My first project - Solar power computer....I need your help

Cariboocoot wrote: »

Welcome to the forum.

Before anyone goes crazy let me point out that you are in Cyprus where the power standard is 240 VAC 50 Hz. I imagine you use the European standard 230 VAC equipment though (they have quite a broad tolerance range for that 'standard').

Well you've missed a couple of finer points on the sizing, so let's check it out.
It would be good if you could measure the computer's actual power usage. I suspect the 115 Watt hours is based on the power supply specs saying "115 Watts". How much it really uses is probably lower, possibly higher; computers use varying amounts of power depending on what they are doing at the time.

But let's go with 115 Watt hours. Now that's AC and inverters aren't 100% efficient so on the DC side the power consumption will be higher. Based on 90% efficiency it will be 128 Watt hours. Then the inverter itself has some draw, depending on what size it is. A small one like the Morningstar 300 will run the computer and use only 6 Watts to do it. Now your total power consumption from the battery is 132 Watt hours.

Divide by nominal 12 Volts and you get 11 Amp hours used. It is usually best to go with 25% DOD so that's 44 Amp hours. Batteries don't come in every size you want so you have to adjust that to what's available, preferably rounding up. Around here that would be a 50 Amp hour 12 Volt AGM.

Now for recharging it. If you went with 50% DOD you'd need twice as much panel to recharge with because of having to bring up the battery from a lower SOC in the same amount of time requiring a higher charge rate. So let's stick with the 25% DOD 50 Amp hour battery.

You want a peak current around 10% of capacity, or 5 Amps. Very easy to achieve. One panel will do it, and it doesn't have to be very big; about 88 Watts @ 17.5 Vmp. Once again you can't get panels in any size you want, so you either have to settle for a little less charge rate (85 Watt panel 4.8 Amps) or a little more panel (130 Watt panel 7.3 Amps). Around here the 130 Watt is actually cheaper, and the AGM could take the higher current without any problem. Of course you'll need a 10 Amp PWM type charge controller.

I can give you an equipment list for this set-up but that may not do you any good as it really depends on what's available where you are.

BB. wrote: »

Welcome to the forum Kellez,

Your calculations and numbers are correct in the "ideal" world. In the real world, there are some issues.

First, you want to discharge the battery at a C/2 hour rate (fully discharging the battery to "dead" in 2 hours)--Flooded cell batteries really only like to be discharged at a C/8 hour rate, or at most, a C/5 discharge rate.

Also, the faster you discharge a battery, the less apparent capacity (a battery discharged at C/20 rate may appear to have a 100 AH capacity, a lead acid deep cycle storage battery discharged at C/2 rate may have an apparent capacity of 60 AH or less).

So--You have two choices, either get a larger battery to drop the discharge rate to C/8 - C/20, or change the battery type. AGM type Lead Acid batteries can be discharged at a much faster rate... Or looking at the various types of Lithium Ion chemistry type batteries (lots of choices, and you get into the whole "Battery Management System" electronics--Some/most Li-Ion battery types need a per cell voltage monitoring/charging adjustment circuit to prevent over/undercharging and damage/battery failure).

And there are two calculations we make when sizing a solar array to battery bank. The first is that there is enough charging current to properly recharge the battery bank based on AH capacity (5% to 13% rate of charge typical). And second is based on the Watt*Hours (or Amp*Hours) you use per day and the amount of sun you get.

The fact you used the "worse" average hours of "noon time equivalent" sun per day is fairly conservative--This is "average" weather--Those values can be +/- 10% over long term average, and day to day, you can have a few days or even a week+ of bad weather where the solar energy is only 5% of system capacity (dark days). So, you either have to be able to turn off the loads during poor weather or have a backup power source (typically a small AC genset).

There are options out there--But you need to pick the correct battery/system to match your needs.

Next, there are other losses--AC inverter conversion losses, battery charging/discharging losses, solar panels+charge controller derating+controller losses, etc... For a simple lead acid solar power system, roughly the end to end efficiency (solar panel to charge controller to battery to ac inverter) works out to be around 52%--Yep, you get only 1/2 of the "rated" capacity of your system.

So, what would a flooded cell lead acid battery system look like? Roughly, the calculations would look like:

• 115 Watt* 1 hour per day = 115 Watt*Hours per day
• 115 WH * 1/0.85 AC inverter eff * 1/12 volts * 2 days of storage * 1/0.50 max discharge = 45 AH @ 12 volt battery bank (long life, good battery capacity for load balance)

Next, double check discharge rate... Use C/8 for maximum continuous discharge (flooded cell lead acid):

• 115 Watts * 1/0.85 inverter eff * 8 hour discharge rate * 1/12 volt battery bank = 90 AH @ 12 volt minimum based on loads

So--Based on common rules of thumbs and battery characteristics (note: we use C₂₀ battery AH capacity numbers for our rules of thumbs/typical off grid system design), a Flooded Cell Lead Acid battery would need to be almost 5x larger than your initial calculations. Is this "the answer"--No just a conservative design with one type of battery chemistry. If you choose AGM or Li-Ion (or other), the calculations will be different (other batteries may cost more, be harder to get, more fragile or less fragile, need a BMS per cell, etc.).

