• Introduction
• A Disaster Resilient San Francisco
• Performance Goals For The "Expected" Earthquake
  • Table 1: Recovery Objectives By Timeframe
• Defining The "Expected" Earthquake
  • Table 2: Earthquake Levels Defined
• Defining Transparent Performance Measures
  • Table 3: Seismic Performance Measures For Buildings
  • Table 4: Defining Resilience For San Francisco
• Next Steps
• Appendix One
  • General Objectives For Response Operations

INTRODUCTION
When a major earthquake strikes the Bay Area, we will face thousands of casualties, hundreds of thousands of displaced households, and losses in the hundreds of billions of dollars. Recovery will take years.

This paper addresses one aspect of the broader policy problem related to making San Francisco resilient in the face of a disaster - the standards we use for deciding when a structure is "safe enough." Our building code embodies hundreds of judgment calls about how strong structures should be, but the public and the policy makers generally have no idea what these standards mean, what the outcomes will be from the "black box" of engineering decisions.

The truth is that when we choose our engineering standards we really are choosing to define how many deaths, how many building demolitions, and how long a recovery time we will have for various levels of earthquakes. Currently, the City of San Francisco has no adopted performance objectives for determining these factors. As a result,

• Design and construction requirements for new construction still focus mostly on preventing the loss of life and in most cases ignore the question of building damage and post-earthquake usability.
• Little is being done to rehabilitate older existing structures, which constitute the majority of buildings and which were built without earthquake-resistant features now required.
• There is no consistent approach to providing, maintaining, and restoring lifeline systems that are needed to support economic recovery.

The overall impact and cost of a disaster is strongly influenced by how long it takes to recover. The time needed to recover depends on the level of damage sustained by buildings, the availability of utilities, and how quickly communities can re-establish usable housing and livable environments.

This paper provides a new framework for improving San Francisco's resilience through seismic mitigation policies. Our goals are to:

1. define the concept of "resilience" in the context of disaster planning,
2. establish performance goals for the "expected" earthquake that supports our definition of resilience,
3. define transparent performance measures that help us reach our performance goals; and
4. suggest next steps for San Francisco's new buildings, existing buildings and lifelines.

A Disaster Resilient San Francisco
Whether hit by a massive earthquake or by a hurricane, tornado, flood, or terrorist attack some communities have shown an incredible ability to recover, that is incredible resilience. Resilient communities have an ability to govern after a disaster has struck. These communities adhere to building standards that allow the power, water, and communication networks to begin operating again shortly after a disaster and that allow people to stay in their homes, travel to where they need to be, and resume a fairly normal living routine within weeks. They are able to return to a "new" normal within a few years. They are resilient communities because such a blow from nature remains a disaster, but does not become a catastrophe that defies recovery.

SPUR believes that San Francisco's disaster planning efforts must stem from a clear understanding of resilience. SPUR defines "seismic resilience" as the ability of the city to remain safe and usable after a major earthquake A resilient city is able to contain the effects of earthquakes when they occur, carry out recovery activities in ways that minimize social disruption, and rebuild following earthquakes in ways that mitigate the effects of future earthquakes. San Francisco's current disaster planning efforts are moving in the right direction, but the lack of public knowledge about the current performance of our buildings and lifelines is hindering our ability to become a resilient city.

Efforts to define the city's seismic resilience objectives must come with a clear and transparent understanding of the hazards and risks that we face. These should be defined in terms of the earthquakes we expect to occur, the damage they will cause to the built environment, and the effects of that damage on response and recovery. What is needed is a set of simple, useful, and expanded performance goals adopted and accepted by policy makers that set reasonable standards for restoring everyday life in the wake of a disaster. The set of general recovery objectives now included in the City and County Emergency Response Plan (Appendix1) is a first good step. SPUR believes that these policies need critical examination and expansion to focus on facilitating recovery.

