Manually adjustable tilt racking for large panels?

[PreppenWolf]

I am in NH and tilting is a requirement. The summer optimal angle is 15deg, the winter angle is 67deg.

I compromised and made custom frames with foldable legs. My frames can be set to 57deg or 25deg which gets me within 3-5% of maximum possible in any given month and I only have to change the angle once in spring (lay them down) and once in fall (extend the legs).

Frames are built from pressure treated (ground contact) 2x4 lumbar with steel corner brackets. The frames and bases (to keep them out of snow) are planted with 16" spiral ground anchors.

 

[Warpspeed]

​

Picking the optimal tilt for each month results in a hair over 5% gain over the course of the year. This would require 8 adjustments during the year. If you allow only 4 adjustments per year this drops a bit to 4.7%.

So, is it worth it? That's up to each individual. If it's a straight economic analysis and you're grid-tied you just balance the extra production against the cost of an adjustable mount. If you're offgrid, the increased winter production may be the decider (or the extra summer production if you live in area requiring air conditioning).

And that sums it up rather nicely.
If you added just one more panel (and forget about tilting) you might easily gain back that missing five percent, and save yourself a lot of messing about constantly adjusting the thing.

Or you can place an each way bet. Mount half your panels for summer tilt, and half for winter tilt.
Its a pretty crappy picture, but this is how I did my north facing panels on a flat roof.

 

[TOU47]

I am in NH and tilting is a requirement. The summer optimal angle is 15deg, the winter angle is 67deg.

I compromised and made custom frames with foldable legs. My frames can be set to 57deg or 25deg which gets me within 3-5% of maximum possible in any given month and I only have to change the angle once in spring (lay them down) and once in fall (extend the legs).

Frames are built from pressure treated (ground contact) 2x4 lumbar with steel corner brackets. The frames and bases (to keep them out of snow) are planted with 16" spiral ground anchors.

View attachment 123426

View attachment 123427

VERY nice & look to be cost effective.

Would love to see your plans & material list. Thank you in advance!

 

[acdoctor]

Now I'm confused. You wrote

which suggests that the output dropped after you changed the tilt. Did you decrease the tilt, or is this just a typo?

I had a discussion similar to this with with someone and checked output at several different tilt angles during a 15 minute period. Like the output at my latitude as a angle.

 

[Llaves]

Frames are built from pressure treated (ground contact) 2x4 lumbar with steel corner brackets. The frames and bases (to keep them out of snow) are planted with 16" spiral ground anchors.

have you done any testing or calculations to verify these will hold up to wind loads? When I look up 16" anchors I'm seeing load ratings of ~400 lbs, which is presumably dependent on soil type. My Osprey mount anchors are load-tested to 1600 pounds during installation. There are 12 anchors for a 36 panel installation.

 

[PreppenWolf]

VERY nice & look to be cost effective.

Would love to see your plans & material list. Thank you in advance!

Per unit
(7) 2x4x8ft
(2) 2x4x10ft
(2) 2x4x12ft
(1) 2x12x8ft
(4) hinges
(2) clasps
(5) joist plates
(4) hurricane ties
(6) L Brackets
(12) solar brackets - 4 per panel
(5) 16" trampoline anchors.
1-1/4, 2", and 3-1/2" deck screws.

We built 14 frames for our 42 panels. I bought the 2x12s in longer lengths because of that and cut them to size. The lumber was $1600 total. I'll have to add up the hardware. Took my son, wife and I a full day to precut all the lumber, four days to build them (two full weekends) and two days to mount the panels and anchor.

I can send you pics of the process if you like and dig up the Amazon hardware.

 

[PreppenWolf]

have you done any testing or calculations to verify these will hold up to wind loads? When I look up 16" anchors I'm seeing load ratings of ~400 lbs, which is presumably dependent on soil type. My Osprey mount anchors are load-tested to 1600 pounds during installation. There are 12 anchors for a 36 panel installation.

