Sunday, March 25, 2018

Design - Rain Harvesting for Toilet Flushing

We looked for almost five years for a suitable location for our energy-neutral passive solar home. Our search was confined to rural areas because it never dawned on us that we would find the perfect site eight blocks from Main Street in sight of the St Louis Gateway Arch (that is, if there were no trees or houses blocking our sight-line). In those early days, I researched water conservation, including recycling graywater and harvesting rainwater, for the day when we would have well water and a septic tank.  (I am still the proud owner of  the book, The Septic System Owner's Manual, which wasn't exactly on the New York Times Bestseller List).   I even researched composting toilets.

Once we found the Collinsville site with its access to cheap "city" water and sewer, I lost interest in all but the usual water conservation practices such as low-flow faucets, shower heads and toilets, a front-loading washing machine and landscaping with native plants that do not need watering.  After all, there is not a lot of incentive for water conservation here in the Mississippi River bluffs, eight miles from the mighty river and only a few miles below its confluence with two other major rivers -- the Illinois and the Missouri.  As a matter of fact, our regional problem is too much water from major floods, never too little. 

An Awakening
Michael Webber's recent book, Thirst For Power -- Energy, Water, and Human Survival, has radically changed my thinking.  He makes the point, for example, that the amount of energy required to provide water and sewer service to the average household exceeds the energy the household uses from the grid for lighting, heating and operating stuff plugged into receptacles.  Here's why.  

It takes energy to pump our water from a deep well on the Mississippi River floodplain to the purification plant.  It takes energy to purify the water and lift it to our homes on the hills above the floodplain. This says nothing about the energy embodied in the manufacture of the materials that go into building a purification plant, the chemicals used for purification, and the maintenance of and upgrades to the treatment plant and the aging water pipes.

Energy inputs do not stop with potable water.  After it goes down the drain, it takes energy to treat it at the waste treatment plant before it is released back into the environment and for infrastructure maintenance and upgrades to our century-old sewer system.

So, thanks to Webber, I  now equate water with energy.  Flushing our toilets with rainwater would be on a par with having a photo-voltaic array. Unfortunately, the Collinsville Public Works Director didn't see it that way; when our design was sketched to him, he disallowed it but said he would kick it upstairs to Illinois State officials and we have heard nothing since.  Rather than pursuing the issue, we are keeping it on hold as an easy future add-on if circumstances permit.  

Meanwhile, I would like to share our design in case there are viewers who are researching  rain harvesting or haven't yet considered it.  Perhaps there is something in it that might be helpful.

System Design
Rainwater Catchment
Fortunately, the 1,400 sq ft roof area of the second story section of the house tilts 
Side view of the second story roof tilted northward

northward towards the concrete back wall that has a short stick-built wall on top of it. I had planned to locate an indoor cistern high inside the storage area (that we sometimes call our "vertical basement") located at the back of the house adjacent to the wall.  The run-off from the roof would be collected by the gutter and directed through the stick-built top portion of the wall into the cistern. A couple of websites provided what I needed to be sure that such a catchment system would yield enough water to flush two toilets year-round. 

The first helpful website was the Texas A & M site, which not only provides generic information on rainwater harvesting but also links to a handy calculator.  The calculator uses the average monthly rainfall, the size of the catchment area and the average monthly water consumption to calculate the size of the storage tank (cistern) necessary to meet demand.  All I needed then was the average monthly water consumption for flushing.

The second website provided a handy water use calculator for estimating the amount of water a household consumes per function -- such as showering, dishwashing and lawn irrigating.  With it, I was able to determine that the amount of water we would need for flushing, as a percentage of our total water consumption, would be 27% or 10,000 gal/year or +/-800 gal/mo.  Plugging this figure into the A & M calculator told me that, because our rainfall is fairly consistent month-to-month (varying only between 2.14" in January to 4.11" May), a +/-500 gal cistern should meet demand well enough that supplementation by city water would be necessary only under special circumstances such as the drought we had in 2012. 

Click on the image for better visualization of the storage area
against the back (north) wall of the house
The cistern would comprise a heavy gauge translucent plastic tank that can be found online and at local farm and home stores for around $500.  We would need a rectangular model that will be narrow enough to pass through our 36" doors.  It would need to have fittings for 4" intake and overflow pipes.  Mounting the cistern high (+/- 12') in the storage area would offer the option of directing any overflow back outside to a swale rather than dumping it into the sewer.

In order to avoid any possibility of cross-contamination between potable water and the water in the tank when city water was added, I would use a dedicated cold water pipe that did not attach directly to the tank -- it would be suspended a safe distance above the tank.  When turned on back at the cold water manifold, the water would drop through a large funnel DIYed into the tank top.  Alternatively, I could wait until the need for supplementation occurred then send water through the gutter using an exterior sillcock and garden hose.  An upgrade to the interior approach could follow if necessary.

