In the last chapter I explored why we offshore voyaging sailors need a battery management system (BMS) that can intelligently control charging sources rather than panicking and shutting down the whole works—totally unacceptable on an offshore voyaging boat.
If you still don’t agree with that after reading the last chapter and the above linked one, that’s fine with me, but let me be clear that as far as I’m concerned, this is a fundamental requirement for a seamanlike lithium-based electrical system, and everything that follows is based on that.
Given that, I will answer any comments that argue with this premise with “fine, let’s agree to disagree”, rather than get into it yet again—that’s as nice a way as I can come up with of saying “you’re welcome to your opinions but please respect mine”.
“Drop Ins”
That said, if you have batteries with a BMS that does not communicate, or that’s the route you want to go—”drop in” batteries with an internal BMS that does not have process control—I will cover how to do that in a seamanlike way in a later chapter, so let’s not go there now, other than to say that such a system will never be as good as one done right…in my opinion.
More Required Capabilities
Before we dig into the simple wire or CAN Bus decision, which I will now cover in a later chapter, let’s explore two other buying criteria that either type should satisfy:
Cell Balancing
Let’s start with cell balancing, because understanding that informs our decision on monitoring—this is sure to get everyone warmed up.
First off, we need a BMS that balances properly and automatically, and a system set up to let it.
Who says? I do…oh, not good enough? You’re right. How about this? Al Thomason at WakeSpeed says.
I spent over an hour on the phone with Al before writing this article and most of that was on balancing and making that happen.
Al has worked with most of the major lithium-battery and BMS manufacturers over some five years to develop profiles that will charge their batteries as efficiently as possible, while not shortening their lives by over-charging.
No one, but no one, and certainly not that handsome silver-tongued presenter on YouTube who purports to make all this simple, or that guy on the forum who has all the answers with absolute certainty, has even a fraction of Al’s understanding.
Here are the key things Al told me, and other credible sources confirm:
Balancing is Real
All lithium batteries, no matter the quality, tend to drift out of balance over time.
For the purposes of this article I will assume that the battery was fully charged and properly top balanced when new, and focus on ongoing balancing while in service.
If you are interested in the initial process, the Victron Lithium Smart Battery Manual has a good explanation (section 4.3) of what needs to be done before new batteries are put into service.
13.5 Volts is Not Enough
One of the biggest myths around lithium batteries is that if we never charge them above around 13.5 volts1 (or close) we don’t need a BMS that controls charging sources since we will never be close to the BMS cut-off voltage of ~14.4 volts.
Yes, that’s true…for a while, but the problem is that at that voltage the battery won’t balance and so will drift over time.
How long before problems show up? That depends on usage and how well balanced the cells were in the first place, how the BMS works, the owner’s astral sign, and on it goes. Could be years, could be months. More likely the latter than the former.
So, yes, just charging at low voltage will work, but it’s far from best practice, and on an offshore voyaging boat only best practice will do, particularly since it’s not that hard or expensive to do it right.
Keep in mind, it only takes one weaker or stronger cell to start causing blackouts if not properly managed.
No, Al and I didn’t make this up. This is just one example of the instructions from a battery manufacturer:
We recommend a minimum absorption time of 2 hours per month for lightly cycled systems, such as backup or UPS applications and 4 to 8 hours per month for more heavily cycled (off-grid or ESS) systems. This allows the balancer enough time to properly balance the cells.
Usage Dependant
Wait, it gets worse. Lithium cell balance drift is usage dependant.
Hi. Great articles! I have not gone the lithium route yet, but I will once my lead acid bank dies, so this is a timely article since I am trying to educate myself as much as I can. Personally I will probably go the full DIY bank route, but I don’t think that matters much for the purpose of this article. Anyway, I just wanted to say that more citations and sources within your article would be nice, since I’m sure you manage to find some things in your research that I don’t. Even if not directly in the article like an academic citation, just a “further reading” list at the end with various sources you’ve found would be appreciated. I realize a lot of this comes from Al and therefore it is hard to add citations to it (which does not mean that it’s any less relevant!!).
Lastly for anyone interested, I would just like to mention that Marine How To have a few good (albeit a bit old and very technical) articles where he also have some tips for further reading (and also HEAVILY advocate for further research).
