Hey All, In Jan this year (2026) I started planning a new solar install with my electrician. He was open to me supplying many of the bigger components and in particular I wanted to find batteries that fit my needs. I looked at a number of options: Deye, Dyness, Esener, Must, SunSynk but the bang for buck on the HinaEss PowerGem Max really had what I was looking for, and I could pick them up for a really good price (R29k) with 10 year warranty / 8000 cycles at 90% DOD – 0.5c.

I hope this helps someone else out there who is looking at these batteries - as there was not a lot of info online about this brand. This is my subjective opinion on these batteries based on my experience since buying them in early March 2026 - you shouldn't use this as a proxy for what your experience will be like.

My setup:

• Deye 12K Inverter

• 2x HinaEss PowerGem Max 16kWh Batteries

• 12x Jinko 580Watt Bifacials (North facing array)

• 10x JA Solar 620Watt Bifacials (East & West facing arrays)

• 6000Watt Ryobi Petrol Generator

• All loads are connected to the Inverter "Load" port.

• All data from the Deye and batteries is pulled into Home Assistant over RS485 -> MQTT.

My previous setup was a Deye 8k, with 2x Hubble 5kWh batteries and 14x JA Solar 460w panels.

Aesthetics

I liked these because they are simple, white boxes, installed on brackets off the floor. The cables are covered by lids, so everything is neat and matches the Deye inverter well.

All the buttons and comms cables are also hidden by the lid, so a clean install is easy.

External Build Quality

 I was really impressed with how they were packaged and how well built (at least from the outside) they seem. They came in wooden crate/ boxes with tons of protective foam, everything nicely wrapped and easy to find.

The steel brackets look and feel sturdy and strong, the same goes for the battery chassis. These things weigh 120kg's each, which meant they were a bit of a mission for my installer to deal with. The included power cables, battery comms cables and ground cables are all pretty decent quality and length. The user manual is also very decent and the whole package feels like it’s worth the R29k – but just.

 The rawlbolts that came with the battery are complete junk though - my installer used some builders warehouse M8 rawlbolts instead which worked well.

 The battery comes with a pre-punched template for drilling holes on the wall which was really a nice touch. The battery is lifted onto the bottom bracket and then secured on the top brackets to the wall. These brackets weren't a nice experience. My installer lost quite a bit of time because you have to perfectly align the brackets on the wall with the battery which didn't happen because putting M8 bolts into an old wall meant that they drifted off centre slightly, which meant the screw holes didn't align with the bracket. Some modifications to the top brackets were required in the end.

 

Software & Wifi

 The battery comes with an app that you use to detect the battery (via Bluetooth) and then connect the battery to your wifi. Then you can use the app to monitor the battery from anywhere.

 Two things I found with this process:

• My wifi password had a special character in it, which meant the battery wouldn't connect because of this. I experienced the same thing with my previous Deye 8k inverter wifi dongle.

• I could only get one battery to connect to the app / wifi. It seems this may be the way that it works as the app detects the other batteries via the "master" battery comm cable. So all batteries show up in the app even if you only get one battery connected to the wifi.

 The battery information and data reported to the Deye seems limited. I have a Home Assistant integration pulling data from the Deye and it really just had the basics (SOC, Temp, Voltage). The app on the other hands gives you 4x temps and 16x individual cell voltages as well as cycle counts. So to fix this I did some research and created a Python app that connects to the battery over RS485 and publishes the stats via MQTT over the wifi to my Home Assistant. This has been very reliable. The app / wifi is flaky and sometimes it loses the wifi settings, so I am happy to not have to use the app for monitoring. The app is useful for firmware updates, so I keep it around for that. My battery is up to date out of the box, so I don't know what the update process is like.

You can check out my GitHub Repo for the battery monitoring / Home Assistant connector: https://github.com/drewzadev/ha-hinaess-powergem

BMS

From very basic observations, I believe the 2 batteries deliver on the advertised 32kWh. Weirdly in the HinaEss app at 100% SOC I usually see 34kWh "available". I don’t know what that means, but I think it’s related to some other observations that I have made with these batteries.

In the first week, Battery 1 would peak at 55.08V and Battery 2 would peak 53.74V at 100% SOC – that’s a difference of 1.34V. Both batteries would show a status of “Standby”, IE: they were not charging. At that time the differences between the lowest cell voltages and the highest cell voltages were also quite large. Battery 1: 245mv and Battery 2: 203mv. I raised these concerns with their support WhatsApp group which is very active and they were responsive, however their response was that this was not a concern and that the voltages would normalise over time. They also said that the big cell delta’s would only show at 100% SOC.

Now that I have been running the batteries for nearly 2 months I have noticed that they are both equal in voltage at 100% SOC (54.2V) and they match each other pretty much their entire SOC curve. I have also noticed that the cell delta’s have been slowly reducing, with Battery 1 at a max delta of 60mv and Battery 2 at 180mv at their highest during 100% SOC charge. 180mv is still high in my mind but the trend is downward and outside of 100% SOC charging the cell delta’s are around 3 – 4mv for both batteries, which seems very good.

Another interesting observation is that at 100% SOC voltages settle at 54.2V and  at 10% SOC is at 51.1V – this says to me that the battery operates in a very narrow voltage band and the BMS does not push the battery to the upper and lower limits of voltages that I have seen in other batteries (56v -  49v). Maybe this is their attempt to ensure the battery delivers the 8000 cycles that they advertise or it could just be the Deye  inverter and the BMS interacting in a weird way.