The cistern would be plenty high to gravity-feed the toilets between flushes but undoubtedly slower than would a connection to pressurized potable water, which for us would be only a minor inconvenience if any.

The water coming off of the roof would need some filtration although less for our metal roof than for shingled roofs that shed the stone granules found in gutters. Debris from trees should not be a big problem because there will be no trees close by at least for quite a few years (as I will be documenting in a couple of months, we are reforesting the land behind the house with over a hundred bare-rooted seedling trees).  So, at least until the trees mature, a pool-type of inline filter between the gutter and the tank, perhaps preceded by a screen, should suffice. (I would have to flesh out the requirements for filtration as the time for installing the system approached.)  In drier climates with long intervals between rains, a "First Flush" approach is used whereby the first rain of the season or the first rain after a dry spell is used to clean the catchment surface before harvesting begins (as I learned from The Greywater Action site). Our rains are frequent enough that first flushing would probably be unnecessary.

Our toilets would have to be on dedicated supply lines or on one dedicated line supplying both toilets.  Otherwise, there would be no way to send rainwater to them without contaminating the potable water system.  As described in prior post, I installed all of the hot and cold water lines for the entire house as home runs, meaning all faucets, all appliances and both toilets have dedicated lines with individual manifold-level cut-offs on the storage room wall next to the water line from the street and the water heater. Without a home-run system, a rainwater system as a later add-on would be impractical.  In our case, it would be a simple matter to dedicate to toilet flushing one small manifold with two lines.  If future residents wished to deactivate the rainwater system, the toilet lines could be disconnected from the manifold, decontaminated and connected to a potable water manifold.  In a worst case scenario, the contaminated lines could be retrieved and replaced since they are encased in PVC conduits (prior post).

Return on Investment
A rainwater system would cost approximately $500 for the tank plus some change for PVC pipes, say, a total of $600.  If toilet flushing represents 27% of our water withdrawals or 27% of our yearly water bill, we would save approximately $54 per year. At that rate, the return on investment will take +/-10 years which we feel is no big deal one way or the other when compared to the golden opportunity to strike an easy blow for sustainability indefinitely.

Thermal Mass
The Annualized GeoSolar system for conditioning our house, that has been covered extensively in dozens of prior posts and quickly accessed by clicking on "Featured Post" in the left column above, depends on the thermal mass of the earth under and around the house and, to a lesser extent, the concrete floor and earth contact concrete walls. Since water is an even better heat sink than earth and concrete, a cistern full of water inside the house would somewhat augment the passive solar conditioning.

Potential Problems
If the water in the tank was much colder than room temperature, the tank would sweat. This might happen after cold rains, during snow-melts or in the unlikely event the cistern is supplemented with a large amount of ground temperature city water.  At this early stage of planning, I would be content to wait and see if the problem exists then install sheet plastic to catch the condensation and a plastic tube to carry it to a nearby floor drain.

Another potential problem might be algae buildup that is common for stagnant water exposed to the environment.  However, the tank would reside in a space without windows and without artificial lighting most of the time so photosynthesis may be minimal if not impossible.  In a worst case scenario, it might be necessary to add vinegar to the tank occasionally.  While a dark-colored tank controls algae outdoors, it would be redundant in the dark storage room; a translucent tank would be better at revealing any buildup.

Still another potential problem is mosquito breeding. The rain barrel on our current house has no openings to the outside similar to the one in the nearby photo; the water from the downspout goes directly into the tank and mosquitoes do not negotiate the downspout in order to breed in the tank.  Our rain harvesting gutter would have a downspout located in the middle of the roof to drain both halves equally then it would extend horizontally quite a ways to reach the cistern located in the storage space near the east wall of the house. The long horizontal pipe would be sloped excessively to eliminate any chance of any puddling so that its length and lack of standing water would discourage mosquitoes.

Zoning Issues
Plumbing codes for graywater, especially graywater for reuse indoors, are ambiguous and inconsistent.  According to Webber, the International Plumbing Code allows toilet flushing with graywater from showers and bathtubs but the Uniform Plumbing Code that is used more widely in the US does not.  Yet the The Greywater Action site shows the inconsistency of regulation between states and says the International code is even less helpful than the Uniform code.  So go figure.

When it comes to rainwater harvesting, regulation seems to be much more lax even with regard to using rainwater inside a dwelling.  The Greywater Action site on rainwater harvesting provides a link to Laws, Rules and Codes for each individual state.  Of the three items listed for Illinois, one leads to the full text of  HB1585 that merely defines rainwater harvesting and specifies that systems must be constructed in accordance with the Illinois Plumbing Code.   Our Director therefore seems to have a lot of discretion but remains cautious by not allowing for now toilet flushing with rainwater in Collinsville. 

