GEO

July 8, 2025

Geotechnical

Geo's geotechnical gold turns into fast results in OPTUM G2

At Geo, they have access to an underground treasure trove – not filled with gold and gemstones like in fairy tales, but with valuable data from the Danish subsurface. Over the decades, more than 300,000 borehole profiles have been collected and analysed, forming the basis for deep expertise. This knowledge has been used in everything from the Metro Cityring in Copenhagen to over 100 offshore wind farms.

Today, Geo combines this massive data foundation with efficient FE calculations in Optum. We spoke with Senior Project Manager Torben Thorsen about how they use the software in practice – and why it has earned a permanent place in their toolbox.

Photo by Jeppe Sørensen

From Borehole Data to Design Power

What do you mainly use Optum for?

“We use it extensively for sheet pile walls, bracing of excavations, and for stability and bearing capacity calculations. That can be in connection with roadworks, railways, embankments, or construction near slopes. For large infrastructure projects, e.g., for Banedanmark, where you may need to excavate near a bridge, it's necessary to verify the bridge structure's stability – and that's where Optum is effective. In general, I also see more and more of my colleagues in the industry using Optum when I review their projects. Five years ago, Plaxis was dominant. Now, Optum is often the preferred program.”

Which features are you particularly happy with?

“The way it handles meshing is incredibly efficient. Optum can automatically refine the mesh where something is happening in the model – for example, around a failure surface. So instead of doing it manually, it optimizes the mesh itself – saving time and making the results more precise.”

When is adaptive meshing clearly an advantage?

“If we need to verify the bearing capacity of an embankment placed on soft soil, it's very useful that the mesh automatically refines precisely where the most critical failure mechanism develops. This provides a more reliable calculation without extra work.”

So, what's the best thing about Optum?

“Speed and stability. It's quick to build models, and calculation errors are rare. I also rarely experience the program freezing, which I sometimes do with other software.”

What do you think of the user experience?

“Optum is intuitive and fast and – without sounding negative – just simpler. That’s a big advantage if you don’t need a ton of advanced features. When the tools you need daily are built into Optum, there’s no reason to use a heavier, more time-consuming program.”

Does it also save materials?

“Yes, perhaps indirectly. It's mostly about using calculation time wisely. The more efficient and precise the model is, the lower the risk of overdesign. And the more we know about the soil, the more accurately we can design. That often means we can save materials. For example, if we can demonstrate that a pile length can be reduced by one meter, we're quickly talking about savings in the million-kroner range – and at the same time, it's a better solution for the climate.”

How do you experience support from Optum?

“It’s really good. They respond quickly, and you get direct contact with developers who know the software in depth. That’s a big plus compared to large international software vendors, where you can feel a bit far removed from support.”

The way it handles meshing is incredibly efficient. Optum can automatically refine the mesh where something is happening in the model – for example, around a failure surface. So instead of doing it manually, it optimizes the mesh itself – saving time and making the results more precise.

Torben Thorsen

Senior Project Manager at GEO

From soil sample to simulation

How do you ensure your soil models reflect reality?

“We take soil samples and conduct lab tests and use the results as input to our calculation models. But it’s not enough just to plug the numbers into the program – we also need to be able to reproduce the lab results in our models. If the model can't simulate what we measured in the lab, we can’t trust it to predict reality reasonably.”

So, you calibrate the model against lab results?

“Yes, exactly. If we have a sample where we’ve measured a certain stiffness and strength in clay or sand, then the same behaviour should roughly be reproducible in a model of the test.”

Is it possible to represent the geology precisely in the model?

“No, we only know the subsurface based on the points where we've drilled. We try to interpret and interpolate between them, but it will always be a kind of sampling. Geology and deposition history are extremely complex – and very local.”

So, each model is essentially an approximation?

“Yes, and that’s why it’s important we model conservatively. For ULS calculations, for example, we use partial factors – reducing the strength of materials and increasing loads – all to ensure the structure holds, even in the worst case.”

Do you sometimes face resistance from clients when you recommend more or costlier investigations?

“Sometimes, yes. Some only see the DKK 150,000 cost of extra investigations – not the DKK 1.5 million you might save in execution. But our task is to ensure the best and most efficient solution, and that sometimes requires more upfront work.”

Reality is simply 3D when it comes to settlement calculations. In 2D, you only get a cross-section, but in 3D, you can see the full spatial distribution of settlements and loads. You have far more data points and a better basis for assessment.

Torben Thorsen

Senior Project Manager at GEO

3D, standards, and untapped bearing capacity

When do you use 3D?

“Primarily for settlement calculations – because they usually have a clearly spatial character – and often in connection with high-rise buildings, where we’ve seen significant development over the past ten years. When building tall structures on soft soil, it's important to understand how the soil settles under load, and that often requires looking at it in three dimensions – whereas 2D is often sufficient for failure and bearing capacity (ULS) calculations.”

Why is 3D necessary for settlement calculations?

“Because reality is simply 3D. In 2D, you only get a cross-section, but in 3D, you can see the full spatial distribution of settlements and loads. You have far more data points and a better basis for assessment.”

Can 3D calculations lead to better – and cheaper – solutions than 2D?

“Yes, if you can utilize the soil’s bearing capacity more precisely, you might reduce material use. But you also need to document it to the client. That requires a model built on good and realistic data from soil investigations – and that also costs money.

But if we look at a train, for example, the load is carried by two wheelsets at each end of the carriage – just like on a car. Normally, the load is modelled as a single line load in a 2D program. But if you wanted to look more closely at where the load actually hits the rails, it would be beneficial to calculate it in 3D. It’s precisely there – under the wheelsets – that the full load is applied. And by calculating it in 3D, you can exploit the spatial bearing capacity effect, which we usually ignore in 2D.

That means there may be a reserve in bearing capacity you don’t capture in a 2D calculation. The reason we typically use 2D – and thus a more conservative approach – is partly because the rail standards require it. But perhaps those rules will change over time.”