The City's inedible golden apple

by David Broadland, July 2010

Has the seismic risk to the Johnson Street Bridge been over-stated? And is the City’s response to that risk an over-reaction?

American bridge engineer Frank Nelson has been asked to travel to Kansas in late July to train a group of that state’s Department of Transportation engineers how to correctly compare rehabilitation and replacement options when they are considering the fate of a bridge. This particular group of engineers were judged to have not given a “creditable evaluation” of rehabilitation while pushing ahead to replace the Amelia Earhart Memorial Bridge across the Missouri River. As part of the required remediation for removing the bridge, they have to take Frank’s course. The engineers currently considering the fate of the Johnson Street Bridge might want to take notice.

Perhaps you remember Frank Nelson. He’s the bridge expert that johnsonstreetbridge.org invited last fall to look at the Blue Bridge and share his considerable experience of evaluating whether a bridge should be replaced or rehabilitated. He’s currently working out of Seattle as Senior Supervisory Engineer at Parsons Brinckerhoff, an engineering firm with offices around the world. But Nelson also spent 14 years managing Oregon’s Department of Transportation Bridge Preservation Engineering Team, which oversaw the rehabilitation of many aging bridges before he left the ODT in 2007.

I corresponded recently with Nelson on the question of whether the $62,000 condition assessment performed by Delcan Corp in two days back in June 2008 was a strong enough foundation on which to build a case for what has recently become, potentially, a $100 million project.

The Johnson Street Bridge Condition Assessment Report, conducted by Delcan and released last year, found a number of problems with the bridge, many of which are readily fixable. The cost of recoating the bridge, renewing it’s electrical system and making mechanical repairs was pegged by Delcan at $4.5 million. The most damning finding was Delcan’s apparent conclusion that timber piles supporting two of the bridge’s piers might fail in an earthquake, and that their failure would quickly lead to the bridge’s collapse.

That original $62,000 study has led to the expenditure of over $2 million in various follow-up studies, cost estimates and public engagement, and now to an updated estimate to replace or fix the bridge that many Victorians feel is vastly inflated and a foolish over-reaction. So the quality of Delcan’s assessment of the expected seismic performance of the timber piles supporting the bridge is as critical to the foundation of all that followed as the piles themselves are to safely supporting the bridge.

For many Victorians, Delcan’s revelation that the immense weight of Big Blue’s steel and concrete was supported by wooden posts came as a shock. It was easy to imagine that after 85 years those posts would have begun to decompose. And if an earthquake shook the bridge, well, of course, it would fall down. As Mayor Fortin said at the time, “Any seismic event will bring it down.”

But is that really the case? Two of the concrete piers supporting the Johnson Street Bridge are in turn supported by row upon row of what are thought to be Douglas fir timbers, probably but not necessarily creosoted for protection against marine borers, driven to what is believed to be bedrock, and completely surrounded by clay and marine silt. These piles are very similar to the ones that still support the Empress Hotel, completed in 1908, and the engineering practice of using driven timber piles to support concrete foundations is still used around the world. Frank Nelson says that Oregon’s Department of Transportation has made no moves toward replacing timber pilings on its historic bridges, one of which dates to 1913. The Oregon coast, where many of the bridges Nelson rehabilitated are located, is in an area considered as seismically risky as Victoria’s.

Normally, there would be no reason to suspect Delcan had not properly assessed the seismic capacity of the timber piles. When an engineering firm makes an assessment, the rest of us usually listen and accept what they say as the objective truth. But a close reading of Delcan’s report reveals two things that are perplexing.

First, the report actually says different things on different pages about the seismic capacity of the bridge. On page 4-1 the report states “In addition, analysis was carried out to determine what level of earthquake the bridge, in its current configuration, can resist without significant damage...this corresponds to a seismic event with a 35 percent chance of exceedance in 50 years.” But in the report’s executive summary, on page E1, it states “Analysis of the bridge in its existing configuration shows that the bridge will experience failure of its foundations and collapse of the counterweight towers under loads from a seismic event with a 35 percent probability of exceedance in 50 years.”

