Advanced Modelling Makes This Possible Without Compromising Safety
Unlike wind turbine towers and rotors, which are fabricated under controlled conditions, in general foundations must be tailor-made. This is because soil conditions (hard or soft) and available space dictate the solutions for the foundation. However, the perception is that optimisation of foundation designs leads to higher risk. Maybe due to this the average foundation designer takes a simple, conservative and conventional design approach. But a client should look at things in a different way. Saving money is one thing, saving materials and reducing CO2 emissions is another. A client should aim to have an optimised design. The skills and experience are available and have shown that substantial design optimisations are possible, without increasing risks.
By Axel Jacobs, Civil Consultant Wind Energy, ABT, The Netherlands
Figure 1 shows an example of a project were something went wrong with regards to the foundation. You can look at it in two ways:
- The design is to the maximum level of risk, so optimisations are not possible. You should strengthen the foundation in order to avoid collapse.
- The concrete base does not seem to be cracked and is sufficiently strong. Apparently an issue in the soil had caused the collapse. So, when this issue is solved, optimisations of the foundation are possible.
Savings
As well as providing cost savings, optimisation can substantially reduce material amounts and the amount of CO2 emissions. When you know that 1 tonne of cement causes 1 tonne of CO2 emissions, you can imagine that making a foundation is not that environmentally friendly at all. So every tonne of CO2 saved has a big impact on the CO2 footprint of the project as a whole.
The statistics listed below are based on a large number of projects where advanced modelling was used. Conventional design is compared with optimised design and the effects of optimisation are as follows:
- 2% total cost benefit on project
- 20% total cost benefit on foundations
- 30% total saved concrete
- 40% total saved CO2 emissions
- 50% total saved reinforcement steel.
Risks
In order to address the main reservation about optimisation – which according to a lot of people is higher risks – we need to look at ‘risk’ more closely.
Risk is doing something that has never been done before. To assess risk you need ‘information’, ‘skills’ and ‘experience’. When one of these tools is missing or incomplete the risks can become too high. When, for instance, due to a lack of experience, certain risks are not identified as risks, these risks can become a real problem. These forgotten risks are the real risks.
You only really know risks when you know where you cannot go further. And this limit can only be found through experience.
A relationship between level of optimisation and risk levels is not as obvious as it might seem. Also, a conventional design can turn out to be of high risk when certain issues are not addressed properly. In the following section this is explained using the general safety concept for design.
Safety Concept
The safety concept generally used for foundation design is that the strength (or resistance R) should be more than the loads (S) in order to have a safe structure (Figure 2). To make it a bit more difficult, both loads and strength have a certain variation. Most design codes introduce so-called partial safety factors (load factors and material factors) in order to achieve a sufficiently low failure probability. As an example, the characteristic load (Sk) is multiplied by a load factor to get the design load (S).
A good example of the variation in strength is soil conditions. We can encounter sites where we have competent rock but also sites where fine sands are to be found. The latter are more sensitive to the influence of groundwater, which can lead to quicksand-like behaviour.
We are now talking about the first tool from our risk toolbox. We are gathering information about the location, materials, loads, groundwater, soil etc. This is important input for our design works and the more precise it is, the better we can assess the risks.
Model
Resistance should be more than load. In order to prove that, we need a model. Let us say that models are the skills and experience tools from our risk toolbox. There are two important laws of modelling:
- A model is not equal to reality. It is always an approximation. That means that the results should also not be treated as reality.
- ‘Garbage in’ equals ‘garbage out’. The designer should be aware that the way a structure is modelled and assumptions made influence the outcome.
The graph in Figure 3 shows the relationship between time devoted to analysis and accuracy. We can say that if we can achieve 100% accuracy, we have exactly reproduced reality. Levels I to IV represent the different ways of modelling, ranging from low to high accuracy.
When we translate this into the safety concept we see that in general the more inaccurate the model is, the more over-dimensioned the structure becomes. The problem is that this effect (see Figure 4) is not visible to the designer. He or she only sees that R is more than S, as shown in Figure 2. The designer could then conclude that the design is highly optimised, forgetting the level of accuracy of the model.
Level I Modelling
When looking at a level I analysis we are talking about the most inaccurate way of modelling. Most conventional designs do not go further than level I (or maybe level II) modelling. This is an analytical design method which uses all kinds of simplifications in order to make it comprehensible. Formulae are used to describe overall behaviour. In order to describe behaviour in a single formula, assumptions and simplifications are used.
It seems that a level I design always leads to over-dimensioned structures and therefore is ‘too safe’. But there we could be wrong. We said before that the real risks are the ones that were not identified as risks. When using an analytical design method it is easy to overlook a possible failure mechanism, or the model that has been used is not able to verify a certain failure mechanism. This is exactly the reason why codes and guidelines are mostly quite extensive. They need to protect the designer, who uses level I modelling against unexpected or forgotten factors.
Codes and guidelines actually translate skills and experience from experts into rules that make it possible for inexperienced designers to design a foundation. These inexperienced designers cannot be expected to know where the ‘edge’ lies. So, in order to compensate for the lack of experience, these rules need to be conservative. On the other hand the codes do also allow for the use of advanced level IV modelling.
Level IV Modelling
In contrast to level I modelling, a level IV model uses the real geometry of the foundation including all internal components (like reinforcement and tower anchorage). Also the real, non-linear material behaviour is modelled. This model actually shows the behaviour of the structure under increased loading and the interaction of all the components. For concrete this means that realistic crack patterns develop (Figure 5) and you can actually see the stress distribution in the reinforcement steel. The real failure mechanism becomes apparent. So instead of guessing what failure mechanism will occur, all failure mechanisms are implicitly verified and the real weak spot can be found. Simplifications are not needed.
Conclusion
A fundamental change of approach is needed to achieve a substantial improvement in the ‘green’ aspect of wind projects as well as cost savings. In order to do that the perception that optimisation also leads to higher risks must change. The tools are already available to achieve an optimisation in foundation design, and these have shown that it is possible to create cheaper, greener foundations without compromising the safety margins required by the codes.
At the same time it has to be recognised that ,while the codes and guidelines are generally conservative, they do allow for deviations from the rules when circumstances permit it, and when expert knowledge can demonstrate an alternative approach. In too many situations, however, the rules are followed strictly and opportunities for optimisation are missed.
Biography of the Author
Axel Jacobs has been active as a civil engineer and foundation designer in wind energy ever since 1995. Because of this long experience he is able to combine his knowledge of structural and geotechnical design with a good understanding of the requirements of wind energy. He has been involved in a broad variety of wind energy projects worldwide, giving him insights into different soil conditions and most wind turbine types and brands.
