| dc.description.abstract | The electric utility industry has been going through a revolutionary change over the past decade. The change in perception of economy and social good by governments has been forcing the electrical industry worldwide to move from a vertically integrated industry into a group of entities. Ensuring that all stakeholders get the benefit of open access to the system requires the design of electricity markets. This also requires developing a wide range of regulations so that there is no discrimination and, more importantly, the reliability and integrity of the system is not impaired. These two issues are extremely important because the electrical industry is integrated with all sectors of society, and any failure of the industry could be catastrophic.
This thesis makes an effort to understand some of the features of the market rules that have been designed in a few markets. It proposes a set of procedures which can be adopted to enhance the effectiveness of system operation. It also suggests a few methods which seem easy to adopt and appear to be beneficial.
Management of congestion in the transmission lines is one of the prime issues in deregulated systems. A comparative study of different congestion management schemes has been undertaken by considering demands of varying characteristics and also the transmission losses. Considering a single network, with an identical set of generators and consumers and identical characteristics for all the markets, the impact of different market rules on electricity prices for consumers and generator profits has been assessed.
Determination of Market Clearing Price (MCP) is a major step for market settlement in pool markets. Two new algorithms are presented which are useful in finding the market clearing prices when both the generators and consumers are required to bid using quadratic cost/valuation functions. Such algorithms are useful to find schedules in systems with no congestion. The first algorithm requires a number of steps depending on the type and number of limit violations, whereas the second one helps determine the final schedule in a single step.
Existing deregulated markets, by and large, base the scheduling on real power considerations alone. However, ignoring reactive power issues could result in sub?optimal solutions. It is shown how a comprehensive market scheduling process can be evolved by considering active and reactive power simultaneously. In the proposed method, the exact capability curves of generators are used, and the impact of reactive power flows on network congestion is incorporated. Three different reactive power pricing mechanisms are compared.
A new scheme of computing adjusted load?flow solutions automatically while taking into account the generator Q limits is proposed here. An alternative way of representing the Q–V characteristics of PV buses is presented, and it is shown how this can be integrated with a conventional unadjusted load?flow method to automatically compute the adjusted solution.
The traditional approach of determining schedules for each interval of the next day independently based on the bids may turn out to be infeasible, requiring re?scheduling by the ISO due to violation of ramp rate limits of the participating generators. It is shown here that it is possible to consider the day?ahead market clearing exercise as a single optimisation process incorporating inter?interval constraints on generator outputs. This scheduling is referred to as dynamic scheduling. A heuristic method of solving this problem is proposed and illustrated considering only active power bids as well as both active and reactive power bids.
A new approach to dynamic scheduling has been proposed, integrating the principles of demand?side management with market scheduling. The consumers indicate their daily energy requirement rather than specific hourly requirements. The scheduling procedure proposed, in addition to automatically taking care of ramp rate limits of generators, also alleviates many congestion?related problems. This approach is demonstrated using two different types of bids currently in use, i.e., bids with quadratic cost functions and bids with several steps. | |
