Most heating systems are designed to supply much more heat than our super-insulated house would need, so a standard system was out of the question.  We considered three options: radiant heat, a central gas burning stove that relies on the ventilation system for heat distribution and some kind of hydronic fan-coil system, most likely combined with the ventilation system.

While we liked the idea of the warm floors of radiant heat, we ruled it out for a number of reasons.  First, it is an expensive system (we didn't get actual numbers, we just went by what people who had installed it told us). Second,  because our house is well insulated, the temperature of the floor can't be all that much above room temperature without overheating the house, because without heat loss, a warm floor soon creates warm air and so on.  In fact the side effect of lots of insulation is that the house has warmer surface temperatures than normal homes, and so you feel warmer at the same air temperature in a well insulated home (see the energy section in the tutorial).  Again, we didn't do actual calculations to see what floor temperature we could have, and we probably should have.  Other less relevant issues combined to cause us to look at other options: we had a less than optimal setup for radiant floor since we had only a framed floor instead of a concrete slab; warm radiant floors combined with passive solar will make the floors too warm (although the limited mass of a wood floor won't hold much heat anyhow); finally, we would have to install all the ducts for a ventilation system anyhow, which really just leads back to the expense issue.  Admittedly if we really wanted radiant, we probably would have found a way to make it work.

We looked at the option of using a gas burning stove (essentially wood stoves designed to burn gas) for quite a while, because, like many people, we liked the idea of being able to "sit around the fire".  The biggest problem again was that most of them put out too much heat, and also that we were fairly skeptical that our ventilation system would distribute the heat well.  This really is a big issue, since air holds very little heat you need a pretty large volume of air to deliver much heat, or the air must be pretty hot.  Our calculations showed that we would need up to 800 CFM of 140 degree F air on the coldest day to keep the entire house warm, and since ventilation systems only deliver 200 CFM, this just wasn't going to work.

In a meeting with the energy consultants (Ecotope), we all agreed that the fan-coil solution would make the most sense, both from a functional perspective and from a cost perspective.  In this system, we would use ducts for both the ventilation system and to distribute heat.  In heat mode, a pump circulates hot water from the hot water tank through a fan coil unit which consists of a coil of pipe with a big surface area for transferring heat, and a fan to move air across the coil and out to the rooms.  In ventilation mode, a small fan circulates air through the house, exchanging some of the recirculating air with fresh air from outside (and using heat from the exhaust air to heat the incomming air).  

The biggest downside to this system is that there isn't much of an option for zoning, since there is only one unit there is only one thermostat.  Each air outlet will have a damper on it, allowing zoning on a more permanent level: for example we can keep the bedrooms cooler the the living area, but just allowing less air in (ventilation requirements are much less than heating, so we'd still get plenty of fresh air). We did consider wall mounted fan coil units (such as the Turbonics) scattered around the house instead of a single unit, but since the fans tend to be a little noisy, Kim vetoed that option.

Because we didn't want cooking odors from one unit going into the other, the ADU has its own ventilation system, and so it needed it own heat source also. The current solution is to use two Turbonics wall mounted fan coil units and just go with the code required ventilation system, which is a slot in the wall (ie a permanent hole) and putting the bathroom fan on a timer set up to run eight hours a day.  While this is cost effective, its seems like its a pretty crappy solution: the is no ability to recover heat from the ventilation process, and experience indicates that people often disable their bath fan timers (we're going to use the Panasonic bath fans which are so quiet, you can hardly hear them, so maybe this won't be a problem).  If possible, we want to upgrade the ventilation system in the ADU to a heat recovery ventilation system.

With the decisions of what heat source to use and a distribution method decided, the only issue remaining is the details of distribution: getting heat where it is needed while also making sure that ventilation air is reasonably well distributed through the house.  Duct layout is often ignored in designs, and left to be designed and fitted in as best as possible by the HVAC contractor when framing is complete.  Needless to say, this often results in large holes being put in the framing (which is designed with to some degree to allow for this): while hardly anyone seems to question whether this makes sense, it strikes us as totally less than optimal.  We had Ecotope do a duct layout, so we had a much better head start than most projects, but since they had to work without talking to the framer or the HVAC contractor, it remains to be seen how closely the final layout will be to the designed one.  Although its unlikely that the distribution points will change much, since each outlet is designed to supply the right amount of heat for the size room it is in, it occurred to us that designing for adaptability solves these problems.

