Home Solar System

[GeraldTheBonzai]

Ooh a real engineering problem

There are so many factors - tilt angle, max wind speed, lift force, slip force....

Thinking.....

As a rough rule of thumb, at a 15deg tilt, i'd go for 100Kg per square meter of panel area. That's total weight - panel, frame plus ballast.

If it's high up or very exposed maybe go 120. If you change the pitch angle to 30 deg, increase everything by 50%.

Ballast needs to be uniformly distributed and consider a tether to the frame.

Superb answer. Thanks.
It's 2 x 8ftx4ft ( sorry) panels, so about 4 sq m
Mounted together at 20 degrees on 4x triangular frames, sat on a flat garage roof. Sealed rubber roof which I don't want to drill through.
I intend to thoroughly load it down with bottled water*, maybe between 100 and 200 kilo. The panels alone are 30 kilo each. Last time we used sandbags, but they droop and don't apply their full weight on the frame.

* Two birds with one stone. Free storage space

There are problems with what you are proposing. You have to consider two forces - the lift force (the wind turning your array into a kite) and the sliding force (the wind pushing on your array). The total weight needs to be sufficient to overcome both. Its the sliding force that I think might be the problem.

With a free standing array on the garage roof, the wind will be pushing on the sides of the array. You say the roof is sealed rubber. I suspect that in the summer, when the roof gets hot, that roofing will become softer. That, combined with thermal expansion of the array and frame, and I think you might find the array starts to wander around the roof. Not a lot but enough that it will apply a shear force to the roofing material. Over time, I think there is the risk that it could end up causing a rip in the roofing material. Think of a washing machine with an uneven load - it only needs a few repeated "shoves" for the washing machine to go for a wander. Those wind gust nudges will have a similar effect. Its going to be worse if the frame is on legs or beams, as now you will have point forces where the frame sites on the roof.

I know you don't want to drill through the roof, but I would seriously reconsider that. Or alternatively, consider an outrigger to the garage - attach something to the vertical walls on the outside, then run a beam across the top of the roof. Attach the frame to that. This will not only anchor the frame, but it will reduce some of the need for ballast - you also have to consider the amount of weight you are putting on the roof as a free standing load. Is it sufficiently strong enough to take the weight.

[jennyjj01]

Ooh a real engineering problem

Superb answer. Thanks.

There are problems with what you are proposing. You have to consider two forces - the lift force (the wind turning your array into a kite) and the sliding force (the wind pushing on your array). The total weight needs to be sufficient to overcome both. Its the sliding force that I think might be the problem.

With a free standing array on the garage roof, the wind will be pushing on the sides of the array. You say the roof is sealed rubber. I suspect that in the summer, when the roof gets hot, that roofing will become softer. That, combined with thermal expansion of the array and frame, and I think you might find the array starts to wander around the roof. Not a lot but enough that it will apply a shear force to the roofing material. Over time, I think there is the risk that it could end up causing a rip in the roofing material. Think of a washing machine with an uneven load - it only needs a few repeated "shoves" for the washing machine to go for a wander. Those wind gust nudges will have a similar effect. Its going to be worse if the frame is on legs or beams, as now you will have point forces where the frame sites on the roof.

I know you don't want to drill through the roof, but I would seriously reconsider that. Or alternatively, consider an outrigger to the garage - attach something to the vertical walls on the outside, then run a beam across the top of the roof. Attach the frame to that. This will not only anchor the frame, but it will reduce some of the need for ballast - you also have to consider the amount of weight you are putting on the roof as a free standing load. Is it sufficiently strong enough to take the weight.

Another great answer.
First thing..... The roof has substantial 9" joists so I reckon it can take plenty of weight. It also has a 3" plastic 'turrett' so the rig shouldn't slide off.

Maybe I'm over thinking this

4x 4" screws should kill the job stone dead and could probably be sealed with mastic?
The drifting about might cause blisters and tears. Perhaps if I spread the weight (a la snow shoes) or even put a rubber 'carpet' under it? As designed, the weight is borne by the narrow edges of 4x 8' lengths of 3x2

Hmmmmm thinking further, the more I spread the weight, the less friction it will have and the easier it might slide ( I was thinking something like some old laminate flooring as a sacrificial base )
Apologies for the imperial measures

[Yorkshire Andy]

Where I drilled my felt roof I put a rubber pad think tractor inner tube the size of the mounting plate the a square of flash band (bitumen backed tape) over the top and warmed it well before driving the screw through.. on a warm day you could see the bitumen ooze out as I cranked it down no leaks yet

[Yorkshire Andy]

Have you seen these Jenny

Put on roof full with ballast

And affix panel job done

https://www.cityplumbing.co.uk/p/fasten ... o_s=gplauk


Or

https://flatroofingsystems.co.uk/produc ... 4sEALw_wcB

[GeraldTheBonzai]

Might have a solution. Get some paving slabs. Create a paving slab base and anchor the frame to the paving slabs. The slabs can provide the ballast, spread the load over a much larger area, and due to the much greater surface area provide by the slabs to the roof, the friction will be very high. This should negate the side loads pushing the array.

