Risk Management Series Designing for Earthquakes
A Manual for Architects
FEMA 454 / December 2006

EXISTING BUILDINGS--EVALUATION AND RETROFIT BY WILLIAM HOLMES
8.1	INTRODUCTION
It is widely recognized that the most significant seismic risk in this country 
resides in our existing older building stock.  Much of the country has enforced 
seismic design for new buildings only recently; even on the West Coast, seismic 
codes enforced in the 1960s and even into the 1970s are now considered suspect. 
Although there are sometimes difficulties in coordinating seismic design 
requirements with other demands in new construction, the economical, social, and 
technical issues related to evaluation and retrofit of existing buildings are 
far more complex.
8.1.1 Contents of Chapter
This chapter describes the many issues associated with the risk from existing 
buildings, including common building code provisions covering older buildings, 
evaluation of the risks from any one given building and what levels of risk are 
deemed acceptable, and methods of mitigation of these risks through retrofit. A 
FEMA program to provide methods to mitigate the risk from existing buildings has 
been significant in advancing the state of the art, and this program is 
described in some detail, particularly the model building types used in most, if 
not all, of the FEMA documents.
8.1.2 Reference to Other Relevant Chapters
The basic concepts used for seismic design or estimation of seismic performance 
are the same for any building. Thus, the principles described in Chapters 4, 5, 
and 7 are applicable for older, potentially hazardous buildings. The development 
of seismic systems as seen through examples of buildings in the San Francisco 
Bay Area is particularly relevant to the issues covered in this chapter, because 
systems typically evolved due to poor performance of predecessors.
Nonstructural systems in buildings create the majority of dollar loss from 
buildings in earthquakes, although the quality of structural performance affects 
the level of that damage. Seismic protection of nonstructural systems, both for 
design of new buildings and for consideration in older buildings, is covered in 
Chapter 9.
8.2  BACKGROUND
In every older building, a host of “deficiencies” is identified as the state of 
the art of building design and building codes advances. Code requirements change 
because the risk or the expected performance resulting from the existing 
provisions is deemed unacceptable. Deficiencies are commonly identified due to 
increased understanding of fire and life safety, disabled access, hazardous 
materials, and design for natural hazards. Thus, it is not surprising that many 
of the older buildings in this country are seismically deficient, and many 
present the risk of life-threatening damage. It is not economically feasible to 
seismically retrofit every building built to codes with no or inadequate seismic 
provisions, nor is it culturally acceptable to replace them all.  These 
realities create a significant dilemma: How are the buildings that present a 
significant risk to life safety identified? How is the expected performance 
predicted for older buildings of high importance to businesses or for those 
needed in emergency response? How can we efficiently retrofit those buildings 
identified as high risk?
The term seismic deficiency is used in this chapter as a building characteristic 
that will lead to unacceptable seismic damage. Almost all buildings, even those 
designed to the latest seismic codes, will suffer earthquake damage given strong 
enough shaking; however, damage normally considered acceptable should not be 
expected in small events with frequent occurrence in a given region, and should 
not be life threatening. Damage may be judged unacceptable due to resulting high 
economic cost to the owner or due to resulting casualties. Therefore, conditions 
that create seismic deficiencies can vary from owner to owner, from building to 
building, and for different zones of seismicity.  For example, unbraced, 
unreinforced brick masonry residential chimneys are extremely vulnerable to 
earthquake shaking and should be considered a deficiency anywhere that shaking 
is postulated. On the other hand, unreinforced brick masonry walls, infilled 
between steel frame structural members, are expected to be damaged only in 
moderate to strong shaking and may not be considered a deficiency in lower 
seismic zones. Seismic deficiencies identified in this chapter generally will 
cause premature or unexpected damage, often leading to threats to life safety, 
in moderate to strong shaking. Buildings in regions of lower seismicity that 
expect Modified Mercalli Intensity (MMI) levels of not more than VII, or peak 
ground accelerations (PGA) of less than 0.10g (g = acceleration of gravity), may 
need special consideration.
Of course, every building with completed construction is “existing.” However, 
the term existing building has been taken to mean those buildings in the 
inventory that are not of current seismic design. These groups of buildings, 
some of which may not be very old, include buildings with a range of probable 
performance from collapse to minimal damage. In this chapter, the term “existing 
building” is used in this context.
8.2.1 Changes in Building Practice and Seismic Design Requirements Resulting in 
Buildings that are Currently Considered Seismically Inadequate
Chapter 7 documents in detail how building systems have evolved in the San 
Francisco Bay Area. This evolution was probably driven more by fire, economic, 
and construction issues than by a concern for seismic performance, at least in 
the first several decades of the twentieth century, but many changes took place. 
Similarly, Chapter 6 gives a brief history of the development of seismic codes 
in the United States. It is clear that for many reasons, building construction 
and structural systems change over time. In the time frame of the twentieth 
century, due to the rapid increase in understanding of the seismic response of 
buildings and parallel changes in code requirements, it should be expected that 
many older buildings will now be considered seismically deficient.
Seismic codes in this country did not develop at all until the 1920s, and at 
that time they were used voluntarily. A mandatory code was not enforced in 
California until 1933. Unreinforced masonry (URM) buildings, for example, a 
popular building type early in the twentieth century and now recognized as 
perhaps the worst seismic performer as a class, were not outlawed in the zones 
of high seismicity until the 1933 code, and continued to be built in much of the 
country with no significant seismic design provisions until quite recently. 
Figure 8-1 shows an example of typical URM damage. The first modern seismic 
codes were not consistently applied until the 1950s and 1960s, and then only in 
the known regions of high seismicity. Of course, not all buildings built before 
seismic codes are hazardous, but most are expected to suffer far more damage 
than currently built buildings
Even buildings designed to “modern” seismic codes may be susceptible to high 
damage levels and even collapse. Our understanding of seismic response has grown 
immensely since the early codes, and many building characteristics that lead to 
poor performance were allowed over the years. For example, concrete buildings of 
all types were economical and popular on the West Coast in the 1950s and 1960s. 
Unfortunately, seismic provisions for these buildings were inadequate at the 
time, and many of these buildings require retrofit. Highlights of inadequacies 
in past building codes that have, in many cases, created poor buildings are 
given below.
	Changes In Expected Shaking Intensity and Changes in Zoning
	Similar to advancements in structural analysis and the understanding of 
building performance, enormous advancements have been made in the understanding 
of ground motion, particularly since the 1950s and 1960s. The seismicity (that 
is, the probability of the occurrence of various-sized earthquakes from each 
source) of the country, the likely shaking intensity from those events depending 
on the distance from the source and the local soil conditions, and the exact 
dynamic nature of the shaking (the pattern of accelerations, velocities, or 
displacements) are all far better understood. These advancements have caused 
increases in seismic design forces from a factor of 1.5 in regions very near 
active faults (on the West Coast) to a factor of 2 to 3 in a few other areas of 
the country (e.g. Utah; Memphis, Tennessee). The damage to the first Olive View 
Hospital (Figure 8-2), in addition to other issues, was a result of inadequate 
zoning.
	Changes in Required Strength or Ductility
	As discussed in Chapter 4, the required lateral strength of a seismic 
system is generally traded off with the ductility (the ability to deform 
inelastically—normally controlled by the type of detailing of the components and 
connections) of the system.  Higher strength requires lower ductility and vice 
versa. The most significant changes in codes—reflecting better understanding of 
minimum requirements for life safety—are general increases in both strength and 
ductility. Many building types designed under previous seismic provisions, 
particularly in the 1950s, 1960s, and 1970s, are now considered deficient, 
including most concrete-moment frames and certain concrete shear walls, steel-
braced frames, and concrete tilt-ups.
	Other buildings designed with systems assumed to possess certain ductility 
have been proven inadequate. Figure 8-3 shows typical steel moment-frame damage 
in the Northridge earthquake caused by brittle behavior in a structural system 
previously thought to be of high ductility.
	Recognition of the Importance of Nonlinear Response
	Historically, a limited amount of damage that absorbed energy and  
softened the building, thus attracting less force, was thought o reduce seismic 
response. Although this is still true, it is now recognized that the extent and 
pattern of damage must be controlled. Early codes required the design of 
buildings for forces three to six times less than the elastic demand (the forces 
that the building would see if there was no damage), assuming that the 
beneficial characteristics of damage would make up the difference.  
Unfortunately, buildings are not uniformly damaged, and the change in structural 
properties after damage (nonlinear response) often will concentrate seismic 
displacement in one location. For example, if the lower story of a building is 
much more flexible or weaker than the stories above, damage will concentrate at 
this level and act as a fuse, never allowing significant energy absorption from 
damage to the structure above. This concentrated damage can easily compromise 
the gravity load-carrying capacity of the structure at that level, causing 
collapse. Similarly, concrete shear walls were often “discontinued” at lower 
floors and supported on columns or beams. Although the supporting structure was 
adequately designed for code forces, the wall above is often much stronger than 
that and remains undamaged, causing concentrated and unacceptable damage in the 
supporting structure. 
	A final example of this issue can be seen by considering torsion. As 
explained in Chapter 4, Section 4.11, torsion in a building is a twisting in 
plane caused by an imbalance in the location of the mass and resisting elements. 
