Causality 47: Hyatt Regency Kansas City

16 July, 2022

CURRENT

The tallest building in Missouri with a large atrium perfect for big bands and dancing, hosted a regular Tea Dance in the summer of 1981. When two walkways collapsed killing over a hundred people, the investigators found multiple fundamental design errors. We look at how assumptions, redrafting conventions and negligence led to an incident that has become the case study in how not to do civil structural design.

Transcript available
Chain of events. Cause and effect. We analyse what went right and what went wrong, as we discover that many outcomes can be predicted, planned for, and even prevented. I'm John Chidgey and this is Causality. Causality is entirely supported by you our listeners. If you'd like to support us and keep the show ad- free you can by becoming a Premium Supporter. Premium support is available via Patreon, through the Apple Podcasts channel subscription and through Spotify. Premium supporters have access to high quality versions of episodes as well as bonus material from all of our shows not available anywhere else. Causality is also a Podcasting 2.0 enhanced show with value support and a new listener submitted sound bite option if you'd like to participate. Just visit https://engineered.network/causality to learn how you can help this show to continue to be made. Thank you! Hyatt Regency Kansas City. Construction of the 40-story high Hyatt Regency Hotel in Kansas City Missouri began in May of 1978 and it opened its doors to the public on the 1st of July 1980 and remained the tallest building in Missouri until 1986. The hotel was situated in the Crown Center commercial complex which was part of an urban revitalisation of an older part of Kansas City near Union Station. The hotel had a revolving restaurant, an exhibit hall, conference facilities and more than 700 guest rooms. One of the other widely publicised features of the new hotel was its multi-storey atrium which had three walkways each at a different floor level, each suspended from the ceiling above. The atrium has a large open area approximately 36m (that's 117 feet) deep 44m (or 145 feet) wide and 15m (or 50 feet) high containing a bar, a stage, and a large open space that would sometimes be used as a dance floor. Each walkway was 47m (that's 121 feet) long and was constructed from steel, concrete and glass weighing in at approximately 29 tonnes (or 64,000 pounds) each. The hotel itself was split into two sections: a high-rise section and a function or events block connected by the atrium. Walkways connected those two sections on levels 2, 3 and 4 with level 3 offset from 2 and 4 such that levels 2 and 4 walkways were directly above and below each other, adjacent to the outside wall, with level 3 independently suspended and adjacent to the others, closer to the center of the atrium. The Hyatt Regency Hotel regularly hosted a Friday night tea dance with live music and dance competitions. a tea dance is a European tradition, though also called "Thé Dansant" or "Dancing Tea" in French, was a dance held in the summer or early autumn, late afternoons before sunset. With that background, let's talk about the incident itself. On Friday evening, the 17th of July, 1981 the Hyatt Regency Hotel atrium was once again hosting a regular tea dance with live music that evening played by the Steve Miller Orchestra. The dance was scheduled to go for 3 hours in total starting at 5pm and finishing by 8pm playing a variety of Jazz swing music. At 3pm local time the first people started arriving and within 90 minutes the first floor of the atrium was now fully occupied, forcing people to move to the atrium terrace and walkways before the official start at 5pm. By 7pm the crowd was estimated at between 1,500 and 2,000 people in the entire atrium area. At 7:04pm the band returned and began playing as part of the dance competition. At approximately 7:05pm the 4th floor walkway, with an estimated 20 people standing on it, buckled in the center and began to fall onto the 2nd floor walkway directly beneath it. As the falling walkway impacted the 2nd floor walkway with approximately 40 people on it, the second walkway then also collapsed, with both walkways collapsing onto the ground beneath, near a very crowded area adjacent to the bar. The number of people on the walkways has been impossible to determine with any certainty, as a television crew covering the event were changing batteries in their equipment when the collapse occurred. Water pipes were severed in the collapse and electrical cables were dislodged leaving the lobby in near darkness with frequent sparks from arcing electrical equipment intermittently lighting the atrium. It took 3 minutes for the Kansas City Fire Department to be contacted in the confusion, followed by the Police Department shortly thereafter. By 7:18pm a total of 7 ambulances had arrived at the hotel. By 7:52pm an estimated 100 firefighters and emergency workers were now actively involved in rescuing people from the wreckage. Due to the weight of the walkways, heavy cranes were brought in to remove the debris, with the first walkway removed by 3:15am the following morning. The last survivor was pulled from the debris at approximately 4:30am. 114 people were killed, and 216 were injured. It remains the deadliest, unintentional structural collapse in the history of the United States. Let's talk about the investigations. Yes...there were multiple. The mayor of Kansas City, Richard L. Berkley, formally requested the National Bureau of Standards to independently investigate the most probable cause of the walkway collapse. Their 378 page report examines in great detail the physics behind the failure and makes