Perlite seems to be a cost effective insulation to use under concrete slabs but there appear to be some issues.  I've found 2 different basic techniques for using Perlite, leave it in the bags or loose fill.   Each presents issues.  The bags are nominally 4 cu ft and typically lay about 8" tall and are about 20"x42".  The light density grade of Perlite typically used to fill CMU cores is recommended for under slab use, it gives about r3.2/'".  If it is left in the bags it means 8" is the thickness everywhere which gives about r25 but the spaces left between bags will lower the actual r value, possibly significantly.  The 'leave it in the bags' method, also,  pretty well excludes a thickened edge slab hence requires significantly more concrete and reinforcement. The other method of use is loose fill (removed from the bags).  With the loose fill method cardboard is layed over the smoothed fill then a waterproof layer then the concrete is poured. The cardboard is needed to allow it to be walked upon, without leaving footprints, before the concrete is poured.  I'm thinking that for my basement floor I'll use a kind of hybrid approach as I have need for extra strength in the slab to support interior walls and my fireplace and cringe at all the concrete wasted if I made the whole basement floor 8" thick.   What I'm proposing is a loose matrix of bags spaced 3"to 12" from each other. in both directions, with loose Perlite poured into the spaces between bags (and possibly compacted).  This way I can have 4" of Perlite under my walls and 8" everywhere else with the gaps between the bags filled with loose Perlite and the bags are like step stones to walk upon while installing a layer of poly after "grading" the loose Perlite or while pouring the concrete.  The stepped slab uses about 1/2 the concrete and the result is still r23+.  FWIW this is the basement floor in a PWF "wood" basement and is poured after the basement's outer walls are up.  The basement's outer walls are 2x8 @16" OC with 3/4" plywood exterior sheathing (all wood basement grade PT) and insulated with r30 mineral wool bats and have a poly sheet on their inside (about r26).  The basement floor is an insulated raft that supports the entire house through an inner basement wall (2x6 @16" OC) sitting on the slab.  By adding (optional) r22 mineral wool to the inner basement wall the effective basement wall can be r44+.  Your comments please.

I would think the bags would get in the way of drainage and use a larger gravel. I think compression load at the wall edge should be the driving factor to determine if you are using 8" of slab vs thicken edge not r-value. Once that edge bends cracks the slab if connected. I'd keep the thickness of aggregate under the slab and walls the same since load is being transferred. Why not get rid of the slab, some of these guys are driving stakes and putting sole plates on them in cold climates they say is proven to last over 40 years, although I like 1000 of years methods better. Is the labor cost of PWF cheaper than concrete, sure has added bonuses if structurally sound and sustainable. Take note they did not insulate this PWF floor. I've seen too many issues with concrete I don't want anything to do with it. Last week we repaired another wall caving in for a client with steel I-beams tied to rafters. Stabilized insulated COB work, why buy into the concrete or wood industry battle just use much cheaper materials on site, dirt. :) I would use large deeper gravel under the walls vs slab to take out compression and drain better. I'd brace the walls against ground source side compression loads, shifts, frost heaving, etc....with steel diagonal wall braces and thick sheathing as possible, block studs to rafters and joist, screws vs nails. Check out the compression crush property of perlite vs gravel if using as load carrying in your design, size matters too, especially if on a sand-silt-clay soil.

https://www.youtube.com/watch?v=AwJlyd-g6X8

TLP,
      I failed to mention,  what is under the Perlite is a drainage layer of 6" of pea gravel topped with 6 mil Poly sitting on "undisturbed " soil. so the Perlite even in bags doesn't interfere with drainage,  I hadn't even considered a PWF floor! Thank you for the suggestion, it may well be better to simply eliminate the slab and go all PWF insulating the basenent floor joist cavities with loose perlite.. 

I just watch a bit of that video no other research on PWF other than seeing it is allowed by code. What are they driving long wood dowels into the ground and fastening sole plates to them somehow? Is there a fastener transferring load or is the sole plate just bearing down on the ground pins at some pitch? What does code say here? If you tell me what that critical # 1 joint looks like we can do some static analysis compared to concrete to get some cost info. I'm curious to, I am just too busy with walls, roof, and HVAC designs right now.

What HVAC system are you using?

TLP,
My understanding of PWF says the stakes you see in the video are simply an aid to getting things level. they are completely optional and have no structural purpose. What holds a PWF in place is simple gravity and soil friction, just like any other type of foundation only PWF doesn't weigh as much. Code hassles are a location dependent problem but PWF is generally accepted.

