Construction - Insulation/Watershed Umbrella for the Annualized GeoSolar System - Cont'd

The first post on the insulation/watershed umbrella delt with its configuration as described by John Hiat in his self-published book, "Passive Annual Heat Storage; Improving the Design of Earth Shelters".  This post is an attempt to explain why the umbrella is so essential for the function of the AGS system.  The "Annualized GeoSolar" terminology, by the way, is taken from a paper by Don Stephens that complements and advances Hiat's original design; the "insulation/watershed umbrella" terminology comes from Hiat.  A thorough discussion of the AGS system is found in three earlier posts:   first post, second post, third post.  The umbrella is covered in the third post.

The Function of the Umbrella As Part of the AGS System

The insulation/watershed umbrella increases the size of the thermal mass that is heated passively by the summer sun via the solar collector and the nine conduits running beneath the floor.  Without the umbrella, the thermal mass would consist of only the concrete floor, the rock sub-base and that part of the underlying soil that remained dry.  The concrete earth contact walls would also contribute but not the earth behind them because they would have to be insulated.  The thermal mass would still be considerable because our ranch house is spread out so much. But, even so, why take a chance on the solar collector not being able to keep up with the heat loss from the mass during the cold months?  Better to risk an overheating problem than an under-performing system.

Heat will be lost from the house mostly in four ways:  (a) the envelope of the house will be the main source of heat loss irrespective of how tight or insulated it is, (b) some heat will be lost directly from the thermal mass outward through the insulated foundation above the umbrella and (c) some heat will be lost through the energy recovery ventilator that uses fresh outside air, even in winter, for indoor air quality. The heat losses through the envelope and ERV will be made up by heat moving out of the thermal mass into the house.  The heat losses through and under the foundation are simply lost until recharged by the solar collector during next warm weather season,

Late Winter Blues
However, there will come a time at the end of winter when the heat that is readily available from the top layers of the mass has been used up and heat from lower down may not migrate fast enough to the surface to keep up with the heat loss through the envelope, foundation and ERV.  In which case, the ambient indoor temperature may drop a few degrees until Spring when the average outside temperature rises above that of the house at a time when the solar collector has already begun to reheat the thermal mass.  So it's all about getting enough heat into the thermal mass before the cold season then miserly keeping it there to compensate for heat losses from the house as long as possible in order to minimize a chillier indoor temperature in late winter and wait out the recharging of the solar collector in the spring.

The rate at which heat leaves the mass during the winter will be slowed somewhat by heat gain through the south-facing windows as well as heat generated by living in the house -- cooking, lighting, human bodies and hot water use.  Both Stephens and Hiat have already warned that the ambient temperature will be lower for the first couple of years until the input from the solar collector is capable of hitting our target of 74 degrees year-round.  At that point, it remains to be seen just how much the temperature drops in late winter/early spring, if any.

How Does It Work?

Heat seeks cold so the exterior winter environment is a prime target for the heat in the house and in the thermal mass.  Heat loss from the house is covered in other posts while heat loss from the mass is our focus here.  It exits the periphery of the mass in the winter as high up as possible.  In our case, the insulated concrete foundation and umbrella-insulated footing force most of the escaping heat to pass under the footing.  With our insulated concrete wall and without an umbrella, the top of the footing would be the departure point for the heat. With the umbrella, the departure point moves laterally to the outer edge of the umbrella.

According to Hiat, it takes 6 months for a unit of heat to travel through 20 feet of dry earth. Once the outside average temperature is lower than that of the thermal mass, heat loss begins.  The first sketch shows what happens with an umbrella in place. The dashed and dotted lines represent the interface between warm and cold earth. Dashed line "A" depicts

Click on image to enlarge for viewing.

the maximum of heat storage possible at the end of the warm weather season.  If the umbrella is 20' wide as Hiat suggests, the dashed line "a" represents the maximum of amount of heat that could be lost in 6 months. Notice that the umbrella virtually eliminates any heat loss from under the house, assuming that the insulation in the umbrella stops all heat loss.  In reality, its R-factor is sufficient to stop enough loss to keep that part of the floor nearest the wall much warmer than it would be without it.

At one point, in order to save a few dollars and a little effort, I toyed with the idea of making the umbrella 16' wide and it wasn't until I sketched out the two options and realized how much heat would be lost from under the house did I forget about the narrower umbrella. Even if the ranch style footprint would still provide enough thermal mass to compensate for the additional heat loss, the floor next to the exterior walls would probably be uncomfortably chilly towards the end of the heating season.

The second sketch shows the maximum amount of heat loss without the umbrella ("C" would be the maximum of heat available and "c" would be the maximum of heat lost in 6 months).  The +/-20 foot floor expanse between the two dotted lines would undoubtedly be uncomfortable and more so the closer you come to the outside wall.

Role of the French Drains
The French drains are 10' below the floor and essentially limit the vertical dimension of the thermal mass to 10" which should be more than adequate for AGS system. Heat migrating into the living space from any deeper would not be very helpful because it would barely make it through the floor before the cold weather season was over.  If the water table were to rise to the level we know it can (by monitoring piezometers for several years before beginning construction), it would bleed heat away from thermal mass. The drains make sure that it stays at the harmless level of 10" below the floor.  For more on the French drains, go to first post, second post and third post detailing the design, fabrication and installation of the drains.

