Summary: | Longyearbyen is an isolated, Arctic settlement, where the energy situation is facing changes in near future. Today s energy system in Longyearbyen is a stand-alone system, consisting of a coal power plant as primary power producer and a diesel reserve system as a backup. In addition, the energy system offers fuel supply to the marine traffic in Longyearbyen, mainly Marine Gas Oil (MGO). The Ministry of Petroleum and Energy is currently assessing the possibilities for new energy carriers in Longyearbyen, and the general focus is that the new system should be more environmentally friendly than today s system. LMG Marin is a company that is currently designing a solution applicable to the changes Longyearbyen is facing, and has given the inspiration for the problem description in this thesis. The energy carriers in the new energy system in Longyearbyen is Liquefied Natural Gas (LNG) and solar energy. The aim of this thesis has been to create an optimisation model intended to minimize the total lifetime costs over a 30-year period for the new energy system in Longyearbyen. The optimisation model in this thesis has been formulated by applying Mixed Integer Linear Programming, and implemented in the linear solver FICO® Xpress. A Rolling Horizon Heuristic algorithm was created to reduce the computational time of the model to an applicable level. The optimisation model was successfully validated in cooperation with the industry partners, and concluded to reflect the energy system realistically. An economic and environmental study were conducted to see the cost and emission effects of different system designs. The results were based on future energy demand predictions in Longyearbyen. The predictions assumed an electricity demand reduction of 25% due to the removal of the coal power plant, and a reduction of the heat demand with 40% due to energy efficiency actions. The LNG ship bunkering demand corresponded to 80% of today s MGO demand, and the solar irradiance level was assumed to remain at today s level. Four ...
|