The initial design has only one return duct, and the fan-coil and H/W tank is located in the attic.  There are a few reasons for this: we planned on having an active solar collector on the roof to augment the gas-fired H/W heater, and the attic was the only area within heated space that the mechanical equipment could easily fit (in retrospect, maybe we should have started our room layout by blocking out eight or so square feet for a mechanical room-at the time we thought the attic was a fine location).  After reviewing the duct layout with the HVAC contractor, who didn't like the idea of forcing warm air downward (counter to the way it wants to go), and who also strongly recommended we have a second return duct on the main floor, we realized that we really hadn't given enough thought to duct placement.  Figuring out where to located ducts in rooms is one thing, but getting them there through a maze of structural components (more on this in the structural section) is a completely different problem.

One possible solution is to make a small well insulated space in the basement (likely under the stairs), and then run the ducts along the basement ceiling (again boxed in with insulation) to wherever they are necessary.  While this trades easy access to the main floor for easy access the the 2nd floor, it at least makes the warm air go upward.  A second return duct can probably be designed in either way-its just a matter of working around structural components.  If we go this route, we can either keep the H/W heater in the attic, and pump the water down (it gets pumped anyhow), or move the H/W heater to the basement also, but keep the solar pre-heat tank in the attic  In any case its easy to run H/W lines from the basement to the attic, and any heat loss than occurs is within heated space.  The downside is that the solar collector doesn't produce much warm water in the winter, and produces a lot in the summer when heat loss from the plumbing is undesirable (although the amount of heat in any case is relatively small).  Note that plumbing considerations leans toward placing the tank in the basement also.

No matter what the duct layout, the ducting follows a number of rules that aren't normally observed::

• The ducts are all with the heated space
• No wall cavities are used as part of the ductwork-this is to prevent dust from entering the ductwork and also to prevent air losses- we want air only coming and going from where we want it.
• The ducts are sealed with mastic instead of duct tape.  Duct tapes has hundreds of uses, but doesn't seal ducts tight enough.
• No insulation is allowed inside ducts.  Again this is to prevent particulates from entering the air.

Active solar

As part of our campaign to be as close to zero energy as possible, we plan on using a solar collector on the roof to generate as much hot water as possible.  In a cool, cloudy climate like Seattle, only evacuated tube collectors can be expected to generate any hot water in other than the summer months, because flat plate collectors let too much heat escape.  After asking around a bit, and looking on the web, we decided that Thermomax is our best choice.  According to their data, a twenty tube Thermomax system can make seven gallons of hot water in the winter and about twenty five in the summer, while in the spring and fall is makes in the high teens. The downside is cost: a twenty tube system is over $2500 and because the collectors can make very hot water (160-180 degrees), a very high quality storage tank is necessary, which is also expensive.  The cream of the crop tank is about $3000 (it not only takes the higher temperature, but lasts for a very long time), while Rheem make an acceptable tank at about $800.

Given that the house has two units (assumed to house three people), we would need at least a thirty tube system, which we priced out (including mounting & pump) to be at least $4000, and likely much more.  Doing a simple economic analysis, we assume three people would use between 15 & 30 gallons of hot water a day (assuming low flow showers, and energy efficient dishwashers & laundry) and guessed that it would average out to about 22 gallons a day (probably pretty stingy compared to the average). Further, assume that on average we'd have to heat cold water about 70 degrees to make hot water, so 22 gallons requires 12300 BTUs of energy, of about 3.6kwh which at 10 cents/kwh is about 36 cents a day or $130/yr. or a payback of at least 30 years.  But the system makes less than 22 gallons of hot water for much of the year, so you have to reduce the yearly output by probably about 30%, and if probably will cost more than $4000 over a thirty year lifetime, so energy prices would have to rise really drastically (beyond regular inflation: for more on economic analysis see the PV section).

Obviously in sunnier climates the payback is much better, and in addition cheaper flat plate collectors are viable there also, further making active solar a more viable alternative. Unfortunately, this disappointing result is just the first of what turns out to be a trend: the analysis for PV electric and rainwater collection comes out about the same.  What these numbers really show is what an incredibly good deal utilities are: not only are they cheaper than you can do on you own, but they're incredibly reliable and require almost no maintenance on the homeowners part.  

To find out about what we actually did, click here.