I am looking at an ICF home with concrete roof and intermediate floors. Yes there are cheaper ways to build, but I am going for durability. The problem is that the concrete roof connects to the concrete walls which in turn connect to the concrete intermediate floors below. Or to put it another way, once you get pass the thermal mass lag, the floors will be as hot/cold as the roof. Not good. I can find lots of solutions for insulating the slab/footings, but nothing for the roof that doesn't have the thermal bridge. I want to at least have a concept of a solution before I bring in architect/structural engineer/energy estimators for how thick the insulation needs to be. What I am considering is to add a "second roof". The ICF walls and concrete deck would go up as normal (e.g. www.quadlock.com) except I would add some U shaped fiberglass re-bar tied to the rebar in the concrete roof and extend above the level of the pour for the roof. After the roof is poured, I would put down some EPS/XPS foam panels or maybe some glass foam (e.g. www.foamglas.com). Then place rebar which would be tied to the top part of the fiberglass rebar U sticking up and pour a second roof of light weight concrete. So at the eaves, it would be outside ICF foam, roof underlayment foam/foamglas and then top concrete to make a completely insulated thermal envelope, have the durability advantage of a concrete roof, the air tight seal between the concrete of the ICF walls joined to the concrete of the first roof and no wood to rot/be eaten by termites. Another alternative would be to use AAC slabs or pre-cast concrete slabs of suitable weight for the top roof, but I don't see a good way to attach them to the first concrete roof so that they would hold in a very high wind. I also know I could use something other than concrete intermediate floors, but I am looking for spans about 36' and any other solution would also have a thermal bridge through joist hangers or whatever. Can anyone suggest a way to solve the thermal bridge between the concrete roof to concrete walls of ICF to concrete intermediate floors or comment on the feasibility of my proposed solution? Any tweaks you can suggest?

Could you add a rough sketch of the house and roof.  A sketch would give us an idea of what you are trying to accomplish design wise.  The attachment of the roof to walls could be very different for a flat roof.  A steeply sloped roof could be challenging.

Here is a quick a dirty drawing of the concept. Sorry for the poor quality, but that's why you hire an architect rather than just hand sketches like this to the contractor. The ICF is unbalanced in size as I am likely to use more insulation on the outside than the inside (something quadlock has the flexibility to do). The exact thickness of the ICF foam, concrete walls, top concrete roof, foamglas (or alternative foam insulation), bottom concrete deck, bottom concrete integrated beam, various rebar, etc. will be determined by the architect/structural engineer, etc. I know that the roof will be pitched, I just don't know how steeply. The proposed house is in snow country so will either need much stronger bottom roof to hold snow build up (already stressed by top concrete roof) or steep enough angle to persuade the snow off the roof. Simple roof with single ridge and no valleys. Of course the steepness of roof that you can pour the concrete is another limiting factor. Considering foamed concrete (e.g. www.alliedfoamtech.com) which would allow for very light weight concrete (e.g. 10 pcf) and very steep angle of application (they show it being applied at 90 degree angle). Go to www.quadlock.com for a much better detail on how the quaddeck integrates to the ICF walls. May cantilever the top concrete roof out slightly from the wall to provide a better drip edge on the roof. Also may try to integrate guttering to the top concrete roof. Key issue is can I provide a slab to roof peak insulated thermal envelope (ICF foam and foamglas) and slab to roof air tight seal (concrete walls and bottom concrete roof ignoring window/door openings) and still have the durability of a concrete roof (top concrete roof) that won't fly off in a strong wind. If this would work, then passive house building standards (air tightness and insulation with minimal thermal breaks) becomes relatively easily with mostly off the shelf components.

With a concrete roof you'll need steel beams to act as rafters. I doubt rebar alone will reinforce concrete adequately over a distance to act as a joist. If you're in snow country you're going to have to cover that concrete with something anyway. The thaw/freeze/thaw/freeze will turn your concrete into sandstone otherwise. There is also the case of breathing which concrete may not do well. At my ski lodge in Europe we used a special one directional material which let's moisture out but not in under the roof cover. ICF's don't let a structure breathe too well. Concrete may hermetically seal the structure the same way. What ends up happening if the moisture has no way to escape is a fungus starts to grow. The one directional material is a two edged sword however. If your roof cover has a leak it is very difficult to figure out the exact location. I have to say it is much easier to build out in the California desert. Moisture and constant rain/snow do not come into play. This quaddeck business may be fantastic in some environments, but not pass the test in others.

