Adding storage to my Enphase system

[400bird]

I have 3 different stud finders, and they all go stupid on my stucco wall.

An 1/8" drill bit (+ a bent coat hanger/mechanic wire) is a good enough stud finder and the holes are small enough that a little silicone and paint, no one will ever see it, especially 8+ feet off the ground.

 

[GXMnow]

I need to just stop dragging my feet and buy some panels.

I just found a place that has used 290 watt Trina, or 295 watt Suntech solar panels for $75 each. They have about 500 that were taken out of service and sold off by an insurance company. I wonder if they were from poor Solar City installs on WalMarts? They are in Santa Paula, CA which 34 miles away. The panels appear to be from 2014, so 7 to 8 years old? Obviously no guaranty. I checked more of their site, and they mostly have large machine tools and even fork lifts etc. Solar panels are not their thing. I think they just want to unload them without losing money on them. They have a check box for "Make an offer", but they want to sell a lot of them. My 12 panels is nothing to them.

They are certainly a little older tech. They are the larger 72 cell size, making less power than my smaller 60 cell panels. But at $0.26 per watt, do I care? I have to do the measurements to see if I can fit 8 or 9 of these on my roof. They are BIG! Full 1 meter x 2 meter size. The 290 watt Trinas do have the UL label on them, the 295 watt Suntech ones are CSA "conforms to UL 1703" on the label. The pictures they have posted look like new, with just a tiny bit of weathering. But I am sure they picked the best looking 3 for the pictures.

 

[TRW]

This is very interesting, it would be so interesting to know why they are available. Next if you buy them it will be fun to see what you can get out of them as far as output, and how it all goes for you.

 

[Hedges]

Some panels from Trina, 2011 ... 2014, were recalled.

BUYER BEWARE: Stay Away From USED Solar Panels (Trina Solar) - Residential Solar Panels System

In this video, we cover why buying used panels is not always the best option over new solar panels. We bought 2 pallets of used Trina Solar panels, not thinking about recalled panels being re-introduced to the used solar panel market. The Trina Solar Panels we bought had been recalled for model...

sunsmartsolarpanels.com

Balance of system is so expensive, I'd rather buy panels for quality not price.

 

[TRW]

Good point Hedges! What I see is it is a bit hard to actually find what panels cost nowadays. Overall it seems the price seems to be good for what you get, as long as the seller is legit!

 

[GXMnow]

Thanks for pointing out the recall. I did not know about Trina having a recall. I guess that makes me lean to the SunTech panels then. But in any case, I will bring a meter and try each panel for VOC and ISC before I accept them. If they don't make at least 75% of spec (assuming a good sunny day) I won't buy them. I did send a message to the seller to ask about where they came from and if I can hand pick from the lot. If there are any cracks, or the leads look damaged, I won't take them either. At least this will only be a 150 volt max system. That is still enough to be an issue, but nothing like a 400 volt string setup.

I just did a bit of searching about the Trina recall, and I can't find any listing of model numbers, but I did find a few things to look for. It seems the issues took 2 years to show up, so these panels being 7 years old, if they have the issue, it should be very visible. Any moisture between the glass and the backsheet are a sign, it has the defective material. Unless I can closely inspect each panel, and measure their output in the sun, I won't buy them.

I am looking at total balance of system cost. For the panels that will be flat on my garage roof, I came up with a very secure but cheap solution. I will probably hold off on the awning and see how much I get with just the group on the roof. But if I do get them at that price, I will get 12 panels, and use the best of the lot. With the savings on these panels, it makes the cost of the RSD units pretty much go away. My biggest cost item is going to be the Schneider MPPT controller. I thought about going with an EPEVER (save about $200) but I really don't want to add yet another separate monitoring app. That cost savings is just not enough to make it worth while. The MPPT60/150 from Schneider is a very good unit with all the adjustability I need to dial it in with my Bolt batteries. And the built in time of use shifting capability of the Schneider system really wants to work with their whole ecosystem. Ideally they want more DC solar than AC solar, but that is not going to happen.

 

[Hedges]

If they don't make at least 75% of spec (assuming a good sunny day) I won't buy them.

