IX. CONCLUSIONIn order to reflect the cost of network security in network charging, this paper, for the first time, incorporates nodal unreliability tolerance and the supply component’s reliability into a distribution network charging model. It works by reflecting how a component’s failure rate and mean time to repair would affect its ability to deliver energy to customers, and how customers’ tolerance to supply interruption under contingencies will impact the component’s time to reinforce. Based on the analysis in the paper, the following observations can be obtained: The proposed model overcomes two disadvantages of the existing pricing models: i) relying on deterministic criteria to reflect the cost of network security; and ii) unable to respect the nodal unreliability tolerance. It works by incorporating both factors into assessing the impact of a nodal perturbation on assets’ investment horizons. The new model can better reflect the actual network planning practice and the stochastic features of network failures. Component reliability, allowed nodal load loss and the failure duration are the three major factors influencing nodal reliability. The new model reduces the nodal charges by considering them: more reliable components and larger unreliability tolerance lead to smaller charges, and vice versa. The new model maintains the merits of the original LRIC model of being able to produce locational and costreflective charges to influence prospective users’ behaviors to maximize the utilization of the existing networks, particularly those with higher reliability level and shorter time to repair. One problem with the new model is that it would need substantial computational effort to analyze network contingencies for large-scale systems. We ran the proposed model on a practical EHV distribution network in the UK,comprising 1,898 busbars: it took the original LRIC methodapproximately 30 minutes to calculate charges for all load busbars but the proposed approach about 27 hours.Although it is a 50 fold increase in time, network charges are calculated on an annual basis and thus such an increase in running time is affordable by network operators to better reflect their investment decisions. Besides, the advance in computational techniques can benefit the application of the proposed method.Future research needs to investigate the correlation between nodal charges, the costs of different reliability improvement strategies, and how customers might respond to the locationlcharges so as to find the economic equilibrium for both network operators and users. Additionally, the benefits in network investment deferral brought about by renewable generators are not considered in distribution network pricing in the UK currently, due to their output intermittency. Future work can also be conducted to examine how it would be morecost-effective to charge them, respecting the impact that they not only can bring forward network reinforcement but also might defer the needed network investment
đang được dịch, vui lòng đợi..