Battery Bank not charging properly

Apica

Good day everyone, I'm having problems full charging 1 string/bank in my 450AH 24V battery bank which I have wired in a series/parallel configuration. I have no problem achieving 1.265 -1.275 SG from each cell in one bank but the no more than 1.250 from the second bank. I installed a Trimetric which will display 100% full but I'm not convinced that this is correct based on my hydrometer readings. What could be causing this? Any information will be greatly appreciated.

Photowhit

Are they the same size, age and type of batteries?

Have they been wired so that they have equal resistance across each string   Since  they are connected together they are 1 battery bank made up of 2 strings. This is usually achieved by being sure that each  connecting wire is the same gauge and length and connecting the bank to the  system with wires from the positive terminal  of one string and  the negative terminal of the other.

There is more information  here;
http://www.smartgauge.co.uk/batt_con.html

mcgivor

Classic symptoms of imbalance, remove the good string and allow the lower string  to reach the SG of the good string, don't apply any loads..Parrallel the two correctly, diagonally as pre link provided above, then equalize to bring everything into alignment before connection of loads. 

Apica

Thank you for your fast replies.

PhotoWhit, The batteries are FLA and are all the same size however there is a 6 month age difference between both strings. The are all connected using 2/0 cable of the same length.

Apica

Mcgiver, I followed your suggestion and separated the strings this morning and left the bank to charge with no load. I'll check the SG later and update.

BB.

Also, use a voltmeter and measure the voltage across each battery (during charging/discharging/resting). You have 8 identical batteries and they should all be nearly the same voltage (something like 0.1 volts or less ideally). If you have one "higher voltage" battery or electrical connection, it will greatly reduce current flow in that string.

And check the battery connections/cabling too. If you have one poor connection in a string, it will stop or greatly reduce current flow and prevent charging (and prevent discharging too).

If you do not have one yet, you might want to get a DC Current Clamp Meter. You can measure the current flow in each battery string during charging/discharging--Again you are looking for "differences" between otherwise identical strings/batteries.

Here is a mid-priced AC/DC Current Clamp meter:

https://www.amazon.com/gp/product/B019CY4FB4

Note--You can DC Current clamp meters from $50 to $350+. And any of them are "good enough" for our needs (debugging/understanding our systems). Note there are many AC only current clamp meters too... You do not want that for working on DC power systems (sometimes the descriptions are a bit confusing because you can have an AC only Current Clamp meter with AC+DC DMM (digital multimeter) to measure AC&DC volts--That is different.

-Bill

Apica

This is my battery bank design. The top string is the newer of the two that does not get any higher than 1.250.

Apica

Thank you BB, I'll check the voltage across each battery and connections as advised. Unfortunately I don't have a DC Clamp meter but I'll definitely look into getting one asap.

Apica

After a full day of charging 1 string with no load the specific gravity has risen to 1.265 in some of the cells, the others still remain at 1.250.

A voltage check across each battery this morning from left to right gave the following readings:

At Rest
6.17, 6.19, 6.18, 6.25

While Charging
6.43, 6.47, 6.46, 6.54

I'll post discharging voltage readings later on today.

BB.

The charging voltage seems low... Either the batteries are quite discharged--Or there is not much current flow (do you know how much current was flowing while charging?).

In general, referencing 12 volt battery--Around 13.6 to 14.8 volts (or 6.8 to 7.4 volts @ 6 volt battery) if the battery string is "actively" charging (something like 5% to 20% of rated capacity--I.e., 10% suggested nominal charge rate on a 220 AH rated battery would be ~22 amps under full sun or on genset/utility charger).

All the battery voltages added together should give you the bus voltage (i.e., 6.43+6.47+6.46+6.54=25.90 volts bus voltage), plus any voltage drop on wiring... Compare that to the voltage measured directly on the charge controller terminals (looking for too much voltage drop in wiring--I.e., small diameter wiring, long wire runs, loose/corroded connections).

25.9 volts on an FLA battery bus is either just after charging is completed (surface charge voltage--Around 29.4 to 29.8 volts for fully charged resting voltage is typical).

