What's a "happy" temperature range for these batteries? I know they don't like it when they're too cold and I had the BMS shut down on me a couple of times in the winter. Since then I built an insulated box around the battery and BMS stack and now the temperature varies between about 15C late at night, and 25-28C peaks if charging overnight for several hours (based on temperature reported via HA). However such peaks are short lived and I'd say the battery spends most of its time between about 17 and 22C.
The box was made to be easily dismantled for the summer months so that's not a problem, I'm just wondering how warm the batteries should be allowed to get (both peak and sustained) for optimum performance, but also longevity.
System is x6 Ver 1/D batteries and an H1-3.7 inverter, installed last November so still getting to grips with how best to manage it!
Ideal temperature for HV2600 battery bank?
Anyone come across any docs for this? There must surely be a "recommended" temperature range for these things to run at?
-
Dave Foster
- Posts: 1305
- Joined: Thu Oct 13, 2022 7:21 pm
Sorry missed this one earlier, FoxESS themselves don’t publish much other than the absolute limits that they work to, and just going off the anecdotal evidence over the last winter the HV batteries seem to be the most tolerant in the Fox range.
Like you i’ve built an insulated enclosure (with a 40w heater controlled by my home assistant), and spent quite a bit of time researching LFP cell technology and reading the various studies that have been published on capacity over time and over a temperature range.
This chart comes from a research paper on LifePO4 batteries and shows how a single LFP cell ‘s capacity responds to temperature
Those are extreme ranges but FoxESS have now confirmed the following - that 100% capacity and cycle life are rated at 25C (ambient), 90%DoD and 0.5C charge and the battery capacity falls because of charge derating between 15C and 0C - at 0C they will not allow a charge, but they will allow discharge down to -10C.
That effectively means if you want to maintain a reasonable autonomy that 15C is the minimum temperature you should consider and if you want best performance >20C is best.
For me I set my battery heater to switch off above 22C and back on at 19C but on a cold winters day it’s on almost all day only switching off when I force charge during Octopus low tariff hours (as the individual cell voltage nears 3.45v the internal temperature rises rapidly)
A few observations, and things i’ve worked out -
The battery temperature reported by the Fox app and the original HA integration comes from the average of the high temperatures of all cells across all packs - my assumption was that the lowest cell temp (nearer the cold floor/wall) would be ~ 5C lower and so that was my reasoning for setting my enclosure minimum temp of 19C.
Since the latest BMS master/slave version it is possible to see all the actual cell temps (if you have agent access), but handily the BMS also reports the lowest and highest cell temps using the latest version of Home Assistant integrations and in warm enclosures the low/high and reported temperatures are within a two degrees of each other.
The latest BMS versions are night and day better than the previous ones, SoC tracking, stability and temperature range are as good as i’ve seen them, but once every month or two, I do a 90% DoD cycle from 100% to 10%, hold the 10% for an hour and then back to 100% again leaving the charger enabled for an hour after it gets to 100% this helps the BMS establish the limits and gives the cells time to bottom and top balance.
In an ideal world finer control of the charger would be preferable to ramp the current down steadily as it passes 90% charge, but maintain the voltage to improve absorption (and limit cell temp rises) - the FoxESS charger control doesn’t do that, it only steps down from high to ~1kw a few minutes before it gets to 100%.
With a little bit of work and the latest BMS/Inverter firmware you can from Home Assistant control the inverter charge current to achieve a charge ramp (Nathan’s EM integration does this) - this is the ideal charge model for LFP batteries.
I think the only thing left to say is that whilst cold is bad for capacity, it’s not harmful to the battery - whereas high temperature will age the battery, reducing its life and capacity permanently.
The batteries are fully tested to operate at 40C ambient but continued operation above 50C will reduce their life - so back to the insulated enclosure, having sufficient air flow or forced cooling is critically important in summer; I made mine out of insulated slabs that are easily removable and will find somewhere to store them from May through November.
Like you i’ve built an insulated enclosure (with a 40w heater controlled by my home assistant), and spent quite a bit of time researching LFP cell technology and reading the various studies that have been published on capacity over time and over a temperature range.
