COMMITMENT • INTEGRITY • LEADERSHIP	View full PDF report
Toll Bridge Seismic Retrofit Program The State Could Save Millions of Dollars Annually by Implementing Lessons Learned Report Number: 2018-104

Audit Results

Lessons Learned From the Seismic Program Could Improve the State's Oversight of Future Projects

The Legislature's decision to shift oversight of the seismic program from Caltrans to the Oversight Committee had positive results. The Oversight Committee provided examples of $866 million in cost avoidance and savings due to its oversight and the use of risk management to inform decisions. Further, the use of risk management resulted in avoiding seven years of potential delays. The seismic program is one of the largest transportation projects in California, and the experience of the Oversight Committee shows that similar oversight of the hundreds of billions of dollars in transportation projects that state and local agencies are either engaged in or anticipating can produce significant benefits.

When the Legislature formed the Oversight Committee in 2005, it also required Caltrans to develop a comprehensive risk management plan for the seismic program. Since then, Caltrans has adopted a policy to make risk management a part of all of its major construction and maintenance projects. Conversely, although MTC has taken a more active role in managing regional transportation projects, it has not yet created a policy to ensure that future projects it directs will benefit from appropriate levels of risk management. While MTC has stated it will create such a policy, there is little to ensure the application of risk management statewide because state legislation does not impose a general requirement on all sponsors of state or locally funded major transportation infrastructure projects to develop risk management plans.

The Oversight Committee's Actions Curtailed the Soaring Costs of the Seismic Program

The Oversight Committee's involvement curbed cost overruns on the seismic program. As we discuss in the Introduction, the Legislature created the Oversight Committee so that it would provide oversight and management to the seismic program. Before the establishment of the Oversight Committee, the seismic program experienced significant cost escalations. However, after the Legislature established the Oversight Committee, the budget for the program remained relatively constant—in fact, the Oversight Committee forecasts that the seismic program will complete its work roughly on budget.

Members of the Oversight Committee provided examples of savings of at least $505 million resulting from its actions and decisions. State law requires it to resolve project issues, perform risk assessments and monitor staffing levels among other duties. According to members of the Oversight Committee, it achieved these savings in part by combining project responsibility with approval authority and by incorporating the multiple viewpoints the committee's members represent. Table 2 provides examples of cost avoidance and savings achieved through Oversight Committee action.

Table 2
Examples of Savings and Avoided Costs Achieved by the Oversight Committee

Area	Description	Estimated Savings and avoided costs
Demolishing the old east span of the Bay Bridge*	The Oversight Committee directed the creation of an alternative demolition strategy, which relied on marine implosions.	$94 million
Encouraging multiple bids on the suspension span of the Bay Bridge	Only one bidder responded to Caltrans' original solicitation for construction of the suspension span. The Oversight Committee offered a $3 million stipend to up to the three lowest bidders to encourage multiple bidders. Increased competition resulted in a bid below Caltrans' estimated cost.	$49 million
Retrofitting the Antioch and Dumbarton bridges	The Oversight Committee voted to remove $353 million saved in the two projects from the seismic program's budget. The Toll Authority made these savings available to other regional projects.	$353 million
Reducing staffing levels	The Oversight Committee reduced a proposed Caltrans fiscal year 2016–17 staff support budget from $32 million to $23 million after in-depth inquiry into project staffing levels.	$9 million
TOTAL		$505 million

In one example, the Oversight Committee took action to encourage multiple bidders on the suspension span, resulting in savings of nearly $50 million. The original solicitation for construction of the suspension span in 2004 resulted in a response from one bidder, and the bid was $666 million higher than Caltrans' initial engineers' estimate. As the State Auditor's Office reported in 2004, Caltrans chose not to disclose program information according to the regular reporting schedule established by law; as a consequence, Caltrans had placed the Legislature in the awkward position of having to try to devise a funding solution six weeks before the bid on the suspension span was set to expire. When the Legislature did not approve funding by September 2004, the bid expired.

According to an Oversight Committee report, when Caltrans rebid the work on the suspension span in 2005, the Oversight Committee used its authority to approve bid stipends of $3 million to each of the three lowest responsive bidders. The stipends acted as partial compensation for expenses the contractors incurred in preparing the complex bids for the suspension span and were designed to prevent having only one contractor bid, as had occurred in 2004. According to the Oversight Committee, the increased competition resulted in two bids, one of which was $49 million below the engineers' revised 2006 estimate. This decrease from the expected cost was especially noteworthy because in 2005 a Caltrans market analysis noted that Caltrans had been experiencing fewer bidders and increasing bid prices for its awarded contracts.

