Natural Cooling and Startup of Steam Turbines: Validity of the Over-Conductivity Function

Abstract

The temperature drop during natural cooling and the way in which the steam turbine restarts have a major impact on the cyclic lifetime of critical parts and on the cyclic life of the whole machine. In order to ensure the fastest startup without reducing the lifetime of the turbine critical parts, the natural cooling must be captured accurately in calculation and the startup procedure optimized. During the cool down and restart, all turbine components interact both thermally and mechanically. For this reason, the thermal analyst has to include, in his numerical model, all turbine significant parts—rotor, casings together with their internal fluid cavities, valves, and pipes. This condition connected with the real phenomenon lead-time—more than 100 hours for natural cooling—makes the analysis time-consuming and not applicable for routine projects. During the past years, a concept called “over-conductivity” was introduced by Marinescu et al. (2013, “Experimental Investigation Into Thermal Behavior of Steam Turbine Components—Temperature Measurements With Optical Probes and Natural Cooling Analysis,” ASME J. Eng. Gas Turbines Power, 136(2), p. 021602) and Marinescu and Ehrsam (2012, “Experimental Investigation on Thermal Behavior of Steam Turbine Components: Part 2—Natural Cooling of Steam Turbines and the Impact on LCF Life,” ASME Paper No. GT2012-68759). According to this concept, the effect of the fluid convectivity and radiation is replaced by a scalar function K(T) called over-conductivity, which has the same heat transfer effect as the real convection and radiation. K(T) is calibrated against the measured temperature on a Alstom KA26-1 steam turbine (Ruffino and Mohr, 2012, “Experimental Investigation on Thermal Behavior of Steam Turbine Components: Part 1—Temperature Measurements With Optical Probes,” ASME Paper No. GT2012-68703). This concept allows a significant reduction of the calculation time, which makes the method applicable for routine transient analyses. The paper below shows the theoretical background of the over-conductivity concept and proves that when applied on other machines than KA26-1, the accuracy of the calculated temperatures remains within 15–18 °C versus measured data. A detailed analysis of the link between the over-conductivity and the energy equation is presented as well.

The impact of gas turbine compressor rotor bow on aircraft operations

“…In steam turbines, temperature gradients driven by natural convection are known to consume LCF life of components, and as such, transient thermal analysis of shutdown and start-up sequences of large steam turbines receive a lot of attention in the field. A substantial body of work using finite element methods to analyse the thermal behaviour of steam turbines has been published in recent years (3)(4)(5)26,27) . Work by Marinescu and associates (3)(4)(5)26,27) studied natural convection heat transfer in steam turbines by manipulating the conductivity function of the 'fluid' and 'solid' domains, using a code developed by Rolls-Royce and Alstom Power.…”

“…A substantial body of work using finite element methods to analyse the thermal behaviour of steam turbines has been published in recent years (3)(4)(5)26,27) . Work by Marinescu and associates (3)(4)(5)26,27) studied natural convection heat transfer in steam turbines by manipulating the conductivity function of the 'fluid' and 'solid' domains, using a code developed by Rolls-Royce and Alstom Power. While this work has developed a useful technique for analysing the thermal behaviour of steam turbines over their significantly long cooldown periods (often in excess of 100 hours), it has yet to be employed to study the contribution of simple parametric design aspects for the purposes of general understanding of the thermal bow phenomenon.…”

Fault Tolerant Thermal Control of Steam Turbine Shell Deflections

“…The rotor can be modeled in a lumped parameter fashion as a single node. It is assumed that the rotor can transfer heat to each shell element via convection [8]. The prediction of the rotor temperature is critical for the proper budgeting and management of the low cycle fatigue impact [9].…”

Multi‐scale time integration for transient conjugate heat transfer

“…The ratio between the two time scales can be up to 10 4 for gas turbine *Department of Engineering Science, University of Oxford, Parks Road, Oxford, OX1 3PJ, UK. [6]). This scale disparity should not affect the steady conjugate heat transfer (CHT) solution behaviour (computational efficiency and accuracy) in any way significant as the domain-dependent time stepping for common CHT solvers can be readily adopted.…”

“…Another example of the practical need for transient CHT solutions is the increasingly flexible operations of conventional steam turbine generators as required to provide top ups matching the emerging renewable power generations (e.g. [6]). When shorter start-up and shut-down processes are pursued, transient temperature distributions in solid domains will be far from being quasi-steady, leading to potentially high-transient thermal loads, which need to be predicted with adequate transient methods.…”