Kindly, Arne
Hi Arne,
Thanks for the kind words.
I did cite several sources. And I will cite more that are relevant in each chapter coming. But adding a whole bunch more sources now is just going to confuse people because so much either does not apply to us or is wrong. In fact, that’s why I’m going through the aggravation of writing this buyers guide, and have gone to all the work to sift the BS using my 40 years of experience in both technical writing/research (yes, before AAC) and offshore voyaging, as well as my trade training as an electronics technician. Bottom line 90% (at least) of what’s out there on lithium is either wrong, over simplistic (which is another form of wrong) or not relevant to us. Sorting out situations like that is what I do, whether it’s lithium batteries or mainsail reefing.
And I agree that Marine How To is a good source and I have often referred to him in the past. That said Rod is not an offshore sailor and as such is more casual about systems that can cause blackouts than I believe is wise for our usage, so when reading his stuff we need to keep that in mind. For example, he recommends drop in batteries with no process control, which, as I have explained repeatedly, is a very bad idea for an offshore boat.
I totally see your arguments John! And after having done some fairly heavy research myself (on both this and other totally unrelated topics), I absolutely get what you mean! And this is one of the reasons I like your site, when I started out as a TOTAL beginner on a sailboat 2,5 years ago (never even set foot on one before), I still managed to singlehandedly sail my new-to-me yacht from UK to northern Norway with no issues, mostly due to this resource. So I (and I believe many others) trust that you are very able to make sense of what actually matters. Just to let you know that at least some of us appreciate all the links to sources/resources you find worthwhile! 🙂
Kindly, Arne
Hi John,
Rod has recently started “recommending” drop-in batteries with no process control, but they’re not real recommendations. He’s just accepting the market power and that this is what part of it is asking for, no matter what he says. So he gives them a decent example of such a product, which also gives him a small revenue, which he really needs due to his health issues.
Still, he makes no secret about his opinions other places on his site, which have been the same for over a decade: Lithium batteries MUST have a properly communicating BMS and an external system that can react to the communication. Anything else he considers swindle. That includes at least one of the batteries “recommended” on his site.
Hi Stein,
Very true, but also note that many people just read one of Rod’s posts, say one on drop-ins, and then extrapolate to “Rod recommends drop ins” (check out a thread in comments to my last chapter), which is why I wrote what I did.
Great post.
I really think lithium is fun right now, but as with many new fads/techniques people sometimes tend to get their temper worked up.
Some reflections in no particular order:
Balancing:
I personally think that for cruising, large lithium banks should have an active balancer. Passive balancers does simply not seem effective enough.
Most “drop-in” batteries have passive balancing.
Individual cell monitoring:
For most DIY systems this is easy to accomplish. It should also be easy to accomplish for “drop-in” batteries, but most does not seem to have this feature right now. I simply think most companies are not adding this “feature” because it hasn’t been requested by the customers, it is mostly a software question.
Voltage:
I have watched/read some info from https://off-grid-garage.com/.
He has an offgrid solar system with diy batteries.
He has done some good videos regarding charging at a slightly slower voltage (13.8 volt for a “12v” system). I think when my system is finished I will try out his settings and see how they work for me as a cruiser.
Hi Henrick,
Whether or not we can get away with charging at 13.8 volts is totally dependant on what voltage the BMS needs to balance properly. As I make clear in the above article it’s vital to know that value and make sure that we are charging at that voltage often enough and for long enough to keep the batteries in balance. Usually that voltage is at least 14 volts and often 14.2. Point being that it’s just not as simple as these YouTubers will have you believe. Also, as I have said repeatedly, taking our information from sources that are not knowledgeable about, and focused on, voyaging yachts is not a good idea. And blanket statements like charge at 13.8 are pretty much always wrong. We need to find out what the right number is for our batteries and BMS.
All that said, sure you could charge at 13.8 most of the time and just go higher when you note the batteries need balancing. However, be aware that you can still overcharge and shorten the life of your lithium batteries at 13.8 if you leave them charging at that voltage for too long. To understand when a lithium battery is full and charging should be turned off we have to be measuring amps going into the battery and act on that.
