ASCE 7-22 Snow load revisions. Big load increases. 4

[manstrom]

I haven't spent much time with ASCE 7-22 yet. I was just made aware that the snow loads provisions are changing from a allowable measurement to an ultimate measurement. Similar to wind load revisions in previous ASCE-7 editions.

Load combination factors are changing:

​

The ASD factor is now 0.7S (down from 1.0S). LRFD is now 1.0S (down from 1.6S)

However, for the mid atlantic region, snow loads are increasing significantly from 25-30 psf to 60-70 psf.

For example, Washington DC is moving from pg=25 psf to pg=62 psf (43 psf ASD). When considering the ASD load factor, that is a 74% increase in snow load.
For LRFD that's 62/1.6/25 = 65% increase.

I wouldn't be surprised at 10-15% bump, but this is rather significant. Is there some other reduction factor that I am missing? I find it hard to believe that any houses or apartment designed before this code won't work with the new code by a significant amount. This has a big impact for wood framed structures and renovations.

https://www.structuremag.org/?p=19724

new map website

https://asce7hazardtool.online/

 

[Lomarandil]

Why? Why do we do this?

Regarding increased loads -- my perception may easily be incorrect, but I haven't heard of roofs collapsing due to snow in the mid Atlantic. Is there some rash of condominium and warehouse failures I'm missing?

Regarding this change from listing service level criteria to listing ultimate values -- is it just to head off an EIT who doesn't realize LRFD includes factoring loads with a factor >1.0? I mean, yes, that happens (unfortunately often), but solve that on the education side!

Don't keep widening the gap between what shows up on my drawings and calculations and how a layperson thinks. "Sure, the drawings list a 60psf snow load, but what that really means is that we expect the roof will function well for the 40psf of snow we often see". Or in the opposite direction, "What do you mean, wind speed of 150mph? We never get winds that high here." (And that's on top of the difference between gust speeds and what they see reported on the nightly weather report)

 

[Celt83]

What gets fun is when you couple the load increase with the reductions in concrete shear strength that happen in ACI 318-19.

I think some of the intent was to try and get jurisdiction away from having local provisions so that the hazard tool can be a one stop shop. As the article mentions this is also a result from the analysis of a significant amount of new data compared to the old maps.

For DC in particular the impact is similar to Baltimore. DC has always mandated a minimum uniform roof load of 30 psf where Baltimore mandated a pg of 30 psf, following the procedure from the article the net effect relative to pf = 30 (DC) vs pf = 21 (Bmore, 0.7*pg) the % delta for DC ends up lower.

 

[canwesteng]

Going from service to ultimate levels makes more sense intuitively, since that is what we really care about. wish Canada/EU would follow ASCE on this. Getting used to something new is always a pain, but I would rather know the actual predicted load with some percent chance of exceedance in the design period (like seismic), as opposed to the 1 in 50 just being increased by 50-60% as a fudge factor. I suspect without this change the snow loads were going up anyway - as the article notes, Baltimore already superseded ASCE 7-16 snow loads with a substantially higher value.

 

[dik]

...maybe trying to accommodate climate change?

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[phamENG]

I don't know - the article does a pretty good job of explaining it. I like the idea of reliability based design, so it doesn't bother me at all. It also probably helps that, being at the southern end of what is considered "mid-Atlantic", even these new snow loads wont exceed roof live load in most cases.

And I don't worry too much about the misconceptions that a layperson might get from my drawings. Those are there to inform other qualified individuals about the design criteria. With that said, I think it would be good to inform people about what these things mean. Like when a hurricane is coming with 110mph sustained winds (1-minute average) to a region with a 115mph 3-S gust wind speed, they should inform people that there is a high likelihood that their houses were not designed to tolerate the winds they are about to see.