2 - 3 Year Time Delay
The umbrella largely solves the problem of conserving as much heat as possible within the thermal mass.  The remaining question is whether our design for the solar collector/conduits will pump enough heat into the thermal mass to fill it out to the maximum.  According to both Hiat and Stephens, one should expect it to take 2 - 3 years to max out.  Afterwards, Stephen says, the problem could be too much heat such that some of the system would need to be mothballed.  In order to speed things up, we may wait a couple of years to install the sun shades over the south-facing second story windows so that the summer solar gain will go into the thermal mass and augment the solar collector heat.  And, if push comes to shove, we may install infrared heaters as supplemental heating for the first couple of winters, particularly if the code requires some sort of thermostatically-controlled heat.

Earth Contact North Wall
The foregoing discussion does not take into account the north wall.  Without the umbrella next to it, it would have to be insulated on both sides for all or nearly all of it height.  By moving most of the vertical insulation horizontally into the umbrella, only the top three feet or so need to be insulated vertically, viz., that part exposed above the grade, that part next to the topsoil covering the umbrella then a foot or so below the umbrella.  And, by reorienting the insulation, most of the two-story wall and the earth behind it creates so much thermal mass as to make an earth sheltered roof discretionary or unnecessary.

Incidental Heat from Other Sources
Heat during the cold months will be lost principally through the envelope of the house, especially through the windows, and through the energy recovery ventilator as it exchanges +/-30% of the indoor air every hour. Most of the replacement heat will migrate out of the thermal mass but there will be incidental heat from other sources as well. Most will come from sunshine through the south-facing windows.  But human body heat, heat from lighting (which we hope to minimize with LED), cooking heat and heat lost from hot water during bathing will also contribute. And, as mentioned above, we may invite the summer sun in through the second story windows for a year or two.

Cooling Season
The function of the AGS system during the cooling season is pretty simple.  Our goal is to maintain a year-round floating temperature in the low 70s, expecting it to rise above the average floating temperature a couple of degrees by the end of the cooling season and fall a couple of degrees by the end of the heating season.  Heat seeks cold so, at these temperatures, any summer heat penetrating the envelope will go immediately into the thermal mass without altering the ambient temperature, or at least not until the last weeks of the cooling season when the cumulative amount of heat uptake is more than the top layers of the thermal mass can absorb and distribute at the 20' per half-year rate.  Heat gain through the envelope will not come through the windows.  The windows will be shaded by overhangs that block the summer sun but admit the winter sun.  The exception could be the delay of overhangs for the second story windows as mentioned above.

The Rest of the Story
At the time of this writing in early winter, I had excavated for the umbrella without knowing for sure when actual installation could begin due to the weather.  The excavation was a matter of removing enough soil to allow the bottom of the umbrella to rest on the footings and slope downhill -- about two feet of soil had to be removed.  

By going ahead with the excavation when the likelihood of installing the umbrella yet this

Insulation protecting the footings under the foundation walls

year was pretty remote, the footings were a little more exposed.  So I purchased 2" thick expanded polystyrene sheets, sawed them in half lengthwise, laid them over the footings and weighted them down with stones.  Eventually, the insulation will be recycled as vertical insulation on both sides of the concrete garage walls.

Hang on, additional posts on the umbrella will follow as it is installed after the spring rainy season.

Construction - Damp-proofing the Earth Contact Wall

Is Damp-proofing or Water-proofing Really Necessary?
Yes and No.  "Yes" because the building code requires it.  But perhaps "No" from the standpoint of needing it long term.

The probability of the earth contact wall getting wet from surface runoff is low. The insulation/watershed umbrella will extend northward from the wall for 20' and the backfill behind the wall will be sloped away from the wall in three directions and end in swales to direct the water around the house and northward to a wet-weather creek at the north edge of the property.

The probability of the wall getting wet from a rising water table is non-existent due to the elaborate French drain system we installed 10' below floor level to lower the water table 
(first post on the French drains, second post on the French drains and third post on the French drains).

Is a Footing French Drain Really Necessary?
Yes and No.  "Yes" because the code requires it and because it may have some advantages until the backfilling is complete and the umbrella is in place.  "No" is probably the correct answer for the long-term because the chances of water reaching the bottom of the wall will be slim and none except perhaps near the ends of the wall. And the bottom of the wall is 7" below the floor level and the footing drain is below the top of the footing, making it a foot or more below floor level.

Ideally, the backfilling should be done in shallow lifts (layers) and each lift compacted.  My intention is to let nature do the compacting while I finish the house. Now that the wall is damp-proofed, I plan to backfill to a depth equal to about half of the final height of the backfill, or 5 or 6 feet, then let it compact itself over time. Eventually, the backfilling will be carried to completion and the umbrella installed.  In the short-term, the footing drain will be useful for draining any water soaking down through the initial lift so as to eliminate hydraulic pressure against the back of the wall.

Rather than encompass the fabric-covered drain with rock, I used sand that not only enveloped the drain but also was several inches deep on top of the footing so as to provide a passageway for water from the wall to the drain under the backfill.

Choosing a Damp-proofer
The fact that water at the wall will not a problem made it tempting to go with an unsophisticated material such as an asphalt based non-fibered roof and foundation coating at 11 cents per square foot.  My concrete supplier however recommended the Platon Waterproofing System at 22 cents per square foot.  The 11 cent differential amounted to $124 to cover 1,128 sq ft which seemed by far to be the best value.

