diy solar

diy solar

Backup system planning

Petrol Head

New Member
Joined
Jul 8, 2021
Messages
13
Hi all, how do I go about planning a solar system to use as a back up for the freezer in the UK ? I have 880w in solar panel potential, 3x 260w & 1x 100w, hopefully 2 more 260w on the way giving 1.4kw. I also have an inverter 2000w PSW. I will buy the battery bank needed, that's the bit I'm not sure about. I have a meter connected to the freezer that gives me the max start up draw 1203w & running draw 67w, does that sound right ?
Any suggestions on battery type, rating & quantity ? Thanks in advance for any advice.
 
Freezer = 250kwh per year (from manufacturers data sheet) = 0.685kwh per day = 685 watt hours per day.
Chosen back up because I'm in cloudy UK 3 days. 685 x 3 = 2055 wh.
2055 / 12v = 171.25 amp hours.
50% discharge = 325.5 ah
25% discharge = 685 ah.
Am I right with the maths in thinking I need somewhere between 325-685 in battery capacity ?
 
You’re on the right track. I will only Comment that with some motors, the start up draw happens so quick your typical meter like a kilowatt meter may not catch what an inverter truly needs. In my planning, its hard to plan how big of an inverter you actually need. For a 67 watt motor, a 2000 watt PSW seems fine.

THe manufacturers spec sheet may be good, but that won’t really cover the summer performance working more versus a winter performance working less. So, this could skew the results. It’s probably working its hardest now since its summer, so a kilawatt meter would give you worst case data versus putting that on in the winter where it’d be best case data. It’s good to build the system to worts case.

In my signature block there’s some math on hooking disimilar panels together in series and parallel. Short story is, if voltage is off by too much, hooking the 100 watt Panel in parallel, you could lose more power than you gain. Can’t really hook it in series either.
 
The numbers generally sound good.
UK is famous for being overcast. How much sun you get is key.
Since you'll get several times as much sun in summer as winter, panels tilted for max winter production probably works best.

No mention of charge controller? Select SCC and design PV array (series/parallel) to fit each other. Voc of series PV string must never exceed max input voltage of SCC. You can calculate exactly using Voc of PV panel, temperature coefficient, and record cold temperature of your location. Or multiply Voc by 1.16 for a safe estimate.

More PV is always good. 5 x 260W is a good amount of power for your need, but can only be configured 5s or 5p. If you can get six, 3s2p and 2s3p give additional voltage options. With multiple panels or series strings, I like multiple panels (or series strings) oriented toward morning and afternoon sun to flatten curve of power production.

I usually suggest powering freezer during the day only and not at night to get by with small battery, but since you expect multiple days without sun that doesn't help.

Your surge current measurement sounds accurate. I've measured other motors (not refrigerator) with scope, and see about 5x nameplate rating. 67W running is nice and low. 685 Wh/day is about half a full-size fridge sold in the U.S.

50% and 25% DoD indicate lead-acid. If 3 days without sun is rare, you could go to 75% on that rare occasion. If power outages are rare you can cycle deeply. I use AGM for my whole-house backup system, with 70% DoD low-battery disconnect. AGM or Gel is less work than FLA (no maintenance) but doesn't last as long.

Lithium gives more cycles, but costs proportionately more unless you DIY, which can cost just 40% of AGM to purchase. But DIY and LiFePO4 have various issues to deal with. For this size bank and fewer than 300 discharge cycles over a decade, I don't think it is worthwhile.

Here is the battery I use, 6V 405AH AGM, so two for about $625 each. You may find similar brands locally.


Here's one of the best FLA


You will be limited to what is actually available with reasonable lead time, due to current market conditions. One forum member was going to get SunXtender but was many months before delivery, so he bought Rolls that was in-stock.

Either type of Lead-acid, but especially FLA, prefers a specific charge current. For FLA, probably 0.12C or 0.15C. Oversize PV array and an SCC that limits current to that target charge rate would be good. Important that SCC has battery temperature sensor and a good charge profile. For FLA, check cell specific gravity periodically and give it an equalizing charge when needed.

