Effects of shading.

Hello.

I'd appreciate your thoughts on my planned installation as I've received opinions from various installers which range from, literally, "No, your system is not viable." to "That's a shading factor of 1 - ie: little to no effect..."

My array will likely consist of 2 rows of 7 panels in portrait format, panels likely to be 190w Suntech or Sharp or similar. The problem I have is that for some winter months, a nearby telegraph pole will cause a moving shadow over both rows. The shadow will only begin to fall on the panels from 1pm onwards - up until then all is ok.

However, during the 3 peak summer months, the shadow won't hit the panels at all as it falls short of the roof sue to the high sun position. For the remaining 6 months, there will be partial shading - eg: only the lower row, and fractions in between!

Everything else is absolutely perfect - exactly south-facing, and roof at a 39o pitch.

I've even contacted the actual panel manufacturers, but they say 'ask the installers'. The installers, as I've said, have given every possible answer covering every possible range.

Thanks!

If you can, get them to split the array L/R rather than top/bottom.

Any reason why 190W panels, cost? have a look at increasing to 250's

Make sure you have a dual tracker in the inverter to reduce shading issues on each half. with 14 panels, there will be a crossover point where the shading affects both simultaneously. As you said, there will be a reduction on output in the winter anyway, but this should make the best of an ackward roof.

Personally, I would say that the effect of shading can never be under estimated. We found that some installers said that it was not a problem and others said that it would kill the output of a whole section of the array on our house. A decent installer will offer an inclusive shading analysis and you can use that to make an assessment.

If shading proves to be a major issue you may mitigate some of the effects via the bypass diodes or, as suggested before, you could (and probably should) put some of the panels on a separate string with a separate tracker. The other approach is to use micro inverters so that each panel is controlled independently and the effect of shading will only drop that panel and not the whole string. That route is more expensive but you may find that the shade mitigation over the 9 off peak months pays for itself over the life of the FIT as every other panel will still be producing near peak and not be dragged down by the lowest performing panel on the string. The jury is out on micro inversion, some love them and some hate them but it cannot be denied that they can be helpful if shade is a problem.

If you number the top row of panels 1 to 7 and the bottom row 8 to 14 which numbers are affected in the winter and which numbers are affected in the spring and summer?

Interesting. We've just put a full 3D shading model and horizon into PV*Sol and run it through both on visualisations and output.

Final result: visualisation exactly as we suspected, looked spot on to what both I and the customer anticipated.

Factoring in the fact that most of the horizon tree shading would be in winter, PV*Sol's output figures were less than 10% different from no shading at all.

Watching the visuallisations time and time again, we're sure it's correct.

So you CAN overstimate the impact of shading.

Gents! Thanks a million for your prompt and thoughtful replies - most grateful & impressed!

(I'll outline more detail as requested by Liquidity further below.)

First, to answer Graeme Harrold, why 190w panels? Yes - it's to 'do' with cost, although that's not as simple as it seems. It's more to do with 'value for money'! For example, my two cheapest options are using either 185W Sharps or 190W cheapo Chinese. The cost difference between these two options is only £400 (the Sharps being cheaper) so my immediate reaction was to think "go for the 190s - it's only 400 squid more". However, a quick calculation showed me that the meagre 70kWh annual difference - woooo, that's £30 - wouldn't pay back the extra £400 outlay for waaay over 10 years - ergo, not worth it.

Similarly with Microinverters. These will add £thousands to the cost - for what gain? 20%? That will take decades to recoup! Worth it? Well, it just doesn't add up for me - unless I'm doing something wrong here.

And here's the other weird thing; I've been checking out Enecsys's website recently (because I thought micros would be the way to go) and followed their test example with interest; you know, the one that shows two identical PV arrays, both of which have dustbins (waaaay over the width of my 150mm post!) completely blocking one panel in each array. Yes, their micros increase overall output by 20%, but - hey - what really stood out for me was that the 'normal' array wasn't doing that badly either! And that was with a muckle great big bin in front of it. How much will that 20% gain cost you? Hmm, around £2-3k. Suddenly, the micro system ends up costing you pretty much what it'll return, if not more...

Also, for my situation, I need to take into account that the shading only begins at 1pm all year - so roughly half the day is absolutely fine in any case.

Worcester, does your software examine shading that falls over a whole panel like mine?

