Let's talk Lithium Ion Chemistries/Physics and Implications

Started by bevans, May 04, 2019, 04:00:16 PM

Previous topic - Next topic

mike90045

If you can stand the higher losses, I've been very happy with my NiFe bank (Edison Alkaline Nickel-Iron) and it's fairly bulletproof.   It does live in the battery shed, but only because the battery shed was built before the house.
http://tinyurl.com/LMR-Solar

Classic 200| 2Kw PV, 160Voc | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph )| Listeroid 6/1, st5 gen head | XW6048 inverter/chgr | midnight ePanel & 4 SPDs | 48V, 800A NiFe battery bank | MS-TS-MPPT60 w/3Kw PV

Ron Swanson

Quote from: RossW on May 05, 2019, 05:21:22 PM
Quote from: Ron Swanson on May 04, 2019, 10:07:26 PM
Somebody prove me wrong, I want to like them but the dollars and cents just don't seem to add up.

On cost, when you add up the cost of new lead-acid technology, plus ALL the other stuff - allowing a maximum DoD of 25% or so

This is where I hang up.  We specify industrial FLA banks for 50% DoD.  They last over 10 years like this.  AGM, L16s etc cannot handle this but industrial seem to do fine.

This really tips the cost equation back to FLA.  Time value of money, it is cheaper to pay someone to come look at water levels monthly than to buy Li.

I assume you made a DIY Li battery system though.  In my case I am trying to figure it on cost delivered by a professional to a customer who wants it to work many years with a warranty.  Which means either battery bank needs to have a big manufacturer behind it.

Please post some dollars and cents if you are able.  Last time I looked FLA at 50% vs Li at 100%, Li was still like 3x as costly total cost of job.  I will get some figgers tomorrow.

bevans

Quote from: mike90045 on May 05, 2019, 06:08:33 PM
If you can stand the higher losses, I've been very happy with my NiFe bank (Edison Alkaline Nickel-Iron) and it's fairly bulletproof.   It does live in the battery shed, but only because the battery shed was built before the house.

Mike, what kind of losses are you talking about - what has been your experience with the Iron Edisons? Cost is about the same as lithium, but I am curious. With their storied history and long life it's a surprise they don't hold a bigger share of the market.
5200w PV array, 16 SunPower 327s
Magnum Energy ms4448PAE inverter/power center
Magnum BMK
Dual Midnite Solar Classic 150s
Midnite WhizBang Jr & shunt
16 kwh Chevy Volt EV lithium ion

bevans

Back to the Volt batteries' LMO-NMC chemistry... If I am able to have custom BMSs made specifically for the Volt batteries and their charging parameters, what glaring challenges remain to achieving a smooth interface with the rest of the system (classics, WBjr, BMK and Magnum 4448pae)? Or is that the main hurdle?

Quote from: RossW on May 05, 2019, 05:21:22 PM
discrete per-cell active balancers (charge-shifting type, not just big fat zener diodes).

RossW, can you explain that difference (charge shifting vs zener diodes)... were these cell balancers a part of the BMS that was factory integrated into your battery bank? What LFPs did you get?
5200w PV array, 16 SunPower 327s
Magnum Energy ms4448PAE inverter/power center
Magnum BMK
Dual Midnite Solar Classic 150s
Midnite WhizBang Jr & shunt
16 kwh Chevy Volt EV lithium ion

RossW

Quote from: bevans on May 06, 2019, 01:03:23 AM
RossW, can you explain that difference (charge shifting vs zener diodes)... were these cell balancers a part of the BMS that was factory integrated into your battery bank? What LFPs did you get?

The common, inexpensive "balancers" seem to be basically a big, precision zener. When the voltage reaches their setpoint, they start to conduct power. So when the whole string is under charge, as each cell reaches its "full" voltage, the balancer starts shunting current past it to stop the cell overcharging.  The problems with that are that firstly, they won't necessarily shunt the full current past - so if you're charging the bank with more current than the balancer can shunt around the cell, you will still end up over-charging the cell. The second thing is, in off-grid use, you may not get to full charge for an extended period, particularly during marginal weather. And if you're "not full" but "holding on" for days or weeks, charging up to lets say 75%, then discharging down to say 40%, day after day... your cells will never get balanced, and in fact will be getting more and more OUT of balance.