Although the Oversight Committee was successful in managing the seismic program, committee representatives we spoke to suggested improvements for future oversight committees. For example, the Legislature amended state law to require that the Oversight Committee be subject to open meeting laws beginning in 2016. The committee had already opted to institute a modified version of open meetings before the legislative requirement. Although the legislative change resulted in increased transparency for the public according to the executive director of MTC, it also increased the difficulty of conducting day-to-day business. In a three-person committee, two members constitutes a majority; thus, when any two members of the committee discuss, hear, or deliberate on any issue pertaining to the Oversight Committee, it would constitute a meeting of the Oversight Committee. This could lead to committee members inadvertently holding a "meeting" when conducting other business. For example, Caltrans and the California Transportation Commission work closely together and it is reasonable to assume that their directors would need to meet from time to time. The executive director of the California Transportation Commission indicated that the Oversight Committee may benefit from having its own attorney to ensure that legal issues that affect the committee as a whole—such as compliance with open meeting laws—are addressed. She stated that counsel from the members' respective agencies currently represents members of the Oversight Committee; however, there is no attorney assigned to represent the Oversight Committee as a whole.

Although increasing the number of Oversight Committee members and assigning dedicated counsel would likely increase costs, the Oversight Committee's current quarterly report notes the cost of Oversight Committee operations from 2005 to 2018 as being about $17 million, or slightly more than $1 million per year. This is a small fraction of the $9 billion cost of the seismic program and significantly less than the cost savings the Oversight Committee achieved. The cost of the Oversight Committee derives mostly from the labor costs of support staff and consultants at each respective Oversight Committee agency.

Caltrans' Risk Management Plan Played a Critical Role in the Oversight Committee's Ability to Prevent Unnecessary Costs and Delays

Caltrans risk management handbook, published in 2003, notes that risk management is most effective when it is performed early and continues throughout the project. However, in 2004 our office found that Caltrans failed to embody these principles early in the seismic program. The Project Management Institute's Guide to the Project Management Body of Knowledge, 6th Edition (PMBOK Guide) recommends that organizations make a conscious decision to manage risk in a controlled and intentional manner in order to prevent deviations from a project plan that could cause a project to fail to achieve its objectives.⁴ In 2004, as the seismic program was experiencing significant cost overruns, we recommended that Caltrans implement a comprehensive risk management plan for the seismic program using its risk management handbook. In 2005, as part of the same legislation that created the Oversight Committee, the Legislature required Caltrans to develop and conduct comprehensive risk management for the seismic program, including identifying project risks and quantifying such risks in financial terms.

Following this legislation, risk management played a critical role in informing the decisions of the Oversight Committee. Qualitative and quantitative risk analysis informed Oversight Committee decisions that led to avoiding hundreds of millions of dollars in potential costs, and seven years of potential delays, as shown in Table 3. The text box describes qualitative and quantitative risk analysis. In one case, the risk management team performed an analysis in 2008 that identified up to $305 million in potential costs for schedule delays and fabrication issues related to the suspension span. As a result, the Oversight Committee approved a $13.9 million mitigation procedure to improve quality control with Caltrans' fabrication contractors, which Caltrans projected would enable the project to avoid between $132 million and $291 million in costs. In another instance, according to the Oversight Committee in 2005, its risk management team conducted an analysis of two contracts to weigh the cost of accelerating one to avoid delay to the other. The team's analysis accurately predicted that the first contract would finish in time for work to begin on the other without needing to accelerate the first contract; Caltrans estimated that this decision avoided $12 million in potentially wasted costs.

Table 3
Examples of Costs Avoided Through Risk Management in the Seismic Program

Example of Risk Mitigation	Description	Potential Costs Avoided (in millions)	Potential Delay Avoided
Demolishing the old east span*	Risk analysis demonstrated the high costs associated with traditional demolition of certain east span components, leading to the Oversight Committee's decision.	$94	4 years
Avoiding underwater construction	In 2009 the Oversight Committee estimated the seismic retrofit of the Antioch and Dumbarton bridges would cost $950 million. The risk management team determined underwater construction would be precarious, prompting the team to devote resources towards developing an alternative strategy. As a result, the potential costs significantly decreased.	$200
Suspension span fabrication	The risk management team identified several risks related to the fabrication of suspension span components. The Oversight Committee used this information to approve a process that would ensure quality and timely fabrication in China.	$132 to $291	2 years
Contract acceleration analysis	The risk management team performed a risk analysis of two contracts to determine whether funds should be expended to accelerate completion on a contract to support work in another area. The analysis revealed less than a 5 percent chance such acceleration would be necessary. The Oversight Committee elected not to accelerate the contract.	$12
Nesting birds permit†	In 2011 the risk management team identified bird nesting as a large delay risk to the old east span's demolition. This led the project team to obtain special permits, which allowed for the safe relocation of birds. As a result, the contract did not experience any delays due to nesting birds.	$17	1 year
TOTALS		$455 to $614	7 years

Risk management on the seismic program has also resulted in reduced environmental and societal impacts. For example, Caltrans noted that in 2011 its risk management team identified nesting birds as one of the largest delay risks to the demolition of the old east span. According to Caltrans, if certain protected birds lay eggs on site, work needs to stop until any chicks hatching from the eggs leave their nests. Caltrans states that federal law prohibits the removal of certain migratory birds—one species, the double‑crested cormorant is pictured in Figure 5 below—or their nests without a permit. Based on Caltrans' data, each day of delay in the demolition project could cost more than $44,000, and a nest could cause a delay of up to 12 weeks. Thus, a single nest could potentially have resulted in costs of more than $2.5 million. These financial and ecological concerns drove Caltrans to develop a strategy for responsibly managing the bird population, which included securing Fish and Wildlife permits, collaborating with International Bird Rescue, and providing alternative nests on the new span before the beginning of nesting season in 2014.