An additional issue that came up in discourse with the Director was that the sewer fee the city collects is pegged to the water meter readings.  He was afraid that the reduction in water consumption due to rainwater flushing would automatically under-report our discharge into the sewer.  He did, however, seem open-minded about our continuing to pay the same sewer fee as we were before rainwater harvesting, or some other amount, that arbitrarily compensated for the discrepancy.

Saturday, February 10, 2018

Construction - The Last Major Dirt Work (Cont'd some more) - Rain Gardens

Anal retention alert:  "Rain gardens" are not exactly on the minds of most people so the detailed information about rain gardens here and in the previous post will undoubtedly test most folks' indulgence.  But I am basing the need for information on my own ignorance about the important sustainability role that rain gardens play.  Despite building into the side of a hill, they were never part of our original design.  It wasn't until a few years ago that our membership in the Wild Ones organization made me realize that our property would be the quintessential beta site for rain gardens.  And, since my green thumbed wife, Dorothy, and I were both hopelessly uninformed, I am assuming that most page-viewers here, despite their interest in sustainability, could be uninformed as well.

When people ask, "Why rain gardens?", I say that our intention is that every drop of rain that falls on our property leaves it underground and purified rather than on the surface carrying silt and pollutants.  What follows is the description of the initial dirt work involved with realizing this goal.

The treatment of rain gardens in the last post was of a generic nature.  The discussion here is about the five rain gardens that we need to control runoff from our hilly property, where the grade falls at least 30' from the highest elevation behind the house to the lowest elevation in front of the house.  At the time of this writing, the gardens were roughed in for a trial run before final contouring and topping off with the rain garden mix in time for planting with native plants next spring.  Out of the five gardens roughed in, two of the berms (dams) failed at the first torrential rain and had to be reconfigured.  

(Remember: the pictures can be enlarged by clicking on them.)
Encircled are the two highest rain gardens next to the
house in which we live

Rain Garden Siting For Our Project
The next highest is beside the garage of the new house
and is one of two that failed
The rain gardens at the highest elevation are on the property where we live east of the building site. They are modest and intended primarily to catch the runoff from impervious surfaces -- house and garage roofs and driveway --  before it can erode the hill sloping down towards the building site.  The next highest is alongside the new house garage but down-slope enough not to threaten the AGS system*. Its perc rate was a non-issue because it lies directly over and is drained by the east-most gravel-filled French drain that catches any deep water flowing towards the AGS conduits.  However, initially the garden failed -- it did not have the intended capacity and had to be dug deeper and wider and the dam raised.  At the time of this writing, it had not yet been challenged by another heavy rain.  Eventually, though, it will have to handle less water (see the caption for the last photo below) and will be reconfigured accordingly.

The two largest rain gardens lie further down-slope. 
The garden in the foreground is the one discussed in the
next section and the second one to fail;  encircled
is the berm that separates it from the street;encircled
 in the distance is the former retention pond
One, with the help of a continuous berm paralleling the street, catches all of the runoff from both properties on the east side of the new driveway. It, too,
 failed, perhaps due to the excess water from the other failed garden above.  It was modified and awaits further testing.  The overflow from it passes through the driveway culvert into last garden that is merely a shallower version of the retention pond that was intentionally maintained during construction.  The pond can be seen in the Google Earth photo in the previous post.  Any water the last rain garden cannot handle leaves the property as it did before construction began -- through a culvert under the street and into the neighbor's lake.

The overflow channels for all of the gardens will be
rip-rapped to control flow rates and to add interest;
large river gravel/stones would be more attractive but
are beyond our budget

Rain Garden Construction and a Battle with Glacial Till (an anecdote) 
Unfortunately, rain gardens for our property were not as simple of scooping out a shallow depression then adding a berm and overflow 
Over-excavation for a rain garden (4-5' deep); the
 big chunks to the right are glacial till
outlet on the downhill side.  I already knew going in that the gardens were likely to lie over the same glacial till (hardpan) that we had encountered throughout construction and during final grading.  Sure enough, hardpan was situated immediately below the topsoil at the most critical site.  Knowing that the hardpan layer would be only a few feet thick, I continued digging with the trackloader until I was below the layer.  The length of the ramp that was necessary for safe digging with the loader, coupled with a reasonably sized flat bottom in the hole, created a long, wide and deep cavity.  In the bottom, I dug perc test holes but an overnight rain rendered them moot.  The deepest part of the excavation filled with a foot of water that had drained away the next time I looked at it, a few hours later.  The water from an inch and a quarter rain a week later was gone reasonably fast as well.