Overview of the damp-proofing membrane; notice the sand
bridging from the membrane to the footing drain (click
on photo to enlarge for detail)

In order for the asphaltic coating to cover well, some areas of the wall would have required parging to cover up surface porosity in the concrete.  Some of the wall also would have had to have been power-washed to provide a reliable bonding surface. And I was not too keen on working with an asphaltic material.  With the Platon system, there is no need to "water-proof" the wall; it provides a drainage plane between the membrane and the wall so that water drains immediately to the French drain instead of going through the wall. And it could be hung without regard to the condition of the wall in less time than it would take to coat the wall, much less parge and wash it.  So, coughing up an extra $124 with the promise of less hassle, made the choice of the membrane a no-brainer.

Installing the Drainage Plane Sheets
Spoiler alert:  Despite researching the installation of the membrane, I misunderstood which side faces the wall only to find out when it was too late that we had installed it backwards.  Fortunately, the side facing the wall was sufficiently irregular as to facilitate drainage.  This is written after living in the house for nearly three years without seeing any water leakage, thanks in part perhaps to the insulation/watershed umbrella overlaying it.

The deadmen divided the membrane into four linear sections. We cut

The washers (left), removed from housewrap nails (top),
were mated with the recycled Tapcon screws (right) to
fasten the top of the membrane

each section to fit then stood it on the footing while fastening. We wrapped the material around the corners onto the deadmen for 9" or so. We used Tapcon screws to fasten the top of the membrane to the wall after mating the screws with the orange washers from housewrap nails. The manufacturer sells an attachment system but the Tapcon screws and washers worked fine at nominal cost, especially since we salvaged the screws from the wood braces that we had used for the insulated concrete forms. The membrane would have been unmanageable working alone so I was happy to have help from stepson, Keith.

Where the membrane got shorter as it

Close-up of the membrane; connecting the dimples on the
wall side are vertical channels to shunt water downward

approached the ends of the wall, the sections were laid flat for cutting.  We also pre-cut the membrane for the top 4' of the 12' high section of the wall and stored it until, several months down the road, the backfilling reaches that level.  It can then be installed while standing on the initial backfill instead of on ladders. At that time, it will be shingled under the lower section (not over as with roof shingles) which means the screws holding the lower section will have to be removed, the upper section tucked in behind it and the screws replaced.  Shingling under means that the edge of the lower section is has to be attached firmly so that it is not pried loose by soil when the backfill is done.

Construction - Concrete Work - Slinging the Gravel Sub-base and Pouring the Slab Floor

This is the sixth and final post on the concrete phase of construction (of the living quarters of the house). The slab floors for the garage and the screened porch will follow eventually. (Click on any photo to enlarge it for detail.)

Preparing for the Gravel Sub-base
The plumbing rough-in bunched the conduits together so tightly, especially those going to the bathrooms, that air space existed under them and any downward pressure on many caused them to flex.  Since the 3/4" clean gravel of the sub-base would not find its way

Ready for the gravel sub-base; the conduits and
waste water pipes have been packed with sand, the
grade beams and piers have been marked with paint
and the grade stakes are in place; black plastic covers
the PEX tubing to protect against ultraviolet radiation
for several months until the roof is in place

under and between most of the conduits, the concrete floor would not be properly supported.  In order to remedy the situation, I hand-packed, as much as possible, sand into

the air spaces and left extra sand on top of the conduits.  As much as I hated to see rain a few months ago, I eagerly awaited a downpour to compact the sand and carry the extra into the crevices below and between the conduits.  Mother Nature could not have been more cooperative -- no sooner than step-son Keith, handyman Pat and I finished distributing the sand, she dumped 3/4" of rain overnight that seemed to do the job just fine.

Under load-bearing walls and posts, the thickness of the slab had to be doubled in order

A wooden form blocks off for the bathtub drain; marking
paint delineates a grade beam on the left and a pier
adjacent to the wood form; tall grade stake in the fore-
ground, shorter spikes mark the corners of the grade
beams and a pier

to form grade beams and piers.  These were laid out and delineated first with 6" outdoor spikes driven into the soil then with marking paint on the surface of the soil.  In order to preserve access for connecting the drain to the tub eventually, the area under the bathtub drain had to be blocked out.  A circumscribed, tall wooden form was securely anchored in place to block out the gravel and the concrete.

Finally, the operator of the slinger truck needed reference points for spreading an even thickness of gravel at the proper depth.  Accordingly, grade stakes in the form of 12" spikes were pushed into the soil to form a grid that more or less positioned them 15' apart.  The rotating laser and a hammer was used to position the tops of the spikes at the +/-5" fill level for the gravel.  The spikes were then sprayed with marking paint to give them visibility.   The spikes that delineated the grade beams and piers were driven much deeper so as not to be confused with the grade stakes.

Slinging the Gravel

It took four truckloads of gravel and a long half-day to sling the gravel into place.

The slinger truck in action

The driver/operator skillfully laid down an even layer without getting much more than a sprinkling in the grade beam and pier areas.  Three of us with garden rakes cleaned out the grade beam and pier areas and finished leveling and smoothing the rest of the gravel.


