Adding storage to my Enphase system
- Thread starter GXMnow
- Start date
There are ways to do it without a PLC, although being comfortable with computers would still be a prerequisite. @400bird uses a Raspberry Pi and NodeRed, which makes most of the programming more of a graphical exercise. Really the functionality should be built into the firmware for the inverter, but unfortunately that isn't the case.
Where it really starts to get complicated is if you try to incorporate load management into the system-- figuring out when and how much to charge the EV, deciding to shut off non-essential loads, etc. If you have a small critical loads panel you can get around that... but if you are putting $XX into a backup battery it is nice to be able to run everything during a sunny day.
GXMnow
Solar Wizard
- Joined
- Jul 17, 2020
- Messages
- 2,710
I got into a bit of an argument with the tech at Schneider back when I first got the XW installed and could not make it stat a charge cycle. I even copied part of their own web page to show him where it said the system could do "time of use", "self consumption", and "AC coupling". But the response I got was the joke about the place that advertises "Open 24 Hours". You go and they were closed. You call them later and ask about it, and they say "Sure, we are open 24 hours, just not all in a row." After a few months, they did re-word the web site a little, but it is still far from clear that power time shifting with only AC coupling does not work with just their internal software.
The "Fix" I was told from Schneider was to remove half of my Enphase microinverters and wire the panels to a Schneider MPPT charge controller. Not exactly a reasonable fix. Other options that sort of work is having a timer take the system off grid until the battery runs below the "Recharge Volts" setting each night. Seems like an odd way to work around the problem. I even gave their engineers several easy things that could be added to their software to make it work, but they basically ignored it all.
In the end, using some form of external controller does work, and the hardware on the XW-Pro is just plain hard to beat. The inverter will crank out all the power you would expect, and a lot more. The 6,800 watt rating is very conservative. It compares well with most competitors 8,000 to 10,000 watt inverters.
The part that I fond the most annoying is that they did code the smaller SW to do "Smart Charge" where it will switch to charge mode and adjust charge current to minimize grid power export. But, that unit can't "grid sell" any power from battery to the grid side input terminals. So there is no way to have it offset loads in the main panel. It can only drive power to the backup panel.
I have thought abut building up a box with the PLC and power meters to sell as a package. The problem is making user friendly software for a non engineer to be able to set it up. And then I realized I could never offer decent customer support.
Hedges
I See Electromagnetic Fields!
- Joined
- Mar 28, 2020
- Messages
- 20,670
They took a grid-interactive battery inverter and added support for AC coupling.
Is it so unreasonable to do what they suggest? Would it work with any MPPT, or just Schneider?
Does it specifically require DC couple >= AC coupled, or can you get the control you want with a small amount of DC coupling?
I have a related but different challenge. With my SMA battery and GT PV inverters, can I do limited/zero export?
If I disconnected from grid whenever production exceeded load, that would zero export, but cycle battery.
No ability to curtail GT PV while grid sets frequency. Unless I add a controller that senses CT and sends commands by SpeedWire.
DC coupled, export only occurs if battery voltage too high (no comms between SCC and inverter). But maybe with inverter told not to "sell" to grid, it would shave power off battery so long as consumed downstream.
So maybe DC coupled would support zero export, goal of zero import. And disconnect from grid prevents GT PV from exporting. DC > AC might make it balanced.
(I'm thinking of a future location, without doing net metering if new regulations unfavorable.)
Ampster
Renewable Energy Hobbyist
I do not remember if you already had some DC coupled panels? I have been under the impression that many hybrids including my Skybox perform better if there is some DC coupled solar. My theory is that the layer of DC coupled solar can react faster to changes in loads. In a sense that is a virtual buffer. I assume this scenerio is only an issue when the grid is down. It should not matter when the grid is up because the grid is the buffer.
GXMnow
Solar Wizard
- Joined
- Jul 17, 2020
- Messages
- 2,710
I have read a few different opinions on where the XW-Pro evolved from. But most seem to suggest the original design of the inverter bridge was an off grid design first. It was meant to be coupled with DC charge controllers. Being able to charge from an AC input was then added as a backup to charge if solar didn't produce. With a lead acid bank, the voltage could fall far enough to kick off a recharge cycle. Then they added being able to sync with the grid to help drive loads from a weak source. Grid support and generator support. Then they added being able to export back to the grid. The design of the inverter bridge is able to work in reverse and pull current from AC and push power into the batteries. They then added the ability to frequency shift. The XW-Pro still uses the same basic inverter block as the original XW and XW+, but they designed a new processor board with a lot more computing power to be able to back feed the grid in compliance with grid codes.
In the idea system that Schneider suggested, the DC coupled solar would be sized to provide all of the power you need overnight, while any AC coupled solar would provide the power you need while the sun is shining. That way the system does not use the XW-Pro as the charger.
When running on grid, the XW-Pro has no way to curtail any AC coupled grid tied inverters. I don't think that even the Schneider grid tie inverters have data control of power output. So if the AC coupled inverters are making too much power, it will export. In the case of my Enphase system, Enphase offers control with consumption CT's to enable zero export limiting by curtailing production. But this causes another issue. It would be throwing away energy production instead of sending it to the batteries, because it does not know I have any.
