Post-quake structural damage isn’t so much a building problem as a ground problem, engineers say

Officials credited strong building codes for the collapse of very few structures in the magnitude 7.0 earthquake that shook Southcentral Alaska on Nov. 30.

But the less obvious structural damage, including cracked foundations and tilted walls, that did occur really isn’t a building problem, based on interviews with a half-dozen structural and geotechnical engineers. It’s more a problem with the ground beneath the buildings, the engineers say — and the ground failed in some surprising ways that may be rooted in the use of man-made fill rather than natural slides.

That’s a growing consensus emerging from the geotechnical community as Anchorage and Mat-Su take stock of quake damage. More than 100 buildings in the Anchorage area, including Chugiak-Eagle River, have been declared unsafe because of structural damage caused by the earthquake. More than 180 additional structures suffered damage.

[New maps and data show widespread Anchorage earthquake damage and more than 100 unsafe buildings]

There are tens of thousands of buildings in the Anchorage Bowl, and officials have expressed relief the damage wasn’t more widespread. But engineers warn the extent of the damage hasn’t yet been fully realized.

Meanwhile, Anchorage’s buildings are the subject of intense interest to both local and out-of-state engineers, as a lesson in how to build — or not.

Out-of-state team flies in for a look

A national team landed in Anchorage several days ago to study the Nov. 30 quake after doing similar work around the world.

The chief goal is to look at what worked and what didn’t, said Kevin Franke, a civil engineering professor with Brigham Young University and the team leader. Called the Geotechnical Extreme Events Reconnaissance Association, or GEER, the team also includes faculty from the University of Nevada Reno, Oregon State University, North Carolina State University and the University of Alaska Anchorage.

[Your support makes stories like this possible. Buy a digital subscription for 24/7 access to the Anchorage Daily News]

Franke personally has traveled to Japan, Central Italy and Mexico City since 2016. Damage from every earthquake takes on its own flavor, he said, speaking by phone Wednesday on the road to meet with officials in Wasilla.

After a 6.6 quake struck central Italy and caused extensive damage, Franke concluded the main cause was the type of structures and construction practices from hundreds of years ago. Shaking a non-reinforced building very hard will cause it to fall down, he said.

The soil beneath the Italian structures fared well, Franke said, because the structures were built on solid rock.

In Anchorage, by contrast, the team has observed that building codes are generally working, Franke said. He noted there had been almost no instances of full-scale building collapses.

Instead, damage can be widely attributed to earth that moved, either sinking or shifting horizontally, Franke said.

“It’s becoming clear to us that this particular event is a geotechnical event,” Franke said.

The amount of ground failure surprised Jesse Gobeli, a senior engineer and principal with PND Engineers, who has inspected a number of businesses and homes since the earthquake.

Residential areas seemed to be bearing the brunt of the damage, Gobeli said. He said he’d seen pockets of cracked or collapsing ground all over the city, from Sand Lake and Jewel Lake to Turnagain and the Dowling and Old Seward area. Some buildings dropped as much as a foot, he said.

“That tends to be the story of this earthquake,” Gobeli said. “Ground failure in areas I don’t think people were expecting.”

‘Seismic hazard maps' show something different

One of the surprises of the quake for Buzz Scher, a longtime Anchorage geotechnical engineer who came out of retirement to work for the city, was what appeared to be very limited damage in the parts of Anchorage most associated with earthquake damage.

U.S. Geological Survey scientists found only small cracks last week in a tour of the sites of big devastation in the 1964 earthquake, such as Turnagain, Earthquake Park, Fourth Avenue in downtown Anchorage and up by Government Hill, Scher said.

But those are the areas labeled as highly susceptible to ground failure in a widely circulated Anchorage risk map labeled “Historic Seismic Zones.” The map has legal weight: New construction in high-risk areas, particularly near Earthquake Park, must undergo additional geotechnical engineering work.

Other areas that saw sinking and foundation issues in the Nov. 30 earthquake, including Sand Lake, are marked as “moderate” on the map but not to the level of needing extra engineering. Meanwhile, the map shows a “moderate-low” seismic risk in Chugiak-Eagle River, the area where a significant amount of the more serious damage occurred.

Experts caution against relying on the map as a guide to the areas of highest quake risk.

Calling it a “seismic hazard map,” for starts, is a bit misleading, said Barrett Salisbury, a state earthquake geologist. He said the map actually reflects the likelihood of ground failure throughout the city.