MY planed HVAC system includes an ERV exhausting from 2 bathrooms with it's return flow to a common hallway, a second ERV (Panasonic bath fan) in the 1/2 bath. And a 15,000 BTU/hr low temperature minisplit with the indoor head located in a closet with 5 "whispergreen" fans also located in the "closet" below the minisplit. The fans drive individual ducts to the perimeter of the house, return air flows through grills at the top of the closet. The fans are turned on or off by a current sensing relay that senses when the minisplit's compressor is on.

Liebler - is there a reason that you are not using a wood footing to support your interior walls?
Are the interior walls actually load bearing?
Why do you need slab thickening at the edges? Are the outside walls not supported on the PWF?
Do you have any concern with multiple vapour barriers in your stack up?

FBBP,
Though I'm no longer planning a concrete basement floor, having adopted the PWF wood floor concept, I'll answer your questions based on the design with the concrete floor. The idea was to make the slab the actual foundation, an insulated raft. The outer PWF basement perimeter would serve as the forms for the slab and not be the actual load bearing wall. In addition a center wall in the basement would support the first floor system. (the load transfer was through the floor system's rim joist which was secured to the inner face of the PWF wall and sat upon a wall supported by the slab) The question about multiple vapor barriers is good. Where I have that issue and limited drying is the PWF basement where there is exterior Poly and interior poly. This is what the PWF design guides recommend, the saving fact is that everything between can be wet without issues as it is all pressure treated ( absolutely no mold growth) or easily drained and water tolerant.

After pondering how to "air seal" the treated plywood of a PWF basement floor, I've switched back to a concrete slab. To avoid supporting anything, vertically, with the slab and needing either a thick slab or thickened edges, I'll carry the PWF down to the gravel drainage layer with very simple "pony walls" The "pony walls have an I beam like cross section with 2x6 flanges and a 2x8 web and sit directly on the gravel (this is how the walls would be supported with a PWF wood floor) . The top of the "pony walls are planar with the top of the thin slab which is reinforced with 15# of Helix per yd. The slab sits on perlite as I described initially, mostly in bags but filled to level with loose perlite. The slab thickness will be about 2 3/8" and it will sit on a 6 mil polyair/vapor barrier.

I'l try and catch up with you over the weekend.....The statement you made above about PWF being the same load transfer to the ground as concrete is incorrect. Concrete and wood transfer load to the ground differently, now you are wanting to look at a pony wall-concrete hybrid which is still different. Think about your design options and load transfer first, not air seals and insulation, they are completely different. The only thing they have in common is transferring load to the ground but, in different ways. There is reason concrete is popular. You have to be careful with wood paying no mind to utube videos and min. code, it is more nodal design through fasteners. Also, your first center load bearing wall design was transferring to the walls cantilevered, they are not just there as forms. It help alot if you could draw a sketch and post. Also look at soil properties at the basement depth wood may be a bad idea because the bearing area to the ground in far reduced, especially compared to a monolithic concrete slab thickened edge. It is a structural system, look at it that way, each component sub-assembly in a composite design plays a role. If you don't transfer and evenly distribute load from the walls to a foundation and soil that can handle it you are asking for issues.

TLP,
Insulation and air barrier continuity are what I'm trying to achieve along with reasonable cost & ease of construction. The choice of double stud walls for the main level was easy, but a suitable basement is a much greater challenge. PWF offers lower cost for a high r basement but requires a very good connection of the upper end of the basement walls to the floor deck ,this led to the choice of a double walled basement with the outer wall extending to the middle of the rim joist. The rim joist rests on the top of the inner basement wall which also supports the floor trusses. . For ease of construction I'd prefer to limit the amount of carpentry before the slab is poured. I was hoping to defer construction of the inner basement wall till after the slab is poured, I'm reconsidering this. The outer wall must extend below the slab to provide lateral support and as a byproduct it insulates the slab edge. Before the inner walls can be erected, the insulation and interior vapor barrier of the outer wall must be completed. If I build both perimeter basement walls before the slab is poured I can minimize the potential for different shrinkage of the inner and outer walls by designing equal amounts of cross grain wood in them. Such shrinkage could compromise the air sealing around windows and doors. With the I beam pony walls supporting the mid basement wall I'll have more shrinkage in the mid wall than the perimeter but no air sealing is affected. Perlite is a very low cost insulator but unfortunately it's structural capabilities are undocumented so I'm on;y depending on it to support the slab, which others have shown it can do properly.

liebler - do you have an engineered detail for that I beam pony wall?

