Almost seems like its better NOT to have inverter communication?

[robby]

I keep mine separate, with no communication.

If five devices all communicate with the same central "hub", then in my mind, I have no redundancy structures. If all five devices are separately monitoring voltage, current, etc, then I would consider that to a level of monitoring redundancy.

The Central Hub or Inverter is not controlling the Battery in terms of shutting it down.
I know a lot of Inverters don't seem to have Max Current Discharge feature but others like Sol-Ark do and they gather this information automatically from the batteries if they are supported by Sol-Arks Comm system and the data is confirmed constantly. If you batteries are not closed loop supported then you have to manually stick in the Packs total Rating and hope for the best.

In your 5 battery scenario if you are 95% maxing out the five batteries and one happens to fail (Blown fuse etc) then the other four will surely fail within a second or two with either Blown fuses or a blown BMS.

In a proper closed loop system the minute one battery fails the Inverter immediately knows this and will limit it's current draw to the Max allowable from the other 4 so that they are never at 100% and being destroyed. This may result in the Inverter drawing more grid power or in off grid system the Inverter starting the Generator and one minute later connecting the load again to see if the Inverter can now handle it. Worst case is that it just trips out the Inverter and it will keep retrying every minute to see if the load is manageable with just 4 batteries.




This is just one example of closed loop in action.

 

[timselectric]

If you are drawing 95% of battery discharge limits. (No matter what size your battery is) you have a poorly designed system. lol

 

[robby]

If you are drawing 95% of battery discharge limits. (No matter what size your battery is) you have a poorly designed system. lol

Yep but there are a lot of people on this forum with just two 5K packs of batteries running their whole house at night.
That is how I started for the first month of operation and closed loop was my main protection, also it was required by Fortress Power. Anything less than three eFlex batteries connected to a Sol-Ark 12K requires a closed loop connection or your warranty is void.

If one battery went out of commission at night and the load was high it would have blown the internal Class T Fuse in the other but with closed loop the Inverter will throttle back battery usage and start pulling from the Grid.

 

[toms]

If you are drawing 95% of battery discharge limits. (No matter what size your battery is) you have a poorly designed system. lol

That’s an odd way of looking at things. I design my systems so that my maximum current draw is 100% of my maximum allowable. (generally 0.5C). If i have large inductive load components i add supercaps to minimise battery size.

The batteries are the largest cost component of the system, it makes sense to minimise their size.

How do you specify your battery size?

My objective is generally to minimise total cost per kwh over the life of the system. Oversized battery banks do not achieve this.

 

[timselectric]

How do you specify your battery size?

3 days of bad weather.

I design my systems so that my maximum current draw is 100% of my maximum allowable. (generally 0.5C)

That's 2 hours of backup. Why even bother buying any batteries?
Unless it's some sort of special circumstances system.

 

[robby]

Yes you really want as low a C rate as you can afford.
The problem is the price.

 

[toms]

3 days of bad weather.

That's 2 hours of backup. Why even bother buying any batteries?
Unless it's some sort of special circumstances system.

Maximum current draw is calculated independently from storage capacity.

If i have a potential maximum current draw of 10kw (approx 200amps @ 48V), for a maximum of 0.5C discharge i need a 400ah bank. In most installs i do, the maximum instantaneous current draw dictates the battery size, the storage capacity is then generally enough. In many cases the storage is over what is required, then supercaps can reduce battery size if there are inductive loads forming most of the peak loads.

As discussed, using integrated systems allows the battery to be easily protected from over current if multiple batteries are in parallel and one goes offline.

(ie two 200ah batteries in parallel for a 200 amp load - if one battery goes offline the system needs to limit current to 100amps)

For me the main advantages of integrated systems are:

- taper current while cell balancing
- limit discharge/charge depending on temperature
- switch additional loads based on SOC.
- ensure multiple charge sources are working together
- single diagnostic / data logging point
- less system components
- by far the cheapest system to put together (may be situation dependent- this is for Australia)

To build a system of non-communicating components that will achieve the same things is far more complex.

If your system doesn’t accurately control charge current relative to temperature and SOC, and you are relying on charge controller setpoints be prepared for short cell life.

 

[upnorthandpersonal]

If i have a potential maximum current draw of 10kw (approx 200amps @ 48V), for a maximum of 0.5C discharge i need a 400ah bank. In most installs i do, the maximum instantaneous current draw dictates the battery size, the storage capacity is then generally enough.