The "Smart Charge" programming in the SW looks like it is well engineered. This is from their manual.
"AC Couple Smart Charge is a feature of the Conext SW that prioritizes battery charging over energy export to local AC loads connected upstream of the Conext CSW battery inverter. AC Coupling must be enabled for this feature to be functional."
"In AC-coupled configuration and with grid present, the Conext SW monitors flow of power from its load port (AC Output) to the utility grid input (AC Input). When power flow to the grid is detected and the battery needs charging after a short delay, the Conext SW initiates a bulk charging cycle by switching to charge mode and drawing AC power to charge the battery. It only draws enough energy to keep the flow of power to the grid to zero. As long as the battery is able to accept the energy, the Conext SW will continue diverting excess PV production to the battery. Once the battery bank is at a level such that not all excess energy from the PV Inverter is being absorbed, the balance will then flow out to upstream loads in the house that are not connected to the Conext SW’s AC output port."
"If PV production is lost for a prolonged period of time and the battery bank discharges below the Recharge Volts setting, the Conext SW will initiate a normal charge cycle and use power from the grid to charge the battery. For information on this setting, see “Configuring Basic Settings” on page 4–7. In all of the above modes of operation, only the PV Inverter exports energy to the grid. The Conext SW itself cannot sell to the grid - it simply passes energy from the PV inverter to the grid. For this reason, the PV Inverter must be fully grid code compliant as it assumes responsibility for anti-islanding protection."
"NOTICE AC COUPLED PV INVERTER COMPATIBILITY AC power generated by AC coupling PV inverters with Conext SW must be consumed by AC loads or used to charge batteries. As an alternative, the excess power produced from a PV inverter can be routed to dump loads. Do not AC couple PV inverters with the Conext SW that are unable to reduce, derate or cease the excess PV inverter power in response to the changes in AC line frequency controlled by the Conext SW. Consult the manufacturer's specifications of your PV inverter and confirm compatibility. Failure to follow these instructions can result in equipment damage."
This is exactly what I have asked Schneider for in the XW-Pro. If energy is flowing back from the output to the grid input, start charging the battery. If the little SW can do this, why can't the far more capable XW-Pro do it? And to make it even worse... This is only in the 230 volt manual. It is not in the 120/240 volt split phase US spec manual. What is the reason for it only being on the Euro model?
All that being said, and how annoying it is to try and work with Schneider tech support. I still think they have some of the best hardware. If I was building another system for myself or a friend, I think I would still use an XW-Pro, but the PLC will be par of the install now. Use it is a basic device that we can control. It responds well to the external commands and has been working great like this. Since I write the code, I can choose exactly how much to let it export or consume. If true zero export is an absolute goal, then I would probably use all DC coupled solar. Then it is a simple matter of having all the loads on the output side of the XW-Pro and setting it to grid sell zero amps. With that setup, you don't even need the PLC. It does just "Work". The DC charge controllers will push all they can into the batteries. If the batteries top out, then the charge controllers curtail and sit in CV output mode to keep the batteries full. As long as the battery voltage stays above the "Grid Support Volts" it will provide all the power to run the loads after the XW-Pro. No need to back feed, all the load is on the output side. With one unit, this does limit you to a 60 amp 120/240 volt panel. IF you need more, you would need a second unit. Going a bit more complex, you could put some large loads in the main panel, but t feed them with zero export, then you need to add the Watt-Node for it to keep doing zero export. Once you add AC coupling it starts to fall apart. When on grid it just can't curtail the AC grid tied solar inverters. But for a system like I have, where my existing solar was already permitted for export, it works just fine as the worst case export is still less than before. And if I had 4,000 watts of DC coupled solar, the system would work without needing the PLC as the DC solar alone would provide enough battery charging. But ALL of the excess AC solar would export to the grid. This is why they suggested taking the microinverters off of half the array.
And Ampster... No, I did not have any DC solar until I added the 10 small NewPowa 100 watt panels and the BougeRV 40 amp charge controller. I would not say that the XW-Pro itself operates better with DC coupled solar, but the system as a whole certainly does. But again, without my PLC controlling the XW-Pro, the system really does not work at all unless you have enough DC solar to provide your desired evening power. That is the big Gotcha! in this whole mess.
In the idea system that Schneider suggested, the DC coupled solar would be sized to provide all of the power you need overnight, while any AC coupled solar would provide the power you need while the sun is shining. That way the system does not use the XW-Pro as the charger.
When running on grid, the XW-Pro has no way to curtail any AC coupled grid tied inverters. I don't think that even the Schneider grid tie inverters have data control of power output. So if the AC coupled inverters are making too much power, it will export. In the case of my Enphase system, Enphase offers control with consumption CT's to enable zero export limiting by curtailing production. But this causes another issue. It would be throwing away energy production instead of sending it to the batteries, because it does not know I have any.
The "Smart Charge" programming in the SW looks like it is well engineered. This is from their manual.
"AC Couple Smart Charge is a feature of the Conext SW that prioritizes battery charging over energy export to local AC loads connected upstream of the Conext CSW battery inverter. AC Coupling must be enabled for this feature to be functional."
