Tuning charge parameters for parallel battery banks

An area for discussing the OutBack FLEXnet DC product.

Moderators: OutBack, OutBack Moderator Team

Post Reply
TCabin76
Forum Whiz
Posts: 38
Joined: Sun Jun 21, 2020 12:36 pm
My RE system: Outback Off-Grid system: 8 - Kilovault 12V/180Ah AGM batteries (2 parallel banks of 4 series batteries for 48V / 360Ah); Kilovault 48 V battery balances - 1 per parallel bank; 3200W PV array (12x270W panels); 2- VFX3648’s (split phase master & OB slave L2) ; 1-FM 80; 1-FNDC; backup genny. MATE3.

Tuning charge parameters for parallel battery banks

Post by TCabin76 » Sun Jun 21, 2020 2:02 pm

We are configuring 2 banks of 4ea 12V/180c20 AGM batteries, configured in parallel to give 48V and approx 360Ah. I’ve been watching the data like a new father and the system seems to be going through the proper cycles: Bulk/Absorb/Float, in what appears to be a reasonable manner, and showing fully charged at the end of the day, based on voltage and SOC. We also see voltages in the range of 51.5V in the morning after 8 hrs of “silence” / rest - no overnight input from genny, and very little load. However, the “Days Since Parms Met” flag is not being reset in the FNDC.

I’ve read a fair amount of discussion on the procedures to tune the system and meet that condition (all parameters met), all of which suggest calculations based the manufacturer’s specs for single batteries. And there is in fact an OB app note for our specific battery model (Kilovault AGM) with suggested improvements to the “stock” specs for better performance, but it is based on one bank, 4-12V batteries in series to get 48V. We have 2 such banks. It seems to me that the Charge Controller looks at the battery as one monolithic whole, but it is in fact 2 separate circuits in parallel. How should one modify the recommended parameters when there are 2 banks, each of which can have it’s own current circuit?

Tried to keep this concise, please let me know if I’ve omitted and key data.

Thanks loads!

RM

SandyP
Forum Guru
Posts: 404
Joined: Tue Jun 21, 2011 4:44 am
My RE system: 8 x 190W 24V Suntech panels (4 strings of 2) July 2011
4 x 325W 24V Suntech panels (2 strings of 2) added Sept2018
Mate2
Outback FM60 MPPT (max output lowered to 55amps)
12 x 2V Hoppecke GEL 660 Ah C100 - 24V System
Outback VFX3024 Inverter/Charger
Victron BMV-602s
Honda 5.5kW Genset
Location: Victoria, Australia

Re: Tuning charge parameters for parallel battery banks

Post by SandyP » Mon Jun 22, 2020 4:37 pm

Can you let us know what parameters you have input for the charge controller and the FNDC?

raysun
Forum Emperor
Posts: 2309
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower One: FX3048T, (2) FM80, MATE3s, FlexNetDC
Outback 200NC batteries (8 @ 48v)
Outback IBR3 battery enclosure
Suniva 330 watt panels (12 - 6 strings of 2 in series)
Hyundai 355 watt panels (6 - 3 strings of 2 in series)
Honda EU7000is gas fuel generator

Re: Tuning charge parameters for parallel battery banks

Post by raysun » Mon Jun 22, 2020 5:29 pm

As far as charging and discharging, a battery made of parallel strings of monoblocks is, in essence, a single circuit.

While there can be variances between strings that cause differences in string currents (an undesirable condition) the battery as a whole is a single circuit. Unless the parallel strings are electrically separated and charged separately, there is no practical way to treat them as separate circuits.

Every so often, some company or other will offer some sort of battery balancing gizmo, but TMK, none have shown to be effective. Other schemes have one rotating monoblocks as if they are tires, but that hasn't shown to be effective.

The Kilovault PLC AGM looks very much like the 200NC battery that Outback used to offer (and that I use.)

Do you have specs for the battery in your system?

TCabin76
Forum Whiz
Posts: 38
Joined: Sun Jun 21, 2020 12:36 pm
My RE system: Outback Off-Grid system: 8 - Kilovault 12V/180Ah AGM batteries (2 parallel banks of 4 series batteries for 48V / 360Ah); Kilovault 48 V battery balances - 1 per parallel bank; 3200W PV array (12x270W panels); 2- VFX3648’s (split phase master & OB slave L2) ; 1-FM 80; 1-FNDC; backup genny. MATE3.

Re: Tuning charge parameters for parallel battery banks

Post by TCabin76 » Mon Jun 22, 2020 8:42 pm

Thank you sandyp and ray sun for your replies.

Here are the settings from the charge controller and battery manager:

FM80-150VDC

Absorb Voltage 57 VDC
Absorb Time 1 Hours
Absorb End Amps 0 ADC
Re-Bulk Voltage 51 VDC
Float Voltage 55 VDC
Current Limit 80 ADC
EQ Voltage 57 VDC
EQ Time 1 Hours
Auto EQ Interval 0 Days


FLEXnet DC:

Battery Ah 360 Ah
Charged Voltage 56.3 VDC
Charged Time 3 Minutes
Charged Return Amps 10.8 Amps
Charge Factor 85 Percent


Here are the specs per Kilovault 2100 spec sheet:

Rated Energy Capacity Watt Hours (Wh) 2160 Wh
Rated Voltage 12V
Nominal Amp-Hour Capacity (C20 Hr Discharge) 180 Ah
Float Charging Voltage 13.7V
Bulk/Absorption Voltage 14.1 to 14.4V
Equalize Voltage 14.1V
Maximum Equalize time 24 Hours
Battery Efficiency 87%
Recharge / Re-bulk Voltage for 50% DoD 14.4V at 77°F (25°C)
Charge End Amps 0.32A
Max Discharge Current 500A (5s)
Max Charge Current 140A (100A recommended)
Max Configuration Up to 4, 48V strings in parallel
Nominal Operating Temperature Range 77° +/- 5°F (25° +/- 3°C)
Operating Temperature Range Discharging: -40 to 131°F (-40 to 55°C) Charging: -4 to 113°F (-20 to 45°C) Storage: -4 to 122°F (-20 to 50°C)
Cycle Life 3000 (50% DoD)


New comments:

The configuration is evolving from the initial install / setup which used the wizard. I am now going back and trying to better learn the behavior of these batteries in the system with the intention of tuning things as needed and knowing what to watch for as they age.

