25 breakdowns in its first 11 months: What happened to the new Flagler Bridge?

Updated Aug 04, 2018
Gordon Meyering tweeted this photo while stuck in traffic when Flagler Memorial Bridge malfunctioned on Nov. 27, 2017.

UPDATE 11:45 a.m. Sunday, Aug. 5: Around 10 a.m. Sunday, the Flagler Memorial Bridge had malfunctioned, according to the traffic alerts from the Town of Palm Beach. 

As of 12:30 p.m., the bridge had reopened to traffic.

ORIGINAL STORY: The chronic malfunctions of the new Flagler Memorial Bridge — more than 20 within its first 11 months — raise a fundamental question for incredulous commuters such as Jere Zenko of Palm Beach:

“How is it possible that something brand new doesn’t work?”’ she asked.

That is the precise mystery that has baffled the bridge’s builders and designers for nearly a year.

While the malfunctions tested the tempers of commuters caught in repeated delays from August to June, behind the scenes they fueled consternation and friction among desperate engineers scrambling after each breakdown to get the new $131.8 million drawbridge going again.

They recalibrated the settings on the motor and machinery. They replaced blown circuit breakers and swapped out switches that direct power. They added more lubrication to the gears and pinions that move the 900-ton spans.

But each time they thought they finally found a fix, it seemed like the bridge built to last 75 years broke down again days or weeks later.

Even as they were internally stymied, project officials gave the public an optimistic but less-than-complete picture of the problems, offering empty reassurances that they were close to finding a fix, a Palm Beach Post investigation found.

In late April, all but admitting defeat, they reluctantly sought outside help. But after a $15,000 consultant issued a two-page memo that found no major mechanical problems, the bridge broke down three more times.

As tempers among Palm Beach residents neared “the boiling point,” a top state transportation official reached out to the bridge contractor’s North American president whose internal inquiry resulted in finger-pointing by a project manager, emails obtained by The Post show.

The records also show engineers questioned the drawbridge’s design and speculated that their attempted fixes to early malfunctions may have contributed to other breakdowns months later.

Although the bridge has worked fine since June 7 — the longest malfunction-free streak since all four lanes opened July 31, 2017 — spotty records kept by the contractor may have hindered efforts to find a solution much earlier, documents reveal.

And other emails show that many of the delays endured by the public were exacerbated because of the failure of the prime contractor, PCL Civil Contractors, to quickly notify Palm Beach about malfunctions and tests.

For commuters who tolerated five years of problem-plagued construction before the bridge fully opened, it all played out like a cruel encore.

“It’s a pretty pathetic situation,” groused Henry P. “Rip” McIntosh IV, a Palm Beacher who described scenes of angry motorists trapped behind the cross gates until police could unclog the mess by directing each car to back out of the lanes one by one. “You’re talking about a hundred-million-dollar bridge that breaks in its first month in use and repeatedly thereafter. That’s disgraceful.”

Meanwhile, PCL will not have to pay any fines or damages if there are no more malfunctions before the company’s contract expires Monday. In fact, the company appears to be in line for a favorable state review, scoring 94 out of 104 points on a grading system.

And while The Post’s review found 25 malfunctions since the bridge opened, only two prompted the Florida Department of Transportation to reprimand PCL.

Company officials declined to be interviewed but issued this statement: “PCL is working closely with the Florida Department of Transportation on mechanical issues that may have arisen on the Flagler Memorial Bridge, and we are committed to delivering a quality bridge to the community.”

Meanwhile, state and town officials are crossing their fingers that the problems have ended.

Unanswered questions

No one disputes the beauty of the new bridge — its Mediterranean-style tender’s booth, ornamental handrails and wide lanes offering postcard views of the Intracoastal Waterway. But it hasn’t won the unanimous confidence of the people who rely on it.

Also known as the north bridge, it’s one of three connections between the West Palm Beach mainland and the barrier island of Palm Beach. About 16,000 vehicles cross it every day.

To get a better understanding of the bridge’s operations and why it was breaking down, a reporter made several requests over the spring and summer for interviews with bridge builders and a tour of the bridge’s internal mechanisms. State officials rejected the tour, saying the contractor controlled the site. The contractor, PCL, declined to provide access.

