Logo, Alpine Research & Engineering, LLC

(415) 265-9525
1976 NW Shevlin Crest Drive, Bend, OR 97703

PAST INVESTIGATIONS


A few past incident investigations are briefly described here.  They illustrate the variety and depth of investigations performed by Dr. Kimble J. Clark.  In some, Dr. Kimble J. Clark was the Principal Investigator.  In others, he was the Project manager or a member of a large multi-disciplinary team of experts.  In the interest of brevity, only a few aspects of these incidents are provided here, and numerous technical details are omitted. Please Contact Us if you are interested in hearing more details on any of these cases.


PROPANE DEASPHALTING UNIT

A fire erupted in a propane deasphalting unit at a large refinery.  The fuel for the fire was propane which leaked under high pressure through a crack in a piping elbow located in a process flow control loop.  This pipe crack resulted from a brittle fracture that grew in many initiation-arrest-reinitiation cycles driven by variations in process flow conditions.  The location in the pipe where the crack occurred should not have been exposed to process flow, but a failed block valve allowed this to happen.  Four operating personnel were injured and property and business losses approached $200 million.



PROPYLENE SPLITTER TOWERS

A major explosion and following fire occurred in a large refinery, centered around the propylene splitter towers.  A pump can in this unit ruptured allowing propylene under pressure to escape and form a heavier-than-air vapor cloud which grew over a several-minute period to approximately 800 feet in diameter. At this point, some of the vapor was sucked into the combustion air intake of a burner on a steam boiler, and the vapor ignited with a flame front traveling back out into the propylene vapor cloud and causing overall ignition.  The magnitude of the unconfined vapor cloud explosion was one of the largest ever encountered by Dr. Kimble J. Clark, with blast wave damage occurring to structures, tanks, and buildings throughout the refinery and outside the refinery. The pump can failed due to design and manufacturing defects.  There were numerous plant personnel injuries, but no fatalities.  Property and business losses totaled approximately $385 million.



AMMONIUM PERCHLORATE MANUFACTURING PLANT

One of the largest accidental explosions to ever occur in the U.S. involved an ammonium perchlorate manufacturing plant.  Ammonium perchlorate (AP) is a major ingredient in solid fuel rocket propellant, and is both a fuel and an oxidizer all in one chemical compound since it contains both hydrogen and chlorine as fuel and over fifty percent by weight oxygen as the oxidizer.  This explosion was initiated by a welder’s torch which started a fire at the plant.  The fire spread quickly to eventually engulf several million pounds of AP stored at the plant.  This stored AP cooked off and resulted in a series of five explosions, with the final explosion being the largest. There were two plant personnel fatalities and many injuries to both plant personnel and fire fighters.  The plant and a nearby commercial establishment were totally destroyed, with property and business interruption loss totaling $100 million.  The cause of this incident was a combination of poor plant housekeeping which allowed the welder’s torch to ignite the initial fire, and improper storage onsite of millions of pounds of AP which sympathetically detonated after being heated by the initial fire.



AMMONIUM NITRATE FERTILIZER MANUFACTURING PLANT

Another very large explosion involving a condensed-phase explosive material similar to AP occurred at an ammonium nitrate fertilizer manufacturing plant.  The explosion occurred when a large cylindrical vessel called the neutralizer, which mixes liquid nitric acid and gaseous ammonia to produce aqueous ammonium nitrate solution, was offline but contained product which was being steam heated to prevent it from freezing during winter conditions.  The neutralizer product began to thermally decompose via exothermic (heat generating) reactions over a several hour period, eventually escalating to a detonation.  The detonation completely destroyed the plant and caused four operating personnel fatalities and over a dozen injuries.  Property and business interruption losses approached $400 million.  The root cause of this incident is still in technical dispute to this day, but the expert team Dr.KimbleJ. Clark worked with concluded it involved a design defect in the neutralizer.




NATURAL GAS EXPLOSION

A natural gas explosion occurred in a single-family residence under construction.  The source of the gas was a leaking buried gas main in the adjacent street.  The fugitive gas migrated to the house through buried electrical conduits running from an above-ground connection box at the street to the meter box at the house.  The natural gas was not odorized so that its accumulation inside the house was not detected by workers there at the time of the explosion.  The ignition source was a lit cigarette, and the resulting explosion caused major structural damage, killed four of the workers, and seriously injured a fifth worker.




PROPANE EXPLOSION

A single-story factory-built house, set on a concrete block foundation with a crawl space three to four feet deep, was completely demolished by a propane explosion.  This incident was the most damaging accidental house explosion Dr. Kimble J. Clark has investigated.  The residence was supplied with propane from an outside above-ground tank.  A leak in the black-iron piping in the crawl space, which routed propane to a furnace and a water heater, was the likely source of the fugitive propane.  This heavier-than-air propane mixed with air and gradually pooled throughout the crawl space.  The resulting explosive mixture was most likely ignited by the furnace which turned on in response to a call for heat.  The subsequent explosion completely pulverized the house and blew the structure fragments and other contents outward in all directions.  Amazingly, the sole female occupant survived, but was severely injured.




NATURAL GAS FIRED RESIDENTIAL FURNACE

A new natural gas fired residential furnace with a special combustion burner design for low NOX emissions was found to be causing fires in numerous homes.  These furnaces were typically installed in the attic where smoldering combustion could occur with out being noticed by the occupant.  These smoldering fires involving wood ceiling joists and the plywood platform that supports the furnace would eventually break out into flaming combustion.  The smoldering combustion was initiated by overheated sheet metal comprising the enclosure of the funace unit.  Extensive analyses of the burner design and testing of exemplar furnaces revealed that the overheated sheet metal was caused by a design defect in the combustion chamber.  In an attempt to lower the NOX emissions from the gas fired burners, a patented technology called “NOX rods” was retrofitted into the old burner tray design without any further modifications to the remaining components in the combustion chamber.  As a result, the balance of convective versus radiative heating of the heat exchanger above the burners was drastically altered, from originally convection dominated (98%) to 50% convection and 50% radiation.  The radiation component of the heating impinged on the inside surfaces of the external sheet metal, thus causing it to be overheated.  The excessive radiation also caused metallurgical failure of the burners and heat exchanger after a relatively short period of operation.  Extensive and lengthy litigation ensued, and eventually completely new integrated burner designs were brought into the market while the old “NOX rods” units were removed.




COAL-FIRED BOILER

In a 500 MW natural-circulation coal-fired boiler at a western powerplant, approximately 20,000 feet of waterwall tubing was found to be severely corroded internally.  These tubes had the internal ribbed design, and the corrosion was located on the hot side of the tubes which face the combustion fireball in the furnace. Remarkably, no throughwall ruptures with steam leaks had occurred, but major portions of the waterwall paneling had to be replaced since ruptures were inevitable due to the rapid rate of tube wall thinning.  An extensive engineering investigation was undertaken, which eventually revealed that the ribbed tube metal temperature was running much too high on the hot side, especially at full-load operating conditions.  This in turn was causing accelerated under-deposit corrosion inside the tube, associated with the sodium phosphate boiler water treatment program in place at the time.  The cause of the excessive tube metal temperature was water/steam mixture flowrates which were too low to adequately cool the tubes at full load operation.  The low flowrates allowed higher-temperature film boiling to occur inside the waterwall tubes, when in fact the original tube metal alloy was selected assuming lower-temperature nucleate boiling inside the tubes.  As a result of litigation,the boiler manufacturer agreed to replace the furnace waterwall panels, redesign the waterwall circulation, and alter the furnace heating profiles.