How to draw large amount of power from batteries through DEYE/Sunsynk inverter?

[Hexados]

Hello everyone!

I hope you are all doing well. I am responsible of some solar projects, mainly from the electrical side of things, plus connecting batteries, BMS and programming, based on what is required. Usually, these work just fine, and the projects are all possible to do and are always met with success with the client being satisfied.

Today, however, I was tasked with something that I truly never came across before, as a demand anyway. Their system is composed of:

- SUN-8K-SG01LP1 DEYE inverter
- 8400 Watt solar panels (using the double MPPT of the inverter)
- 2x 200Ah GreenSun Solar LiFePo4 Batteries (in parallel. NOTE: Those are 1P15S LiFePo4, NOT 1P16S!)

Connections wise, including BMS communication using CAN protocol were all successful. So no problems there.

The problems arise with what the client wants on their "SUN-8K-SG01LP1" inverter (that seemed out of the normal):

1- The possibility to be able to draw out some 25-30 Amps from the batteries alone, or at the very least some 15 Amps on 230V. That's the equivalent of anything between 80-170 Ah on the batteries. I asked to make sure, and indeed they want to have the possibility to be able to empty the batteries in 2 hours if necessary.

2- They want a scheduling system that pretty much acts like a Hysteresis for the LiFePo4. For example, they want to draw only from the batteries even when utility is present (it's connected on the "GRID" side, not the "GEN" side as the frequency is not really stable), and as soon as it reaches a certain set point (say 20% SOC at night), the utility is to slowly charge the batteries back up to another certain set point (say 60%) until morning, where the solar will take over.

3- They want the ability to set a limit to the pull current (internally, not externally with a current limiter, etc.) from the utility, so that they can divide the load between utility and batteries. Yes, utility + batteries, not just Solar + Utility (with priority to Solar), nor Solar + Batteries, but utility + batteries. In other words, they want to set a certain amount of current (say 8 Amps) to draw from the utility, and the rest from the batteries.

To be completely honest, this is the first time I've seen such demands. Usually, I've seen that second option somewhat possible on inverters like Growatt, Blue Sun, Snaterm, Jesudom, Anern, Voltronic, Sako, MUST, etc., where you have the option to switch back from SUB to SBU as soon as the battery charging reaches a certain voltage. But even then, it only applied for Lead-Acid Tall Tubular, AGM or Gel batteries, and LiFePo4 ONLY IF BMS is disabled (user-defined battery settings). I didn't see DEYE having such a feature, to be honest.

For the first demand, we did a bit of testing when the batteries were around 60% SOC, and they began to add more and more load (at first still not realizing that they actually want that high a consumption/load), and when it reached some 80 Amps consumption on the batteries, the voltage had already dropped by more than 2 Volts, going below 47V threshold, which triggered an alarm on the inverter and immediately cut the power off the loads. (KINDLY REMEMBER, these are 15S cells LiFePo4, not 16S, so the max possible voltage to charge them with is actually 54.4V, not 57.6V, so voltages in the high 40s is normal. The batteries were displaying it on their own screens). I thought about it, and the only solution I could find is to go into "Advanced" mode, where I basically remove the BMS and focus solely on the voltage, and based on my calculations, apply the proper Bulk and Floating voltages (which are very close to each other), and a cut-off voltage that is the equivalent of around 20% of the batteries' SOC. I proposed it, but he refused to remove the BMS. Personally though, I'm against the idea of drawing that much power simultaneously.

As for their third demand, I tried checking if I could trick the scheduler in some way to at least stop the consumption from the batteries and switch to batteries instead, but no dice. It was just conflicting with the regular charging process which was set in the "Battery Settings", and simply refused to charge them until I disabled the scheduling altogether.

So, what do you guys think? What is the action that I should take? I can clearly see that there are many security risks as it is, especially on the batteries as they would have a significantly shorter life because of that heavy draw (among many other things), but hey, that's what they want.

Oh, and this is the battery they've bought (two of them):

https://www.greensunpv.com/js/htmledit/kindeditor/attached/20210601/20210601145350_93093.jpg

Thank you all in advance, and sorry for the long thread!

 

[JCSchwarb]

Hello everyone!

I hope you are all doing well. I am responsible of some solar projects, mainly from the electrical side of things, plus connecting batteries, BMS and programming, based on what is required. Usually, these work just fine, and the projects are all possible to do and are always met with success with the client being satisfied.

Today, however, I was tasked with something that I truly never came across before, as a demand anyway. Their system is composed of:

- SUN-8K-SG01LP1 DEYE inverter
- 8400 Watt solar panels (using the double MPPT of the inverter)
- 2x 200Ah GreenSun Solar LiFePo4 Batteries (in parallel. NOTE: Those are 1P15S LiFePo4, NOT 1P16S!)

