The ultimate LiFePO4 battery? Epoch’s 460ah powerhouse
My wife Laura reads most of my entries. She often tells me that I write about lithium batteries too much. Well, sorry but what follows will be another article about another lithium battery. If it helps, this is one hell of a battery and I think everyone considering new deep-cycle batteries for their boat should know about it and consider it. That’s doubly true in light of the significant step forward in battery safety I found when I opened the lid of my evaluation unit.
For the last several weeks I’ve been testing Epoch’s powerhouse 12-volt, 460-amp-hour, 8D sized, IP67, heated, Bluetooth, and Victron communicating battery. That’s a lot of features lined up in the name of the battery, but each of them is worth highlighting. This battery checks the box for nearly every feature I could think of wanting and adds several features I wasn’t even aware I wanted.
Each 12-volt, 460-amp-hour battery costs $1,999. The battery ships with a 10-segment LED display, remote-power switch, Victron communications cable, and mounting brackets. Although my early unit didn’t include it, I’m told that Epoch will include the custom cable required for daisy-chaining batteries to Victron systems. Lastly, my early unit did include an Anderson Power Pole to 4 AWG cable but I’m told that probably won’t be the case in the future.
If, after reading this lengthy review of the battery, you decide you would like to purchase one or more, I encourage you to use this link: https://www.epochbatteries.com/?rfsn=7030270.54d2cb and the coupon code MARINEHOWTO. The link and coupon code will earn you 10 percent off your order as well as help Rodd Collins of MarineHowTo.com. Rodd suffered a stroke and has had to shut down his marine electrical business. For those who haven’t come across his site, it is a wealth of information and Rodd is a frequent source of expertise and analysis for me.
|Amp hours @ 20hr rate
|Useable Amp hours
|Useable watt hours
|$/useable watt hour
|Kilovault HLX+ 3600
|West Marine 8D AGM Dual Purpose
|Lifeline 8D deep cycle
The table above gives an overview of the cost of the Epoch 460 compared to several other options. A couple of years ago I did a writeup on the vagueries of comparing the cost of LiFePO4 batteries to other chemistries. In addition to the considerations I raised, several commenters had good points about the other factors. To keep the discussion short here, I’ll just point out I’m not comparing battery longevity (an area in which LiFePO4 has a significant advantage over any lead-based chemistry), resting voltage differences, charge acceptance rates, or charge efficiency here. The only thing I’m looking at here is a rough comparison of the cost per usable watt hour.
Anyone purchasing 460 amp-hour batteries is likely looking to construct a large house bank. I believe BMS-managed batteries are best deployed in, at least, pairs to ensure resiliency in the event of a BMS failure. So, if that practice is followed, the smallest battery bank possible would be a 960 total amp-hour bank with 768 amp-hours useable assuming an 80-percent depth of discharge. Let’s just compare that for a minute to a bank constructed of GC2 or 8D AGM batteries.
A pair of Epoch 460s would produce 8,832 useable watt-hours. To get the same useable watt hours from GC2 batteries we would need 13.7 GC2s. Let’s call that 14 batteries. If instead, we construct the bank from Lifeline 8D batteries, we would need 5.77 batteries. Again, we can round up to 6 batteries. As we compare costs we should also factor in the labor of installing batteries, the cost of battery cable, battery terminals, heat-shrink, and other miscellaneous supplies. For Epoch 460s, we would need a total of four cables, two positive and two negative. The GC2s would require 28 cables and the Lifeline 8D would require 12.
|wh / cubic inch
|wh / pound
|Kilovault HLX+ 3600
|West Marine 8D AGM Dual Purpose
|Lifeline 8D deep cycle
Another frequent means of comparing LiFePO4 batteries to other chemistries is by weight and size. The table above shows the lithium chemistries all enjoy a significant advantage in watt hours per pound. The Epoch is the leader at over 45-watt hours per pound. But even the lowest density LiFePO4 battery, the battleborn at 32 wh/lb is more than three times as dense as the densest lead-acid battery. The energy density by volume isn’t quite as stark, but the Epcoh 460 is more than twice as dense as the lead acid batteries.
There’s a fuse, in the battery!
Let’s get one exciting thing out early. These batteries contain a 500 amp fuse under the cover of the battery, just before the positive post of the battery. The fuse Epoch uses is from the EV market and carries an amp interrupt capability (AIC) of 50,000 amps. That rating means that with up to 50,000 amps flowing through the fuse, it will still successfully trip to an open and safe condition.