When somebody questions why I'm designing for wind speeds so high, or complain that I'm requiring shear walls in a house, I'm happy to explain the difference between non-hurricane wind reports (2-minute average), hurricane wind reports (1-minute average) and what we design for and why. I also explain the rough statistical relationship - if your 2-minute average wind is X, somewhere in that storm system you have a decent chance of seeing gusts of Y. Something like that. I also explain to them that the sunny, 80 degree day when the contractor slaps up what he thinks is 'good enough' and does a 'shake test' to prove it is NOT what I'm involved for. I care about the day the hurricane arrives, the earth starts to shake, or in this case, the day the temperatures plunge and drop 2 to 3 feet of snow in an area that usually doesn't see more than 2 to 3 inches in a year.

 

[271828]

LOL. Now it'll be harder to tell if S or Lr controls. They have different load factors.

Also, windward and leeward drift lengths are computed differently. If one gives a little higher psf, I'm not sure it's safe to say that one controls. Probably usually would.

Changing the subject to wind: The new Ch. 32 on tornado loads will apply to Risk Category III and IV. That's quite a lot of buildings.

On the bright side, I think the other wind provisions (I think) will give the same pressures, but the presentation is a little clearer now.

 

[manstrom]

Another item I found is the Ct factor goes from 1.1 to 1.2 for a vented attic. That's another 10% load increase.

I hope I'm wrong, but what this means is that if I need to check a warehouse roof for a new 200# mechanical unit under ASCE 7-22, I need to check the roof for the weight of the new unit plus the increase in snow load according to the new code. How will this fly? This will require structural repairs to every structure doing a renovation.

I hope this doesn't carry over to IRC. Builders will lose it.

 

[Celt83]

"controls" does not exist under the current code provisions

 

[phamENG]

This will require structural repairs to every structure doing a renovation.

Not repairs, reinforcement. And so what? The code requires the roof to be more reliable. So it shall be. What part of Virginia are you in? In my neck of the woods, it's not going to matter. I just checked - ground snow load at my house is going from 10psf to 33psf. Once you adjust everything, it's going up about 60%. Okay. That's still 4psf less than roof live load. I'll get a little more drift that I'll need to pay attention to, but it won't break the bank. DC/NOVA/Baltimore will have more impact, but they also get a lot more snow. And remember, this isn't about average snow load, it's about potential extreme snow loads. We don't have frequent roof collapses in this area, but we do have those freak snow storms every 10 to 20 years that dump several feet of snow on the mid-Atlantic and cause a lot of damage as a result.

As for the vented roofs, it makes sense. Insulation requirements are getting tighter. Less heat is escaping from the building. So if less heat is being added to the system and the heat that's there has a way out that doesn't involve going into and melting the snow, you can end up with more snow on the roof.

I hope this doesn't carry over to IRC. Builders will lose it.

I can assure you it will.

 

[Lomarandil]

I'm good with the idea of uniform reliability, improved resolution of design criteria as data sets improve, and if that means a real increase in load, so be it (although I'd rather this be rooted in more than pure academia).

Here's a question. I always think of LRFD incorporating our design margin (or factor of safety) in both load and resistance factors. That we generally get a net reliability index between 2-3.

With the new trend of adjusting load factors to unity, the narrative I hear is that we are designing structures to withstand these ultimate level events without distress. Maybe that's not accurate?

Excluding wood design (grading for 5% chrarcteristic values throws things off a little) -- how do we feel about designing structures with resistance factors not that much less than one?

Put a structure up against a real 150+ mph ultimate wind event, or a 700yr MRI earthquake. How does it hold up? Material variation, geometric tolerances, contractor deviations, design errors and omissions. In my estimation, those have the potential to outweigh a phi factor at 0.75 to 0.9.

 

[Smoulder]

@Lomarandil there's no difference designing for 1.5*5 load or 1.0*7.5. Structural distress depends on how hard you push the capacity calc.

Construction and design mistakes aren't covered by the capacity factor. Except if you didn't need the capacity for its true purpose.

 

[Lomarandil]

Right, but are we selling the idea that our structures now perform against that 7.5 load the same as they have historically performed against 5.0?