Connections wise, including BMS communication using CAN protocol were all successful. So no problems there.

The problems arise with what the client wants on their "SUN-8K-SG01LP1" inverter (that seemed out of the normal):

1- The possibility to be able to draw out some 25-30 Amps from the batteries alone, or at the very least some 15 Amps on 230V. That's the equivalent of anything between 80-170 Ah on the batteries. I asked to make sure, and indeed they want to have the possibility to be able to empty the batteries in 2 hours if necessary.

2- They want a scheduling system that pretty much acts like a Hysteresis for the LiFePo4. For example, they want to draw only from the batteries even when utility is present (it's connected on the "GRID" side, not the "GEN" side as the frequency is not really stable), and as soon as it reaches a certain set point (say 20% SOC at night), the utility is to slowly charge the batteries back up to another certain set point (say 60%) until morning, where the solar will take over.

3- They want the ability to set a limit to the pull current (internally, not externally with a current limiter, etc.) from the utility, so that they can divide the load between utility and batteries. Yes, utility + batteries, not just Solar + Utility (with priority to Solar), nor Solar + Batteries, but utility + batteries. In other words, they want to set a certain amount of current (say 8 Amps) to draw from the utility, and the rest from the batteries.

To be completely honest, this is the first time I've seen such demands. Usually, I've seen that second option somewhat possible on inverters like Growatt, Blue Sun, Snaterm, Jesudom, Anern, Voltronic, Sako, MUST, etc., where you have the option to switch back from SUB to SBU as soon as the battery charging reaches a certain voltage. But even then, it only applied for Lead-Acid Tall Tubular, AGM or Gel batteries, and LiFePo4 ONLY IF BMS is disabled (user-defined battery settings). I didn't see DEYE having such a feature, to be honest.

For the first demand, we did a bit of testing when the batteries were around 60% SOC, and they began to add more and more load (at first still not realizing that they actually want that high a consumption/load), and when it reached some 80 Amps consumption on the batteries, the voltage had already dropped by more than 2 Volts, going below 47V threshold, which triggered an alarm on the inverter and immediately cut the power off the loads. (KINDLY REMEMBER, these are 15S cells LiFePo4, not 16S, so the max possible voltage to charge them with is actually 54.4V, not 57.6V, so voltages in the high 40s is normal. The batteries were displaying it on their own screens). I thought about it, and the only solution I could find is to go into "Advanced" mode, where I basically remove the BMS and focus solely on the voltage, and based on my calculations, apply the proper Bulk and Floating voltages (which are very close to each other), and a cut-off voltage that is the equivalent of around 20% of the batteries' SOC. I proposed it, but he refused to remove the BMS. Personally though, I'm against the idea of drawing that much power simultaneously.

As for their third demand, I tried checking if I could trick the scheduler in some way to at least stop the consumption from the batteries and switch to batteries instead, but no dice. It was just conflicting with the regular charging process which was set in the "Battery Settings", and simply refused to charge them until I disabled the scheduling altogether.

So, what do you guys think? What is the action that I should take? I can clearly see that there are many security risks as it is, especially on the batteries as they would have a significantly shorter life because of that heavy draw (among many other things), but hey, that's what they want.

Oh, and this is the battery they've bought (two of them):

https://www.greensunpv.com/js/htmledit/kindeditor/attached/20210601/20210601145350_93093.jpg

Thank you all in advance, and sorry for the long thread!

Interesting situation with your customer. It reminds me of patients who go in telling their doctor the medications they need instead of describing their symptoms and letting the doctor determine the best plan to health.

It sounds like the customer is giving you conflicting information. They want to maximize their solar, but also want some level of critical UPS backup. I would not get into to all these configurations as it is not feasible nor advisable. Mode SBU is the best mode for almost all situations. We want to use the DC system when it’s available. The only other configuration is do we want to use grid power to charge the batteries? If so, then when the system switches to bypass mode, I have mine set to 46V battery level, you can charge the batteries and supply power until a battery set point of say 52.5V or 60% SOC. I don’t like charging my batteries on grid, so I turned mine to OSO and set the amps to 5A. If they want to keep a minimum battery storage, then bump the bypass cut in to allow enough storage to get them by until the solar kicks in.

I designed my system to be as self-Sufficient as possible, to remain on DC 95% of the time, but when we have gray days it goes into bypass for a short time until the 48V batt set point is reached. I would recommend trying to stay out of bypass mode as much as possible. Sorry I didn’t answer your questions straight on, but tried to give you another perspective to consider. I would highly recommend that they increase their battery capacity to at least 30 kWh for a better running system.