Achieving a desirable level of disaster resilience requires a clear and specific understanding of what is likely to happen in the wake of an earthquake, and specific ideas for how to improve the performance of facilities in order to meet adopted goals. Commonly held misconceptions of expected seismic performance confuse efforts to improve seismic resilience. For example, the common belief that buildings and infrastructure built to code are "earthquake-proof" is incorrect. Instead, many new facilities are designed today with the expectation (at least by the earthquake professionals) that significant damage likely will occur in a major earthquake. SPUR believes that simple and direct descriptions of seismic performance goals should be adopted so that regulations to achieve these goals can be developed. Recommended goals are summarized below for how building performance should be categorized given the major earthquake that is expected to occur.

Performance Goals for the "Expected" Earthquake
SPUR is not recommending that all facilities be upgraded to a level that would make them damage proof without regard to cost. Rather, our intent is to clarify the level of damage that is acceptable and require only those improvements in performance needed to assure a quick recovery, or the level of resilience desired. We believe that most of the current criteria for new buildings are adequate, and the need for strengthening existing buildings is perhaps less extensive than generally perceived. Nevertheless, some important changes are needed.

For purposes of this discussion, seismic performance goals are stated in terms of the general states of damage and repair over an extended recovery period under the assumption that an "expected" earthquake has occurred. We chose to analyze the "expected" earthquake, rather than the "extreme" event, because it is a large event that can reasonably be expected to occur once during the useful life of a structure or lifeline system (see section, "Defining the ‘Expected' Earthquake, below)

We have defined performance goals in terms of three response and recovery phases, four categories or "clusters" of infrastructure types, eight performance categories and selected time frames as shown in the tables below.

Defining the "Expected" Earthquake
A defined level of earthquake performance only makes sense if you also define which earthquake you're talking about. The same damage or disruption that might be forgivable in a large earthquake might be unacceptable in a small one.

Earthquakes are commonly reported in the public press in terms of the Richter Magnitude. That measure was defined in the early 1940s and is useful only for quantifying the energy released overall by a single earthquake. It offers little in terms of how buildings or infrastructure will fare or how they should be designed.

In order to evaluate an existing facility or design a new one, earth scientists and engineers today use different measures to quantify the intensity of the expected shaking at a specific site. These engineering measures account for the possible effects of different earthquakes on multiple faults and therefore are defined in probabilistic terms. For example, a building might be designed for the level of shaking expected to occur with 10% probability over a 50-year time period.

A third way of defining earthquakes involves "scenario events": – specific hypothetical earthquakes defined by the location of the fault rupture and the magnitude of the energy released. For example, a "magnitude 6.9 earthquake on the San Andreas fault near Santa Cruz" would be one scenario event of possible interest in San Francisco, as that would represent a repeat of the 1989 Loma Prieta earthquake. Scenario earthquakes are especially useful for city-wide or regional planning. They also are easier to grasp than probabilistic measures and are therefore effective for communicating earthquake risk to policy makers and to the public. In 2003, the San Francisco Department of Building Inspection's Community Action Plan for Seismic Safety chose four scenario earthquakes as the basis for its planning and mitigation programs.

For the purposes of defining resilience and developing mitigation policies to achieve it, SPUR uses one of the scenario earthquakes also used by the Community Action Plan for Seismic Safety and refers to it as the expected earthquake: A magnitude 7.2 earthquake located on the Peninsula segment of the San Andreas fault.

Our expectations, our goals, and our recommendations are presented with this event in mind. This expected earthquake can be expected - conservatively but reasonably to occur once during the useful life of a structure or system, and more frequently if the structure is renovated periodically (as most San Francisco buildings are) to serve more than one or two generations. Of course, this defined scenario would produce different levels of shaking at different locations, but for most of the city, its effects would be similar, in probabilistic terms, to those with a 10 percent chance of occurring over a 50-year period.

We define resilience in terms of this expected earthquake. Other earthquakes are possible, of course. In a smaller, more routine earthquake, better performance would be expected. For a larger, more extreme event, lesser performance would have to be tolerated. All three events are defined below to encourage uniform definitions.

