A methodology to determine best-suited waiting-time periods for turbine start-up under fluctuating resources

Abstract

— It is not uncommon for engineering process plants to comprise energy recovery. When power generation takes place under such circumstances, the nature of interlinked processes may result in fluctuating resource availability. Such fluctuations may, however, result in turbines tripping due to insufficient availability at times. Power generation under such conditions typically comprises turbine protection measurements in an attempt to prevent unnecessary trip occurrences. One such measurement is to avoid a possible restart-trip scenario in close proximity, by enforcing adequate start-up constraints. These constraints dictate the minimum waiting-time period that needs to be enforced, where resource availability is sufficient to keep the turbine(s) operational. Start-up protection measurements are typically enforced without changing the time-constraint over time. Although such a measurement is required it entails time intervals where potential power generation goes to waste due to a turbine’s non-operational status. As the enforced waitingtime period increases more power generation potential goes to waste; however, reduced periods may result in turbines being restarted, only to trip in near future. A trip does not only necessitate another waiting-time period, but reduces a turbine’s life expectancy. This paper presents two models; the first maximises power generation amongst two turbines where waiting-time periods are incorporated as a variable. These results can then be used in combination of the second proposed model. The second model is a unique methodology that can be used to investigate the effect of turbine start-up waiting-time periods and how it influences the combination of power generation and turbine trips. This methodology is, furthermore, incorporated to determine what the ideal startup period should be for an energy recovery plant that operates under fluctuating resource availability. A case study is presented that demonstrates the working ability of the proposed method and results show that by incorporating this methodology the engineering plant can generate an additional 0.46 MW per annum