Building America Industrialized Housing Partnership

Trip/Testing Report

Location: Seattle Healthy House '97, "The Parks" at Petrovitsky Road and SE 188th St., Renton, WA

Attending:Armin Rudd and Bob Abernethy of FSEC
Nancy Gilbert of ALA Washington

Report Date: 15 April 1998
Review draft on 25 March 1998

Author:Armin Rudd

Distribution List: Rick Finkbeiner, Bill Finkbeiner-Cambridge Homes, Astrid Berg, Nancy Gilbert-ALA Washington, Randy Nicklas-Heat Alaska, Dianne Astry-ALA Minneapolis, George James-USDOE, Subrato Chandra-FSEC, Mark LaLiberte-Shelter Supply

Executive Summary

On 30 January 1998, Armin Rudd and Bob Abernethy of FSEC worked with Nancy Gilbert of ALA Washington at the site of the Seattle Healthy House '97 in Renton, WA to conduct performance evaluations and inspections.

Building air leakage at the Seattle Healthy House '97, at 2.2 ach50, was equal to the lowest of all ALA houses that FSEC has tested so far, which is commendable.

Duct leakage to outdoors, measured at 204 ft3/min and normalized to 7.3% of conditioned floor area, was higher than the average of those tested so far. Approximately 75 to 100 cfm of leakage can typically be attributed to the air handler cabinet itself, which is hard to remedy, and is why we recommend putting air handlers and ducts inside conditioned space whenever possible. Another improvement to duct airtightness would be to caulk all supply and return boots to the wood floor and ceiling drywall, especially at the main return.

Pressure differential mapping was performed with all interior doors closed and the air handler unit fan operating. These measurements showed that using separate return ducts for all rooms was working well to reduce or eliminate inter-zonal pressure differences commonly induced by the central air handler fan and door closure.

A VanEE energy recovery ventilation system was installed in the garage. This system was designed to require continuous operation of the central air handler fan along with the ventilation fan. An energy saving, and functionally equivalent design would be to locate the VanEE fresh air supply in the AHU supply plenum, using separate filtration for the fresh air, and use the FanRecyclerTM control to periodically provide full distribution of ventilation air if the AHU fan had not already been operating for heating and cooling purposes.

Gas combustion appliances for space heating and water heating were appropriately located outside conditioned space in a framed enclosure in the garage, with high/low vents for combustion and dilution air

Access to the air filter at the furnace is very difficult due to limited space. Since it is very important that this filter be frequently inspected and replaced as needed, access should be easy. Also, additional gaskets should be installed to better seal the filter access cover at the furnace.

Observations in the main attic indicated a relatively poor application of the Icynene spray foam insulation. Insulation was very thin in spots, and was entirely missing in some small areas, exposing ceiling drywall, metal supply and return duct boots, and fabricated covers for recessed canister lights and bath exhaust fans, to the vented attic. An intended six inch thick layer of blown rock wool insulation, over the foam, was never installed. An approximately fifteen foot long main return duct, located in the attic, was completely uninsulated. It appears that the insulation contractor is willing to return to finish the job if he can gain access to the house.

Discussion:

From approximately 8 am to 3 pm on 30 January 1998, Armin Rudd and Bob Abernethy of FSEC worked with Nancy Gilbert of ALA Washington at the site of the Seattle Healthy House '97 in Renton, WA. Aric Brown of Oakstar Electric, Inc. was also there working on control wiring for the VanEE ventilation unit, the kitchen downdraft fan, and the central air handler unit (AHU) fan. After working with Aric for some time, we determined that the fan relay on the furnace control board had failed and was the cause of unpredictable results which led to confusion with regard to getting the intended control functions to work. According to Aric, and later confirmed with Randy Nicklas, it was intended that the AHU fan should operate simultaneously with operation of the VanEE ventilation unit. This would provide full distribution of ventilation air supplied through the energy recovery ventilation unit. In addition, it was intended that the AHU fan and the DuroZone outside air damper should operate simultaneously with the kitchen downdraft exhaust fan. The intent was to prevent excessive depressurization of the building when the kitchen exhaust fan was on by allowing outside makeup air to come through the DuroZone damper and be filtered and supplied to the house by the AHU fan. To determine if excessive depressurization would in fact occur when the kitchen exhaust fan operated, we measured the inside to outside pressure differential under that condition. We switched the AHU fan off, then taped off the open dryer vent and ran water to assure that the drain plumbing traps were full. The baseline house pressure with respect to outdoors was -2.0 Pascal, due to stack effect and very light wind. When the kitchen exhaust fan was turned on, the house pressure with respect to outdoors went to -5.0 Pa, resulting in a net depressurization of -3.0 Pa due to the kitchen exhaust fan. With intermittent use of the kitchen exhaust fan, house depressurization to -3.0 Pa will most likely not create a problem with the building envelope or indoor air quality, especially since the vented combustion heating appliances (furnace and domestic hot water heater) are not in the conditioned space.

