Inspector report: Antique boiler explosion in Medina, Ohio

Editor's note: This information is being reprinted in its entirety and as it was received from John D. Peyton, director of the Certified Boiler Engineers for the Commonwealth of Pennsylvania.

I am the director of the Certified Boiler Engineers for The Commonwealth of Pennsylvania and have conducted an inspection of a boiler explosion of an antique boiler at Medina, Ohio. This inspection and evaluation was conducted at the request of Medina County Sheriff Neil F. Hassinger and Lt. John Detchon. (This explosion occurred on July 29,2001)

I was briefed on Aug. 6, about [1 p.m.], by Lt. Detchon and he informed me that the safety relief valve had been removed by Ohio Chief Boiler Inspector Dean Jagger and sent to the National Board for inspection and testing. He stated that the seals were intact on the valve and that the National Board test lab's initial report stated that the valve did not lift with pressures up to 200 psi. This exceeded the set pressure of the valve of 125 psi. 

The pressure indicator was also tested in the same lab and found to indicate 25 psi. lighter, meaning the pressure could be 25 psi higher than the indicator reads. Chief Jagger also removed part of the damaged "crown" sheet and the fusible plug. The crown sheet and fusible plug were sent to Case Western Reserve University for a metallurgical examination. Lt. Detchon also stated that the fusible plug shows slight signs of overheating, however it did not melt and blow out. Lt. Detchon's investigation revealed that the late owner had purchased some type of boiler sealer. It could be safe to assume that the late owner knew that the boiler was leaking and attempted to seal the leaks with this compound. I have accepted Lt. Detchon's statements as fact and started my examination of the boiler parts that were available for inspection and I found the following conditions:

The staybolt holes in the crown sheet showed a maximum engagement of threads of 2.5 threads and often the least amount of engagement was 1.5 threads. The original design of thread engagement was 4.5 threads in a 3/8" crown sheet thickness. There were at least 5 stays that had been welded around the threaded area because of excessive loss of metal in the crown sheet.

The crown sheet shows excessive amounts of corrosion throughout the total surface area. The area around the threaded stays showed more reduction of thickness because this is a high stress area. The problem is compounded, because the less the thickness, the higher the stress. Measurements conducted show a thickness of 0.210 inches, 0.170 inches, 0.125 inches, 0.105 inches and 0.085 inches. The original design thickness was 0.375 inches. The original staybolts were 1 inch in diameter and appear to be 11 threads per inch by measurement with a thread gauge. The condition of the stays in the crown sheet area is one of uniform deterioration and confirms the thread engagement in the crown sheet of not more than 2.5 threads and as low as 1.5 threads of engagement. The diameter of the stays in the corroded area is between 0.600 inches and 0.700 inches. This is a reduction of 64 percent of the cross sectional area of the staybolt. The staybolts position in the wrapper sheet, with exception of approximately 5 staybolts, were in the original position indicating the ease that the crown sheet separated from the stays. The pitch of the stays as was measured in the crown sheet was 4.5" and this was confirmed by measuring the staybolts in the fire box.

The computations using various formulas ASME 1924 and ASME 1998 computed using a thickness of 0.085 inches came out to be between 40 psi and 47 psi.

I also used a carbide type scriber to test surface hardness. I noticed that there was no difference between the steam dome, the barrel and the ruptured crown sheet. This test is not conclusive but an indicator of the metal condition.

The ASME code requires that when the fusible plug is installed, the fusible plug must remain at least 1 inch above the crown sheet. The plug design was proper and met the requirement as confirmed by Lt. Detchon.

Inspection of the hand hole plug above the crown sheet revealed that the plug has not been removed recently as the threads were rusted. The area shows improper inspection of the crown stays. The front tube sheet hand hole plug threads were corroded revealing that the plug had not been removed recently. This plug allows the removal of scale and inspection of the condition of the barrel.

Conclusion

It is my evaluation of this boiler that the crown sheet failure started at the 0.087 inches thickness area, the weakest point in the crown sheet, and this is where the most bending damage is done to the sheet. The rest of the sheet shows signs of being peeled away much as peeling wall paper off of a wall. This was caused by the massive expansion of released steam.

 It is my evaluation that because of the very poor condition of the crown sheet with the reduction of the original thickness from 0.375 inches to 0.087 inches leaving only 23 percent of the original thickness; this was insufficient metal to hold the pressure of the steam resulting in a mechanical failure of the boiler. To further explain, the thinning of the crown sheet allowed the crown sheet to slightly bag in between the staybolts. This bagging allows scale to build up in these pockets or bags that insulate the metal from the cooling of the boiler water. This further compounds the bagging by localized overheating until there is failure of the crown sheet. This explains how the sheet could be slightly overheated without melting the fusible plug and does not support the theory of a Boiling Liquid Expanding Vapor Explosion (BLEVE).

Professor Wallace's Report supports the bagging theory as some parts of the crown sheet shows signs of overheating while others do not.

I estimate the amount of energy released during the explosion at 90 psi. to be around 28,000,000' lbs of force of which approximately 1,280,000' lbs was used to lift the engine and the remaining was dissipated in the blast area around the engine.

The inoperative safety valve had no direct bearing on the explosion as the valve was set for 125 psi. and with the condition of the crown sheet. It is doubtful that pressure was attainable.

I further conclude that, had the boiler been presented for inspection in Pennsylvania, the boiler would have been placed out of service and not allowed to operate.

John D. Payton, Director
Boiler Section