This is not a level of consistency that instills confidence.

www.sterlingfinancial.ca

www.sterlingfinancial.ca

A second issue with the report is that it seems to rely entirely on computer modelling for its prediction about the seismic performance of the timber piles. The 232-page report includes not a single mention of any physical testing of the bridge that could be used to confirm their seismic modelling. Both Delcan and the City of Victoria were asked to comment on this but hadn’t done so by press time.

Frank Nelson says that even though all the bridges he worked on in Oregon had timber piles—which were next to impossible to inspect without damaging them—they didn’t rely on computer modelling to assess their expected seismic performance. “The problem with computer simulations” he said, “is that unless the specific simulation is calibrated through an actual event at the structure there is considerable room to doubt it’s validity. Not the quality of work, just the lack of applicable facts. When we do load ratings—a computer analysis method of confirming that each component of the bridge can carry its expected loading—on historic bridges, we had typically also done a live load proof test with strain gauges and a heavily loaded truck, with its axles weighed. By measuring strain and deflection we could compare this with load predictions. If the math model predicted this live load closely, then it would predict loading at a higher value to a similar accuracy, by extrapolation. Without a connection between math models and real performance, the math model is mostly an indicator of potential concern.” (Emphasis added.)

Delcan seems not to have done the compex physical testing Nelson would have done, so how else could the physical condition of the Johnson Street Bridge’s timber piles be assessed? Nelson suggested it’s not necessary to physically inspect the piles and in fact it could be harmful. Rather, if there is a problem, there will be symptoms that can be observed: “I would personally trust a precision survey, particularly of the railroad bridge, to determine if any movement has occurred since construction,” he said. The railway bridge has experienced much heavier loads—loaded freight trains have crossed it—and would have been more likely to settle or move if there is a problem with the timber piles. Other signs of problematic piles could be found, Nelson says, by “examining the live load shoes for shifts in wear, and alignment of the armor in the bridge deck joints at the free end of the bridge.”

Did Delcan do this work? Again, there’s no indication in their report that they did anything beyond visual inspection. The City of Victoria has not been able—or willing— to confirm this one way or the other. And if Delcan’s condition assessment was little more than a visual inspection, the problem is now being compounded. Various of the technical reports presented to Victoria City Council on June 14 credited Delcan for providing the seismic modelling data that these various engineering firms then ran through their computers and with which they reached, not surprisingly, similar (but more expensive) conclusions to Delcan’s. By the way, the City has estimated the cost of these various studies to be in the neighborhood of $400,000. One has to wonder if the GIGO factor—garbage in, garbage out—might be at play here. “Potential concern” is a long distance from “Any seismic event will bring it down.”

Curious about the extent to which Frank Nelson’s take on computer modelling might be shared by those engineers involved in seismic retrofitting of BC bridges, I contacted Sharlie Huffman, Bridge Seismic Engineer for BC’s Ministry of Transportation. I sent Huffman an exact quote of what Nelson had said and asked her what she thought. She said, “There is a definite skill to modelling although the current programs have improved a lot over the years. Still, whether it is a complex computer model or just a calculator, you need to have an idea what the answer is supposed to be in order to be sure the gizmo is working right and that you have entered everything the way you should. And all structural calculations must be checked. We can do this by running a few hand calculations or using two different computer programs. Of course if you put in the wrong information such as the wrong material properties, then GIGO.”

Huffman is a virtual library of information on the subject of the Province’s program of seismic retrofitting of bridges. MoT has been  retrofitting provincial bridges for a number of years using as a standard a “design earthquake” with a return period of 475 years. In more familiar terms, this is a seismic event for which “the duration of strong shaking...shall be consistent with earthquakes of approximate magnitude 6.5 to 7.0...” MoT gives bridges four different “importance” ratings and for each of these ratings there is a different performance expectation—which describes the hoped-for outcome after the design earthquake occurs.

For instance, the highest importance rating is “Lifeline,” and wthin that category the highest performance expectation currently in use (Safety 2) is “collapse prevention” and “repairable damage.” The recent seismic retrofit of Vancouver’s Lions Gate Bridge was executed to this standard.

