Design parameters for lithium ion batteries

alexj

Hi all,

I am trying to get a better understanding of energy storage system design with lithium ion batteries. I work in developing countries where lithium ion batteries are still very uncommon outside of plug-and-play PV systems, but are likely to become more prevalent in the coming years. I only have experience with BMZ and Sonnen from years ago in the states. A few questions that I hope might lead to an interesting discussion:

1. It seems like LiFePo is the preferred chemistry for off-grid applications. It seems like Discover and SimpliPhi are popular and produce a good product. Are there other recommended brands? Any thoughts for what would be the most appropriate for smaller, very remote installations in the developing world?

2. Is anyone using NMC batteries here? If so, what brand?

3. Are there any other lithium-based chemistries that are currently commonly used for off-grid applications or will be in the future? Any that might be particularly interesting for developing country applications?

4. Is it possible to develop design general guidelines for energy storage system sizing with lithium ion batteries like has been done with lead acid batteries? Lead acid batteries definitely vary in their parameters by manufacturer, but there are some guidelines (DoD/cycles and charging/discharging current) that seem to hold reasonably consistent for AGM/Gel/FLA that can be used for design purposes.

Could this be done with the major lithium chemistries or do you see them as varying too much between manufacturer? For example:

LiFePo:

• SimpliPhi recommends a maximum C/2 charge and discharge rate. Claims a useable DoD of 100%. Recommend an 80% DoD for best cycle life. Claims around 98% roundtrip efficiency.
  
• Discover recommends a maximum C/1 charge and discharge rate. Claims a useable DoD of 90%. Claims around 98% roundtrip efficiency.
• Blue Planet Energy recommends a maximum C/2 discharge rate. Claims a useable DoD of 100%. Claims around 98% roundtrip efficiency.

Would it work to use general sizing parameters for LiFePo like:

• Maximum charge/discharge of C/2. (Is this reasonably consistent?)
  
• Recommended DoD 80% (Would provide a storage buffer. Does this tend to provide best LCOE?)
  
• Roundtrip efficiency 98% (Should this be more conservative?)
• Minimum charging temperature 0C.

Any/all thoughts welcome!

mike95490

1) Lead acid batteries are forgiving, and generally "die" gracefully

2) Li batteries are unforgiving.   1 overcharge or 1 deep discharge, and they are toast.   Might get a few more cycles out of them before they die, but either condition damages the battery.

3) LFP / LiFePo / LP4 (all the same battery, just different abbreviations) are the "safer" batteries that tend to just fail, or swell up and stink.   Any of the other Li chemistry mixes, involve failure with fire or going bang.

4) Some Mfg's include the top and bottom 10% safety margins internal their BMS systems. 
 Some don't and you have to program that into your loads and chargers.  Your Mileage May Vary. Some rate the full capacity, but only deliver 70% of capacity.
 Lead acid batteries like 50% -100%, so same protocols will NOT work for each.

mcgivor

1.  Each lithium type has its advantages and disadvantages, mostly related to energy density, specific power, life expectancy, performance, cost and safety. LiFePo4 is more of a compromise, it isn't the lightest nor can it delivers the current some others can however they are one of the safest, which is why they are a popular choice. The pre-assembled batteries or battery systems are roughly double the price of building a bank using prysmatic cells, some system can be integrated with other equipment, inverters etcetera, able to communicate information from the BMS regarding state of charge.

2. There are some I believe who are using NMC in the form of Nissan leaf batteries 

3. Generally a battery is chosen for a specific application, LIPO for example is extremely light with high performance capabilities, which is why they are the choice for drones, they are however dangerous, since weight is not as important but safety is makes LiFePo4 an ideal choice for offgrid. There are variations even within tha LFP made to enhance certain aspects of performance, according to some articles I've read. Not sure there is anything specifically available for developing countries but one thing is known, lithium batteries do not perform in temperatures below 0°C.

4. There are some manufacturers of drop in replacement type LFP who advertise that their battery can be discharged to 100% of the listed capacity, but in fact the actual capacity is greater, the BMS cuts the load when ~20% to protect against damage. Simpliphi has different warranty periods for differing charge /discharge levels, it is good practice when building a DIY bank to keep within certain parameters, commonly 90% and 20% of capacity. This extends the cycle count expectancy, programmable BMS's are available which can be set to whatever value is desired, but manufacturers will be cautious to preserve their reputation, one would hope. 

alexj

mike95490 said:

1) Lead acid batteries are forgiving, and generally "die" gracefully

2) Li batteries are unforgiving.   1 overcharge or 1 deep discharge, and they are toast.   Might get a few more cycles out of them before they die, but either condition damages the battery.

3) LFP / LiFePo / LP4 (all the same battery, just different abbreviations) are the "safer" batteries that tend to just fail, or swell up and stink.   Any of the other Li chemistry mixes, involve failure with fire or going bang.

4) Some Mfg's include the top and bottom 10% safety margins internal their BMS systems. 
 Some don't and you have to program that into your loads and chargers.  Your Mileage May Vary. Some rate the full capacity, but only deliver 70% of capacity.
 Lead acid batteries like 50% -100%, so same protocols will NOT work for each.

I was watching some training videos from SimpliPhi and they recognized using too high of a charge/discharge rate as the most common failure point in their batteries. It seems like this can be mitigated with proper programming/sizing, just like with a lead acid battery bank, although I guess the cost is far higher if one gets anything wrong or if the parameters were to be cleared for some reason. Lithium ion batteries seem conceptually attractive to me in the developing world as it seems like a system, if properly installed/programmed, can offer many years free of maintenance and issues..although has pretty much been my experience with AGM batteries down here as well. There are many off-grid locations that can take days and significant cost to arrive at in the jungle here..smaller batteries, less weight, more cycles...seems pretty attractive. The cost is still a serious barrier, but I imagine that in the coming years that the BMSs and technology will become more foolproof/refined and continue dropping in price.

