I have a tile roof. It is widely recommended not to go on the roof; the tiles crack easily. How often does a PV array need to be cleaned? Can it be done from a ladder with a power wash unit? I live below the snow line. A ground mount is about 5000-7000 more costly.

For most residential applications you can just let God wash the PV panels for you, unless you experience overflights of flocks of crapping birds on a regular basis. Utility-scale PV arrays are typical on a periodic cleaning regimen, but it's measured in years, based on the cost/benefit of paying & equipping staff for the cleaning. It takes quite a bit of grime to lose even 10% of the total output, unless the gulls or geese are reducing the output of a single panel to the point that the inverter shuts the entire string down.

It's fine to hose 'em down, but pressure washing isn't such a great idea. Using chemical cleaning agents can even violate the warranty. Very soft- bristle brushes are OK for loosening up the bird crap after you've soaked it, but anything that can mechanically abrade or flex the panels put them at risk.

Not sure how you arrived at $5-7K for ground mounts? You can DIY some very nice ground mounts for under $500. Of course you need to have the real estate to do ground mounts.

http://www.builditsolar.com/Projects/PV/TimberMount/TimberMount.htm

http://www.builditsolar.com/Projects/PV/EnphasePV/Mounts.htm

I never clean my roof-mounted (6:12 pitch) PV panels, and I have not seen a significant degradation over time. I do drag a plastic bladed roof rake over them when they get a lot of snow, and I am anxious to get them generating again.

Thanks for your replies. Sailaway; I have several quotes for roof mounted systems and amongst them is a major roofing company nationwide. PD was very competitive for the roof mount system. They were the only ones to offer a ground mount quote; $6000. My county requires ground mount to be engineered with permits; I suspect this is a major cost.

Yes, licensed professional engineering is very very expensive. We charge $75/H and doing this for a very very complicated design could even take as long as 90 minutes.

Borst Engineering Consultation Rates

Have never cleaned mine.

Installed 1Q10.  Dallas environment (rare snow/ice, mostly dust, no birds).  Roof-mounted 8.1 KW.

From real time to 13 months of performance at http://www.welserver.com/WEL0043 .

Best regards,

Bill

Sail; I often think something shuld be easy, just take a few minutes. Then I try to hire it done and 5 different people all bid $500 to install it. I'm speaking of a 9 foot section of guardrail on a staircase. So while a $6000 bid to install ground mount seemed really high, no one was stepping forward to bid otherwise. Of course, I asked 6 people to bid and only 3 sent replies, so maybe people are happy to be out of work. Just wish you and I weren't supporting their lifestyle/health care/education.

“Just wish you and I weren't supporting their lifestyle/health care/education.”
Think about what you just said,
you solicited 5 bids for a simple handrail install. Think about that, 5 people had to look at the project. All of them spent some amount of time and effort examining and biding the job. And they got nothing out of it. All that gets accounted for as a job cost.
The contractor has to bid the job, do the job, insure the job, bill the job. Those are just the direct costs. Add in advertizing, healthcare (lets face it anyone in the building trades is at a high risk of injury and should have health care) When you account for all the overhead a contractor caries, you would be surprised at how much you have to charge to be profitable.
When you are doing small jobs, you either have to charge a lot, or you only do it when you have no other work and the opportunity cost is favorable.

Think through the steps in a ground mount PV system.
Locate where it is going- do a solar survey at that location to verify. Get with the installer to find out what he needs for a rack system
Cost out the job which takes 2-3 hours of my time Bid the Job knowing I have a 25% chance of getting the job.
In a lot of Jurisdictions you need a PE to stamp the plans so I am going to pay 300$ plus my time to get that.
Now you have to permit it, In West Hartford CT that would be a min of 3 hours of my time and at least 400$ by time you go through the process with the building officiall (BO) and the Fire Marshal, and maybe the zoning agent. Woops maybe you are too close to the side yard setback, now I have to have surveyor pull tape from the house corners and show on a plot plan that it meets zoning regs. 7-10 days later I have a permit in hand.
Now I would like to use helical piers for the foundation, but I know the building official will be suspicious of them because he doesn’t like them, so I will have to pour concrete.
So I hire my go to excavator with the bobcat mounted post hole digger and dig holes. This costs me 350$ I pray that the location of the septic and water main is correctly located on the plot plan so I don’t hit anything important. and schedule a inspection by the BO. I get a time frame of noon to 5 pm for the inspection so I sit there all afternoon doing paperwork for other jobs in my van. The BO shows up at 5:10, looks in the hole and says OK. I call the readimix concrete guy, but he already scheduled until the day after.
The concrete gets poured. I pay him 400$ for the readimix. I do a final check that the pin locations are ok.
I order the lumber and structural connectors for 2 days later so the concrete can cure enough. The SE went very conservative with connectors cause he is worried about uplift in a nor’easter cause it is on an exposed hilltop. So I spent an extra 300$ on hardware that I had not expected to, but hey no big deal.
I meet the lumber delivery vehicle and sign for it. I build the actual ground mount in about 6 hours- that part is easy.
I call for another inspection and spend 4 hours the next day waiting on my final inspection.
While I am waiting I reseed the lawn where the holes were dug
Now I have to get paid and I am in good shape. 6K was about right. Nothing went wrong with the job so it was profitable.

