Brisbane Australia's CLEM7 toll tunnel broke after 11 months operation - horrible traffic forecasts
By Peter Samuel
A banking syndicate in Australia last week pulled the plug on RiverCity Motorway (RCM) owners of the CLEM7 toll tunnel after just eleven months of operation, adding to other failures of Australian city toll tunnels - the Cross City and Lane Cove tunnels in Sydney. The 24 bankers refused to renew shortterm loans of $1.3b because of RCM's traffic and revenue about one third of that projected by engineering consultant Aecom in the 2005 traffic and revenue modeling that launched the project. The bankers have put ownership in the hands of a bankruptcy receiver.
A public issue that raised $690m from a stock sale at $1/share in 2006 is now worthless.
As with the two other Australian toll tunnel financial failures revenue is more than sufficient to fund operational costs so motorists won't see any difference. But Aecom which generated the rosy scenario numbers - first for the Brisbane City Council, then even higher numbers for RiverCity Motorway which did a public offering citing the forecasts - is likely to be hit by law suits by investors.
Robert Bain the UK-based rating analyst says recent Australian toll tunnel projects, especially the CLEM7 have been "clearly tainted by optimism bias in spades."
It was not difficult to detect he says but those concerned didn't want to know.
Bain says in an email: "Nobody should be left believing that this (CLEM7 bankruptcy) was some unforeseeable accident. Nor is this symptomatic of the asset class and its inherent risks. It was none of these - by a long shot."
A frequent problem is the lack of 'skin in the game' longterm among those promoting and developing the project. Designers, builders, lawyers, finance brokers, forecasters and indeed the public officials promoting the project get their remuneration by the time the project opens. An unorganized collection of investors then get to carry the can.
Other toll tunnels in Australia have done well - Melbourne CityLink has major tunnel segments which were its most expensive feature. The Sydney Harbor Tunnel and the substantially tunneled M1 Eastern Distributor southeast of the Sydney CBD toll concession between Sydney Airport and the downtown. But the more recent three financial fiascoes of toll tunnels do cast a cloud over the viability of this class of project.
Not that it was unseen. Transurban the biggest investor owned toll operator in Australia declined to bid on the CLEM7 project - then known as the North-South Bypass Tunnel - on the grounds that it was not financially viable. Macquarie bid, but were not competitive with RCM.
Traffic has been running at 22k to 23k veh/day versus 58k forecast for the opening year in the Aecom study for Brisbane City Council - a 60% shortfall.
Brisbane's mayor Campbell Newman who was in ultimate charge of the procurement of the concessionaire to build the tunnel has said he is sad for the investors who lost their investments, but the concession got the tunnel built and will ensure it is operated for 100 years for the benefit of the people of Brisbane.
The state premier, Anna Bligh is also firm that investors took the commercial risk:
"The taxpayers won't be out of pocket here. All of the risk has been borne by the company. The tunnel will stay open and it's a good example of why governments, including the Brisbane City Council, have chosen to go down this path. The commercial risk is not being borne by the taxpayer and that's a good thing."
At no level of government is there consideration of any bailout.
Analysis of a flawed traffic study
We've got the original Aecom North-South Bypass Tunnel Traffic and Transport report of January 2005 conducted for the sponsors of the concession Brisbane City Council, but there are local reports in Australia that even more optimistic numbers (notably 90.7k vehicles/day) were generated by Aecom for RiverCity Motorway before it went for a public issue of stock about a year later. An "Analyst's Briefing Pack" handed out from Sept 21, 2006 included a Maunsell/Aecom forecast with an estimate of 90,676 average daily traffic within six months of opening.
(The authorship of the contentious forecasts was under the name "Maunsell, an Aecom Company," Maunsell being a long established Australian engineering consulting firm acquired by Los Angeles head officed, New York Stock Exchange traded Aecom and now effectively Aecom's Australian arm.)
The report for the City Council concluded with a strongly positive tone: "Based on the traffic modeling outputs, the Project is economically worthwhile, having a net present value of $638 million from quantifiable direct transport benefits (discounted present values, using the discount rate of 6%, and future costs and benefits discounted back to 2004). The analysis result is relatively sensitive to the estimated travel time savings but even if these were considerably lower the project would still be economic, in terms of direct transport benefits."
