Adding storage to my Enphase system

[Supervstech]

Or are you looking for an R-134a psychrometric chart? (@Supervstech can probably recommend better, but here's a quick description on to use such a chart http://www.alephzero.co.uk/ref/circeff.htm if you're not familiar with them)

Home ac units don’t use 134, they use the blend refrigerant 410A and the chart is a little more complicated with a blend. There is bubble and dew pressures... also, pressure charts don’t help when charging, as there is latent (humidity) and sensible loads... also, superheat and sub cooling calculations factor the values.
most modern systems have a vav setup either mechanically or electronically controlled.
Variable speed systems need known line lengths, and the refrigerant weighed in...

 

[GXMnow]

My A/C system is actually R410a Puron refrigerant. It runs at a very high pressure. I think the High side is over 400 PSI. I found this article.

Guidelines For Troubleshooting R-410A Systems

The big push is on to replace R-22 unitary air conditioners and heat pumps, and it looks like the industry has really and truly decided that R-410A systems are “it.” This article offers an overview of some of the issues related to R-22 and -410A in the field: operating pressure differences, oil...

www.achrnews.com

By the graph, it looks like it runs about double the pressure of a 134 system.

 

[GXMnow]

Today was a little interesting.

I was at work and my girlfriend calls me to say the lights flickered and there was a loud boom. But now all the lights seem fine. I had her loo in the garage, nothing out of place, no smell. Light hum from the Schneider inverter still. So it was another hour and a half until I got home. I check the logs, and sure enough, at 5:49 PM the Schneider XW-Pro did a grid frequency high anti island disconnect. It appears to have only gone offline for just the 5 minute qualifying delay. So then I checked the logs on the Enphase Envoy. Over 120 entries starting at 5:48 PM. Grid frequency low, then grid frequency high, then grid instability. All 16 inverters. They all stayed off for the 5 minutes while the backup load panel was running on battery. But on the good side, the inverters did all go back online and were producing power again by 5:54 PM. The Enphase production graph shows a small drop in the 5:45 to 6:00 15 minute time slot, about 25% down. Then the 6 pm on were all back to normal. This bodes well for it working properly with a grid outage. None of the computers in the house glitched at all, just the LED lights flickered.

Still have no idea where the boom came from. I did not see anything on my way home like a car crash into a pol or a crew fixing a transformer. Something big did happen, but my system, did it's job and kept the house powered.

 

[pvdude]

In our rural neighborhood, the grid transformers are mounted on poles, like the old days.
When an animal gets on part of the transformer so as to blow the fuse (or whatever the protection device is), it is a loud boom, and the lights flicker.
Also happens when an overloaded transformer just gives up, “boom” and sometimes the lights flicker, other times the grid goes down until a utility bucket truck/team arrives to effect repairs.
Very glad to hear your system worked as per your design goal!

 

[GXMnow]

I have had 2 more cases where the Schneider XW-Pro detected a grid frequency error. (error id 25 AI Over Frequency) It is odd that each fault has always been the frequency going up too much. Maybe Jack Rickard was right, and the utilities are kicking solar off line with a frequency spike so you have to buy power.

The one on April 29th was at 5:49 PM so the solar was still producing, but they also detected the anomaly and some of them went into the 5 minute hold. I did not get the Enphase log for that day, but it does look like some stayed producing, but it was not enough to run my backup loads panel, so the bulk of the power running the house was coming out of the batteries. The real odd part about this one is that the current graph does not show much of an increase, but the battery voltage dropped nearly 0.4 volts in one minute, from 57.17 to 56.78 volts. In fact, the current went from 0.7 amps discharging, to zero amps at the time of the voltage drop. The XW-Pro had just completed the absorb charge, so it was basically sitting idle, and then it had to form the grid for the Enphase iQ7's.

The May 8th power glitch was clearly much worse. My girlfriend was home and heard a boom and saw the lights flicker bad. At 5:43 pm, the sun had dropped enough that the XW-Pro was starting to help power the house. As the sun drops, the current slowly ramps up, so the power exporting back to my main panel stays at about 2.5 amps or about 600 watts. Right before the boom, the battery discharge current had ramped up to 23 amps at 55.9 volts. That works out to 1,286 watts. Almost half of what the inverter was producing was pushing back to the main panel, and the backup loads were using all of the solar, and the other 600 watts from the battery. Evidently there was still a little solar power going into the system. This is where it seems odd. It reconnected to the grid very quickly. But it stopped inverting, so the house was using grid power. It waited the full 5 minutes before it started pulling any battery current. My backup loads panel too 85 watt hours from the grid in just 5 minutes. That does work out with the math. That is about 1,020 watts for 5 minutes. When the system went back to inverting, it was running 1,370 watts from the battery. But this was running my backup loads panel, AND pushing 600 watts back to the main panel again. The solar was only putting out 100 watts average from 7 to 7:15 on that day. I did see another quick battery voltage drop, but in this case, it does line up with a large current increase when the inverter started pushing power again. The voltage dropped from 55.81 to 55.38 which is a 0.43 volt drop, but the current went from basically zero to over 30 amps.

