Parallel lead acid and lithium circa 2018

[markopolo-ClassB]

Are we going to start seeing parallel lead acid & lfp battery setups now with the drop-in lithium batteries?

There's already a commercial option for this: https://www.bos-ag.com/products/lithium-extension-battery

The 110Ah Trojan Trillium LFP battery - https://www.trojanbattery.com/pdf/datasheets/27_Lithium_12.8-110.pdf - combined with 440Ah AGM would fall into the suggested 4 to 1 range of the bos-ag.com setup.

I'm not thinking of any special setup here, just a typical paralleling of batteries with the full realization that the lithium battery will be float charging the AGM most of the time. No buttons to push, no switches to toggle, no user management of the system, no decisions to be made etc.

We often try for perfect but convenience, economy and extended run-time, not perfect, would be the goal here.

 

[@Michael]

The BOS-Ag setup looks like a great concept for extending battery capacity. Put the BMS intelligence into the cells themselves, and make the whole thing plug-and-play.

Nice.

 

[booster]

Are we going to start seeing parallel lead acid & lfp battery setups now with the drop-in lithium batteries?

There's already a commercial option for this: https://www.bos-ag.com/products/lithium-extension-battery

The 110Ah Trojan Trillium LFP battery - https://www.trojanbattery.com/pdf/datasheets/27_Lithium_12.8-110.pdf - combined with 440Ah AGM would fall into the suggested 4 to 1 range of the bos-ag.com setup.

I'm not thinking of any special setup here, just a typical paralleling of batteries with the full realization that the lithium battery will be float charging the AGM most of the time. No buttons to push, no switches to toggle, no user management of the system, no decisions to be made etc.

We often try for perfect but convenience, economy and extended run-time, not perfect, would be the goal here.

There is probably a pretty high probability that is the system that Hymer was raving about a while ago for the European market. This is what precipitated the question of building similar, which Harry did, but with more charging control.


I think the questions that were asked then would still apply, though, even if we have bit more data now about the system.


It would be interesting to see how the currents flow on charging and discharging, especially out and in of the lithium, as it seems like it would be hard to have the AGM take priority in charging. One would think the lithium would fill first at a 14.3v or absorption voltage that the AGM would want, unless the lithium batteries self limit current going in. It says the lithium would discharge first, which would be expected, and they list a max amps but don't say if it is internally limited to that level. The float they say the lithium gives the AGM might be on the high end of the range, but probably not horrible, but they might also be limiting that internally.


Most of the AGM system chargers have absorption times in the 4 hour range, so the lithium would see that voltage and time essentially every charge cycle. The 4 hours is also not really enough to totally till most AGM discharges so if that was increased by using an amp based charger that time would greatly increase for the lithium. This was why Harry used the 12v to 12v charger in his setup, so the AGM would get fully charged off the lithiums. The other thing the 12v to 12v did was limit the charge current to the AGMs form the lithium, as when I tested that with small batteries the current from the lithiums went way too high ass soon as the charger was shut off if the AGM wasn't completely full or at least over 85%. That my be why this new systems lists max output so low and may limit it to that. If they limit that current, they would eliminate the largest part of the need for the 12v to 12v charger, I think.



At the 4 to 1 ratio, it probably would work best for someone with 300 watts of solar or more and moderate, maybe 50ah, daily use. In that case the AGM would rarely get discharged much at all and the solar would like the low internal resistance of the lithium battery. This would, I think, greatly increase the life of the AGM bank, and you would still have it there for reserve in cloudy weather.


Once power use gets to the point of bigger discharges of the AGM, I think it gets more difficult to predict gains and losses in life for both the AGM and lithium. It will be interesting to see if anyone puts a data logger on one of the Hymers when they get out, and if they are this system, to get some real world information.


The simplicity of the system would be a huge plus, and over the normal time an RV lasts any weaknesses may not be big enough to make a difference. We do know that ARV and Roadtrek both tired full lithium systems with drop ins and quickly went away from them, but don't really know what the issues were for them, if any, but those could have been different control systems.

