Battery wiring for a 24V series / parallell bank

[la7qz]

Hi

Through my work I often come across yachts with 24V battery banks consisting of two or more 12V batteries. Sometimes I see variations of battery voltage withing the bank. For instance, if a battery bank consisting of four 12V batteries in series/parallel is at 26V, one series pair of batteries may be at 12.8 and 13.2 while the other pair are at 13.1 and 12.9V.

Would it not make sense to link the batteries at the 12V level to ensure a more even charge across the whole bank? I have never seen this done on a yacht installation. It would cost a few dollars worth of copper, but not much considering the cost of a yacht large enough to warrant a 24V installation. The biggest 24V bank I've seen built up from 12V batteries on a yacht consisted of 10 Mastervolt 8D Gel batteries.

My experience also makes me curious what kind of cell voltage variations one might see within a 12V battery.

Owen

[dnix71]

If there are two sets of batteries on a yacht, one set is probably for the engine and one for aux power. They are isolated so if one set fails completely, the other can be used to start the engine.

Resting voltage above 12.8 is fully charged anyway. Gels have a lower top voltage limit than flooded since they aren't supposed to gas.

[la7qz]

I'm not talking about two sets of batteries. I am talking about one 24V bank consisting of four or more batteries.

I'm a marine electrician by trade, so I'm used to setting up yachts with more than one battery bank. Usually a house bank and an engine cranking battery. Maybe another cranking battery for a generator and a battery installed up in the bows for the anchor windlass to avoid running really long fat (expensive) cables to a large consumer.

Owen

[dnix71]

No two batteries are exactly alike. When you charge and discharge them in series you will have a voltage difference.

The only way to guarantee equal voltages is to do what cordless tool makers are beginning to do on their high end stuff, have a wire to each battery in the series and a chip that dumps excess current when that battery reaches a full charge. That's less energy efficient, but it allows proper full charging and equalizing every charge cycle.

http://www.rc-truckncar-tuning.com/rc-car-battery.html

If someone can afford a yacht, they can probably afford that kind of charge monitoring.

The better EV car makers also do this.

[scottsAI]

La7qz,

Wiring can make a difference.

Incorrect, yet if the wires are very heavy (bus bars) then does not mater.

Pos-B-B-Neg (ignore the dots)

......|.|

......B-B

The wire resistance causes the two batteries at the power take off to supply more current then the other two batteries. A batteries internal resistance is very low, depending on the connection and wire gauge, the wire resistance is higher than the battery, full load sharing between batteries is not happening.

Better:

Pos-B-B

......|.|

......B-B-Neg

Proper load sharing between batteries. (Assuming wire and connections are the same.)

Some believe the 12v taps should be tied together. (I do not, no way to watch the batteries) If the batteries are healthy it should not matter.

The battery has not been equalized and over time develops an imbalance between the cells.

Equalizing charge will rebalance the cells.

Many vehicle systems charge to 14.4 volts and call it good. (maybe temperature compensated) Equalizing voltage is higher. (depends on battery, SLA do not equalize).

Equalizing is OVER charging the battery (thus bad, not done often) The intent is to get all the cells fully charged, so has a good outcome if not over used.

Consider rewiring the two high voltage in series and low in series. (Like cells in series is goal)

Have fun,

Scott Beversdorf.

[Flux]

I am inclined to agree with Scott on this one. Joining the 12v points will just convert each 12v block into a parallel bank. If all is well it will be fine but as with any parallel bank a rogue cell will drag the rest down. Without this link you have a better chance of picking the rogue up on low volts before it does more damage.

Equalising is the only way to keep things in step without some complex current sharing scheme.

I suspect that schemes that monitor individual cells and share charging or load current are excellent if you have them but more likely to be of benefit towards the end of battery life when things are on their last legs. Starting with new and identical cells things normally work until some cells start to die and it becomes a struggle to keep things going beyond that.