Take a look at the www.quadlock.com site for the integrated concrete beam. It is like a concrete deck for a parking garage or bridge. For the 36' span, you are probably looking at a 4" thick deck and ~20" of integrated beam. I agree that the water issue has to be addressed. Internally, I was planning on humidity sensor controls that would activate the energy recovery ventilation when ever the internal humidity got above a certain level (e.g. 50%). Given the air tightness of the home, this is almost required. The freeze/thaw cycle is a concern. Most of my experience has been in the hot and humid part of the country where it would be unusual to have below freezing temperatures for more than 24 hours and nor more than a single hand full of days that would get that low in a year (and maybe none). On the other hand the top concrete roof would most likely be sealed and then covered with a membrane. So in theory there would be no water actually getting to the top concrete, let alone the bottom concrete. The foamglas is itself a moisture barrier, so if there was a leak in the top concrete that would stop the moisture from getting below that. But you would definitely see the top concrete start to disintegrate if it had a leak and the freeze/thaw cycle started up. If water got past both the top concrete and the foamglas, the bottom concrete should be sealed and it should protect the interior. Bottom line is that the roof should be stronger (and much more expensive) than any other alternative I know of and the redundancy should protect the interior if it starts to fail. There are concrete roofs that have lasted 100's of years in Germany, concrete roads, parking garages, etc, so the potential for the concrete roof to survive the winter is there if the system is engineered right.

I realized I had the concrete beams running the wrong way. It doesn't change the basic idea, but I corrected the drawing to show it correctly (and maybe make it clearer. But now the attachment button seems to be missing, so not sure how to upload the new drawing.

Alpine dwellings are perhaps the most difficult to build and maintain. The atmospheric conditions are a constant destructive force. The focus here is on roofs. For the snow to slide off on its own the roof has to be awfully steep. Snow is extremely heavy. Think of it like having a swimming pool on your roof. I am not sure these quad people certify their deck product to handle snow loads for the region you describe. Concrete overhead requires steel beams for support. Steel beams are very expensive. If you don't use styrofoam quad deck blocks for your roof you will need to insulate your roof somehow. From what I heard 60% of heat escapes through the roof and floor. Only 40% through walls. You won't need any sensors to tell you when you need to airate the place. Once or twice per day you'll need to open the windows and let the air circulate. Moisture still gets trapped in all kinds of nooks and crannies here and there.

My place, link below, was built using wood beams and boards. My only regrets were using the one directional moisture barrier and a fancy tile look-alike metal roof. The mountain man who built my roof told me it would get crushed. First winter, 4 meters of snow, the tile look alike was flattened in several places. A man can walk and jump on it and it won't go flat, but snow is a massive weight. Now when I cover my roof with metal I use only trapezoidal type designs. Links below. The moisture barrier I regret because I had a leak in the roof in one place due to a faulty screw connection and I didn't know it until damage was done. Moisture barrier hides leak locations.

http://vr.jtusz.com
http://www.balex.eu/budownictwo-mieszkaniowe
http://www.balex.eu/budownictwo-mieszkaniowe/blacha-trapezowa

We are talking US midwest, not alpine, but your point is well taken. I have been living in an area that might see 1/8 an inch of snow every 5 years. The place this will be built may see 8 inches in one storm. My suspicion is that we will have to pitch the roof fairly steeply for just the reason you cite. That in turn raises the question of how to pour a steeply pitched concrete roof. Which in turn is why I am considering the foam concrete which can be poured at a relatively steep angle. As far as roof insulation, there are a few options. First, the light weight concrete can have insulating properties. Second, the foamglas insulates. Both of these would be "outside". The quaddeck has insulation (I believe about R-34). But the concrete "I beams" would only have about an inch, so that will be of limited use as the cold will "telegraph" through the beams. What I will likely do is put in an "attic floor" that will run from wall to wall leaving a triangle of space above. In turn, I can put blown insulation however deep as needed. I am hoping the energy engineer can come up with the right balance between slab, roof and wall insulation. But I do know that I need to eliminate/reduce the thermal bridges, hence my original question about the double roof. So it won't be a cold attic because of the insulation in the roof. At the same time, it may be cooler than the rest of the house. This means I am going to have to potentially do some sort of specific ventilation for that attic space. It won't be the soffit and roof ridge because it isn't a cold attic. Probably some sort of gable mounted ERV, again potentially with a humidity sensor since it will primarily be there for humidity control, not fresh air in the attic. The concrete roof loading is going to be determined by the structural engineer. However, a 4" deck with 20" beams should have a 225 psf live load and 15 psf deal load rating with 2 #8 rebar per beam (one high, one low). Go to #6 rebar and the live load drops to 80 psf. So if we can keep the roof pitch steep enough, then we should be able to build to the load as needed. If we have to go with a flatter roof, the 50 year mean recurrence snow load is 20 psf, so we might be okay. Apparently if you put comment and you were the last commenter in the thread, you can't add an attachment. So here is the modified drawing showing the integrated beams running in the right direction. The integrated beams actually look like an I despite my drawing. Again, the www.quadlock.com site has much better detail drawings of that part of the roof.

