Hans Ivar Skjelbred

Planning of Distributed Energy Supply to Suburb

In this paper, a novel model to optimize the design of local energy networks with respect to operation and investment cost and environmental consequences is demonstrated. The chosen case study is based on a real case from a Norwegian municipality. Several alternatives to supply energy to a new suburb with detached houses are considered: Electrical boilers, ground source heat pumps, gas boilers and a simplified demonstration alternative of hydrogen fuel cell based on renewable energy. The alternatives are ranked with respect to operational and investment cost. Emissions from each alternative are also accounted for. Due to an integrated optimization of diurnal operation and investments, each design alternative can be inspected in the full-graphical user interface to verify the utilization of individual components during peak load and low load periods.

Operational hydropower scheduling with post-spot distribution of reserve obligations

As a reliable and flexible electric power source, hydropower can quickly adjust its generation level and provide reserve power to balance the power fluctuations in the system. In this paper, we assume that the day-ahead spot market is cleared and the obligation for each reserve type is already contracted in a preceding market. Then the question is how to make the optimal decision to determine, for each unit and for each time step, the dispatched volume for the day-ahead market and the reserved capacity for various types of reserve. We formulate the constraints added to the optimization problem for distributing reserve obligations. The proposed mathematical formulation is based on an operational hydropower scheduling model used by most large hydro-power producers in Scandinavia.

Operational use of marginal cost curves for hydropower plants as decision support in real-time balancing markets

When participating in the real-time balancing markets, hydropower producers keep making decisions about how much power they are willing to sell or purchase at what prices from the markets. Due to the specific market characteristics, the decisions should be made quickly, placing high demand on the computation time of rescheduling. In this paper, we present a practical method to determine an hourly bidding curve for a plant. It is applicable to both balancing and intraday markets. We heuristically calculate the marginal cost for all the operating points, covering the entire working area for the plant and including all the physical limitations and reserve obligations in other markets. Heuristics are proposed to reduce the computation time without neglecting significant practical features. Numerical results based on a real-world case study demonstrate the advantage of the proposed method in terms of computation time and solution quality.

Finding seasonal strategies for hydro reservoir scheduling under uncertainty

In this paper a new approach to generate strategies for hydro reservoir scheduling is presented. The idea is that the marginal value of water in one reservoir can be calculated by dynamic programming under some simple assumptions on the content in other reservoirs. A benefit of this approach is that one avoids the “curse of dimensionality” by having to calculate water values for all combinations of reservoir content in the different reservoirs. Water values for all reservoirs can be found simultaneously for each reservoir content level. The method is compared to state-of-the art methods in hydropower scheduling and promising results are observed.

Combining start-up costs with non-linear head optimization of hydro storage

In this paper a new method to find optimal strategies for scheduling of pumped hydro storage reservoirs is presented. Allowing start-up costs and non-linear head variations to be handled within a formal optimization algorithm shows increased value for hydro storage operators. Two cases are presented and the proposed solutions are compared to state of the art tools within hydro scheduling.