Solar PV Ground Mount

[byQ]

Well, I'm trying to build a permanent type ground mount in my head. I started with treated lumber but changed my mind and decided to go with galvanized pipe. I'm putting the structure down a dead south facing hill. Two rows of 13 panels. So 45' x 10.5'.

I concrete in some 2 to 3 inch wide vertical pipe (say 3' below and 3' above ground). Two verticals, one towards the top of hill and one in middle of hill.

I do this 11 times - 22 vertical pipes, 4 feet apart. Now I connect the top of the hill pipe to the lower hill pipe with horizontal pipe. I drill holes between vertical and horizontal pipe and bolt them together. Now I run super struts (from Lowes) across this structure - like 4 or 5 horizontal runs. How to connect the super strut to the pipe?

One idea, I drill large enough holes in the pipe so the head of the bolt fits through the hole. Next I make a narrow hole so I can slide the bolt (like a key hole). Now I can bolt the super struts to the pipe. Next, I attach the panels to the struts. Main cost about 18 ten foot super struts, and panel connector pieces. Seems like it should work.

 

[byQ]

Well, I've changed my mind. Now I'm going to put those treated cherry poles (6" semi-round by 8') into concrete vertically and attach treated lumber (10') horizontally with bolts. Next, I'll run the uni-struts across the wood ground mount structure, and attach the panels. My reasoning is that I already have most of these things so I might as well use them. I know some barrier has to go between the treated lumber and the steel.

I have fifty-eight 235 W panels and 2 Sunny Boy 6 kw inverters. I have some 10-guage wire and some conduit to put it in. I was talking to a solar guy and he said it is probably best to have 3 sets of 9 panels each. The 9 panels are hooked in series, and the 3 sets of 9 panels are hooked in parallel. This makes a total voltage of a little over 100 coming out of the array. And this is done twice for each 6 kw inverter. The 2 inverters are hooked in parallel into the net meter. But, I'm left with 2 panels. I guess I need one more panel or I could sell them (or keep for spares).

 

[peakbagger]

Something does not sound right. If you can post the spec sheet on the panels it may be worth it to have someone else look at the calculations. Unless you have low voltage panels, something doesnt line up with your stated string voltage of 100 volts. Most panels for grid tie are nominal 24 volt (and on occasion 32 volts). Nine panels @24 volts is 216 volts which is on the low end of grid tied inverter's input voltage. The voltage in the panels vary with temperature with the voltage dropping below the rating in hot weather. That may get you down below the input voltage on the inverter. Even if it works, most folks go for higher string voltage to reduce line losses in the wiring. Your inverter may be rated for as high as 1000 volts but NEC for residential is limited to 600 volt max. A low voltage array is going to need larger wires than a higher voltage array. The trade off is your inverters may offer dual MPPT inputs so you may want to go with lower voltage string voltages so that you have break your array into 4 circuits. Most companies offer an automated string calculator for their panels, if not the calculations arent that difficult.

So assuming 24 volt nominal panels lets assume you have dual MPPT inverters so you have 4 MPPT circuits. So the logical approach is 58 panels divide by 4 = 14.5 panels per string. You cant cut a panel in half so that is 14 - 24 volt nominal panels in one string which adds up to 336 volts nominal which is a lot closer to the sweet spot on a grid tied inverter. This is where the temperature corrections calculations come in. You need to pick the lowest temperature you can expect in your area and use the voltage correction coefficient to calculate the peak voltage on the string and see if that exceed the input voltage on the inverter. There is normally a never exceed peak voltage that keeps the unit from bursting into smoke and a maximum MPPT tracking voltage that you should stay under. In theory your electrical inspector will ask to see the calculation to ensure that you never exceed 600 volts. That is 3290 watts per string (6580 watts total) which IMHO is bit high for a 6KW inverter but most designers will find that acceptable. Its just means that on rare occasions the inverter may "clip" in ideal conditions. I prefer to oversize inverters as I feel the longevity is a bit better.

Keeping a spare set of panels stashed away is a great idea, especially with ground mount panels. Lawn mowers and kids are not friendly to ground mounted panels. Unless you are in snow zone and mount the panels up quite high, the lower the panel the higher chance of it getting hit by foreign object. The other reason is on rare occasions panels can fail due to factory flaws and realistically within a couple of years of buying the panels, they will be out of production and despite a warranty if the company s still in business, they dont keep a warehouse full of old panels so you are out of luck if you need a new one. Far better to have one in the barn you can swap in. Keep in mind that if the back of the panels can be reached from the ground, you have to protect the wiring from people. Some inspectors may requires a fence with a locked gate, others may allow installing mesh on the back of the panel frames to prevent people from accessing the wiring without tools.