Another interesting feature of the BMS is that it will stop the Deye inverter drawing down below 10% SOC, even though all my inverter settings allow for the battery to go down to 5% SOC. The only time it will go below 10% SOC is if the grid connection is off. This does align with what they suggest in the user manual.

 

Power Delivery

My entire house runs off the Deye inverter. 2x 2kw geysers, stove, 2x ovens, pool, aircons, everything. We have put in the required breakers to ensure we don’t overload the inverter and my Home Assistant will actively shed load if we push too close to the limits. I’ve seen peaks of 11kw for 4 – 5mins without issue. I’ve also seen the batteries charge at 10kw for 10min without issue, although I try to keep the max charge rate to 8kw (~0.25c per battery).

At 10kw loads the 50mm2 cables going from the inverter to the battery busbars gets warm and I do feel like I could have spent more to get 75mm2 cables. Typical loads of 3kw – 5kw are the norm in the house and the cables are room temp at these loads and the batteries also stay pretty cool (20c – 27c depending on ambient air temps).

Usually in the evenings we cook meals on the stove and oven, often adding a microwave and kettle to the mix. Then shortly after that the one geyser will heat water for the evening for about an hour and a half. Which will then leave us with around 90% SOC for the rest of the evening. Most mornings I wake up with around 60% SOC remaining at 500watts average power use during the night.

 

Support

As mentioned support is provided directly from HinaEss via an active WhatsApp group. There are a number of HinaEss people active on the channel and they seem pretty responsive. Their go to is generally to request a battery firmware upgrade for strange problems and it seems that this generally solves the problem. Most firmware upgrades they are able to do remotely for customers, especially if they have a Lux Inverter – they seem to integrate well with the HinaEss batteries.

Overall Review

I would say I am happy with the purchase of the 2 HinaEss batteries. It still blows my mind that I can cook and heat water all from batteries and it feels like it didn’t even put a dent in the power available.

Only time will tell if they actually last the promised 8000 cycles / 10+ years. But so far so good. Please share if you have any experiences with these batteries to help others considering this brand and model.

 

Edited April 27Apr 27 by Drew_ZA
Spelling

Hi Drew

I'm in the market for higher capacity batteries, and was also eyeing these HinaESS 16kw batteries.

I couldn't quite find a local manufacturer website locally, but the international site (https://hinaess.com/contact-us) notes CNBM International South Africa (Pty) Ltd as the local support.

Unfortunately I can only see feedback from hellopeter on these guys (https://www.hellopeter.com/cnbm-international-south-africa) and doesn't look great.

Have you had any experience with them regarding support?

Hey Sam, I purchased mine from SA Solar Experts in JHB. I understand the warranty would be handled by them. I don't know how long they have been around for (I suspect since 2023 / 2024) and I have not had to deal with them in regards to a warranty claim - so I also can't say if they are good or not from that respect.

However their sales service was good and they were very responsive on WhatsApp. Language barriers were a challenge at times, but otherwise they communicated well and were prompt. The HinaEss after sales WhatsApp channel seems to be manned by HinaEss international agents, not local South Africans.

I have a powergem max battery need some to connect it to wifi 😒 just not connecting

On 2026/04/27 at 10:22 PM, Drew_ZA said:

BMS

In the first week, Battery 1 would peak at 55.08V and Battery 2 would peak 53.74V at 100% SOC – that’s a difference of 1.34V. Both batteries would show a status of “Standby”, IE: they were not charging. At that time the differences between the lowest cell voltages and the highest cell voltages were also quite large. Battery 1: 245mv and Battery 2: 203mv. I raised these concerns with their support WhatsApp group which is very active and they were responsive, however their response was that this was not a concern and that the voltages would normalise over time. They also said that the big cell delta’s would only show at 100% SOC.

Now that I have been running the batteries for nearly 2 months I have noticed that they are both equal in voltage at 100% SOC (54.2V) and they match each other pretty much their entire SOC curve. I have also noticed that the cell delta’s have been slowly reducing, with Battery 1 at a max delta of 60mv and Battery 2 at 180mv at their highest during 100% SOC charge. 180mv is still high in my mind but the trend is downward and outside of 100% SOC charging the cell delta’s are around 3 – 4mv for both batteries, which seems very good.

Another interesting observation is that at 100% SOC voltages settle at 54.2V and  at 10% SOC is at 51.1V – this says to me that the battery operates in a very narrow voltage band and the BMS does not push the battery to the upper and lower limits of voltages that I have seen in other batteries (56v -  49v). Maybe this is their attempt to ensure the battery delivers the 8000 cycles that they advertise or it could just be the Deye  inverter and the BMS interacting in a weird way.

I would recommend that you spend few hours/days in balancing mode to remove the imbalance. To achieve that yourself you can skip battery balancing and swap Deye into V-control. If you can have a look into BMS and you should be able see if/when cells are balancing. Set minimum voltage which keeps balancing. After the process return to normal CAN control

54.2V is perhaps stationary reading. Per manual these units should achieve 57.2V and then relax. If they do not, they may have cell overvoltage protection triggering stop of charge.

What imbalance means is that out of 16 cells, one cell(s) will have high voltage reading, while other(s) has low. 180mV would mean that one cell is at 3600mV (overvoltage alert zone, 100% charge) while other is at 3420mV (~98% charge). In long time this causes the high voltage cells to degrade faster than the low voltage cells limiting the useful capacity of the untis.