Fan Pressure Testing for Large, Complex Buildings

Fan Press ure Testing
for Large, Complex Buildings
Terry Brennan
Camroden Associates, Inc.
7240 East Carter Road, Westmoreland, NY 13490
Phone: 315-336-7955 • Fax: 315-336-6180 • E-mail: terry@camroden.com
S y m p o s i u m o n B u i l d i n g E n v e l o p e T e c h n o l o g y • Oc t o be r 2 0 1 4 B r e n n a n • 3 5
AB STRA CT
This paper will present the important lessons learned based on over 30 years of testing
the airtightness of buildings. Common problems encountered while setting up and conducting
tests and their solutions will be presented. Details and locations that are difficult to
air seal and those that are easy to air seal will be presented. Airtightness test targets and
methods needed to achieve them will be covered. Buildings tested include commercial and
institutional buildings, laboratories, hospitals, museums, pools, schools, warehouses, and
industrial applications. They range in size from banana-ripening rooms to 13-story towers
to 800,000-sq.-ft. warehouses. The results of the ASHRAE 1478 (Measuring Airtightness of
Mid- and High-Rise Nonresidential Buildings) research project are included.
SPEA KER
Terry Brennan – Camroden Associates, Inc.
Terr y Brennan is a building scientist and educator and has studied buildings since
the 1970s. He has provided research, investigation of building-related problems, training,
curriculum development, and program support for the EPA, building owners and managers,
architects, engineers, and several state health departments. He is the technical lead for
the ASHRAE 1478, Measuring Airtightness of Mid- and High-Rise Nonresidential Buildings
research project. Brennan chairs the Air Barrier Association of America’s Whole Building
Airtightness Testing Committee. This committee revised the Army Corps of Engineers’
whole-building test protocol in 2012 and is currently an ASTM Task Group (WK35913)
on New Standard Whole-Building Enclosure Airtightness Compliance. He is a member of
ASTM E06 and ASHRAE 62.2, the committee on ventilation for low-rise residential buildings.
Brennan has served as a consultant to the National Academies of Science Committee
on Dampness and Health in Buildings. He is the primary author of the forthcoming EPA
Moisture Control Guide.
3 6 • B r e n n a n S y m p o s i u m o n B u i l d i n g E n v e l o p e T e c h n o l o g y • Oc t o be r 2 0 1 4
S y m p o s i u m o n B u i l d i n g E n v e l o p e T e c h n o l o g y • Oc t o be r 2 0 1 4 B r e n n a n • 3 7
INTRODUCTION
Fan pressure tests are conducted to
test air barrier systems on whole buildings,
apartments or suites of rooms, isolated
special-use areas within buildings, and
air-handling equipment (e.g., ducts, air
handlers, and hot-air solar collectors). This
paper focuses on testing entire building
enclosures or apartments in multifamily
buildings.
Fan pressurization tests may be conducted
for one or more of the following
reasons:
• To measure the airtightness of an
enclosure to determine whether or
not it meets an airtightness target
• To help find air leakage sites using
infrared scanning, theatrical fog, or
chemical smoke tracers
• As part of a quality assurance (QA)
inspection and testing program for
air barrier installation
• To assess the effectiveness of retrofit
air-sealing efforts, either quantitatively
or qualitatively
• To measure air leakage rates or
leakage areas for use in building
energy-use calculations
• To test the ability of installed air
barrier materials to resist specific
pressure loads (typically assemblies
are tested at pressure differences
of 75 and 298 Pascals, and whole
buildings are tested at 50 or 75
Pascals).
Whole-building fan pressurization tests
have been conducted to measure the airtightness
of single-family residential buildings
in the U.S. and Canada since the
1970s. They have been extensively used in
weatherization programs and as part of QA
for high-performance homebuilding programs
(e.g., EnergyStar, LEED® for Homes,
and Passive House). Efforts to extend the
tests to larger and more complex buildings
began in the 1980s. In the past ten
years, North American manufacturers of
test equipment have developed methods
for networking test fans and electronic
micro-manometers so that test data can be
collected and analyzed by computer software
programs. This has made it practical
to test larger buildings than ever before.
The ASHRAE project, 1478-RP, Measuring
Airtightness of Mid- and High-Rise Non-
Residential Buildings, developed a test protocol
specifically for testing large buildings
and conducted tests on 16 buildings
built since 2000 (Anis 2014). The buildings
ranged between four and 14 stories
and between 54,000 and 362,000 sq. ft.
in floor area. The tested buildings were
located in climate zones 2 through 7 of the
International Energy Conservation Code
(IECC) Climate Zone Map. Tests were completed
between 2011 and 2012.
Over the last 30 years, manufacturers of
building products have developed materials
and systems that provide continuous air
barrier systems in commercial and institutional
buildings. During this time period,
the use of air barrier systems in commercial
and institutional buildings has become
common practice.
By limiting airflow through the thermal
enclosure, continuous air barriers perform
two crucial functions in high-performance
enclosures:
• Limiting heat transfer by accidental
air leakage through the enclosure
• Limiting the transport of water
vapor by accidental air leakage (e.g.,
preventing condensation by keeping
warm, humid outdoor air away
from air-conditioned, chilled interior
surfaces; and keeping warm, humid
indoor air from reaching exterior
enclosure materials that are chilled
by cold outdoor air)
In addition to blocking airflow, many
air barrier materials act as capillary breaks
and serve dual duty as both air barrier and
rainwater control. Even in airtight assemblies,
condensation may occur if the vapor
barrier is located in the wrong spot relative
to more permeable insulating materials.