So now the issue became how best to fill 
Same cavity after addition of a truckload of sand
the cavity back to rain garden depth without compromising percolation.  Thanks to input from friend Charlie Pitts, a Certified Naturalist, who had helped us lay out the gardens in the first place, we arrived at a solution for such an atypical situation.  We filled the excavation with a tandem truckload of sand to within +/- 3' of the original grade and plan eventually to top it off with a "Rain Garden Mix", i.e., a special soil formulated from topsoil, sand and compost that is available in bulk from a local supplier.  The remaining question is whether the distribution of the plants as described in the previous post will survive in a rain garden filled this deep with sand.  Maybe a different choice of plants will be necessary.  

The rain garden mix will be added to all of the gardens as soon as we know (a) that the gardens are functioning as planned, (b) the cover crop of grass on the denuded hillsides is holding the soil in place and (c) whether siltration from hillsides occurring before the grass is a factor has either filled the gardens to the proper depth or overfilled them to the extent that partial re-digging is necessary. 

Heavy Rainfall
The garden beside the garage was one of two that that
Rain gardens are not designed to catch all of the runoff from "frog strangler" rains.  Heavy downpours or worst yet, a series of downpours in a short period of time, produce more water than the rain gardens can handle so an overflow must be incorporated into the downhill berm.  Thanks to Charlie, we learned that the overflow does not have to be especially wide or deep because the water passing over it is typically more like that from a gutter downspout than through a roadside ditch.  Accordingly, after the berms were finalized, we sculpted shallow troughs through them which we covered with weed barrier fabric which will control erosion in the short term and weeds later.  At the time of this writing, we had not yet covered the fabric with fist-sized stones (rip-rap) that will slow the flow and prevent erosion downhill.  

Another Month of Dirt Work Should Finish the Job
In the St Louis area, January and February are the two months with the least amount of precipitation.  Our warmer winters with less
The berm has been reconfigured for another test run
frozen ground seem to allow more dirt work but winter cloudiness and what freezing and thawing we do get cause muddy conditions that complicate things. Nevertheless, my goal is to finish the dirt work behind the house in time to plant grass seed in March or April before the spring rains.

The garden beside the garage presently drains half of the
 area north of the house because the 
insulation/watershedumbrella has yet to be installed
 behind the house; after it is in place the final grade
 will direct runoff to the north instead of curling 
around toward the south and into the garden
The remaining dirt work involves mostly the insulation/watershed umbrella and final contouring of the grade north of the house. The goal of the grading is to force as much drainage as possible northward towards a creek valley instead around the ends of the house where it could continue to overload the rain gardens in front of the house. 
The passive solar system for the house is called Annualized GeoSolar.  For information, click on "Featured Post" in the column to the left.

Sunday, February 4, 2018

Construction - The Last Major Dirt Work (Cont'd) - Final Grading, Loess Lessons and Rain Garden Design

The previous post on The Last Major Dirt Work emphasized the retaining wall construction that needed to be done before the final grading, topsoil replacement and seeding could be done.  This post covers two more sections of the insulation/watershed umbrella as well as the grading, top-soiling and seeding. It also introduces the subject of rain gardens by looking at their generic design. The next post will explain how rain gardens were used specifically for our property.

Reminder: clicking on any photo enlarges it for better viewing.

(Nearly three months have elapsed since my last post.  I forsook the computer for a tool belt full time as we family members rallied to rehab my sister-in-law and her husband's Texas home that was flooded by Harvey. Their homelessness ended the first weekend in January so I am home and blogging again.)

More Insulation/Watershed Umbrella
Umbrella installation behind the garage; notice the plastic
 sheeting, two layers actually, emerging from the bottom of
 the retaining wall to be shingled over the sheeting in the
 umbrella; similar flaps lie behind the wall waiting to be
shingled under the umbrella behind the house when it is
At this juncture, the umbrella around the
house is more than half done.  Part of it lies under the concrete floors of the screened porch and garage and the rest that is done lies in front of the house. The two umbrella sections mentioned here lie behind and in front of the garage as well as in front of the main entrance to the house.  The usual layers of sand, insulation, sheet plastic, recycled carpet and topsoil were used as described in detail in the previous post on the umbrella installation in front of the house.

