9 years later... upgrading system... lithium battery math?

couchsachraga
couchsachraga Solar Expert Posts: 84 ✭✭
Almost a decade ago I posted up here and got straightened out for a property sized Lead Acid system for an off grid camp.  It has worked very well, with only one "hiccup" due to an on demand hot water heater bring left on (propane, but with a heating element to prevent freezing... I should have drained it and turned it off as I usually do... but I didn't and it brought the batteries down below 50% for their first time ever.  Thankfully they worked fine still).

Back then this was the formula that was shared:

  • 300 AH * 14.5 volts charging * 1/0.77 solar panel+charger deratings * 0.10 rate of charge: 565 Watt nominal array

I ended up going with a 600AH setup (4 T-1275's), and 1080 of panels.

My first set of batteries is in realistic need of replacement (2 are 10 years old now, the other 2 are 8 or 9).  Though expensive I'm leaning towards lithium (Kilovault at the moment) for a variety of reasons.  The space is heated (kept 40 or above when we're not there, warmer when we are).  Given the extra capacity I'm considering (600AH of batteries, but most of that is useable rather than "only" 300AH (50% DoD) for the Lead Acids) I'm thinking of increasing my array by adding 2 more panels, and having 2 strings of 3  (SolarWorld 265) feeding in to Midnite Classic 200 (I have one already, I believe I'd need a second given the power being produced).

My question is what is the correct math for lithium batteries?  If I understand correctly rather than 1/0.77 it should be more like 1/0.95; perhaps more importantly does the rate of charge for lithium batteries need to be adjusted from the 0.10 rate of charge (C rate) to a higher number?

My "off the cuff" thought, to be clear:
265x6 (really 2 270 and 1 265 in each string; if I understand it correctly that derates them all to 265, but they are at least balanced then) = 1590w of panel

2x Kilovault CHX3600 plus (300Ah usable if I understand correctly; I'm guessing this is 80% DoD) wired in parallel (12v system).

Without knowing the updated figures to use I don't want to move forward with the system; searches have turned about lots of information (a C rate of 1?!), but I'm not sure I'm applying it all correctly, so I wanted to ask here.

Thank you for any and all helpful thoughts and comments.

Comments

  • couchsachraga
    couchsachraga Solar Expert Posts: 84 ✭✭
    Thinking on this overnight I realized I may be better off running the 2 Kilovault batteries in series for a 24v bank - continue using the Midnite Classic 200 I have (with the drop in amps it should handle those panels fine; I might even be able to add 2 more if I wanted), add the inverter I planned anyway and a 24v to 12v DC converter for the items I need to leave 12v (water pump, circulating pumps, and fridge).

    If I follow correctly for a 12v system:
    600AH * 14.5v charging* (1/0.93 (98% efficiency of Midnite * 95% efficiency of Kilovault (an estimate))*.10 rate of charge= 931 watt nominal array.  

    This makes me question if my current system is over-sized (as I am only using 300AH of the 600AH available), and in reality if I need more panels at all going to this setup.  I did note that the "normal" charging for the Kilovaults is 100 amps; if I'm feeding them with the Midnite via a 24v system I suspect it will only be around 55 amps (1590w of panels / 29 volts (?)).   My current Midnight is good up to 78 amps at 24v.
  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    The math works... But you need to look at what you are trying to accomplish too... And respect the limits of the devices/batteries/etc. involved.4

    The charging math:
    • 300 AH * 14.5 volts charging * 1/0.77 solar panel+charger deratings * 0.10 rate of charge: 565 Watt nominal array
    The 0.77 derating was for ~0.81 solar panel deratings (hot panels, Vmp falls) * 5% charge controller losses = 0.77 derating from array to battery charging.

    The 10%... We use 5% minimum suggested. 10%-13% is a healthy middle... And 13%-20/25% for maximum rates of charge... The specific numbers were based on flooded cell lead acid battery charging requirements. 10% was the nominal/minimum suggested charging current for an FLA battery bank (20 hour rate * 0.10 rate of charge = nominal charging current) for best battery health. 5% can work for a weekend/sunny weather cabin system (5% being the minimum suggested rate of charge for Equalize Charging of FLA battery bank. And >13% rate of charge, need to monitor battery bank temperature to prevent thermal run-a-way (battery bank getting hot and depressed charging voltage causing over charging/boiling electrolyte/possible fire).