 

[booster]

This statement is in the "Product Features" and explains things more like would be expected. Less about floating the AGM and more towards daily cycling and topping off the AGM is necessary.


This is much more in line with what would be expected and what started Harry on the path to his system. To do this as stated, I think they must be limiting the discharge current in the lithiums or they would melt things into a big AGM bank that was discharged a ways. If so, this makes the system much more appealing as it takes care of the relatively expensive addition of the 12v to 12v charger and associated relays.

Longer battery lifetime The lead acid battery is not strained by deep charging cycles, as the lithium battery is used for running daily cycles. During its long hours at full state of charge the lead acid battery is floated. Compared to pure lead acid systems both effects positively impact the lead acid battery lifetime. By preventing harmful sulphation, the lifetime can be increased to more than 10 years. This lowers the operating cost of energy storage significantly and has a positive impact on the environment. The lithium battery offers enough charging cycles for more than 10 years of daily use. The lithium battery furthermore recharges the lead acid battery in case that it didn't get fully charged during the day, a case which frequently happens in solar applications.

 

[markopolo-ClassB]

And by keeping them in parallel voltage remains the same.

The 440Ah AGM + 110 LFP combo is likely best suited for dual alternator setups. It would be hard on single stock alternator if current couldn't be controlled.

 

[booster]

And by keeping them in parallel voltage remains the same.

The 440Ah AGM + 110 LFP combo is likely best suited for dual alternator setups. It would be hard on single stock alternator if current couldn't be controlled.

Yeah, the voltage staying the same on both is probably the biggest weakness in it. If the AGMs are full after a charge, and so is the lithium, the AGM will see too high voltage for while. If the AGM is not very full after a charge, the lithium might not get it full before the voltage drops to far. If the charge runs long enough to fill a discharged AGM bank, the lithium will fill quickly and then see higher than desired voltage for some amount of time.



The big question is if the downsides are worth worrying about, and if the current limiting, hopefully both in and out of the lithium would balance a lot of it out, or not.



Definitely intriguing, and we know from Haryy's experience that it does work with the 12v to 12v charger to attain his goal of shortening generator run time. If it will do the same without the 12v to 12v it would be really nice.

 

[markopolo-ClassB]

Voltages seem to match up OK.

14.4V charge voltage seems to be OK for both types.
12.0V seems to be as low as you'd want to ever go on both types.
13.6V OCV for LFP is a higher than typical float for AGM but likely acceptable. 13.6V is the typical Normal mode for PD chargers for example. That 13.6V would trend down.

If I had any need for more battery than what I have now I'd be really tempted to try it.

 

[booster]

Voltages seem to match up OK.

14.4V charge voltage seems to be OK for both types.
12.0V seems to be as low as you'd want to ever go on both types.
13.6V OCV for LFP is a higher than typical float for AGM but likely acceptable. 13.6V is the typical Normal mode for PD chargers for example. That 13.6V would trend down.

If I had any need for more battery than what I have now I'd be really tempted to try it.

Agree, the basic voltages are well within reason. The only issues would in transition type times like the examples I listed earlier. How much issue that would be is anyone's guess, as those time are quite transient.


I am not sure what would be the biggest benefit at this point, if you already have the AGMs so not buying them. The plain capacity increase would be higher cost than more AGM batteries, the weight less, charge times shortened as long as the lithiums covered your normal daily use, AGM life increased if the lithiums cover daily use, the lithiums probably couldn't sit out in below -4*F.



My guess would be that the shortened generator times, be it engine or Onan, would be the biggest benefit, but only a guess. Shore power charging times normally are easier to have plenty of time to full charge AGMs, but generator and driving/idling time shortening would be a big plus for a lot of people

 

[booster]

Just as a note on whatever the possible non optimum things may be with the hybrid system. It is likely that they are probably less of an issue than a less than stellar battery charger in the system. As other discussions have talked about on the forum, most do not do a very good job of charging the batteries so whatever else happens with the hybrid setup might make those issues better or worse depending on what charger is used and use patterns.


