Sie sind auf Seite 1von 13

1994 HEAT RATE IMPROVEMENT CONFERENCE

ELECTRIC POWER RESEARCH INSTITUTE

Heat Rate and Maximum Load Capability Improvements


Through Cycle Isolation

Kevin Coons
Senior Mechanical Engineer
Salt River Project
Coronado Generating Station
P.O. Box 1018
Saint Johns, AZ 85936

Joseph G. Dimmick
President
Leak Detection Services, Inc.
708 Melvin Avenue
Annapolis, MD 21401
Heat Rate and Maximum Load Capability Improvements Through
Cycle Isolation

ABSTRACT

Major improvements in maximum load capability and gross turbine


heat rate were obtained at Salt River Project's Coronado Unit 1,
resulting from work done during the Spring 1993 overhaul. Corrected
maximum load increased by 13.1 MW -- from 403.8 MW prior to the
overhaul compared to 416.9 MW after the overhaul. Corrected gross
turbine heat rate was reduced 270 BTU/kWH -- from 7,920 BTU/kWH
before the overhaul to 7,650 BTU/kWH after the overhaul. Of the work
done, the repair of leaking valves had the largest impact on cycle
performance. The reduction of cycle leakage accounted for an increase
of 9.9 MW in maximum load capability and a reduction to gross turbine
heat rate of 190 BTU.

Weekly maximum load tests, which started in August 1992 with the
installation of an on-line monitoring system, show that maximum load
had decreased approximately 4 MW during the six months prior to the
overhaul. During this time there were no significant changes in HP or
IP efficiencies, or any other directly-measured cycle parameters.
Therefore, this degradation was attributed to cycle isolation valve
leakage.

Acoustic emission leak detection methods were used to identify leaking


valves prior to the outage. Of the 138 valves tested for leakage, 31
valves had medium to very large leaks. Of these 31 leaking valves
identified, 30 were repaired or replaced.

Page i of ii
TABLE OF CONTENTS
................................................................................................................................. Page
TABLE OF CONTENTS .................................................................................................. ii
Introduction.................................................................................................................... 1
Overall Performance Improvements ............................................................................ 1
Table 1 – Impact of Repairs on Cycle Performance............................................................1
Cycle-End Performance Degradation .......................................................................... 2
Figure 2 – Unit 1 Maximum Load Test Results ..................................................................2
Figure 3 – Unit 1 Turbine Efficiency Test Results..............................................................2
Figure 4 – Unit 1 VWO Turbine Pressure Ratios ...............................................................3
Cycle Isolation Valve Leakage Improvements ............................................................ 3
Valve Leakage Testing .................................................................................................. 4
Table 5 – Summary of Valve Leakage Survey Results .......................................................4
Valve Leakage Survey Conclusions ................................................................................................................4
Tested Valves and Traps ...................................................................................................................4
Other Problems..................................................................................................................................4
Valve Position.....................................................................................................................4
Valve Leak Survey Recommendations ............................................................................................................5
Tested Valves and Traps ...................................................................................................................5
Cycle Isolation Checklists .................................................................................................................5
Post-Outage Surveys .........................................................................................................................5
Valve Repairs ................................................................................................................. 6
Table 6 – Valves Repaired..................................................................................................6
Turbine Repairs ............................................................................................................. 6
Table 7 – Turbine Efficiency Improvements.......................................................................6
End Packing Replacements..............................................................................................................................7
Table 8 – End Packing Performance Improvements ...........................................................7
Feedwater Heater Improvements .....................................................................................................................7
Steam Flow Changes .......................................................................................................................................7
Comparison to Other Work........................................................................................... 8
Table 9 – Estimated Individual Valve Leakage Effects ......................................................8
Conclusions ................................................................................................................... 9
Appendix A Instrumentation and Test Procedures .................................................. 10
Instrumentation..............................................................................................................................................10
Test Procedures .............................................................................................................................................10

Page ii of ii
Introduction
Salt River Project's Coronado Unit 1 is a conventional, coal-fired plant rated at 410.94
mw and completed in 1979. It is located in Saint Johns Arizona. Weekly maximum load
tests, started in August 1992, after the installation of an on-line heat rate monitoring
system, showed that the maximum load capability had deteoriated by about 4 MW, with a
commensurate increase in heat rate.
During this time, HP and IP turbine efficiencies remained relatively constant. HP and IP
efficiencies were directly measured by the monitoring system. There were also no
changes in throttle flow nor in enthalpy.
Also, the VWO pressure ratios measured by the on-line monitoring system during the
weekly maximum-load tests had been essentially unchanged. It was therefore theorized
that the degradation was due to unaccounted losses, which meant leaking cycle isolation
valves.