This chart comes from a research paper on LifePO4 batteries and shows how a single LFP cell ‘s capacity responds to temperature
Those are extreme ranges but FoxESS have now confirmed the following - that 100% capacity and cycle life are rated at 25C (ambient), 90%DoD and 0.5C charge and the battery capacity falls because of charge derating between 15C and 0C - at 0C they will not allow a charge, but they will allow discharge down to -10C.
That effectively means if you want to maintain a reasonable autonomy that 15C is the minimum temperature you should consider and if you want best performance >20C is best.
For me I set my battery heater to switch off above 22C and back on at 19C but on a cold winters day it’s on almost all day only switching off when I force charge during Octopus low tariff hours (as the individual cell voltage nears 3.45v the internal temperature rises rapidly)
A few observations, and things i’ve worked out -
The battery temperature reported by the Fox app and the original HA integration comes from the average of the high temperatures of all cells across all packs - my assumption was that the lowest cell temp (nearer the cold floor/wall) would be ~ 5C lower and so that was my reasoning for setting my enclosure minimum temp of 19C.
Since the latest BMS master/slave version it is possible to see all the actual cell temps (if you have agent access), but handily the BMS also reports the lowest and highest cell temps using the latest version of Home Assistant integrations and in warm enclosures the low/high and reported temperatures are within a two degrees of each other.
The latest BMS versions are night and day better than the previous ones, SoC tracking, stability and temperature range are as good as i’ve seen them, but once every month or two, I do a 90% DoD cycle from 100% to 10%, hold the 10% for an hour and then back to 100% again leaving the charger enabled for an hour after it gets to 100% this helps the BMS establish the limits and gives the cells time to bottom and top balance.
In an ideal world finer control of the charger would be preferable to ramp the current down steadily as it passes 90% charge, but maintain the voltage to improve absorption (and limit cell temp rises) - the FoxESS charger control doesn’t do that, it only steps down from high to ~1kw a few minutes before it gets to 100%.
With a little bit of work and the latest BMS/Inverter firmware you can from Home Assistant control the inverter charge current to achieve a charge ramp (Nathan’s EM integration does this) - this is the ideal charge model for LFP batteries.
I think the only thing left to say is that whilst cold is bad for capacity, it’s not harmful to the battery - whereas high temperature will age the battery, reducing its life and capacity permanently.
The batteries are fully tested to operate at 40C ambient but continued operation above 50C will reduce their life - so back to the insulated enclosure, having sufficient air flow or forced cooling is critically important in summer; I made mine out of insulated slabs that are easily removable and will find somewhere to store them from May through November.
Last edited by Dave Foster on Thu Apr 20, 2023 4:52 pm, edited 1 time in total.
As ever, thanks Dave for the very comprehensive response!
Since I started using the USR 410s rather than the LAN port I've been looking at the various high/low battery temp stats that the BMS gives me, I can see quite a big delta between the Low and High cell temps (around 9C), but a fairly small delta between the reported battery temp and the High temp (1-2C). This suggests to me that most of the battery bank is OK with some of it being much cooler - presumably the module closest to the floor. As an example here's the last week:
My batteries are on the concrete slab in the garage (albeit with an air gap under them) and the box basically sits over the top. They don't currently have any active heating, but they do OK with the heat they generate themselves when (dis)charging.
Can I ask what sort of heater you're using for yours? I have a thermometer sat on top of the stack next to the BMS, which usually reads around 20C but putting a small heat source at the bottom of the battery bank seems like it could be a good idea.
I'm still struggling to get my installer to update my BMS / battery slave firmware, I have the impression they are trying to do it by talking to FoxESS themselves, rather than using the higher level of access they (presumably) have. I'm still seeing a usable capacity of around 10kWh vs the ~13kWh I should have, and strange (large!) changes in SoC without the corresponding inflow / outflow of energy that "ought" to accompany them. Probably time to get a little...firmer about it with them!
