Orders of Failure in Building Skin Design and Construction

Orders of Failure in Building Skin
Design and Construction
Jeffrey C.F. Ng, AIA, LEED AP
Intertek
311 W. 43rd Street, New York, NY 10036
Phone: 203-556-0116 • E-mail: jeffrey.ng@intertek.com
Jennifer Keegan, AAIA
and
Matthew Ridgway, PE, LEED GA
Intertek
350 Sentry Parkway, Bldg. 670, 2nd Fl., Blue Bell, PA 19422
Phone: 610-306-7199 • jennifer.keegan@intertek.com
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 S y m p o s i u m • No v e m be r 1 3 – 1 4 , 2 0 1 7 N g et a l . • 1 7
Abstract
The design of high-performance building skins in a 21st-century city demands more
forethought than building codes or zoning ordinances can anticipate. This session will
explore, through forensic case studies, why newly built high-performance skins fail with
the lens of an analytical tool called the “Orders of Failure.” These orders of failures organize
conventional typologies of building skin failures, such as air, water, thermal, structure,
glass, finish, and acoustics, by correlating them with three spatial dimensions and the
fourth dimension of time. Using the information developed by these matrices, the speakers
will discuss common myths of building skin design and the need for alternative bidding and
value-engineering approaches that maintain building skin design integrity and utilize stateof-
the-art prefabrication technologies to assure delivery of durable, high-performance building
skins that enhance comfort and adapt to the seasons in the 21st-century city.
Speaker
Jeffrey C.F. Ng, AIA, LEED AP — Intertek
Jeffrey Ng is an architect with over 35 years of experience
integrating art and science in adaptive high-performance building
skin designs. He has consulted on innovative award-winning national
and international architectural projects. He has presented at numerous
conferences, including the 2017 AIA National Convention, the
2013 Façade Design and Delivery Conference, and the 2010 BESS
Symposium at Cal Poly in Pomona. Prior to joining Intertek, Ng was VP
and lead building skin consultant at Thornton Tomasetti. He has been
associated with leading architectural firms Ehrenkrantz & Eckstut,
Davis Brody, Cossutta Associates, I. M. Pei & Partners, and S.O.M.
Jennifer Keegan, AAIA — Intertek
Jennifer Keegan has 18 years of experience as a building enclosure
consultant specializing in assessment, design, and remediation of
building enclosures. She has investigated failures and provided construction
administration, condition surveys, and design per reviews of
residential and commercial façades; has served as an expert witness;
and offered litigation services. Utilizing her expertise in the built world,
Keegan brings depth and focus to building enclosure commissioning
and a proactive team approach to meet project performance requirements.
Nonpresenting Coauthor
Matthew Ridgway PE, LEED GA — Intertek
1 8 • Ng et a l . 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 S y m p o s i u m • No v e m be r 1 3 – 1 4 , 2 0 1 7
INTRODUCTION
This paper introduces an alternative
way of organizing data of building skin failures
that may facilitate scientific inquiry to
bridge the many disciplines within building
skin science. The intent of this classification
methodology is to help building skin
designers, engineers, and consultants to
clearly communicate scientific findings to
their project teams in order to clarify project
requirements, promote holistic decisionmaking,
and assure long-term performance
and reliability of building skin systems.
Common forms of building skin failures
can be correlated with the four dimensions
of space and time by organizing failures
with dimensional discontinuities, spatial
overcapacity, and degradation due to time.
These Four Orders of Failure create a
hierarchical organization of failures across
conventional building skin performance
typologies. This methodology can be applied
in building skin design, construction, and
forensics to highlight key common characteristics
of building skin failures across
each of the orders, regarding causation,
diagnostics, and repair. Lessons drawn from
these illustrations of the Orders of Failure
can be used to explore alternative methods
for evaluating building skin designs, value
engineering, procurement, fabrication, and
delivery.
BUILDING SKIN FAILURES
The purpose of a building skin is to
filter, moderate, or control the outside natural
environment within which the building
is situated. Building skin failures generally
stem from noncompliance of materials,
components, and assemblies with specified
project performance requirements, manufacturers’
installation instructions, applicable
building codes, and industry standards.
These failures become evident in the course
of the actual use and operation of the building
skin. They can also become evident
through testing under controlled conditions,
intended to simulate some or part of the
natural environment.
The study of building skin failures is a
physical science. Building skin failures are
objective and generally observable, measurable,
and verifiable by established testing
methods. Failures are effects that result
from particular causes. Failures often are
predictable through analysis and investigation.
The environmental conditions that
result in failures are typically discoverable
and repeatable. Similar to any science,
there will be anomalous events that will
be challenging to measure, analyze, test,
or recreate. These anomalies may be difficult
to explain, based on the conventional
paradigm. Studies in these anomalies may
offer insight to the need of a paradigm shift.
The evolution of the study of air barriers is
an example of such a paradigm shift. The
practice of locating the vapor barrier behind
the interior insulation has been superseded
by exterior water-resistive barriers and continuous
insulation.