The CLEM7 tunnel added 2x2 lanes to five existing bridge crossings of the Brisbane River which meanders from southwest to northeast through the middle of this metro area. 6.7km or 4.25 miles long with 4.8km or 3 miles in tunnel it provides a new north-south bypass of the central business district.
The flawed Aecom forecast of Jan 2005 is based on:
- 1.6% assumed annual increases in total river crossing traffic,
- peak period traffic volumes on central area roads to rise annually 2 to 3% (early in the report) 4% annually (in the conclusion)
- a forecast doubling in peak period travel times (from 16mins to 32mins) between 2003 and 2016 in central Brisbane without the tunnel because of worsening congestion
- an estimated 40% of crossing traffic near the city center being through-traffic, that is not having any central city origin or destination, and therefore likely to use the tunnel
- 3.9% annual population growth Inner Brisbane 2001 to 2011
Competing routes are:
- Gateway Bridge and M1 Motorway an established private toll concession by Queensland Motorways about 5 miles, 8km downriver that handles around 94k veh/day, recently increased by duplicating the spans from 6 lanes to 12 lanes and now has all-electronic or 'free-flow' tolling
- Captain Cook Bridge, M3 Pacific Motorway/Inner City Bypass on the west side of the CBD
- an tolled 2x3 lane Story Bridge with foundations almost on top of the route of the CLEM7 tunnel that connects signalized arterials with select grade separations either side and carries around 100k veh/day
This study should have raised all kinds of red flags.
Inner Brisbane was not going to grow by 3.9% a year for a decade, although no doubt there were urbanist planners who wanted that. No inner area of any western city anywhere in the world is growing like that. Most are barely growing, many are losing population.
It just isn't credible that such a huge slice of the Story Bridge traffic would shift to a tunnel serving longer distance traffic when there are far better routes for them - the Gateway to the east and the Capt Cook Bridge and Inner City Bypass to the west - at no toll.
Why would inner city peakhours traffic grow at 2 to 3%/yr (p92) - let alone 4%/yr cited in the conclusion at p181 - when overall traffic is growing at 1.7%?
How in the world did people accept a 58k traffic number produced for the City and a 91k number for investors?
Do you take seriously a report that throws around such divergent numbers almost promiscuously?
The doubling of travel times in peak periods to 2016 is completely out of line with any real world experience in a developed western country in an established central area. Congestion just doesn't get that much worse that fast.
The forecasts of traffic on the crossings with and without the tunnel show a huge transfer of traffic from the Story Bridge to the Tunnel, the Bridge traffic dropping from 101.7k to 62.4k/weekdays for 2018 that makes no sense at all. At 62.4k/weekday this six lane bridge and associated roads would flow freely, travel times savings in the Tunnel would be much smaller, and only those trips with a very high value of time saved would pay the toll.
The no-toll forecasts for traffic in the tunnel 2011 without a toll at 121k/weekday are ridiculous considering the Story Bridge only carries around 100k. The Gateway Bridge a few miles east has lengthy motorways either side and attracts around 100k. Why should a 2x2 lane tunnel with sub-motorway/expressway standard connections attract 120k?
The Aecom report provides no explanation as to how these fantastic numbers were generated, beyond the implausible notion of travel times doubling by 2016 and the tunnel somehow being uniquely positioned to take longer distance traffic - both of which clearly match Rob Bain's "optimism in spades" on behalf of the tunnel project.
Only working from quite absurdly high no-toll numbers could traffic with tolls be projected at 58k/weekdays.
The whole Aecom report has a boosterish tone to its prose from beginning to end. It sounds like salemanship, not analysis.
March 15 2010 the tunnel opened with three weeks of toll-free travel. Traffic averaged 62k weekdays. The first week of tolls at $3.50 for cars saw weekday traffic at only 20.6k and stayed around that level. The toll was cut to $2.00 at the end of June and traffic rose to 27.6k/weekday. By September traffic was reported at 28.4k/weekday.