If it is not obvious, I am no expert so there is a learning curve. Along the curve I came to the “days since parameters met” item in the FNDC readings and found it was not resetting. In trying to trouble shoot that I’ve found information that suggests tweaking some of the current settings, and that was what lead me to my initial question.

As a somewhat random example: the Charged Return Amps setting in my FNDC is currently 10.8 Amps. As I understand it that is one of the factors that determine when the “days since parameters set” flag will reset. As I tried to get my head around that, it occurred to me that the FNDC is looking at the whole battery bank, watching for the 10.8 A condition to be met. It seems to me that that condition could be met by the combined current of the 2 parallel sub-banks, but neither would actually be meeting that condition on it’s own. In other words, if the system is looking for that specific current draw to make a decision regarding where the batteries are in their charge cycle, wouldn’t that be erroneous in terms of the individual parallel banks? E.G., In my case the FNDC sees the current draw of 10.8 A and thinks that the batteries are at a particular state when in fact, judged by current draw they are only half that.

Thanks again for any help / insight you can provide.

RM

TCabin76
Forum Whiz
Posts: 38
Joined: Sun Jun 21, 2020 12:36 pm
My RE system: Outback Off-Grid system: 8 - Kilovault 12V/180Ah AGM batteries (2 parallel banks of 4 series batteries for 48V / 360Ah); Kilovault 48 V battery balances - 1 per parallel bank; 3200W PV array (12x270W panels); 2- VFX3648’s (split phase master & OB slave L2) ; 1-FM 80; 1-FNDC; backup genny. MATE3.

Re: Tuning charge parameters for parallel battery banks

Post by TCabin76 » Mon Jun 22, 2020 8:48 pm

To raysun,

To your comment about balancing gadgets, we have 2 Kilovault 48V battery balancers installed. They are supposed to Monitor the individual batteries and keep them balanced with in their respective series bank (2 series banks of 12V batteries in parallel to get 48V at 2x Ah capacity). Are these the kinds of gadgets you are referring to/ should I be suspicious of them/ test them somehow?

Thanks, RM

raysun
Forum Emperor
Posts: 2309
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower One: FX3048T, (2) FM80, MATE3s, FlexNetDC
Outback 200NC batteries (8 @ 48v)
Outback IBR3 battery enclosure
Suniva 330 watt panels (12 - 6 strings of 2 in series)
Hyundai 355 watt panels (6 - 3 strings of 2 in series)
Honda EU7000is gas fuel generator

Re: Tuning charge parameters for parallel battery banks

Post by raysun » Mon Jun 22, 2020 9:04 pm

I've not seen anything definitive on their efficacy.

I have seen some cautions on wiring needing to carry some fairly heavy current should the balancing unit fault.

If Kilovolt warrants them, and the battery on which they are employed, then it should be OK. I'd read the fine print on that, for sure.

raysun
Forum Emperor
Posts: 2309
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower One: FX3048T, (2) FM80, MATE3s, FlexNetDC
Outback 200NC batteries (8 @ 48v)
Outback IBR3 battery enclosure
Suniva 330 watt panels (12 - 6 strings of 2 in series)
Hyundai 355 watt panels (6 - 3 strings of 2 in series)
Honda EU7000is gas fuel generator

Re: Tuning charge parameters for parallel battery banks

Post by raysun » Tue Jun 23, 2020 12:55 am

I'm going to take a crack at some of this. Responses inline.
TCabin76 wrote:
Mon Jun 22, 2020 8:42 pm
Thank you sandyp and ray sun for your replies.

Here are the settings from the charge controller and battery manager:

FM80-150VDC

Absorb Voltage 57 VDC
Seems low. Did this come from the vendor? The default 57.6 might be more on-target.

Absorb Time 1 Hours
Also low. If I was a betting man, I'd lay my money on 2 hours.

Absorb End Amps 0 ADC
Yes.

Re-Bulk Voltage 51 VDC
Will definitely guarantee a new Bulk cycle.

Float Voltage 55 VDC
Seems a little high...

Current Limit 80 ADC
Those strings are capable of about 50A each?

EQ Voltage 57 VDC
The AGM world calls it a "freshening" charge. Many manufacturers don't recommend one. Others say to do one if the battery hasn't been fully charged for an extended interval. Should equal Absorb voltage.

EQ Time 1 Hours
Arbitrary time, probably insufficient. The FLA metric of specific gravity no longer rising in all cells is not available to AGM. The next best as End Amps declines no further for one hour.

Auto EQ Interval 0 Days
Yes, no Auto EQ.


FLEXnet DC:

Battery Ah 360 Ah

Charged Voltage 56.3 VDC

Charged Time 3 Minutes

Charged Return Amps 10.8 Amps
2% - 3% of the C20 AH capacity is typical.

Charge Factor 85 Percent
This seems very low for modern AGM. I'd think more like 95%.

I should have read below before answering above.


