Batteries - Massive Parallel and Series Issue- 3 questions

BB. · February 2013

Re: Batteries - Massive Parallel and Series Issue- 3 questions

I don't want say you messed up--It is more of an issue that batteries are being pushed into markets that are not ready or able to support those needs (different charging parameters, inverters that can operate at significantly higher input voltages, and improved battery monitor system that may include per cell monitoring, etc.).

A "correctly" sized charging system will be current limited (AC charger, flat out current limited; Solar charger--limited charging current at other than around noon, etc.).

I think you can probably get closer to an "acceptable" charging algorithm with many of the newer MPPT charge controllers. Some have programmable maximum current output and programmable battery temperature compensation...

Things I would play with would include charging closer to 15/30/60 volts or a bit more, and monitoring battery temperature (plot it over a day if you can). For the most part, lead acid batteries are very efficient (90-98% efficient) upwards to 80-90% state of charge, then as you go to 90-95-100% state of charge, you approach 50% of the charging current turning into heat (from my little understanding).

I would look at making for a more "aggressive" battery thermal temperature derating (i.e., instead of 0.005 volts per degree C per cell, perhaps 0.010 V/C*C with zero at 25C/77F). And watch the charging profile. From what I have seen, you want maximum charging current until the battery hits (call it) 85% SOC (5-25% rate of charge), then you want to dial the current back (to ~2.5 to 5% rate of charge maximum) when the battery begins to significantly heat (say in the 100-110F range, and shut down charging above 120F--or whatever your battery mfg recommends). (this is all flooded cell/industrial type battery suggestions--just starting points where/what I would try if I had such a system).

Perhaps when Midnite comes out with their integrated battery monitor/current shunt system (in the "near future"), they could actually have a nice maximum rate of charge vs temperature vs voltage vs estimated capacity as measured at the battery bank). Probably would need some other parameters such as absolute maximum voltage (when batteries go cold--many inverters would over voltage if subjected to "aggressive" charging voltages/current).

Anyway, some thoughts on the subject.

-Bill

ChrisOlson · February 2013

Re: Batteries - Massive Parallel and Series Issue- 3 questions

After thinking about it, I think what I'd do is set bulk (in the 5548's) to 58.8 with finish at 2% of C (54 amps). Then try it and see if they absorb for the recommended time of 4 hours once the bulk voltage setting is reached. If the 5548's cut the absorb time short, then you'll have to bump the voltage a bit to get the batteries to draw more current for longer time until you get the recommended 4 hour absorb.

You won't be able to do the finish stage to 2.58 VPC except by doing a manual equalize about once a week. And that won't be ideal because the RE chargers do a constant voltage EQ against a timer, where industrial chargers do a constant-current stage at C/30 until the 2.58VPC has been reached, and then the EQ is stopped. But it's as close as you're going to come with the charging equipment you got.
--
Chris

Vic · February 2013

Re: Batteries - Massive Parallel and Series Issue- 3 questions

One of the main issues with trying to end an Absorption stage using Ending Current, is that this bank really has an unknown Capacity at this point.

These batts really need to be culled, as described previously, and also by the OP. Some of the lagging batteries may accept some remedial EQing, etc. At that point, there may be some additional information of how much C this bank really has. Or perhaps there might be two good-ish banks, each on a different system (??) and a pile of scrap ... who knows?

Agree with you Chris about needing much more flexibility in CC and Inverter charge profiles. BUT, there is also the aspect of this type batttery not really being the best suited for typical off-grid Solar based charging, IMHO.

Those of us using or thinking of using the MidNite Classic should really try to encourage the MN folks to have a fairly universal accessory (as BB Bill mentioned a Batt Monitor perhaps) that will accommodate advanced charging profiles. It IS possible that no mainstream manufacturer of CCs or Inverters will want to delve into this somewhat noisy/messy support tangle ... dunno.

And rtribble sorry to kinna pick at/on you. As Bill mentioned, often we are forced or encouraged to buy those batteries available to us when we need then. Good Luck, Vic

ChrisOlson · February 2013

Re: Batteries - Massive Parallel and Series Issue- 3 questions

Vic wrote: »

Those of us using or thinking of using the MidNite Classic should really try to encourage the MN folks to have a fairly universal accessory (as BB Bill mentioned a Batt Monitor perhaps) that will accommodate advanced charging profiles.

It's not really an advanced charging profile - the I-E-I constant-current industrial chargers are a lot more common than RE chargers. I believe even Surrette has said they have a guy on staff that is trying to figure out how to apply the charging algorithms used in RE chargers to their battery lines, because their charging recommendations are based on the industry standard I-E-I constant-current algorithm.

But the bottom line is that these batteries in question need to be pulled off the bank, one string at a time after being fully charged (using the current charging scheme) and then individually checked for at-rest voltage, SG of each cell, and preferably a load test on each one. That will weed out the ones that have problems so they can be dealt with. That is the exact procedure I use for servicing our bank twice a year - short of the fact that instead of load testing each battery individually anymore (which takes too much time), I have gone to disconnecting all the strings except for the one I am testing, plug two 1,500 watt milkhouse heaters into wall outlets and use the inverter to load the individual string. The two heaters pull about 65 amps from the string (sometimes more or less, depending on other loads) - I leave it on for about 10 minutes then check the loaded voltage of each battery in the string. If they're within .05 volts of one another I consider them good, move on to the next string and repeat the procedure.

