diy solar

diy solar

Portable off-grid design - please break it before I buy it

GlowInTheDark

New Member
Joined
Oct 15, 2021
Messages
6
Hi,
I’m planning a portable power system as my first solar/off-grid project and I’d appreciate any veteran input to show me my mistakes before I buy them :)

I’m hoping to start buying components in the next two days, so if anyone can give quick feedback I would be very grateful.

Components are all relatively lightweight because the system needs to be easily relocatable.

Circuit diagram.png

Numbered items on the circuit diagram:

1. Household appliances​

a. 3.5KWh of 230V AC required daily
b. 2000W maximum continuous load required (spikes to 2500W)

2. Batteries × 4 (specs for each battery is as follows)​

a. LiFePO4 24V 50Ah in parallel
b. 25.6V Nominal Voltage
c. 1280 WH of Energy
d. 100A Max Continuous Discharge Current
e. 100A Max Continuous Charge Current
f. Cycle Life 3500 (80% DOD)
g. ≤80mΩ Internal Impedance
h. 29.2V Charge Voltage
i. 20.0V Discharge Cut-off Voltage
j. Charge Temperature = 0°C to 60 °C
k. Discharge Temperature = 20°C to 60 °C
l. https://roarpower.nz/off-grid-solar-products-nz.html#battery2450

3. Inverter/charger/solar controller​

a. Victron EasySolar 24V/3000VA/70-50 MPPT150/70
b. 3000VA / 2400W Continuous output power at 25°C
c. Charger
i. 28.8V Charge voltage 'absorption'​
ii. 27.6V Charge voltage 'float'​
iii. 26.4V Storage mode​
d. MPPT
i. 70A Maximum output current​
ii. 2000W Maximum PV power​
iii. 150V Maximum PV open circuit voltage​
e. https://www.victronenergy.com/upload/documents/Datasheet-EasySolar-with-Color-Control-EN.pdf

4. PV panels​

a. Model = DTSB200 (Solar blankets - fold up to briefcase size)
b. Will prop these up using collapsible tent poles
c. Monocrystalline
d. Rated maximum power (PM): 200W
e. Voltage at Pmax VMP: 18.70V
f. Current at Pmax IMP: 10.86A
g. Open-circuit voltage (VOC): 22.51V
h. Short-circuit current (ISC): 11.53A
i. Tolerance ±3%
j. Normal operating cell temp (NOCT): 47 ±2°C
k. https://www.steeringandsuspensionwa...etech-4x4-200w-foldable-solar-blanket-dtsb200

5. Generator (backup for cloudy days)​

a. GT2100ESi
b. 1700W continuous power output
c. https://www.gtpower.co.nz/shop/Inverter+Generators/GT2100ESi.html

6. National electrical grid​

a. If it’s available
b. 230V

7. Wiring for PV panels​

a. 4mm² (12 AWG)
b. If not using the extension cable (item #8), then panels will be 10 meters from the Solar Controller
c. Will want the ability to reconfigure the panel wiring so that they’re all in series (e.g. on cloudy days), so am expecting to have to use MC4 connectors.

8. 10m extension cable​

a. 6mm² (10 AWG)
b. Only used if panels need to be far from inverter
c. Y-connectors will be used to combine any parallel strings into this one cable


As mentioned - I'm new to all this, so am keen to hear of anything I haven't thought of.

Some questions that come to mind:​

  • What happens when my battery is full and the panels are still in the sun? Where does the electricity from the panels go? Does it put strain on any part of the system?
  • I want to ensure the batteries’ State of Charge (SOC) remains synchronised, so I expect to need one or more battery balancer. How many should I use, and where should they be wired? I can’t find an example online for wiring battery balancers into a bank of parallel batteries. https://www.victronenergy.com/batteries/battery-balancer (I've heard a "shunt" may be helpful in this circuit but I haven't had time to learn what that is yet)
  • Is the impedance of these batteries normal? Other similar batteries have an impedance of ≤20mΩ

Many thanks in advance!
GlowInTheDark
 
Your batteries are in parallel, and I think you've properly matched wire lengths.
They are 24V each, no internal nodes accessible.
Can't see any place to connect a balancer.
All will be at same voltage, and each has a BMS inside to balance its cells.

Looks like 3s2p PV is about 70Voc, rises maybe 10% to 20% in freezing weather, still well below 150V limit of SCC.
The power they produce leaks back through their own PV diodes. No harm.

"
Internal Impedance≤80mΩ
"

Normal for the cells people use in DIY is 0.25 milliohm, so 8 in series would be 4 milliohm. 80 sounds unusually high.