Following our normal testing protocol, building airtightness and duct airtightness was evaluated by fan pressurization (Minneapolis Blower Door and Ductblaster). Inter-zonal air pressure differentials were mapped with all interior doors closed and the AHU fan operating. Inspections of the house in general, including the fireplace, and of the attic and crawl space were made.

Building Envelope Airtightness
Two multi-point fan de-pressurization tests were performed on the house. For one test the supply registers and return grilles were open, while for the other, they were taped off. Initial and final baseline pressure measurements for each test (pressure differential between the house and outdoors) ranged between -0.8 Pa to -2.3 Pa. This indicated that wind and stack effects had a small effect. Table 1 lists the results of these tests. The house volume was previously calculated from the house plans at 31,559 ft3 by consultant Randy Nicklas. Using that volume and the measured leakage at -50 Pa, the air change rate at -50 Pa (ach50) was 2.2. The effective leakage area (ELA) was 65 in2. One can visualize this ELA, in effect, as the size hole that would be equivalent to the sum of all the leakage areas in the building envelope, at a house pressure differential of 4 Pa with respect to outdoors.

FSEC has previously tested the building air tightness of six ALA affiliate Health/Healthy Houses. Referring to Figure 2, the maximum building air change rate at -50 Pa (ach50) was 9.0 ach50, the average was 5.8 ach50. Building air leakage at the Seattle Healthy House '97, at 2.2 ach50, was equal to the lowest of all those tested so far, which is commendable.

With the house at -50 Pa with respect to outdoors, the crawl space with respect to house pressure was 49.5 Pa. This showed that the floor between the house and crawlspace was sealed exceptionally well. With the house still at -50 Pa, the attic with respect to house pressure was 45 Pa. This indicated that the ceiling between the house and attic could have been sealed a little better. Attic air leakage was found between the central return grille sheet metal box and the drywall, connecting the house and return duct to the attic. Also, the gasket around the attic access hatch did not seal to the access cover very well. All supply and return boots and pans should be sealed to the flooring and ceiling drywall at all supply registers and return grilles. With the house at -50 Pa, the garage to house pressure was 47 Pa, showing that the interface between the garage and the conditioned space was pretty well sealed.

Evidence of previous efforts to better seal the direct vent fireplace were very effective. No air leakage was found at the fireplace.

Table 1 Building envelope airtightness testing results

	Ducts Open	Error	Ducts Taped	Error
CFM50 (ft³/min)	1175	(+/- 0.4%)	1001	(+/- 0.9%)
ACH50 (1/h)	2.2		1.9
C	94.6	(+/-2.9%)	61.5	(+/-6.9%)
n	0.644	(+/- 0.008)	0.713	(+/--.019)
ELA (in²) at 4 Pa	65	(+/- 1.8%)	47	(+/- 4.3%)
EqLA (in²) at 10 Pa	122	(+/- 1.0%)	93	(+/- 2.5%)
SLA	0.00016		0.00012
R2		0.99953		0.997793

Duct Airtightness
Duct airtightness was measured separately as total duct leakage and duct leakage to the outdoors. Duct leakage measurements were taken at a duct pressure of -25 Pa with respect to outdoors, measured at the large main return grille where the Ductblaster fan was attached. Duct leakage to the outdoors was measured by using two fans to keep the house and the ducts at the same pressure differential, with respect to outdoors, to eliminate air flow between the ducts and the house.

All four ports on the VanEE energy recovery ventilator were sealed off during the duct tests. The duct system showed total duct leakage of 330 ft3/min (cfm), and leakage to outdoors of 204 cfm. As a percentage of conditioned floor area, the leakage to outdoors was 7.3%.

