I thought it made sense but now it doesn't!

[series530]

We are seriously looking at installing a PV system. Our house faces roughly south west and, at the start seemed to be pretty shade free. We've had two surveys done and both have highlighted shading due to the property next door and a large chimney on the front of our property. This is perfectly fine and understood.

The first company has said that it should be possible to get a system of about 4Kwp using 16 panels arranged as a row of 8 along the ridge and a second row of 8 under the dormers at the front. Both are in the centre of the roof line and avoid the chimney and the shade from next door. This system will use a string inverter (SMA).

The second company has said that the maximum recommendation is about 2Kwp and that only the section just below the ridge should be used because the panels under the dormer would be too close to the bottom of the roof line and may be shaded by a telephone and power line that runs diagonally across the front garden (nearest point to the house is about 10 feet, furthest point is about 45 feet). This company also recommends micro inverters as the chimney may have some effect as a shader.

I have a quote from the first company and it is competitive. The quote from the second company will take some time because they want to do some software modelling to decide the best way. No doubt the second quote will be more for, theoretically, less power output. Obviously, the first company could be suggesting a system with a huge on paper output that, in reality, is in efficient.

I feel that both companies have valid points. I don't want to cut corners but I also don't want to pay through the nose for something that may be a perfect engineering solution but which doesn't provide a decent pay back.

So, what are the rules or suggestions about the minimum spacing between a panel edge and guttering or the base of a dormer? If there are no rules, are there any guidelines? Also, what about these power lines, will they make the panels that much more inefficient and bring the whole system down if I go for company 1? If I had a tree in the front garden I could understand the concern, but power lines a fair way from the house ... is that really a big deal or not?

Many thanks in advance.

 

[SolarCity]

Funny you should ask about power lines.

We were interested to find out how much of an effect the power lines have so we carried out a few experiments. Using a 4mm solar cable tied to two rails, we cast a shadow on an array on a job we carried out a few weeks ago. It absolutely does have an impact - surprisingly even if it's just on one panel. The furthest we could get away from the array and still shade the panels was just four metres and this still had an effect on the array. On an install we carried out on Monday, we tried the same experiment. The light was more diffused on this day and the effect on the array was far, far less.

There isn't any regulation for installing panels too close to a ridge or eaves but it I would say it it certainly isn't a good idea to install closer than 200mm - for both wind concerns and the fact that the guttering may become redundant. Can you post a photo of your roof so we can have a better idea?

 

[series530]

very interesting - thanks for relying.

I would post a picture of my roof ordinarily. The problem is that I am now away on a business trip and wont be home until tomorrow. Roughly, from memory, the pitch is about 50 degrees, it's a chalet house, the roof section from the ground floor to the window of the dormer of the 1st floor is about 5 or 6 feet, the window is about 3 or 4 foot (vertical) then there is a flat roof above the dormer and the section above the dormer is probably about 8 foot. Sorry I cannot be any more precise than that.

I'm keen to know how much loss there is from the cable in your experiment. Are we talking fractional percentages or an order of magnitude more? Obviously, I need to consider this in moving forward. I suspect that the minimum distance between the nearest panel and the cable is going to be at least 12 feet (maybe more) and the furthest point from the array will be a good 20 feet - I guess though that the loss will always follow the worst point - hence getting some figures would be interesting.

I suppose that a micro inverter will negate some of the loss but I do worry about fitting micro inverters having read the bad press and in considering the additional cost that they will add to the installation. I guess that an string inverter will suffer far more from a cable shade on the array - hence wanting to understand how much the losses really may be.

The issue of the guttering is interesting too. I was thinking about the problem more from wind effects. However, the issue of the rain and the loss of guttering is much more of a problem now you mention it.

 

[SolarCity]

I'd say with a 50 degree pitch you should be extra worried about the gutter.

Another thing is the orientation of the roof. Being south west facing, does your dormer window shade the area below it in the morning? I'd be very suprised if it didn't.

Regarding the cable, in direct bright sunlight we were getting drops of around 15% when we introduced shade from the cable - this effect came when the shadow was cast on 8 of the panels in fairly close proximity, far closer than the distances you mention. However, when the shadow was held as far from the array as we could get it we still noticed drops of around 5% - 10%. We chose this site as an experiment because we had two strings of eight on different inverters and it gave us an unusual opportunity to check one inverter against the next.

It sounds to me like the second company is preparing some modelling to determine the full effect of the shading and this should give you a more accurate idea of what you should be installing. From the information you have given, I'd say that they seem to be a bit more interested in selling an economically viable system which impresses me.

