The Grumpy Old Man

Index News Home Eco Life Etiquette Big Bro Business Money Contact About Forum


....Why Cost Estimates are Invariably Wrong and Will Continue to be so!....

...We regularly hear about major multi-million-bucks projects, where the estimated completion price keeps rising....

We regularly hear about major multi-million-bucks projects, invariably government funded, where the current estimated completion price keeps rising. Granted not always the case but frequently so. Why is it, you may ask, that the estimating process so often produces an incorrect answer.

If you look at the mechanics of producing one of these high-profile estimates it is easy to see why the production of an estimate will, nine times out of ten, give you an unrealistic figure and significantly less than the eventual outturn.

As a retired engineer, I was regularly involved in the preparation of a number of estimates and can understand how this happens. There are of course some mitigating circumstances which I will illustrate from some high profile, very political projects both past and present.

The estimating process is actually very simple. The job is continually broken down into constituent parts, usually of a size that can be proposed and then controlled by an individual manager. It is acceptable,, and most usual, at this stage to build in some contingency, usually based on the manager's previous experience.

(As an aside, even at this early stage, a human competitive nature kicks in, and the question invariably answered by the estimates is "what is the lowest cost for which you believe this work may be completed?" rather than "how much will it actually cost, given all that we know, to do this work?")

Thus a "tree" is built-up where the leaves are like individual estimates and these come together as a branch, then a trunk and so on until one has a total finished estimate. At each stage, as the estimates are brought together they will frequently be assessed by a manager you may suggest amendments, corrections or straight reductions. Sometimes these proposals are based on technical grounds, sometimes on commercial grounds and sometimes pure ignorance of te potenntial problems!

Hence, there is already a downward pressure on the original Engineer's figure.

Once the full technical proposal/estimate has been prepared, it is assessed by "The Business" who may propose amendments (again, usually reductions) to the figures and most importantly adding a Business profit. Once again at this stage, the financial proposals can be reduced on the grounds of Business competitiveness or it could be decided not to bid for the work!

It is obvious that the process outlined above is very simplified and invariably is underwritten by the experience and knowledge of the original estimator.

Finally the completed estimate is 'bid' to the Customer who in most cases, with the size of these projects, will be a Civil Servant who may or may not have full technical knowledge and experience of the proposed work solution in hand. Thus he may not be aware of any vulnerabilities in the proposals.

In order to get full approval to proceed with the project there is a most complex system in place to secure financing. At any stage during this procedure, staff representing the "Customer" may have similar views as those in "The Business" and approve the estimates, provided certain reductions are made. (These "views" may have nothing whatsoever to do with the "technical" aspects of the proposal.)

Let's look at some examples of estimated costs at work and see why they have been wildly wrong. Some current, some very historic. There are only examples and I'm sure there will be many a clever clogs who would dispute the figures quoted. Most have been gleaned from general publications readily available on the Internet.


HS2 high-speed Rail Link from London to the Midlands and Northern England.


Original est. £30,000,000,000 Current est. £106,000,000,000 and rising

As now, just as work is really getting underway, the estimate is one hundred and six billion pounds with the proviso that there can be no guarantee that this is the final figure! Is a touch of honesty now being seen to prevail?

Grumpy must admit that for once, commonsense seems to be prevailing. In a project of this size and nature, as an example, property purchase must be a prime consideration and any variation in the property market must also affect the final price. (And how often does that market vary?). Much of the proposed route will be through tunnels and looking at other projects involving tunnels, all of them have, cost wise, shot through the roof! I'm also convinced that many of the subcontractors will not have done an exhaustive 'ground survey', on the basis that such a survey would be expensive and with no guarantee of recovering that particular cost. The survey was therefore either not completed properly, or not at all. Clearly, unusual ground conditions will undoubtedly increase the cost as we shall see below.

One must observe that, a threefold cost increase, this early in the project, must question the estimators knowledge of what they're doing, or, did the early adjudicator have the knowledge of likely problems to do his job properly? Or did the customer change the requirements specification? Or did each of these have a bearing on the final result?

And of course, as a taxpayer funded project, are people playing fast and loose with the financial side?

Causes; political indecision, inadequate information collection/collation NIMBY and so on.........