If you do have electrical inspection in your area or if your utility or insurance requires it (many do require a sign off to allow an interconnect), then your mounting system may have to be signed off by a Professional Engineer. Many roof mounting systems calculators can factor in local design conditions and they can generate a PE certification but if you are going homebuilt that could be more of an issue and more costly. You mention treated cherry poles. Unless they are treated by a commercial operation and grade stamped, a PE may not allow them in the design especially poured in concrete. I have couple of arrays built with unistrut, it works well but if I needed it stamped I would have stuck with commercial racking. The typical unistrut bolting arrangement where bolts are slipped in from the back frame flange into a strut nut looks neat but is painfully slow. Unless you go with aluminum strut, you will need to put in galvanic insulation on each and every spot where the panel frames contact the strut. I used handcut HDPE washers with slots in them so I could slide them in after I had the panel bolted in place. Do yourself a favor and use stainless fasteners around the array and make sure you put a dab of neverseeze on each and every thread. I am a fan of nylock nuts as it saves time and reduces the amount of hardware. Stainless fasteners bought in bulk from a marine store (Hamilton Marine ) are not that expensive but stay away from the local hardware stores as they mark up stainless big time. I can buy a box of 100 SS fasteners from Hamilton for 1/4 of what I would buy at a hardware store.

Make sure you understand how to reduce the potential for lightning strikes coming into the house from the array. No guarantees on a direct hit but make sure you follow best practices which means having a large gauge bare copper common ground wire running in the trench outside the conduit tying to the main house ground point. This will be tied to a local ground rod at the array and then the frame will be tied to the local ground rod. A good quality Surge Protection Device ( I recommend Midnight Solar SPDs)should be installed at the array to shunt any charge to the ground rather then back to the house via the DC wires. Lot to be said for a local shut off switch at the array, just make sure it rated for DC as many arent.

 

[byQ]

Thanks peakbagger, this is exactly the stuff I need to learn about. The panels are rated @

• Vmp 30.1
• Voc 36.8V

They must be the higher voltage types. The inverter max is 600 V. So 9 panels @ 30 V = 270 V, so it looks like the voltage needs to go up. I don't want to buy more wire since I have 400' of decent 10 guage wire. So I guess I need to keep an eye on the current.

I'm going to talk to an Electrical Engineer tomorrow. I called the County Building dept. and they don't regulate solar ground mount structures (yet). Probably due to low electricity rates and the fact that Idaho offers no support for solar (except the local power company offers net metering). So there is very little solar around this area.

 

[byQ]

What directions, how many, and what angles the panels are used at is interesting thing to know about. From my understanding inverters are best if you keep them working at 85% or higher their rating. It is better to oversize the amount of panels to the inverter than undersize. The inverter 'takes' power. If there is more then it wants it clips it - so it can never be overloaded.

If solar panels are 30% over the rated KW of the inverter in the long run you will get more energy. Because the inverters are usually not working at peak conditions due to time of day, clouds, etc.... So even if clipping occurs overall you get more energy. It seems to me you could even go over the 30% panel amount if you positioned your panels with some facing SE, most S, and some facing SW. So, you have the early sun panels working hard in the AM, the midday panels (most), and the late day panels working in the PM. This would fatten your bell curve. So your inverter would be running closer to its peak more often.

I have net metering so I guess the best thing for me is to figure out at what angle I will get the most energy over an entire year. I won't be able to move the panels. Well, I guess I could drill 2 holes in the vertical poles and move panels in Spring and Fall. And, since there is more energy to be had in the summer I should favor the higher sun angle. If you had a battery system you might want to favor winter when you needed more power.

As solar panels become cheaper it is feasible to over panel inverters especially if you have some of the panels more towards the east and west. This way you are causing the inverter to work earlier and later with more energy. Of course the inverter will wear out sooner because it is working more.

 

[peakbagger]

If you are not aware of PV Watts, its a handy tool. You can enter the angle for your area and it calculates the yearly output. What it does not do is calculate shading from rows of panels blocking other rows. I use a CAD program for that and use the Naval Observatory sun data for my location to check interference. I expect there is specialized software that could do it quicker but it works for me.

Be careful on your net metering, many utilities have a yearly true up date that can mess things up. My net metering has no true up date so I carry a surplus year to year with the option to sell once a year at some very low rate. Utilities with a true up date reset the account on certain date, some just grab the power others buy it from you at some low rate. If they do have a true up date then the system will look different (smaller) then an open system.

I do adjust my panel angles manually on two of my arrays a couple of times a year. I have a wall mount and pole mount so its easy. The normal optimal angle is your latitude but realize if you get snow, that is too shallow to shed snow. Solar developers tend to go with flatter angles to optimize summer production, they write off winter production as they cant deal with snow.

By the way I skipped the standard advice, before you buy anything get a copy of Solar Power your Home for Dummies.

 

[semipro]

I'm planning (seems like forever) a ground mount installation and am considering superstrut also. Just a caution, consider isolating the aluminum PV panel frames from the steel superstrut or you'll get corrosion.

 

[jebatty]

Solar developers tend to go with flatter angles to optimize summer production, they write off winter production as they cant deal with snow.