ABYC’s electrical safety standards require a fuse with a 20,000 amp AIC. Before disassembling this battery, I’d never seen a class T rated above 20ka AIC. Thus far the only gripes I have about the fuse are that it appears difficult to source a replacement — which I’m hoping Epoch can help with by selling replacements — and I think Epoch should label the outside of the battery to warn there’s a fuse inside. That way, a boat owner won’t find themselves stumped if they accidentally trip the fuse.
But, make no mistake this fuse is a really big deal and a really good thing. By placing a high AIC fuse inside the battery, installers are now relieved of the obligation to place a physically large class T fuse as close to the batteries as possible. The combination of fuse holder size and thick cables can make this a challenge. Knowing the fuse is inside the battery, an MRBF post-top fuse holder can now be safely used. Thus, the entire system is protected via the internal fuse, and the wire off the battery is protected by an appropriately sized MRBF on the positive terminal.
One bit of bad news is that while the fuse in the battery is great, it is a different form factor than the Class T fuses we all source from Bluesea and others. The fuse in the battery has slotted holes at a wider pitch than a typical class C. I’ve been calling the fuse in the battery a class T and did in my video above, but now I think that might not be right.
The basic performance of the 460 is pretty boring, and that’s a good thing. The battery is rated for 460 amp hours of capacity discharged over 20 hours. That means that if the battery is discharged at a fixed current (23 amps) it will deliver that over 20 hours. I’ve run at least a half-dozen such tests and it’s come out over spec each time. Voltage has been rock-solid and shows no issues whatsoever.
Perhaps the more interesting performance tests come from high discharge rates and how the battery performs. Keep in mind, this battery is rated for 300 amps of continuous discharge and 500 amps of discharge for up to 30 seconds. That’s a lot of power! At 25 C (or about 77 Fahrenheit) a Victron Multiplus II 3000 can output 2,400 watts at 93 percent efficiency. So, it shouldn’t consume more than 2,600 watts on a sustained basis. That means it will consume less than 215 amps at maximum capacity. Earlier, I mentioned I think LiFePO4 batteries should be deployed in a minimum of two batteries in parallel. Given the output ratings of these batteries, that means that even in the event of a failure of a battery, the surviving one could power all the boat’s needs.
This battery has both the highest capacity and highest discharge and charge ratings of any I’ve tested. As a result, I had to beef up my load-generating abilities. I’ve typically used the MultiPlus II 2000 on the right side of the picture above. But, the 460 just shrugged at the less than 150 or so amps of load that a single inverter could generate. So, I went digging through my shelves of old gear and pulled out the behemoth Charles / Vanner IQ2600 that I pulled off Have Another Day. The two inverters were able to generate enough load to at least make the battery notice.
There’s a lot of data on the chart above, but shows how the battery stands up to a load. In this case, that load is about 3,500 watts of heaters. That 3,500 watts equates to around 280 amps of current at 12.7 or so volts. In the chart above, voltage is the blue line and it’s holding up nicely. When the load first hit, voltage dipped to about 12.9 volts. After 50 minutes of the load, and a drop of about 70 percent SOC, voltage is still holding over 12.7. In another good sign, the battery’s reported temperature has risen from 72 to 100 degrees during this sustained load.
I took the thermal image above after discharging the battery at just over 300 amps for about an hour and 15 minutes. As you can see, the heat sink on the top of the battery got quite warm. Interestingly, you can also see a hot spot where the fuse resides on the left side of the battery. Lastly, all the connections are pretty warm, not shocking in light of how much current they’re passing. Bear in mind that I’m pulling far more current from this single battery than is likely in most scenarios. Deployed in even a 2P configuration, it’s doubtful the battery would ever see these types of load. My testing represents worst-case scenario loads and the battery has stood up to those loads admirably.
Minimum current sensitivity
Towards the end of testing, I came across a limitation of the 460 amp-hour battery’s BMS’ measurement capabilities. It boils down to this: the battery can’t measure current draws below 1.4 amps. As a result, small current consumption simply isn’t measured by the battery. The screenshot above shows the result of not measuring a small, continuous load. In this case, it’s a 1.2 to 1.3 amp load from my two test inverters idling. I left the battery that way overnight and here’s where I ended up. The SmartShunt on my test bench correctly sees the battery at about 93 percent state of charge. But, because the battery’s BMS didn’t measure the low current draw, it still thinks the battery is at 100%.