Eurotunnel (The Channel Tunnel)


Original est. £2.500,000,000 Final cost £4.600,000,000

As this was an Anglo-French project I would expect the cost overrun to be affected by politics and language, without such annoying trivia as unexpected geological features found by the tunnel boring machines. (Commonly called 'technical problems') I would observe that the final one of those three was perhaps, in Grumpy's view, by far the smallest. Add on Specification changes and one wonders how the cost overrun was so small!

Cause; political meddling and unexpected geology


A Derbyshire Water Supply Project


Original est. £37,000,000 Final cost £107,000,000

This project was to provide drinking water for much of South Derbyshire and parts of Nottinghamshire and Leicestershire.

The idea was to abstract water from a local river, pump it through the new tunnel, to a new reservoir some 10 km away. The reservoir was formed by a new dam built as part of the works. Simple, oh no, what happened was textbook unforeseen project problems!

The delays experienced were twofold (as what one might expect from a two-part project). The proposal for the tunnel was to use a tunnel boring machine and fit sealing rings behind the machine as it moved forwards (one standard tunnel boring practise these days). When the job was complete the boring machine was to be driven off to one side and abandoned as by then it would be worn out! One problem was, I understood t hat half way through the job, the boring machine hit a band of very wet sand and was swept so far to one side that it was lost and a new machine had to be built to replace it!

At the time, I actually saw, round a tunnel access point, fields covered with sealing ring sections. The only problem was that they could not be used, at the time, as there was no boring machine and hence no tunnel in which to fit them. Since they were ordered for "just in time" use, even though work had stopped, they carried on turning up, as per their contract, and had to be stored somewhere.

But building the dam for the reservoir was not without its own problems. When it was more than three quarters complete, much of it collapsed. Fortunately the reservoir was without water at the time, had that not been the case it could have been an horrific experience. I did hear conflicting stories of the cause of this collapse, but they did revolve around a cavity in the ground below where the dam was being built. That cavity I understood to have been formed due to "a chemical reaction on site". Whatever the cause, what was left of the dam had to be totally removed, remedial work done on the basic sub-site, and the whole shebang rebuilt.

Both of these incidents caused significant delays to the project, believed to be in the region of two or three years with all the consequential costs ramifications associated with that delay.

The first of these two major incidences may have been avoided if there had been a full survey, complete with core samples taken at regularly intervals along the tunnel route. I suspect that somebody thought that a borehole every half metre along a ten and a half kilometres line would be a too large an expense. (After all, the bidder may not get the contract and thus this expense would be lost.)

I accept that I have not seen any technical report of the dam collapse but I feel intuitively that such an eventuality should also have been seen as a possibility. Again the use of more core samples in the area of the dam may have shed some light on the situation.

Cause; inadequate planning


Concord Supersonic Transport


Originally est. £70,000,000 Final Development Cost £1,300,000,000

Cost figures for this aeroplane really do reflect the overall problems associated with a large multinational project. The project was executed by a number of companies in the UK and France, although I am sure that control was being executed by both the UK and French governments.

At the time the project is getting underway, although much of the technology was there, the actual implementation of it in a commercial project was a huge technological leap forward and that alone, would have caused problems at each step.

Being such a giant leap and particularly in the realm of aircraft, where safety is of such a huge importance, the testing of the final development project was very extended.

Prime route target of the aircraft was transatlantic from London or Paris to New York. Or, in some instances the Caribbean or South America.

It is accepted that development of the aircraft, being so cutting edge technology in itself caused time extensions. During this protracted development period the rules applying to "aircraft technology" changed significantly, i.e. the product specification changed.

Introduction into passenger paying service was further delayed by political shenanigans with the American authorities. Grumpy believes that the root cause of all of this was American envy (or plain sour grapes) of European technology and through political pressure groups managed to delay the introduction into service of Concorde thus effectively killed the aircrafts commercial viability and profitability as well as significantly increasing the development cost.

Cause; politics and two Bosses, huge technical leap forwards, huge timescale resulting in operating environment changes, technical jealousy from very powerful 'competitor' sometimes called NIH, amongst many others


A New National Children's Hospital for Dublin


Original est. €650,000,000 Current est. "over" €2.000,000,000

This project has commenced construction and, the last I heard of it, it was little more than a hole in the ground.

