Battery voltage plot

Horsefly
Horsefly Registered Users Posts: 470 ✭✭✭✭
edited December 2017 in Solar Beginners Corner #1
Hey guys.  Most of you will recall that with lots of help from you all about a year ago I managed to design a system for our off grid cabin, and got it up and running in July of this year. It's been operating great. One of the things that @Marc Kurth suggested somewhere here was a small little voltage logger. I've had it hooked up since we commissioned the system, just watching how the voltage changes and logging the voltage once per hour. My intent was to see how the batteries did over the winter: No loads, possibly some snow cover on the panels for some time, otherwise float charging most days.

We've had a dry late fall here, so I managed to go up to the cabin today and pull the data for the last couple of months:

Notice the scale on the left in volts. When someone has been there at the cabin for a night, the next morning the CC goes into bulk charging when the sun comes up. Most other days it never actually goes into bulk, but rather starts up in float.  In the chart, there are a few overnight stays by my sister and her husband, and a two week period where my brother was up hunting elk.  There also were some bulk charges in November when no one was there.

Here's what I noticed that I wasn't really expecting, and am hoping some of you experts have some comments. Every evening the voltage dropped down from the float voltage as expected. If there was a bulk charge that day, voltage settled at something over 26V through the night, until the CC brought it back up to float voltage again the next morning. That night the voltage would settle a little lower than it had the night before. This downward trend in low voltage would continue until either another bulk charge or until it reached just under 26V, at which point it seemed to stay level nearly for ever.  So what I'm specifically talking about is the downward slope of a line that would connect the low voltage each morning.

I'm just wondering if this is something that can be explained.  It looks like if the batteries are getting cycled for even a couple of days the low voltage is higher than it would be if the bank stayed in float, and it will drift down over a period of several days.  Any thoughts?

Steve

P.S. I just realized there will be some other questions that are likely to come up looking at this graph, so here's some preemptive answers:
  1. The batteries are AGM, so don't ask me about SG.   :)
  2. Because the logger is only logging once per hour, there may be some lower lows and higher highs than were caught by the logger.
  3. My design was intentionally over-paneled for this batter bank, for more than one reason. Even when someone has used a fair amount of battery charge, absorption pretty much always ends less than an hour after bulk starts. 
  4. I had my low battery cut off set fairly high, at something like 23.7V. This forced my brother to start the generator a couple of times before the battery got very low. That's why my bank mostly didn't get below 25V for my once-per-hour logging.
Edit to Add, in case it matters: Charge voltage is 29.1V, float is 27.2V temperature compensated (-30mV/deg C).


Off-grid cabin: 6 x Canadian Solar CSK-280M PV panels, Schneider XW-MPPT60-150 Charge Controller, Schneider CSW4024 Inverter/Charger, Schneider SCP, 8S (25.6V), 230Ah Eve LiFePO4 battery in a custom insulated and heated case.

Comments

  • Estragon
    Estragon Registered Users Posts: 4,496 ✭✭✭✭✭
    Unfortunately, the graph isn't showing up for me.

    The resting voltage of the bank will be ~25.4v +/- a couple tenths. That's sitting with no loads or charging sources for many hours. The first volts drop fairly quickly from Vabs to Vfloat, and slower fron Vfloat to resting with no load as surface charge dissipates on a curve.

    AGMs tend to have low self-discharge to begin with, and if the bank is cool/cold, it will be even lower. The controller itself is a load too, but if self-discharge and controller load still sees battery voltage at ~26v most of the time, it sounds good to me.
    Off-grid.  
    Main daytime system ~4kw panels into 2xMNClassic150 370ah 48v bank 2xOutback 3548 inverter 120v + 240v autotransformer
    Night system ~1kw panels into 1xMNClassic150 700ah 12v bank morningstar 300w inverter
  • Horsefly
    Horsefly Registered Users Posts: 470 ✭✭✭✭
    Rats! Sorry the image isn't showing up. Here it is as an attachment. 

