Loma Prieta Earthquake: Collapse of the Bay Bridge and the Cypress Viaduct

Published by the Government Accountability Office on 1990-06-19.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                   United States
GAO                General Accounting Office
                   Washington, D.C. 20548

                   San Francisco      Regional     Office


                   June 19,199O

                   The Honorable Jamie L. Whitten
                   Chairman, Committee on Appropriations
                   House of Representatives

                   The Honorable Glenn M. Anderson
                   Chairman, Committee on Public Works
                     and Transportation
                   House of Representatives

                   On October 17,1989, the Loma Prieta Earthquake struck northern Cali-
                   fornia, causing the collapse of a two-level, 1.25-mile-long section of the
                   Cypress Viaduct on Interstate Route 880 in Oakland; 42 people were
                   killed. A section of the Bay Bridge connecting San Francisco and Oak-
                   land also collapsed, resulting in one death.

                   In accordance with House Report 101-301, Further Continuing Appro-
                   priations, 1990, dated October 23, 1989, and subsequent agreements
                   with your offices, this report provides information on (1) what the Cali-
                   fornia Department of Transportation (cdnan~) knew about the struc-
                   tures’ vulnerability to earthquake forces before they collapsed, (2)
                   federal and state funding available and expended in California to
                   strengthen bridge and viaduct’ structures subject to earthquake forces,
                   and (3) funding needed to complete California’s seismic retrofit
                   program.        .

                   The Cypress Viaduct was 1 of about 9,700 bridges in California that
Results in Brief   were built before current earthquake standards. After the 1971 San Fer-
                   nando Earthquake, cal~rans engineers realized that the support columns
                   and decks on all bridges designed before 1971 were potentially vulner-
                   able to earthquake damage. They began what evolved into a three-phase
                   retrofit program to correct the deficiencies. Retrofit work was given
                   lower priority than other safety projects and at the time of the Loma
                   Prieta Earthquake-18     years later--calTrans had completed one of the
                   three phases. This phase corrected what C~~TIXIIS believed to be the most
                   critical problem, the potential for deck sections to separate and collapse
                   on all bridges considered to be vulnerable, including the Cypress

                   ‘While a bridge is an elevated structure over water and a viaduct is an elevated structure over land,
                   we refer to both types of stmctures as bridges.

                   Page 1                                                 GAO/RCED9@177        Loma Prieta Earthquake

             After studying the Cypress Viaduct’s collapse, cal~ransengineers discov-
             ered a weakness in the Cypress’ columns which they believe, when cou-
             pled with the effects of the earthquake on the soft soils underlying the
             structure, led to its collapse. This weakness was in addition to the defi-
             ciencies common to all bridge columns designed before 1971. cal~rans
             engineers believe they would have identified this weakness had the
             retrofit program progressed to phase three. They further believe that,
             had soil conditions been factored into the retrofit program’s priority
             scheme, the Cypress Viaduct would have been scheduled for work

             The San Francisco/Oakland Bay Bridge, also built before 1971, was rein-
             forced in the mid-1970s to protect it against earthquake damage or col-
             lapse. However, cal~ran~ officials did not believe that the section which
             collapsed during the Loma Prieta Earthquake was vulnerable and had
             not reinforced it at this point.

             According to FHWA officials, California has received over $5.5 billion in
             federal-aid highway assistance since 1975 (the first year for which
             information is available) for programs that include seismic retrofit
             projects as an eligible activity. Of the $4.2 billion available to cal~rans,
             cal~ran~officials estimate that they spent about $54 million between
             1971 and 1989 in federal and state funds to complete the first phase of
             the seismic retrofit program-tying      together the road deck section of
             1,26 1 bridges to prevent separation and collapse. The second phase-
             strengthening the columns of 392 or more single-column bridges-is
             scheduled for completion by December 1991 and will cost about $150
             million to complete. The third and final phase-strengthening the col-
             umns of at least 700 multicolumn structures-is still being researched.
             As a result, a cost estimate is not yet available.

             To centralize its control over the retrofit program and to help ensure
             that all work is completed by December 1993, cal~rans has been reorga-
             nized internally. However, the state has not yet committed sufficient
             funding to complete the program. The decision was deferred until after
             June 5,199O. On that date, voters considered and approved a ballot
             measure intended to make additional transportation funds available.

             The vulnerability of bridges to earthquake effects was first brought to
Background   the state’s attention in 1971, when the San Fernando Earthquake
             destroyed five Southern California bridges. The destruction demon-
             strated weaknesses in the existing seismic design standards for bridges,

             Page 2                                    GAO/MXD-W-177   Lmna Prieta Jkthquake

        particularly in road deck and support column design features. cal~rans
        therefore revised its design standards for the construction of new
        bridges in 1974, which were then adopted and periodically revised by
        the American Association of State Highway and Transportation Offi-
        cials. The Association promulgates bridge design standards, which the
        Federal Highway Administration (FHWA) endorses for use on federally
        funded projects. Generally, to receive federal funding, states must
        demonstrate compliance with the Association’s standards.

        Bridges built before the standards were revised, however, needed to be
        strengthened to overcome the design weaknesses identified after the San
        Fernando Earthquake. ca.man~ developed internal guidelines in 1975 for
        retrofitting to prevent deck separation and is currently working on
        guidelines for column retrofitting schemes. The Association, however,
        has not yet adopted retrofitting standards.

        According to FHWA officials, federal-aid highway funds are available to
        states for seismic retrofitting through five programs: Consolidated Pri-
        mary System; Secondary System; Urban System; Interstate Resurfacing,
        Restoration, Rehabilitation, and Reconstruction; and Bridge Replace-
        ment and Rehabilitation. However, F’HWA does not earmark funds to meet
        seismic retrofitting needs; states spend the funds according to their own
        priorities. States generally match federal funds by contributing 10 to 25
        percent, depending on the program, of a project’s costs. States may not
        use federal funds for toll bridges such as the San Francisco/Oakland
        Bay Bridge, but Congress did allow emergency relief funds allocated
        after the Loma Prieta Earthquake to be used for repairing the San Fran-
        cisco/Oakland Bay Bridge.

        In California, the legislature determines the amount of cal~ran~’
        and makes some determinations about how the budget should be distrib-
        uted across the state. The California Transportation Commission is
        responsible for determining how these funds will be allocated among
        transportation programs and approves individual projects at the time
        projects are ready to be advertised.

        Page 3                                 GAO/RCED-99477   Loma Prieta JSarthquake


                         At the time of the Loma Prieta Earthquake, cal~ransofficials knew that
Cypress Viaduct Had      the Cypress Viaduct, a double-deck, multicolumn structure built before
Known Weaknesses         1971, had insufficiently reinforced support columns that were vulner-
but Was Not              able to earthquake damage. Before the Loma Prieta Earthquake, no mul-
                         ticolumn bridge had collapsed during an earthquake. cal~ran~officials
Considered Li .kely to   anticipated that a major earthquake could damage the columns, but did
Collapse                 not believe that the roadway would collapse. They now believe that, in
                         addition to the reinforcement deficiencies common to all pre-1971 struc-
                         tures, the Cypress structure had other deficiencies in its columns, which
                         engineers working on CalTram seismic retrofit program did not know
                         existed, and which CalTram believes may have led to the collapse.

                         The need for seismic retrofitting became apparent when the 1971 San
                         Fernando Earthquake severely damaged five bridges and disclosed the
                         need to tie bridge decks together and to reinforce support columns.
                         After this earthquake, C~IX.IIS revised its standards for new construc-
                         tion and began what evolved into a three-phase seismic retrofit program
                         to strengthen bridges built before 1971. (See app. II for a detailed expla-
                         nation of the program.) Correcting deck weaknesses, the first phase, was
                         given priority because cal~ran~ believed that these weaknesses were both
                         more serious and less expensive to correct. This work was completed on
                         the Cypress Viaduct in 1977 and on a total of 1,261 structures by 1989.
                         Column work was considered less critical and was planned for later
                         phases of the program. In addition, technology was not available for
                         column retrofit work and needed to be developed.

