Disclaimer: I'm new to Solar, and most of what I know about it I've learned in the last week, including lots of good information from this site.

Situation: I'm a homeowner strongly considering a PV solar system of about 10 kW. I've already done much of the low-cost energy efficiency improvements to lower my bills and am down to either expensive new HVAC or solar (which seems to give me more bang for my buck). I've done enough cost-benefit analysis to assure myself it's a rational investment with about an 8 to 9 year payback. Now I'm moving toward some of the more technical aspects of just where to install the panels.

I have plenty of roofspace for multiple options:
- Approximately 92 square meters facing nearly due south (177 degrees) at 6 degree angle, with no obstructions.
- Approximately 126 square meters facing nearly due south (177 degrees) at 17 degree angle, with several vents to work around, and much of which is taken up by a solar pool heater which would have to be moved/replaced.
- Approximately 21 square meters facing nearly due west (267 degrees) at 17 degree angle, with no obstructions.

My research indicates it will cost me 4% of my generation ability to use the 6 degree S-facing rather than 17-degree, which I'm leaning toward to avoid the costs of relocating the pool heater. And I'm sure I can fit most (if not all) of a 10KW system on that south facing roof. However...
I've read that west-facing may make more sense for time-of-use rates, which I will have. (Example argument here, but it's in Australia so switch south for north: http://reneweconomy.com.au/2012/have-we-pointed-2gw-of-solar-pv-in-wrong-direction-17741) During 4 summer months, I pay/earn 4.5 times as much for power from 1pm-7pm PDT as the rest of the day.

My back-of-the-envelope calculation:
- West-facing generates 80% of energy as south-facing
- West-facing shifts peak of generation 2 hours later in the day, giving me 2 more afternoon hours than south-facing
- Assuming 7 hours per day in the summer (Las Vegas NV) centered around 1pm PDT (solar noon) I get 3.5 hours of peak energy with south-facing; I would get 5.5 hours of peak energy with west-facing. 5.5/3.5 is a 60% increase during 4 months of the year (at 4.5 times the payback) compared to a 20% reduction the rest of the year at lower rates.

Obviously it's a bit more complex than that, but the quick calculation indicates there may be some merit to the idea.

I can't (and don't want to) fit all my panels on the west-facing portion of the roof, but I'm curious if it would make sense to put a portion of them there, to increase my (valuable) peak generation.

Any thoughts from you experts out there?

Also, I can't figure out how to make paragraph breaks! Sorry for the run-on paragraph...

During 4 summer months, I pay/earn 4.5 times as much for power from 1pm-7pm PDT as the rest of the day.

That would certainly change the orientation economics. You might also look at some type of storage system(s) so that you never use any power during that period (even on cloudy days) and perhaps even pump more power into the grid during that period than the solar panels generate (ie, store morning energy).

Also, I can't figure out how to make paragraph breaks! Sorry for the run-on paragraph...

Try looking under Suggestions/Feedback for a thread titled "Formatting Paragraphs" and see if that information helps.

Posted By jonr on 17 Oct 2013 05:06 PM

During 4 summer months, I pay/earn 4.5 times as much for power from 1pm-7pm PDT as the rest of the day.

That would certainly change the orientation economics. You might also look at some type of storage system(s) so that you never use any power during that period (even on cloudy days) and perhaps even pump more power into the grid during that period than the solar panels generate (ie, store morning energy).

Unfortunately I can never sell back more than I use. Whatever I over-generate is a kWh credit on my bill (in the same category as I use it, either summer peak/summer off-peak/other)... I can store up and use indefinitely but will never make a profit on it. The point is, the same panel making less energy during a more valuable time of the day may be more cost effective for me (and help my local utility reduce peak loads rather than total energy generated).

No doubt that changing the orientation has merit and that it is worth running the numbers.

An interesting question is what is the best approach when that isn't enough to eliminate your net peak period usage for the summer. That's when you should consider things like 1) do nothing and pay for it, 2) add more panels, 3) store cold water, 4) store morning kwh in batteries and use it in the afternoon.

