Unit Commitment and Thermal Optimization — Problem Statement

Abstract

The demands made on the electrical energy supply are manifold. Consumers expect an inexpensive, adequate supply at all times with high degree of reliability and closely defined quality criteria (voltage, frequency). National and international institutions require attainment of objectives and compliance with regulations in line with the energy and environmental policies. To an increasing extent, long-term quantitative requirements are also bearing an influence on the operation and development of supply systems. In the FRG this applies to coal with precisely defined obligation to take supplies, it applies in general to natural gas with its ‘take or pay rulings and as a rule, it applies to agreements with zoned tariffs and minimum periods of use. Hydrothermal systems with annual reservoirs have long been familiar with the constraints imposed by yearly reservoir management. The inclusion of other regenerative sources of energy, such as the sun and wind, makes new demands on development and operation of the systems. The way in which the supply sector is developing is characterized by a trend for both the systems themselves and the demands from outside made on them to become more and more complex. With the introduction of a more liberal energy and power supply market (e.g. ‘Third Party Access’) there will be an additional increase in the pressure of costs for the energy supply companies. Consequently, the question of optimal system operation is gaining in importance. Heuristic approaches based on the experience of the load dispatcher are no longer adequate. Mathematical, computer-aided approaches are increasingly being used. However, a closed solution to the overall range of problems involved in system optimization cannot be accomplished in the foreseeable future; it is necessary to break them down into individual problems: i) expansion and design planning, ii) revision planning, iii) fuel resource scheduling and reservoir management, iv) weekly and daily unit commitment, v) load distribution, vi) load management, vii) voltage/reactive power optimization. Due to the long periods of time considered in long-term planning (up to 20 years for development planning and 1–5 years for revision and resource scheduling) stochastic influences such as the failure mode of the components and frequency distribution of the load and sources of supply have a strong bearing. This must be taken into account in modelling and method selection. In the case of short-term planning the emphasis is the exact reproduction of the operating behavior of the individual components. A deterministic consideration is adequate. During the course of this presentation the incidental conditions affecting the individual problems will be described. The question of how to link the problem areas will be looked at more closely. Modelling approaches which have been adopted in operational practice will be illustrated.