Figure 5
Cormorants Relocated From the Prior East Span During Construction

Source: Johnson Marigot Consulting.

Risk management on the seismic program has also reduced the impact of the project on the public. Caltrans stated that its risk management analysis led it to plan for a four-day bridge closure over Labor Day weekend in 2009, instead of relying on the contractor's estimate of three days. To provide transportation agencies enough time to arrange alternatives, Caltrans needed to announce the bridge closure months in advance. The decision proved correct, as the contractor completed work just under the scheduled four-day allotment. Considering that almost 125,000 cars crossed the Bay Bridge daily in 2009, an unscheduled closure would have led to traffic delays and frustration for thousands of drivers during a long holiday weekend.

State Law Should Require Oversight Committees and Risk Management Plans for All Major Transportation Infrastructure Projects

The Oversight Committee, combined with Caltrans' risk management, contributed to significant cost savings for the seismic program. The Bay Bridge and the Transbay Terminal faced significant cost increases of $3 billion and $1.1 billion, respectively, before instituting increased oversight.⁵ In order to respond to its increasing costs, the Transbay Terminal project implemented a cost review committee to provide additional financial oversight and enlisted the help of the San Francisco Department of Public Works to provide construction oversight beginning in 2016.

State statutes do not impose a general requirement for all major transportation infrastructure projects to institute similar measures. The importance of a general requirement is particularly evident given that regional plans in the Bay Area, San Diego, and Los Angeles—three of the largest metropolitan regions in California—forecast transportation infrastructure projects totaling more than $600 billion over the coming decades. Although not all cost increases are avoidable, without assurance that large-scale projects have sufficient oversight, California and its communities are at greater risk that these projects will have unplanned cost increases—costs ultimately borne by taxpayers.

While Caltrans may use executive steering committees to provide advice and direction on some projects, it does not always do so. According to the chief deputy director of Caltrans' District 4, Caltrans and its partners may choose to use a steering committee on a case-by-case basis for large projects as it deems necessary. The current replacement of the Gerald Desmond Bridge in Long Beach, for example, has a multiagency steering committee that, according to its charter, provides direction and facilitates actions across partnering agencies. Caltrans has acknowledged the need for additional project oversight: when Caltrans reported to the Legislature in 2014 on lessons learned throughout the seismic program, among its recommendations was the use of multiagency oversight structures on large transportation projects. An investigation completed in July 2014, on behalf of the Senate Transportation Committee, reached a similar conclusion.

Following its use of risk management in the seismic program, Caltrans developed a scalable risk management policy based on the project's size, from $1 million to over $100 million, that applies increasing levels of risk management. In 2012 Caltrans issued a directive requiring staff to apply risk management practices to all capital and major maintenance projects. The directive outlines a scalable policy with increasing levels of risk management depending on the size of the project. Without establishing a general requirement for all major state and local transportation infrastructure projects to develop and conduct comprehensive risk management from the outset of a project, the application of risk management relies on either Caltrans involvement in projects or federal oversight.

However, not all projects involve Caltrans and even when federal oversight is required there is no assurance that it will be sufficient to prevent cost overruns. For example, because Caltrans estimated that the Bay Bridge project would cost more than $1 billion, it was subject to enhanced federal oversight.⁶ This oversight included the assignment of a full-time Federal Highway Administration project manager. However, the oversight proved insufficient to prevent cost increases prior to the establishment of the Oversight Committee. In its 2014 report to the Legislature on lessons learned in the seismic program, Caltrans noted that the project did not get the full benefits from risk management that would have accrued had it been implemented earlier than 2005. A statutory requirement applicable to all major publicly funded transportation infrastructure projects would ensure that projects sponsored by Caltrans, and other state and local entities, would benefit from similar requirements from the onset.

MTC Should Develop a Policy to Ensure That Future Projects Will Also Benefit From Risk Management

MTC lacks a formal policy to ensure that future projects it directs will benefit from appropriate levels of risk management, although it is using risk management on the projects it is currently managing. Under state law, MTC's primary role is that of a regional transportation planning agency but in recent years it has taken a more active role by directly managing some projects. MTC currently manages several projects including the Bay Area Express Lanes Program, for which MTC is responsible for implementing 270 miles of express lanes throughout the region that will use MTC's FasTrak toll collection system. To support this effort, MTC has developed a program-specific risk management plan that clearly identifies roles, responsibilities, and expectations. For instance, the program risk manager's responsibilities on MTC's express lanes program include reporting to the program manager, ensuring that project managers maintain project risk registers, and preparing quarterly risk management reports.

