Need help calculating the wind and snow loading of my array

[ICPete]

Hi JPM,

Wind loading and design for wind-resistant structures, very little. Most of my related work has been FEA for vibration resistance of airframe-mounted electronics and optical instruments. Also large super-stable optical metrology systems. But very little experience in things like highway signs, buildings, silos, etc. I got the conditions (wind speeds) for my area from the ATC Hazards by Location website, which lists ASCE 7-16, ASCE 7-10, and ASCE 7-05 speeds for various MRI year categories and Risk Categories. Also my local AHJ publishes the required wind load and snow load conditions (plus seismic zone etc) to be met here. Their requirement is 90 MPH, Exposure C. For my area the ATC document shows the highest wind speed for Risk Category II (residences) as 115 MPH under ASCE 7-19. I have collected quite a bit of reference information, such as the ANSI/SPRI Wind Design Standard Practice for Roofing Assemblies. There is also an excellent piece by Robert Paullus of Paullus Structural Consultants, his "WIND WEBINAR SERIES #3", from February 2013, that I found online, which is practically a tutorial on designing for wind loading; very helpful! Using Paullus' presentation material (which is available elsewhere in textbooks) I generated a spreadsheet that combines the various factors and gives me the wind load pressure in PSF as well as the total force in pounds for a given area, after entering the wind speed in MPH.

The only design work I've had signed off by a PE to date has been several timber frame home designs back in the 1990s; none of my usual consulting work requires PE. However as I stated above, this solar mounting system will need "an engineered design by an Arizona Registrant for the foundation/footing system. The design shall include attachment to foundation (anchor bolts, size), and number of bolts, embedment depth, size and amount of reinforcement steel, depth and size of foundation/footing." So while the actual statement seems to refer ONLY to the foundation and footing system, I would think they intend for the whole rack or tracker to be so engineered. I need to go speak with someone at Building Safety and get some clarification, before spending real money on either PE design work or the construction itself.

[ICPete]

Hi Bruce,
I certainly will as things progress (if they do progress). I hear you loud and clear on the effects of clouds. I have used both SAM and PVWatts over and over again with various system designs, comparing different fixed angles, tilted single-axis, azimuthal rotation (only available in SAM but not in PVWatts, that I could find). Interestingly both methods give essentially equal modeling results for my set of weather data. Unlike your situation, which you are describing as "cloud central", I am in sunny Arizona at 5000 ft. So yes, some cloudy days of course, but overall I feel like I'm in solar heaven compared to the 24 years I lived in northern Vermont!
Using that data tells me I would get slightly higher total annual output with 60 panels fixed at 35 degrees, vs 40 panels on single-axis trackers with the axis tilted at 35 degrees. A single-axis azimuthal tracker with the panels at 55-60 degrees also gives very good results. Note that I'm attempting to maximize the winter output without taking too much of a hit on the summer output. Our local climate has a very modest 950 CDD load (summer A/C), and about 4200 HDD in the winter. Given that my heat will be supplied by an electric heat pump, I need to favor winter PV output rather than maximizing the total annual output. If I understand APS' latest net metering scheme, the buy-sell differential is calculated (netted out) monthly (obviously not as advantageous to PV owners as was the former annual netting scheme). My goal is to buy very little energy from them even in winter months. This will be a hybrid battery-backed system, so I also plan to take advantage of TOU provisions.
Pete

[J.P.M.]

Hi JPM,

Wind loading and design for wind-resistant structures, very little. Most of my related work has been FEA for vibration resistance of airframe-mounted electronics and optical instruments. Also large super-stable optical metrology systems. But very little experience in things like highway signs, buildings, silos, etc. I got the conditions (wind speeds) for my area from the ATC Hazards by Location website, which lists ASCE 7-16, ASCE 7-10, and ASCE 7-05 speeds for various MRI year categories and Risk Categories. Also my local AHJ publishes the required wind load and snow load conditions (plus seismic zone etc) to be met here. Their requirement is 90 MPH, Exposure C. For my area the ATC document shows the highest wind speed for Risk Category II (residences) as 115 MPH under ASCE 7-19. I have collected quite a bit of reference information, such as the ANSI/SPRI Wind Design Standard Practice for Roofing Assemblies. There is also an excellent piece by Robert Paullus of Paullus Structural Consultants, his "WIND WEBINAR SERIES #3", from February 2013, that I found online, which is practically a tutorial on designing for wind loading; very helpful! Using Paullus' presentation material (which is available elsewhere in textbooks) I generated a spreadsheet that combines the various factors and gives me the wind load pressure in PSF as well as the total force in pounds for a given area, after entering the wind speed in MPH.