Flux

[la7qz]

Hi Scott

Yes, I'm aware of connecting the positive and negative wires at opposite ends of the battery bank. This should be standard practice, but very often isn't done. It is in fact the most frequent modification I make on battery banks on yachts.

Thanks for the input on not joining the batteries within a bank at the 12V posts to make it easier to pick out a rogue battery. There's always something else to think of...

And to dnix71. No, most of my customers are NOT filthy rich. Many of them own only a yacht and nothing else. No house, no car, no mortgage, often a low, or no fixed income. These people live full time on their boats and many earn money by picking up odd jobs as they cruise. They need simple systems that work and can be easily serviced anywhere in the world. Complicated electronics like the charge monitoring system you talk about have no place at sea.

I paid US$ 24000 for my yacht. I will have spent around the same on fixing her up by the time she's ready for the oceans again. I live aboard full time with my wife and dog and work a full 8 hour day earning the money to get the boat ready. The only way I can afford that is by not having a house or car. Having a yacht does not mean you're rich (at least not if riches are counted in money).

Owen

[la7qz]

Hi again

Thanks for the comment.

On equalizing, many yachts, particularly those big enough to have a 24V system now use batteries that can not be equalized. The last one I worked on had sealed wet batteries which is probably the worst possible battery for a boat... I don't like gels much either.

My recommendation is wet (preferably Trojans) if you are willing to maintain them and AGM if you are not willing to maintain or money is no object.

On my own boat,the house bank is four T-105's wired for 12V and the cranking battery is an Optima red top spiral AGM which I got from a megayacht where it was one half of a 24V generator cranking bank and the other battery was bad.

Owen

[scottsAI]

Owen,

The batteries in my UPS are SLA (sealed lead acid)

Equalizing is out, yet printed on the side of several different brands is:

Charge to 14v

Cycled 14.5v

Run time on a UPS was below half new.

Took batteries out, lightly cycled them charging up to 14.5v, twice.

Tested 3/4 of new life! Wow, so charged all my SLA to 14.5.

No SLA charger of mine charges above 14v, one 13.89v?

No UPS charges above 14v, most are dead on.

My SLA were expired when I got them (cheap) in service 3 years so far doing well.

My Dad and I came very close to buying a sail boat. After I got out of the Navy, he was remarried no longer interested. Wonder how that would have turned out.

Have fun,

Scott.

[dnix71]

My parents lived on a converted Japanese trawler in Dunedin for years, but they never called it a "yacht" and they were and still are of modest means.

In south Florida if you live on a "yacht" then you have to have money because that kind of dock space is expensive and hard to find.

My parents boat had a 400 gallon tank for diesel and the aux genny was a Ford Lehman that ran on compression alone, no glow plugs. They had more power away from the dock because of the genny.

It was only a 12 volt system if I remember right.

Wet batteries are a real hazard in rough seas, I can appreciate the safety of gel cells, even though you have to baby them.

A 24volt desulfator running full time might be worth experimenting with.

[Madscientist267]

I'm pretty sure the idea behind UPS charging schemes is based on how the batteries are connected to the final MOSFETs. Since most rely on the freewheel diodes to do the dirty work, regulation is out of the question without switching somewhere on the DC side to prevent overcharge (something I have yet to see).

This being the case, 13.8 is a half way decent balance between gas production and undercharge. Because a UPS sits for long periods with a 'dumb' charge applied, they limit the voltage to reduce gassing and premature death. This comes at the price of not completely charging the battery. From what I gather, UPS designers are not particularly worried about sulfation - UPS batteries are being grossly abused just in normal use. Anywhere there is a 60 or 100A fuse (in the case of 'personal' UPS units) on a 10AH battery, I consider that abuse. Run a couple more of those '3/4 capacity' tests on your UPS, and you will discover what I mean... It will fail you when you need it most. I don't test mine anymore, and the batteries last much longer now (in terms of lifespan).

