As the title suggests, this post is going to be technical, and probably boring for many readers. If you aren't interested in batteries and battery charging, on sailboats, go back to what you were doing. I'll be posting another article before I leave Tonga, which may happen in less than a week.
If you are looking for a general theme for this posting, it could be something like how much there is to learn about this cruising lifestyle and how the learning process is constant. Or, how you think you know a lot about something, only to later discover how wrong that thought was and how much more there is to learn. Or you could think of the theme as being about batteries and power.
While I was sailing between Bora Bora and Beveridge Reef, as on many passages, I had a lot of time to think about random things. For some reason I started reviewing a conversation I had had with a cruising couple a short time before, talking about their battery bank, my battery bank and power in general. Something they said reminded me of something I have heard from other cruisers and seems to be generally adopted as true, which is that solar power is not able to fully charge your ships batteries - every now and then its good to run a generator or plug into shore power in order to fully charge the batteries as doing so is good for them. I had always found these comments somewhat strange as on Luckness my solar was able to fully charge my batteries, on a sunny day, early in the day - I had plenty of power...or so I thought. I had also noticed that my battery bank seemed to be degrading and I thought that I would need to buy a new battery bank when I arrived in New Zealand. The battery bank seemed to not be holding its charge very well, which is one of the signs that the batteries you have are nearing their end of life.
While out in the middle of the ocean, on July 4th, many days from anywhere, I decided to dig my battery manual out and read it, followed by my solar controllers manual, and read that as well. I had read both manuals a few times while back in Seattle and thought I understood everything those manuals had to say, but this time I read the manuals and discovered something new. After understanding what the manuals said, I realized that I had been under charging my batteries for the entire time I've been away from docks, not plugged in. Undercharging batteries is bad for them and can shorten their lifespan.
At this point in this post I have to start getting more technical.
My boat, as with most boats, is powered with 12 volt, lead acid batteries. I have 400 amp hours of battery capacity on board, four group 31 Odyssey batteries. You can think of this as four large car batteries for simplicity, although these are quite a lot larger than car batteries. Being a battery, you can't just look at it, like you can a water bottle or lighter, to see how full it is. You can't weight a battery, like you can a propane tank. The definitive way to learn how 'full' a battery is, is to measure its Open Circuit Voltage, OCV. To do this, you disconnect the batteries, let them sit for at least 6 hours, preferably overnight, and then measure the voltage across the battery.
For my Odyssey batteries, the OCV voltage of a fully charged battery is 12.84 volts. 90% of capacity is around 12.7 volts. 80% is just under 12.6V. 50% is 12.2V. At 12.0 volts the battery is at roughly 35% state of charge - you hope never to see voltages this low.
Nobody ever measures Open Circuit Voltage - its highly impractical to disconnect your battery bank and wait 6 hours on a cruising boat. The reason you need to disconnect the battery and let it sit for a while in order to have its voltage be a meaningful measure of its state of charge is that batteries are not at all like bottles of water in terms of being full or empty. The lead plates inside the battery have two types of charge - a 'surface charge' and a deeper charge. I think of electrons hovering around the surface of a plate being the surface charge, and electrons deep inside the lead plate as being the deeper, or 'real' charge. In order for a battery to be in a fully charged state, the voltage for the entire lead plate, inside and outside, needs to be at the fully charged voltage. You can bring the surface of a lead plate to a high voltage fairly easily, the hard part is to have the charge be absorbed by the entire plate. I'll return to the surface charge being absorbed into the interior of the lead plate a little later.
As measuring the Open Circuit Voltage is impractical, most cruising boats use a battery monitor to keep track of their batteries state of charge. There is a device called a shunt which is able to measure current (amperage) going into or out of a battery. A battery monitor is hooked up to a shunt which is hooked up to your battery bank. The battery monitor monitors the shunt which is constantly reporting back if the battery bank is being charged, depleted, and by how much. Think of a shunt as measuring the flow of electricity into or out of the battery. If everything on your boat is turned off, the shunt will report '0' - no activity. If you then turn a light on, the shunt may report 0.2 amps going out. If the solar panels are charging the shunt may report 13 amps going in. As the various chargers and loads on the system turn on and off, the shunt is constantly measuring the net load in or out. The battery monitor keeps tracks of these loads and essentially counts up and down. Assume that at some point you have a fully charged battery bank and the battery monitor shows its capacity as 100%. As you start using power the battery monitor watches the shunt and tracks the power remaining in the battery. One hour after a 1 amp load is on, the battery bank is down by 1 amp/hour, one 1 Ah. My battery bank holds 400Ah, so it would now be at 399Ah left, or 99.75% of capacity. On Luckness, if the batteries were fully charged in the evening, by the morning the batteries will be down from between 30 and 45Ah, depending on how hard the fridge worked and how much I used my computer. Measuring current over time and counting down, for a battery monitor, is easy.
