This site is now in read-only archive mode. Please move all discussion, and create a new account at the new Victron Community site.
Serious question, please bear with me.
It seems that some boaters are supplementing tired AGM batteries with extra LiFePO4 batteries in parallel. My understanding is that this shouldn't be done. But I hear that many are pleased with the results.
So
Are there any gains?
Are there drawbacks?
Are they fooling themselves?
Are there any dangers involved?
The only issue is that as the lifepo4 has a higher sitting voltage is that it will slowly discharge into the LA. Basically charging the LA all the time. Also until the system voltage drops below 12.7 nothing power wise will be supplied by the LA. So the LA will basically do nothing and the lithium will do all the work.
Now if it's a starting battery then the above will occur but the LA would supply amps during starting or if the lifepo4 bms turned off
If you do it correctly, then you can have Lead Acid and Lithium in Parallel. In fact, as a fortunate happenstance, the chemistries complement each other very nicely.
As far as the Lead being constantly charged by the Lithium, that doesn't really happen even though logic might suggest it does. (I know this as I have independant dedicated Victron BMVs on the Lithium and the Lead, so I can see all currents between the batteries as well as the overall system via a overall system BMV).
I've been running a dual chemistry Hybrid setup for quite a while now and it performs exactly as the maths suggest it should do, and exactly how you would want it to, e.g. The Lithium gets used primarily until the SOC goes down to a point where you would want to stop using Lithium and the Lead takes over. And on recharge, the Lithium takes most of the charge thus maximising the charge efficiency of the system.
Here is an example of this setup in action from last month, currently with 300Ah of Lead Carbon and 100Ah of Lithium ....
You see the Lithium (top) doing all the work and then when it goes to around 15% SOC, the Lead starts doing the work. Then at the end, the Charger is plugged in and the Lithium recharges nice and quick.
In [discharge] use, this behaviour is automatic; When recharging, you need however to take into account the different charge characteristics, and the voltage especially, of the different chemistries. I have a special micro-programmable relay (of my own design configuration) that comes into play when charging to connect/disconnect the Lithium into the Hybrid Bank to allow full charge of both Chemistries with the single charger.
If you want to read more, I have a full writeup and explanation here - https://www.wildebus.com/hybrid-battery-bank-introduction/ .
It would be interesting to try out this setup with some Victron Smart Lithium Batteries, but I have this setup for my own use so cannot afford to just buy those out of curiosity.
No, not at all.
If you look at the maths, Lithium is ~13.3V and Lead is ~12.7V (full battery levels), so IF there were any constant current flow, it would be FROM Lithium and TO Lead, so if anything, it would be a state of permanent discharge from the Lithium.
Now a Lead Battery does not need a constant current to stay charged unless it is faulty, so there is not a flow of current anyway (My BMVs allow me to monitor this).
As per the write up (hopefully you read it to understand further?), the relay is in place to manage the CHARGE aspect. When not charging, the relay just stays on and plays no part other than allowing the two types of battery to be connected together.
Lead generally needs ~14.4V over an extended period to charge (the exact voltage depends on the battery, but 14.4V is a good ballpark number to use to discuss). And in cold weather with temperature compensation that charge voltage can go upto 15V.
Lithium generally never wants to see over 14.4V (and many Lithium Batteries BMSes will disconnect over 14.4V with an over-voltage error) so there is an inherent incompatibility there if the charging defaults are used.
Lithium also does not need a float - You charge .. It takes it ... When full, it wants the charger off. If you monitor the charge characteristics of a Lithium Battery in practical terms, you see they are fully charged at around 14V, so what I can do is have a charger setup for Lead, charge my Lead and Lithium in parallel and disconnect the Lithium (automatically via the programmable relay) from the charger when it is full, at which point you see the Lead continue to charge. The Screenshot shows this charging pattern
As with any non-standard setup, you should understand the characteristics of the products and how they interact within the system. I have the extra monitoring on my system not because it is required but so I can check how everything is working and if it corresponds to the theory and mathematics (which it does). The Relay operates independently of any monitoring or Victron kit . I can however configure and adapt the operational parameters (voltage for connect and disconnect, disconnect and reconnect delays, overrides, etc) via a App over Bluetooth.
Lithium draw down was 85ah, and lead acid was (less than) 30ah?
Not really seeing any conclusive proof that combining lithium and lead is beneficial.
Now a general point....
I am not posting this in an attempt to persuade anyone that a Hybrid setup is something they should have. A question was asked about mixing Lithium and Lead, and I have posted MY answer based on MY experience and MY results.
If this is not of interest, no problem. If someone does not think it is a good idea, no problem. If anyone wants to learn more about the setup I have and have extensively logged results of (so based on actual data on actual products in an actual installation) then they can click on the link I posted earlier.
As the saying goes, Your Mileage May Vary, and if is it not for you, no worries. Not promoting or defending it, just explaining it in the process of replying to the specific question posed.