Dave Angelini

Keep your fingers crossed on the the BMS getting better! There are electronics inside that can be zapped and made useless. The cells being damaged are another huge concern at that cost. I use both lead acid and LFP for my clients. The best of both worlds !
 Wired up to their own inverter/charger,  the ultimate AC bypass offgrid.

I was involved in the failed LG Chem NMC 48V battery for Schneider XW inverters. It was like the Simply phi in that it was easily damaged because of high charge and discharge currents. Special design steps need to be used in large power systems. Beware of  lithium batteries less than 7 KWH unless you are fully up to speed!

Always add in the damage to LFP charging below 32F. I have new customers that blew out 10K$ batteries before we solved their problems. Just one time as Mike said !   Logged in memory for warranty.      For some reason that circuit always survives😉

alexj

It seems like Trojan may have designed a significantly better BMS with their Trillium batteries? It appears to have built-in over-current protection with a contactor to be able to open the circuit as needed based upon voltage/current. Any thoughts?

https://www.trojanbattery.com/pdf/datasheets/27_Lithium_12.8-110.pdf

Dave Angelini

Plenty of companies doing that besides Trojan. 

70 of them in the store here.
https://www.solar-electric.com/residential/batteries-battery-storage/deep-cycle-batteries.html?nav_battery_type=537

To me, it is critical that the battery and the power system communicate closed loop. Not too many doing this. Some say they do and don't also.

alexj

Dave Angelini said:

Plenty of companies doing that besides Trojan. 

70 of them in the store here.
https://www.solar-electric.com/residential/batteries-battery-storage/deep-cycle-batteries.html?nav_battery_type=537

To me, it is critical that the battery and the power system communicate closed loop. Not too many doing this. Some say they do and don't also.

Why does overdischarge (resulting from too high of a current draw) continue to damage/destroy Lion batteries if all of them have an integrated current sensor/contactor? I could see how communication with the inverter would create the best user experience - meaning that the inverter wouldn't nuisance trip regularly to protect the batteries - but is it necessary for this protection mechanism to work properly and prevent damage from overdischarge?

Dave Angelini

Over discharge is one parameter and over current is another in charging or discharging.

Most decent LFP battery systems and be run down to lights out for the inverter and still be charged the next day or from genset. The BMS cuts out before the cells are damaged or unbalanced. We do not do lights out offgrid. Ever !  We don't here in 30 years.

Over current happens with small banks of LFP and is a one time damage event. Instant damage, loss of capacity, blown internal fuse, contactor damage, and logged for warranty by many battery systems. The transients of current from loads, clouds, and big inverter/chargers are too fast for the BMS to protect a small battery.

Closing the loop in a power system with the BMS may still not be fast enough in a small battery system. The LFP choice is expensive compared to LA batteries. People go too small in capacity. They then defeat the long life and other attributes of LFP.   Hope this is helpful

alexj

Hi Dave,

Definitely helpful. Your second two paragraphs get to the heart of what I was trying to figure out: whether properly protecting the batteries from overcurrent events is a question of speed/communications. As soon as I watched SimplifPhi's webinars where they mentioned that this was the primary cause of failure for their batteries (they use the term overdischarge for what we are both referring to as an overcurrent event in their materials/communications...that is why I mentioned it here, despite believing that it isn't the right term) I wondered why this simply couldn't be avoided with the exact solution that I mentioned above - current monitoring and an contactor to disconnect the energy storage system if necessary.

I don't have any intention of working with lithium ion batteries in the developing world for the time being, although they do work well in very small, simple PAYGO systems. Very few locations here have climate controlled locations where the energy storage system can be installed, budgets are exceedingly limited, and there is no recycling of lithium ion batteries avaialble (although it is quite exciting that some LFP batteries do not contain hazardous substances). I just feel that it is something that I should understand better.

Take care!

clockmanfran

A good informative Post, even better when real hands on folk share their experiences.

Mcgivor,   I do know a Norwegian guy that recycles Nissan Leaf batteries and has a heck of allot of experience.  He is on the 'Fieldlines forum' sometimes.  Nissan Norway were unofficially supporting his endeavours and he was putting together good packs specifically for Off Grid use.  Folk from all over Europe were driving all the way to Norway.  As i say, it was all unofficial but he really helped the Off Grid community.  

jakobw

I'm currently deploying Nissan Leaf modules/batteries for off-grid client systems. Super results so far but time will tell. I expect to get 3000 cycles or more out of them before they hit 80% of thier current second-life capacity. Were dirt cheap for a while thier but are now hard to come by. Back to deploying LFP cells from China. Good results with those so far as well.

I recommend reading about/watching some of the testing done with/on Nissan modules with regards to safety. They might be NMC/LMO but they appear to be engineered to be as safe as LFP or close to it from what I have seen.

Over-discharge is a real killer. Nominal overcurrent however is not something I have noticed being an issue so far BUT typically installations for me involve enough LFP that a 1C surge is upwards of 20+ KW. So not much of an issue. The Nissan Leaf modules are rated for 3C surge capacity so even on a small system with 150AH of capacity, I'm looking at a 450A surge capability. @48V, that's 21.6KW. Still far above the surge capacity of the average inverter we install.

We do install pre-charge circuits from REC though so we don't typically expose the banks to the initial capacitor charge surge.

Just my 2 cents...