A furnace shell runs at a temperature you can work out from the lining thickness, and a hot patch means the lining has gone thin, which is a dangerous thing you fix. We have measured that lining ultrasonically and by drilling, and it is full thickness, and the patch is still forty degrees hot.
I am a shift engineer at a foundry and I have the night turn on two melting furnaces.
How the shell tells you about the lining.
A furnace is a steel shell with a refractory lining inside it, and the lining is what stands between the shell and the metal. Heat passes through it at a rate you can calculate, so a known thickness of a known refractory with a known temperature inside gives you a shell temperature.
Ours runs between a hundred and ten and a hundred and thirty degrees on the shell when the furnace is up, which is hot enough to burn you and entirely normal.
When a lining wears, the shell over the worn part gets hotter, and you find it with a thermal camera on a scheduled survey. A local rise of forty degrees is the classic signature of a thin spot, and it is one of the things that can kill people in this industry, because a lining that fails lets metal out.
So a hot spot is not an anomaly you log. It is a reason to take the furnace off and look.
There is a patch on the south side of furnace 2, about the size of a dinner plate, a metre above the bottom, that runs forty degrees above the shell around it.
It has been there since 2017.
What was done the first time, which was exactly right.
The furnace came off at the end of that campaign, three weeks early, at a cost I was told about more than once.
The lining was inspected cold, from inside, by the refractory contractor and by our own engineer.
Full thickness. They measured it in eleven places round that area with a depth gauge through a drilled hole, and the wear pattern was even and normal for the hours it had done.
They relined it anyway, because it was out and because nobody was going to put it back with a question on it.
New lining, new thermal survey. The patch was back within two days of coming up to temperature.
What has been done since.
Three campaigns and three relines, and the patch comes back in the same place every time, within about fifty millimetres by my own measurements with a tape from the same datum.
Ultrasonic thickness testing on the shell plate itself, because a thin shell would also run hot. The plate is within a tenth of a millimetre of nominal all round, including under the patch.
A different refractory in 2020, from a different supplier, on the contractor's recommendation. It did not change.
Thermocouples put into the lining in that area during the 2021 reline, which is the best data anybody has. Three of them at different depths, specifically to measure the gradient through the lining at the patch and compare it with a set on the north side.
The inside face is the same temperature on both sides, which it would be, because it is the same molten metal.
Through the first half of the lining the gradient is the same.
Through the outer half, on the south side, it falls away less than it does on the north, so the heat arrives at the shell hotter, and the arithmetic on that says the outer part of the lining there is conducting about three times as well as the same material does on the other side of the same furnace.
Which is a statement about a material and it is the bit I cannot make sense of.
That refractory is a known product with a published conductivity. It is dry, it is the right density, it was rammed by the same crew in the same shift, and the pieces out of that area at the next reline went off for analysis in 2022.
The laboratory report says the samples are within specification for composition, porosity and density and shows no evidence of slag penetration or metal infiltration.
A refractory that has let metal into it conducts better, which is what I expected and wanted. There is no metal in it.
What happens about it now.
It is a monitored condition. Thermal survey weekly instead of monthly, a trend chart, and a trigger figure at sixty degrees above the surrounding shell at which we take the furnace off.
It has never reached sixty. It sits at between thirty eight and forty three and it has done for eight years.
The insurers know. The inspecting engineer looks at the chart every year and signs the report, and the sentence in it is that the anomaly is stable, trended and within the operator's action threshold.
What I do on the shift.
The survey, the chart, and a walk round with the handheld at the start and end of the turn, which takes ten minutes.
I measure that patch from the same spot on the walkway every time, at ninety degrees to the shell, at the range the camera is calibrated for, because a thermal reading changes with angle and distance and a sloppy measurement would make this look like it was moving when it is not.
The only unofficial thing in eight years is where I stand.
The walkway runs round that side and you can get within two feet of the shell. You do not, because it is a hundred and twenty degrees, but you can be three or four.
I take it from the far side of the walkway now, about nine feet back, and I wrote the new distance into the procedure and calibrated for it properly so the numbers stay comparable.
The reason is a night in 2019 when I was close in with the handheld, and a furnace is a loud place, and you feel the heat on your face the whole time you are on that walkway.
On the patch, and only on the patch, the area about the size of a dinner plate that is forty degrees hotter than everything round it, what you feel when you put your hand towards it at a foot away is cold air, moving, coming off the steel, and I have done that twice in eight years and I am not doing it again.