Does anyone trust anyone's ballast calculations?

Options
I've been trying to write up an installation proposal for a nearby short-stack commercial property. He's adamant about a ballasted install, so we hired a local engineer to do some ballast calculations. The numbers seemed way out of whack -- the methodology seemed fine, but the wind force being applied seemed way too high.

So I started looking around for examples and talking to racking companies. Here's the part that scares me: I received ballast estimates that varied by an average of four times! The high end was our own engineer at about 600 lbs per panel, and the low end from a local racking company was about 150 lbs per panel. Another came in at the high end unless we put on wind shades on the back, and which point it went down 4 times. I'm sure that the mathematics are correct in all of these cases, yet clearly they can't all be correct.

Looking into it, I find that there are three different possible calculations one might use from the ASCE.

The first is the basic load calculation for houses, which returns forces on roofs. Much of the force on a roof is lift due to the speed of the air flowing over the top of it (wind speed) compared to under it (zero, inside the house).

A second version of this basic calculation is similar, but is used for exposed roofs, like on a car park. In this case the wind speed on the top and bottom is similar, so the lift is greatly reduced.

Finally there is the "cladding" calculation, which is basically a way to find out how strong your outmost layer of building material is. You can think of panels on a roof sort of like the roofing membrane.

Basically, the first calculation gives the highest numbers, the second much lower ones, and the last somewhere in between.

Which ones are we supposed to use? From what I can tell, no one can agree! Moreover, when you actually put panels in a wind tunnel, they seem to suggest much lower numbers than any of these calculations!

Comments?

Maury

Comments

  • BB.
    BB. Super Moderators, Administrators Posts: 33,491 admin
    Options
    Re: Does anyone trust anyone's ballast calculations?

    That is a tough one... It is difficult to predict the wind with irregular structures and direction. This coming from a guy whose airplane was probably tied up outside for 50 years--and one storm blew up from behind and snapped a wing off. :cry: (I did find a replacement wing to rebuild and got the plane flying again).

    You might look Solyndra ... There system is a set of tube shaped solar PV collectors and should have less lift/drag in wind--plus their claim to fame is a ballasted installation. I have no idea about their product, how long it will last, how much it costs, and the company's financial health (nobody is doing great right now :cry:). But between their product and possible research into ballasted arrays--perhaps you can find some answers there.

    Engineers tend to be conservative (typically building/structural design uses a safety factor of around 10x). Anything within a factor of 2x with engineers is roughly the same. Anything 10x or greater difference is typically not even in the same ball park with their calculations/assumptions.

    4x difference in design for something like this--probably par for the course (I am not a structure engineer).

    I guess it comes down to a few questions... Will the PEs sign off on their calculations for the ballasting (light through heavy). What is the area like around the building (could panels blow into a school full of children--or is it out in the middle of no-where)?

    Say the light ballast has a 50:1 shot of failing (i.e., 50 year storm). Is the guy willing to take the risk. Can anyone assign risk to light vs heavy ballast. Remembering that we have had some 500 year storms recently.

    In the end, I think looking at risk/reward (expense vs cost) for the mounting options will clarify matters a bit (added weight, snow/water loads/plugged drains, hard mounted arrays with bolts through the roofing material/flashing issues, etc.).

    The array will have at least 5 degree tilt to make them self-cleaning instead of dead flat?

    Sorry I don't have any better answers.

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Cariboocoot
    Cariboocoot Banned Posts: 17,615 ✭✭✭
    Options
    Re: Does anyone trust anyone's ballast calculations?

    Examine the reputations of the various engineering firms involved. See if you can find a company you can trust, then trust their answer. They will have to sign off on the responsibility. The costliest estimate is not necessarily the best one either.
    Also perhaps talk to an insurance company or two. They may have some requirements that must be met in order to cover the structure. If all the "experts" are satisfied, that's the best you can hope for.
  • dwh
    dwh Solar Expert Posts: 1,341 ✭✭✭
    Options
    Re: Does anyone trust anyone's ballast calculations?

    I think the significant question is: How far is the standoff?


    If the panel is stuck down tight on the roof, with no significant airflow underneath, then the cladding calc might be appropriate.

    If the panel is close enough to a surface to have the "roof wind blowing and lifting" effect, then the roof calc would be correct.

    If the standoff is high enough to negate that effect, then the carport calc would be the right one to use.
  • solarix
    solarix Solar Expert Posts: 713 ✭✭
    Options
    Re: Does anyone trust anyone's ballast calculations?