The only design work I've had signed off by a PE to date has been several timber frame home designs back in the 1990s; none of my usual consulting work requires PE. However as I stated above, this solar mounting system will need "an engineered design by an Arizona Registrant for the foundation/footing system. The design shall include attachment to foundation (anchor bolts, size), and number of bolts, embedment depth, size and amount of reinforcement steel, depth and size of foundation/footing." So while the actual statement seems to refer ONLY to the foundation and footing system, I would think they intend for the whole rack or tracker to be so engineered. I need to go speak with someone at Building Safety and get some clarification, before spending real money on either PE design work or the construction itself.

Thank you. If it was me, I'd get more details from the AHJ on exactly what they require and what they will accept. Then, if a P.E. stamp/seal is indeed required, find a P.E. qualified in the area(s) of competence required, and see if that P.E. will review and then stamp your design, including wind and other external loadings. That's usually a lot cheaper than a full design. If you can't find one who will review/stamp your design (and many will not unless you're working in the same outfit under their license as an intern engineer for example), then your stuck with a higher bill for a stamped design done by a P.E. rather than a design that's been reviewed and stamped as fit for purpose by a P.E. The latter may save you a few bucks.

If AHJ doesn't need a P.E. cert./stamp but does needs wind calcs along with other external loadings or engineering design, depending on how long you've been doing responsible charge engineering, it reads to me like you may well have the necessary engineering competence in those areas, and equally importantly the engineering attitude and thoroughness necessary to produce a safe and fit for purpose design. I'd go for it.

How long have you been doing substantial charge engineering ?

Respectfully,

[D0T-C0M]

ICPete, thanks for sharing your thoughts. Here in Eastern Canada we do not require a PE to sign off on the mounting system which keeps the price down but I still want to design this to be safe and last a long time. I am just in the design phase and gathering as much info as possible from other DIYers. I really appreciate you sharing your methodology on how you designed your system. Initially I wanted to do like you and go single axis but my property is not ideally suited for an array along the north / south axis. I am like you very intrigued about the idea of a solar tracker. I was a journeyman Industrial Instrumentation & Controls Tech up to 2000 and now am a journeyman Power Systems Technician working on high voltage transmission and distribution equipment. I will probably fab a custom solar controller using an arduino atmega processor. I plan on using a time base system instead of using a light sensor which, from what I've read can have problems under cloudy skies. My goal is for the controller to connect to my home network through wifi or ethernet link and the human interface will be through a web server run on the controller and will be accessible to anyone on my home network. There will also be local controls at the array.

Seeing you are a Mechanical engineer can you comment on how to properly calculate the moment tilting torque. The slew drive I have been recommended by the supplier has a moment tilting torque of 13.5kN-m on the Azimuth slew drive and the elevation slew drive has a moment tilting torque of 9kN-m. Moment tilting torque is described as being equal to the force/load times the distance from the center of the Azimuth slew ring. Is this simple as using the cosine of the elevation angle of the elevation and distance between the center point the structure supporting the panels? For example if the rack is mounted 12" from the center of the elevation slew ring am I right that the distance away from the center of the Azimuth slew ring is equal cos(angle of elevation) x 12". So if my panels including the support structure above the slew drive equals 2000lbs. At 20 degrees elevation would equal cos20x12=11.28" and Moment Tilting Torque = 2000x(11.28\12)=1879ft/lbs ? I assume this is the correct way to calculate this? If this is the case I would not have much safety factor built in.