Hi John,
Confirming your statement: Even 13,5 Volt “float” charging has resulted in the death of multiple lithium banks. Probably most premature lithium deaths are caused by extended low Amp and often low Voltage charging. Anything above 13,3 Volt is risky if prolonged. Lithium cells should ideally ONLY be bulk charged. They can tolerate some saturation charge as long as the Voltage is a slow as possible. They must NEVER be float/maintenance charged at a Voltage above 13,3 Volt.
Setting a lower limit for the bulk stage is not in any way a sufficient method for cell safety. It’s only a method to reduce cell ageing by staying a bit further away from stressing them. It’s only useful if that strategy fits the rest of the system, which it won’t if there are automatic “balancing” circuits anywhere.
P.S.
I have to nag about this: What the automatic “balancing” circuits do is NOT real balancing. It’s only equalising the Voltage. Real balancing of lithium cells is a specific process that changes the chemistry of the cells. It takes time and precision at higher Voltages. It leaves the cells in far greater “harmony” than the automatic systems can ever get.
Each cell still absolutely needs constant attention for possible drift, but actual rebalancing is usually not needed often on matched banks in use. I know of very large 48V propulsion banks that have stayed within 0,03 Volts between cells for 3 years now, still counting. No “balancing” circuits. These are run daily 100% to below 50%, often further. Harder work than any cruising boat house bank in history. It works, but it has to be good components in a good system.
The vast majority of electric tourist boats in Amsterdam are still better off with, and indeed have, flooded lead acid 2V traction cells. Lithium is only chosen for boats struggling with space, weight or insufficient charging time between running times.
Hi Stein,
That all sounds right with one important clarification. All of this gets way easier to manage and understand if we take a moment to think about ohms law rather than just giving rules of how many volts or amps:
The biggest difference between lithium and lead that explains most of the differences in the way they need to be treated is that the internal resist of lithium does not rise much with state of charge, unlike lead where it does. Once we understand this, and ohms law, the ways in which lithium must be treated to have a long life becomes obvious as does why both current and volts must be managed properly. So to just say low amp and low voltage charging is destructive is overly simplistic. How low? If amps are held to a really low amount, voltage will have to drop too: ohms law.
Conversely at even 13.5 volts the battery will accept more current than is good for it for as many hours as the charging source can supply it. Why? Ohms law: the resistance remains low so a lot of current can flow at lower voltage. This is also why lithium should not be float charged. And why all high current charging sources (alternators) should be controlled by a regulator with a shunt measuring current going into the battery.
My point is that I get a bit concerned when users make pronouncements about what is right and wrong without relating to the basics of how electricity works, and so I always want to bring that in to the discussion.
As to rebalancing, or not. I think that these sorts of broad brush assertions are not the way to go. As I learned from Al, different batteries will behave differently depending on usage profile. So what happens on big banks on ferries that are fully and properly charged every night probably has little or no bearing on what’s right on a cruising sailboat. That will depend on the battery capacity to charge capacity ratio, as well as timing. So it’s not even smart for one sailboat lithium owner to tell another “balance like I do and all will be fine”.
The bottom line is that we must follow the vendor’s instructions for balancing, for example those from Victron (in the article), which are very different than your experience. But even then, we must monitor to adjust balancing to our own usage.
The great news in all of this is that assuming the cells are good, failure to balance, while it may result in the failure modalities I describe in the article, can be easily reversed, which is not always true with sulphating of lead acid. Conversely lead acid are intrinsically more robust because their internal resistance rises with state of charge…Ohms law.
Well done, John. You covered the balancing issue well and made it accessible for all. It’s but one aspect to consider of many and often misunderstood or ignored.
I do have a comment to help others not stray outside the channel. At this stage, only the battery cells themselves can provide the information about their particular behavior and status and thus only a device that can read and understand cell data can provide good advice for cell management. Devices that attempt to infer the information from the pack rather than the individual cells simply don’t have adequate intel. For example, I have two Victron BMVs, one for the lead start battery and one for the house lithium. The BMV does great on the AGM battery, but I have no other reference to compare. It just seems right. However, with my LFP bank, the BMV is useless for state of charge. It isn’t even close. My Orion BMS provides excellent monitoring and it is easy to see that it is correct with data logging and charts. I’m sure Victron is doing the best possible but right now it says my LFP bank state of charge is 92% when it is actually 44%. That’s really awful. I only use it for quick voltage or current status checks.
