-
Introduction to PPP in the infrastructure sector 6
-
Lecture1.1
-
Lecture1.2
-
Lecture1.3
-
Lecture1.4
-
Lecture1.5
-
Lecture1.6
-
-
Chapter 2: Structuring a PPP project 5
-
Lecture2.1
-
Lecture2.2
-
Lecture2.3
-
Lecture2.4
-
Lecture2.5
-
-
Chapter 3: Financing an infrastructure PPP project 6
-
Lecture3.1
-
Lecture3.2
-
Lecture3.3
-
Lecture3.4
-
Lecture3.5
-
Lecture3.6
-
-
Chapter 4 :Documenting the transaction: anatomy of a PPP concession agreement and key risk allocation issues 11
-
Lecture4.1
-
Lecture4.2
-
Lecture4.3
-
Lecture4.4
-
Lecture4.5
-
Lecture4.6
-
Lecture4.7
-
Lecture4.8
-
Lecture4.9
-
Lecture4.10
-
Lecture4.11
-
-
Chapter 5: Documenting the transaction: finance documents 8
-
Lecture5.1
-
Lecture5.2
-
Lecture5.3
-
Lecture5.4
-
Lecture5.5
-
Lecture5.6
-
Lecture5.7
-
Lecture5.8
-
-
Chapter 6:Documenting the transaction: other project documents 2
-
Lecture6.1
-
Lecture6.2
-
-
Chapter 7:Procurement arrangements 2
-
Lecture7.1
-
Lecture7.2
-
-
Chapter 8:Introduction to key sector issues 7
-
Lecture8.1
-
Lecture8.2
-
Lecture8.3
-
Lecture8.4
-
Lecture8.5
-
Lecture8.6
-
Lecture8.7
-
Road projects
(a) Demand or Availability?
How is demand risk allocated in the project? To what extent is the project subject to under-demand risk (in terms of toll revenue) or over-demand risk (which may affect the ability to maintain)?
Some countries launched PPP toll road projects with demand risk allocated to the private sector, only to find that these projects faced difficulties during the global financial crisis as personal income restricted car usage. A number have needed to be restructured and in those countries the trend is for projects to be structured on an availability basis. It is understood that a number of traffic advisers on those projects have faced legal action for over-optimistic traffic estimates.
The success of road tolling is also dependent on public acceptance – willingness to pay. Where alternative travel options exist, the toll chargeable will be a function of users’ value of their own time. In some countries, like France, road tolling is well established. The government has a role in ensuring that the public is made aware of the reason for road tolling and assurance given that the road tolls are a transparent reflection of the cost of building, financing and operating the road.
(b) Performance mechanism
How is performance measured during construction and operation? Is the regime sufficiently objective and usable in practice, or might it give rise to disputes, or even the risk of abuse by the government?
A typical availability based project may use lane availability as a measure of performance, with availability being measured by certain aspects of road quality (e.g. road surface deterioration). An alternative approach (termed “demand management”) is to measure average road speeds as a proxy for road availability.
(c) Tolling interface / competition
What is the interface between the project and any tolls being levied? If the project is subject to demand risk, is there protection against competing roads or other transport modes being developed by the government? How are tolls regulated? How free is the concessionaire to adjust tolls to meet unexpected costs?
Even if the project is an availability-based project, the government may wish the concessionaire to implement the tolling booths / tolling technology; or if this is being implemented by others, it creates an interface that needs to be managed. Where demand risk applies, it may be natural for the concessionaire to request some measure of protection against the creation or substantial improvement of competing routes to allow it to develop a business case, but some requests may not be reasonable – the government should not feel itself prevented from making legitimate improvements to the transport network that are reasonably predictable to meet growing demand. In emerging markets, the development banks may insist that there should also be a toll-free alternative; if so, this should be made clear at bid stage and not introduced as a response to public complaint.
In terms of toll regulation, a balance needs to be struck between the concessionaire’s desire to increase (or decrease) tolls to meet unforeseen changes in costs over time (such as changes in law) or to keep itself afloat as a consequence of reduced demand. But some regulation may be required – either by statute or contract – to ensure that the concessionaire does not abuse a market position, where there is no realistic alternative to using the toll road.