Some air barrier materials accidentally or
intentionally behave as nearly perfect water
vapor barriers, while others are more vaporopen.
Condensation control must be evaluated
for whole assemblies, because condensation
is a function of vapor migration and
the surface temperature of materials.
Continuous air barriers are also
required for buildings or rooms that are
intended to shelter vulnerable occupants,
materials, equipment, or processes from
damaging environmental agents (e.g., bone
marrow transplant wards; art, artifact, or
musical instrument storage or display;
computers and telecommunications equipment;
lithium batteries; solid-state chip
manufacturing). Air barriers are also used
to prevent damaging, annoying, or valuable
environmental agents from escaping from
contained sources (e.g., biohazard level 3 or
4 containment, paint booths, fruit-ripening
rooms).
Continuous air barriers are required by
code in a dozen U.S. states and Canada.
The ICC family of model codes requires
continuous air barriers. One path to demonstrate
compliance is for the results of
a fan pressurization test to be equal to or
less than an airtightness target. The Army
Corps of Engineers (ACE), General Services
Agency (GSA), EnergyStar multifamily
high-rises, Passive House U.S. multifamily
projects, and projects using the Unified
Facilities Guide Specifications are required
to have continuous air barriers and to pass
a whole-building fan pressurization test.
The ACE requirements for air barriers and
whole-building testing have resulted in a
large increase in the use of air barriers and
fan pressure testing in commercial and
institutional buildings.
TEST METHODS
A fan pressure test is conducted on a
building enclosure by using fans to exhaust
air from or blow air into a test enclosure. At
its simplest, a qualitative test can be used
to find air leaks in buildings. Air leaks can
be found by simply feeling the breeze on
the skin. Infrared scanning, theatrical fog,
and smoke tracers can be used to make it
easier or quicker to find air leaks in build-
Fan Press ure Testing
for Large, Complex Buildings
ings. These methods are described in ASTM
E1186, Standard Practices for Air Leakage
Site Detection in Building Envelopes and Air
Barrier Systems.
The fan pressure test becomes a quantitative
test by measuring:
• The indoor/outdoor pressure
induced across a building enclosure
• The exhaust or supply airflow
required to induce the measured
pressure difference as illustrated in
Figure 1
Air is exhausted from the enclosure by
a fan that is sealed into the wall. This lowers
the air pressure inside the box relative
to the outside of the box, drawing air in
through the air leaks in the enclosure (for
simplicity, represented in Figure 1 as a single
hole). The mass of air drawn in is equal
to the mass of air exhausted through the
fan. The airflow and the resulting change
in indoor/outdoor pressure difference are
measured. The tighter the box, the smaller
the airflow required to induce a particular
test pressure difference. Airflow and the
resulting pressure difference are measures
of enclosure airtightness.
Applying this simple concept to buildings,
we run into a few practical problems.
Buildings are not simple, closed boxes
with no interior airflow restrictions. They
have open windows and doors. They are
filled with partitions and floors that restrict
airflow from one place to another. Fanpowered
exhaust and outdoor air fans in an
operating building induce pressure differences
between inside and outside the building.
Indoor/outdoor pressure differences
are affected by wind and stack effect.
To conduct a fan pressurization test, a
building must be prepared so that it resembles
the condition in Figure 1 as closely as
possible. All the exterior windows and doors
must be closed (except those containing test
fans). Interior doors must be opened to allow
air to move freely throughout the building.
Exhaust and outdoor air fans must be shut
off and sealed, at least by closing dampers.
After a building is set up for a test, the
effects of wind and stack effect pressures
must be subtracted out. This is done by
measuring a baseline pressure difference
with the test fans turned off and sealed so
that wind and stack effects are the only
reason there would be an indoor/outdoor
pressure difference. It is not the absolute
indoor/outdoor pressure difference that is
important, it is the amount the indoor/
outdoor pressure difference changes when
a measured airflow is exhausted from or
supplied to the enclosure. If the mass of air
inside the box is an open, single zone, then
the change in pressure difference is the
same at all points on the enclosure. For the
purposes of this paper, the pressure difference
induced between inside and outside by
the operation of the test fan will be called
the enclosure pressure.
There are two methods commonly used
3 8 • B r e n n a n S y m p o s i u m o n B u i l d i n g E n v e l o p e T e c h n o l o g y • Oc t o be r 2 0 1 4
Figure 1 – A schematic illustration of a fan pressurization test.
Figure 2 – Repeated-reference pressure test. A baseline is collected, the test fans
are turned on, the reference pressure is induced, and the airflow required is
measured. Although not shown in this graph, the measured flow for the test was
15,200 cubic feet per minute (cfm) (corrected for temperature and altitude).
to make a quantitative measurement of
building air leakage. The simplest is the
repeated single-reference pressure method.