Add a battery fuse suitable for short-circuit current (or one per battery)
30V/0.08 ohm = 375A per battery, or 1500A for all four in parallel.
Odd, I expect 20,000A from a lithium battery with 0.25 typical, 0.17 milliohm measured IR per cell.
How could it deliver 100A having 0.08 ohms internal resistance? That would be 8V drop, to about 20V give or take.

Better ask them to confirm that IR spec.
 
All the essentials are there except for a capacitator bank on the DC side!
 
All the essentials are there except for a capacitator bank on the DC side!
It’s a good block diagram, but there’s still a bit of details missing like fusing for batteries whether individual or One fuse, and wire sizes. Also, Master off / quick disconnect for the PV and battery side. Devil in the detail stuff. Especially figuring things like where the six inch long class T fuse block will go.

Seems reasonable 1200 watts of panels with 4800 wh of batteries and a 3.5 kwh requirement. This seems like enough energy to get you through one day, and if the next day is cloudy at some point early in the day the generator kicks on or you use grid power.

A 3.5 kwh power requirement for a day is not very much, just a couple hundred more than my RV build. I’ve found for my RV build that requirement varies widely, but I have to have enough power to get me through the coldest, longest night. The same amount of power would normally last me a couple of weeks On other trips.
 
Your batteries are in parallel, and I think you've properly matched wire lengths.
They are 24V each, no internal nodes accessible.
Can't see any place to connect a balancer.
All will be at same voltage, and each has a BMS inside to balance its cells.

Looks like 3s2p PV is about 70Voc, rises maybe 10% to 20% in freezing weather, still well below 150V limit of SCC.
The power they produce leaks back through their own PV diodes. No harm.

"
Internal Impedance≤80mΩ
"

Normal for the cells people use in DIY is 0.25 milliohm, so 8 in series would be 4 milliohm. 80 sounds unusually high.

Add a battery fuse suitable for short-circuit current (or one per battery)
30V/0.08 ohm = 375A per battery, or 1500A for all four in parallel.
Odd, I expect 20,000A from a lithium battery with 0.25 typical, 0.17 milliohm measured IR per cell.
How could it deliver 100A having 0.08 ohms internal resistance? That would be 8V drop, to about 20V give or take.

Better ask them to confirm that IR spec.
Thanks for your quick reply Hedges.

Great to get your input and confirmation about a lot of factors.

You make a good point about the internal resistance. I had noticed that other LiFePO4 batteries generally seem to have an internal resistance of ≤20mΩ but I didn't realise the impact of an 80mΩ IR so I will go back to the supplier on this. Thank you!

After reading your message I've been using an online calculator to get familiar with internal resistance. In case it's useful to anyone else: https://www.translatorscafe.com/unit-converter/he-il/calculator/batt-internal-resistance/


Regarding battery balancers - I thought I would still need one between the batteries somehow, to ensure they stay in sync. Is it not possible to do this when they are in parallel?
 
All the essentials are there except for a capacitator bank on the DC side!
Hi Sunshine, thanks for your speedy response.
I've never heard of adding a capacitor bank. What benefit would this offer? Or what problem does a capacitor bank solve?

Can you give some specifics about what type of capacitors I should use?
And when you say they should go "on the DC side", do you mean between the inverter/MPPT and the PV panels, or between the inverter/MPPT and the batteries?
 
"
Internal Impedance≤80mΩ
"

Normal for the cells people use in DIY is 0.25 milliohm, so 8 in series would be 4 milliohm. 80 sounds unusually high.

Add a battery fuse suitable for short-circuit current (or one per battery)
30V/0.08 ohm = 375A per battery, or 1500A for all four in parallel.
Odd, I expect 20,000A from a lithium battery with 0.25 typical, 0.17 milliohm measured IR per cell.
How could it deliver 100A having 0.08 ohms internal resistance? That would be 8V drop, to about 20V give or take.

Better ask them to confirm that IR spec.
IR sounds about factor of 5 to 10 too high.
300Ah cells have about 0.25mOhm IR
For 50Ah Cell expected IR would be ballpark of 0.25mOhm*(300/50) = 1,5mOhm per cell and 12mOhm per 8S.
 
Capacitor bank would be to supply higher surge current, smooth out ripple. Car stereo guys like to use those.
My measurements with a scope indicate most to all of the 60 Hz current draw to feed inverter comes from battery, because no practical capacitor would be large enough.