Best,

Jay

 

[Hexados]

Interesting situation with your customer. It reminds me of patients who go in telling their doctor the medications they need instead of describing their symptoms and letting the doctor determine the best plan to health.

It sounds like the customer is giving you conflicting information. They want to maximize their solar, but also want some level of critical UPS backup. I would not get into to all these configurations as it is not feasible nor advisable. Mode SBU is the best mode for almost all situations. We want to use the DC system when it’s available. The only other configuration is do we want to use grid power to charge the batteries? If so, then when the system switches to bypass mode, I have mine set to 46V battery level, you can charge the batteries and supply power until a battery set point of say 52.5V or 60% SOC. I don’t like charging my batteries on grid, so I turned mine to OSO and set the amps to 5A. If they want to keep a minimum battery storage, then bump the bypass cut in to allow enough storage to get them by until the solar kicks in.

I designed my system to be as self-Sufficient as possible, to remain on DC 95% of the time, but when we have gray days it goes into bypass for a short time until the 48V batt set point is reached. I would recommend trying to stay out of bypass mode as much as possible. Sorry I didn’t answer your questions straight on, but tried to give you another perspective to consider. I would highly recommend that they increase their battery capacity to at least 30 kWh for a better running system.

Best,

Jay

Hello Jay,

Thank you very much for your reply! I 100% agree with your points. The thing is that the client read the specs on the battery system, that they can pull up to 100% of the load in one hour on LiFePo4 (which, technically, is true), but the BMS system simply doesn't allow it on the inverter. To be able to draw that much power, the client has to buy two more batteries, increasing their system to 800 Ah at least, to be able to compensate for that large voltage drop, but this simply is not going to happen cost-wise, and the BMS system is to stay as it is necessary. Similar to your system, the one I have installed for my own home is pretty much purely based on either solar or batteries, and due to the hyperinflation that my country is currently undergoing, I actually broke even within one year. I calculated how much I use per month to how much solar + battery I'm using, and despite keeping the same consumption (with Lead-Acid batteries no less, though I do personally maintain them all the time, down to acid regulation), I literally saved thousands of dollars. The only difference is that due to lack of electricity altogether, I put my system in SUB mode instead of SBU. This is for the occasional grey days that are inevitable throughout the year. Still saved thousands regardless.

Coming back to the client's system, the BMS on the DEYE, once enabled, does NOT allow for manual configuration of Bulk, Float and Cut-Off voltages. I can only create a scheduler and percentage at which it can cut everything off at best. So, my hands are pretty much tied at that point, which is why I was genuinely wondering if it is even possible to configure the system on the inverter the way that they had asked to begin with. I'm glad that you've given me more clarification on this issue, as I too think it is something impossible. The inverter is basically to act as an inverter, plus a heavy APS system that disregards BMS to begin with.

 

[houseofancients]

Hello Jay,

Thank you very much for your reply! I 100% agree with your points. The thing is that the client read the specs on the battery system, that they can pull up to 100% of the load in one hour on LiFePo4 (which, technically, is true), but the BMS system simply doesn't allow it on the inverter. To be able to draw that much power, the client has to buy two more batteries, increasing their system to 800 Ah at least, to be able to compensate for that large voltage drop, but this simply is not going to happen cost-wise, and the BMS system is to stay as it is necessary. Similar to your system, the one I have installed for my own home is pretty much purely based on either solar or batteries, and due to the hyperinflation that my country is currently undergoing, I actually broke even within one year. I calculated how much I use per month to how much solar + battery I'm using, and despite keeping the same consumption (with Lead-Acid batteries no less, though I do personally maintain them all the time, down to acid regulation), I literally saved thousands of dollars. The only difference is that due to lack of electricity altogether, I put my system in SUB mode instead of SBU. This is for the occasional grey days that are inevitable throughout the year. Still saved thousands regardless.