Older buildings were often designed with a concentration of lateral strength and 
stiffness on one end—an elevator/stair tower, for example—and a small wall or 
frame at the other end to prevent torsion. However, when the small element is 
initially damaged, its strength and stiffness changes, and the building as a 
whole may respond with severe torsion.
	Current codes contain many rules to minimize configurations that could 
cause dangerous nonlinear response, as well as special design rules for elements 
potentially affected (e.g., columns supporting discontinuous shear walls). Olive 
View Hospital featured a weak first story in the main building, causing a 
permanent offset of more than one foot and near collapse; discontinuous shear 
walls in the towers caused a failure in the supporting beam and column frame, 
resulting in complete overturning of three of the four towers. Figure 8-4 shows 
a typical “tuck-under” apartment building in which the parking creates a weak 
story. 
8.2.2 Philosophy Developed for Treatment of  Existing Buildings
Building codes have long contained provisions to update life-safety features of 
buildings if the occupancy is significantly increased in number or level of 
hazard (transformation of a warehouse to office space, for example). As early as 
the mid-1960s, this concept started to be applied to seismic systems. Many older 
buildings contained entire structural systems no longer permitted in the code 
(e.g., URM, poorly reinforced concrete walls), and it quickly became obvious 
that 1) these components could not be removed, and 2) it was impractical and 
uneconomical to replace all older buildings. The “new” code could therefore not 
be applied directly to older buildings, and special criteria were needed to 
enable adaptive reuse while meeting the need to protect life safety of the 
occupants.  In some cases, an entirely new and code-complying lateral system was 
installed, while leaving existing, now prohibited, construction in place. (This 
procedure was used in many school buildings in California after the Field Act 
was passed in 1933—up until the school seismic safety program was essentially 
completed in the 1960s.) This procedure proved very costly and disruptive to the 
building and was thought to discourage both improved seismic safety and general 
redevelopment.
A philosophy quickly developed suggesting that existing buildings be treated 
differently from new buildings with regard to seismic requirements. First, 
archaic systems and materials would have to be recognized and incorporated into 
the expected seismic response, and secondly, due to cost and disruption, seismic 
design force levels could be smaller. The smaller force levels were rationalized 
as providing minimum life safety, but not the damage control of new buildings, a 
technically controversial and unproven concept, but popular. Commonly existing 
buildings were then designed to 75% of the values of new buildings—a factor that 
can still be found, either overtly or hidden, in many current codes and 
standards for existing buildings.
Occasionally, early standards for existing buildings incorporated a double 
standard, accepting a building that passed an evaluation using 75% of the code, 
but requiring retrofits to meet a higher standard, often 90% of the code.
8.2.3 Code Requirements Covering Existing Buildings
As the conceptual framework of evaluation and retrofit developed, legal and code 
requirements were also created. These policies and regulations can be described 
in three categories; active, passive, and post-earthquake. Active policies 
require that a defined set of buildings meet given seismic criteria in a certain 
time frame—without any triggering action by the owner. For example, all bearing-
wall masonry buildings in the community must meet the local seismic safety 
criteria within ten years. Passive policies require minimum seismic standards in 
existing buildings only when the owner “triggers” compliance by some action—
usually extensive remodeling, reconstruction, or addition. Post-earthquake 
policies developed by necessity after several damaging earthquakes, when it 
became obvious that repairing an obviously seismically poor building to its pre-
earthquake condition was a waste of money. It then became necessary to develop 
triggers to determine when a building could simply be repaired and when it had 
to be repaired and retrofitted as well.
 Passive Code Provisions  
As noted above, the development of requirements to seismically update a building 
under certain conditions mimicked economic and social policies well-established 
in building codes. Namely, the concept crystallized that if sufficient resources 
were spent to renew a building, particularly with a new occupancy, then the 
building should also be renewed seismically.  Seismic “renewal” was defined as 
providing life safety, but not necessarily reaching the performance expected 
from a new building. A second kind of trigger—that could be termed “trigger of 
opportunity”—has also been used in some communities. These policies try to take 
advantage of certain conditions that make seismic improvements more palatable to 
an owner, such as retrofit of single-family dwellings at point of sale or 
requiring roof diaphragm upgrades at the time of re-roofing.
Triggers based on alterations to the building are by far the most common and 
will be discussed further here. These policies are somewhat logical and 
consistent with code practice, but they created two difficult socioeconomic-
technical issues that have never been universally resolved. The first is the 
definition of what level of building renewal or increase in occupancy-risk 
triggers seismic upgrading. The second is to establish the acceptable level of 
seismic upgrading.
Most typically, the triggering mechanisms for seismic upgrade are undefined in 
the code and left up to the local building official. The Uniform Building Code, 
the predecessor to the IBC in the western states, waffled on this issue for 
decades, alternately inserting various hard triggers (e.g., 50% of building 
value spent in remodels) and ambiguous wording that gave the local building 
official ultimate power. The use of this mechanism, whether well defined in 
local regulation or placed in the hands of the building official, ultimately 
reflects the local attitude concerning seismic safety. Aggressive communities 
develop easily and commonly triggered criteria, and passive or unaware 
communities require seismic upgrade only in cases of complete reconstruction or 
have poorly defined, easily negotiated triggers. For more specific information 
on seismic triggers in codes, see the accompanying sidebar.

Box 1 Seismic Triggers in Codes
Events or actions that require owners to seismically retrofit their buildings 
are commonly called triggers. For example, in many communities, if an owner 
increases the occupancy risk (as measured by number of occupants, or by use of 
the building), they must perform many life-safety upgrades, including seismic 
ones. However, for practical and economic reasons, seldom does this trigger 
require conformance with seismic provision for new buildings, but rather with a 
special life-safety level of seismic protection, lower than that used for new 
buildings.
The code with the longest history in high-seismic regions, the Uniform Building 
Code (UBC), has long waffled on this issue. Besides the traditional code life-
safety trigger based on clear-cut changes in occupancy, this code over the years 
has included provisions using hard triggers based on the cost of construction, 
and soft language that almost completely left the decision to the local building 
official. The last edition of this code, the 1997 UBC, basically allowed any 
(non-occupancy related) alteration as long as the seismic capacity was not made 
worse.
The codes and standards that will replace the UBC are based on a federally 
funded effort and published by FEMA as the NEHRP Provisions.  These codes 
include the International Building Code (IBC), the National Fire Protection 
Agency (NFPA) and ASCE 7, a standard covering seismic design now ready for 
adoption by the other codes. This family of regulations has a common limit of a 
5% reduction in seismic capacity before “full compliance” is required.  This 
reduction could be caused by an increase in mass (as with an addition) or a 
decrease in strength (as with an alteration that places an opening in a shear 
wall). Full compliance in this case is defined as compliance with the provisions 
for new buildings that do not translate well to older buildings. It is unclear 
how this will be interpreted on the local level.
Many local jurisdictions, however, have adopted far more definitive triggers for 
seismic retrofit. San Francisco is a well-known example, perhaps because the 
triggers are fairly elaborately defined and because they have been in place for 
many years. In addition to the traditional occupancy-change trigger, San 
Francisco requires conformance with seismic provisions specially defined for 
existing buildings when substantial nonstructural alterations are done on 2/3 or 
more of the number of stories within a two-year period, or when substantial 
structural alterations, cumulative since 1973, affect more than 30% of the floor 
and roof area or structure.
The City of Portland, Oregon requires a seismic upgrade of URMs when the cost of 
construction exceeds $30/sf for a one story building, or $40/sf for buildings 
greater than two stories.
Although most jurisdictions leave this provision purposely loose, some also have 
adopted definitive triggers based on cost of construction, on the particular 
building type, and on various definitions of significant structural change.  
Government, in some cases, has been much more aggressive in setting triggers to 
activate seismic retrofit, perhaps to create a lawful need for funds which 
otherwise would be difficult to obtain. The state of California has set a 
definitive list of seismic triggers for state-owned buildings: a) alteration 
cost exceeding 25% of replacement cost; b) change in occupancy; c) reduction of 
lateral load capacity by more than 5% in any story; d) earthquake damage 
reducing lateral load capacity by more than 10% at any story.
The federal government likewise, in RP 6, [NIST, 2002], also has definitive 
triggers: a) change in occupancy that increases the building’s importance or 
level of use; b) alteration cost exceeding 50% of replacement cost; c) damage of 
any kind that has significantly degraded the lateral system; d) deemed to be a 
high seismic risk; and e) added to federal inventory though purchase or 
donation.
The regulations and policies governing any building, private or public, which 
will be significantly altered, should be researched in the planning stage to 
understand the effective seismic triggers, written or understood.

When seismic improvement is triggered, the most common minimum requirement is 
life safety consistent with the overall code intent. However, the use of 
performance-based design concepts to establish equivalent technical criteria is 
a recent development and is not yet universally accepted. As indicated in the 
last section, the initial response to establishing minimum seismic criteria was 
to use the framework of the code provisions for new buildings with economic and 
technical adjustments as required. These adjustments included a lower lateral 
force level (a pragmatic response to the difficulties of retrofit), and special 
consideration for materials and systems not allowed by the provisions for new 
buildings. (Unreinforced masonry, for example, was not only prohibited as a 
structural system in zones of high seismicity, but also could not be used in a 
building at all.) Use of a lateral force level of 75% of that required for new 
buildings became fairly standard, but the treatment of archaic materials is 
highly variable from jurisdiction to jurisdiction.