That same battery monitor measuring the charge going into a battery has a more difficult time. If 10 amps go into a battery for 1 hour you would think that the battery is now 10 Ah more fully charged, right? But it turns out that the charging process is not 100% efficient, there are loses along the way and not all of those 10 Ah end up being stored. Something like 90% to 95% will end up in the battery. The battery monitor needs to account for these loses. The voltage of the current going in also affects things, and over time the battery monitors idea of the state of charge can stray further from the truth. Every now and then the battery monitor needs to be told: the batteries are full - it can then reset its counters and the monitor will be more accurate for a while.
Watts = Volts * Amps.
The amps being measured at the shunt is only part of the power story, the other is voltage. Everybody with solar panels has some sort of solar controller. If you have an alternator, it has a charge controller. Wind generator? It has a charge controller. Battery charger? Another charge controller. Luckness has four independent charge controller, however the most important one is the solar controller as that account for almost all of the charging in my cruise.
A solar controller is responsible for converting the voltage of the power being generated by the solar panels into a suitable voltage for your batteries, and to ensure that the batteries do not become over charged. You can think of the solar controller and the solar charge controller as being the same thing, it probably is.
Now, lets return to the idea of surface charge and the deeper charge. My batteries are fully charged if they measure at 12.84 open circuit volts. However, you do not charge these batteries by holding a 12.84 volt charge on the battery for a long time. Analogies rarely work with electronics, but you can think of the voltage of the charge being applied to a battery as pushing the electrons into the lead plates. If the pressure isn't high enough, the charge isn't pushed into the plate, its left on the surface. In order to spread the charge through the whole battery, you need to raise the voltage. Again, going back to my batteries, there are two charge voltages - one is the voltage you hold the battery at while its being charged and the other is the voltage you hold the battery at once it is fully charged. The first is called the absorption charge, also knows as the acceptance charge. The second charge, the one you apply once the battery is fully charged, is called the float charge.
Battery chargers are typically two or three stage chargers. Three stage chargers are better and more common. Its what I have on Luckness and is what I'll focus on. My solar controller has a three stage battery charger.
The three stages in the battery charge cycle are: bulk; acceptance (or absorption); and float. During the bulk charge the charge controller is basically working as hard as it can to deliver as many amps into the battery. This is the stage where the controller is not holding back at all, its delivering all it can to the battery. This is where I will see my panels delivering 21 amps on a sunny day, which is good to see. During the bulk phase of a charge cycle the battery is trying to accept all the power delivered to it and as it does so its voltage slowly rises. At the start of my charge cycle my batteries are typically somewhere around 12.6 volts, and the voltage slowly rises as the charge is delivered. At some point the voltage rises to the acceptance voltage limit - this is the first trigger for the charge controller. For my Odyssey batteries, once the voltage hits 14.5 or 14.6 volts, the charger changes mode into the second stage, the acceptance phase.
The first phase is characterized by a rising voltage, for example from 12.6 to 14.5. The second phase has a constant voltage. During this phase, the charge controller will try to hold the voltage at a constant level, 14.5 volts for Luckness. In order to do this, the charge controller will need to start throttling the power going into the battery. I typically hit this stage of charging between 10am and noon which is when the sun is starting to deliver its maximum power to my panels, but when the charge phase goes to acceptance, I will see the solar controller start to throttle the amperage going into the battery bank down from 20 amps to 10, then 8, 6, 4 and so on, delivering less and less power as the battery becomes more fully charged.
At some point in time, and this is the whole point of this posting, the charge controller will decide the batteries are full and move into the third phase, the float phase. The voltage drops, for my system, to 13.5 and is held there indefinitely - or until the sun is not able to deliver enough power to keep up with the demand at which point the batteries start depleting and we start another night/day cycle.
So, how does the charge controller decide when the battery bank is full? I have a Blue Sky charge controller with a remote panel on it, which gives me a combined solar/charge controller and battery monitor. You might think that the charge controller would simply be counting the amps going into the battery - when the charge has replaced the amps which left the previous day it would be charged? I.e. if you used 40 Ah overnight and the charge put 40 Ah back, the battery would be charged. This doesn't work.