Keep in mind that while this is fine in e.g. Australia where 'winter' isn't a big deal, in other places in the world, battery size in off-grid situations has to be related to power generation at worst times. In those cases, battery size is driven by the amount of days of autonomy you need and thus daily consumption patterns, not max draw.

As for communication between all parts, I have been doing that with my equipment where I have a central hub (a Pi) which not only logs the data of all the attached hardware (and graphs it with Grafana), it can also be used to control loads, power on water heating when the battery is full, etc. While this is nice to have, it's not exactly critical. The only thing I make sure of is that if the BMS cuts of the battery, the inverter and charge controllers never come back online unless I do so manually. Because of the design, the BMS shutting off the pack indicates a catastrophic failure somewhere, so nothing should move unless I say so.

 

[timselectric]

Maximum current draw is calculated independently from storage capacity.

If i have a potential maximum current draw of 10kw (approx 200amps @ 48V), for a maximum of 0.5C discharge i need a 400ah bank. In most installs i do, the maximum instantaneous current draw dictates the battery size, the storage capacity is then generally enough. In many cases the storage is over what is required, then supercaps can reduce battery size if there are inductive loads forming most of the peak loads.

As discussed, using integrated systems allows the battery to be easily protected from over current if multiple batteries are in parallel and one goes offline.

(ie two 200ah batteries in parallel for a 200 amp load - if one battery goes offline the system needs to limit current to 100amps)

For me the main advantages of integrated systems are:

- taper current while cell balancing
- limit discharge/charge depending on temperature
- switch additional loads based on SOC.
- ensure multiple charge sources are working together
- single diagnostic / data logging point
- less system components
- by far the cheapest system to put together (may be situation dependent- this is for Australia)

To build a system of non-communicating components that will achieve the same things is far more complex.

If your system doesn’t accurately control charge current relative to temperature and SOC, and you are relying on charge controller setpoints be prepared for short cell life.

If it works for you, that's great. Every system is designed around its particular environment and needs.

 

[cinergi]

In cases where the battery may be 300mv out of balance as it approaches full (think: large grade-B DIY battery, such as mine) - it makes a lot of sense. Even though my charge voltage is 56v (3.5vpc), I can very easily cross over 3.65v on a cell which the BMS would shut down and then I have no battery and my system shuts down. In my case, my BMS comms is saving me by ramping down the allowed charge current before it's too late. In my case this extra "thing to fail" is actually preventing failure.

Like most things, depends on the system.

 

[Sanwizard]

In cases where the battery may be 300mv out of balance as it approaches full (think: large grade-B DIY battery, such as mine) - it makes a lot of sense. Even though my charge voltage is 56v (3.5vpc), I can very easily cross over 3.65v on a cell which the BMS would shut down and then I have no battery and my system shuts down. In my case, my BMS comms is saving me by ramping down the allowed charge current before it's too late. In my case this extra "thing to fail" is actually preventing failure.

Like most things, depends on the system.

That is exactly what I worry about. Going away for a couple weeks, and coming back to find the power has been out due to a runaway cell shutting things down, or a rainy week where the BMS shuts things down due to low voltage.
I imagine the second scenario would be managed by the inverters pulling from the grid to charge the batteries though, and the BMS active balancers should prevent the first one. I also see why having independent parallel batteries makes sense, since a failure of one will not bring down the entire system.
I guess there are always trade-offs.

By the way, has anyone besides me been informed that new NEC code calls for zero exposed DC cables inside and outside?
This makes my Aims DC switch illegal, since it has external MC4 connectors. I currently live on Long Island( at least until Thursday, when I move to South Carolina!) I told the buyer he needs to remedy that issue with a new permit.
What are you guys using in the US as an emergency cutoff switch?

 

[nate_syd]

Revive an oldie.... been thinking about this myself...
I cant find much from inverter manufacturers on the benefits & why...
BUT there *could* be many, many benefits - but they come with a risk, its all software, so is the programmer going to make a good decision for you?

Safety/Misconfiguration: If there's some differences in config, the inverter could chose the safer option, i.e. you configure 200A discharge, battery says max 100A, so it sets 100A max. Same for voltages etc. So a safety net feature. Useless if you're doing things right, but maybe insurance for a business installer putting these in houses with kids that wanna fiddle.

Invalid config/Hysteresis: I think this is the name... config is a bit funky... turn off at 51v & on at 51.5v. You're discharging & it hits 51v & turns off. Then the cells go back to 51.7 & it turns on. Then it discharges & turns off... on/off... on/off... on/off...