Many local retrofit provisions are gradually being replaced by national 
guidelines and standards for seismic evaluation and retrofit (e.g. ASCE 31, 
2003; FEMA 356, 2000, etc.). In addition, performance based seismic design is 
enabling a more direct approach to meeting a community’s minimum performance 
standards—although this requires the policy-makers to decide what the minimum 
performance standard should be, a difficult task that crosses social, economic, 
and technical boundaries.
In summary, both the passive triggers for seismic retrofit and the design or 
performance criteria are often ill-defined and, at best, highly variable between 
jurisdictions. Design professionals should always determine the governing local, 
state, or federal regulations or policies when designing alterations or remodels 
on existing buildings.  
  Active Code Provisions 
Active code provisions result from policy decisions of a jurisdiction to reduce 
the community seismic risk by requiring seismic upgrading of certain buildings 
known to be particularly vulnerable to unacceptable damage. For the most part, 
these provisions are unfunded mandates, although low-interest loan programs have 
been developed in some cases. These risk reduction programs usually allow owners 
a lengthy period to perform the retrofit or to demolish the buildings—ten years 
or more.  The standard for retrofit is also normally included in the law or 
regulation and is typically prescriptive, although performance-based design 
options are becoming more acceptable.
Two large-scale examples of active seismic code provisions were started by the 
state of California. The first was a program to reduce the risk from URM 
buildings. The state legislature, lacking the votes to simply require mitigation 
throughout the state, instead passed a law (SB 547-1986) that required local 
jurisdictions to develop inventories of these buildings in their area, to notify 
the owners that their building was considered hazardous, and to develop a 
community-wide hazard reduction plan.  Although not required to do so, most 
jurisdictions chose as their hazard reduction plan to pass active code 
ordinances giving owners of the buildings ten or so years to retrofit them. Over 
10,000 URM buildings have been brought into compliance with these local 
ordinances, most by retrofit, but some by demolition (SSC, 2003).
The second program, created by SB 1953 in 1994 following the Northridge 
earthquake, gave California hospital owners until 2030 to upgrade or replace 
their hospitals to comply with state law governing new hospital buildings. The 
program’s intention is to enable buildings to be functional following an 
earthquake. This law affected over 500 hospitals and over 2,000 buildings 
(Holmes, 2002). Although compliance is ongoing, this law has been problematic 
due to the high cost and disruption associated with retrofitting hospital 
buildings, and the highly variable economic condition of the health system as 
well as individual facilities.
Other examples include local ordinances to retrofit tilt-up buildings, less 
controversial because of the clear high vulnerability and low retrofit cost of 
these buildings. Similar to investigating local regulations regarding triggers, 
it is also wise to determine if any existing building planned for alterations is 
covered by (or will be covered in the foreseeable future) a requirement to 
retrofit. It is generally acknowledged that seismic improvements are easier to 
implement when done in association with other work on the building.
  Post-Earthquake Code Provisions
Following a damaging earthquake, many buildings may be closed pending 
determination of safety and necessary repairs. A lack of clear repair standards 
and criteria for re-occupancy has created controversy and denied owners use of 
their buildings after most damaging earthquakes. Assuming that the earthquake 
itself is the ultimate judge of seismic acceptability, many communities may take 
the opportunity in the post-earthquake period to require strengthening of 
buildings that are apparently seismically deficient due to their damage level. 
However, implementation of this theory incorporating conservative policies that 
require many retrofits may delay the economic recovery of the community. On the 
other hand, standards for repair and/or strengthening which are not conservative 
could lead to equal or worse damage in the next earthquake. It has also been 
observed that owners of historic, rent-controlled, or otherwise economically 
controlled buildings may have an incentive to demolish damaged buildings to the 
detriment of the community at large.
Traditionally, communities (building departments) have used color codes for 
several or all of the following categories of buildings following an earthquake:
A.   Undamaged; no action required. If inspected at all, these buildings will be 
Green-tagged.
B.   Damaged to a slight extent that will only require repair of the damage to 
the original condition. These buildings will generally be Green-tagged, but the 
category could also include some Yellow-tags.
C.   Damaged to a greater extent that suggests such seismic weaknesses in the 
building that the overall building should be checked for compliance with minimum 
seismic standards. This will often require overall retrofit of the building. 
These buildings will generally be Red-tagged, but the category could also 
include some Yellow tags.
C1. (A subcategory of C). Damaged to an extent that the building creates a 
public risk that requires immediate mitigation, either temporary shoring or 
demolition. The ultimate disposition of these buildings may not be determined 
for several months.  These buildings will all be Red-tagged.
The most significant categorization is the differentiation between B and C. The 
difference to an owner between being placed in one category or the other could 
be an expense on the order of 30%-50% of the value of the building, reflecting 
the added cost of retrofit to that of repair.  Earthquakes being rare, few 
communities have been forced to create these policies, but a few have. Oakland, 
California, prior to the Loma Prieta earthquake, set a trigger based on the loss 
of capacity caused by the damage. If the damage was determined to have caused a 
loss of over 10% of lateral force capacity, then retrofit was triggered.  Los 
Angeles and other southern California communities affected by the Northridge 
earthquake used a similar standard, but the 10% loss was applied to lines of 
seismic resistance rather than the building as a whole. These code regulations, 
although definitive, are problematic because of the technical difficulty of 
determining loss of capacity, particularly to the accuracy of 1% (a 1% change 
can trigger a retrofit).
The importance of this issue has been magnified by interpretation of federal 
laws that creates a tie between reimbursement of the cost of repair of certain 
local damage to the pre-existence and nature of these local damage triggers. 
Owners and designers of older existing buildings should be aware of such 
triggers that could affect them should they suffer damage from an earthquake. In 
some cases, it may be prudent for an owner to voluntarily retrofit a vulnerable 
building to avoid the possibility of being forced to do it in a post-earthquake 
environment as well as to possibly avoiding a long closure of the building.
8.3	 THE FEMA PROGRAM TO REDUCE THE  SEISMIC RISK FROM EXISTING BUILDINGS
In 1985, the Federal Emergency Management Agency (FEMA) recognized that the 
principal seismic risk in this country came from the existing building stock, 
the majority of which was designed without adequate seismic provisions. 
Following a national workshop that identified significant issues and potential 
educational and guideline projects that FEMA could lead, a program was launched 
that is still ongoing.  In addition to providing education and technical 
guidelines in the area of high-risk existing buildings, other FEMA programs were 
also significant in enabling communities to understand and mitigate their 
seismic risk, most notably the development of the regional loss-estimating 
computer program, HAZUS.  Most of these activities are documented as part of the 
FEMA “yellow book” series (so known because of its distinctive yellow covers), 
well known to engineers in this country and, in fact, around the world. 
Unfortunately, these documents are less known to architects, although many of 
them contain useful insights into not only the issues surrounding seismic 
evaluation and retrofit of existing buildings, but also into all aspects of 
seismic design. 
8.3.1	FEMA-Sponsored Activity for Existing Buildings
Following is a summary of selected FEMA-sponsored projects beginning in the late 
1980s.  A full listing is given in FEMA 315, Seismic Rehabilitation of 
Buildings, Strategic Plan 2005.
	Rapid Visual Screening
	FEMA 154: Rapid Visual Screening of Buildings for Potential Seismic 
Hazards, 1988, updated 2001
	A method to enable an efficient first sorting of selected buildings into 
an adequately life-safe group and a second group that will require further 
evaluation. The evaluation was intended to be performed on the street in an hour 
or less per building. The first task is to assign the building to a predefined 
model building type and then identify additional characteristics that could 
refine the seismic vulnerability. The method has proven useful to efficiently 
generate an approximate mix of buildings that will properly characterize a 
community’s vulnerability, but not to definitely rate individual buildings, due 
to the difficulty of identifying significant features from the street. Generally 
it is necessary to obtain access to the interior of a building, or, more 
commonly, it is even necessary to review drawings to confidently eliminate older 
buildings as potentially hazardous.
	Evaluation of Existing Buildings
	FEMA 178: NEHRP Handbook for Seismic Evaluation of Existing Buildings, 
1989 and FEMA 310: Handbook for the Seismic Rehabilitation of Buildings: a 
Prestandard.
	A widely used guide to determine if individual buildings meet a nationally 
accepted level of seismic life safety. This method requires engineering 
calculations and is essentially prescriptive, which facilitates consistency and 
enables enforceability. This life-safety standard was adopted by, among others, 
the federal government and the state of California in certain programs. The 
prescriptive bar may, however, have been set too high, because very few older 
buildings pass.  Since its original development, FEMA 178 has been refined and 
republished as FEMA 310, and finally was adopted as a Standard by the American 
Society of Civil Engineers as ASCE 31 in 2003.