Posted By FBBP on 07 Jun 2014 01:00 AM
liebler - do you have an engineered detail for that I beam pony wall?

NO!  It is shown as interior bearing wall support as part of the PWF wood basement floor in the "Permanent Wood Foundation Design and Construction Guide" published by Southern Forest Product Association.  It is on page 46, Alternate 2 in figure A6.  While the above drawing shows  in line joist and sheathing, clearly these elements do NOT contribute to the support of the bearing wall.

•••clearly these elements do NOT contribute to the support of the bearing wall•••
Actually, they have everything to do with the structural support of the detail. Abusing this detail will result in the 2 x 8 "web" overturning. In the detail, the joists prevent the over turn. With your application, when the load is placed on the top 2 x 6, there is nothing to prevent the 2 x 8 from rotating from on edge position to laying flat.
Please don't cheap yourself to death. An engineer is not THAT expensive!

but a suitable basement is a much greater challenge.

I'd just skip the basement and use a FPSF (unheated designs).

Yep the inline joist and sheathing, the entire floor grid(header and joist) keeps load away from the sleeper in Alternate 2. Remember I said think of structure as a system. Because of the floor grid cantilevering load to walls, as I said above, the gravels vertical compression load can be reduced. There is really not much of an over turning moment in the sleeper since it rest on gravel and is not connected so, it is not doing much to take out bending of the joist, the header and sole plate of the bearing wall are. Alternate 1 has more of a potential for one from a lateral seismic or frost heaving load acting at the bottom of the pony wall that is why you see the steel joist straps. I like this option better.

I'm not following what the problem is with air sealing the wood floor per the details shown in this doc. It's not like there are 100 mile hr winds in a basement. You could use a taped Zip sub floor, or depending on the finished floor air sealing should not be a problem.

Keep at it, it is a challenge finding the best solution for the lowest cost most never obtain.

Good one thanks, I'll read it. http://lavellecompany.com/wp-content/uploads/2012/11/PWF-Guide.pdf

Posted By Liebler on 07 Jun 2014 12:14 AM
TLP,
Insulation and air barrier continuity are what I'm trying to achieve along with reasonable cost & ease of construction. The choice of double stud walls for the main level was easy, but a suitable basement is a much greater challenge. PWF offers lower cost for a high r basement but requires a very good connection of the upper end of the basement walls to the floor deck ,this led to the choice of a double walled basement with the outer wall extending to the middle of the rim joist. The rim joist rests on the top of the inner basement wall which also supports the floor trusses. . For ease of construction I'd prefer to limit the amount of carpentry before the slab is poured. I was hoping to defer construction of the inner basement wall till after the slab is poured, I'm reconsidering this. The outer wall must extend below the slab to provide lateral support and as a byproduct it insulates the slab edge. Before the inner walls can be erected, the insulation and interior vapor barrier of the outer wall must be completed. If I build both perimeter basement walls before the slab is poured I can minimize the potential for different shrinkage of the inner and outer walls by designing equal amounts of cross grain wood in them. Such shrinkage could compromise the air sealing around windows and doors. With the I beam pony walls supporting the mid basement wall I'll have more shrinkage in the mid wall than the perimeter but no air sealing is affected. Perlite is a very low cost insulator but unfortunately it's structural capabilities are undocumented so I'm on;y depending on it to support the slab, which others have shown it can do properly.

Is there a detail some where in the doc that shows the double stud wall to floor joint you are describing?

Posted By FBBP on 07 Jun 2014 10:09 AM
•••clearly these elements do NOT contribute to the support of the bearing wall•••
Actually, they have everything to do with the structural support of the detail. Abusing this detail will result in the 2 x 8 "web" overturning. In the detail, the joists prevent the over turn. With your application, when the load is placed on the top 2 x 6, there is nothing to prevent the 2 x 8 from rotating from on edge position to laying flat.
Please don't cheap yourself to death. An engineer is not THAT expensive!

Where I intend to use this detail is in the middle of a concrete slab so instead of plywood sheathing and inline joist there ill be a 3 1/2" thick concrete slab offering far better support for the 2x8.  But your assumption that  the 2x8 could rotate is illogical,  The load wall above has it's center line in line with the center of the 2x8 and with simple vertical loading rotation of the 2x8 is impossible.  

