Most battery manufacturers specify battery life expectancy in terms of number of cycles at a certain maximum discharge level. Transforming this data into the expected or remaining battery life for a house bank is challenging. In measuring the number of cycles, for example, what depth of discharge qualifies as a “cycle”?
Our house bank consists of eight Lifeline 8D batteries. Lifefline reports their batteries are good for 1,000 cycles at no more than 50% discharge, but nearly all batteries go through a few episodes of super-deep discharge where someone forgets to start a generator, is away from the boat longer than expected, a shore-power cord is inadvertently unplugged, or a shore-power pedestal breaker opens. With breakers increasingly sensitive to even tiny current leaks, we expect accidental battery discharge to become more common. For us, we went way more than 10 years never having seen a dock power problem, but have seen five or six breakers pops this year due to water intrusion either from heavy rain or the waves rolling over the shore-side power pedestals.
Our first set of batteries went four years in very heavy use, but as expected, a couple of deep discharges occurred. On the next set, we started an experiment where we did one thing better for the batteries and one that was worse. On the better side, we have a charging, monitoring and automation system that prevents any deep discharges and records battery state throughout their life. This second set of batteries has never been discharged below 50%.
We waited four years to see how they did. To confound the experiment somewhat, we started running the boat as a condo and just use the power and live life as though we were always on the grid. This drove up the cycle count per year dramatically. We have spent the last four years mostly using 16A shore service. And the way we run the boat, our peak draw is way over 16A and sometimes during waking hours is more than 16A (laundry on a cold day while heating the boat). This means we often are cycling the batteries even when at the dock.
The experiment on the second set of batteries was completed in UK at the beginning of 2018, where we replaced that second set. This battery bank had the advantage of better care, but far more cycles, and it also lasted four years. Depending upon how you count, we had 1,500 to 1,600 cycles on these batteries. As long as the batteries are well-cared for and there are no mechanical faults, as the cycle count climbs, the capacity decreases but they otherwise operate unchanged. The capacity of these batteries just kept declining and when we replaced them they were still doing fine but the capacity was down to one third of the rated capacity.
The decreased capacity is apparent in the table below, showing the first few rows of our power state report for while we were at anchor shortly before (left) and after (right) replacing the house batteries earlier this year. Our autostart system starts the generator when the batteries fall to 55% charge and shuts down at 85% charge. Even if we hadn’t noticed it ourselves, the report makes it clear that with the reduced-capacity batteries, the generator was running more frequently and for shorter duration than with new batteries.
The table below summarizes the run statistics across the two anchoring periods. With the old reduced-capacity batteries, the generator was running roughly every two hours for an average charge period of only 1 hour. With new batteries, the average time between generator runs increased to 6 hours and 25 minutes with an average charge period of 2 hours and 4 minutes.
From the above table you can see that we’re running less efficiently on the old batteries compared to the new. The generator produced 41.9 average daily kWH on the old batteries, versus 30.4 on the new. And it was running more, on average 8.4 hours per day compared to 5.2 on the new batteries. But from a livability perspective it really wasn’t noticable. The generator didn’t feel like it was on more often than usual and everything else just worked the same as normal.
Why run the battery bank up to 1,500 to 1,600 cycles rather than the manufacturer’s estimate of 1,000? The easy answer is that at retail price an additional year is worth $1,800. But another factor is with auto-start the reduced bank capacity of older batteries is hardly noticeable. The generator can run infrequently for 2 hours or very frequently for 1 hour. And even with the old bank desperately needing change, the generator is simply running more and otherwise not causing us any inconvenience. While we were anchored at St. Margaret’s Hope shortly before the recent battery bank replacement, we made two day-long trips to Neolithic Orkney and Stromness and were perfectly comfortable leaving the boat during those periods.
Deep discharges are hard on batteries. But based upon our first bank’s life, with multiple deep discharges, and that of the second, with no deep discharges, it appears that one or two deep discharges isn’t a big deal. We suspect four to six probably is, but that’s mostly speculation at this point. We’re hoping our current set has no deep discharges over their entire life.
We are very dependent on auto-start to maintain the battery bank in good condition and to relieve us from the responsibility of always being there when charging is needed. A well-designed auto-start system will shut the generator down if there’s a mechanical problem and, of course, it’s possible the generator might fail to start. At retail price, our battery bank is worth $7,200 USD, so protecting this investment is pretty important. We have indicator lights, alarms, and email notifications if a control system or generator ever fails, but the most reliable defense is a second generator with an independent auto-start system. We didn’t have space for a second generator, but the main engine is capable of delivery 9kW of power, so we put auto-start on it. Effectively we have a 12kW generator and a 9kW emergency backup, both provisioned with autostart.
Abuse like deep discharges or improper charging can destroy batteries in far less than a year. If you can avoid that, then cycle count and depth of discharge is what drives battery life. The cycle count failure mode is usually a slow and steady reduction in bank capacity rather than anything suddenly needing to be changed. Because we live on the boat year-round and are often at hook or plugged into low amperage services, we wind the cycle counts very quickly and 4 or 5 years of battery life is where we end up. I suspect a more typical usage mode with good care could easily yield 8 to 10 years. Telco backup batteries are rarely cycled, usually well-maintained, and they can last 20 years.