Calculating string amperages [RE-wrenches]

[Bill Brooks]

Ray,

I don't think my postings are working anymore, so I haven't been responding
on this and other issues. John does a good job of explaining the issues.
I've never heard that charge controllers are only one-way devices so that
would need to be confirmed by each manufacturer. Be careful about making
broad assumptions about products just because other products (i.e. certain
inverters) have specific features (backfeed protection).

Your calculation of overcurrent size left out a very important point that
John tried to make. The total current to the module is the sum of the 15-amp
breaker plus the 6.375 amps from the paralleled string. This means the
module is exposed to 21.375 amps (decimal places are silly) which exceeds
the 15-amp rating of the module. 

Also, the first 1.25 multiplier from 690.8(A) is not for cloud edge
enhancement. Instead it is a general code fudge factor to adjust the rating
of the module for certain locations (high altitude) and times of the year
(October) when the irradiance can sustain at 1200 W/m2 or more for hours.
I've seen cloud edge enhancement last for 15 minutes in very rare instances,
but not for hours--and edge enhancement can exceed 140%-plus you can have
snow reflection on top of that.

As it turns out, nearly all fuses and circuit breakers we use in PV systems
are rated 100% continuous duty, so the second 1.25 multiplier from 690.8(B)
is not absolutely required. However, the fuse and breaker manufacturers
still recommend only running their overcurrent protection at 75% of
rating--read the design guides from Littlefuse and Bussmann for
confirmation. In the field we have historically had lots of problems with
fuses blowing when only sized at 1.25xIsc. These fatigue failures are due to
high current and high ambient temperature (and fuse holder temperature).

You are obviously a very conscientious and intelligent installer, and I
don't want to come across too strongly on this issue, but I don't want you
to mislead yourself into a design concept that does not comply with the NEC
or with good practice. I agree that is sucks to have to put a breaker on
every pair of 72-cell modules, that is life and it ain't fair.

Bill.

-----Original Message-----
From: Ray Walters [mailto:walters at taosnet.com] 
Sent: Friday, February 08, 2008 11:11 AM
To: RE-wrenches at topica.com
Subject: Re: Calculating string amperages [RE-wrenches]



Well we do have some discrepancies here. William was saying we could  
go to 3 parallel Solar World strings, John Berdner here says only  
one, and I'm still saying 2 is good.
John I have 2 issues with your excellent analysis.
1) 690.8 B) (1) exception: "Circuits containing an assembly, together  
with its overcurrent device(s), that is listed for continuous  
operation at 100% of its rating shall be permitted to be utilized at  
100% of its rating."
This means that with the right breaker, we don't have to use the 1.56  
multiplier for array current, only 1.25 which is for edge of cloud  
effect, etc. I have seen Solar world modules exceed their nameplate  
rating, so this multiplier is appropriate.
2 x 5.1 A x 1.25 = 12.75 Amps on a 15A breaker that is rated for 100%  
duty is acceptable. (most Outback and Midnite solar offerings are  
100% rated BTW)

2) All this argument is based on the possibility of back feed from  
the grid or batteries, but my understanding is that both grid tied  
inverters, and charge controllers prevent back feed. So I don't think  
this is applicable to most work that we are doing. (I am aware of  
relay type controllers APT... that could allow backfeed possibly?) I  
suppose with large enough arrays, the backfeed would be from adjacent  
subarrays?

So unless I'm missing something (please help me out here) I think I'm  
going to keep paralleling the SW165s in lower voltage off grid  
systems. I like to set up 4 modules per breaker, with the modules  
wired 2 in series, 2 parallels. If I'm running series strings over 48  
v nom., I'll take another look at not paralleling.

Ray Walters

On Feb 7, 2008, at 3:10 PM, John Berdner wrote:

>
> William/Wrenches:
>
> The "other" they are talking about in 690.8 is the same as the "n-1"
> parallel strings in the fuses white paper I did.
>
> Consider a system with n strings in parallel feeding a single over
> current device.
> If you were to create a fault in one of the strings the faulted string
> does not contribute back flow current to itself.
> The fault current comes from the other strings that are connected in
> parallel, i.e. n-1, and from current flowing backwards through the  
> fuse.
>
> For the n-1, or other parallel connected, strings we have to add a
> safety factor of 1.25 for high irradiance. We do not have to consider
> the second safety factor of 1.25 for continuous duty.  (Why the  
> fault is
> not considered as possibly continuous still alludes me.)
> The resulting fault current is as described in 690.8: Ifault =
> ((n-1)*Isc*1.25 )+the rating of the over current fuse.
> Since the Code requires the over current fuse to be sized for a  
> minimum
> of n*Isc*1.56 this leads to a minimum wire ampacity / UL series fuse
> rating of >= ((n-1)*1sc*1.25)+(n*Isc*1.56).
>
> Using the example from another recent post with two parallel  
> SolarWorld
> modules with an Isc of 5.1 Amps and a UL series fuse rating of 15  
> Amps:
>
> 1) The fault current from the "other parallel connected strings" is
> 1*5.1*1.25 = 6.375 Amps.
> 2) The minimum over current fuse size / wire ampacity required by Code
> for 2 strings would be 2*5.1*1.56 =  15.91 Amps. Unfortunately the
> standard fuse sizes are 15 Amps (too small) and 20 Amps (next size
> larger) so the Code would require a 20 Amp fuse.
> 3) The total available fault current would then be 6.375 Amps for the
> modules plus another 20 Amps for the fuse or 26.375 Amps.
>
> Since the UL series fuse size for the module is 15 Amps there is no
> Code / UL compliant way to connect 2 parallel strings of this  
> module to
> a single fuse in an application where it is possible for current to  
> back
> feed the fault through the fuse.
>
> The minimum fault current of 26.375 Amps also applies to the wire
> ampacity of the module interconnecting wires per 690.8.  Since most
> modules have AWG 12 leads this would not pass 690.8 regardless of  
> the UL
> series fuse rating.  With a possible fault current of 26.375 Amps  
> AWG 10
> is marginal and likely would not fly after ambient temperature  
> derating
> even with 90C wire.  Modules with AWG 8 leads ?
>
> I hope this helped to add clarity rather than muddy the waters.
>
> Best Regards,
>
> John Berdner
>
>
>>>> wrmiller at charter.net 2/5/2008 01:23:54 PM >>>
>


- - - -
Hosted by Home Power magazine

To send a message: RE-wrenches at topica.com

Archive of previous messages: http://lists.topica.com/lists/RE-wrenches/read

List rules & how to change your email address: www.mrsharkey.com/wrenches/etiquette.php

Check out participant bios: www.mrsharkey.com/wrenches/

Moderator: michael.welch at homepower.com
--^----------------------------------------------------------------
This email was sent to: michael.welch at re-wrenches.org

EASY UNSUBSCRIBE click here: http://topica.com/u/?bz8Qcs.bz9JC9.bWljaGFl
Or send an email to: RE-wrenches-unsubscribe at topica.com

For Topica's complete suite of email marketing solutions visit:
http://www.topica.com/?p=TEXFOOTER
--^----------------------------------------------------------------