It turns out there are two accepted methods of deciding when the batteries are fully charged. For Odyssey batteries, and its different for different manufacturers and battery designs, if you hold the acceptance voltage for 8 hours the batteries are fully charged - so if the controller can hold the voltage at 14.5 volts for 8 hours its finished for that cycle and the batteries are full. This never happens with solar controllers as there isn't enough sun in the day for this to happen.
The other method of triggering the transition between the acceptance and float charge phases is to measure the amperage going into the batteries. Remember that the voltage is constant in the second phase and for that to be true, the controller is reducing the amperage to lower and lower levels. The default setting for the Blue Sky controller is to trigger the transition to float when the charge going into the battery bank is 1.5 amps per 100 Ah of capacity. I have 400Ah of battery capacity, so the controller would go into float when the charge hit 6 amps. When the controller goes to float, it resets its shunt counter so that it measures 0Ah, when you deplete the battery from here it will show amps going out. It also switches to the lower float voltage, and in order to do so the controller will be delivering even fewer amps into the battery. If the controller was delivering 6 amps into the battery at 14.5 volts and hit the trigger to transition to the float charge, it would need to deliver many fewer than 6 amps to hold the battery at the lower voltage of 13.6 - so the controller is shedding more amps and perhaps only delivering 1 amp or less now.
I didn't understand the transition between the acceptance and float phases very clearly when I left on my first cruise in 2011. I would constantly look at my solar controller and see it doing its fast LED blink, which indicates that the controller is in float and that it thinks my batteries were full. I would recall people mentioning how solar controllers can't fully charge a battery bank and think: "well mine can, look at that blink!" During my second cruise, starting in 2012 I noticed that the voltage of the battery bank was lower than I thought it should be for a given state of charge - but since I wasn't measuring the OCV I wasn't too surprised by this. I thought my battery bank was aging and would soon need to be replaced.
As I was reading my Odyssey battery manual, between Bora Bora and Beveridge reef, I came across something talking about the charge cycle:
If the charger has a timer, then it can switch from absorption mode to float mode when the current drops to 0.001C10 amps. If the current fails to drop to 0.001C10 amps, then the timer will force the transition to a float charge after no more than 8 hours.
What's 0.001C10 amps? My eyes glazed over when I read this in 2010. However this time around, I was out in the ocean with nothing else to do.
Batteries have different capacities depending on how much of a load you are drawing from them. My batteries are rated at 104Ah each, but at a C20 charge rate, which is a load designed to fully deplete the battery in 20 hours. My batteries C10 capacity is 97Ah. So 0.001C10 is equal to 0.001 * 97 = .097amps. What the quote above is saying, is that for my battery bank, the charge controller should transition from acceptance (14.5 volts) to float (13.6 volts) when the current being delivered is 0.001 * 400 amps = 0.4 amps. Oops.
The default current trigger on the Blue Sky controller from acceptance to float is 1.5 amps / 100Ah. For my batteries it should have been 0.1 amps / 100Ah. This is why my batteries have been undercharged for so long. The solar controller was moving into float when the charge going into the batteries was 6 amps, the transition should have been when 0.4 amps are going in.
I tried finding this information for other battery types a little while ago. For Lifeline batteries, I found a mention in the manual to the trigger being 0.5% of capacity, which is 0.005C10 which for 400Ah would be 2 amps. For Trojan flooded batteries, the trigger amperage seems to be 1-3% C20. Again, for my battery capacity, 400Ah, this would be 4 to 12 amps (which appears to be a rather large range...)
So. My solar controller was transitioning into the float charge way too early. Before making the change, I would notice that the charge controller would be in the bulk phase for a while, then hit acceptance and very quickly switch into float - it would rarely be in acceptance for longer than half an hour, often much less. I read in the Odyssey manual, that for these batteries, you switch into acceptance at roughly 80% state of charge. This would be when charging a fully depleted battery. However, as I have been undercharging my batteries for so long, the battery monitor would have been resetting its current counter to 0 at lower and lower levels and I may have ended up with my batteries at 80% state of charge while the battery monitor was showing them as being fully charged.