In this method, prepare the building as
described above, measure a baseline pressure,
turn on test fans until a reference
enclosure pressure is reached, and measure
the airflow required to induce the reference
enclosure pressure. Repeat this procedure
at least five times. Historically, reference
test pressures of 50 Pascals and 75 Pascals
have been used for this test. This method
is described in ASTM E1827, Standard
Test Methods for Determining Airtightness of
Buildings Using an Orifice Blower Door, and
Chapter 8 of the Residential Energy Services
Network’s (RESNET’s) National Mortgage
and Housing Home Energy Rating System
(HERS) Standards. Figure 2 shows the
results of a repeated single-reference pressure
test.
This test illustrates the important principle
that it is the change in the indoor/
outdoor pressure difference induced by the
operation of the test fans that is important.
The reference enclosure pressure for the
test is 75 Pascals. The baseline averages
around -8 Pascals. Baseline enclosure pressures
measured at grade are frequently
negative due to stack and wind effects. This
is the case for low-rise and taller buildings
during warm weather in air-conditioned
buildings. Test fans provide supply airflow
to raise the enclosure pressure 75 Pascals
(from -8 to +67 Pascals). The reported result
is 15,200 cfm at 75 Pascals. An airflow and
an induced enclosure pressure are the fundamental
result. Repeating the test provides
the data needed to calculate uncertainty
in the measurements. (Note: Wind effects
during this test seem to contribute plus
or minus 2 or 3 Pascal fluctuations to the
baseline and enclosure pressure measurements.)
If airtightness data at indoor/outdoor
enclosure pressures that are typical of
ordinary operation are desired, a multipoint
regression method is used to collect
and analyze the data. Typical operational
enclosure pressures are in the range of 4 to
10 Pascals. At these pressure differences,
wind and stack effects become a large fraction
of the test signal. For that reason,
data are collected over a large range of
enclosure pressures, and linear regression
analysis is applied to transformed nonlinear
data. ASTM E779, Standard Test Method
for Determining Air Leakage Rate by Fan
Pressurization, details this method. Figure 3
illustrates data from this kind of test.
The relationship between the test airflow
and the enclosure pressure or a range
of enclosure pressures is not linear. It very
nearly follows a power law, as shown by
Equation 1.
Qfan = C(ΔPenclosure)n
Equation 1
Where:
• Qfan = the measured airflow through
the test fan (or fans in the case of
large buildings)
• ΔPenclosure = the change in the indooroutdoor
pressure difference induced
by operation of the test fans (in this
paper called the enclosure pressure)
• C = a flow coefficient that is a characteristic
of the size and shape of the
air leakage sites
• n = is the flow exponent with theoretical
limits between 0.5 and 1.0
Analysis includes calculation of the flow
coefficient C, exponent n, and a 95% confidence
interval. These values allow calculation
of flows at the lower reference enclosure
pressure of 4 Pascals and 10 Pascals, as
well as the higher ones of 50 and 75 Pascals.
The following test methods and protocols
are the most frequently used in the U.S.:
• ASTM E779-10, Standard Test
Method for Determining Air Leakage
Rate by Fan Pressurization
• ASTM E1827-10, Standard Test Methods
for Determining Airtightness of
Buildings Using an Orifice Blower Door
• RESNET Chapter 8 of the National
Mortgage and Housing HERS
Standards
• U.S. ACE Air Leakage Test Protocol
for Measuring Air Leakage in
Buildings
• The Air Barrier Association of
America’s (ABAA’s) Whole-Building
Fan Pressure Test Committee – Standard
Method for Building Enclosure
Airtightness Compliance Testing
S y m p o s i u m o n B u i l d i n g E n v e l o p e T e c h n o l o g y • Oc t o be r 2 0 1 4 B r e n n a n • 3 9
Figure 3 – Multipoint regression method: Regression analysis on transformed
nonlinear data fit to Equation 1. Air leakage at enclosure pressure less than 10
Pascals can be extrapolated using this method.
The two ASTM standards do not include
many of the issues that must be addressed
in order to test large, complex buildings.
They don’t include acceptance criteria for
determining whether or not a building
meets an airtightness specification. All of
the standards and protocols listed above
reference one or both of the ASTM methods
as the basis for collecting and analyzing
data, including calculation of uncertainty.
The techniques used to calculate uncertainty
are different, depending on how the
data are collected and analyzed. Calculated
uncertainty is used to determine whether
or not the data quality objectives have been
met. Data quality objectives depend on the
purpose of the test. For example, if the test
is being conducted to determine whether
or not an enclosure meets an airtightness
specification, and the test result is close
to the target, a small uncertainty interval
is needed. If the test is part of a survey in
which the data will be placed in a histogram,
a larger uncertainty interval may be
acceptable.
The RESNET standard allows both
repeated single-point and multipoint regression
methods and includes acceptance criteria
for determining whether or not a
building meets airtightness specifications. It
does not address preparation of larger, more
complex buildings. The ACE test protocol
allows only multipoint regression, includes
acceptance criteria, and addresses many of
the preparation issues found in large, complex
buildings. The ABAA Whole Building
Test Committee has drafted and balloted
a new standard test method, ABAA Whole-
Building Fan Pressure Test Committee’s
Standard Method for Building Enclosure
Airtightness Compliance Testing. It allows
multipoint regression and repeated singleand
two-point test methods, includes acceptance
criteria, and addresses the issues
found in large, complex buildings, including
many not addressed by other standards.