Coming back to the client's system, the BMS on the DEYE, once enabled, does NOT allow for manual configuration of Bulk, Float and Cut-Off voltages. I can only create a scheduler and percentage at which it can cut everything off at best. So, my hands are pretty much tied at that point, which is why I was genuinely wondering if it is even possible to configure the system on the inverter the way that they had asked to begin with. I'm glad that you've given me more clarification on this issue, as I too think it is something impossible. The inverter is basically to act as an inverter, plus a heavy APS system that disregards BMS to begin with.

telling your clients "no" is an valid answer.
giving them the best possible next scenario after them "no" might land you your best client ever or an ex-client that will turn out to a huge pain in the rear

 

[Hexados]

telling your clients "no" is an valid answer.
giving them the best possible next scenario after them "no" might land you your best client ever or an ex-client that will turn out to a huge pain in the rear

I agree, and I'm currently already at the second half of the latter scenario . This is why I was asking if there was anything that could be done on the inverter itself, even as a compromise. As stated above, upon testing, some 12 Amp (on 230V) pull from the batteries (so, around 75 Amps) through the inverter was still somewhat tolerated, but anything beyond that and the inverter's BMS cut everything off. So, what do you guys suggest as a compromise in this case, one that doesn't include additional battery purchase? (Note that I can suggest adding more electrical components if needed as per your suggestions, but nothing too spectacular and/or expensive of course). I would mainly like to know the best settings to set up on the SUN-8K-SG01LP1 DEYE inverter, even if it will be a compromise.

An extra point that I was thinking about is the generator that they have. Usually, the utility here (if it ever comes) is unstable as power provision, not just scarce as explained above, so that line was connected to the grid directly on the DEYE inverter. The generator, on the other hand, also has the occasional instability, but is very current limited, and the max RECOMMENDED that they can pull from it is 10 Amps. The generator's frequency also tends to be unstable from time to time. For this, I actually suggested that they include an electromechanical interlocking system (which they have), but with only delay timers for each (which they did NOT include), so that in the rare times that the utility is provided, they get to have full advantage of it. In other words, if the generator is online because the batteries are low and there is no solar provision, and suddenly utility is provided, the interlocking system is such that it would give priority to the utility and pass it through. On the inverter, however, just two lines are going in, and they are that of the "GRID". Because of the instability, I had suggested that they connect everything on grid. That sudden shift of phase as a source from GEN to GRID, not to mention a variation in voltage, ESPECIALLY when the electrical installation is such that the generator and utility have common neutral on the contactors, can wreak havoc on the inverter.

So, supposing that the interlocking system is modified such that generator and utility are independently differentiated on the inverter (GEN and GRID inputs respectively), and timers are added to the interlocking system so that there is at least a 10 second delay when transitioning from generator to utility and vice versa, would the inverter be able to handle? Because I personally had one case (not my installation though) where there was this exact same setup (same inverter too), minus the timers, and the inverter's "AC" side basically got shot, and was showing me an input voltage of over 550V, even when there were absolutely no generator or utility incoming at the time. Turning on the generator also did nothing.

Sorry about these intricate questions, by the way. Setups here in my country can get really crazy sometimes, either from the client side, or sometimes from the electrician's side (wrong programming, improper electrical installation, etc.).

 

[houseofancients]

so lets get some thing straight..
these inverters have no bms, they are merely controlled by the batteries bms.
not familiar with these batteries nor their bms, but sure sounds like they have a max discharge limit set .
could be hard bms limit , or merely a setting.
should check that.

second , these inverters have a grid connection ( non essential loads) ,a load connection ( essential loads) and an aux ( or smartload) port.

when the grid is there,best to leave heavy amps drawing applications ( heatpump, boiler/waterheater) that can be missed during a full black out as they are.
the grid will take the high amp draw, and inverter will what ever it can , if properly configured

use a seperate panel for essential loads.

look into saving them high usuage applications

as a last option ... add batteries

 

[houseofancients]

and as a very last option, sunsynk's youtube channel has complete installer course on it..
given your working on alternatives for your client, it would be a good idea to view them (completely) so you know the capabilities and limitations of the inverters you're dealing with

 

[Hexados]

so lets get some thing straight..
these inverters have no bms, they are merely controlled by the batteries bms.
not familiar with these batteries nor their bms, but sure sounds like they have a max discharge limit set .
could be hard bms limit , or merely a setting.
should check that.

second , these inverters have a grid connection ( non essential loads) ,a load connection ( essential loads) and an aux ( or smartload) port.

when the grid is there,best to leave heavy amps drawing applications ( heatpump, boiler/waterheater) that can be missed during a full black out as they are.
the grid will take the high amp draw, and inverter will what ever it can , if properly configured

use a seperate panel for essential loads.

look into saving them high usuage applications

as a last option ... add batteries

and as a very last option, sunsynk's youtube channel has complete installer course on it..
given your working on alternatives for your client, it would be a good idea to view them (completely) so you know the capabilities and limitations of the inverters you're dealing with

Thank you again for your reply. That final option is worthwhile to check out for sure. I'm doing it as we speak.

