Hyperloop is a concept first introduced by Elon Musk. It involves a pod for passenger and car transport that is lifted on air bearings and travels at speeds of up to 740mph (1,191km/h) through an evacuated tube at one thousandth of atmospheric pressure. Costing around US$20 per trip (excluding operating costs), it can allow San Francisco-Los Angeles travel in 35 minutes. The concept is an alternative, cheaper option to high-speed rail.
Hyperloop is considered an opensource transportation concept, allowing companies such as Virgin Hyperloop One and HyperloopTT to construct Hyperloop designs and test tracks. This includes a 500m-length test track in Nevada, constructed by Hyperloop One, a completed 320m ground-level test track, and a one kilometre track sited on pylons of 5.8m that is under construction by HyperloopTT, both in France.
Both companies have developed partnerships around the world, including Mid-Ohio Regional Planning Commission, Dubai’s Roads and Transport Authority, the Indian state of Maharashtra, and many others.
Musk’s SpaceX has also been involved with the concept, conducting annual pod design competitions to stimulate the technology and to promote the mode of transportation. In 2018, WARR Hyperloop, a team from The Technical University of Munich in Germany, reached a maximum speed of 290mph (466km/h) making it the fastest Hyperloop pod to date.
The Hyperloop concept still has a lot of research and development to conduct, only reaching approximately 40% of the maximum speed that SpaceX noted in its concept paper (760mph (1,220km/h)).
Most companies and teams have opted to use magnetic levitation technology, rather than air-bearing technology, due to economic viability and technology advancements in the field.
Physical Infrastructure
The basic elements of a Hyperloop system are:
- Twin parallel inbound and outbound transmission tubes capable of maintaining a near vacuum, dimensioned to suit the capsules or pods that will use them. The pods and tubes are dimensioned to suit the intended cargo.
- A system to create / maintain a partial vacuum in the tubes.
- Capsules or pods with propulsion, guidance, braking, control and communication systems.
- Pods with passenger or cargo facilities including air, light, heating, seating, luggage space, passenger restraint and luggage/goods restraint.
- Loading arrangements for people and/or goods to and from pods at atmospheric pressure.
- Connection or launch arrangements for pods at atmospheric pressure to and from vacuum tubes.
- Loading arrangement/launch capacity matched to tube capacity/pod frequency.
- Maintenance facilities for the tubes, the systems within the tubes and pods.
- Emergency arrangements for incident response to allow evacuation of passengers and removal or recovery of pods.
Note that the envisaged, relatively small pod sizes and capacities demand a high frequency, short headway service to provide useful system capacity. The high speed and the need to control g-forces requires an alignment with a very high radius of curvature and a high dimensional accuracy.
Physical Infrastructure for Tunnels
There are two options: the tunnel as vacuum tube; or the tunnel as envelope that contains one, two or more vacuum tubes. If functioning as the vacuum tube, the tunnel needs to meet the alignment’s geometrical and tolerance requirements. All linear services for maintenance, operation or emergency functions need to be contained within the tube. The tunnel lining needs to be virtually impermeable to air or vapour. There needs to be a capability to maintain the vacuum and also provision for whatever emergency access/evacuation is necessary at whatever intervals are necessary. A benefit of tunnel as vacuum tube is that thermal expansion, contraction and braking loads can readily be taken back into the ground.
Hyperloop Geometry and Tunnel Geometry
Transport systems tend to require rectangular cross-section vehicles. This is the case for seated passengers or the vehicle to carry them. It is also the case for most freight. Most long tunnels tend to be circular. Structurally, this is generally the most efficient cross section as it allows any tunnel lining to be designed efficiently, principally for compression forces. It permits a variety of construction techniques, including the use of tunnel boring machines (TBMs).
The ‘spare’ space generated by a rectangular vehicle envelope in a circular tunnel can be used, but only in a very inefficient way for services, traction supplies, walkways, ventilation, etc.
Hyperloop Tube and Tunnel Configurations
The next consideration is the arrangement of tubes in tunnels. This will generate possible tunnel sizes, which can vary: from very small if the tunnel acts as the tube for a one-way pod, to very large if the tunnel acts as an envelope at atmospheric pressure and contains two vacuum tubes and space within the atmospheric envelope for evacuation, maintenance access and services.
Tunnelling Efficiencies
Progress rate and cost are affected by tunnel diameter, but the relationships are not simply scalable. There are many contributors to progress rate and cost.
Something simple like spoil has a disposal cost that is closely related to the second power of diameter and the first power of total length of tunnel but is also related to market forces in terms of availability and cost of a suitable disposal site.