-Bill

Apica

BB, the Flexmax 80 was putting out 13.3amps while charging. The measurements where done at 7:45am. 

I'll do measurements at various times of day on Sunday when it will be more convenient.

BB.

Apica, also, double check your voltmeter (check car battery, check AC house voltage, etc.). I have seen DMMs (digital multimeters) with weak/bad batteries give very consistent/weird results.

Those battery voltages are showing low voltage when charging... Typically that is because they are at a lower state of charge... Typically when a FLA battery hits >~80% state of charge, you should be much closer to ~14.4 to 14.75 volts (7.475 volts @ 6 volts) when you have significant charging current.

Here is a nice graph that estimates the voltage of a FLA battery vs various current flows (charging/discharging) and battery state of charge:

http://www.scubaengineer.com/documents/lead_acid_battery_charging_graphs.pdf

Note that these are estimates and may not apply to your setup... For example, a battery with C/20 (5%) rate of charge, a 13 volt @ 12 volt (or 6.5 volts @ 6 volt bank) should be around 50% state of charge.

Obviously, I am not measuring your system and I don't know about your meter/tools/conditions/age&conditions of batteries/etc..... But when something does not "look right", I suggest going into the details and get independent confirmation when possible (2nd meter, using current clamp meter to confirm sharing of current in parallel battery strings, etc.).

-Bill

petertearai

Hi . What size is your pv .. maybe not enough amps for your battery bank .

mcgivor

petertearai said:

Hi . What size is your pv .. maybe not enough amps for your battery bank .

In the signature bottom of posts. 6 × 260W + 2 × 255W

petertearai

Ahhh . I don't see a signature on my phone . Must check on laptop .

Apica

mcgivor said:

petertearai said:

Hi . What size is your pv .. maybe not enough amps for your battery bank .

In the signature bottom of posts. 6 × 260W + 2 × 255W

petertearai, I have always had doubts about my array size. I live in the northern hemisphere and ideally my panels should be facing south, I have not been able to do that because of shading issues. I do get adequate sun hours, usually between 7am - 3pm, enough to achieve the absorb voltage of 28.8V for 2 hours most days but very little float time when charging the 450AH bank. Now that I'm charging one string  225AH I am able to float for at least 2 hours if I discharge to 50% SOC. I have <600W continuous loads. I have been switching over to grid power at 9pm for the last 3 days to prevent my bank from discharging  below 70%.

Yesterday I purchased 1 more 270W Intenergy Panel. Now I have 6 x 260W panels (1560W) connected to the FM80 and 2 x 255W & 1 x 270W (780W) connected to the FM60 for a combined total of 2370W.

Apica

At 6:00am these are the measurements I gathered :

At Rest

Trimetric: Battery Voltage 24.5, SOC 80% Power being drawn from bank 11.5W

FM80 LED Display Voltage 24.5V, Voltage at terminals 24.5V
FM60 LED Display Voltage 24.5V, Voltage at terminals 24.5V

Individual battery voltages: 6.10, 6.12, 6.11, 6.18 
Total string voltage 24.51

at 7:00 I will switch power from grid to PV and update during the course fo the day.

Apica

BB, my DMM seems to functioning well. Thank you for sharing that lead acid battery charging chart.

mcgivor

Apica said:

mcgivor said:

petertearai said:

Hi . What size is your pv .. maybe not enough amps for your battery bank .

In the signature bottom of posts. 6 × 260W + 2 × 255W

petertearai, I have always had doubts about my array size. I live in the northern hemisphere and ideally my panels should be facing south, I have not been able to do that because of shading issues. I do get adequate sun hours, usually between 7am - 3pm, enough to achieve the absorb voltage of 28.8V for 2 hours most days but very little float time when charging the 450AH bank. Now that I'm charging one string  225AH I am able to float for at least 2 hours if I discharge to 50% SOC. I have <600W continuous loads. I have been switching over to grid power at 9pm for the last 3 days to prevent my bank from discharging  below 70%.

Yesterday I purchased 1 more 270W Intenergy Panel. Now I have 6 x 260W panels (1560W) connected to the FM80 and 2 x 255W & 1 x 270W (780W) connected to the FM60 for a combined total of 2370W.

