Accident Example 2

Consultancy and Engineering of Infrared Monitoring Systems for the Chemical Plant Industry


Images taken by Rainer Sobottka

Avoiding Accident Example 2: Expanding Flanges at Reactor

Reference is made to actual accidents with partly death causalities and disturbances on POX plants 1982 to 2008 and watched by myself.
Please refer also to the internet under search words of “gasifier accident” “flange rupture”, “gasifier explosion”, etc. One accident happed like this and the next few days many infrared cameras and camera system monitors were ordered by the plant management…..

This infrared image left show a man hole flange at a gasifier, left picture is the infrared image, the right one is an image from a digital camera. The right flange shows the manhole as we normally see it, no notice of any outside skin temperatures. This is only one example of many other possibilities of flange rupture on a reactor.

Certain reactors with refractory have the problem, that the refractory sometimes get worn out and the hot burning gases (approx. 800 to 1500 degr. C) enter into the gaps between the single brick layers. This process is continuing till the hot temperature reaches the outer steel shell of the not insulated vessel. However, the outer steel vessel has a design temperature of only 450 degr.C, for example).

The hot gases reach the out shell from inside and expand the material. Flanges leak all of a sudden, gaskets burn, thermo element construction gets damaged and the hot gases leak into atmosphere, not recognized during day time (H2 gases) but at night.

At night there is a light flame coming out of the thermo element flange with inner damaged gasket. In infrared image however, you can see the flames and leakages during day time, as well as, during night time very good. Only the infrared camera can detect easily such deviations to normal operation. I have watched such flames in header lines to flare and in thermo element construction. Means, such incidents can happen, and without any infrared camera detection installed no one can ever recognize it in time. Normally and as per earlier conventional construction design such areas are not manned and not supervised. In extreme conditions such happening will proceed till disaster if not using infrared cameras for detection, supervision, alert and shut down.

The image on top left show a man hole with a thermocouple flange and a conduit cable connected. The flames came out of the cable joint at the thermocouple flange and later out of the conduit union. I must express my very concern for such incident, as the dangerous gases can lean into undetected into the control room if further damages occur on cable joints and conduit casket assemblies, due to high pressure from the reactor gases.