The original data for the map was collected in 1979 and digitally compiled and republished in 1997, likely based on observations made after the 1964 earthquake, Salisbury wrote in an email. He said the map highlights areas where native soil is likely to slide away or collapse with strong shaking.

“The map doesn’t really provide any information about the likelihood or intensity of shaking,” Salisbury wrote. “Just, when it shakes again, what natural materials are likely to shake loose.”

Seismology data showed that some of the consistently hardest shaking hit Chugiak-Eagle River, which the map does not reflect, Salisbury pointed out.

Meanwhile, the Nov. 30 earthquake did not shake hard enough or long enough to cause slides in the more historic areas, Scher said. The quake recorded only about 30 seconds or so of the hardest shaking, as opposed to several minutes in the Good Friday quake.

Common practice

Engineers also agree that the ground that collapsed or cracked was largely placed by humans, also known as “fill.”

John Thornley, who is the co-lead of the Earthquake Engineering Research Institute’s reconnaissance team, said human efforts to level out slopes seemed to be linked to a significant number of issues.

“I wouldn’t point necessarily to bad contracting, but some people cowboy in their own work, and those results may be showing,” Thornley said.

Scher said that in his observations in Eagle River, he had only seen one apparent slope failure on native soil. He said the vast majority seemed to be taking place on human-placed soil.

Ross Noffsinger, the city’s acting building official, said it’s very common for trucked-in soil to be placed improperly. Homeowners who live on a slope and want a flat backyard can go to the big dirt pile off O’Malley Road, pick up truckloads of dirt, and dump it in their yard to level it off, Noffsinger said.

But there are detailed geotechnical engineering requirements for doing that. Proper placement involves removing organic material, meeting the right specifications and compacting it properly. That doesn’t always happen, Noffsinger said.

Noffsinger said the problem is generally homeowners who don't know the rules. While city inspectors regularly discover instances of improper ground placement, there are likely many more the city isn’t aware of, Noffsinger said. He said the discoveries usually happen because of complaints called in by neighbors

Keith Mobley, the president of Northern Geotechnical Engineering, said he’s actually seen less damage than he expected so far in his inspections in Chugiak-Eagle River. But he said he observed between 30 and 40 instances of what appeared to be cracking in human-filled soils, particularly on steep hillsides.

He said he suspected bad fill, arising from the efforts of homeowners or builders to save costs.

“It’s a money issue,” Mobley said. “It’s 100 percent money.”

Danny Graham is a senior structural engineer with LCG Lantech. In the past week alone, he’s tagged seven homes in the Chugiak-Eagle River area as uninhabitable. He also said it appeared human-placed organic materials were causing the more substantial issues, including big lateral shifts of the earth.

Franke, the BYU professor and member of the national reconnaissance team, said solid geotechnical engineering practices are essential in the Anchorage area, where there are challenging soil conditions.

“Don’t cut any corners, and then generally, it’s going to perform OK,” Franke said.

Vine Road likely a fill problem

Vine Road, a shortcut between Big Lake and Wasilla, was built several decades ago over a bog. A widely-circulated photo from the Nov. 30 quake showed a section of the road shattered into pieces. (On Sunday, the road reopened to traffic.)

Kyle Brennan, the vice president of the engineering firm Shannon Wilson, has been looking at buildings, roads and bridges for state and local officials. He said the problem of poorly placed organic material applies to roads, including that segment of Vine Road.

Brennan did some geotechnical work on Vine Road about five years ago as part of a state upgrade project. He said the embankment in the section of road that collapsed was built at a time of limited stabilization methods.

“Going into that project, we knew there were a few problem areas,” Brennan said. “And that was one of them.”

In general, that embankment was built over a lot of water and poor soil quality, Brennan said.

But the tremendous expense of redoing the embankment on a relatively remote, rural road in anticipation of a major earthquake would have killed the entire upgrade project, Brennan said. Much of civil engineering, he noted, is a cost-benefit dance.

Brennan is also puzzling over why other, similar embankments in the area held up just fine.

Those are the types of questions that engineering groups are starting to dig into, but conclusive answers likely won’t be available for months. Noffsinger, the city’s acting building official, said his teams are still backlogged with inspection requests. Several months into the repair process, the city will piece together data on what worked and what didn’t, he said.