Congestion and concerns over environmental protection point to better use of underground space as a natural choice in future developments. This offers opportunities for the underground construction industry, and new infrastructure might not necessarily mirror the forms constructed by previous generations. Recognising this direction, the Highways Agency is exploring the concept of “limited facility tunnels” (LFTs) or “minitunnels”, and “limited facility bridges”. It commissioned a study to evaluate the concept and to develop planning and design guidance with a view to publishing a design standard.

The project spanned disciplines including structural design and maintenance for cut-and-cover, bored tunnel and bridge options, and a range of highway transport engineering skills to cover matters such as highway geometry, traffic flow and signing.

Input was sought from emergency services, UK tunnel and bridge operators, French LFT operators and technical specialists from Transport Research Laboratory (TRL) and the UK’s Department of the Environment, Transport, and the Regions (DETR). However, because the structures (and traffic systems) are not in common use in the UK, there was the problem of predicting future performance, which is safety critical and dependent to some degree on the behaviour of the travelling public.

The high cost of building full-scale prototypes can be addressed in part through computer-based simulation techniques, enabling the designer to “see” what has been designed before construction. Desk-based systems can provide a valuable resource for testing the function of control systems and allow navigation of proposed structures to identify geometric conflicts.

More sophisticated virtual reality systems offer greater scope to evaluate and develop possible designs, enabling the engineer to “construct” the proposed works quickly and at low cost, and allowing the end users to test them in a virtual world. Such technology can thus yield insights into the likely performance of unfamiliar infrastructure and systems without the safety risks associated with full prototype construction. Additionally, the relatively low costs mean that a large number of alternatives may be tested and potentially hazardous situations incorporated for experimental purposes.

The minitunnel concept was subject to virtual reality technology as part of the Highways Agency’s study. The term limited facility tunnel describes a highway tunnel or underpass which is limited to use by certain types of vehicle. The typical restriction is a height limit, usually of 2.6m. Such structures are relatively unusual in the UK and sometimes originate from changes of use of old tunnels. Examples include the former tram tunnel under the Aldwych in London and a former rail tunnel in Caernarfon.

In France, however, purpose-built LFTs are well established for bypassing congested urban junctions. There, LFTs are generally described as minitunnels. Although this carries other associations within the UK tunnelling industry, the two terms are used synonymously in this discussion.

Potential applications

Highway minitunnels fundamentally provide a means of distributing road space between vehicle types, for a number of ends. At existing junctions, for example:

– to improve operational efficiency by smoothing traffic flow and maximising the use of space. This is the purpose of the minitunnels on suburban commuter routes in Lille and in central Paris.

– offer safety and efficiency benefits at junctions by segregating conflicting traffic streams where a full grade separated junction might not be viable because of space or cost.

– provide a means of segregating road user types on a route to maximise the “high quality” (generally at-grade) space for high priority road users. For example, diverting cars into a tunnel in an urban area might allow an increased provision of bus and cycle lanes at the surface.

Other possible applications for limited facilities are at transport nodes where travellers leave their cars and transfer to public transport. In areas of high land value or environmental sensitivity, interchange parking facilities are likely to use underground space and will have limited headroom to suit the permitted vehicle categories.

The concept may also be applied to new corridors. For example, the new A86 highway to the west of Paris (Figure 1) is being developed to include a 10km section of cars-only tunnel. This is arranged as two decks, each carrying three lanes of traffic within a single large bore. Large vehicles are carried on a separate route with reduced gradients.

In the UK, outline proposals have been developed in the past for schemes such as a second Channel Tunnel for cars only. While the scale of these major projects is quite different, many of the fundamental design and operation questions remain the same.

Barriers to use of minitunnels

A number of significant possible disadvantages with light vehicle-only tunnels were considered in the study. Many of the concerns relate to safety. All highway tunnels are at risk from fire and associated problems of smoke control. Current ventilation and lighting specifications for highway tunnels are based on typical full-size tunnel profiles and so may not be applicable to tunnels of smaller cross section. Furthermore, a small tunnel offers little working space around disabled vehicles and is more likely to be blocked after even a simple breakdown.

Another set of issues relates to traffic engineering. In particular, the need for effective systems for detecting and diverting over-height vehicles from entering limited clearance areas. This raises questions about signing, automatic detection systems, escape lanes, height restriction barriers and the associated law. The matter is complicated because available traffic signs in the UK do not include provision for signing a route as “cars only”.

Finally, it is known that a proportion of motorists suffer from some degree of “tunnel phobia” which might prove a more acute problem in smaller diameter tunnels.

Constructing and testing virtual tunnels

While likely construction methods were fairly readily identified, it was felt that questions of signing and likely driver behaviour issues could only really be answered by developing some form of prototype structure for experimentation.

By using a simulator it was possible to construct

a range of prototypes at the same time to test different ideas. The driving simulator was used to give the best possible impression of what these structures would be like for the end user and so to test likely driver behaviour. It also provided a means of demonstrating the concept to possible developers.