As physical phenomena in a spatial
world, building skin failures can be understood
and categorized in relation to the
three spatial dimensions and the fourth
dimension of time. This form of categorization
of building skin failures into four orders
can be used as a methodology to organize
building skin failures and to reveal common
“dimensions” or norms in phenomena, causation,
diagnostics, mitigation, remedies,
and repair of building skin within each of
the four orders.
Building skin failures are typically studied
as deficiencies within discrete typologies
of building skin performances, as noted
below:
1. Water penetration
2. Air infiltration
3. Thermal performance
4. Condensation resistance
5. Structural performance
6. Glass performance
7. Finish performance
8. Acoustic effectiveness
9. Fire protection continuity
This conventional organization of common
building skin failures can result in a
large compendium of information, myriad
industry standards, product literature,
and lengthy checklists that we as building
designers, engineers, and consultants must
refine to be useful communication tools
during design or as quality control or quality
assurance programs during construction.
This paper offers another approach to educate
and communicate with project teams
during the design and construction process,
so that decisions made in fast-track, costdriven
projects are, at least, fully informed.
When building skin systems are
designed, installed, and then fail, camps
are often divided between those claiming
systemic improper construction and those
claiming improper design. The truth is often
a combination of these two. Building skin
systems, such as cavity walls, double skins,
pressure-equalized cavities, and continuous
exterior insulation, fail due to poor installation
and improper design. These failures
have been compounded by poor communication
and poor contractor coordination.
Building skin professionals must continue
to evaluate extensive historical and available
test data to develop technical refinements
to better adapt these building skin
systems to their specific building locations
and use.
THE FOUR ORDERS OF FAILURE
The Four Orders of Failure is a taxonomy
or classification system of building
skin failures based on four dimensions of
space and time, as related to component
interfaces, system capacity, and temporal
degradation.
First-Order Failures
First-order failures consist of a linear
opening, discontinuity, or nonuniformity
within a component of the building skin
assembly that is either in noncompliance
with product specifications or may result in
failure of the assembly. Generally, a firstorder
failure is a local defect resulting in
the failure of the building skin component.
The location of a first-order failure can
be identified as a local discrete point in the
surface of the component(s) that can be
geometrically described by a line segment
associated with a starting point on the exterior
surface of the component and an end
Orders of Failure in Building Skin
Design and Construction
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 S y m p o s i u m • No v e m be r 1 3 – 1 4 , 2 0 1 7 N g et a l . • 1 9
point on the interior surface of the component.
A first-order failure in terms of water
infiltration is a penetration or gap that
allows a linear stream of water to enter the
building skin assembly in an uncontrolled
manner, such as an unsealed penetration
in the weather-resistive barrier membrane
(Figure 1).
First-order failures are localized component
defects or deficiencies. They are generally
noncompliant with installation, manufacturing,
or design requirements caused
by poor workmanship or improper quality
control, such as gaps in sealant application
or improper depth sealant joints. First-order
failures can be repetitive, such as penetrations
through exterior insulation with
thermally conductive fasteners. First-order
failures can also be a result of accidents on
site or in the fabrication plant, such as cuts
in a membrane or damage to finished work.
First-order failures could be observed
during the installation of the components by
on-site inspections, or during plant manufacturing
by in-line inspections. They may
become progressively more difficult to discover
after installation, without probes and
diagnostic testing. Repetitive first-order failures
could be identified during shop drawing
reviews, prior to installation. Accidental
first-order failures could be prevented with
proper sequencing of work by adjacent
trades, QA/QC programs,
and protection
of finished work.
First-order failures
can be managed
or remediated by using
specified and approved
components, observing
manufacturers’
requirements, implementing
project-specific
quality control programs,
proper training
and checklists of critical
installation requirements,
and third-party inspections.
Second-Order Failures
Second-order failures consist of a planar
gap or joint discontinuity between components
of the building skin assembly.
Generally, a second-order failure is a defect
or deficiency in the assembly of components
within the building skin assembly. The
location of a second-order failure can be
identified at the interface between adjoining
components that can be geometrically
described by the shape of the planar area of
discontinuity between components.
Second-order failures in terms of water
infiltration consist of a planar gap within
the water management system that allows
water to enter the building skin assembly
in an uncontrolled manner, which may be
caused by poor substrate preparation for
sealant applications (Figure 2).
Other examples of second-order failures
include improper sequencing of
flashing materials and failed weld joints
in metal fabrications.
Second-order failures are defects or
deficiencies in the assembly of components
within the building skin system.
They are generally noncompliant with
design documents, shop drawings, manufacturers’
assembly requirements, or
industry standards.
For field-installed assemblies, second-
order failures can be managed or
Figure 1 – First-order failure:
unsealed penetrations in the
WRB membrane.
Figure 2 – Second-order
failure: unadhered sealant
to concrete substrate.
2 0 • Ng et a l . 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 S y m p o s i u m • No v e m be r 1 3 – 1 4 , 2 0 1 7
remediated with retaining qualified installers,
pre-installation meetings, construction
supervision, coordination of work between
trades, field mock-ups, field testing of first
installations, and third-party inspections.
