The Official Magazine of the Bluewater Cruising Association
SharePrint

Lithium House Battery - Impressions After One Year

Barb Peck & Bjarne Hansen

Hoku Pa'a
Niagara 35
March 25th, 2021

We’ve previously recounted our experiences using a lithium battery for our dinghy engine, and now want to share our impressions about our lithium house battery. In December 2019, we exchanged our two tired lead-acid 6 Volt golf cart batteries for a single 12 Volt lithium battery. After six months of cruising Mexican waters, followed by six months back in BC, our overall impression is positive, though with a battery there always has to be some negative too.

Golf cart batteries are popular with cruisers: chosen because they are relatively inexpensive, widely available, and intended for deep-cycle use. That last point is worth expanding upon. Lead-acid batteries contain stacks of lead plates immersed in a sulfuric acid solution. Charging and discharging transfers lead compounds between the solution and the plates. This chemical process isn’t perfect; each charge/discharge cycle leaves unconverted residue, which over time builds up and diminishes the remaining battery capacity. The deeper the discharge cycle, the less able the battery is to be restored upon charging. This relationship between amount of discharge and the number of cycles is illustrated in the following graph from U.S. Battery.

The graph indicates that XC2 model batteries should last 500 cycles at a 100% depth-of-discharge (DOD) and 1150 cycles at 50% DOD. A convenient way of comparing the usefulness of a battery is to calculate how many total amp-hours can be supplied by that battery over its whole life. Let’s say you have a battery with capacity of 135 A-Hr. Each 100% DOD cycle would get you 135 A-Hr (ignoring the decline in capacity as the battery ages). 500 cycles would get you 500 * 100% * 135 A-Hr = 67 500 A-Hr total from the battery. Now if you only discharged the battery to 50% DOD instead, you get 1150 cycles * 50% * 135 AHr = 77 625 A Hr, i.e. about 15% more cumulative amp-hours over the life of the battery. This is why it is commonly recommended to limit the discharge of a lead-acid battery to about 50% or less of its capacity; you will get more A-Hr over its life. Different manufacturers will have different numbers, but in general one expects about 1000 cycles of useful life at 50% DOD from a lead-acid battery that has been well cared for.

In contrast, lithium batteries deteriorate less with each charge/discharge, so their cycle life is much longer than lead-acid. Several variations on the lithium chemistry currently exist. Lithium-Polymer (LIPO) and Lithium Cobalt Oxide (abbreviated LiCoO2 or LCO) are widely used in cellphones and laptops, where the highest energy in the smallest package is paramount. Because this chemistry is subject to thermal runaway, it has been known (rarely) to spontaneously ignite when damaged or abused. Another chemistry, called Lithium Iron Phosphate (LiFePo4 or LFP), is not quite as energy-dense but is not subject to thermal runaway and is therefore safer. This type of battery is now available for boats in smaller and lighter packages than lead-acid, and with a longer cycle life. See this graph of Trojan’s Trillium line of Lithium batteries:

At 100% DOD, the expected life is 4400 cycles: this equates to 12 years at one full discharge and charge each day. Since we are not actually daily draining our battery fully, I expect it to last 15 or 20 years.

When new, our golf cart batteries had 215 A-Hr capacity. After 5 years of use, I’d estimate their capacity had diminished to about 90 A-Hr. In trying to keep to a 50% DOD, we were left with only about 45 A-Hr of usable capacity. Our replacement lithium house battery has a capacity of 92 A-Hr and we are able to use all of that before charging. So, with this single house battery, we now have more usable capacity than from the two lead-acid batteries.

Of course, one has to pay for this extra performance. Our new battery cost $1300, compared to about $400 for two golf cart batteries (attracting extra scrutiny from the ship’s purser). The following graph compares lead-acid and lithium batteries on the three characteristics discussed so far: cost, lifetime energy, and life cycles.

Battery Management System

Lithium batteries need strict monitoring of temperature and charging/discharging rates to prevent damage. You may be familiar with the danger of shorting the terminals of a lead-acid battery: so much current can flow that it may melt or weld the terminals and even cause a fire. Well, the lower internal resistance of a lithium battery means that it can supply even higher surge currents than a lead-acid. This is one reason why lithium battery packages incorporate internal electronics – a Battery Management System (BMS) – to ensure that dangerous conditions don’t occur. It is possible to purchase bare lithium cells and assemble your own battery, but it is much easier and safer to purchase a battery that already incorporates a BMS.

In addition to preventing excessive currents, the BMS also balances individual cell voltages and monitors battery temperature. Because the BMS is such an important component, lithium batteries should be sourced from a reputable supplier who has invested effort in selecting the right lithium cells and developing a proper BMS. Some North American manufacturers include: Lithium Battery Power, Battle-Born, and Trojan. Our lithium house battery is designed and made by Trojan, the same company that has supplied reliable lead-acid batteries for decades.

