The false labelling of meat products (horsemeat being passed off as beef) has made the headlines recently in Europe, with repercussions from the UK to Romania. The food industry isn't the only one where you might not get what you ask for. In the construction industry those enforcing specifications need to be especially vigilant. Structural failure, fatalities, and charges of negligence and manslaughter are entirely possible, as a spate of cases around the world has made abundantly clear.
In Japan in December 2012, about 180 concrete slabs – comprising part of the ceiling to the Sasago Tunnel – fell, crushing vehicles and their occupants, killing nine ('Fatal tunnel collapse blamed on aging bolts' , The Japan Times, 4 December 2012). Each slab weighed 1.2 tonnes. The slabs were supported by metal rods anchored by bolts. The bolts failed. The tunnel had not had any major repairs since it opened, in 1977. In particular, none of the ceiling support bolts had been replaced. A routine inspection on the tunnel in September had shown no irregularities, but hammer tests had not been carried out on the section that collapsed. Subsequent inspection showed that 632 of 670 problems identified involved the bolts, of which two were easily pulled off by hand, 608 were loose and the rest had been affected by corrosion. Fourteen other tunnels were found to have similar problems (C. Billones, 'Defects found on 670 sections in Tokyo-bound Sasago Tunnel' , The Japan Daily Press, 14 December 2012). The tunnel re-opened on 8 February 2013. Families of the victims have filed a criminal complaint against the tunnel's operator.
In another case, in August 2012, a month before the building was due to open, an engineering firm discovered (accidentally?) that 23,550 bolts used to fasten pre-weathered steel cladding panels to the structure of the Barclays Center , the new home of the Brooklyn Nets, were only half as strong as they were supposed to be (C. Bagli, 'Problem with weak bolts has complicated the Barclays Center's early days' , New York Times, 1 January 2013). The company advised the New York City Buildings Department, which then denied it had been told and issued a certificate of occupancy! Fortunately tests showed that only about 8% of the bolts needed to be cut out and replaced. The fabricator that had supplied the bolts (and the 12,000 unique panels, in a $32.4 million contract) had defaulted on a bank loan and closed for business during the project, which may explain the error. The plant eventually re-opened to finish the contract. The completed building, with the replacement bolts, survived Hurricane Sandy, so there was a happy ending.
In California, there has been quite a scandal about the testing of bridges undergoing seismic retrofit, with a chain of claims and counter-claims, beginning with an article in the Sacramento Bee, about a 'rogue' testing technician, who admitted to falsifying tests on other projects in 2008 and 2009. Caltrans issued a press release on 8 June 2012, calling for a full retraction of the original news story – this didn't happen. The Sacramento Bee reported (C. Piller, 'Caltrans' records show problems with tests on Bay Bridge, other bridges' , Sacramento Bee, 4 August 2012) that a California Department of Transportation (Caltrans) team had uncovered use of suspect (e.g. fabricated) radiation and sonic test data for some of the Bay Bridge tower piles. Other bridges were also affected.
A recent article in the Bee, 'Report: California's bridges, overcrossing safe' (7 February 2013), advised that Caltrans analysed more than 23,000 tests and found only 11 that were 'demonstrably unreliable'. Re-tests were carried out in these cases and the structures were found to 'withstand the stresses for which they were designed'. Nevertheless, Caltrans is changing its testing practices – it will subject test data to quality control assessment to verify the accuracy of the results, and will require greater oversight of testing operations in the field.
The most recent article, 'Caltrans experts, despite testing problems, say structures are safe' (8 February 2013), advises that there will be an independent review of the new Bay Bridge, the Federal Highway Administration released a report critical of Caltrans' testing lapses and materials procurement, and the official report does not address a number of key concerns. The article notes that 'an unknown number of data problems might have gone undetected'. This story will continue to run.
There aren't many testing regimes that have had to withstand this level of scrutiny.
Type test certificates
Two nuclear reactors in South Korea have been shut down to replace 'non-core' components – more than 7,600 fuses, switches, cooling fans and other parts – that had been provided with 60 forged quality certificates, by eight suppliers. Most (95%) of the products were made in Europe and the USA. Investigations were initiated to identify who faked the certificates and to look at possible collusion by officials in Korea Hydro and Nuclear Power (KHNP) . Twenty-two people had been arrested in a previous incident concerning corrupt KHNP procurement processes (WNN, ''Corruption' arrests in South Korea' , World Nuclear News, 12 July 2012). KHNP's CEO has promised to resign once 'the mess' has been rectified. The company will tighten its audit of nuclear components (S. Han, 'Korea Hydro CEO to quit once nuclear parts 'mess' rectified' , Bloomberg, 7 November 2012), though in this case safety had not been compromised. The shutdown could add $276 million in costs for KHNP's sole owner and customer, Korea Electric Power Corp.
In the UK, Confidential Reporting on Structural Safety (CROSS) Newsletter 29 (January 2013) carried five reports on the quality of documentation accompanying products. One advises that steel plates, structural sections and metal castings sourced from outside the UK are, on occasions, supplied with paperwork certifying compliance with the specification, but destructive testing shows them to be non-complying. Another is about 60 mm structural steel pins in a walkway bridge, specified to the uncommon grade S355-J2-G3. The certificate confirmed this grade, and noted that the steel was sourced from outside the UK. But it also noted that the material was tested normalised and supplied as rolled, i.e. the pins did not have the required ductility, did not accord with the certification, and would have to be heat-treated (normalised) to meet the original specification.
In another case, the German government advised that some imported steel products purported to conform to EN 10210-1:2006 and EN 10219-1:2006 were made of unkilled or killed steel, rather than fully killed steel as specified in the standards, and so could only be welded in limited applications, and that welds might fail. A RAPEX notification was issued to this effect.
Steel sheet piles produced outside the UK came with CE marking certificates which were not authentic – the certification company was a 'shell company' with no base in the EU, and other information was found to be false. Fortunately this was spotted before the piles had been purchased. The final case reported by CROSS also involved false CE marking certificates, produced by an organization based in East Asia. Its European office was closed by the authorities, but a website offering certification services still exists.
Cumulatively, these reports (and others by CROSS, e.g. in Newsletter 27) are very worrying. It appears that inspectors cannot necessarily trust certification. In particular, CE marking may not be as robust as traditional third party product certification schemes, such as those run by the UK Certification Authority for Reinforcing Steels (CARES) .
Offsite and onsite inspection and test regimes clearly need to be tighter, particularly for critical (e.g. structural) components. The regimes need to be fully described in the project specification, and must be carefully enforced.