I've been looking at several Xantrex efficiency ratings on Mike90045's site, here.



Efficiency #s vary from 54% to 95+% between installations. Can any one explain this? Maybe the low efficiency sites (wennfred and magneto) don't have their inverters set up optimally.

Guppy is the number one person to answer the question--but I will take a stab at it...

With the inverters, you have two major portions of energy loss (typically as heat)... Some losses are related to current flow (also known as I squared R or P=I^2 * R). As current increases through a resistor, the power loss (as heat) increases (double the current, get four times the heat).

So this first loss is going to be most at high power, and quite small at lower power. But, in any case, the power loss is related to the amount of current going through the inverter/system/wiring).

The second is the "tare" loss... This is the amount of energy simply to run the computer, switching electronics, etc. inside of the modern inverter. This power is, roughly, a fixed amount of power loss.

So, to use a (guessed at numbers) example. My Xantrex GT 3.0 (3kW peak) is about 95% efficient. Assume 5% loss of 3 kWatts would be 150 watts. Say that 100 watts is loss through the I^2*R losses of the FETs (electronic switches and internal wiring). And 50 watts is used to run the electronics (computer, switchmode power supply, etc.).

As the power falls from 3 kWatts, the I^2*R part will fall pretty quickly (current goes down by 1/2, loss to heat is now only 1/4). So at 1.5 kWatts, your losses here are 25 watts.

But, the tare loss is still 50 watts. So now, your losses are 25+50=75 watts . EFF = (1,500-75)/1,500=95%

But, as the power drops even more, say you have clouds (or morning/evening operation) and you are now down to 300 watts:

You would get about (300-1w+50watts)/300=83%... And the lower you get, the worst the efficiency becomes.

At 100 watts, you would have 50 watts of loss and only about 50% efficiency.

So, you really need to graph the efficiency vs power through the inverter and see where each system is operating at to understand if there are any problems or not.

For example, I just took a quick look at your link and see that they are all at 86% or better efficiency now.

Note, the above numbers are pure guesses picked for easy, round numbers (I think the real efficiency drops faster with declining output than I have shown above)... The actual numbers are different and the losses more complex--but I think it helps you to see what several major contributors to the efficiency are and how they change over operating ranges.

-Bill

The Sunpower inverter on Mikes list is "cooked" and I'm assuming its calabration is way off, what ever the reason, it is not normal performance all the other sites are in the mid 90% range which is normal

for actual, UL verified performance go here

http://www.consumerenergycenter.org/erp ... summaries/

select you inverter and look at the graphs, these as VERY accurate, real world performance, at low/mid/high PV operating voltage from min to max power.

The GT series is the most efficient inverters available, with the most features standard and are the lowest cost ... a tough combination to beat!

My hat's off to you SG! That is a very impressive efficiency curve--and at the best prices. Great engineering!

Looks like my guesstimates were not too far off for first order efficiency effects.

-Bill

This is what Xantrex told me:

Please keep in mind the GT is not a lab quality power analyzer and is spec'ed to be +/- 1% on the readings. Efficiency Percentage is calculated from power in / power out, then the display accuracy. If the input power reading is off by 1%, the output power reading is off by 1% and the display is off also by 1% if would definitely account for the noticed discrepancy and then some.


I really would like to get my Eff up, not sure what to check or do. Could it be the angle of the panels?



Fred

I'm wondering if wennfred would chime in, something looks pretty
http://members.cox.net/wennfred/GTVIEW/GT_Screen_Save.png
tired on his array, as it's currently running just over 1,000 watts, 81F, and 89% efficiency [ 3193W peak ]. That seems unusually low, and should be well above the low efficiency floor when running low power (< 200W). I know if my setup was doing that, I'd be screaming !

It not real ( 89% ) it’s just what the unit is reporting. The Spec is +/- 1% + 30 watts, which is another 3% at 1kw. The factory calibrates the unit, but it not likey this is the same operating conditions as the inverter runs for the customers ( units stuck in garages come to mind )

This is the downside of having GT-View, everyone wants to see lab-quality generated numbers but one needs temperature stabilized, power corrected and very accurate circuits to do this.

All units perform in the real world very close to the CEC graphs. I know how you’re all feeling, but the truth is it has no effect on what you’re selling into the grid.

Its been 2 UGLY May Grey days here in San Diego, the Cloud cover is very thick, a few hours ago the Sun peeked in for a few minutes and the efficiency shot up to the normal 93.3 percent.

Can't wait to get this May Grey & June Gloom weather over so I can see the real numbers.


Fred

Amen. June Gloom is in full force, but yesterday was one of the first mostly-sunny days since I installed my array.

(22 Sharp 208W panels, 2x11, Xantrex 4.0GT)



However, even with the clouds, the system seems to get enough peaks to even things out.


I'm actually surprised that the panels are able to get this much out of a very cloudy, overcast week.

Efficiency seems to be stable around 95% for the daylight hours. It's just morning and night when the system is milking those last "100W" minutes.

Nice graphs, esp the bottom one with eff & temp. I'd be interested in a copy if you have a template for release.

I see you and I have the same array, 2 strings of 11ea, Sharp 208's

Solar Guppy, what's the difference, if any between GT3.8 & 4.0 ? Just a software setting, or any real hardware difference ?
(I also see you've been busy shuffling strings around, and have the new inverter fired up.) Does the GT-5 have 2 MPPT trackers in it for 2 strings, or are the 2 arrays just combined in the wirebox ?