Tech Blog: The 48vDC Lithium Electrical Refit – Part 2: Outcomes After 1 Year

Everyday Usage

After cruising full-time for 1 year aboard Surface Interval, we can confidently say the electric refit has been a massive improvement for livability, and are very happy we went through with the considerable expense and time of this project back in Seattle. It has been essentially trouble-free the entire time, with a few minor exceptions which we’ll get into below.

Every day we can easily check the house battery state of charge through the Victron GX Touch display, or through our mobile devices and the Victron apps. The ability to see live, at any time, with ease, how much power is being consumed and generated cannot be understated when you’re living in very remote areas of the world for extended periods. Having the fully integrated system to allow this visibility into what’s going on is essential to us in our decision making for power use and generation.

Generally the entire system is very simple to use and maintains itself without much attention required. The original goal of being able to reliably power a SCUBA tank compressor off the batteries and inverter has successfully been met, and has been working flawlessly. Each morning we review how full or empty the battery bank is, how much solar we are producing for the area we are in and how much we expect to produce given the forecast, how much potential motoring we’ll be doing for our travels in the next day or two, and how much power consumption we’re anticipating. The primary deciding factors as to whether we need to run the engine in ‘generator-mode’ is based on our anticipated power consumption and the battery state of charge. Questions we ask ourselves, include: Are we staying on the boat using Starlink and laptops most of the day? Are we away from the boat most of the day exploring and therefore not using power? Is it sunny all day or cloudy? Do we need to fill multiple SCUBA tanks and/or run the water maker for several hours? So far, we have only run the engine for less than 2 hours to add another 50-55% to the batteries. This need arises typically only when we’re anchored for a week or more where we are SCUBA diving regularly (filling tanks) and using a lot of fresh water for rinsing gear and showering post-dive (running the watermaker).

Victron Remote Monitoring (VRM) Dashboard (overcast/rainy morning)
VRM Past 30days (May 8-June 7, 2025) System Overview

Performance stats of the system have ebbs and flows depending on where we are and what we’re doing. We’ve divided our usage over the past 12 months into these regional areas: Pacific Northwest, US and Mexico Pacific Coast, Pacific Ocean Crossing, and French Polynesia. To give an idea how how we use the system and how it performs depends on where we are and how we’re using the sailboat. For context, our usage and performance in the Pacific Northwest when we left in summer time had lots of sunshine, but we were often motoring in Canada due to the tight schedule to get around Vancouver Island and then down south. Down the US West Coast we had a mix of motoring and sailing, and mostly sunny days that would start late (10-11am) due to the marine layer that takes good bit of the morning to burn off. We also spent significant time in marinas in California, using shore power. Down the Mexican West Coast (Baja and Mainland) we were mostly sailing and had lots of sun, but we were shifting into winter time, so the sun angle was not improving as we moved south. We were never plugged into shore power from our time in San Diego until arriving in La Cruz, so we were dependent on solar and our alternator. Lots of SCUBA diving and watermaker usage in Baja meant heavy power consumption there. The boat never saw rain from when we arrived in San Francisco (late Aug) until part way through our Pacific Crossing in mid-April (about 8 months), meaning we had a lot of solar! The Pacific Crossing from Mexico to French Polynesia had tons of sun directly overhead, and only a few days of motoring. Since we were sailing, we did not generate power from the engine and were relying primarily on solar to recharge the battery bank. We also did not need as much power since we were not diving at all, and not showering much! We rarely had panel shading issues during the crossing. This was partly due to luck because of our route and angle of the boat, but also due to being at the equator in April with the sun angle is as good as it gets. And lastly, in French Polynesia where it’s the rainy season currently, there is a lot of cloud coverage and rain but still a fair amount of sunshine as the squalls pass through quickly. Good wind in French Polynesia means mostly sailing, but also lots of SCUBA diving and watermaker usage.