However, according to MTC's deputy director, MTC does not have a formal policy ensuring that it will develop similar plans for future projects it directs. Instead, it has historically made risk management decisions for the projects it directs on a case-by-case basis. He also stated that risk management costs could sometimes exceed what MTC believes to be fiscally prudent for a project. However, it is possible to develop a scalable risk management policy, as Caltrans has done, to accommodate projects of varied complexity and cost. For example, Caltrans' policy encourages risk documentation on extremely small projects of under $1 million but requires increasingly in-depth risk management as costs increase up to and above $100 million. Without a formal policy, MTC may apply risk management inconsistently—or neglect to apply it altogether—on future projects, which could increase costs and ultimately the financial burden on toll payers. As Caltrans has discovered, ensuring judicious and measured risk management on all projects requires a policy that allows flexibility but ensures the universal application of best practices. According to its executive director, MTC intends to develop a policy for projects it directs, which will improve efficacy and ensure the systematic application of risk management.

MTC's risk management policy should include the asset management plan the Toll Authority is currently creating.⁷ State law requires the Toll Authority to fund maintenance and rehabilitation programs on the Bay Area toll bridges after the seismic projects are completed. However, due to the long lifespan of Bay Area bridges, MTC cannot reasonably rely on the expertise of its current staff to ensure it applies appropriate engineering considerations to future maintenance funding decisions. To address this issue, the Toll Authority is assembling a panel of independent engineering consultants to assist it in creating an asset management plan that supplements similar Caltrans efforts. The asset management plan will assist in MTC's current efforts, which prioritize projects based on a scoring system that heavily weighs life and structural safety while considering current bridge conditions. Including the asset management plan as a component of an overall risk management policy will help ensure that MTC appropriately manages critical assets throughout their projected lifespans.

The Remediation of Defective Components on the Bay Bridge Represented a Small Portion of the Project's Total Costs

State statutes do not define what constitutes a defect for major transportation infrastructure projects. Similarly, Caltrans and MTC have not created a definition for defects generally, or on the Bay Bridge in particular, and therefore have generally not identified items as defective. For the purposes of this report, we used the PMBOK Guide definition of defect as detailed in the text box.

As indicated in Table 4 below, we found that Caltrans spent at least $66.6 million on remediating construction defects and at least another $19.7 million resolving fabrication problems in the contract change orders that we reviewed. Construction defect costs are those that indicate Caltrans paid to repair or replace a component that had already been installed on the bridge, while fabrication problems involve repairing or replacing a component before installation. Costs associated with Table 4 do not indicate the existence of current defects, but rather detail the cost of work conducted after Caltrans identified an issue and resolved it to the satisfaction of the engineers on the project. As we discuss later, multiple peer reviews and engineer panels have reviewed critical elements of the Bay Bridge and noted that they were safe.

Table 4
Cost of Remediation in Change Orders We Reviewed
(In Thousands)

Area of the Bridge	Construction Defect	Fabrication Problem	Total
Suspension span	$56,550	$17,591	$74,141
Yerba Buena Island	2,315	620	2,935
Skyway	1,746	755	2,501
Oakland touchdown	6,017	726	6,743
TOTALS	$66,628	$19,692	$86,320

Although Caltrans has developed specifications in its contracts which utilize technical industry standards to define certain defects, this information is specific to each contract and would not provide a general definition for the project as a whole. Because contractors are generally responsible for the cost and effort to remediate defective work if they fail to meet contract specifications, our review focused on instances in which Caltrans compensated a contractor to repair or replace a component. While Caltrans and MTC may disagree with respect to work we identified as defect remediation, the seismic program nevertheless incurred expenses while repairing or replacing bridge components. In total, our review of a selection of 800 contract change orders—out of about 2,500 total change orders for permanent work—identified $86 million of defect remediation costs on the Bay Bridge, or 1.5 percent of the project cost. Our review did not include temporary structures or work that was tangential to the bridge itself, such as electrical substations providing power to the bridge.

Caltrans has procedures to identify and address problems experienced during construction, such as work that does not meet specifications. Federal law requires Caltrans to develop a quality assurance program for federal-aid highway construction projects and according to Caltrans, its quality assurance program generally seeks to improve production and avoid defects. When Caltrans personnel identify an issue, they are to document it and communicate with both the resident engineer and the contractor on the project to resolve it. For example, in one instance, Caltrans personnel found that a contractor working on the east span allowed the use of a protective coating technique that was out of compliance with the contract's specifications. Caltrans staff worked with the contractor to resolve the issue at the contractor's expense.

Resolution may sometimes result in a contract change order, depending on the nature and severity of the issue. For example, quality assurance personnel detected a problem and issued a report when metal plates failed to meet quality requirements, which prompted Caltrans to work with the contractor to develop a solution that could be implemented without additional cost. However, there were other cases in which Caltrans incurred significant expenses remediating defects. For example, resolving problems surrounding well-publicized broken bolts—originally intended to anchor a segment of the bridge—required developing and implementing an alternative anchoring method. Caltrans' procedures allow it to identify a variety of issues and remediate defects when necessary, but not all defects can be proactively prevented.

