Seismic Retrofit Ordinances Part 2—Understanding Earthquakes

To view the figures and tables associated with this article, please refer to the flipbook above.

I have heard that many engineers from non-seismic regions are scared by seismic design. I feel the same about wind design, and after listening to other presentations at the Alabama conference I attended, that feeling has only been reinforced. However, to put the non-seismic regions at ease, if you can understand the basics of the vibration of a single-degree of freedom, lumped-mass system and follow a load path from the roof to the ground, you know enough to design most structures for earthquakes.

In terms of behavior, the underlying basis of seismic design is that certain elements in the building need to survive the earthquake with a degree of damage permitted (called deformation-controlled elements or perhaps better, yielding elements) and some elements in the building (called force-controlled or non-yielding elements) cannot be damaged because they are relied upon to deliver the seismic load generated by the mass of the building to the yielding elements. This simple concept exists implicitly in the provisions of the building code, but the code unfortunately never states it explicitly as its foundational basis.

Like windstorms, earthquakes come in all “shapes and sizes” with different durations, ground shaking intensity, frequency content, etc. But the goal of the building code is for the building to hang together until the shaking stops in what might be considered a repairable condition in the design basis earthquake. The design basis earthquake (DBE) is nothing more than the envelope of seismic design forces for all structures, from short, stiff structures (short fundamental period of vibration, say 1 second or less) to tall, less stiff structures (long fundamental period of vibration, say 3 seconds or more) in what is called a spectrum. The engineer can achieve the repairable condition goal with either (1) massive strength so that no or little damage occurs or (2) reliably ductile elements that are damaged during the earthquake and then repaired after.
The engineer, in consultation with their client, can select from several commonly defined seismic performance goals, unless the building code otherwise establishes a minimum goal based on occupancy:

Frankly, an engineer needs to be my age or older to personally remember any of them. Even the Northridge earthquake occurred 30 years ago. Prior to 1971, there hadn’t been a major earthquake in a heavily populated area in the continental U.S. in a very long time, and there really hasn’t been one since.

The remainder of this article, and the next part in the series which will be published later, will focus on seismic retrofit ordinances. However to really understand the seismic engineering field and what constitutes a complete retrofit, it is important to note that in any retrofit, besides the structural issues i.e the building itself, geologic effects need to be considered (fault rupture, soil liquefaction, settlement, landsliding, etc.) as well as non-structural hazards (contents, ceilings, walls, MEP systems, etc.). See Figures 1 and 2 for ground effects at bridge abutments and soil liquefaction.

A very good book that contains a great collection of historic earthquake damage photos, basically showing everything that “could go wrong,” is Earthquakes, Volcanoes, and Tsunamis – An Anatomy of Hazards by Karl V. Steinbrugge, published in 1982. Used copies are available for about $50.

Mandatory Seismic Retrofit Ordinances

Besides modifying the building code provisions for new buildings, over the years, California has enacted seismic retrofit ordinances to address seismic deficiencies in existing structures that were originally designed according to the standards of the day or more recently to the building code. To some degree this explains my feelings about relying on the building code (a dose of healthy skepticism), remembering that the code represents a minimum standard and the complexity of the building code, which often obscures the goal (too detailed and probably not entirely correct anyway in terms of achieving the desired performance objective, despite our best efforts as noted previously).

The adopted retrofit ordinances are typically phased in terms of implementation, starting with an initial inventory and screening phase performed by the local building department to gauge the extent of the “problem” in their jurisdiction, an evaluation and reporting phase performed by the building owner to decide whether their building actually falls under the purview of the ordinance, and then construction documents and construction phases for buildings requiring retrofit. To allow for adequate planning and financing of the project by the building owner, the overall schedule can be 20 years or longer. To the dismay of many, sometimes it is a better decision to demolish an old building than to retrofit it, since when retrofitted, it is still an old building.
California has had four mandatory seismic retrofit ordinances:

Seismic retrofits of a mandatory nature have an implied performance target, but the structural engineer needs to educate the owner and determine if a higher goal should be targeted.

Two additional mandatory seismic retrofit ordinances are under development, except as noted below:

Tall cantilevering parapets have been known seismic hazards for a very long time, probably as long as parapets on buildings located in high seismic regions have existed (Fig. 3). Dynamically, an unbraced, cantilevering element can generate four or five times the lateral force of a similarly configured braced element. In terms of California earthquake history, the 1906 Great San Francisco Earthquake and Fire produced such extensive damage and devastation that the contribution from parapets themselves was probably not seen as noteworthy. For the next half-century, except for earthquakes in Santa Barbara (1925 – M 6.6, MM IX) and Long Beach (1933 – M6.4, MM VII) there were not any other damaging earthquakes in densely populated areas.
Truth be told, earthquakes are rather rare events compared to strong wind events such as tornados and hurricanes. During the middle decades of the 20th century, as California experienced tremendous growth, particularly in the south with the Los Angeles region surpassing the San Francisco Bay Area in population, retrofitting older buildings didn’t appear to be a priority. One might reasonably speculate that the construction industry and building officials reasoned that older buildings would eventually be demolished making way for new and larger modern buildings. Also, people naturally tend to forget or lose interest as time passes. It takes a really concerted effort to effect a change.