Posted By TLP on 07 Jun 2014 06:06 PM

Posted By Liebler on 07 Jun 2014 12:14 AM
TLP,
Insulation and air barrier continuity are what I'm trying to achieve along with reasonable cost & ease of construction. The choice of double stud walls for the main level was easy, but a suitable basement is a much greater challenge. PWF offers lower cost for a high r basement but requires a very good connection of the upper end of the basement walls to the floor deck ,this led to the choice of a double walled basement with the outer wall extending to the middle of the rim joist. The rim joist rests on the top of the inner basement wall which also supports the floor trusses. . For ease of construction I'd prefer to limit the amount of carpentry before the slab is poured. I was hoping to defer construction of the inner basement wall till after the slab is poured, I'm reconsidering this. The outer wall must extend below the slab to provide lateral support and as a byproduct it insulates the slab edge. Before the inner walls can be erected, the insulation and interior vapor barrier of the outer wall must be completed. If I build both perimeter basement walls before the slab is poured I can minimize the potential for different shrinkage of the inner and outer walls by designing equal amounts of cross grain wood in them. Such shrinkage could compromise the air sealing around windows and doors. With the I beam pony walls supporting the mid basement wall I'll have more shrinkage in the mid wall than the perimeter but no air sealing is affected. Perlite is a very low cost insulator but unfortunately it's structural capabilities are undocumented so I'm on;y depending on it to support the slab, which others have shown it can do properly.

Is there a detail some where in the doc that shows the double stud wall to floor joint you are describing?

The direct answer is NO.  However this is one detail I have discussed with a licensed, registered Civil engineer who specializes in PWF designs.  

Just wanted to make sure I understand the upper floor to wall joint. If I understand correctly the rim joist is on the outer-inner wall and the inner wall upper sill is acting as a sleeper for the floor joist. Both the inner and outer walls are 2x8’s 16 OC. Primary floor loads are being distributed to the ground through the inner wall, primary roof loads through the outer wall. This is an excellent joint. You say you talked to a licensed PE about this so what are the vertical loads that the inner and outer walls are taking to the ground? How much positive margin is in your wall design? How thick does concrete have to be to take the same load as wood? What compression mix of concrete with helix? What soil compression allowable are you building on? What is the compression strength of perlite in bags and loose fill and the hybrid mix you are looking at? You have two double outer walls that are load bearing, and from what I can tell by your post an center bearing wall to span floor joist to. If you don’t know the answers to these question, since you have given no indication that you do, you need an Engineer. You have a large outer wall foot print, smaller center, that need to get load to the ground effectively.

As YOU said, you are finding it a challenge to design a basement effectively. You say the walls were easy but, failed to realize that at the time you were designing the walls(since they have to provide load paths to the foundation and ground) they had to be designed to. As I said, it is a system you don’t isolate. I’m not sure how you expect others out here to help you do so without having all the facts, no prints, no design guides, that should have been in the OP. To make matters more ‘illogical’ you appear to have up most concern for r-value and build sequence to cut cost. Cutting cost is great if the structure works or trust me you will see sustaining cost down-stream you or/a licensed PE may or may not see now. You are changing major foundation structure at the in-expense of cheap insulation material and robust build sequencing, when it should have been set and integrated with the PE discussions you had with the floor-to-wall and roof-to-wall joints to take out live a dead loads. To ask members for help with insulation and r-values without knowing the structural impacts is “illogical” to say the least, and a very good indicator you are over your head and need help as FBBP said. If your PE did not inform you on the above I'd find another one, they'll be gone by the time you have issues or have a contractual way out if they have a good attorney, or your state statue has a limitation.

TLP,
Not quite, the inner basement wall is 2x6 16" OC and for design purposes the inner wall is presumed to carry all loads except the exterior cladding. However the outer and inner main floor walls have the same details (headers etc.). so either the inner or outer main floor wall are, individually, capable of supporting the roof loads. The fact that they will share the roof loads is a safety margin. Yes I have a loading schedule and the presumed soil strength is 3000PSF. FWIW at the time of discussion with the PE I was planning on PWF walls straight through the slab with the perlite only supporting the slab itself and the inner and outer basement wall shared a 14" wide "footing plate". The major issue with perlite is it's unknown compressive strength and stiffness. It may be both simpler and safer to simply use EPS which is well understood, even if it is slightly more costly.