diy solar

diy solar

Does Micro Charging decrease Battery Life

Joined
Jul 1, 2021
Messages
6
Here is my setup: I have an Ampere-Time 24 volt 100 AH battery connected to a 3,000 watt 24 volt inverter. I use a 1,200 watt solar panel to charge the battery as needed via an MPPT charge controller.
  • Under normal conditions, the load on the inverter varies through out the day. When the load is low and the sky is sunny, the solar panels charge the batteries. When the load is high or clouds are passing over the solar array, the battery discharges through the inverter.
  • Under normal conditions the charge and discharge rates are below 20 amps. However, I have a circuit breaker in place to ensure that the load never exceeds battery-rated-capacity of 100 amps. I also limit maximum possible charge current so that it cannot exceed 60 amps.
  • Under normal conditions the system will draw a few watt-hours from the battery for a short time. Then when that load switches off, the solar panels restore those watt-hours to the battery. Normally the battery is operating above 95% state of charge because it never uses more that 5% charge in any given interval. This is what I call micro-charging.

Here is my question: The battery is rated for a life of 4,000 cycles where the depth of discharge is 90% between charges. Is the LiFePo battery limited to only 4,000 cycles for micro-charges in which the depth of discharge is only 5% between charges?

jim
 
  • Under normal conditions the charge and discharge rates are below 20 amps. However, I have a circuit breaker in place to ensure that the load never exceeds battery-rated-capacity of 100 amps.

What is the trip curve for your breaker?

The ones I use for most AC circuit will carry 50% over for 10 or 15 minutes. Thermal/magnetic so time to trip at moderate load varies with amount of overload. Fast trip occurs at 5x rating.

I'm not sure about the magnetic/hydraulic such as most DC breakers, but they still have a time delay.
 
This is what I meant by 100 amps: I put a breaker between the inverter and the load to limit the AC current pushed by 120 volts so that the battery current pushed by 24 volts never exceeds 100 DC AMPs. So I was talking about an AC breaker not a DC breaker. I use fuses on the DC side because in my experience DC breakers are a dreadful way to protect your system.

jim
 
This is a common question and concern when people see a batteries cycle life rating estimates. But luckily, "micro cycling" is not an issue at all. (There are quite a few terms that people use to describe this. I call it micro cycling)

Case in point is hybrid vehicles, which do hundreds or thousands of small cycles every time you drive it. If small cycles hurt these batteries, they would lose significant capacity in a few weeks. Which obviously does not occur. It will not hurt the rated cycle life, and most hybrid vehicle batteries (depending on chemistry/design/application) will last for decades even with 100'000's of cycles.

Considering the fact that most hybrids are using lithium ion packs (which are typically pouch cell lipoly which have an awful cycle life rating), I do not see this issue as a cause for concern. Especially when using packs with higher cycle life ratings and better thermal stability.

A "Cycle" constitutes charging to 100% and discharging to 0% at a specified temperature. When a pack is rated for 4000 cycles with a 100% DOD (as given in my example), you can typically pull exponentially more "micro cycles". But again, depends on chemistry and other factors.
 
Last edited:
I am also very interested in micro cycling of batteries, but I have not been able to find any conclusive studies that have done long term real world testing. I started typing this before Will's answer popped up. From what I have been able to find, I totally agree. I drive a Ford C-Max hybrid. I see the battery charge state gauge go from 70% to 40% many times on every drive. I would say they average more than 20 short cycles every day. I have been driving it for 8 years, and it has 180,000 miles on it, and the battery is still working just fine. It should do a bit over 2 miles of EV only mode with the small battery, but with a slight downhill slope, I can squeeze it to nearly 5 miles before it has to start the engine still. So I think the cells must still have over 80% capacity.

My JK BMS has a battery cycle counter. After watching it for a few months, it is only adding a cycle when I have charged and discharged the total battery pack amp hours. My battery is 360 amp hours, but I am only cycling about 140 amp hours a day, so it only adds a cycle every 2.5 days.

One thing I would suggest for your setup though is to adjust your charge controller down a bit. Keeping any lithium battery near it's maximum absorb voltage for a long time will degrade the cells and shorten the life. The cells are much more stable in the 50% to 80% state of charge area. So if you are only using a small amount of the battery capacity anyways, set the charge controller to stop bulk charging at only 3.4 volts per cell on LFP cells.
 
This is what I meant by 100 amps: I put a breaker between the inverter and the load to limit the AC current pushed by 120 volts so that the battery current pushed by 24 volts never exceeds 100 DC AMPs. So I was talking about an AC breaker not a DC breaker. I use fuses on the DC side because in my experience DC breakers are a dreadful way to protect your system.

jim

Same issue with trip time.

Ignoring inverter efficiency and battery voltage variations, 24V x 100A = 2400W. 2400W/120V = 20A
A 20A breaker in the AC circuit takes about 10 to 15 minutes to trip at 30A. I've used multiple space heaters to test that.
At over 100A, it trips instantly. I've also tested that.

A 20A breaker on AC side would let 150A be drawn from battery for 10 or 15 minutes, 300A to 400A for a much shorter time (maybe minute or two). Current draw of 500A from the battery is what it would take to trip instantly.

If you want to limit DC amps to 100A you would need a current sensor with fast-trip mechanism such as BMS. Maybe 200A should be allowed for a couple seconds, for starting surge of motors.
 
Considering the fact that most hybrids are using lithium ion packs (which are typically pouch cell lipoly which have an awful cycle life rating), I do not see this issue as a cause for concern for any chemistry we use for solar packs.