Notable usage stats with this system and the higher loads we see:

  • Our house 48 volt bank is 20 kWh, so 3 kWh of solar power = +15% charge.
  • On lighter usage days with good solar production, solar is at least able to maintain the state of battery charge, sometimes it can actually recharge more than we use if it’s a sunny day. This doesn’t happen on days when we run the watermaker or SCUBA compressor.
  • Filling a HP100 SCUBA tank from ~1000 psi to 3400 psi takes about 30 minutes. The compressor uses around 2 kWh. Filling 2 tanks = 10% battery discharge.
  • Our CruiseRO Watermaker makes 30 gallons per hour of fresh water. We have about 150 gallons of fresh water tank storage between 2 tanks. We typically will make water once 1 tank is close to being empty (every ~ 5 days) and leave the other full in case of emergency. [Side note: The average person in the US uses 82 gallons of water per day at home (see https://www.epa.gov/watersense/statistics-and-facts). At our current refill rate, our high estimate is that we are each using 7.5 gallons of water per day.] To fill the 75 gallon tank takes at least 2.5 hours, if not closer to 3 depending on output rates. The watermaker uses 1.1 kWh, so to fill one water tank uses around 3 kWh (15%) of the battery.
  • Starlink uses around 100 Wh. So if we leave it on over night (~10 hours), we’ll discharge 1 kWh (5%) of battery for no useful purpose. It adds up. So we turn off Starlink (and the whole inverter) when we’re not using it at night or when we leave the boat.
  • The fridge and freezer compressors can utilize 100 Wh or more at times, especially when the lids are open often with items moving in/out which requires them to work harder. We also noticed a big difference in this power usage in Cabo San Lucas where the ocean water (finally) warmed up. Warmer surrounding water is great for swimming, and not so great for the keel coolers that the compressors utilize to dissipate heat from the refrigerant. Compared to Canada & Seattle, we now easily see double the power usage from these compressors (1.5-2 kWh per day). With most all other loads disabled overnight, we typically see about a 4-5% (.8-1kWh) drop in battery charge overnight, which we mostly attribute to the fridge and freezer.
  • During passages when we’re using autopilot 24×7, it is a major power consumer. It will typically utilize 200-250 Wh (10-12% battery discharge over night). Our autopilot is like a third crew member though, so this is a worthwhile use of energy for us.
  • Inverter idle/low power draw of the Quattro II 5kVA is quite good for its size, but it does still make a difference to shut the inverter off at night when we don’t need it. Per the spec sheet, the Zero Load Power usage of the inverter is 18w. 10 hours (generally overnight) equates to 200 Wh or 1% of the battery. Leaving the inverter on does not draw a ton of power, but it does add up with no purposeful use of the power. Plus it’s easier to just turn the whole inverter off at night, which shuts down Starlink and any other small AC loads.
  • Over the past year, we were plugged in to shore power for 77 days in total. This breaks down to about 3 weeks in Seattle, 2 weeks in San Francisco and Monterey, 4 weeks in San Diego, and 2 weeks in La Cruz, Mexico (the other 3 weeks in La Cruz we disconnected from the metered shore power, as solar kept up with our usage).
  • Bonus: running the engine in generator-mode to recharge the batteries quickly also makes hot water for us, as the engine coolant has a loop through our hot water heater. We haven’t had to use electricity to make hot water since leaving Northern California!
VRM: Generating power through Solar and the 48V alternator, and running the watermaker

Solar Production

Solar is our primary and only power generation source outside of running the engine/alternator, and it has been performing pretty close to our expectations. Of course we wish we had more solar, as all cruisers will say. The large rigid panels from SunPoweredYachts.com that charge our 48 volt house bank directly have produced almost 1 Megawatt Hour (932 kWH) in the past 12 months from Seattle to the Tuamotus in French Polynesia! That averages out (when not plugged into shore power) to about 3.2kWH per day of solar power from the two 475 watt rated (MAX6-475W-COM) panels. This also aligns to what we typically see in a day when there’s good sun out, with occasional clouds and sub-optimal sun angles. We also get some panel shading from our radar mounted above and between the panels.