The devices I got have one balancer module per cell and an AC bus between all the balancer modules. They strive, constantly, to keep the cells within a few millivolts of each other. Any cell(s) above average voltage will "donate" power onto the bus, and any cell(s) below the average will "take" power from the bus. This happens when the cells are charging, discharging or just sitting quietly by. It is typically a low current - only a few amps at most - but it's happening ALL the time, so the cells don't get out of balance in the first place. I've never caught mine more than about 12mV difference across an entire pack.

I have one pack of 16 series-connected 300AH prismatic LFP cells, and a second pack of 16 series-connected groups of 3-parallel 200AH prismatic LFP cells. I only charge them to a maximum of 55.5V (3.468V/cell). As soon as the sun goes and the surface charge is knocked down, they settle to 53.0V (3.31V/cell) and after a full night of running all the computers, water pump, lights, TVs, etc, they're usually at 52.4V (3.275V/cell). The generator is set to start if they get below 35% SoC, or down to 48V (3.000V/cell) and the inverter will shut down a little below that. I've had the generator start a few times on SoC, but never on low volts.
3600W on 6 tracking arrays.
7200W on 2 fixed array.
Midnite Classic 150
Outback Flexmax FM80
16 x LiFePO4 600AH cells
16 x LiFePO4 300AH cells
Selectronics SP-PRO 481 5kW inverter
Fronius 6kW AC coupled inverter
Home-brew 4-cyl propane powered 14kVa genset
2kW wind turbine

tecnodave

Ross,

Please provide link to your cell balancer, I need to read more on this approach......Thanks so much for the high quality information that you provide....
David
#1 Classic 150 12 x Sharp NE-170, 2S6P, 24volt L-16 Rolls-Surette S-530, MS4024 & Cotek ,  C-40 dirv.cont. for hot water
#2 Classic 150 12 x Sharp NE-170, 2S6P, 24 volt L-16 Interstate,Brutus Inv.
#3 Kid/WBjr 4/6 Sanyo 200 watt multilayer 4/6 P
#4 Kid/WBjr 4/6 Sanyo 200 watt multilayer 2S 2/3 P

mike90045

Quote from: bevans on May 05, 2019, 08:16:19 PM
Quote from: mike90045 on May 05, 2019, 06:08:33 PM
If you can stand the higher losses, I've been very happy with my NiFe bank (Edison Alkaline Nickel-Iron) and it's fairly bulletproof.   It does live in the battery shed, but only because the battery shed was built before the house.

Mike, what kind of losses are you talking about - what has been your experience with the Iron Edisons? Cost is about the same as lithium, but I am curious. With their storied history and long life it's a surprise they don't hold a bigger share of the market.

A couple % overnight loss
Wide operating voltage range from 50% to Absorb cycle while charging, not all inverters can handle w/out faulting
about 70% recharge efficiency
consumes lots of water
http://tinyurl.com/LMR-Solar

Classic 200| 2Kw PV, 160Voc | Grundfos 10 SO5-9 with 3 wire Franklin Electric motor (1/2hp 240V 1ph )| Listeroid 6/1, st5 gen head | XW6048 inverter/chgr | midnight ePanel & 4 SPDs | 48V, 800A NiFe battery bank | MS-TS-MPPT60 w/3Kw PV

bevans

Quote from: tecnodave on May 07, 2019, 12:31:21 PM
Ross,

Please provide link to your cell balancer, I need to read more on this approach......Thanks so much for the high quality information that you provide....
David

Ross, I second that. Awesome info. Very interested to see what BMS you’re running.This is helping me understand the function and process of a BMS, and give me a healthy respect for how little I know. Do you have further recommendations regarding what to look for in a BMS, or manufacturers/models that have not yet been mentioned in this thread? (Specifically for LG chem cells - lmo-nmc chemistries, if possible)
5200w PV array, 16 SunPower 327s
Magnum Energy ms4448PAE inverter/power center
Magnum BMK
Dual Midnite Solar Classic 150s
Midnite WhizBang Jr & shunt
16 kwh Chevy Volt EV lithium ion

RossW

Sorry for the slow response - been extra busy with work and other projects!

