Why do cost overruns and schedule slippage happen

[jis]

A study on why cost overruns and schedule slippage happen in projects and what can be done to reduce the chances of such happening...

https://www.constructiontechnology....r-key-ingredients-for-success/8028979.article

 

[AmtrakMaineiac]

When I was doing software development at Honeywell one of the project managers had an expression for the optimism mentioned in the article. He called it "green lights through town". In other words the way managers sometimes estimate projects would be like estimating how long your commute was by assuming that every traffic light going through town would be green.

 

[cirdan]

The article mentions that repeatability makes costs more controllable, citing the example of a wind park due to the high level of repetition from one turbine to the next.

I was thinking, surely this is also true of a high speed line. Most of these projects use pretty much standard designs for everything. For example bridges over minor obstacles do not get designed from scratch for every instance but there will be a standard design or small portfolio of standard designs that gets adapted to the peculiarities of the location. Electrical feeders and substations are also pretty much repeatable if not off-the-shelf. Fencing, drainage solutions, all ditto.

When I toured Argentina I was quite surprised to discover there are so many rural stations that look like perfect clones of the sort of stations you might find in the parts of the UK such as the Cotswolds or Yorkshire, often down to the smallest details such as door handles and window catches. At the time I wondered why this was and wondered if it could really have been economic to ship architects and parts from England to the remotest corners of Argentina when they could easily have found craftsmen locally, probably willing to work for much less. But this appears to be the solution. It was all about de-risking and repeatability.

Maybe the problem with high speed lines is that there aren't very many of them so the carrying over of standard designs and solutions, as is done for example on highways, is just not occurring. Maybe we just need to take into account that pioneering projects need to be more costly as they do the learning and establish the blueprints for those that will follow.

In other words, the solution to cost overruns on high speed rail is more high-speed rail?

 

[jis]

One also needs to plan to develop an infrastructure for re3peatedly building the same thing over and over, instead of treating each project as a start from scratch project. It has been posited that the reason electrification costs so much more in UK and US when compared to many other countries - as much as an order of magnitude more than those in countries with a standing electrification establishment, like India and China, is exactly because of this effect.

This is not limited to Civil Engineering construction. One of the predictors of cost and success rate in Cardio-Thoracic Surgery is how many such procedures are carried out at the facility. Practice indeed makes for perfection and incidentally cost reduction.

 

[Metra Electric Rider]

When I toured Argentina I was quite surprised to discover there are so many rural stations that look like perfect clones of the sort of stations you might find in the parts of the UK such as the Cotswolds or Yorkshire, often down to the smallest details such as door handles and window catches. At the time I wondered why this was and wondered if it could really have been economic to ship architects and parts from England to the remotest corners of Argentina when they could easily have found craftsmen locally, probably willing to work for much less. But this appears to be the solution. It was all about de-risking and repeatability.

Apparently exporting cast-iron buildings was big business for Britain, such as churches that could be built from pre-fabricated kits of parts, which were shipping to the far corners of the globe, not just to the corners of the Empire. Argentina is interesting since there was even a branch of Harrod's in BA at one time and there was/is a large Welsh population in the country too (I believe only a few still keep up with the Welsh language).

 

[cirdan]

and there was/is a large Welsh population in the country too (I believe only a few still keep up with the Welsh language).

the interesting thing about the welsh in Argentina is that in contrast to many other groups who emigrated and just happened to take their language with them , the thing in Wales in the 19th century was that the signs were that welsh language and culture were on their last legs and would soon be extinct . Many of those who went to Argentina were welsh traditionalists who believed the only way to save and protect their way of life and language was to find a safe place and emigrate there . The welsh wrote to the then Argentina government who granted them land and the assurance they would not be interfered with in return for the settlers swearing allegiance to Argentina . This was a clever move by Argentina as it permitted them to settle disputed territories with loyal settlers .

In later years settlers from many other countries came there too and eventually diluted down the welsh culture . Today there are only a few thousand welsh speakers left and the number is declining further year by year as many people are failing to pass the language on to their children . This is ironic, seeing the original purpose of the settlement, and as in wales proper the welsh language has undergone a massive revival over the last 50 years or so .

 

[cirdan]

.

This is not limited to Civil Engineering construction. One of the predictors of cost and success rate in Cardio-Thoracic Surgery is how many such procedures are carried out at the facility. Practice indeed makes for perfection and incidentally cost reduction.

I expect that in the next couple of years AI and robotics will gain traction in medical treatment and lead to further streamlining and gains in efficiency and quality . Especially in highly repeatable operations . Although, in similarity to self driving cars there are still big questions over liability and who is responsible if the AI gets something tragically wrong that a human would have got right .

Maybe one day we will see AI also enter architecture and civil engineering but I think this is still further away right now . AI is of course already playing a big part in the automation of inspection and maintenance of ROW as this is pretty much a classical strength of AI , comparing what is to what should be and evaluating the difference .

