Common Geotech Questions from Contractors
Written by Mike Smith
January 26, 2023
Contractors depend on geotechnical engineers. You need a basic understanding of geotechnical concepts, so you can ask the right questions and make the most of the soil reports. But let's be real, geotech confuses most of us.
Randy Blount of Blount Contracting sat down with Mike Smith of Smith & Annala Engineering Company (SAECO) to ask the burning questions contractors have about geotech.
Here's what they had to say.
Question #1: Why are swell test results different?
Randy asks: I have a project where I'm trying to sell the material I remove. A few labs did swell tests to make sure we can use the soil at another site. I'm frustrated because two labs say the soil passed their swell tests, but one says it failed. How did that happen?
Mike and his team see this a lot—when multiple firms test a site, and nobody's results match. To answer Randy's question, Mike starts with swell testing basics.
What swell testing is and why it matters
Geotechnical engineers are concerned about how much clay is in soil, because clay changes volume. "You might be able to build something today, but a year or two from now, if moisture gets in that material, it's going to expand. Some materials can collapse,"Mike says.
Geotech engineers run swell tests to gauge how the soil responds to heavy moisture. For most sites, you want 1.5% swell or less.
The swell testing process
Swell tests are done in labs, so contractors can't watch the geotech team perform them onsite like field density tests. "We'd welcome anybody to come watch a swell test [in the lab], but it's pretty boring—even more boring than most of the stuff we do," Mike jokes.
For a swell test, the technician simply takes a soil sample and wets it down to see how much it swells.
Why labs disagree on swell test results
The jobsite the soil came from and the site where you want to use it as fill might have different geotechnical reports. Labs may interpret these reports differently and then disagree on swell test results.
Every firm also has its own test parameters. That's because technicians have to bring the soil to a certain level of compaction and moisture content in the lab. However, they don't know exactly how much moisture is in that lab-prepared specimen until after the test.
During the test, they estimate the moisture content. If the technician gets the soil too dry, it swells more during the test than onsite. Too wet in the lab, and it'll swell more onsite.
"To the credit of the industry, guys are pretty good at telling they're close to optimum moisture," Randy points out. Mike agrees: most techs can get extremely close. It's just a matter of whether techs from different labs get close enough.
How to confirm the swell test ran correctly
After the swell test, some firms dry the soil completely in an oven. Then, they compare the soil's density before and after it dried to learn how much moisture it contained during the test.
However, not all firms do this. Some companies say "eh, close enough"because they don't want to spend the time or money confirming the test.
That said, being even half a percent off from the optimum moisture content can mean the difference between a failing or passing test. Mike encourages contractors to ask for the test data or have their own geotechnical consultant review it to ensure the test was good.
Optimum moisture content for swell testing
Optimum moisture content depends on the job. Many jobs have an acceptable range called a moisture band. For instance, imagine the optimum moisture content is 2%. The moisture band could be 1-3%, and you can use soil in that range.
Randy says that sometimes, "We know there's some clay in the soil. So we say it can be used, but it has to be optimum moisture plus." That means the soil can only be at or above optimum moisture content, not below.
Soil that's too dry during construction could swell later and make the structure unstable. If the soil is at optimum moisture content or higher, it won't swell much post-construction, so you can feel more confident in the structure's stability.
How to handle existing soil reports
Some owners give the contractor soil reports and say, "We really don't think you need six feet of over-excavation and re-compact"or "We've worked right next to the site, and we think it's good material."
You as a contractor are not stuck with that soil report. Some geotech consultants don't do enough borings and testing, and their recommendations are off. You can get another opinion if something seems odd in the initial report.
Some commercial building contractors hire Mike's firm to do additional testing and reports. "It might be $5,000, but it could save you $100,000,"Mike says. Keep in mind those extra tests can take three to four weeks.
Randy adds, "Maybe right away saying, 'We're gonna do another soils report' isn't the right call. But [you need] somebody you can call and say, 'Does this seem out of the ordinary?' That's where collaboration and partnership really pay off."
As a contractor, you need good relationships with geotech firms who can help you make sure the job gets done properly.
Question #2: Why do nuclear density and sand cone tests produce different results?
Randy asks: I'm installing base material for a roadway, and the nuclear density test fails. My superintendent asks for a sand cone test, which is another density test we can do here in Phoenix. The lab does a sand cone, and we pass. How's that possible?
According to Mike, field density testing is geotech companies' bread and butter. He offers a few reasons different field density tests yield different results.
Incorrect math or too few proctors
Percent compaction is your in-place density divided by a laboratory density of the same material. (To get the lab density, the technician does a proctor. The proctor determines the soil's maximum dry density and optimum moisture content.)
Lots of numbers feed into the calculation for percent compaction. Each number has potential testing and subjectivity issues, so you could get different results that way.