A BMS can be designed such that the lithium battery is generally compatible with an existing 12V electrical system. You’ll see batteries advertised as “Drop-in Replacements”, or “12V Compatible.” As usual, the marketing department is guilty of some exaggeration. You will definitely notice some differences when switching to lithium, and it’s likely you will have to make changes to your electrical system other than just dropping in your new battery.

Electrical System Changes

One change relates to the charging and resting voltages of lithium batteries. Whereas a lead-acid battery might swing from 12.2 V (significantly discharged), to 12.7 V (charged), to 14.4 V (charging), a lithium battery exhibits lower variation over its operating cycle. Our battery, for example, reads 12.9 V when at 25% charge, 13.3 V when 100% charged, and 13.7 V when charging at 10 A. The following graph shows the voltage of our house battery recorded over a one-month period.

The benefit of the smaller voltage variations is that some of your electronics (e.g. VHF radio) will be happier. The downside is that any battery monitor or automatic switch, if it relies on battery voltage to determine the battery state, will be incorrect. In our case, we have a Blue Sea Battery Combiner/Isolator that automatically connects the house battery to the engine starting battery when the voltage indicates that the batteries are charging. This threshold is set to 13.2 V, which is too low: the starting and house batteries were connected to each other even when the lithium battery was not being charged, so we needed to change our battery combiner.

The charging voltage of our alternator (like many alternators) is set to about 14.3 V. This is fine for our lithium battery, as the internal BMS can handle up to 14.8 V and it regulates the charging current to prevent overcharging. This may not be true of all lithium batteries – check the documentation and ensure the maximum charging voltage is not exceeded by your alternator, solar controllers, and other charging devices.

The lower internal resistance of lithium batteries allows them to charge at higher currents than lead-acid, as well as more efficiently. We notice that our alternator runs hotter than before, due to the about 30% higher charging current. So, if you switch to lithium batteries, ensure that your alternator has plenty of cooling airflow, that the belt(s) are in good condition, and that the electrical connections are clean and tight. Provided your alternator can deliver the current, you’ll likely notice a decrease in charging time. If your present alternator is maxed-out when charging, you may benefit by upgrading to a higher-current system.

One additional note related to alternator charging: once the BMS decides that your lithium battery is full, it may quickly stop any additional current from flowing into the battery. Our Trojan battery behaves that way; we have seen the charging current drop from 10 A to zero instantly. Most stock alternators do not appreciate sudden drops in the load – it causes a voltage spike that can destroy the alternator’s regulator (this is why manual battery switches may show the warning “Do Not Switch OFF While Engine Running“). Advanced external charge regulators should be able to handle the sudden drop in current – but check their documentation. If not, then a workaround is to ensure that you always have a lead-acid battery (e.g. engine start battery) connected to the alternator. Then there will always be a load drawing a few amps at minimum. One can also attach a surge suppressor to the alternator (as we have) to absorb the voltage spike caused by sudden load changes.

Other Battery Options

There are several options other than lithium for replacing a standard lead-acid battery. While we don’t have direct experience with the following, here are two that are popular, along with points of comparison based on the manufacturers’ literature.

Firefly (Carbon Foam Lead-acid) Battery

This type of battery can withstand greater depth-of-discharge (DOD) than regular lead-acid, with a published life of 1150 cycles at 80% DOD. It also supports higher charge rates, similar to lithium at temperatures > 20°C, which means shorter engine run times. At low temperatures, they can charge faster than lithium. For example, Trojan’s Trillium battery can only charge at 15 A or less at temperatures between 0°C and 5°C (between 5°C and 23°C the rate is 56 A). However, most cruising boats would rarely have their batteries at temperatures less than 5°C.

Carbon foam lead-acid batteries have a similar charge profile to other lead-acids, except long-term floating is not recommended unless the float voltage is limited to 13.5 V. This may necessitate adjustment on some charge controllers. A Firefly battery rated at 116 A-Hr 12 V costs about $766, or about 60% of the cost of a Trojan Trillium for about 1/4 of the cycle life.

Jeff Cote at Pacific Yacht Systems (Vancouver) has informative videos about Firefly batteries (and other battery topics) on his website.

Silicon Dioxide (SiO2 or Lead Crystal) Battery

SiO2 batteries have great low-temperature performance – retaining 60% of their capacity at -30°C. However, as noted earlier, this may not be relevant for most cruising boats. The electrolyte in SiO2 batteries is non-corrosive and not liquid, so is safer than lead-acid. They have a lifetime of about 2800 cycles at 50% DOD – a bit more than twice that of lead-acid – and while they are more expensive than regular lead-acid, the manufacturer states the per-cycle cost is lower. The charge profile is similar to lead-acids except they have a maximum charging rate between lead-acid and that of Firefly and lithium batteries; engine run times should decrease.