The cost of excavating spoil depends on the design and the configuration of excavation equipment (often a TBM), but also the cost of the logistics inside and outside the tunnel. The cost of the head of a TBM may be related to the third or fourth power of diameter. Spoil logistics are an integral part of the overall logistics but are not independent. The overall logistic arrangements are affected by the tunnel diameter and length. As tunnel diameter increases the range of logistics arrangements for spoil, segments and personnel access changes. At some nominal diameters, a small increase in diameter does not change the logistics options but with others a small diameter alteration will enable a significant change to be made to the logistics. Safety and practicality considerations at small diameters have a great influence on logistics and are greatly influenced by the size of a human being.
Efficiencies in progress rate and cost will rise initially as tunnel length increases. A significant part of programme time and cost is the initial set up for the tunnel construction, particularly if using a TBM. Although procurement time can often run in parallel with physical site preparations for tunnelling, there is site preparation time, TBM erection and launch time, commissioning, system testing and learning before tunnelling can reach full efficiency. Segment design, casting, supply and stockpiling also need to run in parallel. The cost and time of these initial activities need to be spread back into the overall cost and time per unit of length.
Excavation
Excavation of rock tunnels may be by drill and blast in some ground conditions. Removal of ventilation fumes increases the ventilation requirements. In weak rock, open-face excavation using road headers or excavators may be possible. However, in most ground conditions including many rock types, a TBM will probably be used for long tunnels.
Spoil handling
For most TBM-driven tunnels of medium size and upwards, a conveyor will be used for spoil removal. For small and very small tunnels, spoil is likely to be removed using spoil skips on a construction railway.
In some ground conditions, spoil will be removed using a slurry system with pipes used to pump slurry to the TBM face and return pipes taking the slurry containing the spoil to the surface for separation and treatment The use of slurry TBMs is to suit the ground conditions and is generally applicable to tunnelling in non-cohesive soils or chalk.
Tunnel lining – segments
In soft ground and some types of rock, a tunnel lining will be required. This is likely to be precast segments installed within the protection of the TBM shield or tail skin. In rock tunnels and some weak rocks, such as chalk, and strong soils such as stiff clays, a sprayed concrete lining may be used combined with rock bolts, if necessary. Following any such form of initial lining, a permanent lining is likely to be required for long-term durability.
Therefore, during construction, the tunnel will need segments or materials for a temporary lining to be followed by concrete for a long-term lining solution.
Segments can be manufactured offsite and delivered to the job site. For long tunnels, it may be practical and economical to manufacture segments at or very near the TBM launch site. Concrete for segment manufacture or any other form of lining is likely to be mixed on site for projects with high volume demand.
Transport of segments
Transport generally uses a construction railway of 600mm- or 750mm-gauge for very small tunnels, with a limitation on both vehicle and locomotive size. For small tunnels, battery locomotives are generally preferred but currently-available types constrain tunnel lengths. For tunnels over about 4.5m-diameter, 900mm gauge track with a 1,600mm wide locomotive and rolling stock envelope are almost an industry standard.
Locomotives can be battery electric for shorter tunnels, diesel for longer tunnels subject to ventilation capacity and can possibly be trolley electric with batteries.
For medium and large tunnels, rubber-tyred vehicles can be used but these tend to be used mainly for shorter tunnels only and for carrying segments and other materials, with spoil going out on conveyors. They need to be able to pass in the tunnel
Tunnelling Constraints
Alignment
A tunnel for Hyperloop will need to have very large radius curves in the horizontal and vertical planes. This means very shallow gradients for any launch arrangement. They also require high accuracy in relation to the planned alignment. The model tunnel considered in detail [in the full report] will be based on a minimum 25km horizontal or vertical curvature.
High accuracy can be achieved with modest lengths of tunnel but long, blind headings (perhaps meeting a long, blind heading constructed from the opposite direction) can be a significant constraint if there is no independent way of ‘closing’ a survey traverse. ‘Closing’ is the process of verifying the true position of the tunnel once a connection to the surface or to another tunnel has been achieved and correcting incremental errors in the survey that have arisen during tunnelling.
Geology
Geology is generally one of the first constraints to be considered in tunnel design, although in urban areas with significant existing underground infrastructure, the need to find an available alignment may take precedence. Any attempt to choose a good geological horizon for a long-distance tunnel is likely to conflict with the operational requirements of a Hyperloop alignment. Therefore, any tunnel for Hyperloop must be capable of being constructed in the prevailing geology rather than benefitting from the choice of the most suitable geology.
Effective length of tunnels
The preceding text has generally considered the tunnel under construction to be a blind heading.
Where there are two parallel tunnels, they can be cross connected at intervals with a cross passage and a flow and return ventilation system implemented to reduce the length of blind heading to be ventilated. It can also permit a degree of survey traverse closure, improving the accuracy of tunnelling.
A cross-connected second tunnel can also be used to provide a refuge tunnel in the event of incident. This can reduce the effective length of the leading blind tunnel heading. A crossover connection between two parallel tunnels can simplify logistics sharing a common route for spoil and segments, if required.
