Testing a new design
After reducing Unit 3's ramp rate during dispatch, Sim Gideon began investigating other ways to reduce cyclic life expenditure. One was to apply both a modified sliding-pressure (MSP) approach and a multivariable steam temperature control approach to achieve higher ramp rates.
In MSP, throttle pressure is varied toward a target value under the supervision of a superheat steam temperature control system with sufficient range. If the range is exceeded (for example, if spray valves are nearly wide open or closed), throttle pressure is pulled back until steam temperature conditions are favorable. In multivariable control, the throttle pressure setpoint is adjusted based on HP rotor temperature, superheat steam temperature, and superheat spray demand variables. For typical dispatch operation, there should be adequate opportunities to make throttle pressure adjustments while meeting load and ramp rate objectives.
Testing during the investigation demonstrated that using MSP at low loads reduces HP rotor temperature change during ramping and allows faster ramp rates consistent with CLE guidelines. However, MSP operation also requires coordinated use of multivariable steam temperature control to maintain superheat steam temperatures near the setpoint. The number of burners in service and their tilts, superheat sprays, and IFGR damper position must be precisely controlled. Due to the slow temperature response at low load and the multivariable requirements, Sim Gideon staffers chose a POMS as the control methodology.
To further increase ramp rate potential, engineers recommended reducing the maximum load for fast ramping from 330 MW to 315 MW. This would provide a zone from 60 MW to 315 MW at the fastest possible ramp rate. Load increases from 315 MW to 330 MW would be at a very low ramp rate. This restriction is due to the high first-stage metal temperature gradient at high loads.
The table presents calculated average ramp rates consistent with the 10,000-cycle HP rotor CLE curve. Note that in MSP mode at low load, there is an increase in superheat steam temperature from 995F to 1,000F.
The electric slide. Performance improvement achieved by sliding-pressure operation. Source: LCRA
Tight curves ahead
Achieving the desired ramp rate requires close control of superheat steam temperature. By coordinating control of the throttle pressure setpoint, the number of burners in service and their tilts, superheat sprays, and the position of the IFGR damper, the POMS system can provide adequate superheat steam temperature control while maintaining first-stage metal temperatures within desired constraints.
Sim Gideon engineers then developed a test plan for investigating the ability of control methodologies to increase unit ramp rate while maintaining the rate of HP turbine metal temperature change within acceptable limits. The unit ramp rate is currently limited by HP turbine rotor CLE considerations, as determined by the turbine vendor's starting and loading instructions (SALI) curves (Figure 2).
2. Limiting factor. Unit ramp rate is limited by the high-pressure turbine rotor's cyclic life expenditure. Source: LCRA
The asymptotic shape of the SALI curves suggests that a small decrease in total HP rotor temperature change provides a large potential increase in the allowable rate of change of temperature for the same CLE. If the total HP rotor temperature change could be reduced from the current maximum of 170F to a range of 115F to 135F, the ramp rate could be maintained at a higher value without exceeding the desired CLE.
The testing also investigated the use of MSP to increase HP rotor temperature at low loads. By operating at lower throttle pressure at low loads, throttling losses are reduced and higher steam temperatures are attained at the turbine's first stage. With increasing load, the throttle pressure is ramped back to normal, producing a net lower temperature change of the first-stage metal over the load range.
Temperature changes in the HP rotor cause the stress expressed as CLE by the turbine vendor. Each temperature cycle consumes a percentage of cyclic life. LCRA would like to operate on the unit's 10,000-cycle curve, which corresponds to the 0.01% CLE/cycle curve in Figure 2. The red vertical line at 170F corresponds to the maximum HP rotor temperature change from minimum load to full load. To remain within the 10,000-cycle curve, the rate of temperature change should not exceed 280 degrees F/hr. The dispatch rate and load end points dictate both the change in HP rotor temperature and the rate of change of temperature. Figure 3 presents first-stage and superheat temperature data during typical load dispatch operation.
3. One day at a time. Load, turbine first-stage temperature, and superheat temperature during dispatch of Sim Gideon Unit 3 on a typical day. Source: LCRA
Testing confirmed that there is considerable variability in the relationship between first-stage metal temperature and load due to the slow response of turbine metal temperature, particularly at low load. Figure 4, developed by LCRA engineers, represents the steady-state relationship between load and first-stage metal temperature.
The steep slope of the overall curve in Figure 4 suggests that restricting fast ramping at very high loads may have the beneficial result of reducing HP rotor stress. At lower loads, where the curve isn't as steeply sloped, the benefit may not be as great. However, data collected during the tests do indicate the potential benefit of reducing temperature variation even at low loads by raising the first-stage metal temperature by operating the unit in sliding-pressure mode.
4. Hot topic. LCRA engineers determined the steady-state relationship between load and turbine first-stage metal temperature, a key for controlling ramp rates. Source: LCRA
Email
Comments
Print
Reuse
