The Modeling And Solution For Operation Adequacy Of Power Systems Considering Primary Energy Congestion

The traditional research on power system adequacy is based on the assumption that the supply of primary energy is adequate. It considers the adequacy degree of generation and transmission capacity relative to load, which could be categorized as generation adequacy and transmission adquacy. Generation adequacy could be categorized as installation capacity adequacy and operation adequacy. The former refers to generation planning; and the latter refers to the allocation and dispatch of operation capacity.However, the power system is just a part of the whole energy system, and primary energy supply is increasingly influencing the power system adequacy. For instance, the large-scale integration of intermittent renewable energy resource (RES), such as wind and solar, introduces more uncertainties to operation adequacy; and the short supply of fossil fuel, such as coal and natural gas, could decrease the actual installation capacity adequacy. Therefore, it is extremely essential to study the power system adequacy considering the primary energy congestion.The decision-making of operation adequacy control with the integration of large-scale intermittent RES is an optimization problem. The traditional solution is to consider spinning reserve capacity (RC) constraint within the unit commitment problem. The requirement of adequacy and economic is fulfilled by minimizing the objective function, which includes start-up cost, generation cost, spinning reserve cost and risk of potential power outage. In this dissertation, the main contributions based on the existing optimization model are as follows:(1) In the existing optimization model, the net load demand (the load minus the total power output of RES) forecast eror is assumed to follow Gaussian distribution, and the "generation cost" and "spinning reserve cost" in its objective function are set to be uncorrelated with the differences among scenarios. However, Gaussian distribution performs poorly for describing the forecast error in actual operation. Therefore, we, model the forecast uncertainty by discrete scenarios of net demand load, instead of any distribution. Accordingly, the "generation cost" is modified to be the weighted sum of corresponding cost of each scenario, which includes the "generation cost of spinning reserve"; and the "spinning reserve cost" is modified to be "capacity cost of spinning reserve", which is uncorrelated with the differences among scenarios.(2) The spinning RC of generation side is the only control objects in the existing optimization model. Its economic efficiency could be improved espacilly for the system with significant RES penetration. Therefore, the complementary features of multi grade of RC from generation side and demand side are analyzed. And then, one kind of spinning RC of demand side (interruptible load, IL), and other non-spinnning RC from generation side and demand side are introduced into the optimization model successively. Thereinto, the "passive power outage risk" based on historic offline statistic data is replaced by the sume of "cost of active IL shedding" and "remaining risk of passive power outage" in the objective function when IL is introduced. The "cost of active IL shedding" is further replaced by "cost of post-event control" when various grade of RC are introduced. The "start-up cost", "spinning reserve capacity cost", "generation cost" is enabled to describe the situation for multi grade of RC. The constraints for IL and other non-spinning RC are added into the optimization model as well.(3) It is no longer to control the RC through administrative instruction in the market environment. The operation adequacy control decision-making is divided into two parts:to obtain the control permission in the market named "decision of contract" (or contract decision) and the actual dispatching/control of RC named "decision of implementation" (or implementation decision). The optimization models for both of them are modeled respectively. And the differneces and relations between them, such as objective function, decision variables, the space of optimization, constraints, etc, are discussed.(4) Based on the "decomposition method" using in the optimization of large systems and its application in the optimization of security and stability control decision-making of power systems, a solution for optimization of the operation adequacy control is proposed. Operation adequacy control is divided into preventive control for adequacy (PCfA) implemented before the occurrence of power shortage, emergency control for adequacy (ECfA) implemented after the occurrence of power shortage immediately, and correction control (CC) implemented when system frequency or voltage is out-of-limit. CC is shared by adequacy control and security control. Take this decomposition based on different timing as core, a three-level decomposition idea is proposed, by which the whole problem is be decomposed into several different sub-problems. And the global quasi optimum is obtained by coordinate the sub-problems which are optimized separately.(5) To accomplish the expanded optimization model and proposed solution preliminary, the optimization of the implementation decision of operation adequacy control for power systems with significant wind power (the representative of intermittent RES) penetration is studied. The control is decomposed into PCfA (including the dispatching of spinning, hot, and cold RC of generation side) and ECfA (IL shedding) based on timing of implementation. Three different search methods are used to obtain the optimal decision, for which the sum of PCfA cost and ECfA risk is minmum. Based on the simulation results, the effectiveness of these three search methods are compared, the improvement of expanded optimization model relative to exsiting model are demonstrated, and the superiority of risk-based method for reserve allocation compared to deterministic and probabilistic method is verified.Generation and transmission expand planning, as an essential measures for power system long-term adequacy, are increasingly influenced by the status of fossil fuel supply. However, the capacity and price of fossil fuel transportation are not the key factors considered in the mainstream researches. In this dissertation these factors are considered to address this issue:(6) If the location of power coal (the representative of fossil fuel) and load demand centre are distributed adversely over a long distance, it is needed to make tradeoff between grid congestion of the power node and its distance to power coal base when selecting the geographical location of new generation capacity. Moreover, the new transmission line will change the situation of grid congestion, which further increases the complexity of this problem. The total risk cost of power supply (includes purchase cost of electricity, congestion risk, and investment cost) during the planning period is taken as the index to estimate the economical efficiency of different generation and transmission expansion palnning with different transport price of power coal.