<html><body><span style="font-family:Verdana; color:#000000; font-size:10pt;"><div>Kelly- You've got phase-balancing issues that will not be entirely solved with 3-phase breakers, but that's a good place to start. As Jarmo wrote, some inverters will talk to each other to shut them all down when one stops. Many manufacturers are now building 3-phase inverters to solve this phase-balancing issue. <br></div><div>So your proposed solution with 3-phase breakers on both the PV and grid side is a necessary but not sufficient upgrade. </div><div>Your real problem is overloaded neutral, as indicated by the melted insulation. If you put an OCPD on the neutral (not allowed to break neutral if it is grounded) you would discover that it is running a lot more current than it is supposed to. <br></div><div>It's not clear to me why you chose delta connection on the PV side. A delta-wound transformer has the neutral as a center tap on one of the windings. The neutral is not balanced on all windings. If the neutral is center-tapped between legs A and B, the voltage from either A or B to neutral is the same, but from C to neutral is much higher. Each of A or B sees neutral as a half-winding away, while C is a winding-and-a-half away from neutral. It is called the High Leg or Stinger Leg. </div><div>See the excellent WikiPedia listing with pictures for Delta windings. <br></div><div><br></div><div>Then while an inverter across A-B is balanced on neutral, the inverters across B-C or A-C are not. Those inverters will have a lot of neutral current, leading to overloaded neutral as you are seeing. </div><div>So you've got to get rid of that Delta wound transformer. Replacing it with a step-down transformer in reverse is not a fix, because that still has a delta winding. <br></div><div>You could replace it with a 208-480 step-up with both sides in Wye configuration. Maybe you can find a straight 2-to-1 240-480 3-phase transformer with both sides in Wye; I'm not sure if these are made in 3-phase. This would allow all the inverters to see the same neutral. </div><div>Check the inverter spec sheet-- They can work with 208 Vac as well as 240 Vac although with a bit more AC current to get their power rating. </div><div>The spec sheet will also show that many inverters (such as Sunny Boys) can also connect 277-to-neutral direct to the 480 grid without a big 3-phase step-up transformer at all. You did not specify which inverters you are using, but with your system configuration using several 3.8kW inverters this would seem like the best re-configuration. And it has the added advantage of not having the transformer parasitic load on the grid side all night. </div><div>Be sure to use 3-phase breakers as well. </div><div>If you send me your system schematic back-channel I'd be happy to suggest a least expensive reconfiguration. </div><div>Don Barch</div><div>Energy Solar<br></div><blockquote id="replyBlockquote" webmail="1" style="border-left: 2px solid blue; margin-left: 8px; padding-left: 8px; font-size: 10pt; color: black; font-family: verdana;"><div id="wmQuoteWrapper">
Message: 2<br>
Date: Thu, 2 Apr 2015 15:19:11 -0700<br>
From: "Kelly Keilwitz, Whidbey Sun & Wind" <<a href="mailto:kelly@whidbeysunwind.com">kelly@whidbeysunwind.com</a>><br>
To: RE Wrenches listserve <<a href="mailto:re-wrenches@lists.re-wrenches.org">re-wrenches@lists.re-wrenches.org</a>><br>
Subject: [RE-wrenches] PV step-up transformer design<br>
Message-ID: <<a href="mailto:7010DA9F-54AD-477E-B855-6DEC65CF890A@whidbeysunwind.com">7010DA9F-54AD-477E-B855-6DEC65CF890A@whidbeysunwind.com</a>><br>
Content-Type: text/plain; charset="utf-8"<br>
<br>
Wrenches,<br>
We have issues with a 1-1/2 year old 25 kW grid-tied PV system utilizing 2 step up transformers from the array to grid. Actually it?s two identical, 12.5 kW systems (for incentive and financing reasons) on each transformer. Note that the inverter size and voltage selection was also dictated by incentives. Each of the two systems are designed as follows:<br>
<br>
- 3, 3.8 kW, 240 Vac inverters (16A max output current) across each phase of the 240V, 3-ph Delta primary of a 15 kVA transformer. Each inverter 1-ph output has a fused disconnect. The output of all three inverters is combined into 3 phase delta configuration with terminal blocks before the transformer. Other than the inverter AC disconnects there is no OCPD on the 3-ph primary side of the transformer.<br>