 

[svetz]

Maybe Jack Rickard was right, and the utilities are kicking solar off line with a frequency spike so you have to buy power.

They don't have to, via UL 1741 they have the right to throttle back your production.

The real odd part about this one is that the current graph does not show much of an increase, but the battery voltage dropped nearly 0.4 volts in one minute, from 57.17 to 56.78 volts. In fact, the current went from 0.7 amps discharging, to zero amps at the time of the voltage drop. The XW-Pro had just completed the absorb charge, so it was basically sitting idle, and then it had to form the grid for the Enphase iQ7's.

Possibly some equalizing?

It reconnected to the grid very quickly. But it stopped inverting, so the house was using grid power. It waited the full 5 minutes before it started pulling any battery current.

5 minutes is a curious number, I'd call them about that.

 

[GXMnow]

5 minutes is a curious number, I'd call them about that.

It is the same 5 minutes that any grid tie inverter must wait before it can "export" power. But evidently, if it sees good power, it will still go into pass through right away. After the 5 minute delay, it went back to pushing power back to my main panel.

 

[svetz]

I misread your post, I thought that it stopped inverting for 5 minutes and the house had no power. On the grid side, nothing to worry about.

 

[Hedges]

It is the same 5 minutes that any grid tie inverter must wait before it can "export" power. But evidently, if it sees good power, it will still go into pass through right away. After the 5 minute delay, it went back to pushing power back to my main panel.

Somewhat different behavior with my Sunny Island, which serves as UPS (sort of) for downstream load.

If power fails, Sunny Island watches power go down briefly, maybe a couple cycles (don't think there is a guaranteed response time) then disconnects from grid and supplies AC to downstream loads from battery. This is long enough outage for downstream GT PV to go offline. Most loads will ride through the dropout.

While grid is down, the unprotected loads (garage) are off, or suffering through any brownout or over-voltage that might be occurring.
The downstream protected loads are up. After a delay, GT PV reconnects and provides power to the UPS system so it can operate indefinitely.

When grid returns, unprotected loads see it immediately. If power returns/goes away repeatedly (a function of some automatic utility switches), these loads experience that.
When grid returns, Sunny Island continues to supply downstream protected loads from battery. It watches the utility grid with a wary eye until it has seen in-spec power for 5 minutes. It then synchronizes phase of the AC it produce with the grid and connects without a glitch.

 

[Ampster]

They don't have to, via UL 1741 they have the right to throttle back your production

Yes, that was Jack's conspiracy theory about UL1741. I think he was using that to scare people into buying his systems that would sit behind the meter and allow people to self consume solar energy rather than sell it to the grid.
I saw UL 1741 as a standard that made AC coupling feasible.

 

[GXMnow]

Somewhat different behavior with my Sunny Island, which serves as UPS (sort of) for downstream load.

If power fails, Sunny Island watches power go down briefly, maybe a couple cycles (don't think there is a guaranteed response time) then disconnects from grid and supplies AC to downstream loads from battery. This is long enough outage for downstream GT PV to go offline. Most loads will ride through the dropout.

While grid is down, the unprotected loads (garage) are off, or suffering through any brownout or over-voltage that might be occurring.
The downstream protected loads are up. After a delay, GT PV reconnects and provides power to the UPS system so it can operate indefinitely.

When grid returns, unprotected loads see it immediately. If power returns/goes away repeatedly (a function of some automatic utility switches), these loads experience that.
When grid returns, Sunny Island continues to supply downstream protected loads from battery. It watches the utility grid with a wary eye until it has seen in-spec power for 5 minutes. It then synchronizes phase of the AC it produce with the grid and connects without a glitch.