In 1964, the Great Alaska Earthquake and Tsunami, the second largest instrumentally recorded in history, occurred (M 9.2). According to the book titled San Francisco’s Parapet Ordinance (Paul Newman, Jay Turnbull and Sarah Haugh—published 1977) the Alaska earthquake generated local interest in requiring the bracing of parapets in San Francisco. San Francisco’s older building stock, many of which are historic structures, have tall and often elaborate parapets and cornices, constructed structurally from clay brick. In addition to structural engineers, some of the city’s leading architects were involved in the effort partly out of concern that architecturally significant buildings would be demolished, or the parapets removed as the most cost-effective solution to hazard reduction. One can only imagine what the parapets might look like if engineering and cost were the only concerns.

The Parapet Ordinance was enacted in 1969 and became the “original” seismic retrofit ordinance. Like most of the ordinances discussed herein, the authors started small by focusing on just a part of a masonry building, although they clearly knew the larger hazard presented by brick masonry buildings themselves. Passing the ordinance was the start, but it was not until 1975 that San Francisco budgeted sufficient funds to start the screening and inventory process and enforce the Ordinance.

The Ordinance was very successful. If one walks around San Francisco today, and looks closely, you will see evidence of braced parapets on almost all brick buildings. In the early 1970s, the UBC prescriptively required that bracing be designed for 0.20g, while a true cantilever parapet had to be designed for 1g (five times as much), demonstrating the benefit of bracing. Diagonal steel braces, spaced roughly 6 to 8 feet, are the most common design connecting the back of the parapet with the roof structure.

As the architects feared, the least expensive bracing design involves steel rod anchors that pass through the parapet wall and anchored with a steel plate washer. Structurally, this is better than an anchor that is only grouted into the thickness of the brick, since the through-anchor generates tension through bearing while the grouted anchor relies on the less reliable break-out in the brick matrix to develop the tension force. Aesthetically the through-bolt approach is inferior but is mostly used on ordinary commercial buildings or in locations hidden from public view rather than on the fronts of historically significant buildings located on the major streets. San Francisco’s parapets performed well in the 1989 Loma Prieta Earthquake.

Just like masonry parapets, un-strengthened brick masonry structures have been collapsing and falling onto sidewalks in earthquakes as long as masonry buildings and sidewalks have existed. Throughout the former Roman world (east and west and farther east than that) when archeologists talk about several destroyed cities stacked on top of one another, I think this is what they are referring to.

Prior to the parapet ordinance, brick walls and brick parapets were nominally connected or not connected at all (as engineers think about it) to the horizontal roof or floor framing. If the walls were connected to the horizontal structure, it was often by means of what is often called a “government anchor.” Legend has it that these anchors were called “government anchors” as in the “government made me do it.” I can’t vouch for this bit of history, but it makes sense as a prescriptive requirement. Such anchors consisted of a round steel rod attached to the wall and fitted into a hole in the wood joist or rafter by means of a 90-degree bent tip, basically a J-bolt (Fig. 4). The attachment to the wall was either by means of a nut and plate washer on the exterior (and sometimes a steel rod, see Figure 5) or a nut and plate washer embedded into the middle of the brick as the wall was constructed. The later approach didn’t perform as well during earthquakes due to its lower natural strength and often the poor quality of the mortar used in the masonry.

Earthquakes such as those which occurred in Santa Barbara (1925) and Long Beach (1931) demonstrated the vulnerability of brick masonry buildings. However, it wasn’t until 1981 that Los Angeles became the first city to enact a retrofit program (Division 88 of the Los Angeles Building Code). To quote the code,

“The purpose of this division is to promote public safety and welfare by reducing the risk of death or injury that may result from the effects of earthquakes on unreinforced masonry bearing wall buildings constructed before 1934. Such buildings have been widely recognized for their sustaining of life hazardous damage as a result of partial or complete collapse during past moderate to strong earthquakes. (Figures 6 and 7 show typical damage the ordinance was designed to address.)

"The provisions of this division are minimum standards for structural seismic resistance established primarily to reduce the risk of loss of life or injury and will not necessarily prevent loss of life or injury or prevent earthquake damage to an existing building which complies with these standards. This division shall not require existing electrical, plumbing, mechanical or fire-safety systems to be altered unless they constitute a hazard to life or property.

“The owner of each building within the scope of this division shall cause a structural analysis to be made of the building by a civil or structural engineer or architect licensed by the state of California, and if the building does not meet the minimum earthquake standards specified in this division, the owner shall cause it to be structurally altered to conform to such standards or cause the building to be demolished.”

In terms of seismic performance, the intent appears to be somewhere between life-safety and collapse prevention. The State of California enacted a statewide plan in 1986. San Francisco followed with its own plan in 1992.

The overall intent was for the structure to be able to withstand a certain minimum horizontal base shear, to have adequate anchorage of the walls to the horizontal structure and for there to be a “complete, continuous stress path from the part or portion of the building to the ground.” In common parlance, this and similar ordinances came to be called “Bolts Plus” meaning anchoring the walls to the roof and floors with bolts (the Bolts part) and making sure the walls had adequate in-plane and out-of-plane bending strength (the Plus part). See Figure 8 for a retrofitted building in San Francisco’s Chinatown.

Part 1 of this series appeared in the February 2025 issue of STRUCTURE.