Final Grading
The dirt work discussed herein lies in
Topsoil over the umbrella behind the garage ready for
 seeding; the last major section of umbrella yet to be
 installed lies beyond the retaining wall, i.e., behind the
front of the house and east of the garage.  Since we 
own (and live in) the old farmhouse east of the building site, we have the option of carrying the grading onto that property as far as necessary to flatten contours and funnel surface runoff from both properties into a series of rain gardens.  After moving superfluous soil out of the way of smooth contours, berms and rain gardens, and, after laying down +/- 3" inches of topsoil, we were able to plant grass seed in time for germination before cold weather -- barely.  Grass is an interim crop to hold the soil then, as discussed in a
Friend Kerry helps with the umbrella installation in 
conjunction with setting forms for the concrete
 sidewalk and garage apron
previous post on native plants, the grass will be replaced 
almost entirely by environmentally-friendly (no mowing, no fertilizing, no watering) plants that are native to the Midwest.

Harnessing the Impetuous Loess
Building in the Mississippi River bluffs has been a mixed blessing.  The rough terrain provided the south facing hill that we needed for a passive solar build and, quite unexpectedly, within a community rather than out in the country, but the wind-blown loess soil that covers the bluffs has taken us to school.  We have learned that it is quite stable when left undisturbed and covered with vegetation because the grains of soil are elliptical in shape and, when lined up like soldiers by wind action long ago, were locked together. There are places nearby, for instance, where cuts through loess for interstate highways have shown very little erosion.  When disturbed, however, the loess (aka silt) is unstable in both wind and heavy rains --  it blows like talcum powder and erodes like lumps of sugar.  Controlling erosion during construction has been a major challenge; the retention pond that catches runoff before it can leave the property has been re-dug and cleaned out at least once a year since construction began. Parenthetically, plants thrive in the loess despite its being almost plaster-of-Paris-hard when dried out. 
View of building site before footings
 were poured; the denuded areas are
where dirt was excavated; the
somewhat denuded areas north of
the house are topsoil and subsoil
storage piles; notice the solar
collector in front of the house; the

denuded area surrounded by 
greenery in the middle bottom of the
 photo is the retention pond catching
 runoff during construction that was
 converted into one of the rain

Unavoidable Site Disturbance
Unfortunately, it is impossible to build an earth contact, Annualized GeoSolar* conditioned house without moving a lot of dirt, which violates one of the major tenants of green building -- that of minimizing building site disturbance.  Not only was it necessary to excavate into the side of the hill for the house itself, we also had to install the insulation/watershed umbrella that extends sub-grade +/-20 from the floor of the house in all directions.  We had to use soil-disturbing measures to amend the water table so that it did not compromise the AGS system and we had to install the large solar collector and conduits for the system.  And the final grading has been extensive in an attempt to direct runoff into strategically located rain gardens so that it does not soak into the soil where it would compromise the AGS system or stay on the surface and carry silt and contaminants onto the street or into the neighbors' lake.  The solution to both problems is a series of strategically-place berms and five rain gardens.

How Do Rain Gardens Work?
In a nutshell, the purpose of rain gardens is to corral surface runoff long enough for it to soak into the soil and, if flowing, flow underground where it is purified instead of on the surface where it carries pollutants. The plants in and around a garden are tiered such that those liking wet feet, such as sedges, occupy the bottom, those who can tolerate a wet soil, but do not like wet feet, rim the slope of the garden and those liking it a bit drier form the periphery along the top edge of the garden and often spilling over onto the surrounding grade.  In our case, the rain garden plants will be native rather than cultivars and non-natives.  For an authoritative resource on rain gardens, link to the Missouri Botanical Rainscaping Guide.

Percolation Test
In order to avoid stagnation that would kill plants and incubate mosquitoes, a percolation test is done before digging to be sure that the retained
Downloaded photo of a rain garden containing
what looks to be mostly sedges in its deepest
section; notice the graveled overflow to the left
 water will soak away fast enough.  At least two post-hole-size holes are dug 24" deep a few feet apart in the lowest area of the proposed rain garden, usually just inside of the proposed berm that will serve as the dam.  The fresh holes are filled with water and the drop in water level -- the "perc rate" -- is timed over the span of a few hours.  Then the process is repeated the next day.  The first test simulates rainfall on relatively dry soil; the second simulates rainfall on soil that is already saturated by prior rain.  A perc rate of 1/2" per hour for the second test is considered minimum for a functional rain garden. The rate for the first test is typically faster.

The perc rate per se is not the entire story however.  A good share of the water collecting in a rain garden is utilized by the plants before it can soak into the deeper soil making the water disappear much faster than if the water-loving plants were not present.

The exact manner in which rain gardens are incorporated into our landscaping is the subject of the next post.
*Click on "Featured Post" in the column at the left for information on AGS

Wednesday, October 18, 2017

Construction - The Last Major Dirt Work - Two Retaining Walls and More

It is late September in the lower Midwest and the first frost could be a month away. Even with global warming (which seems to have pushed back the deadline for planting grass and expecting it to germinate before cold weather), we might have at the most three weeks to finish the dirt work and plant seed.  Consequently, construction is now on hold until the dirt work is done.