Anyway, once I get that sorted, I'll look into the kind of cell balancing exercise you mention. I'm also interested in what you mention about the EM integration, presumably you mean this? I'm very curious how this manages to change the behaviour of the charger in the inverter - unless I misunderstand!
Since I started using the USR 410s rather than the LAN port I've been looking at the various high/low battery temp stats that the BMS gives me, I can see quite a big delta between the Low and High cell temps (around 9C), but a fairly small delta between the reported battery temp and the High temp (1-2C). This suggests to me that most of the battery bank is OK with some of it being much cooler - presumably the module closest to the floor. As an example here's the last week:
Can I ask what sort of heater you're using for yours? I have a thermometer sat on top of the stack next to the BMS, which usually reads around 20C but putting a small heat source at the bottom of the battery bank seems like it could be a good idea.
I'm still struggling to get my installer to update my BMS / battery slave firmware, I have the impression they are trying to do it by talking to FoxESS themselves, rather than using the higher level of access they (presumably) have. I'm still seeing a usable capacity of around 10kWh vs the ~13kWh I should have, and strange (large!) changes in SoC without the corresponding inflow / outflow of energy that "ought" to accompany them. Probably time to get a little...firmer about it with them!
Anyway, once I get that sorted, I'll look into the kind of cell balancing exercise you mention. I'm also interested in what you mention about the EM integration, presumably you mean this? I'm very curious how this manages to change the behaviour of the charger in the inverter - unless I misunderstand!
-
Dave Foster
- Posts: 1305
- Joined: Thu Oct 13, 2022 7:21 pm
This is the heater I use heater, basically a greenhouse tubular heater that sits on the floor in front of the stack.
My batteries sit on a cold garage floor and they are up against an external wall, you can see my individual cell temps from the attached - nice and warn at the top (M1), and cold at the bottom (M7) with a bit of variance (I assume front to back).
Weirdly the battery temperature sensor seems to be nearer the average of the highest cell temps (not of all the cells) which seems strange as we're so worried about the cold here, but I guess these storage systems are much more prevalent in hot countries Australia etc.. where it's the fear of damage which comes from over heating cells which is driving the data.
Fingers crossed you get the latest battery firmware soon, I think it will resolve a lot of the issues you are seeing, the capacity is better and the soc is so much more chilled than that it was.
My batteries sit on a cold garage floor and they are up against an external wall, you can see my individual cell temps from the attached - nice and warn at the top (M1), and cold at the bottom (M7) with a bit of variance (I assume front to back).
Fingers crossed you get the latest battery firmware soon, I think it will resolve a lot of the issues you are seeing, the capacity is better and the soc is so much more chilled than that it was.
I'm assuming from your other posts that you have "installer" access to your Fox Cloud account when I ask this, but do I understand correctly that with that level of access, you can install the upgraded firmware to a remote inverter or BMS unit with "just" a few mouse clicks?Fingers crossed you get the latest battery firmware soon
I'm just trying to find out how it "should" work, I can't fault the physical / electrical work done for my system but they don't seem to be very well organised on the firmware upgrade side of things. In extremis, can they "hand over the keys" to me to do myself?
-
Dave Foster
- Posts: 1305
- Joined: Thu Oct 13, 2022 7:21 pm
Yes installer and agent access (strangely it’s agent that gives you the real control) - 6 months ago agent access was quite important as there were to be fair a few issues, particularly with battery firmware, probably not as important now the f/w is better and with the HA integration.calum wrote: ↑Thu Apr 20, 2023 5:55 pmI'm assuming from your other posts that you have "installer" access to your Fox Cloud account when I ask this, but do I understand correctly that with that level of access, you can install the upgraded firmware to a remote inverter or BMS unit with "just" a few mouse clicks?Fingers crossed you get the latest battery firmware soon
I'm just trying to find out how it "should" work, I can't fault the physical / electrical work done for my system but they don't seem to be very well organised on the firmware upgrade side of things. In extremis, can they "hand over the keys" to me to do myself?