Here's something you might add: Precharge resistor.
When you first connect lithium batteries to inverter (either close switch/breaker or just touch cable), voltage on inverter's capacitors jumps to battery voltage as close to instantly as possible. That burns the thing that made contact and stresses the capacitor.
What people do is connect a resistor (or light bulb, or very long thin wire like 50' of telephone wire) momentarily, to gradually charge the capacitor. Then make solid connection.

Lots of equipment was designed and used with lead-acid batteries, which have 5x higher resistance so 5x lower current charging capacitors as compared to lithium batteries. (your lithium batteries, according to their documentation, is 20x higher resistance than typical lead-acid)
 
When you relocate the system, do you plan to disconnect the batteries? If "no", I would suggest you get one 24v battery of the proper size.

Battery current and internal resistance go hand in hand, as internal resistance causes a voltage drop (across that resistance0 under load, so the load is limited by the battery capacity and C rating. Don't try to compare internal resistance unless the battery capacity and C rating are the same. Then it might make some sense. (edit) If you fully charge the batteries before you connect them in series, that is about the best you can do to "keep them in sync. Of you worry, disconnect them from time to time and fully charge each one.
 
It’s a good block diagram, but there’s still a bit of details missing like fusing for batteries whether individual or One fuse, and wire sizes. Also, Master off / quick disconnect for the PV and battery side. Devil in the detail stuff. Especially figuring things like where the six inch long class T fuse block will go.

Seems reasonable 1200 watts of panels with 4800 wh of batteries and a 3.5 kwh requirement. This seems like enough energy to get you through one day, and if the next day is cloudy at some point early in the day the generator kicks on or you use grid power.

A 3.5 kwh power requirement for a day is not very much, just a couple hundred more than my RV build. I’ve found for my RV build that requirement varies widely, but I have to have enough power to get me through the coldest, longest night. The same amount of power would normally last me a couple of weeks On other trips.
Hi chrisski, thanks for getting back to me so quickly.

Yes I agree the devil is in the detail. I'm keen to iron out the details before I buy components so I appreciate your help.

A few follow-up questions:
  1. Regarding the wire-sizes between each battery and between the battery bank and the inverter: with a 2400W inverter and 24V batteries the maximum current would be 100A so it seems best to use a cable thickness of 13mm² (6 AWG). Does that seem right?
  2. Regarding fuses: yes this is a good point that I'll need some fuses. A few questions:
    1. Is it best to have them between each battery, or just one for the whole battery bank?
    2. From Hedges message, it looks like I'll need to get confirmation about the batteries' internal resistance before the size of the fuses can be determined, is that right?
    3. Do I need fuses between any other components?
  3. Regarding the master off / quick disconnect.
    1. Are these two names for the same component, or is "Master Off" something different to "Quick Disconnect"?
    2. Do I need two quick disconnect devices? (one for the battery and one for the PV array)
  4. Regarding the class T fuse block:
    1. I've never heard of this type of fuse. Why is it important to have this one specifically?
    2. Which components should it go between?
    3. Should it be on the positive wire? (I recall somewhere that fuses should always go on the positive wire, but I'm not sure if that's true)

Yes 3.5kWh is pretty small. This is because it doesn't cover any heating.
 
Can you give some specifics about what type of capacitors I should use?
24v version of these maxwell S Caps

Copy & Paste from a Australian Super cap company-

Features:  High power output to support peak current loads  On-board energy storage to handle power surges (high capacitance and energy density)  Long cycle life Applications:  Energy Harvesting for wireless sensors  Peak power support for GSM/GPRS transmission  Last gasp power for remote meter status transmission  Peak power support for locks & actuators  Peak power support for portable drug delivery systems  Short term bridging power to ride through power Interruptions or for battery hot swap

At the moment Super Caps are seldom included in initial DIY solar plans... PV--CC--Bat--Inv.
It is possible that as Super Cap advantages become more well known the default initial plan would be this....PV--CC--Bat--Super Cap--Inv and it is interesting to see other responses-
My measurements with a scope indicate most to all of the 60 Hz current draw to feed inverter comes from battery, because no practical capacitor would be large enough.
How about one of these-
48V 165F Maxwell Super Capacitor 48V 165F Maxwell Super Capacitor 48V 165F MAXWELL SUPER CAPACITOR

Edit- found a super cap thread-
 
Last edited:
Capacitor bank would be to supply higher surge current, smooth out ripple. Car stereo guys like to use those.
My measurements with a scope indicate most to all of the 60 Hz current draw to feed inverter comes from battery, because no practical capacitor would be large enough.