"Dirt work" means several things:  constructing the
retaining walls that are essential for earth sheltering, finessing the exposed mouths of the drainage tiles, final contouring of the subsoil and redistributing the top soil that was removed for construction. Secondarily, a series of rain gardens need to be installed for controlling the run-off originating from our present residence next door and continuing down through, and augmented by, the hillside configuration of the new building site -- a total run of several hundred feet.  And some remaining sections of the insulation/watershed umbrella has to be installed in the process.

As for retaining walls, the plans call for two major "umbrella walls" plus two of smaller size that did not involve the umbrella.

Reminder:  To enlarge any photo for clearer detail, click on it.

The Umbrella Behind and Under the East Retaining Wall (one of two "umbrella walls")
As explained in a triad of prior posts, the umbrella comprises layers of sand, foam insulation board and 6 mil plastic sheeting plus a couple of layers of recycled
North elevation; notice transitions from 12' center wall and
8' walls on either ends; the dark line is the architect's take
on the final grade, including two retaining walls
carpeting.  (The design of the umbrella surrounding the house is based mostly on that described by Hiat in his must-read self-published treatise, Passive Annual Heat Storage.)  At the walls, the umbrella drops about 4' from the height directly behind the house to a lower level NE and NW from the house, which means that the umbrella runs behind and under the walls. 
The section of umbrella associated with the walls must be built first and the walls backfilled before the rest of the umbrella can be installed. 

Most of the north earth contact wall of the house is 12' high then it transitions abruptly to eight foot walls near each end (see drawing).  The most
Excavation complete, metal posts in
place and a sand bed for the
horizontal insulaiton
efficient way to 
maximize earth contact for the house while dealing with the transitions is with retaining walls.  Both the east and west umbrella walls have to be about 5' high next to the house and, by running them diagonally east and west, can be as low as a couple of feet 20' out from the house.  The excavations for the two walls were about half track loader work and about half hand work and took me the biggest part of two days. Then it was time to install the vertical and horizontal insulation and plastic sheeting that allows the umbrella to drop beneath the walls.

The horizontal insulation under the walls is merely a matter of making sure the excavation drains well, is smooth and
First layer of plastic lies under the horizontal insulation
and behind the vertical vertical insulation
covered with enough sand that the insulation can be nestled into it to handle better the load of the wall.  The vertical insulation is another matter.  It must be supported so that it is not damaged by laying

rocks against it on one side and backfilling against it on the other.  So I wired the insulation to metal fence posts that were shortened as necessary to stay below the
Horizontal insulation over two additional layers of
6 mil plastic sheeting
final height of the insulation.

I proceeded by setting two posts, one near the house and the other about 20' from the house. Their height matched the linear slope that the insulation will take from a high end next to the house to a low end next to the French (the French drain that will drain water from the top of the umbrella -- grist for another post). Intervening posts were driven along a mason line stretched between the first two posts (east wall) or, better yet,the edge of the horizontal insulation (west wall). I stood the vertical insulation on the horizontal insulation and stabilized it with clamps and props while I snapped a chalk line
Vertical insulation wired to posts

designating the sloping height of the insulation then used a wood-cutting blade in a reciprocating saw to cut the insulation to height.  I then set all of the insulation aside while I covered the soil on which the horizontal insulation will rest with a layer of sand and screeded it smooth for a solid foundation for the insulation and the rock wall above it.

A 6 mil plastic sheet was laid down over the sand such that there was 5- 6' of excess running in both directions for later shingling with the plastic sheeting of the umbrella above and below the wall.

After the horizontal insulation was returned to position and the vertical insulation was wired
to the posts and secured with more wire as needed between the posts, two equal-sized pieces of plastic sheeting were then draped over
Two more layers of 6 mil plastic sheeting covering the
the top of the vertical insulation to be spread out in two directions: (a) down the front side of the vertical insulation and over the top of the horizontal insulation plus plenty of excess running horizontally to be shingled over the plastic sheeting of the umbrella below the wall and (b) an equal amount of excess running horizontally off of the top of the vertical insulation eventually to be shingled under the plastic of the umbrella above the wall.

The horizontal insulation under the wall loosely follows the plan for the rest of the umbrella behind the house, i.e., 4" thick for the first 8' out from the house, 3" thick for another 8', then 2" thick for the final 4' -- 20' in all.  The vertical insulation is 4" thick throughout with extruded polystyrene on the stone side and the weaker (but cheaper)
Generous layer of sand over the plastic to protect it from
injudicious stone placement; boards protect loose edges
 of the plastic from foot traffic during stone placement
expanded polystyrene on the backfill side.

