic Example
15
Intersection Approach Delay, continued

• Upstream unsignalized intersection divides the
  southbound approach into two links.

Unsignalized Driveway Intersection
Link 2
Signalized Intersection
Link 1
SimTraffic Example
16
Intersection Approach Delay, continued

• Output report for signalized intersection shows
  delay for Link 1 only.

SimTraffic Example
17
Intersection Approach Delay, continued

• Adding delay for Links 1 and 2 to get the entire
  southbound approach delay.

SimTraffic Example
LOS Results Link 1 Links 1 2 SB Approach E
/ 59 F / 86 Intersection D / 40 D / 44
18
Ramp Influence Area

• HCM defines the ramp merge analysis area as the
  two right shoulder lanes plus the auxiliary lane
  for 1,500 feet downstream of the ramp gore
• Ramp influence area does not necessarily match
  model links

19
Ramp Influence Area, continued

• Build model network to account for ramp influence
  area and volume collection locations.

VISSIM Example
20
Ramp Influence Area, continued

• Ramp merge area with acceleration lane

Link 191
Link 190
VISSIM Example
On-Ramp
21
Ramp Influence Area, continued

• Collect statistics for ramp influence area.

VISSIM Example
22
Ramp Influence Area, continued

• Collect statistics by lane for each link.

VISSIM Example
23
Ramp Influence Area, continued

• Calculate volume-weighted average density first
  across links, then across lanes

VISSIM Example
24
Ramp Influence Area, continued

• Analysis Results

Location Density (vpmpl) LOS Link 190 (all
lanes) 42.4 E Link 191 (all lanes) 37.2 D Ram
p Influence Area 42.1 E
Link 191
Link 190
VISSIM Example
25
Recommendations

• FHWA - Modify NGSIM priorities to better reflect
  user needs
• TRB - Update the HCM to provide specific
  guidelines for simulation software vendors and to
  reflect a wider range of performance measures
• Software Vendors Provide technical support and
  training on HOW to perform simulation modeling
• Users Invest time to understand HOW to perform
  simulation modeling

26

• Questions?

The Secrets to HCM Consistency Using Simulation
Models