For plant-fabricated assemblies, secondorder
failures can be managed or remediated
by observance of the engineered
shop drawings; use of proper, well-maintained
equipment; staff training; and inline
inspections with state-of-the-art testing
methods. Coordinated design details
and specifications and field verification of
as-built conditions will facilitate the development
of accurate shop drawings by the
contractors for assembly of the building
skin system. Active clarification of details
and ongoing review of work between trades
by contractors, manufacturers, architects,
engineers, and consultants can play a key
role in mitigating second-order building
skin failures.
Third-Order Failures
Third-order failures are failures of size,
volume, or design capacity of the building
skin system. Generally, a third-order
failure is a failure in compliance with
performance requirements of the building
skin system that may be
a result of improper engineering
or design.
A third-order failure in
water infiltration consists
of an overload of the capacity
of the building skin
water management system
resulting in uncontrolled
water intrusion, such as
employing a 1-inch-high
sub-frame flashing assembly,
designed for 6.24-psf
water resistance, for a window
system that encounters
regular severe rain
events with wind speeds
in excess of 50 mph. Other
design defects that can
lead to third-order failures
include the improper spacing
and sizing of weeps
and the omission of vents
for mitigating pressure differential
in the air space of
a masonry cavity wall.
A third-order failure
may be due to forces that
cause an overload of the
design capacity of the building skin that
was not anticipated by prescriptive building
codes, such as extreme wind loads
at the corners of high-rise building skins
with complex geometries. A third-order
failure can also be the result of conflicting
performance requirements, such as ADAcompliant
doors installed within a curtainwall
system that requires 15-psf water
resistance.
First- and second-order failures can
appear to be a third-order failure in the
water management system. Failures such
as improper flashing material, improper
application of barrier membrane, the
obstruction of weep holes (first order) or the
improper lapping, configuration, and termination
of flashing (second order) can initially
appear as a capacity issue. These apparent
escalations of the Orders of Failures
can generally be remediated by compliance
with the system design documents.
However, a third-order failure is a failure of
a system that was installed in accordance
with design documents, whose capacity or
performance capability is exceeded by actual
environmental conditions, user expectations,
or regulatory requirements. Thirdorder
failures can also be a result of limitations
in capacity or performance caused by
aesthetic requirements, such as low or no
curb height for proper termination of roof
membranes, oversized panels, and complex
geometries.
Third-order failures can be managed or
anticipated by computer simulation modeling,
physical scale model testing, laboratory
mock-up testing, and field performance
verification testing (Figure 3). Remediation
of third-order failures will become progressively
more difficult as the project moves
from the design phase and further into
the construction phase. Solutions may
require alternative designs and applications
of innovative materials, assemblies,
and fabrication technologies. Peer review
of design documents can assist in assuring
consistency in performance requirements
for different building skin system
and related components. Single-source
responsibility of the building skin assembly
can be key in assuring compatibility of all
components, compliance with performance
requirements across all the interfaces of the
entire building skin, and comprehensive,
project-specific warranty of assemblies and
components.
Figure 3 – Third-order failure: field-testing storefront that had water intrusion.
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 S y m p o s i u m • No v e m be r 1 3 – 1 4 , 2 0 1 7 N g et a l . • 2 1
Fourth-Order Failures
Fourth-order failures are failures due to
degradation of components and assemblies
over time. Generally, a fourth-order failure
is a progressive decline in performance of
the building skin system, caused by repeated
usage; exposure to vapor, moisture, UV
or airborne debris; seasonal change; freeze/
thaw cycles; severe weathering; or extreme
environmental conditions. A fourth-order
failure in water infiltration consists of degradations
of sealant or gasket materials
from UV exposure, limiting joint movement
and causing failure of flashing membranes
due to extreme temperature variations,
such as within parapets and copings.
Another form of fourth-order failure can
be due to the limitations of the building
skin system to adapt to changing building
codes and industry standards. Building
code changes may require a completed
building to add insulation, change windows
and glass types, or improve the air barrier
system to meet the revised R-value requirement
of building skin.
Fourth-order failures can result in the
occurrence of lower-order failures. Longterm
exposure of the roof coping to ultraviolet
light can result in
first-order cohesive failure
and second-order
adhesive failure of the
sealant joints. Freeze
thaw cycles can cause
second-order cracking in
the mortar joint below
the coping. Weathering
and movement can cause
third-order cracking and
displacement of the coping.
High temperature
variations, combined
with radiant energy
reflected from neighboring
structures, can
cause deterioration of the
through-wall flashing,
which can result in firstorder
water intrusion
around the reinforcing
bars supporting the coping.
Similarly, deterioration
of a concrete coping
can result in corrosion
and failure of structural
steel reinforcing bars.
This can lead to secondorder
separation of the concrete from the
reinforcing bars and first-order water intrusion
(Figure 4). These de-escalations of the
Orders of Failure can be characterized as a
“cascade of the Orders of Failure.”