    What I'm seeing is that the CC did a bulk & absorption on 29-Oct. The next morning, the voltage had dropped to 26.3V before the sun came up enough that the CC started bringing the voltage back up to float, with no bulk / absorption. The next day the voltage dropped to 26.25V before it started back up. The next day it was 26.2V. The following day it was 26.15V, and again on the day after that. The next day it dropped to 26.1V for a couple of days, then to 26.05V for a couple of days. After a few more days it seems to stabilize at about 25.95V - 26.0V.  It's this daily drop in the low point that I don't understand.

    Based on what your are saying, my batteries are not generally getting to a true resting voltage before the CC starts up again in the morning.  I guess that's OK. 
    Off-grid cabin: 6 x Canadian Solar CSK-280M PV panels, Schneider XW-MPPT60-150 Charge Controller, Schneider CSW4024 Inverter/Charger, Schneider SCP, 8S (25.6V), 230Ah Eve LiFePO4 battery in a custom insulated and heated case.
  • Estragon
    Estragon Registered Users Posts: 4,496 ✭✭✭✭✭
    Can see the plot now - that slight downward trend is sorta odd. The only thing I can think of offhand is the days are getting shorter, so more time for voltage to get to resting? Would be interesting too see if it continues until ~ Dec 21st then trends slowly back up.
    Off-grid.  
    Main daytime system ~4kw panels into 2xMNClassic150 370ah 48v bank 2xOutback 3548 inverter 120v + 240v autotransformer
    Night system ~1kw panels into 1xMNClassic150 700ah 12v bank morningstar 300w inverter
  • Horsefly
    Horsefly Registered Users Posts: 470 ✭✭✭✭
    Estragon said:
    Can see the plot now - that slight downward trend is sorta odd. The only thing I can think of offhand is the days are getting shorter, so more time for voltage to get to resting? Would be interesting too see if it continues until ~ Dec 21st then trends slowly back up.
    Well, it will probably be late May or early June before I'll be able to pull the log again. I'm going to be pretty anxious to see how everything does through the winter. The log will be my only way to see if the panels were covered with snow.

     As far as the graph, I'm mostly concerned if the low voltage continues to drop all winter. It would be bad if it really needs a good discharge now and then to stay healthy, since that won't happen until next summer.
    Off-grid cabin: 6 x Canadian Solar CSK-280M PV panels, Schneider XW-MPPT60-150 Charge Controller, Schneider CSW4024 Inverter/Charger, Schneider SCP, 8S (25.6V), 230Ah Eve LiFePO4 battery in a custom insulated and heated case.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,728 ✭✭✭✭✭✭
    High error of the logger, temp comp changing and settling in? Extreme cold? Does not seem to be that big of a deal as long as the battery is occasionally getting some charge. The cost of leaving batteries for the winter does have to be paid. It is easy for me to say don't worry about it. Good Luck  
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • westbranch
    westbranch Solar Expert Posts: 5,183 ✭✭✭✭
    I got the same feeling as Dave did....  those post charge values are quite high by eyeballing it  high ~27.5V  Low ~25.8V Started at ~26.3V.... another trend I see is the Highs are creeping up  AND the Lows are descending all so slowly.. interesting to see what the graphs look like after the batts are broken in after a year.

    I would try to calibrate the logger with the V reading at the CC and compare that to the @ Battery Volts and try to 'normalize' the logger values  to the CC...
     
    KID #51B  4s 140W to 24V 900Ah C&D AGM
    CL#29032 FW 2126/ 2073/ 2133 175A E-Panel WBjr, 3 x 4s 140W to 24V 900Ah C&D AGM 
    Cotek ST1500W 24V Inverter,OmniCharge 3024,
    2 x Cisco WRT54GL i/c DD-WRT Rtr & Bridge,
    Eu3/2/1000i Gens, 1680W & E-Panel/WBjr to come, CL #647 asleep
    West Chilcotin, BC, Canada
  • Horsefly
    Horsefly Registered Users Posts: 470 ✭✭✭✭
    edited December 2017 #8
    One of the first reviews of the logger on Amazon was someone doing a detailed analysis of the accuracy of the logger, and he determined it was very accurate. I guess that still could be a source of the problem, but since it was a trend and not a random error, I don't think it was an accuracy issue.  Still, it is probably worth me doing some checks next spring.  Cross-checking with the reading at the CC could be a little time consuming, but I suppose I can do it. I don't have another logger (kinda wish I would have purchased the COMM Box for the Schneider stuff, but...), so I would have to just right down the voltage at the CC each hour and see how it compares when I pull the log.