Here are the specs per Kilovault 2100 spec sheet:

Rated Energy Capacity Watt Hours (Wh) 2160 Wh
Rated Voltage 12V

Nominal Amp-Hour Capacity (C20 Hr Discharge) 180 Ah
360AH battery

Float Charging Voltage 13.7V
54.8V Float

Bulk/Absorption Voltage 14.1 to 14.4V
With AGM, overcharge runs the risk of losing electrolyte to off-gassing. Undercharge runs the risk of sulphation. Both permanently lose capacity. With the Absorb range of 56.4 - 57.6 they are punting IMO. The tradeoff is time to completion v.s. higher overcharge voltage. The 'sweet spot' may depend on how long the Absorb phase can run in your locale.

Equalize Voltage 14.1V
Maximum Equalize time 24 Hours
A kinder, gentler EQ charge. It would minimize the risk of off gassing. The 24 Hour EQ-athon may be prescribed for commissioning the battery or before a capacity test. Most EQ durations should be measured by target end Amps condition.

Battery Efficiency 87%
This seems really low, but they're the authority.

Recharge / Re-bulk Voltage for 50% DoD 14.4V at 77°F (25°C)
I don't get this spec at all. Rebulk is usually the voltage the battery drops to, triggering a new charge cycle.

Charge End Amps 0.32A
Seriously? That's the world's record low End Amps. Typically pegged at 2 - 3% of C20 - an End Amp of 3.6A would be more typical. Two strings = 7.2.
It would be an interesting experiment to set the lowest Absorb voltage and run the cycle indefinitely until the lowest possible End Amps is reached.

Max Discharge Current 500A (5s)

Max Charge Current 140A (100A recommended)
That's a huge charge current - even 100A. Implies exceptionally low internal resistance which is a hallmark of AGM, but not that low, IME.

Max Configuration Up to 4, 48V strings in parallel
Three strings is the typical rule of thumb.

Nominal Operating Temperature Range 77° +/- 5°F (25° +/- 3°C)
AGMs usually have a narrow temperature compensation range during charging.

Operating Temperature Range Discharging: -40 to 131°F (-40 to 55°C) Charging: -4 to 113°F (-20 to 45°C) Storage: -4 to 122°F (-20 to 50°C)
Cycle Life 3000 (50% DoD)

For me, these specs would trigger a call to tech support asking for the rationale behind the somewhat unusual settings. (Then again, when it comes to batteries, I'm easily triggered.) I'm sure there is rational explanations based on this battery's internals, and I'd be keen on hearing them.

Getting on to the question about Charged Parameters Met. IMO, it's a figure of merit for battery charging. It's best use is to help identify when charge should be terminated according to Outback algorithms. There are only three factors that apply: Charged Voltage, Charged Time, and most important Charged Return Amps. Setting of all three is arbitrary, and should be derived from the battery specs. Looking at the specs, especially End Amps, I'm not sure how Charged Parameters could ever be met, but the latter is likely due to my inexperience with this battery.

How was the Charged Return Amps of 10.8 arrived at?

New comments:

The configuration is evolving from the initial install / setup which used the wizard. I am now going back and trying to better learn the behavior of these batteries in the system with the intention of tuning things as needed and knowing what to watch for as they age.

If it is not obvious, I am no expert so there is a learning curve. Along the curve I came to the “days since parameters met” item in the FNDC readings and found it was not resetting. In trying to trouble shoot that I’ve found information that suggests tweaking some of the current settings, and that was what lead me to my initial question.

As a somewhat random example: the Charged Return Amps setting in my FNDC is currently 10.8 Amps. As I understand it that is one of the factors that determine when the “days since parameters set” flag will reset. As I tried to get my head around that, it occurred to me that the FNDC is looking at the whole battery bank, watching for the 10.8 A condition to be met. It seems to me that that condition could be met by the combined current of the 2 parallel sub-banks, but neither would actually be meeting that condition on it’s own. In other words, if the system is looking for that specific current draw to make a decision regarding where the batteries are in their charge cycle, wouldn’t that be erroneous in terms of the individual parallel banks? E.G., In my case the FNDC sees the current draw of 10.8 A and thinks that the batteries are at a particular state when in fact, judged by current draw they are only half that.

Thanks again for any help / insight you can provide.

RM

SandyP
Forum Guru
Posts: 404
Joined: Tue Jun 21, 2011 4:44 am
My RE system: 8 x 190W 24V Suntech panels (4 strings of 2) July 2011
4 x 325W 24V Suntech panels (2 strings of 2) added Sept2018
Mate2
Outback FM60 MPPT (max output lowered to 55amps)
12 x 2V Hoppecke GEL 660 Ah C100 - 24V System
Outback VFX3024 Inverter/Charger
Victron BMV-602s
Honda 5.5kW Genset
Location: Victoria, Australia

Re: Tuning charge parameters for parallel battery banks

Post by SandyP » Tue Jun 23, 2020 12:58 am

Based on the Outback integration sheet (for Radian/FXR systems) the numbers are :

Charge Controller Item
Absorb Voltage 55.6 Vdc / 0.5 hr (I'm unsure why this is lower than the battery spec sheet?)
Float Voltage 54.4 Vdc
Re-bulk Voltage 51.0 Vdc
DC Current Limit 80 Adc
Absorb End Amps 0 Adc

FN-DC Item
Battery Ah 150 Ah/string (So 300Ah for your 2xstring battery bank)
Charged Voltage 54.8 Vdc (ie the just above the Float voltage setting)
Charged Return Amps 6.0 Adc (So 12.0Adc for your 2xstring battery bank?)
Battery Charge Efficiency 96% (The 87% in the specs is very low for AGM ?typo?)

Assuming it would be similar for a VFX system, I suspect the differences in the FN-DC settings are preventing the parameters being met.

