| dc.description.abstract | The primary focus of this thesis is to develop models for oil and gas separation processes. These models form the basis for estimation of unmeasured variables and the control of important process variables. The secondary focus of the thesis is to use the developed models (gravity separator, hydrocyclone and compact flotation unit) for optimization, control or estimation. Key results include in-depth analysis of the different models and optimal operation of a separation system that combines these models. The tertiary focus of this thesis is on state and parameter estimation, which includes results on estimation of unmeasured variables by the use of simplified models and estimation algorithms, such as Kalman filter or Moving horizon estimator. To cater to these focuses, the thesis is divided into four parts.
Part I demonstrates the need for models via a case study on determining optimal operating points for an oil-water separation system. The maximization of oil content in the oily product is chosen as the objective.
Part II showed the use of estimation methods in combination with estimation-oriented models for estimation of unmeasured variables. In addition, a chapter is included on a method called pathfollowing that alleviates some of the computational challenges encountered by advanced estimators, such as moving horizon estimator.
Part III: This part presents three dynamic models, one each for - inline deoiling hydrocyclone (HC), compact flotation unit (CFU) and gravity separator. These models are able to predict important variables, such as oil in water and water in oil, which are difficult to measure, especially subsea. The optimal operation of the CFU is studied with an objective to minimize the use of flotation gas. In addition, a dynamic subsea separation system is constructed using the three models. The optimal operation of this system is studied under changing disturbances in order to propose a simple control structure.
Part IV: This part concludes the thesis with some final remarks and the way forward. It covers two additional topics - guidelines for model development and use of process models in industry. | nb_NO |