A good sunny day in the middle of January won't make as much as in April or July.
If I recall correctly, PTC output of Trina may have been a lower percentage of STC output than some other brands.

But in any case, I will bring a meter and try each panel for VOC and ISC before I accept them.

Some bad panels show up as reduced VOC and ISC. Others are OK for those parameters but deliver reduced current Imp at Vmp.
There are electronic testers. I used a couple electric radiator heaters and detected the problem. In an operating array, a thermal imaging camera showed uneven patches (for most panels, but more so for a bad one.)

Some resellers (Santan) say they test. Perhaps that is with a PV panel test meter and 1 standard sun electric illumination? Don't know, no details given.

Measuring current of each panel into a resistive load or lead-acid battery bank and rejecting outliers might be the easy low-tech solution without dedicated meter. I think a high enough operating voltage is needed so diode-bypass of a low current section isn't missed. Suitable MPPT charge controller into a partially discharged car battery might be a good tester

I think I read issue with some Trina was leakage to frame, not reduced output. Don't know about the others.

 

[GXMnow]

At my home, I peaked at 3,300 watts out of my Enphase inverters today. They are 97% efficient, so I will just leave that out of it. My panels raw STC rating is 300 watts x 16 = 4,800 watts. So I got a little under 69% of rated power. The sun in Santa Paula should be about the same, it's just 34 miles away.

I have a pair of 8 ohm load resistors. If I put just one across the panel output, it should produce about 69% of 7.9 amps = 5.45 amps.
5.45 amps x 8 ohms = 43.608 volts That is a little too high. VMP is only 36.4 volts, and VOC is 44.9 volts. So the current may be a bit low and it should hit between 36 and 43 volts. That test will prove all of the cells are working. If I parallel the 2 resistors, to 4 ohm, the current might go a little higher, but let's stick with 5.45 amps x 4 ohms = 21.8 volts. If the panel can push my 4 ohm load resistor past 21 volts, I think I can call it good. I know that is only a little over 114 watts, but that is due to the voltage being pulled down with too low of a load resistance. The ideal load would be about 6.6 ohms. I'll see if I can come up with that, but it needs to be able to handle 200 watts. My 8 ohm load resistors are 500 watts each.

From everything I could find on the web, I will opt to check out the SunTech panels. I have not found any recall issues on them. And I even found one article where SunTech ripped on Trina for their failing back sheets. They are the same price and rated 5 watts more. They are the same size at 77" x 39" x 2" thick. And they weigh in at 60 pounds per panel.

 

[pvdude]

Here is what was discovered so far, when upgrading our existing single, Schneider Conext XW Pro 6848 to a dual, stacked system.
The upgrade consists of:
-A new big PDU to replace the mini PDU.
(The mini is for a single inverter, the big PDU connects many inverters)
-A second XW Pro 6848 (GREAT deal from altestore)
-Two conduit boxes for cable routing between all the equipment.

The existing Schneider 100-600 MPPT controller is retained.

- A special tool is needed to calibrate the inverters, about $600 for the **^%$# thing.
Conext™ Configuration Tool - 865-1155-01

- 5 more AC breakers are needed, beyond the 3 supplied w/ the new PDU.
The AC breaker cabling and connections are bewildering, much study required to understand it.



Schneider Electric Conext XW+ Split Phase 2-Pole 60 Amp AC Circuit Breaker Kit for Parallel Inverters (for PDP) Part Number: 865-1215-01


I also took RV10flyer's advice and am installing a "Micro-air easy start" on the heat pump, so as to lower the LRA from 132A to some number that will tax the inverters less.

And on the recent subject of adding PV panels:
Completely separate project that must be completed before it gets too hot here for me to work on the roof of the shed - I bought 8 more REC365AA.


Soon will have 24 in production.
Some helpful person(s) on the forum explained to me that if I had some panels facing West,
more energy could be obtained in the late afternoon.
The roof of the shed where the electronics are installed faces due West, and has excellent insolation for a few hours in the afternoon.


