Average Expected Generation

[FB.]

I decided to look in more detail at my PV system's generation stats. It has only been operational for a few months, so hasn't seen a full year.

60% of days are below target for that day.
40% of days are above target for that day.
The system total generation is comfortably outperforming SAP.

But those days which are above target (when the sun shines), tend to be spectacularly above target and appear to be pulling up the averages; just as one millionnaire in a workplace would pull up the average wealth of the group; yet most would not be "average".

My data suggest that the data used for SAP calculations, when divided into daily stats (and adjusted for day length and sun position, so that, for example, the end of February should generate more than the beginning) probably quite closely estimate the median generation, but not the mean generation, which is pulled dramatically higher by a few really sunny days.

Does anyone else's data confirm this?
Any thoughts?

 

[TedM]

I would suggest you don't get too hung up on SAP figures. Their calculation is pretty crude and if you are located anywhere south of Sheffield then you should expect to outperform them comfortably.

The next version of SAP (to be called '2012' but not coming into force until 2013 and then probably another couple of years before it filters down to the PV world) will go some way to addressing this by applying different base insolation values by geographic region.

 

[FB.]

Hi Ted.

It's not so much the total generation, but the individual day-to-day variation.

A typical run of five days might be something like: 2kW, 2kW, 2kW, 4kW, 10kW.
Average of that would be 4kW per day, but actually most days are below the average; and the mid-point if we rank from high-to-low is 2kW.
The average is very much influenced by the occasional really high reading when there's direct sun on the panels.

 

[wstevew]

This month my target is 6kWh per day

Currently

Cloud and rain(worst) = 2.55kWh
Clear and Sunny (best) 9kWh

Falling behind at the moment. But as the month passes these will increase.

No sunny days = below target (not happened yet).

Feb was 150% of monthly target.

 

[pwboston]

extra 50% thats a big difference, mine was above target around 10% in Lincolnshire.

 

[FB.]

wstevew

"....Falling behind at the moment. But as the month passes these will increase...."

I have a spreadsheet which gives values for each day of each month; when the day length is in the accelerating phase in springtime - or in the rapidly-shortening days of autumn - there is a big difference between the start and end of each month.
Rather than step-ups, it gives a smooth trendline for the generation target.

So my "trend" for the target generation figure would be:
01 March: 7.1kWh
08 March: 7.9kWh
16 March: 8.8kWh
24 March: 9.9kWh
31 March: 10.9kWh

Whole-month target: 275kWh.
Average per day: 8.9kWh.

 

[jerriais]

I would consider that (meaning FBs figures) fairly normal. Our UK weather patterns mean that our weather is usually governed by either a high or low pressure system. A normal pattern for Feb would be roughly 25% high pressure and 75% low pressure days. It's rare to get a day with a bit of both so a typical generation pattern will be lots of cloudy days with low generation and a few bright sunny days that will bring the average up to around the predicted level.

 

[wstevew]

extra 50% thats a big difference, mine was above target around 10% in Lincolnshire.

Feb target 94. Actual was 146 I think.

March target is 186.


December +20%
JANUARY +30%

 

[liquidity]

If you haven't already done so enter your data on pvoutput.org - this will allow you to compare results with people near you which should either put your fears to rest or give you reason to investigate further.

 

[FB.]

Feb target 94. Actual was 146 I think.

March target is 186.


December +20%
JANUARY +30%

Hi Steve

What is the system size (panels and inverter), what is the roof pitch, which direction does it face and what source of data are you using to set the targets?

My system clocked-up 182kWh in February, but that needs to be put into the context of how big and which way it faces.

I also think that there are a lot of local micro-climates that will be seen to make a mockery of the SAP figures or those from other solar data sources.
The following link to the Met Office interactive maps of historical weather averages shows that there is a lot of variation, even within a short drive of a particular location.

Link here > Met Office Mapped Averages <

.

 

[pwboston]

I did 202kw in February, my "prediction" by the European site was 184kw

4kw system south facing, 30 deg slope in Lincs.

 

[wstevew]

3.84kWp.90 degree East/West. 50/50. 22.5 degree pitch in Essex. Little to no shading

2850 annual sap figures

Monthly targets taken from a sap calc Phone app. Phone app monthly figures add up to my annual target . Also very close to a web link I have seen on here that has 2 different irradiation databases.