Bridges Collapsing during Construction

My research assistant, who I thought had been slacking off recently, sends me this story:

On June 26, just two weeks after an “inspection” by President Uhuru Kenyatta, a $12 million Chinese-built Sigiri bridge in Western Kenya collapsed before it was completed.

Built by the Chinese Overseas Construction and Engineering Company in Busia County, the bridge connects a region that has historically lacked government investment and development. Around a dozen people died on the river after a boat capsized while attempting to cross in 2014.

I suspect she forwarded it to me thinking I was involved in the design – a reasonable assumption, given the results. But alas, nothing to do with me.

I doubt this is much of a story, actually. The first thing that strikes me is $12m is rather cheap for a bridge, even a relatively small one. And bear in mind this is in the middle of nowhere, which would account for a lot of the costs. Sure, the Chinese have screwed up but at that price, who cares? Just build another one a little further down.

Bridges are prone to failure during construction as, depending on the method, they see stresses during construction they would not normally be subject to once built. I have no idea what happened to the bridge in Kenya – from the photo it looks as though it’s simply sheared off at the far end – but at university I studied the case of the Cleddau Bridge in Pembroke Dock:

The bridge was expected to be completed by March 1971, however on 2 June 1970 a 70 m (230 ft) cantilever being used to put one of the 150-tonne sections into position collapsed on the Pembroke Dock-side of the estuary. Four workers died and five were injured. Construction was halted until October 1972.

(More pics here)

I took an interest in this case study because I used to cross the Cleddau Bridge often as a kid (the lecturer pronounced it Cledd-ow; correctly, it’s Cleth-aye). A section of the box-girder bridge was being extended in cantilever towards its permanent support when it buckled, resulting in catastrophic failure (this word document explains it well). Once installed on both piers, the completed section would have easily been strong enough, but in cantilever it is subjected to much higher stresses.

While this incident is largely forgotten, the collapse of Melbourne’s West Gate Bridge during construction a few months later brought about a global rethink on box-girder bridges. The cause of failure were different, but the failure modes similar (i.e. buckling of a cantilevered span during construction). I’ve both driven over and sailed under the West Gate Bridge, and each time I thought of the Cleddau Bridge back home.

Thankfully, we seem to have learned how to build large bridges without having them collapse during construction. Cheap bridges in remote parts of Africa? We’re getting there. At least nobody was killed (well, we hope: none was reported, anyway).


25 thoughts on “Bridges Collapsing during Construction

  1. 12 million pounds for the Welsh bridge as well. That was a pretty poor performance there on behalf of the design engineers, I am pretty surprised they made that gross error, there is no excuse for that in Britain. Telford would turn in his grave.

    Just in the finishing stages of signing a contract for a job in Africa myself, with a bit of a home brew design of ours, makes you think twice about it when you read stories like this. Not much Hi-Vis going on either.

    At times like this I like to pull the Hegigio Gorge Bridge out that I was on in PNG, it was the world’s highest bridge when we built it and was only pipped recently. It is a cable stayed structure, local guys actually had excavators down there on the side cutting the abutments, I wasn’t game to walk down there, everybody suffers from vertigo on unprotected slopes like that, never mind track an excavator down and cut into the rock, the spoil was just dumped over the side as you can see. Vegetation doesn’t take long to recolonise in that jungle. It was a day’s drive to get between the abutments, I choppered between, no accidents on the project, good Hi-Vis compliance. Russian chopper pilots, quite often very low visibility and they fly on memory as its too near the sides for instruments. The big A Frame was made in Brisbane, I held it back for trial assembly and pissed everyone off because of the shipping delay. Most and maybe all of the white guys in the pictures are British and Irish Lions and some Frenchies from Freycinet done the cabling. During the trial assembly in oz we discovered that the main bushes in the A Frame hinges were the wrong fit, took them out, drove them round to a machine shop and got them fixed, no worries. If we had discovered this problem on site it would have been a six to eight week delay plus costs.

  2. My knowledge of “structures” is pretty thin and pretty old (and learnt entirely by being required to teach the subject!) but you probably have more experience; do correct me if this is wrong. The thing with “structures”, at least at the level I was involved with, is that your intuition can carry you a long way, though you then have to do the sums to get the quantitive results you need. If the structure is complicated enough this description might fail, but in simple cases you could see if you faced hogging or sagging, whether a component would be in compression or in tension, and so on.

