Choosing a small solar panel

[NW-Bound]

For a place where electric cost is as high as $0.35/kWhr, it is true that the $1/watt panel cost is a godsend. I can see one would want to cover his roof with cheap panels, and screw the low efficiency of the panels on the east-facing roof.

But, but, but Hawaii is at a lower latitude. In the mainland, I have seen terrible installations that would really hurt the payback.

It is also true that when panels are this cheap, it is absolutely not cost-effective to have a mechanical system that automatically tilts to track the sun through the day. It may not even be worthwhile to lower or raise the panels for the seasonal sun angle.

I just looked up my electric costs. Over the last 12-month period, I use a total of 23,162 kWhr, and paid a total of $2370. So, it works out to $0.10/kWhr, including all taxes and service charges.

As a rough back-of-envelope calculation, I use an insolation calculator that shows that, when averaged out over a year, my location provides the equivalent of 6 hrs/day of maximum solar energy. The above is for a panel that is fix-mounted, and tilted to the same angle as the latitude. By the way, the calculator uses actual measured data to reflect average cloudiness of a location. My location in the SW has the best insolation for the continental US, period.

So, a $1 worth of panel would give me an average of 1W * 6 hrs * 365 days = 2.190 kWhr/year. That's $0.22 worth of electricity/yr, or 22% return. However, it does not include the installation cost, nor the cost of the grid-tie inverter.

The true payback would be a bit more difficult to compute, because I pay by a demand-rate schedule, which is up to more than $0.20/kWhr in the late afternoon in the summer. This is the time where A/Cs are cranking all over town, but also when one would orient his solar panel for maximum output. So, the payback would be even faster.

Being a geek that I am, I would even spend $10K-$20K even without the subsidy to build something to play with, even if I just break even. If I had a large lot so that the array could be built on the ground for maximum orientation, ease of maintenance, and no risk to the roof structure, I would do it. But with my suburban home with a bad orientation of the roof surfaces, I just cannot see it as something worthwhile.

I do have more land, actually a nice south-facing hillside up in my boonies home, but I hardly use any electricity up there to make it worthwhile. Thieves might also cart off all this stuff while I am not up there.

In short, though I just want some excuses to play with this stuff, I do not see myself going further than solar for the MH.

 

[ERD50]

But, but, but Hawaii is at a lower latitude. In the mainland, I have seen terrible installations that would really hurt the payback. ...

I was curious, so I went back to the calculator I used and plugged in the Honolulu zip code. Adjusting to get the same 3656 watt panels, a Hawaii location only pumps out about 19% more annual KWh compared to those panels in N IL. I thought it would be a bigger delta.

The 2.5x $/KWh rate is a much bigger factor.

-ERD50

 

[ratto]

However, it does not include the installation cost, nor the cost of the grid-tie inverter.

NW-Bound is spot on. A grid-tie or off-grid inverter will be expensive, especially the pure sine wave models. For an off-grid setup, this does not count the super heavy deep-cycle battery bank which you might need help from forklift to move them around. Even for the grid-tie setup, different utility companies will have different rate policies at different time. So the excess electricity generated by solar panel might not even get a fair buy-back price. Without heavy subsidy by tax dollars, it will be difficult for the small scale homestead setup to ever break even. Even if and when it does, the panels and/or battery bank will need replacement as well by then.

The installation cost for any grid-tie device, like inverter or transfer switch will not be cheap either. This is kind of like buying a backup generator for home will not cost you a fortune, but hiring an electrician to install and hook it up with a multi-circuits transfer switch might cost you an arm.

 

[Nords]

In short, though I just want some excuses to play with this stuff, I do not see myself going further than solar for the MH.

NW-Bound is spot on. A grid-tie or off-grid inverter will be expensive, especially the pure sine wave models.
The installation cost for any grid-tie device, like inverter or transfer switch will not be cheap either.

Some comments:
1. Tilting the solar panel to match the latitude? Isn't that a sine function, and even if you're 20 degrees off aren't you still pretty close to 100%? Again, I'm not sure that the efficiency rolls off that quickly for most factors.
2. I don't know how long the payback periods are, but it'd be worth including a reasonable 3% inflation and then an unreasonable 6% in those calculations. It'd give you an upper/lower bound and help consider the impact of $200 oil.
3. Photovoltaics are really really cheap today. They have a payback that could be deemed not unreasonable. Do you see them getting cheaper in the future? In other words, what price per watt do you need?
4. In late 2004 we paid $2500 (full retail) for a 3000-watt grid-tied MPPT inverter. I've virtually ignored it for over seven years. I haven't checked prices, but it seems difficult to believe that today's inverters would cost more money or be less reliable.
5. When we built our array in late 2004, we did all the mechanical work. The electricians installed the inverter, hooked up the wires, and did all the paperwork. The total cost of their labor was $750, and over half of that was to pay the runner to stand in line at the city permit desk.
6. Most grid-tied inverters for photovoltaic systems depend on the utility's input voltage to function. If the utility goes down then the inverter goes down too. You buy a grid-tie system for cheap power, not 100% up time.
7. Are you including tax credits in your calculations? Our numbers worked for us at panel prices around $4/watt, and our system paid itself off in about six years after the state/federal tax credits.

 

[NW-Bound]

Regarding the effects of the panel orientation, the maximum power is of course obtained when the sun ray hits the panel at 90 degrees, and it falls off as the sine of that angle (or the cosine of the off-axis angle). And it adds up to a lot more than people imagine.