"In AC-coupled configuration and with grid present, the Conext SW monitors flow of power from its load port (AC Output) to the utility grid input (AC Input). When power flow to the grid is detected and the battery needs charging after a short delay, the Conext SW initiates a bulk charging cycle by switching to charge mode and drawing AC power to charge the battery. It only draws enough energy to keep the flow of power to the grid to zero. As long as the battery is able to accept the energy, the Conext SW will continue diverting excess PV production to the battery. Once the battery bank is at a level such that not all excess energy from the PV Inverter is being absorbed, the balance will then flow out to upstream loads in the house that are not connected to the Conext SW’s AC output port."
"If PV production is lost for a prolonged period of time and the battery bank discharges below the Recharge Volts setting, the Conext SW will initiate a normal charge cycle and use power from the grid to charge the battery. For information on this setting, see “Configuring Basic Settings” on page 4–7. In all of the above modes of operation, only the PV Inverter exports energy to the grid. The Conext SW itself cannot sell to the grid - it simply passes energy from the PV inverter to the grid. For this reason, the PV Inverter must be fully grid code compliant as it assumes responsibility for anti-islanding protection."
"NOTICE AC COUPLED PV INVERTER COMPATIBILITY AC power generated by AC coupling PV inverters with Conext SW must be consumed by AC loads or used to charge batteries. As an alternative, the excess power produced from a PV inverter can be routed to dump loads. Do not AC couple PV inverters with the Conext SW that are unable to reduce, derate or cease the excess PV inverter power in response to the changes in AC line frequency controlled by the Conext SW. Consult the manufacturer's specifications of your PV inverter and confirm compatibility. Failure to follow these instructions can result in equipment damage."
This is exactly what I have asked Schneider for in the XW-Pro. If energy is flowing back from the output to the grid input, start charging the battery. If the little SW can do this, why can't the far more capable XW-Pro do it? And to make it even worse... This is only in the 230 volt manual. It is not in the 120/240 volt split phase US spec manual. What is the reason for it only being on the Euro model?
All that being said, and how annoying it is to try and work with Schneider tech support. I still think they have some of the best hardware. If I was building another system for myself or a friend, I think I would still use an XW-Pro, but the PLC will be par of the install now. Use it is a basic device that we can control. It responds well to the external commands and has been working great like this. Since I write the code, I can choose exactly how much to let it export or consume. If true zero export is an absolute goal, then I would probably use all DC coupled solar. Then it is a simple matter of having all the loads on the output side of the XW-Pro and setting it to grid sell zero amps. With that setup, you don't even need the PLC. It does just "Work". The DC charge controllers will push all they can into the batteries. If the batteries top out, then the charge controllers curtail and sit in CV output mode to keep the batteries full. As long as the battery voltage stays above the "Grid Support Volts" it will provide all the power to run the loads after the XW-Pro. No need to back feed, all the load is on the output side. With one unit, this does limit you to a 60 amp 120/240 volt panel. IF you need more, you would need a second unit. Going a bit more complex, you could put some large loads in the main panel, but t feed them with zero export, then you need to add the Watt-Node for it to keep doing zero export. Once you add AC coupling it starts to fall apart. When on grid it just can't curtail the AC grid tied solar inverters. But for a system like I have, where my existing solar was already permitted for export, it works just fine as the worst case export is still less than before. And if I had 4,000 watts of DC coupled solar, the system would work without needing the PLC as the DC solar alone would provide enough battery charging. But ALL of the excess AC solar would export to the grid. This is why they suggested taking the microinverters off of half the array.
And Ampster... No, I did not have any DC solar until I added the 10 small NewPowa 100 watt panels and the BougeRV 40 amp charge controller. I would not say that the XW-Pro itself operates better with DC coupled solar, but the system as a whole certainly does. But again, without my PLC controlling the XW-Pro, the system really does not work at all unless you have enough DC solar to provide your desired evening power. That is the big Gotcha! in this whole mess.
GXMnow
Solar Wizard
- Joined
- Jul 17, 2020
- Messages
- 2,710
The post was getting too long, so I am continuing a bit here.
For whatever reason, the people at Schneider just don't see "AC coupling as a primary power source" as being something they want to support. As we have proven, it can work, and work vey well. I even had it working without using any external sensing. That means the functions I was doing to make the XW-Pro "Smart Charge" could be done with just code in the XW-Pro and/or the Gateway or Insight boxes. Before I added the grid side meters, I was only reading the Grid Watts and Output Watts from the XW-Pro, and then having a little code decide if it should charge and at what rate to reduce the grid export to what ever level I decided. And if the system had the Schneider approved Watt-Node connected, it could also add everything else I am doing since I added my main panel power meters. This is not a technical or hardware issue. Schneider is making a deliberate choice to not support this in their own software.
I have to wonder if this is some form of an OEM deal they got stuck with???