    Mind you, there may be further practical problems if you build with Chinese steel.

  3. dearieme,

    I’m no expert either, but behaviour of simple structures, e.g. cantilevers, bridges, etc. are fairly simple to predict in theory. Certainly, the calculations that go with the simple models can be done by hand, e.g. second moments of area, stress calcs, etc. This will get you the approximate size of your sections, or I-beams, or whatever.

    Of course, you then build the thing and Tacoma Narrows happens.

  4. I know naught of what you engineers say, but I do recall after the Welsh bridge failure the box-girder viaduct carrying the M1 across the Don valley in east Sheffield was partially closed for what seemed years. Probably was years as no one knew what to do about it.

    By the way this viaduct has a curve in it so it avoided a pair of (long disused) cooling towers. Recently amid much fanfare of local trumpets these ‘twin towers’ were removed lest they should fall on t’Motorway. The towers are gone but the kink remains to puzzle speeding motorists.

  5. Oh yeah, Galloping Gertie. Caused by resonant frequencies or something. I forget the details.

  6. Even hunting and gathering people know that it’s more dangerous to do hunting and gathering than to actually eat the stuff.
    Likewise digging a hole in the ground or fixing a roof more problematic than the finsihed product.
    If engineers haven’t noticed this then they think they are architects, or have pretensions to be.

  7. “The thing with “structures”, at least at the level I was involved with, is that your intuition can carry you a long way”

    Absolutely, there are many simple ratios of span length, section height, sag and hog that yee just cannae change the laws of. Obviously, we don’t see this in play much these days but they still apply.

    I used to work for the Queensland Main Roads and done a stint with bridge branch, this is when CAD was just starting to get take up. Draftys in those days just about done the bridge design. The other thing was that we used to build timber bridges right up until the seventies and we had many bridges under management from a maintenance point of view. We even had them in the Brisbane metropolitan region (probably still do), solid as and going nowhere, they were and still are great examples of how simple a bridge structure is.

    In another company I worked with we had an old salt of the earth type estimator, he used to wear a blazer and a tie and done his estimates on a typewriter, in the nineties! He had some great stories about building road and rail bridges literally in the middle of nowhere, real wilderness and total isolation. They would go there, make a camp, cook their own food, scout the area for hard wood trees suitable for bridges, fell them, trim them, use them for piles, headstocks, girders, decking and railing the whole nine yards. Main Roads had standard designs that were actually issued on regional basis based on the tensile strength of the local trees. He said that his design drawings never changed during the whole time he was doing this kind of work.

    See here how simple a bridge really is.

    Also see this new composite fibre (CF) bridge material in use, including the simple meccano style members, this company delivered the first ever CF road bridge in Rochester NY. Good use of Hi-Vis as well although I don’t know how effective that hardhat will be for the bloke kind of under the load would be if it dropped on him, I know its lightweight but…………

    Believe it or not they used to be big concrete producers, sold the business and were restricted from making concrete again and since pioneered the use of cementless concrete and CF. They have a big beautiful polished hardwood table in their boardroom in Toowoomba, except it’s not wood its CF, it is amazing what they are doing. They also built a commercial airport recently with all the material sourced from within or from scratch, I hope they win the contract for the construction of the Sydney second airport, they truly deserve it.

  8. I love the Hard hat lark, “they can save your life” or “prevent injury” is what you are told! One safety presentation showed a Steam engine (90+ Tons) being lifted on the Indian Railways, with a load of workers under the thing without hard Hats, as an example of bad practice… Sadly no one asked what a hard hat is supposed to do if 90 Tons falls on you!

    I used to be in charge of a mobile railway crane max lift 12 Tons, If we had a audience we wore hard hats, the audience would no doubt believe we were safer… When we did not have the audience we did not bother with the useless hats.

  9. Sadly no one asked what a hard hat is supposed to do if 90 Tons falls on you!

    You’ll be fine, provided you’re wearing the chin-strap.

