Renewables: "talking 'bout my generation"

Martyn1981

EricMears wrote: »

I'd be interested to know how NG get their figures for PV generation.

They use PV Live. I know you've always completely dismissed the industry standard of kWh/kWp (Os on here) as useless, but with some sampling nationwide and knowledge of the pitch/orientation of those samples, you can get a pretty good estimate of national PV energy generation via knowledge of installs and averaging, and at any given time a very decent estimate of power.

Also, whilst no central metering of PV exists in the UK, I'd assume/suspect that virtually all grid tied PV generation has some sort of metering, so data collection can be done to verify and revise estimates as time goes.

Martyn1981

Martyn1981 wrote: »

National Grid reported a record yesterday with 8GW of PV generation, supplying 25% of demand at times.

Gridwatch have changed the relevant graph from 7GW to 8GW as PV was going into the headings at 7GW, but is now doing the same at 8GW.

With regard to today, it must be a very low CO2 (gCO2/kWh) day, as we have strong PV, the wind has really picked up since late morning, and it's a Sunday, so low demand. Coal is barely on at the moment, and gas is a surprisingly low 7GW for 3pm.

Plenty of PV gen here too, and washing and vacuuming done. :cool:

legoman62

old_grouser wrote: »

6.1 Os yesterday, and 6.3 Os today, and a forecast of sun all day tomorrow on the sunny south coast. Boring! I made a suggestion in a post on St Patrick's Day that 25 Os over five days would really be something to celebrate, as opposed to the occasional Hawaii. Now I realise I'll be there by about 9am tomorrow, and the dream score ought to be 30 Os over five days - a really high five! I wonder if I or anyone else will manage this.

A/an Hawaii is a dream score for me with my SW facing panels:D

I would do well to beat a S facing roof:eek: most of the year.

EricMears

Martyn1981 wrote: »

I know you've always completely dismissed the industry standard of kWh/kWp (Os on here) as useless

That' a considerable over-simplification hence I'm afraid just plain wrong.

What I have often pointed out is that when comparing two properties, the simple kWh/kWp ratio is seriously flawed in that it doesn't take into account the many potential differences between the two. e.g. they could be in different parts of the country, the panels may point in different directions and one set may be more efficient than the other. The only accurate way to compare two individual properties would be to express each one's daily output as a percentage of its pvgis forecast for that month (or ideally for that day but the estimates are harder to come by).

If one knows the total kWp of the 244 individual sites and the total installed kWp over the whole country then it would be a meaningful calculation to try and scale up their daily output to get an estimate of the total daily output over the whole country. That of course makes the assumption that the 244 sampled sites are truly representative of the whole country. I have to say that there are a number of serious flaws in that assumption :-

244 isn't a very large number compared with the hundreds of thousands (millions even ?) of installations over the country.
The 244 are probably unchanged from year to year whereas the whole country adds a few more every year. Newer installations will probably use more efficient panels.
A very large sample population randomly sampled would probably include a 'fair' share of optimally & non-optimally aligned panels. A small sample population is unlikely to do that accurately.

I'd hate to even try and estimate the sources of error in such a calculation but would expect it to be at least 10%. The only real hope is that there are likely to be errors in both directions so they might balance out to some extent.

zeupater

EricMears wrote: »

... If one knows the total kWp of the 244 individual sites and the total installed kWp over the whole country then it would be a meaningful calculation to try and scale up their daily output to get an estimate of the total daily output over the whole country. That of course makes the assumption that the 244 sampled sites are truly representative of the whole country. I have to say that there are a number of serious flaws in that assumption :-

244 isn't a very large number compared with the hundreds of thousands (millions even ?) of installations over the country.
The 244 are probably unchanged from year to year whereas the whole country adds a few more every year. Newer installations will probably use more efficient panels.
A very large sample population randomly sampled would probably include a 'fair' share of optimally & non-optimally aligned panels. A small sample population is unlikely to do that accurately.

I'd hate to even try and estimate the sources of error in such a calculation but would expect it to be at least 10%. The only real hope is that there are likely to be errors in both directions so they might balance out to some extent.

Hi

The logical approach would be to apply the sampled live dataset performance to the full MCS installation database on a geographic installation density basis in order to estimate generation with a reasonable degree of accuracy ... I'd be very surprised if this isn't the approach they're employing, therefore there's no fundamental reason to assume that their figures either aren't representative or not within a tolerance well within a +/- 10% band, with the probability of intentional underestimation being the case on credibility grounds ... if this is the case and the installation dataset is continually reassessed, then it's possible that accuracy could be maintained at within +0/-3% for the majority of the time ... be interesting to have official comment from Sheffield University on this though !....

HTH
Z

rugbyleaguesmate

5.35 o's today 260kw for the month already fabulous morning afternoon still strong

Merlin139

Nice sunny day but the clouds came in too early to spoil a maiden.

Hawaii + 0.9 O's

ASavvyBuyer

Just 3.585 O's today.

legoman62

Not quite the day we were promised. Still OK. Made 4.625 O's:T

old_grouser

old_grouser wrote: »

6.1 Os yesterday, and 6.3 Os today, and a forecast of sun all day tomorrow on the sunny south coast. Boring! I made a suggestion in a post on St Patrick's Day that 25 Os over five days would really be something to celebrate, as opposed to the occasional Hawaii. Now I realise I'll be there by about 9am tomorrow, and the dream score ought to be 30 Os over five days - a really high five! I wonder if I or anyone else will manage this.

Well, it happened! 6.15 Os today, making 30.3 Os in total for the last five days.;) I may have to change my username.

On the other hand, I find myself wondering how often this has happened in the past five and two thirds years since the system we call Rufus came to live on the roof. I don't really do data. Young Mrs Grouser, the data queen, is going to show me how to do the analysis, but first I have to combine sixty something CSV files into one in order to crunch it all. So, as you were....:cool: