Speaking as a lead engineer in a study of the Port of Portland, Oregon State University’s Dr. Armin Stuedlein sat down to talk about blasting into the soil to test the solidity of silt, as well as the different types of waves an earthquake can make.
He also talks about being better prepared for the “Big One” that we’ve been hearing about for decades, and which structures we might not be able to rely on when it hits.
TCA: Hi, I’m Sally Lehman and I’m with The Advocate. Today, I’m speaking with Dr. Armin Stuedlein of Oregon State University about a remarkable feat of engineering which will save the Port of Portland up to $50 million in their aim to be more earthquake ready. Thank you for being here, Dr. Stuedlein.
Stuedlein: My pleasure. Thank you.
TCA: With a master’s degree from Syracuse and a doctorate from the University of Washington, you have spent a large part of your career looking at how soil turns “liquid” during earthquakes. Can you tell us how that happens?
Stuedlein: Well, I’d like to start off by saying, although we’ve made great strides in the last 50 years in studying soil liquefaction, there are many aspects of the phenomena that require continued study. And in part, that’s what triggered the work at the airport, particularly when soil has the potential to liquefy at very large depths, such as the depth of interest at the Port of Portland along the Columbia River.
But the process of soil liquefaction is essentially triggered by the soil being subjected to shear waves, which are generated by an earthquake fault zone. So when that fault begins to slip, it sends several kinds of seismic waves through the parent bedrock, eventually reaching the soils overlying that bedrock, and then ultimately finding the ground surface. But of the types of body waves that an earthquake produces, it’s the shear waves that will turn a soil into liquid under the right conditions.
So when the soil is sheared, you can imagine particles that are in some kind of loose arrangement trying to find the denser arrangement under the loading. When soils are saturated, you’ll find that the porewater within the void prevents the movement of the particles into that denser state. And so when the particles are rejected by the porewater, which is incompressible, the pressure within that porewater rises. And when the pressure gets large enough, it can lead to partial or complete loss of soil strength.
And anything supported by the soil, such as structures or a runway, will begin to deform and sink and potentially move laterally under sloping ground as the soil loses its strength and the entire mass and supported structure begins to move.
TCA: When you talk about shear waves, you’re talking about S waves as opposed to P waves, correct?
Stuedlein: That’s correct.
TCA: And what are P waves?
Stuedlein: Along the rupture surface are compression or P waves. These move very quickly and can pass through both solids and liquids.
The second event that happens during a rupture of an earthquake is the emanation of S waves from the fault zone, and these move more slowly and will only travel through solid materials. Soils will experience loading due to both P and S waves, but compression waves, from earthquakes themselves, tend to be of lower amplitude, and perhaps attenuate more quickly than the shear waves.
The reason why we can have early earthquake warning systems put into place, which are just now coming online along the West Coast, is because those P waves move so much quicker and they give us time to shut down hospital operations, kick on generators, and give enough warning to perhaps exit buildings prior to the damaging arrival of these S waves or the shear waves.
TCA: How does sound impact soil strength and consistency?
Stuedlein: That’s a really interesting question. There are two ways I think I can respond to that.
Number one, geotechnical engineering is as much of an art based on experience as it is rooted in science and engineering. And that’s because our materials are geologic in nature and they are not manufactured or fabricated like steel or concrete. So we tend to use the word sound when we think about sound geotechnical engineering practice, which is an implied reference to the art and experience side of our field.
On the other hand, I can think of P waves as being acoustic waves or acoustic signals and so we can think of waves as being sound waves because these compression waves can be heard. On the other hand, shear waves are more difficult to be heard. But certainly, when we think of listening to an album or listening to a concert, we are experiencing these kinds of waves that travel through the air.
TCA: You were looking at events in the 8.0 to 9.0 point earthquake range and being prepared for that. The Loma Prieta earthquake in San Francisco registered at 6.9, and the Northridge quake in Los Angeles registered at 6.7. Are we really expecting an earthquake as large as an 8.0 or 9.0?
Stuedlein: Oh, my, absolutely.
Although Northern California will be impacted by the Cascadia Subduction Zone, most of the faults that are present within the state of California are shallow crustal faults, and these faults tend to be of shorter length. We can think of the length of the fault as to be correlated to the magnitude of the earthquake that can be produced.
So the Cascadia Subduction Zone runs about 1,000 kilometers from Northern California up into British Columbia. And because there’s a potential for the entire fault zone to rupture simultaneously, we can expect significantly greater energy to be released as a result of that.
The main question that researchers have been looking at from the seismological point of view is whether we would expect a partial rupture of the Cascadia Subduction Zone or a full rupture of the Cascadia Subduction Zone. My colleague Chris Goldfinger here at Oregon State University has attempted to quantify the events associated with partial and full rupture.
So depending on whether we do have a partial rupture, we would expect an earthquake to lie at the lower range and in magnitudes perhaps 8.0, [8.5] or 8.7. And as the entire portion of the fault zone ruptures, we could expect even as large as 9.3 or perhaps 9.5 magnitude earthquake. So it all kind of depends on what the level of stress that has been built up along this fault zone has been and how that built up shear stresses distributed along the fault, and whether the fault would rupture over the entire length or a portion in order to relieve the most highly built up stress-built-up area.