Conversely, lower-order failures can
result in the appearance of fourth-order
failures. First- and second-order water
infiltration in laminated glass can cause
third-order delamination. This can result
in fourth-order yellowing or bubbling of the
interlayer.
While fourth-order failures, in general,
cannot be completely avoided, they can be
Order Air Infiltration Water Penetration Thermal Structural
1 Gaps in corner or in Unsealed gaps or penetrations; Conductive fastener; Damaged fasteners
improperly installed gaskets holes in WRB membranes damaged insulation or members
2 Improperly sized or missing Unadhered sealant to Gaps between Failed connections;
gaskets; improperly substrate; improperly insulation batts adjustable anchors
installed doors lapped WRB membranes
3 Improperly designed Low flashing height Insufficient insulation Excess deflection
weatherstripping at thickness of members
doors/windows
4 Gasket UV degradation Sealant UV degradation Insulation sagging Corrosion of
or degradation members
Table 1 – Typologies of building skin failures.
Figure 4 – Fourth-order failure: deteriorated concrete coping with exposed rebars.
2 2 • Ng et a l . 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 S y m p o s i u m • No v e m be r 1 3 – 1 4 , 2 0 1 7
mitigated by the use of durable materials
and high-performance assemblies, design
redundancy, post-installation performance
verification testing, proper maintenance programs,
and timely repairs. Maintainability
can be improved with custom window washing
systems, modular panels, and providing
for accessibility. Designing for adaptability
or exceeding current industry standards
and building codes can contribute significantly
to the long-term performance of
the building skin system. Nature can be a
source of inspiration for innovative design
such as simulating a natural ventilated system,
developing self-cleaning surfaces, and
using self-crystalizing concrete aggregate
that may stand the test of time.
The matrix in Table 1 lists representative
Orders of Failure for various typologies
of building skin performance.
APPLICATIONS OF THE FOUR
ORDERS OF FAILURE
The Four Orders of Failure in building
skin can be used as an analytical tool for
presenting information in matrices to assist
decision-making during design, value-engineering,
bidding, and construction phases
of a project.
Design Phase
The Orders of Failures can also provide
insights in developing and analyzing building
skin design solutions that bridge across
individual modes of building skin failures
within a common order, such as providing
an air cavity that mitigates pressure differentials
to manage water and air infiltration
in the building skin assembly.
Table 2 shows a design solution matrix
developed for a 12-story luxury condominium
project with a 13,000-sq.-ft. building
skin that was designed to include articulated
limestone walls and a window-wall
system with alternating glass planes. The
matrix identifies the performance requirements
and proposes design options for
mitigating first-, second-, third-, and fourthorder
failures.
Order Air Infiltration Water Penetration Thermal Structural
0.060 cfm/sf 12 psf Opaque: 15-R 50-psf Wind Load
Window: 0.38 U
1 Self-healing WRB membrane QC sealant joint installation Nonconductive fasteners. Observe edge
to assure proper geometry. Sealed penetrations limit dimensions
2 Vulcanized corner gaskets; Manufacturer-approved flashing T&G insulation; Engineered adjustable
seal all gaps assembly; adhesion testing seal all gaps anchors to slab
and use of primer, if needed
3 Continuous WRB membrane Field-test sealant; 4-in. continuous Structural window
with interior air seal proper height window gutter insulation calculations per
ASCE-7
4 Air-permeable WRB Two-stage silicone seals Hydrophobic Separation of
membrane; silicone gaskets with closed-cell backer rods insulation dissimilar metals
Table 2 – Design solution matrix.
Order Conventional Field-Installed Prefabricated Windows Unitized Curtainwall Prefabricated Stone
Window and Veneer Stone and Cubic Stone and Windows
1 Susceptible to poor Fewer joints with large cubic Fewer joints in Factory-fabricated
workmanship stone; field-installed windows factory-fabricated composite stone
and flashing assemblies CW panels; separate panels and window
unitized stone panels panels
2 Field-installed joints between Field-installed joints between Field-installed joints Field-installed
stone/stone and stone/stone and between CW/stone joints between
window/stone windows/stone panels panels
3 No assembly test; stone wall Lab test of similar windows Pre-engineered lab Single source
cannot be field-tested; only; stone wall cannot be tested CW and pre-engineered, tested
field-performance verification field tested; field-performance stone panels; total stone/window
tests on limited windows verification tests on limited field-performance assembly; fieldwindows
verification test on performance
field-installed joints verification test on
field-installed joints
4 Long installation time Lead time > 6 mos.; Long lead time; Longest lead time;
>12 months; installation <8 mos.; installation <4 mos.; installation <4 mos.;
5-20 years life expectancy 10-20 years life expectancy 20-50 years life 50-year life
expectancy expectancy
Table 3 – Building skin fabrication alternatives.