Protecting the Vertical Insulation
The first retaining wall was erected a year and a half ago with the help of weekend volunteers using salvaged stones from a 19th century barn foundation. Similarly, we did the new retaining walls with weekend volunteers using more of the same stones . I either piled sand next to the wall or parked the track loader with sand nearby so that a layer of sand could be used between stones vertically and horizontally to help situate and stabilize them, considering their discordant sizes and shapes.  

Using sand as a filler meant that the space between the vertical insulation (covered with plastic sheeting) and the stones was also filled with sand.  But, without backfilling, the weight of the sand against the insulation or any pressure from the stones would distort the insulation -- a serious problem that partially collapsed the insulation and compromised the R-value of the wall built previously. Consequently, I fastened with 12" spikes a 2 x 6 over the horizontal insulation such that it abutted and stabilized the bottom of the vertical insulation.  I then was able to backfill the vertical insulation with sand, a shovel-full at a time, to a depth of a foot or so. The 2 x 6 was then removed and the spikes were reinserted to stabilize the insulation during wall-building.  The rest of the backfilling was done gradually in tantum with building the wall.

Building the Wall
I feel blessed to have had enough weekend volunteers, not only to build the wall we have
Stone wall well under way
been discussing here, but to do all of the other "stone work" that had to be done before I could finish the dirt work on on the critical east and south sides of the house.

Whereas the first retaining wall went up too fast for a high-quality outcome, the new wall took all of a half-day to build and with a satisfying result.  I had already stached nearby what I thought were enough stones and sand to build the wall. My role then was to give general instructions as to its size, height and the need to be careful not to damage the plastic sheeting or the stuccoed house wall.  The volunteers were free to place the stones as they saw fit and they did a great job.  Delegating paid off because I had under-estimated the number of stones and sand necessary and was kept busy with the track loader ferrying in more materials.

The Second Day
The second day was as much about hiding
Ugly French drain mouths
drain tiles as about building retaining walls.  The exposed mouths of the French drains and the north concrete wall foundation drains were so ugly as to beg camouflaging with something that would eventually enhance the native gardening that will surround them.  I had pinned down weed barrier fabric in the areas to be covered with stone and asked the crew to be
Stones added for aesthetics and to control erosion;
notice original retaining wall in the background 
creative in installing the stones.  Again the result exceeded expectations.

Then on to a couple of retaining walls. The first was next to the garage doors, which, by now, was hardly a challenge for the volunteers.  The second involved adding height to the first retaining wall at its house end.  Again, all I did was bring in the stones with the loader -- the crew did the rest un-micro-managed.

With the stones in place, I needed to take advantage of the relative dry fall to finish the
Raising the original wall next to the house
dirt work and plant seed -- the subject of the next post.

The Second Umbrella Wall
The retaining wall detailed above is the east-most umbrella wall.  The west-most umbrella wall was completed only to point of receiving stones.  With the weather window for the rest of the dirt work closing fast, the umbrella behind and to the west of the house became lower priority.  The wall will have to be finished in the ensuring months in conjunction with installation of the rest of the umbrella.

Sunday, September 17, 2017

Construction - The "Tall" Roof Revisited

The north-facing roof over the taller part of the house was begun a few months ago and covered with a 6 mil plastic until the time was ripe to complete it.  A prior post details its construction -- a cathedral roof using trusses -- and a recent post talks about temporizing it with sheet plastic.  Still another post discusses air and moisture barriers and uses the tall roof as a cathedral roof example.  

The design of the two main roofs, south-facing and north-facing, calls for cathedral roofs with "mini-attics".  The rationale for a mini-attic is detailed in the recent post, ventilated cathedral roof, showing construction of the south-facing roof from the truss phase to the temporization phase.  Now it is time to build out the mini-attic for the north roof and temporize the roof in a more substantial manner than with 6 mil plastic. 

Completion of the Mini-Attic
The mini-attic comprises two layers of sheathing separated by 2x4s on
The first layer of plywood sheathing was air-sealed with
flashing tape previously; the 2 x 4s on edge support the
second layer of OSB sheathing and provide for the over-
 hanging eaves and ventilation bays  
edge so as to create a ventilated space between the insulation and the roof cladding.  As rationalized in the post on the south-facing roof, I used plywood for the first layer of sheathing because it "breathes" better than OSB for drying purposes but used OSB for the second layer because it is cheaper and breathing was not an issue. We
sealed the cracks in the plywood layer, as well as its junction with the wall sheathing, with flashing tape to provide the definitive air and moisture barrier for the roof assembly (as opposed to relying solely on a barrier on the interior plane of the ceiling). The two-by-fours were fastened to the underlying trusses with construction screws to create the 3 1/2" tall ventilation bays, to provide the overhanging eaves and to support the OSB sheathing.