Let me explain my thoughts. SunPower is a huge company. They sell ultra premium solar panels at a much higher price than most of the competition. But having just solar panels was not enough. They went to Enphase for microinverters. They are a specific OEM version. They run at a higher voltage than the normal ones as the SunPower panels are 96 cells in series. And they have their own monitoring portal which is not compatible with other Enphase systems like the Envoy. Fast forward a couple years, and SunPower realized they needed storage to compete. Why didn't they just go to Enphase again? I have no idea, but the Ensemble system from Enphase is a nice setup, it is just expensive. That alone is not a problem for SunPower. So what did SunPower do? They decided to build their own storage system, the SunVault. But if you look inside the SunVault, it has a Schneider XW-Pro mounted inside. It is AC coupled only to the SunPower panels that have Enphase microinverters on them. And SunPower uses their own (Envoy/Gateway) box that controls it, very similar to what my PLC is doing, but with a nice polished web UI. Is there a chance that their deal with SunPower is keeping them from offering this functionality?
For whatever reason, the people at Schneider just don't see "AC coupling as a primary power source" as being something they want to support. As we have proven, it can work, and work vey well. I even had it working without using any external sensing. That means the functions I was doing to make the XW-Pro "Smart Charge" could be done with just code in the XW-Pro and/or the Gateway or Insight boxes. Before I added the grid side meters, I was only reading the Grid Watts and Output Watts from the XW-Pro, and then having a little code decide if it should charge and at what rate to reduce the grid export to what ever level I decided. And if the system had the Schneider approved Watt-Node connected, it could also add everything else I am doing since I added my main panel power meters. This is not a technical or hardware issue. Schneider is making a deliberate choice to not support this in their own software.
I have to wonder if this is some form of an OEM deal they got stuck with???
Let me explain my thoughts. SunPower is a huge company. They sell ultra premium solar panels at a much higher price than most of the competition. But having just solar panels was not enough. They went to Enphase for microinverters. They are a specific OEM version. They run at a higher voltage than the normal ones as the SunPower panels are 96 cells in series. And they have their own monitoring portal which is not compatible with other Enphase systems like the Envoy. Fast forward a couple years, and SunPower realized they needed storage to compete. Why didn't they just go to Enphase again? I have no idea, but the Ensemble system from Enphase is a nice setup, it is just expensive. That alone is not a problem for SunPower. So what did SunPower do? They decided to build their own storage system, the SunVault. But if you look inside the SunVault, it has a Schneider XW-Pro mounted inside. It is AC coupled only to the SunPower panels that have Enphase microinverters on them. And SunPower uses their own (Envoy/Gateway) box that controls it, very similar to what my PLC is doing, but with a nice polished web UI. Is there a chance that their deal with SunPower is keeping them from offering this functionality?
Interesting theory.
It’s always mystified me and frustrated me that the XW-Pro does not offer SW-like battery charging capability and that the SW does not offer XW-Like export capability.
It’s like: ‘we can sell you this 90% of a perfect product or this 90% of a perfect product, but we don’t there is really any market demand for a product that is 100% perfect.’
Any idea of what % of Schneider’s Conext sales are through SunRun?
Ampster
Renewable Energy Hobbyist
Could be. I feel a similar situation happened at Outback when they were acquired by Enersys, a battery supplier. I could not get any cooperation about using CAN communications with my Orion BMS.
Xmasterfunk
New Member
- Joined
- Apr 25, 2022
- Messages
- 13
""people at Schneider just don't see "AC coupling as a primary power source" as being something they want to support. ""
Maybe some people need to start purchasing the Mojave?
I have watched a bunch of their contractor videos now and they are the only company I see explicitly signing on to the AC-coupling scheme. Its refreshing to see the use cases explicitly shown in their presentations. That's the only reason i have started looking at Outback instead of Schneider. Especially with the regulations in CA. I personally believe California's codes will start propagating in other districts, and those companies that don't jump on the band wagon will get left behind.
This would be unfortunate for Schneider as i believe they are the few inverter OEM that is not trying to use the inverter to sell their batteries. As Enphase, Outback, and i think Generac show, you pay about 2-3 times much per kWhr for batteries that come with an inverter; as opposed to buying the inverter and sourcing from 2nd hand batteries, or even some first hand batt from battery only OEMs.
GXMnow, although i have not been able to read all your threads, i do enjoy reading you accounts through it all. It is really hard to find so much detail on these subjects and the industry is really lagging behind where we need to be, and where i think consumers are willing to put their wallets. I hope that people like you, myself and others on forums like this can help nudge the industry to get out want the people need.
GXMnow
Solar Wizard
- Joined
- Jul 17, 2020
- Messages
- 2,710
I just got my monthly bill for Oct. 18 to Nov. 17 from SCE.
Sorry if this is rambling a bit.
I ended the billing month with a net energy export to the grid of 80 KWHs total. They applied a credit of $17.22 for the energy and delivery charges. That works out to $0.215 per kilowatt hour. on the credit. Even though I opted for the monthly billing, they are still keeping the energy charges and the taxing as separate items. I already had a credit of $12.80 from the previous month, so my energy credit is now up to $30.02 after 2 months into the next full year cycle. On the taxes and non bypassable charges, they charged me $10.52 but I still had $47.63 credit because of the climate credit I got last month. So my balance on the taxes side is still at $37.11 that they owe me. If they did just total up each month, I should have just gotten a credit of $6.70 for this month. I really don't get why they still want to run the two separate totals.