My installer had an issue setting up all their accounts so they aren’t able to update the firmware and chasing Fox regularly for updates, isn’t that fair and frustrating - there is a less well known way to correct this, which is to buy a new wifi datalogger, (~£30) create an agent account, fit the data logger to your inverter /connect to wifi - add it to the agent account and your inverter will now show up in that account with full access, if you switch back to the old logger it will still be in there as a user. BUT with the agent account you can update the inverter to latest firmware yourself and when done just revert back to the old data logger which keeps your installer happy, until the next time you want to update. - where are you located?, if in the North West you can borrow my wifi logger, update your firmware and revert back (i’ve done this a couple of times on my friends systems)
I'm in Stockport so smack dab in the "NW" region 
I'm wondering how much more often the firmware will need updating - there is definitely an argument for getting to a stable version of a firmware (on anything!) and sticking to it, it feels like (from what I've read on here and elsewhere) that the FW for the Fox H1 units is getting to the state it probably should have been at launch, and we're probably close to the "corner case bug fixing" stage of maturity. All of which is to say your offer would be much appreciated if the logistics can work out!
Edit: The module is this thing, right? That's very tempting to be able to "dual stack" the inverter...
I'm wondering how much more often the firmware will need updating - there is definitely an argument for getting to a stable version of a firmware (on anything!) and sticking to it, it feels like (from what I've read on here and elsewhere) that the FW for the Fox H1 units is getting to the state it probably should have been at launch, and we're probably close to the "corner case bug fixing" stage of maturity. All of which is to say your offer would be much appreciated if the logistics can work out!
Edit: The module is this thing, right? That's very tempting to be able to "dual stack" the inverter...
-
Dave Foster
- Posts: 1305
- Joined: Thu Oct 13, 2022 7:21 pm
Lol, face palm when I looked at your locationcalum wrote: ↑Thu Apr 20, 2023 9:11 pm I'm in Stockport so smack dab in the "NW" region
Edit: The module is this thing, right? That's very tempting to be able to "dual stack" the inverter...
Yes that’s the module, it’s such a low cost relatively (after a solar/battery install) to give you the extra control and freedom, and doesn’t break your data or your installers access when you switch back to the old dongle.
Have a think which way you want to go, most weeks i’m usually somewhere between Lymm and North Wales so no problem for me to arrange to pop over.
That's a very kind offer and one I might yet take you up on
I was able to find a 'new' WiFi logger dongle delivered for £20, which arrived today. I've registered an Agent account and added the datalogger by its serial number, but it looks like it was already registered to another device in Brazil
Interestingly, not a Fox device, but a "Linha SIW200G" - unless they sell their kit under other names...
Do you know if there's any way for me to break the link with this other device, or is that permanent once set up? I'm wondering if I connect it to my Fox H1 it'll recognise that it's Fox hardware and let me upgrade the f/w, I don't actually need it to work properly for datalogging purposes!
That said, if you think this seems like a bad idea I'll get a refund from the (eBay) seller and get another one from a more...legitimate source
(I thought I cured myself of trying to save relatively small amounts of money and wasting time in doing so, apparently the cure was not completely successful!)
I was able to find a 'new' WiFi logger dongle delivered for £20, which arrived today. I've registered an Agent account and added the datalogger by its serial number, but it looks like it was already registered to another device in Brazil
Interestingly, not a Fox device, but a "Linha SIW200G" - unless they sell their kit under other names...Do you know if there's any way for me to break the link with this other device, or is that permanent once set up? I'm wondering if I connect it to my Fox H1 it'll recognise that it's Fox hardware and let me upgrade the f/w, I don't actually need it to work properly for datalogging purposes!
That said, if you think this seems like a bad idea I'll get a refund from the (eBay) seller and get another one from a more...legitimate source
(I thought I cured myself of trying to save relatively small amounts of money and wasting time in doing so, apparently the cure was not completely successful!)
-
Dave Foster
- Posts: 1305
- Joined: Thu Oct 13, 2022 7:21 pm
Ouch, i’ve never seen one registered to one of those before - unfortunately the link you can’t break (without FoxEss support) is the link between datalogger and a users account but it would throw and error and not let you add it if it were still registered to a users account.
