Realistic Cycle life?

[Jonathan86]

Are there really thousands of cycles difference between batteries or is it just marketing bs? I'm seeing some server rack batteries with 4000 cycle life, up to 7000+ cycle life at the same DOD% and C rate, and then the new eg4 power pros listed as high as 8000 cycles, and yet they're one of the least expensive batteries, with heaters. From what I've found, all or most seem to be using the same cells. Has anyone seen a significant difference in reliability between battery brands? or anything else to make one an obvious choice over another? I'm building a new system and having trouble deciding on which battery to go with.

Also, how important is communication? I see some say they have communication and some don't and then some specifically have closed loop communication. I'm not sure if it's necessary to spend hundreds of dollars more per battery for the communication, for example, between the two SOK server rack batteries, or if I can use some other device to get the communication feature? I'm building a Victron system.

Thank you.

 

[OffGridInTheCity]

I just completed cycle 1,936 on my oldest 18650 14s battery and no sign of degradation - which is nearly 6 years of operation. Lifetime average of 35.1% DOD. LifePo4 chemistries are reported to have more cycles at 80% DOD than 18650 INR chemistries.

What's important in my mind is a *low stress* operation (not communications necessarily). A properly sized battery bank for mild charge/discharge C levels, mild DOD, and mild ambient temps/humidity should lead to the best life span. Based on my experience so far, I'm inclined to believe 7,000 cycles for LifePo4 is possible.

But of course it takes a long time (on the order of 21yrs) for actual proof of 7,000 cycles at 335 cycles/year. May enough of us be on this forum long enough to prove or disprove the lifespan numbers

 

[timselectric]

I don't think that I will live long enough to know how many cycles my batteries last.

 

[NCislander]

If I can get 10 years of usage, I'm good. Money & Efforts well spent.

 

[Jonathan86]

If I can get 10 years of usage, I'm good. Money & Efforts well spent.

I got twelve years out of my lead acid. Going to spend twice as much, 12 years later, on new lithium. 4k cycles would be about ten years, 8k about twenty. If I could get ones that are still very usable after ten years, that would be great.

 

[Jonathan86]

I just completed cycle 1,936 on my oldest 18650 14s battery and no sign of degradation - which is nearly 6 years of operation. Lifetime average of 35.1% DOD. LifePo4 chemistries are reported to have more cycles at 80% DOD than 18650 INR chemistries.

What's important in my mind is a *low stress* operation (not communications necessarily). A properly sized battery bank for mild charge/discharge C levels, mild DOD, and mild ambient temps/humidity should lead to the best life span. Based on my experience so far, I'm inclined to believe 7,000 cycles for LifePo4 is possible.

But of course it takes a long time (on the order of 21yrs) for actual proof of 7,000 cycles at 335 cycles/year. May enough of us be on this forum long enough to prove or disprove the lifespan numbers

Have you been completely off grid the last six years?

What exactly do you mean by low stress? The rate of discharge? The depth of discharge? Even though most brands say not to go beyond 80% DOD, is it much better on the batteries to only go down, say 50% or 25%? Would it actually be better overall to buy enough batteries to always keep the SOC high?

 

[ScrotusGobbleBottom]

A properly sized battery bank for mild charge/discharge C levels, mild DOD, and mild ambient temps/humidity should lead to the best life span.

This! Keep your C rate at .2 or below on average, dont go under 20%, charge to no more than 3.5 / cell and give them a loving home.

 

[Rain@]

I think different batteries have different cycle times. For EVE cells, the number of cycles of 280k is usually more than 6,000 (according to marked in the specification written). I learned that HITHIUM 280Ah has more than 7000 cycles, which is higher than EVE 280K, but the price is lower than 280K. I guess it may be that HITHIUM 280ah may have lost or weakened some performance, and then strengthened the cycles. And then this high cycles become its promotional feature.

 

[rhino]

I think different batteries have different cycle times. For EVE cells, the number of cycles of 280k is usually more than 6,000 (according to marked in the specification written). I learned that HITHIUM 280Ah has more than 7000 cycles, which is higher than EVE 280K, but the price is lower than 280K. I guess it may be that HITHIUM 280ah may have lost or weakened some performance, and then strengthened the cycles. And then this high cycles become its promotional feature.

I don't really trust comparing cycle counts between battery companies since there really isn't a standard they follow. Even between different EVE cells they use different testing methods making it hard to compare. On top of that a properly sized solar installation will typically not stress the batteries nearly as much as the test methods do.

 

[sunshine_eggo]

Cycle counts are bullshit.

 

[Brucey]

I just completed cycle 1,936 on my oldest 18650 14s battery and no sign of degradation - which is nearly 6 years of operation. Lifetime average of 35.1% DOD. LifePo4 chemistries are reported to have more cycles at 80% DOD than 18650 INR chemistries.