An intermediate shaft can reduce ventilation requirements and possibly reduce travel times at shift change. If it is to provide full logistics for spoil removal and segment supply, the shaft and related logistics will be a major undertaking.
Drainage
Tunnel vertical alignments usually address drainage needs, whether from water ingress through the lining, from shafts, from utility leaks. condensation or the potential use of fire hydrants. To achieve self-cleansing flow rates in pipes, gradients of 1:500 or greater are usually required, lower gradients risk siltation and blockages unless the systems are routinely maintained. However, the length of the tunnels and speed of Hyperloop transport makes the routine use of such gradients impractical. Drainage provision will be a major design and maintenance issue.
Model tunnel for hyperloop
The many variables make it difficult to define a model tunnel for Hyperloop. Any choice of model will be subjective. Noting the foregoing comments on very small tunnels, only those of nominal diameters 4.5m, 6.5m and 8m will be considered as long tunnels. Tunnels of 4.5m-diameter will be considered as a pair of twin tunnels, while 6.5m and 8m tunnels will be considered as single tunnels containing twin vacuum tubes.
For simplicity, it will be assumed that a model Hyperloop tunnel is 100km long, with 25km tunnels driven from a ramp portal at each end, an intermediate drive point at the 50km halfway point. At the halfway point will be major underground development and surface connections to facilitate the driving of tunnels 25km in each direction from the halfway point to meet the tunnels driven from the ramp portals. The assumed tunnel methodology will be EPB in pressurised mode from one portal, slurry TBM from the other portal, EPB in open mode with the facility to switch to closed mode if necessary, for one of the drives from the halfway-point and a rock drive with an open face TBM for the other drive from the halfway-point. All drives will be assumed to be segmentally lined.
The underground development at the halfway point is assumed to include a 200m-deep shaft and inclined adits. The removal of TBMs from their meeting points is assumed to be achieved by driving the TBMs offline for burial in situ and making a final connection with SCL methods. There will be an intermediate shaft roughly twothirds of the way along each drive, i.e. approx. 17km from launch to allow the ventilation route to be shortened and to allow a survey tie in to be made.
Ramped portal boxes to achieve a running surface depth of around 25m with a vertical curvature of 25km (as noted in the section describing alignment) will be long and in the model tunnel they have been assumed to be 750m long. They can be shortened if the speed is reduced. This may be possible if the tunnel portal is near the beginning or end of a Hyperloop route rather than at the transition from a surface alignment to a tunnel alignment.
Spoil removal from the portals is assumed to be from a new railhead on an existing adjacent railway line to a suitable rail connected disposal site 100km away. Segments are assumed to be manufactured close to each portal site with aggregate and cement supply by rail from a distance of 100km. The sites are assumed to have enough storage to allow stockpiling required to avoid programme risks due to shortfalls in segment supply.
For the halfway point site, a remote location is assumed with a need for a 10km conveyor tunnel to a new railhead on an existing adjacent railway line connected to a suitable spoil disposal site 100km away. Segments are assumed to be made on or near site with aggregate supplied by rail and via a conveyor in the conveyor tunnel. Cement is assumed to be delivered by road.
Costing a model tunnel for hyperloop
Initial cost estimates have been prepared for 100km-long twin 4.5m, and single 6.5m- and 8m-diameter tunnels. The costs are for civil engineering elements only and include a normal level of waterproofing only. They do not allow for an air- and vapour-impermeable lining.
Vacuum containment is assumed to be provided separately with a steel or other tube(s) inside the envelope of the tunnel.
Costs include direct tunnelling costs, tunnelling establishment costs, preliminaries for all civil engineering construction, civil engineering design costs and off-site costs for the removal of excavated material. Risk and uncertainty has been addressed using [UK] Treasury norms for optimism bias. An allowance is made for client project management costs. All estimates are based on the use of currently established methodologies and technologies.
For each tunnel configuration, a cost has been derived. In Table 1, the total cost for each configuration has been divided by 100km to give the cost per km. It is considered reasonable to extrapolate these costs upwards to provide an indicative cost for substantially longer tunnels. However, a significant reduction in length would require a revision to the cost model and may lead to an increased cost per unit length. It should also be noted that the model tunnel includes two portal boxes that increase the overall length of ‘path for Hyperloop’ beyond the nominal 100km used for deriving the cost per km.
The costs tabulated below may seem high and [in the full report] they are compared with benchmarking data from past projects. Production rates are not aggressive, and the supporting infrastructure that is needed for tunnelling 100km in the UK is considered to be extensive. It includes very long portal boxes, deep shafts, a halfway point underground development and access arrangements, extensive surface infrastructure and the construction of a 10km conveyor tunnel to get excavated material to a railhead.