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 S y m p o s i u m • No v e m be r 1 3 – 1 4 , 2 0 1 7 N g et a l . • 2 3
Fabrication Alternatives
The same 12-story condominium project
has an aggressive eight-month installation
schedule. The matrix in Table 3 compares
four options for installation of the window
and stone panel wall systems. The matrix
highlights the different risks in the Four
Orders of Failure associated with each of
the four fabrication options. The prefabricated
and unitized options minimized the
risk of failures, reduced installation time,
and provided the best assurance for longterm
performance.
Value Engineering
Value engineering is a necessary form
of cost control in most projects—large or
small. These value-engineering exercises
are generally conducted by the construction
manager or general contractor and not led
by the project building skin design team
who initiated, developed, and coordinated
the integrated building skin system design.
These sessions are often conducted under
high-stress conditions when the project is
either about to go out to bid or is in the middle
of construction. Consequently, there is
often not sufficient time and priority to fully
understand the impact on the performance
requirements of the project, or the “ripple
effect” on the design and performance of
adjacent systems. Value engineering is generally
based on standard conventionally
built alternatives and often performed in a
piecemeal fashion. This singular focus on
cost reduction can not only compromise
building skin performance but can have
a major unexpected impact on the project
schedule due to out-of-sequence work, conventional
field-installed assemblies, use of
less-durable materials, and modifications
of adjacent systems. The quality control
and assurance programs that were an integral
part of the original single-source wall
assembly, specified by the design architect,
are often foregone by piecemeal value engineering.
Value engineering requires a more comprehensive
organization and presentation
of the cost and consequences of each alternative
to facilitate an informed decisionmaking
process that will not compromise
performance and adversely affect the project
schedule and long-term durability of the
building skin. Organizing relevant information
about potential building skin failures
in a clear, succinct, and hierarchic way
would significantly contribute to limiting
short-sighted cost savings at the expense of
long-term performance.
The example below relates to a project,
where midway into construction, in order to
reduce construction costs, it was proposed
to install ribbon windows with horizontal
precast concrete spandrels on the threestory
penthouse in lieu of the original unitized
curtainwall system.
The matrices in Tables 4 and 5 compare
the potential first-, second-, third-,
and fourth-order failures for this value
engineering option and the original unitized
curtainwall. The tables illustrate the significant
risks in third-order failure through
performance requirements with this value
engineering option. They highlight the need
for proper redesign of the exterior wall system
to address these potential deficiencies.
Particularly, the entire water management
system needs to be modified to provide for a
proper WRB membrane, flashing assembly,
and two-stage sealant system to reduce the
risk of first- and second-order water and air
leakage. The impact of thermal bridging on
thermal performance needs to be addressed.
Alternatives to the proposed addition of
interior building insulation must be considered.
These matrices also inform the owner
of the requirement for annual inspections
and maintenance.
CONSTRUCTION CASE STUDIES
The orders of failure reveal themselves
during the construction phase and present
inherent risks when best practices are
bulldozed by cost and schedule-focused
projects. Fabrication before shop drawings,
installation prior to successful mock-ups,
and projects without enclosure consultants
are three scenarios that result from lowcost
and schedule-driven projects, which,
in turn, create opportunities for failure. The
following case studies will discuss the risks
associated with each of these scenarios.
Owners are pulled in many directions,
and advice received from their architect,
façade consultant, general contractor, subcontractor,
and fabricator is often biased.
While we are all working towards the same
objective, we all have different priorities
and ideas on what is necessary to complete
the project. There is tremendous pressure
Order Air Infiltration Water Penetration Thermal Structural
0.10 cfm/sf 12 psf 15 R-value 50-psf Wind Load
1 Gaps in field-installed Water intrusion at intersection Thermal-bridged Check for load
weather and air seal of gasketed vertical and interior building imposed by windows
field-installed window joints insulation on concrete spandrels
2 Leaks at field-installed Damage to WRB membrane Discontinuity of Field-installed
window/concrete during installation insulation at floor slabs adjustable anchors;
spandrel joints check spandrels are
within tolerances
3 Continuity of air barrier Primary seal has no backup; Difficult to install Check structure for
difficult to install and flashing difficult to install. continuous insulation load imposed by
and properly terminate in concrete spandrels concrete spandrels
4 High maintenance cost of High risk of water intrusion and Interior insulation may Concrete may
air/weather seals; damage to interior finishes condensate; risk of mold require spall repair
<10 years life expectancy and structure; and corrosion to in 10-20 years;
<10 years life expectancy structure; <20 years <20 years life
life expectancy expectancy
Table 4 – Risks associated with ribbon windows with horizontal precast panels (value-engineered option).
2 4 • Ng et a l . 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 S y m p o s i u m • No v e m be r 1 3 – 1 4 , 2 0 1 7
to turn buildings over on schedule and
on budget. Therefore, many decisions are
made from a time-and-cost perspective,
increasing the risk of poor performance
and inefficiency. In the long run, many of
these decisions result in additional time
and cost to compensate for proceeding out
of sequence.