Conclusions

Are we happy that we swapped out our 120 lbs of lead-acid batteries for 27 lbs of lithium? Yes indeed! Based on one year of using our house battery, and 2-1/2 years of our lithium dinghy battery, we are impressed with their efficiency (i.e. faster charging), lack of needing maintenance (no topping-up of water, as with flooded lead-acids), and steady output voltage (consistent performance from voltage-sensitive devices like the dinghy motor). We have had to make some adjustments but so far this has been a good change.

Comments


  1. Glen Wilson says:

    Thanks for the report Bjarne. It was well written, clear and concise. You covered the topic very well and I found the information very helpful.

  2. Daragh Nagle says:

    Excellent explanation of the issues and options for high performance batteries. Thanks!

  3. Al Kitchen says:

    Very well presented Barb & Bjarne. Valuable information for cruising community in an elegant format.

  4. Per Thyrrestrup says:

    I refer to your article: Build your own inexpensive VHF-AIS antenna
    Dear Barb and Bjarne,
    Thank you for your interesting article.
    I will make an antenna for my AIS Transponder meaning that I will be transmitting also. Did you use RG58 for the antenna corresponding to your measurements? Since I will use it also for transmitting the impedance of the antenna must 50 ohm. Have you measured the input impedance of your antenna? (It is difficult to read the figures on your measurement.)
    On commercial VHF antenna I have seen that they have a coil which together with the whip is forming kind of and impedance matching( to 50 ohm it must be), Is such impedance matching not necessary for your antenna?

  5. Bjarne Hansen says:

    Hi Per,
    Glad that you enjoyed our articles!

    With respect to the VHF-AIS Antenna, I’m happy to address your questions. Yes, RG-58 (or any other 50-ohm coax, like CA-195R) will be fine for making your antenna. I’ve made two antennas, one from RG-58 and one from CA-195R, and they both perform similarly. CA-195R is a less lossy coax than RG-58, but that only becomes noticeable if you use it for a long run back to the radio. When making your antenna, if you make it with a “tail” (the section before the “T” junction, to which you connect the mast feedline coax) of only a few feet, you won’t notice any difference.

    Sorry that the published graph is a bit hard to read: the actual impedance of my antenna measured as 44.6 ohms at 162 MHz, which as indicated in the comparison table, corresponds to an SWR of 1.12:1 and a power loss of 0.3%. That’s quite good, and is better than the commercial antenna I tested, which had an SWR of 1:1.64 at the same frequency. Either antenna is fine for both transmitting and receiving.

    The commercial antennas with the coil at the base that you describe, are of the type called “end-fed” – that is, the RF energy is fed into the antenna at the end of the radiating element. And yes, that coil is there to make the impedance be 50 ohms from the perspective of the radio. The antenna design I have described in my article is a “center-fed dipole” – that is there are two radiating elements and the RF energy is fed into the antenna at the center (the “T” junction). Its impedance is close enough to 50 ohms at resonance that it doesn’t need a loading coil.

    I hope that clarifies things for you! Feel free to ask if anything else comes up, and I’d love to hear how your antenna turns out when you build it.

  6. Per Thyrrestrup says:

    Hi Barb and Bjarne,
    Thank you very much for your detailed answer.
    I will make an experiment in which I use your dipole construction, but I will try to build it into a plastic tube. This means that the end tail coax cable will be placed parallel to the “shielding” wire (red wire on your picture). The end tail coax will be connected to a PL239 glued into the end of the tube. But I must be carefull since I do not have a network analyzer for verification.

  7. Bjarne Hansen says:

    Hi Per,
    I suggest avoiding running the coax tail close to either of the legs of the dipole, as it will affect the antenna’s radiation pattern (i.e. the signal strength will likely be significantly reduced in some sectors around the antenna). Also, the nearby presence of the coax to one leg will cause the impedance to shift away from ~50 ohms. This could be compensated for by a loading coil, but that would complicate the design.

    Ideally you want the tail to exit the antenna at approximately 90 degrees. The closer it gets to parallel to either of the legs, the less optimum the performance. Once the tail (or any other metal, like the mast) is about 1/2 the antenna length (40 cm in this case) or farther away, then it will have a diminishing effect and can run in any desired direction.

    The good thing about making one’s own antenna is that it’s relatively cheap and easy to experiment with different arrangements. By all means try assembling it and see how arranging the tail in different orientations affects the performance. A network analyser does certainly make testing easier, but by observing how well reception works one can get a good idea of relative performance. You can also enlist the help of a friend by having them listen to your test transmissions (but use low power on an appropriate channel. Definitely not channel 16!)

  8. Per Thyrrestrup says:

    Dear Barb and Bjarne,
    Thank again for your answer.
    I fully understand your reservations – especially the 16 watt one 🙂 !
    I’m not ready yet to make the experiments. I will revert with results by then.

    Best regards
    Per

Leave A Comment

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