Referring to Figure 2, FSEC has previously tested duct leakage to outdoors for seven ALA affiliate Health Houses. The maximum leakage as a percentage of conditioned floor area was 10.7%, the average was 5.7%. Duct leakage at the Seattle Healthy House '97, at 7.3% of conditioned floor area, was higher than the average of those tested so far. Approximately 75 to 100 cfm of leakage can typically be attributed to the air handler cabinet itself, which is hard to remedy, and is why we recommend putting air handlers and ducts inside conditioned space whenever possible. For gas-fired furnaces, only sealed combustion, direct vented units should ever be located inside conditioned space.

Observations of the duct system were as follows:

Metal duct was used throughout. The joints were coated with mastic.
In unconditioned spaces, the metal ducts were wrapped with foil-faced or poly-faced fiberglass insulation, except for the main return plenum in the attic which was sprayed with Icynene foam insulation. It was noted in the crawl space that the insulation wrap was pulled away from the main supply duct as it came through the floor; this was put back in place as best as possible.
The main return duct, located in the attic, that connects the main return plenum to the top of the AHU unit was completely uninsulated. A photograph of this is shown on an attached page. This duct should be insulated to R-6.
Many return and supply duct boots in the attic were not fully insulated, leaving bare metal duct exposed to the attic. Photographs of this is shown on an attached page. Make sure that all return and supply duct boots are fully insulated in the attic.
Air leakage was found between the supply and return boots and the ceiling drywall. Use low VOC latex duct mastic (such as RCD #6) to seal the boots and pans to the wood flooring and ceiling drywall at all supply registers and return grilles.
The DuroZone motorized outside air damper was evaluated for bypass and was found to be exceptionally airtight. There was a 99 Pa pressure drop across the damper out of 100 Pa on the downstream side. However, referring to the attached photographs, the outside air duct was terminated with an interior-type supply register on a side wall at the front entry. That supply register will likely rust and be unsightly in a short time.
It is possible that the hole that was cut in the bottom of the main return plenum to conduct the first duct leakage test in June 1997 was not sealed before the ceiling drywall was installed. If the builder thinks this is likely, it would be possible to make a repair by going through the side of the duct to seal the hole in the bottom, then re-seal the side of the duct.

Description	Floor Area (ft²)	ach50 (1/h)	Duct leak to Outside @-25 Pa (cfm)	Duct leak as % of floor area (%)
1995 Orlando	3750	5.6	88	2.3
1996 New Orleans	3900	4.8	188	4.8
1996 Jacksonville	2370	5.4	71	3.0
1996 Birmingham	3160	6.9	288	9.1
1996 Huntsville	2880	6.7	278	9.7
1997 Orlando	3520	2.2	0	0.0
1997 Dallas	4500	9.0	482	10.7
avg:	3440	5.8	199	5.7

1997 Seattle	2791	2.2	204	7.3

Pressure Differential Mapping
Pressure differential mapping was performed with all interior doors closed, and the central air handler unit fan operating, and the ventilation fan off. In typical residential construction, closed rooms will go to a positive pressure with respect to the central area, and the central area will go to a negative pressure with respect to outdoors, due to restricted return air flow from closed rooms. Inter-zonal differential pressures should not exceed 3 Pa. If they do, uncontrolled air exchange to outdoors will increase, with a possible negative effect on indoor air quality, and energy use for space conditioning will increase.

Pressure mapping for the Seattle Healthy House '97, with the central AHU fan on and all interior doors closed and the ventilation fan off, showed +3.4 Pa between the central area and outdoors (as an aside, this went to 0 Pa when the kitchen exhaust fan was turned on). This showed that the most significant duct leakage was on the return side which was tending to pressurize the building. With the AHU fan on, pressure differentials between the house and closed rooms ranged from a high of +4.4 for the Office to less than +0.9 for all the other rooms. This shows that using separate return ducts for all rooms was working well to reduce or eliminate inter-zonal pressure differences commonly induced by the central AHU fan and door closure.

Controlled Ventilation System Observations
A VanEE energy recovery ventilation system was installed in the garage. This ventilation unit could be energized in a number of ways:

a timer on the unit was set to periodically energize the unit;
an adjustable dehumidistat in the central area could energize the unit;
a switch in the bathrooms could energize the unit and increase the fan speed to the maximum.