Case Study 1 – Fabrication Prior
to Approved Shop Drawings
Almost 30% of nearly 300 architects,
owners, and consultants polled in a number
of symposia at which we have recently presented
along the East Coast of the United
States have admitted to falling victim to this
practice on at least one project. This is a
disturbing trend that speaks to the prioritization
of schedule and the willingness to
take on the risks caused by out-of-sequence
work.
The project in this case study began
with a failed laboratory mock-up. The large
window assembly, which was approximately
14 feet wide and 10 feet tall, could not properly
manage water. The window assembly
was redesigned into a curtainwall assembly.
The first set of shop drawings was produced
within four months. Four months later, as
the shop drawing revision process continued
to resolve the identified failures, including
a lack of weeps (first-order failure),
discontinuity in the plane of airtightness
between the window assembly and adjacent
metal panel (second-order failure), and a
water capacity issue (third-order failure),
the manufacturer convinced the owner to
release fabrication, promising to make up
lost time in the schedule due to the redesign
process. Despite our recommendations and
continued identification of potential first-,
second-, and third-order failures, fabrication
was released with no approved shop
drawings or valid laboratory test data.
Despite the promises of improving
the schedule, fabrication was continually
delayed. After one year of waiting for units,
the first fabrication plant visit revealed first-
, second-, and third-order failures including:
1. First-order failures: Gasket failures
2. Second-order failures: Gaps and discontinuities
in the insulation
3. Third-order failures: Glass roller
wave distortion and omission of
insulation
The manufacturer promised to make
up lost time in the schedule and recommended
on-site fabrication in lieu of plant
fabrication. Despite our recommendations
to the contrary, the owner agreed to allow
shipment of crates of parts to the project for
on-site fabrication, still without approved
shop drawings.
Four months later, a small crew was on
site to fabricate the first unit. Fabricating a
unit on site without shop drawings proved
more difficult than anticipated. It took two
months to fabricate the first window, which
was to be a “first-install” mock-up. Over
the next three months, the mock-up failed
three times before another installer was
hired to fabricate on site and install the
next mock-up.
One month later, the new first-install
mock-up was tested and failed. The mockup
unit was retested three times over the
next five months, and the following failures
were identified: leaks at fasteners (firstorder
failure), sealant joint adhesion failure
(second-order failure) and excessive air
leakage (third-order failure). However, in an
effort to make up some of the lost time in
the schedule, unit installation continued.
Through a series of isolated testings, an
issue with the gaskets was identified on
the operable windows that required fabrication
of custom gaskets, which resulted
in another three-month delay. When the
custom gaskets arrived, nearly 80% of the
units were fabricated and installed. All of
the operable lites needed to be retrofitted
and reinstalled. The laboratory mock-up
test will be scheduled after all 550 units are
installed in the condominiums, more than
two years behind schedule.
The decision to fabricate without shop
drawings, which became further complicated
by the decision to fabricate on site,
resulted in a project delivered nearly two
years behind schedule at an undisclosed
financial loss. This project will be completed
without approved shop drawings and with
a laboratory mock-up test that will hopefully
be successful upon completion of the
project.
Case Study 2 – Installation Prior
to Approved Mock-ups
Another scenario that results from lowcost
and schedule-driven projects is installation
prior to successful mock-ups. This
scenario is a more common occurrence
and was touched upon in the previous case
study.
Freestanding mock-ups are often
excluded from fast-track schedules and
budget-focused projects, leaving project
teams to rely on the first-install mock-up.
Order Air Infiltration Water Penetration Thermal Structural
0.10 cfm/sf 12 psf 15 R-value 50-psf Wind Load
1 QA/QC of factory-installed Check-field install of QA/QC of Check for proper
gaskets bridge flashing between units factory-installed glass fasteners and locations
2 Gasketed joints to be Gasketed joints to be Field-installed 3-in. Check tolerances of
field-installed within field-installed within mineral wool with field-installed
tolerances tolerances no pins or gaps adjustable anchors
to slab
3 Lab-tested, pre-engineered Field-test multi-chamber, Computer model of Lab-tested,
single-source system; gasketed CW to verify condensation risks; pre-engineered
>20 years life expectancy performance; >20 years life system, install per
>20 years life expectancy expectancy shop drawings;
>20 years life
expectancy
Table 5 – Risks associated with unitized curtainwall system (original design).
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 S y m p o s i u m • No v e m be r 1 3 – 1 4 , 2 0 1 7 N g et a l . • 2 5
Delays due to fabrication, weather, cure
times, and mock-up failures often result in
installation prior to the completion of a successful
mock-up.
As is common for many projects, the
first-install mock-up in this example failed.
While the team took the time to understand
the failure mode, identifying first-, second-,
and third-order failures, the contractor continued
to install the windows while the diagnostic
evaluation continued.
After the second failed mock-up test,
the window was removed so the team could
more closely scrutinize the installation.
Upon review, it was noted that an errant
hole was drilled through the thermal break
during fabrication, which resulted in the
leak. Unfortunately, 50% of the windows
were already installed and needed to be
removed, repaired, and reinstalled.
If the project team agreed to work
through the failed mock-up and identify
the first-order failure, all of the windows
could have been repaired prior to installation.