The thing to remember is that the rated wattage of a panel  is under perfect conditions, anything less will subtract, so facing  anything other than somewhere in the southern direction in the northern hemisphere there will be a deficit, even facing south the best you could expect is ~75% of rated output in most conditions.

Secondly the bulk/absorption voltage of 28.8V is probably  too low for FLA, assuming there is temperature compensation, it should be somewhere in the mid 29V regon, depending on manufacturers recommendations, which may themselves be based on grid supply with extended charging times available, something not applicable to solar applications. Basically it's a fine ballancing act, using a higher voltage to force the current into the bank in the limited time available and "one size doesn't fit all".

Thirdly,  if  depth of discharge is 50% on a regular basis it only reinforces the need to increase the charging rate by increasing the voltage.

I'll leave now to see what others have to offer.

Apica

Excellent points Mcgivor.  I recently purchased 2 Outback BTS but haven't gotten around to installing yet, so no temperature compensation at the moment.

My greatest challenge has been obtaining charging parameters for these Trace batteries, my research online has turned up nothing. I'm just being cautious using 28.8V for absorption. I have increased the bulk/absorb voltage to 29.6V before but was concerned about the high rate of water loss.

Apica

Apica said:

At 6:00am these are the measurements I gathered :

At Rest

Trimetric: Battery Voltage 24.5, SOC 80% Power being drawn from bank 11.5W

FM80 LED Display Voltage 24.5V, Voltage at terminals 24.5V
FM60 LED Display Voltage 24.5V, Voltage at terminals 24.5V

Individual battery voltages: 6.10, 6.12, 6.11, 6.18 
Total string voltage 24.51

at 7:00 I will switch power from grid to PV and update during the course fo the day.

Update

10:20 am and Batteries are now in absorbing

Trimetric : Battery Voltage 28.6, SOC 91%

FM80 LED Display Voltage 28.8V, Voltage at terminals 28.8V, Amps Out 43.1A, @ 1130KW
FM60 LED Display Voltage 28.8V, Voltage at terminals 28.8V, Amps Out 24.9A @ 680KW

Individual battery voltages 7.05, 7.06, 7.05, 7.16
Total String voltage 28.32 ( How can this be possible?)

Estragon

It looks to me like there's about a half volt of voltage drop between CC and bank.  What size and length is the wire? 

That much drop means the bank likely isn't getting proper absorb voltage, so settings should be increased to compensate.

Temp compensation is quite important for proper charging if ambient is more than +/- a few degrees off 25°C, and to limit charge related heating (and excess water loss) if warm.

BB.

Measure the voltage at the FM 80 terminals vs Battery Bank bus bars with the same DMM.

Note that if your temperatures are greatly different than ~75F, I have seen some charge controllers give a "corrected" temperature reading to ~20C/70F... Don't know about Outback specifically.

Assuming the FM 80 gives actual battery bus voltage reading, many controllers can be adjusted to the match a reference DMM reading (always raises the question is your DMM or FM 80 more accurate).

Wikipedia: Segal’s law 
Segal’s law is an adage that states: 

“A man with a watch knows what time it is. A man with two watches is never sure.” 

-Bill

Apica

Estragon said:

It looks to me like there's about a half volt of voltage drop between CC and bank.  What size and length is the wire? 

That much drop means the bank likely isn't getting proper absorb voltage, so settings should be increased to compensate.

Temp compensation is quite important for proper charging if ambient is more than +/- a few degrees off 25°C, and to limit charge related heating (and excess water loss) if warm.

Wire size from CC to E-Panel is 6 AWG and 4 feet long. Room temperature is 84°F and using a very basic infrared thermometer to check each battery, temperatures ranged from 88.7 - 90°F.

I just installed the Outback RTS to the FM80, left the default setting (WIDE) and noticed the absorb voltage dropped from 28.8V to 28.6V but the FM60 was still absorbing at 28.8V. Should I attach the second sensor for the FM60 to the same battery or wait until I connect the other string and attach to anther battery?

