Extraction of oil and gas can cause reduction in pore pressure, occasionally resulting in subsequent compaction that forms a surface subsidence bowl, especially in shallow reservoirs. In the last 10 years, there has been over 10 feet of subsidence in parts of the Lost Hills oil field in California (Bruno et al.,1992). The surface subsidence at Lost Hills not only causes damage to surface facilities and wells, but also reactivates faults and reduces rock permeability. Subsidence makes reservoir optimization difficult. Hence, it is important to assess or predict the surface subsidence and the reasons for subsidence early in the life of an oil field to make an optimization plan. We use jointly the capacitance resistance model (CRM) (Alberoni et al., 2002 and Yousef, et al., 2006) that relies only on injection and production data, and the InSAR satellite imagery of surface subsidence. From CRM simulations, we estimate the connectivity between injectors and producers as well as general water flow directions from individual injectors. We then superimpose well connectivity and InSAR imagery to diagnose the reasons for the subsidence. Using new surface subsidence models, which are based on the continuity equation of CRM and rock mechanics, we are able to predict the average surface subsidence at Lost Hills from the injection and production rates. Our work shows that there was significant volumetric rock damage at Lost Hills and the well connectivity changed dramatically with time because of reservoir compaction and the rock damage. We conclude that for a soft, fragile and nearly- impermeable rock such as the diatomite, high injection rate weakens the rock and creates dynamic water flow tubes or ‘channels’ without providing good pressure support to the reservoir. These high permeability ‘channels’ re-circulate most of the injected water between the injectors and producers. Our CRM/InSAR approach is new and gives insights into the time-dependent and spatially variable fluid flow fields in a relatively shallow waterflood. Consequently, we may be able to suggest optimum water injection strategies to enhance oil production, while minimizing rock damage and surface subsidence. In addition, the proposed surface subsidence models are convenient and reliable to predict the average surface subsidence. / text

To better understand the issues about the surface movements in the coal mining region Lugau-Oelsnitz, Germany, small-scale numerical models are firstly utilized for verifications via analytical solutions, to explore the simulation schemes, and for parameter sensitivity analysis. 1D rock column numerical models shows that simulated surface movements are consistent with analytical solutions. The investigations via 2.5D profile numerical models also show that uplift is linear related to water level rise under confined mine water conditions, while a quadratic function is valid for unconfined mine water. Geodetic survey in the Lugau-Oelsnitz district shows that at the end of the active mining period (1844 to 1971), general subsidence is about 5 - 10 m, with a maximum of 17 m in the southern mining area. General uplift velocity after abandoned mine flooding between 1972 and 2014 is about 0.5 - 2.0 mm/year. Based on numerical simulation results, predicted general uplift velocity vary between 0.5 - 3.0 mm/year, while maximum uplift position is moving toward south.:1 Introduction 2 State of the art 2.1 Overview 2.1.1 Coal mining induced settlements 2.1.2 Flooding induced uplift 2.2 Approaches to predict subsidence 2.2.1 Empirical approaches 2.2.2 Influence function methods 2.2.3 Physical models 2.2.4 Numerical simulation methods 2.3 Approaches to predict uplift 2.3.1 Empirical approaches 2.3.2 Numerical simulation methods 2.4 Comparison and conclusions 2.4.1 Comparison of research methods 2.4.2 Conclusions 3 Numerical simulation approaches 3.1 Continuum mechanical simulations with FLAC3D 3.1.1 Mining induced subsidence 3.1.2 Flooding induced uplift 3.2 Discontinuum mechanical simulations with 3DEC 3.2.1 Self-weight induced settlement in jointed rock column model 3.2.2 Uplift for jointed and fully saturated rock column 3.3 Parameter sensitivity study 3.3.1 Parameter effect on subsidence 3.3.2 Parameter effect on uplift 3.4 Interface and volume element representation of faults 3.4.1 Simulation schemes 3.4.2 Parameter sensitivity analysis of fault 3.4.3 Discussion 3.5 Conclusions 4 Case study: Coal mining region Lugau-Oelsnitz 4.1 Background information 4.1.1 Mining background 4.1.2 Geological and hydrogeological situation 4.2 In-situ monitoring data 4.2.1 Groundwater level data 4.2.2 Surface movement data 4.2.3 Discussion of data analysis 4.3 Continuum based numerical modelling 4.3.1 Introduction 4.3.2 Model set-up 4.3.3 Calculation results 4.3.4 Surface movement predictions 4.4 Discontinuum based numerical modelling 4.4.1 Model set-up 4.4.2 Calibration results 4.4.3 Surface movement prediction 4.5 Conclusions 5 Conclusions and prospects 5.1 Conclusions 5.2 Main contributions of thesis 5.3 Inadequacies and prospects