To get back to Liquidity, say top row is number 1 to 7 left to right, and bottom 8 to 14, also l to r. Typical winter scenario is that at 13.30, panels 1 and 8 will have complete vertical shadows on them, 150mm wide. At 15.30, the panels affected will likely be a lower corner of 3, and almost completely upper-half diagonal of 4, a complete lower diagonal of 9 and an upper (top left) diagonal corner of 10. (Phew...)

From mid-Nov to mid-Jan, the shadow will cover the full roof height. From mid-Feb to mid-Oct, only half the upper row will be covered at worst. Mid-March to Sept, only the lower row will be affected, and from mid-May through June, July to latish August they should all be in full sunshine.

Hence my thinking that a cheapo system will provide the same return on my outlay as an all-singing type.

Or am I more confused than I first thought...:angel_smile:?

Thanks again, folks - it's all helpful info.

The "factor of 1" nonsense with shading calculations is one of my biggest issues with the SAP calculations. It is utterly nonsensical to suggest that a little bit of shading is going to have zero effect.

I started off by taking a short term view. I looked at a number of installers and began to make my decision based upon solar yield capacity vs price. I thought that the FIT was paid based upon the KWp output. A very helpful member on here then pointed out that the FIT is paid based upon generated electricity and not upon a SAP calculation. Had my roof been totally unshaded it would still have made sense for me to put as many cheap panels on the roof as I could because ROI would be fastest. However, my roof did suffer from some shade and this simple formula suddenly didn't stack up any more.

One of the three installers, who had expressed what I thought was excessive caution about shading on my roof suddenly began to make some sense. There really was no point in just throwing a load of panels up there if, for a fair bit of the time, one or more of them was killing the string. My cost would go up and my ROI time would increase.

So I started to think more about the longer term. One of the issues with this is that there are so many variables over which we have no control: how much sun there will be, if the FIT is killed or stunted through political intervention and so on. However, taking a pragmatic view and crossing fingers I took a look at our life style: I work from home (when I am not away on business) so we do consume power during the day. We have a number of garden features that pump water around and its a shame to turn them off. We could run the washing machine and dish washer during the day, etc, etc. So, with this in mind and the fact that the FIT is going to pay me for the energy that I produce at quite a decent rate and whatever I use during the day is free (and should get to be even better value as power prices increase over the next 25 years) and I will get something for what I export, maybe there is something in the idea of maximising my harvest over the longer term.

Now, in my case, and time will tell because it isn't installed as yet, my array would be affected by next doors chimney and aerial. A single string array would be knocked out until early afternoon for much of the day. A multi string tracker would mitigate some of this but, a dual string would certainly stunt half of the capacity. Interestingly, the chimney goes away quickly, but the aerial would shade a couple of panels for quite a while. It's amazing how a 2" diameter pipe has such a major effect.

Even this didn't convince me that micro inversion was worth the extra. One of the installers (the cautious one) put me in contact with a few of his customers and they, in turn, gave me access to their web pages so I could see real time energy production. One had little or no shading and the other had a tree obscuring a couple of panels later in the afternoon. The tree aside, it was a revelation to see how the panel production varied across the array. While the overall average yield per day was roughly the same for all of the panels (and be aware that maximum panel output varies - the worst will drag the string to some extent) , the instantaneous yield was going up and down almost randomly. In fairness, I haven't visited the sites but I am assuming that the variation is down to moving cloud and reflections. Theoretically at least, a micro inverter based system will track the variation and try to harvest the maximum yield on a per panel basis as opposed to a single string inverter that gets dragged down to the lowest common denominator. You have to offset this increased harvest against the slightly lower efficiency that micro inverters offer though.

I was still not convinced and this was mainly because of the increase MI cost vs a string inverter. They do cost more and there is more to go wrong. However, the Enecsys ones come with a 20 year warranty. Nobody knows if they will last 20 years but you will get a replacement in the event of a failure. True, there is an installation cost but, chosen wisely, better installers will offer an independent labour warranty to mitigate some of that risk. If one dies outside of that warranty, with care, they can be replaced without a huge labour and infrastructure cost. A string inverter on the other hand does not come with a 20 year warranty unless you pay extra. Statistically, a string inverter will not last 20 years so you have to build the operational cost into the equation (unless you intend to move house before it breaks). Considering this, the MI route, while still more expensive, is not as big a delta as it may first appear.