With lead acid, you only need peukert and coulomb counting to get close. Not so easy with LFP with only those and bank capacity added on. This also means a smart regulator cannot know exactly what the battery needs. It’s actually much more of a problem than you may think. Beware.
There are two schools of thought here. One has the BMS as the authority and it must instruct all charging sources including alternator, solar, shore power, wind, generator, etc. Those devices need to be told what to do by the BMS and if they don’t behave, they get cut. Same for loads with the final straw being contactors that should never have to open but will do so to protect the batteries. The final straw for bad behavior of all the charge and load centers.
The other approach is that the BMS can only be trusted for that last protection and maybe some opaque balancing behind the curtain. In that approach, each big charge and load source – and of course a very nervous operator – must attempt to understand what the battery needs sans advice from the BMS. That’s the drop in world where you need smarter charge regulators and battery monitors than we have today. I would love it if we could buy some Battleborns, a Wakespeed, and a Victron BMV with other blue boxes for solar and be off to the races. It might even work fine for light duty local and coastwise cruising. That tech is just nowhere near ready for heavy duty offshore cruising. I hope it gets there because the alternatives of high cost integrated systems or high geek DIY ones are not attractive to most.
Hi Robert,
Thanks for the kind words.
I agree that a Victron BMV is not adequate alone and that we must also monitor down to cell level. I say that repeatedly in the article above. However, if your BMV is that far out it’s a calibration and charge reset problem, not intrinsic to the device. We have an article on how to fix that. It’s for lead acid (on my list to update) but the principles are the same: https://www.morganscloud.com/2018/08/30/battery-monitors-part-3-calibration-and-use/ Follow the steps in that article, adjusted for lithium, and your BMV will be very accurate as long as you fully charge your bank say once a week so the monitor can reset and sync.
One point: it’s way easier to get a BMV properly calibrated for lithium than lead because the former does not discharge at different rate dependant on current (Peukert Exponent), as you say, so if it’s not working right on lithium, something is wrong in the settings.
As to a smart regulator not knowing when a lithium bank is charged. That’s true for regulators that say they are smart, but are actually stupid, like those from Balmar. But a regulator that measures amps going into the battery like the Wakespeed 500 can absolutely know when it should cut off charging with lithium or lead, although it’s even better if the BMS communicates with it. I explain all that in the article and here: https://www.morganscloud.com/2020/07/04/stupid-alternator-regulators-get-smarter-finally/
As to the rest of your comment, it duplicates stuff I have already explained in the last two chapters.
And that leads me to a friendly request: you and I are wasting a lot of time discussing things that I have already covered in detail in other chapters. So I would be grateful if you would at least have a scan through the Online Book to see what’s already covered: https://www.morganscloud.com/category/electrical/online-book-electrical-systems/
That said, if you then think I have got something wrong, I’m interested to discuss that on the relevant chapter.
Hi. Great article. As usual, very informative.
I have one comment regarding the drawing of the connections between WS500 and the Victron smart Lithium Battery Bank.
The Victron Smart Lithium Batteries has got two dedicated BMS signal cables, connecting the batteries to the BMS. If a system has multiple batteries they are to be daisy chained to the BMS. Those cables doesn’t show on the drawing and as far as I have understood the Victron system, that has to be a foul since the BMS cables are vital to have the BMS controlling the charging of the batteries from the alternator and WS500.
Hi Henrick,
Thanks for the kind words.
You’re right. But note that that diagram is provided by WakeSpeed to help people understand how to hook up the WS500 to Victron Lynx system, not as a complete system schematic of the Victron side. Also note that I cut that diagram out of a complete instruction sheet.
Wakespeed are assuming that users will read the Victron manuals and act accordingly and anyone who does not has only themselves to blame, in my view.