                         After studying the collapse, C~~TMSengineers now believe that the
                         Cypress Viaduct had an additional, more serious weakness in its col-
                         umns which, when combined with the amplified shaking motions in the
                         soft soils underlying the structure, led to its collapse. (See app. III for a
                         discussion of earthquake force measurements.) According to camram
                         engineers, the reinforcement provided at the point where the columns
                         joined the lower deck-the pedestal section-was insufficient to pre-
                         vent the columns from breaking. (See app. I, figs. I.2 and 1.3.) It was at
                         this point that many of the columns sheared off, causing parts of the
                         upper and lower decks to collapse. (See app. I, figs. 1.4. and 1.5.)

                         This design feature, according to cal~ran~ engineers, was only used on the
                         Cypress Viaduct and six other viaducts built in San Francisco about the
                         same time, which were severely damaged but did not collapse. CalTrans
                         engineers working on the retrofit program after 1971 were not aware
                         that any bridges contained this design feature and thus did not address
                         it in the retrofit program. However, they now believe that an inspection

                         Page 4                                    GAO/RCJD-9&177   Loma Prieta Earthquake

                      of the design and construction plans would have revealed the potential
                      for collapse. cal~ran~had not inspected these plans before the Loma
                      Prieta Earthquake, but planned to do so during the third phase of the
                      statewide retrofit program which, at the time of the earthquake, had not
                      yet begun.

                      C~~TWSengineers were aware that soft soils were layered throughout the
                      San Francisco Bay Area, including the area under the Cypress Viaduct
                      and the Bay Bridge, but had not systematically gathered the information
                      or incorporated it into the process used to determine the need for and
                      criticality of retrofit work. They now believe that, if soil information
                      had been used as a priority factor in the retrofit program, the Cypress
                      Viaduct would have been given a much higher priority.

                      Because of its exceptional size and complexity, the Bay Bridge is consid-
Bay Bridge Damage     ered a “special structure” to which standard engineering codes and stan-
Was Not Anticipated   dards do not apply. As a result, its retrofit needs were independently
                      assessed during the 197Os, and the bridge was equipped with deck
                      restrainers and other devices at points considered susceptible to earth-
                      quake damage. The additional reinforcing of several of the piers has
                      been planned since 1984. cal~a.t-t~did not consider the work urgent, how-
                      ever, and it was not scheduled for completion at the time of the Loma
                      Prieta Earthquake. According to C~~TWSofficials, the section that col-
                      lapsed was not previously considered vulnerable and an independent
                      engineering expert, studying the collapse under a National Science Foun-
                      dation grant, told us it would not have been affected by the additional
                      reinforcement planned for the piers.

                      The collapsed section broke off of one of the strongest piers supporting
                      the east bay section of the bridge. This pier was designed to absorb the
                      energy of an earthquake and prevent shocks to other points along the
                      bridge. According to cal~ran~engineers, the unanticipated magnitude of
                      the horizontal motion caused by the Loma Prieta Earthquake broke the
                      bolts holding the span to this pier, causing it to fall. (See app. I, fig. 1.6.)

                      cal~ransofficials believe that a detailed structural analysis could have
                      detected this section’s weakness. However, the bridge was not consid-
                      ered vulnerable and, before the Loma Prieta Earthquake, such analyses
                      were not performed because of their cost and complexity and because of
                      the limited funds available. The analysis involves a computer simulation
                      of how various shocks would affect about 27,000 possible stress points
                      along the bridge’s 8.3~mile length. cal~ran~is now planning to conduct

                      Page 6                                     GAO/RCED-90-177   Loma Prieta Earthquake


                        detailed analyses of all seven San Francisco Bay crossings, as well as
                        three large bridges in southern California. CdTrans officials estimate that
                        the Bay Bridge analysis will cost about $300,000 and take a year to com-
                        plete. The contract was awarded in April 1990.

Funding Available for   complete the first phase of its seismic retrofit program. Neither cal~rans
Retrofit Work           nor FHWA officials can readily estimate the federal share of these costs.
                        cal~ran~did not separately account for funds spent on seismic retrofitting
                        until after the Loma Prieta Earthquake. However, cal~rat~ budget offi-
                        cials normally estimate the federal share at 85 percent when preparing
                        budgets. This ratio would place the federal share of the seismic retrofit-
                        ting work at about $46 million.

                        FHWA  has allocated more than $11 billion in federal-aid highway funds to
                        California over the last 16 years to support transportation projects. Of
                        this amount, over $5.5 billion was allocated for four programs which
                        included seismic retrofitting work as an eligible activity: Interstate
                        Resurfacing, Restoration, Rehabilitation, and Reconstruction; Consoli-
                        dated Primary System; Secondary System; and Urban System.
                        According to FHWA officials, CalTram could use the $4.2 billion available
                        under the Interstate 4R, Primary, and Secondary programs to support
                        its activities. The $1.3 billion made available to the state under the
                        Urban System program was passed on to local governments.

                        Funds from the Bridge Rehabilitation and Replacement Program may
                        also be used for retrofitting work. However, it was not until 1990 that
                        FMWA, in response to the Loma Prieta Earthquake, allowed bridge funds
                        to be used on projects specifically designed to correct seismic deficien-
                        cies. Previously, these funds could be used for seismic retrofitting work
                        only when it was done in conjunction with other rehabilitation work.

                        Although seismic retrofitting work is eligible for funding under four fed-
                        eral-aid highway programs, the federal government does not specify the
                        amount of funds to be spent for such work. States determine how much
                        to spend for seismic retrofitting work on the basis of their own needs
                        and priorities. According to FHWA officials, California has been a leader
                        in researching and performing seismic retrofitting.

                        In response to the Loma Prieta Earthquake and Hurricane Hugo, the
                        Congress approved $1 billion in emergency relief funds which is avail-
                        able to repair highways and bridges damaged by these disasters. Some

                        Page 6                                   GAO/RCED-9Cb177   Loma Prieta Earthquake

                       of these funds will be applied to the Bay Bridge repair costs and will be
                       used to repair and retrofit other bridges damaged during the earth-
                       quake, including the Embarcadero, China Basin, Southern, Alemeny,
                       Terminal Separation, and Central viaducts in San Francisco. The emer-
                       gency relief funds can also be used to rebuild the Cypress Viaduct.2

                       Before 1987, some retrofit projects were included in CalTrans’ State
                       Transportation Improvement Plan, but funding was approved on a pro-
                       ject-by-project basis after the projects were ready to be advertised for
                       contract. After the 1987 Whittier Earthquake, the California Transpor-
                       tation Commission approved cal~ran~’    request to set aside a lump sum for
                       the second phase of the seismic retrofit program. cal~ran~proposed to use
                       $64 million of its budget (which included funds from both state and fed-
                       eral sources) over 4 years to strengthen single-column bridge structures.

                       After the Loma Prieta Earthquake, the state legislature appropriated
                       cal~ran~$60 million from the state’s disaster relief fund for retrofitting
                       needs.3The legislature anticipated that this appropriation would be sup-
                       plemented by federal funds. To raise the $60 million and to provide
                       funds for other emergency relief needs, the legislature imposed a 13-
                       month, quarter-cent sales tax increase. The legislature also mandated
                       that all retrofit work be completed by December 1991.

                       cal~ran~has not yet estimated the cost to complete all phases of the state-
Cost to Complete the   wide retrofit program. Officials have estimated it will cost $150 million
Retrofit Program Is    to complete most of the second phase of the retrofit program-strength-
Unknown                erring the support columns on 392 single-column bridges which need
                       retrofitting. This amount could include over $127 million in federal
                       funds and could increase as CalTrans continues to screen structures for
                       retrofitting needs. CalTrans anticipates that this work will be completed
                       by December 1991.

                       cal~ran~does not know how much funding will be required to complete
                       the third phase of the program-strengthening      the support columns on
                       multicolumn bridges. Although screening has not been completed, cal-
                       crankofficials estimate that at least 700 such structures will need

                       ‘We are currently reviewing the cost and status of the emergency relief program for California and
                       its impact on the Highway Trust Fund.