Defining Transparent Performance Measures

Declaring in simple, concise and understandable terms the expected seismic performance of structures and systems given a specific earthquake size requires first the adoption of terms that are recognizable, consistent with other performance rating systems, and useful is establishing policy. While the earthquake professional community has learned that expected seismic performance is best measured and expressed in terms of casualties (lives lost and serious injuries), dollar loss (cost of repair as a percentage of replacement value), and the length of time required to repair for reuse or restore service, no standard set of performance measures is in use.

The various engineering professions use a wide variety of terms to describe their performance expectations and, in some cases, do not clearly state what they are. In the case of new buildings, the code requires different design standards for essential facilities, high occupancy facilities, and facilities containing hazardous materials, but stops short of declaring what performance is expected. Rehabilitation standards for existing buildings that are undergoing refurbishment are often designed to an explicit performance level but many voluntary retrofits are done to no specific standard. Utility systems are often designed with employee safety goals and expected days to restore service to users, though there is no consistency.

Most important, even when performance objectives are explicitly defined, they are almost always stated in terms of safety only – that is, in terms of whether people nearby would be injured during the earthquake itself. While safety is essential, this limited perspective fails to address the larger questions of resilience: Will the buildings be usable when the shaking stops, and how soon will full services be restored?

 

For purposes of understanding and working toward resilience, SPUR defines categories of the expected performance of buildings and lifeline systems in terms of both safety during the earthquake and usability during the response and recovery periods that follow. Lifeline performance is further defined in terms of the time allowed to resume service.

SPUR uses the following terms in developing new building design standards and mitigation programs needed to achieve San Francisco's resilience objectives. The levels of performance they describe are to be paired with the effects of the Expected earthquake. Table 3 presents these performance measures in a tabular form.

In addition, SPUR defines the expected performance of all utility and transportation systems, or portions of systems, serving the City in terms of the days required to restore service to 90 percent, 95 percent and 100 percent of the defined customer base.

Category I—Resume 100 percent of service levels within four hours.
Critical response facilities - including emergency housing centers – need to be supported by utility and transportation systems critical to their success. This level of performance assures that these systems will be available within four hours of the disaster. It requires a combination of well built buildings and systems, provisions for making immediate repairs as needed, and redundancy within the networks that allows troubled spots to be isolated.

Category II—Resume 90 percent service within 72 hours, 95 percent within 30 days, 100 percent within four months
Housing and residential neighborhoods require utility and transportation systems be restored quickly so that these areas can brought back to livable conditions. There is time to make repairs to lightly damaged buildings and replace isolated portions of the networks or create alternate paths for bridging around the damage. There is time for parts and materials needed for repairs to be imported into damaged areas. These systems need to have a higher level of resilience and redundancy than the systems that support the rest of the city.

Category III—Resume 90 percent of service within 72 hours, 95 percent within 30 days, 100 percent within three years
The balance of the city needs to have it's systems restored as buildings are repaired and returned to operation. There is time to repair and replace older vulnerable systems with new. Temporary systems can be installed as needed. Most existing lifeline systems will qualify for Category III performance.

Table 4: Defining Resilience for San Francisco
As mentioned above, SPUR defines San Francisco's "seismic resilience" as its ability to contain the effects of earthquakes when they occur, carry out recovery activities in ways that minimize social disruption, and rebuild following earthquakes in ways that mitigate the effects of future earthquakes. This chart represents our best thinking as to how a resilient San Francisco ought to perform in the event of an expected earthquake, and how far away we are right now from meeting that goal.



Next Steps
While this paper defines the goal of a earthquake-resilient San Francisco, new papers on new buildings, existing buildings, and lifelines recommend actions to achieve this goal. Overall SPUR's specific recommendations for action have been developed by considering:

1. the goals for seismic resilience for each component of our city
2. the gap between current seismic performance and the goal, and
3. the cost of making the necessary improvements or retrofits.

Priority is given to those actions that provide the best improvement to seismic performance and resilience with the least amount of cost and disruption.