I have never installed panels on a roof before, am using the snapNrack system.
Trying to do it safely, but not sure yet how to get the panels from the ground to the roof w/o me participating in a 911 emergency call.
Thinking about building a ramp to slide them up, like I did for the ground mount panel install.

After going through the first Summer with the new solar system, I learned that at this latitude, Summer means clouds.
So I am hoping that with almost 9kw of direct-coupled PV, charging the batteries will work.

 

[GXMnow]

I looked at the inverter stacking instructions. Yeah, it seems overly complex. Most of it is to ensure the two inverters evenly share the load and don't fight.

The 8 365 watt panels should make for a nice battery charger. That is right about the 3000 watts I am trying to do. It seems the DC MPPT is quite a bit more efficient at putting the sun into the batteries when compared to even 97% efficient micros and then using the XW as the charger at 92% or so. I am probably just going with the MPPT 150/60 charge controller. 150 volt works pretty well for the shorter wire runs, and 60 amps at 48 volts will handle over 2,880 watts before it will start to clip. Actual battery volts runs over 52 volts, so 3,000 watts of panels it not likely to hit clipping in any normal sun conditions here.

Look forward to your install report.

 

[pvdude]

Finally some warm weather here!
Made progress on the upgrade from a single inverter to two.
Removed the MPPT controller & mini PDU.
I had hoped to simply slide the original inverter into position on the 3 aligned mounting brackets,
but the bracket design made that impossible.
The crane & cradle did not work either, no room.
Did manage to brute force inverter#1 into position, and install the conduit box.
Tomorrow will try to get inverter #2 in place.
Then the big PDU and MPPT controller will be mounted (not heavy).
At that point, I can begin staring @ cabling schematics and wiring it up.

 

[Shale MacGregor]

I am still fairly new on the forum, tried to answer a few questions as I asked some, and have been talking about my plans and purchases. I figure this is the right place to talk about what I am doing and where it is going.

My existing system is a very typical Enphase setup. I have 2 arrays of 8 panels each. The panels are Sil Fab 300M 60 cell units with an iQ7 inverter on each one. All 16 inverters are combined in the attic and go to an Envoy iQ3 on the wall of my garage near the power meter. I am at the limit of a single run with the 16 inverters. I have space to add more, but my existing main breaker panel is only a 100 amp unit, which limits me to a 20 amp feed in from the solar power system.

The system has now been running for just over a year and has been performing great. My south facing roof is at 210 degrees, so I stretch the usable sun a little into the evening, and the tilt of the roof is about 20 degrees which give very good exposure with the peak power in spring and fall. The 300 watt panels were actually able to clip the inverters for over an hour in May, pushing over 3,840 watts into my breaker panel. My production averages over 25 KwH a day and topped out at almost 32 KwH with all 16 panels making close to 2 KwH each. That is well more than I expected to get out of 300 watt panels. That works out to 6.7 sun hours. In summer (now) the panels are running a bit hot and the sun is higher than my 20 degree tilt, so I am not quite topping out the panels and inverters like it was in spring. Today, each panel was close to 1.7 KwH of production, for 26.4 KwH total. That is still 5.5 sun hours. Peak power was down to 3,460 watts.

During sunny cool days, I am pushing over 10 KwH back on the grid only to buy most of it back at a much higher rate. Even in these hot summer days, I still push over 5KwH out and have to buy much more to run the A/C in the evening. I knew I wanted to do "Time of Use" shifting with storage at some point. I was waiting on the Enphase Ensamble system, and looking into their batteries, but I just could not accept the cost for the rated performance. And the fact I would have to pay the installer almost as much as I did for the original solar gear install. So a few months ago, I really started looking into other storage options. The Tesla Powerwall 2 was looking like a fair deal, but it came with more than a few strings, and few installers would work with my existing system, and my original installer, does not work with Tesla. And trying to get a Powerwall 2 for a DIY install is not going to happen here in So Cal.