                       3The legislature appropriated an additional $20 million to retrofit local bridges.

                       Page 7                                                  GAO/RCED+ML177 Loma Prieta Earthquake

                         retrofitting. In addition, the technology to retrofit multicolumn struc-
                         tures is still being researched. As a result, CalTram will not be able to
                         complete this work by the legislature’s December 1991 deadline, and has
                         asked for an extension until December 1993. Because multicolumn
                         bridge structures are more complex than single-column structures, cal-
                         crankofficials expect the costs for multicolumn bridges to be higher. cal-
                         crankofficials said an estimate for phase three work would not be
                         available before December 1990.

                         In past years, retrofitting work has been given lower priority than other
CalTrans Has             safety projects. The first phase took 18 years to complete. Retrofit
Upgraded Priority of     projects were identified by calnan~ headquarters and assigned to the
Seism ic Retrofitting    appropriate districts for completion. However, the districts received no
                         additional resources specifically for retrofitting work. As a result,
Work, but Future         retrofit projects competed with other locally identified projects for pri-
Fundir lg Is Uncertain   ority and resources. According to calnan~ officials, safety and rehabilita-
                         tion projects were assigned priority based on the number of lives lost in
                         previous incidents. Thus, retrofit projects were accorded a relatively
                         low priority because, until the Loma Prieta Earthquake, earthquake
                         damage to state bridges had resulted in the loss of only two lives. Other
                         projects, such as installing guard rails and median barriers, tended to
                         take precedence because they were viewed as more critical to improving
                         highway safety.

                         As the phase one screening was completed and the appropriate district
                         offices were notified of which bridges needed to be retrofitted, neither
                         cal~ran~nor the California Transportation Commission monitored the
                         retrofit program’s progress. Some districts completed the work quickly,
                         while others did not.

                         Although C~~TW-LS  received approval in 1987 to reserve $64 million for
                         retrofitting single-column bridges and viaducts, projects were not for-
                         warded for funding until after the 1989 Loma Prieta Earthquake.
                         According to C~~TWSofficials, some project designs were completed by
                         June 1988, but funding problems and engineering resource constraints
                         prevented their completion. As a result, the districts did not complete
                         the site work needed to forward the projects for funding approval. Con-
                         sequently, as of April 1990-30 months after the 1987 Whittier Earth-
                         quake-no single-column retrofit projects had begun construction.
                         However, three contracts were awarded in March 1990, and eight others
                         had been advertised.

                         Page 8                                   GAO/RCEINO-177 Loma Prieta Earthquake


                          According to cal~ran~officials, seismic retrofitting work will no longer
                          compete with other projects for resources. To ensure that all retrofitting
                          work is completed by December 1993, calnan~ has shifted engineering
                          resources from the districts to headquarters to work solely on retrofit
                          projects. cal~ran~has also established a separate budget for seismic
                          retrofit projects and appointed a Seismic Retrofit Program Manager,
                          who will oversee the projects to ensure that all necessary work is
                          accomplished. Further, cal~rans is incorporating soil information into the
                          data base maintained on the state’s bridges, and is working on a plan to
                          incorporate the information into the seismic retrofitting program . In
                          addition, cal~ransis providing the California Transportation Commission
                          with monthly progress reports.

Funding I to Complete     cal~ran~new emphasis on retrofitting leaves open the question of where
                          it will find the funds necessary to complete its program . Although cal-
Retrofit Program Is Not   ~ransis responsible for identifying state highway projects and estab-
Assured                   lishing priorities, the California Transportation Commission determ ines
                          how available funds will be spent and which of cal~ran~’ proposals will
                          be approved for funding. As of May 1990, the Commission had not
                          determ ined how it would make funds available to supplement the $60
                          m illion appropriated by the legislature for seismic retrofitting work.

                          When the state legislature appropriated the $60 m illion from its emer-
                          gency fund, it required that these funds be used to match federal funds
                          to support the retrofit program . Commission officials anticipated that
                          the emergency relief funds appropriated by the Congress after the Loma
                          Prieta Earthquake could be used for this purpose. However, according to
                          FHKA officials, federal emergency relief funds cannot be used for routine
                          retrofitting work. Instead, the $60 m illion could be used to match fed-
                          eral funds from the state’s annual federal-aid highway allocation.
                          Assuming a 15-percent matching share, the $60 m illion state share could
                          provide a total of about $340 m illion in federal-aid highway funds for
                          the program . The proposed 1990-91 budget includes $291 m illion for
                          seismic retrofit projects: $25 m illion in state funds and $266 m illion in
                          federal funds. Because the costs to retrofit multicolumn bridges are
                          expected to be higher than those for single-column bridges, $400 m illion
                          may not be sufficient to complete all retrofitting work needed on the
                          state highway system.

                          In any case, an anticipated shortfall in California’s revenues could pre-
                          vent the state from completing many transportation projects, including
                          seismic retrofit work. State budget analysts have projected that by June

                          Page 9                                  GAO/RCJZD-90-177 Loma Prieta Earthquake

              1994, the cumulative shortfall in the transportation budget will reach
              $4.5 billion in state and federal funds. To generate additional funds, the
              governor and legislature have approved an increase in the state gasoline
              tax. This increase will take effect as a result of voters’ approval of a
              June 1990 ballot measure to increase the state’s spending limit, which
              was passed by the voters in 1979.4 Raising the cap will allow the state to
              spend the additional gasoline tax revenues on highway projects.

              But even if the additional state revenues are realized, funding the bal-
              ante from California’s normal federal-aid highway program would
              require a substantial reallocation of funds already planned for other
              construction projects. According to the California Transportation Com-
              mission, the proposed 1990-91 budget would use $266 million, or 40 per-
              cent, of the state’s federal-aid program allocation. To prevent the delay
              of other construction projects, the Commission has recommended that
              the legislature appropriate additional state funds to the retrofit pro-
              gram. These funds could come from the state’s general fund or from the
              emergency funds generated by the temporary sales tax. To date, the leg-
              islature has not acted on this request.

              Although California’s three-phase seismic retrofit program was started
Conclusions   18 years ago, only the first phase has been completed. CMTIWS gave
              higher priority to other highway projects considered more critical to
              highway safety. We recognize that making decisions regarding such pri-
              orities is difficult because of the number of important projects which
              compete for the funds available. However, had the retrofit program
              progressed to the third phase before the Loma Prieta Earthquake, cal-
              ~ransengineers believe they would have identified the flaw which they
              now think was a major contributing factor in the Cypress Viaduct’s col-
              lapse. With respect to the San Francisco/Oakland Bay Bridge, cal~rans
              officials thought that it had been sufficiently retrofitted prior to the
              earthquake and did not consider it susceptible to collapse.

              In the wake of the Loma Prieta Earthquake, California has begun to
              focus more attention on completing its retrofit program. The state legis-
              lature has appropriated some of the funds needed for this work, and
              cal~ran~has created a separate budget and dedicated staff specifically
              for this effort. These are the types of actions needed if the retrofit pro-
              gram is to be completed by 1993. Actions taken after both the 1971 San

              4The cap limits expenditures by stats and local governments to 1979 levels which have been @usted
              annually accmding to various factors, including changes in population and the cost of living.

              Page 10                                              GAO/lWEIMk177       Loma Prieta Earthquake


                  Fernando and the 1987 Whittier Earthquakes were not sufficient or sus-
                  tained long enough to ensure the program’s completion, Only time will
                  tell whether future actions will be sufficient to ensure that the retrofit
                  program is completed as soon as possible.