Instead, FDOT requested questions in writing, a fairly common tactic that allows officials to limit their answers, often with incomplete information. Answers were provided by a “project team” — comprised of officials with the state, contractors and engineers. Some questions were never answered, and it sometimes took weeks to get responses to follow-up questions.

For example, when asked why the new multimillion-dollar bridge was breaking down so often, the project team referred a reporter to a malfunction log kept by bridge tenders working for a PCL subcontractor. But the log didn’t offer a clear answer to that question, an opinion echoed by FDOT consultants who reviewed the log in late April and found it lacked key information. The log also didn’t cite three malfunctions reported by Palm Beach.

After months of back and forth over unanswered technical issues, FDOT allowed consulting engineer Geoffrey Parker, the town of Palm Beach’s liaison on the project, to speak to The Post. He maintained that the string of malfunctions was an “unfortunate” series of issues.

“There was no one clear, defining reason,” he said, adding that the drawbridge is made of “thousands of parts” that work in sync to raise the 900-ton spans. “There are things that do slip up.”

Parker also worked on the Royal Park Bridge, known by locals as the middle bridge, which uses the same “rolling bascule” design as the Flagler bridge. Replaced in 2005 by Holland-based contractor Ballast Nedham, the Royal Park Bridge had issues when it first opened, he said.

There might have been “a handful of issues” in the Royal Park Bridge’s first year, said Tom Bradford, Palm Beach’s town manager from 2015 until February.

“But not to the extent of what they had on the Flagler Memorial Bridge. What was experienced on the Flagler Memorial Bridge was an aberration. It’s outside the norm.”

‘A hallelujah day’

Henry Flagler, the bridge’s namesake, built an overseas railroad on 42 bridges across 128 miles of water in seven years from 1905 to 1912, according to the Flagler Museum.

It has taken nearly six years for teams of subcontractors to rebuild the Flagler Memorial Bridge, which is a little more than half a mile long.

Work started in September 2012 after a debate over the design; town officials wanted to make sure the bridge, then priced at $94.2 million, reflected Palm Beach standards.

Plans called for replacing the deteriorating two-span drawbridge built in 1938 for $725,000. It was supposed to be a three-year project.

But the town insisted on keeping the bridge open while the new one was built next to it. Vibrations from the construction led to problems on the existing bridge — from cracks in the aging foundation to settlement issues with support beams. That led to delays and partial shutdowns, sparking loud outcry from residents. Nearly $10 million was spent to repair the bridge that was being replaced.

Finally, two lanes of the new bridge opened in December 2016. Eight months later, all four lanes opened, an event embraced by commuters eager to forget the traffic headaches of the previous five years.

“It is certainly a celebration of the smooth movement of traffic once again,” Palm Beach Mayor Gail Coniglio said at a ribbon-cutting ceremony on July 31. “It is a hallelujah day.”

The celebration didn’t last long.

Like ‘a Swiss watch’ 

Bascule — a French word for “rocking” or “seesaw” — is the basis for a drawbridge design that has been used for centuries. The Flagler Memorial Bridge is what’s known as a rolling bascule bridge, an enhanced design dating to the 1890s, that both pivots and rolls, the base of each span like a rocking horse.

The Flagler bridge has four spans, also known as leaves. These leaves are made of concrete and steel. Each one weighs 900 tons and measures 98 feet long by 44 feet wide. Two connect the eastbound lanes and two connect the westbound lanes, with each pair coming together roughly in the middle to formsurface over which cars drive.

Raising and lowering the massive spans is an intricate engineering feat that’s lost on most motorists as they wait for the crossing gates to rise. They don’t see the action beneath the bridge tender’s booth, where a vast network of machinery performs a delicate but complex dance.

“I liken it to a Swiss watch,” Parker said. “It’s a series of engineered moving parts, electrical, hydraulic and mechanical, all integrated to move together.”

It all starts when the traffic light turns red. Once the crossing gates drop, the bridge tender pushes a button that releases pins from the front and back of each leaf — pins that lock down the spans while cars and trucks are crossing. Also known as locks, each pin is a couple of feet long, square-shaped — roughly the thickness of a loaf of bread — and fits into an interlocking receiver on the end of the adjacent span.