Compacting the  Gravel Sub-base?
Our plans called for 5+" of 3/4" clean gravel under the concrete.  Apparently, usual protocol calls for pouring concrete over the the slung gravel without compacting it first.  Our

Compacted gravel sub-base; burlap covers the black
plastic over the PEX tubing to protect it from wind damage

undisturbed soil under the gravel was more than dense enough to have supported the concrete without the gravel so it made no sense to top it off with gravel that was not compacted.  And the slinger truck driver confirmed our conviction by saying that the amount of shrinkage that could be expected from compaction would be 3/4".   I immediately had him top off the original layer with another 1" of gravel to give a finagle factor for compaction . Thank goodness there was a plate compactor in the family and it immediately saw some serious use.

Pre-pour Tasks
The designated pour day was too windy due to an approaching cold front and had to be postponed for a couple of days in the span of which rain was forecast.  In order to keep the soil dry for the concrete trucks, we spread 6 mil plastic sheeting and weighted it down with enough junk to withstand the stiff winds . Unfortunately, the effort proved fruitless because the rain was minimal. 

Two of Jamie Schulte's crew appeared the day before the pour and used a laser to snap chalk lines on the concrete wall and set grade pins designating the height of the pour.  As soon as Jamie and his crew had unloaded their equipment on the day of the pour, they used the plastic sheeting to fit a

The inaccessible areas were poured using a conveyor
truck; the accessible areas were poured directly from
the chute; notice the plastic moisture barrier in place

moisture barrier over the entire gravel sub-base.  There were three floor drains that also needed to be suspended at the correct height over the roughed-in PVC pipes.  By using metal drains for two of them, I was able to drill, tap and insert three set-screws in each one for securing them to the PVC.  For the shower stall, it was only necessary to wrap foam around the PVC as a space holder for the brass drain to be installed later.

Pouring the Floor
The depth of pour was targeted to be 5" and level with the top of the foundation wall. The square footage was 1,680.  In lieu of metal reinforcing mesh, the mix was

Beginning the final finish 6 hours after beginning the pour

strengthened by the addition of fiberglass fibers and using 8 bags of cement per yard of concrete for a crushing strength 5,000 psi.  Some of the concrete was poured from a conveyor truck, the rest directly from the chutes on regular ready-mix trucks. The amount of concrete needed -- 29 cubic yards -- surprised even Jamie.  The considerable square footage of grade beams and piers, which double the thickness of the concrete, was probably the major reason for underestimating the volume.  In any case, the cost of the concrete, along with the surcharge for the conveyor truck, was about double what I estimated originally.

Wrapping Up the Floor
Concrete shrinks as it sets, causing random cracks.  Control joints are used in an attempt to direct the cracks to where they will be least problematic.  They are sawed into the concrete to a depth of about 1/4 of the floor thickness after the the concrete has had time to set but before cracking.  Jamie and a helper snapped chalk lines and used both a walk-behind and portable concrete saw to make the cuts.  Fortunately, we were able to hide most of the control joints under future partitions and a couple could be partially hidden under the future stairway and kitchen island.  Since it will be quite a while before the finished floor goes in, I am hoping all the cracking will have occurred, whether in the control joints or not, and the ceramic or porcelain floor tile will hide them without the crack expanding more later and cracking the tile.

The pour was done in mid-November and would be susceptible to cold weather damage if not protected.  Accordingly, Jamie recommended a product called "Cure & Seal" as a temporary coating lasting about three months before letting loose, which will get us through the winter, and one that does not compromise the bond of the thin set to the concrete when the tile is laid.  I tried applying it with a cheap garden sprayer but found it faster and easier to roll it on.  Five gallons cost $120 and was only a tad more than we needed for 1,680 sq ft.

Insulated Slab Edge
A final note.  The advantage of pouring the floor flush with the tops of the insulated concrete forms is that the slab edge is separated from the exterior environment by 5" of expanded polystyrene insulation.  Without the slab edge insulation, heat would be conducted in and out through the exposed concrete (thermal bridging) and alter the floor temperature adjacent the wall accordingly.  All energy certification programs require slab edge insulation (see first post on certification and second post on certification  (the latter is more germane to slab edge insulation)).

Construction - Concrete Work - Plumbing and Electrical Rough-in Under Slab Floor

This is the fifth post on the concrete phase of construction and I apologize for its length but countless surprises challenged my DIY-ness and I would be remiss if I did not pass them along for whatever they may be worth.

With regard to the concrete per se, all that remained was to "sling" in the gravel sub-base and pour the slab floor. However, both tasks had to wait until the plumbing that resides below the concrete had been roughed in. The rough-in comprised not only the typical waste/sewer system but, in our case, the complete water supply system as well. And a considerable amount of electric could be roughed in before the pour. (Reminder:  Click on any photo to enlarge it.)

Waste Plumbing Rough-in
In conjunction with pouring the footing under the front wall of the house, a short section of 4" PVC pipe was inserted below the footing forms with a foot or so extending

Waste plumbing rough-in; notice the that it connects with
and exits through the section previously installed
under the footing

beyond the forms on both the outside and inside to facilitate running the sewer line. The same was done with 2" PVC for the water line. The outbound sewer line connected directly to the 4" but the water line merely used the 2" to hold space until a 1" water supply pipe between the house and the water main at the street could be installed.