                  In commenting on our report, cal~ran~said that (1) our report should
Agency Comments   offer recommendations to help the Congress resolve the major policy
                  decisions regarding a nationwide seismic safety program for transporta-
                  tion; (2) California is committed to completing an aggressive, expedited
                  seismic retrofit program; and (3) the only outstanding question is the
                  federal government’s commitment to a comprehensive seismic retrofit
                  program. Of these issues, this report addresses only California’s commit-
                  ment to an expedited seismic retrofit program. We were not asked to
                  review the need for a nationwide seismic safety program or the federal
                  government’s commitment to California’s retrofit program. Specifically,
                  our work focused on what cal~ran~knew about the condition of the two
                  collapsed structures, what funds were available and have been spent on
                  seismic retrofit work in California, and what funds will be needed to
                  complete California’s seismic retrofit program.

                  We recognize that California has renewed its attention to seismic retrofit
                  work and has taken steps to complete it. However, past commitments
                  have not resulted in the actions needed to correct deficiencies identified
                  more than 18 years ago and, even with its current interest in the pro-
                  gram, the commitment of funds needed to complete the program remains
                  an outstanding issue.

                  With respect to cal~ran~’comment that the only outstanding question is
                  the commitment of the federal government, the federal government has
                  provided cal~ran~over $4 billion since 1975 for federal-aid highway pro-
                  grams that included seismic retrofit work as an eligible activity.
                  Although it is not realistic to suggest that all of these funds should have
                  been used for seismic retrofit, the fact that CalTram spent about 1 percent
                  of these funds, about $46 million, on seismic retrofit projects reflects
                  how the state set its priorities and commitment to seismic safety during
                  this time period. The complete text of cal~ran~’comments are contained
                  in appendix IV; our comments follow.

                  We also discussed the contents of this report with FWWA officials respon-
                  sible for seismic safety activities. They said that, in their view, the
                  report provides a fair and accurate account of the issues addressed. At
                  their suggestion, a few minor technical changes were incorporated

                  Page 11                                 GAO~CED-90-177   Loma Prieta Earthquake

where appropriate. Official comments from the U.S. Department of
Transportation were not obtained.

We conducted our work from October 1989 to May 1990 at the Cali-
fornia Department of Transportation and FHWA in Sacramento, Cali-
fornia, and Washington, D.C.; interviewed agency officials; and reviewed
official files, policies, and federal-aid highway program information. We
also reviewed reports and studies on the Loma Prieta Earthquake and
the resulting damage. We performed our work in accordance with gener-
ally accepted government auditing standards. Details of our objectives,
scope, and methodology are contained in appendix V.

As agreed with your offices, unless you publicly announce its contents
earlier, we will make no further distribution of the report until 30 days
from the date of this letter. At that time, we will send copies to other
interested congressional committees, the Department of Transportation,
and the state of California. We will also make copies available to inter-
ested parties upon request. If you or your staff have any questions on
this report, please contact me on (415) 556-6200 or Kenneth M. Mead,
Director of Transportation Issues, on (202) 275-1000. Other major con-
tributors to this report are listed in appendix VI.

Thomas P. McCormick
Regional Manager

Page 12                                 GAO/RCElHM77    Loma Prieta Earthquake

                                                               .;   i

                                                       i                 :
      Page 13   GAO/BCED&%l1?   LomaPlietaFhrthquake       .

Appendix I                                                                                             16
Pictures and
Illustrations of
Cypress Viaduct and
Bay Bridge Damage
Appendix II
California’s Seismic
Retrofit Program
Appendix III                                                                                           23
Earthquake Force
Appendix IV                                                                                            24
Comments From the        GAO Comments                                                                  32
California Department
of Transportation
Appendix V
Objectives, Scope, and
Appendix VI                                                                                             36
Major Contributors to
This Report
Figures                  Figure 1.1: Cypress Viaduct Prior to Loma Prieta                               16
                             Earthquake of October 17,1989
                         Figure 1.2: Illustration Depicting Reinforcement in                            17
                             Cypress Viaduct Columns

                         Page 14                                 GAO/ICED-90-177   Loma Prieta Earthquake

Figure 1.3: Illustration Depicting Typical Cypress Viaduct                   18
    Support Column Failure
Figure 1.4: Cypress Viaduct After October 17, 1989,                          19
Figure 1.5: Cypress Viaduct’s Upper Deck Road Surface                        20
    After October 17, 1989, Collapse
Figure 1.6: Collapsed Section of the San Francisco/                          21
     Oakland Bay Bridge


cal~ran~ California Department of Transportation
FHSVA    Federal Highway Administration
RCED     Resources, Community, and Economic Development Division

 Page 16                                GAO/WED-90-177   Lana Prieta Earthquake
pictures and Illustrations of Cypress Viaduct
axd Bay Bridge Damage

Figure 1.1: Cypress Viaduct Prior to Loma Prieta Earthquake of October 17,1999

                                           Source Cahfornla Department of Transportation

                                           Page 16                                         GAO/RCED-96-177   Loma Prieta Earthquake
                                     Pictnres and Dlustrations   of Cypress viaduct
                                     and Bay Bridge Damage

Figure 1.2: Illustration Depicting
                                                                                        Upper Deck
Reinforcement in Cypress Viaduct

                                                                                        \\ 3      \


                                     Source: California Department of Transportation.

                                     Page 17                                            GAO/IWED~l77       Loma Prleta Jkthqnake

                                             Appendix I
                                             Pictures and Dlnstmtiom     of Cypress Viaduct
                                             and Bay Bridge Damage

Figure 1.3: Illustration Depicting Typical
Cypress Viaduct Support Column Failure

                                                                                                                         2. Upper deck
                                                       tI         Upper Deck
                                                                                                                            drops down.


                                                                                                                         1. Column drops
                                                                                                                            down and out
                                                       *1         Lower Deck

                                             Source: California Department of Transportation.

                                             Page 18                                                    GAO/BCED-9blT7     Loma Prieta Earthquake
                                      Appendix I
                                      Pictures and lllustratioua   of Cypress Viaduct
                                      and Bay Bridge Damage

                                      Source: Callfornla Department of Transportation

                                      Page 20                                           GAO/RCED~177   Lmua Prieta Earthquake

~-   -..-   -.-- q, -,.-   _ .--_---.----         ---     - -
                                           Appendix I
                                           Pictures and Illustrations       of Cypress Viiuct
                                           and Bay Bridge Damage

Figure 1.4: Cypress Viaduct After October 17,1989, Collapse

                                                                                                                   I”          ”


                                                              *   1     I   ’

                                           Source: Cahfornla Department of Transportation

                                           Page 19                                              GAO/RCED-W177   Loma Prieta Earthquake
                                            Appendix I
                                            Pictures and Illustrations   of Cypress Viaduct
                                            and Bay Bridge Damage

Figure 1.8: Collapsed Section of the San Francisco/Oakland      Bay Bridge

                                            Source California Department of Transportation

                                            Page 21                                           GAO/WED-90-177   Loma Prieta Earthquake
Appendix II

California’s SeismicRetrofit Program                                                           c

                  The need for seismic retrofitting became apparent when the 1971 San
                  Fernando Earthquake severely damaged five bridges in Southern Cali-
                  fornia. As a result, California began what evolved into a three-phase
                  seismic retrofit program to strengthen structures located in areas of fre-
                  quent seismic activity to prevent similar damage from future earth-
                  quakes Three types of weaknesses were identified for correction:

              . Deck sections that were not adequately tied together at the hinges had a
                tendency to separate and collapse.
              . Single columns that supported some structures were likely to suffer
                damage and possible collapse because of insufficient reinforcement,
                which allowed concrete in the columns to crumble.
              l Multicolumn structures, including double-deck structures such as the
                Cypress Viaduct (see fig. I. 1), were also considered vulnerable due to
                insufficient reinforcement. But because of the additional stability pro-
                vided by multiple support columns, they were not considered as vulner-
                able as single-column structures. Before the Loma Prieta Earthquake, no
                multicolumn bridge had collapsed during an earthquake.

