Sudden SoC drop to 0% when below 20%

Is this a calibration issue? I normally don’t use the battery enough to get below 20% SoC.

I use Time Control UPS mode. It charges from 00:00 to 06:00, and Battery Level Preset is 15% - 90%

Yesterday when the battery got near my low preset of 15%, it suddenly dropped to 0% SoC and the AC output was disabled. The battery started charging from the grid, and only when it got to 1% SoC I was able to turn on AC power again.

Expected: when the battery SoC reaches 15%, it changes to grid power passthrough when in Discharge Time.

Actual: battery suddenly dropped to 0% SoC and AC power was shut off.
The unit started emergency battery charging until 3% (during Discharge time)

Notes: at 0% and charging, the battery voltage was at 47.7

Firmware versions:
IoT: 9014.06
ARM: 4004.09
DSP: 4003.11
BMS: 1108.05

If it’s been several months since you’ve done a charge to 100%, then drain to 0%, then charge back to 100% I’d definitely recommend that. It’s pretty common for the SOC indicator to get out of whack after awhile - and it’s especially true when the unit is not ever reaching the true “full” and “empty” states. So yes, my guess is it’s a calibration issue.

1 Like

Thanks, I’ll run a full cycle this weekend.

A bit concerning that it also shut off AC power - I think a firmware update could cover this case: when connected to grid and SoC is 0%, enable AC bypass as first priority.

Though as described above it may have been an unexpected emergency state. OR it could be legacy behavior since during development there was no AC inverter bypass

agree! Would very much like to see some new features/improvements to the firmware! (I have an AC300 - but most of the features would carry over!) Definitely feel like it’s the area in need of most improvement in these products!

Yep. I really like the products, just need some refinement once their insane manufacturing cycle calms down a bit.

Hi @ed209

Sorry for the inconvenience.
How long have you used your unit? Is the first time the problem has happened? You can try to calibrated it by full charge and full discharge. If it is still abnormal, please keep in touch with us.

I’ve only had it in operation about a month and have not yet run a 100% - 0% - 100% cycle.

Will update results within a week.

Hello, I did testing over the weekend and found an issue that should be corrected.

The full charge-discharge correctly calibrated the unit :white_check_mark:

When the battery SoC is 0%, grid AC or DC power does not work even when connected to the grid :x:

After fully charging the EP500, I set Time-Control UPS Battery Level Preset to 0%-100%, and changed the time to Discharge. Then I ran AC power off the battery, until 0% SoC. The battery percentage was correct (no sudden drop). The unit was plugged into the grid the entire time.

After reaching 0% SoC, the EP500 will switch over automatically to grid passthrough.

After reaching 0% SoC, the EP500 shut off DC and AC power, including active loads. The battery began charging until 3%. The DC and AC power was disabled until 1%.

The Issue:
If functioning as a UPS, the battery must always provide power if grid power is available.

Part 1: if the battery falls to 0% while in UPS mode (either from configuration or the battery is out of calibration), and the unit is connected to the grid:

  1. If AC Inverter is active, switch to AC grid power passthrough
  2. If DC power is active, maintain DC power
  3. (current behavior) Charge battery from grid back up to acceptable level (right now 3%). Note this also assumes that 0% SoC still has a decent buffer for operation of the system. This might need adjustment to ensure #1 and #2.

Thanks for your time, I hope this helps.

Hello, I’m still noticing some sudden drops in measured SoC when it gets low. I did a full calibration about a month ago.

In my case: I ran fairly high loads (~1500 watts) and the SoC drained from 80% to 30%. At that point I turned off most of the load, leaving about 250 watts running.

Then in less than an hour I heard the relay switch from battery to grid. SoC had dropped to my configured low of 10%.