	Techniques Used in Seismic Retrofit
	FEMA 172: NEHRP Handbook of Techniques for Seismic Rehabilitation of 
Existing Buildings, 1992
	In recognition of the lack of experience in seismic upgrading of buildings 
of most of the country’s engineers and architects, this document outlined the 
basic methods of seismically strengthening a building, including conceptual 
details of typically added structural elements. The material recognizes the FEMA 
model building types, but is primarily organized around strengthening of 
structural components, perhaps making the material less directly accessible.The 
document also preceded by several years the publication of the analytical tools 
to design retrofits (FEMA 273, see below). For whatever reason, the publication 
went relatively unused, despite the fact that it contains useful information, 
particularly for architects unfamiliar with seismic issues.
	Financial Incentives
	FEMA 198: Financial Incentives for Seismic Rehabilitation of Hazardous 
Buildings, 1990
	To encourage voluntary seismic upgrading, this document described the 
financial incentives to do so, ranging from tax benefits to damage avoidance.
	Development of Benefit-Cost Model
	FEMA 227: A Benefit-Cost Model for the Seismic Rehabilitation of 
Buildings, 1992
	Due to the expected high cost of seismic rehabilitation, the need to 
provide a method to calculate the benefit-cost ratio of seismic retrofit was 
identified early in the program. This requires estimation of financial losses 
from earthquake damage resulting from a full range of ground-shaking intensity.  
Financial losses include direct damage to structural and nonstructural systems 
as well as business interruption costs.  A controversial feature was the 
optional inclusion of the value of lives lost in the overall equation. The 
project was primarily to develop the model rather than to provide new research 
into expected damage or casualty rates. Thus, approximate relationships 
available at the time were used.  However, the documentation concerning 
contributing factors to a benefit-cost analysis is quite complete, and a 
computerized functional spreadsheet version of the method was developed. 
	Although the use of benefit-cost analysis never became popular in the 
private sector, trials of the method indicated that very low retrofit costs, 
high business-interruption losses, or high exposure-to-casualty losses are 
required to result in a positive benefit-cost ratio. For example rehabilitation 
of tilt-up buildings is usually fairly inexpensive and usually proves cost 
effective.  Similarly, buildings with high importance to a business or with high 
occupancy in areas of high seismicity also result in positive results. Despite 
the apparent overall results from this program, considerable rehabilitation 
activity continued, both in conditions expected to yield a positive benefit-cost 
and in other conditions (in many cases due to extreme importance given to life 
safety).
	Typical Costs of Seismic Rehabilitation
	FEMA 156: Typical Costs for Seismic Rehabilitation of Existing Buildings, 
1988, 2nd edition, 1994.
	The second edition of this document collected case histories of 
constructed rehabilitation and completed reports judged to have realistic costs. 
A database was created to separate costs by primary influence factors: model 
building types, rehabilitation performance objectives, and seismicity. A serious 
difficulty in collection of accurate data was the inevitable mixing of pure 
rehabilitation costs and associated costs such as life-safety upgrades, the 
American Disabilities Act (ADA), and even remodels. Although a large amount of 
data was collected, there was not nearly enough to populate all combinations of 
the factors. Nevertheless, a method was developed to use the data to make 
estimates of costs for given situations. The major problem was that the 
coefficient of variation of rehabilitation costs, for any given situation, is 
very high due to high variability in the extent of seismic deficiencies. The 
information collected is probably most useful to estimate costs for large 
numbers of similar buildings where variations will average out. Use of the 
method to accurately estimate the cost of a single building is not recommended, 
although even the ranges given could be useful for architects and engineers not 
familiar with retrofit issues.
	Technical Guidelines for Seismic Rehabilitation
	FEMA 273: NEHRP Guidelines for the Seismic Rehabilitation of Buildings, 
1997 and FEMA 356 Prestandard and Commentary for the Seismic Rehabilitation of 
Buildings.
	This document, developed over five years by over 70 experts, was the 
culmination of the original program. Previously, the most common complaint from 
engineers and building officials was the lack of criteria for seismic retrofit. 
FEMA 273 incorporated performance-based engineering, state-of-the-art nonlinear 
analysis techniques, and an extensive commentary to make a significant 
contribution to earthquake engineering and to focus laboratory research on 
development of missing data. The document broke away from traditional code 
methods and in doing so, faced problems of inconsistency with the design of new 
buildings. Improvements were made in a follow-up document, FEMA 356, but the 
practical results from use of the method indicate that considerable judgment is 
needed in application. Work is continuing to improve analysis methods and in 
methods to predict the damage level to both components of a building and to the 
building as a whole. Even given these difficulties, the document has become the 
standard of the industry.
	Development of a Standardized Regional Loss Estimation Methodology—HAZUS
	National Institute of Building Sciences, Earthquake Loss Estimation 
Methodology, Technical Manual (for latest release).
	A good summary of this development is contained in a paper by Whitman, et 
al. in EERI Earthquake Spectra, vol 13, no 4. FEMA also maintains a HAZUS 
website, www.fema.gov/hazus.  
	The development of a standardized regional loss estimation methodology was 
not in the original FEMA plan to reduce risks in existing buildings. However, 
this development has had a major impact in educating local officials about their 
seismic risks and estimating the level of risk around the country in a standard 
and comparable way.
	In 1990, when the development of HAZUS began, the primary goals were to 
raise awareness of potential local earthquake risks, to provide local emergency 
responders with reasonable descriptions of post-earthquake conditions for 
planning purposes, and to provide consistently created loss estimates in various 
regions to allow valid comparison and analysis. The loss estimation methodology 
was intended to be comprehensive and cover not only building losses, but also 
damage to transportation systems, ports, utilities, and critical facilities. A 
technically defensible methodology was the goal, not necessarily an all-
encompassing software package. When it became obvious that the methodology was 
far more useful and could be more consistently applied as software, HAZUS was 
born. The program uses census data and other available physical and economic 
databases to develop, on a first level of accuracy, a model of local conditions. 
Expected or speculated seismic events can be run and losses estimated. Losses 
include direct damage, business interruption, and casualties, as well as loss of 
utilities, loss of housing units, and many other parameters of use to emergency 
planners. The building inventory uses the FEMA model building types and an 
analysis method closely tied to FEMA 273, linking HAZUS to other FEMA-sponsored 
work regarding existing buildings.
	Subsequent to the original development activity, HAZUS was expanded to 
create loss estimates for wind and flood.  
	Incremental Rehabilitation
	FEMA 395, Incremental Seismic Rehabilitation of School Buildings (K-12), 
2003.
	This is the first in a series of manuals that FEMA (U.S. Department of 
Homeland Security) intends to develop for various occupancy types including, for 
example, schools, hospitals, and office buildings.  The concept is based on the 
fact that seismic strengthening activities are more efficiently accomplished in 
conjunction with other work on the building, and such opportunities should be 
identified and exploited even if only part of a complete rehabilitation is 
accomplished. This is perhaps most applicable to K-12 school buildings because 
of their relatively small size and ongoing maintenance programs.  FEMA model 
building types are again used to categorize potential opportunities in different 
conditions. As is pointed out in the manual, this technique has to be applied 
with care to avoid an intermediate structural condition that is worse than the 
original.

8.3.2	The FEMA Model Building Types
Most of these developments were part of the integrated plan developed in 1985.  
As such, FEMA coordinated the projects and required common terminology and 
cross-references.
The most successful and virtually standard-setting effort was the creation of a 
set of model building types to be used for the characterization of existing 
buildings. The model building types are based primarily on structural systems 
rather than occupancy, but have proven extremely useful in the overall program.  
Model building types are defined by a combination of the gravity-load carrying 
system and the lateral-load carrying system of the building. Not every building 
type ever built in the country and certainly not the world is represented, but 
the significant ones are, and the relative risks of a community can well be 
represented by separating the local inventory into these types. Of course, there 
was no attempt to represent every “modern” building type because they are not 
considered hazardous buildings.  However, with minor sub-categorization that has 
occurred with successive documents, the majority of buildings, new or old, now 
can be assigned a model building type.  The test of the usefulness came with the 
successful development of HAZUS using the model building type because this 
program needed a reasonably simple method to characterize the seismic 
vulnerability of inventories of buildings across the country.
Currently, no single FEMA document contains a graphic and clear description of 
the model building types, although engineers can generally determine the correct 
category.  Because of the ubiquitous FEMA-developed documents, guidelines, and 
standards regarding existing buildings, and their common use by engineers, such 
descriptions are included here to facilitate communication with architects. The 
types are illustrated on pages 8-23 through 8-31. Table 8-3, at the end of the 
chapter, presents a summary of the performance charactoristics and commom 
rehabilitation techniques.  
8.4 SEISMIC EVALUATION OF EXISTING BUILDINGS
Not all older buildings are seismically at risk. If they were, the damage from 
several earthquakes in this country, including the 1971 San Fernando and the 
1984 Northridge events, would have been devastating, because much of the 
inventory affected was twenty or more years old. Often, strong ground shaking 
from earthquakes significantly damages building types and configurations well 
known to be vulnerable, and occasionally highlights vulnerabilities previously 
unrealized. For example, the Northridge earthquake caused damage to many wood-
frame buildings—mostly apartments—and relatively modern steel moment-frame 
buildings, both previously considered to be of low vulnerability. It is natural 
to catalogue damage after an earthquake by buildings with common 
characteristics, the most obvious characteristic being the construction type, 
and the secondary characteristic being the configuration. Both of these 
parameters are central to processes developed to identify buildings especially 
vulnerable to damage before the earthquake. In fact, the categorization of 
damage by building type is primarily what led to the development of the FEMA 
Model Building Types discussed in Section 8.3.2.