I changed the setting in the Blue Sky controller to 0.1 amps/100Ah on July 4th (what were you doing that day?) and started watching the state of charge. The next day, the controller moved into acceptance at noon and stayed there all day. Same the next day, and the next. By July 8th, a series of bright sunny days, the charge controller finally reached the float trigger and transitioned into float at 2pm, the first time my batteries had been fully charged since leaving my slip at Shilshole.
Since then, I've noticed that the voltage of my batteries is higher in the morning and the voltage is held higher during the day. For a given state of charge, as reported by my battery monitor, the battery bank voltage seems to correspond to the OCV voltage more closely. I think my battery bank condition is fine - I am no longer planning on replacing the battery bank in New Zealand.
Returning to the folk lore I had heard about solar panels not being able to fully charge a battery bank - its not true. My battery bank is truly being fully charged now - it no longer happens every day but does normally happen every few days. If its cloudy I get close to fully charged but not into float, which is now a very high standard. If its sunny, and I start the day less than 30Ah down, I'll enter float that day. Occasionally the system will enter float when the charge is lower, but rarely if its down 40Ah or more.
This has been a good learning experience for me. Understanding your boats batteries and charging/monitoring system is important.
Here are a few more random points:
- Luckness is an energy efficient boat. All of the lights are LED, except the mast mounted steaming/deck light. My chart plotter draws 0.75 amps with its backlight on full. My AIS/anchor alarm is on constantly, 24/7, and draws around 0.25 amps
- At night, at anchor, I have a constant draw of 0.9 amps (handheld VHF 0.2, AIS/Anchor alarm 0.25, anchor light 0.3, computer plugged into 12v trickle charging, iPhone plugged in, various monitors and controllers on - fridge, solar controller). When the fridge comes on the load goes up of course
- On an offshore passage, during the day, my power draw is 1.3 amps: chart plotter 0.75, AIS 0.25, handheld VHF 0.2, timer, etc the rest. At night add 0.3 for the tri-color and subtract some for the chart plotter backlight being set to its lowest level
- On a coastal passage, my base station VHF adds 0.7 amps.
- After my first cruise, while in Seattle, I found a 12 volt charger for my MacBook online at Amazon. This has saved me a lot of power and allows me to run my computer as long as I want. If you are planning on using your inverter to charge anything routinely, try to find a 12 volt replacement for that charging system. The only thing I'm charging on my inverter now is my toothbrush, every two or three weeks
- The Frigoboat Capri 35 fridge I installed before leaving Seattle, without a keel cooler but with a digital thermostat and the adaptive controller, has been very efficient. I would be down a little over 40Ah in the morning when I was in the hotter tropical waters months ago, but I'm routinely down a little over 30 Ah now (with a lot of computer use.)
- My wind generator has been useful and I like having it, but I rarely turn it on if its sunny or will be sunny the next day. If the wind generator is making power during the day when the solar controller is charging the batteries, there can be a case of dueling charge controllers. Once the solar controller has entered acceptance it tries to hold the voltage constant by lowering the current it delivers. If the wind generator is making power in this phase it confuses the solar controller and the solar controller may throttle the panels or move into float too early. I have also seen the voltage rise above 15 volts for periods when the wind generator is making power into a system which is already fully charged.
- I only use the wind generator when I need it. I think of it as a mechanical system which has bearings that will wear out and I try to save it for when its really needed
- You need the system to enter float every now and then, as doing so resets counters in your battery monitor to the zero point. If you don't enter float the value shown by the battery monitor for how many Ah you are down will become less and less meaningful. This is partly conjecture on my part - but it was my experience while I wasn't fully charging my batteries. Every now and then the monitor needs to know that the batteries are at a full state of charge
- If you plan on using a generator to charge your batteries every now and then, do so at a time when the battery charger can deliver the maximum amount of current possible. This is the bulk phase of charging. Once the batteries have entered acceptance, the controller will throttle the current going in which is a poor use of a generator. So perhaps run the generator first thing in the morning, bring the state of charge up to a reasonable level and then let your solar take it the rest of the way. This implies that if you only use a generator, or alternator, to charge batteries that they will rarely be fully charged - unless you don't mind running your generator as it delivers less and less current for hours and hours, finishing the charge into float. Solar seems ideal for this
- I left Mexico with a working water maker and had problems with having enough power to run the boat and make water in the Marquesas. The conventional wisdom is that the South Pacific is sunny and you should make lots of power. The reality is that there are cloudy periods. Many anchorages have little wind. If I decide to replace my water maker, I'll also be buying a generator
Questions? Comments? Email me.