SOC, efficiency, and Battery capacity temperature compensation

[rmak]

I am trying to "fine tune" the SOC reading and want to see if I did this right.

8-Deka L16 wired for 24V for total 740AH @25C
Chart says 70% capacity at 0C
222AH lost divided by 25 (temp change) = 8.88A per degree C.
8.88A / 740A x 100 = 1.2% change in amp per 1 degree C temp drop.
I have 1.2% as the "Battery Temperature Compensation" in the WBjr. and 90% battery efficiency (default setting).

I took SG readings with a Hydrovolt temp compensated meter when the SOC read 73% and they avg. 1.260SG which would be ~90% charge. Do I just need to adjust that 1.2% compensation to fine tune the SOC% reading the WBjr shows? Does the efficiency % have correlation to the SOC reading? What data does the battery efficiency change?

Thanks
Ray

[ClassicCrazy]

The way I did it - leave the default for the %change
Adjust the efficiency setting  to reflect when your amp hours count back to zero .
So after an evening when it is 70% you want to watch the Whizbang  amps decrease the negative number back towards zero . At that point is where your efficiency setting will determine how many more amps it will take for the SOC to say 100% .  So when it starts getting closer to full then you should be taking your SG readings to make sure that  the batteries are full when they say 100% . You tweak the efficiency and possibly the capacity readings to get to something close to what you observe. This would also be the point  you would set your ending amps to go from Absorb to Float. You need to make sure you have  the Absorb time set long enough to find this point so it doesn't time out and go to float too early.

Hope that makes sense. The SOC is something of an estimate and can be pretty close if you tweak it all right - though sometimes you may need to readjust for colder battery temps or age of batteries, etc.

I just reread  your post - there are two temperature settings not to be confused . One is estimate for how much battery capacity the pack has as it gets colder.
The other is the temperature compensation for charging when the batteries get colder or warmer - this changes the battery charging setpoint and is very important to have set correctly . usually it is -3 or -5 mv - check  your battery manufacturer specs.

Larry

[ClassicCrazy]

Also look through the knowedge base tab on this link
Lots detailed info
http://www.midnitehelp.com/

Larry

[boB]

HI guys and happy new year. 

My brain is not working too well so far (cold/bronchitis, etc) but am I hearing the words "amp" when I should be reading "amp-hours" ?

Yes on the two different temperature compensations.  Normal temp-comp is just absorb charge voltage based on battery temperature and
then there is the temperature which I'm pretty sure is the 1.2% figure which adjusts the amp-hours-remaining (where SOC comes from)
based on battery temperature.  And 1.2% may very well be in the right ball park.

The charge efficiency number (90% here) only tells the SOC and remaining amp-hours how many more amp-hours are required in charging
the battery.  In this case of 90%, you will need 10% more amp-hours coming in than if it were set to 100% efficiency.

100 amp-hours needs 110 amp-hours to get full.

There is no adjustment percentage for discharging the battery.  You pull out 100 amp-hours and the remaining amp-hours goes down
by 100AH. 

[rmak]

So use the efficiency for tweaking the SOC% change.
Starting with a full charge (verified by SG), would I discharge a certain amps out (say 100AH) then observe how many amps in to reach full charge again? Would a longer discharge be more accurate?

Hypothetical: 100AH out and 150AH in=50% efficiency. 100AH out and 125AH in=75% efficiency.

Should SG readings be taken with batteries disconnected? If so for how long? EastPenn says to leave disconnected for 24 hours for proper voltage readings.

Ray

[rmak]

Am I correct in that the parameters "Capacity", "Efficiency, "Temperature reference", and "Battery capacity compensation" that are put into the setup for the WBjr does not affect the function of the charging? It only affects the SOC% and capacity readings.

Ray

[boB]

Am I correct in that the parameters "Capacity", "Efficiency, "Temperature reference", and "Battery capacity compensation" that are put into the setup for the WBjr does not affect the function of the charging? It only affects the SOC% and capacity readings.