I have 45 degree pitched roofs on my ranch east covered with 24 gauge sheet metal. Snow slides off only during a significant thaw at that angle. If you want snow to slide off you'll need a roof that looks like an upside-down ice cream cone. Either way, look at the pictures of roofs on the links. You'll notice there are rails/guards installed on the roof to prevent the snow from sliding off. You do not want the snow to slide off of your roof. Why? It will take the gutter with it. Remove the gutter and the bottom 3-4 feet of your house will be splattered with an ugly and destructive film of debris. A rain gutter keeps the bottom 3-4 feet of your house clean and free of rot.

How much slope is an open question. Snow on the roof is not a bad insulation. But the architect/structural engineer have to agree that the roof will be able to stand the snow accumulation. But the real question I am asking is about avoiding thermal bridges. ICF is great to avoid thermal bridges on the walls. Great air barrier with ICF so you can control air exchanges with an ERV. I can find lots of suggestions/solutions about how to insulate around/under a foundation/slab. But what I can't find is suggestions on how to avoid thermal bridges on the wall to roof connection for ICF that have a concrete roof (for an even stronger structure, like putting a lid on a box). Especially if you also have concrete for intermediate floors. I don't want my concrete roof be a giant radiator to heat/cool the house. I've proposed a solution. I'll attach a copy of it again. The basic concept is a second concrete roof joined to the first concrete roof by fiberglass rebar and separated from the first concrete roof by some sort of insulation (e.g. foamglas). What I am hoping is that someone can point me to an alternative solution, point out a flaw in my proposed solution or have someone with some experience in the matter tell me that it looks reasonable and is ready for me to propose to the architect/engineers. Given that I am having trouble finding any solution, I suspect that the architect won't have this as a standard drawing he just needs to update. I appreciate any thought you can give to the problem.

Probably one of the reasons you're not coming up with much is because what you are proposing very well may not work. With the pitched roof, concrete weight and snow load are going to have create some hellish shear loads at the roof/wall junction. I really doubt rebar will be anywhere near enough to handle the shear, especially if you put an insulation layer between them. That's akin to adding a layer of grease under a heavy box. It will slide a lot easier.

The second concrete layer is questionable, IMO. It's kind of like a slab on top of a teflon blanket. It's going to want to slide downhill, especially with snow on it, and I don't think fiberglass rebar is anywhere near enough to handle that shear load.

Concrete walls and roof snow loads add up to a lot of load on the footers, particularly for a two story house. The concrete roof will add a whole lot more.

On my ICF one story house the tallest wall was 13'. With no roof load bearing on it I still needed a 16" wide footer. In four places where roofs and wall met at a corner I had to add 36" square pads because of all the load from the wall and roof.

You really need to have an engineer look at your idea and rough out the expected loads, both downward and lateral shear. From that you can get a better idea on the feasibility of not having a solid concrete roof/wall junction.

Could you use something thin and light like StuccoMax or like the GigaRoof coating instead of the second slab?  I already have that question to the GigaCrete distributor so I do not know if their thin coating would work on a steep pitch.