Installation of the waste lines was a matter of trenching into the excavation just deep enough to have the correct fall for proper drainage.  The downstream pipes were well below the grade and the upstream most pipes were almost at grade. The only thing that complicated the installation was trenching through soil that was more like stone than dirt, thanks to a layer of what old-timers called "hardpan" that presented even more of a challenge by not having seen a rain for almost a month.  The good news was that it will make a great base on which to rest a house.


The installation of the waste system was less complicated than our iteration of the water supply system but it does give meaning to the old saying, "Trifles make perfection but perfection is no trifle".  It requires a lot of know-how and attention to detail.  Fortunately, between two friends and I, we had enough composite knowledge to put together a system that passed inspection the first time.

PEX Rough-in for Water System
Our water supply system comprises PEX (cross-linked polyethylene) tubing as much as possible.  PEX was used extensively in Europe for potable water several decades before finding its way to the US in the mid-80s (it was used in the US before then for radiant floor heating but not for potable water).  Even though it is not used yet by many tradesmen, it meets code in most places and has many advantages over copper and plastic pipes. (For a thorough discussion, go to design guide for PEX plumbing.) Its major advantages for our project are simplicity, significantly lower cost and DIY-friendliness.

The myriad design options for PEX are well covered in the above link.  Suffice it to say that, for our project, only two major decisions had to be made. First, do we run separate "home run" lines to each appliance and fixture or do we run trunk lines to areas like the kitchen and bathrooms then branch off to the individual appliance or fixture?  Second, do we run the lines beneath the concrete floor or run them more conventionally through the framework of the house?  We opted for home runs and below the slab.  Actually, the two are linked in that only continuous tubes of PEX, free of any fittings, can be buried.  By marrying home running with burying, the rough in for the supply system was essentially complete before pouring the floor, much less waiting for the framing to be done, and at a fraction of the cost of doing it later.

Underground Installation
If PEX is buried, it needs to be protected from injury while the rock sub-base is slung in place and the slab itself is poured.  Moreover, it must be protected from chafing where it enters and exits the concrete.  These issues are usually handled by trenching into the dirt, covering the PEX with dirt, sand or pea gravel and providing short sections of conduit for the tubing where it leaves the dirt, sand or gravel and passes through the concrete. My lack of experience made me cautious to the extent that I decided to use conduits for the entire length of each home run -- from a couple of feet above the floor level at the upstream end near the water main and water heater to a couple of feet above the floor at the downstream end in wet walls near the appliances and fixtures.  The loosely-fitting conduits gave me the security of being able to retrieve the PEX if, for some reason, it wasn't working out as intended.  Also the conduits obviated the need for trenching in that they were robust enough to lie on top of the grade and withstand the slinging of the rock sub-base and the weight of the concrete.  Most were buried in the gravel, a few showed after the gravel was slung but barely.  Using so  much PVC did increase cost but not significantly.

When the decision was made to enclose PEX in conduits, I experimented in the aisle of my favorite home center using a length of 1/2" PEX, various diameters of Schedule 40 PVC pipe and 45 degree fittings.  As a result, I decided that 1 1/2" PVC pipe would be the smallest possible diameter for ease of insertion of the PEX for the 20 - 30' distances required and for the bends necessary to enter and exit the concrete. And, supposedly, the tubing could just as easily be removed from the conduits in the future if necessary.

Wrong!  As a trial, I glued together the PVC for one of the longest runs, tried to push the 1/2" PEX tubing through it and found that it hung up and couldn't be budged past the rough edge of the second 45 it encountered.  I tried to run an electrical fish tape in from the other end so as to pull the PEX and it too hung up immediately. Consequently, it was necessary for all 20 home runs to push the PEX past each fitting before gluing the fitting.  Fortunately, the PEX was not effected by the PVC primer or the adhesive so there was no danger of gluing the PEX to the conduits.  If a PEX line needs to be retrieved and replaced in the future, the thought comes to mind that a plumber's snake would negotiate the rough fittings and could be used as a fish line for pulling the tubing through the conduit or, better yet, the old tubing could be used to pull the new through the conduit.  

As it turns out, 1 1/2" PVC was the right choice.  PEX is a stiff material with a strong memory that wants to stay coiled up. Over distances of 35", 1 1/2" barely allowed the PEX to be pushed through.  Even the 10' sections of pipe between fittings offered sufficient resistance that I think a smaller diameter conduit could not handle easily. In one instance, I tried pushing two PEX tubes through in tandem and quickly decided that the effort was not worth the savings on conduit.  

Controlling the Entry and Exit Points for the Conduits
The biggest challenge to running the PEX system was to have it enter and exit the concrete at a perpendicular angle and be properly located inside the wet walls adjacent to fixtures and appliances.  To have any part of the conduit exit outside the wall would be unacceptable. 

The supply system rough-in completed:  Bathrooms to the
left, kitchen island sink and laundry to the right; future
manifold area in the background; second kitchen sink out
view to the right

Consequently, to control the ends of the conduits,  I shop-made salvaged 2 x 4 supports.  One set had 2 1/4" holes for the conduits and an identical set had 7/8" holes for the PEX.  Five foot sections of 4" PVC pipe were used as posts to hold the 2 x 4s in place.  Since the buried portion of the posts would remain in the concrete, I chose plastic posts over metal to eliminate the possibility of the metal rusting, expanding and cracking the concrete.  Wood posts were out of the question for termite reasons. Eventually, the PVC posts will be sheared off flush with the floor and filled with concrete. In many cases, the stub-outs of the waste plumbing rough-in could be used in lieu of or in tandem with temporary posts.  I chose 4" posts only because 2-hole metal straps were available for fastening the 2 x 4s to them and the pipes were left-overs.