Perhaps you could show a battery voltage plot that indicates the power handover form the Li to the Pb battery.
A nice vrm voltage plot would be easier to follow.
This may be on interest ... kind of shows how the charging works when you have Lead and Lithium connected together and what happens with the charge cycle. The graphs show the overall Battery Bank, and the Lithium Battery below.
You see the Voltage increasing as the batteries get charged. Usual straight line for the Lithium Charge SOC. You see the Li current (bottom left) flattish and then as the voltage hits around 14V, the current into that battery falls away as the battery is approaching full - confirmed by the BMV (bottom right) showing 100%. The Relay doesn't switch off until the voltage gets to 14.2V and then all the charge current goes to the Lead (but as usual for lead, when nearly full, it gets slow to finish which is why it drops away. The Lithium tends to drop to the low 13Vs initially when disconnected and then recovers back to normal lithium rest voltage.
Unfortunately it is not possible to just use VRM for this as it is necessary to combine different data sources to see the picture so I imported the data into an Excel Pivot Chart.
I hopefully still have those charts in a Excel workbook so if you tell me what colour combos work well, I will see if I can replicate using them.
A few things ....
The amount of Power drawn is the amount of power needed. That is just a fact, yes? So around 115AH, then. Agreed.
How is combining beneficial?
1. Ohms Law - P=IV. If you need say 500W, a 12.8V (nominal voltage of a Lithium) will have a current discharge of 500/12.8 = 39.06A; a 12.0V (nominal voltage of a Lead) will be 500/12 = 41.67A. Lithium is more efficient and a 100Ah Lithium Battery has more Power than a 100Ah Lead (not many people actually realise that)
2. Recharge Time. A Lithium Battery will recharge significantly faster than Lead and will accept a much higher C rate. It also does not tail off in charge rate as it gets to >75% of SOC
3. Charge Efficiency. Recharging Lithium is much more efficient and it converts nearly all the available current into battery charge. This is not the case for Lead.
All the above points to Lithium being better, which is pretty well the case, but we live in the real world where a Lithium Battery is significantly more expensive than Lead. Cost comes into play....
4. My 300Ah of Lead Carbon (what Victron would call "immortal" in their Lead Carbon range) costs 75% in total the cost of the single 100Ah Lithium Battery. If I were to fit the same amount of Lithium Battery to give me the equivalent usable capacity (remember what the possible safe discharge level of Lead Carbon is. A clue: it is NOT 50%). it would cost a great deal more money to have an all-Lithium bank
In general day to day use, the usage is such that the Lithium is sufficient and the Lead is in the background. But if I need more capacity than the Lithium has, then it is no problem as the Lead is available. If solar is very poor on a trip, then I know my use is not particularly compromised as the Lead is there to provide power, which would not be the case on a budget-limited 100% Lithium install.
5. Because the Lead Is called on much less, the charge cycles for it is greatly reduced and so will last a lot longer than a 100% Lead install.
The charge cycles for Lithium are so much higher to start with there is a question about if it really matters if you discharge them lower - you might take a few years off them, but by the time they need replacing even then, Lithium will be even better and a lot cheaper likely, so changing might be something you would do anyway at that point. The lifetime cycle count for the Lithium Battery model I have is around 4,000, and drops to about 2,000 if you discharge very low. So halving the life! but does that really matter in reality? 2,000 cycles is a lot of cycles. IF (and it is a big IF) you discharge to that level on a DAILY basis, the battery would still last 5.5 Years. My Battery has been installed around 2 years I think? This is a screenshot I just took of the BMS data in the battery ....
81 cycles recorded. That battery will outlive me!
Just curious what the manufacturer of the lithium batteries thinks of all this?
Sounds like you toss your warranty away and lose Victron support, all to save a few bucks. looks like short-term thinking to me, but each to their own.
@Wildebus approach, which can save money and work well.
Ideally you want lithium with BMS integrated that can manage lead charging profiles for the lithium.
@Nickdb, Lithium... In what way would you be impacting the warranty? you are using a battery and the BMS is not reporting any issues so there is no problem. This next bit may be of interest as an aside .... the specific Battery I am using in my Hybrid install used to report Overvoltage when used with a Victron B2B and a Victron IP22 charger, both set to Li-on profiles. It is a little over-sensitive in the BMS. We swapped the battery out for the customer and in MY configuration with the VSDR, I don't get any BMS errors, so actually this setup is BETTER for the battery :)
Victron Support ... There you may have a point in that you are using a device in an usual way so if you have a question when using that setup, your answers may be limited as outside of their experience or design parameters. However, I doubt that is the case even here, although I have come across this recently with what I consider to be a programming glitch but I might be the only one ;) (namely, MP not going into Low Power mode if AC Detected, but I will just have to live with that. it is not a serious glitch, just an annoyance).