Additionally, a five-acre site might only have one or two proctors. The technician may not have sampled all material types, so they may not have the right information to calculate the percent compaction. A firm that takes more samples could come back with different test results.
Labs determine maximum dry density differently—especially when testing rocky soil. Mike says, "If you have a certain amount of rock, that changes the way we do that laboratory proctor. If the laboratory isn't following or paying attention, they may run the wrong proctor."
The field technician has to know what method the lab used to find the maximum dry density. Then, they have to do a rock correction to make the proctor match the rock conditions onsite. They need that info to accurately calculate the soil's in-place density and percent compaction.
Mike adds, "Some testing labs . . . just use an assumed value of how much rock is there based on the lab sample, as opposed to actually doing a percent rock test on location." Companies using assumed values will get different test results than ones that actually know how much rock is in the soil.
Not all rock is created equal.
For example, Mike explains, "In the Phoenix area, we've got some pretty dense, hard aggregate. It's great for concrete and asphalt. It has a specific gravity of 2.6, but in other parts of the state, it might be as low as 2.2 or 2.4. That sounds pretty small, but when you work through the math, that could be the difference between a failing test and a passing test."
Different soil gravity may mean you need a different proctor for that part of the site. Mike suggests a one-point proctor, which is faster than a full-blown lab proctor because the technician can determine maximum dry density and optimum moisture content onsite.
Contractors and superintendents often ask for one-point proctors when compaction suddenly becomes an issue. However, not all geotechnical technicians know how to do that.
A one-point proctor occurs in the field, so technicians can't confirm the soil's moisture like they can in a lab. Instead, they have to be able to feel the soil and tell that it's close to optimum moisture content. That takes a really good technician.
According to Mike, one-point proctors are "pretty simple," but just like anything in life, they require the right tools and practice. At SAECO, Mike wants his team to do one-point proctors once a week. They also keep samples in the lab so technicians can practice before they test in the field.
"Especially our newer folks, they need to prove they can do that, because I have to sometimes put my professional engineering license on one-point proctors," he explains.
Nuclear density tests are common because they're easier and quicker than many other tests. Geotechnical engineers call them the "gold standard."
But most markets have an alternative density test. In Phoenix, that's the sand cone test, and it's the go-to for some agencies. "It can be frustrating when . . . all my QC tests were good with the nuclear gauge. Now QA comes in, and they're going to use the sand cone method. All of a sudden, they're not [passing],"Randy says.
According to Mike, these disagreements can happen simply because one firm made a miscalculation or didn't do a rock correction.
Miscalculated sand density
The sand cone test uses volume replacement. A technician excavates a small amount of soil, weighs it, and determines the moisture and rock content. Then, they fill the hole with sand. Since they know the sand's density, they can calculate the density of the excavated material.
But all sand was not created equal, either. Geotechnical engineers need to know the sand's density, which can change.
SAECO buys sand in 40 to 50 pound bags. "We empty those out in a garbage pail, and somebody calibrates the sand to get its actual pounds per cubic foot. If that number is off by just a couple pounds . . . you could get a false positive, false negative, or just a value that doesn't marry up to what you got using a nuclear density gauge,"Mike explains.
Another thing to know: sand cone tests don't account for moisture content, so geotech firms have to use a separate test for that. Adding another test to the mix can increase the chances of errors.
With so many variables to calculate density and percent compaction, it's easy to make mistakes—even if the testing company tries to do everything by the book. As a contractor, knowing why test results differ helps you spot potential issues.
Question #3: What should contractors know about proctors for big jobs, like dams and runways?
Randy asks: I'm working on a dam project. The labs already did two proctors. Is it okay to ask them to help me understand the results, and when should I ask for more proctors?
Simply put, big jobs need more testing. Mike and Randy discuss what contractors should know when they're working with geotech firms on large projects.
Proctor results and material changes
It's important for contractors to understand the soil reports. If you need help with that, ask. "It's not offensive in any way to say, 'Hey, I see we have two proctors. Could you help me understand what those materials are?'" Randy says.
It's especially important to ask questions when materials change. The top and bottom layers of soil that you excavate on the jobsite can be pretty different. For example, Phoenix is in a floodplain, so it's common for contractors to dig down five feet and find alluvial material that contains more sand or silt.
When materials change, you might need additional proctors to ensure that the soil density, moisture content, and compaction calculations are still accurate.
Timing and amount of testing
Geotech firms usually collect soil samples before earthwork starts. They may ask you to help them dig down below the surface with a backhoe or other equipment. They may also blend several layers of soilâ€”which will likely happen during construction anyway—to get the most accurate maximum dry density.
Contractors sometimes think all soils onsite are the same. But the bigger, more linear, or deeper the site is, the more likely you are to encounter different soils. Then, you'll need more samples to find the true range of maximum dry densities and optimum moisture content.