Our four 100 watt 12 volt flexible panels, also from Sun Powered Yachts, have been performing pretty well, but of course are no match for the rigid panels. We can quantify their performance in our system, but that performance would likely be different from other systems without a dual voltage design. Since there’s no real battery storage bank for the 12 volt side (the single 12 volt buffer battery is always kept charged up by our 48v-12v chargers), there’s no ability for the 12 volt solar to charge a discharged battery, and thus there’s less efficiency gained with our 12 volt panels. This was a known downside to the design. That being said, they do power 12 volt loads during solar production hours so we try to optimize this aspect by running heavier 12 volt loads when we can make more power (namely our watermaker low pressure boost pump that consumes 300+ watts). This is an area we’ve been thinking of options to possibly improve in the future.

In the past 12 months, the two 100 watt panels on top of our hard dodger have produced 67 kWh, and the two 100 watt panels on top of the bimini have produced 160kWh, for a total of 227 kWh or .62 kWh per day. The discrepancy between the two sets of panels is due to the boom that extends over the hard dodger. We often forget, or are being a little lazy, and fail to move the boom to the side when at anchor so the panels can have full sun. We’re becoming more in tune with moving the boom, as solar production in French Polynesia has been much less than in Mexico (lots of clouds and squalls, especially in the Marquesas).

Regional notes:

  • Pacific Northwest: Typically we would see about 3 kWh daily when the sun was out most of the day. Max 4 kWH on an all-sunny day with good sun angle and minimal panel shading.
  • US/Mexico West Coast: Typically wewould see 3-3.5 kWh, and sometimes closer to 4 kWh. Further South in Mexico we would see 4 kWh more regularly, sometimes up to 4.5 kWh. Baja in the winter between La Paz and Loretto was back to 3 kWh days mostly due to the lower sun angle. Mazatlan and La Cruz we were back into the 4.5 kWh days.
  • Pacific Ocean Crossing: Between latitudes of 10N and 5S, we were regularly seeing 5+ kWh days, with several hours of the panels at max rated output and slightly over (seeing 900-950 W charging for several hours!). Most days we would produce enough power to be back at 100% by 1pm, but we were very cognizant of our power usage since we were so remote. For example, the inverter and Starlink were only on for a few hours a day. The autopilot was the biggest power draw of 200-250 Wh, using about 10-12% battery capacity.
  • French Polynesia: Output has been less than we had hoped, and we are seeing 2.0-3.8 kWh days. In the Marquesas, the tall mountainous islands are often surrounded by rain clouds, and some days we only saw 1.5 kWh days or even less for the really rainy days. But when the sun does shine and the clouds move away we are seeing 3.3-3.8 kWh days. There are no marinas in the Marquesas or Tuamotus, so we’re trying to get all the solar we can get here!
VRM Past 30 days in French Polynesia. 1st half is in the Marquesas, 2nd half in the Tuamotus.

48v, Lithium, Alternator, and Arco Zeus Regulator Performance

We’re very happy with our move to 48 volt and LFP batteries, as they have been trouble-free. Our system provides 4 major improvements over 12 volt lead acid batteries: fast recharging with 5-6 kWh from the alternator; no need to keep the batteries at 100% state of charge; ability to use the full capacity of the battery without fear of damaging it; and low current and low voltage drop for our heavy power usage on the SCUBA compressor, watermaker, and kitchen appliances. The 48 volt alternator, solar, and LFP combination allowed us to remove our Nextgen 5 kW generator from the boat in La Cruz, Mexico without any hesitation.

In the past 12 months, the 48 volt system has had 1020 kWh discharged from the batteries. Note that during sunlight hours the 48 volt solar will power loads directly (additional 454 kWh), similarly for when the alternator is running (rough est. 120 kWh). In total, ~1600 kWh was consumed from the 48 volt side, averaging out to 5.5 kWh per day over the past year when not plugged in to shore power. Adding the estimated 12 volt solar production when not on shore power (180 kWh) to this total, gives us the total boat power usage when not on shore power of 1781 kWh and a total daily average of 6.2 kWh.