If it lets you add it to your agent account and you see another inverter, it sounds as if it was registered to a user but it isn’t anymore in which case plug it into your inverter and add it to wifi and within 5-10 minutes your inverter should show up in the cloud account.
Assuming it does show up you can go to the remote firmware update tab, start with inverter remote upgrade, you need to copy and paste your inverter into the search box, and it will populate, select it and click upgrade and create the job, you can only do one task at a time, always start with Master, latest is 1.54(H) and set a timeout of 45 then click upgrade… make a cup of tea as its takes a while, you can check on progress by clicking on upgrade status. When complete upgrade your manager as soon as possible to 1.57 (H) again wait for completion. I always do a full shutdown and restart at this point.
Then onto battery, select HV2600 search and you’ll see your inverter the little note under battery info will show current versions - again select and upgrade, create the job starting with Master 1.018 let it complete then onto slave 1.15 which will take a while as it goes to each pack in turn - once complete I do a full shutdown and restart again and that’s it done.
If you can’t add the datalogger / inverter to your account - message me your mobile and i’ll arrange to pop over.
If it lets you add it to your agent account and you see another inverter, it sounds as if it was registered to a user but it isn’t anymore in which case plug it into your inverter and add it to wifi and within 5-10 minutes your inverter should show up in the cloud account.
Assuming it does show up you can go to the remote firmware update tab, start with inverter remote upgrade, you need to copy and paste your inverter into the search box, and it will populate, select it and click upgrade and create the job, you can only do one task at a time, always start with Master, latest is 1.54(H) and set a timeout of 45 then click upgrade… make a cup of tea as its takes a while, you can check on progress by clicking on upgrade status. When complete upgrade your manager as soon as possible to 1.57 (H) again wait for completion. I always do a full shutdown and restart at this point.
Then onto battery, select HV2600 search and you’ll see your inverter the little note under battery info will show current versions - again select and upgrade, create the job starting with Master 1.018 let it complete then onto slave 1.15 which will take a while as it goes to each pack in turn - once complete I do a full shutdown and restart again and that’s it done.
If you can’t add the datalogger / inverter to your account - message me your mobile and i’ll arrange to pop over.
Thanks for the info, I'll give it a go this weekend. Physical access is currently "challenging" because we are having a new heating system installed and the heating engineers have filled my garage with boxes of their gear!
I've just checked on the app and my H1 firmware versions are Master 1.54 / Slave 1.02 / Manager 1.57, so it looks like I'm on the latest there. Presumably I can just go straight to updating the BMS / battery firmware and then use the power button on the front of the BMS to restart that when it's done? Or would you advise restarting the whole system? Presumably I would use the Inverter Shut Down and Startup procedures in the manual (sections 6.8 and 6.7 respectively) if so.
-
Dave Foster
- Posts: 1305
- Joined: Thu Oct 13, 2022 7:21 pm
Yes you can skip the inverter and go straight to the batteries, start with master first wait for it complete and then upgrade the slave (that will update all of them so give it time to complete).
When you’re done definitely shutdown the BMS, but it’s best not to switch it off when it is supplying load - so a full shutdown / restart is worth it.
Basically turn off the PV (DC) isolator, turn off the AC Isolator, then the Battery Power button, and finally the battery breaker.
After a few seconds the inverter will go blank - you’ll hear the clunk as the batteries go off but the lights do still flash so ignore that. Wait 10 seconds and then Battery breaker on, Power button on, AC Isolator on, DC Isolator on - the inverter will do it’s checking countdown and then you’ll be good to go.
Will has done a video which is very helpful to guide you through the restart viewtopic.php?t=5
When you’re done definitely shutdown the BMS, but it’s best not to switch it off when it is supplying load - so a full shutdown / restart is worth it.
Basically turn off the PV (DC) isolator, turn off the AC Isolator, then the Battery Power button, and finally the battery breaker.