However, only in the most vulnerable building types does damage occur relatively 
consistently.  For example, at higher levels of shaking, the exterior walls of 
unreinforced masonry bearing-wall buildings have relatively consistently fallen 
away from their buildings in many earthquakes, ever since this building type was 
built in large numbers in the late 19th century. More recently, a high 
percentage of “pre-Northridge” steel moment-frame buildings have received damage 
to their beam-column connections when subjected to strong shaking.  Even in 
these cases, the damage is not 100% consistent and certainly not 100% 
predictable.  In building types with less vulnerability, the damage has an even 
higher coefficient of variation.  Engineers and policymakers, therefore, have 
struggled with methods to reliably evaluate existing buildings for their seismic 
vulnerability.
As discussed in Section 8.2, the initial engineering response was to judge older 
buildings by their capacity to meet the code for new buildings, but it became 
quickly apparent that this method was overly conservative, because almost every 
building older than one or two code-change cycles would not comply—and thus be 
considered deficient.  Even when lower lateral force levels were used, and the 
presence of archaic material was not, in itself, considered a deficiency, many 
more buildings were found deficient than was evidenced in serious earthquake 
damage.  Thus, policymakers have generally been successful in passing active 
retrofit provisions (see Section 8.2.3) only in the most vulnerable buildings, 
such as URM and tilt-ups, where damage has been significant and consistent, and 
individual building evaluation is not particularly significant.
The evaluation of existing buildings typically starts with identification of the 
building type and damaging characteristics of configuration (e.g., soft story).  
This can be done rapidly and inexpensively but, except for a few vulnerable 
building types, is unreliable when taken to the individual building level.  
Engineers and code writers have also developed intermediate levels of evaluation 
in which more characteristics are identified and evaluated, many by calculation.  
In the last decade, more sophisticated methods of analysis and evaluation have 
been developed that consider the nonlinear response of most structures to 
earthquakes and very detailed material and configuration properties that will 
vary from building to building.
 8.4.1 Expected Performance by Building Type
As previously mentioned, damage levels after earthquakes are collected and 
generally assigned to bins of common characteristics, most commonly the level of 
shaking, building material and type, and configuration.  Combined with numerical 
lateral-force analysis of prototype buildings, this information can be analyzed 
statistically.  The three primary parameters -  building type, shaking level, 
and damage level - are often displayed together in a damage probability matrix 
similar to Table 8-1.  The variability of damage is such that for any shaking 
level, as shown in the columns, there is normally a probability that some 
buildings will be in each damage state.  The probabilities in these tables can 
be interpreted as the percentage of a large number of buildings expected to be 
in each damage state, or the chances, given the shaking level, that an 
individual building of this type with be damaged to each level.
Statistical information such as this is used in several ways:
	Identification of clearly vulnerable or dangerous buildings to help 
establish policies of mitigation
Many extremely vulnerable building types or components can be identified by 
observation without statistical analysis, including URM, soft-story “tuck-under” 
apartment buildings, the roof-to-wall connection in tilt-up buildings, 
residences with cripple wall first-floor construction, and connections of pre-
Northridge steel moment-frames.  The clearly and more consistently dangerous 
building types have often generated enough community concern to cause the 
creation of policies to mitigate the risks with retrofit.  For a combination of 
reasons, URMs and tilt-ups currently are the targets of the most active 
mitigation policies.
	Earthquake Loss Estimation
Regional earthquake loss estimates have been performed for forty or more years 
to raise awareness in the community about the risks from earthquakes and to 
facilitate emergency planning.  Given an approximate distribution of the 
building inventory and a map of estimated ground motion from a given earthquake, 
damage-probability matrices (or similar data) can be used to estimate damage 
levels to the building stock.  From the damage levels, economic loss, potential 
casualties, and business interruption in a community can be estimated.
Starting in 1991, FEMA began a major program to develop a standard way of 
performing such loss estimations to facilitate comparative loss estimates in 
various parts of the country.  This program resulted in a computer program, 
HAZUS, described briefly in Section 8.3.1.
	Formal Economic Loss Evaluations (e.g. Probable Maximum Loss or PML)
Since consensus loss relationships became available (ATC, 1985), a demand has 
grown to include an estimate of seismic loss in “due-diligence” studies done for 
purchase of buildings, for obtaining loans for purchase or refinance, or for 
insurance purposes.  An economic loss parameter, called Probable Maximum Loss, 
has become the standard measuring stick for these purposes.  The PML for a 
building is the pessimistic loss (the loss suffered by the worst 10% of similar 
buildings) for the worst shaking expected at the site (which gradually became 
defined as the shaking with a 500-year return period, similar to the code design 
event).  Although a detailed analysis can be performed to obtain a PML, most are 
established by building type and a few observable building characteristics.  
Because of the high variability in damage and the relatively incomplete 
statistics available, PMLs are not very reliable, particularly for an individual 
building.
	Rapid Evaluation
As foreseen by FEMA’s original plan for the mitigation of risks from existing 
buildings, a rapid evaluation technique should be available to quickly sort the 
buildings into three categories: obviously hazardous, obviously acceptable, and 
uncertain.  The intent was to spend less than two hours per building for this 
rapid evaluation.  Under the plan, the uncertain group would then be evaluated 
by more detailed methods.  The results of FEMA’s development efforts, FEMA 154 
(Section 8.3.1) is fairly sophisticated but, because of the large amount of 
unknown building data that is inherent in the system, for an individual 
building, is unreliable.  The sorting method is probably quite good for 
estimating the overall vulnerability of a community because of the averaging 
effect when estimating the risk of many buildings.
8.4.2 Evaluation of Individual Buildings
Engineers have been seismically evaluating existing buildings for many years, 
whether by comparing the conditions with those required by the code for new 
buildings, by using some local or building-specific standard (e.g. URMs), or by 
using their own judgment.  These methods are still used, as well as very 
sophisticated proprietary methods developed within private offices, but the 
majority of evaluations are now tied in some way to the general procedures of 
ASCE 31-03, Seismic Evaluation of Existing Buildings (ASCE, 2003), that in 2003 
became a national standard.  There are three levels of evaluation in the 
standard called tiers, which, not accidentally, are similar to the standard of 
practice prior to the standardization process.  These levels of evaluation are 
briefly described below, as well as similar methods that fall in the same 
categories.
However, before beginning a seismic evaluation, particularly of a group of 
buildings, it is logical to assume that buildings built to modern codes must 
meet some acceptable standard of life safety.  Due to the large number of older 
buildings, the effort to eliminate some from consideration resulted in several 
well-known milestone years.  First it was compliance with the 1973 Uniform 
Building Code (UBC) or equivalent.  After study and reconsideration of 
relatively major changes made in the 1976 UBC, this code was used as a 
milestone.  Primarily caused by life-threatening damage to various building 
types in the 1994 Northridge earthquake and subsequent code changes, a 
relatively complex set of milestone years was developed.  ASCE 31 contains such 
a set of code milestone years for each of the several codes used in this country 
over the last thirty years.  Although ASCE 31 suggests that compliance with 
these codes is only a recommended cut-off to not require evaluations for life 
safety, in all but very unusual situations, the table can be accepted.  This 
table, Table 3-1 of ASCE 3,1 is reproduced as Figure 8-7.
	Initial Evaluation (ASCE 31 Tier 1)
The ASCE Tier 1 evaluation is similar to FEMA’s Rapid Evaluation in that it is 
based on the model building type and certain characteristics of the building.  
The significant difference is that structural drawings, or data equivalent to 
structural drawings, are required to complete the evaluation, and the evaluation 
will take several days rather than several hours.  After identifying the 
appropriate FEMA Building Type, a series of prescriptive requirements are 
investigated, most of which do not require calculations.
If the building is found to be noncompliant with any requirement, it is 
potentially seismically deficient.  After completing the investigation of a 
rather exhaustive set of requirements, the engineer reviews the list of 
requirements with which the building does not comply, and decides if the 
building should be categorized as noncompliant or deficient.  A conservative 
interpretation of the method is that any single noncompliance is sufficient to 
fail the building, but most engineers exercise their judgement in cases of 
noncompliance with only a few requirements. Historically, this method has 
developed with the pass/fail criterion of life safety, but the final ASCE 
Standard includes criteria for both Life Safety and Immediate Occupancy, a 
performance more closely related to continued use of the building.  Because of 
the importance associated with the Immediate Occupancy performance level, a 
building cannot pass these requirements with only a Tier 1 analysis.
	Intermediate Evaluation (ASCE 31 Tier 2)
The ASCE intermediate level of evaluation, called Tier 2, is similar in level of 
effort of historical nonstandardized methods.  Normally, an analysis of the 
whole building is performed and the equivalents of stress checks are made on 
important lateral force-resisting components.  This analysis is done in the 
context and organization of the set of requirements used in Tier 1, but the 
process is not unlike seismic analysis traditionally performed for both 
evaluation and design of new buildings.  ASCE 31 includes the requirements for 
both the LifeSafety and Immediate Occupancy performance levels for a Tier 2 
Evaluation.