TCA: You conducted three days worth of testing at the airport. How was the testing conducted and what led you to that type of testing?
Stuedlein: The exciting portions of our work were indeed conducted over three days of testing, but our work actually initiated some five months prior to the actual fun stuff.
So our program consisted of exploring the subsurface and quantifying the stratigraphy both through what we call “in situ” tests, as well as by retrieving samples and performing laboratory tests. Then it took several weeks to install our instruments and to test the ground and the instruments for, number one, understanding and verifying that we had indeed installed our instruments in the proper location. We had to then back out the orientation of our instruments in terms of azimuthal coordinates and orientations.
When you measure shear waves – or P waves for that matter – you’re basically measuring two things. You’re measuring the velocity, but you need to know the time between your measurement points and you need to know the distance between your measurement points. The timing comes from our blast’s, but the distance is something we have to establish well beforehand in order to have good controls on our measurements.
So, there was five months of solid field work that was conducted in order to set the stage for those three days of blast testing. But ultimately, the first test blast occurred on the 3rd of October, 2018. And that was ostensibly conducted to check our data acquisition system, make sure that we were getting everything we needed to get, adjust the gains and sampling details, and also to, perhaps most importantly, establish how energy would attenuate at the site in order to make possible adjustments to our experiment the following days.
We did have several structures within five to six hundred feet of our test site, including the FBI headquarters. And so we had to pay special attention in order to, number one, assure structural integrity of those adjacent structures, but also not to alert or otherwise alarm the residents of those structures when we conducted the larger tests.
So after that first test, on that first day, we were able to make refinements to our blasting program at the Port of Portland, which we did, and we were able to take some really good high level data to help us inform how we would interpret the remainder of the tests and then simply gave us confidence to do forward with its 30 second blast on the 4thand 5th of October.
Blast then occurred the next day after that test blast, and that consisted of 30 seconds of charges of alternating size and distance from our array. And this allowed us to put seismic energy into the ground, make the observations of these shear waves, and understand how the soil would interact with the shear waves in terms of the buildup of what we call excess pore pressure or the water pressure within the soil that leads to liquefaction. And so we were able to then relate the shear strain due to the shear waves to these excess pore pressures that were generated, allowing us to quantify how a very large volume of soil would behave to these damaging S waves.
And so that first main task was focused at a depth of 90 feet within liquefiable, medium dense sand deposits, which are saturated and extend to some great depths. As I said, we tested at 90 feet, but they can locally extend deeper at the Port [of Portland] properties.
And then on the third day of testing, we repeated a very similar test program, but at a depth of 40 feet. And the plastic silt materials that overlie those sands, which can also behave poorly in an earthquake under the right conditions, or perhaps you might say the wrong conditions.
[TCA: As a result of the tests, it was decided to use deep soil mixing rather than vibro-compaction. Why was that?]
I would like to make sure and point out to our readers and listeners that the work that OSU provided was basically a fundamental or basic study of how soil behaves to dynamic loading. Our test results were provided to the Port and used by the Port engineers and their consultants to select a particular ground improvement methodology. And although we have engaged in discussions with the Port about the advantages and disadvantages of certain mitigation techniques, OSU really couldn’t claim responsibility for selecting that particular ground improvement methodology.
In my role as a researcher, I don’t make recommendations for engineering solutions, but rather present the facts of how the soil might behave during an earthquake based on the results of our findings. So you can imagine one aspect that we engaged in was characterization of soil behavior, whereas the engineers of record, which is a legal term required for those making recommendations for construction, would select the appropriate mitigation strategy.
But, being relatively well versed in ground improvements, I can imagine why deep soil mixing could be elected. It is a more expensive ground improvement methodology for sure, but it provides a higher level of risk management because the ground improvement technique produces a stronger post improvement condition. Stiffer will provide better performance, which is critical when you’ve got a very thick – potentially five to six foot thick – reinforced concrete runway that needs to overlie these very weak and liquefaction susceptible soils. So, if you can use a stronger ground improvement technique, you can provide greater tolerances on performance, greater post earthquake performance predictability. And I imagine that that led the Port and their engineers to select that particular approach.
It’s worth pointing out that the runway will be our number one key resource in order to bring in rescue supplies and materials for our entire western Oregon region. So if there’s one place to spend money to ensure community resilience, the runway certainly is it.
TCA: You’ve said that this shows that blasting experiments are useful for soils that are not easily sampled. Where else in the Pacific Northwest would you expect to be able to put this process to work?
Stuedlein: One of the reasons that geotechnical engineers are faced with such uncertainty in our profession, and hence the reliance on experience more than in other civil engineering disciplines, is because it is very difficult to obtain samples of our material that are sufficiently undisturbed that we can obtain reliable engineering behavior in the laboratory. Again, our specimens aren’t fabricated or manufactured like steel and concrete. So we need to take great care in order to obtain samples of material.