"Writing off winter production" might be good advice in some locales and especially for installs that do not allow for snow clearing. But ground mounts typically make snow clearing easy and fast, and winter production is far too great to sneeze at. Output yesterday, Feb 4, a clear, cold, very low humidity day in northern MN, with production enhanced by snow bounce, total 56 kWh for the day. Panels are at 35*, latitude is 47*N. My 4-yr average Feb production is 2/3 of July production, the highest average production month. Also for Feb 4, peak system production was 11,249 watts, 247 watts/panel, which nearly maxed out the 250 watt rated micro-inverters for my system.

 

[peakbagger]

I ran PV watts for Boise Idaho PV watts admittedly does not deal with snow ( I am not sure if any program does). This introduces error as snow cover means lost production although its offset a bit by an increase in insolation from sun reflecting off the snow.

Boise is 43 degrees latitude but I rounded it to 45 degrees

Using the 15 degree plus or minus rule of thumb

Annual output for 10 KW fixed mount at 4 different angles

Tilt Angle - 90 - 10043 KWH
Tilt Angle - 60 - 13998 KWH
Tilt Angle - 30 - 14957 KWH
Tilt Angle - 45 - 14881 KWH

Where horizontal is 0 degree tile and vertical is 90 degrees

Per PV watts even without snow factored in, the flatter angle panel puts out the most power on a year round basis. It there are multiple rows of panels one set of panels will shade the one behind it, the steeper the angle the better chance of shade. A developer wants to maximize the panel density so they try to minimize the space between panels which also leads to a flatter angle.

Snow covered panels reduce the annual output especially with string inverters. The problem comes in, where does the snow go after the snow slides or is raked off? I have to run my snowblower in front of my pole mount on occasion as the snow build up a pile level with the bottom of the array. My drive bys past a few large solar fields in Mass after a snowstorm is acres of white. Only once have I seen a crew attempting to clean between panel rows. A dedicated homeowner may be able to support it but developers look at bottom line and with FITs and SRECs it doesnt make a difference what time of the year the production occurs, its just the total. In some areas it may make a difference and it means the system may look different.

 

[jebatty]

... snow cover means lost production although its offset a bit by an increase in insolation from sun reflecting off the snow.

Insolation also increases in winter due to less humidity and less dust in the air, which directly increase insolation, and by colder temperatures, which increases electrical efficiency. Do you know whether PV Watts takes these into account? Since my system was installed in 2013, smoke from forest fires in the west and northwest US and Canada have been considerable and have resulted in a very noticeable haze in the summer sky, which likely has reduced summer insolation. Based on recent weather and climate patterns, forest fires may be a regular event each year. I doubt that PV Watts data accounts for forest fire smoke.

The problem comes in, where does the snow go after the snow slides or is raked off?

Yes, a consideration. Where I live we do not get a lot of snow. The bottom of my panels are about 4' off the ground, and now in the 5th winter, snow has never reached closer than 2' from the bottom of the panels, mostly because walking along the panels to pull the snow off packs the snow down.

 

[peakbagger]

PV Watts in based on long term measurement of sun insolation. So its looking at the average amount of sunlight received each day over many years. In theory that accounts for days with clouds which could be related to rain, snow and sleet. If there is seasonal fog it should be factored in. If there are seasonal dust storms over the long run that too should be accounted for. Obviously if climate change makes a long term change in the environment, the past will not predict the future. The overall amount of sun entering the atmosphere also changes , a year of volcanoes will put a lot of particulate in the upper atmosphere and will increase the atmospheric albedo meaning less light getting to a solar panel. There are also global variations,North America has reduced fine particulate and aerosol pollution so the amount of insolation has been increasing, China and the far east have higher particulate loading so they get less insolation. Reportedly when the US shut down air traffic after September 11th, the insolation in the region actually increased due to less impact from jet aircraft. In the long run increasing CO2 in the atmosphere will reduce the amount of insolation.

I dont know if PV watts takes into account the change in efficiency of panels due to temperature, its predictable so it could be factored in. PV Watts has an overall efficiency factor that probably covers a lot of things including the local microclimate. Generally the claim by many is the efficiency factor is too conservative. A lot of new solar owners dont factor in the "burn in" factor with PVs where the panels are more efficient when new and then relatively rapidly it drops off to the nameplate efficiency. Seasonal variation probably hides it but line up a good year with new panels followed by a bad weather year and I expect some folks may get a surprise. I see that in my annual production, it varies by year and i expect someone who keeps track of daily production over the long term could build up statistics pretty well. All they need are three years of data and then its time to hit the statistics books to fit the data to a bell curve.

 

[Karl_northwind]

For string and inverter selection use the tool that you find when you google blue oak PV. that's what plenty of the pros use. you can find nearly and inverter and panel and with your location it will help you protect your inverter from high voltage in cold weather full sun conditions, plan for hot panels in the summer and make sure they'll work OK with the low voltage you get in the summer.

PV watts is ok for long term production estimation, but not a string sizing tool.