for interesting reading. In 1983 a grand jury was convened in Kansas City to investigate if the collapse was as a result of any illegal actions of those involved. In 1984 the state of Missouri convened an administrative hearing to determine whether there had been any violation of state licensing laws by those involved. In 1985 the ASCE, that's the American Society of Civil Engineers also held a disciplinary hearing. The relevant findings from all of these will be discussed in due course, but before we get to that let's talk a little bit about the construction of the hotel. The hotel was built in the late 1970s when there was a period of high inflation, high interest rates, and high unemployment. The number of available construction projects at that time was limited and hence the contracting companies at that time priced very aggressively to win work so that they could stay in business. As projects increased in availability, those same companies then pushed to close out their projects they had on the books so they could start on the next project as soon as they could. This era of construction also saw the popularisation of the so-called "Fast Track" method of design and construction. Projects delivered by Fast Track are characterized by construction leading ahead of the final completed design and in the context of civil engineering specifically, structural design precedes architectural's final design. There's reliance on conceptual sketches, preliminary drawings and early component orders with red-lining of shop drawings and multiple re-issues of design drawings during the construction phase. Sounds risky? Well, if you have good quality assurance it can still work and on large projects can reduce the time taken to deliver a project by 25%, with the trade-off of increased cost overall in many cases. Sometimes time is money too so the sooner it's built, the sooner it can make income and those trade-offs are considered to be worth it in the end. So for this hotel construction...who are the players involved? There are a few. Eldridge and Son Construction Company was the general contractor on the project. Eldridge subcontracted the steel work fabrication to Havens Steel Company as the fabricator in December, 1978. Gillum Colico Structural Engineering Consultants Incorporated (or GCE) was selected to perform all structural engineering services for the design and construction of the hotel in July of 1976. Jack Gillum was one of the principals of GCE. GCE subcontracted all structural engineering services for the project to Jack D. Gillum & Associates Limited and designated Jack Gillum as the professional engineer for the project, noting that Jack Gillum was also the president of that firm and hence he was the Engineer of Record. Jack Gillum appointed Daniel Duncan from Jack D. Gillum & Associates as well as the project manager for daily work execution, as well as a project engineer and a senior project designer in late 1976, along with a full project team. Patty, Berkebile & Nelson, Herbert Duncan, and Monroe & Lefebvre Architects Planners Consortium, Inc., (or PBNDML for short) served as associate architects and during 1977 developed the basic design and some of the detailed designs for different parts of the hotel. In June of 1978, Jack D. Gillum & Associates lost both the Hyatt Regency Hotel project's, project engineer and its senior project designer in very quick succession, leaving only the project manager Daniel Duncan with any history of the design decisions that have been made to date. During the construction phase of the hotel, there was an incident regarding the atrium. On Sunday the 14th of October, 1979 during construction, a section of the East atrium roof on the North side of the building above the restaurant area fell four stories onto the lobby floor. Fortunately it happened on a Sunday when no one was working on site and hence there were no injuries. A spokesman for the Crown Center Redevelopment Corporation, the then owner of the hotel under construction, stated that a: "...beam fell because of an installation problem..." Following the incident it came to light that the general contractor Eldridge & Son was fined $5,920 for 17 violations during the hotel's construction. Having said that, none of these issues were related directly to the collapse, however are suggestive of a fast-paced construction process with an inadequate amount of quality assurance. There's quite a lengthy analysis of this specific incident in the show notes by Gregory Luth: "The Chronology of Hyatt Regency Collapse" if you're interested. The key point is that the so-called installation problem was actually more of a lack of a cohesive design problem that during installation was worked around, rather than questioned and fixed properly. Had the designers taken that opportunity to stop the job and review the design from end to end they would have more than likely uncovered the walkway design issues we're about to go through...but they didn't. Let's talk about the walkway design. The walkway is comprised of four spans in total with each span denoted by its interconnecting joints and hanging rods starting at 7 through to 11. Hence the first span is designated Span 7-8 then 8-9, 9-10 and 10-11. Before you wonder why the numbering starts at 7...the numbers line up with the building's column numbers, hence the walkways hung between column 7 and 11 of the overall building space. Because the walkways on levels 2 and 4 were joined together by hanging rods, the 4th floor walkway was referred to as upper and the 2nd floor as lower. Finally the hanging rods were positioned on either the East or West side of the walkway. For example the eastern centre-most rod connection point on the 