I looked into a few other ideas, but for safety reasons, I narrowed it down pretty quickly to Outback, SMA, and Schneider. I looked at Growatt and Sol Ark also, but they didn't have all the features I wanted. The Outback Skybox almost won, and has had some great reviews, but looking over the manuals, I was not totally convinced. The "one box does it all" system sounds great, but in the end, I liked the flexibility of the Schneider a bit better. So today, I placed my order for a Schneider XW-Pro 6848 inverter with the Conext Gateway box. It went on sale for a good price and a free shipping offer, so I clicked go. I should have it here by this weekend.

I have already talked about my battery on a few other posts, but I will summarize it here. Our friends at Battery Hookup had a great deal a few weeks back, so I picked up 3 LG Chem battery modules from a Chevy Bolt. I got 2 of the 10S 3P and one of the 8S 3P. I only had to cut one buss bar to split the 8S into two 4S groups. Then wired each 10S with a 4S and then paralleled the two 14S groups. The end result is 14S 6P of LG 3.7 volt 60 amp hour NMC pouch cells. 51.8 volts at 360 amp hours total or 18,684 watt hours. I figure about 17 KwH useable, but I do not plan to run them below 30% or charge them above 90% unless I am in a power fail condition. That gives me about 11 KwH to use every day for "time of Use" shifting which should just about eliminate any extra solar going to the grid and doing all self consumption. I am using the JK BMS with active cell balancing and the Bluetooth App. I have the 200 amp version that will do up to 24 cells and 2 amps of balance current. These nearly new LG cells are testing very well balanced so far, but without the inverter charger, I have only been able to push and pull 10 amps into the pack. Once the inverter is up and running, I will be able to give the pack a real test.

For the initial setup, I will just do time based control and have it charge at 1 KW for 10 hours while the solar is producing, and then push back 2 KW into my home during the peak rate 5 hour window. I may add a "WattNode" later to have it do true "Zero Sell" and the Schneider "Batt Mon" to give it more acurate battery status for "State of Charge" control of the store and sell as well as more accurate control if I have to run off grid during an extended power failure. Without the Batt Mon, it only uses battery pack voltage. These Li NMC cells have such a flat discharge curve, it can barely tell between 75% and 50% state of charge without measuring the current going in and out of them. My BMS reports a very accurate SoC, but for now, the Schneider inverter can't get the data from the JK BMS. That may happen in the future as the BMS does have both a CAN port and RS485 port for wired communication. I may be able to program an Arduino to read the BMS and report the data as if it is a Schneider Batt Mon.

Since this is the "Show and Tell" area, I will post a few pictures.View attachment 19680View attachment 19681

I will add a few more soon. I do have a few pics of the array etc. And even a few of the panels going up that I took with my drone.

I am currently banging my head on a similar project, I only want to charge the batteries while the enphase are actually producing. This means sensing and controlling whether or not to charge... I considered making the battery large enough to handle an entire day of charging so that on slow days I use what is a available while charging with either grid or solar... I need to sit down and think this through some more but appreciate your post. Cheers

 

[GXMnow]

Looking good PVdude. The XW-Pro is a heavy beast. My brother and I managed to muscle mine onto the wall bracket, but it was touchy. Once on the bracket, it did not want to slide side to side at all, it grabs on tight. It was just 1/4 inch off center, and it took us 30 minutes to get it moved over to line up the locking screws holes.

 

[GXMnow]

I am currently banging my head on a similar project, I only want to charge the batteries while the enphase are actually producing. This means sensing and controlling whether or not to charge... I considered making the battery large enough to handle an entire day of charging so that on slow days I use what is a available while charging with either grid or solar... I need to sit down and think this through some more but appreciate your post. Cheers

I think everyone of us with an AC coupled XW-Pro is running into this same issue. The XW-Pro software does not directly support any automated charging. The SW has "Smart Charge" that does exactly what we want, but the XW does not support it.

400bird has a Raspberry Pi controlling his. The data he has posted looks great. He took it a step further and even has it controlling his EV charger as well. I am on the fence about going the route he is using. His Pi is reading an energy monitor at the main panel and adjusting the charge current based on the current going out to the grid.

