Private road tunnels

Posted: February 28, 2013 in Transport
Tags: , , ,

The financial collapse of Brisbane’s airport link tunnel last week adds to a growing list of failed road projects funded as a Private Public Partnership (PPP). Brisbane’s Clem7 tunnel, as well as Sydney’s Cross City and Lane Cove Tunnels have all gone into receivership following lower than predicted patronage and thus an inability for the operator to pay its debts. In each case the road continued to operate and was sold to a new owner at a discount price.

NOTE: You can also add Sydney’s Airport Line to that list of failed PPPs, but it’s a different case and so won’t be discussed here

As this is a private company, the financial burden is on the investors and lenders, rather than the government. Queensland Premier Campbell Newman commented that “one thing I’ll say about Airport Link is this is one decision…that the taxpayers won’t end up paying for, thankfully”.

There is nothing stopping the Queensland Government from being the one to buy the road, something which previous Queensland or NSW Governments have not taken up the option of doing despite the . Even if it doesn’t, the infrastructure is now built, and will eventually revert back into public ownership.

The problem appears to be with the traffic forecasting. This is roughly how it works.

First, the private company works out how much the project will cost to build. Although there can be cost blow-outs, this is still the easiest bit to estimate so let’s assume it’s 100% accurate every time. They then multiply this amount by a percentage deemed to be a sufficient return on investment (usually the going rate of interest plus a few percent) to gives the necessary revenue.

The government then provides the company with traffic forecasts, their estimate of roughly how many cars are expected to use it. The company then divides the revenue figure by the traffic forecast to get a toll amount.

EDIT: It’s been pointed out in the comments section that these estimates can sometimes come from a private consultant, rather than the government. Where it comes from isn’t really all that important, and there was a reason that the description above was “this is roughly how it works”. But it’s a valid point and is noted.

Where it all went wrong is that, for a number of reasons including poor forward planning, freeway construction in Australia has been requiring more use of tunnels (all 4 examples above were tunnels). These cost about 4 times as much as surface construction. So, 4 times the cost means 4 times the necessary revenue, and 4 times the necessary toll. This assumes static traffic forecasts.

In reality, they are not – they are sensitive to price. Higher tolls mean lower traffic. And what happens if you revise that traffic forecast downward? The required toll goes up even further! A vicious cycle if there ever was one.

There are only really 2 solutions to this.

One is to go back to having the government build roads. Michael Pascoe points out that this also has the added benefit of allowing access to cheap debt, as government loans have a lower interest rate than loans to private businesses. However, government debt is now a dirty word, and so this is an unlikely option to be chosen.

The other is to reduce the cost of construction. Good planning can do this in the long term. The M2, which is entirely on the surface thanks to the maintenance of a dedicated reservation for it decades in advance, made an after tax profit of $38 million last year (Source: page 74, Transurban Annual Report). Whereas plans for the M4 East portion of WestConnex have it slotted like the Eastern Distributor rather than tunnelled in order to cut down on costs. While there remain a number of concerns over WestConnex itself, the aim of finding ways to bring down its cost of construction is definitely a welcome one.

  1. Ray says:

    I cannot see the current plan for WestConnex being feasible. Assuming it will be a 6-lane motorway (which would be the minimum you’d expect) it would take up the whole of the existing Parramatta Rd corridor reservation which would deny access to adjoining properties without wholesale resumption along its route, which would be prohibitively expensive. The disruption to existing traffic flows during construction would be monumental and you would have to question whether it would be worth the pain. I just wonder whether Infrastructure NSW has factored in the practicality of the disruption and costs associated with a slotted motorway along Parramatta Rd compared with the cost of a tunnel alternative.

  2. My understanding is that WestConnex will only be 4 lanes wide, which is how wide the M4 is at the Strathfield end.

    Parramatta Road is 7 lanes wide, including the turning lane, and one lane in each direction could overhang over the slot. That would allow 4 lanes on Parramatta Road and 4 lanes in the slot.

    Details have yet to be released (which is the real concern, in my opinion). But the slot option doesn’t appear impossible.

  3. Ray says:

    Except for turning lanes at major intersections, Parramatta Rd is mostly 6 lanes wide. The previous M4 East Motorway proposal was for 6 lanes from Homebush Bay Drive/Centennial Drive, continuing in tunnel from Concord Rd to Sydney Airport/Port Botany, with connections to Gladesville Bridge, Anzac Bridge and Parramatta Rd at Broadway. To build a 4 lane slotted motorway would be a huge mistake, completely ignoring the folly of building the M5 East as 4 lanes, when the government is now faced with the prospect of doubling the capacity to 8 lanes.