To continue with the calculations using Flooded Cell... First need to size the solar array to battery bank AH capacity. Nominally, we recommend 5% minimum rate of charge (for good battery health) and "low usage system" (emergency backup, use system at weekend/seasonal cabin) to 13% rate of charge (full time off grid, used daily, maximum "cost effective" solar array for typical applications).

• 90 AH * 14.5 volts charging * 1/0.77 panel+controller deratings * 0.05 rate of charge = 85 Watt array minimum
• 90 AH * 14.5 volts charging * 1/0.77 panel+controller deratings * 0.10 rate of charge = 169 Watt array nominal
• 90 AH * 14.5 volts charging * 1/0.77 panel+controller deratings * 0.13 rate of charge = 220 Watt array "cost effective" maximum

And based on hours of sun per day and your loads:

• 115 WH per day * 1/0.52 system efficiency * 1/2.78 Hours of sun per day (winter minimum average) = 79 Watt array minimum

So, for this fictional system design, the size of the battery bank is setting the array capacity (large battery bank, needs a large solar array to meet minimum charging requirements). For a daily use system, I would suggest around 169 Watt array for this system. Of course, this is way over-sized for your loads and would give you a minimum nominal "useful AC output" of:

• 169 Watt array * 0.52 system eff * 2.78 hours of sun winter day = 244 WH per day

Next, the solar panel. You looked at a 28 cell solar panel. The typical Vmp (voltage maximum power) rating would be around 14.0 volts for a 28 cell panel (~0.5 volts per cell). To charge a flooded cell battery you need ~14.5-15.5 volts minimum. And when running a solar panel under full sun on warm to hot days, the cells Vmp is depressed (by as much as 20%)... When all the losses are taken into account, you need a Vmp (standard test conditions) of ~17.5 volts minimum. Or a 36 cell panel for charging a 12 volt battery system (minimum/nominal for a PWM type charge controller).

There are ~28 cell panels sometimes out there--But this is from the "olden days" where people wanted "self regulating" panels to charge batteries without a battery charger. Today, we always use battery charge controllers as this is more efficient and will give you better battery life/more energy (overall) from your system.

Next, the "other losses". AC inverters, even small ones, take something like 6 watts just to "turn on". If you had an AC inverter to supply your 230 VAC power that you left running 24 hours per day (did not turn off the DC power to your inverter when the AC was not required)--You would have to add:

• 6 watts * 24 hours = 144 Watt*Hours per day to your load calculation

Or pretty much double your system capacity. There are AC inverters with "search mode" and remote on/off inputs. A very nice/small AC inverter that has these features is the MorningStar 300 watt 12 volt TSW (true sine wave) AC inverter--Available in both 120 Volt 60 Hz and 240 Volt 50 Hz versions. Highly recommended for your application.

Anyway--Those are the basics. Not to say my design is "perfect" and the only answer--But it is a starting point for a reliable system that should give you many years of use with minimum maintenance/surprises.

-Bill

Cariboocoot wrote: »

You're building your own panel? Okay there's a few problems with that. One of them being that the cells you buy on the Internet tend to be one rejected by companies for not meeting specs, including shunt resistance.

Figure 0.5 Volts per cell under load; the 0.57 reading you got would be Voc which doesn't reflect actual power output. Therefor a 32 cell panel would be 16 Volts, which can be too low for a 12 Volt system. When the panel heats up (and this does not mean 'is used in desert temperatures') the Voltage goes down. More Voltage will be lost in the wiring. So with 16 Vmp by the time you measure Voltage at the battery under load the panel may not produce enough to fully charge the battery; you can easily be down 3 Volts at that point.

This is why standard "12 Volt" panels have 36 cells and a Vmp around 18. Notice that 36 cells times 3.9 Watts each comes to 140 Watts; pretty much a typical "12 Volt" panel with an Imp of 7.8.

You have to have both sufficient current and Voltage or the battery will not charge properly. Since PV is a current source it isn't difficult to come up with that; any one of those cells would produce the 7.8 Amps (3.9 Watts / 0.5 Volts) but if you don't have enough of them in series to make the proper charging Voltage under operating conditions it doesn't work.

BB. wrote: »

Yep, you are OK to post links when asking/answering questions... We try to avoid xyz.com is having a 1/2 off sale today type posts and spam (boy, there are lots of spammers out there).

Here is a similar looking controller:

Wellsee MPPT

Don't know if it is the same internals or not.

And down past the middle in this thread, one person apparently has your same controller:

Tri star charge controllers

In the end, just accept that you are experimenting and that some of the experiments will fail.

Charge controllers (and power electronics in general)--To be long lasting need good components & design, as well as good sized heat sinks/forced air cooling. High temperatures and thermal cycling are the death of electronics.

As the systems become larger and more mission critical (i.e., cabin/home in the wilderness, emergency backup power after an ice storm) and hours from any stores/suppliers, you will probably start looking for more reliable (and more expensive) components. At some point, saving a few hundred dollars is not worth the time, effort, and expenses to constantly repair the system/run a genset for backup power.

-Bill