My post may come over as an advert for micro inversion. This is not the intention. I'm gambling on this route in my case because of the shading and I'm try to embrace an approach which is a longer term view that should give me an ROI of about 8 years and a healthy profit for the remainder of the FIT - somewhat under the assumption that the servicing costs of the installation will be much lower overall. Part of me also feels good about being even a little bit green. My feeling is that PV is a huge investment but the FIT, provided I get in early, pays a healthy return. Best then to invest a bit more (because of my shading) and hope to reap a reward overall. I wouldn't take the MI route with no shading, for sure. My plan may be full of holes, it may prove overkill. I have no idea. Only time and continuation of the FIT will tell.

Ian

Again, thanks so much - this is really helpful.

BiggsSolar, I have to agree - and I'm only using 'common sense', not any technical installer's knowledge. What is 'light shading' for heaven's sakes, and is a bit of soft shadow caused by a few dappled leaves the same as a stark, crisp shadow caused by a pole? It always struck me that the 'shading factor' used was too broad a brush. But, hey.

Ian, many thanks for your thoughtful post. It makes a lot of sense. No doubt in my mind that Microinverters will make a significant difference, but the jury in my head is still out on whether it would be cost effective in my case - taking into account that its benefit would only be needed for half of each day, and then for only roughly half the year.

I'm still trying to get my head around Enecsys's own test on their website - a black bin, which must be around 600mm wide, is completely blocking a panel (or two?) in an array. And yet the Micros only give a 20% increase in output? Hmmm.

It isn't that the Micros are poor performers, not at all; It's more that it appears that a fully blocked panel will only drop an array's output by around 20%! And I'd have thought that Enecsys would want the 'normal' array's performance to tumble as much as possible to make their system look good, so they'd have tried to get the worst possible situation with that bin... Surely? I really don't understand this.

If it's really only a ~20% improvement I can expect, then - taking a broad guesstimate - in my situation it would only amount to - ooh - 10%. Is this worth £2-3k extra outlay for me? It doesn't appear to.

Can I ask installers with software to answer this hopefully simple situation: A single straight array with 14 typical panels connected to a single Sunnyboy in full sun. Output is? Now, one panel in that array has full shading over its whole height - what drop in total output will occur. (Panel manufacturers cannot give me a straight answer to this :angel_smile

Thanks, all.

A supplementary question, please! With, say, two rows of 7 panels, can each row be connected as usual - in series - and then both rows together in parallel to a Sunnyboy? So, that each row can work independently of the other? Or does this need a special inverter?

Scooby said:

Can I ask installers with software to answer this hopefully simple situation: A single straight array with 14 typical panels connected to a single Sunnyboy in full sun. Output is? Now, one panel in that array has full shading over its whole height - what drop in total output will occur. (Panel manufacturers cannot give me a straight answer to this :angel_smile

Click to expand...

I can't get a figure of how much a shaded panel will affect the overall output at a certain point in time but I could let you know how much of an annual effect it will have If you give me some dimensions, location, pitch of your roof and a photo I could try and give you an idea.

That's really kind, Biggs. This shows my garage roof at EX39, Devon. It is exactly south-facing, at around 39o pitch, and the max area is 8m x 3.9m (this includes eaves overlaps at each end, so usable space is more like <7m x 3.5m. This photo was taken in January, where the shadow height is at its worst. It's like this for around 3 months, but the shadow doesn't land on the roof at all for 3 to 4 months in summer - with every stage in between!

Not even panel manufacturers will tell me how much a thin, vertical, straight shadow will affect a panel and the array it's in! That's really the answer I need...

Okay, estimating the size of the pole, I have run this through PV Sol Expert.

In the layout shown, it calculates an annual loss of 14% through shading.

There's an interesting illustration of shading here - more real world than with dust bins (Enecsys need a better marketing department!)

http://www.builditsolar.com/Projects/PV/EnphasePV/Shading.htm

It also shows the mppt effect of optimising each panel.

BiggsSolar, thanks very much for that info. Does your annual loss calc take into account the variation in shading height over the year - ie: the fact that there will be no shading at all over the 'highest' months?

Series530 (Gary?), it's good to see a real life example. Mind you, what I'd really love to see is your output performance with a series string! The professional answers you received mainly go to show how there still doesn't appear to be a definitive answer... Jeepers.

Is this info out there?!

Cheers.