Current flow through the battery is measured using current sensing resistors (CSR). Due to the high current handling of the battery, Epoch has to use a CSR that is capable of detecting large current. Unfortunately, especially in PCB mounted CSRs, those that have high current handling capabilities also have relatively high sensitivity floors.
I think this is a real limitation and one for any user of these batteries to understand. The battery won’t register any current below 1.4 amps. That showed up in my testing with one battery. But, imagine you’ve built a house bank of four of these batteries. Now, your loads are divided across all four batteries meaning you will need 5.6 amps to hit the sense floor of each of the four batteries. I think that’s a real enough issue to demand a Coloumb counting shunt in the mix like a Victron SmartShunt or BMV.
One of the major headline features of this new battery is built-in Victron communications. The 100 amp-hour batteries I tested previously have communications ports described as CANBus, but we never saw external integration with them. In fairness, I don’t believe Epoch ever promised communications with those batteries, though I thought it would come. Unlike the 100 ah batteries, this series launched with external communications working. As soon as I plugged the RJ-45 plug into the BMS.CAN port on my test Cerbo, information began flowing.
Bluetooth communications are also present and utilize the same app earlier Epoch batteries have used. The app provides the same detailed information and quick at-a-glance core health information. As I encountered with earlier batteries, the warnings and alarms the app displays are somewhat cryptic. I am fairly certain the DOC Alarm above indicates the discharge over-current alarm is active. I also triggered a COC Alarm which, I think, indicates charge over current. Because the label for both of those faults is in yellow, they’re warnings. The last screenshot shows a red fault for, I’m pretty sure, cell under voltage protection. That is a fault and has resulted in the BMS disabling discharge.
Victron integration opens many possibilities
With battery data available and feeding to a Victron Venus OS device like the Cerbo, the boater has a wide-ranging choice of how to display the data. The screenshots above show several of those ways. The system overview and alarm screen use the HDMI output on the Cerbo and HDMI input on a Garmin 9219. With the display 9219 connected to the Cerbo, and a USB cable sending touch commands from the MFD to the Cerbo, you gain full display and control capabilities of the Cerbo. The second screenshot above shows Victron’s MFD app displayed on the 9219. The app leverages Garmin’s OneHelm app system to send data from the Cerbo via HTML5. Lastly, the final screenshot shows data displayed in Garmin’s native power view. This display is leveraging NMEA 2000 data delivered from the Cerbo to the MFD via Victron’s VE.CAN to NMEA 2000 cable.
Comparing available data
The two data boxes above compare data available via the battery’s BMS.CAN connection and via a Victron SmartShunt. The SmartShunt connects inline in the negative conductor to the battery. Both sources provide core data like voltage, current, and state of charge. But each source also provides data the other doesn’t. For example, the BMS.CAN connection from the BMS provides cell minimum and maximum voltages as well as battery temperature without an additional temperature sensor. Currently, the SmartShunt is providing runtime estimates and consumed amp hours the CAN connection isn’t. The lack of runtime data is under investigation by Epoch because they expect the battery should be providing it.
I greatly appreciate the Battery History information VRM and a SmartShunt or BMV provide. Via VRM, that information isn’t available leveraging the battery’s CAN connection. The Epoch app does provide a cycle count, but the detailed information on charged and discharged energy, time since last full charge, maximum and minimum voltages, and more are unique to Battery History. In light of the low cost of a SmartShunt, when I design systems I’m likely to still include one.
The construction of the Epoch 460 amp-hour battery is very similar to the 100 amp-hour battery. That includes a high-quality, water-resistant battery case. In the case of the 460, the top is secured by 14 hex head screws spaced around the top.
With the top screws removed, the top itself flips out of the way to reveal the BMS, bus bars, and aforementioned fuse. I’ve torn down a range of LiFePO4 batteries and Epoch’s batteries compare very favorably. The layout is clear and simple, the component quality is very high, and heck, there’s even a 500 amp class T fuse thrown in for good measure.
Every screw has a witness mark on it in permanent marker. These marks provide easy visual confirmation that the screws haven’t moved and are as they were at assembly. The marks on the screws securing the bus bar to the cell pack are fairly clear, but a little subtler are the marks top left above on the screws holding the top deck to the lower assembly.
With the bus bars removed and the screws securing the upper half to the lower half, I removed the top altogether. That provides access to the cell pack which we will cover in a minute. But first, take a look at the block of thermal paste covering the FETs in the center of the image above. That paste provides a thermal pathway from the BMS’ FETs to the aluminum heatsink built into the cover. As we saw in the thermal image earlier, under heavy load, the heat sink does an efficient job removing heat from the FETs.