A "Cycle" constitutes charging to 100% and discharging to 0% at a specified temperature. When a pack is rated for 4000 cycles with a 100% DOD (as given in my example), you can typically pull exponentially more "micro cycles". But again, depends on chemistry and other factors.

This post has link to life testing of many batteries.
Only a couple reached the expected cycle life without failure.
(To me, all the more reason to DIY and be prepared to Repair It Yourself, too.)
Most of them were cycled relatively deeply, so doesn't show whether exponentially more micro cycles can be performed, or just up to the same total Ah of cycle life. (Obviously avoiding going deep into the charge/discharge knees allows longer life.)

 
Thanks Hedges. You comments support jim's first law: if it aint broke it will be. However, my inverter has saved my butt several times so far because it shuts down when an over-power condition exists. If not, my solar days would have been short and I would be perusing some other hobby right now :)

jim
 
Does the "micro-cycling" occur at a float voltage, or does it sweep back through a bulk/absorption cycle at higher than 3.45V/cell?
 
This post has link to life testing of many batteries.
Only a couple reached the expected cycle life without failure.
(To me, all the more reason to DIY and be prepared to Repair It Yourself, too.)
Most of them were cycled relatively deeply, so doesn't show whether exponentially more micro cycles can be performed, or just up to the same total Ah of cycle life. (Obviously avoiding going deep into the charge/discharge knees allows longer life.)


Disregard those test centre results if you are interested in actual battery lifespan testing.

It is chemically impossible to extrapolate accelerated current/temperature tests.

It may still be decades before some SEI formation techniques and things like AC charge ripple / micro cycling can be given definitive effects on lifespan.

There is a lot of testing underway, but the main problem is that they need to be tested in real time. I know of testing that began in 2007 - still ongoing.
 
Disregard those test centre results if you are interested in actual battery lifespan testing.

It is chemically impossible to extrapolate accelerated current/temperature tests.

What was extrapolated or accelerated?
Some batteries like Sony were cycled 6kWH to 8 kWh per cycle, "Capacity appears to have decreased linearly over time with a SOH of ~84% after ~2,965 cycles apparent"

Sure, for tests that weren't yet completed, they made an attempt at extrapolating/projecting where it was going. At 3 cycles per day it will be there after just a few years.

What wasn't extrapolated was that the vast majority failed before completion of testing.
The take-away for me was "don't expect the battery to last as advertised", except for the handful of brands/models which did.

"While many battery packs have experienced faults and/or failed prematurely, the Sony battery pack from Phase 1 has proven highly reliable to date, alongside the Pylontech and GNB Lithium battery packs from Phase 2. The Sony battery pack (Phase 1) has retained over 80% of its initial capacity after nearly 3,000 cycles. The Pylontech battery pack (Phase 2) has also retained over 80% of its initial capacity after nearly 2,000 cycles."
 

Mike
 
@mikefitz beat me to it! ;-)

Micro-cycling might possibly cause some problems, but it's a weird edge case (for example when lithium is used as a stand-by battery with a charger that holds them at a lower capacity). From the wiki entry:

Lithium has a small memory effect that builds up with micro-cycles over a small voltage range.

During the course of a day the charging/discharging isn't enough to impact them if they reach full charge. But, if you don't commonly use your Batteries and the BMS upper limit is 90%, then they might be microcycling around that value and that can strand capacity.
 
As I understand it, the claimed memory effects only manifest when cells are NOT fully charged. If anything, this situation would seem to minimize the chances of memory effect via multiple charges to full per day.
 
What's described is similar to an issue in the public-safety sector regarding 1st-responder radios with improper charging - regardless of chemistry, li-ion or even older nimh:

Motorola instructs users to turn OFF their radios when placed back in charger, as the small load presented when they leave them running in the charger, prevents the radios from ever reaching a full charge, and held at an endless *attempt* to reach the designed CV voltage, and never reaching it.

Yet degradation still occurs! Because even when not reaching the designed CV voltage because of the radio's "on condition" if the loaded voltage is not too low, given enough time - a full 100% SOC can still occur! And instead of shutting off waiting for the designed CV to be reached, a condition of 100% SOC - at a lower voltage, is held forever sitting in their charger stands. (1st responder lurkers - turn your radios OFF when charging - thanks, and don't use them as stands.)

Using the charger as a "holder" while in use is the biggest tempation to avoid, but it's done anyway. It has taken a decade or more to have the $marter chargers watch for a constant-current load while under the designated CV and take action, but that's another forum.

Long story short - the randomness of what you are doing with your bank should not be an issue, although personally I like to avoid extremes. Let it fully charge once in a blue moon without a load. But that's just me.
 
I agree micro cycling probably causes wear on your batteries and decreases cycle life. How bad is it, I haven't seen tests, so I can only say "probably not very bad". Will gives a great example with EVs, many of them "hide" the very top end of the charge range (as well as the bottom). So, it is much less likely to be a problem if you do as the manufacturer recommends, and avoid the top 10% and bottom 10% of the battery cycle. Try using 3.4v as the top of your charge, and then float at 3.35v. 27.3v and 26.8v. Might even float a bit lower, but that should substantially reduce the micro cycling.
 
This does make me wonder though - how many have balanced or tried to commision their banks the first cycle out with a load on during charge?

I've always assumed a no-load condition when initially balancing or commissioning, but I may be wrong in some cases.
 
Back
Top