Edit: Raysun beat me to it :-)

raysun
Forum Emperor
Posts: 2309
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower One: FX3048T, (2) FM80, MATE3s, FlexNetDC
Outback 200NC batteries (8 @ 48v)
Outback IBR3 battery enclosure
Suniva 330 watt panels (12 - 6 strings of 2 in series)
Hyundai 355 watt panels (6 - 3 strings of 2 in series)
Honda EU7000is gas fuel generator

Re: Tuning charge parameters for parallel battery banks

Post by raysun » Tue Jun 23, 2020 1:03 am

SandyP wrote:
Tue Jun 23, 2020 12:58 am
Based on the Outback integration sheet (for Radian/FXR systems) the numbers are :

Charge Controller Item
Absorb Voltage 55.6 Vdc / 0.5 hr (I'm unsure why this is lower than the battery spec sheet?)
Float Voltage 54.4 Vdc
Re-bulk Voltage 51.0 Vdc
DC Current Limit 80 Adc
Absorb End Amps 0 Adc

FN-DC Item
Battery Ah 150 Ah/string (So 300Ah for your 2xstring battery bank)
Charged Voltage 54.8 Vdc (ie the just above the Float voltage setting)
Charged Return Amps 6.0 Adc (So 12.0Adc for your 2xstring battery bank?)
Battery Charge Efficiency 96% (The 87% in the specs is very low for AGM ?typo?)

Assuming it would be similar for a VFX system, I suspect the differences in the FN-DC settings are preventing the parameters being met.

Edit: Raysun beat me to it :-)
Thanks for adding this info Mate.

These specs are making my head swim (and before I just thought it was the wine.)

I'm calling a Ronald Regan here: Trust, but Verify.

SandyP
Forum Guru
Posts: 404
Joined: Tue Jun 21, 2011 4:44 am
My RE system: 8 x 190W 24V Suntech panels (4 strings of 2) July 2011
4 x 325W 24V Suntech panels (2 strings of 2) added Sept2018
Mate2
Outback FM60 MPPT (max output lowered to 55amps)
12 x 2V Hoppecke GEL 660 Ah C100 - 24V System
Outback VFX3024 Inverter/Charger
Victron BMV-602s
Honda 5.5kW Genset
Location: Victoria, Australia

Re: Tuning charge parameters for parallel battery banks

Post by SandyP » Tue Jun 23, 2020 1:25 am

raysun wrote:
Tue Jun 23, 2020 1:03 am
../..
I'm calling a Ronald Regan here: Trust, but Verify.
Strange why the Outback and Kilovault numbers are very different.

@TCabin76 : In summary for the parameters to be met the FN-DC wants to have seen :

1) Around 5% more amps put back into the battery than you have taken out since the last full charge
2) The Absorb charge phase completed and the battery bank now in the Float phase.
3) The charging amps while "Floating" being below 12A dc.

raysun
Forum Emperor
Posts: 2309
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower One: FX3048T, (2) FM80, MATE3s, FlexNetDC
Outback 200NC batteries (8 @ 48v)
Outback IBR3 battery enclosure
Suniva 330 watt panels (12 - 6 strings of 2 in series)
Hyundai 355 watt panels (6 - 3 strings of 2 in series)
Honda EU7000is gas fuel generator

Re: Tuning charge parameters for parallel battery banks

Post by raysun » Tue Jun 23, 2020 4:44 am

raysun wrote:
Tue Jun 23, 2020 1:03 am
SandyP wrote:
Tue Jun 23, 2020 12:58 am
Based on the Outback integration sheet (for Radian/FXR systems) the numbers are :

Charge Controller Item
Absorb Voltage 55.6 Vdc / 0.5 hr (I'm unsure why this is lower than the battery spec sheet?)
Float Voltage 54.4 Vdc
Re-bulk Voltage 51.0 Vdc
DC Current Limit 80 Adc
Absorb End Amps 0 Adc

FN-DC Item
Battery Ah 150 Ah/string (So 300Ah for your 2xstring battery bank)
Charged Voltage 54.8 Vdc (ie the just above the Float voltage setting)
Charged Return Amps 6.0 Adc (So 12.0Adc for your 2xstring battery bank?)
Battery Charge Efficiency 96% (The 87% in the specs is very low for AGM ?typo?)

Assuming it would be similar for a VFX system, I suspect the differences in the FN-DC settings are preventing the parameters being met.

Edit: Raysun beat me to it :-)
Thanks for adding this info Mate.

These specs are making my head swim (and before I just thought it was the wine.)

I'm calling a Ronald Regan here: Trust, but Verify.
Wait.... Is this the same battery? I don't have the integration guide, but this looks like maybe its Lithium instead of VRLA?

provo
Forum Guru
Posts: 403
Joined: Thu Jul 27, 2006 4:34 pm
My RE system: Sixteen Evergreen EC-120
(4 strings, total 1920W)
Eight Rolls S-550 (2 strings, total ~800Ah @ 24V)
One FM80
One VFXR3524A
Hub 10
Mate3s
FNDC and Trimetric
Location: Sierra foothills

Re: Tuning charge parameters for parallel battery banks

Post by provo » Tue Jun 23, 2020 9:51 am

TCabin76 wrote:
Sun Jun 21, 2020 2:02 pm
...We also see voltages in the range of 51.5V in the morning after 8 hrs of “silence” / rest - no overnight input from genny, and very little load. However, the “Days Since Parms Met” flag is not being reset in the FNDC.

I’ve read a fair amount of discussion on the procedures to tune the system and meet that condition (all parameters met), all of which suggest calculations based the manufacturer’s specs for single batteries. .....

There's a certain satisfaction in seeing the "Days Since Parms Met" go to zero every day ... when I got my FNDC I spent a few days tweaking the End Amps till it dependably flipped every day, but I have FLA's so I knew from SG's that they were really getting charged. I just use a timed absorb.