AquaDyne-

It sounds like you have done a careful analysis and come up with a reasonable conclusion. I would point out two things:

1. Solar PV panels have a fairly significant temperature coefficient, about -0.5% per deg. C, and in Las Vegas, the temperature swings during the day are large. That may add to the penalty for a western orientation relative to a southern orientation. I don't know if you factored that into your calculations.

2. If you mount the panels on two different sections of roof facing different directions, you will need a separate inverter for each section. Otherwise, shaded panels tend to short out the active panels.

I've had good results with an Enphase microinverter. But it does cost more per watt.

Posted By jonr on 18 Oct 2013 10:53 AM
I've had good results with an Enphase microinverter. But it does cost more per watt.

Yep, that is certainly one solution to the problem of panels facing in different directions.

Any thoughts

Simple calculus can probably supply an exact answer.

A real world solution is found if you observe what your maximum payback is. Once you have more generation capacity than maximum allowable payback, it's a simple matter of losing overall efficiency to get generation when it is most useful to you. In Las Vegas, that's probably when your A/C is running the hardest. I'm guessing that is the 1-7 pm time slice that has the highest rates and remittance.

I suppose that in some cases, adjustable racks would be worth it. You can change orientation with the season. On the other hand, panel prices keep coming down and so just using more of them looks attractive.

This is what you need for these "what if" propositions:

http://rredc.nrel.gov/solar/calculators/PVWATTS/version1/

Looks like a good scenario for microinverters (Enphase) or optimizers (SolarEdge). Especially for any partial shading issues you may have from those vents. You can customize a system nicely with those different roof facings.

I have a follow up question here. Is there room for a ground mount or pole system on your property? If so, does a tracking system make sense? I am asking the experts, as I am not familiar with the costs. But summer is when the variable pricing kicks in. A whole heck of a lot of extra revenue is generated if the panels face directly at the sun from 7 AM to 7PM, especially during peak hours. Would the extra revenue offset the cost of the trackers?

Posted By McFish on 12 Jan 2014 01:27 PM
I have a follow up question here. Is there room for a ground mount or pole system on your property? If so, does a tracking system make sense? I am asking the experts, as I am not familiar with the costs. But summer is when the variable pricing kicks in. A whole heck of a lot of extra revenue is generated if the panels face directly at the sun from 7 AM to 7PM, especially during peak hours. Would the extra revenue offset the cost of the trackers?

As a general statement, the cost of the PV panels themselves has gotten so cheap that it is usually more practical to buy more panels than to invest in a tracking system. I suppose there could be exceptions.

Lee; I had heard that previously. I'm curious though; just as prices have come down much in recent year, maybe the math as to relatice merit of tracking changes with the variable pricing of time of generation? I'm just trying to update my assumptions and input new data. I would try to do the math myself, but i lack the understanding of how to do it.

NREL has updated their production estimating tool. So play with this tool to help narrow down the production numbers for your different senarios.

http://pvwatts.nrel.gov

That being said. Here in Ohio, admittedly a very different climate, we recommend more solar modules rather than a tracker. As has been said already the price is sooo much lower than it used to be for the solar modules that trackers almost never make sense. Especially if you have good roof space. As trackers are typically done as a pole mounted system and the added cost of excavation, foundations, heavy duty galvanized steel pipe, trenching, AND tracking motors/sensors that require maintenance... I think you get the idea here.

I'm not sure what you utility bill tells you in terms of breaking down your peak kWh vs off-peak on a monthly basis, but once you run your PV watts simulation, just below the monthly summary there is an option to download the simulated hourly production data. That way you can get a pretty good idea of how many kWh you get during peak rate times.

Also SMA America has an online tool called sunny design. If you can learn to use it (not too difficult in my opinion) you can put in more detailed information on your utility rates, I believe they allow you to set time of use rate info. Then you may get an even better idea of what your options are. With sunny design you have to use their inverters as the design basis. But in reality it will not effect the production numbers by more than a few %, which is less than the spread from pv watts. (Pv watts uses an average of 30yrs weather data to create its simulation) so any given year can be up to 10% off the simulated #)

I realized that I was wrong about the sunny design software allowing time of use info.

NREL has a piece of software called the System Advisory Model (SAM) that does let you specify time of use utility data.

It is a more involved piece of software, but still within the competent DIY'ers ability.