Manifolds with cut-offs for each home run

Looking ahead, the beginning ends of the PEX lines will be joined to copper manifolds with individual cut-offs for each line, much like a circuit breaker for each electrical circuit in a service panel. Eventually, copper pipes will run from the tankless water heater to the hot water manifold(s) while PEX will suffice between the water main shut-off and the manifold(s) for the cold water supply. 

The PEX at the fixture/appliance end of the lines will be supported by brackets attached to the framing and stubbed out of the wall for uninterrupted connection to faucets and appliances.  Or perhaps, I will use a copper stub out. In either case, only three or four fittings per line will be necessary -- two at the manifold end and one or two at the fixture/appliance end. And, of course, there will be no need for cut-offs adjacent to the fixtures and appliances because cut-offs are built into the manifolds.

Bracket-supported PEX for direct
connection to faucets

PEX with copper connection for faucet

Materials Cost
The rough-in cost for the waste and supply materials was only $300. However, when the cost of PVC conduits was added, the total cost was $841. I consider the $541 for the conduits to be cheap insurance against damaged PEX lines and not having the option of retrievability in the future.

Other Advantages to Home-running
Compared to trunk and branch systems, the home run system requires more PEX tubing but far fewer fittings. Since the cost of tubing is much less than the cost fittings, home running is cheaper. And, a traditional copper pipe system would cost even thousands more due to the higher cost of copper and, for underground installation, someone who is capable of soldering with silver solder. Another advantage to an underground home run system is quicker hot water. Compared to a home run system run through the framing, the quickness derives from the shorter distance between the water heater and the fixtures.  A trunk in a trunk and branch system requires a larger pipe (usually 3/4") than the branches (1/2").  This means that, compared to a home run system with all 1/2" pipes, there is more cold water in the system to be pushed out by the hot water.

Home Run Electrical Conduits

Electrical conduits from future breaker panel area to areas
that would be hard to reach by running wire through the
 framework of the house, e.g., the island sink in the kitchen
to the right in the picture or the front wall of the house to the
left

After the water supply rough-in was complete, one last job remained before pouring the floor -- running electrical conduit.  Being able to make long runs sub-concrete simplified getting service to areas remote from the future breaker box, especially where the open floor plan (living room, dining room and kitchen) and cathedral ceilings would significantly complicate running wires through partitions and ceilings. I ran 3/4" conduit for two reasons -- being unfamiliar with pulling wires through conduits, I figured bigger was better and 3/4" gave the option of  pulling two or three sets of wires when only one or two might really be necessary.  If one or two sets remain unused in the future, there would be no downside

Stub-in for future floor receptacle in the form of a space-
holding tin can; the black pipe in the background, the water
 supply line coming in from the street, is wrapped with pipe
 insulation to keep the water cool as it passes through the
thermal mass under the floor when it is warmed by the AGS
 system eventually

since I fell heir to several spools of wire when helping to ready our former church for a remodel. 

The need for a floor receptacle was met by using a tin can to hold space during the pour.  A hole saw in the drill press, provided access for the conduit  through the bottom.  The rotating laser located the top of the can 2" above the pour, a stone supported it and a length of rebar piercing the bottom the can and driven into the soil at an angle stabilized it.

*     *     *     *     *     *     *     *
October 2019 Update
The final rough-in for the plumbing was completed four years later.  Here is the link to the later post on the plumbing rough-in.  Of special interest is its description of the manifold, tankless water heater and the automatic overflow prevention, all part of the PEX home-run system.

Construction - Concrete Work - Insulated Concrete Forms

This is the fourth post on the concrete phase of construction.  It details the unconventional foundation under the stick-built walls.  The first post on concrete work covered the excavations for the footings.  The second dealt with pouring the wide footings and concrete walls.  A third detailed the narrow footings under the stick-built walls.  Here we dwell on the rather unique and unconventional foundation wall, poured in insulated concrete forms to support the stick-built walls.  (Click on any photo to enlarge it for closer inspection).

Shallow Frost Protected Foundation
Our foundation is based on a recent "green" innovation that has actually found its way into the International Building Codes.  It is called a "shallow frost-protected foundation" (see the earlier post on shallow frost-protected foundation).  Instead of the top of the footing for the foundation being below the frost line, it is protected by insulation so that it can be much closer to the surface.  Our plans call for a 20" foundation wall, i.e., the distance from the top of the footing to the top of the foundation is 20".  For termite reasons, the code calls for an 8" separation between the soil and the wood elements of the structure.  If this dimension is subtracted from the height of the foundation wall, the top of the footing is only 12" from the grade instead of the +/-30" needed to get below the frost line.  This represents a 46% dividend in terms of embodied energy in the concrete as well as a few dollars saved on the volume of concrete.  However, the forms do not come cheap (a little over $1,200 for the 164 linear feet  20" high).  It would have been almost twice as much if the forms were used to pour a conventional foundation that reached below the frost line.  However, it must be said that, for a first-timer, the installation was unexpectedly time-consuming and frustrating, due largely to the unevenness of our footing.