Now If I was doing this for myself, my actual out of pocket expenses would only be around 1200 $, the PE would do it for ½ cost as a favor, the excavator would do it for free because he is a riding buddy and I did a blower door test on his house last year gratis. See the difference?

I fully concur with your perspective on this Eric. DIY is as much about knowing who to barter with as in doing the actual work yourself... It is often preferable to barter with a known competent party than soliciting bids from unknown parties and hoping to pick a good one. Of course, you have to do what you have to do given your personal circumstances. However, you clearly didn’t barter well Eric, as we often PE stamp drawings for free for some of the people who support us :-)

We have a 5500W single axis tracking array that is ground mounted. Our county planning and building department require no permits for ground mounted arrays less that six feet tall, but permits and engineering are required for roof mounts. Odd how counties differ, and regarding cleaning, it's interesting to note the varied approaches. Ours is located close to a heavily traveled gravel driveway, so I have to clean it weekly during the summer due to dust accumulation, and that may have been avoided had we mounted it on our roof, but washing it off with a soft, hose connected brush takes only a couple minutes, so after five years, I still don't find it troublesome. Bird droppings and some small patches of mold or similar growth are problems and have to be dealt with regularly. I let the array go unwashed for two weeks one summer and then cleaned one string. By turning off the breakers to the three strings in turn, I found that it lost 3% generation in one week of dust collection, so I wash it once a week. I built the ground mount system, designed the tracking mechanism, and hired a local solar company to install the panels. The cost of the mounting system, which is far more elaborate than it needs to be, was about $2500, most of that going for aluminum channel, I beam and other metals.

You raise a very good point. One has to consider all the issues associated with where you locate the panels. Vehicle induced road dust is one that I had not considered. What tracking methodology did you select and what performance benefit would you estimate that you achieved?

When I proposed to the panel installer that I wanted to build a tracking array, he replied that I should just add extra panels for the same effect, but we had space constraints, and I held on to the opinion that I could make the array for less money than extra panels would cost.  My early calculations led me to believe I'd gain 25% with a single axis tracker.  I don't know how to check for accuracy, but I suspect my figures were on the high side.  Our 5.5kW array generates an average of 8900 kWh per year; it's in its fifth year and it's comprised of 24-230kW Sunpower panels.  I designed and built my own tracking mechanism.  You can view an old YouTube video of it that will show a bit of how it works if you Google "TrackingQT1.".  I used a CoolMuscle programmable motor to drive a linear actuator I bought on eBay and reconfigured to accept the motor.  I built a circuit using a Parallax microcontroller that stores hourly tilt angles for every day of the year, and with the help of my son and some generous online friends, I wrote the code to run it.  The controller wakes in the morning, checks with a few GPS satellites for an accurate reading of the date and time and begins the day's tracking sequence. At the end of the day it stores the array in a near vertical position and sleeps.  Since the array is long, I installed some motion dampers at the far ends to control wind induced motion and general bounce.  The linear actuator acts at the center of a 65' long span of panels.

I have two 4kw systems near a busy four lane road in a dusty city (Los Angeles).
The panels get truly filthy.
Had my panels cleaned recently after a year or two of accumulating crud, and it increased total daily power output by 10%.

I'm interested in the efficiency gain of a single N/S horizontal axis controller as compared to manual adjustment (based on the season). I suspect that there is little difference for places not near the equator. And specs make no sense if it isn't clearly spelled out what axis is moving, what direction it is and what is done with the other axis.

I expect that the best way to verify the value is to remove one panel from the array and put it in a fixed position.

Sounds like a nicely executed system mmoreland!

The performance gain going from fixed to single axis does vary by site (e.g., does the site receive full, unblocked solar exposure) and the time of the year (e.g., there is more gain during the longer solar day summer months). 20 -25% is the typically expected performance gain for a good site in the lower 48 states. Lots of good data and info on Build it Solar.com:

To Track...or Not Track

Thanks for the kind words, sailawayrb. I've spent quite a few hours reading on Build it Solar. I'll be building a second ground mount array on a parcel adjacent to ours with equal exposure to the sun. I intend to begin making it manually adjustable and will change it monthly to an angle that makes the most of the mid six hours of exposure of the middle day of the given month, and I should be able to reasonably accurately compare the two forms of tracking. Here's a little easier way to have a quick look at our array if anyone cares to: http://youtu.be/gjWTwdLGGhs How does one insert a link into one of these replies?

Mmoreland, the instructions for inserting an URL may be found here:

Link Instructions

BTW, I like your user name as it is so true. They are not making any more land...

Thanks...I'll try it.  Here goes:


http://youtu.be/gjWTwdLGGhs

As I said...very well executed...really a thing of beauty!