When the pins are released, a small device called a “limit switch” signals a green light on a control board in the bridge tender’s booth. The green light tells the bridge tender to press the next button, which triggers a 25-horsepower motor that sets the mammoth spans in slow motion.

At the center of this technological marvel, beneath the tender’s booth, in a cavernous room that houses the machinery, are two circular wheels, one on each side of the span. The key to the wheels are toothed edges called pinions, which roll along horizontal tracks that are spiked with matching teeth. The pinions connect with and, in turn, roll a large half-moon shaped section on the underbase of the span.

As the pinions roll, the bridge spans gently rise.

When a bridge ‘walks’

To close the drawbridge, the entire process goes in reverse: The spans lower back down, the pins are locked into place and the crossing gates go up.

While it’s critical for all of the machinery to work in precision, it’s not unusual for there to be “infinitesimal horizontal movement” of the spans, the project team said. “This is expected and bridges of this type automatically recenter themselves. This is a time-tested design.”

But too much movement — from the drawbridge’s motor or outside sources such as wind and heat or from a design flaw — can cause jams in the locks, preventing the leaves from opening or closing.

That’s why the teeth on the gears and tracks must be designed with appropriate shape and tolerances tight enough to control the side to side movement, called transverse movement, at the end of the leaf.

And the torque, the amount of rotational force exerted by the electric motor, must be set properly so the pinions roll in precise alignment with one another. If not, one pinion on one of the tracks can slide forward and the other can slide backward, an unintended result referred to by engineers as “walking.”

If the bridge walks, it can lead to all sorts of problems: The ends might not match up when the spans try to open or close and the locking pins might not engage. The switches and sensors will recognize when the alignment is incorrect and in turn refuse to proceed to the next step of the bridge operation until the issue is acknowledged and rectified. 

The refusal of the system to continue is an intended failsafe known as the “safety interlock” and exists in the programming in the logic of the bridge. This refusal to continue with the opening is often referred to as a bridge malfunction.

“There are some issues that do tend to present themselves,” Parker said.

But he also agreed that it isn’t normal for a new multimillion dollar drawbridge to malfunction 25 times within its first 11 months.

‘These malfunctions are concerning’

“Yesterday, the new bridge malfunctioned again,” Bradford, the town manager at the time, wrote Aug. 24 in an email to FDOT, one of the first public records that mention a problem.

But details about what happened aren’t clear — the purpose of Bradford’s email was to point out that the bridge tender had failed to alert the police department, “per approved protocol,” when the bridge operations were restored.

The bridge broke down again on the mornings of Sept. 6 and Sept. 15, according to “traffic alerts” posted on the town’s website and on social media.

None of those first three breakdowns is mentioned on PCL’s malfunction log.

The first breakdown in the company’s log was at 2:33 p.m. on Sept. 21. The cause was listed only as “limit switch.” Operations were restored just after 3 p.m.

That was followed by malfunctions on the mornings of Sept. 26 and Sept. 27. Both times a “tail lock” in one of the southern spans failed to drive, preventing the spans from fully closing. The bridge was shut down for an hour each day.

The Sept. 27 malfunction occurred on the same day bridge builders were holding their weekly meeting to discuss the project’s progress as they prepared to hand the bridge over to FDOT.

Minutes from that meeting include a section called “Bridge tail lock issues & repeated malfunctions” — one of the earliest public records mentioning FDOT concerns about accepting the bridge, “considering the issues it is having,” from the contractor.

Santiago Zuleta, PCL’s project manager, “agreed that these malfunctions are concerning,” the minutes show.

‘Confident the issue will be rectified’

On Oct. 4, the mayor of Palm Beach was on the phone voicing her concerns to Parker, a senior engineer with FDOT consultant New Millennium Engineering.

“Concerning the reliability of the bridge, you are absolutely correct: There is no reason why a brand new bridge should be plagued with any issues regarding operation,” Parker said the next day in an email to Coniglio.

“We have determined what the ‘repeat offenders’ are and the contractor has scheduled a number of field meetings to properly sort and correct these issues with finality. The department will not accept a bridge that exhibits such behaviors.”