Ray

The temperature reference will affect different things.  It is typically set to 25C.  If that is set too much different, then readings will be off (SOC)
and the Absorb voltage will not be correct if you have a battery temp sensor plugged in.

Capacity entered is just the rating in AH of the battery.
Battery capacity compensation is how much more or less capacity (AH) the system will think you have available based on the battery temperature relative to the reference temperature (25C usually)

It seems like you are doing things correctly although charging efficiency is, as far as I know, typically a bit higher than 90%.  I thought that with new batteries, it was more like 95% but I don't think that having it set at 90% is a problem.  Probably is better for when it reaches 100% to know it is really there.

Disconnecting for 24 hours huh?  That may be true.  Not sure.  You can also watch the battery voltage after the battery has been sitting for a while after the surface charge starts to drop off.  Right after a good absorb charge, the voltage might show a bit high and that may also reflect in the SG readings.

[rmak]

[/quote]
The temperature reference will affect different things.  It is typically set to 25C.  If that is set too much different, then readings will be off (SOC)
and the Absorb voltage will not be correct if you have a battery temp sensor plugged in.
[/quote]

So the "battery reference" and "battery capacity compensation" in the WBjr setup menu does affect the charge parameters? I thought only the "T-Comp" (-3mv/degree is my current setting) from the main menu in the controller changed the voltage for the Bulk, Absorb, and EQ compensated for temperature from the sensor.

Confused
Ray

[boB]

So the "battery reference" and "battery capacity compensation" in the WBjr setup menu does affect the charge parameters? I thought only the "T-Comp" (-3mv/degree is my current setting) from the main menu in the controller changed the voltage for the Bulk, Absorb, and EQ compensated for temperature from the sensor.

Confused
Ray

That's OK, Ray.  This is why it is not always good to have so many parameters available and hopefully the defaults are set close to most systems from the start.

The reference temperature is the temperature that the batteries are rated for.  That is, their capacity and their Absorb, Float and EQ voltages, etc.

It is the deviation from that reference temperature that controls all of these things.  If the battery temperature and the reference temperature are both 25 degrees C, then NO adjustments are necessary.  And it would not make any difference if you had the battery temp sensor plugged in or not.

Turns out that SOME (few) batteries are referenced to 30 degrees C believe it or not.  I think I saw some that were 20C also so that is why that adjustment is even in there to begin with.  For 95% of customers, we could have just keep it at 25C and not even say anything about it and that would most likely keep any confusion down a notch. 

[Vic]

Hi Ray,

Some of us believe that the Ah charge efficiency of Flooded Lead Acid (FLA) batteries is roughly about 70-ish%.   Sealed batteries are considerably more efficient  --  in the low 90 percent range:

Here is a technical paper:
https://xtronics.com/uploads/batpapsteve.pdf

Here is another Thread on this topic:
http://midniteftp.com/forum/index.php?topic=4115.0

When one gets down to it,  the SOC measuring devices are approximations of actual SOC,  as there are SO many variables involved in the calcs.

The WbJR is a very useful and inexpensive device.   It does a great job,   and SOC approximations are often within about 10 points of actual SOC,  when setup correctly,   but then DOD,   discharge rate,   charge rate,   etc vary,   the needed settings may well change.

All,   IMO,   FWIW,    Vic

[rmak]

Thanks for taking the time to answer my questions Bob but I have a feeling we are not talking about the same thing. I have the charge parameters set up correctly for the batteries now and am trying to get the SOC% to read a little more accurately which I believe is done with the setting in the WBjr.

The Whizbang Jr. has a setup menu. Figure 6 in the manual has "Battery capacity" and Battery efficiency." Figure 7 has "Reference Temp" and "Battery capacity temp comp."

https://www.midnitesolar.com/pdfs/Whizbang_manual.pdf

Am I correct that these parameters do not control charging voltage?

