Modelling toll impacts using congested highway assignment software – a case study of the proposed N2 - R72 link road in east London, South Africa

• Main Author: Duze, Fezekile
• Other Authors: Zuidgeest, Marcus
• Format: Dissertation
• Language: English
• Published: Faculty of Engineering and the Built Environment 2021
• Subjects: civil engineering
• Online Access: http://hdl.handle.net/11427/33657

Traffic congestion has become a growing burden on society. Various approaches to model transportation do not scale efficiently due to its complex nature. The estimate of the locality and quantity of traffic that passes through, leaves or enters a study area are arrived at by assumption. There could be an increase in congestion if investments in infrastructure are done based on inaccurate transport modelling. Furthermore, outside the study area, the effects of changes in infrastructure are unknown. Hence, it is necessary to model transportation networks at a larger scale than previously required. Governments all over the world are faced with continual challenges of providing infrastructures with basic amenities to their citizens. In every financial year, government budgets are always overstretched, which leaves very little for infrastructure maintenance, especially in less developed countries such as South Africa. Tolling road is one option to overcome this challenge. This thesis, therefore, studies toll road modelling options for the proposed N2-R72 link road in East London, South Africa. The proposed N2/R72 Link Road is approximately eight kilometres long and includes a longspan bridge across the Buffalo River as well as a number of interchanges/intersections within an urban environment. Tolling in urban areas is not uncommon in South Africa and is typically associated with mainline plazas supported by several ramp plazas. The strategy presumably for the N2/R72 Link Road involves a mainline plaza constructed on the N2/R72 Link Road just before or after the Buffalo River Bridge. Traffic will be attracted from the existing roads depending on the benefits and toll on the proposed link. The research problem addressed in this thesis arose from the fact that TomTom Traffic Index in South Africa identified East London as the third most congested city in South Africa. The morning and evening commuters experience a massive increase in travel times. Globally, it has been shown that congestion level does not necessarily get reduced as a result of an increase in road capacity. At the same time, there is a need for a possible of diversion to the proposed link road, which can be a choice of a road user's willingness to pay toll fees. The challenge that arises is that diversions can result in gains and losses in terms of revenue for toll agencies if travellers' chose alternative, lower class, existing routes. Now, based on the above the Buffalo City Metropolitan Municipality (BCMM) approached SANRAL to construct a new link with a bridge. In order to finance that (with the benefit of managing traffic as well) the idea is to toll the new link. Knowing that the road user preferences are different. especially there is a need to understand the impact of tolling on traffic flows in the area. The main research question is What the impact of tolling is on the use of the upgraded link knowing that there is heterogeneity amongst users in terms of sensitivity to user costs for transport? The effect of tolling on the proposed N2-R72 link roads is being tested using the congested highway assignment software called SATURN. The way the problem is although preferably VISSUM as a ‘proper' micro model in that they model on a real-time, behavioural level. EMME models on a macro level and could have been used to accommodate all four steps in conventional models. It is tricky to use SATURN on a meso and micro level, as the micro level can distort the trip assignment if only a small (but key) part of the network is simulated (i.e. the assignment results in trips diverting around the simulation by using the buffer network). SATURN as employed in this study can operate at the meso level (buffer networks) with the junction simulation component operating at the micro level (somewhat). This report cautions against confusing the concepts of micro, meso and macro traffic models. In order to understand how the transport network may react to the proposed changes, a modelling approach is proposed. To achieve this, a traffic model was developed to represent the existing situation. This model provides the benchmark against which any proposal will be compared. This study will compare results between the existing and proposed situation in order for an informed decision can be taken on whether to proceed with the proposal based on the impact it will have on the existing network. In transport planning, various transport models are used to forecast impacts and evaluate options. This study investigates and reports on the impact of tolling should SANRAL construct a new road linking two national roads knowing that there is heterogeneity amongst users in terms of sensitivity to user costs for transport. To undertake this study a congested highway assignment model will be used based on a known case study and available old model. The literature review illustrates that with a growth of Public-Private Partnership (PPP) projects, toll roads would increase incessantly. Route choice by users is greatly influenced by toll and can sometimes also have effects on trip departure time and choice of mode. To model toll roads, users' willingness to pay (WTP) or Value of Time (VOT) has an important role, and generally, worker's wages is considered to be equivalent to VOT. This study also acknowledges that there is a distinguished difference between urban toll schemes and congestion pricing. The GFIP e-toll scheme is not congestion pricing, but rather a way to generate revenues for road upgrading and network expansion. It is not meant to suppress trip demand; in fact, it has the opposite effect when upgrades are made, and the network is expanded. The “user-pay” principle is viewed as a traffic demand measure, not as a means by which to raise funds for road building. The obvious questions are: What is the “user-pays” principle? How is the amount that the user must pay determined? Congestion pricing is meant to reduce congestion by suppressing demand. The objectives of congestion pricing are to reduce congestion, to reduce the environmental impact of vehicles by reducing harmful exhaust emission, and to improve the space for public and Non-Motorised Transport (NMT) modes. Revenue from congestion pricing should be used for public transport provision and not upgrading and/or building new roads. The results of the model analysis show that, since traffic growth on toll roads is increasing somewhere in the range of 2% and 6%, the impact on income cannot be negative. There is, notwithstanding, a risk of heightening maintenance charges if heavy vehicles continue to increase at an expense twice that of light vehicles. While not a single verification utilized in sight in the literature that toll roads in South Africa are looking to alleviate clog, it very well may be presumed that growth of light vehicle (Class 1) exacerbate congestion. It is the research candidate's view that research on the perception of Value of Time (VOT) by road-user needs to be found and on toll diversion, being able to classify and predict future traffic and volumes will be essential for the national roads agency and other provincial road authorities on roads they are in control for.