<table cellspacing="0" cellpadding="0" border="0" ><tr><td valign="top" style="font: inherit;">Allan's point is well taken.<br><br>Resistive connections within the PV could easily be another cause for<br>excessive DC-side voltage drop, and clearly merits closer examination.<br><br>Dan<br><br><br>--- On <b>Sun, 7/22/12, Allan Sindelar <i><allan@positiveenergysolar.com></i></b> wrote:<br><blockquote style="border-left: 2px solid rgb(16, 16, 255); margin-left: 5px; padding-left: 5px;"><br>From: Allan Sindelar <allan@positiveenergysolar.com><br>Subject: Re: [RE-wrenches] Low power production at a winery: DC voltage drop upon SB6000 start-up<br>To: "RE-wrenches" <re-wrenches@lists.re-wrenches.org><br>Date: Sunday, July 22, 2012, 12:25 PM<br><br><div id="yiv485511530">
<div>
<div class="yiv485511530moz-cite-prefix">Eric,<br>
Without challenging any of Dan's calculations, I will add that the
BP SX170 has proven to be prone to failures. You didn't indicate
which model of BP170 you have. I think that the SX170s are the
model that was the source of BP's conditional recall of modules
that could cause roof fires - no, I'm not making this up, but
neither is my memory sharp about this. <br>
<br>
Here's the text of an email dated 10/10/07 from Phil Undercuffler,
now at Outback, then at Conergy:<br>
<blockquote>
<pre>
BP reports that these failures tend to manifest early in the product
life; modules that will have problems will generally fail in the first
few years of service. Failure rates are reported as .02% of product
distributed in US, .03% in European market
Cause:
The solder joint inside the low-profile potted J-box on MC connected
modules is what is failing.
Changes in solder formulation (due to the phase out of lead solder
for environmental reasons) and move to heavier gauge wire made it
harder to make a good solder joint.
Failure of the solder joint causes heating and arcing, which can in
certain cases ignite the epoxy material surrounding the joint.
BP's concern is the potential for the burning epoxy to ignite
flammable materials in the close vicinity of the junction box.
BP does not believe that standard roofing materials such as asphalt
shingles, tile or metal roofs will be affected by this issue.
A dealer can perform tests with a relatively inexpensive infrared
thermometer or contact thermal probe. It is best to measure the
module from the front, while in operation.
Problem modules will show 10-20 C difference between the solder
joint and the module field just before failure.
For solder connection, gather the following info:
Measure VOC and ISC
Physical examination, looking for brown or black spots on
solder traces
Model number
Serial number</pre>
</blockquote>
<br>
They are rated at +/- 9%, one of the worst specs around at that
time. We used quite a few of them at that time and have replaced
many of them under warranty, including one entire 40-module
commercial array, identical to the one you described.<br>
<br>
I would still encourage testing individual string outputs for Vmp
and Imp under load, to look for significant variances. At worst,
it gives you a baseline reading. Also, if you can, look for
browned hot spots, visible beneath he glass in the area over where
the leads leave the module back. <br>
Allan<br>
<br>
<div class="yiv485511530moz-signature">
<title></title>
<font face="Times New Roman, Times, serif"><b>Allan Sindelar</b></font><br>
<small><a rel="nofollow" ymailto="mailto:Allan@positiveenergysolar.com" target="_blank" href="/mc/compose?to=Allan@positiveenergysolar.com"><font color="#000099" face="Times New Roman, Times, serif"><u>Allan@positiveenergysolar.com</u></font></a></small><font face="Times New Roman, Times, serif"><br>
<span style="font-size:10pt;">NABCEP Certified Photovoltaic
Installer<br>
NABCEP Certified Technical Sales Professional<br>
New Mexico EE98J Journeyman Electrician<br>
Founder and Chief Technology Officer<br>
<b>Positive Energy, Inc.</b><br>
3209 Richards Lane (note new address)<br>
Santa Fe, New Mexico 87507<br>
<b>505 424-1112</b><br>
<a rel="nofollow" target="_blank" href="http://www.positiveenergysolar.com/"><u>www.positiveenergysolar.com</u></a></span></font>
<p class="yiv485511530MsoNormal" style="margin-right:722.25pt;"><span style="font-size:10pt;"><font face="Times New Roman, Times,
serif">
</font><b style=""><br>
</b></span></p>
<span style="font-size:10pt;"><br>
</span> </div>
On 7/22/2012 10:56 AM, Exeltech wrote:<br>
</div>
<blockquote type="cite">
<table border="0" cellpadding="0" cellspacing="0">
<tbody>
<tr>
<td style="font:inherit;" valign="top">Eric,<br>
<br>
I checked my magic spreadsheet for the BP SX170.<br>
<br>
Using 800W/m^2 irradiance, and a presumed ambient of 37C
(98F),<br>
everything appears to be working within specified
tolerances. The<br>
presumed conditions create exactly your measured Voc,
which is<br>
why I used them. Other temperature and irradiance
combinations<br>
will also work...<br>
<br>
The voltage delta you measured between Voc and Vmpp is
within<br>
BP's specifications. No PV has a perfect "fill factor".