After a few seconds the inverter will go blank - you’ll hear the clunk as the batteries go off but the lights do still flash so ignore that. Wait 10 seconds and then Battery breaker on, Power button on, AC Isolator on, DC Isolator on - the inverter will do it’s checking countdown and then you’ll be good to go.
Will has done a video which is very helpful to guide you through the restart viewtopic.php?t=5
Awesome, thank you, particularly for the link to the video, I'd forgotten about that one 
We have flattened our battery today as we're using the jury rigged immersion heater in the new hot water cylinder until the heatpump is wired in and commissioned. So I'll hopefully do the BMS f/w upgrade tonight, do the full restart, and then charge the batteries back up to full overnight, which will hopefully give the new firmware chance to "get familiar" with the performance of the battery bank.
We have flattened our battery today as we're using the jury rigged immersion heater in the new hot water cylinder until the heatpump is wired in and commissioned. So I'll hopefully do the BMS f/w upgrade tonight, do the full restart, and then charge the batteries back up to full overnight, which will hopefully give the new firmware chance to "get familiar" with the performance of the battery bank.
Success!!!!! 
All system components have the latest firmware, and as of now the battery is at its SoC floor, programmed to charge to 100% overnight. Hopefully the numbers will see it hit 100% after accepting around 13kWh, as per spec!
Many thanks once again for your help, I owe you at least one bad hangover's worth of beers
All system components have the latest firmware, and as of now the battery is at its SoC floor, programmed to charge to 100% overnight. Hopefully the numbers will see it hit 100% after accepting around 13kWh, as per spec!
Many thanks once again for your help, I owe you at least one bad hangover's worth of beers
So, overnight 11.8kWh went into the battery, which is an improvement over what I've seen before in similar situations but still a bit sort of the 12+ I was hoping for.
I'm guessing at least part of this is down to temperature based de-rating; at the start of the charge period the Low Cell Temp was around 14C, whereas the Battery Temperature reading was 21C or so. It'll be interesting to see how this tracks as the weather warms up. I'll look into the greenhouse heater you posted above, my only concern is how I actuall fit it into the enclosure which is pretty snug around the battery/BMS stack!
I also wonder if it takes the BMS a few charges to get a handle on the batteries and get the most out of them, but that's maybe just wishful thinking on my part!
I'm guessing at least part of this is down to temperature based de-rating; at the start of the charge period the Low Cell Temp was around 14C, whereas the Battery Temperature reading was 21C or so. It'll be interesting to see how this tracks as the weather warms up. I'll look into the greenhouse heater you posted above, my only concern is how I actuall fit it into the enclosure which is pretty snug around the battery/BMS stack!
I also wonder if it takes the BMS a few charges to get a handle on the batteries and get the most out of them, but that's maybe just wishful thinking on my part!
-
Dave Foster
- Posts: 1305
- Joined: Thu Oct 13, 2022 7:21 pm
Great, it should be much better behaved now 
The kWh remaining does take a few charge cycles to get it’s bearings and you do occasionally see the odd jump in SoC as it approaches full but typically when it’s charging from solar - but on the whole, much better
The kWh remaining does take a few charge cycles to get it’s bearings and you do occasionally see the odd jump in SoC as it approaches full but typically when it’s charging from solar - but on the whole, much better
13.6kWh went into the battery last night
Technically this is more than "should" fit but of course charging a battery isn't a perfectly efficient process.
Either way, to anyone browsing the forums with "I don't seem to be getting the proper capacity out of my batteries" issues, or seeing strange SoC changes - get your kit onto the latest firmware version, and then do a few "full to empty" cycles. If my experience is anything to go by, you should see a significant improvement.
I'd have to back over some of my old data to confirm this but I think I was missing out on about 25% of the usable capacity of my batteries so well worth it.
Technically this is more than "should" fit but of course charging a battery isn't a perfectly efficient process.
Either way, to anyone browsing the forums with "I don't seem to be getting the proper capacity out of my batteries" issues, or seeing strange SoC changes - get your kit onto the latest firmware version, and then do a few "full to empty" cycles. If my experience is anything to go by, you should see a significant improvement.
