[RE-wrenches] Combining Multiple Inverters

[Brian Teitelbaum]

Corey,



The wire ampacity has to be sized for the OCPD, as this is potentially a
bidirectional wire in a fault. While the inverter is current limited, the
grid it’s connected to is only limited by the Service Main, and that 50A
OCPD, which would exceed the ampacity of #10 wire.



Brian Teitelbaum

AEE Solar







*From:* RE-wrenches [mailto:re-wrenches-bounces at lists.re-wrenches.org] *On
Behalf Of *Corey Shalanski
*Sent:* Monday, June 30, 2014 11:38 AM
*To:* re-wrenches at lists.re-wrenches.org
*Subject:* Re: [RE-wrenches] Combining Multiple Inverters



Thank you to all who responded to this thread - particularly Jason/Dave/Jay
for pointing out the maximum OCPD requirements.



Getting back to my original question.. I would like to use Dave's
"extremely small subset" example of two SB5s operating at 277V. In this
case, if we were to parallel the output of each inverter onto a 50A OCPD
(#6 wire), could we downsize the wire to #10 from the split going back to
each individual inverter? or would we be better served (required?) to run
#6 all the way back to each unit? Since we are dealing with current-limited
devices I would tend to think we can downsize but wanted to get Wrenches'
feedback on this.



--

Corey Shalanski

Joule Energy

New Orleans, LA

ᐧ



On Fri, Jun 27, 2014 at 12:41 PM, <re-wrenches-request at lists.re-wrenches.org>
wrote:


Message: 5
Date: Fri, 27 Jun 2014 11:32:02 -0400
From: Dave Click <daveclick at fsec.ucf.edu>
To: re-wrenches at lists.re-wrenches.org
Subject: Re: [RE-wrenches] Combining Multiple Inverters
Message-ID: <53AD8E72.4080009 at fsec.ucf.edu>
Content-Type: text/plain; format=flowed; charset=ISO-8859-1

This is an interesting conversation, but I'm not sure it's going to
really impact anything.

When an inverter is listed, it is tested with some "maximum output
overcurrent protection" as part of its rating per UL 1741. Typically
inverter manufacturers set this to be just above the inverter's max
current + 25%. As you know, many inverters can work at more than one
voltage, so for the SB5s their max current ratings at 277/240/208V are
18/20.8/24A, respectively. The max overcurrent protection is 50A. This
figure often appears in the manual but not in the data sheet. For SMA,
it's in the manuals. For some manufacturers, it's only available from
tech support.

For the Sunny Boy line, the only way you could have more than one
inverter on a shared OCPD is if you have two SB5s operating at 277V
(18.1A rated current each, so 2x18.1x1.25 < 50A). The SB6, SB7, and SB8
all have currents too high to make this work, as they share the 50A max
OCPD. The SB3/3.8/4 have a max OCPD of 30A, which is too low to
double-up since each SB3 requires an OCPD of at least 20A. The TLs don't
work either, as the max for any unit is too low. I don't see it in the
SMA manual, but I know that in other manuals I've seen a manufacturer
requirement for a dedicated OCPD. The only inverters I know of that have
a max OCPD rating far beyond their rated current are micro-inverters.

As for the shared disconnect, since at least the 2005, 690.15 has noted,
"A single disconnecting means in accordance with 690.17 shall be
permitted for the combined ac output of one or more inverters or ac
modules in an interactive system." That being said, the NEC requires ac
modules and micro-inverters to have ac disconnects (connectors,
typically) per 690.6 and 690.15(A). 690.15 notes that an inverter be
able to be disconnected from all sources, but if doubling up these SB5s
you could argue that this shared disconnect accomplishes that via the
anti-islanding protection... so I agree now that this is a grey area.

Anyway, it comes down to whether it's worth worrying about this for what
is an extremely small subset of compatible inverter configurations, when
you can just do something like install a $50 Eaton BR24L70RP (70A bus,
240V, 4 spaces) with two 30A breakers and be on your way.

DKC
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.re-wrenches.org/pipermail/re-wrenches-re-wrenches.org/attachments/20140630/a31091e0/attachment-0002.html>