Caltrans incurred increased costs remediating fabrication problems and construction defects depending on the complexity of the issue identified. For instance, Caltrans paid more than $400,000 to make modifications to a fabrication process for certain beams after the original process caused the steel to behave in unanticipated ways. Further, in correspondence with the requestor of this audit, the executive director of the Toll Authority and MTC stated that the well-publicized broken bolts, part of the anchoring system on the eastern side of the suspension span, were unquestionably a construction defect. Experts noted that a confluence of environmental factors and metallurgical conditions caused the bolts to fail, despite meeting industry standards. Caltrans spent more than $22 million remediating problems surrounding the broken bolts and developing an alternate anchoring method. Another significant expenditure, $13.9 million, provided for the development of a quality control process improvement after Caltrans personnel identified welding defects during fabrication of support structures. Caltrans also spent more than $12 million to make repairs, modifications, and adjustments to resolve fit issues between various complicated components.⁸ Examples similar to these account for the vast majority of remediation costs we identified and were paid for using contingency funds.

In spite of increased costs associated with remediating defects, according to multiple peer review panels from as recently as 2014, remediated defects such as those identified in our review do not indicate the presence of safety issues on the bridge. Caltrans designates the Bay Bridge as a lifeline structure, which requires that the bridge be able to provide a high level of post-earthquake transportation service for emergency response and support for the economic livelihood of the Bay Area. According to Caltrans, the east span is designed to withstand a once‑in‑1,500‑years seismic event. To ensure that problems have been resolved effectively, and that the Bay Bridge is ready to function in a seismic event, the project has undergone multiple peer reviews, in which internal and external engineering experts have conducted in-depth evaluations and determined that critical components are safe. For example, the Toll Bridge Seismic Safety Peer Review Panel stated that the tower foundation, part of the suspension span, was designed and constructed in a manner that meets or exceeds structural safety standards.

The Seismic Program Will Cost About $9 Billion, but Debt Service and Ongoing Maintenance Costs Will Continue

The seismic program will end in 2019 at a cost of about $9 billion, roughly on budget. Although costs associated with debt on bonds backed by toll revenues as well as ongoing maintenance and rehabilitation costs on all of the Bay Area bridges will continue into the future, the Toll Authority forecasts that it will have sufficient revenues to cover those expenses.

The Seismic Program Will Have Cost About $9 Billion at Completion

The Oversight Committee estimates that the seismic program will end by 2019 having cost just under $9 billion, not including ongoing maintenance and rehabilitation. As of August 2018, the Oversight Committee estimated that the final cost of the program will be about $34 million under the program's current budget. Nevertheless, as we discussed previously, the history of the seismic program has been one of escalating costs. Table 5 below shows the evolution of the program's budget by bridge, as well as the anticipated final costs of the program.

Table 5
Seismic Program Budgets and Costs by Bridge
(In Millions)

Bridge Retrofit or Replacement	1997 Cost Estimate	2001 Cost Estimate	2005 Budget	2009 Budget	Current Approved Budget (as of June 2018)	Costs to Date (as of June 2018)	Projected Costs at Closeout (2019)
Bay Bridge, east span replacement	$1,285.0	$2,600.0	$5,486.6	$5,486.6	$6,509.0	$6,428.3	$6,533.0
Bay Bridge, west span	553.0	700.0	736.9	736.9	757.9	757.7	757.9
Benicia–Martinez	101.0	190.0	177.8	177.8	177.8	177.8	177.8
Carquinez*	83.0	125.0	114.1	114.1	114.2	114.2	114.2
Richmond–San Rafael	329.0	665.0	914.0	914.0	811.9	794.8	794.9
San Mateo–Hayward	127.0	190.0	163.5	163.5	163.4	163.4	163.4
Vincent Thomas	45.0	62.0	58.5	58.5	58.5	58.4	58.5
San Diego–Coronado	95.0	105.0	103.5	103.5	103.2	103.2	103.2
Antioch				267.0	71.1	71.1	71.2
Dumbarton				483.0	112.4	111.8	112.2
Contingency†		448.0	900.0	900.0	46.7
Misc program costs			30.0	30.0	26.0	26.0	26.0
Net programmatic risk							5.6
TOTALS	$2,618.0	$5,085.0	$8,684.9	$9,434.9	$8,952.1	$8,806.7	$8,917.9

The cost estimates of the seismic program increased substantially from 1997 through 2005. In 1997 the Legislature allocated $2.6 billion in funds from various sources to finance what was at that point projected to be the entirety of the seismic program—five Bay Area toll bridges and two toll bridges in Southern California. A 2001 Caltrans report on the seismic program cited several factors leading to increased cost estimates including rising construction costs and delays in selecting a design for the east span. The Legislature responded in 2001 by allocating additional funds, bringing the new total for the seismic program to $5.1 billion. By 2005 Caltrans' estimates of the cost of the seismic program had grown well beyond $5 billion to $8.7 billion. In that year, the Legislature identified an additional $3.6 billion of funds to pay for the estimated cost overruns, with $2.2 billion to come from toll revenues and the remainder from several state funding sources. The Legislature further indicated that the Toll Authority would be responsible for covering any excess costs associated with the east span of the Bay Bridge through toll revenues.