                  Correcting deck weaknesses became the first phase because the Cali-
                  fornia Department of Transportation (cal~ran~)believed these weak-
                  nesses were both more serious and less expensive to correct. cal~ran~
                  planned to reinforce the hinges between deck sections by tying them
                  together with cables or other reinforcers. This work was completed on
                  the Cypress Viaduct in 1977 and on a total of 1,261 structures by 1989.

                  Next, cal~ran~planned to strengthen single-column structures by
                  encasing the concrete columns in steel jackets. Although column damage
                  was originally observed in the 1971 San Fernando Earthquake, the
                  October 1987 Whittier Earthquake caused further bridge damage, rein-
                  forcing the need to strengthen columns on bridges built before 1971. The
                  California Transportation Commission approved csrrans request to
                  spend $64 million on these projects in December 1987. Research began in
                  1987, and the first contract was awarded in March 1990. cal~ran~plans to
                  have this phase completed by December 1991.

                  Finally, cal~ran~planned to retrofit multicolumn structures-the   third
                  phase. This phase, which included the Cypress Viaduct, had not been
                  scheduled at the time of the earthquake. The technology needed to
                  retrofit multicolumn structures is being researched and is not yet fully
                  developed. cal~ran~plans to have the third phase completed by December

                  Page 22                                 GAO/RCED-90-177   Loma Prieta Ehthquake
Appendix III

Earthquake Force Measurements

               Earthquake effects can generally be categorized by two measurements.
               The Richter scale measures the energy release of an earthquake at its
               epicenter, while ground acceleration measures the gravity, or “g” force,
               of the ground-shaking motions. The g force can vary considerably
               according to the type of ground material involved and is more mean-
               ingful than Richter scale measurements to engineers designing bridges
               or buildings. The Loma Prieta Earthquake measured 7.1 on the Richter
               scale at its epicenter, near Loma Prieta peak in the Santa Cruz Moun-
               tains, where the ground acceleration ranged from 0.47 g to 0.64 g. No
               measurement devices were placed on the Bay Bridge or the Cypress Via-
               duct, about 60 miles from the epicenter. Devices near the Cypress Via-
               duct measured the ground acceleration at between 0.26 g and 0.29 g.
               This is in contrast to measurements of 0.11 g and 0.13 g in San Jose,
               located between the epicenter and the Cypress Viaduct, and 0.06 g at
               Yerba Buena Island, which anchors the middle of the San Francisco/
               Oakland Bay Bridge.

               Geologists attribute the variations in ground acceleration to differences
               in soil composition; soft soils experience stronger ground shaking than
               hard soils. The collapsed sections of the Cypress Viaduct and the Bay
               Bridge were built on relatively soft soils, while Yerba Buena Island is on
               bedrock. The effects of earthquakes on soft soils were not fully realized
               by Caltrans until 1985, when the Mexico City Earthquake caused severe
               ground shaking and liquefaction, the loss of ground strength when
               loosely compacted, water-saturated sediments liquefy, in parts of the
               city built on landfill.

               Page 23                                 GAO/RCED-3&177   Loma Prieta Ihrthqnake
Appendix IV

Comments From the California Depaxbnent                                                                                                                   *
of Transportation

Note: GAO comments
supplementing those In the
report text appear at the    ATE OF CALIFORNIA-BUSINESS.       TRANSPORTAnON   AND HOWNG   AGENCY                            GEORGE DEUKMEJLAN.       Garamor

end of this appendix
                                     OF TRANSPORTATION
                             WCE OF THE DIRECTOR
                             I20 N STREET
                             0 BOX PA2873
                             KRAMENTO    CAUFORMA      94273    0001

                                     May 16,          1990

                                     Thomas P. McCormick
                                     Regional Audit Manager
                                     General Accounting  Office
                                     Suite 900, State Fund Building
                                     1275 Market Street
                                     San Francisco,   CA 94103
                                     GAO Report                on the      Loma Prieta          Earthquake

                                     After      the    Loma Prieta       earthquake     on October     17,    1989, the
                                     California        Department     of Transportation     (Caltrans)     welcomed the
                                     General Accounting         Office auditors     in the hope that they would make
                                     recommendations         to Congress on the appropriate            way to address
                                     seismic       safety    concerns    as they relate      to the transportation
                                     We believe      the report should offer recommendations               to help Congress
                                     resolve     the major policy        decisions     regarding    a nationwide         seismic
                                     safety      program     for    transportation.        Unfortunately,          the report
                                     suggests       by implication         that     the Caltrans       program,         although
See comment 1                        acknowledged to be the leader in the country,                   is lacking       in scope,
                                     timing,      funding,    priorities       and commitment.        Additionally,         other
                                     than to note the most obvious of facts that the advanced California
                                     program did not prevent the severe damage experienced                         in the Loma
                                     Prieta    earthquake,       the report provides        absolutely      no guidance for
                                     California,        for other states        or for the nation       as a whole on the
                                     development of an appropriate             seismic retrofit      program. No existing
                                     standards       for such a program are identified,                no comparisons          are
                                     drawn and no suggestions             are made for the development               of needed
                                     Contrary  to comments                       repeatedly  made in the report,     California   is
                                     committed to complete                       a most aggressive,  expedited   seismic retrofit
See comment 2                        program.   It will   be                      completed  within a few years at a cost of
                                     hundreds of millions                        of dollars.
                                     The a         outstanding     question      is the commitment of the federal
                                     government.      Will    it persist     in its current         policy   of hiding   from
                                     the problem and requiring             the normal federal           aid programs to be
                                     diverted     to the seismic safety          effort?        If it does, California's
See comment 3                        no--ma1 federal-aid          program      will      suffer     very    seriously,     but
                                     California's       seismic   retrofit     program will        be completed.

                                                   Page24                                                    GAO/RCED-90.177LomaPrietaEarthqualce
                        Comments From the California Department
                        of Transportation

                Mr. McCormick
                May 16, 1990
                Page 2

                What does the federal         policy    of ignoring        the issue mean to other
                states?     Should they likewise       continue       to ignore the issue? Should
                they     shut     down their     normal      federal-aid          programs      to pursue
                aggressive      seismic retrofit     efforts     like California's?           Should they
                take something         from the federal-aid            programs      to start      a well-
                planned      long range retrofit         program at a level              that     does not
                interfere      with the basic priorities           (i.e.    traffic     safety)      of the
                federal-aid       program?
                The report provides          no guidance on the most important,           fundamental
                questions.        In fact,    it does a disservice       by refusing    to recognize
                how complex and difficult            these public     policy    choices are, and by
                criticizing         by implication      the nation's        most advanced seismic
See comment 4   retrofit       program because it tried        to make those choices and tried
                to balance,        within   constrained    funding seismic retrofit          work with
                other       safety      and performance       objectives      of the      federal-aid
                California      has been a national       and international       leader  in the
                structural     and seismic research area for many years. Seismic design
                criteria     used by the other       49 states    have come from Caltransl
                advanced seismic research and design programs. The GAO report does
                not explain     the background and reasoning       for decisions     made during
                the past 20 years regarding         the seismic     strengthening     of bridges
                in California.     We believe  this    document should provide a framework
                by which the Congress of the United States can develop guidelines
                to help the states adequately         deal with the seismic safety problem
                across the country.
                I would suggest the following          points must be considered so that the
                Congressional       Oversight    Committees can make informed decisions     in
                developing     directions     to improve the seismic safety of our nation's
                transportation        system.
                        * Over the years,           California      - and to          the best of our
                        knowledge every other state has followed                   the principle        that
                        we could expect          and accept      earthquake        damage on highway
                        structures     but not accept          collapse     that       could     result     in
                        serious    injury      or death.       This policy         was based on the
                        national      and     international         research,          state-of-the-art
                        technology     and the knowledge gained by                 Caltrans       engineers
                        from actual       experiences       in earthquakes          in California         and
                        around the world.
                        * Prior to 1971, the department had developed a standardized
                        design     based on the available           knowledge     of the time.
                        Structural     engineering      is a dynamic discipline,       constantly
                        evolving    and changing.      Advances in technology       and knowledge
                        are continually       incorporated    into designs.     This applies      to
                        both buildings      and highway structures.      A building    or highway