Overall Performance Improvements


Results from routine testing done before and after the Unit 1 overhaul indicate that
maximum load capability increased by 13,148 kW (403,693 kW to 416,840 kW), while
gross turbine heat rate, GTHR, decreased by 270 BTU/kWH (7,923 BTU/kWH to 7,653
BTU/kWH). The maximum load and heat rate improvements for major equipment are
tabulated in Table 1 below.
Cycle Maximum Gross Turbine
Parameter Load Heat Rate,
Increase, kW (BTU/kWH)
Cycle Isolation 9,885 193
Valve Leakage
HP Efficiency 2,680 29
IP Efficiency 960 28
N1 Packing 133 3
N2 Packing 738 6
N3 Packing 12 1
FW Heaters 420 7
Flow Change -1,680 3
Totals 13,148 270
Table 1 – Impact of Repairs on Cycle Performance

Coons and Dimmick, Page 1 of 9


Cycle-End Performance Degradation
Figure 2 shows the slow degradation apparent in the trend of the results of maximum load
tests over the six-month period prior to the outage.

Figure 2 – Unit 1 Maximum Load Test Results


Figure 3 shows that the directly-measured Unit 1 turbine efficiencies [HP & IP] remained
stable over the six-month period prior to the outage, while there was a loss of 4 MW in
maximum-load capability.

Figure 3 – Unit 1 Turbine Efficiency Test Results

Coons and Dimmick, Page 2 of 9


Figure 4 corroborates Figure 3, and shows that the VWO turbine pressure ratios were
unchanged during the period that the degradation of maximum-load capability and heat
rate occurred. It was therefore concluded that the losses were in LP Efficiency/Cycle
Isolation. The LP turbines were not opened during the outage, and no repairs to the LP
turbines were undertaken.

Figure 4 – Unit 1 VWO Turbine Pressure Ratios

Cycle Isolation Valve Leakage Improvements


As seen from Table 1, the repair of leaking cycle isolation valves had the strongest impact
on cycle performance. The effect of cycle isolation valve leakage was obtained from the
reduction in unaccounted losses following the overhaul.
Weekly maximum load tests showed [See Figure 2] that maximum load had increased
approximately 13 MW after the overhaul. During this same time [See Table 1] the slight
increases in HP or IP efficiencies and other cycle parameters were measured directly.
Therefore, the remainder of the improvement to maximum load capability was attributed
to repairs to cycle isolation valves.

Coons and Dimmick, Page 3 of 9


Valve Leakage Testing
Acoustic emission leak detection methods were used to identify leaking valves prior to
the outage. Of the 138 valves tested for leakage, 31 valves had medium to very large
leaks. All 31 were opened during the outage, and all but one were were repaired or
replaced because of internal damage. The one valve not repaired was found to have a bad
bypass valve that had given a false indication.
Prior to the overhaul, Leak Detection Services, Inc. was contracted to identify leaking
valves for repair during the scheduled overhaul. Valves important to cycle isolation were
identified by the contractor from Piping and Instrumentation Diagrams (P&ID's). Draft
valve lists and test procedures were prepared in advance and submitted for approval by
Salt River Project personnel. The database classified each valve in terms of the effect of
its potential leakage on overall heat rate. The database includes many valves that are
listed for purposes of test procedure preparation, but are not intended to be tested.
Testing was conducted over a 5-day period a few weeks prior to the regularly scheduled
overhaul. Instrumentation and test procedures are described in Appendix A. Table 5
summarizes the results of the valve leakage survey.

CONCLUSIONS LRG MED SML Totals


Leaking 18 13 16 47
Tight 92
Table 5 – Summary of Valve Leakage Survey Results

Valve Leakage Survey Conclusions


Tested Valves and Traps
Most of the traps and valves identified for testing, plus some others, were tested.
Of the 138 valves and traps tested during the survey, 47, or 34%, were leaking. The
survey found a total of 92 tight valves or traps, of which total, 65 were valves or traps
important to heat rate. The survey also identified 16 small leaks, of which 10 were
important to heat rate.
The percentage of leakers [34%] was lower than the average of 53% [provided from
records of the contractor] over many plants. This was probably because of generally
good maintenance and operation practices at Coronado.
Of the 47 leakers found during the pre-outage suvey, the most serious heat rate
problems were presented by the 25 valves or traps that had large or medium leaks and
are important to heat rate.
Other Problems
Valve Position
One valve was found open when it should not have been. The Hi-Level dump on
the third heater was regulating and the normal drain was wide open. When the
dump was isolated, the heater level went up into the alarm condition. This