But that’s a considerably lower standard of seismic performance than the standard Victoria city councillors (excepting Geoff Young) voted for at a June 17 council meeting. Their choice, which would add $10 million to the tab city taxpayers would pay, would see the bridge “open to all traffic after the design earthquake.” But in this case, the “design earthquake” would be a magnitude 8.5 earthquake as compared with the provincial standard of 6.5 to 7.0.

So why has Victoria City Council second-guessed the Provincial standard? Critics suspect engineering staff and Council of trying to make the rehabilitation option so unattractive that citizens will vote “yes” to replace the bridge in the November referendum. But the councillors and the mayor would probably say it’s because “The bridge is located in the most seismically active zone in Canada where it is estimated that there is a 30-35 percent probability of experiencing a major earthquake (in the range of M7.0 - M7.9) in the next 50 years.” That claim has appeared in two “Decision Request” documents authored by the bridge project’s Project Manager, Mike Lai. Lai attributes the “30-35 percent probability...” claim to Natural Resources Canada.

Focus contacted Dr John Cassidy, Earthquake Seismologist with Natural Resources Canada to see if that was an accurate assessment of the Johnson Street bridge’s seismic risk. Cassidy wrote back and said, “I would probably say ‘One of the most seismically active zones in Canada...’ As for the 30-35 probability—that is correct... but not necessarily for a M7.0-7.9. Rather that 30-35percent probability is for intensity VII shaking level. Intensity being the Mercalli Scale.” Dr Cassidy advised those interested in knowing more about the Mercalli Scale and what “Intensity VII” means to visit: http://earthquakescanada.nrcan.gc.ca/info-gen/scales-echelles/mercalli-e.... You’ll find this line as part of the Intensity VII description: “Damage negligible in buildings of good design and construction, slight to moderate in well-built ordinary buildings, considerable in poorly built or badly designed buildings, abode houses, old walls (especially where laid up without mortar), spires, etc.”

For those worried about the City’s dire prediction, it might be comforting to know that there are five levels above the Intensity VII level. Dr Cassidy added that, “Describing shaking in terms of earthquake magnitude is very challenging, as the shaking level for a magnitude 7 earthquake will be very different depending on how far from the earthquake you are... 10 km, 50 km, 100 km.... the strength of shaking drops off quickly.”

“Intensity VII” on the Mercalli Scale converts to roughly M6.0 on the scale that we like to think of as the “Richter” scale but is actually the “Moment Magnitude” scale (don’t ask). So what the City should be saying is that “There is a 30-35 percent probability of a magnitude 6.0 earthquake occurring somewhere in the surrounding seismically active zone.” Depending on the distance this earthquake occurs from the bridge, and the actual condition of its timber piles, it may fall down or it may not even blush.

That’s not to say the “big one” will never happen. It’s coming. According to Sharlie Huffman, the current thinking is that a Cascadia Subduction Zone earthquake of magnitude 8.7 to 9.2 has a 14 percent probability of occurring in the next 50 years. The epicentre could be anywhere along the 1200 kilometre length of the fault and the average time between events is about 500 years. The shortest interval is believed to be about 200 years and it’s been 310 years since the last “big one.”

The other kind of earthquake Victoria experiences from time to time is called a “crustal” earthquake. Huffman says these are the “trickiest as we don’t know where the crustals will hit or if they will be shallow (bad) or deep (like Nisqually).” The Nisqually earthquake, a magnitude 6.8 event 18 km northeast of Olympia, Washington, struck on February 28, 2002. Many Victorians felt it, but no real damage was done. The Johnson Street Bridge did not fall. Huffman notes the upper limit of this kind of quake was defined by the magnitude 7.2 earthquake near Strathcona Park in 1946, another seismic event Big Blue survived.

The City’s decision to reach into the stratosphere for a standard of seismic retrofit for the Johnson Street Bridge that goes far beyond that being applied by the provincial Ministry of Transport has earned that option the moniker “Golden Apple” from the City’s critics. That over-reaching, along with unanswered questions about the true nature of the bridge’s seismic risk, and City Hall’s penchant for over-stating the bridge’s seismic risk, have done little to improve the probability that civic politicians will resolve this problem before the next civic election. Which may prove to be the biggest seismic event in each of their political careers.

David Broadland is the publisher of Focus Magazine.