Environmental constraints in stochastic hydropower scheduling for long planning horizons

Abstract

The transformation of the European power system to a climate-friendly one by 2050 involves a shift from fossil fuel generation to renewable energy sources. As a part of this transition, flexible assets are needed to balance out the increasing variability in electricity supply and demand to ensure stability in the system. Hydropower can be an enabler for the green transition because of the technology’s unique ability to provide both short-term operational flexibility and long-term energy storage in the reservoirs. On the other hand, hydropower plants may negatively impact surrounding ecosystems in several ways. To mitigate the negative impacts of hydropower, environmental regulations are normally defined in the licences of hydropower plants. Environmental regulations are necessary to protect local ecosystems and to respect the needs of other stakeholders. Nevertheless, such regulations may reduce the operational flexibility of the hydropower plants and are therefore also associated with a cost. Good utilisation of renewable energy resources contributes to lower system costs and security of supply. To achieve efficient use of water for power generation, hydropower producers rely on decision support tools to schedule the short- and long-term operation of hydropower plants and reservoirs. Accurate representation of environmental constraints in hydropower scheduling models is required to make correct assessments of the operational flexibility of hydropower plants and the influence of environmental regulations. Understanding the implications of environmental constraints on the operation of hydropower plants and their capability to provide flexibility to power systems is imperative to effectively plan the operation of hydropower-dominated power systems with high shares of variable renewable power generation. The work conducted in this thesis investigates the implications of environmental constraints on flexible hydropower plants in stochastic scheduling models with long planning horizons. The impacts of different types of environmental constraints have been assessed from the perspective of a profit-maximising power producer operating in a competitive market and from a cost-minimising system perspective considering a wind- and hydropower-dominated region of a power system. A special emphasis was put on the modelling and evaluation of reservoirfilling constraints that are formulated as reservoir-level dependent discharge limitations (soft reservoir-filling constraints). The results are disseminated through five scientific papers, where three are published and two are under review at the present time. The publications constitute the core of this thesis and substantiate the discussions and results presented here. The work in this thesis contributes to the overall understanding of environmental constraints on the operations of hydropower plants. Two stochastic optimization models have been developed, one for the scheduling of a hydropower system from the perspective of a single producer and one for the scheduling of a windand hydropower-dominated region in a power system. The models are based on stochastic dynamic programming (SDP) and include non-convex reservoir-level dependent environmental constraints. The models are used to investigate the implications of environmental constraints on the operation of hydropower plants, the importance of including such constraints in the strategic scheduling of hydropower plants with reservoirs and, finally, the interplay between environmental constraints and reserve capacity requirements. Four different types of environmental constraints are considered in this thesis: a soft reservoir filling constraint (reservoir-level dependent discharge limitation), a reservoir ramping constraint, a minimum release constraint and a ramping constraint on discharge. The findings imply that environmental constraints may have considerable impacts on seasonal reservoir management and that, under certain conditions, there is an economic benefit in planning for soft reservoir filling constraints (reservoirdependent discharge limitations) in advance. Furthermore, the results show that reservoir-level dependent (i.e., state-dependent) constraints may induce a nonconcave expected future profit function and significantly change the expected marginal value of storing water in some periods. The work also investigates and discusses the impacts of three different types of discharge constraints (i.e., reservoir-level dependent discharge limitations, minimum release requirements and ramping restrictions on discharge) on the available flexibility in a hydrodominated region of a power system. The results show that the impacts on the capability to meet the demand for electricity and reserve capacities requirements depend on the characteristics of the environmental constraints, such as if the constraint mainly reduces the available power capacity or the amount of regulated energy production, and if the constraint includes state- and time-dependencies.