Slow Progress

So it looks like the monthly target of 170% of predicted output is going to be a stretch. 5 of the last 6 days have been cloudy and produced little more than 1 kWh per day. At this rate we're going to struggle to get to 160%. 

I shouldn't moan really, when we have a good day the output is excellent, but in times of prolonged cloud (as now and forecast for the next few days) it's less impressive. Still, the sun's angle is improving, the days are slowly starting to get longer and I'm still making some minor inroads into out electricity consumption - roll on spring and some decent weather.

It'll be interesting to see the hours of sun stats for the month when they're published in January - I wonder how well we've done when compared with the seasonal average. Time will tell.

Shortest Day

So I was having breakfast with a good friend of mine near Heathrow, watching the sun rise at about 8am and it looked like it would be a good day today. It was. 

Despite it being the shortest day of the year we still generated nearly 8.5 kWh today, the system seemed to top out at 2200 watts since there was a little wispy cloud around that never really went, but at this time of year you cant complain when good weather shows up. So that makes 112kWh for the month so far, and 146% of the expected yield for the whole of December, with 9 days still to go. If the current run rate continues (5.14 kWh/day) we could come in at over 200% of expected. I can't see it happening with the current weather forecast, but something well over 170% is a distinct possibility.

Playing

OK. So since I have a few day's off work for Christmas I've had some time to play with the Google docs links. I have a few more things working, have imported my original tracking spreadsheet and so will crack on with the graph conversion. This should allow a more automated update process in the future, however there are one or two teething problems to be resolved. 

One addition to the blog are the key sunrise/sunset stats etc at the top. This is linked in to a today() spreadsheet function along with a vlookup() which should automatically populate the table above for 5 days, centred on today. Well that's the theory, I shall see if it works automatically over the next few days. If not then I will remove the gadget and publish the full table elsewhere. 

Earlier I lost all my charts on the right and so have rebuilt the links. Hopefully this wont be a regular occurrence, time will tell. 

In the mean time, it's been a dull, damp day with drizzle after a promising sunrise. I don't have the figures yet, but I suspect I've generated less than a kWh today. But that's in line with the weather forecast, so I suppose I shouldn't complain. Any units at this time of year are welcome - it all helps.

Update at last

Lots to say, but no time to type it all in - hopefully with a few days off in the coming weeks I'll be able to remedy that soon. 

In the mean time, I had a bright idea - I wanted to find a better way to publish the charts and stats from the panels, but I was looking for a smarter way, one not requiring screenshots and Photobucket. After a little thought and research I came up with a reasonable smart solution. 

1) Store the data on Google documents (a bit of a fiddle)

2) Create the charts you require (again fiddly, but possible)

3) Change the blog design to include a HTML gadget

4) Select publish from the chart, select and copy the script

5) Paste the script into the gadget

6) Sit back and bask in your own crapulence

(Step 6 not compulsory). 

I like this approach, it took a bit of fiddling and you don't have the chart control option you do in Excel. But it means you can add charts to your blog that update in real time - change the data, the chart changes in Google docs and so do the charts on the blog. I'm quite impressed with how easy it was to do, I shall have to give some thought as to how I might exploit it more. But for the time being version 1.00 is up and running on the blog (hopefully upper right) and I may allow myself a little smile of self satisfaction.

There is a difference

So having finally extracted some data from the Owl monitor (more on this another day), I'm in a position to compare the system output from before and after the scaffolding was removed.

The results are pretty clear. The chart below shows the comparative data for the cumulative generation for the 12 days with and without the scaffolding.

It's fairly clear here that there is a difference, the scaffolding was reducing output by about 50%, which corresponds with the level of drop I was expecting to see from my research.

To be really nerdy, I also put the individual days of data through a t-test, based on a single-tail test of un-paired data. The result again was clear, with p<0.005. I.e statistically there is 0.5% chance that the data difference was accidental and a 99.5% chance it was a real difference. That's well beyond normal statistical tests and good enough for me.

Working again

After a brief visit from the electrician, all is well again.

Turns out there was a bad connection on the AC isolator in the loft, with the connection on the cable insulation rather than the copper - all appears to be well now. We shall know for certain tomorrow when the sun comes up - it was dark when the system went live again. However, I'm fairly confident since the Owl meter came back up to it's usual reading straight away (0.084kw), which is a good sign.

On that note, I have so say how useful having the Owl meter is. It may not be perfect and may not be absolutely accurate, but it's a great indicator of relative output and a fantastic early warning system. Well worth the forty quid or so it cost to buy from amazon.