Hi John,
Just catching up on your latest lithium articles which I applaud you for trying to enforce good system design with, it isn’t as simple as it may seem. I have largely ignored lithium batteries in boats as I fall relatively cleanly in the group of people who are served perfectly adequately by lead acid and do not see the effort and expense as being worth it. On the other hand, I do follow it for cars as it matters directly to my car. And this is where I see an interesting similarity in that the information around cell balancing for cars is all over the place and kind of poor just like you have found with boats. This seems rather alarming as cars are much larger investments and done by much larger companies with huge engineering, training, etc. teams and not single person stuff like Al. On our car, I could find no real documentation on what to do and only some internet forum posts which are generally much more concerned with only charging to 80% or so than to balancing occasionally. There doesn’t even seem to be a mention of how the cells are balanced and only by using diagnostic software can you find that they are top balanced (pretty predictable). Some of the other cars out there have documentation that mentions balancing and some even suggest doing it as often as monthly.
It is going to be really interesting to see whether they either automate these functions well enough or do a good enough job of educating consumers. Given how poor people are at maintaining things, I fear that the educating method will fail. It certainly seems possible to convert much of this from battery speak (“if you have x chemistry, charge to 80%”, “top balance”, etc.) to normal speak (“press for maximum range/capacity when needed”, “do an extended charging session soon”). Cars provide a lot of data points and while what they are doing is not perfect, the more modern packs do seem to hold up pretty well.
In our case, I have a quarterly calendar reminder to top balance as we usually charge to 75%. I certainly hope that the boating world figures this out more quickly than the automotive one is but the odds seem stacked against it.
Eric
Hi Eric,
Wow, that’s interesting and surprising. I would have thought that the car guys would have had optimal charging and balancing fully automated by now.
By the way, as far as actual telling the user when top-balancing is required, Victron do exactly that (see the screen in the article above) and seem to be one of the leaders in thinking about this.
As to boats, the more I research this the more I realize that things are way more complicated around good lithium charging practices than most people think. I think the result will be that many (perhaps most) owners of lithium batteries on boats are going to get way less cycles before failure than they think they will, or the manufactures promise. My guess is that this problem is still hidden because most cruisers who have installed lithium have done so in the last few years and so have used few cycles so far.
Probably the biggest problem is solar controllers that start a new cycle every morning, and shore power chargers that do the same every time they are turned on. Unless managed, both then stay in acceptance for a fixed time each cycle. And for people who motor a lot, the average “smart” (read stupid) alternator regulator is as bad.
Again, Victron seems to be a leader in actually allowing the BMS to manage solar and shore power chargers automatically, but only with their top end systems.
Bottom line, until we have full systems that measure and act on tail current going into the battery (ex loads) these systems will need a lot of manual managing to have long lives. For example turning off solar when not using the boat.
But then it was ever so. It was only 15 years ago that we at AAC realized, (and wrote about) that most-all common “wisdom” about charging of lead acid batteries was rubbish and most all charging sources were as much battery wreckers as chargers. https://www.morganscloud.com/2010/08/02/agm-battery-test-part-1/
Heck, there are still to this day boat yards that insist on putting lead batteries on trickle charge over the winter at far too high voltages. And people still insist that lead acid batteries benefit from being charged very slowly, even though the opposite is true.
Hi John,
Good to hear that Victron are thinking about it as more user centered by asking the user to do something but not alarming them than battery centered. And agreed on how long it has taken people to understand how to charge lead acid, heck most of the stock charge programs are still terrible. I learned how to take care of those from building electric cars that used golf car batteries but still killed plenty of house batteries in boats until solar came along.
To be fair, longevity studies on recent BEV’s have actually shown great battery longevity to the point where I have no worries about needing to replace a battery pack due to mileage and while gains are possible, they may not be deemed to be needed. The news does love to dwell on stories of ride share drivers killing packs in 100k miles but when you dig a bit deeper, the pack was badly abused and an internal combustion engine would not have lasted with comparable abuse either. However, some of the older BEV’s had quite poor longevity. The most well known case I believe was due to using passive thermal management primarily which proved definitively that active thermal management is key to longevity in that use case. For example, our car when unplugged will use the battery to heat the battery at <0F and if you are plugged in, it starts doing it at 30F with a similar routine for hot temperatures, it actively manages while driving, and most of the cars now preheat the battery before fast charging. I don’t know if balancing has had anything to do with the better longevity of recent batteries but it is certainly possible.