4th floor was denoted 9UE. The box girders connecting the walkway to the hanger rods were made of 200mm wide (that's 8") of c-channel, welded together along their length hence creating the so-called box. How heavy was it though? Because during the investigation they found significant variability in the thickness of the concrete decking and topping materials, so they confirmed the final weights by weighing several walkway spans as the design drawings did not align to what was constructed in that regard. Analyzing the fallen walkway sections led to an estimate of the mass of the walkway to within a worst case error of +/- 136kg (or 300 lbs). The addition of gypsum board to meet fire endurance requirements added approximately 1,050kg (that's 2,310lbs) to each walkway span. We'll get to that shortly. The investigators therefore concluded that the average final weight of a single walkway span as built was 8,050kg (that's 17,750lbs) which exceeded the as-designed span weight by about 8%. Not a good start. So what went wrong? We know that the walkways were 8% heavier than their design but that's only one issue. It was determined that the most likely connection point that failed first was 9UE with transferred load to spans U8-9 and U9-10 on the East connections, then causing 8UE and 10UE to rapidly fail with the other side of the walkways joint 9UW failing as well. The method of failure was that the nut and washer pulled through the center of the box beam under high load. Given the construction method of welding x2 c-channels together, the investigators spent significant time confirming that the quality of the welding job on those channels was not a factor in the incident. In addition, whilst there were people present on the walkways it was found in the investigation that the mass of the people on the walkways was not a significant contributing factor to the failure that they saw in the incident. The two centre-most spans then rotated downwards pulling span U10-11 off its bearing seats at the building column number 11. As the upper walkway then fell onto the lower walkway, it followed a similar failure progression. The investigation determined that there were two design changes that contributed to this incident, as well as a design gap. We'll talk about each in turn. The first: Fire-proofing in March, 1978. A routine design review by the Codes Administration Office, Kansas City Public Works Department resulted in a 6 page handwritten assessment of the fire endurance of the walkways structural steel, dated the 10th of March, 1978. A meeting was held on the 16th of March, 1978 between representatives of the Codes Administration Office and the architects PBNDML with 2 pages of minutes which included an agreement to clad the walkway structural steel with gypsum board to address fire resilience concerns. Specifically x2 layers of 16mm (that's 5/8") thick gypsum board supported by 0.8mm metal studs and nailing strips, then attached to the structural steel by power driven fasteners. The structural drawings were reissued on the 30th of March, 1978 and it is unclear in the documentation at the time if any changes were made to the structural details as a result of gypsum board being added to each walkway span. The investigation concluded that the walkway design did not get adjusted to account for the additional load from retrofitting gypsum board as a fire retardant material during design. That said, they also concluded that even if the load from the gypsum board wasn't present, there would still have been insufficient redundancy (also called reserve load capacity) in the hangar rod design to resist the failure of a single rod. Hence, if one breaks, they'll all break. The second design change, also in March 1978, related to the hangar rod connection. In March, 1978 drawings for the walkway hangar showed a single rod continuously threaded from top to bottom, that was suspended from the beams on the atrium ceiling, passing through the fourth floor walkway and second floor box beams with securing nuts and washers at each walkway level. In early January, 1979 the drawings were now at the fabricators, Havens Steel Company, and their engineering manager called the structural engineers project manager. Havens engineering manager raised concerns about the durability of a continuously threaded rod over that length during construction. The original design called for a nut 6.1m up the hanger rod and didn't use sleeve nuts. That, coupled with the higher cost of manufacture, they proposed to change the design from a single continuous rod to two rods instead, with threads on each end. According to Havens testimony the structural engineer checked the turning moment and the sheer force presented at the box beam for the offset condition while on the phone and Daniel Duncan accepted that change over the phone. The structural engineer then asked the fabricator to submit the change request through the "normal channels for approval." The shop drawings were then marked up to now have two hanger rods with a 102mm (that's a 4") inset at the fourth floor walkway connection point, in place of the original single continuous hanger rod arrangement. To put it another way, the ceiling to upper walkway rods were on the outside and the upper to lower walkway rods were offset on the inside of those. On the 12th of January, 1979 Havens Steel Company pulled the Hyatt Regency Project out of their engineering department and then subcontracted the for-construction drafting of the partially completed shop drawings to an external engineering firm. This was due to Havens winning a large project and they required their engineering resources to be focused on the new job instead. The outsourced drawings included