dmaceld

Thanks for the feedback. I agree that the architect/engineer is going to be the final say here, I just want to have the best idea I can some up with to start the ball rolling.
I see what you mean about shear. The weight of the bottom half of the top roof will want to slide off away from the top half of the top roof.
But isn't that what the rebar in the top roof for? For that shear to happen, the rebar in the top roof (not just the fiberglass rebar connecting the two roofs) would have to shear. While an unreinforced slab of concrete would easily shear, I am thinking that the steel in the rebar wouldn't shear. Then it becomes a question of the concrete shearing off from the rebar, which should be a relatively straight forward engineering question. The fiberglass rebar would help with the shear, but I was thinking of it primarily in terms of anchoring the top roof to the bottom roof in a high wind.
If the shear is an issue, and adding more rebar doesn't solve the issue, I can think of two possible solutions. One is to transition from concrete on the top roof to a metal roof. I could use fiberglass J bolts or the like to attach the metal roof to the bottom concrete roof. I don't like this as well as I have seen too many rust streaked metal roofs (I am looking for durability with the minimum practical maintenance). The second possibility would be to add some sort of battens to the top of the insulation between the two roofs. This would help with the slippery slope part of the shear problem.
I'm not sure how much the fact that the top roof would be on both sides of the slope impacts the problem, but it does need to be considered. It isn't just a matter of a concrete slab sliding down a slick surface (because the inverted V of the roof would prevent that), but the concrete shearing away from itself.
As far as the footings go, I agree. Starting from the basement and going up to the roof eave, we are looking at about 33' of concrete. Add two concrete intermediate floors and a couple of concrete roofs (and the loads they are transferring to the walls) and you have a lot of weight on the footings. It gets even more interesting when you start looking at insulating the foundation. One possibility is just monster footers. Expanding the footers at the junctions where the walls meet makes sense. The other possible solution is to use lighter weight concrete to help reduce some of that weight (e.g. the foam concrete mentioned above in the thread). I am also not sure what the soil is like, but I suspect that it isn't the limestone bedrock I am more used to.

Alton
Interesting idea. If I used the heavy 20 oz mesh that was draped from one eave, over the peak and down the other side to the other eave, then the stuccomax shouldn't shear. I am not sure how durable the stuccomax would be when faced with direct rain on the roof vs rain hitting it on a wall. I am also not sure how well it would work in a high wind. I wouldn't want the whole stuccomax membrane lifting off the roof, so it needs to be anchored to the bottom roof somehow. I am having trouble visualizing how you get that connection from the stuccomax to the concrete roof.
One thing I like about the idea is that I could integrate concrete rain gutters into the top roof (which makes the shear issue worse) and extend the stuccomax from wall, over the gutters, over the roof and down the other wall. That would make the entire envelope of the house one monolithic membrane. The downside is that now you have the question of the stuccomax having to stand up to the running water in the gutter. That is really going beyond the original intent of the wall covering. If you get a response on using the stuccomax on surfaces other than the walls, I hope you post the answer. I have thought about using the plastermax on the bottom of the quaddeck in place of drywall. It would make a fine fire resistant cover for the quaddeck foam and a good durable surface. But I don't know how easy it would be to apply it on a ceiling.
Gigacrete does show their stuccomax on foam panels for a pitched roof in conjunction with their gigapanels concept, so the idea of it being on a roof shouldn't be a totally new idea to them. Perhaps fiberglass J bolts to the steel frame of a gigapanel used as roof insulation and then covered with stuccomax would be an idea worth exploring.
Thanks for all the feedback.

I'm not familiar with all types of roofing, but off the top of my head I would think the most maintenance free roof you can have is corrugated steel roofing made of Corten steel. Once it rusts, which is fairly quick, it stays that way until it's gone! Fasteners may have to be redone every 10 or 20 years, or so, but that would be about it.

Concrete won't be maintenance free. You have to coat it with something, as you show, but that coating will need renewing periodically. Don't count on stuccomax lasting forever either.

I didn't think about rebar passing over the ridge. That might be enough to keep the slab up there. You must be planning on some hellish winds if you're concerned about holding down a concrete slab so the wind won't pick it up!

Obviously it's your house and your money but I just can't for the life of me see any real benefit for using concrete for floors and roof for a residence. I'd stick to ICF for the walls, hang joist or truss supported wood floors inside, and use a truss roof with heavy sheathing and the whole roof strapped down with dozens of hurricane straps.

The other option is to use Steel SIPS for the roof if you don't want a wood roof.

Concrete roofs are better suited for hurricane and tornado areas but for 90% of the rest of the USA, they are probably overkill. As mentioned, it can get pretty pricey with a concrete roof and depending on the design, can get pretty complex.

I would check with Florida based ICF companies and see what they offer and have to say in regards to concrete roofs.

I'm not sure a steel SIP or metal roof is going to get past the thermal bridge issue. The thermal path I see would be metal on roof to concrete walls and then through floor joists to the interior. You could potentially cut down on this with a fiberglass sill of some sort that was between the roof and the walls, but for that matter, you could do the same with a concrete roof.