With the top half out of the way, we gain visibility to the cell pack below. One of the few complaints I had with the 100ah battery was the amount of wasted space in the case. The cells barely filled half the volume and left tons of empty space. That made the batteries pretty big compared to many other 100 amp hour batteries out there. There’s no such problem with the 460!
Though they’re hard to spot, there are four long bolts through the stiffeners holding the cell pack to the case. I assumed I would remove those bolts and pop the cell pack out of the case. You know what happens when you assume, right? I soon learned that in addition to the four bolts, an extremely effective adhesive holds the cells in place.
There’s nothing like seeing your shop in a video and realizing what a mess it is. I’m in the process of cleaning things up in here, but that’s another story. With the cells finally out of the case I was able to confirm the 460 seems to have the same high build quality. The aluminum stiffeners at the ends of the cells paired with tough metal strapping provide cell compression.
I’m not sure what the glue is that’s used to secure the cells to the case, but it holds extremely well. Especially considering the relatively small amount in the bottom of the case. Also, in addition to the glue, four long bolts secure through the stiffening plates at the end of the cells into the brass threaded inserts visible in the bottom of the case.
The cell QR codes are covered up by foam, but once I cut the foam away I found unobscured codes. The unobscured codes in the proper placement suggest these aren’t used or B-grade cells.
Up next I tore into the BMS to get a better look at the overall design and FET placement. There are a total of 56 FETs, 28 on the front and 28 on the rear. I believe half the FETs handle discharge loads and the other half handle charging duties. The board’s design appears robust and the components appear high quality. It’s also nice to see, like the 100 ah battery, the care taken with things like securing cables, insulation and jackets, labeling, and gluing all the connectors in place.
I’m not sure my traditional BMS comparison photo tells the full story. The current handling and durability of the FETs on this BMS appear to dwarf any others I’ve seen before.
It appears from the spec-sheet that each FET carries a max rating of 320 amps and 100 volts. But, I don’t think there’s any way the traces on the board or the FET’s ability to dissipate heat could keep up with that amount of energy. It appears each FET can dissipate the heat from a maximum of 313 watts. I’ll be the first to admit I’m over my skis trying to interpret this data. So, if any readers have a better understanding, please chime in!
The 460 amp hour battery preserves many of the same features I appreciated in the 100 ah battery. Chief among those features are the temperature sensors on the positive and negative terminals of the battery. Prior to the 100 ah model, I’d never seen temperature sensors used like this. But once I did it seemed like such an obvious feature with so many benefits. I’ve seen the aftermath of a poor connection at a battery’s terminal. When you’re dealing with batteries that have the potential to deliver the current these do, the damage can be significant.
The app displays the temperature of both terminals, the temperature of the MOSFETs, and an ambient temp. I think that ambient temp is probably coming from the temperature sensor on top of the cells, but I don’t know that with certainty.
Documentation and ABYC-E13 compliance
ABYC’s E-13, the recently published safety standard for lithium batteries, spells out several requirements for compliance. In my entry on E-13, I noted that much of the standard relies on the battery manufacturer’s recommendations and requirements. Beyond extensive reliance on the battery’s manual, E-13 requires listing or approval from one of several agencies including UL, a BMS in place, and adequate water protection. There are several installation requirements and quite a few recommendations made in the notes.
So, for a battery to be able to be installed in an E-13 compliant manner, it must have a BMS, it or its cells must be approved, it must be protected from water, and the installation must follow the recommendations and requirements from the battery’s manual. The 460 meets the first three of these requirements but the current documentation doesn’t provide enough information to clear the last requirement. Several sections of E-13 reference information in the manual. In Epoch’s case, that information isn’t in the current manual. Epoch tells me they are near publishing updated literature that will provide all the required information.
As I said at the beginning of this review, the Epoch 460 is one hell of a battery. It isn’t perfect and I’ve highlighted several of those imperfections. Most notably, I think the current sensing floor and data availability issues take some of the shine off the Victron communications. But, I still believe this battery provides pretty phenomenal value, safety, and performance. I would strongly recommend that if you’re going to install these batteries you add a Coloumb counting battery monitor.
I’m about to revamp the house bank on my RV and plan to leverage four 460/12s for that project. Currently, to my eye, these are the most compelling batteries available in the market. I’m excited to put a set to work and be able to report on their real-world performance.