KB's Design

[kb58]

Oh I know all about entropy, how when energy is converted from one form to another, the new form has less energy than its original (and is why every perpetual motion machine on YouTube is put forth by someone who's trying to trick you, is delusional, or both, but I digress).

I agree with you in principle, but it's not that simple for end-users to know which is actually "better." For your example above, say the person across the street from the power plant needs to buy a new hot water heater. He sees two in the store, both the same size and same price, one gas, one electric. One says it uses $250 a year to run while the other says it takes $200. Which should the buyer purchase? I didn't say which was which, and that's the point. One may have a more efficient design, or may have better (or worse) insulation. Then there's the case of extenuating circumstances. Buying a heat-pump style electric hot water heater makes since if it sits in a hot garage which you wish was cooler.

I get what you're saying, but how a piece of equipment is designed can undo much - or all - of that energy's natural advantages.

[kb58]

...Some of the questions you may have:
Have you looked into inverter types yet, be they string micro inverters or optimizers. Given your shade, I'd suggest that string inverters may mot be the best choice, maybe optimizers ?
How about panel choices ?
Start looking for a PV vendor/installer.

Good luck.

All are up in the air due to conflicting issues.

Inverters: In a perfect world I'd want an inverter that:
1. Doesn't affect ham radio reception, meaning no microconverters or optimizers
2. Has an "always live" plug on the side of it so that I could power a refrigerator or cellphone during daylight hours in the event of an outage.
3. Has an EV charge point - or at least as an option.
4. Has a great warrantee and customer support

None offer all the above. The closest I can get is an SMA unit with Jinko panels (with their built-in non-RFI-generating "optimizers"), and it has a 2000W plug on the side of it, but no EV option.

SolarEdge's website, applications, and instructions, are much better than SMA's though. OTOH, I hear of poor SolarEdge customer support. SMA, who knows.

That said, and playing devil's advocate:
I'm a ham radio operator, yet how much should I care about RFI? Being solar-related means there won't be any at night. OTOH, as retirement approaches, I may well want to operate during daylight hours. Six one way, half dozen the other.

Regarding the 2000W always-on plug, how important is it, really? How often have we had an outage lasting more than an hour? Once, ever.

These topics get dangerously close to the slippery slope known to survivalists, where the end of society is always "right around the corner". In SoCal at least, our most likely apocalypse is an earthquake and in that situation, I don't think having electricity is all that helpful. Water is the big thing here and it doesn't matter how many guns, bullets, and beans the tin-hat crowd has stored away once they get thirsty. But hey, with their guns, they can take all the water they want... again, I digress.

On the topic of vendor/installers, I'd do all the work myself. It's a way to both save money and speed up the ROI payoff.

[J.P.M.]

Oh I know all about entropy, how when energy is converted from one form to another, the new form has less energy than its original (and is why every perpetual motion machine on YouTube is put forth by someone who's trying to trick you, is delusional, or both, but I digress).

I agree with you in principle, but it's not that simple for end-users to know which is actually "better." For your example above, say the person across the street from the power plant needs to buy a new hot water heater. He sees two in the store, both the same size and same price, one gas, one electric. One says it uses $250 a year to run while the other says it takes $200. Which should the buyer purchase? I didn't say which was which, and that's the point. One may have a more efficient design, or may have better (or worse) insulation. Then there's the case of extenuating circumstances. Buying a heat-pump style electric hot water heater makes since if it sits in a hot garage which you wish was cooler.

I get what you're saying, but how a piece of equipment is designed can undo much - or all - of that energy's natural advantages.

I figured you know, but my response wasn't aimed at you. It was respectfully aimed some at AZ66 in response to his post, but mostly for other readers not as technically savvy as you.

As a simple, real world , or at least most of the U.S. practical and applicable example, if one fuel source, nat gas, is about 1/3 as costly as another, electricity, as is common in the U.S. and, in addition, installation and operating costs aside from the fuel cost are comparable, seems a done deal for nat. gas if cost is the determining factor. Besides, and to reiterate, that was an example to illustrate a point.

But, not my money, house, life, etc.

I'll save my venom about how stupidity has become a lot more common than it once was with respect to the lack of common sense, much less critical thinking skills people have lost with respect to intelligent decision making of the type you write of.

A heat pump water heater that's purchased partially with the idea of cooling a garage, but costs more to operate than a gas, tank type water heater to meet the primary duty of heating water is, or may well be, false economy, especially in a cold environment where heat pump water heaters are probably never as cost effective as nat. gas for the water heating application.

A well designed piece of equipment will use the energy source attributes to best advantage for the application. That's why it's called design and that's why God made engineers. Any design that doesn't do that is a POS and ought not to be done in the first place. That's what DIY'ers who think they're innovation geniuses because they may know which end of a hammer to pick up engineers, and U-tube are for.

