During an annual survey of the switch rooms at the cement processing plant a very slight temperature increase was seen on a pair of parallel neutral conductors where they were connected to the outgoing terminals of the panel. The armoured cables fed out to a remote switch panel feeding mostly three phase plant and a small amount of single phase equipment. Temperatures were not excessive although the survey was taking place in an external switch room in February so the ambient temperature in the switch room was low.
Although the temperature seen was not high at only 19°C, we were surprised to see any heat at all on the neutral connection as most of the plant was three phase machinery and the neutral conductor would be carrying very little load. The temperature pattern was also strange in that the two parallel conductors had different temperatures leading us to think that it was not a connection issue at the lugged termination. Access to the chamber next to the cable termination chamber was restricted due to large storage cabinets standing behind the panel These were emptied and removed to allow us to open some additional covers to carryout some further thermography. With the cabinet removed and the covers open the cause of the heating became clear. The removable neutral link on the incoming breaker to the switch panel was loose and had an incredibly high temperature of 232°C. The heat from this poor connection was spreading along the copper busbars of the panel to the connection we could see.
We reported the issue to the site staff who were not immediately concerned as the temperature was so low. When we explained the possible causes and that we were likely not to be looking at the source of the problem they agreed to move the cabinets. When we were able to show them the temperature at the loose connection they were glad that they had agreed to give us greater access.
It was not possible to isolate the panel at the time of our visit but a shutdown was arranged for the following weekend when the link and nuts and bolts were replaced. The faces of the connection were cleaned and the bolts were tightened to the correct torque.
With a relatively short planned shutdown we were able to prevent a failure of this connection which would have affected the entire switch panel and would have likely caused a prolonged interruption of the supply to large parts of the plant. Indirect access to the cause of issues can lead to them being missed and should always be a factor when considering the severity of thermal anomalies. This was compounded by the highly reflective copper busbars reflecting the low ambient temperature in the switch room giving an artificially low apparent temperature.