Anyway, you probably know that meeting Charge Parameters is not necessary unless you're using End Amps to terminate the absorb. If instead your absorb phase ends with the timer, it doesn't matter -- the accumulated Ah's go to zero anyway when absorb ends, and the SOC also goes to 100%, even if Charged Parameters are never met.

TCabin76
Forum Whiz
Posts: 38
Joined: Sun Jun 21, 2020 12:36 pm
My RE system: Outback Off-Grid system: 8 - Kilovault 12V/180Ah AGM batteries (2 parallel banks of 4 series batteries for 48V / 360Ah); Kilovault 48 V battery balances - 1 per parallel bank; 3200W PV array (12x270W panels); 2- VFX3648’s (split phase master & OB slave L2) ; 1-FM 80; 1-FNDC; backup genny. MATE3.

Re: Tuning charge parameters for parallel battery banks

Post by TCabin76 » Tue Jun 23, 2020 9:01 pm

@raysun - thank you for taking the time to add you comments to my spec sheet. It will take me a bit to digest what you have said there. As Arnold said: I’ll be back.

TCabin76
Forum Whiz
Posts: 38
Joined: Sun Jun 21, 2020 12:36 pm
My RE system: Outback Off-Grid system: 8 - Kilovault 12V/180Ah AGM batteries (2 parallel banks of 4 series batteries for 48V / 360Ah); Kilovault 48 V battery balances - 1 per parallel bank; 3200W PV array (12x270W panels); 2- VFX3648’s (split phase master & OB slave L2) ; 1-FM 80; 1-FNDC; backup genny. MATE3.

Re: Tuning charge parameters for parallel battery banks

Post by TCabin76 » Tue Jun 23, 2020 9:12 pm

@sandyp: Re that integration guide that you reference, I think it is the same app note that I looked at:

http://www.outbackpower.com/downloads/d ... p_note.pdf

It got me at first too, until I noticed it referenced an HLX battery which is lithium. - doh!

TCabin76
Forum Whiz
Posts: 38
Joined: Sun Jun 21, 2020 12:36 pm
My RE system: Outback Off-Grid system: 8 - Kilovault 12V/180Ah AGM batteries (2 parallel banks of 4 series batteries for 48V / 360Ah); Kilovault 48 V battery balances - 1 per parallel bank; 3200W PV array (12x270W panels); 2- VFX3648’s (split phase master & OB slave L2) ; 1-FM 80; 1-FNDC; backup genny. MATE3.

Re: Tuning charge parameters for parallel battery banks

Post by TCabin76 » Tue Jun 23, 2020 9:33 pm

@ Provo
There's a certain satisfaction in seeing the "Days Since Parms Met" go to zero every day ... when I got my FNDC I spent a few days tweaking the End Amps till it dependably flipped every day, but I have FLA's so I knew from SG's that they were really getting charged. I just use a timed absorb.
I find the timed absorb confusing. How can you know how long of an absorb phase will be required to top things off? I have an off grid system and unless I kick in the genny I’m at the mercy of the solar gods as to how much power I get from my PV’s. Seems like I could have a sunny morning and just about the time absorb kicks in the clouds roll in. In that circumstance wouldn’t I want to be watching the Amps to know where the batteries were in the cycle? Maybe I just don’t really understand how that absorb timer works...
Anyway, you probably know that meeting Charge Parameters is not necessary unless you're using End Amps to terminate the absorb. If instead your absorb phase ends with the timer, it doesn't matter -- the accumulated Ah's go to zero anyway when absorb ends, and the SOC also goes to 100%, even if Charged Parameters are never met..
I’m not following your comment about accumulated Ah’s going to zero. I’m still struggling with all the nomenclature, is there another term or concept associated with accumulated Ah’s? Where is that listed.

Thanks, I’m getting smarter, but very slowly... :grin:

raysun
Forum Emperor
Posts: 2309
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower One: FX3048T, (2) FM80, MATE3s, FlexNetDC
Outback 200NC batteries (8 @ 48v)
Outback IBR3 battery enclosure
Suniva 330 watt panels (12 - 6 strings of 2 in series)
Hyundai 355 watt panels (6 - 3 strings of 2 in series)
Honda EU7000is gas fuel generator

Re: Tuning charge parameters for parallel battery banks

Post by raysun » Tue Jun 23, 2020 11:50 pm

TCabin76 wrote:
Tue Jun 23, 2020 9:33 pm
@ Provo
There's a certain satisfaction in seeing the "Days Since Parms Met" go to zero every day ... when I got my FNDC I spent a few days tweaking the End Amps till it dependably flipped every day, but I have FLA's so I knew from SG's that they were really getting charged. I just use a timed absorb.
I find the timed absorb confusing. How can you know how long of an absorb phase will be required to top things off? I have an off grid system and unless I kick in the genny I’m at the mercy of the solar gods as to how much power I get from my PV’s. Seems like I could have a sunny morning and just about the time absorb kicks in the clouds roll in. In that circumstance wouldn’t I want to be watching the Amps to know where the batteries were in the cycle? Maybe I just don’t really understand how that absorb timer works...

The Absorb process is at once simple and complicated. Trust me, we all share a degree of your confusion. ;)

In its fundamental concept, Absorb is a controlled overcharge designed to "top off" the battery without causing undue "wear and tear" on the internals. It is a constant voltage charging phase in which the charge current steadily declines as the battery internal resistance rises. So how long should Absorb be executed? How is "topped off" determined? For flooded lead acid batteries, the end state can be determined directly by measuring specific gravity, though nobody but the ultimate OCD system operator is going to do that every charge cycle. For sealed batteries, measuring SG is not possible, so another method is needed.