I expected the 250 watt load (in an hour) to drop SoC about 6-8%, instead it went down 20%

I want to say without the high load, it has been calculating correctly at lower SoC. But even at 1500 watts it’s… 0.35C, not really that much. (1500/4800 * efficiency loss)

My unit did cut over to grid correctly at my configured 10% SoC, so everything is still fine. Just making and reporting my observations, thanks.


Hello, there’s definitely an issue with the SoC calculation… last night, on battery, the AC cut out when it was at around 30% SoC, with a load under 300 watts.

I understand some drift that requires recalibration, however this is really far off considering the recent full recalibration.

I must reiterate: as a UPS function, it should NEVER cut off AC power while in use. If a low SoC is detected which requires charging back to 3%, it should immediately prioritize AC passthrough, and as secondary function restore charge of the battery if within total power limits.

I also did see a recent email regarding a new AC charging cable, is this related?

Please let me know if I can provide data for troubleshooting, thanks.
In my setup, I charge only from AC, and use the Time Control UPS mode.

@ed209, have you received any reply yet? This is a great writeup. This actually is a critical issue. I have been using my EP500 as supplement supply to my EV Charger. It works fine in the Customized UPS mode that consist of Standard UPS Offline, along with the SOC parameters and Charging Time parameters together. It works great while it applies the AC Grid power passthrough during non-peak hours. AC Grid power cuts off at the start of Peak Hours at 16:00. However, a big issue is when the battery discharged to 0%. The EP500 automatic shuts off. Just as you said, it will not resume until the battery recharged above 2-3%. I afraid this consistent low power auto shut off will eventually damage the battery.
As I have originally posted to question about the function of parameter SOC Low. If this works as it describes in the User Manual. The discharge should cut off at the SOC Low Preset setting. I would add to your suggestion should integrate working together with the SOC Low Preset setting. For example; if I have the SOC Low set at 35%, the EP500 should stop the battery discharge when it hits 35% capacity and switch over to the Grid AC passthrough. So, I know it will have 35% reserve capacity for emergency grid power outage use.

@randychan38 there isn’t much danger of battery damage, I believe. There’s some padding built into the battery so accessible power is 4800Wh, I’ve read somewhere.

The power shuts off to protect itself. When working, the first battery bank to fall below a certain low voltage (independent of user configured option) will cut off power. From there healthy batteries should be able to sit for a long time without issue.

The behavior that is an issue for me is sudden loss of AC power even when connected to the grid. I’m ok with the batteries protecting themselves during unexpected imbalance.

I’m not ok with a UPS function that is easily interrupted during normal operation.

@ed209. That are exactly my issue. Although I understand (was told) by Bluetti Admin on this board that there are 10% buffer to protect the battery from going total empty. Anyway, I also understand from Bluetti Team that the design of the SOC Low setting is a setting of a percentage at users’ choice to reserve the battery for emergency power outage. However, I do not see that since the battery continues to power the loads until it reaches 0% and auto shut off. In this case, what if, a power outage occurs during the time the EP500 were shut off. We would then have no power until enough solar energy or grid power back on to charge the battery to at least 2%.
An additional fact from Bluetti Team states an associated design of the EP500 never stop powering loads unless the AC power is off, AC Grid in the option was switched off. Also the battery was too low. If none of these conditions were false, I agree that when the battery capacity reaches the SOC Low setting (let’s say 35%), the EP500 should switch the powering source from battery to AC Grid provided AC Grid power present.

I’ve now experienced this at over 50% SoC. I woke up this morning to see all of my equipment off. Last night around 10pm it was over 50%. From 10pm to 12 am, total draw is less than 100 watts. It charges from 12 am to 6 am, and this morning SoC was around 68%, matching 6 hours of charging.

I have my charging limit set for a while now to 20-95%, so it should be within voltage variation range for balancing.

@BLUETTI can you give me some insight in the the issue? I would prefer not to start an RMA.

This happened to my first unit. I know you may think that getting replacement will be a pain but it actually isn’t that bad. I went through it. In the long run it will save you a lot of frustration to request a new unit.