The more I think about this moment tilting torque and ways to minimize it. Maybe I'll change the design so that I come straight out of the elevation slew ring with a pipe on either side and mount 6 PV panels on either side with a gap in between. That way the load will rotate around the centerline of the elevation slew ring thereby practically eliminating the moment tilting torque, instead of suspending it 12" above and having all that torque being put on the Azimuth slew drive ring as I approach 0 degrees elevation. In this design the top of the elevation slew ring will protrude above the PV panels.

[ICPete]

D0T-C0M,
I think you already know this, but moment or torque is calculated as a force times a distance, where the distance is measured perpendicularly from the force vector to the axis around which you are calculating said moment. So the necessary input is to know the force and the point at which it is applied, and its direction. The challenge of understanding all the load cases and how to set up the calculations for the resulting stresses on members, joints, and components is why getting help from a PE is recommended. That is, a PE with experience in similar structures and mechanisms.
As I stated up front, I'm not a registered professional engineer, so I can't really help you with the detailed calculations.

[ICPete]

Hi JPM,

To answer your question:
"How long have you been doing substantial charge engineering ?"

My career started in 1976. Being in a role equivalent to "responsible charge engineering" (albeit not in fields that required the PE license), since the early 90's; say 25+ years.

So I'm definitely an "old timer"! Started full time consulting in 1989. Back to employee status from 2005-2012, then self-employed consulting since then.
Much of my work involves designing one-off or prototype systems, which really need to work correctly the first time. So I do what someone once described to me as "Design by Analysis". I use FEA quite a lot, as well as optics modeling. I got started with 3D solid modeling on an HPUX workstation in 1993 (purchased myself for my consulting work); what a boost to productivity compared to board drafting or even 2D Autocad! I moved to Solidworks gradually over the period from 2001 to 2004.

BTW I appreciate your attention to my situation, consideration of the alternative possibilities, and advice on proceeding.
I'll let you know what I learn from the county Building Safety department about the requirements for the PV support structure.
Pete

[D0T-C0M]

D0T-C0M,
I think you already know this, but moment or torque is calculated as a force times a distance, where the distance is measured perpendicularly from the force vector to the axis around which you are calculating said moment. So the necessary input is to know the force and the point at which it is applied, and its direction. The challenge of understanding all the load cases and how to set up the calculations for the resulting stresses on members, joints, and components is why getting help from a PE is recommended. That is, a PE with experience in similar structures and mechanisms.
As I stated up front, I'm not a registered professional engineer, so I can't really help you with the detailed calculations.

I edited my post above, so maybe you didn't see it. To reduce the tilting torque I'm going to change the design so that I come straight out of the elevation slew ring with a pipe on either side and mount 6 PV panels on either side with a gap in between. That way the load will rotate around the centerline of the elevation slew ring thereby practically eliminating the moment tilting torque, instead of suspending it 12" above and having all that torque being put on the Azimuth slew drive ring as I approach 0 degrees elevation. In this design the top of the elevation slew ring will protrude above the PV panels.

Also I believe this design should also result in less wind loading by dividing the 12 panels into two 6 panel array with a 12" gap between them. IF I am correct in my calculations using a single 12 panel array (230sqft) should be able to operate at 0 degree elevation (panels vertical) in direct winds up to 70 mph before exceeding the tilting torque specs. I will certainly be parked horizontal well below that wind speed.

[ICPete]

I like your idea for reducing the moment and getting the CG of the panels close to the elevation drive rotation axis. I'm not sure the wind loading will be much lower by having the gap between the two sets of 6 panels however.

Also take a look at what happens if you have an asymmetrical snow load. Since you are presumably in snow country, it could happen that snow melts off one side before the other. Even though we don't get a lot of snow here, and the panels should be parked at about 60-degrees tilt from horizontal, I am analyzing the cases where either one-half or one-quarter of the array is loaded by gravity with 30 PSF of snow (that is roughly five feet of snow I believe, so highly improbable here). The off-center load will create quite a high moment in either one or two planes of my structure; it is quite a severe test. For instance, looking at several commercially available two-axis trackers online, there is no way, IMHO, that they could withstand this sort of asymmetrical loading.

But I would think it becomes more possible in Canada, especially if you happened to have the tracker parked with the panels level overnight. Maybe you simply avoid snow buildup by always parking the panels plumb when it's snowing?? It sounds like you're going to have vastly more range of motion than my single-axis design.

