Novice Questions About Complex Multi-Array, Multi-Inverter, ESS Implementation

[6Bridges]

Hi Folks, this is my first time posting. I have read a bunch of postings and it is clear there are some very experienced people on this forum. I had some basic questions I wanted to pass by some experienced practitioners regarding the target equipment I am considering for my installation.

I apologize in advance for the length of the posting. If you want to jump right to the questions without the background, please scroll to the scenarios described at the bottom of this posting.

Here is the background…

I am a computer scientist by training, and a solar enthusiast due to my desire to be resilient in the face of the increasing risks of climate change. I am not concerned about an armageddon, but I am interested in keeping the buildings on my property including my currently being built high-performance (built to Passive House standards), all electric (no fossil fuels) house operating when the grid is down for 10 days due to an ice storm and there is little to no PV production due to cloudy conditions.

When I purchased my property in 2022 it came equipped with a roof-mounted array on the barn made up of 61 modules each rated at 300 watt DC for a total of 18.3kW DC. This array is connected to two Fronius Inverters (8.2kW AC watts and 6kW AC watts) for a total AC output of 14.2kW. The current system has no batteries.

My goal is to install an energy management system based on Savant Power Systems solution. Although their website now shows different tech (Savant purchased POMCube), their solution historically utilized Sol-Ark inverters, HomeGrid batteries, and the Savant Power Modules, which allow for software-controlled load centers. Since they use Sol-Ark inverters, they can also accommodate a backup generator in the solution. I will be using the Sol-Ark/HomeGrid version of the Savant solution.

Because of electrical needs of the buildings on the property, will add a second ground-mounted array spec’ed at 23 kW DC and 16 kW AC. Will be combining everything with three Sol-Ark 15k inverters and four HomeGrid stacks each with four 4.8 kW batteries providing a total ESS of 76.8 kWH. And, as mentioned, the system will include Savant Power Systems for smart load centers and overall control of the system.

We are also installing a 600 Amp new mainline from our local electric provider. Since we want to be able to passthrough all 600 Amps of grid power to the load centers, we will need three Sol-Ark 15k inverters (each with a 200 Amp bypass). We will also have a propane-fired generator connected to provide additional resilience when the grid is down.

I am trying to make sure I understand the logic that controls which energy sources the inverter will choose to use in particular situations, along with a few edge cases.

Here are my assumptions about my configuration (which might be wrong):

1. Each Sol-Ark 15 can supply up to a max of 62.5A of output from the sum of the AC-coupled solar (from the two existing Fronius inverters) and the new DC-coupled ground-mounted system I will install.
2. One of the people associated with my project said it would be only safe to assume we will get 50A of PV output per inverter.
3. 50 A is the maximum continuous battery output from a single inverter.
4. I have been told that the PV output and the battery output are not additive (so, if we can get a max of 50 A PV and 50 A batteries, that does NOT mean we can get 100 A of combined power from the inverter). So, the max continuous inverted power output we would ever get from a single inverted is 50 Amps. Is this a correct assumption?
5. Assuming 50 A is the max out, then with three inverters we would have a max output of 150 A (meaning that is the max power that would be available to the load centers).
6. Even though the inverters support Time of Use (ToU), I have a symmetric net metering agreement and therefore I don’t see a need to use ToU, but I am still interested in how that parameter affects energy sourcing within the inverter.

Based on those assumptions, here are my additional questions posed in five scenarios:

First scenario:

1. Assume grid power is available, and we are generating PV.
2. Inverters will use up to 150 A of available PV to offset load demand
3. Inverters will supplement with grid power to meet any demand over 150 A.
4. Now what if ToU is enabled and the inverter is told to use battery power and not grid.
5. In this case the inverters will check to see if the ESS capacity is above the low state of charge set in the inverters.
6. If ESS is above low SoC, then the inverters will provide battery power up to the combined 150 A limit of PV and batteries (assuming there truly is a 50 A limit on what can provided from a single inverter).
7. Here is the edge-case question, assume things are reasonably correct to this point, what happens if the load demand is in excess of 150 A? Remember in this scenario grid power is available, we have just told the Sol-Ark inverters to use battery power. Will the Sol-Arks go ahead and satisfy the residual demand using grid power, even though it was told to use batteries?
8. What would happen if the inverters did not use grid power (and I didn’t shed any loads), would the inverters shut down?