 

[jerriais]

It sounds to me like the second company is preparing some modelling to determine the full effect of the shading and this should give you a more accurate idea of what you should be installing. From the information you have given, I'd say that they seem to be a bit more interested in selling an economically viable system which impresses me.

Seconded, we do shade modelling for all cases where shade could be an issue and even with 8 years experience in the solar industry I'm often surprised by the effects of shading so any company that can tell you that shading will or won't matter without doing proper 3D analysis is suspect in my book.

 

[kristoff84]

I think you should run proffesional simulation in industrial packages, such as PVsyst in order to estimate the potential shadow impact on PV arrays. If the shadow is expected to be there during all day I would recommend to move the modules out of the shadow to avoid high losses in the system. Without doing a proffesional modeliing it is difficult to estimate the shadow effect.However, if you need to install modules in the shadowed area try to use either 2 MPPT trackers inverter, such as SMA 4000TL or Microinverters,whcch are still quiet expensive compared to standard inverters. Bear in mind that when you connect your modules in series, shadow which effect one module will automatically impact other modules in that array. Howevery, most of PV manufacturers use bypass diods which help to prevent the PV module/arrays by the effect of shadow.

 

[jim1976]

completely of subject how do you put a new thread on here ???

 

[series530]

I'm back home now: to try to answer some of your questions:

The distance between the bottom of the dormer and the bottom edge of tile roof (where it joins the gutter) is 1.9M. The panel being advocated by the first company is the REC peak energy series. It has a length of 1.65M giving about a foot of difference between its length and the length of the sloping roof.

Our house has a width of about 15M. Both company A and B said that only the middle section of the house is suitable for PV because, at one end, our neighbours house is about 2M in front of our building line, is two storey and has a pitched roof and a couple of aerials on the chimney. All of this renders the left hand third unsuitable due to shading. The third on the other side has our chimney (which is a rather massive limestone clad affair which extends about 1.5M above the top of one of the two dormer sections). Both companies said that no units should be located between the chimney and the end of the house (something like about 4M).

Company A, who are suggesting the REC units and an SMA 4000TL inverter said that we could have 8 panels in the viable area in the middle just below the ridge (but above the dormer) and a further 8 panels below the dormer. I don't think that the dormer really adds much shade because the shade is caused predominately by the house next door and our chimney. Company A said that we may consider running each of the two sets of array using a dual string array system rather than a single one. The peak KWp was quoted as about 4000 with this system. Company A would do the micro inverter if I wanted it but the price would be about £2000 more overall.

Company B, suggested that the space below the dormers was a definite no no because of the panel size (wind lifting them) and because of the shading that would most likely be caused by the telegraph cable. I measured the distance between a theoretical place where the panels might be and the nearest point of the cable. It was 5.5M or there about. I would estimate that, given the cable runs diagonally away from the house, at the point at the other end of the array, the distance would be in the region of 8.5M (maybe more). Company B were suggesting Sanyo units on grounds of higher efficiency and most probably using micro inverters. They estimated that we wouldn't get any more than about 2KWp using their system with 8 panels.

I haven't had a quote from company B because they want to run the simulation. I want to get the data before I make any decision. I have the feeling that company B will be tendering a quote that is at least 30% more than company A - for a system with a theoretical output that is much lower. Obviously, the Sanyo units are more efficient, the company plans to use thicker wiring and micro inverters (the jury is still out for me regarding these if I am honest). Company A will be using the string inverter approach, panels that are not as efficient as the Sanyo ones and, unless the lower array is hopelessly inefficient, it should offer more output power (unless I missed something) for less money.

I want to keep an open mind. It's all too easy to be swayed by a lower (but not bargain basement) price and then find that what you buy isn't as good as you expected. On the other hand, I don't want to spend Rolls Royce money and end up with something that is performs like a mid range Mercedes (which is still a perfectly fine but cheaper car).

 

[series530]

A quick update: we now have a third company who understood the issue of shading at both ends of the house. They suggested that we could get away with 4KWp worth of panels using arrays above and below the dormer. I asked them what they thought about the shading from the cables. A look of surprise appeared. We went outside and took a look. It was 12.00 noon and what was interesting was that this was the first sunny day where I had taken an opportunity to see the shading. One of the earlier posters asked if the dormer caused any shading? I can confirm that, at that time of day there is a nice shade line along the top of where the lower array of panels would be. In fairness, it would probably go but a portrait orientation would render them useless until the sun passed over. The chap suggested fitting the panels landscape instead. This would get around the issue of the shading but it does drop the output somewhat because the loss in surface below the dormer cannot be recovered by adding more panels above (assuming I kept all the panels in landscape mode). So, the best we could go for landscape is 3KWp.