RiverCity Motorway realizing it had got hopeless forecasts from Aecom hired new traffic consultants, Integrated Management Information Systems (IMIS). Their forecast is that the tunnel will run 40.4k/weekday by 2020, about half the number Aecom had provided Brisbane City Council in 2005, and less than half the traffic for 2011 the Analysts' Briefing Pack mentioned around the time of the stock offering, and 128k Aecom cited for 2021.
Tolls have since been raised to $3.00 and traffic is this year running 22k to 23k/day.
The CLEM7 tunnel is named after Clem Jones (1918-2007), Brisbane's longest serving mayor and the M7 motorway to which it is linked at the southern end. Construction is under way on a separate investor toll concession of Airport Link another toll tunnel that will extend the M7 northeast to Brisbane's major airport.
The CLEM7 is an impressive piece of engineering and construction. Using the world's largest tunnel boring machines (TBMs) - two Herrenknechts with cuttting heads of 12.34m or 40.5ft diameter the tunnel took only 3.5 years from start to finish. It opened nearly 8 months early. The large diameter enables generous overhead height and 2x3.5m, 11.5ft travel lanes plus 2m, nearly 7ft of shoulder. The tunnel has three connections - to a motorway on the south end and to signalized arterial at one intermediate point and at the northern end. Built in hard igneous bedrock with precast concrete liner segments it reaches a depth of 60m, 200ft at its lowest point, 10m or 35ft below alluvial soils of the Brisbane River.
The tunnel turns a 10 to 20 minute journey into about 4 minutes. By all accounts the tunnel provides a great ride, has excellent signage, lighting, ventilation and traffic monitoring and management. It has all-electronic or 'free flow' tolling using transponders or license plate imaging.
A spokesman for the bankruptcy receivers has said: "Operationally, the tunnel is performing well and we do not intend to make any significant changes to operations."
At $3b (treating US$ and $A at par) it ended up more than twice as expensive to build as was initially estimated - about $1.2b.
Tolls with a transponder are currently $3.00 for cars, $4.95 for medium commercial vehicles and $9.90 for vehicles over 30t weight, 66k pds. There's a 47c surcharge for a license plate toll.
BACKGROUND: Brisbane with 2m population is a modest sized metro area - a far third after Sydney and Melbourne in Australia. It is about the size of San Antonio TX, the 3rd metro area of Texas, Orlando FL, Kansas City MO, Las Vegas NV, Columbus OH, Charlotte NC, Indianapolis IN, Norfolk VA, around the 30th rank in US metro size. Standards of living and car ownership and usage are comparable with America.
COMMENT: Our job is to report failures because failures are news. And people need to learn from failures - clients, investors, officials and the companies themselves.
Aecom has clearly done a dreadful job on the CLEM7 tunnel forecasts and cost many investors heavily. Something was very wrong in their Australian operation. Their reputation at least for traffic forecasting takes a major hit. That doesn't mean they're a lousy company overall. When we first started writing about tollroads Vollmer Associates now part of Stantec did a lousy job for the Dulles Greenway with CLEM7 style overestimates of traffic and revenue 15 years ago. Many other forecasts they've done have been good.
Wilbur Smith screwed up bigtime over the Greenville Southern Connector projections. One or two other places their forecasts have been badly off too. But very often they produce good forecasts, and sometimes they're on the conservative side.
Parsons Brinckerhoff, Bechtel and Modern Continental, who have often done good and sometimes excellent engineering work, performed abysmally for the Mass Pike on the Big Dig project. These companies committed the ultimate sin of a consultant - they committed clienticide. Their incompetence or worse on this job killed the client. The Turnpike Authority in Boston became politically delegitimized.
HNTB at the Turnpike in Maine has behaved appallingly, cultivating a near-corrupt relationship of planning and then performing the planned work on a contract never recompeted since first awarded in 1985. Talk about asking for trouble! And they encouraged the Turnpike to waste $40m on an unnecessary and non-permittable new mainline toll plaza that has made the toll authority deadly enemies at the Turnpike's south end in York.
That means HNTB has done a horrible job in Maine, and that the head office in Kansas City hasn't supervised their people in Portland. It doesn't mean that HNTB doesn't have excellent people or that it doesn't do an decent job for hundreds of other clients in scores of other places.
We could go on... but you get the drift. All organizations are fallible. Many do good and bad work. Care is needed before generalizing from single examples - editor.