Estragon

6awg over a 4' run at 40a current should only drop maybe 1/10th volt.  Like Bill suggests, maybe try measuring at both ends (with good current flowing).  Maybe a connection needs retorquing?

My classics are networked, so they share a RTS (also charge stage changes).  I assume FMs can be networked, but maybe yours aren't set up to share?  I have my RTS stuck to a middle battery about 1/2 way down.  If room temp gets much different than batteries, some people tape a chunk of insulation over the sensor to get a better reading.

If not networked, I'd put the other sensor next to the first on the same battery.  Ideally, you want both controllers in the same state at about the same time, which they likely won't do if they have different target absorb voltages.

mcgivor

Definitely install both if not networked,  the controller without a RTS  will raise the voltage above the target of the other with the RTS essentially shutting it down. They need to both act as a team with the same goal plan, the compensated voltage being target, the RTS should be programmed to reduce the voltage by 0.005V (5mV) per cell per °C or by 60mV for every °C on a 24V nominal system.

The reverse occurs as temperature drops below 25°C, increasing the voltage inversely proportional to the > than 25°C value.

As an example let's say the bulk/absorb setpoint is 29.6V, at 35°C (95°F) the target voltage would be reduced to 29.0V, this is important because it's a dynamic function, the battery temperature is not constant, self heating whilst under charging conditions due to the inherent inefficiencies of lead acid batteries. Conversely the battery temperature will drop overnight meaning the target voltage value may well differ throughout the day, it really takes a lot of guesswork out of the equation when using a RTS.

Apica

Today was very sunny so I took the opportunity to do some tests still using 1 string (225AH)  which were discharged to 54% SOC the night before. I decided to use only the FM80 CC  which now has the RTS installed to charge the batteries and increased the absorption voltage to 29.6V for 2 hours.

After a full day of charge with no load and a maximum current of 41.2A. I observed that the absorption voltage had decreased to 29.2V  At the end of day I took measurements across the battery bank at rest, FM80 Led Display & Terminals, Bus bars and Trimteric voltage and all were the same for the first time... 26.1V. However a SG check of the battery cells left me disappointed. They were either 1.265 or 1.250. I'm right back where i started. I'm beginning to wonder if these batteries are defective. 

Tomorrow I will do a Equalization at 31V and measure again.

mcgivor

The difficulty with monoblock batteries is if there is a weak cell (low SG ), is the other cells reach a higher voltage not allowing the low cell to increase sufficiently to gas so the SG remains low at the expense of the other cells. This phenomenon is even more difficult whth long strings, 12 cells in your case, more cells to form the aggregate voltage, this is where equalization is used to provide enough voltage to overcome the deficit seen during normal charging. Cycling of the battery will only make the process take longer, two steps forward, one step back.

Before equalization the battery should be fully charged, if the cells in question begin to gas, that means the voltage is high enough, check  the SG every 30 minutes, when it matches the other cells or fails to increase after 3 hours  there is little point in continuing as it would be more harmful to the ballance of cells. Use the temperature sensor, the battery temperature will increase, the compensated voltage is sufficient.

FWIW, when I switched to lithium my lead acid bank had low SG cells, nothing outrageous just something like yours, they remained essentially unused but on charge for weeks, only load was an on demand 100W water pump. After this time had passed all cell SG's were perfectly matched without any EQ, time it seems is all it took. 

Apica

Just an update.

After equalizing the problem string for 4 hours I was able to get SG on each cell in the bank to 1.275. Both strings have been paralleled, absorb time and voltage has been increased to 3 hours and 29.2 Volts and things have been going smoothly. My only concern is that 1 string (i.e the one that previously had no issues achieving high SG readings) uses up much more water.  

mike95490

Apica said:

Just an update.
...... My only concern is that 1 string (i.e the one that previously had no issues achieving high SG readings) uses up much more water.  

I suspect that the current in your 2 strings, is not equal.  Get a clamp-on DC meter, and check that both strings are charging equally. There may be a bad connection at one of the jumpers somewhere.   Or get the system charging hard, and use an IR thermometer to read the temps of each connection, the warm one is the bad one.