It is useful to be able to compare test
results between buildings. In order to do
so, a variety of metrics have been used over
the past 30 years. The most commonly used
metrics for large buildings are cfm at an
enclosure pressure of 75 Pascals divided by
the total surface area of the test enclosure
(cfm75/ft2). The surface area of the enclosure
includes the roof, walls, and foundation
floor. Residential programs often normalize
by the building volume, resulting in
air changes per hour at 50 Pascals (ACH50).
ACH50 is the cfm at an enclosure pressure
of 50 Pascals times 60 min./hr. divided by
the volume of the building.
Neither of these metrics is perfect for
comparing buildings. Normalizing to the
surface area makes low-rise buildings with
large footprints appear more airtight than
taller, more complex buildings, because the
large roofing membranes and concrete floor
slabs dominate. ACH50 runs into problems
comparing buildings over a wide range of
sizes because as size increases, surface
areas and volumes change at different rates.
Figure 4 shows the results for the
tests conducted during the ASHRAE 1478
research project. In this project, the buildings
were tested in pressurization and
depressurization mode. The HVAC penetrations
had fans off, dampers in the closed
position, and HVAC penetrations temporarily
sealed. The results are normalized by the
enclosure surface area (top, bottom, and
sides included). Figure 4 shows the results
for a pressurization test, a depressurization
test, and the average of the two for each
building. The results ranged from a low of
0.06 cfm75/ft2 to a high of 0.74 cfm75/ft2.
The average is 0.29 and the median is 0.24
cfm75/ft2. Half of the buildings in the study
met the ACE’s target of 0.25 cfm75/ft2, and
75% met the GSA target of 0.40 cfm75/ft2.
All of the buildings in the study were volunteered.
All of them were owner-occupied or
institutional buildings. Six of the buildings
had air barriers specifically included in the
drawings and specifications or had an air
barrier consultant as part of the design
or QA team. It is suspected that the study
sample is biased toward tighter buildings.
Here is an abbreviated list of codes
and building programs that include wholebuilding
fan pressure tests and their compliance
targets:
• 2012 IECC – Continuous air barriers;
0.40 cfm75/ft2 enclosure is one
path to compliance.
• GSA – Continuous air barriers and
the enclosure is less than 0.40
cfm75/ft2 enclosure (all six sides)
P100.
• ACE – Continuous air barriers,
and the enclosure is less than 0.25
cfm75/ft2 enclosure (some ACE projects
use 0.15 cfm75/ft2).
• Passive House requires continuous
air barriers, and the enclosure is
less than 0.6 ACH50 (around 0.03
– 0.12 cfm75/ft2 enclosure [all six
sides]).
• British code for museums and archi-
4 0 • B r e n n a n S y m p o s i u m o n B u i l d i n g E n v e l o p e T e c h n o l o g y • Oc t o be r 2 0 1 4
Figure 4 – Test results normalized to enclosure surface area for the 16 mid- to
high-rise buildings tested in ASHRAE project 1478.
val storage is less than 0.08 cfm75/ft2
of enclosure.
The results for the 1478 project include:
• 12 buildings that meet the GSA target
• 8 buildings that meet the ACE target
• 5 buildings in the range to meet the
Passive House target
• 1 building that meets the British code
for museums and archival storage
The following data are needed to successfully
conduct tests in accordance with
E779 or E1827 and normalize the results to
the building surface area:
• Airflow through test fans
• Inside/outside pressure difference
during baseline condition
• Inside/outside pressure difference
induced by test fans
• Temperature of air passing through
the test fans
• Altitude of the building
• Test enclosure surface area or building
volume
• Interior pressure differences to monitor
single-zone condition
TEST EQUIPMENT
There are three North American manufacturers
of test equipment for conducting
whole-building fan pressurization tests. The
equipment includes variable-speed test fans
with integrated flow-measuring sensors,
adjustable frames and panels to seal test
fans into door or window openings, and
electronic micro-manometers for measuring
air pressure differences. Two of the
manufacturers have developed a combination
of computer software and networking
hardware so data can be collected from
multiple micro-manometers measuring test
fan air flows and enclosure pressures. The
tests can be controlled from a computer,
and the data are automatically analyzed in
accordance with test standards. One of the
manufacturers produces a large, trailermounted
test fan capable of providing test
flows equal to that of several smaller test
fans. In the author’s experience, networking
smaller fans provides the simplest solution
for buildings with many interior partitions
and less than 100,000-cfm test flows. In
larger, more open buildings, the trailermounted
fans have the advantage of a single
setup location.
It is sometimes possible to test a building
using the ventilation fans. A depressurization
test can be conducted using
the exhaust fans in the building (with all
outdoor air intakes closed or sealed). A
pressurization test can be conducted using
the outdoor air ventilation flows (with all
exhaust outlets closed or sealed). The major
barrier to this method is making accurate
measurements of airflow through the ventilation
systems. In the author’s experience,
this process takes far more time than using
manufactured test equipment. This could
be overcome in new buildings by designing
the ventilation systems to conduct a fan
pressure test.
SPECIFYING A FAN
PRESSURE TEST
Whole-building fan pressure tests are
being specified for new buildings with
increasing frequency. Many of the specifications
crossing the author’s desk lack crucial
information or include requirements that
are impractical or inconsistent with other
requirements. For example, the specifications
may not include the surface area of
the test enclosure or even clearly identify
the bounding walls, floors, and ceilings of
the test enclosure. Some specifications prohibit
testing on rainy days or require the
testing agency to shovel snow off the roof
and two feet away from the ground floor.