This really is a fantastic example of what can happen when you can't make your mind up!

In 1993, proposals were made to amalgamate three Dublin Children's Hospitals. There was so much argument over this that two of the hospitals concerned prepared their own development plans to go their own way.

In 2005 a government commissioned report advised that there should be one such hospital in Dublin, providing a National Children's Hospital service, rather than three.

In 2006 a site for this hospital was selected and in 2011 plans were prepared and submitted for planning permission. Approval was given for these plans in 2012 but this approval was overturned on public appeal.

In 2018, yet another committee recommended a different site! New plans were prepared and submitted and these were approved with perhaps less of the public hoo hah than there had been with the other proposals.

Somebody seems to have been able to "kick ass", as work actually started on site that year.

So here we are fifteen years after the proposals were first made, three planning hearings no doubt each requiring very different detailed plans to be drawn up, but at last, somebody has got a shovel out.

It is actually no wonder that the costs have escalated when not only has it taken fifteen years (of spending) to get this far, but we had the banking crash in 2008 and the subsequent financial bailout of Ireland.

It is common sense that over such a long period costs must have ballooned. And even now, people are warning that they could continue to rise. This is hardly surprising when I'll bet that over 10% of the original budget was spent before anyone even purchased a shovel, let alone use it shift soil.

Cause; political shilly shallying coupled with an inability to progress any decision, pressure from NIMBY (or trying to please too many people), multiple planning applications, hence protracted timescales even prior to the actual plans being finalised


Carillion


Is not possible to take any overview of 'estimating problems' without looking at Carillion. So far as I can see, they were a multifaceted public service company seemingly doing most tasks from school dinners to building hospitals and other infrastructure.

Carillion collapsed with reported half-year losses of £1,500,000,000, debts of £900,000 and a potential deficit (unpaid company liabilities) of £587,000,000. In addition to this it was reported that investors in the company had been paid a dividend total of some £80,000,000 in the previous financial year. (How in the world, this lot was run up I do wonder. After all, what were the banks playing at particularly when, if I as a customer inadvertently go 10p overdrawn, they go ape sh*t).

I admit this next few paragraphs are based solely on media reports.

As part of the contract bidding process I have no doubt that the original Carillion estimates were put together at a relatively low level within the company. I would guess that "Management" reduced such things as "contingency", "profit mark-up" and various other things such as "capital equipment depreciation" "Chronos" and many other weird 'Business Expenses' that management are prone to adding to estimates. As a result of this, it was suggested in some forums that the price Carillion was charging, for much of their work was so low that even the competition could not see how they could ever do the work, let alone make a profit. (It seems they were right!)

This tinkering with the estimates, I would think, was done to ensure that Carillion won the contract whatever the cost! See where that got them.

It was suggested in the media that many Carillion contracts had a significant "start work down payment" (paid when work commenced). It was alleged that the latter of these payments were being used to support work on earlier contracts when it was found that they were financially inadequate. Let's face it, this is very like a "Ponzi" scheme where the interest paid to previous investors came from the current investor. These schemes work, on paper, provided there is an ever increasing number of new investors joining the scheme. Sooner or later, they run out of new investors, and thus funds of their own and in simple terms, are criminal and illegal.

The eventual outcome was that Carillion were unable to pay their day to day expenses. Hence, overnight, work on hospitals and schools stopped, leaving emergency provision needing to be made for other contracts providing essential daily services (such as school dinners). Larger projects just stopped and many of them now have a two or more year delay in their programme just to get them finished. Additional funding has been required on top of that already paid to Carillion. Not only have Carillion workers lost their jobs, probably without being paid their final salary but their prospective retirement pensions have also been blasted to smithereens.

And all for what? One might say, a big boys lust for growth at the expense of the toil of the workers. Perhaps overall an object lesson in why not to tamper with estimates!

Cause; greed and management incompetence and possibly illegal practises


Festioniog Railway (an historic example)


Never mind, the sort of cock ups mentioned above have been the subject of many a project ever since man tried to "see the future" or in other words make an "estimate"

For people unfamiliar with North Wales, the narrow gauge Festiniog Railway is frequently dubbed "the world's first commercial railway" with work starting as early as 1831. Starting as an industrial operation, it is now a major tourist attraction.