    Although we still have to see the details, I suggest that with the limited access along its route and the greatly reduced surface road capacity, the slot option is not practicable as it would deny inner west residents of an effective arterial road link to the CBD and destinations beyond. Considering the disruption caused to existing traffic flows during construction, in my opinion there is no option but to build the M4 East component of the WestConnex proposal in tunnel.

  4. Greg says:

    I think that the plan was for 6 lanes (with widening to 6 lanes back to Parramatta), and my expectation based on the (limited) material published to date is that property resumptions will be required on at least 1 side of the road.

    The part which is intended to be a slot motorway is from Strathfield to Taverners Hill, with a bored tunnel beyond that. This is an area which still has a high portion of low value land uses (such as car yards) and has relatively big lot sizes, leading to easier acquisition, and a larger amount of resulting land left over post construction.

    This is part of the genius of the plan though – my suspicion is that they have done this as an excuse to compulsorily acquire as much land as possible along the corridor and then redevelop it has high density housing once the road is complete to offset the cost of construction of the tunnel. Kind of like a blunt form of land value capture.

    The other advantage from their perspective is that a slot is an excuse to reduce the above road capacity, something which always causes controversy.

    The problem is – who wants to live next to a slot motorway? I wouldn’t be surprised if it morphs into true cut and cover, if not at construction then over time as housing is built, allowing them to increase the value of the land along the road. It will be a true shame if provisions for this are not made during construction, which could also allow public transport (light rail or busway) to be built on the new ‘lid’.

    Anyway, perhaps a bit off topic, but bringing it back… I don’t necessarily support WestConnex, but I certainly support the aim of coming up with innovative ways to fund and offset the cost of construction of major transport infrastructure and the focus on ‘value engineering’ that seems to have happened with that proposal.

    With a similar focus for the ECRL perhaps we would have ended up with a bridge over the Lane Cove river as originally planned, rather than the tunnel that we ended up with. I would hope that we see a similar value engineering focus on future rail project.

  5. Dudley Horscroft says:

    A few corrections are needed.
    1. There is no moral hazard in Public Private Partnerships. They are legitimate ways of funding a project. They may be well set up or badly set up, but that is the case whether state funded or independently funded or jointly funded.

    2. Whether or not the “government then provides the company with traffic forecasts” I don’t know, but in each of the four failures in question the company relied on non-governmenbt traffic forecasts. So much so that in at least one case the forecast provider is being sued.

    3. The general rule of thumb for railway construction is that the cost of elevated construction is four times the cost of at grade construction, and the cost of tunnelled construction is ten times the cost of at grade construction. I cannot think that this would be any different for road tunnels.
    Specific cases may be cheaper, but how much cheaper could it be possible? Consder the additional costs of ventilation, lighting, and supervision, not needed for rail tunnels.

    4. According to the published “Have your say” document re WestConnex, there are four sections of the M4 to be considered. The western end is already four lanes in each direction, and the central section is to be widened from three lanes to four owing to the existing heavy congestion. The fourth section, Strathfield in, is to be six lanes. As Ray says, this will take up the entire existing road way. On the basis of plans published so far, of the new six lane section, four will be under the restored Parramatta Road and two, in the centre, will be in the slot. The “artist’s” rendering with the document shows that the entire strip of land either side of Parramatta Road is to be resumed, everything knocked down and replaced by high rise housing. In the “economic case” for the road, no doubt the cost of this has been netted our in the belief – plausible – that the value of the new high rise housing will be at least equal to the value destroyed.

    5. This leaves the question: if the inmost existing section has to be widened to four lanes in each direction to remove congestion, how is it that the new tunnel need be only three lanes in each direction? Will this not lead to congestion?

    6. In building a new project there need be two important queries to be answered. Is it economically sound, and is it financially sound? In the case of WestConnex the BCR is reported as 1.5, hence economically sound. The business = financial case is yet to be completed, but the above mentioned document suggests the preliminary answer is no, in that user charges will only cover about 70% of the costs. leaving 30% to come from government. This suggests that user fees have been estimated deliberately low as any higher fees would reduce the demand so much that there would be less revenue. But about 90% of the estimated benefits are due to time savings. So if demand is less than expected (think Cross City tunnel, etc) the benefits will be less. And if the resistance of motorists to tolls is greater than expected (think all tunnels, ) demand will certainly be down, approximately 50% in line with the other examples. Demand only needs to be 30% less than currently estimated to bring the benefits from time savings down by 30%, which would reduce the BCR to something not statistically different (given the amount of estimation in costs, etc) from 1.0.