Amvic Insulated Concrete Forms

The brand of  insulated concrete forms (ICFs) that I selected was the Amvic System because it were available from a local supplier and was manufactured in Missouri, both of which help to minimize embodied energy which is an important sustainability issue.

Installing the Insulated Concrete Forms on the Narrow Footing
Let me say upfront that our installation had little in common with that detailed in the "ICF Technical & Installation Manual" for the Amvic System.  The need for improvisation was entirely my fault by not getting the footing perfectly level throughout.  In retrospect, it would have taken rented metal forms installed with huge precision to have met Amvic's spec of 1/4" variance for levelness.  But salvaged  2 x 8s, despite being a lot more work to install, not always straight-edged and impossible to install precisely, were free and produced a leveler footing than would have been possible by pouring directly against the soil in a trench.

The Amvic ICFs are available for wall thicknesses of 4" to 10" of concrete with 2 1/2" of expanded polystyrene ("styrofoam") on either side,  The styrofoam serves first as the outside and inside forms into which the concrete is poured.  The "Dixie Cup" walls are strong enough to handle the concrete by virtue of internal webs that hold the two sides together.  Secondly, the forms stay in place after the pour as R-22+ insulation.  Our ICFs were for 8" walls such that the overall thickness, styrofoam included, was 13".  The 20" height for the wall was achieved by using one course of blocks 16" high then adding a 4" so-called "height adjuster" course to the top.  The smooth top of the height adjuster also facilitated screeding the top of the wall when the concrete was poured.

Our house plans give the dimensions of the foundation wall in terms of the exterior plane of an 8" wall and the batter boards had been set to these exterior dimensions. The ICFs then had to be installed on the newly-poured footing so that the concrete poured into them matched the batter boards.  To this end, mason lines were strung one last time between batter boards so that the equivalent of corner posts could be established and marked with concrete screws on top of the footing.  A chalk line was then tensioned between screws and snapped to give the position of the inside edge of the exterior styrofoam, which is to say, the outside edge of the concrete when it is poured inside the styrofoam.

Chalk line wrapped around a corner screw and
 ready to snap for a layout line.

We used a technique for aligning the forms that the installation manual recommended, viz., 2 1/2" steel drywall tracks for precise positioning of the ICFs along the chalk line. The tracks were fastened with Tapcon concrete screws on the chalk line so that the outside styrofoam of the ICFs, when seated in the track, would give a perfectly straight wall.  But "straight" is not the complete answer -- the wall has to be level if the foundation is to be level.  Despite working so hard using wood forms to control the levelness of the footing, there was a variance of about an inch over 164 linear feet. Therefore, the variance had to be finessed as the ICFs were installed through a combination of shimming (raising) under the low spots and shaving the bottom of the ICFs (lowering) over the high spots (all of which was much more time-consuming than expected).

Leveling the ICFs Longitudinally 
Thank goodness for a rotating laser when working alone.  I used it first to survey the top of the entire footing at +/- ten foot intervals to see how much variance there was and where the highs and lows were located.  Based on that data, I picked a default height that was weighted towards shimming (raising up) rather than trimming the bottoms of the ICFs.  As a consequence, the highest shim was +/-3/4" and the most that the bottoms of the ICFs had to be trimmed was +/-1/4" and the latter occurred only over 10-15' linear feet distributed over three areas.  

Installing the Metal Track
The first task was to drill holes in the tracks so they could be screwed to the footing with Tapcon screws. The easiest way to do so was ahead of time with a drill press.The installation was started where the track could be fastened without shims at the default height.  As the installation proceeded around the footing, the following steps emerged as the best approach ("proximal end" means the end next to the last section installed and "distal end" means the end farthest from the last section):

-  Lay the track on the chalk line
-  Fasten the proximal end first guessing (based upon laser readings) as to the amount of

Friend, Glen, leveling
the track with laser
 guidance.  Click on
 the photo to see the
 shims under the track

 shimming necessary and placing the shim(s) adjacent to the screw rather than drilling through them 
-  Since part of the shims will remained in the concrete avoid using wood shims to avoid attracting termites; use plastic shims, either store-bought or made from styrofoam trimmings
-  Fasten the distal end of the track on the chalk line without regard to shimming
-  Check the proximal end with the laser; loosen the screw, adjust the shimming as necessary and retighten the screw
 -  Check the distal end with the laser, loosen the screw and shim as necessary
 -  Through each of the intervening pre-drilled holes in the track, drill holes in the
 concrete
 -  Add the screws; shim beside each screw, guessing as to the thickness; tighten the screw and check with the laser; if necessary, loosen the screw and finesse the shim height
-  Helpful hint:  Before shimming, press down on the track above a screw hole with the laser measuring rod to get a read on the shim thickness required then add the screw and the shim; doing so eliminates some of the trial and error and tends to speed up the process.

Leveling the ICFs Transversely 
So the above efforts leveled the outside styrofoam but the inside styrofoam was still a mess. The ICF blocks are sufficiently rigid that the inside foam goes along for the ride longitudinally but they still need leveling transversely. And, of course, transverse leveling also plumbs the wall.  We tried two methods but the one that worked best was to wait on the inside shims until the forms were leveled longitudinally.  Then, in conjunction with plumbing and leveling the forms, shims were added as necessary.

