Tweaking ductless mini-split heating and cooling performance in conjunction with mechanical ventilation

Tweaking ductless mini-split heating and cooling performance in conjunction with mechanical ventilation

I found this past winter to be particularly cold and overcast, providing a good number of days consistently below the heat-load design temperatures of our region.  This had an adverse impact on our ability to benefit from free solar heat gains in our interior as did the fact that the house had yet to be fully occupied (we didn't benefit from heat gains of occupants, household electronics, or human activity).  But these circumstances did allow us to look at the performance of our building under, what I felt were, more extreme heat-loading conditions.

Using mechanical ventilation set at a level to replace our indoor air volume, of approximately 60,000cf, eight times a day--with a nominal 3-4 degree fahrenheit deficit--meant that we needed active heating from our ductless mini-splits during these times.

Unlike a typical passive house, which I believe tends to be smaller--in the realm of 1300 square feet TFA--our home is greater than five times that size, in excess of 6600SF.

The house has four bedrooms: a master bedroom and bathroom on the 2nd floor; two additional bedrooms and a bathroom on the 2nd floor, and an additional master bedroom and bathroom on the 1st floor (for age-in-place consideration).

The house is generally symmetrical, sectioned into thirds with the center section consisting of a two story foyer and great room separated by a bridge connecting the two sides of the home.

The first floor also has a study, dining room, kitchen, and dinette as well as an extended section consisting of a mud room and a laundry room that connects to the attached three car garage.

Our walkout basement is quite large in its own right, clocking in at an excess of 2000 square feet.  At the moment it is about 70% finished.

With respect to our Zehnder ERV mechanical ventilation system, all of their comfotubes get distributed from the two ERVs location in the basement's mechanical room.  This means that a good portion of all of our ventilation duct work is subject to the temperature conditions of our basement.

Each of the four bedrooms have a supply register located in the ceiling in the furthest corner of each room as do the study, the dining room, and the family room.  The return ports are located in each of the three full bathrooms, the kitchen, and the laundry room.  The basement consists of three returns and three supplies as well.

Transfer grills are in place between the two master bedrooms and their adjacent bathrooms; the standard bathroom on the 2nd floor which is shared between the two additional bedrooms; and each of the two additional bedrooms leading to our connecting bridge.

Having analyzed the floor plans, Barry Stephens of Zehnder and I decided to place our backup heating and cooling ductless mini-splits at a central location and rely on the MHVRs to assist in distributing the conditioned air around the home.  Barry suggested that this approach was being used abroad with a measure of success.

Conservatively estimated, we figured the house would require approximately 19,000 Btu/hr of backup heating at our heat load design temperatures.  With the COPs of the Fujitsu RLS2 minisplits approaching 2.0 at those outside temperatures, we figured we would need two of them.

We selected the Fujitsu 9RLS2, their smallest and most efficient unit, because our air conditioning needs were expected to be either non-existent or minimal--since we are using earth air tubes--and the heating capacities of the 9RLS2, 12RLS2, and 15RLS2 are essentially the same.

I decided to install both units below the bridge area facing the family room towards the back North side of house.  We ultimately determined that given the sheer size of and layout of the home, this setup was not optimal and required tweaking.  I was finding the basement consistently ran 10 degrees cooler and the bedrooms, study, mud and laundry rooms were typically six degrees cooler (when without the benefit of solar heat-gain) than the set temperature of the minisplits located in the family room.

To elevate the temperatures of these rooms to a desirable level, I had to set the thermostat of the mini splits to 74-76 degrees.  Clearly, this one not a good situation.  Not only did we find the family room to be too warm, we spiked our electric consumption because we had to "over drive" the mini splits.  We had hoped the the main ceiling fan would have been sufficient to assist in getting the air moved around and balancing this out sufficiently, but this was not the case.

I then started documenting the temperature readings of the "cold" rooms and the air temperatures at each of the supply ports in those rooms.  This is what I found.  On average the air temperatures at the supply ports were consistently six or more degrees lower than current room temperatures.  In other words, I was "cooling" the rooms.