A skeptical Coniglio shared the email with town officials, one of whom forwarded it to the town engineer with a request: “Can you find out what the reliability issues really are?”

While no malfunctions were reported in October or November, a commuter on Nov. 27 tweeted “New Flagler bridge stuck West bound” above a photo of the two west spans in the upright position.

Tripped circuit breakers prevented the southeast leaf from fully closing, once in December and twice in January, according to the log, including a 90-minute delay during the afternoon rush on Jan. 3.

“The issue has occurred approximately three times within the last month and a half,” an FDOT spokesperson said Jan. 9 in a statement to a reporter. “As this is an intermittent issue, it is challenging to pinpoint. We are confident the issue will be identified and rectified soon.”

Deputy chief’s ‘grave displeasure’

On Feb. 20, the bridge was out more than 2½ hours during the afternoon rush. But that wasn’t because of a malfunction, according to the log, “just extended closures while new settings were tested.”

Those “new settings” would be suspected of factoring into future malfunctions.

The next day, Deputy Palm Beach Police Chief Ann-Marie Taylor emailed the contractors and FDOT to “express my grave displeasure” about how the testing forced motorists “to sit in traffic, unnecessarily for an extended period of time.”

“Why would calibration (maintenance) be scheduled during rush hour traffic? Why wouldn’t you notify the police department in advance or at least ask for input? Why didn’t the bridge tender notify us this was occurring, rather we had to hear from the public there was a problem at the bridge?” she asked.

On Feb. 22, Scott Ryder of American Engineering Group, a PCL subcontractor, emailed an apology. “Our intent was to have the testing completed long before rush hour began,” he said.

The next day, the bridge broke down again for 20 minutes. The culprit? A bridge tender’s error, the PCL log said. FDOT sent an email to the contractor, demanding they use “only experienced bridge tenders.”

By the end of February, the bridge had broken down 12 times in seven months, public records show. But some of the most frustrating delays were yet to come.

“WHAAAAAT??” Coniglio wrote to Parker on March 26 after what the FDOT called “an electrical issue” prevented the southeast leaf from properly closing twice in the same day, causing delays of nearly four hours.

“Gentlemen. Please. It is now a serious long-standing problem built at a cost of 125 million dollars and completed less than 9 months ago,” she wrote after the second malfunction.

The bridge had malfunctioned seven times in two weeks in March prompting more angry emails from the mayor, including this one on March 27:

“Holy Smoke….. AGAIN???!!!

“Twice???!!!

“What are we going to do?”

Faulty switch?

Parker replied with another apology and said, “Our saving grace is that the actual design engineer was here checking out the malfunctions of yesterday.” Attached to the email was a grainy photograph of the bridge’s inner machinery: “For now, though, the position of this guy” — a limit switch – “seems to be a great interest.”

But the design engineer’s presence that day apparently didn’t solve the problems. On the night of March 30, the bridge broke down twice again — at 8:30 p.m. for 90 minutes and at 11:22 p.m. for a little more than an hour — because of more issues with the span locks.

“Oh my heavens! It Just Doesn’t Stop,” Coniglio wrote in an email to Courtney Drummond, the FDOT’s chief engineer.

Desperate for answers, engineers wondered whether some of the latest problems might’ve been caused in part by a faulty limit switch that was used during the adjustments made on Feb. 20.

“We believe that the…limit switch may not have been triggering consistently while we were fine-tuning the system on February 20, masking the effects of our adjustments,” according to minutes from a design meeting.

Engineers adjusted the torque and the timing of the machinery back to the original settings and swapped out the limit switch, changes that seemed to work for a while.

But just after the start of the afternoon rush on April 11, the northeast span failed to close properly, shutting down the bridge for nearly 90 minutes.

“According to reports from the site, the northeast leaf is misaligned transversely to the north,” potentially the result of the adjustments made on Feb. 20, according to an email from Mike Sileno, a principal engineer for Hardesty & Hanover, which designed the bridge.

‘DOT proper is lacking expertise’

That malfunction, the 20th in the eight months since the bridge’s four lanes opened, instilled a new sense of urgency in project officials.