Here's something you might add: Precharge resistor.
When you first connect lithium batteries to inverter (either close switch/breaker or just touch cable), voltage on inverter's capacitors jumps to battery voltage as close to instantly as possible. That burns the thing that made contact and stresses the capacitor.
What people do is connect a resistor (or light bulb, or very long thin wire like 50' of telephone wire) momentarily, to gradually charge the capacitor. Then make solid connection.

Lots of equipment was designed and used with lead-acid batteries, which have 5x higher resistance so 5x lower current charging capacitors as compared to lithium batteries. (your lithium batteries, according to their documentation, is 20x higher resistance than typical lead-acid)
Thanks Hedges, that makes sense about the capacitor bank - so if I understand correctly, it helps protect the batteries by smoothing out the load that they see from the inverter/charger.
How does that affect things when the inverter/charger is charging the batteries? I.e. I think the inverter/charger monitors the batteries' voltage to determine their state of charge (SOC) and to decide when to stop charging or when to switch from constant current (CC) to constant voltage (CV). By putting a capacitor between the charger and batteries, could this interfere with the charger's ability to monitor the batteries?

Good point about the Precharge resister. I'd heard of that idea but I'm not sure about the details. Can you give an example of what type of resistor I would need? Hopefully there's something that will minimise my chances of giving myself an electric shock :)
 
  1. Regarding the wire-sizes between each battery and between the battery bank and the inverter: with a 2400W inverter and 24V batteries the maximum current would be 100A so it seems best to use a cable thickness of 13mm² (6 AWG). Does that seem right?
I use these to answer how thick a wire would be. Honestly though, Inverter runs are more of don't exceed ampacity rather than voltage loss. I believe if you check the ampacity chart, 6 AWG is a no go:


Voltage loss is here:


Regarding fuses: yes this is a good point that I'll need some fuses. A few questions:
  1. Is it best to have them between each battery, or just one for the whole battery bank?
  2. From Hedges message, it looks like I'll need to get confirmation about the batteries' internal resistance before the size of the fuses can be determined, is that right?
  3. Do I need fuses between any other components?

Try this:


Regarding the master off / quick disconnect.
  1. Are these two names for the same component, or is "Master Off" something different to "Quick Disconnect"?
  2. Do I need two quick disconnect devices? (one for the battery and one for the PV array)
My Master off switch:


I have that switch to disconnect the battery, and Midnite SOlar DC Circuit breakers for the panels.

A quick disconnect can be a lever visible for firefighters to remove power. I don't have one. May be necessary for NEC code.
 
When you relocate the system, do you plan to disconnect the batteries? If "no", I would suggest you get one 24v battery of the proper size.

Battery current and internal resistance go hand in hand, as internal resistance causes a voltage drop (across that resistance0 under load, so the load is limited by the battery capacity and C rating. Don't try to compare internal resistance unless the battery capacity and C rating are the same. Then it might make some sense. (edit) If you fully charge the batteries before you connect them in series, that is about the best you can do to "keep them in sync. Of you worry, disconnect them from time to time and fully charge each one.
Thanks DThames, I think I understand that. It's taking a while to get the spec-sheet for these batteries, I'll have a look at the C rating when I get that.

I do plan to disconnect the batteries when relocating. Even so, I would much prefer two 24V 100Ah batteries, however there are significant limitations on supply at the moment so I may have to go with the 24V 50Ah ones. FYI they will be in parallel (not series) :)
 
6 awg with 90 degree insulation good for 105 Ah if single conductors in free air, up to 30 degrees C.


2400W/20V/90% x 1.12 x 1.25 = 187A fuse --> 200A fuse


20V low voltage and IR drop (pick another number if too low for your lithium batteries, might be common for lead-acid)
90% efficiency of inverter at 2400W (may be even lower)
1.12 multiple because 60 Hz ripple drawn from battery is 12% higher RMS (causes heating) than mean (delivers power)
1.25 multiple is standard oversize to avoid nuisance blowing.

In that case, 1 awg is sufficient, 2/0 cable is great (300A, more than sufficient).
Adjust he math, pick an available fuse, cable with sufficient ampacity.

Typical LiFePO4 cell I've seen specs for was 0.25 milliohm, measured 0.17 milliohm.
3.5V/0.00017ohms = 20,588A

With such figures, each battery in parallel should have its own class-T fuse, because they are rated to interrupt 20kA.
Each one 50A rating, four in parallel 200A to protect cable?
But depending on your battery's actual internal resistance and short circuit capability, maybe you can parallel four batteries into one fuse.
There are also other fuses rated 50kA, from ferraz-shawmut. Also breakers DC rated 50kA from Midnight.
 
Back
Top