To answer your question: 

Potentially you might shorten the life of the panels by keeping them off.

Why?  Look up the arrhenius equation. The hotter something is, the quicker it degrades. 

A solar panel is something like 20% efficient. That means that energy is pulled away from the panel. Or in other words, a 300 watt panel that's off will effectively be heated by an additional 300 watts vs one that's operating. 

I've never measured it, but I'd bet it's probably worth another +20degf vs a running panel. 

Is that meaningful in aging terms? Maybe? Rule of thumb is chemical reaction rates double for every 10degC of temperature increase. This means thermally induced aging will therefore increase.

As a percentage of the damage over the life of the solar panel, I'm not sure where thermal damage vs thermal cycling vs vibration vs UV damage stand. But it's probably a decent percentage of effect on total life.

That's not true at all.  My 9.5kw array on the worst rain + cloudy day so far (since Feb), generated about 16kwh, and right now is averaging about 60kwh per day / 1.8Mwh/mth.  

 Say you double that to a 20kw array, that gives you 32kwh on a really bad day. 

That would still generate 960kwh, which is way above the average household use.  20kw*$2.5 is $50k. 

Add 40kwh of batteries for $25k, and you're still at only $75k, half of your lower estimate. 

 More sensibly, you could probably do it for $50k if you planned on a backup generator for the worst of the winter storms.

Gotcha. Yeah, I think most inverters are able to prioritize like that. Basically, if there is available solar, prioritize charging batteries before export.

You can get more fancy if you have time-of-use rates:

For example, Sol-Ark: https://diysolarforum.com/threads/time-of-use-settings.55624/

The term 'hybrid' normally is used in comparison to 'grid-tied' or 'off-grid' inverters. A hybrid can become either grid-tied or stay off grid completely.

Enphase is not really a 'hybrid' inverter in the traditional sense. But I guess functionally it can be if you have their batteries + system controller box.

If you don't care about off-grid usage, you don't need a hybrid inverter, just a grid-tied one. My understanding is hybrid inverters tend to be more expensive than just straight grid tie.

Stuff like Enphase can be set into self-consumption / zero-export mode. In fact, many (most?) inverters have a function like this. Also, most (all?) battery charger / MPPT / inverter combo units can prioritize charging the battery vs exporting.

You'll need to read up in the manual / app to see if you can do that. I've heard of people trying this with some success. Still feels a bit dicey though. To me, it's a big unknown as to how their MPPT algo works, and whether it does anything like a full sweep to Isc.

Do you have a battery bank? Why not just use an 12-48V charger on that battery bank in parallel? If the demand is more than solar, the AIO will start drawing from the batteries, which in turn will start drawing through the external charger.

That's really cool!

It's really interesting to see the angled cells. And the fact they use the whole disk rather than chopping it up. I guess wafers were a lot more expensive back then, and the space inefficiency vs. cost / waste generated by dicing them up didn't make sense.

400-ish is the standard now.

Big panels mean lower incremental costs (each panel has a rapid shutdown module, for example, a set of 4 mounting bolts, etc). If you're doing microinverters, same story. It's probably cheaper to do 2x 400 watt panels than 3x 280W panels for example, since 3x low power microinverters might be (say) $150*3 = $450 vs $190*2 = $380 for the higher power version.

On the other hand, if you have a complicated roof, it might mean you can't fill it as easily, and may have to leave bigger gaps.

Grid-tied solar acts as a supplement to the grid. If you're using less power than you're producing, it's going to the grid or your batteries.

Right now, going completely off-grid is probably not going to make sense financially. You have to design a system big enough to carry you through the worst cloudy winter day, for example. The rest of the time, that energy potential is just wasted.

Sizing batteries and solar depend on many things -- your usage, your roof, your location on earth, etc.

You can DIY, yes, depending on state / country.

I had actually been on my way to pickup some stuff from that HD when it burned. By the time I got there, the flames from the roof looked like they were a wall of flame 20ft high. It was an insane spectacle to see. Not exactly a small building.

What's strange is that this isn't even the first time I've gone to pickup parts and discovered the store burned down. Central Computers was the other store I showed up to and found a smoking pile of embers. It's not really something you expect to have to check for before you head out to buy something...

PVWatts allows you to download the hourly data. Throw that into Excel and have at it.

Edit: took a look for Bangkok. It's kinda of wild how much it varies day to day. I guess rainy season / haze party a big part. To answer your question, it looks like you could see anything from like 3.6kw to 1.4kw at noon, depending on conditions.

He's interested in the shaded roof measurement, not so much the panel.

While the emissivity of shingles might not be exactly correct vs what his thermal camera has as a default, I'm betting the relative delta is going to be pretty much correct (shaded vs unshaded). E.g, it might be in reality his two measurements should have been 95 and 115 (20 deg delta), rather than (say) 84 and 104 due to an incorrect emissivity setting, but I'd bet the delta is still pretty close.

I did a similar measurement on the inside of my attic. https://imgur.com/wDOglMr 

I saw something like a 15-18 deg difference as well. So it looks like we're seeing similar numbers.

For my array (30 panels), I figure the heat load difference is the equivalent of running something like 4 space heaters in my attic (3.3kw)! Definitely debatable, and something I wish I had actual attic temperatures for before/after panels.

https://diysolarforum.com/threads/solar-panels-as-attic-insulation-check-my-math.83523/

In hindight, it's actually a smaller area than I was thinking, once I remove the tree/ planter beds, walkways etc. Realistically, close to half that - 2000 ish sq ft. edited the post.