I am aware of that "T-Comp" in the main menu will change voltage in reference to the battery temperature but I did not think the WBjr controlled any of the charge parameters. (except for referencing for using "end amps" on the CC when set to "shunt")

Ray

[boB]

The "Reference temperature" of 25C (usually)  sets the battery reference temperature for the whole Classic.
It is THE reference temperature for T-Comp and also battery capacity calculations in the WB Jr. calculations

It is just a setting that previously did not have a place to adjust in the Classic MNGP menu system.
i.e. that setting (Reference T) only came out when the WB Jr. and SOC% came out.  Before that, the
Ref Temp was always set to 25C and could not be changed.

Another problem with this State Of Charge thing and batteries is that it requires the battery capacity (at full charge), be accurate.  Do you really know what the capacity is ?  If you can find that (and you can !) under normal discharge conditions, then the SOC% accuracy can be really close I think !  As long as you don't need to get the Peukert discharge efficiency calculations involved...   Should be fine as long as accumulated short-timed large current draws don't reduce the capacity too much compared to a straight line of SOC% Vs Amp-Hours removed.

Would be nice to add a way to measure the battery capacity in the software.  That still can be done manually with the display (I think) or remotely by computer.

[rmak]

The "Reference temperature" of 25C (usually)  sets the battery reference temperature for the whole Classic.
It is THE reference temperature for T-Comp and also battery capacity calculations in the WB Jr. calculations

It is just a setting that previously did not have a place to adjust in the Classic MNGP menu system.
i.e. that setting (Reference T) only came out when the WB Jr. and SOC% came out.  Before that, the
Ref Temp was always set to 25C and could not be changed.

Another problem with this State Of Charge thing and batteries is that it requires the battery capacity (at full charge), be accurate.  Do you really know what the capacity is ?  If you can find that (and you can !) under normal discharge conditions, then the SOC% accuracy can be really close I think !  As long as you don't need to get the Peukert discharge efficiency calculations involved...   Should be fine as long as accumulated short-timed large current draws don't reduce the capacity too much compared to a straight line of SOC% Vs Amp-Hours removed.

Would be nice to add a way to measure the battery capacity in the software.  That still can be done manually with the display (I think) or remotely by computer.

Bob,
Thanks for clarifying. I was not aware that the "reference temp" was used by the Classic.

Maybe use a SG sensor to get an accurate SOC reading and help control charging parameters?

http://www.electrosense.com.au/sg-electrode-lead-acid-battery-specific-gravity.htm

Ray

[ClassicCrazy]

Yeah thanks for clearing that up Bob - now that I look at it I see that 25 degrees is in the Temp compensation setting.
I was thinking it was in the capacity settings area - been awhile since I set those.

Larry

[boB]

Glad it's clear now.

Checking the SG  might tell if it is fully charged but just giving the battery a good long absorb charge and making sure that the
battery current is low is probably about as good.

But the actual capacity has to be measured by running the battery down and measuring how much energy it can deliver over time.
I think that can be done by drawing it down to 50% or maybe 25% by using combination of voltage (unloaded) and SG measurements based on the manufacturers' graphs or possibly general lead acid voltage vs. SOC% graphs.  Measuring that voltage or SG might require the battery to be unloaded at least for a little while.  That's the part I don't know about.

Maybe the easiest way to tell what the discharged capacity is is to let it discharge all the way down to almost 0% ?  Under normal usage, I think that would be a real good measurement.  I have a feeling that battery capacities are lower than people think they are or lower than the specifications.  Some may be higher though for part of the batteries life.

But some batteries, and Rolls will admit that you do not get full capacity when the batteries are new.  It takes a while for them to "form" or whatever.
Then, the batteries are good for a while and then start to deteriorate, capacity wise, over time.  Slower than it took them to come up to full capacity though.

I do think it is possible to get a good idea on the capacity of the battery automatically by having the computer look at battery voltage at low current draws after some time after full charge (absorb to float) and after AH have been removed.

It might make sense to just see a graph of battery voltage vs. Amp-Hours removed  over a couple days.  I'd love to see that graph just because.