    The float value was creeping up, and I assume it is just the temperature compensation. When I pulled the logs yesterday the control panel said the BTS was reporting 40 deg F for the battery temp. I didn't do the math for what that would mean for a -30mV/deg C temp compensation, but it would clearly be higher than the 27.2V that I programmed it for.

    I guess I won't worry about it, but it is curious.  Most of the stuff about lead acid batteries seems kinda magic to me, which caused me to read more about them than I probably really needed to.  I never saw anything that would explain this behavior.  It doesn't look like @Marc Kurth  visits the board much any more, or else I would expect him to chime in with a good explanation.

    I'll post an update here (if I can remember) next spring / summer when I pull the updated log.
    Off-grid cabin: 6 x Canadian Solar CSK-280M PV panels, Schneider XW-MPPT60-150 Charge Controller, Schneider CSW4024 Inverter/Charger, Schneider SCP, 8S (25.6V), 230Ah Eve LiFePO4 battery in a custom insulated and heated case.
  • mcgivor
    mcgivor Solar Expert Posts: 3,854 ✭✭✭✭✭✭
    The changes could be due to many factors, other than battery temperature, which will have an effect on voltage, these include logger temperature, as temperatures drop the components within, resistors for example, change in value which can alter the readings. Often voltage loggers have a temperature logger due to the correlation between the two, it's accuracy too is dependent on ambient changes, the changes could be natural deviations in tolerance. Personally I wouldn't be overly concerned about subtle changes especially when it's known that temperatures are dropping, the days are getting shorter, which means the controllers self consumption, albeit small 2.5W, is drawing for more hours. Understandably there is a certain amount of anxiety with a new system, to get a more accurate indication of trends, variables have to be compensated for, perhaps looking at trends over the course of a year would be less concerning. The voltage is still very good even at the lowest reading, in fact I'm surprised at how good they actually are, my guess is there will be a reverse as weather warms up, days get longer etcetera.
    1500W, 6× Schutten 250W Poly panels , Schneider MPPT 60 150 CC, Schneider SW 2524 inverter, 400Ah LFP 24V nominal battery with Battery Bodyguard BMS 
    Second system 1890W  3 × 300W No name brand poly, 3×330 Sunsolar Poly panels, Morningstar TS 60 PWM controller, no name 2000W inverter 400Ah LFP 24V nominal battery with Daly BMS, used for water pumping and day time air conditioning.  
    5Kw Yanmar clone single cylinder air cooled diesel generator for rare emergency charging and welding.
  • westbranch
    westbranch Solar Expert Posts: 5,183 ✭✭✭✭
    Horsefly, Mcgivor, all true, but to strip it down I see that the top (avg) line is ~ >1 V above what a fully charged battery would be and marginally increasing while the bottom (avg) line is ~ where you would expect the top line to be +/-  .. I would not expect a  large Temp adjustment at this time of the year, more when things get down to -20*F and below.  What is your temp adjustment value?.  Is there a 'local' meteorologic station to refer to ?
    The ~ 1 volt drop overnight, when not inhabited, is a  bit more than I would expect but may be, probably is, the tare rate of that large capacity CC you have.  My classic is ~ 4 watts or less IIRC
     
    KID #51B  4s 140W to 24V 900Ah C&D AGM
    CL#29032 FW 2126/ 2073/ 2133 175A E-Panel WBjr, 3 x 4s 140W to 24V 900Ah C&D AGM 
    Cotek ST1500W 24V Inverter,OmniCharge 3024,
    2 x Cisco WRT54GL i/c DD-WRT Rtr & Bridge,
    Eu3/2/1000i Gens, 1680W & E-Panel/WBjr to come, CL #647 asleep
    West Chilcotin, BC, Canada
  • Horsefly
    Horsefly Registered Users Posts: 470 ✭✭✭✭
    Thanks @westbranch. As I mentioned in the first post, my settings are per the battery manufacturer: 27.2V float, 29.1V charge, and temp compensation of -30mV/°C. When I pulled the logs the BTS was reporting the battery temp as 40°F. That would make the float voltage go up by about 0.62V, to about 27.82V.  Looking at the spreadsheet behind the graph, that's pretty close to what it was: 27.75V mostly, with one reading up to 27.8V. So, I think that part is working right.  It may be that my AGM batteries are taking a higher float / charge than whatever you are used to.