[kb58]

After seeing the self-shading thread I thought I'd check what SAM is telling me for my proposed setup.

As background, I plan the array to consist of two-panel subassemblies in landscape-mode, ground-mounted, with a total of 9-11 subassemblies oriented toward 220 degrees, each one directly behind the other.

I noticed that SAM's calculated row spacing never changes based on tilt angle. Maybe that's because it's already figured out what it needs to be to have zero shaded at all times of the year. I'd have thought though that if I set tilt angle to, oh, zero, row spacing should change to panel width but it does not. Maybe I input something wrong, but that led to the realization that having multiple rows in-line pointing south(ish) could be problematic. Until recently I thought I could model the latitude and desired tilt angle of 30 degrees and get a quick answer. Picturing how the sun moves through the sky though makes me think it's not that simple. That is, with row spacing based upon the tilt angle of 30 and winter solstice here being about 33 degrees, that seems to indicate that regardless of time of day, the array will always experience self-shading anytime the sun is lower than 33 degrees elevation. That's a problem, especially during the summer when the proposed TOU means it's really important to generate power for as late as possible. Because the site is clear nearly to the horizon in the summer afternoons until sunset, I may have to consider pointing the panels further west, and/or increasing spacing. The problem with the latter is that the array's footprint get huge.

I'll do some more modeling, but I thought I'd bring it up for others to contemplate or comment on.

Here's a repost of the site shading (the spike at 240-260 isn't real):

[J.P.M.]

After seeing the self-shading thread I thought I'd check what SAM is telling me for my proposed setup.

As background, I plan the array to consist of two-panel subassemblies in landscape-mode, ground-mounted, with a total of 9-11 subassemblies oriented toward 220 degrees, each one directly behind the other.

I noticed that SAM's calculated row spacing never changes based on tilt angle. Maybe that's because it's already figured out what it needs to be to have zero shaded at all times of the year. I'd have thought though that if I set tilt angle to, oh, zero, row spacing should change to panel width but it does not. Maybe I input something wrong, but that led to the realization that having multiple rows in-line pointing south(ish) could be problematic. Until recently I thought I could model the latitude and desired tilt angle of 30 degrees and get a quick answer. Picturing how the sun moves through the sky though makes me think it's not that simple. That is, with row spacing based upon the tilt angle of 30 and winter solstice here being about 33 degrees, that seems to indicate that regardless of time of day, the array will always experience self-shading anytime the sun is lower than 33 degrees elevation. That's a problem, especially during the summer when the proposed TOU means it's really important to generate power for as late as possible. Because the site is clear nearly to the horizon in the summer afternoons until sunset, I may have to consider pointing the panels further west, and/or increasing spacing. The problem with the latter is that the array's ground area get huge.

I'll do some more modeling, but I thought I'd bring it up for others to contemplate or comment on.

Here's a repost of the site shading (the spike at 240-260 isn't real):

Can't respond at this time, more later, but I noticed the DR- SES rates and ties you included are the current rates and not the times or rates of the new DR - SES. Which did you use in coming up with your production $ numbers ?

[kb58]

My thinking and numbers have evolved a lot over the last few weeks, so if the numbers you're referring to we're more than 10 days old or so, they're out of date.

[J.P.M.]

My thinking and figure have evolved a lot over the last few weeks, so if the numbers you're referring to we're more than 10 days old or so, they're out of date.

They are not currently out of date, but soon will be. Part of the point of the exercise is that because the DR - SES rates are what all SDG & E NEM 2.0 users except probably those with EV's will likely default to, and, because the new DR- SES rates and times, which seem to be on a track that will make them effective ~~ 12/01/2017 or so, those using the old rates for planning purposes will be in for a surprise. It looked from your attachments that you were using the old rates. I know not to use the old rates. From your attachments, I wasn't sure you had the same knowledge.

[J.P.M.]

After seeing the self-shading thread I thought I'd check what SAM is telling me for my proposed setup.

As background, I plan the array to consist of two-panel subassemblies in landscape-mode, ground-mounted, with a total of 9-11 subassemblies oriented toward 220 degrees, each one directly behind the other.