Concerning your first point, I'm actually a bit surprised that you mentioned that even these inverters don't really have a built-in BMS system within them, and they can only communicate with the battery's BMS. That would mean that pretty much none of the inverters that come from China to our country (like Growatt, Huawei, Voltronic, Sunsynk, etc.) really don't come with a BMS! I make this statement since Deye/Sunsynk is pretty much the most expensive of the bunch. You also have even "weaker" inverters that don't even come with a BMS port for CAN communication here, but whatever. As for the batteries, I checked the settings, and it really looks like that it's a hard BMS current limit. This begs the question then: Was it the inverter itself that cut off the battery, or the battery itself? After all, when the disconnection did occur, the inverter was still running on batteries alone. This was still during the testing phase and parametrization, so the Grid and the PV's breakers were still manually tripped.

As for your second point, the AUX/GEN port (based on one of the videos I watched from SunSynk, minding the comment section), what I had mentioned above seems to hold true. This particular port seems to be very sensitive to frequency fluctuations. I do need additional input from users and other installers about this particular point, but this is what I've noticed thus far. Anything else concerning heavy loads is connected to both PV and Grid anyway. The availability of GEN would simply add more options for me of what I can do as a compromise, which is why it was mentioned. As for the heavy load part on the GRID, it's all already set up like that, as intended. Pretty much every system installed here is like that, protected with the regular circuit breaker, often rated slightly lower than the max current passthrough of the inverter in question.

 

[Hexados]

Update: After tinkering with the inverter, I was able to find a way around the BMS. It's a strange and somewhat unconventional way, but here it goes:
- The very first thing to do is to switch off the circuit breaker of the load, PV, Grid and Gen inputs respectively, and only leave the battery powering the inverter alone.
- When connected to the LiFePo4 battery with the appropriate communication cable using the CAN protocol, go the "Battery Settings" and instead of Lithium (while previously connected with Lithium!), choose "Use Batt V".
- Apply it and then exit the setting. You will most likely get a communication fault here, which is normal!
- Go back to "Battery Settings", and now go down a couple of pages with the on-board keys, and find the settings where you find low voltage and battery cut-off voltages. These are now accessible and do apply EVEN WHEN REVERTING BACK to Lithium option!!
- The battery that the client was using happened to be a 1P15S system, so a lower overall voltage from the conventional 1P16S, maxing out at a bulk charge of 54.4V instead of the standard 57.6V. The limit that I put to the cut-off and low voltages are 42V and 43V respectively.
- Apply the settings minding the other parameters of course (you don't want to overcharge your LiFePo4 batteries by accident), and then go back to the first page aforementioned.
- Select "Lithium" again, and test it out now.

I have tested this and it worked marvelously! I was able to pull a whopping 150 Amps from the batteries for a small period of time (since it was just a test). Today, the client needed that large power pull, and was able to pull over 110 Amps for over two hours. The batteries held well. The inverter was ventilating and emitting hot air from the sides, but this was normal. What gave, however, were those DC circuit breakers ?. They simply weren't rated for this type of task, so they tripped. A minor inconvenience for a major victory. I did advise them not to always go for such large current consumptions all the time, to which he was understanding (it's just that today we had a crazy storm (happens to us around this time of year for a few days), and to the area that they were in, A LOT of heating was required).

 

[toni999]

Hello, I have seen this thread and I am interested because I have a similar problem

I have a 10KW three-phase DEYE and some batteries with 16 cells of 280 Ah, the BMS is 400A, the circuit breaker is 200 and the cables are 50mm2, as you can see, everything is sized for a large current.

Well, I can't get the battery to charge as I want.

The problem is that I want to maximize self-consumption because they pay very little for injected energy, and purchased energy is very expensive.

In this country the electric companies function as a monopoly.

In addition, in my area in winter there is a lot of fog and many days there are only 2 or 3 hours of sun

I have 18 450W panels and they produce about 6 KWh, but the inverter prefers to inject current into the network rather than charge the battery

Example battery at 20% charging 12 amps and exporting 3KW I do the test of disconnecting the outside line and suddenly it starts charging at 45 Amps.

I reconnect the outside line and it returns to the previous state, that is, charging 10 amps and exporting the rest.

In this situation, after two hours the sun goes down and nothing has been charged.

This is the behavior I don't like it seems to prioritize import export over battery usage.

Another test on a sunny day, night comes and the battery is at 95%, because it starts to buy energy from the network instead of using the battery.

I have done many tests, but I can not optimize it, the only thing that seems to affect the loading speed is the export limit parameter if I limit it, the battery charge increases.

Now, my goal is to build a board with Arduino or similar, which reprograms the export limit parameter depending on the state of charge of the battery, but I'm not really into this type of board and I'm still looking for an example in " C" to make the communication.