Shell Haven

Baffle wall in furnace

What better way to start shining a light into the archives than to recall the fiery glare of Furnace 722 at Shell Haven. Long since gone, Shell Haven was the youngest of the Thames Refineries, sitting on the north bank near to the picturesque old village of Fobbing and the less than picturesque sprawl of Corringham.

A routine maintenance shutdown of the Platformer III facility at the refinery was programmed for the start of March 1987. Part of the facility included a large furnace and there was some concern that the internal baffle wall, some 11 metres long, was in poor condition. A monitoring camera had been installed to watch the wall and it was considered that a collapse might occur before the scheduled shutdown date. This would result in the immediate shutdown of the plant and loss of product being processed. The additional costs would run into seven figures.

In late November 1986, Survey Control Services were asked to look at the situation and see if an accurate monitor could be set up to predict the collapse date. The furnace had a series of inspection spyholes, fitted with thick heat resistant glass and heavy metal covers. The internal temperature was at around 700 degrees centigrade. On the face of the baffle wall were a series of firebrick protrusions, designed to cause turbulence in the hot gases and create mixing to promote an even temperature. It was decided that the position of these could be monitored and thus give an indication of the wall movement.

Because of the hot air blast which came out when the spyhole glass was removed, it was apparent that each observation would have to be made within only a few seconds, if both instrument and observer were not to be roasted. An intermediate protection plate of heat resistant glass was mounted so that it could be swung over the spyhole. This enabled the main protective glass to be removed during the observation sequence, the intermediate plate providing heat relief between the actual observations. As the project was being carried out in mid-winter, this gave the surveyor the interesting sensation of standing in an airspace at -5 degrees to the rear and +50 degrees to the front.

Two of the spyholes were close to the line of the wall, which facilitated monitoring by measuring angular deflections. Test measurements showed that an agreement to better than 5mm. could be obtained by measuring angles to the face of the protruding firebricks. Each brick was observed in turn from two set-ups, one on each side of the furnace. Twenty six reference points were included in the monitor. As the spyholes were accessed from high-level walkways, a reference system was set up on the ground to re-fix the observation positions for each visit.

Calculations indicated that the wall was likely to move a maximum of about 400mm. at the central position before collapse occurred. Monitoring started at the beginning of December 1986, initially every two weeks. By the end of January, the movement graphs showed a consistent and progressive acceleration. A predicted collapse date of 1st. March was deduced, the start date for the shutdown. Several of the burners were turned off and output reduced in mid-February. The observations at the start of the last week of February indicated about 6 days life left in the wall. The shutdown date was brought forward two days. Less than twelve hours after the last product was output, the wall collapsed. Mission achieved.

After major maintenance and rebuilding, Platformer III went on to provide a further 10 years of useful production.