Actually is is quite similar.
In both of these recent cases, the power glitch was actually very short, maybe just a few cycles. With both reporting as grid frequency too high. As soon as it sees any glitch like that, it also disconnects from the grid and goes into invert mode to run the backup loads panel. It does it quick enough some of my Enphase iQ7's have stayed working. But in the worst case, they should all be back online in 5 minutes after the inverter is making a stable local grid. The odd part was the reconnect. Since the power was not actually out, it was just a glitch, the XW saw good power right away. It actually reconnected the backup loads panel back to the grid and stopped using battery power. But due to the anti islanding rules, it then had to wait 5 minutes before it could start inverting and export power. The backup loads never saw an outage.

 

[GXMnow]

So the last few days have ben a grid torture test here in So Cal. Right now it is is just dropping back under 100F, after a high of 106F. A little over and hour ago I noticed the lights flicker a tiny bit, and the A/C shut off. I thought maybe just the change in load made the LED's flicker, so I didn't think much of it. The house is cool, even with the blazing heat outside. Then I get a text from my girlfriend. She works at Six Flags Magic Mountain a few miles away. The whole park lost power. So I checked my logs, and sure enough, the Schneider XW-Pro saw a grid instability and switched to battery power. The A/C probably shut off from the glitch, and it times out and won't start for a minimum of 3 minutes. And the Enphase log also shows all 16 panels shut down as well due to "grid instability". None of my PC's or any other gear in the house even noticed. The XW went back to AC pass through after just 30 seconds, and started exporting power again in 5 minutes, right as the Enphase also went back to producing again.

This was about the worst power disturbance I have had since the Enphase software update. I have to say the iQ7's recovered without a problem at all this time. That is a big improvement over the previous event. I am not sure how accurate the clocks are synced between the Enphase and the Schneider, but if they are dead on, it looks like the Enphase solar inverters actually went offline a little before the XW-Pro detected the problem and disconnected. The first log on Enphase is stamped at 4:31 pm, and the grid frequency error is logged in the XW at 4:32 and 53 seconds, almost 2 minutes later. Now the Enphase does not have seconds, so it could have been 4:31 and 59 seconds, and the clock in the XW could be over a minute fast. I have less than 100 watts of load other than the A/C compressor in the main panel now. Almost all of my house load is in the backup load panel, so it is possible I could have a power failure and not even notice other than the A/C stops blowing cold. The blower in the furnace is on the backup power.

On another not, heat certain does drop solar panel output. Peak power today only hit 3,500 watts under a scorching sun. The air temp was 106F, so I bet the panels were well over 130F. I normally peak over 3,800 watts on normal 80-90 days, and have topped 3,900 when it is sunny and cool, but that is where the iQ7's max out. Back in May, I had days where output was over 3,800 watts for 3 hours straight again the iQ7's clipping. Today only just touched the 3,500 watt mark at perfect solar noon.

 

[Frank in Thailand]

Thanks for the update!

It seems to me that we talking about a few watts, 3500 Vs 3800

Usually, without iQ7, people would just slap on an additional solar panel if they want more energy.

Fine-tune 10% is usually not worth the effort of investment.

Same as cooling with water does give 10 to 25% extra, it's a lot of work

To me it looks like all worked as it should, despite the abnormal grid conditions, and all recovered.
I call that a win!!

With temperatures above 100F, airconditioning is "most" important, together with the refrigerator / freezer.
You don't want your food or yourself go bad

Good to hear your backup system was functioning as it should!

 

[GXMnow]

Doing the math, the power drop does fit with the panel temp. Not really complaining, it is making good power, but of course, we always want more. The Enphase Enlighten App graphs the output over the previous day, and the same day last year, and the drop off from last year looks dramatic. But it is less than 10% off. With clouds in the morning, and the 5 minute shut down at 4:30 pm, it still made 24 KWH's today, but it made 27.3 yesterday, and 29.9 last year. The big drop was mostly the morning clouds. Output was down to just 300 watts at 11 am. Yesterday it was 2,900, and last year, it was 3,100 for the same time of day. But the clouds cleared out and it hit the 3,400 watts before noon. And it didn't fall under 2,000 watts until 5 pm.

I do want to make a little more power to help run my A/C in the summer, so I am looking at adding about 2,000 more watts of PV panels. Either 6 72 cell panels, or 8 or 9 60 cell panels. But those will DC charge my battery bank, and make the power I use in the evening when the sun goes down.

 

[Frank in Thailand]

That's always good.

You are aware there is other option?