Second scenario:

1. This is just a minor variation on the first scenario.
2. We have grid power, and there isn’t sufficient PV to offset load demand
3. In this case the ESS is below the low SoC
4. I assume inverters will respect the low SoC and not try to pull power from the ESS
5. Which means that the inverters would have to use grid power or shut down (assuming I didn’t shed loads).
6. What will the Sol-Arks do in this case?

Third scenario:

1. Grid power is available, some PV is available, but not enough to meet demand load.
2. Grid power will satisfy whatever load power needs that are not met by PV
3. Here is my question, will the grid power also be used to recharge the ESS?
4. And, if so, to what state of charge will the grid power be used to achieve? 100% charge? Charge to the high SoC set in the inverter (say 80%)?, BTW I was assuming high SoC only applied when the generator was running, is that correct?

Fourth scenario:

1. The 150 A (from PV power) available for load satisfaction is more than needed for the current loads
2. The inverters will use the residual PV power to charge the batteries.
3. Related to scenario 3, what is the target charge level of the batteries when using PV as the source, is it 100% or is it high SoC?

Fifth scenario:

1. Grid is down, not enough PV to satisfy load demand, batteries are being used.
2. ESS energy capacity crosses low SoC parameter.
3. Inverters turn on generator
4. Generator powers loads up to capacity of the generator
5. Any excess generator power is used to recharge the ESS
6. When ESS reaches high SoC, generator is turned off
7. Is this a proper description?

Sorry again for the long posting. I look forward to your thoughts!

 

[DougfromdaUP]

Here are my assumptions about my configuration (which might be wrong):

1. Each Sol-Ark 15 can supply up to a max of 62.5A of output from the sum of the AC-coupled solar (from the two existing Fronius inverters) and the new DC-coupled ground-mounted system I will install.

No. Each inverter can invert 15,000 watts continuously with pv and battery, or 12,000 continuously with battery alone. It doesn't have to invert the AC coupled input. That is directly passed through. It can handle 19kW of AC coupled I believe.


2.One of the people associated with my project said it would be only safe to assume we will get 50A of PV output per inverter.
People say lots of things that don't make sense. You'll get the full rated current and wattage if you have enough panels and batteries. Lots more for short periods. Look at the spec sheet.




3. 50 A is the maximum continuous battery output from a single inverter.
No. 12,000 real watts is the maximum continuous battery only output. 50 amps at 240v if you have a perfect power factor.


4.I have been told that the PV output and the battery output are not additive (so, if we can get a max of 50 A PV and 50 A batteries, that does NOT mean we can get 100 A of combined power from the inverter). So, the max continuous inverted power output we would ever get from a single inverted is 50 Amps. Is this a correct assumption?

No. See answer to 1

 

[zanydroid]

No. Each inverter can invert 15,000 watts continuously with pv and battery, or 12,000 continuously with battery alone. It doesn't have to invert the AC coupled input. That is directly passed through. It can handle 19kW of AC coupled I believe.

I don’t think you can really rely on 15kW though since that is dependent on solar conditions. If on grid then sure the extra solar might be useful for offsetting demand charges for burst loads during the day.

Usually AC coupled is limited by the storage charger capacity. I have heard SolArk wants DC couple size to be comparable to AC coupled size but I am not sure what quirk of their design causes this.

People say lots of things that don't make sense. You'll get the full rated current and wattage if you have enough panels and batteries. Lots more for short periods. Look at the spec sheet.

Hard agree on this. You just need to overpanel enough, within reason of the MPPT limit. In the summer you can easily max out MPPT.

The power architecture of these AIOs is perhaps a little confusing to people. There are multiple places where there could be bottlenecks, and the bottlenecks are different size. In this case the inverter/AC charger has different size bottleneck from the MPPT to DC charger path. Going from battery to AC inverter as two bottlenecks — DC step up converter and inverter. While going from MPPT goes through a different up/down converter (could be internal to the MPPT, could be after it) , followed by shared inverter. You need a better electrician or EE to grok this (and it’s not clear they really need to know the edge case quirks of individual equipment to this extent to design a good system).

 

[6Bridges]

Here are my assumptions about my configuration (which might be wrong):

1. Each Sol-Ark 15 can supply up to a max of 62.5A of output from the sum of the AC-coupled solar (from the two existing Fronius inverters) and the new DC-coupled ground-mounted system I will install.

