Swelling Retaining Walls

The challenge

Queen and David Bowie sang about the pressure bringing the building down, but we’re guessing they weren’t big on the engineering detail behind why. Maybe if Freddie and co had been tasked with keeping said building erect, they might have felt the need to dig a little deeper? That’s what happened last year when we were tasked with saving a collapsing retaining wall in West Melbourne. Before we could propose a solution, we needed to determine exactly why the wall was bursting to begin with.

In this article

  • Wrestling with a retaining wall under pressure
  • Exploring basaltic clay and how to design for its characteristics
  • Using first principles to solve a new problem

Last year, in West Melbourne, an underpass retaining wall began to show signs of strain, separating and cracking at the corners. The basaltic clay behind it was absorbing moisture and swelling. We were asked to help brace it but we had a problem: we didn’t know how much pressure was coming from the swelling clay.

Basaltic clay is found all over Melbourne. In fact, we have the largest basalt plain in the world which extends from near the middle of Melbourne to Warrnambool in the State's West. Derived from igneous rock, the dark clay soil is dense and heavy, and swells when it gets wet, creating pressure against whatever is holding it. This problem is commonly understood and the conventional wisdom is that you can’t design around it because the force of the swell is too strong. Your best bet is to try and keep the clay dry.

At the West Melbourne underpass, we discovered that a drainage pipe behind the wall had cracked and was slowly leaking water. The clay was absorbing the water and expanding outwards, putting pressure on the wall as a result. If the wall were straight it could probably bend to absorb the pressure, but where it turned a ninety-degree corner, the two sections of the wall were beginning to separate.

Ed Freeman, a geotechnical engineer, figured there were two options. The first was to plug the pipe and wait a few years for the swelling to settle down. The second option was to create enough counter-pressure to stabilise the wall.

“We looked at putting tensioned anchors in front of the existing retaining wall. To achieve this, we’d have to build a new wall in front of the old one and drill the anchors through both walls at angle. These would then be connected to the rock below, putting tension into those anchors, before we locked them off” Ed explains. “If we could design the anchors to take the right load, it would hold the wall behind in place.”

But what was the right load to design for? How big, and expensive, would those anchors need to be?

Despite how common basaltic clay is in Melbourne, there isn’t much data on its swelling pressure. Before Ed and his team could propose a remediation plan, they had to figure out how the clay would behave when exposed to moisture They took a sample of the soil, and sent it to the laboratory to do a constant volume one-dimensional consolidometer test, where the sample is placed in a confined container, plunged into a bath, and the resulting swell is measured.

We have a reasonable idea about swelling pressure for most clays, but in this case, we needed to take a sample for testing. The test we used was a consolidometer test, which plunges the sample into a bath of water, and measuring how much it swells.

While they were waiting, they scanned the academic literature to find out what was known about the swelling behaviours of different soil types. In their research, they came across an equation developed by two engineers in Egypt to measure swell pressure in a different type of clay. The equation used plasticity index (a measure of how soil behaves when you add water), the initial moisture content and the clay content to calculate the pressure — all commonly known metrics that they had from their standard geotech investigation of a site.

“These are parameters we know and can easily measure with basaltic clay, so we used this equation, designed for a completely different soil, and plugged in the metrics from our clay samples.” Ed says. “We wanted to see if the equation gave us the same or similar results to what we got in the lab and it turns out they correlated pretty well.”

The team compared the swell pressure they observed in lab with the swell pressure predicted by the equation and found a close enough correlation to make the equation useful for predicting the swelling pressure of basaltic clay in future.

Ed and the team also found that the swell pressure decreases dramatically as the clay becomes saturated. This may have some useful applications in design. When building a structure on a foundation of basaltic clay you can allow for some expansion of the clay — by using a compressible filler material perhaps — and the structure may be able to cope with the remaining, diminishing force.

Applying basic principles like these to make decisions about real problems is where engineering really comes into its own, and where Ed and his team most enjoy their work. The crucial thing is to have the maths at your disposal, to provide a solid rationale for your work.

“You can do stuff with maths” Ed says. “You can write a report with maths; you can’t write a report with the accepted wisdom. When we’re brought in to do a remediation design, or design for any kind of basaltic clay scenario, we can now use these numbers to show how it might work practically to our client. We draw on this testing and while the proposed solution is more expensive, we can give our client confidence that it’s going to work and reduce the risk of the wall blowing out to a much lower level. Or, if it so happens that the maths turns out the other way, we can give them a definitive answer that the solution won’t work. This is also useful information to inform future design.”

We’ve since shared our findings at industry conferences to help other geotechnical engineers design for the swelling basaltic clay problem. Ed was also shortlisted for the Jack Morgan Award for his work in 2017. Although it seems to be difficult to suppress the swelling of this clay, we hope our contribution will mean fewer retaining wall failures and expensive remediation works for basement retention, reinforced soil walls and retaining wall design.


  • The swelling pressure of basaltic clay can be roughly estimated using the Nayak and Christensen (1971) equation, based on correlation between the equation and laboratory tests.
  • The swelling pressure of basaltic clay was found to decrease significantly at relatively low levels of volumetric expansion

This story is written by Simone Ubaldi, as part of the Research Review series. The series is produced by the Arup Australasia Research team; Alex Sinickas, Bree Trevena and Jeff McAllister with contributions from Sheda and Noel Smyth.

Lead Arup Researcher

Ed Freeman
Ed is a geotechnical engineer in our Melbourne office.

Ask Ed about:

  • Basement retention, reinforced soil walls and retaining wall design.
  • Designing for basaltic clay, including theory and testing of swelling behaviour of different clay types.
  • Getting a copy of his paper from the AGS Victoria Symposium.


Research TEAM

Associate, Transport & Resources, Brisbane office
Sergei is a geotechnical engineer in our Los Angeles office.
Lucie leads the geotechnical team in our Melbourne office.

Have a problem or a project?

We work with industry partners, governments, universities, startups and community organisations. We do this through research partnerships, and as consultants and facilitators for foresight, research, storytelling and technical writing workshops.

research with us