    It turns out too that the above numbers work well as a starting point for "any" battery bank for solar... 5% rate of charge--You take out 25% of Amp*Hours on weekend, you have ~5 days to recharge in nominal sun. 10%-13% in a reasonably sunny location, 25% discharge overnight, and you get the bank pretty much recharged the next day... And >13% rate of charge--More solar than you might really need (although, solar panels are cheap, and keeping battery well charged even through winter is a good thing).

    The above is based on rates of charge... We also need to look at how much energy you pull from a bank and how much sun you have in a day.

    The typical starting point for a full time off grid system was to start with 2 days of storage (no sun) and 50% max "planned" discharge (for longer battery life). 4x your daily loads--And 25% daily discharge. That formula works out for FLA batteries very nicely... Because the sun is in the sky for limited hours per day (3-6 hours or so typically)--To fully recharge an FLA battery needs a couple hours of "full charging" current (from ~75% State of charge to 90% or so Soc), and another 2-6 hours of "absorb" charging (holding 14.75 VDC for 2-6 hours) to fully recharge to near 100% SoC.

    The sizing works nicely... cycle from 50% to 100% normally... Can discharge to 20% SoC at times when needed (and need to quickly recharge >75% SoC to reduce sulfation) over a few days of sunny weather/genset backup charging. And that 50% to 20% "buffer"... Works nicely with FLA batteries when you assume a loss of 20% Capacity as they age/cycle... The bank cycling from 50% to 100% will still work fine even with 20% loss of capacity.

    Note--For most rechargeable chemistries--There are limits that we don't want to exceed... FLA batteries--Too high of voltage/too long of charging (such as "excessive" EQ charging) and damage the plates and overheat the batteries. And we don't want to go below ~20% SoC because if you have a "weak cell"--That weak cell can actually go from +1.8 volts to zero to -1.xx volts and begin to "reverse charge"... For most types of rechargeable battery chemistries, that is the "kiss of death".

    So--Looking at LiFePO4 batteries (common Li Ion chemistry that is "more stable"/much less likely to catch fire/explode)... We look at the State of charge range... Typically, suggested 10-20% at the bottom SoC (reduce chance of cell damage) and ~90% SoC at the top end (batteries have a longer cycle life if 100% is avoided)...

    So that gives us a working range of 20%-90% SoC = 70% "useful" capacity. And many folks do use this for their sizing calculations. Li Ion batteries recharge much faster and more efficiently, and no "absorb" cycle than lead acid batteries--So drawing below 75% SoC is not an issue (unlike FLA batteries). Also, while FLA batteries are very efficient below 80% SoC--The efficinecy eventually falls to near zero % during the end of Absorb and during EQ cycles)... Using 80% SoC is handy for FLA battery average charging efficiency. AGM~90%. And Li Ion pretty near 100% efficiency.

    Note the "missing" capacity "fudge factor" here... We have a somewhat "useful" capacity for FLA from 50% to 20% (basically another day of storage) and allowing for age/cycle related capacity loss. We have not done that here. Capacity loss in Li Ion--I have not seen anybody taking about that (do cells slowly have a loss of capacity, or do they work until they fail?).

    So--For FLA, the 2 day / 50% rule of thumb gives us a solid 2 days "no sun" capacity and a "soft" 3 day "no sun" capacity before needing to start the genset/pray for good weather. With a 2 day / 70% rule, 2 days of "no sun" and then sun or backup genset--At least.

    There were some other reason for FLA batteries 2/50% rules... FLA batteries have limitations in charging current (>13% rate of charge, the batteries can get hot/over heat), the 2-6 hour "absorb" time (slowly reducing charging current--No "fast charging" here). And another hidden reason--Battery Bank maximum operating current (around C/8) and max starting surge (around C/2.5)--Need bank large enough to supply high current/power to start well pumps, refrigerator compressor) and run larger loads for periods of time (microwave, induction stove, etc.).

    For Li Ion, most of them seem to have a C/1 ability to charge/discharge and even (depending on cell/etc.) surge discharge rating... So--For a "normal cabin/home"--We almost don't even need to worry about current capacity as long as it is less than C/1 (one hour discharge rate). For homes/cabins and such, 5 hours in the evening for two nights is a 10 hour discharge rate--Pretty lazy for a Li Ion bank. (for a 2/50% FLA bank, that is a 20 hour (C/20) discharge rate).

    So--No 5%/10%/13% minimum charging rule for Li Ion--Vs Lead Acid which needs the spec. charging current for good life/operation).

