Steam turbine designers and researchers agree that for large units (about 300 MW and up) the ramp rates during start-up, large load changes, and other transients are mainly limited by unsteady thermal stresses in the rotors of high-pressure (HP) and intermediate-pressure (IP), or integrated HP-IP cylinders. Cycling thermal stresses or low-cycle fatigue that exceeds prescribed limits can lead to cracking of the rotors.
The allowable ramp rates can be found with the help of charts that quantify the turbine life "used up" by each cycle (Figure 1). These "lifetime consumption" charts are calculated for each specific type of HP and IP rotor, taking into consideration the rotor's radial size (the external and internal radii of the rotor body in the most thermally stressed section), the thermophysical and strength properties of the metal it is made of, and the stress concentrators on the heated surface. The ramp rate for varying steam temperatures is derived from the expected transient operating conditions and from the permissible number of cycles that will give birth to a crack, or from the specific lifetime expenditure, measured in percent per cycle. For less frequent transients, the admissible number of cycles is set lower, resulting in greater rates of heating steam temperature changes. The least number of cycles is assumed for cold start-ups; the greatest number is assumed for load changes in the governed range. Figure 1 also shows a shaded zone that is considered "out of bounds" based on the threat of a brittle fracture of the rotor under the combined action of tensile (centrifugal plus thermal) stresses in its depth, near the central bore or rotor axis.
1. Setting the ramp rate limits. Permissible turbine ramp rates can be derived from ranges in rotor surface metal temperature changes and specific low-cycle fatigue limits. Source: General Electric Co.
What to monitor?
The conditions of heat transfer from steam to a rotor's surface are much more intense than those for stator components. In addition, the ratio of a rotor's heated surface area to its mass is much higher than for stator components. As a result, rotors are much more sensitive to variations in heating steam temperatures caused, for example, by load changes or excursions of main or reheat steam temperatures at the turbine entrance.
Operators can therefore expect that if they keep the maximum thermal stresses on a rotor's heated surface within the allowable range, the maximum thermal stresses in the stator components of the same section will not exceed their design limits. However, the converse may not be true. That's why it is essential to continuously monitor the actual temperature and thermal-stress states of the rotors to ensure turbine reliability. Because the main and reheat steam temperatures vary independently during transients, it is important to monitor both the HP and IP rotors or, for turbines with integrated HP-IP cylinders, both the HP and IP steam admission rotor sections.
Some turbine designs may impose additional limitations on the handling of transients caused by thermal stresses in casings of the HP and/or IP cylinders. This mainly refers to cold and warm start-ups, as well as shutdowns with forced cooling down of the turbine. These thermal stresses are commonly produced by temperature differences within the casing flanges (across their width and between the wall, flanges, and stud-bolts). Axial temperature unevenness in the casing wall also can be substantial in the neighborhood of intercasing chambers with significantly different heating steam temperatures or the conditions of heat transfer from steam to the casing's surfaces. In this case, temperature monitoring of high-temperature rotors should be supplemented with metal temperature measurements of the corresponding parts of the high-temperature casings. Also, for some turbine types it is necessary to monitor the temperature and thermal-stress states of the HP valve steam-chests during start-up.
For some large steam turbines it is also desirable to monitor the temperature state of low-pressure (LP) rotors. This is true for turbines with welded or forged LP rotors and elevated reheat steam temperatures, as well as for some large wet-steam turbines of nuclear power plants. Monitoring enables control of the reheat steam temperature (or the steam temperature after the moisture-separators-and-reheaters, in the case of wet-steam turbines) during start-ups without the appearance of inadmissible tensile stresses in the LP rotors (in their steam admission parts) caused by superposition of centrifugal and thermal stresses.
The maximum thermal stress on a rotor's heated surface is commonly considered to be proportional to a so-called "effective" metal temperature difference across the rotor's radius in its most-stressed section: between the surface temperature and the integral average temperature in the section. The tensile thermal stress in the rotor depth (on the central bore surface or near the rotor axis) is commonly derived from the entire radial temperature difference between the rotor's heated surface and depth. The aforementioned metal temperatures and temperature differences should be the object of on-line temperature monitoring for steam turbine rotors. However, because their direct thermometry is too complicated and unreliable, the problem must be solved indirectly by means of physical or mathematical modeling.
Email
Comments
Print
Reuse