I also have a problem with all the penetrations you have to have with a metal roof/seams on a steel SIP roof. The difficulties of getting a good air barrier, the dangers of leaking, etc. all tend me away from it. Personally, I don't like the rust look, though that is pure aesthetics.

But in researching the metal roof, I did come across a potential solution. closed cell spray polyurethane foam (ccSPF). This can be applied above the roof (metal, concrete, etc.) and then have a membrane applied to the ccSPF. The ccSPF is good at sealing any penetrations (e.g. those associated with a metal roof). The ccSPF is insulating so a coat as thick or thicker than the ICF foam helps make a continuous insulating barrier around the whole house. The roof version of ccSPF tends to be about 3 psf, so minimal roof loading. The ccSPF can also be applied to a flat roof to create a slope (i.e. apply more foam to the center and less to the edges, then rasped smooth). All of this tends toward the possibility of a roof that is less steep, simplifying the concrete roof issues, etc. A 1/12 or 2/12 roof may make a lot of sense then as the weight of the snow and the ccSPF can be calculated and accounted for. The snow that lands and stays (since the well insulated roof will minimize melting) can act as a further insulating barrier during the winter.

The ccSPF does require a membrane on top. The membrane has to be renewed ~15 to 20 years. Depending on the membrane used, it can help keep the roof cooler based on the reflectivity and emissivity of the membrane. While I would love to find a maintenance free roof, I don't expect to find anything made by man to fit the bill. But I do want to find a roof that degrades nicely (e.g. when the membrane starts to wear thin, the ccSPF below continues to keep water out). You just have to periodically check the roof.

Let me tell you the advantages I see of the concrete roof/intermediate floors. First it makes for a stronger building. Just as the concrete walls of the ICF are going to be stronger than wood or steel framed walls, the same is true for the concrete decks. Plus because the concrete decks are integrated to the walls, racking, wind, etc can better be resisted. The open spans I am looking for also tend me towards a concrete floor. A steel truss floor that is the same depth of the concrete with integrated concrete Ibeam (e.g. 24" depth truss on 12" centers) would be in the same ball park as far as span, but it would be much less live load bearing capability (almost an order of magnitude less). So putting a cast iron claw foot soaking tub, a spa, pool table, etc. higher up in the house becomes a chancier proposition. Plus, we want the laundry facilities up near the bedrooms, not in a basement. Just listen to a washing machine enter the spin cycle on an upper floor to realize why you want the stability of a concrete floor. Further, just as the ICF walls helps cut down on noise, so to do the concrete floors between levels of the house. The floors won't creak either.
One advantage of the steel truss to the concrete floor is the ability to route ducts, etc. easier. The steel truss, like the concrete, is an engineered system so the issues of as designed vs as built can at least somewhat be minimized.
The cost of a concrete roof is higher. But if cost was my number one issue, I wouldn't be looking at ICF in the first place. I also see it as buying insurance. While the insurance rates can be lower, I am also talking about what happens if a tornado or earthquake comes to visit. It is hard to argue that a house with the additional strengthening of concrete roof and intermediate floors isn't going to be safer than the alternatives. Which in turn raises the question of what do you value you family's life at? Saving a few thousand dollars can be costly.

Thanks for the replies. The feedback is inching me toward a better solution.

I admit to not being an experienced home builder since I've only built one, the ICF house I now live in. I can't help but think you are reflecting your experiences with conventional construction done in usual fashion, not the high quality that can be done. I have a Jacuzzi sitting on top of a floor that is constructed of 9 1/4" deep I joists, 12" on center, with one layer of 3/4" Advantech subfloor, 1/2" gypsum underlayment, and ceramic tile. There are no cracks in the floor tile and grout.

The laundry room is the same except the joists are 16" o.c. The floor does not shake throughout the house when the washer is spinning. There are no cracks in the tile. Both showers are mortar base and sitting on 7" LVL joists. Again, no cracks. But I do have the advantage of short spans. The longest is about 14'.

Sound transmission can be mitigated with insulation and isolation.

I'd have to look at load tables but I have a hard time believing concrete will carry more load than steel trusses, unless the beams are pre-tensioned, something that's not practical for a residence, IMO. Concrete also flexes, but obviously not as much as wood or steel. The load carrying capacity of concrete beams in commercial buildings and highway bridges comes from the pre-tensioned cables, not from the concrete or the rebar.

For tornado protection build a safe room. Less costly than making the whole house a safe room. For earthquake protection I'd rather have a wood floor above me than a concrete floor. More flexibility and a lot lighter. Not so likely to come apart and fall.