Although we might take great care, we will never be able to obtain a perfectly undisturbed sample without extraordinary measures such as freezing the ground prior to sampling. And it’s worth pointing out that in the early discussions with the Port, we pointed to ground freezing and sampling as an alternative. However, for the cost of our in situ tests at the Port properties, you would only be able to freeze maybe two or three boreholes, and get maybe a half a cubic foot of soil material in the areas of interest for that effort. So we were able to test a much larger volume of soil in situ – undisturbed – for the same amount of money that such an extraordinary approach would have taken.
What other soils can we perform our technique with? Well, in addition to the sands that we described in the silt at the Port, we’ve also applied our approach to the silts of the Port of Longview. And additionally, at that location, we tested much more shallow because we used another source of seismic energy called T-Rex, which is a very large vibersized mobile shaker from the University of Texas at Austin. And we collaborated with those folks to evaluate the seismic performance of those materials. And we were able to validate our approach by a direct comparison with the response of this T-Rex mobile shaker.
But anywhere along coastal estuarine environments where you have very soft nonplastic silts. The nonplastic silts can be sometimes difficult to sample, particularly when there’s no element of clay mineralogy within the soil matrix, then sampling becomes very difficult.
Likewise, under certain conditions of rock, it may be very difficult to understand how a rock mass might behave because we can get core samples, but if the rock mass is jointed or fractures, you’re not going to be able to replicate that kind of large scale behavior in the laboratory very easily. So I can think of a number of geological settings in which samples are not going to be representative and in which our test technique can be useful for understanding fundamental, dynamic behavior.
I think the key thing is that we test large volumes of material in situ without disturbance, whereas in the laboratory we’re kind of subjected to the whims of sample size and sample disturbance and other issues associated with laboratory test apparatuses.
TCA: What do you feel are the most at risk structures in the Portland Metro area when the Cascadia earthquake hits?
Stuedlein: That’s another great and really relevant question. You know, anywhere where we have liquefiable soils, I think this is where I can speak the most profoundly about.
We’re going to have issues. So I can think of the fuel storage tanks along the Willamette and Columbia River. Any kind of hydraulic fill material – these are human placed materials that were largely done between 1900 and maybe 1940 or so – these were fill materials that were dredged from the Columbia River and then sluiced into place. Any structures that are supported on these materials without engineering protocols to help support them will be subject to significant seismic damage, or at least the potential for seismic damage. I’m not a structural engineer, but my structural colleagues are really concerned about unreinforced masonry structures in the Portland Metro area as well as here in Corvallis.
TCA: Are there structures at risk in Corvallis?
Stuedlein: In Corvallis, we see that there is a range of buildings that were constructed in the late 1800’s up to the present day. Our seismic building codes really did not even consider such elements of seismic loading until maybe the early 70s, and then the recognition of the Cascadia Zone being active really didn’t take off until the early 90s or so.
And so our building codes have to follow the science and the work of our seismologist colleagues and geomorphologist, and essentially our geological sleuths who pieced out the seismic history of our area. So without being able to point to a particular structure, I’m fairly certain that here in Corvallis we do have buildingstock that could be subjected to poor seismic performance during the Cascadia.
TCA: What level of isolation would Corvallis likely face if a large earthquake occurs?
Stuedlein: I guess that depends in part as to when our earthquake comes. We know it will come. It is uncertain as to when it will come.
If the Van Buren Bridge replacement can be put into place prior to the Cascadia Subduction Zone [earthquake], we should have a structure crossing the Willamette River during or in the time immediately following the Cascadia Subduction Zone earthquake.
But at present, it’s likely that the Harrison Bridge and the overpass south of downtown may be out of service following the Cascadia Subduction Zone. At that point, we would head north to Albany, but we would likely find that the bridges crossing the Willamette in Albany might also be out of service.
So the potential for being somewhat isolated is large. And the community should prepare for weeks of isolation prior to being able to expect significant recovery efforts to put these bridges back into service. So we need to all do our part and prepare for the eventuality and help each other should that event actually occur.
TCA: What do you recommend our readers do to prepare for a large earthquake?
Stuedlein: I would suggest several things.
For those with natural gas hookups, there are seismic shutoff valves that we can each put on our homes for about $150. That will prevent fires which are extremely common after earthquakes and is shown time and again in, for example, the ‘89 Loma Prieta, the ‘94 Northridge, and even the most recent Ridgecrest earthquake in California. So the seismic shutoff valves for our residential natural gas or even commercial natural gas is a really cheap mitigator to prevent fire and the spread of fire in our neighborhood.
Having a water supply for probably three to four weeks on the safe side would be highly recommended, as would stockpiling of batteries for flashlights and radios, dry foods that can last a good portion of time, or canned goods. Having those ready to go would be very good.
My colleagues and I, we tend to all have our gas tanks always more than half-full so that we have fuel in the tank should we need to travel for whatever reason following an earthquake. Something just as simple as always having a full tank of diesel or gasoline in your car could be quite helpful.