Examples of Panels Weighing Too Much for House?

[emartin00]

There is a need to change codes and processes if they are unnecessarily preventing solar installations. I've worked with AHJs a number of times on unnecessary requirements and got them changed in the past.

The uplift requirement is not the issue here. The AHJ is worried about the actual weight on the house to the point that, in many cases, they are requiring a full engineering analysis clear down to the footings and foundation of the home. They want to know that a system isn't going to compromise the structural integrity of a home, even in an earthquake. But I'm pretty sure that a solar system isn't going to determine whether a house will collapse in an earthquake. We have engineers stamp our system designs all the time but this is a whole other level of analysis and costs.

Where are you located?

[J.P.M.]

exactly. Ballasted systems always need a stamp (and someone to review the structure to see that it can handle the load). This is why it is difficult to put ballasted on residential. Often the structure can not be certified as it is too old of uncertain integrity. Commercial often has specs on the large buildings to easily calculate to.





It is possible to do railed systems on flat (Horizontal) as well as both railed and railless on pitched roofs.

Railless on a flat roof though possible would result in a flat array which is not desirable.



Railes have structural integrity of their own. They tie the rafters together and are considered to increase integrity more than the lag bolts would decrease the integrity, Because they tie the rafters together and spread the load and uplift force out, especially with staggered penetrations. Drilling holes into the rafters to put lag bolts in decreases integrity.
Railless on the other hand do not have any integrity of their own, they have more lag bolts and add weight but no integrity.



exactly, and why I stated tilted systems should have a stamp.



many jurisdictions do not require stamps in any of the conditions I suggested should have them but most of the larger well established do. We stopped doing ANY residential ballast system as it is too difficult to get anyone to review and stamp for the older ones and in jurisdictions that don't require it, we didn't feel comfortable either.

Ironridge has standards for revers tilt but that is for their equipment not for the structure it is bolted to.

Ballasted systems in and of themselves do not need a P.E. stamp on the design simply by virtue of being a ballasted design as you first sentence seems to read. Is stamping in an uncertain situation a good idea ? Probably. Required by some localities? Yup. Inherently required by the simple fact of being ballasted ? Nope.

All that said, I'd suggest that as a practical matter, ballasted systems ought to be the last method of mounting to be considered, if at all. IMO, the reason it's usually attempted at all is because it looks simple with most folks being clueless about the need to consider loadings imposed by a large weight, and the need to ensure that weight says put.

If stamping is done for any reason, part of any good design/review will include consideration of what the array's imposed loads will be transferred to, such as any supporting or mounting structure, and what those loads will do to that structure. I agree that such things are often difficult or impossible to check due to lack of drawings/calcs or the ability to properly inspect load bearing members. If a ballasted system is put on a structure of unknown or unverifiable construction, expect a hard time finding a P.E. willing to take a chance on it, or if you do, expect to pay a hefty premium as you seem to be saying. That also probably makes a system more costly and harder to sell.

On rail vs. railless racking, I can't, off the top of my head, think of a reason why railless systems on flat roofs require an array to be limited to a horizontal orientation as you seem to be implying. I don't believe the railless police are out and about on that one. Maybe a tilted array using railless systems is not practical on horizontal roofs for a lot of reasons, but I could sure design one.

Think of a one or two panel tilted mount as for a solar flat plate water heater on a flat roof. I drive by several each day with tilts that are somewhere between 20 and about 45 deg. or so. Just make the supports that are farther from the equator longer. That may not be as practical for large arrays as railed systems and so they are probably not all that common for large arrays, but there is no inherent reason why railless systems need to always be in the plane of the mounting structure or roof. One reason they may not be common might have to do with marketing and maybe also because there is little demand for such support systems. Just get longer standoffs for the supports farther away from the equator, or as long as necessary to fit the application (nothing says you can't have 4 legs/panel with all 4 legs having different lengths from one another), X brace the high side for transverse moment(s), and depending on the support length(s), maybe check for Euler buckling, flutter instability for compressive wind loading and a few other things, and call it done. Not sure what it would cost, but that's different from being not

As for rail systems having structural integrity on their own : A rail has little ability to resist bending moment on its own and little structural ability beyond tensile strength to resist an axial tensile load compared to a rigid 2 dimensional structure like an array. Rail systems on arrays, depending on how panels are connected, may and probably will result in greater array rigidity than railless systems due to the rigidity of the resulting railed 2 dimensional array.

But rigidity in and of itself is simply a characteristic of the array/mounting system and may, or may not, be beneficial or even desirable to an array's ability to stay put in a wind or seismic event. As emartain00 seems to be writing, as for whether or not an array stays put on a roof has more to do with the number of attachments to the mounting structure (the roof) than it has to do with array rigidity. As a matter of some opinion and circumstance, a rigid array may fair worse in an event, not only because it may well have fewer connections to the support structure (for example, a roof), but also because of the possibility and perhaps likelihood of the entire rigidly connected array coming to a bad end rather than one or two unconnected panels being lost. For example, a railess system that had 4 supports per panel with the panels not rigidly connected might be better able to withstand an external loading event simply by virtue of having more attachments to the structure, although I can see where such a system would have other less desirable design considerations like being able to easily hit a rafter every time and lots of other considerations.