the box beam detail but no rod connection was shown in its detail, which unfortunately is common practice in the industry. Different drawings show different details based on their final intended use cases. To understand the overall design of the connection configuration to the box beams, a person would need to consider all of the drawings and specifications collectively... not individually. The offset rod design change was therefore never submitted for a formal review in its entirety and the final connection to the box beams was only implied on the shop drawings. They assumed that the connection design was complete and specified elsewhere, so they simply added the "weld" symbol at the connection point, before sending the drawings out for final approval. Ordinarily if there are any design changes required to the shop drawings they would be submitted under a separate cover sheet along with a request for engineering validation of the shop drawings. This did not occur. On the 7th of February the outside detailer made their final consistency and completeness checks and the drawings were sent to the structural engineer for engineering approval. With various communication delays the drawings didn't reach the engineer's desk until Thursday the 16th of February and the contractor had requested expedited approval due to the construction timetable pressures under the contract. At that time both the project engineer and designer had long since departed and the project manager Daniel Duncan was too short of time to review the drawings himself so he delegated their checking to a senior technician. The senior technician was trained in structural engineering and had two decades of experience but wasn't licensed as an engineer. Whilst he queried about the rod specification (more on that in a moment) there were no other significant findings and the drawings were returned approved for construction on Sunday the 27th of February. That's right...they were working the Sunday to get it done. This modification essentially doubled the load transferred from the fourth floor walkway box beam to the holding nut for the upper hanging rod from 91kN to 181kN. Another way to describe this is the nut holding the weight of the fourth floor walkway in the original design only had the weight of its own floor, in that segment, applied to it with a continuous rod. Once a second rod was added, instead of the load from the lower walkway being transferred up through the rod to the ceiling, since it was hanging from the upper walkway now (not a common rod) the entire mass of the lower walkway segment was applied to the upper walkway as well. Therefore the mass on the holding nut on the fourth floor walkway was effectively doubled. Having said all of that, with the correct redundancy factors applied and under the original design, accounting for the correct weight of the walkways as they were built, the minimum load value needed to be 151kN even with a single rod design, and this still didn't meet the AISC that's the American Institute of Steel Construction specifications requirements. Specifically the AISC Specification for "Design, Fabrication and Erection of Structural Steel for Buildings" as well as the "Code of Standard Practice for Steel Buildings and Bridges" as required by the Kansas City Building Code. Another key problem was the design of the walkway support beam's connection to the hanger rod. The original design sketch for the rod connection showed a bracket extending off of the web of the W8 (that's a Wide flange I-Beam). W8 (that's the vertical section of the beam where 8 in W8 is the height of that beam in inches) and that had an indicative spacing for the rod connection with the rod sizing and forces all drawn on the engineer's sketch. In March, 1978 the project manager replaced the W8 beam at the hangar location with a pair of 8" channels we previously spoke about, turned toe to toe, and extended these past the interconnecting W16 beam. This was done to eliminate concerns about the eccentricity of the original suggested bracket connection with the I-Beam. The revised design was in effect symmetrical where the original design wasn't. The revised engineers sketch also showed the rod size force and rod grade. When the draftsman transcribed the engineer's sketch to a for-construction drawing, they omitted the note specifying the strength of a 413MPa (or 60kips) for the hanger rods. In civil engineering in that era, writing the load detail on the sketch indicated to the fabricator that the connection design still needed to be completed, and this was normally part of the fabricator's scope of work. The engineer had assumed that the fabricator would complete the design detail for the connection to the box beam hanger. It is not clear why the additional details were not transcribed by the draftsman, however the fact that they weren't, led to a mis-assumption by the fabricator about whether the connection design was actually complete. They believed it was, when it wasn't, and equally the engineer believed it was being handled by the fabricator when it wasn't. Placing the rod through the weak point of the welded c-channels may have looked the most pleasant to the eye however the nature of c-channels is that they taper from each corner to each tip furthest from the back of the "C" such that the tips are the narrowest parts of the channel. Placing those tip to tip and then running a weld seam along them to join them together and then grinding that weld flat for assembly, resulted in a very weak connection. The welders on site recognised this and attempted to weld the inner join line inside the box section, but they could only reach as far in as the end of their welding rod allowed them