I'd have to back over some of my old data to confirm this but I think I was missing out on about 25% of the usable capacity of my batteries
so well worth it.-
Dave Foster
- Posts: 1305
- Joined: Thu Oct 13, 2022 7:21 pm
Fantastic result
You’ve got the battery chemistry to contend with and an RTE of around 95% so 13.6kw in will be nearer your 12.8kw out.
You’ve got the battery chemistry to contend with and an RTE of around 95% so 13.6kw in will be nearer your 12.8kw out.
I've always been a bit bothered by the large disparity between the 'High' and 'Low' cell temps once I've put the lid on the battery box for winter. It's always about 9C, the temperatures just move up and down depending on ambient as well as charging, discharging etc. "High" temps get a little bit high after charging, even though the "low" temps are still some way off the ideal range in the 20s C. I could of course put some trace heating at the bottom of the box, but that would make the top even warmer.
But then I had an idea, what if I used some of the spare PC fans I have here to circulate the air around the box? I rigged up a slightly janky set of adaptors to let a pair of 92mm 12V DC fans blow downwards from just below the BMS unit. They fit quite well there as that's where I left a gap for all the cables at the front of the stack.
Aaaaand - it works pretty well! In the temperature trace below, the yellow line is where I put the fans in, where the usual ~9C delta narrows considerably, it widens when the battery is charging then narrows over the following hours. Best it's down to about 2C, but even at its widest it's only about 3.5C, ie just over a third of what it was.
We're charging and discharging the battery completely every day at the moment, and we pretty much see the same ~13.5kWh going in every night, hopefully this will help that amount hold up as we go through the winter months. I am still thinking about adding a greenhouse heater since I have space to do that at the bottom of the box, and the fans would help spread that warmth around.
Here's a couple of pics to show off the full janky glory
Edit: For some reason it's displaying the thumbnails upside down, but if you click through it's the right way up
But then I had an idea, what if I used some of the spare PC fans I have here to circulate the air around the box? I rigged up a slightly janky set of adaptors to let a pair of 92mm 12V DC fans blow downwards from just below the BMS unit. They fit quite well there as that's where I left a gap for all the cables at the front of the stack.
Aaaaand - it works pretty well! In the temperature trace below, the yellow line is where I put the fans in, where the usual ~9C delta narrows considerably, it widens when the battery is charging then narrows over the following hours. Best it's down to about 2C, but even at its widest it's only about 3.5C, ie just over a third of what it was.
Here's a couple of pics to show off the full janky glory
Nice neat battery box and fan setup.
Have you used the battery box over a winter yet? Did you find the 50mm of polystyrene worked well enough to keep batteries warm enough to accept decent charge?
Mine are in the loft and we had them fitted last Feb so didn't really see much of the winter last year. I have been thinking of just making a polystyrene box round the batteries but unsure if I need more than 25mm thick or would it be best to go 50mm?
Cheers for any suggestions.
Have you used the battery box over a winter yet? Did you find the 50mm of polystyrene worked well enough to keep batteries warm enough to accept decent charge?
Mine are in the loft and we had them fitted last Feb so didn't really see much of the winter last year. I have been thinking of just making a polystyrene box round the batteries but unsure if I need more than 25mm thick or would it be best to go 50mm?
Cheers for any suggestions.
We had our system fitted almost exactly a year ago, and the unprotected batteries stopped working several times during the cold snaps we had last winter. I built the box in mid January, and after that everything worked pretty well, although we probably weren't getting the absolute best capacity out of the batteries. I suspect a lot of that was more down to firmware than temps, though.
If I were doing it again (or advising someone else!) I would say make sure you insulate all six sides, allow for internal air circulation, and ideally allow that air circulation to become air throughput if needed - potentially allows you to duct in cooler air from outside if elevated battery temps are a concern.