  10. I do remember that when I wore a hard hat I banged my head more often: it cut down my peripheral vision a lot. On the other hand, each collision was less damaging, and it was claimed the hat would protect me from, say, a falling spanner. Anyway, it was a neat way to carry my goggles and ear muffs.

  11. @Bardon: “It was a day’s drive to get between the abutments…”

    Incredible. I was about to ask about potential corrosion – well liquids tend to be aggressive; how do you fix a hole in a pipe on a bridge like that? – but it looks like oil production from Mananda stopped in 2014.

  12. Hello TJ,

    Rail and bridges now there is a boom industry, if you are still in it. I see that China will now supply the bullet trains for the Moscow Kazan High Speed Rail confirming China as the world leader in high speed rail products and services. Plus I don’t like Alstom as they didn’t award me a contract on Sydney Light Rail recently. The amount and magnitude of rail projects coming on line is truly mind boggling. We are on the cusp of the biggest building boom the world has ever seen by a country mile, a new Gilded Age of sorts.

    Check out these rail bridges in action.

  13. I see that China will now supply the bullet trains for the Moscow Kazan High Speed Rail confirming China as the world leader in high speed rail products and services…

    …at prices Russians are willing to pay.

    I took the Kazan to Moscow train once, it was fairly rapid by Russian standards. I’d have said I took the Moscow to Kazan train, but I arrived from Saint Petersburg. Thems were the days.

    Plus I don’t like Alstom

    Nor do the French guys I know who work for them.

  14. @ Alex, there was man access provided for repairs. I wasn’t on the pipeline for that one although I have been involved in many pipeline in that area and from memory the product wasn’t overly corrosive. All the flow lines are above ground, black (uncoated) pipe, it is cheaper to up the wall thickness and just lay it out on top of the ground than coating and burying it. Being inland and elevated atmospheric corrosion is negligible and like I said I don’t recall any internal corrosion problems.

    Also done the Napa Napa Oil Refinery downstream and there were no special corrosion measures taken on any of the process work in the plant either, never seen NACE material specified anywhere in PNG either.

    By the way that abutment without the A Frame (with decent access road), to build it we started by dropping guys from a chopper with chainsaws, then dropped components of a dozer, assembled it on the cleared patch and started clearing out on a radius. Russian pilots were the king of the clouds, great set of blokes.

    Done some freaky work in Irian Jaya next door on Freeport Copper Mine as well, put in a jaw crusher at 4,500m, this was above the Grasberg glacier and near the Equator! You had to take a chairlift to get up that high, all gas bottles etc had to be half full to allow for expansion and staff needed a few days altitude training, that was another story that place. The airport was owned by Freeport and the main clock was set on Louisiana time, yes its true, some Indonesian Kopassus (purple beret) ran amok there and killed something like 17 and wounded 12 with 30 shots, just before I arrived a crazy, crazy place

  15. Yes, unless you’re dealing with H2S in high concentrations or some weird fluid, you’d just use carbon steel and bump up the wall thickness for internal corrosion. External corrosion can be more tricky, depending on the environment.

  16. A lot of corrosion happens internally at the bends. This is not because of chemical properties but because of the sand mixed in with what you’re pumping.

  17. @Bardon: I didn’t know Russian helicopter pilots ever worked in PNG. Where did they learn piloting over mountainous terrains? Most of Russia’s oil regions are flat.

    I’ve read that the Napa Napa refinery runs on the local Kutubu blend, unusually light and sweet (0.04% sulfur vs. 2% for Dubai), so corrosion from H2S shouldn’t be a problem. The reason I asked about the Hegigio bridge is that the pipeline was supposed to move well liquids – a mixture of crude, water, sand and gas – to a processing facility. But if it’s only a 15-mile flowline meant to last for 10 years, I agree a rupture is unlikely. Once the field is depleted, will the jungle gobble up the pipeline like some ancient temple?

  18. There is a saying in PNG that white men are either missionaries, misfits or mercenaries. I think the Russian pilots fit nicely within that lot, plus they were probably the only ones that had the expertise and balls to fly in that terrain on visuals.

    As for the jungle recolonising, absolutely, I can just picture the creepers entangling around the bridge cables already, wild pigs defecating on the pipe supports,the laughing picaninny playing drop the stone from the new short cut between villages with the older men spitting out betel nut on the shutdown vale. If you stood still long enough their lichens would start forming on your feet.