Financial figures for the project are extremely sketchy as they are based on the fact that some of the "ground" work had already been done. I have found that in 1833 a contractor's estimate (for major work) of £24,185 and 10 shillings was accepted for fifteen months work. (Taking into account inflation that figure would be of the order of £3,165,000 in 2020).

The contractor went bankrupt during the work and the Festiniog Railway Company used their own labour to complete it. Thus the eventual outturn is very difficult to compute as much of the work costs would not have been separately documented. The overall time taken to complete the work was eventually four years.

There is ample evidence that the work was very poorly specified as even while it was progressing there are arguments over whether it should run through a tunnel or over a number of inclines! Either way, the estimating was up the creek, probably because nobody had ever done anything like this before and as the specification was changed it was clear that the contractor didn't know what he was doing either!

Cause; inadequate specification, Contractor's inexperience in new (unusual) work, estimate based on invalid assumptions


Great Western Railway electrification


Original est. £874,000,000 current est. £10,000,000,000 (with bits missing)

As with many straight government funded projects there has been much analysis of why this particular project estimate has ballooned. High-level descriptions of the causes of the overrun are quite classic and can be seen in many of the projects in this small paper. There are openly admitted to be

inadequate Project Definition at a sufficiently early stage

inadequate detail planning in the early stages of the project, resulting in unforeseen surprises as work progressed

poor detail cost estimating

as is not unusual in a major project, financial and program risks were identified quite early in the project. These would appear not to have been addressed with due alacrity or the urgency with which they should have been analysed.

Sufficiently to say that all of these problems can be seen in many of the earlier projects discussed. It seems that Project Managers never learn or Business Executives fail to follow their advice for whatever reason.

As an example of the lack of foresight and understanding of the potential problems staring the Project Managers in the face is the recent realisation that a large spring in the middle of an existing tunnel under the River Severn has made the tunnel so damp that problems have now arisen with electrical earthing. I hate to say it but it ain't rocket science that damp (and particularly potentially saline) and high-voltage electricity don't mix very well! It is appalling that the project had progressed this far) to suddenly find that this can be a problem when the spring was there pre day one and even the 19th-century engineers who built the tunnel, recognised the (water) spring as a potential problem.

Cause inadequate project definition, failure to recognise address problems when they are found, poor detail cost estimating.


A personal experience


My own experience underwrites many of the aspects outlined above. Resulting from experience gained on earlier contracts, the company I was working for was invited to submit a single action tender for a similar product. (Single Action Tender meaning only one company was invited to bid)

The product we were bidding for used a similar technology as a project recently completed. The new product, however, significantly pushed the technical boundaries well beyond anything we knew had been achieved before.

The project was split into three parts with myself as the Project Manager of one of those parts. My estimating process was straight out of the textbook as it was clear what individual building blocks were required. It was standard practice at this stage to recommend a percentage contingency that should be applied to the work. In simple terms this is usually based previous experience and how big a step into the unknown the contract would be taking. In this instance I recommended 100% contingency for a particular subset of the work.

As the saying goes, management saw red as they thought the contingency, was huge when ten or fifteen percent was the norm. I fought tooth and nail (on the basis that we would be totally in the dark in our steps forward, eventually winning the case for this level of contingency on this (small)part of the contract. Towards the end of the contract, after initial testing had apparently proved the satisfactory performance of the equipment, final, more exhaustive and detailed analysis of the results suggested that there may be a technical problem. It transpired that there was an unidentified (possibly design) fault with a very small, frequently used but absolutely crucial part of the equipment. Fixing this new (and previously unknown) problem caused significant delays to the project.

Cause; unforeseen and unknown technical problems, not even identified by other users of a small piece part.

So, as can be seen in many of the examples above, estimating is a nightmare, no better than a best guess frequently bearing little or no resemblance to the basic source data. Having said that, Grumpy is actually surprised that anybody gets anywhere as near as they do to the eventual outturn!

Which would suggest to me that much of the fuss that is generated, particularly if it is in parliament or the media, is pure trouble making. With the best will in the world or even the best crystal ball, anyone could hazard a better guess, because, after all is said and done, that is what an estimate is.