Total System Effieciency

[Jaco De Jongh]

I know that this can be a lengthy discussion with a lot of technicality's, but i will try and keep it short. My main aim of this post is to find out if some of the Forum members has done these tests before and if the results I got is in line with the industry norm.

First of, my pocket calculator isn't  halve as big or clever as @plonkster machine, so this is figures I see combined with assumptions I make.

Today is the first hot day with clear skies since I started using AICC, and being at home today i decided I am going to push the system a bit and manage the load manually to try and stay as close as possible to full capacity without sacrificing battery power.

Some of the figures as read from the Inverters: Lets assume they are accurate (witch I seriously doubt)

• With 18 x 260Watt panels (4680 Watts) the max power I could draw was 4000watts, and I could maintain the production between 3900 and 4000 watts for the best part of 3.5 hours. I guess Solar irradiance for my area, temp, and connection losses might explain the difference. Does this look in line with your figures or is 680 watt difference from rating to actual a reason to worry?
• If i take a snapshot of the Battery Watts, Load watts and PV watts, witch I did a couple of times, I could see that there is conversion losses of 300-400Watts at 3500-4000 watt PV production. That is in the region 10% conversion losses. So in theory my two Inverters in parallel is only 90% efficient in converting PV to where it is needed. Does that sound right? 

I know there might be a lot of reasons for losses, I just need to know if this kind of figures are in line whit what you have experienced.

[Mark]

That sound very good. 90% for this time of year is very good.

Sent from my SM-G800F using Tapatalk

[___]

7 minutes ago, Jaco de Jongh said:

only 90% efficient in converting PV

The efficiency normally quoted for inverters (around 96%) is best case scenario, usually at 90% of its rated load or some value that is rarely used (hence the whole discussion of not oversizing an inverter too much)...

So it sounds about right to me.

[Guest]

+1 - sounds about right.

But from what I have learned over the years,the overall system efficiency can be as low as 75%, 85% at best.

From panels to AC out other side with batts in the mix.

[Deonsr]

Nico you are lucky - 85% efficiency with only 15% losses. There are at least 6 types of losses in every PV installation:

1. Manufacturer’s power tolerance.  (1%) All panels have a power tollerance. Normally this is -1%/+3%. So worst case is that the specified power output is actually 1% lower than spec.

2. Temperature Loss. (10%) Solar panels don’t like to be hot. Most solar panels lose about 10% of their rated power on a 25°C day, more if it is hotter. Let’s assume 10% for this estimate. (1% for every 2 °C the panel temp is hotter than STC 25°C) On a 25°C ambient temp, the temp at the panel in full sun could easily be 40-50 °C)

3. Dirt (5%) Soon after installation dust will settle on the panels’ glass. These  block the amount of sunlight reaching the solar cells behind the glass reducing your power. The reduction in power from dust build up typically lies in the 5%-15% range.

4. Wiring Losses (voltage drops) (2%) All the solar panels on your roof are interconnected with wires, then a long pair of DC wires connects the final solar panel to your inverter. All these wires have a small electrical resistance, which means the electricity flowing through them will suffer a voltage drop. This will reduce your power proportionally, typically by around 2%. (keep your wiring runs as short as possible)

5. Inverter Efficiency (4%) Everything goes through your inverter so the inverter efficiency will directly affect your system output. Depending on your inverter you can expect an inverter efficiency of about 96%, giving a 4% loss.

6. Converting

   (3%)
   This is a difficult one to determine I like to think there is 1% loss from converting DC from the panels through the inverter. Charging batteries have another 1% loss, discharging from the batteries another 1%. The list actually just goes on and on. Even connectors not soldered or crimped correctly could be the reason for massive losses but that is a story for another day.

   Multiply all those together and you are looking at 25% total losses which is not uncommon but reality. Giving a real world peak power of 4000w from a 4680w array is a loss of 680w - or 14.53%....you are very lucky.

[Jaco De Jongh]

@Deonsr, I have read that the panel production will decrease quite a lot over the first year and then quite linear thereafter, this is still a fairly new installation, maybe the figures will look a bit different next year.

@The Terrible Triplett, I will take that 75-85% any day over the figures I saw from Eskom this week. Our Lead energy advisor on the mine showed me a document earlier this week that shows, generating electricity from coal is only 25% efficient.

EDIT: I can and do adjust my Panels every month to face the sun better, I think that might help me to achieve that 4000watt.  

[Jaco De Jongh]

I wiped snagit when i wiped my pc, did not look for software or a way to make a screen shot in Linux yet. 

Today I manually manipulated my system to see what i can get out of it, I missed it sometimes and wasn't prepared with the right size loads to keep it on maximum, so at times the system under performed.

Made a total of 28.7Kw with an efficiency of 6.13. 

Now I can start managing my system accordingly.

[Chris Hobson]

For me a good indication that wiring extra is optimal is to see whether your system can go over 100%. Need a cool day with good sunshine and occasional clouds. My reasoning being that if your losses due to orientation, cabling etc is too great one will never hit the magic 100%.  On a ordinary cloudless day I hit between 80 and 90%.  With largely cloudless skies we only go over 100%  ½ dozen times a month. 

[Guest]

3 hours ago, Jaco de Jongh said:

I wiped snagit ...

Windows has a Snipping Tool app since Win7. Not as cool as Snagit, but it does the job for free.

[___]

4 hours ago, Jaco de Jongh said:

software or a way to make a screen shot in Linux yet. 

The package name is "shutter". Really nice tool to grab random areas of the screen. This is an area where Linux was -- frankly -- a little lacking until recently. I always made my screenshots with gimp... which was cumbersome :-)

[Jaco De Jongh]

8 hours ago, plonkster said:

"shutter". Really nice tool t

Thanks sir, quick install and really easy to use. I like it

[Chris Hobson]

13 hours ago, Jaco de Jongh said:

I wiped snagit when i wiped my pc, did not look for software or a way to make a screen shot in Linux yet. 

Today I manually manipulated my system to see what i can get out of it, I missed it sometimes and wasn't prepared with the right size loads to keep it on maximum, so at times the system under performed.

Made a total of 28.7Kw with an efficiency of 6.13. 

Now I can start managing my system accordingly.

That is a good curve for an inverter that is not a GTI. I have noticed being in a valley my curve does not increase gradually. It ramps up very quickly climbing by 1000 to 1500w in ten minutes or so.