Although once a topic of great debate by others (in some of my early posts), I have had much better luck with FLA on (modified) UPS systems than I ever did with SLA. I attribute this to  the difference in each chemistry's "preference" for a given voltage. I still maintain that FLA hits full charge at a lower voltage than SLA does, although many disagree with me entirely and emphatically. Shrug.

Steve

[scottsAI]

Steve,

Capacity test took about 2 hours, 4 amp load on 12ah battery.

Trying to be nice. Killed test at 11.49v.

Lately we get enough power failures to test without adding more.

Size and load UPS for 20 min run time, turn off 5 min or so after power failure.

Should connect data cable and have computer tell it to shut off. One more wire...

Doing well to get 5 years from SLA, one past 7.

Expired SLA were cheap, with all the electric scooters and stuff the price is much higher. (cant find them) May switch over to FLA .

You could be right about the charging voltages, cant find schematic to see how they work.

I remember you doing something different to modify? (looked could not find it) Did reread OS poll.

Have fun,

Scott.

[Madscientist267]

Yeah definitely go easy on them if you are going to test them... They dont hold up well otherwise, I promise.

As for schematics, if they are APC units, they released the diagrams at some point for several models. I actually kept a few of them because it took a little digging... try a google search, if you don't have any luck, let me know what model you have and I can see if I have it...

The main 'modification' (if it can really be called that) was to reroute the battery connection out of the rear of the unit so that it can be connected to an external battery. For convenience, on some of them I introduced an extra connector. A 30A inline fuse along with a fan helps protect things. They have the ability to run much longer now than they originally did, and as you probably know, the heat sinking on the MOSFETs isn't really designed for extended run. The 24V units already have a fan built in, so all I needed to do was reroute the cabling.  

5 (and 7) years is quite a feat from my experiences with these things... Must have got your hands on a good batch or something

For my 12V units, I'm using cheapy U1-3 lawn 'n garden variety FLAs for the batteries. While not exactly ideal for this application (they're cranking batteries), they give me between 2 and 4 years depending on the usage they see, and only hit me for about $25 a piece after core. Much cheaper insurance than replacing them with the real-deal replacement versions from APC, and they have a longer runtime than their internal counterparts if I really need it!

Steve

[scottsAI]

Steve,

Wow if you have the schematics... spent several hours over the last few years without finding them.

Just looked again, schematics are for smaller units. Need bigger than 1kw.

Have APC SmartUPS: 5k, 3k, 2.2k, 1.5k, 1.4K, 1k and smaller not so interested.

Want to backup home with 4 of the 3k. Standby power is 47w each.

200w 24/7 is 10% of my electric bill, too costly for stand by power.

Realized most of the power is going into the transformer, gets warm without load 115F. (Battery current is near zero).

Magnetizing currents in a transformer account for several percent of full load. (about right).

Want to remove the transformer from the circuit, except when needed.

The AVR function can still switch in/out if the power line goes out of spec.

Have fun,

Scott.

[Madscientist267]

Scott -

I'm not ignoring you... things have been busy lately. I tried to look around on my server from work, but the connection to the web from here is so slow that I can't get much hunting done (filenames aren't all descriptive..) - I'll take a look when I get home. I don't think the ones I have are for anything bigger than the SmartUPS 1400, but I'll see. A friend of mine might have some too, he's into this kind of thing; I'll check with him and get back to you.

Steve

[scottsAI]

Thanks Steve!

SmartUPS 1400 is good!

Looking at the parts in the 1k and up, all look the same, maybe bigger.

Scott.

[Madscientist267]

I apparently dont have the 1400 schematics - and Im having trouble even finding the smaller ones. I know I had them, several different models even, and used them to mod one of my 650VA units... sigh.. Organizational skills apparently need to be refined LOL

I had a server crash about a month ago, and I'm still sorting out the mess - I'm sure they're around somewhere; I'm paranoid about backups... now the key is... where?!

If I find them, I'll let you know.

Steve

[scottsAI]

Thank you!