Use this URL to cite or link to this record in EThOS:
Title: Analysis of market incentives on power system planning and operations in liberalised electricity markets
Author: Xu, Meng
ISNI:       0000 0004 5994 7581
Awarding Body: Durham University
Current Institution: Durham University
Date of Award: 2016
Availability of Full Text:
Access from EThOS:
Access from Institution:
The design of liberalised electricity markets (e.g., the energy, capacity and ancillary service markets) is a topic of much debate, regarding their ability to trigger adequate investment in generation capacities and to incentivize flexible power system operation. Long-term generation investment (LTGI) models have been widely used as a decision-support tool for generation investments and design of energy policy. Of particular interest is quantification of uncertainty in model outputs (e.g., generation projections or system reliability) given a particular market design while accounting for uncertainties in input data as well as the discrepancies between the model and the reality. Unfortunately, the standard Monte Carlo based techniques for uncertainty quantification require a very large number of model runs which may be impractical to achieve for a complex LTGI model. In order to enable efficient and fully systematic analysis, it is therefore necessary to create an emulator of the full model, which may be evaluated quickly for any input and which quantifies uncertainty in the output of the full model at inputs where it has not been run. The case study shows results from the Great Britain power system exemplar which is representative of LTGI models used in real policy processes. In particular, it demonstrates the application of Bayesian emulation to a complex LTGI model that requires a formal calibration, uncertainty analysis, and sensitivity analysis. In power systems with large amounts of variable generation, it is important to provide sufficient incentives for operating reserves as a main source of generation flexibility. In the traditional unit commitment (UC) model, the demand for operating reserves is fixed and inelastic, which does not reflect the marginal value of operating reserves in avoiding the events of load shedding and wind curtailment. Besides, the system-wide reserve constraint assumes that the operating reserve can be delivered to any location freely, which is not true in real-world power system operations. To recognize the value and deliverability of operating reserves, dynamic zonal operating reserve demand curves are introduced to an enhanced deterministic UC model for co-optimizing the day-ahead schedules for energy and operating reserves. In the case study on the RTS-73 test system, comparisons are made between the choices of reserve policies (e.g., single, seasonal or dynamic zones) and of different reserve zonal partitioning methods. Results suggest that the enhanced deterministic UC model produces on average lower operational cost, higher system reliability and higher energy and reserve revenues than the traditional one. Finally, we discuss future directions of methodological research arising from current energy system challenges and the computer models developed for better understanding of the impacts of market incentives on power system planning and operations.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available