“I think we all need to come together as a group and find a permanent solution to what is taking place (bridge malfunctions). We all rode the limit switch reason as a possibility for a while only to have nothing come of it,” Parker wrote in an email April 12 to FDOT officials.

“If what I am hearing is correct and the inherent ‘walking’ of the bridge is the issue, then we need to find out why this bridge has an intolerance to such movement and how we deal with it for the next 50-75 years,” he said.

“While I’m not typically an advocate of my next suggestion, I think scrutiny from an outside engineer and certainly central office expertise may be of benefit here.”

The email also pointed out that PCL’s contract with FDOT was ending on May 8, “so we have to act quickly.”

Sardinas agreed that “DOT proper is lacking expertise to help you on this matter.” Parker reached out to Thomas Andres, FDOT’s assistant state structures design engineer, about a plan to gather experts “to come up with a best course of action and to discuss the design.”

“For obvious reasons,” Parker added, “I think experts outside of those who have already looked at and touched this bridge would be best to take an impartial look at all of the facts.”

FDOT tapped the engineering firm WSP, formerly Parsons Brinckerhoff.

On April 20, four days before WSP’s site visit, the bridge broke down for 73 minutes because of the same issue with the leaf. That prompted another flurry of complaints from angry commuters, including this one:

“It is a BRAND NEW bridge. It should be working perfectly. Perhaps there should be a ‘lemon law’ for highway structures,” resident Jim Reidy wrote to the town.

‘Eroding public trust’

Reidy’s email was forwarded to David Sadler, director of FDOT’s office of construction in Tallahassee. Sadler passed the complaint on to Shaun Yancey, head of PCL Construction Enterprise’s North American division.

“I’m writing to ask you if you are aware of the mechanical malfunctions that have been plaguing the Flagler Memorial replacement bridge and if so what can be done to remedy the matter? These malfunctions are significantly impacting public users of the bridge (as indicated in the attached) and eroding public trust in the bridge,” Sadler said in the April 23 email to Yancey.

Yancey replied two hours later, “This is the first I have heard of this issue” and promised to look into it. 

His query produced a dialogue with top state officials and a response from Zuleta, PCL’s project manager, who seemed to place blame on an unnamed state engineering consultant:

“These malfunctions started to occur after the (bridge’s engineer of record, Hardesty & Hanover) and the system’s subcontractor (Electro Hydraulic Machinery) adjusted the control settings to comply with a request/preference from the owner’s engineer. Prior to these recent control setting adjustments we did not have a malfunction of the bridge for over two months.”

However, PCL’s own log lists a malfunction on Jan. 31, just three weeks before those settings were adjusted on Feb. 20.

Zuleta added that the “team is working diligently to resolve” the problems and had “significantly reduced the frequency of these events.”

Yancey forwarded Zuleta’s email to Sadler and apologized “for the inconvenience this issue is causing the public of Palm Beach. You have our commitment to continue to work this problem until it is resolved.”

‘Problem may be with operators

With three WSP engineers watching from the control room in the bridge tender’s booth, the bridge worked fine during the firm’s April 24 site visit. Although the engineers reported no major problems, they noted that an “alarm monitoring system reflected a host of errors ranging from oil temperatures to lock malfunctions,” WSP supervising engineer Todd Mitchell wrote in a two-page memo.

“During the visit, a PCL foreman informed WSP that it appeared certain operational issues occurred when the bridge tender attempted to proceed to the next step without waiting for the completion of the previous action,” Mitchell wrote. But such actions, he said, “should not affect the performance of the bridge or cause malfunctions. There should be a series of safety interlocks that prevent the operator from advancing to the next step.”

The report also criticized PCL’s record-keeping, noting that the contractor’s log didn’t detail specific causes of each “malfunction or what bridge operation or alarm preceded the malfunction. The list was also lacking details on how the bridge operation was resumed after the malfunction.” WSP also didn’t know “if all malfunctions are being reported.”

After the WSP site visit, the bridge malfunctioned at least four more times starting on April 27 when it went down at 4:01 p.m. for more than an hour because of what the log described as “weather event.”

