Comments on 7.2kw system in San Diego

[sensij]

The inverter output at the end of the day is exactly the same as the monitoring sw. But if you take the 15-min reported wattages, sum and divide by 4 you get a slightly less number eg 3%. Without any proof I am guessing it has something to do with transmission loss or how the 15 min wattage is reported. If you look at the lcd on the inverter the wattage changes instantaneously. Somehow an average has to be calculated and reported.

Agreed, it is not the monitoring software, but the multiple ways that the inverter does calculations that leads to the discrepancy. I've been working with SolarEdge support on this, and they have not dismissed my observations as lightly as some forum members have. The discrepancy is not simply due to averaging, and there are a couple ways that this can be proved, which I intend to present once I have the whole process more clearly understood. My recommendation is to trust the power readings, and sum the power to get energy, and have less faith in the direct energy readings, but until I can prove why, skepticism is welcome and warranted.

[thejq]

I don't think is the inverter itself. Inverter actually report at actual output without any loss or power consume by inverter itself. It is the monitoring system.

For those who have solaredge system, at end of the day, check both reading from inverter and monitoring software.

The inverter output at the end of the day is exactly the same as the monitoring sw. But if you take the 15-min reported wattages, sum and divide by 4 you get a slightly less number eg 3%. Without any proof I am guessing it has something to do with transmission loss or how the 15 min wattage is reported. If you look at the lcd on the inverter the wattage changes instantaneously. Somehow an average has to be calculated and reported.

[silversaver]

I've got a sneaking suspicion that most inverters (string anyway) overestimate array output by something like that amount but I don't have #'s to back it up yet. I suspect mine (rebadged PowerOne) reports high by something like that amount.

I don't think is the inverter itself. Inverter actually report at actual output without any loss or power consume by inverter itself. It is the monitoring system.

For those who have solaredge system, at end of the day, check both reading from inverter and monitoring software.

[ericf1]

With regard to the daily energy reported by solaredge inverters, I've been collecting evidence that they systemically inflate the total by about 3%. On clear days, a more accurate total can be generated by summing the reported 15 min power readings. More to come.

I'm seeing the same 3% difference between what the inverter reports and 15 minute averaging by PVOutput over the 4 day period I have that data (159kwh vs 154kwh).

Using a single day as an example, I summed all of the 15 min power readings for the day and multiplied by .25. This gave me the exact same result as PVOutput. I'm not sure what SE is doing to arrive at their figure.

[J.P.M.]

With regard to the daily energy reported by solaredge inverters, I've been collecting evidence that they systemically inflate the total by about 3%. On clear days, a more accurate total can be generated by summing the reported 15 min power readings. More to come.

I've got a sneaking suspicion that most inverters (string anyway) overestimate array output by something like that amount but I don't have #'s to back it up yet. I suspect mine (rebadged PowerOne) reports high by something like that amount.

[silversaver]

Same size system with Enphase microinverters at different orientation with different result:

https://enlighten.enphaseenergy.com/pv/public_systems/KWpC417458#null/production/graph/days

Now, is the Solaredge monitoring offering correct output result?

[sensij]

With regard to the daily energy reported by solaredge inverters, I've been collecting evidence that they systemically inflate the total by about 3%. On clear days, a more accurate total can be generated by summing the reported 15 min power readings. More to come.

[thejq]

Just another point of reference from Carlsbad, CA. Capture.JPGCapture.JPG

My 16x300W LG produced 211.47 KWh in the same period. But the orientation is slightly different. My array points to SW vs yours S. So the morning overcast would affect you slightly more. The peak power matches perfectly though as expected. Cheers on your new system!

[J.P.M.]

After a short delay (plans were returned due to insufficient documentation of panel/racking fire rating), I finally turned the system on last week. Here's the output for the first full week. Had a little marine layer the morning of the 9th, then clouds came in the afternoon of the 10th and stayed all day on the 11th. Total AC produced over 7 days was 295kwh. Best day was 46.3kwh on the 8th, average day was 42.1kwh.