+72 cells?
"All" the +400 watt have different size, no longer 195*95cm (roughly) but like 210*105...
How to get more watts out of a panel? Make it bigger

It would require own MPPT,
As the voltage is also higher then the standard from 72 cells.

If it would be separate installation anyways, this might be a good solution.

Especially the half cell double sided seem to handle heat well, and partial shade.

With +14kw/h (if there is enough sun) I'm not looking to expand.
Greatly oversized but we live off grid, and need to account for rain season, not the sunny days.

Solar panels price have dropped significant the last few years, here we pay about $65 for 345w poly panel.
No brand, warranty is "not existing" in Thailand.
So that doesn't matter.

 

[GXMnow]

Just to make monitoring nice, I will probably get the Schneider MPPT 60-150 charge controller, so it will be on the same data buss as the XW-Pro inverter. With a 150 volt maximum, the 72 cell panels can only be wires 2S to be safe. At 3S the VOC would be too close to the 150 volt limit. With the 60 cell panels, I can safely go to 3S without going over the 150 volt limit. So I would likely fit in 6 72 cell panels at 2S3P or go to 9 60 cell panels at 3S3P. 400 x 6 = 2,400 watts, or 9 x 300 = 2,700 watts. With 5 sun hours, the 2,400 watts gets me 12 KWH's a day. My current panels at this time of year are doing better than that. My array is 16 x 300 watt panels, for 4,800 watts and a typical summer day is doing over 27 KWH's or 5.6 sun hours, and that is after the inverter to AC but the Enphase inverters are 97% efficient. The Charge controller will be close. I am currently only cycling about 8 KWH's in and out of my battery bank, but I could easily bump that up to use the full 12 KWH from the DC array even with my current 18 KWH battery bank, but I also bought another set of cells. I am deciding if I want to add just 50% and have a spare 9 KWH pack, or just use it all and double the online battery. In either case, I think I will put them in 2 separate cases and make them portable, so I could bring 9 or 18 KWH's and an inverter to someone else during a power failure. The only pain of doing that is having to balance the batteries before they can be connected back to the system. My full charge voltage is 57.4 volts, and I current run down to 51 volts. A 6.4 volt difference would be an extreme amount of current in 2/0 cables. Just 0.1 ohms would drop it to 64 amps, which would be fine for them to balance each other. But that does still need to be a 400 watt resistor. OF course, i do not ever plan to plug it in at complete opposite ends of my charge range.

 

[svetz]

...looks like the Enphase solar inverters actually went offline a little before the XW-Pro detected the problem and disconnected...

The frequency probably started to go before the voltage. The bad thing about trying to manage a renewable resource is that when the frequency starts to crumble all the renewables suddenly go out and that makes a bad situation worse for the grid operators.

 

[Hedges]

The frequency probably started to go before the voltage. The bad thing about trying to manage a renewable resource is that when the frequency starts to crumble all the renewables suddenly go out and that makes a bad situation worse for the grid operators.

Because that's what they were ordered to do.
Original UL-1741 spec required disconnect if voltage or frequency outside standard range, OK at < 1% PV penetration. Now some markets are around 25%.
Newer UL-1741SA requires rid-through moderate excursions outside spec for 299 seconds, and disconnect by 300 seconds. Or, frequency-watts or voltage-watts adjustments to output power.
If they got that right, it will give them time to stabilize or assist with stabilization. I believe communicating with loads to adjust power demand should be part of the solution. (Air conditioners often the biggest issue.)
The new inverters are supposed to make tuning parameters accessible, so in the future may be internet connected allowing commanded operation (for better or worse.)

My grid-backup system of older equipment is designed to isolate and play in its own sandbox until the system is stable again; my loads are removed too.

 

[pvdude]

I was originally going to go w/ the Schneider 80/600 MPPT charge controller, but was told by the vendor that they were either discontinued or on long BO.
As an option, they offered the new Schneider 100/600, which they had in stock, for a small up-charge.

Glad to have the 100A capacity, able to run two parallel strings of 8-series REC Alpha 365 @ 44.6vdc
total of about 356vdc. The PV array is about 230 cable-feet from the building where the electronics is mounted, so the numbers worked out.

The PV maxes out @ just about 100A.

 

[Hedges]

$1300 for 6kW SCC

Amazon.com

Not much more money and I get an AC coupled GT PV inverter. (Which battery inverter gets power from and controls.)
If the AC power is used right away, avoids double-conversion to/from 48V.
That's the way I prefer to go, at least for a large PV / small battery system. If most power is stored in battery for later use, then SCC could make more sense.