Some require the testing agency to verify
that all air barriers have been installed
correctly. At the time of testing, most air
barriers are covered by interior or exterior
finishes.
Here is some guidance for specifying an
airtightness test:
• Make it clear who hires the testing
agency.
• Do not repeat or contradict items
that are adequately covered in the
referenced test methods.
• Include a rationale for any specification
that contradicts requirements
in the referenced test method.
• Provide the purpose of the test.
• Require qualifications of the testing
agency.
• List conditions that must be met
before the building can be tested.
• List contractor and owner responsibilities.
• Designate test methods (E779,
E1827, ACE protocol, RESNET,
ABAA).
• State the airtightness target (cfm75/
ft2 enclosure or ACH50).
• Name test enclosure boundaries.
• Test enclosure surface area or building
volume.
• Designate treatment of HVAC penetrations,
trash chutes, gas meter
rooms, mechanical rooms, coiling
doors and dock levelers.
• List acceptance criteria (how the test
result will be interpreted).
• Describe what happens if the building
fails.
• Include reporting requirements.
CONDUCTING A FAN
PRESSURE TEST
Planning
The basics of testing covered in the
introduction are the same for all buildings.
The logistics, setup, and testing for
large, complex buildings include many more
issues than when testing single-family residences.
There are more people involved:
owners, building management, building
facilities, contracted HVAC management
and maintenance, and security. The building
may have multiple occupants. Some
areas of the building may be inaccessible for
safety or security reasons (e.g., they’re not
going to let you into IRS offices, banks, law
offices, or the executive suite). In the case of
new construction, the owner is usually the
general contractor, special health and safety
issues are in play, and the HVAC systems
belong to the HVAC subcontractors. The
HVAC systems are more numerous and
more complex. Everything is bigger–more
stairs to climb, more corridors to walk, more
windows and doors to close and latch or
block open, as well as more test equipment
to move, set up, and calibrate.
Here’s a list of tasks that must be
accomplished to plan a test:
• Determine whether or not it is a
new, unoccupied building or an
occupied building. It is much easier
to test an unoccupied building than
an occupied building. Access is easier,
and there are fewer nonparticipants
who may accidentally interfere
with data collection.
• Identify parties. Parties involved
must include someone authorized
to approve conducting of the test;
someone who can provide access to
all rooms, mechanical rooms, and
safe access to all locations where
HVAC-related penetrations have to
be temporarily sealed; and someone
S y m p o s i u m o n B u i l d i n g E n v e l o p e T e c h n o l o g y • Oc t o be r 2 0 1 4 B r e n n a n • 4 1
authorized to place the HVAC equipment
into the proper status so that a
test can be conducted and to restore
the equipment to operational status
after the test. In some cases, security
personnel, an air sealing crew,
or health and safety personnel will
be involved.
• Select date. Select a date when
there will be the fewest people in the
building. In new buildings, this is
often easier than in occupied buildings.
In occupied buildings, testing
may be most practical during evenings,
weekends, or holidays.
• Identify test enclosure boundaries.
Test boundaries must be identified
and agreed to before testing.
Decisions must be made about how
mechanical rooms, basements, crawl
spaces, attics, and vestibules will be
treated. The surface area of the test
enclosure should be included in test
specifications. If the surface area is
not provided, conduct and document
the calculation clearly.
• Identify adequate power for test
equipment. Most test fans use
electric motors. Each fan typically
requires a separate 115-volt, 15-amp
circuit.
• Identify areas that are not clearly
inside or outside the test.
Mechanical rooms, attics, and crawl
spaces are examples of spaces that
may or may not be inside the test
area. Generally, if they are vented
to the outdoors through permanent,
nondampered openings, they are
considered outside the test enclosure.
If the air barrier systems and
insulation layers enclose these spaces,
they are considered inside the
space.
• Identify HVAC equipment that
must be turned off, dampers that
must be closed, and penetrations
that must be temporarily sealed.
How the HVAC penetrations are
treated depends on the test’s purpose.
If the goal is to test the quality
of design and installation of air barrier
systems in the building enclosure,
then all HVAC-related openings
typically have their associated
fans turned off, their dampers in
the closed position, and outdoor air
and exhaust louvers and grilles are
temporarily masked. If the intention
is to test the building in a more
operational condition, then HVAC
penetrations will have associated
fans turned off and dampers closed,
but they will not be temporarily
air-sealed. The ABAA test standard
contains a relatively complete list of
treating HVAC equipment and related
penetrations; the list is included
in Table 1. Table 1 includes treatment
for coiling overhead doors,
trash disposal systems, and loading
dock levelers, as well as HVAC penetrations.
• Calculate how much test air is
needed. How much test capacity is
needed is fairly simple to calculate
for buildings that have a test specification
that includes an airtightness
target—e.g., area enclosure (ft2) x
0.25 cfm/ft2 at 75 Pascals for ACE
projects. For other projects, the airtightness
target per square foot of
enclosure 0.25 cfm/ft2 at 75 Pascals
must be replaced by the specified
target. If it is an existing building
with no specified airtightness target,
then the airtightness must be
guessed. For ordinary construction,
0.40 cfm75/ft2 is a good rule of
thumb. If the building is experiencing
problems associated with leaky
enclosures, a good rule of thumb
is 0.6 cfm75/ft2. Bear in mind that
some problem buildings the author
has tested are greater than 1.5
cfm75/ft2. If it is anticipated that the
building is relatively airtight, bring
enough fans to meet 0.25 cfm75/ft2.