 

[Metra Electric Rider]

I expect that in the next couple of years AI and robotics will gain traction in medical treatment and lead to further streamlining and gains in efficiency and quality . Especially in highly repeatable operations . Although, in similarity to self driving cars there are still big questions over liability and who is responsible if the AI gets something tragically wrong that a human would have got right .

Maybe one day we will see AI also enter architecture and civil engineering but I think this is still further away right now . AI is of course already playing a big part in the automation of inspection and maintenance of ROW as this is pretty much a classical strength of AI , comparing what is to what should be and evaluating the difference .

I think it'll come first in civil engineering and then architecture, mainly in repetitive things, like tract houses and franchise type stores/restaurants. I suspect it'll be a combination of scanning of existing spaces and then generation of design. It will probably see things in 3D better than some people can and catch little things that the eye misses, but the human touch of unexpected creativity will still be needed in design.

 

[jis]

I think it'll come first in civil engineering and then architecture, mainly in repetitive things, like tract houses and franchise type stores/restaurants. I suspect it'll be a combination of scanning of existing spaces and then generation of design. It will probably see things in 3D better than some people can and catch little things that the eye misses, but the human touch of unexpected creativity will still be needed in design.

Bringing this back to railway electrification, if there is a standing organization for doing it, they have institutional knowledge that they can bring to bear on mundane activities like drilling the holes for foundations for the posts, a seemingly simple thing that has flummoxed both the UK and US electrification project managers delaying projects by years, or requiring abandonment of part of it due to exhaustion of funds. Even in order to meaningfully use AI there has to be institutional knowledge of what works and what does not. Yes some or much of it can then be incorporated in AI training, but the institutional knowledge has to be generated to teach the AI system for helping in future projects. That won't happen unless there is an ongoing establishment.

 

[cirdan]

Doing things like electrification is on the one hand pretty much a do and repeat exercise, where you have a certain set of elements and solutions that you repeat over and over, and that should therefore be quite well controllable cost-wise and not require too much intervention by management. Then on the other hand there is variability in the details. Every location is slightly different, even if to the untrained eye it may look the same but there will typically be some minor thing that requires a minor tweak in the approach, often requiring inter-disciplinary considerations ranging from geology to electrical engineering to mechanical and structural engineering to legal considerations.

This is precisely the strength of classical AI: do the same thing over and over but with minor variations adapted within pre-defined limits, while flagging up anything outside of those limits for a human to handle.

On a different matter, I have a friend in Switzerland who is now retired but used to work for Kummler & Matter, the electrification company who design, build and install the actual catenaries, masts etc, that is, the mechanical component of the electrification which is the bit everybody sees and whose installation is the most disruptive and hence requires careful planning, often using night-time slots to go in and change out components, even put in wholly new masts with new foundations etc (they don't just do new stuff but a large part of their business is in replacing existing electrification as it reaches the end of its lifespan - much of Switzerland's electrification installation is pretty old, some of it pre-war even). This friend told me they never have this type of problem with cost overruns or problems drilling holes. Its pretty much a do and repeat operation that they cover with standard solutions and standard parts. In fact the most unpredicatble bit of the job was removing the old foundation blocks as you didn't always know how deep they went or how hard the concrete was. Swiss railways demanded a complete removal of this type of thing, whereas in other countries I often see masts just getting acetylene torched off at ground level and the foundation blocks left in-situ, with the new mast being placed a little to the side.

 

[George Harris]

One of the jokes about engineer's estimates as done by the Owner's engineer/consultant is that it is the cost of the job if it was being built in heaven.

Reality is that many times the increased cost is due to changes in scope, changes in conditions, unanticipated conditions and events, etc. Sometimes these can be substantial. One outstanding example in the CAHSR is in the line through Fresno. Originally the plan was to be elevated. Then the changed city government decided that they did not want to look at it, so the line was changed to be at or below ground level, which substantially increased costs due to major increases in costs for street and utility relocations. Yes, a mile of railroad at grade in open country is cheaper than a mile of elevated structure in open country, but in urban areas the reverse is true due to road crossings and utility relocations.

Having spent a few months in a contractor's estimating office many years ago, here are a couple of things that also serve to increase costs. One is a careful review of the quantities of each item as found on the plans versus the quantities in the bid form. The contractor's unit prices on the bid form are then adjusted to be slightly high on those that are undercounted in the bid form and slightly low on those that are overcounted. The target here is to achieve the desired total price on the bid form, but in such a manner that it will result in an increase in total as-built price. Note, this is called unbalancing the bid and is generally illegal, but nigh impossible to prove. Another is that the plans and specs are carefully analyzed by some of the most experienced of the contractor's people and a list of potential claims developed along with what the anticipated related price would be. Depending upon the level of experience and competence of the designer the costs relative to these can be substantial.