We installed a 1-by ridge board at the top of the roof in anticipation of the overhang for the second story clerestory windows extending southward from the ridge. Ventilation of the mini-attic will be by natural convection from eave vents at the lower end and an uninterrupted ridge vent at the upper end.  Accordingly, we left the sheathing 2" short of the ridge to provide an airway for the ridge vent.

Vapor Control
As discussed in the links in the first paragraph, the slope of the roof was less than originally planned by about 6" in 12', i.e., 12/2 instead of 12/2.5.  The slope is the minimum allowed by code for asphalt shingles and for standing seam metal roofing. The unfortunate decrease in slope was was due to building the first interior wall for 2 x 12 rafters instead of the 16" trusses that I eventually used and getting it a little too tall even then for 2 x 12s.  The low slope means that our moisture control strategy must use a moisture impermeable material that favors wet prevention over drying under the roof cladding (for details see the post on vapor barriers).  This strategy works because the mini-attic drys the OSB sheathing from below. Without it, any moisture that frustrated the drywall side of the roof assembly or found its way under the roof cladding would wet the sheathing and, over time, cause wood rot, mold growth and diminished R-value of the insulation.

Provisional Roof Cladding
For the other low-slope roofs, I plan to use 30# felt as the wet prevention material under standing seam metal roofing.  However, for the tall roof that is far above street level and faces away from the street, I am compromising in favor of asphalt shingles for the final roof in order to save money.  And it would be convenient to postpone installation of the shingles indefinitely for two reasons: to defer their cost and, since our dry months are August - October, to free up time to get on with the dirt work that needs to be done before it is too late for turf seeding.

A few years ago in anticipation for building the house, I bought on Craigslist 12 rolls of roofing.  It was cheap and I thought that it might somehow come in handy during construction.  As it turned out, I used it on the tall roof to double as the wet prevention layer instead of 30# felt and to serve as a provisional roof until such time shingles, or even a metal panels, are installed over it.

Ice and Water Shield
The code calls for a dedicated ice and water shield extending
Two courses of ice and water shield adhere directly to
the sheathing covering the drip edge at the eave
up-roof from the eave 3' beyond the plane of the interior wall.  For us with our wide overhang and thick wall, that spec meant two 3' courses. While shopping for the shield at the local home center, I lucked onto a roofing contractor who was buying it for himself. He recommended the cheaper of the two brands in stock as being the easiest to install and, indeed, it was a snap. (For installation details, go to the Tarco website.)  It is applied sticky side down directly to the sheathing.

Drip Edge
By covering the junction of the sheathing and the facia, the drip edge protects the junction from water penetration.
We installed abbreviated pieces of 15# felt at the rakes
so the drip edge could be installed over them; the felt
overlaps the ice and water shield down-roof; the rest of
the felt was installed just ahead of the roll roofing in order
to prevent wind damage to it before it could be covered
 As dictated by code, we fastened it under the ice and water shield at the eave and over the 15# felt under-layment at the rakes using roofing nails. I used aluminum drip edge but, if the final roof were likely to be steel, I would have used steel drip edge in order to minimize the possibility of galvanic deterioration of the aluminum in the presence of moisture.  In fact, I probably should have used aluminum nails instead of steel roofing nails for the same reason.

Imperfections in the Craigslist roofing were caused by
 improper storage but appearance was not critical since
 a definitive roof will cover the roll roofing eventually

Installation of the Roll Roofing
There are two common ways to install roll roofing. The simplest method is simply to fasten the overlap between courses with roofing nails but moisture penetration can occur around the exposed nails unless they are covered with roofing cement and horizontal rain might undermine the overlap and reach the sheathing. The better approach is to nail the top edge only and seal the overlap between courses with roofing cement such that the nails remain hidden and the lap is sealed.  Although it was much more time-consuming and costly (by the amount of four pails of cement), I opted for the leak-proof second method.
Only a couple of courses to go then its clean-up and 
on to a month's worth of dirt work

Due to the fact that the Craigslist rolls had been improperly stored lying down before I purchased them, they were imperfect and would not have passed muster if appearance had been important.  But, since the roll roofing will be covered with shingles eventually, protection was all that counted. Eighteen rolls total were required, ten of which were the Craigslist rolls.  We found the online installation instructions for roll roofing to be helpful but we deviated from them to the extent of using 15# felt as under-layment instead of fastening the roofing directly to the sheathing.

My thanks to friend, "Pat the Plumber", who, though being a journeyman plumber, helped with the roof.  And to Keith, my step-son, who pitched in as well.  Thanks, guys.