Attaching the Forms to the Footing
The disadvantage of opting for more shimming and less shaving to achieve levelness was that there was a space between the metal track holding the outside styrofoam and even more space under the inside styrofoam that did not have a track taking up space.  The manufacturer recommends using a proprietary foam and sprayer to fill the space and "glue" the forms to the concrete. The customer service person at the local supplier suggested, since our project was a one-time and small project, not to invest in the sprayer but to use a spray adhesive instead. It wasn't until we were ready to use it did we realize that the styrofoam must be in intimate contact with the concrete for the adhesive to work.   So how do we DIY a way to fasten the forms to the footing and obliterate the gap under them so that the concrete does not raise the forms or leak out under them? 

(This then is the point of departure from the technical manual).  First we ripped enough 1 x

The 1 x 2s were fastened to the concrete.  To the right,
one layer sealed off the bottom but, to the left, two
layers were necessary due to unevenness of the footing.

2s from salvaged lumber to circumvent the entire ICF setup on both sides, butted them against the forms and fastened them to the footing with Tapcon screws.  This stabilized the bottom of the forms from the lateral pressures of the concrete during the pour and largely sealed off the space under the forms (there was the equivalent of 70 linear feet, randomly situated, that needed a second layer of 1 x 2s and 30 linear feet that sat directly on the concrete). However, the 1 x 2s did not keep the forms from lifting as would be expected from adhesive foam recommended by the technical manual.

While actually exceeding the manufacturer's specs for spacing of vertical bracing, we used 2 x 4s resting on top of the 1 x 2s and stopping just short of the top of the wall. The verticals were drywall-screwed into the plastic webs that tie the two sides of the forms together -- the same ones that serve as "studs" for fastening the exterior and interior finishes later. Then we used drywall screws and Tapcon screws to anchor to the footing by toenailing the verticals to the 1 x 2s or, where they did not exist, directly to the footing.  As recommended by the manufacturer, the basic 16" x 48" ICF straight blocks in the first course were tie-wired together then extra verticals were added where adjacent blocks were still a little unstable or unlevel.

Strongbacks and Lateral Bracing
In additional to vertical bracing, there needs to be lateral bracing.  We chose to do it on the interior of the forms so as to have it out of the way of pouring the concrete. The verticals on the interior of the forms provided solid fastening for the braces.The verticals on the exterior of the forms, though anchoring the forms and supporting the strongbacks, did not receive

View showing the verticals, strongbacks and braces

any braces since bracing was necessary only on one side of the forms.  The exception was a "T-wall" that intersected the exterior wall.  Here exterior bracing was necessary because, without the bracing, the weight of the concrete in the leg of the "T" might cause a blow-out.

Strongbacks would have been overkill if the wall had not been topped off with the 4" height adjusters that gripped the course below rather tenuously and looked suspiciously vulnerable to displacement during the pour.  So, as an insecure amateur, I needed the strongbacks as insurance..

Garage Door Cutouts
The garage floor will overlap the foundation below the two garage doors.  Therefore, the

Cutouts for the garage doors 

ICF forms needed to be 9" shorter under the doors, which is to say the forms had to be shortened by that amount.  Despite leaving the 4" extensions off, the blocks below needed to be shorten 5" but doing so presented a dilemma in that the all-important tops of the webs in the forms would have to be cut which would severely weaken the walls of the forms.  I elected to

Garage door area - final configuration

cut away only the unsupported foam between the webs, thinking that I could finesse the cutouts after the concrete cured.  And the vertical rebars that were fine for the rest of the foundation were too long at the cutouts.  It was a simple matter to shorten them after they were stabilized by the concrete.

Rebar extending into existing wall

In order to transition the concrete from the higher foundation walls to the lower cutouts, 2 x 8 pressure treated blocks-outs were cut at 8" to fit inside the ICFs. They were anchored in place with 6" TimberLok construction screws that pierced the vertical braces (existing or added specifically to support the block-outs) and the foam before screwing into the treated members.  The block-outs were held at least 1 1/2"  back from the eventual rough door opening so as to accomodate two-by framing for the door opening and still maintain a 9' opening.

Tying Walls Together
The joint between the existing concrete walls and the new foundation wall was a "cold joint", meaning they merely abut each other but are not physically joined. In order to keep the relationship between them from shifting, the rebar from the foundation wall was extended into the existing wall by drilling holes a couple of inches deep for the rebar in the foundation wall. 

Pouring the ICFs
It took 5.5 yards of concrete, a conveyor truck and five volunteer workers to pour the forms

Forms poured and anchor bolts in place

with relatively little effort -- definitely a DIY-capable endeavor.  And I am happy to report that the installation of the ICFs was sufficiently robust to eliminate any potential blow-outs.

Quality Control Issue with the ICFs
The ICFs are understandably fragile. Some components were damaged in transit and were unusable.  The height adjusters that were 1/8 to 1/4" shorter than the straight blocks and had to be split in half

Poured foundation after bracing removal

and, even then, somewhat forced to place. The design of the corner blocks have been upgraded for better retention of cladding materials but the height adjusters have not been ungraded.  Consequently, both the adjusters and the corner blocks had to be modified with a knife to maintain levelness. Cutting and fitting two halves of the adjuster separately left space that I filled with minimal expanding spray foam, which actually may have been an advantage in that the adjusters were glued together and perhaps more stable.

Would I use ICFs again?  In a heartbeat, but only after more research before buying.