Steam Accidents & Forensic Investigations

Wayne F. Kirsner, P.E.
Principal / Kirsner Consulting Engineering
MS/Physics/1973/Georgia Tech
BS/Physics/1972/Georgia Tech
BSME/1980/Georgia Tech

Professional Engineer/Mechanical/1983

ASHRAE Distinguished Lecturer:
June 2001 through June 2009

Feb. 1997 investigating valve explosion at Ft. Wainwright, Alaska

Steam Investigations & Accidents, 2017 - 1983

For Newcrest Lihir Mining located on Papua New Guinea, April 2017: Investigated waterhammer incident in the steam piping coming from one of three Autoclaves during warm-up of the Autoclave when cold water was introduced to an isolation loop for instrumentation. The incident startled the Operator but did not rupture any piping component. The incident motivated Operations to invite me to site to investigate the accident and do two 8-hour waterhammer seminars for Supervisors, Engineers, and operators. I left a power point slide presentation for management explaining their incident along with another potential incident found during a plant walk-thru.

Autoclave slide of waterhammer

Remains of 90 gallon non-ASME SS Tank blown apart by BLEVE.

July 2009, Expert for Spence Engineering (who was not responsible for this Accident) in law suit arrising from the BLEVE of the stainless steel tank shown at right. For a hot water BLEVE to occur: (1) the tank must be pressurized so water can be superheated with respect to atmospheric pressure, (2) there must be steam atop the water in the vessel, and (3) the vessel must suffer a steam release through a crack, hole, or perhaps a safety valve relieving. Thus, the tank must generally be pressurized beyond it's rating, or in a weakened condition.

The water inside this 304 SS tank non-ASME which was supposed to operate at no greater than 15 psig (but was tested to 100 psi by the manufacturer) was inadvertently overheated to a temperature approaching 288oF by uncontrolled application of 41 psig steam to water passing through a heat exchanger. (An inoperative control convinced us the temperature limiting control on the heat exchanger must have been bypassed). A small water leak at a pump seal allowed water to leak out of the closed system without being made-up by the tank's water regulating valve. That's because the pressure in the tank exceeded the pressure setting of the regulator. Thus steam filled the void left by the leaking water.

Stress Corrosion Cracking concealed by insulation severely weakened the tank so it could not withstand the pressure of the hot water. We believe a crack developed near the top of the tank which released steam so that a rapid pressure drop initiated the the BLEVE which proceeded to blow the tank apart causing an instantaneous total loss of containment. The superheated water in the tank, upon release, partially flashed to steam immediately releasing expanding vapor which created a shock wave in the air of the surrounding space. A nearby worker was knocked to the floor and badly scalded by the hot water entrained in the steam blasted from the tank. A 75 psig safety relief valve provided atop the tank never actuated.