The good thing about this Hybrid setup is that you don't actually need Lithium Compatible chargers even - so the exra cost of the VSRD device could easily to paid for by potential savings in not needing to upgrade older limited chargers?
@kevgermany, that ideal would indeed be ideal :) I doubt there is any kind of volume to make that viable however (and it could limit the battery selection as well).
There is a (Germany?) company that sells a pre-packed Lithium Pack add-on to Lead I think? but their approach is rather pricey as the batteries are quite small. As I undertstand it from their info,if you want the capacity of say 2 packs,you end up with buying the extra Li/Pb management system twice. Want 3 times the capacity, you end up with 3 management systems. I think that is wasteful compared to my approach, which is unlimited batteries either end, and the only limit is a maximum 200A (continuous) current between the two, which is a pretty decent operating level.
Now any installation should not be based on "hope" but solid data and knowledge. Having done tests on the Lithium Batteries and their charge and discharge patterns in a standalone configuration, and then again when operating in a Hybrid installation, I am pretty confident "hope" is not part of my install ;)
On my own Hybrid installation, I have individual BMVs on each side so I can continue monitoring in detail but it is not an essential part by any means. On other Hybrid installs done, they don't have a BMV dedicated to one battery type, but the Lithiums installed have a built-in SOC Monitor accessible via a Bluetooth App so the users can see the status of the Lithium as well as the overall picture with a BMV or Smartshunt for the overall Battery Bank.
The VSDR device I developed will also manage the situation well. If you have that, the Lithium BMS becomes less important. Without the VSDR or similar functionality, you can get an over-voltage situation which is generally not critical as long as the batteries BMS provide a disconnection feature, but I think it is better to preempt issues with design, rather than hope.
With the new Custom Widget in VRM allowing you to combine different devices in the one graph, I thought I would show the two BMVs (one on Lead, one on Lithium) together and how the Lithium disconnects and reconnects from the DC system.
The Green is the Lithium, the Blue is the Lead....
Very interesting! I came here looking to possibly combine a 6kW 48V LiFePo CANBus battery with 4x 12V AGM batteries in series. I don't need more capacity, just better peak current tolerance as the BMS as you mentioned can be "sensitive".
I read the conclusions from the provided link at wildebus.com and it is great but has a slightly different priority from what I need with just peak currents
The system works very well, apart from rare shutdowns by the battery with ugly ripple and low battery warnings. That is what I want to address.
My tests indicate this happens in two scenarios:
1. when attempting to start up the MultiPlus II 48 5000 when there is no PV at all.
2. when you get "unlucky" with several peak currents happening at the same time.
Freezer, waterpump, somebody using a handtool etc.. Discipline by users will also help ;)
The main battery is as mentioned a 48V 6kW CANbus battery that is handled by DVCC from a Cerbo GX. I assume the DVCC part may create some problems.
The battery developer says it is fine to charge the battery with standard LA/AGM algorithms.
Grateful for any comments! Many thanks!
I have a server rack Lifepo4 battery with a senstitive bms that trips wiht fast big current like a compressor or well pump kicking on. Keeping the FLA in parallel as described prevents that, improving the functioning of the Lifepo4. Also, I see no significant discharge into the lead at night, but what little does keeps them both at 25.6. Also, the lead in the system keeps my wind turbine (think alternator) safe in case the bms on the Lifepo4 were to kick off. The lower resistance of the lifepo4 prevents the turbine from shutting down too because it just accepts the current without voltage spikes that occur with FLA from large gusts. Lots of benefits for my system.
I have seen where this works well as stated. https://www.youtube.com/watch?v=tAuPfgZgXec Basically in a nutshell is a cost savings for not having a full Lithium house bank. How I understand it one should size there banks so that Lithium will be the work horse and the AGM/ Lead acid will be the backup or reserve power if needed. Lead acid rarely gets drawn down and charging is accelerated do to the Lithium. Neat idea! And if I am think correctly when charging from an alternator one would be fine if BMS shuts down, voltage spike would be absorbed by Lead or AGM. I think I would try it first on the RV then if all worked well then the SV. I have talked to a large manufacturer of LiFePO4 batteries and they didn't have much to say as they are interested in selling their product.
Interesting avenue by Kevgermany and thanks Wildebus for the graphics. What is interesting is that the Victron Smart BMS 12 200 does exactly that by having the AGM start battery combined (through the BMS) with the Lithium battery for charge - not for discharge. So it led to the question, how to configure the alternator regulator so as to provide valuable energy to the Lithium while not overcharging the AGM since they are on the same wire.
From my understanding of the charge profile of the Victron AGM and of the Victron Lithium, having a short bulk around 14.5 would be an happy moment for the AGM and rapidly go into absoption at 14.2 for the Lithium which will draw all the amperes. If the charge with alternator is less than 2 hours then chance of gassing the AGM are limited i guess. Your thoughts ?