(d) Dealing with demand growth
Will any KPIs or maintenance costs be affected by demand growth? If there is an upside, does this have to be shared with the government and is this commensurate with the downside risks?
Demand growth may affect the concessionaire’s ability to meet demand-management KPIs (if they exist) and will increase maintenance costs. The contract may need to allow reasonable adjustment for this so that the concessionaire’s position is manageable. On a revenue-risk project, demand growth should increase the level of profit on investment; this could be managed through toll regulation (limiting tolls to a fair level of return, reducing the cost to the user), or the excess profits could be captured through a super-profit mechanism, e.g. a formula that requires a share back to government if the concessionaire’s rate of return exceeds a certain threshold. Alternatively, as used in some countries in Latin America, the contract could provide that it terminates early when a certain level of equity return has been reached.
(e) Accidents and emergencies
How are risks allocated when dealing with accidents and emergencies? To what extent is the concessionaire reliant on (or at the mercy of) the government emergency services? How do these events get addressed in the payment and performance mechanism?
If an accident requires the emergency services, the police or fire service may expect to close part of the road until such time as injured parties have been safely removed and it is safe to bring in recovery vehicles to clear the road. The concessionaire may have control over the latter activities but not the former, so it may be reasonable for the concessionaire not to be penalised under any availability KPIs unless it fails to respond in good time when the relevant service has given control back to the concessionaire.
(f) Weather and other events
What level of interference by weather and other events does the concessionaire have to assume? Are there levels of weather or flooding risks that cannot be predicted (e.g. 1 in 50 year events) that should constitute a compensation event?
This is a question of design resilience. The government’s output specification may require the concessionaire to design the road surface, bridges and tunnels to be able to withstand a certain level of natural events, and this can be priced. If a much more severe event occurs that was not required to be designed for, and the consequence of this is not fully insured, it may be reasonable for the risk to be shared.
(g) Maintenance and replacement
Maintenance costs are strongly influenced by vehicle types and loads; the relationship between vehicle weight and road surface damage is not linear. Changes in vehicle mix could be influenced by government policy and the concessionaire should be able to adjust tariffs to reflect any material changes in costs.
Excess vehicle loading (e.g. too much weight per axle) can cause disproportionate damage to the road surface. Government has a role to play in policing this properly and enforcing against offenders, using powers which are likely to be beyond the control of the concessionaire. If so, it may be reasonable for the concessionaire to be protected if there is a material flouting of the rules, or even if there is a material change in traffic mix (for example if the government diverts all heavy traffic from urban routes onto the concessioned road).
(h) Interface with third party rights
New or upgraded roads will have an interface with roadside developments, and may also travel over existing facilities such as railways and pipelines. It is not always possible to predict the level of disruption (or indemnity) that will be necessary to deliver the intended road and it may be difficult to obtain permits to work from unconnected entities.
The risk of obtaining suitable permits from third party agencies and property owners is always difficult to predict, and delayed receipt of permits can have a material impact on timetable if the float (contingency) in the project programme is used up. Government may have a role in ensuring that any agencies under its control ensure that permit requests are not unduly delayed if applied for properly, and accepting a commensurate risk in the concession contract. For third parties outside government control (such as private railway companies or pipeline owners), the concessionaire will most likely need to give open-ended indemnities in favour of the relevant parties in order to have the right to work over or move the infrastructure. This may create difficult to ensure obligations which also need to be brought into the risk allocation discussion.
(i) Contamination
Oil spills and shed loads are an occupational hazard for roads projects and could lead to significant environmental damage and/or significant revenue loss due to road closures. To what extent is the concessionaire protected against historic contamination on brownfield projects, and to what extent is insurance available to cover this risk during the operational period?
Ideally some level of environmental survey will already have been completed and this can be provided to the concessionaire. In many cases, government will except the concessionaire to form its own opinions on such surveys and take the risk, but if there is insufficient time or ability to verify this information during the bid stage, this may not be a cost-effective risk approach. In many projects, the government recognises this by accepting the environmental survey as a baseline and providing a level of protection to the concessionaire if environmental issues are discovered which were not described in the report or inferable from it.