Just a shame to lose 4 hours or so generation on what was a very nice day for December. However tomorrow, Wednesday and Friday are all supposed to be good weather too, so here's to a good week and breaking the generation prediction for the month.

Nowt

So as of 11am today, on a lovely bright December day, generation has dropped to zero. Absolute zero. Nada. Nothing. El zilcho.

So after a quick dash back from work and a look around the system, the fault is not apparent. The trip in the fusebox is ok, the isolators are on and the inverter is on. However, the inverter is complaining about having no AC connection. So now we wait for the installers to come back and give the system a hearty kick (later today or early tomorrow).

Interestingly enough, the system has dropped to absolute zero (with no background noise at all on the OWL - it normally reads 0.084kW even at night), which makes me wonder whether the power is not going beyond the generation meter.  

Also the Owl record stops abruptly at 11am, but has three distinct drops to nothing in the previous 30 mins or so, with each drop lasting 5 mins or so.

I'm hoping this is a loose connection in the generation meter, or potentially a meter fault. If it's the inverter, then so be it - but I hate to think how long I might have to wait for a replacement :-(

Time will tell. 

Watch this space... 

Early Stats

It's a little early to start putting up banks of stats, but the impact of removing the scaffolding is pretty clear. 

Whilst in place: 11 days, average generation 2.47 units, best 4.36kW , peak 1300W

When removed: 4 days, average generation 6.23 units, best 8.92kW, peak 2400W

Whilst I cant fully normalize these figures to allow for weather differences, they have been pretty comparable, with the second batch on slightly shorter days, although there have been a couple of almost perfect days this month (Dec 2011), especially good for this time of year. 

The predicted daily production for December is about 2.5 units per day, so I'm well up on this. Two or three really good days and I might have completed most of the monthly target well in advance. 

The forecast for today is not great, but there are a few good days in the week ahead, at least according to the Met Office, so here's to a good week ahead. 

Once I have a few more observations I will publish a chart or two - it's a little early yet.

No more climbing frame

30/11/2011

Finally, let me say that again, finally the oversized climbing frame had been removed by the scaffolders and it feels like we have the house back to ourselves. (NB muggins here probably has some mopping up to do and a little washing down at the weekend, but let’s let that ride for now). This happened almost two weeks after the work was finished, making the entire project nearly three weeks long! I think this may be some sort of record.

Apart from the obvious aesthetic benefits, and the increase in upstairs light, there is one major benefit from it going – no more partial shadowing of the panels!

The subject of shade on solar panels is widely discussed on the internet in qualitative (i.e. wishy washy) language, but true facts are harder to find. However an interesting research article posted by the physics department of Arizona University (here) sheds some light on the subject. (No pun intended, well, maybe a little).

In their tests (see figure 1 in the paper) they show that partial shading (say 5-10%) can reduce the output of a panel by up to 94%. This is quite startling. Clearly as the sun move to its highpoint during the day, the shade goes and returns later in the day, you can see a big difference between the output of the shaded panels and the control array. The lower chart show the control experiment in the summer when no shade occurs. Obviously the construction of the panels has a great deal of bearing here, and whilst bypass diodes are a common solution, it does little for the wasted energy of the 90% of the panel that is getting light.

This would appear to correlate well with my findings. At first glance, it appears that the shadow from the scaffolding appeared to be minor; however on closer inspection it’s not so clear. Four uprights and one plank above guttering level will cast a show up to 3 inches wide impacting as many as 6 different panels at a time on the lower row, with smaller overlap onto the top row. 3 inches width of shade per panel out of a total width of 39 inches are pushing on to 10%, so this could be significant for 6 out of the 14 panels in the array. Multiply that proportion by an 80% loss and effectively I could be losing 50% of output due to ‘just’ a few metal poles in the way. Swapping figures with other local users seemed to confirm this – while I was generating 4 units/day similar systems were doing between 8 and 10. Big thanks here to MSE.

Proof of the pudding was revealed almost immediately – within two hours of the scaffolding going, the Owl meter was showing values in the 2000-2300W range, well above the 1300W peaks observed before. Thankfully, with today starting clear and bright, it makes a fair comparison possible with previous best days. (I shall probably trawl through the data in more detail later).

This has really put my mind at rest, now that I have some decent weather, light and no shadow I can really see what the system is capable of, even in low light levels, and suddenly meeting, or hopefully exceeding, the target output figures seems a distinct possibility.

I shall continue to monitor for the next few days, but I suspect the difference will be clear. Watch this space for more information and hopefully a t-test or two.