to. Hardly precision work and pretty much unconstructable in that way...laws of physics and all that. A final note though about the rod design change that also occurred in March 1978. The architect requested the rods be changed from 44.5mm (that's 1-3/4") to 32mm (or 1-1/4") to "lighten up" the appearance of the bridges. The original drawing showing the rod detail did not specify the material on the drawing though the other specifications suggested standard A36 grade rods (that's 36ksi tensile strength) should be used based on the original design. To achieve the load-bearing requirements for the narrower rod, a grade 60 rod (therefore with a 60ksi tensile strength) would be required and this was marked up on the revised engineer's sketch. Just a quick note...I talk about minimum yield strength... that's the amount of stress a material can withstand before it succumbs to a permanent deformation and doesn't return back to where it started. In imperial units it's usually expressed in ksi which is kilo pounds per square inch of pressure. Metric units is normally in MPa or Mega Pascals of pressure. If you recall, the final drafting for the for- construction drawing set was outsourced by the fabricator to another engineering firm that had a long-standing business relationship with the fabricator. During the course of checking the drawings the technician from the outsourced back-drafting company involved asked the project manager about the strength of the hanger rods. The technician's own calculations found that 36ksi steel rod would not fully support the load indicated on the other available drawings. The project manager responded from memory that it was a high strength rod however no mention of the specific rod grade was provided. Further than that, no attempt was made to verify this by looking at a complete set of drawings and specifications and the drawings were issued without this information detailed on the for- construction drawing set. Some analyses of this incident call this out as a cause of the incident but that's not really true as the rods themselves didn't actually fail. Having said that, I have no doubt at all that had the box beam joints not failed first with the passage of enough time and cumulative deformation of the rods eventually would have led to a collapse in the future. When exactly that could have happened who can say? Technically...not a cause of this incident but clearly an error that could have led to a different incident had the other issues not beaten them to it first. Let's talk a little bit about the legal fallout. In 1984, Missouri's Board of Architects, Professional Engineers and Land Surveyors commenced disciplinary proceedings against Duncan, Gillum and GCE. After a 27-day hearing and weeks of compiling their report the commission eventually issued its findings which were 442 pages long and found all three parties grossly negligent and revoked their licences. Judge James B. Deutsch, an administrative law judge for Missouri's Administrative Hearing Commission found the structural engineers guilty of gross negligence, misconduct and unprofessional conduct. They claimed that the design flaws had resulted from a mis-communication between Jack D. Gillum & Associates and the Havens Steel Company. Mr Duncan, Mr Gillum and GCE unsuccessfully appealed their decision up to the Missouri Court of Appeals. The Court of Appeals also rejected the more substantive attacks on the sufficiency of the evidence in the initial proceedings, stating the following: 1) "...Mr Duncan was responsible for designing and approving the building structure..." of which the walkways fell within that scope; "...[T]he walkways offered a potential of great danger to human life if defectively designed..." 2) "...Mr Duncan approved the fabricator's change, recommending it to the architect, and approved shop drawings reflecting it without confirming its acceptability; (noting that) the change effectively doubled the box connection load..." and 3) "...Mr Duncan never reviewed the shop drawings even though such a review is an 'Engineering Function' that even GCE's in-house policies required that he do..." The Court of Appeals upheld the original conclusion that: "...[t]he conduct of Duncan from initial design through shop drawing review and through the subsequent requested connection review... supports the Commission's finding of conscious indifference to [non-delegatable] professional duty." The Court of Appeal also upheld the original conclusion that Mr Gillum had displayed gross negligence because he failed as the engineer of record: "...to assure that the Hyatt engineering designs and drawings were structurally sound...prior to impressing their upon his seal..." and failed "...to assure adequate shop drawing review." During the trial the detailer, architect, fabricator, and technician, all testified that during construction they had contacted the project engineer regarding the structural integrity of the walkway connection detail. They claimed that each time it was raised the project engineer assured them that the connection design was sound and whilst he claimed to have checked the detail, in reality there was no evidence found that he had ever performed any calculations at all. The legal repercussions for the Hyatt engineers firmly established the Engineer of Records responsibilities for the structural integrity of an entire building during construction...including the shop drawings. Hence in the civil engineering context, the Engineer of Record should design and detail all non-standard structural connections. More broadly all new designs need to be thoroughly checked and all modifications to design details require a formal written approval from the