If you have an inverter already, probably doesn't have AC transfer switch. When power fails you can move power cord to inverter and turn inverter on. Or rig up automatic transfer switch. Inverter running continuously? That drains battery a bit at night, cycling batteries unnecessarily. Could automate remote on/off switch. Or have a float charger plugged into grid power to keep it charged at night.

Oops, we forgot something: inverter no-load drain, which could be as much as refrigerator's 67W but 24/7, never turns off. If 50W, then 1200 Wh/day, twice the refrigerator's 658 Wh/day. Look up no-load consumption. If possible, measure. A 12V timer could turn inverter on/off by its remote switch (if it has one), so only operated during 12 daylight hours, cutting drain in half. Freezer needed about 10 hours operation (685W/67W), but my guess is it runs even fewer hours, lower kWh consumption, in cool weather. Wire an extra freezer thermostat to inverter remote on/off switch?

Need to recalculate battery capacity to cover inverter no-load consumption and hours of operation per day.
I like the idea of thermostat turning inverter on/off best. Set slightly colder than freezer's thermostat so inverter only runs when freezer does. (note that if refrigerator/freezer, thermostat is in refrigerator and freezer runs colder based on air valve.)
 
Last edited:
You’re on the right track. I will only Comment that with some motors, the start up draw happens so quick your typical meter like a kilowatt meter may not catch what an inverter truly needs. In my planning, its hard to plan how big of an inverter you actually need. For a 67 watt motor, a 2000 watt PSW seems fine.

THe manufacturers spec sheet may be good, but that won’t really cover the summer performance working more versus a winter performance working less. So, this could skew the results. It’s probably working its hardest now since its summer, so a kilawatt meter would give you worst case data versus putting that on in the winter where it’d be best case data. It’s good to build the system to worts case.

In my signature block there’s some math on hooking disimilar panels together in series and parallel. Short story is, if voltage is off by too much, hooking the 100 watt Panel in parallel, you could lose more power than you gain. Can’t really hook it in series either.
Thanks chrisski. The meter still has a high of 1205 & also an overall reading of 1.5kw usage in 43 hours, @ 837 per day, a little higher than my 685 per day from the manufacturers guidelines, probably the summer performance as you mentioned. I'll keep the meter going. I don't think I'll add the 100w panel to the 260w string, will use separately for other charging.
 
Depending on how your build goes, a second charge controller for the 100 watt panel could be an option. Could even choose to add more 100 watt panels.

I’ve got a spare 100 watt panel I’m turning into a a portable USB charging station. Overall, USB charging is a negligible impact on my charging needs, but its an interesting project. I may charge a top of 100 watt hours a day on USB, but probably closer to 20 or 30 watt hours most days.
 
The numbers generally sound good.
UK is famous for being overcast. How much sun you get is key.
Since you'll get several times as much sun in summer as winter, panels tilted for max winter production probably works best.

No mention of charge controller? Select SCC and design PV array (series/parallel) to fit each other. Voc of series PV string must never exceed max input voltage of SCC. You can calculate exactly using Voc of PV panel, temperature coefficient, and record cold temperature of your location. Or multiply Voc by 1.16 for a safe estimate.

More PV is always good. 5 x 260W is a good amount of power for your need, but can only be configured 5s or 5p. If you can get six, 3s2p and 2s3p give additional voltage options. With multiple panels or series strings, I like multiple panels (or series strings) oriented toward morning and afternoon sun to flatten curve of power production.

I usually suggest powering freezer during the day only and not at night to get by with small battery, but since you expect multiple days without sun that doesn't help.

Your surge current measurement sounds accurate. I've measured other motors (not refrigerator) with scope, and see about 5x nameplate rating. 67W running is nice and low. 685 Wh/day is about half a full-size fridge sold in the U.S.

50% and 25% DoD indicate lead-acid. If 3 days without sun is rare, you could go to 75% on that rare occasion. If power outages are rare you can cycle deeply. I use AGM for my whole-house backup system, with 70% DoD low-battery disconnect. AGM or Gel is less work than FLA (no maintenance) but doesn't last as long.