Better yet, if a freestanding mock-up
had been tested prior to fabrication, the
machining error could have been corrected
in the factory.
Case Study 3 – Design and Construction
Without a Building Skin Consultant
A third outcome resulting from low-cost
and schedule-driven projects is the exclusion
of the building skin consultant from
design projects or delegation of responsibility
for the enclosure to the wrong team
member. All inherent responsibility and
risk for the enclosure are delegated to the
architect. At times, this leads to the need
for an enclosure consultant to resolve performance
issues after the project is complete.
One example of this scenario is a
12-story condominium that was completed
just two years prior to our engagement.
Claims of tenant comfort issues and high
utility bills resulted in a preliminary assessment.
Results of a whole-building air leakage
test in conjunction with an infrared
scan validated their concerns. The air leakage
through the enclosure was more than
double the code limit, and the continuous
insulation was not installed in a codecompliant
manner. Discontinuities in the
air barrier (a second-order failure) resulted
in a third-order failure—a capacity issue in
air leakage. Discontinuities in the insulation
(a second-order failure) resulted in a
third-order failure—a capacity issue in thermal
performance. Both negatively impacted
energy efficiency and tenant comfort. Both
of these cascading failures could have been
addressed in the design and construction
phase of the project, had an enclosure consultant
been involved.
FORENSIC C ASE ST UDIES
The forensics of building skins are
replete with cautionary tales of building
skins that were designed with the highest
of intentions, but failed to perform as
intended. They deteriorate over time due
to seasonal changes, severe weathering,
repeated usage, or exposure to UV. These
temporal degradations generally become
evident in multiple dimensional interfaces
of components—from linear gaps where air
can infiltrate, and planar adhesive failures
of sealants, to three-dimensional overload of
a system’s ability to resist bulkwater intrusion.
Without reconstructing entire portions
of the building skins, many repairs of these
fourth-order failures involve patch solutions
that degrade the original state-of-the-art,
multi-chamber wall system to a traditional
barrier wall system. Lessons learned from
these dynamic fourth-order failures may
assist in developing better-performing remedies
and preventive maintenance of building
skins. Additionally, these lessons can
inform design engineering and construction
management of new building skins.
Not all building skin failures are immediately
evident after construction. Some
failures may not be noticed until well after
post-installation testing has been performed
and performance has been duly verified.
Third- and fourth-order failures are failures
of design capacity and degradation over
time. In order to fully understand these
higher-order failures and their effects on
building skins, a designer must take a
holistic view of building systems, components,
and assemblies.
For example, a mock-up of an exterior
wall system designed to be a perfect barrier
may pass water infiltration criteria during
initial performance verification testing, only
to later reveal the same fully installed system
performs poorly over time. More often
than not, such problems can be traced back
to lower-order failures. Failed or improperly
installed sealants (first order) will become
point sources for water infiltration, while
linear discontinuities at expansion/control
joints (second order), which were not tested
during initial mock-up phases, can further
contribute to water infiltration. These unanticipated
water infiltrations can significantly
increase the rate of deterioration if unmitigated.
Corrosion of internal reinforcing
steel might advance deterioration of exterior
walls to the extent that spalling or distortion
occurs or causes mold on interior framing/
finishes (fourth order). A barrier wall system
does not provide any drainage capacity
and, therefore, would fail to meet the design
criteria for water resistance. This lack of
redundancy in the water management system
is one reason barrier wall systems generally
do not meet criteria for industry best
practices. When similar failure modes are
applied to a rainscreen system, a properly
designed and implemented drainage plane
and flashings will be capable of managing
water leakage. However, if these components
are undersized or improperly sloped,
they could be overwhelmed during normal
operation and permit third-order water
intrusion to the interior of the building.
Of primary importance to the higherorder
failures are issues that are often sorted
out early in the design process via identification
of the owner’s project requirements
and design intent. While cost and schedule
may dominate the construction process, it
is recommended that the project requirements
and design intent be referenced.
Efforts to drive cost and schedule may
adversely affect previous decisions related
to service life, durability, and redundancy.
Designers and fabricators have a responsibility
to understand how building skins
will fail or at least anticipate how decisions
might affect performance requirements and
other failure typologies.
Unanticipated performance and/or
serviceability requirements often result in
cascading Orders of Failure defined by a
compounding effect that lower Orders of
Failure might have in causing higher order
failures such as decreasing the service life
of enclosure components and assemblies.
When failures cascade, characterization of
deficient conditions may be complicated, as
the relationship between cause and effect
of related assemblies and failures can be
convoluted. It may be useful to summarize
symptoms in a matrix in order to help identify
root causes.