Fill factor is the<br>
aspect that determines the shape of the V/I curve
established when<br>
Voc and Vmpp are plotted. The BP you mentioned have a
fill factor of<br>
0.769, which is very typical of polycrystalline PV. A
perfect fill factor<br>
would be 1.0 (Vmpp equal to Voc), which obviously doesn't
happen.<br>
Thus, all PV have an operating voltage at maximum power
that's below<br>
the open circuit voltage -- some more than others.<br>
<br>
<br>
Measured:<br>
375Voc<br>
280V at inverter at mpp<br>
14-15A (fluctuating) at mpp<br>
<br>
Calculated:<br>
375Voc<br>
301V mpp<br>
14.96A mpp<br>
<br>
Assumed:<br>
800 w/m^2<br>
37C (98F) amb<br>
<br>
4,503W array power under above conditions<br>
<br>
280V measured at the inverter<br>
14.5A<br>
<br>
Presuming 301V is correct Vmpp, this equates to:<br>
<br>
1.38 ohms total system R, and 290W loss<br>
<br>
290W/4503W = 6.44%<br>
<br>
6.44% total loss --higher than a 2.5-3% best-practices
target, and likely<br>
due to slightly oxidized and/or loose connections in the
DC-side of the<br>
system. Could also be due to long conductor runs,
slightly undersized<br>
conductors for the amperage and length, or both.<br>
<br>
As Jeff Quackenbush mentioned, Vmpp is slighly lower than
one might<br>
expect, but doesn't appear to rise to the level of
something that would<br>
yet be indicative of something wrong.<br>
<br>
<br>
Dan<br>
<br>
<br>
<br>
<br>
--- On <b>Sat, 7/21/12, SunHarvest <i><a rel="nofollow" class="yiv485511530moz-txt-link-rfc2396E" ymailto="mailto:eric@harvesthesun.com" target="_blank" href="/mc/compose?to=eric@harvesthesun.com"><eric@harvesthesun.com></a></i></b>
wrote:<br>
<blockquote style="border-left:2px solid rgb(16, 16,
255);margin-left:5px;padding-left:5px;"><br>
From: SunHarvest <a rel="nofollow" class="yiv485511530moz-txt-link-rfc2396E" ymailto="mailto:eric@harvesthesun.com" target="_blank" href="/mc/compose?to=eric@harvesthesun.com"><eric@harvesthesun.com></a><br>
Subject: [RE-wrenches] Low power production at a winery:
DC voltage drop upon SB6000 start-up<br>
To: "RE-wrenches"
<a rel="nofollow" class="yiv485511530moz-txt-link-rfc2396E" ymailto="mailto:re-wrenches@lists.re-wrenches.org" target="_blank" href="/mc/compose?to=re-wrenches@lists.re-wrenches.org"><re-wrenches@lists.re-wrenches.org></a><br>
Date: Saturday, July 21, 2012, 3:19 PM<br>
<br>
<div id="yiv485511530">
<style></style>
<div>
<div><font face="Arial" size="2">Hello Wrenches,</font></div>
<div> </div>
<div><font face="Arial" size="2">I am helping a
client troubleshoot a complaint about
underproduction of his on-grid system installed
about 6-8 years ago. At first I thought I might
encounter the bad Kyocera KC120 issue but it
turns out they have BP170's, 40 of them for an
STC power rating of 6.8kW. </font></div>
<div> </div>
<div><font face="Arial" size="2">Ratings of the BP
modules: PTC=150.7, Vmp=35.4, Voc=44.2, Imp=4.8,
Isc=5</font></div>
<div> </div>
<div><font face="Arial" size="2">The system has four
panels of ten modules wired in series for a
nominal rating of about 350v, 4.8a per string.</font></div>
<div> </div>
<div><font face="Arial" size="2">At the DC
disconnect (first accessible combiner) I
measured: All strings right at about 375v(oc),
4.7a(mp). After the strings are combined, at the
DC input of the SB6000 (with the AC power to the
inverter OFF) I measured about 375V as expected.
Once the AC power is connected, the
SB6000 starts up, and MPP operation is
initiated, the DC voltage drops to about 280V,
and amps sit between 14 & 15A. I didn't
think the voltage was supposed to drop upon MPP
tracking...especially this much. This voltage
drop would account for the observed power loss
between actual and rated production values. I'm
going to check with SMA but I wanted to ask the
experts here too, as someone may advise
something like, "Oh yeah, BP modules have a
similar defect as the Kyocera..."</font></div>
<div> </div>
<div><font face="Arial" size="2">My questions:</font></div>
<div> </div>
<div><font face="Arial" size="2">Is the DC voltage
supposed to drop significantly in MPP mode on
these inverters?</font></div>
<div><font face="Arial" size="2">If not, does this
indicate a bad inverter?</font></div>
<div><font face="Arial" size="2"><br>
Anyone know of problems with BP modules
manufactured around 2006-2008?</font></div>
<div> </div>
<div><font face="Arial" size="2">The Kyocera modules
showed good volts and amps until a load was
connected. Seems like a similar issue here.</font></div>
<div> </div>
<div><font face="Arial" size="2">Eric Stikes<br>
SunHarvest Solar<br>
A Sustainable Energy Group Partner<br>
+1 (530) 798 - 3738<br>
<a rel="nofollow" target="_blank" href="http://www.harvesthesun.com">www.harvesthesun.com</a></font></div>
</div>
</div>
<br>
</blockquote>
</td>
</tr>
</tbody>
</table><br></blockquote></div></div><br></blockquote></td></tr></table>