No. Each inverter can invert 15,000 watts continuously with pv and battery, or 12,000 continuously with battery alone. It doesn't have to invert the AC coupled input. That is directly passed through. It can handle 19kW of AC coupled I believe.


2.One of the people associated with my project said it would be only safe to assume we will get 50A of PV output per inverter.
People say lots of things that don't make sense. You'll get the full rated current and wattage if you have enough panels and batteries. Lots more for short periods. Look at the spec sheet.




3. 50 A is the maximum continuous battery output from a single inverter.
No. 12,000 real watts is the maximum continuous battery only output. 50 amps at 240v if you have a perfect power factor.


4.I have been told that the PV output and the battery output are not additive (so, if we can get a max of 50 A PV and 50 A batteries, that does NOT mean we can get 100 A of combined power from the inverter). So, the max continuous inverted power output we would ever get from a single inverted is 50 Amps. Is this a correct assumption?

No. See answer to 1

Thank you for your responses, they are very helpful.

It makes sense that the AC coupled solar from my two Fronius inverters would pass right through the Sol-Arks.

You made it clear that the 15000 W and 12000 W limits are on inverted power, not passed-through AC power.

So, if I were able to max at my PV generation on both my existing AC-coupled array and the new DC-coupled array, I would have available (combine from the three Sol-Ark inverters) at my load centers: 14200 W (from the AC coupled) + 16000 W (from the DC) or 30200 W, which if I could get pure 240V would be ~125 A.

Of course, I would also have power available from the ESS. Since each inverter can invert a maximum of 15000 W of combined DC-coupled PV and DC Battery, it would appear that the max theoretical power I would have available from AC-coupled solar, DC-coupled solar, and batteries would be 14200 W + 45000 W = 57200 W, which at 240V would be 238 A.

Does this seem like a correct derivation?

Thanks!

 

[zanydroid]

It makes sense that the AC coupled solar from my two Fronius inverters would pass right through the Sol-Arks.

So let's be clear, there are two levels/concepts of AC coupling here.

What a lot of forum regulars call AC coupling is the notion of allowing something like a Sol-Ark to grid form for the Fronius, when the grid goes down. This is more complex form, and it is the kind that leads to the sizing ratios (EG AC Coupled solar <= AC charger size on the grid former).

Another flavor is the notion of pass-through as you said. Ideally the AC coupled batteries will seek to absorb the excess power instead of letting it flow out to the grid. There is some equipment (at present not super popular on this forum) that work this way.

AC coupling in general is more stable when there is a grid that can help mediate various kinds of hiccups. There is more complexity/more things that can go wrong when off grid.

So, if I were able to max at my PV generation...

I didn't go over this math in detail but it sounds rather overly optimistic. You are adding output from grid-tie inverters to your AC output from batteries & ESS. When you can only count the AC output from the batteries/ESS. Because the GTI might disappear at anytime. As well, you probably are not aware that AIO like SolArk typically use the same set of transistors/hardware for both AC charging and discharging. If a GTI's output stays constant but a load drops off, the AIO will have to ramp up their charger to maintain balance in the system. In some topologies reversing from charger to AC actually has some latency (milliseconds?).

Solar -> MPPT -> "DC battery" is pretty well behaved. It is easy to stop and start. It behaves in a "pull" fashion. The more you are able to achieve with a stack of SolArks (and if they had compatible MPPTs, which I don't think they do) the more simple/stable things will be.

Solar -> AC -> "AC battery" requires satisfying numerous constraints to keep the flow from the solar going. Grid tie inverters behave in somewhat of a "push" fashion (but not strictly push, it's complicated). If you are on-grid, the grid will keep the AC up all the time. Great. If you are off grid however, the "AC battery" has to form a grid that the AC-coupled solar is happy with. And if there are things like, battery not able to deliver power to inverter (learned of a new case just yesterday involving bidirectional DC DC converters inside certain types of batteries), the grid may be interrupted and the inverters will go offline. Another case is if the battery is not able to absorb the power, in this case the AC coupled solar has a high probability of shutting down, and there is a low probability that the excess power will cause a problem. In many AC coupled systems, including SolArk, the AC coupled stuff that is meant to stay up when off-grid, goes into the system via a relay. If there is an unexpected power imbalance the relay opens, just in case.

(FWIW I'm also a software engineering veteran that just recently got into power engineering for fun)