​​​​Callum O'Dwyer

In 1984 Syncrude Canada suffered a devastating fire caused by the fracture of a 6” diameter line containing bitumen (a hydrocarbon) at 550°F (287°C). The fire started around midnight on August 12th 1984 and blazed unabated for some hours. The fire melted some 600 lb/in2 steam lines which made firefighting more difficult. The fire was extinguished by the following morning.

 

The fire was centred on the No.2 coker and its auxiliary pipework and smaller vessels. Another large low pressure reactor was also damaged by the fire.

The cause of the fire was relatively simple. The 6th line that had ruptured was specified to be constructed from a 5% Cr – 1% Mo alloy. During an earlier operation problem in 1979 a section of the line had been removed to gain access to the line. Unfortunately when the line was reassembled, the replacement piece was only made from a plain carbon steel similar to ASTM A516-grade 70.

 

This material over a five year period had thinned to around 0.010” thick under the action of sulphidation and scouring from the liquid bitumen and coke being carried in the line.

Once the excitement had died down the engineers of the plant were faced with the problem of what to do. The fire had buckled structural steel and there were traces of thermal damage to adjacent pressure vessels and piping systems.

 

The first thing that was done was to produce a map of the thermal contours around the centre of the fire. This identified where there were likely to be problems. The oil industries’ solution to problems is usually to through money at it. But where to throw the money?

The first thing the metallurgy group did was to arrange for in situ surface metallographs to be carried out on critical components. One was the coker, but two other pressure vessels were involed. One of which was a “large low pressure reactor” whose “bottom” had been severely “cooked”. The other vessel was the H-Oil reactor and it was quite small.

 

Examination of the microstructures revealed they were abnormal in terms of typical microstructures of the microstructures seen in production steels. Whilst it was possible to obtain approximate hardness values, it was not possible to obtain tensile test or Charpy impact test data without cutting samples from the pressure wall. Owners of pressure vessels are normally reluctant to cut sections out of the vessel wall. In the case of the large low pressure reactor, the microstructure varied right around the circumference and the engineers had no idea what these meant in terms of tensile and toughness data.

 

Since winter temperatures reach -40°F (-40F) in Fort McMurray, the toughness of the vessel was deemed to be important even though the vessel normally operated at temperatures around 550°F (287°C).

 

Had the engineers been able to correlate the microstructures with mechanical properties the decisions would have been easier to make.

Basically the engineers were faced with the following choices:

 

• Scrap the vessel and bear the cost of a new one and the cost of the loss in refinery production.

• Run the damaged plant at lower pressures and temperatures. But what lower temperatures and pressures?

• Ignore the damage (!)

 

In fact what was done was the following:

 

• All piping in the areas that the thermal contour mapping had shown to be possibly damaged was removed.

• The H-Oil reactor was scrapped. The insurance company subsequently refused to pay for it on the grounds that Syncrude could not technically justify its scrapping.

• The first 20 feet of the “large low pressure reactor” was removed and replaced on site. As far as the author is aware no investigations were carried out on the mechanical properties of the bottom 20 feet of the reactor.

 

It seems to the author that there is basically a need for some work to be carried out to correlate abnormal microstructures with mechanical properties.

 

Hence this project – starting with the simpler low carbon steels.