Most of the world's trillions of watts of electrical power is generated by steam from the burning of coal and the fission of uranium. The steam exiting the turbines contains a large amount of energy from which it is impractical to extract electrical energy. In the case of such densely-populated cities as New York, the exhaust steam may be piped to buildings where it is used for both heating and running turbines for air conditioning. This use is known as co-generation, a hot topic for environmental advocates.
Scene of the Crime
The thermal expansion coefficient of steel is about 10 millionths of an inch, per inch, per degree Celsius. This sounds trivial until multiplied by a couple of hundred degrees and the some 10,000 inch height of a skyscraper. Fix the bottom of a steam riser and the top will bob up and down by several feet as the temperature is cycled. Such motion is intolerable, so expansion joints are inserted. In the current case the joint was a stainless-steel bellows about 24 inches long by eight inches in diameter which could accommodate about five inches of expansion. The pipe was fixed on the 24th and 37th floors, with the joint installed on the 36th floor.
The design was fine, but the execution went terribly wrong. The joint split and 360F steam spewed into a business office with tragic results. Huge lawsuits resulted.
An MIT colleague in mechanical engineering and I were retained and flown to New York by one of the plaintiff lawyers. The evidence was sequestered and examined by a New York City Board of Enquiry before people like myself were allowed to see it. The board seems like a good idea, but much of the mounting hardware, plus several coupons cut from the bellows, simply disappeared. Such disappearance is outrageous. A private consultant responsible for such a loss could easily be ruined by a lawsuit and lose his Professional Engineer's license. My examination was further impeded by the crush of people in the examination room. I associate such mobs with something like the King Tut exhibit, not the examination of a piece of leaky stainless steel.
The failed joint was in bad shape. The top and bottom flanges were parallel, but had a lateral offset of about eight inches. The joint had a large circumferential rip that had vented the fatal steam. It also had a smaller circumferential crack and two longitudinal cracks. One of the latter showed discoloration due to having leaked steam for weeks or more. Unable to do any sort of microscopic study on the bellows material, as the samples taken by the board had disappeared and I was not allowed to cut any new ones, all I had to go on was the unaided eye.
I concluded, very tentatively, that the joint had been installed in a skewed configuration, so the normal few inches of expansion and contraction would cause the observed cracking. But why do such an installation? I suspected the pipes above and below the joint did not match up, and the joint was skewed to absorb the discrepancy. Such installation implies dereliction of duty by a number of people, including the professional engineer, who signed off on the job.
Why didn't someone notice the leaking steam? The city insisted the joint be in a sealed compartment that prevented any kind of routine inspection. Only the bureaucratic mind could fathom the reasoning behind such a rule.
I later received a report of an earlier inspection stating that the welds at the upper floor had been torn loose and raised about four inches and deflected laterally by the same amount. Also, an expansion joint on a lower floor had guide channels to preclude lateral movement, whereas the subject joint did not. This later information supports my hypothesis of skewed installation. The guide channels could hardly have been installed on a skewed joint. Also, the skewed bellows could not compress readily during heating and therefore exerted more than the expected force on the welds, probably causing their fracture.
I was paid and heard nothing more of the case. Perhaps some readers can recall this 30-year-old case and add to the discussion.