    Yes, quite right on the line, upstream of processing and different stream than Kutubu, I wouldn’t be surprised if they went for very heavy wall on the bridge crossing to mitigate this. I nearly came a cropper on sour gas in a big way on another project in the Bass Strait. It was an offshore/onshore project, with the semi processed gas piped 85 km to an onshore processing plant and then sales gas and liquids to market. The offshore unmanned platform Yolla manufactured in Batam Island and the pipeline were both NACE, the linepipe was being manufactured in Japan and the metallurgy was marginal on the NACE hardness, my client broached this with me and relayed their concern. I spoke to an expert with a Dr in front of his name who was also a pipeline manufacturer and doing our onshore line (non-NACE). He said that in his expert opinion it was all bullshit and as long as his arse points to the ground we would never have gas that sour in Oz that it could become an issue. I didn’t take his advice and upped the ante with the Japs to move the bell curve, which they did. The actual gas once producing was not only sour but far higher than anticipated. Look it up, I wouldn’t mind knowing what the levels were. If I had taken the Doctors advise I may have brought the whole company down as the 85km sub sea line would have been condemned.

    As it happens this contract nearly did bring the whole Clough empire down and gas quality being out of spec on the unmanned processing platform was one of the things that saved it. If it wasn’t for Harold Clough, a man that I looked up to throughout my career, a visionary that made it possible for me to do all these ball tearing jobs, generous such that he gifted me a massive number of wealth creating shares as a young man then I wouldn’t be where I was now professionally and the company would have definitely folded on that project. Even back then in the noughties and despite his age and walking stick with his personal fortune on the line he was still as sharp as any of the QC’s in the room that we were dealing with in that high stakes game.

  19. A lot of corrosion happens internally at the bends. This is not because of chemical properties but because of the sand mixed in with what you’re pumping.

    That’s erosion: avoid if possible, if not increase the wall thickness.

  20. I didn’t know Russian helicopter pilots ever worked in PNG. Where did they learn piloting over mountainous terrains?

    I expect they learned it first day on the job. I knew a sat-diver in Sakhalin, who was also a potato farmer, who made quite a lot of money offshore Italy working subsea pipelines because he and his fellow Russians didn’t really bother with decompression intervals. Young Russian men seem bewilderingly unaware of things like their own mortality, even more so than young men generally.

  21. Watcher: I’ve been across Tinsley Viaduct in Sheffield fairly regularly for the past 30 years and it has always been 2 lanes either side with the nearside lane blocked off. Until a few months ago when I was surprised to find 3 lanes again. They have worked on the road so it is now 4 lanes either side & perhaps they’ve strengthened the bridge as well.

  22. @Bardon: Many thanks for this mini-memoir. I was on vacation last week and then it took me some time to digest it and read up on the gas project. Curiously, the one impurity at Yolla I’ve seen mentioned is mercury – not a common one, it seems.

  23. Well spotted there Alex, I am impressed that you identified this. This was what I was eluding to earlier when I said that it was the gas quality being out of spec that saved our bacon at the end of the day. Our client being a cheapskate like most of them are went ahead on the whole project based on a sample from a single “bottle” of gas and that was their contractual input to us for the entire design process and contractual obligation in that regard.

    We were running way behind schedule, the offshore component alone had massive constraints being an unmanned platform. For example, we unexpectedly ran into winter time and were choppering our guys in from Burnie in Tasmania to the platform in the Bass Strait, the Helicopter flying regulations are tight as you have to meet the first and last light requirements take-off and landing requirements, no Russian PNG protocols there I am afraid. That meant that by the time we got there, desuited, started work, finished, cleaned up and suited up we only had about 3 hours productive work a day max, and three hour productive work is an oxymoron in itself.

    So, we were on a hiding to nothing with liquidated damages, can you imagine our joy when we eventual commissioned the platform and discovered the mercury content in the first gas, obviously there was no mercury processing allowed for on the platform and the client directed us to “Stop Work”.

    I remember that day very well, it was a Friday afternoon and it was just before Christmas.

    “Non adimplet contractus” as the QC’s like to call it.

Comments are closed.