PVI-5000-OUTD-US 5000 watt Inverter

Re: PVI-5000-OUTD-US 5000 watt Inverter

LucMan,

Thank you for your comments.

You have pointed out an advantage of having two inverters. Everything else being equal, that provides a good way to check some of the system operation. If you read my long chain on my Fronius inverter, you may notice that it took me quite a while to notice the problem.

Because I was so easily fooled by the Fronius, I have been watching the Power-One very closely. Everything seems right on with it and all indications are that it is performing better than expected taking both the Fronius oscillation and the modest increase in efficiency into consideration. However, it will be a few more months before I can make a direct comparison since my system didn't come online until about June 1 of 2008. I am basing my conclusions on the daily readings of my non-billing AC meter, not the on board inverter measurements.

Of course, that begs the question. Do you have two non-billing meters? Did you take independent voltage and current measurements? Or, did you use the internal measurements of two very different inverters?

I haven't heard any comments specifically about Power-One or SMA, but I have been warned that inverters have poor and often optimistic measurement hardware.

If you are using a single non-billing meter as I suspect did you record daily readings for your old mismatched inverters in June 2007? Did you record the readings again with the matched inverters in June 2008? If you did both of those things, how about taking a look at that data and posting again?

You mentioned that your two strings are identical, but why would they ever have had different inverters, and why did you end up with a SMA 3800 and a SMA 2500? I would have expected identical inverters all along, unless your arrays have different orientations. Do they?

I didn't know that the PVI 3000, which is transformerless, was certified in the USA in 2007. The SMA 3800 and 2500 are transformer type inverters. These two inverter types use very different wiring on the DC side. How did your installer deal with this? Did (do) you have two DC disconnects? How were the disconnects wired? How were the strings wired?

Thanks,
Mike

Re: PVI-5000-OUTD-US 5000 watt Inverter

LucMan,

Thank you for your reply and added information.

I have come to the same conclusion as you did that the P1 was not harvesting the same amount of energy as the SMA. However, since you should have been well within the P1 specifications, I think that it bears further study to figure out what was going on. Do you still have the voltage and current readings of solar noon from 2007? What were they for the two array / inverter combinations? Did you and your installer try to figure out what was going on?

I asked about the disconnects and the DC wiring because, while the two types of inverters both require the array wiring to be floating, the transformer design requires only the positive leg to be disconnected in the DC disconnect and the transformerless design requires both legs to be disconnected. In addition, the transformer inverter internally grounds the negative through a fire protection fuse. Were the DC disconnects integrated into the inverters, or were they external? If the DC disconnect was external, was there one or two?

Where are you located? The US electrical code requires some special things be done with the array wiring for a trnsformerless inverter. Did (does) your panel wiring meet code?

Thanks,
Mike

Re: PVI-5000-OUTD-US 5000 watt Inverter

Bill,

Thanks for your note.

I was asking LucMan for the paired P1 and SMA voltage and current readings that he took back in 2007 to determine that the P1 wasn't performing as well as the SMA. These should have been taken over just a few minutes on a clear day so should be pretty good.

My system is producing about 3700 watts in the mid day sun these days which seems correct for the dissimilar orientation of my arrays and the time of year. However, those cloud edges push it up past 5000 watts. Too bad cloud edges come with clouds.

Mike

Re: PVI-5000-OUTD-US 5000 watt Inverter

Hello LucMan,

Thanks again for the information.

I apologize for causing some confusion. I was trying to cover 2 subjects. One was why did the P1 not, or appear not to, perform to specifications. The other was how the unit was wired to meet code since transformerless inverters are yet very rare in the USA. Of course, you are correct that the DC wiring has no effect on the AC power.

What I was hoping to get from you were voltage and current readings for the two inverters at the time that you observed the problem. I would really like to understand what was going on because if there is some design flaw in the P1, I would very much like to know about it. So, if you still have those readings, please post them. If not, then do you remember approximately what they were ... not the DC power, but the DC voltage and current.

For comparison sake, my AC output as measured on my non-billing meter was 26 kWh yesterday. My system has 28 SuntTech 180 panels, or 5040 DC watts. There was only a small amount of high cloudiness yesterday.

By the way, how do you know the AC power output for yesterday? Earlier, you mentioned that you do not have a non-billing meter. Others have pointed out that the inverter internal readings are not very accurate.

I apologize for being so anal about this. I respect your decision to quickly reject the P1 and move on. I should have done that with my Fronius. However, I disagree that you should call the P1 inferior without knowing what the problem was. I see at least the following possibilities:

1) Your P1 had a defect.

2) The P1 model has a design defect.

3) The measurements that you based your decision on were faulty.

4) Your panels are not rated for a transformerless system.

5) The installation / inverter setup was faulty.

Item #4 brings up why I keep asking about the DC wiring. I have pointed out that the use of the P1 inverter requires unique electrical code considerations. I am not an expert on this, but in 2007, I believe that the inter panel wiring and the panel feeds all needed to be in conduit. The building inspector may not have realized that the P1 was transformerless since this inverter type was very rare in 2007. Are your panels wired with conduit? (The 2008 electrical code changes some of this and is not relevant to this discussion.)

Or putting it another way, if your contractor chose the unique P1 inverter and then did not follow through with wiring it correctly, why should it be a surprise that it didn't (or appear to) meet expectations?

Again, I apologize for being so persistent and you if don't want to spend any more of your time on this, I'll understand. Perhaps another reader of this chain will join us and can help explain just what was going on.

Thanks again,
Mike

Re: PVI-5000-OUTD-US 5000 watt Inverter

LucMan,

Thanks again for the response.

I think that you are tired of this conversation since you didn't address several of my questions. I don't blame you. Thank you for taking the time to post your experience and taking the time to follow up yet again.

The only measurements that you supplied that are complete enough to draw any conclusion are those for 10/21/07. There are only two conclusions possible for these results. Either your two strings were not identical at the time (shade?), or your measurements are wrong. No inverter malfunction can account for your measurements. Here is my reasoning.

Below are your 10/21/07 1 pm measurements again with the currents computed using your watts figures:

SMA 285 vDC 7.09 amps (2020 W)
Mag 248 vDC 6.73 amps (1670 W)

Please assume:
The measurements that you supplied are relatively accurate, and
Your two strings are similar in every way, including illumination and temperature.

While your system is in operation, the arrays act as energy sources. From any PV panel manufacturer's product graph, it can be observed that the slope of the current / voltage curve of an illuminated panel is negative. In other words the more current that is drawn from the panel, the lower the voltage. Since the panels are placed in series for an array, the array acts the same way as an individual panel.

An inverter acts as a load on the array. It has the complicated job of determining exactly what current to draw from the array to harvest the maximum energy, but that is not relevant to this discussion. What is relevant is that for the PV array, higher currents draw MUST result in lower voltages. Your measurements show a lower voltage with a lower current. Therefore, one of the starting assumptions must be wrong. There is no evidence of an inverter problem.

A "red flag" to me was your installer's use of dissimilar inverters on a symmetrical system to begin with. If your installer fixed your system when he replaced the inverter, he may have actually done something else that was the real fix and perhaps he "forgot" to mention it to you.

Thanks again,
Mike