All of this got me thinking:

By far the biggest ROI comes from the FIT. The saving in locally generated power is probably second but miles behind and the export portion is worth two bob and an orange. I could, in theory, fit 4KWp worth of panels on my roof and each and every one of them could remain in shade for most of the day if the sun never shone. I would still get my FIT based upon the theoretical output and the minimal generation would power some or all of my appliances during the day. If I fit less panels, my FIT is lower so it will take longer to get my ROI.
Provided I could afford the outlay in the first place, I may as well go for a bigger array even though some of it may suffer from very minor shading for part of the day (the base of the dormer for instance) because the FIT pays based upon theory rather than actual power produced.
There really is no benefit in buying more efficient panels or micro inverters or anything else to get the efficiency up because it is simply throwing money down the drain - as long I can generate enough to cover usage during the sunny part of the day, I can forget about the export portion because it will never cover the cost of smaller and more efficient panels or micro inverters to address minor shading issues.

Obviously, all of this ignores the green element - but looking purely at the financial perspective, fit reasonable quality and don't get too hung up about minimal shading because, as long as the arrays have a reasonable output and aren't too large for the roof a larger capacity installation typically comes with a quicker pay back time and more absolute output for a given level of shading.

Have I missed something fundamental in my thinking?

thanks

 

[SolarCity]

I think you've been confused somewhere down the line. FITs are paid on generation readings, not theoretical output. If you're not producing, you're not earning.

It sounds to me like you have a complicated set-up and it would really benefit from some 3D modelling.

 

[series530]

Obviously, I'd got myself mixed up with the hype of quoting. Thanks Biggsolar for clearing that one up. The 3d modelling results are due any day from company B (possibly even tomorrow). So, the shading does play a very significant role because it affects the true payout. Thus, it is better to install a smaller array that is less or unaffected by shading than a bigger array that is pinched for a fair portion of the day and outputting under par (ignoring bypass diodes or micro inverters) because the payout is really on energy produced and not what is written on the quote. Am I back in reality land again now?

 

[liquidity]

The 3D modelling for both configurations (maximum kWp but more shading issues vs less kWp but little to no shading) is critical in your circumstances. Many companies will gladly cram a 4 kWp system into your roof and walk away - without good modelling you won't know for a few months or longer whether or not the system will have a reasonable ROI. Microinverters will help but they generally drive up the cost and even with a micro inverter there's little point in installing a panel that gets a lot of shade.

 

[jerriais]

The 3D modelling for both configurations (maximum kWp but more shading issues vs less kWp but little to no shading) is critical in your circumstances. Many companies will gladly cram a 4 kWp system into your roof and walk away - without good modelling you won't know for a few months or longer whether or not the system will have a reasonable ROI. Microinverters will help but they generally drive up the cost and even with a micro inverter there's little point in installing a panel that gets a lot of shade.

Quite right, in all the 3D modelling simulations I've run with microinverters I've usually found that they really make a difference when a thin shadow passes over a module such as when a telegraph pole or antenna is casting shading but they are no better than coping with solid shading than any other type of inverter.

Matt

 

[solarfred]

QUOTE
Funny you should ask about power lines.

We were interested to find out how much of an effect the power lines have so we carried out a few experiments. Using a 4mm solar cable tied to two rails, we cast a shadow on an array on a job we carried out a few weeks ago. It absolutely does have an impact - surprisingly even if it's just on one panel. The furthest we could get away from the array and still shade the panels was just four metres and this still had an effect on the array. On an install we carried out on Monday, we tried the same experiment. The light was more diffused on this day and the effect on the array was far, far less.

There isn't any regulation for installing panels too close to a ridge or eaves but it I would say it it certainly isn't a good idea to install closer than 200mm - for both wind concerns and the fact that the guttering may become redundant. Can you post a photo of your roof so we can have a better idea?
UNQUOTE

Could you confirm that the panels in your experiment were in portrait and therefore the shadow was cast across the the narrowest widths and did you try the experiment from "top to bottom" or in other words based on a landscape format?? The results should be quite different and potentially determine the layout plan.
Thanks

 

[SolarCity]

The panels were in portrait - interesting point about the landscape format though. I'll have to give it a try.