My first concern was that my HVAC installer should have not placed the supply ports at the ceiling in the corner of each bedroom on the 2nd floor.  I believed that even with 20 plus inches of blown cellulose insulation in the attic, that I was losing heat to it.  Fortunately, as I later found out, this wasn't really the case.  However, in the future, I will insist that supply ports always be placed high on the farthest wall of each room and not the ceiling to eliminate the potential of heat loss or gain by keeping them entirely within the confines of the conditioned envelope.

As I saw it, there were two challenges.  One was this large temperature delta with the supplies and the other was the "cold" basement.

The next thing I tried was closing up the three returns in the basement with the hopes of causing the ventilation system to effectively provide more net warm air to the basement.  This was a wrong headed approach as it took the ERVs out of balance which reduced the efficiencies of the ERVs' heat exchange. 

I shared my experiences with several people including Jason Morosko (of UltimateAir) and Barry as well as the installer of our minis, in hopes of getting some ideas. 

Jason believed it was better that I evacuate the cold air from the basement through the ERV's return ports instead of closing them off and that I should place a small low-voltage fan (like a PC cooling fan) somewhere near the mini splits to direct warmer air into the basement in that manner.

Barry believed that the configuration we used may have been a bit too ambitious (in its simplicity) given the sheer size and layout complexity of the house and was going considering other options.

I searched online and found an article about other passive house builders that had experienced similar disparities in room temperatures who experimented by substituting the supply ports in the "cold" rooms with returns and having some measure of success.  However, Barry suggested that Zehnder had not found this approach to be an optimal solution either and he discouraged me from making those changes.

I then began to think that the two issues I was experiencing were interconnected.  I thought of my earth air tube system and the temperature moderating that occurs with the air flow through them.  I then suspected that I was partially heating the basement directly by the airflow of supply and return ducts the ran throughout the basement ceiling.  And since we opted to use a Stiebel-Eltron heat-pump hot water heater located in our mechanical room, instead of a solar hot water system, I believed our basement temperature disparity issues were only going to get worse because of the cooling effect of the hot water heater.

With this in mind, I had one mini-split moved from the family room and placed into operation in the basement.  I am pleased to report, while still not perfect, this revised placement has made a substantial improvement to the balancing out of temperatures throughout the home.

Not only is the basement extremely comfortable, the temperature differences between the supplies and the rooms are much much smaller.  In fact, I am finding that the temperatures at the supply registers are warmer than what they are at the ERV.  It is clear to me now that the basement temperatures have a direct impact on moderating the temperatures of the rooms throughout the entire house.

I am finding when the doors are open on all of the rooms, the temperatures differences from room to room are minimal.  They do get more pronounced when the doors are closed, though.  When I turn the ERVs off whether the mini splits are on or off, the room temperatures appear to be even more similar suggesting that more tweaking of our ventilation system is in the offing.

I am also considering elevating the temperatures in the basement a little bit more to see if we could pick up a degree or two in the supply temperatures of the rooms on the 1st and 2nd floor.  It's going to take some experimentation and empirical analysis, either way.

Barry will be making another visit to our home to tweak the balancing of the ERVs further.  My goal is to have the rooms experience minimal temperature variation even with their doors closed.  We'll see what comes of Barry's visit, but I am considering the idea of placing very-low powered small fans within several transfer registers in the hopes of distributing the indoor temperature more evenly.

I am certainly open to new ideas from anyone else who has shared a similar experience.

One final observation I'd like to make:  The operational costs of the minisplits have been lower in this configuration than when they were originally.  Even when I was running only one of the two in the family room.  I suspect this is because each one now works less hard to maintain their respective room temperatures and because the mini splits have variable refrigerant flow (VRF) heat pumps.

This experience has taught me that strategic placement of your mini splits within the context of your mechanical ventilation configuration is as important as properly sizing your heating and cooling system.