Some time ago, while searching the internet for information on how to properly wire together 12V batteries in multiple banks, I stumbled upon this quite unique article by SmartGauge Electronics: How to correctly interconnect multiple batteries to form one larger bank.
The bottom line of that article is that there are commonly used ways to wire together multiple batteries in parallel, which would result in a current imbalance between them. In particular some of the batteries in a wrongly wired bank will end up providing more current to the load than the others and therefore will discharge faster. Also while being charged, some of the batteries will receive more current (and thus charge “more”) then the others, resulting in an unevenly charged bank. This imbalance between the batteries may lead to faster ageing and eventually a failure of the whole battery bank.
The same article also mentions the proper way to wire together multiple batteries so to reduce or completely avoid any imbalance. The author states that these results are based on his 20-years experience and on some computer simulations he had run to confirm his intuition and calculations by hand.
Fast forward to July 2014. While sailing in Croatia on Kismet one morning we woke up with a “warm” surprise. The two main fridges (Kismet has 3 of them) were not working and the temperature inside was above the expected 4 degrees Celsius. At first we thought the compressors were gone bad, but both at the same time?!? After some coffee to fully wake-up :-), a quick trouble-shooting showed that our battery banks had a voltage of around 23.5V, while normally it should have been well above 24V. Not good!
It turned out that control unit for the fridges has an embedded safety feature that shuts off the compressors if the battery voltage drops below 23.7V, so that was why they were not working any longer. Luckily the third fridge has a lower safety threshold, so we still had some cold beers ready for the post-fixing. But I am digressing…
We disconnected all the loads from the batteries and went on the measure their voltage. What we found was that each 24V bank (we have 3 of them) showed a different voltage and even the two 12V batteries, which are in series, within each bank were not at the same voltage. A mental alarm went immediately off. Imbalanced batteries!!! What to do?
We noted that within each 24V bank, the “front” 12V battery, which means the one closer to the load, had a lower voltage than the other one. This was the same for all banks. Suddenly my mind went back to the article from SmartGauge. A quick check to the ways the 3 battery banks are wired on Kismet, especially the way the load is attached to them, confirmed the supposition. Our batteries were not wired correctly, which over time (they were replaced in 2010 according to the records) caused so much imbalance that the front ones were about to fail.
As a temporary workaround we swapped the front and back batteries in each bank, which improved the situation (no more fridge shutdown during the night), and made a note to solve the issue once Kismet was in the shipyard for some major work on her.
Once back home I read again the article from SmartGauge and decided that it would make sense to re-run those computer simulations to check whether I could get the same results. After all I have an ME in Electronics and I should live up to the expectations, right? Just kidding! 🙂
Anyway, I did run those simulations using iCircuit for Mac and found out the same results mentioned in the article. Duh! Here some evidences.
This is for an asymmetrical wiring, a common setup found in many battery banks.
All the resistors that are in the circuit are meant to simulate either the resistance of the connection cables or the internal resistance (series and parallel, the latter also called self-discharge resistance) of the batteries.
Look at the currents shown at the bottom. The battery closer to the load (100 Ampere in this case) provides 35A, while the farther one provides only 17A! So the closer battery provides double the current than the farther one. The situation reverses for the charge, where the closer battery gets more charging current than the farther one. As a consequence the battery closer to the load and the charger cycles more deeply than the others, which also means it ages much faster.
Here is the same simulation but with a proper wiring. Note the full symmetry of the wiring (beautiful, isn’t it?). The current from each battery travels paths of equal resistance, which makes the current even from all batteries. In other words, in this setup each battery provides (and receives when charged) the exact same amount of current. As a consequence all batteries cycle in the same way and while they will eventually unbalance due to some small internal differences, this will take much much longer than with the previous setup.
Here is the sad conclusion then. Kismet’s batteries have been suffering from wrong wiring and we did not even know! Here is how the current cabling, the one that eventually lead to failing batteries, looks like.
Note how the load is connected closer to Bank 1 than to Bank 2 and 3. This is very similar to the simulation first case above. Bank 1 will end-up providing much more current to the load than Bank 2 and 3.
Surprisingly enough the fix is quite simple. Here is how we are going to re-wire the banks.
First, the negative cable from the load will be moved to the negative pole of Bank 3. Second, we will make sure that the connecting cables P1 and N1 have the exact same length and characteristics (cross-section in particular), so they have as much as possible the same resistance per unit of length. The same goes for the N2 and P2 pair, and for the B1, B2 and B3 trio.
Note that with this new wiring the current from each battery toward the load travels exactly the same total cable length.
For Bank 1, from the positive pole close to the load the current travels B1 + N1 + N2
For Bank 2, from the positive pole on the bottom right the current travels B2 + N2 + P1
For Bank 3, from the positive pole on the bottom right the current travels B3 + P1 + P2
Since the three paths above all have the same total cable length and characteristics, the same exact current will be drawn from each bank and each battery. Cool, isn’t it?
Now back to you. How are your batteries wired? 😉
As always, fair winds and following seas!