With how low I was able to keep the consumption and export numbers, the total taxes were actually just $1.60 which is well below my $9.85 minimum charge, so they tacked on "balance of minimum charge" of $8.92 It seems odd that I net exported 80 KWH, only consumed 11 KWH and exported 91 KWHs. But my Non Bypassable charges are based on 24 KWHs. I have no idea where that number is coming from.
Sorry if this is rambling a bit.
I ended the billing month with a net energy export to the grid of 80 KWHs total. They applied a credit of $17.22 for the energy and delivery charges. That works out to $0.215 per kilowatt hour. on the credit. Even though I opted for the monthly billing, they are still keeping the energy charges and the taxing as separate items. I already had a credit of $12.80 from the previous month, so my energy credit is now up to $30.02 after 2 months into the next full year cycle. On the taxes and non bypassable charges, they charged me $10.52 but I still had $47.63 credit because of the climate credit I got last month. So my balance on the taxes side is still at $37.11 that they owe me. If they did just total up each month, I should have just gotten a credit of $6.70 for this month. I really don't get why they still want to run the two separate totals.
With how low I was able to keep the consumption and export numbers, the total taxes were actually just $1.60 which is well below my $9.85 minimum charge, so they tacked on "balance of minimum charge" of $8.92 It seems odd that I net exported 80 KWH, only consumed 11 KWH and exported 91 KWHs. But my Non Bypassable charges are based on 24 KWHs. I have no idea where that number is coming from.
GXMnow
Solar Wizard
- Joined
- Jul 17, 2020
- Messages
- 2,710
The last two days have been horrible for solar power here in not so sunny So Cal.
For the last 2 weeks of Nov. I was still managing to produce nearly 16 KWHs a day from the Enphase panels, and the DC panels were also pushing over 2.5 KWHs to the battery bank. That was enough that I was still managing to zero out the electric usage at the utility meter. We had a few low production days, but when the following day was decent again, the system was able to cover it by using a bit more battery one day and recovering the next. But now December first and second have depleted my reserve. Dec. 1 only produced 1.6 KWHs and Dec. 2 was only a tick better at 2.7 KWHs. The battery bank covered through the 4 to 9 pm high rate time of Dec. 1, but then I put it into grid charge mode and used 12 KWH trying to get some juice into the battery again. Then I used that stored battery power and covered 4 to 9 pm again of Dec. 2. Today is looking a little better, but I have it once again charging a bit from grid on the cheap rate to be sure I can cover the evening high rate. With minimal usage in the house, I need to get the battery bank up to about 52.5 volts to make it through 4 pm to 9 pm. That is only a 1.5 volt drop while pushing out about 16 amps on average. 52 volts x 16 amps = 832 watts x 5 hours = about 4.2 KWHs. Due to losses in the charging and inverting, I need to use around 5 KWHs minimum to cover the high rate block.
For the last 2 days, the DC solar certainly helped, especially when you look at what made it into the battery bank. Most of the time, the Enphase panels were not even able to cover the load, so they did not provide any charge power to the battery bank. At the same time, even the small power the DC panels made was still going into the battery. Dec. 1 made 479 watt hours and Dec. 2 made 644 watt hours. It is just after 8 am, and the DC panels are already pushing over 70 watts. The parasitic loss seems to be a little lower on the DC charge controller. While the production is very low, the DC panels are able to do better, relative to the Enphase array. While I am seeing 70 watts on the DC panels, the Enphase array is only making 350 watts. Yes, that is more power, but it is coming from 4,800 watts of panels. 350 / 4,800 = 7.3% production. The DC system is only 1,000 watts of panels. Doing the same math we get 70 / 1,000 = 7.0% production. When there is decent sun, the DC panels are typically only at 92% of the production of the large array, this is 96% of the production. I have pretty much accepted that the NewPowa panels are really just 92 watts each, not 100 watts when comparing them to my SilFab 300 watt panels.
If we go by full day production, let's see how the math works out.
On Dec. 1, the Enphase / SilFab system produced 1.6 KWHs from 4,800 watts of panels 1.6 / 4.8 = 0.3333 Sun Hours.
On the same day, the BougeRV / NewPowa system produced 479 watt hours from 1,000 watts of panels 0.479 / 1 = 0.479 Sun Hours.
That is much better for the DC system. Instead of the more typical 92%, the DC system produce 144% of the sun hours for that day.
On Dec. 2 we did much better, getting 2.7 KWHs from the Enphase system. 2.7 / 4.8 = 0.56 Sun Hours.
The DC system also did much better at 644 watt hours. That is 0.64 Sun Hours.
Still better at 114% of the production rate of the Enphase, but you can start to see that as the sun increases, the Enphase takes off and does better. Going back and picking a random Nov. day, Nov. 18th, let's see what we got.
Nov. 18 Enphase production was 17.9 KWHs. 17.9 / 4.8 = 3.73 Sun Hours
Nov. 18 BougeRV production was 3.482 KWHs / 1,000 = 3.48 Sun Hours
The BougeRV/NewPowa system only did 3.48 / 3.73 = 93.3% of the production of the Enphase/SilFab system.