    We are (mostly) left with how many days of storage do you want and how much sun do you have. With FLA batteries, 1 day of storage was not good for full time off grid--Because 1 day of "good sun" would not fully recharge the bank... And >3 days (and 50% max discharge)--Was a rather large bank, and possible self discharge issues (very large FLA battery bank needing more solar charging as the bank tends to self discharge--Older/Aged FLA storage batteries and especially Industrial/Traction/ForkLift batteries seem to have higher self discharge-->2% per day--Probably need a new bank).

    With Li Ion--A 1 day / 70% capacity bank will run what you want--Just need sun the next day, and possibly a genset boost. And it will recharge in 1 day (or even ~1 hour with enough panels) very nicely.

    Lets start with making some assumptions... 600 AH @ 12 volt bank, 25% discharge, and Albany NY "sun"...
    • 600 AH * 12 volt bank * 0.25 daily discharge * 0.85 AC inverter eff  = 1,530 Watt*Hours per day
    And looking at sun:

    http://www.solarelectricityhandbook.com/solar-irradiance.html

    Albany
    Average Solar Insolation figures

    Measured in kWh/m2/day onto a solar panel set at a 47° angle:
    (For best year-round performance)

    JanFebMarAprMayJun
    3.09
     
    3.89
     
    4.40
     
    4.46
     
    4.67
     
    4.92
     
    JulAugSepOctNovDec
    5.02
     
    4.91
     
    4.56
     
    3.88
     
    2.83
     
    2.73
     

    Guessing non-winter camp with 1,530 Watts of panels or ~4 hours a day of sun:
    • 1,080 Watt array * 0.52 off grid Lead Acid AC system eff = 2,246 WH of average harvest per day
    I will use the 1,800 WH number as your "minimum" needed energy per day (genset?). Sizing a Li Ion battery bank:
    • 0.81 derating for "hot panels in summer sun"
    • 0.95 derating for MPPT charge controller
    • 0.99 derating for Li Ion charging eff
    • 0.85 AC inverter efficiency
    • 0.81 * 0.95 * 0.99 * 0.85 = 0.65 "end to end" off grid Li Ion solar AC inverter system eff
    • 1,530 WH per day * 1/0.65 Li AC system eff * 1/4.0 hours minimum average per day = 588 Watt minimum 4 hour "break even" array
    So, your 1,080 Watt array is >> 588 Watt "break even" array--You took care of your batteries and they lasted you a long time (plus cool/cold weather helps).

    And sizing the Li Ion bank:
    • 1,530 WH per day * 1 day * 1/0.70 useful Li Ion battery capacity * 1/0.85 AC inverter eff * 1/12 volt battery bus = 214 AH @ 12 volt for 1 day storage
    • 1,530 WH per day * 2 day * 1/0.70 useful Li Ion battery capacity * 1/0.85 AC inverter eff * 1/12 volt battery bus = 429 AH @ 12 volt for 2 day storage
    • 1,530 WH per day * 3 day * 1/0.70 useful Li Ion battery capacity * 1/0.85 AC inverter eff * 1/12 volt battery bus =  643 AH @ 12 volt for 3 day storage
    The typical max charging current (cool, clear day, near solar noon):
    • 1,080 Watt array * 0.77 panel+controller derating * 1/14.5 volts charging = 57 Amps "more or less best case charging current"
    I have not given you "the answer" here... But you can figure out what is your desired "safety factor"... More storage/no genset. Less storage/use genset. Or simply do not need much power during cloudy weather...

    In general, we suggest using between 50% to 65% of "predicted harvest" for "base loads" (lighting, refrigerator) which have to run 24x7. And run "other loads/tools" during sunny weather.

    Based on your present system--Do you have an estimates for your daily loads? Sounds like it has been working well for you.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • couchsachraga
    couchsachraga Solar Expert Posts: 84 ✭✭
    Bill,  THANK YOU for taking all the time to write that up!  I managed to follow it the first time through (though I'll likely re-read it).  The system you guided be to years ago has indeed worked very well; I've only needed a generator twice in all these years, and both times in the dead of winter (it can be "a bit" cloudy up here at times...).   A good part of that has been relying on 12v appliances and getting the most efficient ones we could (within $ reason) - SunDanzer chest refrigerator is our largest constant draw, and that isn't much.  The biggest draw is the water pump, which is a 12v Shurflow RV style pump - it can draw up around 240 watts when it is almost up to tank pressure.  Thankfully that is a fairly short time.  We do use the place year round, so I did indeed try and over-panel it a bit and add more storage.  If everything was staying the same I'd just replace the batteries (perhaps with a single lithium battery admittedly), however there are a few things afoot that I can't quantify yet load wise:
    1). I'm hoping to get a Starlink for internet service when they open it up to our area.  Right now I use a cell amplifier which is nice... when it works.  I'm fairly certain the Starlink dish is going to draw more than my current amplifier.... but I don't know how much yet as there isn't a lot of information on them
    2). We've been using a battery ignited range for the past 10 years and may upgrade it to something that isn't as finicky.  When we purchased it we were warned against it for every day use, and I can see why.  We've gotten good at scrubbing pots though!  I'm hoping to find a range that only needs a little 110v AC to work rather than a glow bar needing 500 watts or more of continuous power.
    3). I'm adding a shop and will likely get an inverter to run 110 AC to it given the distance and voltage drop with 12v DC.  Having a bit more power available means less generator use.

