Aging paper insulation in power cables

[ItAintMe]

I'd appreciate any ideas on places to look for info on determining the condition of paper lead power cables (33kV mostly) in aged cables. I'm trying to develop some guidelines for determining when a cable is approaching the end of its economic life, and how to prioritise replacement programmes. We have quite a lot of old cable (50 years or so) which is generally giving good performance (not many faults) but when repairs do arise, we are starting to find the papers are drying out and getting brittle.

We tried some partial discharge mapping a year ago. I like the concept but the system we tried wasn't really developed enough to be giving good results. Also, we have a lot of mixed cable (part paper, part gas pressure or oil) where it's less effective. Cable lengths tended to be longer than it's reach too. Any comments appreciated

 

[cuky2000]

I have afraid to tell you that there is not specific answer for your case to determine the cable replacement in a 50+ years old plant.

<sub>Few years ago, I was part of a team to study the MV underground network for a large utility in the East Coast. The problem was to target the feeder population that will be more likely to fail.

There were available a large historical database that include the stresses information in the system such as overvoltage, overloading, short circuit and environment conditions. A failure mode was prepared and continuous adjustment were implement to increase the prediction index.

There were a generous budget were a lot of resources were invested every year.

Since this is random phenomenon the solution approach was to use probabilistic models to predict the failure population.

Those that were expecting a deterministic result raised some concern. There were few cases that was not possible to explain how a new cable with minimum stresses fail and an old cable with a large history of overstress remain in service for more than 50 years.</sub>

For your case probably a good advise is to use the common sense implementing few steps considering that probably other equipment could be as old as the 50 old cable mentioned in your post and resource limitation may be expected.

Steps recommended are as follow:

1-Develop a strategy for cable replacement program prioritizing base on degree of importance in accordance manufacturing process and the overall electrical one line diagram.

2- Provide redundancy path feeders to reconfigure your network in case of failure

3- Quantify the cost-benefit implementing step 2 versus downtime cost.

4- Request a capital budget for maintenance. For sure your management would be interested in know the return of the investment.

5- Check how far can you go investing capital budget before someone question if the plant is arising the end of life.

Good luck

 

[jbartos]

Suggestion: Perhaps you would be interested in &quot;Impulse tests&quot; as described in Reference:
Westinghouse Electrical Transmission and Distribution Reference Book, by Westinghouse Electric Corporation, 1964,
Chapter 4 &quot;Electrical Characteristics of Cables&quot; (by H.N. Muller, Jr.) and some cited References therein.

 

[cuky2000]

Dear Jbarto:

The test recommended in the T&D Handbook is a destructive go-not-go test not recommended for cables in service for more than 5 years. I am under the impression that there are cable in service for more than 50 years.

[i}“The DC test voltages are applied to discover gross problems such as improperly installed accessories or mechanical damage. DC testing is not expected to reveal deterioration due to aging in service”
[/i]
See the enclose reference:

http://www.ewh.ieee.org/soc/pes/icc/subcommittees/subcom_e/education/2000/fall00_goodwin.ppt.

ItAintMe mentioned that a none destructive test (Partial Discharge Mapping) were performed with limited practical results.

 

[ItAintMe]

Thanks cuky and jbartos.

Jbartos - we actually use dc tests for proving joints and repairs on higher voltage cables (33kV and above). Impulse tests we really only use for acceptance testing. We are looking for criteria we can use to prioritise for replacement of a large network of power cables.

Cuky, I work for a power company, and it's our sub-transmission network I'm concerned with. I'd be interested to hear more about the work you did with the east coast utility. We don't have a large database as you say they had at their disposal. With some difficulty, we can searth paper records and some outage report databases to get a history of failures and repairs, but that's about it.
What do you mean exactly, when you say &quot;A failure mode was prepared and continuous adjustment were implement to increase the prediction index.&quot; What data did you use to do this? it seems like additional data was being regularly added to your information? Thanks again both of you