Firstly, as has been mentioned many times, "BMS" is a confusing term.
Some people interpret it as a "Battery Management System" while others think "Battery Monitoring System".

So I am using 16 discrete cell-balancing modules on each of my two battery banks.
They're similar to these: http://tinyurl.com/y5yhzcyr and are available with or without the LEDs.
I have one bank with them, and one without. They perform the same. The LEDs let you see at a glance if any are taking from or donating to the rest. The downside is a small extra current draw.  Not that 16 LEDs is much in the greater scheme of things!

Additional to the balancing system I have a cell MONITORING system. It gives me confidence the BALANCING system is doing its job. There's a rough page describing it at http://support.rossw.net/batterymonitor/

The balancers are available in various nominal voltages, they have 2V units for lead-acid, and 3V (or so) for Lithium.
I don't believe there is any significant difference between them except perhaps the current limiting resistors on the LEDs :)

I spent a good while researching them and they appear to be basically a free-running bi-directional inverter. The turns ratio of the transformer is the only critical part. It looks like they'll all synchronise as soon as there's anything on the AC bus, and by virtue of the H-bridge, if the AC side is higher than the DC side (taking turns ratio into account), then power flows into the cell. No smarts, no fancy communications or calibration. Just plain, simple, efficient medium-power stuff with very little to go wrong.

3600W on 6 tracking arrays.
7200W on 2 fixed array.
Midnite Classic 150
Outback Flexmax FM80
16 x LiFePO4 600AH cells
16 x LiFePO4 300AH cells
Selectronics SP-PRO 481 5kW inverter
Fronius 6kW AC coupled inverter
Home-brew 4-cyl propane powered 14kVa genset
2kW wind turbine

myercreek

After 30 years of using Lead Acid batteries (L16s) for our small home, we took the plunge and replaced the last failing FLAs with 200 amp Lithium Ion batteries from Re-Lion in South Carolina. At my age I had no desire to think about moving 120 pound batteries again in 6 years or so. The new batteries have been in use for 4 months and we are thrilled - no maintenance, no bubbling, no odor - and only 62 pounds each instead of 120 pounds. My only concerns have to do with monitoring the system to know when to use the generator to give them a boost after several days of inversion with little solar gain. (Location is east slopes of the Cascades). Even after several days, the voltage readout goes down to only 12.7 or 12.8 from 13.1, the SOC to 82%. With my FLAs I knew it would take 0.2Kwh gain to raise the SOC 1%. Now the SOC jumps from 87% to 100% with only 0.1 Kwh gain and the V at 12.8 or 12.9. The settings on the Midnite Classic are set such that the generator produces only a float charge to the batteries, yet an hour of generator time bumps the SOC to 100%. My conclusion is to pay more attention to the Volt reading and less to the SOC readings. Suggestions?

boB

Quote from: myercreek on December 07, 2019, 08:21:03 PM
Even after several days, the voltage readout goes down to only 12.7 or 12.8 from 13.1, the SOC to 82%. With my FLAs I knew it would take 0.2Kwh gain to raise the SOC 1%. Now the SOC jumps from 87% to 100% with only 0.1 Kwh gain and the V at 12.8 or 12.9. The settings on the Midnite Classic are set such that the generator produces only a float charge to the batteries, yet an hour of generator time bumps the SOC to 100%. My conclusion is to pay more attention to the Volt reading and less to the SOC readings. Suggestions?

When the Classic goes from Absorb to Float, assuming the Absorb time has been met at or above the Absorb set point voltage, the Classic will reset the SOC to 100%.   That is mainly done for lead acid batteries. 

If Absorb does not terminate to Float, then the amp-hours alone, into the batteries, taking amp-hour efficiency into account, will bring the SOC to 100%.

Is it that the generator is running when the Classic goes from Absorb to Float ?   For lithium batteries, you would most likely have a very short Absorb time, is that correct ?  If so, then the SOC would easily jump to 100% after the Absorb voltage is reached.

Please add to this if you like.




K7IQ 🌛  He/She/Me

Steve_S

Relion has a Battery Monitor module available for their packs.  It uses a separate shunt for readings.  Here is a YT Video on it:  https://www.youtube.com/watch?v=LUilz14vkzY  it's one option.