    Given the Lithium's faster charging I may just swap one battery in and see how it goes for a year, then add another (and more panels) if needed.  Partially I'm wondering if between it being a new battery and faster charging if I'll have plenty of capacity anyway; as you noted if not I can always start the generator when needed.

    Going through this has given me a good "kick" to update my load spreadsheet and better understand how much excess capacity I have now.  In "normal" usage we cycle between 92-100 during sunny weather (even in winter) and upper 70's to 100 in with more clouds.  Amusingly this has lead me back to focussing on one of my questions, which there may not be an answer to - how much of a difference will the faster charging capabilities of a lithium battery make (vs FLA).  I guess I'll have to find out:)

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    If you need longer distances from battery bank to DC loads (or, I would suggest >800 AH @ 12 volts), 24 VDC might be a good step up for you... Many RV type pumps and DC fridges are available in 24 VDC or 12/24 VDC (for DC fridges).

    If you ever need to upscale to a full size fridge (roughly 1-1.5 kWH per day)--Then 120 VAC and a 600+ AH (at least for FLA) @ 24 VDC would be a good start (around 3,300 WH per day @ 120 VAC).

    Look real closely at your loads... Especially those that might run 24x7... I could not find a quick Cell Booster spec... But lets say it is 5 volts @ 1 amp vs StarLink (15-30+ Watts) and a Router @ 25 Watts (pure SWAGs):
    • 1 amp * 5 volts * 24 hours per day = 120 WH per day (Cell booster)
    • 30 Watts * 24 hours per day = 720 WH per day (15-30 Watts StarLink?)
    • 25 Watts * 24 hours per day = 600 WH per day (average router?)
    • 720+600=1,320 WH per day for (maybe 1,000-1,300 WH Per day) Starlink+Router
    1,320 WH per day (if running 24x7) is as much as a typical very efficient full size refrigerator (Energy Star) 120 VAC unit takes..
    The 24x7 Digital Stuff that is not battery powered (cell phone, etc.)--They can be pretty ugly energy hogs.

    The ranges--Many of the "home type" that don't have standing pilots seem to use the Glow Bar (around 400 Watts or so when the "flame is on")... I asked a repair guy once why this is so--His understanding was that people got nervous hearing the Click Click of spark igniters in ovens/dryers/etc... Spark Igniters (with flame sensor) do work fine on AC inverters--But they have to be PSW/TSW type (not sure MSW would work)--And the PSW/TSW has to have a grounded neutral (flame sensor detects "flame reification" of the sensor current when flame is present). My 10+ year old stove had the Neutral/Hot wired backwards and the flame sensing/spark was very erratic.

    The faster charging of the Lithium does a better job of solar harvest--Basically accepting all current from 20% to 90% SoC--Then full/stop.

    Lead Acid will accept full current to ~80-90% SoC, then hold 14.75 VDC and ramp current from 100% to less than 1% current (over the 2-6 hours)... With solar and declining afternoon sun--Not always obvious that this is a big deal in summer... But in Winter with sun in sky only a few hours and deeper discharge, can be a bigger issue.

    Lithium is great with generators/alternator type charging... You have a 400 AH battery and 40 Amp charging... Discharge to 20%:
    • 90%-20%= 70% capacity to charge
    • 400 AH * 0.70 = 280 AH to recharge
    • 280 AH / 40 Amp = 7 Hours to recharge
    If you have a 100 Amp charger:
    • 280 AH / 100 Amps = 2.8 Hours to charge
    Pretty much that simple. (solar being variable is much more difficult to predict/model simply).

    It sounds like you might be taking a step up to a 3-6x larger system (StarLink, add a shop AC and "real" refrigerator)?

    If so--Depending on what you have now--You might be looking at starting a new system (keep the present system for backup/guest house/gift to neighbor)...