Those are small panels (150W ish?). It's probably fine.

I'm just curious -- did you read the install manual for the brackets? I feel like that would be a mandatory thing to do before just yolo-ing a novel project with the potential to cause leaks / damage...

Those wires should ideally be secured, and put in conduit. Some are rated for UV exposure. But lying on the roof deck isn't great -- they may abrade over time as the wind moves them around.

Far from me to question the choice of materials (hah), but: Why? Tiny panels with inefficient brackets (look at the wasted space between panels) don't really seem like a good use of time or roof space.

A single 300-400 panel would smoke those two, and you'd only need a single set of brackets, for example.

+1. What he said.

Enphase makes each panel + inverter its own independent solar generator. They don't care about their neighbors. This is unlike a string inverter, in which they're all daisy chained together.

If I were you, I'd expand with Enphase, rather than a separate string inverter, unless there's a compelling reason not to. You definitely could use a string inverter. Again, it would look like it's own independent solar generating system. It's just another generator attached to the grid. Heck, minus your main breaker, it could be at your neighbors place for all that the enphase system cares.

The only potential issue might be if it also has some sort of PLC communication system to the RSD modules. I don't know if you get interference between different vendors, but I assume given the small amounts of data, it's probably insignificant if there is.

Incorrect, incorrect.

His statement is accurate. Regardless of clipping, a panel will produce less the hotter it gets.

If you're clipping during cold but bright days, you'll generate less -- and therefore clip less -- during hot, bright days).

The top poster has a real world example of that.

I mean, you could get a flat rate out of someone in theory. It's just that they'll pad that flat rate to the point that you'd effectively be paying for a max-complexity board.

If you're willing to spend a flat rate of $20/sq inch, I can shop that out to fabs for you...

You can potentially estimate their costs for a given board. Some of the online vendors have different options you can adjust, and you can see the price delta right there. That gives you some idea of what it costs them.

We had a good idea at the BigCorps(TM) that I worked at, given the number of designs we'd made, but there was also always negotiation around quantity. When you're ordering 1 million of something, you get good prices... :)

But think about it. A single layer board without soldermask or vias / drill holes / plating is like a single step in the process.

By comparison, a 12 layer, ELIC / microvia, 4 mil trace, hard gold plated contact / ENEPIG plated with impedance control is many, many, many more process steps. Each with its own chance of failure / yield issues / chemical costs. Panelization efficiency is another big one.

Board area is like one characteristic out of 20+. If you knew all the features you want, you could probably extrapolate to a cost per sq-in, within reason. So if you have a design, and you want to answer the question "If I spend 8 hours to shrink it by 10%, is that worth my time?", this might make sense. In my case, one design I worked on I decided to run the numbers on shaving the circular PCB into one that looks more like a rounded square (more efficient panelization). I saved probably about 5% of the area (about $0.05/pcb). Times 1 million, and that works out to a decent chunk of cash :)

Yeah. Suddenly that grim sense of satisfaction at totaling someone's car for damaging your property probably doesn't quite seem worth it anymore when they're badly hurt / dead.

Your breakaway system is frankly the best solution. Not a big deal to reset, and you don't have to deal with the potential for really hurting (an admittedly stupid) person.

In the event that the concern is for cars continuing on into the house, building a solid concrete wall (freeway median style, but maybe with some nice rock veneer) is probably a better idea. They'll either slide along it, or if they do hit it head on, at least its spread across the entire front of the car.

That still seems like a pitiful amount for a really high risk, visible crime. At any point, someone in a vehicle with like 6 cameras could show up to use a station, and it'd be really obvious what you were doing.

From what I can tell, a catalytic converter goes for like $75 scrap. For the same work as a single $10 supercharger cable, you get 7x the value. And you get to choose vehicles that are out of the way and secluded. And a catalytic converter is like 1/10 the volume of the equivalent value in cables...

You can definitely DIY them. I've talked about this before, but I found someone to do them for $250, and frankly it was well worth the money.

The thing you're getting is someone knowledgeable about all the standards and requirements. It would suck to make some sort of mistake, and have to redo work because of it. Honestly, I think it's easily worth the money.

You can definitely do it. I'm in the Bay Area and did it myself. If you look at my post history, you can see my writeup. I did a 9.5kW system, enphase based. I saved myself probably $10-15k.

I'd recommend hiring someone to do the plans -- I think I paid $250. Well worth it. I also paid someone $150 to do the PG&E permit stuff, but I don't think you'd need to bother with that. I was against the wall for NEM 2.0, so I had to make sure everything was 100% correct on the first try.

I did all the electrical work myself, all the lugging up of panels, conduit, etc. Lots of work, and there's some stuff I wish I did better. But all in all, pretty rewarding. And you can break it up in multiple weekends. Racking one week, electrical the next, etc.

Passed the city inspection on the first attempt :)

My recommendation: Do an enphase system. Yes, it's more expensive, but It's much more plug & play, and I think the AHJ like it since there's little room for weirdness.

Think of how you're going to spin "sorry your kid died, but I didn't want to leave a screw hole in the frame of my window"

RTFM. I can guarantee all the questions you could have would be answered by doing so.

Call it 5000 kg.

Google says electricity generation average in US is about 0.5kg CO2/kwh. Call that a 4kW system, per year it's probably 6500kwhr. That means it saves 3250 kg/yr.

So, call it 2 years, and it's paid back for the sequestered carbon in that tree. And that's CO2 that will never be released, rather than that tree which would at some point die / rot / convert back at least partially into CO2.