All that aside, as for an array or it's mounting system adding strength to a support structure, it's usual to calc out/design or proof calc a support structure on its own, without consideration of any strength that may be added by what's being supported. So, conservative and usual design practice, at least as I learned and practiced it makes the idea that an array with a rail system may add structural integrity to a support structure moot.

BTW, structural integrity does not always mean rigidity. It has more to do with a structure's ability to withstand imposed design loads and still be as safe as required, and fit for purpose and service.

[J.P.M.]

By the way, there is now a convenient service call greenlancer.com that is providing solar engineering services in just about all states. In most cases they can provide 24hr turnaround on standard stuff. I've pretty much quit using my local PE friend that my jobs were small potatoes to...

From their website, as much as I looked I saw no mention or example or reference to what, if anything, Greenlancer does with mechanical design. Maybe it's like dating and expecting to get lucky: Some things you have to ask for.

[ButchDeal]

Ballasted systems in and of themselves do not need a P.E. stamp on the design simply by virtue of being a ballasted design as you first sentence seems to read. Is stamping in an uncertain situation a good idea ? Probably. Required by some localities? Yup. Inherently required by the simple fact of being ballasted ? Nope.

Well they are required inherently because of the definition of ballast. You are adding weight and most places, particularly the ones with residential flat roofs do require them because of the fact you are adding weight.

All that said, I'd suggest that as a practical matter, ballasted systems ought to be the last method of mounting to be considered, if at all. IMO, the reason it's usually attempted at all is because it looks simple with most folks being clueless about the need to consider loadings imposed by a large weight, and the need to ensure that weight says put.

residential often ask for it as it is low penetration and thought to leak much less but we have had very poor luck getting them approved and have as a matter of course stopped any attempts to do so in the future for residential (exceptions are possible of course).

If stamping is done for any reason, part of any good design/review will include consideration of what the array's imposed loads will be transferred to, such as any supporting or mounting structure, and what those loads will do to that structure. I agree that such things are often difficult or impossible to check due to lack of drawings/calcs or the ability to properly inspect load bearing members. If a ballasted system is put on a structure of unknown or unverifiable construction, expect a hard time finding a P.E. willing to take a chance on it, or if you do, expect to pay a hefty premium as you seem to be saying. That also probably makes a system more costly and harder to sell.

exactly.

On rail vs. railless racking, I can't, off the top of my head, think of a reason why railless systems on flat roofs require an array to be limited to a horizontal orientation as you seem to be implying. I don't believe the railless police are out and about on that one. Maybe a tilted array using railless systems is not practical on horizontal roofs for a lot of reasons, but I could sure design one.

Well mounting systems are designed as a unit. None that I know of have a tile component There would be considerable torsional force on the mounting points which are usually simple clamps. Further it would require a pivot system particularly at the base which none of them have. You could likely cobble some pivots together along with the standoffs but I would highly doubt it would warrantied or covered by the pre-calculations that the manufacturers provide.

Think of a one or two panel tilted mount as for a solar flat plate water heater on a flat roof. I drive by several each day with tilts that are somewhere between 20 and about 45 deg. or so. Just make the supports that are farther from the equator longer. That may not be as practical for large arrays as railed systems and so they are probably not all that common for large arrays, but there is no inherent reason why railless systems need to always be in the plane of the mounting structure or roof. One reason they may not be common might have to do with marketing and maybe also because there is little demand for such support systems. Just get longer standoffs for the supports farther away from the equator, or as long as necessary to fit the application (nothing says you can't have 4 legs/panel with all 4 legs having different lengths from one another), X brace the high side for transverse moment(s), and depending on the support length(s), maybe check for Euler buckling, flutter instability for compressive wind loading and a few other things, and call it done. Not sure what it would cost, but that's different from being not

again you would need a pivot and standoff which the railless systems do not provide. The array would have no structural integrity at the joint (your example of two modules, the mid point between them). The railless systems have no way to hold the modules in a flat plane other than the bolts to the roof, a little twisting and the clamps would pop off.

As for rail systems having structural integrity on their own : A rail has little ability to resist bending moment on its own and little structural ability beyond tensile strength to resist an axial tensile load compared to a rigid 2 dimensional structure like an array. Rail systems on arrays, depending on how panels are connected, may and probably will result in greater array rigidity than railless systems due to the rigidity of the resulting railed 2 dimensional array.

Well they have a bit of rigidity and most have a latteral and vertical loading capability as in this example:

https://www.youtube.com/watch?v=jzLB06KY0i0

The rails are attached to the truss system. If you have an annal retentive installer (like me) that likes the array to be perfectly flat even when the roof isn't you might see what I mean. Walk on a common construction grade roof with a string pulled across from end to end. just the weight of a man and flex the roof 1/4 to 1/2 or so (worse on a poor roof). We like to align the rails with a string end to end. Once they are all tightened up, you can walk across and there will be no sag ( or greatly reduced) and the rail will be dead on to the string or laser line.