 

[RWW0002]

I can get behind reliability-based design, but I still feel like the "design loads" that are shown on drawings and are conveyed to owners/contractors/general public should be service-level. Always.

@phamENG I disagree that design criteria is only intended for "qualified individuals" I think it is very much meant for the end user, and effectively communicating these loads is very much a part of our jobs. I get tired of having to explain "well yes your last building was 90 mph, and this one is 115, but it means the same thing.. but next year it can be 105, but that is really closer to 81 under the old standard." etc. Now we are going to have to do the same thing with snow, and it all just moves the general public and end-users farther away from being able to have a "feel" for the limits of their buildings.

As an example, after some high winds came through our area a few years ago (tornado-level - beyond current design-levels) there was the usual outcry about infrastructure and building practices. The news and city council's takeway - Blame it on the adopted code and engineers. Here is a paraphrased quote from the news and council "Currently IBC 2012 requires buildings to withstand 90 mile per hour winds. But IBC 2018 requires buildings to withstand 105 mile per hour winds."
Ha! Their "fix" was to adopt a lower design pressure in the building code!
I could go on all day about what they should have been focusing on - but at the end of the day by constantly changing the basis of the standard and getting farther away from not only what the general public is used to seeing but what can be expected to be actually measured over the life of a structure, we are only bringing confusion into the conversation when we should be offering clarity.

Floor live load is not communicated at an ultimate level (I assume so people do not get the wrong, (and unconservative) idea about floor capacity), and neither should snow, or even wind for that matter. This is why I am still a fan of putting Vasd on drawings, and I assume we will soon be putting Pg(asd)on them as well..

 

[phamENG]

Lo - I would say yes.

In the days of old, we thought we had a good idea of the load but the material and workmanship were less predictable. So we designed for a certain load but hit the capacity of our design with a penalty to provide an adequate factor of safety.

Now, we feel much more confident in the strength of our materials (except wood, as you mentioned) and workmanship through frequent testing and inspection procedures. But we're starting to realize that maybe we shouldn't be so confident in the applied loading.

So where before we might have said that a building can support 30psf snow load but it could actually support 50psf, we're now saying it can support 45psf and it can still support 50psf. It'll still be able to do what we say it will, but it won't do as much above and beyond what we say it will.

 

[phamENG]

RWW0002 - I agree that there needs to be more clarity. Perhaps local ASCE/SEI chapters should be more proactive about explaining things? The reason I say that about the drawings is because the typical end user never sees them. Or if they see them, they don't attempt to read or understand them. If there were a way to start from a simple, understandable wind speed, and end at a reliability based design I'd be all for it. But I don't think there's a good way to do that. And what about the weather broadcasts? They are inconsistent in their wind speed reporting. Do people understand the difference between normal weather report wind speeds and reported hurricane wind speeds?

Listing Vasd or some sort of Pg(asd) wouldn't change the way the building's used - the end user can't do much to increase the speed of the wind hitting their house or the amount of snow accumulating on it. They can increase the live load, though. So if you go from saying you can have 40psf on your living room floor to saying you can have 64psf on your living room floor, you'll also go from having occasional overloads into the 50 and 60psf range to actual overloads into the 70 and 80psf range that could cause a collapse. So in cases where reliability based design and hazard based design result in the same loading, stating the load as service level and factoring it makes sense. But where they don't align (which, conveniently, tend to be environmental loads that a clever human can't do much to change), designing to the extreme to ensure reliabilty makes sense.

Maybe, instead of listing a wind speed, we should just list an average wind pressure for MWFRS design? Or a sort of base shear value for the entire MWFRS? That would make it like snow is now. We provide a pressure, not much else. So unless you know what your snow density is, you can't translate it to a certain depth anyway. So you wouldn't be able to incorrectly compare it to a forecast or weather report.