    I've been in your shoes on this one and agree that the "science" of wind loading is poorly developed for solar arrays. Unirac has a study they put out supporting the ballasted array they sell and it is very mushy calculations with lots of fudge factors and numbers pulled out of not so well adapted ASME reference standards. Then they take that result and multiply it by a factor of about 1/2 based on their own "wind tunnel testing" and then add on a 50% safety factor to get a final number I can't say has much connection to reality. The PE involved with my project ended up doing a little yank test on our array and said "that feels pretty good". By the way, we used 150# per 17sqft module on 30 deg tilt racks. Makes me wonder a bit when driving over a suspension bridges!
  • Maury Markowitz
    Maury Markowitz Registered Users Posts: 9
    Options
    Re: Does anyone trust anyone's ballast calculations?
    solarix wrote: »
    By the way, we used 150# per 17sqft module on 30 deg tilt racks.!

    Ok, that's definitely in-line with one of the two sets of numbers I'm seeing. The other is 600# for the same module. That's just too much, our customers won't accept that. What I need to do is convince myself 150 is enough!

    Maury
  • solarix
    solarix Solar Expert Posts: 713 ✭✭
    Options
    Re: Does anyone trust anyone's ballast calculations?

    For what it is worth, here are my calcs for wind load ballast that we used a couple years ago.
  • Dave Angelini
    Dave Angelini Solar Expert Posts: 6,800 ✭✭✭✭✭✭
    Options
    Re: Does anyone trust anyone's ballast calculations?
    Ok, that's definitely in-line with one of the two sets of numbers I'm seeing. The other is 600# for the same module. That's just too much, our customers won't accept that. What I need to do is convince myself 150 is enough!

    Maury


    I don't use a roof for obvious reasons but if I had to I would do what a sailor would do. Calculate the loading and then obtain materials that far exceed all calculations. I suppose your calculation takes into account the loss of strength due to exposure? If the customer won't accept it and your worried about your reputation then walk away. Better yet, do what a sailor would do, take the customer out in a storm.

    In the end there is only two kinds of weather! Good weather & Sailing weather!
    "we go where power lines don't" Sierra Nevada mountain area
       htps://offgridsolar1.com/
    E-mail offgridsolar@sti.net

  • BB.
    BB. Super Moderators, Administrators Posts: 33,491 admin
    Options
    Re: Does anyone trust anyone's ballast calculations?

    Another way to look at it:
    • 150 lbs/17 sqft = 8.8 lbs per sq.ft. ballast
    Is that enough in your area (on that roof) to keep things from blowing around in a storm? Panels are designed for ~50 lbs per square foot wind loading... What will be the weight per sqft of the structure+panels?

    Wind is a very local event. In my area where we live--pretty sheltered by hills and trees.

    A few miles away (where my poor plane was) open draw, flat land, no trees, on the coast, facing North/South just funneled wind (perhaps even a local twister) right at the wing.

    If yours is a very windy area (no natural wind breaks) and not much wall to break up wind from getting under panels (from any direction)--Will the planned ballasting/mounting make you feel comfortable.

    Perhaps the customer would allow you to bolt down the perimeter of the array. May help prevent the array from picking up at the edges?

    -Bill
    Near San Francisco California: 3.5kWatt Grid Tied Solar power system+small backup genset
  • Maury Markowitz
    Maury Markowitz Registered Users Posts: 9
    Options
    Re: Does anyone trust anyone's ballast calculations?
    solarix wrote: »
    For what it is worth, here are my calcs for wind load ballast that we used a couple years ago.

    Ok, one mystery solved. This equation calculates the force of the wind through what other systems call the "effective area", which is simply the exposed area in the direction of the wind. In the case of a panel at 30 degrees, the drag force is acting on the sin(30) of the panel, so it's going to be 0.5 of the wind force.

    That explains some of why the previous calculation I was given was so much higher. He scaled the wind force by the angle (from figure 6-2), but did not change the effective area. In other words, he calculated the drag force on a vertical panel, and then used THAT force to calculate the lift and drag. I believe that is wrong -- that your calculation is correct (after all, lift should be zero for a flat plate, aerodynamics 101).

    But then you lose me...

    For one, how did you get Cn from the chart on 6-18A? For the 30 degree line, I don't see a 0.9 figure anywhere. And shouldn't we select the worst case from this chart? It goes from a maximum of +2.6 to -2.5. Or is this simply the combination of some of the figures?

    Then you move onto the forces. You consider both lift and drag, but I do not see any attempt at overturning forces. Is the overturning force always low?

    Maury