Lithium gives more cycles, but costs proportionately more unless you DIY, which can cost just 40% of AGM to purchase. But DIY and LiFePO4 have various issues to deal with. For this size bank and fewer than 300 discharge cycles over a decade, I don't think it is worthwhile.

Here is the battery I use, 6V 405AH AGM, so two for about $625 each. You may find similar brands locally.


Here's one of the best FLA


You will be limited to what is actually available with reasonable lead time, due to current market conditions. One forum member was going to get SunXtender but was many months before delivery, so he bought Rolls that was in-stock.

Either type of Lead-acid, but especially FLA, prefers a specific charge current. For FLA, probably 0.12C or 0.15C. Oversize PV array and an SCC that limits current to that target charge rate would be good. Important that SCC has battery temperature sensor and a good charge profile. For FLA, check cell specific gravity periodically and give it an equalizing charge when needed.

If you have an inverter already, probably doesn't have AC transfer switch. When power fails you can move power cord to inverter and turn inverter on. Or rig up automatic transfer switch. Inverter running continuously? That drains battery a bit at night, cycling batteries unnecessarily. Could automate remote on/off switch. Or have a float charger plugged into grid power to keep it charged at night.

Oops, we forgot something: inverter no-load drain, which could be as much as refrigerator's 67W but 24/7, never turns off. If 50W, then 1200 Wh/day, twice the refrigerator's 658 Wh/day. Look up no-load consumption. If possible, measure. A 12V timer could turn inverter on/off by its remote switch (if it has one), so only operated during 12 daylight hours, cutting drain in half. Freezer needed about 10 hours operation (685W/67W), but my guess is it runs even fewer hours, lower kWh consumption, in cool weather. Wire an extra freezer thermostat to inverter remote on/off switch?

Need to recalculate battery capacity to cover inverter no-load consumption and hours of operation per day.
I like the idea of thermostat turning inverter on/off best. Set slightly colder than freezer's thermostat so inverter only runs when freezer does. (note that if refrigerator/freezer, thermostat is in refrigerator and freezer runs colder based on air valve.)
Thank you Hegdes, a lot of good info to digest there.

I have a couple of cheap MPPT controllers, a 100a, 1300w @12v, 2600w @24v also a 40a.

I'm still undecided as to which route I take from here, purchase new quality components at much more expense or carry on & try to put a system together with 2nd hand/free stuff (but safe of course). I'll see how many panels I end up with, an even number gives me more options as you've said but my inverter is 12v so would dictate a series string ?

Batteries ... still undecided & thanks for the info.

Much to think about :)
 
Depending on how your build goes, a second charge controller for the 100 watt panel could be an option. Could even choose to add more 100 watt panels.

I’ve got a spare 100 watt panel I’m turning into a a portable USB charging station. Overall, USB charging is a negligible impact on my charging needs, but its an interesting project. I may charge a top of 100 watt hours a day on USB, but probably closer to 20 or 30 watt hours most days.
'interesting project' absolutely, especially making something that works, is safe, beneficial & at very little cost. All my USB is solar & a few gadgets too, it's not the saving in energy usage, which would be negligible, it's the satisfaction of doing it.
 
I think for this application the best solution is a UPS type setup where the system is always feeding the freezer, not just in emergencies. My preference would be an all in one (charge controller/inverter) with grid assist so that if you have solar, that is what the system is using, otherwise you are just using pass through from the grid. When the power goes out, the system switches to solar (if available) and/or battery(s). This way the system is always working and you don't get any surprises if (when?) the power goes out, but you still have the reliability of the grid when the sun is not out. A small generator would be a good idea to have as well. You don't need much to keep the battery(s) charged. A 1000w generator would do it since the draw is pretty low.
 
I'll see how many panels I end up with, an even number gives me more options as you've said but my inverter is 12v so would dictate a series string ?

With MPPT (not PWM, and not PWM made by brand "MPPT"!) you can have higher voltage panels or series string, and the power is converted to battery voltage at higher current. Some MPPT accept up to 100V, others 145, 150, 200, 250, 600V. Prices vary, as does efficiency vs. input voltage.

