We challenge TfL Costings

Following  his “key manifesto commitment”,  in a press release issued on 4 October 2016,  Sadiq Khan said the bridge would be going ahead, to be delivered as soon as 2020.

Three years after the Mayors unequivocal go-ahead for the bridge, TfL and the Mayor have done a U turn – now saying that the bridge is unaffordable,  based on the headline worst case scenario of £602M.

We have the following comments:

  • This bridge was promised to be potentially delivered this year.  Instead,  TfL have achieved nothing tangible in all that time.  Questions should be raised regarding why the huge gap between expectations and reality with a focus on judgement and competence
  • Why after several years of deliberations did the cost suddenly rocket from £330M to £602M in a few months?
  • According to our figures we believe there is a MASSIVE amount of bloat in the TfL figures amounting to £428M as illustrated here in orange and detailed in the text:
  • We think the headline £602M figure is open to question. Some are saying this has been engineered as a face saving exercise to stop the bridge for political reasons.
  • TfL must explain why this is 4 times the actual cost of the French Jacques Chaban-Delmas bridge which is a MASSIVE 575m vertical lift bridge with 4 traffic lanes, 2 monorail tracks and 2 sidewalk/bikeways

Although we were allowed confidential access to the breakdown some months ago which prevented us discussing it here,  as a result of a question from Caroline Pidgeon at Mayors Question Time this is now in the public domain

In our view the TfL figures should be reviewed because we believe they are fundamentally flawed.  The following table was published in the MQT minutes:

Concept design for central lifting bridge @ 12m height May 2019 Estimate Comments
Construction costs £183m Lifting bridge on latest alignment, including 90m high towers, 550m approach spans, 171m main span, landscaping and public realm works to Durand’s Wharf, concrete ship impact protection with eco concrete details.
Indirect costs £43m Design, surveys, supervision and associated costs
Risk £91m Allowance on construction, indirect costs and inflation at 40 per cent in accordance with Treasury & TfL guidance
Land (including risk) £46m  
Inflation £100m Based on the latest BCIS indices of the Royal Institution of Chartered Surveyors
Point estimate outtrun prices £463m  
Range £370m to £602m A cost range of -20 per cent to +30 per cent has been applied based on design maturity and market sounding

Our critique of the budget

SInce TfL have adopted an extreme pessimistic attitude to budgeting,  we are openly playing devil’s advocate by going to the opposite extreme with an optimistic assessment;  we invite TfL to meet us half way with a more realistic budget with a view to “getting the bridge done” – to borrow a phrase!

TfL have responded to an advance copy of our critique with this summary (full text here):

TfL has a robust cost estimating process, which follows the principles accepted and applied by other major clients and market leaders in our sector. We have robustly engaged with the market to ensure the estimate is as robust as it can be for the stage of development.

In fairness to TfL we accept that they are constrained by a framework which we regard as unduly oppressive and which stifles innovation, but there is still an enormous gap between our model and theirs.

Judge for yourself!

Construction costs £183M

Of all the costs in the TfL estimate,   we are not so very far apart on this element

Noting that this is for an ugly vertical lifting bridge following a more costly and bizarre “L” shaped route across the river bed.

The designers and engineers of the elegant Boomerang Bridge have estimates of around £150M which have been verified independently.  This includes construction, profit, inflation and optimism bias.  Noting that the TfL figures account for inflation separately we are applying an ex-inflation construction cost estimate of £120M in our alternative figures

Indirect costs £43M

“For Design, surveys, supervision and associated costs”

In the absence of a breakdown of these figures they are difficult to challenge.  However if we assume 250 working days per year over 4 years and with a professional charge rate average of £1000 per day this equates to an average of 21 people per day – just working on “Design, surveys and supervision” – not actually building the bridge!  We struggle to accept this is realistic and have invited TfL to provide justification.  In the meantime we will assume this figure contains significant bloat and propose £20M under this header

Risk £91M

“Allowance on construction, indirect costs and inflation at 40 per cent”

This exposes the level of diligence of the TfL estimate because inflation is already (generously) allowed for elsewhere – not difficult to miss, under the line item called “Inflation”!

And in what world is an inflation figure of 40% realistic?

Adding a separate blanket  “risk” item must always be subjected to the closest scrutiny because there is a high level of probability that it is cumulative because the underlying costs probably already include an allowance for risk.

In addition,  the procurement process should mitigate risk by requiring fixed price contracts from suppliers.

Regardless of this we accept that there may be unpredictable costs so we allow £20% on the construction costs – £24M

Land (including risk) £46M

Again,  we challenge the diligence with which TfL’s figures were compiled because this appears to be another example of double counting – this figure includes risk even though there is a separate category for £91M of risk!