And back when it was very sunny, it was around the 92% mark all the time. At first I was thinking it might be the charge controller was less efficient, but if that was the case, the low solar production would take the bigger hit. This is proving that at very low sun, the BougeRV charge controller is actually very efficient. If anything, it is the Enphase system that has some parasitic loss when running at low power. And they does make some sense. It has to power the inverter electronics and the whole data logging and grid code compliance software. But once the solar production covers that low power loss, the efficiency just get's better and better as the sun increases. The DC charge controller seems to have very little power overhead to cover, so it does well with small input power.
For the last 2 weeks of Nov. I was still managing to produce nearly 16 KWHs a day from the Enphase panels, and the DC panels were also pushing over 2.5 KWHs to the battery bank. That was enough that I was still managing to zero out the electric usage at the utility meter. We had a few low production days, but when the following day was decent again, the system was able to cover it by using a bit more battery one day and recovering the next. But now December first and second have depleted my reserve. Dec. 1 only produced 1.6 KWHs and Dec. 2 was only a tick better at 2.7 KWHs. The battery bank covered through the 4 to 9 pm high rate time of Dec. 1, but then I put it into grid charge mode and used 12 KWH trying to get some juice into the battery again. Then I used that stored battery power and covered 4 to 9 pm again of Dec. 2. Today is looking a little better, but I have it once again charging a bit from grid on the cheap rate to be sure I can cover the evening high rate. With minimal usage in the house, I need to get the battery bank up to about 52.5 volts to make it through 4 pm to 9 pm. That is only a 1.5 volt drop while pushing out about 16 amps on average. 52 volts x 16 amps = 832 watts x 5 hours = about 4.2 KWHs. Due to losses in the charging and inverting, I need to use around 5 KWHs minimum to cover the high rate block.
For the last 2 days, the DC solar certainly helped, especially when you look at what made it into the battery bank. Most of the time, the Enphase panels were not even able to cover the load, so they did not provide any charge power to the battery bank. At the same time, even the small power the DC panels made was still going into the battery. Dec. 1 made 479 watt hours and Dec. 2 made 644 watt hours. It is just after 8 am, and the DC panels are already pushing over 70 watts. The parasitic loss seems to be a little lower on the DC charge controller. While the production is very low, the DC panels are able to do better, relative to the Enphase array. While I am seeing 70 watts on the DC panels, the Enphase array is only making 350 watts. Yes, that is more power, but it is coming from 4,800 watts of panels. 350 / 4,800 = 7.3% production. The DC system is only 1,000 watts of panels. Doing the same math we get 70 / 1,000 = 7.0% production. When there is decent sun, the DC panels are typically only at 92% of the production of the large array, this is 96% of the production. I have pretty much accepted that the NewPowa panels are really just 92 watts each, not 100 watts when comparing them to my SilFab 300 watt panels.
If we go by full day production, let's see how the math works out.
On Dec. 1, the Enphase / SilFab system produced 1.6 KWHs from 4,800 watts of panels 1.6 / 4.8 = 0.3333 Sun Hours.
On the same day, the BougeRV / NewPowa system produced 479 watt hours from 1,000 watts of panels 0.479 / 1 = 0.479 Sun Hours.
That is much better for the DC system. Instead of the more typical 92%, the DC system produce 144% of the sun hours for that day.
On Dec. 2 we did much better, getting 2.7 KWHs from the Enphase system. 2.7 / 4.8 = 0.56 Sun Hours.
The DC system also did much better at 644 watt hours. That is 0.64 Sun Hours.
Still better at 114% of the production rate of the Enphase, but you can start to see that as the sun increases, the Enphase takes off and does better. Going back and picking a random Nov. day, Nov. 18th, let's see what we got.
Nov. 18 Enphase production was 17.9 KWHs. 17.9 / 4.8 = 3.73 Sun Hours
Nov. 18 BougeRV production was 3.482 KWHs / 1,000 = 3.48 Sun Hours
The BougeRV/NewPowa system only did 3.48 / 3.73 = 93.3% of the production of the Enphase/SilFab system.
And back when it was very sunny, it was around the 92% mark all the time. At first I was thinking it might be the charge controller was less efficient, but if that was the case, the low solar production would take the bigger hit. This is proving that at very low sun, the BougeRV charge controller is actually very efficient. If anything, it is the Enphase system that has some parasitic loss when running at low power. And they does make some sense. It has to power the inverter electronics and the whole data logging and grid code compliance software. But once the solar production covers that low power loss, the efficiency just get's better and better as the sun increases. The DC charge controller seems to have very little power overhead to cover, so it does well with small input power.
GXMnow
Solar Wizard
- Joined
- Jul 17, 2020
- Messages
- 2,710
From the morning to about 2 pm, shadows are not a problem here. This is the array view today, at 1:44 PM
The very bottom left corner is just starting to get the shadows from the palm trees. Here is the entire day from Nov. 29, the last decent production day before the storm came through.
If you compare the two plots, you will notice that the lower roof, especially the far bottom left corner, does not get all the much more energy after 2 pm, while the upper roof on the right side pulls in another 300 watt hours per panel. By 4 pm, basically the entire lower roof is in shadow. If I totally chopped down the last of my palm trees, it will help a little, but only in the dead of winter with the very low sun angle. And the local tree guys want over $1,000 per tree to take them down, and it is 4 more trees. Ouch. In the summer, when I need more power to run the A/C, I don't get any shade issues from those trees.