 

[400bird]

Adding Schneider XW Pro

My system is “off grid”, so I can’t sell back to the utility. DC coupled PV. What I need is a “software diode”, so I can supplement our loads from the grid, but never send energy to the grid. The “Grid Support” option works perfectly, the inverter decides when it needs to extract energy from the...

diysolarforum.com

It's up and working still, but today was overcast.
Here's a random day last week:

 

[400bird]

Finally some warm weather here!
Made progress on the upgrade from a single inverter to two.
Removed the MPPT controller & mini PDU.
I had hoped to simply slide the original inverter into position on the 3 aligned mounting brackets,
but the bracket design made that impossible.
The crane & cradle did not work either, no room.
Did manage to brute force inverter#1 into position, and install the conduit box.
Tomorrow will try to get inverter #2 in place.

Wow, that's a lot of inverter to move around! The lift frame you've got is great for a single, but as you discovered, it doesn't work when they are side-by-side. Good luck!

 

[GXMnow]

I was just on my So Cal Edison account to check how well I am doing at nulling out my grid usage, and I noticed the expected end of month bill looked a bit higher than I would expect. My usage numbers were in line, why the difference? Well, it looks like Edison snuck in another rate hike. My low tier on time of use is now 30 cents, and the high tier 4 pm to 9 pm is current 43 cents for winter, but jumps to 49 cents starting in June for summer. OUCH!. I guess on the good side, that does effectively shorten my pay off time for the solar panels. The summer rate change, makes the battery pay off about the same, as I will still get 19 cents for every kilowatt hour I move from off peak to peak time of use. But right now in the winter, it is only 10 cents per KWH I move. Moving 9 KWH is only 90 cents a day. So I am only saving $200 all winter, and I'll save about $280 through the summer, moving 12 kwh each day to help run the A/C.

Making 8 megawatt hours a year, at 30 cents a KWH = $2,400 worth of electricity coming from my existing solar panel in the off peak time. If I add the new DC panels that will make about half the energy, but at 49 cents, since I will use it all to run my A/C from 4 pm to 9 pm in the hot summer, what does that work out to? 4 months x 30 = 120 days. 6 sun hours average, 2,700 watts of panels. That is about $950 worth of peak rate electricity, just for the 4 summer months.

With the base rate going up to 30 cents, I actually feel a little better letting the Enphase inverters back feed while I DC charge the batteries. Now I NEED to get the DC panels up and running before June. It makes less sense to take the 30 cent power to charge he battery. I'll take the NEM credit, but still not buy any peak rate power ;-)

 

[400bird]

Wow, sounds like SCE made the decision for you.

But that is prob good news.

Have you decided on how or if your going to address arc fault detection on the charge controller?

 

[GXMnow]

I am still having trouble finding add on arc fault protection. The one combiner I found that supports it is way over priced. Midnight solar now has a DC charge controller with built in Arc Fault protection. I am still leaning towards the Schneider charge controller, but this is one feature that could null out the cost difference and make it worth having a third monitoring app. As much as I want to be totally by the book, I may run without arc fault for a while and see if the industry comes up with something more reasonable.

 

[Hedges]

Here is a device which does that, but don't know if still available

Uses | Sensata Devices for Solar Photovoltaic (PV) Applications

Sensata Devices for Solar Applications, Alternative Energy Power Solutions, Battery

w.sensata.jobs

Stand Alone Solar Arc-fault protection devices.

I have been studying the NEC requirements for wiring solar panels and one of the requirements is arc-fault protection if the PV array is at 80V or more. However, when I search for PV arc-fault protection product.... I find no stand-alone products. Some SSRs have it built in. Most grid-tie...

diysolarforum.com

I bought a couple surplus for possible future use. I would run all rooftop PV strings through the CT, and have the control output break AC to all the GT PV inverters.

Arc Fault has had issues with false trip. SMA message board someone said earlier inverters were unusable for that reason and he had to use other brands. I imagine later models were OK. Part of the difficulty is that arc energy is about 60 dB below switcher harmonics. So can't just use signal strength out of high-pass filter. Need to run FFT, ignore periodic harmonics in frequency domain, look at non-coherent signals. Also have to ignore radio transmissions (PV panel wiring forming a good or bad antenna would contribute to that.)