IET logo
Decrease font size
Increase font size
Topic Title: Solar DC Cable run through field
Topic Summary:
Created On: 15 March 2014 08:03 PM
Status: Post and Reply
Linear : Threading : Single : Branch
Search Topic Search Topic
Topic Tools Topic Tools
View similar topics View similar topics
View topic in raw text format. Print this topic.
 15 March 2014 08:03 PM
User is offline View Users Profile Print this message


Posts: 67
Joined: 17 December 2008

Went to view a site I may potentially be carrying out the installation for.

When I got there I was told by the main contractor that the installed 25mm four core was pre installed by the customer but didn't know the length.

Its a three phase 20KW system ground mounted in the field. So I took my measuring wheel to (amongst other things) measure the cable run for the calc to see whether the armourings would suffice for the ELI.

And was promptly told the cable was for the DC run back to the inverter which was to be installed in the house.

150M cable run.

So have been looking in to this since. I cannot see a problem except:

1: The issue of wether to bond the armourings.
2: No inverter I have found will take two 10KW strings (around 1200Vdc)
3: As the cable is four core we are limited to two strings
4: Should the strings be fused? There are internal fuses for the strings in the inverter

Has anyone encountered this situation before?

The engineer who is having the work done is an electronics engineer and has demanded the DC run back to the house to minimise losses. However there is no inverter that can take 10KW strings!

ANy thoughts welcome!
 15 March 2014 08:15 PM
User is offline View Users Profile Print this message


Posts: 649
Joined: 17 September 2001

Perhaps if the engineer thinks he has already done all the design work on the system, you should ask him which inverter he wishes to install!

Is the cable rated for 1200V?

S P Barker BSc PhD MIET
 15 March 2014 08:18 PM
User is offline View Users Profile Print this message


Posts: 524
Joined: 11 January 2006

DC along 150 meters it's sounding like a fire starter already

Keep up with the antics of the UK,s most humoured electrical company
Twitter: @rtindustria
 15 March 2014 08:46 PM
User is offline View Users Profile Print this message


Posts: 67
Joined: 17 December 2008

Yes exactly! Which inverter can handle two 10KW strings!

Thanks for highlighting the working voltage of the cable. For a BS5467 cable the given voltage is 600/1000V.

I couldn't find anything which stated whether the stated voltage 600/1000 was ac or dc voltage. I assume that the 600/1000v means a.c or d.c as the regs current carrying capacity tables refer to ccc and the installation methods as a.c or d.c
 15 March 2014 08:50 PM
User is offline View Users Profile Print this message


Posts: 67
Joined: 17 December 2008

Originally posted by: jamieblatant

DC along 150 meters it's sounding like a fire starter already

Ill be pulling up with Fire starter blaring out the van at this rate!
 16 March 2014 12:33 AM
User is offline View Users Profile Print this message


Posts: 6435
Joined: 27 December 2005

20kW inverters are available, however at this rating you have crossed from the usual domestic solar inverter into the commercial offerings. Assuming that the system is to run in parallel with the public electricity supply, you may well need to have G.59 protection (depending on the DNOs requirements), and it is likely that you will also need a three-phase connection to the DNO. Does the property already have a three-phase supply?


 16 March 2014 06:48 AM
User is offline View Users Profile Print this message


Posts: 8452
Joined: 15 January 2005

I'd simply ask the design engineer for how he wished his design and specified materials to be installed.
As Alan implies, it sounds like a 'commercial' installation, will need the co-operation of the DNO (now that will be fun and expensive) and is usually forgotten by designers.

Being very curious I always ask what the client is hoping to achieve once someone presents me with a project outside my comfort zone so I can understand the situation. You could start there.

 16 March 2014 12:21 PM
User is online View Users Profile Print this message


Posts: 1910
Joined: 07 August 2007

As the existing cable is 4 core, could not 2 DC circuits be run, one for each string of PV modules, and each string connected to its own inverter at the house end.
The DC voltages and currents would then be reasonable relative to the cable.
2 inverters each one feeding back into one phase of a 3 phase supply wont be balanced, but only 10KW per phase should be fine.

There is little point in fusing the DC circuits since PV modules cant produce fault current. The short circuit current is typicly AT MOST 115% of the normal load current, 110% typicly.
No fuse exists that will reliably carry a current of 100X but reliably open at 110X.

In practice standard SWA armoured is fine for 1,200 volts DC since the insulation is designed for 1,000 volts AC, RMS which is about 1,400 volts peak. Still not a good idea though unless the cable supplier will confirm this.

Another option would be to redesign the system with the inverters in a suitable housing near the array, and the existing cable used for standard 3 phase AC to the house.
This has the merit that a standard AC supply is available near the array to facilitate any future repairs etc.
 27 April 2014 12:51 PM
User is offline View Users Profile Print this message


Posts: 67
Joined: 17 December 2008

Well its been a while since I asked for some advice on this but thought I should follow up on what happened.

As one of the work force put it: "There's been a lot of fire fighting on this job" )

When I was recruited half way through the install I didn't have all the information but the reality was this:

The DNO had only authorised 15KW of generation on to the grid. However the customer wanted (and had already ordered) 20KW of modules to deal with the anticipated shading - I had no input about that and at that stage it was way too late to come in and rubbish the design.

And of course a 135 metre 4 core armoured, buried and covered with no input from the company I now work for!

So the major design hurdles to get the system working as efficiently as it could was how to string the system when we were limited to four cores, how to limit the amount of energy exported to the grid and how to keep the system voltages from exceeding 800 volts.

All solar equipment at present is only able to take a 1000volts. The modules and inverters all state that the voltage must not exceed 1000Volts. However most inverters have a voltage mppt tracking range up to usually around 800 volts so once you go over that the inverter will cut out.

It wasn't easy at all but in the end we used an SMA Tripower ee-10 15KW inverter - It was the only inverter that would do the job and took an eternity to get to.

There was three phase onsite as its a mansion not a house. The inverter has two MPPT's.

In the field we built a combiner box and a frame out of unistrut bolted to a concrete pad under the array mounting frame.

Three strings were combined in parallel in the field and one string is just 12 modules wired in series. Kept the positives and negatives in separate boxes to avoid any unwanted fires.

It was a job to prove to the DNO that the inverter would limit the power to 15KW even though it was a 15KW inverter. (Because we had 20KW of modules) and because the DNO latched on to the output current of the inverter - 24amps.

SMA UK couldn't help so we called Germany and the engineer there explained that 24amps was what the current would rise to when the grid voltage was very low.

So the inverter current output varies depending on the grid voltage.

Three strings of 22 series connected modules connected in parallel meant that on the one string we had 719 volts DC and the current for each string was 8.21 amps x3 = 24 amps

The inverters second input was only able to take a small amount of power but it worked out perfectly as the remaining 14 modules were then connected in series on their own.

Certainly proved challenging but we got there in the end!

Thanks for the input from the original posters.


See Also:

FuseTalk Standard Edition v3.2 - © 1999-2016 FuseTalk Inc. All rights reserved.