The budget for the seismic program increased again after 2009 legislation that added two bridges to the program. The seismic program did not initially include the Antioch and Dumbarton bridges because they met seismic standards established in the early 1970s. However, after a series of earthquakes both in California and abroad, in 2008 the Toll Authority and Caltrans determined that the two bridges needed retrofitting. The Legislature added both bridges to the seismic program and directed the Oversight Committee to pay for the projects through cost savings or with toll revenue from the Toll Authority, authorizing the Toll Authority to increase tolls to complete work on the bridges. Early the following year, the Toll Authority approved adding $750 million to the seismic program's budget—a figure based on cost projections for retrofitting the Antioch and Dumbarton bridges. Because the Toll Authority was responsible for costs exceeding prior legislative appropriations, the $750 million would come from toll revenues. This brought the total budget for the seismic program to $9.4 billion.

Much of the increasing cost of the program was due to rising costs for replacing the east span. In 2005 Caltrans estimated the budget for the east span at $5.5 billion, an amount roughly equal to the prior legislative appropriation for the entire seismic program. As described in the bill analysis and supported by Caltrans estimates, this increase absorbed most of the $3.6 billion for overruns appropriated by the Legislature in 2005. Further, the cost of the east span would continue to grow; the Oversight Committee projects that the final cost of the east span will be about $6.5 billion. The additional $1 billion was the result of increased costs in three areas: the diversion of traffic from the old bridge to a detour structure at Yerba Buena Island, the suspension span, and Caltrans' support costs. Of these projects, the Yerba Buena Island detour accounted for over $340 million due to contract changes that combined the detour with other foundation work related to the Yerba Buena Island portion of the Bay Bridge. According to Caltrans, it advanced the foundation work on the Yerba Buena Island contract as a risk mitigation measure to get foundation work completed on schedule. It incurred more than $260 million in additional costs on the suspension span in an attempt to mitigate delays, resolve complex construction issues, and keep the Bay Bridge opening on schedule.

Caltrans' support costs accounted for an additional $390 million of the costs of the east span; this amount represents the cost of support personnel involved in developing and delivering the project, including staff such as project managers, engineers, and others. Any delays in the project, as well as contract changes, will increase costs in this area. The Oversight Committee projects that support costs on the east span will amount to about 26 percent of capital costs upon project completion. By comparison, in fiscal year 2015–16 Caltrans reported that the ratio of support to capital costs across all of its projects was much higher at about 38 percent.

Despite the continued growth in the cost of the east span, overall the seismic program will end in 2019 roughly on budget, due to the Oversight Committee and the application of risk management, as we discussed previously, as well as cost savings in other parts of the program and a $900 million contingency in the program budget. In fact, the Oversight Committee reduced the overall program budget after achieving cost savings on other bridges. The final cost of retrofitting the Antioch and Dumbarton bridges was about $180 million, nearly $570 million less than projected. These reductions were largely the result of Caltrans' assessment that it could reduce project scope and risk by eliminating underwater work on the bridges. Further, Caltrans received significantly lower bids than anticipated for the work on both bridges. As Figure 6 below demonstrates, in 2010 and 2013 the Oversight Committee chose to remove funds from the seismic program's budget entirely, reducing the budget from $9.4 billion to about $9 billion. State law gives the Toll Authority latitude in the use of toll revenues, and according to the Toll Authority's chief financial officer (CFO), MTC returned some of the Antioch and Dumbarton savings to the Toll Authority's toll account for use on other projects but generally did not direct them to a specific use. However, in 2013, MTC redirected $130 million of savings from the seismic program to other Bay Area transit agencies for replacing buses and rail cars.

Figure 6
Effect of Oversight Committee Actions on Seismic Program Budget and Costs
(Billions of Dollars)

The inclusion of a contingency fund in the budget has allowed the Oversight Committee to manage cost increases to the east span without going over budget. In 2001 Caltrans included a $448 million contingency in the program budget. By 2004 Caltrans had increased the estimated contingency to $900 million. According to a 2004 report on the seismic program from the Toll Authority, Caltrans arrived at the $900 million amount through a quantitative analysis that established a contingency range of between $500 million and $900 million. That amount was part of the estimate Caltrans provided to the Legislature in 2005 showing an increase of nearly $3.6 billion in the seismic program. According to Caltrans' breakdown of contingency expenditures, the bulk of the $900 million contingency has gone to the east span.