This is just my opinion. It may be worth about half of what you paid for it!

dmaceld,

I'm not sure that the concrete intermediate floors are that far out of line. The can be poured at the same time as the walls, so the cost of the concrete pump truck isn't going to be much if any additional. The labor is probably going to be in the same ball park as putting in the steel frame and deck. You do have to pay for the scaffolding to brace the walls, but that shouldn't be to outrageous as the bracing can be reused (if nothing else reused on this job). This gives you a floor ready for flooring and a ceiling ready for plastermax.

Budgetary material pricing for the deck, concrete and rebar is ~$5.33/sq ft. I don't know what the steel truss for a 2 feet deep x 34 ft span x 4" (span table recommendation for that length), decking (above and below), connections to the wall, additional insulation for sound proofing, etc. but I bet it isn't that much less.

If you say the labor is the same (I suspect that the concrete labor may actually be less) and the steel truss is half the material cost, then I am looking at a ~$4K up lift for this home. For that $4K I get a much better product (in my opinion). I also get a product that is less dependent on quality of the construction crew. Any time you build a house, you have to make trade off's somewhere. For me, I would rather put my money here than some of the other alternatives. Your mileage may vary.

As far as the span tables, these are from span tables for manufacturers of the deck and steel truss. The deck can be built with up to 800 psf live load vs the same span with a steel truss would e 80 psf. Given a minimum building code of 40 psf live load, I may be able shave off on the concrete/rebar and steel truss. Part of this is figuring out the loads in the house (e.g. pool tables, hot tubs,, etc). I'll have better budgetary numbers after dealing with the structural engineer.

I don't know if there would be any savings in reducing the number of trades involved.

Posted By yetanotherjohn on 13 Mar 2012 12:33 PM
I'm not sure a steel SIP or metal roof is going to get past the thermal bridge issue. The thermal path I see would be metal on roof to concrete walls and then through floor joists to the interior. You could potentially cut down on this with a fiberglass sill of some sort that was between the roof and the walls, but for that matter, you could do the same with a concrete roof.

I also have a problem with all the penetrations you have to have with a metal roof/seams on a steel SIP roof. The difficulties of getting a good air barrier, the dangers of leaking, etc. all tend me away from it. Personally, I don't like the rust look, though that is pure aesthetics.

If you get good quality Steel SIPS, like those that are G90 rated and then cover it with a standing seam metal roof, there will be no "rust" with the metal SIP roof.

One thing you have to remember is that while a concrete roof is very strong, in an earthquake scenario (which I don't know if it applies to you), a concrete roof overhead is much more deadlier than a lightweight roof. If it should fail, the concrete would crush you. A Steel roof will resist lateral loads as a SIP has tremendous compression strength.

In regards to air barriers, furring the interior ceiling with 4" channels, adding R-13 fiberglass, and then 5/8" drywall will create an air and sound barrier. With a 12" Steel SIP (R-50) and then R-13 fiberglass, you will have a R-63 roof. All while only being under 18" in width.

A standing seam metal roof sitting on top of a steel SIP, when properly done, should never leak and be good for 30+ years of maintenance free work. Concrete roofs require sealing every so often to keep from water penetrating.


Just my 2 cents...

The point on the heavier roof is well taken. Of course, a steel SIP panel falling on you isn't going to make your day either. The home would be in the new madrid earthquake fault zone. I hadn't really been focusing on earthquakes much. From what I can find out, properly reinforced concrete can be pretty good at resisting earthquakes. A lot of the assumed features of the building (e.g. shape) are already optimized for earthquakes. The problem is that "properly reinforced" essentially means more rebar which increase cost of material (more rebar) and labor (to install the rebar).
Ultimately, the structural engineer is going to have to lay out the options here. Part of this is the calculation of the chances of a fire (probably highest) vs tornado vs earthquake. But I appreciate the point you make as one more dimension to consider.
As far as the air barrier goes, you right that a steel SIP can be sealed correctly and make a good air barrier. But that is assuming it does. The cast in place concrete roof and wall naturally makes an air tight barrier that depends a lot less on the level of craftsmanship of the installers.

On thermal bridging, my understanding was that the steel SIP had steel connections between the two steel panels of the SIP. As such, to the extent there is a steel connection between the two sides of the SIP, temperatures on one panel will bleed through to the other panel no matter how much insulation is between. O the other hand, I haven't study metal skinned SIPs much, so they may have a way to beat the thermal bridge.