You do the same thing with railless and it will still sage as you walk across and work because they flex at the edges of the array where they are attached. Further the nature of the railless means you can't stagger the mount points, they have to be at the edges or corners of the PV modules.

But rigidity in and of itself is simply a characteristic of the array/mounting system and may, or may not, be beneficial or even desirable to an array's ability to stay put in a wind or seismic event. As emartain00 seems to be writing, as for whether or not an array stays put on a roof has more to do with the number of attachments to the mounting structure (the roof) than it has to do with array rigidity. As a matter of some opinion and circumstance, a rigid array may fair worse in an event, not only because it may well have fewer connections to the support structure (for example, a roof), but also because of the possibility and perhaps likelihood of the entire rigidly connected array coming to a bad end rather than one or two unconnected panels being lost. For example, a railess system that had 4 supports per panel with the panels not rigidly connected might be better able to withstand an external loading event simply by virtue of having more attachments to the structure, although I can see where such a system would have other less desirable design considerations like being able to easily hit a rafter every time and lots of other considerations.

I disagree. In a railless rigidity is exactly as you say, a characteristic of the mounting system and roof it is mounted too. But a railless system has its own rigidity, that is the point of the rails. in ground mount you mount the PV modules to the rails then carry the unit to the ground mount and bolt it as a unit to the ground mounts. you can do the same thing on a roof mount tilted system (same for an untitled one if you had a skinny installer

tilted systems without rails, once one mount fails the units will fly apart. in a railed system, the mounts share the load making a single mount less likely to fail but any one mount can be removed with redundancy with the rails.

All that aside, as for an array or it's mounting system adding strength to a support structure, it's usual to calc out/design or proof calc a support structure on its own, without consideration of any strength that may be added by what's being supported. So, conservative and usual design practice, at least as I learned and practiced it makes the idea that an array with a rail system may add structural integrity to a support structure moot.

yes the rails should not be relighed on to support the structure but the point is on a typical roof mount the load is now partially on the array, which in a railed system is distributing the load as well as helping to keep the truss all aligned. In both railed and railless you have to add lag bolts which weaken the truss. The railed system the lags are staggered, with a railless they are in line every other truss only. Thus the weight is now born by HALF the truss or rafters.

[J.P.M.]

Well they are required inherently because of the definition of ballast. You are adding weight and most places, particularly the ones with residential flat roofs do require them because of the fact you are adding weight.




residential often ask for it as it is low penetration and thought to leak much less but we have had very poor luck getting them approved and have as a matter of course stopped any attempts to do so in the future for residential (exceptions are possible of course).



exactly.



Well mounting systems are designed as a unit. None that I know of have a tile component There would be considerable torsional force on the mounting points which are usually simple clamps. Further it would require a pivot system particularly at the base which none of them have. You could likely cobble some pivots together along with the standoffs but I would highly doubt it would warrantied or covered by the pre-calculations that the manufacturers provide.



again you would need a pivot and standoff which the railless systems do not provide. The array would have no structural integrity at the joint (your example of two modules, the mid point between them). The railless systems have no way to hold the modules in a flat plane other than the bolts to the roof, a little twisting and the clamps would pop off.



Well they have a bit of rigidity and most have a latteral and vertical loading capability as in this example:

https://www.youtube.com/watch?v=jzLB06KY0i0

The rails are attached to the truss system. If you have an annal retentive installer (like me) that likes the array to be perfectly flat even when the roof isn't you might see what I mean. Walk on a common construction grade roof with a string pulled across from end to end. just the weight of a man and flex the roof 1/4 to 1/2 or so (worse on a poor roof). We like to align the rails with a string end to end. Once they are all tightened up, you can walk across and there will be no sag ( or greatly reduced) and the rail will be dead on to the string or laser line.

You do the same thing with railless and it will still sage as you walk across and work because they flex at the edges of the array where they are attached. Further the nature of the railless means you can't stagger the mount points, they have to be at the edges or corners of the PV modules.



I disagree. In a railless rigidity is exactly as you say, a characteristic of the mounting system and roof it is mounted too. But a railless system has its own rigidity, that is the point of the rails. in ground mount you mount the PV modules to the rails then carry the unit to the ground mount and bolt it as a unit to the ground mounts. you can do the same thing on a roof mount tilted system (same for an untitled one if you had a skinny installer

tilted systems without rails, once one mount fails the units will fly apart. in a railed system, the mounts share the load making a single mount less likely to fail but any one mount can be removed with redundancy with the rails.



yes the rails should not be relighed on to support the structure but the point is on a typical roof mount the load is now partially on the array, which in a railed system is distributing the load as well as helping to keep the truss all aligned. In both railed and railless you have to add lag bolts which weaken the truss. The railed system the lags are staggered, with a railless they are in line every other truss only. Thus the weight is now born by HALF the truss or rafters.

For my part, While I believe I understand what you're writing, I feel discussing this further is a waste of my time. Respectfully, from what you write and how you write it, I don't believe you have the structural engineering background beyond what you've been told to understand what I'm writing about. I suggest we simply agree to disagree as this is going nowhere.

We probably lost the OP somewhere along the way anyway.

Respectfully,

J.P.M.