With 6 panels you might do 6p, 2s3p, 3s2p, 6s depending on how voltage fits MPPT specifications.
At 12V, current gets pretty high so fat wires to battery, and fat wires after connecting several PV panels in parallel.
Having some panels (or series strings of panels) at different angles can reduce peak current by as much as 30%.
 
I think for this application the best solution is a UPS type setup where the system is always feeding the freezer, not just in emergencies. My preference would be an all in one (charge controller/inverter) with grid assist so that if you have solar, that is what the system is using, otherwise you are just using pass through from the grid. When the power goes out, the system switches to solar (if available) and/or battery(s). This way the system is always working and you don't get any surprises if (when?) the power goes out, but you still have the reliability of the grid when the sun is not out. A small generator would be a good idea to have as well. You don't need much to keep the battery(s) charged. A 1000w generator would do it since the draw is pretty low.
Thanks, I'll look into that setup, I do also have 2 generators tucked away.
 
With MPPT (not PWM, and not PWM made by brand "MPPT"!) you can have higher voltage panels or series string, and the power is converted to battery voltage at higher current. Some MPPT accept up to 100V, others 145, 150, 200, 250, 600V. Prices vary, as does efficiency vs. input voltage.

With 6 panels you might do 6p, 2s3p, 3s2p, 6s depending on how voltage fits MPPT specifications.
At 12V, current gets pretty high so fat wires to battery, and fat wires after connecting several PV panels in parallel.
Having some panels (or series strings of panels) at different angles can reduce peak current by as much as 30%.
The max pv voltage is 50v for the 100a 'MPPT' if using that would it dictate a parallel string is the only option ?

Battery - I've just calculated the cost of a DIY Lifepo4 12.8v 310ah (4x 3.2v 310ah cells) & blue tooth BMS for under £600 ($825), that has got my interest, 310ah Lifepo4 should be more than plenty based on 171ah requirement, maybe even 200ah because of the better discharge %, hope I've not over looked anything, but still looking at all battery options. Thanks for your help Hedges.
 
The max pv voltage is 50v for the 100a 'MPPT' if using that would it dictate a parallel string is the only option ?
Usually SCCs are rated volts higher than amps like a Vitron 100/50 which is 100 volt max input, and 50 amp max output.
 
The max pv voltage is 50v for the 100a 'MPPT' if using that would it dictate a parallel string is the only option ?

Look over the specs closely or post a link. Could be MPPT with low headroom.
PWM accept a range of input voltage, but should always be close to battery voltage, so need to make sure this one isn't PWM.
 
Look over the specs closely or post a link. Could be MPPT with low headroom.
PWM accept a range of input voltage, but should always be close to battery voltage, so need to make sure this one isn't PWM.
The label has been bleached by the sun but you can just make out the figures, the other image is the same as I bought, not that exact item, it was a couple of years ago.
No worries if not suitable, I can get another.
 

Attachments

  • 20210724_113239.jpg
    20210724_113239.jpg
    333.9 KB · Views: 2
  • Screen Shot 2021-07-24 at 11.53.09.png
    Screen Shot 2021-07-24 at 11.53.09.png
    517.7 KB · Views: 2
That's a well-known style, many pictures of them.
It is "MPPT" in brand name only, but PWM in function.

Here's a selection of cheap and of quality MPPT charge controllers:


Quality MPPT charge controllers, some quality PWM, some ("DIY" series) that may not be quality (suspicious of re-branded rather than developed by Midnight)

 
That's a well-known style, many pictures of them.
It is "MPPT" in brand name only, but PWM in function.

Here's a selection of cheap and of quality MPPT charge controllers:


Quality MPPT charge controllers, some quality PWM, some ("DIY" series) that may not be quality (suspicious of re-branded rather than developed by Midnight)

Not MPPT function ... my fault for not checking. Thanks for the links, will look into.
 
Head of Midnight just posted, and says the "DIY" line really are different from the original design. So you might want look into them if you are price-conscious.

 
LiFePo4 bank seems ideal for me, I'm going to set up a decent system & use it.
 
Last edited:
Back
Top