Also along the same lines is that very many folks with non engine generator systems that use a the van alternator through a separator can easily be putting full charge voltage on full batteries during an entire day's drive, so up to maybe 12 hours, which is very hard on AGM batteries.


Sometimes it is easy to not think of where in big picture different weaknesses appear, and how the rank of consequences order is in the real world.

 

[hbn7hj]

I can have a parallel system just by flipping the switch to both with a 1 to 1 ratio, not 1 to 4.

The RV is in the shop due to a deer strike but let me know what you want to know. It will get new lead acid batteries in January.

Current read outs with AGM at 50% and lithium at 100% may be a place to start. I'll let you know. It isn't getting out of the shop anytime soon.

The system works very well with all the switches. Power is no longer a problem. I do not have a compressor fridge.
Harry

 

[booster]

I just noticed on the website for the batteries that Hymer is shown as one of the two sales affiliates in Germany. One other in Europe and the rest in less developed areas.

 

[booster]

It looks like they are also partnering with Victron, which should be able to give them near instant market penetration if Victron starts to sell/push their systems.


https://www.bos-ag.com/news/news-archive

 

[markopolo-ClassB]

The Hymer connection likely now means a Thor connection.

https://www.bos-ag.com/news/news-ar...stem-hat-das-goldene-reisemobil-2018-gewonnen

Description of Hymer system: https://www.hymer.com/en/models/technology-plus/hymer-smart-battery-system.html

It's not the 4 to 1 ratio, it can be more lithium than lead so I guess it is very flexible.


Harry's 1 to 1 ratio likely offers great price/performance particularly at around 100Ah AGM & 100Ah LFP. There would be no problem running a microwave oven for example.

 

[booster]

It is interesting on both the manufacturer and Hymer explanations they talk about priority charging (as in first) for the AGM so it is sure to get fully charged. In a some ways that seems a bit contrary to being able to get the most benefit from the system. The bos-ag site does state another place that the lithium will be used to charge the lithium so a bit of contradiction there.



They say all the timing of charging is built in to the control system, but I would think that would be probably not be all that needed. The AGMs are essentially always be at lower voltage the lithiums are, so they will get charge when the lithium won't, unless the charger has enough capacity to get to a high enough voltage by satisfying the acceptance of the AGMs first until they get to the lithium voltage. At that point the AGMs wouldn't be full, though. They could delay turning on the lithium charging until a higher voltage is seen on the AGMs, perhaps at 14.3v which would be at the end of the bulk stage, but that is going to extend the charging time quite a bit before the lithiums charge. I guess I would think you would want to get charge into the lithium batteries if you want to cycle off them as they mention. Especially with solar, they could wind up cycling in the mid range of the AGMs all the time and never get them full enough to charge the lithiums.


It seems like it would make more sense to charge the high acceptance lithium batteries first and then cycle of that and use them to charge the AGMs. The lithium should get a bit more out of the solar and also shore charging if the charger is fairly large, in any given amount of time. The difference would probably be even more if on the van alternator or standalone generator where you really want to minimize charge/run time like Harry does.


It is going to be very interesting to see how it actually functions in the real world, and especially if Victron starts putting out systems or AGM enhancement systems.

 

[markopolo-ClassB]

The AGM would want to be at a lower voltage but if the AGM and LFP are connected via the same bus then both will be at the same voltage. Keeping them connected seems safer to me - no potentially damaging huge current outflow or inrush on/with the AGM.

..............................
It seems like it would make more sense to charge the high acceptance lithium batteries first and then cycle of that and use them to charge the AGMs. The lithium should get a bit more out of the solar and also shore charging if the charger is fairly large, in any given amount of time. The difference would probably be even more if on the van alternator or standalone generator where you really want to minimize charge/run time like Harry does...................

That's what I think too.

It would seem to be an unnecessary complication to try to charge the AGM first. The AGM will be charging (float or otherwise) maybe 90% of the time (when connected to the LFP battery). Could that be what they mean by priority to the AGM?

 

[booster]

The lithium keeping the AGM at float, albeit probably a fairly high float voltage, is another interesting thing in that they say the priority discharge the lithium first. This could put the lithium into low voltage shutdown before the AGM would start to discharge at about 12.8v. If this is what happening, the priority is happening all by itself so no control of it really needed by the BMS.