Cell inbalance during charge/discharge - if 1 cell is about to go high/low voltage, the inverter could throttle the charge/discharge so that it doesnt trip the BMS. So adjust the max power given the conditions.

BMS Mosfets: if the config is a little out between the Inverter/BMS (similar to safety above), the BMS could trip a protection. You're potentially turning the mosfets on/off under load which should be fine, but it could be better to gently adjust power instead of a hard shut.

Now - the last two are great features, however you could do those pretty easily with some external software controls & adjusting the config of the inverter. It also only happens when you've got less than ideal setups.

 

[Nobodybusiness]

Lots of threads with issues with inverter comms and batteries, Almost seems like it would be better not to setup communication? Hakuna-Matata?

I have the option to do it either way.

Have done it both ways but when doing closed loop the inverter only sees my SOKs instead of both SOK and DIY.
In closed loop the inverter caters its charge to the SOKs.

I just run open loop now and set the max voltage in the inverter.

Does it matter?
Dunno.

The batteries get charged daily.

 

[curiouscarbon]

Properly integrated communication
No communication
Improperly integrated communication

descending order list of my preference.

using JBD BMS and victron inverter, i am still figuring out an optimal solution.

using a raspberry pi is the easiest "glue" for now maybe (hub)

one day it could all be RJ45 ethernet. although CANbus seems more popular now.

the devices seem to have various "dialects" and prebuilt batteries seem to be supporting more dialects over time.

choosing graceful failure modes ahead of time is a big goal for me. thanks for all the insights in this thread.

 

[Nobodybusiness]

Properly integrated communication
No communication
Improperly integrated communication

descending order list of my preference.

using JBD BMS and victron inverter, i am still figuring out an optimal solution.

using a raspberry pi is the easiest "glue" for now maybe (hub)

one day it could all be RJ45 ethernet. although CANbus seems more popular now.

the devices seem to have various "dialects" and prebuilt batteries seem to be supporting more dialects over time.

choosing graceful failure modes ahead of time is a big goal for me. thanks for all the insights in this thread.

I just use the set points in the inverter.

charge voltage
Float voltage
Low voltage
Shutdown voltage.
Restart voltage .

As long as the inverter can read the voltage and act accordingly I don’t see any real advantage other than maybe tapering current.

 

[Skypower]

I’m quite satisfied with my open loop system. I don’t believe I’d be able to do this without the active balancers that I’ve always had in one configuration or another. The time at bulk, absorb, full or whatever you want to call it is not configurable in my AIO, it has its own algorithm which seems to do some time adaptation, however it is undoubtedly too short for a passive balancer. But I look at it as not a bad thing because of the active balancers the batteries spend only a fraction of the time at a high state of charge, hence much less stress. The inverter doesn’t have to wait for the last slow to passive balance battery to say “hey I’m done with my thing up here, let’s all float”.

 

[timselectric]

Current automatically tappers, as the batteries reach the set voltage. Closed loop, offers me no benefits.
If you have low quality batteries/cells. Or don't know how to setup your charging/discharging profile. Closed loop, can do it for you.
But I prefer that my system does what I want, when I want.

 

[cinergi]

Current automatically tappers, as the batteries reach the set voltage. Closed loop, offers me no benefits.
If you have low quality batteries/cells. Or don't know how to setup your charging/discharging profile. Closed loop, can do it for you.
But I prefer that my system does what I want, when I want.

This is exactly where closed loop shines and is helpful in my system. As I approach 56v (3.5vpc), I usually have 150+ mv of difference and the closed-loop comms helps back off the current so that I don't trip the cell high-volt and disconnect my battery.

 

[webbbn]

IMO, the charger should be one unit and not require "closed loop" to complete it. Either the batteries should be able to be charged by just applying a reasonable voltage/current, like pretty much every battery operated devices works these days, or the charger in the inverter should have all the charging smarts, like the chargers used in most (older?) RC drones and such. Plug in the battery and balance leads, tell the inverter/charger what type of battery it is and how many cells in parallel, and it does the rest. Those have been around for close to 20 years...

The former is obviously preferable, and will likely happen soon as BMS technology morphs into a complete, intelligent charger, balancer, conditioner.

IMO, arguing that you don't need/want closed loop is similar to arguing that you want to go back to the old carbureted engines of the '70s, which were great when they were tuned properly, but I much prefer just letting the car figure out what it needs to do to run optimally.

 

[toms]

Too many people gloss over the main advantage of closed loop which is the ability to taper current based on temperature and SOC.

They are generally the same people that are satisfied to get ten years lifespan out of a battery that would last fifteen years treated properly.