We've already had two wind storms of 50mph before the ground was frozen. They don't even wobble.

 

[TOU47]

Per unit
(7) 2x4x8ft
(2) 2x4x10ft
(2) 2x4x12ft
(1) 2x12x8ft
(4) hinges
(2) clasps
(5) joist plates
(4) hurricane ties
(6) L Brackets
(12) solar brackets - 4 per panel
(5) 16" trampoline anchors.
1-1/4, 2", and 3-1/2" deck screws.

We built 14 frames for our 42 panels. I bought the 2x12s in longer lengths because of that and cut them to size. The lumber was $1600 total. I'll have to add up the hardware. Took my son, wife and I a full day to precut all the lumber, four days to build them (two full weekends) and two days to mount the panels and anchor.

I can send you pics of the process if you like and dig up the Amazon hardware.

Very nicely done...thx so much for sharing! More info & more PICTURES are always better. ?

 

[Zwy]

Why did the power drop as you increased the tilt, which should have better aligned the panels to the sun elevation?

Production increased by 400W. 6 hours production would be 2400W.

Times 30 days, not too bad.

Of course this is for people who look at photos and recognize what they are seeing. If you don't see it, carry on.

 

[Zwy]

And that sums it up rather nicely.
If you added just one more panel (and forget about tilting) you might easily gain back that missing five percent, and save yourself a lot of messing about constantly adjusting the thing.

Doesn't always work that way. Take the 400W @acdoctor produced. One 400W panel would cover that production, however a few things would complicate the easy scenario. Current string configuration, PV mount, VOC and SCC specs.

Where does one add this odd panel in an array composed of 20 panels with 10S2P? So we add 2 panels? What about VOC? Where will that put things? If the SCC is maxed for VOC, then another SCC would probably need to be added.

Not quite as easy as "just one more panel" gets you there. For my 16 panel array with top of pole mount, I'd have to add 4 panels, not just 2 for balance. But I run 8S to get VOC high where I don't suffer from voltage drop. So now it's an 8 panel string?

Might be cheaper and easier to make do with what is already in place. 400W per hour is 2400W in a 6 hour day but we all know that isn't full production.

But it's free just for tilting the panel.

Or you can place an each way bet. Mount half your panels for summer tilt, and half for winter tilt.
Its a pretty crappy picture, but this is how I did my north facing panels on a flat roof.

You could just mount all halfway between the max tilt for each solstice and it would produce the same.

 

[Warpspeed]

You could just mount all halfway between the max tilt for each solstice and it would produce the same.

That is quite true, but I wished to see how much difference there was between the two different elevation angles under various conditions.
As its turned out, very little.
In this game, a bit of experimenting, can sometimes produce some quite unexpected results.

 

[Zwy]

That is quite true, but I wished to see how much difference there was between the two different elevation angles under various conditions.
As its turned out, very little.
In this game, a bit of experimenting, can sometimes produce some quite unexpected results.

I experiment now and then, wife was commenting last night about me making mistakes. https://diysolarforum.com/threads/safe-grid-use-of-the-5000es-and-transformer.34121/post-656147

I think my experimenting is a little different than yours.

 

[Llaves]

Production increased by 400W. 6 hours production would be 2400W.

Times 30 days, not too bad.

Of course this is for people who look at photos and recognize what they are seeing. If you don't see it, carry on.

1. Some of us read what the OP actually writes, rather than relying on pictures.
2. Math/physics doesn't lie. If you tilt a panel 7° off the optimum tilt, the output will change by cos(7), which is less than 1%. To get the nearly 10% change observed, some other conditions changed. It's not enough to say the sky "looked the same". Of course, you can believe otherwise, just don't be disappointed when the total output doesn't increase 10% as suggested by those two point readings.

There's a reason the NREL went to the trouble of posting a calculator that is backed by a vast database of real weather data. If you want to know how changing your system (eg, tilt) will impact your annual output, this is the way to find out. Measuring one moment in time tells you nothing, esp. when there are clouds moving across the sky.