We invited TfL to provide a detailed breakdown for this item because there is no way of knowing how this is split between the two landing points but they declined to do so.

What we do know is that because the Durrands Park landing point will be more extensive (due to the longer length of the ramp compared with the higher landing point at Westferry Circus) there will be a larger land acquisition on that side.  But this land is already in the ownership of the local authority and simply requires a change of use!    So far as Southwark Council are concerned,  the use of the land is simply changing from one legitimate public amenity (park) to another,  more valuable public amenity (combined bridge and redeveloped park) so there is no justification for paying acquisition costs.

There must surely be an expectation that since land owners on both side would benefit hugely from the new infrastructure the cost of the land should not be included

On this basis we are proposing this figure be deleted.

Inflation £100m

This figure must surely have zero credibility,  for the following reasons:

  1. Inflation is running at around 2% per annum.  On that basis it is arithmetically impossible to reach £100M!  Even if the whole cost was incurred at the end of year 4 and (generously) taking that figure as £300M,  this would amount to only £6M pa x 4 = £24M
  2. Realistically,  the costs would be spread over the lifetime of the project,  so the expenditure at risk of inflation would rapidly decrease with time,  so even the £24M would be far too high
  3. The inflation figure seems to assume that all expenditure will be subject to inflation,  but normal procurement practice would seek to eliminate this through fixed price contracts with suppliers.  Because only a very small proportion of the total cost would be subject to inflation our view is that the TfL calculation should be rethought.

On this basis we beleive that if the project is properly procured with fixed price contracts wherever possible a generous inflation allowance would be £10M

Range +30% (£139M)

The figure which we believe was instrumental in causing the project to be stopped was the headline £602M,  which assumes a “range” of +30%.  so that is what we will focus on.

But what is “range” exactly?  It is another word for “Risk” – and that is already accounted for!

So a 30% risk allowance is added to the £91M risk line item, and the £100M Inflation line item as well as all other line items which probably already include a risk element!

We propose that this figure has no realistic basis and should be dropped

The optimistic budget

ItemTfL EstimateOur estimate
Construction costs£183m£120M
Indirect costs£43m£20M
Land (including risk)£46m£0M
Point estimate outtrun prices£463m£174M
Headline worst case (includes range)£602m£174M

TfL’s current plan

Since the announcement that the bridge development would be “paused”,  TfL have been developing the alternative of a ferry.  Again,  progress is disappointingly slow.

TfL previously considered and rejected a ferry in favour of the bridge and we believe that since the fundamentals are unchanged the justification for now proceeding with the ferry option is not soundly based.

12 Disadvantages of a ferry:

  1. We would have to pay a fare for the ferry. Using the bridge was planned be free of charge, like any other bridge in central London.  Even the Woollwich Ferry which carries road traffic is free! Why should TfL exploit us on this crossing when all others are free?
  2. Constant interruptions and delays due to need to give way to normal river traffic
  3. Chaos of 3 ferries using a narrow stretch of the river
  4. Manpower costs for the three ferries proposed would be significant – we assume 18+ personnel for three vessels and two shifts x 7 days a week
  5. Unlikely to be 24/7 operation
  6. Potential for interruption of service at low tide and in adverse weather
  7. Greater potential for strike action (look at issues we get periodically on the tube)
  8. Delays during embarkation and disembarkation and waiting for the ferry
  9. The existing ferry fails to keep to the timetable and there is no reason why this would change with a new ferry
  10. Electric ferries are unlikely to be viable and operation of diesel engines (on the proposed three ferries) would cause constant noise and pollution. This would add to environmental damage at exactly the time when other transport solutions are going green
  11. The proposed 3 ferries will need to be replaced periodically.  Coupled with the higher running costs,  the whole of life bridge costs are likely to be substantially less than the ferry
  12. Transit time with a bike on the bridge is likely to be a predictable* 3 minutes.  The Ferry will inevitably be slower,  allowing for wait time, crossing time,  docking,  loading/unloading and delays due to river traffic – we estimate 10 minutes

The proposed bridge would have none of these disadvantages.  Compared with waiting and frequent delays on a ferry service,  the bridge could operate 24/7 and offer completely green and quiet operation with predictable* and short journey times

*We assume bridge openings will be scheduled, predictable, infrequent and only allowed outside busy times

Click here for our appraisal of bridge vs ferry pros and cons

Our Proposal

We propose that Tfl reappraise their figures in view of our comments to reset the budget and get the bridge project back on track.