Submersible pump cable joints

[ashtree]

I have seen lots of cable joints on submersible pumps last for years often outlasting the pump themselves. In recent times though we have had a string of failures with epoxy filled joints some only lasting 12 months before a fault occurs. I have had some robust discussions with our electrical teams on the issue and it would seem that they think they are doing the right thing. These are motors in the 20-100hp range typically.

Years ago i was involved in one emergency situation where we had to improvise on joining a cable on a 45hp pump motor. The conductors were stripped back, soldered together, heavily wrapped in tape. A piece of 2 inch conduit was slid over the joint, one end packed with mud, then taped and the conduit was stood upright and filled with a pipe repair 2 pack epoxy, which was all we had. A couple hours later when the epoxy had hardened we threw the cable and pump in the water and started pumping. That joint done in non ideal circumstances with compromise equipment went for years and was only cut out when the pump was replaced.

Now using crimped connectors with hydraulic crimping tools, self sealing heat shrink, non conductive epoxy and proper jointing systems we get failures regularly. The autopsy on those failures normally indicate that there is a failure(overheating/burnt) of the epoxy around one of the joints. This would make you think that the crimp connectors do not have enough crimping pressure applied. I have checked this and is in line with the recommended settings. Eventually this failure gets big enough that it gets to the outside, allowing water in and a fault occurs.

I am interested to hear if other people are having the same issues and if this is a gradually worsening issue, and if anybody has a solution that has worked.

I have attached a picture of the failed joint that was cut open.


Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[MikeHalloran]

The problem with investigating power circuit failures is that the fire destroys the evidence.

Can you get sections through the actual crimps that remain?

It might be illuminative to section a few recently completed crimps that haven't failed.

I'm not convinced that cutting a few pieces of cable and crimping the freshly cut ends would be representative as a test.

I'm conjecturing that, maybe, in order to not waste expensive cable, the crews have been stripping and preparing the cut ends of stored cable, without cutting off and discarding the first foot or two, that may have oxidized in storage.

Do your splicing standards suggest spreading the stripped ends and scraping the copper down to bright metal before assembling the crimp? Maybe they should.






Mike Halloran
Pembroke Pines, FL, USA

 

[jraef]

No disrespect intended, but just to be thorough. You are using the special crimp sleaves designed for high strand count cables, right?


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[ScottyUK]

I can't imagine any credible way of cleaning individual strands and then putting the conductors back into a form which would be crimp-able. The suggestion to trim a few inches back is sensible.

What type of crimp die did you use? The hexagonal type can be quite fussy about the wall thickness of the crimp tube or lug, so that tubes from one manufacturer may not be fully compatible with the die set from another. The tool goes appears to go through its full cycle and achieves required crimp force, but if the tube wall is thin the joint is still poor. We found this out to our expense a few years ago - both crimp tool and lugs were well-respected brands, but from different manufacturers. We had a lot of suspect connections which had to be re-made. Now we only buy lugs and tubes from the manufacturer of our crimp tool.

 

[ashtree]

Thanks for the replies.

Just to answer the questions.

Our teams do not normally use the hexagonal compression type but a style that has a single compression point along a cylindrical sleeve. One joint i actually observed being done appeared to have a matching brand between crimp tool and sleeve but i cannot guarantee and i would expect that this is not always so.
Whether the crimp sleeve is specifically designed for high strand count cable is an excellent question and i do not take any offence whatsoever. The short answer is that i do not know but i am pretty certain that the same equipment would be used for standard multicore cables as well as the flexible high strand count cable. I will ask that question but i know i will be told that there is no difference.

In terms of cleaning the strands etc, all that is normally done is that the cable is cut back enough to get to clean conductors. This was something i had specifically asked about as i had suspected that at times the cables were hardly even being cut back enough to get away from the heat damaged section.

I have disposed of the sectioned joint but we seem to have regular failures across a large fleet of pumps so the opportunity may arise again shortly.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[EdStainless]

In an effort to make resins more electrically insulating they have also made them better thermal insulators.
I would not trust crimps in this service. They are heavily insulated, and they can see significant cycling.
The lack of cooling makes the thermal issues with crimps even worse. They will loose contact over time.
We would crimp solid conductor cables, but we were actually pressure welding the two conductors together, no power went though the crimp sleeve.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[ashtree]

Just an update.
We are using Rhino tools hydraulic crimper which uses a hexagonal crimp dies with preset hydraulic pressure based on the die in use.

The crimp sleeves are a CABAC product. CABAC recommend using increased pressure on high strand count flexible cables but then say for conductors under 16mm2 to use normal procedures. The Rhino crimper does not allow any pressure change.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[ScottyUK]

Sounds like you are on the same road we were: in our case it was a Cembre (pronounced 'Sharm-bray') tool with what I think were Elpress lugs. Both good brands, just not 100% compatible.