At this time, the only 2 loads on the 48 volt side are the AC Inverter (Quattro 5k), and the 48-12v chargers that power the 12 volt side. The 48v-12v chargers (3 Victron MPPT 100/50 solar controllers) haven’t had any issues using them in this [unofficial] capacity. Additionally they have had no issues when all 3 are loaded up when the electric windlass or winch are in use. Typically only the first charger/controller in the array is actually sending power since the normal utilization is ~100 watts. In the past 12 months, Charger 1 has sent 614 kWh to the 12 volt system, Charger 2 has sent 3 kWh, and Charger 3 has sent 6 kWh. If we add the 12 volt solar production (227 kWh) to this output (623 kWh) we can see the boat used 850 kWh for all the 12 volt loads in the past ~12 months (about 2.3 kWh per day). Subtract the charger output corrected for non-shore power time (492 kWh) from the 48 volt output (1600 kWh), and we can calculate that our average daily AC power usage is 3.9 kWh, when not connected to shore power.

  • 12v DC loads = 2.3 kWh per day
  • 120/240v AC loads = 3.9 kWh per day
  • Total electrical load per day = 6.2 kWh
  • 3.8 kWh per day of solar – 6.2 kWh daily load = -2.4kWh power deficit on average, or about a 12% drop in the battery state of charge per day. This means we can go around 6 days until we get to 25% state of charge on the battery bank, and start the engine to recharge. 2 hours of alternator power gets us back to 75-80% state of charge.

There haven’t been any issues with the alternator or the Victron BMS/LFP batteries (or any of the Victron equipment for that matter). Regarding the 48 volt alternator, the only side affect it has is the amount of heat it can generate when outputting >3.5 kW. For the first few months of cruising we had no extra blower fans installed for the alternator or engine bay, and it became quickly apparent that to generate an output >3.5 kW would require a blower for the alternator. We saw alternator temperatures exceeding 240F when generating beyond 3.5 kW. Luckily we had one blower fan installed already for the generator we never used. While in San Diego we rerouted the generator blower fan to the alternator, which now allows us to keep a continuous output of around 5-5.5 kw with alternator temperature around 200-210F. In La Cruz, we added a relay to the blower fan power circuit tied to the engine ignition, so now it automatically starts cooling whenever the ignition is on. To get any higher output (7 kW max) we’ll need to add a blower fan for the engine compartment to extract heat, as after a couple hours of running 5-5.5 kW output it will get up to 140-150F in there (the Fireboy fire suppression system will automatically discharge fire suppressant chemicals around 170-180F).

The Arco Zeus has been mostly working great, but we have a love-hate relationship with it. We do like how the company is continuing to develop the product to introduce more functionality, including the full integration with Victron through DVCC (which we converted our install over to), and we love the generator-mode option to max out the alternator field to generate the most power possible. But the software developers have introduced a couple of problems in firmware upgrades in the past 6 months that have caused some short-lived stress. There was one release that broke communication between the Victron BMS that Mark actually discovered and isolated for the developers, who then got a fix pushed out within a couple days. Luckily we could fall back to the non-Victron integrated mode of our original install to stay up and running. A few weeks ago there was a configuration change for how Victron DVCC should be configured that wasn’t communicated to users (and still isn’t officially documented). I was able to get the configuration corrected after raising the problem on the Zeus Facebook user group and other installers were able to point me to change a setting in the configuration. The current biggest flaws of the Zeus are: lack of detailed documentation; and the upgrade requirements for the Zeus that once your phone app updates, you are required to update the Zeus before you can enter its configuration mode. There should be an ability to lock the firmware versions down so configurations don’t inadvertently break when you are forced to upgrade firmware versions.

Other than those 2 instances, the Zeus has worked very well, and overall we are still very happy with it. We are constantly logging into the app to see all the live alternator stats, using the generator-mode at anchor, and tweaking the configuration through the simple app. Compared to reports of what we hear about the only other alternative the the Zeus, the Wakespeed WS500, it’s far simpler to maintain and tweak as we see fit for our use.