 

[RWW0002]

@phanENG - I agree that simply providing some sort of pressure value is likely better than wind speed. Although I maintain that providing a 50-year MRI (around the design life of many structures) 3-sec gust (basically what most weather stations are reporting) wind speed is a pretty good stand-in. People could expect that their structure would have a pretty good chance of withstanding this wind without "cutting into the safety factor".

I have done this thread a disservice by moving away from snow though, and it is just as relevant, if not more, to the Ult vs Service conversation. New construction in areas with the "new" ultimate snow loads may think they are building way stronger than past builds, when in reality the capacity is similar, or maybe even lower. - Similar to the living room floor you mention above.

Engineer's design terminology and design criteria, for better or worse, makes its way into the standard vernacular within the building community. Our design standards become benchmarks to an extent for past and future designs. ex. "That roof was a 40 psf snow roof and it has stood the test of time.. The hurricane blew down that building that was built to 120 mph but this one that was designed for 140 mph fared better.."
Whether the general public really understand the difference in 90 mph vs 120 mph or 40 psf vs 20 psf (Snow, Live, lateral.. regardless) is important, but not super relevant (for what its worth, engineers kind of work to the same standards to an extent). What they understand is "this building was designed for X in the past and has worked or has not worked so I want to build my new one for X, or Y"

Whether right or wrong, reliability-based or not, by constantly changing the stated design criteria we are making the above comparisons difficult.

I maintain that the best way to communicate the stated design criteria is to keep it as close as possible to loads reasonably expected during the lifetime of a structure, and to keep them as close as possible to being "on the same page" as past design criteria. Surely there is a way to still have reliability-based design and report service-level loads - similar to floor live load.

 

[canwesteng]

The question with service level loads is valid - what do we expect the loads on our drawing to be user for? And what does ASCE expect them to be used for? I've previously told clients that their existing structures are good for X wind or X snow (both service level loads), and establish controls to prevent the areas from being occupied in case these are exceeded. If you were to do that with the new true snow/wind loads, I don't think there would be enough margin that this is comfortable anymore. That being said, the true loads can also avoid clients panicking if they see the snow load or wind load exceeded, say if 90 mph wind is forecasted and drawings show 85mph.

 

[phamENG]

The only reason I can think of that design criteria are listed on drawings are for code review. The building official can at least check and see if we entered the load calculation from the right place. Whether or not we did the rest of it right...some care, some don't. Unless the client is an institution with a dedicated engineering department, very few people ever look at that or ask me about it. I've had a few people say "I want my house to withstand a category 5 hurricane" and I have to explain to them what I can and can't design for. But that's about it.

I don't think there is a way to keep reliability based design tied to "average" loading or service level loading. Not directly. I guess there could be a service level map and then another map with location specific conversion factors that take that service level load up to a local ultimate. But that feels like it just gets even more convoluted. But maybe not.

It all comes down to communicating what we're doing. The stereotypical engineer isn't great at communicating - especially with people outside the profession. But I say it's worth it to step up the communication about why the design is better this way than to settle for a less reliable design to make the communication easier.

 

[RWW0002]

Building owners may not say "I want my building to withstand x" But they sure do say " Why did my steel package just increase by 15% or what do you mean this rafter won't span 15' - it sure did last week!" I work with some pretty smart non-engineers and they are definitely more than capable of understanding design criteria and have a good discussion on reliability and building performance.

I disagree about design criteria only being for code reviewers. Building owners and insurers have pretty vested interest in design criteria also, especially when Factory Mutual is involved and loss of production or equipment is as or more important than building loss. It also plays a part in post-disaster discussions about general design/construction and infrastructure. These discussions often get lost in the weeds of code iterations lately though..

I want to be able to simply say "snow/wind loads increased to X due to recent data" but instead or having a meaningful conversation with smart owners/contractors about why load increased I am having to explain the semantics of what how to compare loads between code cycles before the discussion even starts.

I guess I am going to have to just blame everything on inflation and leave it at that. Ha!