I'm not familiar with Rhino as a brand but if they make their own range of lugs and tubes then give them a try, or at least ask whose they recommend. The photos of the Rhino die set I found online look very like the Cembre brand - I wonder if one of the companies is re-badging the product? Might even be worth trying Cembre tubes with your tool.

 

[ashtree]

Thanks Scotty

Rhino do not appear to make lugs nor do they recommend any it seems. The Rhino tool is made in China although that in itself is not necessarily an issue.

I am concerned though that our crew are saying that the unit is preset to the right pressure and you just change the dies to the lug size you need and everything is okay.

Rhino are a bit vague on the detail and do not provide any info apparently on what that pressure is , but apparently it can be adjusted. When you are using crimping sleeves from different manufacturers and then you don't know what the right pressure is and are just relying on some preset (which itself can be adjusted) i am not convinced that we can be confident that these repairs are being done correctly.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[LionelHutz]

Switch to a crimp tool that is listed for certain lugs then you know testing was done. It will also have some means of verifying it is working correctly. We've been using a Greenlee hydraulic crimp head at work and it is UL listed for the Panduit lugs we use. It uses slugs and a go/no-go gauge to determine if the crimp pressure is correct.

 

[ScottyUK]

"I am concerned though that our crew are saying that the unit is preset to the right pressure and you just change the dies to the lug size you need and everything is okay."

That's exactly how the Cembre tools work, and I'm pretty sure that the Klauke / Greenlee designs are more-or-less the same. The old BICC tools also worked this way. That's a lot of industry experience and company reputation at stake. Something which is important to understand is that the tool will register full crimp pressure on an open die set by traveling until the dies make contact. With the correct tube or lug in the dies then the required deformation will be achieved during the crimp and the dies will not quite make contact with each other.

If the manufacturer's tube or lug which was tested with the die set is used for the joint then I would have confidence in the joint, but that's based on major brands with a reputation to protect and a solid business in own-brand lugs and tubes to suit their tools. Rhino don't have any reputation in my part of the world - neither good nor bad that I can find - but the fact that they don't make tubes and lugs doesn't fill me with confidence.

 

[ashtree]

Scotty
The Rhino is fairly similar to the Cembre but at least the Cembre publish the crimping force. Rhino do not that i can find.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[djs]

We have a UL508 panel shop. We do use some lugs although on smaller wire. UL insists that we use the same brand crimper as the lugs we are crimping. I would say see if you can find CABAC dies that fit it your Rhino crimper.

 

[LionelHutz]

ScottyUK - The dies in die-type crimp tools I have observed always make contact when the crimp is complete. For example, a 10-ton crimp tool should fully compress the lugs with less then 10-tons. The dies finally making contact still cause the pressure set-point to be reached which "pops" the pressure release.

We use crimped lugs on high strand cable all the time. We haven't had a failure in a crimp yet. We use a combination of normal and high stranded lugs. But, the manufacturers just make high strand lug with a flare on the end so it's easier to insert the wire. Otherwise they are the same lug. We size the lug ID to the wire. Basically, pick the smallest ID that will go over the wire. We never just pick a lug that matches the wire size.

Personally, I like the dieless crimpers for a stranded wire application because stranded wires are not as consistent a diameter as solid wires. So, the dieless crimper will continue to apply the force until the lug has compressed around the wire the proper amount vs a die which will only compress until the dies come together.

 

[ScottyUK]

Lionel - I will pay closer attention to the next crimp we do. The tool develops 132kN or about 13 tonnes so much the same as the one you describe. I've always believed that there was a few thou of daylight but I may be wrong.

 

[Compositepro]

I agree with Scotty. If the dies actually touch each other, it would seem that you really cannot know if the crimp is actually any good. It would indicate that there is not enough wire filling the connector, or no wire at all.

 

[LionelHutz]

The dies definitely still come together on the hand and hydraulic die type crimpers we have. It even happens on the small control wire ring/fork crimpers. The dies have the correct inside dimension to provide a full crimp when the dies are fully together.

 

[roydm]

Maybe not related but I'll share it anyway
I have seen instruments fail where you have a cable joined inside epoxy and came to the conclusion the rigid epoxy doesn't bond well to a flexible insulation and the moisture creeps into the joint.
This is particularly noticeable with pH probes where the impedance is typically well over 100 Megohms, the signal shorts out in just a few hours.

We have used silicone sealant with good results.

 

[ashtree]

Thanks everybody for the discussion.

I know this question will open a can of worms in itself. For maximum reliability underwater should we be crimping then soldering the joint?

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[ScottyUK]

In my opinion a correctly prepared and crimped joint which forms a gas-tight metal-to-metal connection will be as good or better than a soldered joint. Many soft solder alloys will melt at a temperature lower than than the final conductor temperature under a worst-base through fault - which could be up to 250°C for a thermosetting cable - so the joint can effectively fail under service conditions. If the joint were subject to movement then the soldered joint would also be inferior from a strain relief perspective, although in this particular case of a resin-encapsulated joint that probably isn't a problem.