08-12-2024, 01:28 PM
Peter, the failure was due to a number of issues, all associated with the manufacture and testing.
The equipment was pushing the boundaries of known technology of that time, nearly 60 years ago.
The failure initiated in two separate locations at weld defects in the circumferential weld seam between the end closure forging and the shell. Whilst all known NDT was applied, this did not pick up these defects.
The pre heat applied before welding and the local post weld heating was insufficient which allowed the welds to cool too fast causing hydrogen cracking.
It was also found that the thermal sensors in the furnace used to carry out the final post weld heat treatment were inaccurate which lead to the heat treatment being ineffective.
Finally the test was carried out over the Christmas break when the shop would have had very few employees on site, but it was cold and the water temperature for the test would have been just a few degrees above freezing, which would increase the probability of brittle fracture and hence an instant catastrophic fracture.
The report says that the temperature of the test water was around 7 degrees but our engineer whom I worked for was told they broke the ice on the canal to fill the vessel!!
Google John Thompson Converter failure and see the photo.
Fortunately this was the only known failure we ever had, but so much was learned from this and instigated many new practises used in the industry to this day.
The failed sections were cut off and removed and new sections fabricated causing much delay but the vessel finally went into service and survived through to the plant being decommissioned.
The equipment was pushing the boundaries of known technology of that time, nearly 60 years ago.
The failure initiated in two separate locations at weld defects in the circumferential weld seam between the end closure forging and the shell. Whilst all known NDT was applied, this did not pick up these defects.
The pre heat applied before welding and the local post weld heating was insufficient which allowed the welds to cool too fast causing hydrogen cracking.
It was also found that the thermal sensors in the furnace used to carry out the final post weld heat treatment were inaccurate which lead to the heat treatment being ineffective.
Finally the test was carried out over the Christmas break when the shop would have had very few employees on site, but it was cold and the water temperature for the test would have been just a few degrees above freezing, which would increase the probability of brittle fracture and hence an instant catastrophic fracture.
The report says that the temperature of the test water was around 7 degrees but our engineer whom I worked for was told they broke the ice on the canal to fill the vessel!!
Google John Thompson Converter failure and see the photo.
Fortunately this was the only known failure we ever had, but so much was learned from this and instigated many new practises used in the industry to this day.
The failed sections were cut off and removed and new sections fabricated causing much delay but the vessel finally went into service and survived through to the plant being decommissioned.