Its generally agreed that when the Absorb charging current falls within 2% - 3% of the C20 Amp Hour (discharge) capacity, the battery is fully charged. For a 100AH @ C20 battery, the End Amps would be between 2A and 3A. In charging systems that don't have a way to measure End Amps (aka Charged Return Amps) a timer is used to estimate when the Absorb phase is complete. The Absorb time should be stated by the battery manufacturer in their charging specifications.

That's the simple example of Absorb. Now, the real world and how we attempt to deal with it.

Absorb is not a phase that can be optimally fixed in time. While the Absorb battery voltage is generally reached when the battery is 80% - 90% full, there are several factors that can impact that. Running a static Absorb time can run the risk of under charging the battery, or overcharging the battery. Sophisticated charging systems use a more complex algorithm to decide the amount of time the Absorb phase should execute.

After a static timer, the next simplest method is to execute Absorb until the End Amps value is reached, then terminate the Absorb cycle. Using a "coulomb counting" battery monitor like the FNDC makes such a method possible. It can track net charge flowing into the battery and signal the Outback chargers when to end the Absorb cycle. The Auto Charge Termination Control in the Mate works in conjunction with the FNDC to terminate charging when the Charged Return Amps are reached for the Charged Time duration. (If the Absorb Time is reached before the Charged Return Amps are reached, the Absorb phase terminates without reaching the Amps target.)

That particular set of parameters, coincidentally, factors into setting the Charged Parameters Met flag, and blinking the green LED on the FNDC. Each time, and every time? No. There's more twists to this tale.

Batteries are less like a Swiss watch, and more like your significant other getting ready for a big night out. It's ready when it's ready, and tracking the variables that go into it are beyond the ken of mortal man.

Outback attempts to accommodate the variables in a variety of ways. The FlexMax series of charge controllers have an additional layer of "smarts" to manage charging. Since the FM CCs can operate autonomously, without the aid of the Mate/FNDC, they have an additional algorithm that manages Absorb timing. Since the FM CCs CCs can't determine net charge current into the battery (only gross charge current, which is the sum of charge current and load current), End Amps becomes a tricky factor to use for charge termination. Set too low, the End Amps may never be met. Set too high, Absorb can terminate prematurely. We almost always set the FM CC End Amps to 0 to disable it.

The charge controllers are reduced to using voltage as the metric for determining how long the Absorb cycle should be. How does it work? The first factor is the Absorb timer. If nothing else influences the charge cycle, Absorb starts as soon as the Absorb voltage is reached, and proceeds for the amount of time set in the Absorb timer. (Note, the Absorb Time is accumulated. If the charge controller goes in and out of Absorb, due to sun or load, the timer increments only while in Absorb. If the full Absorb time is not met during a charge cycle, the "left over" time is added to the Absorb Time of the next charge cycle.)

Other things do, however, influence the Absorb Time. In an effort to not run the Absorb phase overly long (a very bad thing for VRLA batteries), FM CCs track another usage factor: the REBULK voltage. REBULK is the parameter that, should the battery voltage go below its value, triggers a new charge cycle. For the FM CCs REBULK plays another, more sophisticated role. After a charge cycle is complete REBULK serves to modify the Absorb Timer. When the FM CC goes Silent, every minute the battery voltage stays above the REBULK setting, a minute is subtracted from the Absorb Time. For every minute the battery voltage is below the REBULK setting, a minute is added to the Absorb Time. (There is an additional factor, for which I cannot recall the trigger point, but if the battery voltage falls far enough below REBULK, each minute adds four minutes to the Absorb Time.)

So now, without external controls of the Mate/FNDC, the modified Absorb Time can be greater, or less than the value set in the FM CC parameters. In theory, a lightly discharged battery may have the CC run a full Absorb charge once, and never again. In practice batteries are generally discharged enough to trigger an Absorb phase.

All this long winded explanation is brain cramp inducing enough without adding the complexities of rationalizing it with Charged Parameters Met requirements, so that will be saved for another post. Absorb (pun intended) this info, and we'll dive into that pool later.
Anyway, you probably know that meeting Charge Parameters is not necessary unless you're using End Amps to terminate the absorb. If instead your absorb phase ends with the timer, it doesn't matter -- the accumulated Ah's go to zero anyway when absorb ends, and the SOC also goes to 100%, even if Charged Parameters are never met..
I’m not following your comment about accumulated Ah’s going to zero. I’m still struggling with all the nomenclature, is there another term or concept associated with accumulated Ah’s? Where is that listed.

Thanks, I’m getting smarter, but very slowly... :grin:

provo
Forum Guru
Posts: 403
Joined: Thu Jul 27, 2006 4:34 pm
My RE system: Sixteen Evergreen EC-120
(4 strings, total 1920W)
Eight Rolls S-550 (2 strings, total ~800Ah @ 24V)
One FM80
One VFXR3524A
Hub 10
Mate3s
FNDC and Trimetric
Location: Sierra foothills

Re: Tuning charge parameters for parallel battery banks

Post by provo » Wed Jun 24, 2020 5:28 am

TCabin76 wrote:
Tue Jun 23, 2020 9:33 pm

I’m not following your comment about accumulated Ah’s going to zero. I’m still struggling with all the nomenclature, is there another term or concept associated with accumulated Ah’s? Where is that listed.

Accumulated Ah's is calculated by the FNDC by counting electrons (coulombs) into the battery during charge and out of the battery during discharge. The charging Ah's are "discounted" by a settable number (the "charge factor") as compared to the discharging Ah's, to account for charging inefficiency. When the result is zero, that's when the FNDC reports 100% SOC. If you have an FNDC, the SOC is reported in many places. As far as I know, the accumulated Ah's with the charge factor correction is only shown in OpticsRE (called Corrected NET Battery toward the bottom of the Status page of the FNDC Device Map.)