                        Page 25                                            GAO/RCED-90-177LomaPrieta Earthquake

Mr. McCormick
May 16, 1990
Page 3

      bridge designed in 1990 contains seismic             design standards and
      features that are not found in structures              20, 30 or 50 years
      * The 1971 San Fernando earthquake                   was a watershed       event
      which Caltrans         engineers used to expand their           knowledge base
      and to make improvements               to structures       with the goal of
      preventing       collapse.     In this earthquake,       columns experienced
      unanticipated         damage and five        bridges     collapsed     when the
      decks were pulled           off their    supports.     From that quake, new
      design     criteria      were developed       to strengthen        columns and
      special     priority     was given to a retrofit          program to prevent
      collapse       by securing       bridge decks to their         support columns
      through the use of hinge restrainers.
      * Factual evidence from subsequent earthquakes,    including   the
      1987 Whittier  Narrows event, confirmed that hinge restrainers
      should be the top priority    in a retrofit program.
      * After     the 1987 Whittier         Narrows quake, the department
      accelerated      its research     program into the strengthening       of
      columns, with an emphasis on single             column structures,   as a
      way to minimize         damage so repairs         could be made without
      closing    vital    thoroughfares     to traffic.      This approach was
      confirmed     by the 1988 earthquake        in Armenia where thousands
      died when emergency vehicles            were unable to reach victims
      because the road system was unusable.
      * California's        seismic program has been a major undertaking.
      It involved       extensive    new research,    review and screening  of
      thousands      of structures       and thousands     of hours of work by
      Caltrans     staff     and consultants.
      * Multi-column    structures  have as a class performed very well
      in quakes prior      to Loma Prieta,  including  San Fernando and
      Whittier.    There had been some damage, but no collapse.
      * Loma Prieta    produced a number of expected and unexpected
      occurrences.    Hinge restrainers      were effective      and the new
      column design criteria    incorporated     into structures      designed
      after   1971 performed   as intended.      As expected,      repairable
      damage occurred    to some structures.      Collapse   of the Cypress
      Viaduct and the section     of the Bay Bridge was not expected.
      * The Cypress Viaduct        was designed    and constructed   to a
      strength   exceeding   the standards    of the day. The structure
      was built     to those     standards   and was meticulously      and
      properly   maintained    over the 32 years since its opening in
      1957. The same is true for the San Francisco-Oakland             Bay
      Bridge,   which was opened to traffic      in 1936.

       Page26                                          GAO/RCED-90-177LomaPrietaEarthquake
                       Appendix N
                       Comments From the California Department
                       of Transportation

                Mr. McCormick
                May 16, 1990
                Page 4

                       * Caltrans       engineers     were unaware of the potential            for
                       collapse    at Cypress. There was no indication        that the Cypress
                       Viaduct was uniquely        vulnerable.  Of course with the benefit
                       of hindsight,      we have been able to identify        a design detail
                       that contributed       in combination   with  other factors        to this
                       catastrophe.      But before the October 17 earthquake,         there was
                       absolutely       no evidence       that would    have      prompted     the
                       department     to conduct an exceptional      detailed      study of the
                       * As a result    of the Loma Prieta      Earthquake,     Caltrans               has
                       taken the following    steps:
                       1. Examine all bridges    in California     for their   potential                for
                       vulnerability   in an earthquake.
                       2. Embark on an accelerated    seismic retrofit       program for               all
                       3. Develop a response spectra       for soft soils.
                We believe    that the installation        of hinge restrainers                 kept many
                bridges from collapsing        during earthquakes    in California             since 1971
                including    Whittier     and Loma Prieta,    saving hundreds of              lives.    Not
                only did the hinge restrainers         keep the structures      from          collapsing
                on vehicles,    but they also helped keep these structures                    intact    and
                sehriceable,      allowing    emergency vehicles      to get aid             to damaged
                areas of the community.
                Were the Cypress and Bay Bridge strong      enough to withstand     the
See comment 5   forces   experienced during this   earthquake?    Obviously   not.    To
                suggest that there was a "design   flaw"   or that Caltrans    ignored
                the problem is not backed up by the facts.
                The GAO report      has three     major    areas   that    need clarification:
                       1. SCOPE OF THE RETROFIT PROGRAM
                       Following      the 1971 earthquake,          the department       embarked on a
                       seismic     retrofit        program     focused      on efforts      to prevent
                       collapse      of structures.          The problem      at San Fernando was
                       corrected      through the use of hinge restrainers.                Further,       it
                       was identified         that    additional       column strengthening          would
                       help     reduce      damage and not            require     closure     of vital
                       thoroughfares         while     repairs      were made. Only after               the
See comment 6          Whittier      earthquake       did we make any distinction                between
                       single and multi-column           bridges.      That distinction      was driven
                       by the state of seismic research which offered                    some possible
                       approaches that could be tested and used to retrofit                        single
                       column bridges.        Even then, there was no technology              available
                       for retrofitting         multi-column       bridges.
                       There was NO evidence or indication    prior to October 17, 1989
See comment 7          that   Cypress    was  uniquely     vulnerable    to   collapse.

                       Page 27                                            GAO/RCED-90-177 Loma Prieta Earthquake

                Mr. McCormick
                May 16, 1990
                Page   5

                           Consequently,     there was no reason to undertake        a study in
                           which Caltrans     engineers   might have noticed   the structure's
                           unique design detail      or that the amount of reinforcing      steel
                           in the columns was less than current       requirements.    In fact,
                           hinge restrainers      had been installed   at Cypress to correct
                           the problem that had been well documented at San Fernando and
                           subsequent earthquakes.
                           The question       of how best to use the resources             to address the
                           wide range of problems on our transportation                  system presents
                           a most complex issue of public             choice.      How much is enough
                           for seismic       research   and retrofit     when (a) You are already
                           addressing      the vulnerability       of bridges       slipping     off their
See comment 8              supports     that had been identified         at San Fernando,           (b) You
                           have already       updated   your design criteria           and (c) There is
                           no indication         of the possibility        that     any structure         was
                           vulnerable     to collapse     due to column failure.           In California,
                           the retrofit       work to prevent collapse          (for user safety)         had
                           been completed.        The choice facing California,             based on what
                           was known prior        to the Loma Prieta      earthquake,        was a choice
                           between investment         to provide      new services          and increased
                           operational       safety   or investments      to protect         the services
                           relying     on older facilities.
                The Loma Prieta         Earthquake was a tragedy,         but it would be an even
                greater    disaster       if we fail    to learn from this experience,          expand
                our knowledge and make improvements to the nation's                 transportation
                system.      To reach         that    goal,    the   California     Department       of
                Transportation         believes    the Congress needs to provide        guidance on
                funding     priorities        and the scope of the program             needed on a
                nationwide      basis to strengthen          the transportation      system in the
                event of a major earthquake.                Earthquakes     are more than just        a
                California      phenomenon. They are a national            problem.
                I have attached          some additional  comments to            the GAO report  which
                I believe   will         help the Congress deal with              this very important

                            Page28                                           GAO/RCED-90-177LomaPrietaEarthquake