PREPARING THE BUILDING
For most tests, the intention is to prepare
the test building so that it resembles
as nearly as possible the simple box in
Figure 1.
• The box is sealed except where test
fans are inserted into door or window
openings.
• The interior is wide open, creating a
single-zone condition as far as enclosure
pressures are concerned.
• Manometers with pressure taps
measure representative enclosure
pressures.
• Test fans are placed to provide good
enclosure pressure distribution.
In practice, making the building a sealed
box means closing all the doors, windows,
hatches, and HVAC-related penetrations
in the enclosure. Decisions must be made
about how to handle situations such as
mechanical rooms with doors to both the
interior and exterior. Should both doors be
closed? Should the exterior door be open
and the interior door closed? Or should the
interior door be open and the exterior door
be closed?
Creating an open single zone is often
achieved by simply opening all the interior
doors. Some situations may create barriers
to single-zone conditions:
• Rooms that cannot be opened for
security reasons. Often these rooms
are so well-connected by ductwork
that during testing they are within
10% of the enclosure pressure and
meet single-zone conditions.
• Portions of the building that are
within the test enclosure but isolated
from the rest of the building
by firewalls with no access doors
penetrating the firewall. These
have doors opening to the outside
and can be tested simultaneously by
maintaining zero pressure difference
between them and the main part
of the building using separate test
equipment. (See Figures 5 and 6.)
• Ceiling or floor cavities with large
air leaks to the outside.
The following list includes items that
cause problems for enclosure measurements:
• Wind pressures, especially fluctuating
wind pressure
• Water in tubing that connects outdoorenclosure
pressure taps to manometers
• Direct sunlight on long lengths of
tubes
• Small air leaks in tubing greater
than 100 feet in length
• Someone stepping on a tube
Although E779 and E1827 recommend
wind speeds less than 4 mph, the introduction
of computerized data acquisition and
analysis test equipment has made it possible
to collect data simultaneously from all
enclosure pressure and fan sensors averaging
over longer time periods and across
multiple façades, providing high-quality
data over a much wider range of wind conditions.
The ACE test protocol has no upper
4 2 • B r e n n a n S y m p o s i u m o n B u i l d i n g E n v e l o p e T e c h n o l o g y • Oc t o be r 2 0 1 4
limit on wind speed.
Place outdoor enclosure pressure taps
away from exterior corners, as close to the
intersection of the exterior wall and grade
as possible. Protect the open end of the tube
from water (e.g., tape tubing to well with
open end pointing down, and use an open
vessel that shelters the tubing end from
rain). Use tubing lengths less than 100 feet
when possible. Check tubing for air leaks as
part of quality assurance, and place tubing
where it is hard for someone to step on it.
In a wide-open building, the test fans can
be placed anywhere there are doors to the
exterior. (Note: On windy days, do not set up
a test fan on the windward side of the building.)
If interior walls and floors divide the
building into spaces connected by a single
door or a small number of doors, then the
test fans must be distributed among these
somewhat-isolated areas. Some examples:
• A single entry door into an area
that is closed off from the rest of
the building except for a single
door. Imagine three fans set up in
the entry door exhausting 15,000 cfm
from the area induces an enclosure pressure of -75 Pascals
relative to the outside. Some of the 15,000 cfm will come
through the exterior walls of the area from outside, but a lot
will come through the connecting door from the rest of the
building. If 13,000 cfm of the air comes through the connecting
door, and 2,000 cfm comes from outdoors and leaks
through the demising walls, an 18-Pascal pressure drop will
be induced across the connecting door. The portion of the
building outside the area with the fans will be at -58 Pascals
relative to the outside. In this case, single-zone conditions
will not be met.
• A six-story building in which the major air leakage site is
the roof/wall connection on the sixth floor. Two stairwells
with single doors to each floor connect the top five floors to
the ground floor. Test fans can only be installed on the ground
floor (short of removing a panel of glass from a curtain wall
on upper floors). If 30% of the test air comes through the air
leak on the sixth floor, when the rest of the building is at an
enclosure pressure of 75 Pascals relative to the outside, the
sixth floor may have a considerably lower enclosure pressure.
Single-zone conditions are not maintained.
Creating a single-zone test condition in large, complex buildings
is something that cannot be taken for granted. For that
reason, the ACE and ABAA test standards require that pressure
differences between the area where the test fans are located, and
portions of the building separated by walls or floor with few interior
door openings in them must be measured. This allows singlezone
conditions to be verified. For these standards, single-zone conditions
mean that the pressure difference between interior zones must
be less than 10% of the enclosure pressure. (See Figures 5 and 6.)
S y m p o s i u m o n B u i l d i n g E n v e l o p e T e c h n o l o g y • Oc t o be r 2 0 1 4 B r e n n a n • 4 3
Figure 5 – An illustration of testing a building that is divided in two by a
firewall with no openings through it. Test fans are set up in each portion of the
building. The fans in the red-outlined portion of the building are controlled to
induce the correct enclosure pressures on the building while the test fans in the
green-outlined portion of the building are controlled to maintain zero pressure
difference between the two. This provides results for each section separately and
for the overall building.