Too bad I don't get anywhere near this kind of information from the DC array. But I have manually watched it a few days by checking it from time to time and it does give the total energy and peak power for each day. Again, in the dead of winter with the low sun angle, the DC array on the garage is getting a morning shadow from my neighbor's palm trees. But this is only in the very early morning, and it clears by about 9:30 am. And the DC panels still make more sun hours in that time than the Enphase array. At the same time, the entire lower roof is still shaded by the second floor of the house. The sun does not fully hit the lower roof array until 10 am or so.
The bad part about the DC array is that I am running series strings. Currently it is 2 parallel strings of 5 panels each in series. I did a little experiment, and just the shadow of my arm over one panel drops the DC array output by over 40%. And if I put that shadow on both rows, the output is 70% less than with no shadows. It's a good thing that the DC array has virtually no shadow from 9 am to 5 pm, even with the low winter sun. In summer, it should be shadow free from 8 am to 6 pm.
After the last 2 days of lousy production, it is so nice to have sun again. It is now 2 pm. The Enphase/SilFab array has already produced 12.2 KWHs, and the BourgeRV/NewPowa array has produced 2.22 KWHs. Already, the DC array alone has made more energy than the whole system did all day on Dec. 1. The lower roof of the Enphase system is now getting shadows and the output has dropped to 1,340 watts (28% of rated power), while the much smaller DC array without shadows is still putting out about 450 watts (45% of rated power). Looking at each panel shows what is going on here.
The shaded panels are down to just 10 to 12 watts while the panels that are getting full sun are still making 148 to 159 watts even with the low sun angle. So the panels getting full sun are still doing better than the DC array. And you will notice it is 8 panels making good power and 8 panels not doing well. That is the upper vs lower roof. Putting the panels 10 feet higher gets them out of the shadows. from the low evening winter sun.
The very bottom left corner is just starting to get the shadows from the palm trees. Here is the entire day from Nov. 29, the last decent production day before the storm came through.
If you compare the two plots, you will notice that the lower roof, especially the far bottom left corner, does not get all the much more energy after 2 pm, while the upper roof on the right side pulls in another 300 watt hours per panel. By 4 pm, basically the entire lower roof is in shadow. If I totally chopped down the last of my palm trees, it will help a little, but only in the dead of winter with the very low sun angle. And the local tree guys want over $1,000 per tree to take them down, and it is 4 more trees. Ouch. In the summer, when I need more power to run the A/C, I don't get any shade issues from those trees.
Too bad I don't get anywhere near this kind of information from the DC array. But I have manually watched it a few days by checking it from time to time and it does give the total energy and peak power for each day. Again, in the dead of winter with the low sun angle, the DC array on the garage is getting a morning shadow from my neighbor's palm trees. But this is only in the very early morning, and it clears by about 9:30 am. And the DC panels still make more sun hours in that time than the Enphase array. At the same time, the entire lower roof is still shaded by the second floor of the house. The sun does not fully hit the lower roof array until 10 am or so.
The bad part about the DC array is that I am running series strings. Currently it is 2 parallel strings of 5 panels each in series. I did a little experiment, and just the shadow of my arm over one panel drops the DC array output by over 40%. And if I put that shadow on both rows, the output is 70% less than with no shadows. It's a good thing that the DC array has virtually no shadow from 9 am to 5 pm, even with the low winter sun. In summer, it should be shadow free from 8 am to 6 pm.
After the last 2 days of lousy production, it is so nice to have sun again. It is now 2 pm. The Enphase/SilFab array has already produced 12.2 KWHs, and the BourgeRV/NewPowa array has produced 2.22 KWHs. Already, the DC array alone has made more energy than the whole system did all day on Dec. 1. The lower roof of the Enphase system is now getting shadows and the output has dropped to 1,340 watts (28% of rated power), while the much smaller DC array without shadows is still putting out about 450 watts (45% of rated power). Looking at each panel shows what is going on here.
The shaded panels are down to just 10 to 12 watts while the panels that are getting full sun are still making 148 to 159 watts even with the low sun angle. So the panels getting full sun are still doing better than the DC array. And you will notice it is 8 panels making good power and 8 panels not doing well. That is the upper vs lower roof. Putting the panels 10 feet higher gets them out of the shadows. from the low evening winter sun.
GXMnow
Solar Wizard
- Joined
- Jul 17, 2020
- Messages
- 2,710
The height difference between the little wall around the garage roof and the panels is only about 8 inches. For that to cast a shadow, the sun has to be very low in the sky. At this time of year, the morning shadow from the neighbor's trees is much higher than the wall, so the wall is well out of the picture before the tree shadows clear. Since my house is turned about 20 degrees east of due south, the evening shadow that hits first is actually from the 2 story house across the street. And it does not hit the panels until about 5 pm. By then, the angle is already so bad that each panel is down to about 10 watts before the shadow hits. Production cut a little early today as heavy clouds rolled in about 3 pm. The clouds dropped the Enphase system output from 1,100 watts to just 400 watts.
GXMnow
Solar Wizard
- Joined
- Jul 17, 2020
- Messages
- 2,710
Here is my SCE usage graph so far for this billing period. We are just past half way to the next bill.