When we arrived on-site, the VanEE was running continuously. It was found that the de-humidistat was set at approximately 30% relative humidity. Outside temperature conditions were mild and the relative humidity was above 30%. Under those conditions, the VanEE energy recovery ventilation unit would have continued to run since there was no potential for outdoor ventilation air to bring the interior relative humidity below 30%. For the Seattle climate, it may be better to limit the minimum dehumidistat setting to no lower than 30% or 40% relative humidity to avoid over ventilation and excess energy use. A common misconception is that energy recovery ventilation systems can be used to dehumidify the interior space. When outside air is dry with respect to inside air, the moisture transfer effect in the energy recovery process transfers moisture from the outgoing exhaust air stream to the incoming fresh air stream, limiting the drying potential of the outside air. Thus, energy recovery ventilators can dehumidify the interior space when outside air is much drier than inside air, but the interior space will be dehumidified less than if there was no energy recovery. To put it another way, energy recovery ventilation will keep the interior space more humid than it otherwise would have been without energy recovery when outside air is dry relative to inside air, and, energy recovery will keep the interior space less humid than it otherwise would have been without energy recovery, by reducing the incoming moisture load, when outside air is wet relative to inside air.

Nancy also pointed out that the fan-speed boost switch did not appear to have any effect in at least one of the bathrooms. Randy Nicklas mentioned that this was later fixed.

Fresh air was routed from the VanEE energy recovery unit to the return plenum of the AHU. Exhaust air was taken from the return plenum at the AHU a few feet upstream from where the fresh air was supplied. In this case, if the AHU fan did not operate with the VanEE, the VanEE exhaust and fresh air supply streams would likely short-circuit in the AHU return plenum. It would be better to locate the VanEE fresh air supply in the AHU supply plenum, using separate filtration for the fresh air. If this was done, instead of operating the AHU fan constantly with the VanEE, the FanRecyclerTM control could be used to periodically provide full distribution of ventilation air if the AHU fan had not already been operating for heating and cooling. This would reduce the AHU fan operational time, and save energy, while still providing sufficient distribution of fresh air. An additional benefit of the FanRecyclerTM control is that temperature and humidity conditions throughout the house would be periodically re-averaged if the AHU fan has not operated after a period of time, improving occupant comfort especially in closed rooms.

Combustion Appliance Venting

An induced-draft gas furnace and a natural-draft gas domestic hot water heater were located outside conditioned space in a framed enclosure in the garage. High/low vents for dilution air and combustion air were provided.

Air Filter Access
Access to the air filter at the furnace is very difficult due to limited space. Randy Nicklas has mentioned that in a later attempt to remedy this filter access problem, the access cover has been split, increasing return duct leakage dramatically. Additional gaskets should be installed to better seal the filter access cover at the furnace.

Attic Insulation
Observations in the main attic indicated a relatively poor application of the Icynene spray foam insulation. The average insulation thickness sprayed on the flat ceiling ranged between 3" to 5". However, the insulation was very thin in spots, and was entirely missing in some small areas, exposing ceiling drywall to the vented attic. Many return and supply duct boots in the attic were not fully insulated, leaving bare metal duct exposed to the attic. Boxes were constructed to cover recessed canister lights and exhaust fan housings. Some of these boxes were not completely insulated, leaving whole sides of some boxes exposed. Photographs shown on an attached page provide examples of these deficiencies.

It was later determined that the intended thickness of Icycnene spray foam insulation was 3.5 inches. This was to achieve a good seal between the attic and the house. Then, an additional 6 inches of blown rock wool insulation was to be installed on top of the foam to achieve a code required R-value of 30 h-ft2-F/Btu. The installation of rock wool never occurred. The insulation contractor is willing to return to finish the job if he can gain access to the house.

Concluding Recommendations:

Insulate the main return duct, located in the attic, that connects the main return plenum to the top of the AHU unit. Use R-6 insulation.
Make sure that all return and supply duct boots are fully insulated in the attic.
Make sure that the intended thickness of attic insulation covers the entire attic floor. This requires an additional six inches of blown rock wool insulation.
Use low VOC latex duct mastic (such as RCD #6) to seal all boots and pans to flooring and ceiling drywall at all supply registers and return grilles.
Additional gaskets should be installed to better seal the filter access cover at the furnace.
Use closed cell foam gaskets on top of the stop molding to better seal the attic access cover.
On future houses, consider eliminating the DuroZone system and the operation of the central air handling unit fan with the kitchen exhaust fan, unless additional airtightening was done and the house would intermittently depressurize more than 5 Pa.
On future houses, consider using the FanRecyclerTM control on the central air handling unit to periodically provide full distribution of ventilation air if the AHU fan hasn't already been operating for heating and cooling. This would be instead of operating the AHU fan continuously with the operation of the VanEE ventilation unit, yielding energy savings and longer equipment life.