Modelling and optimisation of the Otahuhu B combined cycle gas turbine power station

摘要

The generation of electrical power in New Zealand is currently complicated by the governments desire to sell 49% of the state-owned existing power stations, the fact that Rio Tinto are threatening to sell Tiwai Point Aluminium smelter which currently consumes 15% of the national electrical power, and the geographical peculiarities of New Zealand where the hydro is generated in the south, and the energy consumed in the north, transmitted along a thin and narrow corridor. Contact Energy, which run a large 400 MW combined cycle gas turbine (CCGT) power plant in Auckland, bid, as do all electrical generators in New Zealand, on the national electricity market. To be profitable, the station must closely follow the time-varying electrical market, and be able to produce sufficient energy on demand. Optimising such a production requires models that accurately predict steam thermodynamics and the heat transfer within the boiler, and models that predict the combustion thermodynamics in the gas turbine. The Industrial Information and Control group have developed a comprehensive heat-recovery steam generator (HRSG) package that can be use to predict the steady-state operating conditions of the power station over a wide operating range. In addition, the group have developed a widely-used optimisation platform that can be used to establish optimal operating conditions for given external environmental conditions such as electrical closing price and gas prices. However the one thing missing to date is consideration of the dynamic response of the plant. Currently it is known that the combined plant is relatively slow to respond to the quickly changing market demands, especially when the steam boiler is used. If however, the less efficient gas turbine is used alone, (with the boiler switched off), the dominant time constants of the plant are considerably reduced. This complicated the optimisation problem since using the boiler restricts the ability of the plant to respond quickly to market demands, but if used, improves the overall energy efficiency. This paper describes the application of an optimal dynamic modelling project applied to an actual 400 MW power station. The paper validates the first-principle steady-state models, and develops simple dynamic models of the boiler and the gas turbine using historical plant data. The paper then explores various optimisation scenarios and illustrates the possible benefits.