Our review of the Oversight Committee's most recent budget and cost projections indicates that the Oversight Committee anticipates $112 million in remaining work before project closeout. The vast majority of the remaining expenditures relate to work on the east span. For instance, the largest item, $76 million, is to dismantle the old east span and to retain and improve piers to allow for public access, collectively the final major phase of project closeout. Further, Caltrans' most recent risk management estimate show a reduced draw on the program's contingency fund, which will result in a surplus of roughly $34 million at project closeout. Due to the resolution of several issues, including a recent settlement with the firm that built the suspension span, costs have come down slightly and Caltrans staff predict the seismic program will finish slightly under budget.

Based on contemporaneous Caltrans and MTC documents, cost overruns on the east span of the Bay Bridge did not cause significant delays on other bridges in the seismic program. Before the creation of the Oversight Committee, state law required Caltrans to issue annual reports on funds spent on the seismic program. These reports include information on anticipated completion dates for each of the bridges. After examining those reports and comparing them to project completion dates, we determined that of the six bridges in the initial seismic program other than the Bay Bridge, Caltrans completed two on time, and it completed another—the Richmond–San Rafael Bridge—on time after revising its initial estimate. Caltrans finished the retrofit of two bridges within a year of its initial estimates, and it finished one within two years. According to Caltrans, the delays on the Richmond–San Rafael Bridge were the result of complications related to underwater foundation work and environmental restrictions. Further, Caltrans completed work on the main portion of the west span of the Bay Bridge earlier than its initial estimate, although it completed the approach to the west span on time after revising the initial estimate by one year. In fact, Caltrans had completed work on most of the bridges, aside from the east span, by 2005, when the Oversight Committee began managing the seismic program. MTC's executive director stated that MTC's increased responsibility after 2005 did not negatively affect other projects that were in progress. Caltrans finished retrofitting the Antioch and Dumbarton bridges, added to the seismic program in 2009, on time and well under budget.

The Toll Authority Forecasts Sufficient Toll Revenues to Meet Maintenance Expenditures Over the Next Decade

In addition to funding toll bridge retrofitting and eventual replacement, the Toll Authority funds bridge maintenance. Bridges are complex structures that require both routine and long-term maintenance throughout their lifespans. State law specifies that Caltrans is responsible for maintaining the Bay Area toll bridges, and once seismic work is complete on each bridge, the Toll Authority will be responsible for funding that maintenance. According to Caltrans' records, the cost of routine maintenance—which includes activities such as graffiti cleanup, deck repair, and inspections—on the Bay Area toll bridges ranged from about $8 million to $13 million annually from fiscal years 2012–13 through fiscal year 2016–17. Costs associated with rehabilitation—which includes all nonroutine activities such as replacement of damaged or worn components, improved toll collection systems, and studies—totaled about $110 million in fiscal year 2016–17. The Toll Authority already pays for maintenance and rehabilitation costs for most of the Bay Area toll bridges, and when Caltrans completes work on the Bay Bridge, it will begin billing the Toll Authority for maintenance and rehabilitation costs on that bridge as well.

The Toll Authority forecasts that it will collect sufficient revenue to meet the bridges' maintenance and rehabilitation needs over the next 10 fiscal years, projecting that toll revenues will increase from $724 million in fiscal year 2016–17 to $889 million in fiscal year 2026–27. This revenue would also be sufficient to cover the Toll Authority's expenses, including payments on bonds. After reviewing the revenue forecasts for the previous 10 fiscal years, we determined that actual revenues have consistently exceeded revenue forecasts by small amounts, suggesting that future forecasts are reasonable. Before the passage of Regional Measure 3 in June 2018, the Toll Authority had projected the need for a $1 toll increase in fiscal year 2026–27. According to the Toll Authority's CFO, this was to prepare for the replacement of the region's oldest toll bridges, several of which are over 50 years old. However, because Regional Measure 3 authorizes a series of toll increases and transportation projects beginning in 2019, the Toll Authority's most recent projections are outdated. The increased revenue from Regional Measure 3 may make the $1 increase in fiscal year 2026–27 unnecessary. According to the Toll Authority's CFO, the Toll Authority will update its revenue projections in the next fiscal year. In fact, it will need to do so when it prepares to issue bonds to fund future projects.

Finally, the Toll Authority maintains a reserve for maintenance and rehabilitation costs to help ensure that it is able to meet future needs. The Toll Authority retains consultants to help it predict future maintenance and rehabilitation needs. Based on the consultants' work, the Toll Authority created a reserve for maintenance and rehabilitation. That reserve is currently $120 million and is based on costs of $60 million per year. Although this is below the maintenance and rehabilitation costs for fiscal year 2016–17, these costs fluctuate from year to year. According to the Toll Authority's CFO, the authority builds the rehabilitation budget on a "life-to-date" basis—that is, it is less concerned about a single fiscal year's expenditures, and focuses instead on whether the expenditures do not exceed the authorized budget over the life of the project. Nevertheless, according to its CFO, the Toll Authority is planning to reevaluate all of its reserves sometime next year in order to take into consideration changes due to the passage of Regional Measure 3 in June 2018. Further, an updated consultant's assessment estimated that annual maintenance and rehabilitation costs would increase to an average of slightly more than $90 million a year beginning in 2017, although this has not yet resulted in an update to the reserve.