The earth spin-axis is tilted at 23.5 deg from its orbital plane. This coupled with the higher latitude of Northern America means that the sun is very low in the winter. For example, Northen Illinois where our friend ERD50 lives is at 42 deg latitude, where at the winter soltice, the sun is only 24.5 deg above the horizon (90 - 23.5 - 42 = 24.5).

In fact, the Web site that I use confirms that in January, for all of the upper 49 States of the US, a vertical south facing panel would collect more power than a horizontal plate.

If one does not care to adjust the panel angle to track the seasonal sun, tilting it up at an angle equal to latitude would give the best overall annual power output.

To optimize the power for the winter, it is recommended that the tilt angle is set to "latitude plus 15 degrees". To optimize power for the summer, it is recommended that the tilt angle is set to "latitude minus 15 degrees".

Source: U.S. Solar Radiation Resource Maps

The current low price of the panels does make it very tempting, as I explained. I have just priced the cost of grid-tie inverters, and found that it is as low as $0.64/watt for a conventional design, and $1/watt for the new microinverter type. The latter is a new distributed design that uses one small inverter per panel, and allows one to add incremental capacities as needed.

Yes, as I said, the price is at the point where it is economically feasible even without tax incentives, particularly if the home owner installs it himself. It is not that hard to mount and wire up these panels. The problem is that I have a tile roof. It also does not have the right orientation for a roof-flat-mounted panel to work optimally. I also have problems with shading.

Regarding shading, as I said, I have seen terrible shading problems with installations around my neighborhood. If 5% of a panel is shaded, one would lose 5% of power, right? Wrong!

A panel is made of cells connected in series. Each cell contributes a small voltage of 0.4V when illuminated. But a cell getting less light than the rest in the string acts like a Zener diode with a breakdown voltage from more than 12V down to 5V (the lower the better). Dark cells not only do not contribute, but block and absorb 10x to 30x the power of a producing cell (the ratio of the 12V or 5V to the 0.4V). The cells that are dark may even fail due to heat stress!

Any partial shading is a big no-no!

Note that it is not feasible to add bypass diodes at each cell, which would theoretically solve the partial shading problem. No panel maker does that. I guess the multitude of diodes would even add to the failure modes of the panel.

 

[Nords]

I can tell you that our PV production varies about 25% between January and July. We have some shading problems in January for the last 90 minutes of the day. But I don't bother to change the panel angle or do anything else to improve their production. It's good enough, and the benefits of having them on the roof have been greater than the drawbacks of sub-optimal performance.

I guess the question is whether you think photovoltaics has any payback on your rooftop.

But judging from your laser-keen focus on optimum power-producing characteristics, I think the losses would annoy you more than the gains would benefit you.

For any other PV wannabes cursing their stylish tile roof, try this mount:
Quick Mount PV - waterproof Universal Tile Mount for curved and flat tile roofs, steep slope
I have the "classic comp" version, and they're easy to install.

 

[ratto]

Nords, it's great that you could do most of the work by yourself to save a bundle. Before I put a transfer switch (TS) panel in side by side with my main load center about 2 years ago, I got 4 quotes ranging $1.2-1.7k (just labor & minimum parts only, no TS). I could tell their strong displeasure immediately after I told them that I already bought a TS. In the end, I did it by myself and passed code inspections. The total cost for me was only about $120 in about half day effort. The permit and inspector cost was about $170.

I did a quick Google search on a 3000-watt grid-tied MPPT inverter. You're right that its price hasn't been changed much since you bought yours.

6. Most grid-tied inverters for photovoltaic systems depend on the utility's input voltage to function. If the utility goes down then the inverter goes down too.

This makes a perfect sense to me now. Previously I was wondering PV cell voltage fluctuates based on input light intensity and how a grid-tie type inverter handles that. Even if the whole PV system voltage could be kept relatively stable via a heavy duty regulator, the AC output from a grid-tie inverter should be kept in sync with the utility grid AC phase to avoid causing interference or cancelling each other out. Also, it's dangerous for the linemen if a grid-tie inverter keeps working during a power outage. Now I see these practicality and safety issues have been addressed.

 

[ERD50]

... The current low price of the panels does make it very tempting, as I explained. I have just priced the cost of grid-tie inverters, and found that it is as low as $0.64/watt for a conventional design, ...

I hadn't thought about $0.64/watt inverter cost for $1/watt panels. That's a considerable adder. And that is pretty mature technology, I don't expect it will drop too fast.

It got me thinking if a big industrial install couldn't just use the DC directly. Imagine a bunch of loads that could be instantly/automatically switched from AC to DC (LED lighting? Universal motors?). So let's say the morning sun hits the panel, the voltage pops up to say 120V, so the first small load kicks over from AC to solar DC, dropping the voltage to maybe 110V. That enables the second (slightly larger) load for switching, and as the sun comes on stronger and the voltage pops back up to 120V again, load # 2 kicks in, and on, and on. Simple on/off switches, lower loss than inverters; the loading provides the regulation.

Regarding shading, as I said, I have seen terrible shading problems with installations around my neighborhood. If 5% of a panel is shaded, one would lose 5% of power, right? Wrong!

A panel is made of cells connected in series. Each cell contributes a small voltage of 0.4V when illuminated. But a cell getting less light than the rest in the string acts like a Zener diode with a breakdown voltage from more than 12V down to 5V (the lower the better). Dark cells not only do not contribute, but block and absorb 10x to 30x the power of a producing cell (the ratio of the 12V or 5V to the 0.4V). The cells that are dark may even fail due to heat stress!

I was aware that any shading dropped the power on the whole series string, but hadn't thought about how that could back-bias and possibly damage the shaded cells. Same thing with a weak battery cell in series gets back-biased. Interesting.


-ERD50