    In my system I have the Inverter (SW4024) in standby for the winter. Although I can't find it in the specs, I would think there is some small power consumption when it is in standby (vs. Off). The tare loss / night power consumption of the CC (MPPT-60-150) is spec'd at 2.5W.  The System Control Panel (SCP) also consumes something less than 3W.  So I'd say a pretty conservative estimate is that my system has a total tare loss of something less than 8W. That's enough that I guess I'm not surprised by the drop during the night.
    Off-grid cabin: 6 x Canadian Solar CSK-280M PV panels, Schneider XW-MPPT60-150 Charge Controller, Schneider CSW4024 Inverter/Charger, Schneider SCP, 8S (25.6V), 230Ah Eve LiFePO4 battery in a custom insulated and heated case.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,728 ✭✭✭✭✭✭
    Steve,  if there ever was a candidate for a combox, there would be a slot with your name on it. The screen shot is just one page of many that could log enough and for long enough to keep an analytic up all night. All you would have to do is have someone pull/replace the SD card and you could peak at it if the snow melts.
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • Horsefly
    Horsefly Registered Users Posts: 470 ✭✭✭✭
    edited December 2017 #13
    Yeah Dave, I think you have me pegged.  :)   I've been worried that I won't have enough stuff to analyze during the winter.  I happened to make a pitch for the Combox to my partners in this effort, so maybe that will tenderize them for a purchase next summer.
    Off-grid cabin: 6 x Canadian Solar CSK-280M PV panels, Schneider XW-MPPT60-150 Charge Controller, Schneider CSW4024 Inverter/Charger, Schneider SCP, 8S (25.6V), 230Ah Eve LiFePO4 battery in a custom insulated and heated case.
  • westbranch
    westbranch Solar Expert Posts: 5,183 ✭✭✭✭
    Tis the Season.....
     
    KID #51B  4s 140W to 24V 900Ah C&D AGM
    CL#29032 FW 2126/ 2073/ 2133 175A E-Panel WBjr, 3 x 4s 140W to 24V 900Ah C&D AGM 
    Cotek ST1500W 24V Inverter,OmniCharge 3024,
    2 x Cisco WRT54GL i/c DD-WRT Rtr & Bridge,
    Eu3/2/1000i Gens, 1680W & E-Panel/WBjr to come, CL #647 asleep
    West Chilcotin, BC, Canada
  • Horsefly
    Horsefly Registered Users Posts: 470 ✭✭✭✭
    Tis the Season.....
     :)  Good idea! Maybe I'll give it to myself for Christmas! I can even promise not to peek until Christmas day!
    Off-grid cabin: 6 x Canadian Solar CSK-280M PV panels, Schneider XW-MPPT60-150 Charge Controller, Schneider CSW4024 Inverter/Charger, Schneider SCP, 8S (25.6V), 230Ah Eve LiFePO4 battery in a custom insulated and heated case.
  • westbranch
    westbranch Solar Expert Posts: 5,183 ✭✭✭✭
    w ;)
     
    KID #51B  4s 140W to 24V 900Ah C&D AGM
    CL#29032 FW 2126/ 2073/ 2133 175A E-Panel WBjr, 3 x 4s 140W to 24V 900Ah C&D AGM 
    Cotek ST1500W 24V Inverter,OmniCharge 3024,
    2 x Cisco WRT54GL i/c DD-WRT Rtr & Bridge,
    Eu3/2/1000i Gens, 1680W & E-Panel/WBjr to come, CL #647 asleep
    West Chilcotin, BC, Canada