The Toll Authority's Payments on the Debt It Incurred as a Result of the Seismic Program Will Continue for Decades

Although state law allocated state funds for the seismic program, the total funding was a mix of state, federal, and, overwhelmingly, bond money backed by toll revenues. In fact, of the anticipated $9 billion final cost of the seismic program, more than $6 billion will have come from toll revenues paid by drivers. The appendix on page 37 provides a breakdown of revenue sources for the seismic program.

The Toll Authority's debt service on bonds related to the seismic program will continue for decades. Based on the Toll Authority's power in state law to issue revenue bonds and use the proceeds for financing projects, it has generally issued bonds for both seismic and non-seismic projects. According to the Toll Authority's CFO, due to this pooling of bond funds it cannot separately identify interest payments for the seismic program. However, according to MTC's most recent financial statements, it will pay $9.3 billion from fiscal years 2017–18 through 2055–56 in interest on various MTC projects. When outstanding principal is included, the Toll Authority's remaining debt service payments through fiscal year 2055–56 will total $18.7 billion.

The Toll Authority's practice of pooling bond proceeds from both seismic and non-seismic projects is within its authority and benefits the region. Not only does state law give the Toll Authority broad powers to issue bonds and use bond proceeds, its CFO stated that doing so creates several advantages such as reducing risk for investors and creating additional funding to cover debt. He also said that pooling has led to more favorable financing terms that allow the Toll Authority to maximize the amount of funding it receives for the bonds it issues.

Recommendations

Legislature

To ensure that large transportation infrastructure projects throughout the State benefit from appropriate oversight, the Legislature should require that all publicly funded major transportation infrastructure construction projects estimated to cost $500 million or more, have oversight committees subject to open meeting laws. When practical, each oversight committee should include individuals from at least three major agencies involved in the project, with roles that reflect financial interests as well as project execution and oversight. Further, when possible, each committee should include at least five members to support its ability to conduct day-to-day business without violating open meeting law requirements. The oversight committees should act as the authorities for critical decisions and have sufficient staff to support their decision-making roles.

To ensure that oversight committees perform their duties in a manner commensurate with the demands of large transportation infrastructure projects, the Legislature should require that the oversight committees have duties similar to those of the Oversight Committee, including the following:

• Providing project direction.
• Reviewing project status, costs, schedules, and staffing levels.
• Resolving project issues and evaluating project changes.
• Developing and regularly updating cost estimates, risk assessments, and cash-flow requirements.

To ensure that oversight committees effectively address both the fiscal and project management elements of large transportation infrastructure projects, the Legislature should require consolidated reporting at least annually detailing cost savings, cost overruns, and updates on project completion.

To ensure that oversight committees and the agencies involved in large transportation infrastructure projects engage in sufficient and appropriate risk management, the Legislature should require all publicly funded transportation infrastructure projects with a total estimated cost of $500 million or more to develop risk management plans that use both qualitative and quantitative risk analyses throughout the course of the projects.

MTC

To ensure that future projects have adequate risk management, MTC should formalize a scalable risk management policy by June 2019 so that the projects it directs benefit from sufficient and ongoing risk management.

---

We conducted this audit under the authority vested in the California State Auditor by Government Code 8543 et seq. and according to generally accepted government auditing standards. Those standards require that we plan and perform the audit to obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions based on our audit objectives specified in the Scope and Methodology section of the report. We believe that the evidence obtained provides a reasonable basis for our findings and conclusions based on our audit objectives.

Respectfully submitted,

ELAINE M. HOWLE, CPA
California State Auditor

Date:
August 28, 2018

Staff:
John Lewis, MPA, Audit Principal
Nicholas B. Phelps, JD
Christina L. Downard
Joseph S. Sheffo, MPA
Ashley Snyder

Legal Counsel:
Heather Kendrick, Sr. Staff Counsel

For questions regarding the contents of this report, please contact Margarita Fernández, Chief of Public Affairs, at 916.445.0255.

---

Footnotes

⁴ Recognized for its development of standards for project management, the Project Management Institute publishes the PMBOK Guide, which provides guidelines for managing individual projects. Go back to text

⁵ The Transbay Joint Powers Authority originally estimated the Transbay Terminal to cost $1.2 billion in 2007, but its budget grew to $2.3 billion in 2016. Located in downtown San Francisco, this terminal will serve as the primary bus and rail terminal for the Bay Area. Go back to text

⁶ Previously, projects over $1 billion were required to submit annual financial plans. However, in 2005 Congress revised this to define a major project as a project with an estimated cost of $500 million or more and required such projects to submit project management and annual financial plans. Go back to text

⁷ As noted in the Introduction, MTC and the Toll Authority are legally separate entities that share the same board, staff, and facilities. Go back to text

⁸ Fit issues occurred when a component required modification to fit properly with another component. Go back to text

Back to top