This chapter provides first a short discussion about the limitations and advantages of the methods applied in this PhD thesis and it concludes with a summary of the results.
7.1. DISCUSSION
The work completed with the TRT in Paper A provides a significant amount of experimental data which allows to carry out inverse modelling yielding consistent results with the independent laboratory measurements. However, the inverse modelling of 3D finite element models for TRT data interpretation is not recommended in practical applications and more simple models, such as semi-empirical models that consider the geometry of the energy piles are recommended.
Paper B proposes a design method based on semi-empirical models of multiple energy piles. The experimental data in this aspect is more limited. Further soil temperature data at several radial distances would assist a more thorough assessment of the accuracy of the models. Even though the development of semi-empirical models is simple, the model does not account for groundwater flow and, therefore, its applicability is limited, and it must be used with care when seasonal thermal storage needs to be considered.
Aggregation of errors derived from spatial superposition techniques found in Paper B have also been reported in literature. Nevertheless, it has been demonstrated that for time varying thermal loads (non-constant thermal pulses in the long term), the semi-empirical models yield average fluid temperatures similar to corresponding full 3D FEM and the errors are within acceptable limits for design purposes.
In Paper C, the multiple pile g-functions are used to reproduce operational average fluid temperatures for heat extraction periods in an actual energy pile foundation in Denmark. Regarding the proposed optimisation strategy based on the desirability function approach has strong features: it is flexible and allows new conditions to be easily implemented, such as more restrictive upper and lower temperature limits. For design purposes, the proposed tool seems very practical. More case studies should be tried to for further validation of the model.
To analyse the thermo-mechanical implications, a tailored finite element model has been developed considering a time varying thermal load in an energy pile, where the soil was considered homogeneous and isotropic. However, according to the latest literature, models that account for the different mechanical properties of the soil layers
are more convenient. Besides, the analysis of a monotonic thermal load, i.e., the temperature change calculated from the multiple energy pile model here proposed, should suffice for design. For this purpose, the load transfer method [44,76,77] seems appropriate since it allows the tuning of the soil-pile and pile-structure interaction parameters.
7.2. CONCLUSION
The main objective of this PhD thesis is to create a framework for the analysis and design of GSHP systems based on precast quadratic pile heat exchangers to cover the heating and/or cooling needs of a building, without compromising the structural role of the piles.
To analyse the thermo-mechanical implications of the geothermal use of energy piles, a numerical study has been carried out. The results show that a typical geothermal utilisation of the energy foundation does not generate significant structural implications on the geotechnical capacity of a single energy pile. However, ground thermal loads need to be considered in the design phase to account for potential extreme temperature changes. These findings are in line with the literature.
The temperature disturbance in the pile-soil system, resulted from the heating and cooling of the piles, depends on the thermal properties of the concrete, the surrounding soil and the pile arrangement. Hence, an assessment of the induced temperature changes with respect to the initial undisturbed temperature needs to be carried out in order to estimate the induced thermal stresses and strains. Therefore, there was a need to develop a tool that considers the peculiarities of precast energy piles and that calculates the temperature changes occurring in the energy piles given a thermal load of a building.
First, the suitability of different heat flow models to interpret thermal response test data of pile heat exchangers is investigated. Interpretations based on semi-empirical pile models yield soil thermal conductivity estimates similar to those obtained from the 3D finite element inverse modelling, given minimum testing times of 60 hours. It is highlighted the relevance of using appropriate models to describe the thermal behaviour of precast energy piles in the short and long term.
Semi-empirical models show a promising potential to account for thermal interactions between piles due to its simple implementation. Hence, semi-empirical dimensionless temperature g-functions for multiple piles are developed by utilising 3D finite element model heat transport simulations with temporal and spatial superposition techniques.
Multiple pile g-functions yield fluid temperatures similar to those obtained with full 3D modelling, at minimal computational cost.
As a further step, the multiple pile g-functions are applied for estimating operational average fluid temperatures in an existing energy pile foundation in Denmark. The multiple pile g-functions reproduce fluid temperatures similar to what is observed.
The thermal model is then utilised in an optimisation algorithm that yields the minimum number of energy piles required by simultaneously maximising the pile spacing (constrained by the foundation pattern) and taking into consideration the thermal load of the building. The optimisation tool shows that the number of pile heat exchangers needed for this case study could have been reduced by 32%. The model can also be used to create design charts that would support the early design process.
Additionally, the tool facilitates the implementation of additional conditions for the optimisation, that can be tailored for specific cases.
Energy piles comprise a real option for space conditioning. Their design needs to consider thermal and geotechnical aspects and this PhD has developed a frame to ease this holistic design. The multiple pile g-function based temperature model combined with the proposed optimisation strategy offers a reliable basis for feasibility studies and for the safe dimensioning of energy pile foundations.