If you find you need more ballast the you can free stand it on the slabs.

[jennyjj01]

Have you seen these Jenny

Put on roof full with ballast

And affix panel job done

https://www.cityplumbing.co.uk/p/fasten ... o_s=gplauk


Or

https://flatroofingsystems.co.uk/produc ... 4sEALw_wcB

I considered them, but nothing quite big enough. I want both panels on the same mounting.
Those links do give some help though: First linked to epdm** gasket washers at 18p *
https://www.cityplumbing.co.uk/p/fasten ... l/p/666739

The second linked to a data sheet of suggested ballast weights. Looks like I should be looking at 100kg per sq metre.

** that's what I was trying to remember. It's an epdm rubber roof
* I can stretch to a few times 18p

[jennyjj01]

Might have a solution. Get some paving slabs. Create a paving slab base and anchor the frame to the paving slabs. The slabs can provide the ballast, spread the load over a much larger area, and due to the much greater surface area provide by the slabs to the roof, the friction will be very high. This should negate the side loads pushing the array.

If you find you need more ballast the you can free stand it on the slabs.

Good idea. Concrete snow shoes

[jennyjj01]

Until last week, the Ecoflow Stream micro inverter could be had for <£100
Currently It's £129 on the website... Currently playing chicken with their shopping cart to see if they send me a discount code.

OK. Another day, another discount code. Currently £119 on the Bay.

I'm not biting yet '
I'll probably bite at £110. (tightwad)

Well, I Double Dog missed out.
As of this morning, with various discount codes, the Stream was £122 from EcoFlow.com or £129 from them via the river site or £134.10 from EcoFlo via the Bay

The Bay site shows 10 available but with estimated deliver as late as 5 May

I'll get my panels mounted ready.

Today the Ebay Listed price plunged to £97 ( Still 28 days lead time)
So I purchased one. Thanks YA for the heads up on the price drop.

[Yorkshire Andy]

OK. Another day, another discount code. Currently £119 on the Bay.

I'm not biting yet '
I'll probably bite at £110. (tightwad)

Well, I Double Dog missed out.
As of this morning, with various discount codes, the Stream was £122 from EcoFlow.com or £129 from them via the river site or £134.10 from EcoFlo via the Bay

The Bay site shows 10 available but with estimated deliver as late as 5 May

I'll get my panels mounted ready.

Today the Ebay Listed price plunged to £97 ( Still 28 days lead time)
So I purchased one. Thanks YA for the heads up on the price drop.

Your welcome

[GeraldTheBonzai]

After my little hot spot incident on my home solar rig, thought I would write up a little bit of info. Welcome to

Compressive Resistance
So, in my system, I found that I had a cable that was getting really hot. At the time I was pulling 600W through my inverter, at 12V, which translates to around 54A going through the DC side. On inspection, the connectors where the cables connected to a terminal had oxidised. They were tinned copper lugs, that on inspection looked "dull" and pitted. Moisture had got in and create a thin film of oxide on the surface. On cleaning them up down to bright copper, and refitting the cables, everything cooled down. So why was this tiny amount of oxidation causing so much heating?

When two metal surfaces touch to form a circuit, they don't touch uniformly. Instead, at a microscopic level, they form lots of point contacts. It's through these contacts that the electicity "flows". When you have good clean contacts, then the resistance between the surfaces should be very low - it should be micro Ohms. Lets call it 0.0005

You calculate the power using the formula P = (I)2 * Rc. Power = current squared * Compressive Resistance.
Doing the sums in my systems I should be (54*54) * 0.0005 = 2916*0.005 = 1.45W
So the is 1.45W being applied across that junction, which is being converted to heat - this is easily dissipated by the metal in the joint.

But what happens if there is oxidation, or the bolts have not been tightened down, so that there is a poor contact.
Assume the compressive resistance is now 0.02 ohms, which is still very little, but substantially more than a good connections. Lets do the sum again: (54*54) * 0.02 = 2916 * 0.02 = 53.8W which is the equivalent of holding a large soldering iron against the terminal. Continually.

Unfortunately, on 12V systems, the current is necessarily high - this is why you need fat cables. If I had been running at 24V, the impact would have been a lot less. Halving the current would reduce the heating effect by a factor of four (Amps squared remember). The bad contact would only have been generating 14.5W.

So if you have a home solar rig, that you have built yourself, and you are pulling a fair amount of power:

• Check your contacts on a regular basis - look for corrosion / oxidation
  Check that everything is tight
  Use the right size cables. On my 12V system I use 35mm welding cable, rated at 240A. Technically, this is pushing it at peak power, but I never run the rig at the full 2Kw.
  Use dielectric grease on the terminals - this is non conductive grease that prevents oxidation
  Use the right size lugs for your terminals - if you use the wrong size, you get point contact, not surface contact which leads to heating
  Consider going to 24V - which comes with its own problems, like battery balancing