Please use this identifier to cite or link to this item: http://hdl.handle.net/1946/3032
The objective of the study is to make a complete finite element model of a geothermal well. Three casings, cement around the casings, a liner and the formation surrounding the well were modeled with three different types of finite element models. A two dimensional thermal model was used to analyze the temperature distribution of the well and a two dimensional structural model was used to estimate the rise of the production casing along with the well’s stress distribution. Finally, a three dimensional buckling model was built to examine the buckling behavior of the production casing where the results from the two dimensional models were used to define loads and constraints.
The maximum upward displacement of the production casing for a 2300 m deep well is 0.05 m according to the two dimensional structural model, but a rise of around 0.13 m has been detected in casings at power plants in Iceland [Gretarsdottir, 2007]. Different contact behavior was defined between the cement and the steel in the production pipe and it greatly affected the results. The stresses were also examined and compared to the yield strengths of the materials.
The Von Mises stresses in the steel did not reach the steel’s yield stress but the y-component of the compressive stress in the cement reached the ultimate strength of the cement and well over. Different values of the Young’s modulus for the cement were also examined in order to estimate its effect on the expansion. The Von Mises stresses were considerably higher when the Young’s modulus was increased, but it did not affect the displacement results of the production pipe. When the well was modeled with a three dimensional buckling model, buckling did not occur for the loads defined. However, a load magnitude of 85.3% of the loads defined in the model, caused the well to buckle when a 20 MPa sideward pressure was modeled. The sideward pressure area was increased and the buckling load stayed the same but the displacements where the well buckled were larger for the increased area.
The models built in this project can be useful when designing geothermal wells, to estimate the effects of different properties of the well defined. According to the case study performed, buckling is not unlikely to occur in real geothermal wells and the buckling model could be used to estimate whether changing some parameters would result in decreased risk of failure due to buckling.