                                   GAO REPORT ON THE MMA PRIETA EARTHQUAKE

                 The following       specific     comments regarding            the text       of the report
                 are provided       in    addition    to the points             presented       in the cover
                 Pages 1 and 2 regarding              "RESULTS IN BRIEF"
                       For all the reasons discussed in our cover letter    we feel                             your
                       summary does not fairly   present the State's  program.
                 Page 4,     the first         full     paragraph      is     incorrect      and     should        be
                 rewritten    as follows:
                       In California,          Caltrans     prepares        a multi-year          estimate       of
                       state     and federal        funds projected          to be available            for its
                       capital       program,        and    forwards         it       to    the     California
                       Transportation         Commission, a board appointed                 by the Governor,
                       for review and approval.             The Commission reviews and approves
                       the fund estimate.           Based on the final           fund estimate,        Caltrans
                       and the several           regional      transportation            planning      agencies
See comment 9          nominate      projects      for inclusion          in a State          Transportation
                       Improvement Program (STIP), which lists                     specific     projects,      and
                       submits it to the Commission for approval.                         Starting     in 1990,
                       Caltrans      also prepares       and submits a separate               Highway System
                       Operation       and Protection      Plan (HSOPP), which includes                  safety,
                       rehabilitation,         and operational         projects        on the state highway
                       system.          An annual      budget      proposal         is also prepared             by
                       Caltrans,          and submitted          through        the      Governor       to     the
                       legislature,        which appropriates         funds for the capital            programs
                       set forth         in the STIP and HSOPP. However,                      funds are not
                       allocated       to individual      projects      by the Commission until              they
                       are ready to be advertised              for bid.
                 Page 4, middle      paragraph        which begins     with    "Earthquake       effects...":
                       Recommend consistent      terminology      for reference   to "gl' force.
                       For example, in fifth       and sixth lines      from end of paragraph,
                       reference  is made (first)      to "....between      .26 and -29 percent
See comment 10         of gravity   . ..'I and (second) to II....      measurements of .ll and
                       .13 in San Jose . . ..'I    We believe     the convention    would be -26
                       gf .29 g, .ll g, and .13 g respectively.
                 Pages 5 and 6, beginning              with   "Correcting       deck weaknesses          . ..'I.
                       A major factor     in the decision     to give the "hinge restrainer"
                       phase highest      priority      was that    there   was no available
                       technology     for either     single  - or multiple-column    retrofit
                       strengthening.      When the "Phase II" program was presented          to
See comment 11         CTC in 1987, it was specifically         made clear to the Commission
                       that the technology        did not exist   even then.


                        Page29                                                 GAO/RCED-90-177LomaPrietaEarthquake

                         Appendix N

                       The statement that "... no work had been done as of April 6,
                       1990, on the 392 single-column              structures..."        is wrong.
                       Clearly,    as the balance of the paragraph indicates,                     both
                       research     and design work had in fact              been done on such
                       structures,     granted that no actual construction             contract was
                       yet underway. In addition,          the Department has never withheld
                       from allocation      or advertising      u   retrofit     project on which
                       design was complete.          The statement that I*... some (single-
                       column retrofit)      project designs had been completed by June,
                       1988, but other projects were given priority                in funding" is
                       wrong.      The statement on pp lo-11 that "...some                    project
See comment 12.        designs were completed by June 1988, but the districts                 did not
                       complete the site work needed to forward the projects                       for
                       funding approval....      (I helps to clarify     what the auditors were
                       told,    but is still          incorrect.        In these cases, the
                       nstructuresn      element of some project designs may have been
                       considered complete, but the final Plans, Specifications,                   and
                       Estimates needed to advertise the projects were not complete
                       with respect to items such as utility              relocations,      right of
                       way acquisition,      permits, etc., and the projects could neither
                       be presented to the CTC for funding,                  nor advertised        for
                  Page 8, "Funding has been available           in the past     . . ..I'.
                       The statement "although California         began its seismic retrofit
                       program in 1971, it did not begin earmarking funds for this
                       purpose until      1987" is incorrect        on two counts.      First,
                       California   does not normally t@earmarklVsny funds. In 1987 the
                       CTC simply accepted the Department's                recommendation to
                       establish a lump-sum reservation         as a matter of convenience,
                       because the specific      individual    projects needed to implement
                       the program had still      not been identified     (in part, at least,
See comment 13.
                       because the appropriate        retrofit   technology hadn't yet been
                       identified).     Second, nmajortl seismic retrofit       projects were
                       in fact included in STIP's prior to 1987, either as part of
                       reservations   or as line items projects ("minor" projects, with
                       a few exceptions,     have never been included in the STIP in any
                       program area).       Through the 1986/87 FY, 1260 bridges were
                       retrofitted    with hinge restrainers         at a cost of about $54
                  Page 9, "California     has not earmarked funds..."
                       Again, this paragraph is incorrect,          for the reasons given
                       above regarding the similar reference on page 8. Furthermore,
                       the sentence "Funding for each project              was taken from
See comment 14.        Caltrans' State Transportation       Improvement Program: funds for
                       the seismic retrofit      program were not specified      within the
                       budget" implies that funds for other kinds of projects m
                       specified   in the budget.     In fact, with very few exceptions,
                       the budget nev~l specifies      funding for individual   projects or
                       even project types, and until 1990 funding for every project,
                       not just seismic retrofit     projects,   was "taken from" the STIP.

                         Page 30                                       GAO/ICED-!40-177 Loma Prieta Earthquake

                         Appendix IV
                         Chnmenta kom the California Department
                         of Transportation

                       (Now, because of recent changes in statute,  funds are "taken
                       from" the STIP, the HSOPP, the TSM Plan, etc.)
                  Page 10, second para under "Competing priorities            . ..I*.
                       The statement that II... the districts     received no additional
                       resources specifically    for retrofitting    work" is incorrect.
                       The resources assigned to the districts         were based on the
                       project workload from the STIP, which included the seismic
                       retrofit  program.     In fact, where the capital      program is
                       concerned, the districts     have not in recent years received
                       support resources snecificallv     for any project type.
                       The statement that (I... safety and rehabilitation               projects
                       have been assigned priority        based on the number of lives lost
                       in previous incidents!"          and the attribution        to Valtrans
                       officialstl   are misleading.       Accident histories     are m basis
See comment 15.        for preliminarily      identifying    potential    project locations in
                       certain     categories     of lVsafety8* projects.         Likewise    the
                       conclusion that "retrofit        projects were accorded a relatively
                       low priority     because . . . earthquake damage to state bridges
                       had resulted in the loss of only two lives...tt            is misleading
                       in its      suggestion of a single,           direct    cause-and-effect
                       We believe it would be more accurate to say the Department
                       considered there was no appreciable risk of loss of life due
                       to catastrophic   failure  of bridges in any maximum credible
                       earthquake, particularly    once the hinge restrainers   had been
                       installed.    To the extent guardrails,    median barriers,    and
                       other safety projects     were accorded high priority,      it was
                       because of the certainty    that such projects would reduce the
                       future loss of life.
                  Page 11, "Funding for all      retrofitting     work is uncertain":
                       Third and Fourth lines of first       paragraph, the CTC does m
                       U1control Caltrans' overall budget."        The audit report needs
                       to reflect   the distinction    between the program, the budget,
                       and fund allocations.       While the CTC does, indeed determine
                       which projects    are included in the STIP (and, now, HSOPP,
See comment 16.
                       which is the relevant document where the retrofit        program is
                       concerned), the budaet is determined by the Legislature         and
                       Governor.     The CTC does also allocate         budgeted funds to
                       specific projects, including retrofit      projects, but within the
                       controls and constraints     contained in the budget.      In fact,
                       the CTC can only allocate            funds appropriated     by the
                       Legislature,   which considerably limits the options to
                       determine *I... how it would make funds available to supplement
                       the $60 million    appropriated by the legislature      for seismic
                       retrofitting   work."