<br>
- Transformer secondary to grid is 277/480 wye. The secondary output has two fused disconnects (one at each end of the 600-ft run between array and line-side connection) each with 20-A fuses.<br>
<br>
After more than year of operation we noticed one inverter, in one of the two identical systems, was down (monitoring showed it happened past December), indicating a grid problem. One fuse (of the 3) in the secondary (grid-side) fused disco had blown. Replaced that fuse, but then 2 fuses at the disconnect near the meter, 600 feet away, popped. <br>
<br>
Systematic checking of the system reveals no shorted conductors, but a problem at the transformer. Close inspection revealed that the insulation on our 90?C neutral conductor had slightly melted where it touched the X-O neutral wire of the transformer (the splice looked fine). Now it gets more interesting?.<br>
<br>
In the last week, while troubleshooting the first 12.5 kW system, its adjacent twin system had an inverter quit with a ground-fault error (these inverters commonly will display ground fault errors for other reasons). One of the 20A fuses in the secondary (grid 277/480) side had popped. And, yup, the neutral wire had signs of overheating. We shut all three inverters down in that system. The transformer doesn?t seem damaged, hopefully because we caught it in time. But we?re looking at a replacement transformer for the first system.<br>
<br>
Here?s what we think may have happened: The grid went down and, on coming back on, the in-rush current to the secondary side of the transformer popped a (undersized) fuse, taking out one leg of the 480. One inverter on the primary side saw a bad grid and dropped out, but the other two connected and kept producing. This unbalanced production resulted in the neutral of the secondary wye (grid) side of the transformer, being overloaded. Over 3 months this unbalanced operation damaged the transformer.<br>
<br>
Initially we replaced the fuses near the line-connection with 60A (it?s a hot-swap and we don?t want those to go again unless the conductors need it), and the fuses near the array transformer secondary with 30A, thinking that will handle the inrush current to the transformer after a grid outage better than the 20A fuses. However, after a lot of discussion, research, and consideration of the risk (of replacing another transformer) we?re wondering if we need to alter the design more substantially. <br>
<br>
The main design change we?re considering is to replace the fused disconnect on the secondary side with a 3-phase breaker, that will shut off all phases of the grid supply simultaneously in case of an over-current fault. But, what if one inverter just plain quits and that breaker stays on? Will the neutral on the secondary still be overloaded from the other two?<br>
Other changes we?re pondering include:<br>
- Do we need a ganged 3-ph breaker on the inverter side, so that all three inverters go off at once? This will only be useful in case of over current on an inverter output, which isn?t likely, and is now addressed with the fused 240 AC 1-ph disconnects on the inverter outputs.<br>
- Use a step-down transformer in reverse. I.E. use a 277/480 wye primary and 240 delta secondary and back feed the secondary with the PV. Some info from other sources suggest that this will reduce the inrush current from the grid after an outage. I think this would be inefficient at transmitting the PV power.<br>
- Can/should we remove the neutral connection to the transformer on the secondary side? There are no 277 loads or sources.<br>
- Can/should we put OCP on the neutral to the transformer?<br>
<br>
Your suggestions and experience will be appreciated. This has additional importance as we are preparing to install another 75 kW at the same site utilizing 240Vac inverters to the 277/480 grid.<br>
<br>
Thanks,<br>
-Kelly<br>
<br>
Kelly Keilwitz, P.E. <br>
Principal<br>
Whidbey Sun & Wind<br>
Renewable Energy Systems<br>
NABCEP PV Installation Professional<br>
WA Electrical Administrator<br>
<a href="mailto:kelly@whidbeysunwind.com">kelly@whidbeysunwind.com</a><br>
PH & FAX: 360.678.7131<br>
<br>
</div>
</blockquote></span></body></html>