Just my 2 cents on the issue...

[J.P.M.]

Hi JPM,

To answer your question:
"How long have you been doing substantial charge engineering ?"

My career started in 1976. Being in a role equivalent to "responsible charge engineering" (albeit not in fields that required the PE license), since the early 90's; say 25+ years.

So I'm definitely an "old timer"! Started full time consulting in 1989. Back to employee status from 2005-2012, then self-employed consulting since then.
Much of my work involves designing one-off or prototype systems, which really need to work correctly the first time. So I do what someone once described to me as "Design by Analysis". I use FEA quite a lot, as well as optics modeling. I got started with 3D solid modeling on an HPUX workstation in 1993 (purchased myself for my consulting work); what a boost to productivity compared to board drafting or even 2D Autocad! I moved to Solidworks gradually over the period from 2001 to 2004.

BTW I appreciate your attention to my situation, consideration of the alternative possibilities, and advice on proceeding.
I'll let you know what I learn from the county Building Safety department about the requirements for the PV support structure.
Pete

Thank you for the reply. IMO, both complete and informative. FWIW, I think we might see a lot of our engineering the same way(s).

I'd be interested to hear (read) about progress on your project.

Regards,

J.P.M.

[D0T-C0M]

Also take a look at what happens if you have an asymmetrical snow load. Since you are presumably in snow country, it could happen that snow melts off one side before the other. Even though we don't get a lot of snow here, and the panels should be parked at about 60-degrees tilt from horizontal, I am analyzing the cases where either one-half or one-quarter of the array is loaded by gravity with 30 PSF of snow (that is roughly five feet of snow I believe, so highly improbable here). The off-center load will create quite a high moment in either one or two planes of my structure; it is quite a severe test. For instance, looking at several commercially available two-axis trackers online, there is no way, IMHO, that they could withstand this sort of asymmetrical loading.

But I would think it becomes more possible in Canada, especially if you happened to have the tracker parked with the panels level overnight. Maybe you simply avoid snow buildup by always parking the panels plumb when it's snowing?? It sounds like you're going to have vastly more range of motion than my single-axis design.

Just my 2 cents on the issue...

The elevation slew drive has a tilting torque of 9kN which is ~2000lbs. That would have to be quite a snow drift on one side of my tracker I would be cleaning them after every storm so its not like it will have time to build up. Thanks for your input.

[bcroe]

I like your idea for reducing the moment and getting the CG of the panels
close to the elevation drive rotation axis.

Pivot near center of gravity is what I would do. Bruce Roe

18ArrayR.JPG

[J.P.M.]

I like your idea for reducing the moment and getting the CG of the panels close to the elevation drive rotation axis. I'm not sure the wind loading will be much lower by having the gap between the two sets of 6 panels however.

Also take a look at what happens if you have an asymmetrical snow load. Since you are presumably in snow country, it could happen that snow melts off one side before the other. Even though we don't get a lot of snow here, and the panels should be parked at about 60-degrees tilt from horizontal, I am analyzing the cases where either one-half or one-quarter of the array is loaded by gravity with 30 PSF of snow (that is roughly five feet of snow I believe, so highly improbable here). The off-center load will create quite a high moment in either one or two planes of my structure; it is quite a severe test. For instance, looking at several commercially available two-axis trackers online, there is no way, IMHO, that they could withstand this sort of asymmetrical loading.

But I would think it becomes more possible in Canada, especially if you happened to have the tracker parked with the panels level overnight. Maybe you simply avoid snow buildup by always parking the panels plumb when it's snowing?? It sounds like you're going to have vastly more range of motion than my single-axis design.

Just my 2 cents on the issue...

Just my $0.02 as well and definitely not a knock as I think I understand what you write and I'm with your opinion(s), but FWIW, my guess is you may be a notch or two above most of this crowd on mechanical design.

[RShackleford]

Came across this thread in search, and maybe asking here is better than starting a new thread ...

I am not doing a tracking system, just a simple ground mount, clamping panels to a pressure-treated frame with SS hardware. My panels (Talesun TP660P) can apparently be clamped on the long sides or the short sides, with allowable loading of 5400Pa and 2400Pa, respectively. Any simple way (without hiring an M.E.) of figuring out if I can get away with assuming 2400Pa loading ?