"A lot of projects might just have one or two laboratory proctors, but we've had some where we've had as many as 80 or more laboratory proctors,"Mike says. Some projects (like earthen dams) even call for daily one-point proctors to back up the lab tests. Mike adds, "If you've got a five-mile-long structure with a bunch of fill coming from all over the place, you may need that level of testing."
Randy's company has built dams that were miles long and required numerous proctors. "Those proctors all start to blend into a range," he says. That range paints an overall picture of the soils across the jobsite.
Without enough testing, you may not be able to meet the project specs. For instance, you may have a 98% compaction spec and 10-pound density range. "But if the geotech firm was using the wrong proctor, you could beat the ground to death and never get there," Randy says.
Modified vs. standard proctors
Sometimes, geotechnical firms use standard proctors, and other times, they use modified proctors. Modified proctors test the same material as standard proctors, but they're designed for projects that will bear unusually large loads. If the geotech firm uses the wrong proctor, your soil may not get the compaction it needs to bear the structural load.
"We see modified proctors in airport construction, specifically under runways, because runways have some of the highest loading of any structures out there," Mike says.
For modified proctors, technicians use heavier hammers and more blows to compact the soil. A standard proctor might call for 110 pounds per square foot, but a modified proctor could call for 118 or more. Modified proctors also reduce the optimum moisture content. You need drier soil because you're hitting it so much harder.
As a contractor, you need to know when to use these two tests. Then, if the soils report references a standard proctor but the specs call for a modified proctor, you can straighten it out before the project starts and avoid delays.
Geotech companies outside the area
Mike often sees project owners in Phoenix hire geotech companies from Dallas or other areas to do the soil reports, because they've got relationships with firms there.
This can lead to more testing issues. As Mike points out, "In Dallas, they might use modified proctors everywhere, so they come here and do a soils report in a modified proctor. Well, you better make sure everybody on the team is aware of that, because everybody's rolling out there assuming it's a standard proctor just like every other Phoenix project."
Be proactive about these little things during the pre-construction phase so they don't turn into big things. If you wait and find a mistake when the job is supposed to start, you may have to delay the project for days while the laboratory retests the site.
Question #4: Why can I suddenly not get compaction?
Randy asks: When there's more aggregate in the soil, my densities increase. Then they come down at some point. When compaction suddenly comes down, you'll hear terms like, "We broke." What does that mean, and what's happening?
Suddenly struggling with compaction is pretty common with asphalt, but it can happen with soils, too. Here are a few reasons why.
Too much or too little moisture
Every material has a liquid component and a soil and aggregate component. When there's too much liquid, it displaces the soil and rock particles. That makes it harder to achieve compaction because as you push down, the water pushes the soil particles apart.
Just like you can't compact mud, you can't compact dust. There's just not enough moisture to hold the soil together.
You can also over-compact soil. When that happens, the soil starts to break or crack—hence the term "we broke."
Over-compaction is easy to do with asphalt. Asphalt needs a controlled amount of air voids to be flexible. "If you start taking all those air voids out . . . you've really compacted it more than you should," Randy says.
Mike agrees and adds, "We sometimes think, Well that's got to be bad to have air in asphalt. It should be totally solid, right? No, no, no. We want a few air voids in there. It's called flexible pavement for a reason; it needs to move a little. And that amount of air voids is pretty small, 5-8% percent or so."
In Phoenix, some contractors use super pave with more aggregate, bigger aggregate, and less oil. Randy has noticed that sometimes it's harder to get compaction with three-quarter inch or 19-millimeter aggregate mix. That can be because the nominal aggregate size—or controlling aggregate size of the mix—needs a certain thickness of lift in order to compact.
With soils, you might be able to get compaction on a six-inch lift, but with asphalt, you're often trying to compact two or three inches of material. You don't want a bunch of large aggregate particles in there, because they hinder compaction.
For example, a common spec in Arizona is that the lift must be one-and-a-half times the nominal aggregate size. That's a one-and-one-eighth inch lift for a three-quarter inch aggregate mix. However, if you have aggregate that big in a lift that thin, the rocks will bounce off each other and won't compact properly.
The wrong design or aggregate can make achieving compaction tough—or impossible.
Aggregate mix and equipment
Modern super pave has less oil content and more rock, making it harder to compact. Mike adds, "The rock itself has to be crushed in a way that's going to be more stable once you achieve compaction."
Randy has seen crews do a few compacting passes with a vibratory roller, switch to a static roller, then use a pneumatic roller to finally get proper compaction. "You might have to put a lot of different equipment out there to figure out what's the best way to work with this specific mix," Mike agrees.
That's especially true for airport runways. Mike says, "You have to show you can achieve those air void levels and compaction before you can go into full production on a runway asphalt project. I've been involved in some projects where we might have to do 15 control strips or more before they've got the right equipment to deal with that specific mix."
Written by Mike Smith
January 26, 2023