I bought the 25mm sheets of EPS in a pack because a) they would fit in my car easily and b) it was the cheapest way to get the amount of insulation I needed. I cut them up to fit on the OSB panels. The back panel of the box (not visible in the photo) only has 25mm of insulation, because otherwise it doesnt fit! It maybe that is part of the reason for the variance in my case, there are individual cell temp values available in the Foxess website, and I daresay I could work out which cells are at the front or back of the banks, but we'd be well into diminishing returns at that point. But I put the box together in the garage and all the other panels got 50mm of insulation from the off, so I can't really say how much difference it would have made vs 25mm (or 50mm!) all round.
If you're going to the trouble of getting up into the loft and building an enclosure, I'd go with 50mm to begin with. How warm did your batteries get in the summer? if you expect to need to dismantle the thing to allow them to cool, I'd aim to add some hard materials to the outside, otherwise just making it out of insulation will likely fall to bits after a few seasons, especially with EPS. The Kingspan polyurethane stuff might do better but I'd still suggest you do something structural even then. I did 9mm OSB for lightness, with battens on the corners and threaded inserts and bolts so it won't mind being disassembled and reassembled.
If I were doing it again (or advising someone else!) I would say make sure you insulate all six sides, allow for internal air circulation, and ideally allow that air circulation to become air throughput if needed - potentially allows you to duct in cooler air from outside if elevated battery temps are a concern.
I bought the 25mm sheets of EPS in a pack because a) they would fit in my car easily and b) it was the cheapest way to get the amount of insulation I needed. I cut them up to fit on the OSB panels. The back panel of the box (not visible in the photo) only has 25mm of insulation, because otherwise it doesnt fit! It maybe that is part of the reason for the variance in my case, there are individual cell temp values available in the Foxess website, and I daresay I could work out which cells are at the front or back of the banks, but we'd be well into diminishing returns at that point. But I put the box together in the garage and all the other panels got 50mm of insulation from the off, so I can't really say how much difference it would have made vs 25mm (or 50mm!) all round.
If you're going to the trouble of getting up into the loft and building an enclosure, I'd go with 50mm to begin with. How warm did your batteries get in the summer? if you expect to need to dismantle the thing to allow them to cool, I'd aim to add some hard materials to the outside, otherwise just making it out of insulation will likely fall to bits after a few seasons, especially with EPS. The Kingspan polyurethane stuff might do better but I'd still suggest you do something structural even then. I did 9mm OSB for lightness, with battens on the corners and threaded inserts and bolts so it won't mind being disassembled and reassembled.
-
Dave Foster
- Posts: 1305
- Joined: Thu Oct 13, 2022 7:21 pm
I’ve done something similar this year, not quite as neatly as you and with a lot more duck tape and ty-raps - but I settled on the same principle with fans at the top blowing down as it’s always a lot hotter at the top and the concrete floor anchors the temperature.
I’ve got a 40W tubular heater at the bottom which is turned on when the average of coldest and highest cells drops below 20C but with the mild weather at the moment it’s not needed the batteries are keeping themselves warm on their overnight charge.
There’s now just 3-4C difference between the highest and lowest cells, I can’t say i’m looking forward to a cold snap but it should handle it better than last year.
@Calum, there is new firmware out for the HV2600’s (this week), but i’m really reticent to try it as what i’ve got is working very well and I think i’ll be sticking with what it.
I’ve got a 40W tubular heater at the bottom which is turned on when the average of coldest and highest cells drops below 20C but with the mild weather at the moment it’s not needed the batteries are keeping themselves warm on their overnight charge.
There’s now just 3-4C difference between the highest and lowest cells, I can’t say i’m looking forward to a cold snap but it should handle it better than last year.
@Calum, there is new firmware out for the HV2600’s (this week), but i’m really reticent to try it as what i’ve got is working very well and I think i’ll be sticking with what it.
I really wish that Fox issued even brief notes for their firmware updates. I've not touched my firmware since April or so, it seems to be working fine but at some point I probably do need to bring the inverter firmware up to the latest, and that's probably a good time to do the batteries as well. But yeah, always nervous about regressions with these sorts of updates!@Calum, there is new firmware out for the HV2600’s (this week), but i’m really reticent to try it as what i’ve got is working very well and I think i’ll be sticking with what it.