   Detailed Evaluation (ASCE 31 Tier 3)
The most detailed evaluations are somewhat undefined because there is no ceiling 
on sophistication or level of effort.  The most common method used in Tier 3 is 
a performance evaluation using FEMA 356, Prestandard and Commentary for the 
Seismic Rehabilitation of Buildings (FEMA, 2000), based on simplified nonlinear 
analysis using pushover analysis (see Chapter 6).  This method approximates the 
maximum lateral deformation that the building will suffer in a design event, 
considering the nonlinear behavior created by yielding and damage to components.  
The level of deformation of individual components is compared with standard 
deformations preset to performance levels of Collapse Prevention, Life Safety, 
and Immediate Occupancy, with the probable damage state of the building as a 
whole set at that level of the worst component.  Efforts are being made to more 
realistically relate the damage states of all the components to a global damage 
state.  This method can be used either to determine the probable damage state of 
the building for evaluation purposes, or to check the ability of a retrofit 
scheme to meet a target level.
With the advancement of computer capability and analysis software, nonlinear 
analysis techniques are constantly being improved.  The ultimate goal, although 
not expected to be an everyday tool in the near future, is to simulate the 
movements of the full buildings during an entire earthquake, including the 
constantly changing properties of the structural components due to yielding and 
damage.  The overall damage to various components is then accumulated and the 
global damage state thereby surmised.
8.4.3 Other Evaluation Issues
There are several other issues associated with seismic evaluation that should be 
recognized.  Only three will be discussed here.  First is the data required to 
perform competent evaluations at the various levels, as discussed above.  
Second, it is important to understand the performance expectation of the pass-
fail line for various evaluation methodologies.  Last, the reliability (or lack 
thereof) of the methods and of evaluation and/or performance prediction in 
general,should be recognized.
  Data Required for Seismic Evaluation
Obviously, for methods depending on the FEMA building type, the building type 
must be known.  In fact, there are other similar classifications of building 
types also used to define building performance at the broadest level.  Using 
data as discussed in paragraph 8.4.1, crude expectations of performance and 
therefore comparative evaluation can be completed.  Most such systems, however, 
are refined by age, physical condition of the building, configuration, and other 
more detailed data, when available.  Most “rapid” evaluation methods, based on 
building type and very basic building characteristics, do not require structural 
drawings.  Responsible evaluators will insist on a site visit (in many cases to 
make sure the building is still there, if nothing else).
The more standardized evaluation methods discussed in paragraph 8.4.2 
essentially require drawings.  If detailed structural drawings are not 
available, simple evaluations of some model building types (wood buildings, 
tilt-ups, and sometimes URM) can be performed based on layout drawings or from 
data prepared from field visits.  However, when reinforced concrete, reinforced 
masonry, or structural steel is a significant part of the structure, it is most 
often economically infeasible to reproduce “as-built” drawings.  Practically in 
those cases, with rare exceptions, the building is deduced to be in 
nonconformance and, as a retrofit, a new seismic system is introduced to render 
the unknowns of the existing structure insignificant.  Even in those cases, 
however, extensive field work is necessary to produce enough structural data to 
create a reasonable set of construction documents.
If original structural drawings are available that are confirmed to be 
reasonably accurate from spot checks in the field, most evaluation techniques 
can be employed.  However, material properties are often not included on the 
drawings and must be deduced from the era of construction.  Deterioration can 
also affect the material properties of several building types.  Often the 
potential variability in the analysis due to different possible combinations of 
material properties requires in-situ testing of material properties.  The 
techniques for this testing are well established, but cost and disruption to 
tenants are often an issue.
As explained in Chapter 3, “Site Evaluation and Selection”, many areas of the 
country are mapped in detail for seismic parameters related to design, although 
such parameters continue to be investigated and updated.  When warranted, site-
specific studies can be performed to obtain timely and locally derived data.  
However, other seismic site hazards, such as liquefaction, landslide, and 
potential surface fault rupture, are less well mapped and may require a site-
specific study, if there is reason to suspect their potential at a site.  
Perhaps a less obvious important characteristic of a site is the detail of 
adjacent structures.  Particularly in urban settings, adjacent buildings often 
have inadequate separation or are even connected to the building to be 
evaluated.  Although legal issues abound when trying to deal with this issue, it 
is unrealistic to analyze and evaluate such a building as if it were 
freestanding.  Formal evaluation techniques, such as ASCE 31, have addressed 
this issue, at least for buildings that are not connected, by highlighting the 
conditions known to potentially produce significant damage.  First, if floors do 
not align between adjacent buildings and pounding is expected, the stiff floor 
from one building could cause a bearing wall or column in the adjacent building 
to collapse.  Secondly, if buildings are of significantly different height, the 
interaction from pounding has been observed to cause damage.  See Figure 8-8.
  Performance Objectives and Acceptability
Traditionally, evaluation techniques have been targeted at determining if a 
building is adequately life safe in an earthquake, similar to code goals for new 
buildings.  However, as discussed in section 8.2.2, a standard different and 
less than that used for new buildings evolved, but was still termed life safety. 
Only with the development of performance-based engineering did evaluation 
methods aimed at other performance standards emerge.  Even life safety has 
proven to be amorphous over the years and often has been defined by the 
evaluation technique du jour.  Seismically, life safety is a difficult concept, 
due to the huge potential variation in ground motion and the many sources of 
damage that could cause injury or death.  However, the term is well embedded in 
public policy and continues to persist in seismic codes and standards.
FEMA, in sponsoring the development of FEMA 273 (and later FEMA 356), wanted a 
more specific definition of a suitable goal for seismic safety, and thus the 
Basic Safety Objective (BSO) was defined.  This performance objective consists 
of two requirements: the building would provide life safety for the standardized 
code event and, in addition, the building would not collapse in the Maximum 
Considered Event (MCE), a very rare event now defined by code.  Since these FEMA 
documents are non-mandatory (unless locally adopted), the BSO has not become a 
widely accepted standard (the BSO also includes mandatory nonstructural minimum 
requirements, which also may delay its wide acceptance).
Chapter 6 contains a detailed discussion of performance-based engineering, which 
is gaining acceptance for evaluations at any level. But performance 
characterization in various forms has been used for some time, primarily to set 
policy.  Such policies require descriptions of various performance levels, even 
if the technical ability to define or predict the various levels often lagged 
behind.  Table 8-2 shows several such performance descriptions, many developed 
decades ago, which have been used to set policy—most concentrating on life 
safety.  The table is set up to approximately equilibrate levels of performance 
across horizontal lines.
The first columns in Table 8-2 describe a system used by the University of 
California.  GOOD, the best performance, is defined as the equivalent life 
safety as that provided by the code for new buildings—but without consideration 
of monetary damage.  The next level below was set at the acceptable level for 
evaluation, while retrofits are required to meet the GOOD level.
The next column, labeled “DSA”, is a Roman numeral system developed by the 
California Division of the State Architect for use with state-owned buildings.  
Each level has a description of damage and potential results of damage 
(“building not reocccupied for months”) but no reference to engineering 
parameters.  The state used an acceptance level of IV, but set the goal for 
retrofits to III.
The levels described in the next columns come from one of the early developments 
of performance based earthquake engineering, Vision 2000, developed by the 
Structural Engineers Association of California.  It is a relatively 
comprehensive scale using five primary descriptions of damage, each with a 
“plus” and “minus”, resulting in ten levels.  
Finally, to indicate a perhaps more commonly recognized standard of performance, 
are the three occupancy tagging levels of Red, Yellow, and Green used for 
emergency evaluation immediately after damaging earthquakes.
  Reliability of Seismic Evaluations
The most significant characteristic in the design of buildings for earthquakes 
is the variability of ground motions.  Not only do magnitudes and locations 
vary, but also the effects of fault rupture, wave path, and local site soils 
create a literally infinite set of possible time histories of motion.  Studies 
have shown that time histories within a common family of parameters used for 
design (response spectrum) can produce significantly different responses.  This 
variation normally dominates over the scatter of results from analysis or 
evaluation techniques.
However, codes for new buildings can require many limitations of material, 
lateral system, configuration, and height that will reasonably assure acceptable 
performance, particularly the prevention of collapse.  These same limitations 
can seldom be applied to existing buildings, so the variation of actual 
performance is expected to be much larger.  In addition, the cost of retrofit is 
often high, and attempts have been made to avoid unnecessary conservatism in 
evaluation methodologies.  It is probable, therefore, that a significant number 
of buildings may fail to perform as evaluated, perhaps in the range of 10% or 
more.  No comprehensive study has been made to determine this reliability, but 
ongoing programs to further develop performance-based seismic engineering are 
expected to estimate the variability of evaluation results and refine the 
methods accordingly.
8.5 SEISMIC REHABILITATION OF EXISTING BUILDINGS
There are many reasons why buildings might be seismically retrofitted, including 
renovations that trigger a mandatory upgrade, a building subjected to a 
retroactive ordinance, or the owner simply wanting (or needing) improved 
performance.  The reason for the upgrade may influence the technique and 
thoroughness of the work, because owners faced with mandatory upgrade may seek 
out the least expensive, but approvable, solution, whereas an owner needing 
better performance will more likely be willing to invest more for a solution 
that addresses their particular concerns.  There are many other factors that 
shape a retrofit solution, such as the type of deficiency present, if the 
building is occupied, and the future use and aesthetic character of the 
building.
