Soil-structure interaction of horizontally loaded piled-raft foundations

Abstract

Wind turbine power generation has gained significant popularity over the past few decades as an option for cleaner energy production amidst growing climate change concerns. However, the design of a turbine’s foundation, capable of supporting tall structures subject to large horizontal forces and overturning moments, remains challenging. The current focus with many new wind farm constructions is on taller wind towers, allowing for the same generation capacity from fewer wind turbines. Complex dynamic wind loading, which is amplified for taller wind turbines, and intricate soil-structure interaction between the foundation and the supporting soil require consideration to obtain foundation solutions that are both economical and sustainable. Although raft foundations are preferred for supporting onshore wind turbines due to cost and ease of construction, many researchers have recently favoured the use of piled-raft foundations. Not only do these foundations adequately support wind turbines when these towers are constructed on less favourable soils, susceptible to large settlements or low bearing capacities, but they also provide a substantial and more economical solution for resisting the significant overturning moment acting on the foundation. Yet, the response of these support structures is not well understood, especially considering that vertical loading is no longer the driving force in determining the size of these foundations, with loads dominated by the dominant horizontal load and overturning moment from the wind. Thus, given the increase in the dependency on these renewable energy structures and their size, the need to investigate this is important, given the strict design criteria and allowances. In this thesis, a full-scale onshore wind turbine piled-raft foundation supporting a 117 m high wind turbine located on a newly constructed wind farm near Wesley in South Africa was instrumented and monitored for an extended period of time. The data presented includes those obtained during construction, turbine installation, and during turbine operation. In addition, finite element (FE) modelling was also conducted on piled-raft foundations under these unusual load combinations, considering soil-structure interaction and foundation rigidity. The full-scale testing showed that the foundation response was dominated by the dynamic horizontal load and overturning moment, compared to the vertical self-weight of the turbine, as expected, with the loads shared by both the raft and the piles. As the number of wind cycles increased, the results from the instrumented foundation socketed into bedrock indicated that a greater portion of the applied loads were distributed amongst the piles. Given the significant rigidity of the pile connecting raft, the response of the piles was dominated by the push-pull effect. Seasonal temperature changes also affect foundation response, which is usually neglected due to the foundations being buried. Additionally, from the FE modelling, apart from the soil-structure interaction concepts that were considered, the relative stiffness between the pile and the raft proved valuable towards analysing the rotational stiffness of the foundation for wind turbine application, also allowing for the potential axial forces in the trailing piles to be limited, given the large horizontal loads and the significant overturning moments. Based on the responses observed from the full-scale testing and the results from the relevant FE models, it is clear that the upper limit has been reached regarding our current approaches to designing these foundations. In addition to the regular checks for restricting foundation settlement, differential settlement, horizontal displacement and meeting the minimum rotational stiffness requirements of the foundation, larger wind turbine models have presented additional critical design checks that cannot be ignored. These include the potential cracking of the raft under loading and the development of significant tensile forces in the trailing piles, both of which must be limited. Especially for larger turbine models, considering a balanced soil-structure interaction approach was shown to be beneficial. However, as mentioned previously, consideration should still be given to the constructability of these foundations, as larger foundations might result in more significant thermal gradients within the concrete section.