1 Like

@Felton except for lugging it back up and down the stairs :frowning:

I’m trying an experiment now by setting max charge to 100%. At the moment it’s charging at ~575 watts, but it’s been showing 99% charge for the past 80 minutes. (edit: finished at 100% after ~85 minutes) I’m conjecturing that it is filling in the missing capacity right now, and recalibrating.

Could be that balancing or calibration only is occuring at 99%+?

What was the issue on your first unit, @Felton ?

The first three following pargraphs describe my initial problem and resolution of the firt unit. The subsequent paragraphs explain what i use the unit for(skip if not interested)…

When I first received the unit I set it up in the bedroom and used it with a roll around a/c unit for a few days. I vaguely remember that on one occasion I woke up to find that the ep500 had shut off. It took me a while to get it to reboot. At the time I chalked that up to my lack of knowledge operating the unit. I didn’t have any issues immediately after rebooting it so I went to the next test.

I hooked it up to a newer model fridge using an extension cord. I woke up one morning to find the fridge with no power. I was able to get the 500 rebooted and bring the fridge to temp. Eventually, over time, the 500’s lcd display went out and that was when I sought help from Bluetti

I eventually contacted the manufacturer (in China, I believe) and they were very helpful. Through emails, they walked me through various steps to check the cables hooked up to the motherboard. It was deemed necessary to replace the unit. The replacement process was relatively painless, living in a single story house.

My current unit is working fine. I live in the Los Angeles area and I charge it through three 380 watt solar panels that I installed on the roof with no a/c input from the grid. This morning I was pulling in around 930 watts from the sun. The 500 is hooked up to the house through the use of my gas generator transfer switch. It powers my 2 fridges, alarm security system, wifi and LED TV, and the office and bedroom. We run the portable electric heater in the bedroom for an hour as we go to sleep for the night. On the average, the unit will drain down to about 30 to 35 percent before the sun is strong enough to charge it back up.

I mentioned in a previous post that the replacement process isn’t bad. I didn’t have to lug it up and down a flight of stairs but I did have to transport it home in my Subaru wagon and get it in the house by myself, not to mention packing and shipping the defective unit back. Now that it is working I would say that I am happy with the product and it is performing as expected. It was worth getting the replacement ordered nd delivered.

Good luck with yours. If you have questions that I may be able to answer, reach out to me.



Update on this: further testing confirms that calibration only seems to occur at 99% to 100%. This isn’t ideal, but in the last 3 weeks been able to avoid any sudden power loss.

I also think I may be in the minority in that I only use it for Time of Use with the grid, with no supplemental solar power to constantly top it off (to my knowledge solar charging max SoC cannot be configured).

The “workaround” by setting max grid charge to 100% is acceptable – because 100% appears to top out around 55.3V, which at 16S is 3.45V per cell which is very conservative.

Assuming no other issues, I’m okay with it. It’s too bad the charge levels can be so inaccurate, and that it can suddenly drop unexpectedly and cut off output power.

I hope that this can be addressed in future firmware updates. For now, critical devices should definitely be routed through a real UPS.

Firmware versions on my unit:
IoT: 9014.10
ARM: 4004.09
DSP: 4003.11
BMS: 1108.05

1 Like

Hi @ed209,

Sorry about this issue.
If the charge levels still drop suddenly, pls contact our customers service at

Update on this, I’ve been charging to 100% (device reading) once a week, and also increasing my daily max SoC to 95-99% so the cells are are in the range where voltage differential can be detected. (due to flat voltage curve)

As a result I’ve not experienced any unexpected shutoffs in the last two months.

The takeaway is: I still hope a future firmware update will address the issue of power loss as a “UPS” if the cell voltages get too low. AND I’m hoping that by charging to a higher SoC, there will be some balancing going on that doesn’t require charging to displayed 100%.

Second takeaway: I should have purchased the pro version, such that the minimum PV voltage is lower.