                         Page 31                                      GAO/RCED-90-177 Loma Prieta Ebthquake
               Appendix IV
               Comments From the California   Department
               of Transportation

               The following are      GAO'S   comments on cal~ran~’
                                                                  letter dated May 16,199O.

               1. We were asked to examine what cal~ran~knew about the condition of
GAO Comments   the two collapsed structures, what funds were available and have been
               spent on seismic retrofit work in California, and what funds will be
               needed to complete California’s seismic retrofit program. We were not
               asked to address safety concerns for the nation as a whole or to deter-
               mine the adequacy of seismic safety standards endorsed by the federal

               2. In the past, cal~rans’commitment has not been sufficient to ensure
               completion of the seismic retrofit program. Although cd~rans has
               renewed its commitment, it will not be fulfilled unless funds are diverted
               from other projects or the legislature appropriates additional funds.

               3. cal~ran~has received over $4 billion since 1975 for federal-aid
               highway programs that included seismic retrofit work as an eligible
               activity. The state chose to spend about $46 million-about      1 percent-
               of this on seismic retrofit projects, reflecting state-determined priorities.
               In addition, if the state had initiated column retrofit research and per-
               formed some of the column retrofit work during the last 18 years, the
               impact on any single year’s budget would have been considerably less
               than it will be between now and the end of 1993.

               4. We recognize that California made choices between retrofit work and
               other safety programs. However, we found that when these choices
               were made, seismic retrofit work received relatively low priority.

               6. We do not say nor do we suggest that cal~ransknew that the Cypress
               Viaduct contained a flaw or that they ignored it. However, we revised
               the report language to use the term “weakness” to describe the hidden
               defect which, when placed under stress, had the potential to cause col-
               lapse. We did not evaluate whether the Cypress Viaduct met the design
               standards and codes of its day but, as we point out, cal~ransengineers
               now believe this design detail was a major factor in the structure’s

               6. Only one of the problems identified at San Fernando was corrected
               through the use of hinge restrainers. The potential for column damage
               and structural failure was identified, according to ca.mans, in 1971 and
               reemphasized in 1987. In contrast to cal~rans’ assertion that no distinc-
               tion was made between single-column and multicolumn bridges until
               after 1987, a 1978 cal~ran~publication describing the 1971 earthquake’s

               Page 32                                      GAO/RCED-90-177   Loma Prieta Earthquake
Appendix Iv
Comments Fkom the California   Department
of Transportation

effects portrayed single-column bridges as being “particularly” vulner-
able to damage. Further, C~~TKUIS officials considered multicolumn
bridges as less vulnerable because of the additional stability provided by
the additional columns. The technology to retrofit multicolumn bridges
is not available.

7. We do not say that cal~ransengineers knew that the Cypress Viaduct
was uniquely vulnerable to collapse or had any reason to specifically
target it for review. cal~ran~engineers told us, however, that if the pro-
gram had progressed to the third phase, engineers reviewing the struc-
ture’s plans would have discovered the column weakness.

8. Retrofit work was not finished with the completion of phase one in
1989; cal~ransengineers had long recognized the need to retrofit bridge
support columns and had in fact planned to deal with these deficiencies
in phases two and three of the retrofit program.

9. This paragraph was deleted from the final report because it was not
considered necessary to include a detailed discussion of California’s pro-
cess for funding transportation projects in order to understand that the
California Transportation Commission approved cal~ran~’   funding
request for the seismic retrofit program.

10. It was decided that the discussion of the measurement of earthquake
forces would best be presented as an appendix to the report. Therefore,
this information is now incorporated as appendix III to the report. Per
cal~ran~ suggestion, references to the percentage of gravity were changed
to “g” force for consistency.

11. To say that the technology to retrofit either single- or multicolumn
structures was not available in 1971, when the weaknesses were first
identified, or in 1987, when the phase two funding program was
presented to the California Transportation Commission, ignores the fact
that cal~ranswas responsible for developing the needed technology.
Although a 1978 cal~ranspublication stated that a contract would be
awarded for column retrofit research in “the near future,” the contract
was not awarded until 1987.

12. We did not find that the Department withheld from allocation or
advertising any retrofit project on which design was complete. However,
we were told by several cal~rans engineering officials that retrofit

Page 33                                     GAO/RCXZIM@177   Loma Meta   Edhcwke
Appendix IV
Commenta Prom the California   Department
of Transportation

projects were not finalized, even though the structural designs were fin-
ished, because funds were not available to perform the construction

13. This paragraph was deleted from the final report because it dupli-
cated information provided later in the report on funds provided for the
seismic retrofit program.

14. This paragraph was revised to delete the reference to the
earmarking of funds for the seismic retrofit program and to eliminate
the implication that other types of projects were specified in the budget.

15. Although C~~TIXIIS uses a formula, which includes lives lost as well as
other elements, to determine how projects in the state transportation
plan will be ranked, cal~rans officials told us that the determining factor
was the lives lost element.

16. C~~TXUIS budget is in fact determined by the legislature and approved
by the governor, but the responsibility for allocating budget funds
among projects has been delegated to the California Transportation
Commission. As a result, if the legislature does not decide to increase
C~~TWIS’  budget to complete seismic retrofit work, the Commission will
need to decide whether to reallocate funds already dedicated to other

Page 34                                     GAO/lEEIM@177   Loma Prieta JSuthquake
Objectives,Scope,and Methodology

              On October 23,1989, the House Appropriations Committee, in House ’
              Report 101-301, Further Continuing Appropriations, 1990, requested
              that we provide information on the collapse of the Cypress Viaduct sec-
              tion of Interstate 880 and the San Francisco/Oakland Ray Bridge during
              the October 17,1989, Loma Prieta Earthquake. In accordance with sub-
              sequent agreements with the Committee, our objectives were to deter-
              mine (1) what the California Department of Transportation (cal~rans)
              knew about the structures’ vulnerability to earthquake forces before
              they collapsed, (2) what federal and state funding was available and
              expended in California to strengthen bridges subject to earthquake
              forces, and (3) what funding is needed to complete California’s seismic
              retrofit program.

              To accomplish our objectives, we interviewed officials from cal~rans;the
              Federal Highway Administration in Washington, D.C., and Sacramento;
              and the California legislature, as well as various experts within the
              seismic engineering community. We attended all sessions of the Gov-
              ernor’s Panel on the Loma Prieta Earthquake, at which a broad spec-
              trum of engineers, seismologists, geologists, and budget experts
              presented information on the Bay Bridge and Cypress collapse, the
              retrofit program, and the California and federal budget processes. We
              reviewed maintenance files for the two failed structures, state and fed-
              eral budget documents, program priority policies and plans, and federal-
              aid highway program information.

              We did not independently evaluate the design or construction plans to
              determine the structural adequacy of the Cypress Viaduct or the San
              Francisco/Oakland Bay Bridge or the technical reasons for the struc-
              tures’ collapse. Rather, we relied on cal~ran~’
                                                            information and opinions,
              views of experts presenting information to the Governor’s panel, and
              various reports and studies published after the earthquake.

              Page 35                                 GAO/IEEMO-177   Loma Prieta Earthquake
Appendix VI

Major Contributors to This Report

                        John W. Hill, Associate Director, Transportation Issues
Resources,              Jacquelyn L. Williams-Bridgers, Assistant Director
Community, and          Benjamin E. Worrell, Assignment Manager
                        Steven L. Cohen, Senior Evaluator
Development Division,
Washington, D.C.

                        Karen S. Zuckerstein, Assistant Regional Manager
San Francisco’          Larry J. Calhoun, Issue Area Manager
  V      Office         Mary C. Bufkin-Smith, Evaluator-in-Charge
                        Judy K. Hoovler, Evaluator
                        Belinda F. Jones, Evaluator
                        Jonathan M. Silverman, Writer/Editor

(342807)                Page 36                                GAO/IWEB~l77   Lmna Prieta Jhrthquake