Figure 6 – Two separate zones tested as one building. The
first zone is the ground floor and is shaded in purple.
The second zone consists of the unshaded floors—the
basement and the second through the sixth floor. There
are no intentional openings between the two zones. A test
fan is located in each zone, and zero pressure difference is
maintained between zones during the test.
MORE ON HVAC PENETRATIONS
Preparation for the HVAC-related openings
in the building should be in accordance
with the test specification. The ABAA test
standard provides guidance for two default
test purposes:
• Air barrier systems enclosure test
(HVAC-related openings excluded)
• Operational enclosure test (air barrier
systems and HVAC-related openings
included)
In the first, the purpose of the test is to
determine whether the installed air barrier
systems in the walls, ceilings, roofs, and
foundation that form the thermal enclosure
have met a specified airtightness target. The
second test purpose is used if the intention
is to test the overall airtightness of the
buildings in a way that is more representative
of air leakage during more ordinary
4 4 • B r e n n a n S y m p o s i u m o n B u i l d i n g E n v e l o p e T e c h n o l o g y • Oc t o be r 2 0 1 4
Intentional Openings Air barrier systems Operational enclosure
enclosure test (HVAC-related test (air barrier systems
openings excluded) and HVAC-related
openings included)
Doors, hatches, and operable windows inside the test enclosure Open Open
Fire dampers Remain as found Remain as found
Windows, doors, skylights, and hatches in the bounding enclosure Closed and latched Closed and latched
Windows, doors, hatches, and operable windows in ancillary spaces Treat in accordance Treat in accordance
as identified in section 10.6.3 with 10.6.3 with 10.6.3
Dryer doors and air-handler access panels Closed and latched Closed and latched
Vented combustion appliance Off, unable to fire Off, unable to fire
Pilot light As found As found
Chimney or outlet for vented combustion device in a separate As found As found
mechanical room
B-vent or other insulated chimney serving a vented combustion Sealed* As found
appliance located within the test enclosure
Solid fuel appliances (fireplaces, wood-burning stoves, pellet stoves) No fires; dampers closed; No fires; dampers closed
chimney sealed*
Exhaust, outdoor air, make-up air fans, air handlers that serve Off Off
areas inside and outside the test enclosure
Clothes dryers Off Off
Air intake inlet with motorized dampers Dampers closed and sealed* Dampers closed
Air intake inlet with gravity dampers Sealed* As found
Air intake inlet with no dampers Sealed* Open unless fan(s) serving
inlet is operated >8000
hrs./year, then sealed*
Exhaust or relief air outlet with motorized dampers Dampers closed and sealed* Dampers closed
Exhaust or relief air outlet with gravity dampers Sealed* As found
Exhaust or relief air outlet with no damper Sealed* Open unless fan serving
outlet is operated >8,000
hours/year, then sealed*
Active or passive smoke control systems – air reliefs and intakes Sealed* As found
Intended powered or nonpowered openings for vented Sealed* As found
shafts/stairwells
Waste or linen handling systems and equipment Sealed* at rooftop chute Rooftop chute vent open,
vent opening. chute intake doors closed,
chute intake room and chute
discharge room doors
closed and latched, fire
dampers left as found
Clothes dryer outlets Sealed* As found; sealed* if
dryers are not yet installed
Exhaust, outdoor air, or make-up air fan that runs >8,000 hours per yr. Sealed* Sealed*
Ductwork that serves areas inside and outside the test enclosure Sealed* at supply and return Sealed* at supply and
return
Floor drains and plumbing Traps filled Traps filled
*Sealed means that an opening has been temporarily masked airtight (e.g., covered with self-adhering plastic film, taped polyethylene
film, or rigid board stock). See Annex A. Also see article 10.12.4.
Table 1 – Default conditions for building preparation.
operating conditions.
Table 1 from the ABAA
standard is shown herein.
It includes the HVAC-related
penetrations, the operational
status of certain HVAC
equipment, treatment of
doors and windows in the
test enclosure, treatment of
doors and hatches inside the
test enclosure, treatment of
some specialized spaces that
are isolated from the test
enclosure (e.g., trash conveyance
and compaction systems),
and HVAC systems
that serve the test enclosure
and zones outside the test
enclosure.
The ASHRAE 1478 project
used a simple test procedure
to estimate the air leakage
through HVAC-related
penetrations. The HVAC penetrations
were prepared by
turning off the exhaust and
make-up air fans, closing all
the outdoor air and exhaust
air dampers, and temporarily
sealing all the HVAC penetrations.
The building’s enclosure pressure
was then held at 75 Pascals, and a
baseline with HVAC penetrations sealed
was collected. Next the HVAC penetrations
were unsealed (but motorized dampers
remained closed; gravity dampers may
have been drawn closed or blown open), and
undampered openings (e.g., kitchen range
exhaust) become open holes. The results of
these simplified measurements are shown
in Figure 7.
THINGS THAT CAN GO WRONG
DURING A TEST
This is an abbreviated list of things that
can go wrong during a test. (Note: At this
point, the author fervently hopes he has
listed every conceivable error. However, this
list does not include the really dumb stuff,
like forgetting the fan controllers, going to
the wrong building, or conducting a test
with a window on the sixth floor open.)
