Electricity Part 3 – Battery management systems

Are you a total battery nerd, or setting up your own lithium iron phosphate battery system? If so you will probably find this post interesting. If not I’m betting you will find this all pretty dry. I think this stuff is cool, but I am deep (way deep) into setting up the systems upon which we will rely for the next couple of years of cruising.

If you remember from my last excellent post Electricity Part 2, we now have a fancy house battery that is basically a large cell phone battery with slightly different chemistry (LiFePo vs LiPo–that iron (Fe) makes it much less likely that the battery will spontaneously catch on fire). In case you were not sure, I will refer you to one of Einstein’s lesser known formulas: fire + boat = bad. A LiFePo battery does the same job as the lead-acid battery under the hood of your car, but goes about it in a very different way.

While researching this new battery technology I ran into the debate about whether or not to actively manage the battery bank on the per-cell level. There are people out there who say “Pshaw, just plug ’em in and go, a series string is self-balancing.”

Self balancing? Huh, that sounds important.

DSC_0323
The copper links connect the yellow batteries into a series string of four cells.

Our house battery is really four three-volt batteries linked in series, negative to positive. This has an additive effect, so the bank ends up at a nominal 12 volts. Balancing comes into play because each cell can have slightly different voltages. A low cell could be at 2.8 volts while another one could be at 3.4 volts, with the remaining cells at 3.0 and 3.2 volts, which all add up to 12.4 volts overall. So what?

Charge Curve
Not science. My best guess.

Well, here is the problem. In lead-acid, the voltage of the battery varies in direct proportion to the amount of energy in the battery. When LifePo absorbs (or discharges) energy, the voltage does not vary by much until right at the full (or empty) point; then the voltage increases (or drops) very quickly. If this goes too far, then the battery is ruined, flushing between $500 – $2000 down the drain. When you charge a battery, or a battery bank, it is generally charged to a set voltage. In our case 14.4 volts would be ideal because it works out to 3.6 volts per cell. Fully charged is 3.8 volts per cell; 3.6 gives a little safety factor. The problem is, the charger does not know the voltage of each individual cell–only the total voltage of the pack. If one cell manages to hit 5.4 volts while the others are all at 3 volts, it still adds to 14.4. The charger thinks, Great–we made it to 14.4 volts, time to shut down (or more accurately, it can’t push can’t more electrons in); but that one cell that hit 5.4 volts is now garbage.Cell overcharge

Hey, those guys on the internet said the series is “self-balancing”, so what is the big deal? The other thing they said was, it is necessary to very carefully manually balance the battery pack before you start to use it. So, naturally I took the batteries out of the crate and stuck them directly in the boat without balancing them and with no active management system. Believe it or not, this worked out ok in the short term.

As the battery bank charged and discharged over a three week family boat vacation last summer, I obsessively checked the overall voltage as well as each individual cell. Here is what I noticed: the series string is self balancing, but only when it is discharging. Once I thought about it, this made sense. The cell the the highest voltage is the cell with the most “push” and it discharges at a slightly faster rate until it no longer has the highest voltage; then the top two cells (now holding hands) are discharging at the same, slightly faster, rate. Eventually all four cells will be the same.

The problem is charging. Due to small variations in manufacturing, each cell holds a slightly different amount of energy. This means that when we charge the system at 14.4 volts one of the cells will “win the race” and be charged up first. Practically speaking, with my three-week experiment, this means that when you discharge your battery bank by 2-3% and then neurotically charge it back to full every couple of hours, the fast cell ends up getting further and further ahead. At the end of our three week trip, the high cell was at 4.2 volts (not yet at self-destruct) and the low one was at 3.2, with the whole bank at 13.8 volts. I think that if I had done the correct thing and carefully balanced the bank, things would have stayed much more in line and maybe, just maybe, the battery bank may have started to “self balance”–if I’d have ever let it get below 95% charged.

IMG_2908
Balancing the battery bank in my basement. Simple incandescent light used to draw down the power in the fullest cell.

Still, the one battery cell at 4.2 made me a little nervous. So, I spent a little more money (it’s a theme) and bought a basic battery management system (HousePower BMS). With the batteries in our basement, I started the slow process of making the four cells match with exactly the same voltage. I started by using a light bulb and the new BMS (battery management system) to get things all in line. I then realized I could simply wire all four cells in parallel and leave them alone for a couple of days, and they’d all end up at the same voltage.

IMG_2915
Red wires between each terminal are the shunts/cell monitors.

The BMS does a couple of different things. First, there is a shunt on each cell. The shunt reads the individual cell voltage and will apply a small load (0.7 amp) to the cell when it gets over 3.7 volts. This has the effect of “pulling down” the cell, or cells, that are too far ahead. Secondly, I can wire the BMS to sound a buzzer if any cell voltage should go to high or low. Lastly, I can wire it to trip a relay to disconnect the battery from the system in case of an overcharge or critical discharge. So far I have only used the shunts. I have a buzzer and a disconnect relay, which I will wire into the system when I re-install the batteries in the spring. In our basement, the system seems to work well and stay in balance. Charging at 16 volts, I get two of the cells to start shunting a little before the others, but this is pretty extreme. I think once it is in the boat, the system will behave itself; maybe after a year or so I will have to wire the whole thing in parallel again to get it back in balance.

Wow–you have nearly made it though this whole, very boring, post!

IMG_2794
Battery monitor. There it is. Do you see it?
IMG_2794
The grey circle is the Victron monitor thingy.

 

 

 

 

 

 

The final item we have for monitoring the state of charge for the house and starting batteries is a Victron smart battery monitor. The monitor measures the amount of amps flowing into and out of the battery bank. The monitor is programmed with the particulars of our battery; then it does the math and tells us what percent of charge we are at. It also tells us how much current is flowing in or out at any given time, plus the voltage of either the starting battery or the house battery. It is like a gas gauge for our batteries.

Up next is the final installment in the Electricity series: Electricity Part 4 – Getting the Electricity In There. That one will be a little more interesting, and will lead us to the new and exciting Engine series, in which I finally explain the origins of the Frankenstarter.

 

Leave a Reply

Your email address will not be published. Required fields are marked *