Although strong thunderstorms did hit the area that day, the notes from a May 2 bridge design meeting, under a section called “Malfunction on 04/27/18,” offer a different explanation:

“Department has a concern with all the adjustments to numerous overrides when malfunctions have happened that perhaps the original logic is no longer applicable,” the minutes say. “Point being: Bridge tenders are overriding logic controls to seat the bridge that should not be able to be overridden. … The problem may be a problem with operators.”

If operator error was a significant factor in the bridge’s many breakdowns, the project team downplayed that notion.

When asked by The Post, they blamed the operator just once — a 21-minute delay on Feb. 23. They also said that none of the malfunctions listed in the log were exacerbated by inexperienced bridge tenders.

PCL picked the company, ISS Facility Services, that employs the Flagler bridge tenders. When the state takes ownership of the bridge, ISS will continue providing bridge tenders, DOT spokeswoman Barbara Kelleher said.

‘Not a pleasant situation’

On May 8, after the mayor said frustrations among town residents had reached a “boiling point,” Parker appeared before the Town Council to explain the chronic breakdowns. He introduced himself as “the engineer with a very humbled and sheepish look on his face.”

“It is somewhat embarrassing that we’ve given you a landmark bridge and it doesn’t operate more than a couple weeks at a time thus far without something taking place,” he said, adding that he understood the public’s frustration: “I’ve seen the faces of people making turnarounds on the bridge and it’s not a pleasant situation at all.”

He said inspectors found no misalignment problems or issues with the bridge’s machinery. And he defended the bascule bridge as having a “dependable design.”

“A hundred years ago,” he said, “it was a bridge tender flogging a team of mules to go ahead and pull the pulley that operates the bridge.”

“Do we possibly need that?” Coniglio quipped.

Parker, a native of London, used a British idiom as he continued his explanation: “Now we have motors and sensors and generators. Although it’s the same kettle of fish, it’s a different kettle of fish and we just need to get that kettle of fish boiling correctly.”

Parker told the council that engineers were focusing on “the controls and the operations, which is as simple as the software that runs the bridge and … the people who run that software.” He said PCL and the DOT would make sure “through repeated training, if necessary, that the bridge tenders know how to operate the bridge properly.”

His answers didn’t offer much solace to the council.

“It is a bit daunting not to know when likely this thing is going to be fixed, since all of this seems to be within your power to correct by basic training measures or replacement of software or what have you,” Councilman Lew Crampton said.

Parker replied, “I won’t rest until the bridge acts and functions as it should. There is no reason why a brand-new bridge should be exhibiting these types of symptoms.”

The bridge malfunctioned three more times since that meeting, the last on June 7 when a faulty relay switch prevented the southeast leaf from lifting at 7:25 p.m., causing a nearly two-hour delay.

Ever since, the bridge has opened and closed without delay.

Perfect score on traffic maintenance

On Monday, PCL is expected to hand the bridge off to the state, officially ending its involvement with the six-year project. And the company is in line to score 94 out of 104 possible points on the FDOT’s Contractor’s Past Performance Rating, a nine-category system that factors into a company’s chances of winning state contracts.

PCL scored a zero out of four possible points in a category about the use of subcontractors owned by minority and women. It scored 4 out of 10 possible points on a category about coordinating and cooperating with “construction engineering personnel, property owners and utility companies.”

The company got perfect scores on all other categories, including 12 out of 12 on “Proper Maintenance of Traffic & Minimize Impacts to the Traveling Public.”

An FDOT spokeswoman did not respond to requests for comment about why the state gave PCL high marks for the company’s performance on a bridge that chronically malfunctioned over its first 11 months.

Now, instead of pointing them at the state and the contractor, town officials are crossing their fingers that the bridge will work fine for the next 75 years.

“There was some frustration expressed, but I think we have moved past that, knock on wood,” Taylor, now the town’s acting police chief, said in an interview.

“It’s a fantastic bridge,” she added, “when it’s fully operational.”

Still, for some commuters, the damage from its dismal debut has been done.

“It’s always in the back of your mind as you are about to cross the bridge — am I going to make it over or not?” said Zenko, a Palm Beach Snowbird. “It interferes with your life in that way.”