If the AGM was discharged far enough and the charger not particularly large so it couldn't get the charge voltage up very quickly, the AGM would also charge first to some point determined by the acceptance of the AGM vs charger size vs what voltage the lithium comes out of low voltage shutdown.


Personally, I would like to see pretty big charger on a system like this for shore power and engine generator. Probably sized so that you wouldn't hurt the AGMs by too fast charging if they were low, so something in the .4C of the AGM. That would give you a 120 amp charger for the theoretical 300ah of AGM/100ah of lithium system. The large output would allow the charger to reach the turn on voltage of the lithium quickly for faster recovery. That leaves the solar, that would probably never charge the lithium until the AGM was near full and the voltage finally came up. This may be the priority charging they are talking about, which also will happen all by itself with no active BMS actions required.


It appears they may be making it look like their BMS is doing all kinds of charge allocating, when it is really just the way it happens based on the size of AGM bank and chargers, and the wake up voltage of the lithium pack. Nothing wrong with doing it that way, as it is cheap, effective and reliable, but a bit of a marketing fib if that is what is happening. I am sure they wouldn't want to tell potential buyers about all the things that may affect the results they actually get.


If I were to get a system like that, it would have two battery monitors on it before it ever got powered up :whistling:

 

[markopolo-ClassB]

With a parallel LFP AGM setup I think you could run the lithium from 90% SOC OCV down to 10% SOC under moderate load. That looks like it would be from about 13.4V down to just above 12.0V. The AGM could handle that. Right sizing the capacity of the combo bank so that those deep discharges aren't occurring all the time would be good.

On a 300AGM/100LFP or even 400AGM/LFP combo bank exit absorb amps could be set as low as 5A at 14.4V if no loads present or maybe as high as 8A.

I agree about using a big charger. I think I'd size it so that it is not running at 100% rated output. And yes to monitoring it all to know what is going on.

 

[peteco]

The simplest LFP add for me would be to my current two 6-volt lead acid bank at 225 Ah. So I would be at 225LA/100LFP ratio. If LFP costs came way down I might consider going to 200 Ah LFP. I see a higher ratio is recommended (4/1). What is the problem with a lower ratio: 2/1 or even 1/1?

 

[booster]

With a parallel LFP AGM setup I think you could run the lithium from 90% SOC OCV down to 10% SOC under moderate load. That looks like it would be from about 13.4V down to just above 12.0V. The AGM could handle that. Right sizing the capacity of the combo bank so that those deep discharges aren't occurring all the time would be good.

On a 300AGM/100LFP or even 400AGM/LFP combo bank exit absorb amps could be set as low as 5A at 14.4V if no loads present or maybe as high as 8A.

I agree about using a big charger. I think I'd size it so that it is not running at 100% rated output. And yes to monitoring it all to know what is going on.

I did a quick look and couldn't find a decent discharge curve for the Lifepo4 batteries. Do you happen to have one? We know the charge curve has a steep dogleg at the finish end, so it would be interesting to see what the lower end of the discharge curve looked like, as that would be the determining factor in how each of the battery styles got charge as the cycle progressed. If the acceptance voltage for the lithium doglegs up quickly it could get above the AGM acceptance so most or all the charge would be going to the AGM. Superimposing the charge and discharge curves for the two types would be very interesting, especially when charger size gets thrown in.

 

[booster]

The simplest LFP add for me would be to my current two 6-volt lead acid bank at 225 Ah. So I would be at 225LA/100LFP ratio. If LFP costs came way down I might consider going to 200 Ah LFP. I see a higher ratio is recommended (4/1). What is the problem with a lower ratio: 2/1 or even 1/1?

Harry is happy, and I think he is at 1 to 1 now on his setup. I don't think ratio if much of a big deal unless it got to be more than the 4 to 1. That 25% of capacity in the lithium would be just about enough to top off the AGM bank if a fast charge brought it up to the bulk/absorption transition, so I think that would be an important element to keep.