TfL Traffic Modelling Guidelines – Version 4.0
In November 2021, Transport for London (TfL) presented their latest version (Version 4) of their Traffic Modelling Guidelines at a virtual launch event.
Whilst those who are regularly working for TfL on modelling projects will be familiar with the updated guidelines, we thought we would produce a short post that details the main changes since Version 3. We will also highlight the key changes for those using LINSIG & VISSIM, before reviewing the new chapters on Cyclist and Pedestrian modelling for London projects.
The latest guidelines can be found at the following link: https://tfl.gov.uk/corporate/publications-and-reports/transport-modelling-guidance
Updates from Version 3
Version 3 of the guidelines was originally released in 2011 and over time, policies, strategies and software has evolved. Version 4 seeks to bring the guidelines up to date with the following revisions:
The most recent political, physical and demand-led changes in London
The Mayor’s latest Transport Strategy – which focuses more on sustainable means (active travel & public transport)
A consistent modelling methodology – Base, Future Base and Do-Something scenarios.
Requirement to use TfL’s ONE Model for obtaining Future Base & Do-Something scenario flows.
Business Case Development manual & TUBA analysis guide.
Updates in Version 4
The updates in Version 4 are briefly summarised as follows:
Part A – Modelling Considerations
o Updated to include the latest policies
o Updated to capture new developments in modelling and software
Part B – Modelling Guidelines
o Updated modelling best practice
o Covers new software versions
o Reflects new data sources
Part C – Healthy Street Approach
o New section that advises on emissions, pedestrian and cyclist modelling.
o Interactions with microsimulation and best practice
Key Changes for LINSIG
The most notable changes within the LINSIG modelling section (Section 4) are as follows:
General re-working of the section, with additional detail on the various modelling elements and following a more logical order – Introduction, Preparation & Graphical User Interface (GUI), Model Development, Base Calibration, Base Model Validation, Proposed Modelling & Outputs
Additional guidance on flow allocations (e.g., lane based or O-D matrices) and how to incorporate public transport (Section 4.4.6).
Matrix estimation is recommended not to be used for creating O-D matrices and if it is to be used, should be agreed with TfL in advance (Section 4.4.6.3).
New section on modelling pedestrians and cyclists in LINSIG (4.4.8 & 4.4.9).
More defined model validation section, with DoS validation now also including 10% criteria for lanes upstream of pedestrian crossings (Section 4.5.5).
New section on how to appropriately develop Proposed Models (4.6), which includes details of layout changes, scenarios, flow changes, signal time optimisation and outputs.
New section on LINSIG model outputs (4.7), which covers network results, cyclic flow profile, uniform queue & storage graphs and travel time/delay matrix view. The Report Builder and Model Audit View outputs are also considered.
Part C – Section 2 – Cyclist Modelling
This new section contains a range of information and guidance on how to model cyclists in an acceptable manner for TfL projects.
The section includes:
Details of a case study for the Cycle Superhighway route, which was used to highlight the previous and updated process for including cyclists in models.
Information on TfL research on cyclists, which include looking at interactions at Advanced Cycle Stoplines (ASLs), at segregated stoplines, dispersion along a network of junctions and the impacts of cyclists on saturation flows.
Guidance on the suggested model parameters, data collection and how to model the flows in the different modelling software packages is included.
A dedicated chapter for modelling cyclists in microsimulation software, covering the different behaviours, link types, compositions, speed & acceleration and saturation flow considerations.
Guidance on how to validate the cyclists in the models, including flow checks against surveyed counts and travel time analysis against GPS data.
Part C – Section 4 – Emissions Modelling
The final new section contains details on the methodology and requirements for collecting emissions data from microsimulation models.
The section provides a case study in Putney High Street, which was a pilot for identifying an appropriate modelling methodology for collecting vehicle emissions outputs. This helped to develop an emissions modelling methodology that provided details on the following:
The appropriate use of microsimulation software (either VISSIM or AIMSUN).
Considerations with regards to the model boundaries, road gradients, vehicle types and vehicle dynamics.
The model calibration and validation requirements, model outputs and model runs.
The choice of emissions modelling software, which also includes the fuel types, engine standards, running of the software and the output reporting.
The most important comment is to determine if emissions outputs are required at the very early stages (during scoping discussions), as this then has an impact on the data collection and model development tasks.
The section then goes into detail on two emissions modelling software programs – EnviVer (Environmental Vissim-VERSIT+ simulations) and PHEM (Passenger car and Heavy duty Emission Model). Details on the inputs, vehicle configuration and outputs are provided for these two specialist software programs.
Key Changes for VISSIM
The most notable changes within the VISSIM modelling section (Section 7) are as follows:
General re-working of section, with additional detail on the various modelling elements and following a more logical order – Introduction, Preparation & GUI, Model Development, Base Calibration, Base Model Validation, Proposed Modelling & Outputs
Acknowledgment of new versions and introduction of scenario management in Version 6 onwards (7.1.2)
GUI section (7.3) provides guidance on using new features e.g., lists & scenario management
If using scenario management, it should be noted that TfL require each scenario to be submitted as a single *inpx file for auditing purposes (Section 7.3.4).
For the saturation flow validation (Section 7.4.2.4), Figure 1.27 shows a useful image demonstrating how to set up Reduced Speed Areas at signalised stop-lines.
The Priority Rules section (7.4.2.5.1) includes useful guidance of how to set-up gap time and headway priority rules.
Section 7.4.2.6 provides guidance on how to create ‘Externally Generated Congestion’, to account for queuing that occurs outside of the agreed model extents.
New sections for Cyclists/Motorcyclists (7.4.3.5) and Pedestrians (7.4.3.6) are included, which detail how to appropriately model these in VISSIM.
For calibration and validation, VISSIM models need to be run for 20 different random seed runs to reduce any potential bias in the seed runs (Section 7.5.1).
The Proposed Model Development section (7.6) includes much more detail on the use of scenario management and how to make network changes effectively.
New section on VISSIM model outputs (7.7), which includes guidance on extracting both numerical and visual outputs.
Part C – Section 3 – Pedestrian Modelling
This new section contains a comprehensive range of information and guidance on how to model pedestrians in an acceptable manner for TfL projects.
This includes:
Relevant Guidance documents
Modelling Approach – covering static/dynamic modelling, pedestrian comfort levels & pedestrian crossing behaviour.
Preparation – including data collection (surveys, flows, public transport, signal timings), pedestrian types & compositions.
Model Building & Development – including routing, different levels, interactions with public transport, pedestrian crossings, signalised junctions, zebra crossings and validation (which should be agreed with TfL prior to project commencement).
Model Outputs & Analysis – providing details on they key outputs required – pedestrian comfort levels, Fruin’s level of service, cumulative high density, space utilisation, journey times, social costs & screenshots/videos.
Summary
We hope this post has helped to provide some insight into the updated TfL Guidelines and provide a brief overview of the main changes for LINSIG & VISSIM modelling, as well as the new chapters for Cyclist and Pedestrian Modelling. Hopefully there are some elements you can take away and apply to your own models.
Thanks for reading!