  6. Dudley, to be fair what you’re offering is an opinion not a correction. And the opinion that PPPs in Australia have contained moral hazard is self evidently true. Projects like the NSW 35-year Reliance Rail PPP represent a massive political risk to Government, and as such are “too important to fail”: Government contributing $135m to propping up the contract when it looks shaky is a case in point. This is in my mind completely analogous to the moral hazard that underpinned institutions like RBS during the UK financial crisis. Once an institution knows it is “too big to fail” they are no longer carrying the risk. This effectively supports/ rewards risk taking behaviour, giving the participant all the upside (profit, success, career advancement) of successful risks, and a reducing/ eliminating downside (loss, failure, career setback) of unsuccessful risks. Over cooking a bid to win a 35-year PPP is such a risk. Can you imagine how the loosing tenderers, who perhaps judged the margins better and priced (higher) accordingly must feel? Much like responsible banks during the UK financial crisis who more conservatively managed their affairs I imagine.

    The response of Premier Newman to Brisbane’s tunnel is clearly one with a very different, and very important message.

  7. Ray says:

    I agree with your sentiments David, but I still believe there is a place for PPP’s provided they are rigorously assessed by the government to ensure that they are based on sound economic outcomes.

    Just digressing from the subject of this post for a moment, it will be interesting to see how the PPP for the operation of the NWRL pans out, as any proponents tendering for this project will be obliged to cap their fare structure to the same as the government controlled sector (Sydney Trains). As by their nature, they will have to make a profit for their investors, this suggests to me that the level of the government contribution will have to effectively subsidise the private sector, otherwise we would have another Airport Rail Link debacle on our hands. Is it worth it?

  8. The Airport Line model is generally considered a failure and unlikely to be repeated. It’s much more likely that the NWRL model will resemble that of ferries and buses, where the government pays the private operator a fixed fee to operate a set level of service and keeps the fares.

  9. Dudley Horscroft says:

    Thanks David for your comment on PPP and moral hazard. Perhaps I should, in the light of your comment, have expressed myself a little bit more clearly. “There is no inherent moral hazard in Public Private Partnerships.”

    What you have described is the moral hazard of a project that “is too big to fail”. This is where moral hazard lies. This is exactly analogous to, say, the “WasteConnex” project should it be entirely built as a Government project (hopefully unlikely). You get to the stage where contracts for road widening have been let, and there has been massive destruction of buildings along Parramatta Road. Then the contracts for the “slot road” and new tunnels come in at double the original (grossly-underestimated) price. Does government give up? Does it cancel the road widening contracts (as it should) and return the land along Parramatta Road to the original owners? No, this is a Public Project and the Public Service and Roads Minister cannot be wrong. So, many more billions are sunk into the scheme, and we end up with a cess-pit full of money.

    At least we don’t give Public Servants knighthoods for such blunders nowadays.

  10. Simon says:

    I think we’re getting carried away with comments about Moral Hazard. E.g. with the Boeing 787 program has had to buy out some of its risk sharing partners because it could not allow them to go belly up.

    Point is that there is risk sharing that might be a good idea, but lots of risk sharing that is certainly a bad idea. Reliance Rail is an example of the latter.

  11. Post has been edited to reflect the fact that traffic forecasts don’t always come from the government. The description was more of a rough guide to the process, and who does the forecasts makes virtually no difference to the point being made.

    But it’s a fair point, and part of a detailed comment.

  12. […] failure of PPPs like the Airport Line as well as the Cross City and Lane Cove Tunnels mean that government’s […]

  13. enno says:

    It seems to be a very questionable “rule of thumb” that an elevated railway costs four times as much as a railway at grade. Who made that up ?
    All of china’s new railways are elevated, and when you consider how easy it is to put in a three metre thick pillar every 60 metres and then put prefabricated tracks on top, it’s easy to see why. It avoids the need for thousands of drains, underpasses, fences, roads and stuffing around for months with bulldozers.

  14. Enno –

    Keep in mind that rules of thumb aren’t exact, that’s why they’re called rules of thumb. The first comment in the post on the link below has the x4/x8 rule of thumb used in the post above. The post in the link itself concludes multiples that are less than that, around x2/x4.

    But again, considering these are meant to be ballpark figures, it’s still a fairly good estimate as far as rules of thumb go.

  15. […] is rather ironic, as the financial collapse of the Cross City Tunnel actually represents a benefit, not a disadvantage of the PPP model. Despite the financial collapse, for the driving public the tunnel will continue to operate as […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s