The continuing improvement of analysis techniques and the emergence of 
performance-based design are also having a large effect on retrofit schemes, by 
enabling engineers to refine their designs to address the specific deficiencies 
at the desired performance level.  In many cases, however, the retrofits are 
becoming controlled by the brittleness of existing components that must be 
protected from excess deformation with systems that may be stronger and/or 
stiffer than those used for new buildings.  Some older retrofits, done to 
prescriptive standards or using now-outdated strengthening elements borrowed 
from new building designs, may themselves be deficient, depending on the desired 
performance.  Seismic retrofit analysis, techniques, and components, similar to 
new building technology, are not static, and applications should be regularly 
reviewed for continued effectiveness.
8.5.1 Categories of Rehabilitation Activity
In most cases, the primary focus for determining a viable retrofit scheme is on 
vertically oriented components (e.g. column, walls, braces, etc.) because of 
their significance in providing either lateral stability or gravity-load 
resistance.  Deficiencies in vertical elements are caused by excessive inter-
story deformations that either create unacceptable force or deformation demands.  
However, depending on the building type, the walls and columns may be adequate 
for seismic and gravity loads, but the building is inadequately tied together, 
still forming a threat for partial or complete collapse in an earthquake.  It is 
imperative to have a thorough understanding of the expected seismic response of 
the existing building, and all of its deficiencies to design an efficient 
retrofit scheme.  There are three basic categories of measures taken to retrofit 
a building:
1)	Modification of global behavior, usually decreasing deformations (drifts);
2)	Modification of local behavior, usually increasing deformation capacity;
3)	Connectivity, consisting of assuring that individual elements do not 
become detached and fall, assuring a complete load path, and assuring that the 
force distributions assumed by the designer can occur.
The types of retrofit measures often balance one another, in that employing more 
of one will mean less of another is needed.  It is obvious that providing added 
global stiffness will require less deformation capacity for local elements (e.g. 
individual columns), but it is often less obvious that careful placement of new 
lateral elements may minimize a connectivity issue such as a diaphragm 
deficiency.  Important connectivity issues such as wall-to-floor ties, however, 
are often independent and must be adequately supplied. 
 Modification of Global Behavior
Modification to global behavior normally focuses on deformation, although when 
designing to prescriptive standards, this may take the form of adding strength. 
Overall seismic deformation demand can be reduced by adding stiffness in the 
form of shear walls or braced frames.  Addition of moment frames is normally 
ineffective in adding stiffness.  New elements may be added or created from a 
composite of new and old components.  Examples of such composites include 
filling in openings of walls and using existing columns for chord members for 
new shear walls or braced frames. 
Particular ground motions have a very specific deformation demand on structures 
with various periods, as discussed in Chapter 4.  Given an equal period of 
vibration, this deformation will occur, whether distributed over the height of 
the building or concentrated at one floor.  If one or more inter-story drifts 
are unacceptable, it may be possible to redistribute stiffness vertically to 
obtain a more even distribution of drift.  A soft or weak story is an extreme 
example of such a problem.  Such stories are usually eliminated by adding 
strength and stiffness in such a way as to more closely balance the stiffness of 
each level, and thus evenly spread the deformation demand over the height of the 
structure.
Seismic isolation is the supreme example of the concept of redistribution of 
deformation.  Essentially all deformation is shifted to bearings, placed at the 
isolation level, that are specifically designed for such response.  The bearings 
limit the response of the superstructure, which can be designed to remain 
essentially undamaged for this maximum load.  The feasibility of providing 
isolation bearings that limit superstructure accelerations to low levels not 
only facilitates design of superstructures to remain nearly elastic, but also 
provides a controlled environment for design of nonstructural systems and 
contents.
Global deformations can also be controlled by the addition of passive energy 
dissipation devices, or dampers, to the structure.  Although effective at 
controlling deformations, large local forces may be generated at the dampers 
that must be transferred from the device to structure and foundation, and the 
disruptive effect of these elements on the interior of the building is no 
different than a rigid brace.
  Modification of Local Behavior
Rather than providing retrofit measures that affect the entire structure, 
deficiencies also can be eliminated at the local component level.  This can be 
done by enhancing the existing shear or moment strength of an element, or simply 
by altering the element in a way that allows additional deformation without 
compromising vertical-load carrying capacity.
Given that in most cases, that certain components of the structure will yield 
(i.e., become inelastic), some yielding sequences are almost always benign: 
beams yielding before columns, bracing members yielding before connections, and 
bending yielding before shear failure in columns and walls.  These relationships 
can be determined by analysis and controlled by local retrofit in a variety of 
ways.  Columns in frames and connections in braces can be strengthened, and the 
shear capacity of columns and walls can be enhanced to be stronger than the 
shear that can be delivered.
Concrete columns can be wrapped with steel, concrete, or other materials to 
provide confinement and shear strength.  Concrete and masonry walls can be 
layered with reinforced concrete, plate steel, and other materials.  Composites 
of glass or carbon fibers and epoxy are becoming popular to enhance shear 
strength and confinement in columns, and to provide strengthening to walls. 
Another method to protect against the collapse risk posed by excess drift is to 
provide a supplementary gravity support system for elements that might be 
unreliable at expected high-deformation levels.  For example, supplementary 
support for concentrated wall-supported loads is a requirement in California 
standards for retrofit of unreinforced masonry buildings.  In several cases, 
supplementary support has also been used in concrete buildings.
Lastly, deformation capacity can be enhanced locally by uncoupling brittle 
elements from the deforming structure, or by removing them completely.  Examples 
of this procedure include placement of vertical saw cuts in unreinforced masonry 
walls to change their behavior from shear failure to a more acceptable rocking 
mode, and to create slots between spandrel beams and columns to prevent the 
column from acting as a “short column” prone to shear failure.
 Connectivity
Connectivity deficiencies are within the load path: wall out-of-plane connection 
to diaphragms; connection of diaphragm to vertical lateral force-resisting 
elements; connection of vertical elements to foundation; connection of 
foundation to soil.  A complete load path of some minimum strength is always 
required, so connectivity deficiencies are usually a matter of degree.  A 
building with a complete but relatively weak or brittle load path might be a 
candidate for retrofit by seismic isolation to simply keep the load below the 
brittle range.
The only location in the connectivity load path at which yielding is generally 
allowed is the foundation/structure interface.  Allowing no movement at this 
location is expensive and often counterproductive, as fixed foundations transfer 
larger seismic demands to the superstructure.  Most recently developed retrofit 
guidelines are attempting to provide simplified guidance to the designer on how 
to deal with this difficult issue and minimize foundation costs.
8.5.2 Conceptual Design of a Retrofit Scheme for an Individual Building
There are many specific methods of intervention available to retrofit designers, 
as previously discussed.  The selection of the specific type of element or 
system is dependent on local cost, availability, and suitability for the 
structure in question.  Any system used to resist lateral load in new buildings 
can also be used for retrofit.  It is thus an extensive task to develop 
guidelines for such selection.  In addition, as in the design of a new building, 
there is usually a choice of where to locate elements, although it is generally 
more restrictive in existing buildings.  However, in the end, there are 
nonseismic issues associated with each building or project that most often 
control the specific scheme to be used.
The solution chosen for retrofit is almost always dictated by building  user-
oriented issues rather than by merely satisfying technical demands.  There are 
five basic issues that are always of concern to building owners or users: 
seismic performance, construction cost, disruption to the building users during 
construction (often translating to a cost), long-term affect on building space 
planning, and aesthetics, including consideration of historic preservation.
All of these characteristics are always considered, but an importance will 
eventually be put on each of them, either consciously or subconsciously, and 
these weighting factors invariably will determine the scheme chosen. 
	Seismic performance
	Prior to the emphasis on performance-based design, perceived qualitative 
differences between the probable performance of difference schemes were used to 
assist in choosing a scheme.  Now, specific performance objectives are often set 
prior to beginning development of schemes.  Objectives that require a very 
limited amount of damage or “continued occupancy” will severely limit the 
retrofit methods that can be used and may control the other four issues.
	Construction cost
	Construction cost is always important and is balanced against one or more 
other considerations deemed significant.  However, sometimes other economic 
considerations, such as the cost of disruption to building users, or the value 
of contents to be seismically protected, can be orders of magnitude larger than 
construction costs, thus lessening its importance.
	Disruption to the building users during construction
	Retrofits are often done at the time of major building remodels, and this 
issue is minimized.  However, in cases where the building is partially or 
completely occupied, this parameter commonly becomes dominant and controls the 
design.
	Long-term effect on building space planning
	This characteristic is often judged less important that the other four and 
is therefore usually sacrificed to satisfy other goals.  In many cases, the 
planning flexibility is only subtly changed.  However, it can be significant in 
building occupancies that need open spaces, such as retail spaces and parking 
garages.
	Aesthetics
	In historic buildings, considerations of preservation of historic fabric 
usually control the design.  In many cases, even performance objectives are 
controlled by guidelines imposed by preservation.  In non historic buildings, 
aesthetics is commonly stated as a criterion, but in the end is often 
sacrificed, particularly in favor of minimizing cost and disruption to tenants.