The TRL Driving Simulator used for the LFT trials features a real medium-sized saloon car (a Rover 414Sli). The car is mounted on hydraulic rams, which supply motion to simulate the tilt and roll experienced in normal braking, acceleration and cornering. Three screens surround the car, providing a 210° front/side image, and one rear screen providing normal rear vision using vehicle mirrors. The trial route images are projected onto the screens to give continuous, linked image between front, side and rear screens. A sound system provides engine and road noise and noise from passing traffic.

The simulation also includes other vehicles whose behaviour can be programmed to include driving styles ranging from passive through “normal” to aggressive. The trial is controlled by an external operator who can intervene to give instructions.

One of the primary concerns about the use of a driving simulator is the validity of applying the results to the real world. But a comparison against drivers’ performance of the same tasks in a real, instrumented vehicle confirmed that the simulation is fully representative of behavioural effects.

Experimental design

In the initial stages of the project a range of options for the construction method and general arrangement of limited facilities were identified. In addition, photographic and video records of existing limited facility routes in Britain and France helped provide the information from which the programmers developed the basic simulator model.

The virtual world created in the model consisted of a highway route with one full-size tunnel and a range of junctions incorporating optional limited facility routes. Most of these were geometrically compatible with the proposed design standard but one was constructed to a minimum specification as a single lane tunnel, with lower headroom and without space for emergency walkways.

A range of signing configurations, road markings and escape lane provisions were employed to allow comparison of driver response. The final form of the model is summarised in Table 1.

Around 40 randomly selected drivers took part in the trials. They drove for 20 to 30 minutes through the tunnel model. The participants were asked to drive normally and to follow the main road. They were advised that at complex junctions the tunnels generally represented the quickest route. The drivers were observed during the trial and were then asked to fill in a questionnaire about the journey.

Additionally, quantitative data for the vehicle was recorded in each run for later analysis. A few HGV drivers, local authority highway engineers and fire service personnel were also asked to drive in the model and provide comments.

Experimental observations

The majority of drivers found the simulated environment sufficiently convincing and were generally confident to comment on the tunnels and features they had only encountered in the virtual reality world. Drivers tended to be cautious on entering the minitunnels for the first time, with 50% reporting some feelings of anxiety when driving through the minimum facility tunnel, compared with a third for the unidirectional, one-lane LFT, and just over 20% for bi-directional, two-lane LFTs. In subsequent encounters with the LFTs the drivers became noticeably more relaxed.

The subjective observations were directly supported by measurable parameters such as speed and the mean distance of the car from the kerb and the consistency of the car’s lateral position, which all vary with the driver’s concentration and comfort. These observations give some indication of the likelihood of problems with tunnel phobia.

The simulations also provided evidence on the merits of central barriers in bi-directional LFTs. This was further supported by observation of driver behaviour in the two-lane bi-directional tunnel where the left lane was deliberately blocked with a broken down vehicle. Several drivers crossed double white lines to proceed on the other side. One of the drivers then drove out of the tunnel on the wrong carriageway into the path of oncoming traffic.

Just over 85% of subjects felt that the signing used in the trials was a helpful indicator of a limited facility, and a similar proportion felt that a green approach lane was helpful. However, the four HGV drivers all responded that markings painted on the road surface (which were used in some trial runs) would also be helpful. They also felt that escape lanes on the approach to LFTs should be better indicated.

Overall, 80% of the drivers who expressed a clear view felt that the concept of limited facility routes to bypass relatively complex junction areas was attractive. Among those who raised objections, there were concerns about cost of construction, consequences of an accident within the LFT, and stated preference for funds to be spent on public transport rather than highway construction.

Of the drivers who chose not to use the LFTs, over a third mentioned confusing signing, or uncertainty about lane choice. This high level may have been influenced by the image on the simulator screen being slightly out of focus, which tended to make details on signs unclear. This problem was partially overcome by upgraded software in later simulation runs. More than 20% of subjects claimed they chose not to use the tunnels simply to make the route more interesting.

Conclusions

The study has demonstrated the feasibility of the development of minitunnels within the highway network. At the same time it has shown that minitunnels may present a number of potential hazards, including the need for unusual road layouts and signing and the risk of adverse driver response.

It is intended that the design standard and associated guidance developed under the project will enable decisions to be made on appropriate location and design of limited facility grade separated junctions, giving due consideration to all the economic, environmental and safety issues.

It is also clear that limited facility roads may be viewed as a tool which could achieve a number of aims within an overall transport strategy. These aims may have as much to do with best practice in environmental management and transport integration as with traditional highway planning objectives such as reducing congestion.

The study demonstrated the value of undertaking feasibility stage design works for unfamiliar structures and systems within a virtual reality world. Although not a perfect representation of reality, the simulator proved a successful means of exploring the minitunnel concept and its applications.

Improving technology will make such simulations an increasingly important part of all future design for transport systems. Most particularly, it will have a role in design of underground works where planning and optimising the ambience of public access areas is important.

Related Files
Figure 1