Watching the automotive world is going to be really interesting as there are many more data points and they are better at collecting data than the marine or RV world. It is obviously a different use case and sometimes different chemistry but still instructive. The 2 that I am really interested in watching are BEV’s with no home charging as they will tend to do much deeper cycling (which has been shown to hurt longevity although not catastrophically like lead acid) and they won’t keep their batteries in as narrow a temperature range, and plug in hybrids which are designed to really cycle the battery hard unlike full BEV’s that tend to use <20% of capacity on a day to day basis.
I hope this wasn’t too much off the subject of boats. We will likely replace our 9 year old house bank in another year or 2 and that will be lead but I could see the replacement after that going lithium and I will need to educate myself then.
Eric
Hi Eric,
No worries about being off topic, particularly given that we can learn a lot from electric cars. I was particularly interested to learn that BEV preheat the battery for fast charge.
Also, useful on a personal note given that Phyllis and I have been thinking about electric cars. Right now in Nova Scotia we would be buying a coal burning car, but that will change as hydro power from Labrador comes on line so we are getting ready and even wired our new garage for although did not actually buy the plug etc.
Hi John,
Interesting comparing your views on “must haves” for monitoring Li Batteries vs the claims for the “Juice” Li batteries assembled (they buy in the cells, package them and fit their proprietary BMS) and marketed by Enertec Marine in NZ
https://juicelithium.co.nz/technology/
In particular they say the BMS is monitoring voltage (3), there is active cell balancing (8) and there is data logging (11).
Question for me is how much of this info is directly accessible to the user, ie does not need a technician with a laptop to interrogate the BMS. I will be catching up with these guys in a couple of weeks so will be asking these questions.
cheers
Alan
Hi Alan,
Given that they have a NMEA 2000 output it is probably possible to get some data out, but that may or may not include cell state. Also, given that they have a relay designed for generator starting one might be able to use that to control charging sources. The only way to know is get hold of their manual and have a thorough read through while armed with a list of required criteria. Talking to the vendor, while useful, is rarely a total answer because they will always skew their answers to make the sale. Just look at the crap spewed by “drop in” battery sales people.
This challenge of sifting marketing claims for reality is exactly why I’m writing this series and will end it with a check list of required capabilities.
All that said, after all of two minutes on their site, their products look like a good possibility.
The Juice BMS is stated to have specific outputs for Charger and Load Control per (6) in
https://juicelithium.co.nz/features/
This is a requirement from “AS/NZS 3004.2 (Part 2)2014, Electrical Installation Standard for Marine Installations”, the controlling/obligatory standard in our part of the world.
If I can source it I will post a link to their manual.
Hi Alan,
Sure, each of us needs to read the manual of any batteries we are considering. That said, just to be clear, while I may read the manual out of interest and part of my learning, I will not be opining on the suitability of a given piece of gear for a given use. Rather my goal is to write a complete buyer’s guide with a criteria check list so that members can make their own decisions.
Hi Alan,
Enertec installed our Mastervolt electrical system; https://www.morganscloud.com/jhhtips/while-converting-to-lithium/
back when they were the MV agents in NZ. MV subsequently bought BEP electronics in NZ so establishing MV NZ direct. This prompted Enertec’s move to create their own Juice range of products.
Anecdotally, a friend was fitting out a new power-cat for fishing and diving – he wanted to fit Lithium batteries and asked me who we had used and I referred him to Enertec. When I saw him a few weeks later he told me Enertec had persuaded him against installing lithium, as he didn’t have a dedicated battery locker that could keep them dry and cool, meaning Enertec didn’t secure any business.
We average over 100 nights aboard our yacht each year and as of last month, since commissioning in 2016 we have used 165 cycles without any system glitches. The MV integrated electronics and BMS handle the issues John describes above. So Enertec will fully appreciate these issues and as I understand it, worked with Auckland University to design and test their own BMS functionality.
If you ask good questions, I am 100% confident you will get straight and helpful answers.
Hi Rob,
Great to hear that Enertec talked your friend out of lithium when it was not appropriate. I would call, that the ultimate endorsement.