Engineer of Record. The day before that decision was handed down by Judge Deutsch, the American Society of Civil Engineers had announced a new policy of holding structural engineers responsible for all elements of structural safety in the buildings they design. Let's talk about the aftermath. The hotel reopened 3 months after the incident, after $5M USD of reconstruction. A single walkway was reinstalled, only this time supported from beneath by concrete pillars. Other than the 3rd floor now being left without a connecting walkway the lobby generally retained its original design. The hotel was renamed the Hyatt Regency Crown Center in 1987, but when Starwood took over the hotel in November of 2011, it was renamed the Sheraton Kansas City Hotel at Crown Center. Mr Gillum and Mr Duncan had their engineering licences revoked, as did the company GCE, and the two had their memberships to the ASCE revoked as well. GCE survived only by merging with a Denver engineering firm: Ketcham, Konkel, Nichol, Barrett. No criminal charges were filed and no one was prosecuted. Compensation claims from the courts awarded victims around $140M USD paid by the Crown Center Redevelopment Corporation, the owner of the hotel, with multiple significant insurance payouts on top of that. In 2015, over three decades later, a memorial was finally built adjacent to the hotel and was paid for by the Skywalk Memorial Foundation. It had taken the foundation a decade to raise the $550,000 USD to pay for the memorial and associated maintenance endowment. The names of each of the victims are etched in the memorial of which the centerpiece is a six meter or twenty feet high abstract sculpture of a couple embraced as they dance. So what do we learn from all this? There are three points I'd like to go over here. Responsibility of the engineers during design; Complete design reviews rather than piecemeal; and personnel churn. The Engineer of Record is a commonly used term in North America which has in the past and particularly prior to this incident, had varying interpretations as the ultimate responsibility of that job role in a construction project. The Hyatt Regency walkway collapse has become the legal precedent against which the role of EoR has ever since been defined in the USA. The buck stops with the Engineer of Record. They are responsible. In Australia we have legal requirements under the Professional Engineers Act and for many engineering disciplines today, if you aren't a Registered Professional Engineer in the state you're operating in and/or a Chartered Professional Engineer in Australia then you aren't allowed to sign off on a design drawing. I've been an RPEQ in Electrical Engineering for 17 years and also an RPEQ in ITEE and a Chartered Professional Engineer in both as well and I take those responsibilities extremely seriously. When I stamp and sign a drawing it means I've checked every detail of that drawing to the best of my ability which is in effect what the Engineer of Record is required to do as well. When I'm asked to RPEQ a drawing, there's also additional requirements. I need to have been involved with the design from start to finish...an aspect that was lacking in the case of this specific incident. I've been involved in projects where we've spent significant amounts of money subcontracting to find the right RPEQ from other organizations in order to get the design correctly validated after that specific engineer had left the design organization mid-project. We tracked them down to their new job and we got them back. Design reviews for complex systems aren't as simple as you'd like sometimes, especially when we fractionalise our designs. I like to call it "Fractional Engineering." The idea that reviewing the design or change as a whole just takes too long or has too many hurdles when it's considered in its entirety, so we break it into fractions of the whole and review those, one fraction or one slice at a time...in isolation. There's often too much detail to put on a single drawing or in a single specification which means the engineer needs to often read, ingest and comprehend a significant number of drawings and information in order to determine if the design in its entirety is safe and correct. That takes more time and as they say, "It takes as long as it takes, if you have to do it right." The time pressures and distractions from other elements of the building's construction clearly affected the outcome in this incident. "It's just a walkway...I have bigger concerns..." I have no doubt that kind of mindset was a factor. Finally though... personnel churn. It's insidious and it comes from a belief that people can be plug and play. "We're all just numbers." I've worked for companies that push hard on documenting everything...turning everything into a process, with evidence provided for every single step and then when someone leaves the organization the management layer hires someone new and think everything can just pick up from where it was...no risk, no problem. Thinking is...any new person can just read all the documents, comprehend the design and finish the job. If only it worked that way. It is simply not possible to write down every last detail, every conversation, every side note, every corridor conversation, site meeting, inspection note, and even if it was...could the next person that wasn't present early in the project then absorb all of that information and comprehend its importance and context in any reasonable amount of time? Enough to finish the job without an incident? It's complete fiction. We are not plug and play. We are the sum of our experiences and we are unique. So few things in engineering are truly copy and paste widgets and even if they are