Buried cable temperature monitoring

I am not aware of any product made for this purpose.

Two possibilities occur to me.


Firstly the conductor temperature may be calculated from load current and voltage drop, may need correcting for power factor, and only calculates the AVERAGE temperature and not any hot spots.


Secondly, several suppliers offer remore temperature sensors, primarily intended to monitor outside air temperatures, one of these could be attached to the cable before burial. Not hugely reliable as it measures the SHEATH temperature and not the CONDUCTOR temperature. Also other parts of the cable may be hotter than the measured point.


Better practice though would be to use a larger cable and avoid any doubt and worry about cable overheating.

For a short length the cost is not that great.

For a greater length the extra cost may become significant, but remember that the larger cable will save energy and give better voltage regulation at the load.


EDIT TO ADD, it appears that there IS a product made for this purpose, see following post.

Hi Jam,


We have used a retrofit Cable Temperature Monitoring solution for 15 years, connected to our SCADA. The cable is 300mm2 three core, and we use real time monitoring with a doped fibre optic sensor cable in two locations. The fibre allows measurement along a length, rather than taking spot temperatures. Although this is sheath temperature, we do have a calculation that gives an approximation of core temperature. Using the stored results, we have a new agreed power level with our insurers that gives us another 50% capacity based on its original installed capacity. I don’t want to discuss the technical detail on an open forum, but if you PM me a phone number or email address I would be happy to discuss my experiences, 


Regards,


Alan.

Thanks both, much appreciated.

Agreed that the better answer would be a larger cable for a new build, but the application is an existing cable in multiple locations about a site and the client is keen not to have to replace them all only to find that they have been providing adequate service; this would be part of a wider package including more detailed retrospective calculations and electrical testing.
broadgage‍ interesting idea to indirectly infer temperature from resistance. Might not work in this case because there are several different installation conditions in play, but certainly one for the memory bank.

I have tried to send you PM Alan Capon‍ but its failed (have struggled to reply to them in the past so apologies if its me)

If already existing cables are suspected to be marginal or at risk of overheating, It might be worth considering the following.


1) What is the power factor like ? 0.85 is usually considered acceptable, but improvement to unity might be worth considering as it will reduce the amps for a given load in KW.


2) Is the load balanced ? an unbalanced load will increase cable heating.


3) Could grid tied solar be installed on the buildings supplied by the cables. This will reduce the net load during daylight and assist in reducing cable temperatures.


4) General energy saving measures, and in particular LED lighting can reduce loading and be worthwhile. What sort of loads are supplied ? any other oportunities for power saving ?


5) What is the actual voltage at the load, if unduly low this may increase the amps used for a fixed load in KW.

Is there an obvious 'pinch point' where the cable is least well cooled ? It is far easier to monitor a single point than along a length.  In tricky cases one can insulate a short section that is accessible to mimic the section of concern that you cannot reach, and instrument that instead so long as the environment  to be mimicked is known.  (I have not done this for power cables, but the physics should be the same.)


In a live system you can really  only measure the external jacket temp, and make some assumptions about how the temperature offset between core and outer varies with the external insulation.

For extreme examples you could have a cable in running water, and the external jacket temp would be very low, as the water is a good coolant,  but at (the very high) full power, the core would be 70C or whatever, and all the temperature drop is across the plastic of the cable. The other extreme would be  a cable that is very well lagged, or in still air, with a lower full power level. Here the core and jacket temps are much closer to each other because most of the temperature gradient is external to the cable. Another way to look at this, is that in the 2nd case  the plastic of the cable jacket is a better thermal conductor than the external lagging.


Personally I have always found that real cables run a bit cooler than expected, assuming that grouping and routing methods have been correctly included in the  prediction, but that may just be my luck.

M.

Hi Jam,


I believe that I have sent you a PM with my email address on it. 


Regards,


Alan.