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • mcgivor
    mcgivor Solar Expert Posts: 3,854 ✭✭✭✭✭✭
    Without going into extreme detail, here are some general guidelines. When making calculations for LFP, don't assume they will have far greater capacity without sacrificing life expectancy, if a given capacity with lead acid worked well in the past replace with the same capacity in LFP, there will be an increase in overall capacity available but use this as a buffer to promote extended life expectancy as well as occasional shortcomings in solar input.

    With regards to charging, if the array was sufficient to satisfy the LA bank, it will be more than capable of supporting the same capacity in LFP, given their inherent efficiency advantage, accepting full current available, without the need for an inefficient absorption stage. The typical 1C charging current allows for a larger array to compensate for inclement weather 

    Moving away from 12V is advisable for anything other than an extremely small system, the higher the better, to reduce current related voltage drop.

    Many manufacturers publish charging/discharg voltage based on maximum capacity, this is not necessarily the best approach, it's better to stay within the 20 to 95% window to reduce unnecessary stress on the cells.


    LFP have significant advantages over LA  with the exception of being able to perform in extreme low temperatures, using the correct  charge and discharge settings is critical  in achieving success. My advice is to study the chemistry to understand it before making any changes, for the most part they are entirely different from LA, so it's nessersary to wipe the slate clean and start from scratch, it's not difficult but don't correlate aquired knowledge fom LA and expect it to work out well.

    This link    http://nordkyndesign.com/assembling-a-lithium-iron-phosphate-marine-house-bank/   is helpful in explaining much of the information required, although focused on a DIY build, it explains what the manufacturer of a monoblock had to do, understanding this may seem trivial however it is, in my opinion, vital to understanding the details.

    After extensive study I made the switch to LFP  bank 3 years ago despite much negative feedback, it's a DIY bank and has been outstanding in comparison to the FLA bank it replaced, my motivation to switch was heat related, a tropical climate which  limit LA to 3 years tops before degregation sets in, personally I'd never look back given my circumstances.

    Overall, if low temperature is not a factor, you will be pleased with the results.

    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.
  • Ralph Day
    Ralph Day Solar Expert Posts: 1,019 ✭✭✭✭
    I noticed the Takagi's (on demand water heater) heat resistors too, and removed them before the first winter.  On the first -25C night the exchanger was full of slush, but not hard frozen.  There's only about 18 inches from the heater to the outside termination.  

    What I did was rig a 120v muffin fan to blow air through the heater constantly in the winter.  Not onto it, through the heat exchanger and out the exhaust stack.  Now after the coldest night the exhaust stack is room temperature.  An HVAC duct boot is screwed over the intake port on the heater cover and the fan blows into it.  Works great.  


  • couchsachraga
    couchsachraga Solar Expert Posts: 84 ✭✭
    Ralph - thank you for the idea - I may try that at some point.  When we use the place it is usually every weekend or every other weekend and draining it has worked well but on those cold nights when we are there I do keep an eye on it to see if we have an extra drain on the battery bank from it.
  • couchsachraga
    couchsachraga Solar Expert Posts: 84 ✭✭
    Thank you Bill and mcgivor as well for the extra comments and link!  On the one hand I don't plan on changing much (other than starlink), on the other if I go to 24v for the battery bank I'll need an inverter (right now I only have a SureSine 300w unit to run the electronics for the on demand heater and a small fan) and would likely change all my 12v bulbs out for 120v AC (all LED; right now I have a mix of LED and CFL)), and add a 24v to 12v converter for the existing fridge and water pump.

    To complicate matters part of the reason I'm contemplating this is moving my 4 panels from ground mount to roof mount (7/12 pitch standing seam; stays clear of snow).  IF I move the panels up there I will add at least 2 more.  Otherwise I'll leave them where they are and stick to 1080 watts of panels.

    The thoughts and sizing for the lithium batteries has helped tremendously - thank you all!!

  • BB.
    BB. Super Moderators, Administrators Posts: 33,431 admin
    Regarding LED lights... There are actually quite a number of 24 and 12/24 VDC lamps out there-If you want to stay with DC for lighting but go to 24 VDC battery bus:

    https://www.amazon.com/s?k=24+vdc+led+bulbs&ref=nb_sb_noss

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Ralph Day
    Ralph Day Solar Expert Posts: 1,019 ✭✭✭✭
    And just be sure to drain the exchanger completely.  A friend missed draining where the flow measuring wheel is housed and it fractured.  I think it's a large plastic thumb knob.