Anyway, that sum of in and out can get off over time, so the FNDC stays synchronized by resetting SOC to 100 and Ah's to zero whenever absorb ends and float begins. That feature can cause problems, of course, if your absorb time is set wrong. But End Amps can cause problems too, because it includes the effect of loads.

User avatar
EA6LE-ONE
Forum Guru
Posts: 151
Joined: Tue May 05, 2020 7:51 am
My RE system: System 1:
MATE3s, 3 x Radian GS8048A,
4 x FLEXmax 100, HUB10,
2 x EnergyCell 48V 2700RE,
1 x FLEXnet DC, 4 x FLEXware ICS Plus,
54 x PANASONIC VBHN325SA17 325W

System 2:
MATE3s, 2 x VFX3648,
2 x FLEXmax 100, HUB10
1 x EnergyCell 48V 2700RE,
1 x FLEXnet DC, 2 x FLEXware ICS
21 x PANASONIC VBHN325SA17 325W
Location: Providenciales, Turks & Caicos Isl.

Re: Tuning charge parameters for parallel battery banks

Post by EA6LE-ONE » Wed Jun 24, 2020 6:32 am

provo wrote:
Wed Jun 24, 2020 5:28 am
Anyway, that sum of in and out can get off over time, so the FNDC stays synchronized by resetting SOC to 100 and Ah's to zero whenever absorb ends and float begins. That feature can cause problems, of course, if your absorb time is set wrong. But End Amps can cause problems too, because it includes the effect of loads.
If you use only one shunt from negative bar to the battery and have the negative of the inverter and CCs before the shunt on the negative bar, will not have the correct end amps?
I have mine setup like that and works very well with end amps.

provo
Forum Guru
Posts: 403
Joined: Thu Jul 27, 2006 4:34 pm
My RE system: Sixteen Evergreen EC-120
(4 strings, total 1920W)
Eight Rolls S-550 (2 strings, total ~800Ah @ 24V)
One FM80
One VFXR3524A
Hub 10
Mate3s
FNDC and Trimetric
Location: Sierra foothills

Re: Tuning charge parameters for parallel battery banks

Post by provo » Wed Jun 24, 2020 9:05 am

EA6LE-ONE wrote:
Wed Jun 24, 2020 6:32 am
provo wrote:
Wed Jun 24, 2020 5:28 am
Anyway, that sum of in and out can get off over time, so the FNDC stays synchronized by resetting SOC to 100 and Ah's to zero whenever absorb ends and float begins. That feature can cause problems, of course, if your absorb time is set wrong. But End Amps can cause problems too, because it includes the effect of loads.
If you use only one shunt from negative bar to the battery and have the negative of the inverter and CCs before the shunt on the negative bar, will not have the correct end amps?
No, that's fine. End Amps is the net current into the battery, and that's what is measured by the single shunt. I don't use End Amps, so I'm guessing, but I suppose if solar was limited and a big load came on it could pull the charging current below the End Amps setting for long enough to end the absorb prematurely.

raysun
Forum Emperor
Posts: 2309
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower One: FX3048T, (2) FM80, MATE3s, FlexNetDC
Outback 200NC batteries (8 @ 48v)
Outback IBR3 battery enclosure
Suniva 330 watt panels (12 - 6 strings of 2 in series)
Hyundai 355 watt panels (6 - 3 strings of 2 in series)
Honda EU7000is gas fuel generator

Re: Tuning charge parameters for parallel battery banks

Post by raysun » Wed Jun 24, 2020 9:32 am

provo wrote:
Wed Jun 24, 2020 9:05 am
EA6LE-ONE wrote:
Wed Jun 24, 2020 6:32 am
provo wrote:
Wed Jun 24, 2020 5:28 am
Anyway, that sum of in and out can get off over time, so the FNDC stays synchronized by resetting SOC to 100 and Ah's to zero whenever absorb ends and float begins. That feature can cause problems, of course, if your absorb time is set wrong. But End Amps can cause problems too, because it includes the effect of loads.
If you use only one shunt from negative bar to the battery and have the negative of the inverter and CCs before the shunt on the negative bar, will not have the correct end amps?
No, that's fine. End Amps is the net current into the battery, and that's what is measured by the single shunt. I don't use End Amps, so I'm guessing, but I suppose if solar was limited and a big load came on it could pull the charging current below the End Amps setting for long enough to end the absorb prematurely.
You are correct. If a single shunt is used, it can show net amps, but not differential amps.

In the former case, if the charger was capable of 200 amps, and the load demanded 199 amps, the net to the battery is 1 amp. There would be no practical way for a monitor to discern whether the net flow to the battery is due to battery impedance or the load/charger dynamics.

User avatar
EA6LE-ONE
Forum Guru
Posts: 151
Joined: Tue May 05, 2020 7:51 am
My RE system: System 1:
MATE3s, 3 x Radian GS8048A,
4 x FLEXmax 100, HUB10,
2 x EnergyCell 48V 2700RE,
1 x FLEXnet DC, 4 x FLEXware ICS Plus,
54 x PANASONIC VBHN325SA17 325W

System 2:
MATE3s, 2 x VFX3648,
2 x FLEXmax 100, HUB10
1 x EnergyCell 48V 2700RE,
1 x FLEXnet DC, 2 x FLEXware ICS
21 x PANASONIC VBHN325SA17 325W
Location: Providenciales, Turks & Caicos Isl.