As you can see, these last 4 days have been bad. This view does not show it, but I did have to buy a little of the high rate time of use power. I have been grid charging the battery a little to try and cover it, but I fell short by 5 KWHs, not too bad. Today started looking better, but the clouds are back once again. The weather forecast for today is again off and on clouds and fog pretty much all day. I will kick it into slow charge again tonight as they are once again predicting clouds tomorrow. But then they say sunny for Wed, Thur, and Fri.
I just fell under producing as much power as I am using, making me a net consumer for the month at this point. That was starting to annoy me, but then I looked at the same period last year.
Needless to say, I am doing a lot better than last year. There are two big differences. The DC panels are producing about 2.2 to 3 KWHs each good sunny day. That alone almost zeros half the days, but not quite. So where is the other savings? My son is away at college. He would leave his TV, Xbox, and PC running virtually 24/7. That probably accounts for around 2 KWHs a day that I am no longer consuming. This plot shows that my bill for this period last year (just for the energy) was $56.67, and that was at a lower rate, we had 2 increases over the last year. How low can I cut it this year? So far, the site is estimating my monthly energy bill will be about $15. That is quite a difference. Between the PLC program doing it's job, the DC panels adding more energy, and my son not being here, my bill could be less than 30% of what it was last year.
As you can see, these last 4 days have been bad. This view does not show it, but I did have to buy a little of the high rate time of use power. I have been grid charging the battery a little to try and cover it, but I fell short by 5 KWHs, not too bad. Today started looking better, but the clouds are back once again. The weather forecast for today is again off and on clouds and fog pretty much all day. I will kick it into slow charge again tonight as they are once again predicting clouds tomorrow. But then they say sunny for Wed, Thur, and Fri.
I just fell under producing as much power as I am using, making me a net consumer for the month at this point. That was starting to annoy me, but then I looked at the same period last year.
Needless to say, I am doing a lot better than last year. There are two big differences. The DC panels are producing about 2.2 to 3 KWHs each good sunny day. That alone almost zeros half the days, but not quite. So where is the other savings? My son is away at college. He would leave his TV, Xbox, and PC running virtually 24/7. That probably accounts for around 2 KWHs a day that I am no longer consuming. This plot shows that my bill for this period last year (just for the energy) was $56.67, and that was at a lower rate, we had 2 increases over the last year. How low can I cut it this year? So far, the site is estimating my monthly energy bill will be about $15. That is quite a difference. Between the PLC program doing it's job, the DC panels adding more energy, and my son not being here, my bill could be less than 30% of what it was last year.
GXMnow
Solar Wizard
- Joined
- Jul 17, 2020
- Messages
- 2,710
Looks like it might be time to have Enphase look at my Envoy. I have now had no data from the iQ7 inverters for about 3 days. It is obvious that all 16 are producing power, but I can't see any panel data. Even enlighten is only showing total system power, no per panel array data now.
I just power cycled only the Envoy. I will wait about 30 minutes and see if it reconnects. If it does not, I will power cycle the Envoy tonight after the sun goes down and see if they all wake up tomorrow. In either case, I think I need to contact Enphase and see what they say.
I just power cycled only the Envoy. I will wait about 30 minutes and see if it reconnects. If it does not, I will power cycle the Envoy tonight after the sun goes down and see if they all wake up tomorrow. In either case, I think I need to contact Enphase and see what they say.
GXMnow
Solar Wizard
- Joined
- Jul 17, 2020
- Messages
- 2,710
The problem is not on the internet side. The Envoy is responding on my network and also reporting the total energy to Enlighten. The only problem is the Envoy is not talking on the powerline comm to the iQ7's. It was intermittent for a few months, it would lose comm from just minutes to an hour or so, but it always came back. Well, it's not coming back now. Rebooting the Envoy board did not help. I don't want to power cycle the panels while they are producing power, so I will do that after sun down tonight. I sent an e-mail to Enphase support, they have not replied yet. Production peaked at 3,300 watts today which is excellent for all 16 panels in December.
The DC panels show a peak of 700 watts, but was actually floating between 500 and 600 watts. And now we have another band of clouds so the power is all over the place. The peak of 700 watts is 70% of rated power, and 3,300 watts is 69% of the rated 4,800 watts from the Enphase array. So I do not see anything wrong with production at all. It is just the communication to the micros.
What frequency range is the Enphase powerline communication? I really doubt it is an interference issue, but I will try turning off a few things to see if it helps. The fact that I lose all 16 together, every time this happened, it has to be the Envoy itself. If I just lost one of the roof arrays, I would be more inclined to think it could be interference, but it loses both arrays every time.
The DC panels show a peak of 700 watts, but was actually floating between 500 and 600 watts. And now we have another band of clouds so the power is all over the place. The peak of 700 watts is 70% of rated power, and 3,300 watts is 69% of the rated 4,800 watts from the Enphase array. So I do not see anything wrong with production at all. It is just the communication to the micros.
What frequency range is the Enphase powerline communication? I really doubt it is an interference issue, but I will try turning off a few things to see if it helps. The fact that I lose all 16 together, every time this happened, it has to be the Envoy itself. If I just lost one of the roof arrays, I would be more inclined to think it could be interference, but it loses both arrays every time.