What's important in my mind is a *low stress* operation (not communications necessarily). A properly sized battery bank for mild charge/discharge C levels, mild DOD, and mild ambient temps/humidity should lead to the best life span. Based on my experience so far, I'm inclined to believe 7,000 cycles for LifePo4 is possible.

But of course it takes a long time (on the order of 21yrs) for actual proof of 7,000 cycles at 335 cycles/year. May enough of us be on this forum long enough to prove or disprove the lifespan numbers

So rounding up to 2000 cycles and 35% DOD around 700 full cycle equivalents?

 

[Jonathan86]

So rounding up to 2000 cycles and 35% DOD around 700 full cycle equivalents?

Is that really how it works? I was told by a tech support guy at a solar retailer that every time the batteries discharge, no matter how far down, and then get recharged, that's a cycle.

 

[AshleyL]

Is that really how it works? I was told by a tech support guy at a solar retailer that every time the batteries discharge, no matter how far down, and then get recharged, that's a cycle.

100% to 0% = one cycle

100% to 50%, recharged to 100%, then used down to 50% again = one cycle

 

[Jonathan86]

This! Keep your C rate at .2 or below on average

This means don't discharge faster than 20% of my total battery bank AH, right? So, if I need to discharge faster, it's best to buy more batteries? and the cost of more would be offset by the total bank lasting longer?

 

[Jonathan86]

100% to 0% = one cycle

100% to 50%, recharged to 100%, then used down to 50% again = one cycle

Wow, ok, thank you very much. That makes sense.

 

[Brucey]

Is that really how it works? I was told by a tech support guy at a solar retailer that every time the batteries discharge, no matter how far down, and then get recharged, that's a cycle.

Yeah they told you some nonsense there.

 

[Jonathan86]

Yeah they told you some nonsense there.

I haven't received much help and the answers I have received didn't leave me feeling very confident. I've called three different companies. They all basically told me I need to figure it out and then call them back to tell them what I want to buy. I was really hoping the company I'm planning to spend $30-40k with, was going to put together a system for me. Apparently I was wrong. I'm glad I found this site, as it's helping me understand a lot. The company I bought my first system from, 12 years ago, put everything together for me. Unfortunately, they don't seem to be in business anymore.

 

[mciholas]

Is that really how it works? I was told by a tech support guy at a solar retailer that every time the batteries discharge, no matter how far down, and then get recharged, that's a cycle.

If that was true, regen on EVs would kill the battery in about one thousand miles since every stop light or stop sign is a charge followed by a discharge.

The only polite thing to say about that tech support guy is that he isn't in engineering.

Mike C.

 

[Brucey]

I haven't received much help and the answers I have received didn't leave me feeling very confident. I've called three different companies. They all basically told me I need to figure it out and then call them back to tell them what I want to buy. I was really hoping the company I'm planning to spend $30-40k with, was going to put together a system for me. Apparently I was wrong. I'm glad I found this site, as it's helping me understand a lot. The company I bought my first system from, 12 years ago, put everything together for me. Unfortunately, they don't seem to be in business anymore.

I would suggest Current Connected if you are still looking, 10 year warranty on Victron from them if you go that path and they sell eg4 as well if that's your fancy.

 

[OffGridInTheCity]

Have you been completely off grid the last six years?

I'm off-grid with grid-assist via a set of ATSs - but yes. For the last 6 years, the system has generated 15,000kwh/year of consumed power.

What exactly do you mean by low stress? The rate of discharge? The depth of discharge?

Both. The larger the battery relative to the load means that the C rate is less per cell. In my case, it's pretty common to think of a laptop type 18650 cell with 2600mah capacity as doing 500ma (0.5a) or ~0.2C as 'reasonable' for sustained operations. My powerwall is large enough I have it down to <150ma (0.15a) / cell for both charge and discharge which is ~0.06C - e.g. low stress.

Even though most brands say not to go beyond 80% DOD, is it much better on the batteries to only go down, say 50% or 25%?

I'm 'lithium-ion INR chemistry' which is ~2.8v -> 4.2v range and this is the famous Battery University chart I designed against... showing SoC ranges vs cycles. I'm operating roughly between purple (75-45% SoC) and blue-green (75-45% SoC) and literally shooting for 7,000+ cycles and hoping for 85%+ original capacity remaining. The chart below is found on this web page - https://batteryuniversity.com/article/bu-808-how-to-prolong-lithium-based-batteries


Haven't researched LifePo4 as much but I assume that 'in general' the same principles apply. The problem with LifePo4 is the shallow charge/discharge curve so it's impossible to do 'fine tuning' SoC operation by voltage alone whereas with 18650 it's not.

Would it actually be better overall to buy enough batteries to always keep the SOC high?

It's not about keeping SoC high - its about using the middle portion of the battery's charge/discharge range. Avoid the extremes for lower stress. Look at the chart above and you can see 100-40% Red line (60% DOD) is predicting significantly less cycles that the same 60% DOD between 85-25% SoC Green line to illustrate my point.