the larger the scale, the differences location and materials makes can vastly impact the outcome. It's so important to keep the core team consistent in projects and if that means paying more, giving people what they want to get them to stay, that's not pandering, that's a long-term vision about what's the best for the project and the company overall and that's what it means to be a good manager. Not trying the sell of plug and play, "We'll roll on, it'll be fine..." that's not management. Building and maintaining a good, stable, competent team that communicates well together makes an enormous difference. Creating time and space to execute thorough design reviews does too. Somewhat of an unusual footnote however. When something traumatic like this occurs to a person, they can handle it in a multitude of ways. We saw the tragedy of the Challenger Space Shuttle engineers with Bob Ebling retiring from engineering, unable to practice, consumed by guilt, although admittedly he didn't cause the Challenger explosion but he did fail to stop the launch, but in this case something altogether different happened. Jack Gillium was born on the 21st of November, 1928 and graduated from the University of Kansas with a degree in Architectural Engineering before being drafted into the army to serve in the Korean war between 1950 and 1952. In the early 1960s he found his own structural engineering company, Jack D. Gillum & Associates and was known for many prominent engineering designs around the world until Hyatt Regency. Whilst he initially fought the allegations against him he ultimately came to a different conclusion about his involvement, and accepted that he was in fact, partly to blame for the incident. As he lost his engineering license he instead decided to spend the rest of his working career advising, teaching and publicly speaking about how he and his company had got it wrong and how we can all learn from it. He wrote multiple papers including the excellent: "The Engineer of Record and Design Responsibility" paper that's linked in the show notes. He passed away on the 4th of July, 2012 aged 83 years old. I've gone a bit back and forth on how I feel about him but in the end Jack Gillum could have gone into seclusion, wracked by guilt, but he chose to stay out in front and essentially do exactly what I'm doing here with Causality. He tried to educate people about how it can go wrong as it had for him personally in his case, to try and prevent future incidents from occurring. And that, on balance, I think that was the right call to make, and I'm glad he chose that in the end over alternatives. At the risk of being a touch bit flippant, the old and mostly untraceable saying is: "When you assume, you make an ASS out of U and ME." The "Fast Track" method of construction employed here led engineers and fabricators leaning on assumptions out of a lack of available time with horrible results. For those engineers listening to this now, whether you're reviewing a design, even if you're not the EoR, or your CPEng or RPEng aren't on the line, take that review with the same level of vigor as if they were. Gather the information you need to be sure if the design isn't clear and push back... if you need more time...you NEED MORE TIME! If you need to speak with the original designer...find them! Ask them. Get clarity. Get it right. Being an engineer can be a stressful job but can also be a rewarding one too. I regularly think back over my 25 year career to date and I wonder if someday an error I made in the past, might turn out with negative consequences - ones that I didn't foresee at the time in the past. I try not to dwell on it but that's not really the point. I use those thoughts, those fears, to keep me sharp. A bit of fear isn't a bad thing. It's a reminder to take what I do, what many of us do for a living, very, very seriously. Had the structural engineers and fabricators done the same on this project, those 114 people might not have died. Make the time. Do better. If you're enjoying Causality and you'd like to support us and keep the show ad- free you can by becoming a Premium Supporter. Premium support is available via Patreon, through the Apple Podcasts channel subscription, and through Spotify. Just visit https://engineered.network/causality to learn how you can help this show to continue to be made. Thank you. A big thank you to all of our supporters, a special thank you to our Silver Producers: Mitch Biegler, Kevin Koch, Lesley, Shane O'Neill, Hafthor, Jared, Bill, Joel Maher, Katharina Will and Dave Jones, and an extra special thank you to both of our Gold Producers: Steven Bridle and our Gold Producer known only as 'R'. Causality is heavily researched and links to all materials used for the creation of this episode are contained in the show notes. You can find them in the text of the episode description of your podcast player or, on our website. Causality is a Podcasting 2.0 enhanced show and with the right podcast player you'll have episode locations, enhanced chapters, and real-time subtitles on selected episodes and you can also stream satoshi's and Boost with a message if you like...there's details on how along with the Boostagram leaderboard on our website. Causality now has support for listener submitted sound bites from any episode! You can create and Email in for inclusion on the website and in Podcasting 2.0 compliant apps. If you'd like to make a credited submission visit https://engineered.network /createsoundbite/ to learn more...there's also a link in the notes. You can follow me on the Fediverse @chidgey@engineered.space on Twitter @johnchidgey or the network @Engineered_Net. This was Causality. I'm John Chidgey. Thanks so much for listening.