I noticed that SAM's calculated row spacing never changes based on tilt angle. Maybe that's because it's already figured out what it needs to be to have zero shaded at all times of the year. I'd have thought though that if I set tilt angle to, oh, zero, row spacing should change to panel width but it does not. Maybe I input something wrong, but that led to the realization that having multiple rows in-line pointing south(ish) could be problematic. Until recently I thought I could model the latitude and desired tilt angle of 30 degrees and get a quick answer. Picturing how the sun moves through the sky though makes me think it's not that simple. That is, with row spacing based upon the tilt angle of 30 and winter solstice here being about 33 degrees, that seems to indicate that regardless of time of day, the array will always experience self-shading anytime the sun is lower than 33 degrees elevation. That's a problem, especially during the summer when the proposed TOU means it's really important to generate power for as late as possible. Because the site is clear nearly to the horizon in the summer afternoons until sunset, I may have to consider pointing the panels further west, and/or increasing spacing. The problem with the latter is that the array's footprint get huge.

I'll do some more modeling, but I thought I'd bring it up for others to contemplate or comment on.

Here's a repost of the site shading (the spike at 240-260 isn't real):

Row spacing as SAM does it has never been my best friend. The user inputs a Ground Coverage Ratio (GCR) and the row dimensions. That sets the row spacing. SAM then calculates the irradiance looses from the shading (differently for beam, and diffuse irradiance) that the row spacing will produces. I don't use it much and I could be wrong, but I don't believe SAM will calculate a row spacing that will result in zero self shading of more equator distant rows by rows nearer the equator, mainly because there will always be some shading of the diffuse radiation. Perhaps the GCR can be varied as part of a parametric study, but I'm not sure that SAM will do the GCR variation parametrically. See the help screens for gory details and some cryptic references to stuff done by Paul Gilman.

I've done a fair amount of row shading in the past, way before SAM. Dirty short cut method. Look up something called the "profile angle". Then, calculate the profile angle of the shadow cast by the top of the "front" row to the bottom of the "back row at 0800 solar time and 1600hrs. solar time Dec. 21. adjust the row spacing for the worse case. Once you know the height from the ground to the top of the front row and the height from the ground of the bottom of the back row, and the profile angle on that date at those times, calculate the profile angles and set the row spacing from that. The method is cheap and dirty if you learn/understand something about solar geometry.

Tan (Profile angle) = tan( solar altitude)/cos(solar azimuth angle - array azimuth angle)

See "Duffie & Beckman", Solar Engineering of Thermal Processes for angle conventions - be careful, they can be tricky. I use 2d ed., ISBN 0-471-51056-4

Hope that helps.

[sensij]

Row spacing as SAM does it has never been my best friend. The user inputs a Ground Coverage Ratio (GCR) and the row dimensions. That sets the row spacing. SAM then calculates the irradiance looses from the shading (differently for beam, and diffuse irradiance) that the row spacing will produces. I don't use it much and I could be wrong, but I don't believe SAM will calculate a row spacing that will result in zero self shading of more equator distant rows by rows nearer the equator, mainly because there will always be some shading of the diffuse radiation. Perhaps the GCR can be varied as part of a parametric study, but I'm not sure that SAM will do the GCR variation parametrically. See the help screens for gory details and some cryptic references to stuff done by Paul Gilman.

Hmmm.. GCR is one of the selectable inputs to be varied in the parametric simulation. Translating optimal GCR back into actual row spacing would take several reads through the help files, for me.

[J.P.M.]

Hmmm.. GCR is one of the selectable inputs to be varied in the parametric simulation. Translating optimal GCR back into actual row spacing would take several reads through the help files, for me.

Kind of what I figured but haven't done parametric in a while and never done self shading, so not sure. I appreciate & understand getting P.O.A. adjustments for the diffuse component is not as straightforward as getting a profile angle, but the way SAM does self shading just seems cryptic to me.

[kb58]

Using Google Sketch up, I modeled the panels and solar paths. Good thing I did because there is significant shading unless the rows are spread apart further or angled more westerly. Doubling up (two high x two wide) doesn't fix anything because then they're twice as tall and have to be spaced twice as far apart. As said previously, I'm trying to maintain a very low profile for aesthetic reasons.

During the winter, the sun's low enough, and the trees outside our yard tall enough, that I don't have to worry about self-shading below 30 degrees.

Oh decisions, decisions.

[J.P.M.]

During the winter, the sun's low enough, and the trees outside our yard tall enough, that I don't have to worry about self-shading below 30 degrees.

To be clear, are you saying that, in the winter, the array will be shaded by the trees, and so any self shading becomes a moot point ?

[kb58]

Yes, around the shortest days of the year, the sun's path skims the tops of the neighbor's trees, at around 30 deg elevation.

[J.P.M.]

Yes, around the shortest days of the year, the sun's path skims the tops of the neighbor's trees, at around 30 deg elevation.

Understood. Thank you.

[Mike90250]

Yes, around the shortest days of the year, the sun's path skims the tops of the neighbor's trees, at around 30 deg elevation.

That's going to GREATLY affect solar harvest (in a bad way)