Case Study – 15-Year-Old
Institutional Building
An institutional building has had significant
leaks since the day it opened approxi-
2 6 • Ng et a l . 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 S y m p o s i u m • No v e m be r 1 3 – 1 4 , 2 0 1 7
mately 15 years ago. The building enclosure
consisted of traditionally built masonry cavity
walls and EPDM roofs that were topped
with an integrated 3½-acre unitized skylight
system with a clear span over the interior
space. The system is framed over 5-in. steel
tubes to form the primary support structure
for the glazing. Preassembled unitized
glazing panels are set on an extruded aluminum
sub-frame installed over the steel,
and typically include four lites of 7/16-inches,
stacked and then captured by extruded
aluminum ribs. Individual lites were originally
factory-glazed with structural silicone
sealant to intermediate aluminum frames.
Field-applied perimeter sealant joints were
installed at the joints between adjacent
unitized panels and the extruded aluminum
capture system. At end conditions, the unitized
panels turn 90 degrees from vertical to
horizontal where the skylight terminates at
a structurally supported horizontal extruded
aluminum frame that is installed over
horizontal EPDM roofing.
An initial post-occupancy evaluation
attributed leaking of the building primarily
to improperly installed sealant joints at
locations within the skylight, which had over
125,000 linear feet of sealant. Subsequent
analysis years after the completion of the
building revealed more systemic water infiltration
issues that are all interrelated:
1. First-order failures: Several joints
at shallow setting blocks and a few
adhesion issues of the sealant joints
related to workmanship were noted.
2. Second-order failures: The extruded
aluminum capture system butt
joints and panel end-cap butt joints
were sealed to one another such that
the primary weather seal had very
little (approximately <¼-inch) bite
when installed in accordance with
the manufacturer requirements,
which was found to create linear
sealant joint failures in situ.
3. Third-order failures:
a. A value management exercise
during construction changed
skylight end units from an offset
sloped panel to a horizontal
panel. This caused the weep
system to be unable to manage
water infiltration cascading
down from the large field of the
skylight.
b. The sub-frame was designed to
anticipate drainage of condensation
but did not anticipate future
failures of sealant joints and
bulk water infiltration.
4. Fourth-order failures: Unanticipated
water infiltration within the skylight
system has led to the premature
degradation of the PVB interlayer at
the laminated glass edges.
The above list is not comprehensive to
this case study, but is an accurate representation
of how issues can be interrelated.
Following characterization of the leak activity,
it was determined that the second-order
failures were causing the most systemic
problems related to the volume of water
infiltration. Addressing the issues of transition
joints resolved approximately 95% of
the most significant leakage. However, the
damage sustained on the interior of the
building steel, drywall, and other finishes
remains an ongoing operations and maintenance
problem, due to the inability of
the weep systems to facilitate serviceability
requirements of the aging skylight sealants.
Anticipating durability and serviceability
requirements is necessary for any building to
be sustained into the future. When possible,
it is useful to design and construct buildings
of components and assemblies with similar
maintenance and service cycles, but it is not
unusual in modern construction to use materials
of varying service requirements adjacent
to one another. Third- and fourth-order failures
can be mitigated by providing formal separation
between systems, creating efficiencies
for future maintenance and capital projects.
For example, a modern tower was constructed
of a combination of a custom unitized
curtainwall, along with prefabricated EIFS
panels at a podium deck. There is available
service history of the prefabricated EIFS panel
system to anticipate the maintenance requirements
for the EIFS panel joints. However,
the lack of service history of the custom
unitized curtainwall precludes fully anticipating
its future maintenance requirements.
In this case, a formal articulation and flashing
between adjacent EIFS and curtainwall
systems would serve to reduce the risk of
cascading Orders of Failure between the
systems.
It is also important to note that aging
buildings exist in a world of changing
requirements where building code is often
the lowest common denominator. In the
restoration of older stock or historical buildings,
designers can be faced with resolving
issues related to air infiltration, thermal
resistance, and condensation risk. A review
of NY State Building energy code requirements
for a historical building where an
energy code variance was not available,
revealed the following fluid changes in the
thermal performance requirements for exterior
walls:
1. 1959 NYS BC:
a. Vapor barrier required between
floor and sub-floor
b. No insulation requirement
2. IECC 2009: R-20ci
3. NYS ECC 2010: R-20c
4. IECC 2012: NYS ECC 2015: R-25ci
5. IECC 2015: R-30ci
In our drive for energy efficiency, an
under-insulated building would be considered
a fourth-order failure. More generally,
serviceability requirements of building skins
are changing. Just as a building constructed
in a flood plain would be deemed unacceptable
if it did not anticipate future floods,
our building skin designs need to address
our rapidly changing environments.
CONCLUSION
The Four Orders of Building Skin
Failure is a classification system that spatially
correlates the multi-dimensional
interfaces of building skin components
with the building skin failures. This classification
system can assist building skin
professionals in presenting relevant information
about project requirements and
potential failures in a clear ordered form
to our owners, contractors, and project
teams, and thereby contribute to holistic
decision-making to assure long-term performance.
This methodology can be used
as an analytical tool to promote innovative,
adaptable, high-performance building skins
that can address changing environmental
and regulatory requirements.
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 S y m p o s i u m • No v e m be r 1 3 – 1 4 , 2 0 1 7 N g et a l . • 2 7