Improving Passive House (Passivhaus) Temperature Performance

Having taken the steps of cleaning both mini splits and the ERVs, identifying appliances of high phantom loads, temperature variances at the intake side of both mini-splits, one mystery remained to be solved: how to figure out a way to extract the greatest efficiency of the mini-splits as we could as we believed our placement was compromising their performance.

I could not understand how a house could maintain an interior temperature of 62F (at its lowest point) when we experienced an extended power outage that lasted 36 hours which spanned two evenings of temperatures in the single digits.  Nor could I understand how the home's interior temperatures could rise to as high as 75F in certain rooms when sunny outside regardless of the outside temperature and yet we struggled to raise the temperatures about 72F when the minis were operating.  .

Since we already determined a connection between our Stiebel Eltron heat-pump water heater and the mini-split located in the basement, I decided to install some foam material--that we had used for insulating our basement slab-- between the floor joists located in the ceiling directly above the mini-split.  The hope was that by preventing the cooler air in the mechanical room from supplying the intake of the mini-split, its temperature sensor would be more accurate than before and hence the heating performance would be more accurate and eliminate the need for installing a remote thermostat.

I am pleased to report that this effort was successful.  The temperature setting of the basement mini-split is now more inline with the actual temperature of the entire basement.  Furthermore, no longer does the mini-split kick on whenever the Accelera's heat-pump engage.

Having corrected the basement, the next area that we needed to address was the family room mini-split.  This one was especially important because this was the unit that provided heating for the 1st and 2nd floors.  The solution turned out to be very simple.  What we did was create essentially a vent that connected the closet area of the 1st floor master bedroom--an area located directly behind the curved staircase leading up from the foyer to the 2nd floor bridge.  The temperature of the air there is more along the lines of the temperature of the rest of the house.

I cut a small rectangular area into the drywall of about one inch in height and as long just above the intake of the family room mini-split.  I was thrilled to find the the supply temperatures of the air that were now being supplied dropped to within 2F (down from 12F) and set temps are much closer now to actual rooms temps.

I also determined that the apparent COP increased from a low of about 1.85 to approximately 2.45 at a given outside and interior temperature.  The reason for this was the mini-split was no longer being asked to raise the temperature from 82F but a more realistic 68F.  Looking at the performance chart of the Fujitsu RLS2s it is clear that heat production, in terms of BTUs, noticeably drops as the interior temperatures rise and hence heating efficiency as well.

At a low last night (Feb 16th, 2014) of about 18F (very nearly at the temperatures used for determining heat-loads) throughout the night and a high of 24F with overcast conditions, the day earlier, the observed improvements have been dramatic. No longer are the temps falling to 66-67 in the family room through the night regardless of the mini-split temperature settings.  Family room set temperature was 70F.

Following are the temperature profiles of the rooms taken at 7.00-7.30 with an outside temperature of 18F at the time of readings.

Family room set temp: 70F
Average nightly usage: 1400 watts

Master bathrom (1st floor): 69.2/48%
Master bedroom (1st floor):  69.4/48
Family Room: 71.7/45%
Kitchen: 70.8/45%
Laundry: 69.4/47%
Bird Room: 69.0/47%
Dining Room: 70.5/48%
Study: 70.1/46%
Bridge: 71.9/44%
Master bedroom (2nd floor): 70.8/47%
Master bathroom (2nd floor, south facing): 68.3/49%
Back bedroom (2nd floor, north facing): 68.9/47%
Front bedroom (2nd floor, south facing): 69.6/48%
Shared bathroom (door closed): 68.9/50%

Basement mini set temp: 66F
Nightly usage:  ~700  watts continuously

Basement: 68.0/39%
Mechanical Room: 67.8/42%
Pet room 67.1/41%
Bar Room: 67.6/41%
Movie Theatre Room: 68.0/42%

Total average about 2.2kw between both units.


Family Room:

According to Fujitsu, electric consumption of 2000 watts results in approximately 17,000 BTU/hr at a COP of 2.49 (improved over 1.8 historically) at the temperature conditions experienced.

Therefore, at about 1400-1500 watts this equates to about 12,300 BTU/hr or 12,000 BTUs--its nominal heat production rating.


Electric consumption of  700 watts equates to about 6,000 BTU/hr.

Consequently at about 18,300 BTU/hr production to maintain or slightly raise temperatures throughout night suggests a heat-load of about 2.8 (per square foot).  Our PHPP modeling predicted a similar number, which goes to show how accurate and thorough the PHPP modeling software is.

These results also suggest that we actually did have the heating capacity to handle the entire conditioned interior space (TFA: 6600SF) with only one mini-split (even with the heat-pump DHWH robbing some interior heat from us).  All we need to do is take install a blown ducting connection between the first floor and the basement (which could be reversed during the summer months), which is our next project to undertake.

The observed results also suggest that room-to-room temperature variation has noticeably improved and we do not appear to need the supplemental heating coils installed in the ERV supply ducts as we had believed we may.

I am kicking myself for having not discovered our design "flaw" earlier as we have now been through two summer and winter seasons.  Perhaps had it not been for the fact that this winter has been the coldest one in thirty years and one that has given us plenty of days and evenings below the design temperatures of our climate, we may not have caught this.  Without question, eMonitor (now called Sitesage) has played an integral part in assisting us in our resolution.