The only downsides that we’ve seen so far with 48 volt: lack of 48 volt native solar panel options; 48 volt nominal (60 volt) rated fuses and circuit breakers are a bit more difficult to source; and slightly more system complexity since we still require a 12 volt side (though the dual voltage setup also gives us another layer of redundancy).

Quattro II 5kVA and Autotransformer Performance

Not a whole lot to mention for this section, as it just works. We haven’t had a combination of loads yet that have tripped it up. Very recently we were running the SCUBA compressor and watermaker simultaneously without even a hiccup (3.5 kw continuous load). Heat buildup in the inverter/battery compartment is kept to a minimum with all the fans and vents installed. While in Mexico, the one time we were able to connect to shore power in La Cruz (120 volt), the shore power voltage was too high for the Isolation Transformer (that transforms 120v shore power into 240v) and Quattro to accept (270 volt vs. 265 volt max) during the night when there was less load on the whole dock from all the boats (no aircon needed at night—we’re not running aircon, but many people on the dock in La Cruz do). It turns out having shore power voltage too high for inverters to handle was a common issue for folks throughout Mexico, since Mexico’s AC voltage standard is actually 127 volts meaning voltage levels can be up to 130-135+ volts (with our Isolation Transformer doubling that voltage towards the Quattro).

One issue with the Autotransformer (AT), if you can call it that, was back in Seattle where we had a small electric space heater. The AT would make a buzzing sound while the space heater was running, unless you fiddled with the heat output knob and/or fan speed. Usually by decreasing the heat output or fan speed you could stop the AT from buzzing. Possibly some noise feedback from the heater or a funky way it was cycling power utilization was causing the AT to do this, but it hasn’t happened for any other device we have (including our induction cooktop and electric kettle).

System Management

The central brain for most of the system is the Victron Cerbo GX (with GX Touch display). We mostly have not had any problems with managing all of the Victron equipment, and the management apps (VictronConnect and VRM) are really nice. The only issue we’ve had with the Cerbo GX was the GX Touch display glitched out and wouldn’t respond to touch until Mark exercised the HDMI port (not just the USB power port). This only happened once in Mexico, and has not reoccurred.

From our system design perspective, one of the downsides we learned about the Cerbo GX was it will only work for connecting devices on the 48 volt side, even if we just want basic monitoring of other 12 volt devices. Any device that can communicate with a Cerbo via VE.Direct/etc will assume the Cerbo should have full control. This would cause the Cerbo to become super slow whenever we tried to connect our 48v-12v MPPT chargers (since they’re considered 12 volt devices). The one 12 volt device that did work ok with the Cerbo was a SmartShunt in energy monitoring mode. The the 48-12v chargers and 12 volt solar chargers can only be monitored locally through the VictronConnect app – unless we were to install a second Cerbo for the 12 volt side (which we don’t plan to do). The Epoch 12 volt buffer battery has its own app to connect to the battery over Bluetooth, which is pretty basic, but it works fine. And as mentioned already the Zeus app also works great over Bluetooth for managing the alternator.

Also to note here, we installed some Ruuvi environmental monitoring ‘tags’, as they can connect to the Cerbo through Bluetooth. The Cerbo will upload all of the Ruuvi tag stats to VRM to so we can see it’s data remotely (see the first image on this blog post), but it does not appear we can connect more than 1 Ruuvi tag to it. We installed a second Ruuvi tag in La Cruz to monitor the engine compartment temperature to have better insights as to what is happening in there when running the alternator in generator-mode for long periods. The Ruuvi tags also have their own Bluetooth app, and its a great product/app that we would recommend.

Maintenance

Maintenance of the whole system has been very minimal, especially compared to the lead acid battery days. There is no need to run the engine everyday for an hour (as needed to top up lead acid batteries), since LFP doesn’t care about being at a partial state of charge. It is recommended by Victron to top up the batteries to 100% every week, if used heavily, to keep the cells balanced; but we probably average closer to 1-3 weeks. We’ve had no issues with cells balancing in each of the batteries.