A severe duty load test like that weeds out the weak ones really quick.
--
Chris

Vic · February 2013

Re: Batteries - Massive Parallel and Series Issue- 3 questions

ChrisOlson wrote: »

It's not really an advanced charging profile - the I-E-I constant-current industrial chargers are a lot more common than RE chargers. I believe even Surrette has said they have a guy on staff that is trying to figure out how to apply the charging algorithms used in RE chargers to their battery lines, because their charging recommendations are based on the industry standard I-E-I constant-current algorithm.

Sorry Chris. I have been wanting a NUMBER OF DIFFERENT, additional charge profiles in some of the newer CCs as refinements..

You ARE correct that the profile you referenced is NOT "advanced", although it would be an advancement for any Solar CC. Personally, at this point, I have no use for that specific profile.

What I would like to see, as an additional "refinement", is the ability to only start a recharge of my off-grid banks after N-days, or after the measured or inferred SOC is below a settable value. In addition, trying to include some factor of forecast solar availability on successive days would be even neater. Some solar sites are remote and unattended for extended periods. Do realize that if i-net access is available to these sites, that just having the complete range of CC functions under manual remote control would go a long ways to accommodating these desires ... and so on.
That is all. Vic

Cariboocoot · February 2013

Re: Batteries - Massive Parallel and Series Issue- 3 questions

You all do realize that one of the inherent problems with solar charging is that there is not full charging power available all the time, right? The bell curve of a PV array output will force the charging profile to fit the power potential. You can't get around it. This makes it much different than charging from mains where you have an almost unlimited amount of power available that can be adjusted to suit any Voltage-Amperage-Time profile you wish.

ChrisOlson · February 2013

Re: Batteries - Massive Parallel and Series Issue- 3 questions

Vic wrote: »

You ARE correct that the profile you referenced is NOT "advanced", although it would be an advancement for any Solar CC. Personally, at this point, I have no use for that specific profile.

I think the XW's Boost Charging is a neat thing - even better than the constant-current algorithms. It pushes the battery early on in the absorb stage for the first hour at higher voltage, then backs the voltage down to prevent excessive heat and gassing - getting the charging done in a shorter time. That's one thing I'd like to see the Classic be able to do, and I would use that with our battery type. It would get the bank up more during the peak of the solar day so it requires less power to finish absorbing as the available solar power tapers off in the afternoon. All it would take is an extra setting for the bulk voltage, which is higher than absorb voltage - and I'll bet it's something boB could do in firmware in the controller with no changes required to the hardware.

--
Chris

Vic · February 2013

Re: Batteries - Massive Parallel and Series Issue- 3 questions

Cariboocoot wrote: »

You all do realize that one of the inherent problems with solar charging is that there is not full charging power available all the time, right? The bell curve of a PV array output will force the charging profile to fit the power potential. You can't get around it. This makes it much different than charging from mains where you have an almost unlimited amount of power available that can be adjusted to suit any Voltage-Amperage-Time profile you wish.

YES, good point, I certainly DO, just one of the reasons that this type batt makes the most sense in other application, NOT Solar-charged.

However as PV arrays become larger vs battery capacity, it becomes less unreasonable to expect that getting much closer to being able to perform the charge profile mentioned by Chris.

Yesterday, we produced 20.0 Kwh from our 3.15 Kw STC array. The batts were ALL smiles! Thanks, Vic

rtribble · March 2013

Re: Batteries - Massive Parallel and Series Issue- 3 questions

I vary rarely see the MX60's in absorb mode !! whats up ?

Cariboocoot · March 2013

Re: Batteries - Massive Parallel and Series Issue- 3 questions

rtribble wrote: »

I vary rarely see the MX60's in absorb mode !! whats up ?

Well let's see.
12kW of array on a 48 Volt system = 192 Amps of current (48 Amps per MX).
Batteries are 220 Amp hours 12 Volt times twelve strings (hear the alarms going off? It is not possible to keep twelve strings in balance) for 2640 Amp hours @ 48 Volts. That's a 7% peak charge rate. These are big industrial batteries and if your loads are running while charging (pretty much a given) your charge rate will be even worse than that.

Disconnect half the batteries and see what happens.

Then re-evaluate loads, battery capacity, and charge rate. Check every battery cell for SG and see how many of them are low, off-average, or just plain sulphated to death. Look over the wiring for size, length and configuration.

Frankly that battery bank is a mess. You'd be better off with forklift batteries or 2 Volt cells rather than all those 12 Volt units tied together.

BB. · March 2013

Re: Batteries - Massive Parallel and Series Issue- 3 questions

Back to basics... What is the battery bank voltage/specific gravity (if flooded cell). What is the battery bank capacity (I am not sure I found anything with that Deka part number.

What is the average current you get from the solar array (Vmp/Imp) and what is the battery bank power (Vbatt*Ibatt). etc...

-Bill

rtribble · March 2013

Re: Batteries - Massive Parallel and Series Issue- 3 questions

12kW of array on a 48 Volt system = 192 Amps of current at full sun
Batteries are 220 Amp hours 12 Volt times twelve strings = 2640 Ahrs
FDNC says we are full most days by 1pm

Photowhit · March 2013

Re: Batteries - Massive Parallel and Series Issue- 3 questions

rtribble wrote: »

FDNC says we are full most days by 1pm

What is a FDNC???

Photowhit · March 2013

Re: Batteries - Massive Parallel and Series Issue- 3 questions

I wonder in such a large battery bank, with so many strings, if you might be able to allow a clamp meter to lead you to a bad string by shutting down all charging and loads, throw the breaker so nothing is going in or out. Then the battery bank should be trying to send current to the weak cells or at least the weakest strings. I don't know if a clamp meter is sensitive enough to detect this, but it might be worth a shot.

Batteries - Massive Parallel and Series Issue- 3 questions

Comments

Categories