 

[peakbagger]

I have posted this before. Its a 2 KW array. I am in Northern NH and my tilt angles are 30 degrees off vertical for winter, 45 spring and fall and 60 degrees for summer. Its balanced so I can change angles in about 5 minutes solo.

With respect to wintertime gain by tilting , PVWATTS is not of much use as it does not attempt to take into account snow buildup and reflection from snow. Snow buildup on a 45 degree array is significant issue and far less at 30 degrees. If the right type of snow its still and issue but it will definitely slide off quicker. The All Earth trackers will go vertical sometimes when they are covered with snow and I have seen a couple of nearby sets going through a full days rotation vertical with a layer of snow on it and repeat for a few days until a warm up. Snow reflection is big, Vertical panels act like they are tilted on a sunny day as the sun reflecting off the snow will make the panel put out more than rated wattage on a cold day. Its obviously a combination of the temperature coeeficient and sun reflection off snow. Anyone who hikes in the winter above treeline or ice fisherman usually learn by painful experience that snow blindness and sunburns can occur when out on the snow all day. I probably have cataract surgery in the future for too many winter hikes without sunglasses

The array is all unitstrut available at home depot. The two black poles are 4 by 6 rectangular tubing poured in block of concrete. I did the complete installation solo including mixing and pouring 50 bags of sakrete. it pivots on 3/4" SS shaft. I could definitely cut down on strut use by using the double strut that can be ordered by an electrical supplier but the markup on unistrut is steep at a supply house and I lost my friends electrician's discount. Standard strut is cheaper at HD .

 

[Llaves]

this is how I did my north facing panels on a flat roof.

Are you south of the equator? If not, a north-facing panel would be self shading, esp at high tilt like that.

 

[Warpspeed]

Are you south of the equator? If not, a north-facing panel would be self shading, esp at high tilt like that.

Yes thirty eight degrees latitude, down south.
Melbourne Australia.

 

[Llaves]

PCWATTS is not of much use as it does nto attempt to take into account snow buildup and refelection from snow. Snow buildup on 45 degree array is significant issue and far less at 30 degrees.

I'll agree that PVWatts doesn't attempt to model snow shedding, but a panel covered with snow generates essentially nothing regardless of tilt. You can't use a model without understanding the assumptions, which is this case is that the panel is clear. Likewise the model assumes the panels aren't shaded. If they're shaded, of course the model is off. If you know the conditions at which snow is shed and you know the frequency of snow falls, you can model different scenarios and estimate the output. It beats just throwing up your hands and saying it's unknowable.

It is NOT true that PVWatts doesn't model reflection from the snow (or any ground cover, more generally). From the technical manual -

The total POA incident on the module cover is the sum of the three components (Eqn. 2).
I_poa = I_b + I_d,sky + I_d,ground (2)

I_poa = incidence in the plane of the array
I_b = beam normal input * cos(angle of incidence)
I_b,sky is the diffuse sky component
I_d, ground is the diffuse ground component

If your area has a TMY3 file available for weather data, the file contains hourly readings of the albedo, which is a component of the diffuse ground model. While snow is a good reflector (ie, high albedo), the ground contribution is quite a bit smaller than the direct component, in large part because the panels are pointing away from the ground.

 

[Llaves]

Yes thirty eight degrees latitude, down south.
Melbourne Australia.

that explains a lot

 

[peakbagger]

Interesting on the ground albedo factor as it can vary significantly with the local ground cover. Folks I know with softwoods on the ground and large paved driveway in front of the panels do not show the reflection effect anywhere near what folks like myself that have a fairly flat clear area of snow in front of the panels. Mine is definitely not perfect as I have a 14' wide driveway running east to west in front of my south facing panels. I switched to a snowblower so no major snow banks associated with the driveway.

 

[Warpspeed]

Snow is unbelievably reflective.
I spent a couple of years in the Antarctic, it can be well below freezing and you still get sunburn very easily.