Re: Tuning charge parameters for parallel battery banks

Post by EA6LE-ONE » Wed Jun 24, 2020 9:45 am

provo wrote:
Wed Jun 24, 2020 9:05 am
EA6LE-ONE wrote:
Wed Jun 24, 2020 6:32 am
provo wrote:
Wed Jun 24, 2020 5:28 am
Anyway, that sum of in and out can get off over time, so the FNDC stays synchronized by resetting SOC to 100 and Ah's to zero whenever absorb ends and float begins. That feature can cause problems, of course, if your absorb time is set wrong. But End Amps can cause problems too, because it includes the effect of loads.
If you use only one shunt from negative bar to the battery and have the negative of the inverter and CCs before the shunt on the negative bar, will not have the correct end amps?
No, that's fine. End Amps is the net current into the battery, and that's what is measured by the single shunt. I don't use End Amps, so I'm guessing, but I suppose if solar was limited and a big load came on it could pull the charging current below the End Amps setting for long enough to end the absorb prematurely.
I use 10 min time for the end amps setting, and will not end prematurely. the FNDC will try to meet all three conditions: absorb voltage (0.4v lower than CC setting), time 10min (default 1min but that can create the end too soon) and 2% for end amps ( check battery documentation)

User avatar
EA6LE-ONE
Forum Guru
Posts: 151
Joined: Tue May 05, 2020 7:51 am
My RE system: System 1:
MATE3s, 3 x Radian GS8048A,
4 x FLEXmax 100, HUB10,
2 x EnergyCell 48V 2700RE,
1 x FLEXnet DC, 4 x FLEXware ICS Plus,
54 x PANASONIC VBHN325SA17 325W

System 2:
MATE3s, 2 x VFX3648,
2 x FLEXmax 100, HUB10
1 x EnergyCell 48V 2700RE,
1 x FLEXnet DC, 2 x FLEXware ICS
21 x PANASONIC VBHN325SA17 325W
Location: Providenciales, Turks & Caicos Isl.

Re: Tuning charge parameters for parallel battery banks

Post by EA6LE-ONE » Wed Jun 24, 2020 9:48 am

raysun wrote:
Wed Jun 24, 2020 9:32 am

You are correct. If a single shunt is used, it can show net amps, but not differential amps.

In the former case, if the charger was capable of 200 amps, and the load demanded 199 amps, the net to the battery is 1 amp. There would be no practical way for a monitor to discern whether the net flow to the battery is due to battery impedance or the load/charger dynamics.
I don't see the need of differential amps. if only one shunt is used connected to the battery and the the CCs and inverters are connected before the shunt then you will see only what is going to the battery and what is coming out.

raysun
Forum Emperor
Posts: 2309
Joined: Tue Jul 26, 2016 5:57 am
My RE system: Flexpower One: FX3048T, (2) FM80, MATE3s, FlexNetDC
Outback 200NC batteries (8 @ 48v)
Outback IBR3 battery enclosure
Suniva 330 watt panels (12 - 6 strings of 2 in series)
Hyundai 355 watt panels (6 - 3 strings of 2 in series)
Honda EU7000is gas fuel generator

Re: Tuning charge parameters for parallel battery banks

Post by raysun » Wed Jun 24, 2020 9:56 am

EA6LE-ONE wrote:
Wed Jun 24, 2020 9:48 am
raysun wrote:
Wed Jun 24, 2020 9:32 am

You are correct. If a single shunt is used, it can show net amps, but not differential amps.

In the former case, if the charger was capable of 200 amps, and the load demanded 199 amps, the net to the battery is 1 amp. There would be no practical way for a monitor to discern whether the net flow to the battery is due to battery impedance or the load/charger dynamics.
I don't see the need of differential amps. if only one shunt is used connected to the battery and the the CCs and inverters are connected before the shunt then you will see only what is going to the battery and what is coming out.
If you don't see the need, then don't bother with it.

User avatar
EA6LE-ONE
Forum Guru
Posts: 151
Joined: Tue May 05, 2020 7:51 am
My RE system: System 1:
MATE3s, 3 x Radian GS8048A,
4 x FLEXmax 100, HUB10,
2 x EnergyCell 48V 2700RE,
1 x FLEXnet DC, 4 x FLEXware ICS Plus,
54 x PANASONIC VBHN325SA17 325W

System 2:
MATE3s, 2 x VFX3648,
2 x FLEXmax 100, HUB10
1 x EnergyCell 48V 2700RE,
1 x FLEXnet DC, 2 x FLEXware ICS
21 x PANASONIC VBHN325SA17 325W
Location: Providenciales, Turks & Caicos Isl.

Re: Tuning charge parameters for parallel battery banks

Post by EA6LE-ONE » Wed Jun 24, 2020 10:20 am

raysun wrote:
Wed Jun 24, 2020 9:56 am

If you don't see the need, then don't bother with it.
Raysun, sorry to annoy you :)

From your experience, if you need only to monitor the health of the battery with FNDC, 2 shunts will do it better than one?
Since I installed the FNDCs, I like more the way I can monitor the battery with only one shunt. I had it before on the old system with 2 and never bother much with it, a bit confusing to read it on the old mate.
For now I can say the batteries are working well and the float is triggered at the right time with the FNDC using the return amps.

provo
Forum Guru
Posts: 403
Joined: Thu Jul 27, 2006 4:34 pm
My RE system: Sixteen Evergreen EC-120
(4 strings, total 1920W)
Eight Rolls S-550 (2 strings, total ~800Ah @ 24V)
One FM80
One VFXR3524A
Hub 10
Mate3s
FNDC and Trimetric
Location: Sierra foothills

Re: Tuning charge parameters for parallel battery banks

Post by provo » Wed Jun 24, 2020 11:04 am

EA6LE-ONE wrote:
Wed Jun 24, 2020 10:20 am

Since I installed the FNDCs, I like more the way I can monitor the battery with only one shunt. I had it before on the old system with 2 and never bother much with it, a bit confusing to read it on the old mate.
I went 13 years with only a Trimetric battery monitor, which just uses one shunt. Got used to that, am an old fart, and one shunt is still OK with the FNDC :grin: .

Post Reply