[ATTACH=CONFIG]6049[/ATTACH]

Eric:

FYI and FWIW: For the most part, sounds like we have similar orientations and tilts. Your array is 1.37 times larger than mine. You're LG, I'm Sunpower. Your daily output in mostly sunny weather for 03/06 -03/13/15 is approx. 1.45-1.5 times my output for the same days.

Even accounting for my 18 mo. older array burn in/degradation allowances, and some weather variation for about a 15 mile separation, it looks like your LG outperforms my S.P by a few %. More anecdotal information that most decent systems, including Sunpower, produce approx. equal output in approx. equal locations and conditions. BTW, your greater LG output/kW is about what SAM predicts for LG vs. S.P.

[ericf1]

After a short delay (plans were returned due to insufficient documentation of panel/racking fire rating), I finally turned the system on last week. Here's the output for the first full week. Had a little marine layer the morning of the 9th, then clouds came in the afternoon of the 10th and stayed all day on the 11th. Total AC produced over 7 days was 295kwh. Best day was 46.3kwh on the 8th, average day was 42.1kwh.

first week.jpg

[Hillsider]

Lockable Safety Switch-Reviewsd

Yep, it looks like I'm screwed! I reviewed SCE's net metering handbook, & unless my main breaker is on the same panel as their revenue meter, a separate lockable manually operable safety switch must be provided. I will be replacing my two "ancient" breaker-boxes ( no main-breaker, only 12 branch breakers ) with a new outdoor load center. I originally was considering a 200A Square-D load center, but though convertible, it was not "reversible" ( suitable for bottom-feeding the main-breaker ). The unit I settled on is a Siemens PW2040L1200CU, which will be de-rated with a 150A Main Breaker kit. The PV array will be back-fed through a Siemens Q240 40A 2-pole breaker, at the top of the panel. For some reason, Siemens does not offer a lockable cover with their PL Series load centers.

[sensij]

SDG&E does not usually require a lockable switch. From their net metering terms and conditions:

A description and location of the visible, lockable AC disconnect switch if present.

Effective January 01, 2010, customers installing inverter-based systems will no longer be required to include an AC disconnect switch when the facility has a self-contained electric revenue meter (i.e., 0-320 amp socket-based meters or 400 amp K-based meters). This type of meter is used by the vast majority of all SDG&E customers.

[ericf1]

Hi Eric,

Looking at your pix from 2/5/15, the shot showing the wiring inside the inverter safety switch, with the main service panel to it's right did not show an additional box, known as a lockable safety switch.

In the pic labeled "Safety Switch", the center component is a DC disconnect switch. Once the cover is installed, the switch can be locked in the OFF position. Once this switch is in the OFF position, the max voltage of a string is 1vdc per DC optimizer. This switch is included with the SE7600A and meets the lockout requirement as far as SDG&E is concerned. On the AC side, the service panel door is lockout compatible.

[Hillsider]

Lockable Safety Switch

Hi Eric,

Looking at your pix from 2/5/15, the shot showing the wiring inside the inverter safety switch, with the main service panel to it's right did not show an additional box, known as a lockable safety switch. Does SDG&E not require one? Where I live, SCE requires a separate switch that the Fire Department can access in an emergency, & lock-out the PV system to prevent injury to emergency personnel. BTW, I have been tracking your posts, as I am designing a PV system very close to your install ( 24 LG305N1C-B3 Panels, 24 SE P400 Optimizers, SE7600A Inverter, SE1000-WIFI-01 Zigbee & Midnite MNPV-4 Combiner ). Because my roof-top is small and horribly shaded, I will be ground-mounting the array & combiner behind ( due South ) the house on an acre of bare ground, using schedule-40 2" GRC pipe & IronRidge racking components.

Garry

[ericf1]

I'll go to the meter and see if the arrow is going backward. If it does, then you can offically turn on your solar system because your net metering has started.

Net metering was enabled at the meter on the 6th. It now counts backward and shows "delivered" kwh, "received" kwh and -xxxxkw during production hours on the meter. When I look at my usage online at SDG&E, they just show 0 power used during production hours, no credit. My meter is saying one thing, my SDG&E account info another. Maybe this is normal, it's all new to me. At least they didn't bill me for the power I produced...