Soundbites
Many Assumptions
Fractional Engineering
Get Clarity

Duration 41 minutes and 31 seconds

Show Notes

This show is Podcasting 2.0 Enhanced

Create your own Credited Soundbite

Links of Interest:

Engineering Analysis:

News Articles:


Episode Gold Producers: 'r' and Steven Bridle.
Episode Silver Producers: Mitch Biegler, Kevin Koch, Shane O'Neill, Lesley, Hafthor, Jared, Bill, Joel Maher, Katharina Will and Dave Jones.
Premium supporters have access to ad-free, early released episodes with a full back-catalogues of previous episodes
SUPPORT CAUSALITY PATREON APPLE PODCASTS SPOTIFY PAYPAL ME
STREAMING VALUE SUPPORT FOUNTAIN PODVERSE BREEZ PODFRIEND SPHINX TRIBE
CONTACT FEEDBACK REDDIT FEDIVERSE TWITTER FACEBOOK
LISTEN RSS PODFRIEND APPLE PODCASTS PANDORA GOOGLE PODCASTS INSTAGRAM STITCHER IHEART RADIO TUNEIN RADIO CASTBOX FM OVERCAST POCKETCASTS CASTRO GAANA JIOSAAVN AMAZON YOUTUBE

People


John Chidgey

John Chidgey

John is an Electrical, Instrumentation and Control Systems Engineer, software developer, podcaster, vocal actor and runs TechDistortion and the Engineered Network. John is a Chartered Professional Engineer in both Electrical Engineering and Information, Telecommunications and Electronics Engineering (ITEE) and a semi-regular conference speaker.

John has produced and appeared on many podcasts including Pragmatic and Causality and is available for hire for Vocal Acting or advertising. He has experience and interest in HMI Design, Alarm Management, Cyber-security and Root Cause Analysis.

You can find him on the Fediverse and on Twitter.