I can use the ground-mount design tool that IronRidge provides; even though I'm not using their hardware, I imagine their loading calculations would depend only on the size, height, and tilt of the array. For a design, they quote shear, moment, and uplift in pounds; perhaps I can simply divide the uplift by the array's area, and then convert to Pascals ? However, I come up with something 200Pa, so that's probably not right.

[J.P.M.]

Came across this thread in search, and maybe asking here is better than starting a new thread ...

I am not doing a tracking system, just a simple ground mount, clamping panels to a pressure-treated frame with SS hardware. My panels (Talesun TP660P) can apparently be clamped on the long sides or the short sides, with allowable loading of 5400Pa and 2400Pa, respectively. Any simple way (without hiring an M.E.) of figuring out if I can get away with assuming 2400Pa loading ?

I can use the ground-mount design tool that IronRidge provides; even though I'm not using their hardware, I imagine their loading calculations would depend only on the size, height, and tilt of the array. For a design, they quote shear, moment, and uplift in pounds; perhaps I can simply divide the uplift by the array's area, and then convert to Pascals ? However, I come up with something 200Pa, so that's probably not right.

To your question about simple ways: Short answer is "maybe". However, if the AHJ requires a set of wind calcs or other external loadings for the structure, I'm guessing you're probably unqualified to do them and, depending on what the AHJ requires, you may need some help so the answer would probably be "no".

I'd first find out what required in the way of calcs by contacting the AHJ and ask:

1.) Are wind or other design calcs for external loadings required ?
2.) How detailed do such calcs need to be in terms of which components and what types of loads need to be addressed ? What codes/standards (if any) need to be followed ?
3.) Does the design require a P.E. stamp for approval ?

Then proceed based on the AHJ's guidance.

Your imagination is leading you astray. External loadings require more than you mention and probably more than you are aware of to be complete and have a better probability of being safe.

From your comment about fixed arrays, I infer you are under the misconception that fixed arrays are inherently simpler (stronger?, safer ?).The loadings for fixed arrays will, in all likelihood be different than those for a tracking system, but may well be no less involved or lower in magnitude - All of which makes not one iota of difference. Most any design, except the most simple (maybe) needs to be calculated, and the calcs checked to be a good design. That includes all components, including anchorage and fixations like the posts and maybe even soil characteristics.

The AHJ will inform you if you ask.

BTW, I didn't fall off a turnip truck last week. I'm aware the above sounds like useless overkill and bureaucratic B.S. designed to do little more than give some political hack's otherwise unemployable brother-in-law something to do besides make love to his fist, and that most of the time a lot less, if any serious, much less safe design gets done by shade tree mechanics and redneck engineers, but that's the way I learned my engineering and so far it's served me well.

See what the AHJ says.

Take what you want of the above. Scrap the rest.

[RShackleford]

The AHJ just said I need an engineer-sealed drawing, and I talked to a guy who said if I gave him some preliminary drawings, he would do it for $250; but as far as #1 and #2, I don't know (yet).

Your imagination is leading you astray. External loadings require more than you mention and probably more than you are aware of to be complete and have a better probability of being safe.

Not sure what you mean. More than I mention, meaning in addition to wind and snow ?

Seems like ground-mount is simpler than tracking, simply because there are no moving parts. But yeah, I am having my consciousness raised about the fact that ground-mounts are much more subject to wind loads - just from the IronRidge calculator for ground mounts - those 6-7ft deep footings got my attention !!

I am an engineer, but of the electrical/computer variety (I've only dabbled in ME), so I might be capable of comprehending some of the necessary calculations. Mainly I just want the drawings I give this local ME to be somewhere in the ballpark; and I believe the footings are gonna be the primary issue.

Maybe I should suck it up and just buy the IronRidge, about $1000, plus locally sourcing about 60ft of 3" pipe and contracting those deep footings; I'm gonna have to do the footings anyhow, my lumber looks like it's gonna be $600 or so. And there's the $250 engineering fee, which IronRidge claims I can forget, since their stuff is "pre-engineeered" (or whatever).