• An entry door, a gravity damper, or
masking on an HVAC-related opening
may blow open during a pressurization
test. Catch this by noticing
that the test result is changing.
• Someone may open a door or stand
S y m p o s i u m o n B u i l d i n g E n v e l o p e T e c h n o l o g y • Oc t o be r 2 0 1 4 B r e n n a n • 4 5
Figure 7 – The triangles are test results with HVAC penetrations sealed, and the dots are with
the HVAC penetrations unsealed. In some buildings, it makes a significant difference; while in
others, it makes almost no difference.
Figure 8 – Depressurization and pressurization curves for a building with a large
gravity damper blowing open during the pressurization test.
on a tube during the test. This is
obvious if continuous monitoring is
taking place on-screen or if a data
point is automatically rejected by
internal software quality assurance.
• An interior door may be blown
closed, changing the test boundaries
partway through a test (noted by
change in the sound of the fans’ flow
or enclosure pressure data).
• Heavy wind or rain can make a test
impractical (pretty obvious).
The building depicted in Figure 8 was
tested in depressurization first. The data
have a small 95% confidence interval and fit
the power curve well. The lowest three pressurization
data points match the depressurization
data points well. However, each succeeding
pressurization data point increasingly
deviates from the depressurization
result. The building is getting progressively
more leaky. This is typical of what it looks
like when an entry door is unlatched and
the pressure in the building is preventing
it from closing completely. As the test
pressure increases, the door opens further
but not all the way because the door closer
provides resistance. A large gravity damper
and self-adhering duct mask blowing off
and HVAC louver also appear the same way.
INTERPRETING RESULTS
How test results are interpreted depends
on the purpose of the test. If one is testing to
determine whether or not a building meets
an airtightness specification, a test can
return one of three results:
• It clearly passes.
• It clearly fails.
• I can’t tell.
It is, of course, the “I can’t tell” result
that causes problems. The situation where
the smallest confidence interval is the best
is when the test result is within the uncertainty
of the measurement. No matter how
small, acceptance criteria must be used to
determine whether it is a pass or a fail.
ACE and ABAA test protocol specify:
• Calculate 95% confidence intervals
(in accordance with test method)
— Pass: If test result ≤0.25 cfm75/
ft2 and 95%CI ≤ 8%:
— If test result ≤0.25 cfm75/ft2 and
95%CI > 8%:
• Pass: If test result + 95%CI ≤
0.25 cfm75/ft2
• Fail: If test result + 95% CI >
0.25 cfm75/ft2
— Fail: If test result >0.25 cfm75/
ft2
• ABAA allows failing a building if the
specified airflow produces less than
75% of the reference test pressure.
None of the standards include acceptance
criteria if the test is being done to
compare to other buildings or to compare a
post-airtightening test to a pre-airtightening
test. Criteria would need to be developed to
meet the needs of each project.
REPORT
A test report should include, at minimum,
the following items:
• Date
• Weather conditions
• Building testing agency
• Building description
• Test target
• Test results
• Location of test equipment
• Identification of test enclosure
boundaries
• Test configuration of each intentional
opening in the building enclosure
• Test environmental conditions
• Any departures from test standard
or specifications
• Measured test results in tabular
form
• Conclusions and recommendations
FINDING THE BIG AIR LEAKS
There are a number of methods used to
find air leaks in buildings. Several of these
methods are described in ASTM E1186,
Standard Practices for Air Leakage Site
Detection in Building Envelopes and Air
Barrier Systems. Many of them depend
on operating a building under negative or
positive pressure. As such, after a wholebuilding
pressure test has been conducted,
it is a good time to seek the leaks. The methods
include:
• Listen—Some air leaks whistle, buzz,
flap, or whoosh when the building is
pressurized or depressurized.
• Use your skin to feel for air leaks
when depressurizing.
— Stand in stairwell doorways at
each floor: The floors with the
greatest breeze have the largest
leaks.
— Stand at the ends of corridors
where wings join other parts of
the building.
• Use chemical smoke bottles or theatrical
fog to track air flows.
• Use infrared scanners when there
is a large enough indoor/outdoor
temperature difference to identify air
leaks.
Using these techniques, the author has
found that the largest air leaks in large
buildings occur where enclosure systems
meet but don’t connect. This is the order
of most commonly encountered large leaks:
• Where roofs meet walls at the top of
the building
• Where lower roofs meet walls that
rise above them
• Where there are exterior soffits
beneath lower walls, canopies, and
plazas
• Where walls meet foundations
• Where utility shafts connect two or
more of the above items
CONCLUSIONS
Using modern test equipment, it has
become practical to conduct fan pressurization
tests on large, complex buildings. A
final whole-building airtightness test will:
• Document compliance or failure to
comply with an airtightness specification
• Provide induced pressure difference
to aid in finding remaining air leaks
(E1186)
• Provide motivation to effectively
install air barrier assemblies and
identify air leakage sites not detailed
in drawings and specifications during
construction
• Not otherwise make buildings more
airtight
REFERENCES
Wagdy Anis, 2014, “ASHRAE 1478-RP,
Measuring Airtightness of Mid- and
High-Rise Nonresidential Buildings
Research Results and Conclusions,”
WJE, 2014.
4 6 • B r e n n a n S y m p o s i u m o n B u i l d i n g E n v e l o p e T e c h n o l o g y • Oc t o be r 2 0 1 4