There’s the occasional Victron firmware upgrade to do on each component (except non-smart AT and Isolation Transformer), but you must dive into the VictronConnect app and click on every single device and see if it reports an update is available. It would be nice if the app would alert the user about upgrades available for each device without having to drill down into each device.

The Zeus appears to be on a 2-month cycle for firmware upgrades, and as mentioned before, you are forced into the upgrade if the mobile app updates and you need to get into configuration mode. The Epoch battery had 2 firmware upgrades that were close to each other (probably a bug in the first they had to fix), but nothing since.

The only mechanical type of maintenance was to re-torque all of the terminal connections with a torque wrench, which we did before leaving for the Pacific crossing. No issues with connections becoming loose!

Lastly, the Epoch battery is LFP and is constantly sitting near 100% state of charge, which isn’t great for LFP batteries. We have a couple times now disabled the 48-12v charger array to forcibly discharge that battery to keep its health decent. We need to do it more often. A project for the future to automate!

Post-Install Changes and To-Do List

There are a few loose ends that either we recently changed up or are researching for future improvements. Since the initial install, we have made the following changes: removed Nextgen 5kW generator (woohoo!); added alternator blower fan with ignition relay; converted the Zeus configuration to use DVCC with the Lynx BMS/Cerbo GX (so it’s getting charging data and control directly from Victron); and upgraded the Cerbo to the next major software/UI version.

Still on the to-do list (not in any particular order):

  • Finish wiring in the old Victron Battery Monitor (BMV) for the AGM starter battery that was leftover from the old install.
  • Add a second blower fan for the engine compartment to exhaust hot air out to the stern of the boat. Newer Outbounds have this from the factory.
  • Move the 48-12v charger manual disconnect switch to the System Panel. Add a manual disconnect switch for the Orion XS charger to the System Panel.
  • Rewire the original AC Voltage and Amp (now unused) meter displays on the factory breaker panel to monitor the two 120 volt AC legs (L1 and L2). Since our Quattro inverter is actually a 230 volt model, and we use the ‘dumb’ Autotransformer to make 120/240v, the Cerbo has no way to monitor those legs. This is more of a nice to have and would be fun to repurpose the old displays that we originally didn’t think were reusable, until we recently rediscovered the original manuals.
  • Add 240 volt outlets in New Zealand in the galley and garage.
  • Add a 48 volt circuit breaker panel for future expansion.

Some areas Mark is researching for future improvements:

  • What other 48 volt power generation sources can we add? 48 volt wind generator? More solar? We have a big investment in our current 48 volt alternator and brand new spare, but the next generation of alternator generators using modern tech are also very appealing (nearly 60-100% increased charging rate, and possible to use as a hybrid-assist electric motor). There are also some interesting 48 volt generator options on the horizon (prop shaft driven) that can put out massive amounts of power, see below for more info.
  • We have an unused shore power outlet on the stern of the boat (previously used for a TV antenna), that we possibly could use for shore powering an independent 48 volt battery charger that can accept a 90-264 voltage range from shore power anywhere in the world. This has an added benefit to protect the rest of the boat’s electrical equipment from any power surges or dirty power on shore. But currently the only known charger of this type (EG4 Chargeverter) isn’t designed for the marine environment. Waiting for industry developments here.
  • How to optimize our 12 volt solar panels to fully utilize their capability? More 48 volt solar panels? Add a 12v-48v boost converter for 12 volt panels?
  • How to automate the Epoch battery maintenance to keep it in better health? Along with the manual disconnect switch, possibly install a relay that is connected to a Cerbo GX GPIO port, and control it with Node-Red automation.

Some additional 48v generator/alternator technology on the horizon:

One thought on “Tech Blog: The 48vDC Lithium Electrical Refit – Part 2: Outcomes After 1 Year

  1. For electrical and boat novices, this is a hefty blog. But I really like this documentation to help others and to remind you of what you have done!!

    As you can see, I am slowly catching up on this very important reading…

Leave a Reply

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

Proudly powered by WordPress | Theme: Sprout Blog by Crimson Themes.