SPUR's recommendations⁵ for existing buildings focus on dealing with the city's most seriously deficient buildings through mandatory programs where the needs are clear, and the development of additional programs where further definition is needed. Soft story wood frame buildings need to be retrofitted. Buildings providing shelter and other essential City services need to be evaluated and retrofitted as needed. San Francisco's non-ductile concrete buildings – structures that have frames of steel reinforced concrete, as differentiated from the typically brick built unreinforced masonry buildings, and that are vulnerable because their frames do not bend in earthquakes – need to be inventoried, and a program should be developed to mitigate their potential for collapse. Gas shut off valves need to be installed to prevent massive fires from consuming the city following an earthquake. And the adequacy of the 1992 Unreinforced Masonry Building Retrofit Ordinance needs to be assessed.

SPUR's recommendations for new buildings focus on adding transparency to the building code requirements; incorporating new, near term cost-effective improvements into the code; creating a certification system for voluntary seismic upgrades akin to the LEED system for green building standards; and adding strong incentives for owners to build to higher seismic standards. Owners and tenants need to understand what seismic performance to expect from the buildings they own, lease and construct, and the options they have for improvement.

SPUR's recommendations for lifelines begin with the formation of a lifelines council that brings the various owners and operators together to assess the vulnerability of the systems that serve San Francisco. The council also should develop performance standards and retrofit priorities. The City needs to incorporate the needed retrofits into its capital improvement plans, give specific attention to managing the city's gas system, and form partnerships with the regional and state public and private providers to secure the necessary modifications to their systems. The City also should become an advocate at all levels for the development of consistent performance standards nationwide.

Achieving disaster resilience is critical to the survival of San Francisco. Resiliency is not an impossible nor economically infeasible goal. Success requires a clear understanding of what will happen, and what steps must be taken now – before the disaster - coupled with an achievable implementation schedule and a rational program of incentives to make programs feasible. These recommended first steps serve to bring clarity and transparency to the problems we face, and set the stage for the development of detailed solutions.

SPUR believes that we must embark on what will likely be a 50-year process to make San Francisco a resilient city. The recommendations within this paper help us get started.

APPENDIX ONE
San Francisco has maintained workable and exercised emergency operations plans for decades. The CCSF 2005 Emergency Operations Plan represents the latest complete document that outlines the organizational structure and operating procedures that are placed in effect in the event of a major earthquake. That plan is presently undergoing a complete revision and transformation into an all hazards, Emergency Response Plan.

At present, the CCSF 2005 Emergency Operations Plan – Earthquake Response Plan Enhancement includes the follow general objectives and time frames that have been used as a basis for the Resilient City recommendations.

General Objectives for Response Operations
Immediately following a major earthquake, and for as long as a state of emergency exists within the CCSF, the response to the earthquake will be the first priority of all CCSF departments and agencies. All available resources will be directed to achieving the following objectives:

• Save lives.
• Reduce immediate threats to life, public health and safety, and public and private property.
• Provide necessary care, shelter, and medical services to CCSF residents and other members of the general public.
• Restore the operations of facilities, whether public or privately owned, that are essential to health, safety, and welfare of the community, including critical CCSF facilities, hospitals, utilities, and transportation infrastructure.
• Assess damage to infrastructure, public facilities, and the built environment.
• Expedite the restoration of services, the economy, and the community at large; and begin the process of recovery.
• Keep the public informed.

ENDNOTES
¹ Shelter in place is used by emergency response professionals to mean the place in a building where people can seek safety during a life threatening incident. SPUR uses "shelter in place" to mean that a building is disaster resilient enough for people to safely remain in their home during both the earthquake itself and subsequent needed repairs, even though the public utility systems may not be working.
² Building inspected and deemed safe for occupancy.
³ Building inspected and found to be damaged enough to warrant restricted access.
⁴ Building inspected and found to be unsafe to occupy.
⁵ The following paragraphs summarize the policy recommendations of SPUR's three other "Before the Disaster" policy papers: "The Dilemma of Existing Buildings: Private Ownership, Public Risk," "Building it Right the First Time: Improving the Seismic Performance of New Buildings," and "Upgrading Infrastructure to Enhance San Francisco's Earthquake Resilience." All three papers can be found on SPUR's website at www.SPUR.org.