These parameters can merely be recognized as significant influences on the 
retrofit scheme or can be used formally to compare schemes.  For example, a 
comparison matrix can be developed by scoring alternative schemes in each 
category and then applying a weighting factor deduced from the owner’s needs to 
each category.
Figure 8-9 describes the evolution of a retrofit scheme based on several changes 
in the owner’s weighting of these five characteristics.
8.5.3 Other Rehabilitation Issues
  Inadequate recognition of disruption to occupants
It is unfortunately common for the extent of interior construction and 
disruption to be underestimated.  In many cases, occupants who were originally 
scheduled to remain in place are temporarily moved—at a significant increase in 
cost of the project—or the work is required to be done in off-hours, also a 
premium cost.  Figures 8-10 and 8-11 indicate the level of construction 
intensity often required in retrofit.
Similarly, “exterior solutions,” where strengthening elements are placed on the 
outside of the building are often more disruptive and noisier than anticipated 
and often require collector members to be placed on each floor within the 
building.  Figure 8-12 shows the result of an exterior retrofit of adding towers 
on the outside of a building that, in fact, did not cause a single lost day of 
occupancy.  High-strength steel rods were epoxied into horizontal cores, drilled 
twenty feet into the existing concrete beams to form the needed collectors.
 Collateral required work
As previously mentioned, retrofit work is often performed in conjunction with 
other remodeling or upgrading activities in a building.  Such work normally 
triggers other mandatory improvements to the building, such as ADA compliance or 
life safety updating—all of which add cost to the project.  However, even when 
seismic retrofit is undertaken by itself, the costs of ADA compliance, removal 
of disturbed hazardous material, and possibly life safety upgrades must be 
considered.
8.5.4	Examples
It is impossible to include examples that show the full range of structural 
elements and configurations used in seismic retrofit.  There are definitely 
patterns, usually driven by economics or avoidance of disruption to occupants, 
but depending on the particular mix of owner requirements, as discussed in 
Section 8.5.2, thoughtful architects and engineers will always come up with a 
new solution.
Photographs of retrofit buildings, although often interesting, seldom can tell 
the full story of the development of the scheme, and if the majority of retrofit 
elements are inside or hidden, tell almost nothing.  Some photos are shown here, 
but are not intended to demonstrate the full range of buildings that have 
successfully undergone seismic retrofit or the full range of solutions to 
individual problems.  In addition, due to limited space, only one or two points 
are made with each photo, rather than a full case study.
8.6   SPECIAL ISSUES WITH HISTORIC BUILDINGS
Seismic evaluation and retrofit of historic buildings generate complex public 
policy issues for which few general rules can be identified.  Restoration, or 
renovations of large and important historic buildings usually have considerable 
public and jurisdictional oversight, in addition to employing an experienced 
design team that includes a special historic preservation consultant.  The 
control and oversight for less important buildings that have historic status at 
some level, or that may qualify for such status, are highly variable.  Designers 
are cautioned to locally investigate approval procedures for alterations on such 
buildings as well as seismic requirement, for them.
8.6.1	Special Seismic Considerations
It has been recognized in most areas of high seismicity that local public policy 
concerning seismic retrofit triggers must include special considerations for 
historic buildings.  As discussed in Section 8.2, initial seismic safety 
criteria for existing buildings were focused on requirements for new buildings, 
which were marginally appropriate for most older buildings, but completely 
inappropriate for historic buildings.  Special allowances were therefore created 
for archaic materials that were not allowed in new buildings, and the overall 
seismic upgrade level was lowered to reduce work that could compromise historic 
integrity and fabric.  These kinds of technical criteria issues have been 
somewhat mitigated by the completion of FEMA 356 and the emergence of 
performance-based earthquake engineering, because consideration of archaic 
materials and fine-tuning of performance levels are now part of the normal 
lexicon.
8.6.2	Common Issues of Tradeoffs
Many buildings in this country that qualify for historic status are not 
exceptionally old and can be made commercially viable.  The changes that are 
needed for successful adaptive reuse will often conflict with strict 
preservation guidelines, and compromises are needed in both directions to 
achieve a successful project that, in the end, could save the building from 
continuing decay and make it more accessible to the public.  These tradeoffs 
occur in many areas of design, but seismic upgrading work often requires 
interventions that are not needed for any other reason.  These interventions 
often fall under historic preservation guidelines that call for clear 
differentiation of new structural components, or that discourage recreation of 
historic components that are removed.  As previously indicated, there are no 
rules for these conditions, and the most appropriate solution for each case must 
be determined individually.
Another common conflict is between current preservation of historic fabric and 
future preservation of the building due to the chosen seismic performance level. 
Typically, a better target performance in the future, possibly preventing 
unrecoverable damage, requires more seismic renovation work now.  Most historic 
preservation codes allow lower expected seismic performance to reduce 
construction work and minimize damage. Like many seismic policies, there have 
not been enough earthquakes with seismically damaged historic buildings to test 
this general philosophy.  In an ever-growing number of cases of important 
buildings, this dilemma has been addressed using seismic isolation—which by 
reducing loading to the superstructure, reduces required construction work and 
also reduces expected damage in future earthquakes. Typically, however, 
installing isolation into an existing building is expensive and may require a 
significant public subsidy to make viable. Several high-profile city hall 
buildings such as San Francisco, Oakland, Berkeley (after the 1989 Loma Prieta 
earthquake), and Los Angeles (after the 1994 Northridge earthquake) have been 
isolated, with FEMA assistance as part of post earthquake damage repairs.
8.6.3 Examples of Historical Buildings
The following illustrations show samples of seismic retrofit of historic 
buildings with brief descriptive notes.  Complete discussion of the preservation 
issues and rehabilitation techniques of each case would be extensive and cannot 
be included here.
8.7	Conclusion
Table 8-3 summarizes common seismic deficiencies stemming from various site and 
configuration characteristics as well as those that might be expected in each 
FEMA model building type.  See Section 8.2 for a discussion of “seismic 
deficiency” as used in this chapter and this table.  Also included in Table 8-3 
are retrofit measures that are often used for each situation.

8.8  References

8.8.1 References from Text
American Society of Civil Engineers (ASCE), Seismic Evaluation of Existing 
Buildings, ASCE 31-03, 2003.
Applied Technology Council (ATC), Earthquake Damage Evaluation Data for 
California, ATC 13, 1985.
California Seismic Safety Commission (CSSC), Status of the Unreinforced Masonry 
Builidng Law, 2003 Report to the Legislature, SSC 2003-03, 2003.
Federal Emergency Management Agency (FEMA), Prestandard and Commentary for the 
Seismic Rehabilitation of Buildings, FEMA 356, 2000.
Holmes, William T., Background and History of the California Hospital Seismic 
Safety Program, Proceedings, Seventh National Conference on Earthquake 
Engineering, Boston, 2002, Earthquake Engineering Research Institute.
Hoover, Cynthia A., Seismic Retrofit Policies: An Evaluation of Local Practices 
in Zone 4 and Their Application to Zone 3, 1992, Earthquake Engineering Research 
Institute.
National Institute of Standards and Technology (NIST), Standards of Seismic 
Safety for Existing Federally Owned or Leased Buildings, ICSSC RP 6, 2002.
U. S. Department of Interior (Interior a), The Secretary of the Interior’s 
Standards for Rehabilitation, Department of Interior Regulations, 36 CFR 67, U. 
S Department of the Interior, National Park Service, Preservation Assistance 
Division, Washington, D.C.
U. S. Department of Interior (Interior b), Guidelines for Rehabilitating 
Historic Buildings, U. S Department of the Interior, National Park Service, 
Preservation Assistance Division, 1977, Washington, D.C.
8.8.2 To Learn More
Applied Technology Council, Seismic Evaluation and Retrofit of Concrete 
Buildings, Report No. SSC 96-01, California Seismic Safety Commission, 1996. 
A contemporary of FEMA 356 that features many of the same methods and 
performance terminology. This document contains a good description of retrofit 
strategies.

Earthquake Engineering Research Institute Ad Hoc Committee on Seismic 
Performance, Expected Seismic Performance of Buildings, SP 10, Earthquake 
Engineering Research Institute, 1994.
This document was published slightly before FEMA 356 and thus contains slightly 
different performance terminology.  However, photo examples and extensive 
description of damage states are contained.  In addition, estimates are given 
for the approximate number of various buildings that would be expected to be in 
various damage states for different ground motion intensities.

Federal Emergency Management Agency, Region 1, Safeguarding Your Historic Site, 
Boston, MA.
This document contains an extensive bibliography covering renovation and repair 
of existing buildings.

Freeman, John R.  Earthquake Damage and Earthquake Insurance.  McGraw-Hill Book 
Company, Inc., 1932, New York.
Extremely interesting from a history standpoint, this book Includes discussion 
of seismology, geotechnical engineering, structural engineering, codes, and loss 
estimation, and excellent history and available data on earthquakes up to 1932.

Holmes, William T., Risk Assessment and Retrofit of Existing Buildings, 
Proceedings Twelve World Conference on Earthquake Engineering, Auckland, New 
Zealand, 2000.
This paper contains a more technically oriented description of the methods of 
FEMA 356 and strategies for design of retrofit systems.

[End of Chapter 8]