Coons and Dimmick, Page 4 of 9


condition usually indicates a tube leak. [A tube leak was confirmed during the
overhaul].
Valve Leak Survey Recommendations
Tested Valves and Traps
All 25 of the large and medium leakers important to heat rate were recommended for
repair or replacement during the next scheduled outage. All of the leaking traps were
recommended for repair even if the leaks were small because experience indicates that
traps deteriorate quickly.
Cycle Isolation Checklists
It was also recommended that Cycle Isolation Checklists be prepared and be checked
completely on every shift during normal operation so small problems can be identified
before they become big and expensive.
Post-Outage Surveys
Another survey was recommended to be conducted immediately after the outage to
check the effectiveness of the repairs and to prevent progressive damage to valves
which are undamaged, but simply not stroked completely shut yet. This survey should
include all of the valves that will have been repaired, replaced, or repacked during the
outage.

Coons and Dimmick, Page 5 of 9


Valve Repairs
Thirty-one valves and steam traps were repaired or replaced during the overhaul, as listed
in Table 6.
2nd Heater Hi-Level Dump HP Feed Pump A Recirc
HP Feed Pmp A Recirc Bypass 1st Ext. Trap Byp before NRV
1st Ext. Trap Man Byp before NRV 1st Ext. Trap Byp after NRV
1st Ext. Trp Man Byp after NRV 2nd Ext. Trap Byp before NRV
2nd Ext. Trp Byp after NRV 2nd Ext. Trp Man Byp after NRV
rd
3 Ext. Trp Man Byp after NRV MS Stop Valve 1 After Seat Drain
Cold Reheat Header Drain Cold Reheat Header Drain Trap Bypass
Cold Reheat Header Drain Feed Booster Pump A Recirc
Feed Pump Discharge Return to Deaerator Feed Booster Pump B Recirc
(Bypass)
1st Heater Hi-Level Dump Feed Pump Discharge Return to Deaerator
(Equalizer)
HP Feed Pump A Recirc Isolation 6th Ext. Trap after NRV
7th Ext. Trap after NRV 2nd Ext. Trap after NRV
4th Ext. Trap Byp after NRV MS Lead Drains to Condenser
Stm Air Htr A Dump to Cond. Rear Side Conv, HDR Drain
Lower Side Conv. HDR Drain MS Line Drain to Condenser
Hot Reheat 1 Drn to Cond.
Table 6 – Valves Repaired

Turbine Repairs
No work was done on the LP turbines. Work done during the overhaul improved
efficiency in both the HP and IP sections of the turbine, as shown in Table 7.
Major contributors to improved performance were:
• Replacement of radial spill strips in the HP/IP sections,
• Installation of Retractable interstage packing,
• Repair of Solid Particle Erosion (SPE) damage to the stationary diaphragms and,
• Repair of stationary diaphragm flow path SPE damage in the first stages of the IP
section.
Cover deposits were removed from the first three stages of the IP section, but these
deposits had very little impact on turbine performance.
Test Date: 15 Feb 1993 4 May 1993 Efficiency Increase
HP Efficiency 82.4 84.5 2.1%
IP Efficiency 91.5 93.2 1.7%
Table 7 – Turbine Efficiency Improvements
A steam path audit was done on the HP/IP sections using a computer program from
Encotech, Inc. The program quantified the impact of turbine component degradation.
Opening measurements were taken by an Encotech field engineer. Closing measurements
were taken by CGS engineering personnel.

Coons and Dimmick, Page 6 of 9


Improvement of HP efficiency was predicted to be 2.6%, compared with 2.1% actual.
The steam path audit predicted no change in IP efficiency, but actual IP efficiency
increased by 1.7%.
The reason for the discrepancy is probably due to having different people take surface
roughness measurements in the opening and closing audits. The surface roughness
measurements are highly subjective and vary from individual to individual. Since surface
roughness accounted for 50% of the total losses in the IP section, a small change in
measured surface roughness had a relatively large effect on predicted performance.
End Packing Replacements
The conventional N2 packing was replaced with retractable, positive-pressure packing.
The conventional packings in the N1 and N3 seals were replaced with new conventional
packing. Packing leakage values and performance improvements were calcualted as part
of the steam path audit and checked with a PEPSE® model. Both the steam path audit
and the PEPSE model agreed closely on the assessed penalties for packing leakage. The
PEPSE results are shown below in Table 8.
Packing Seal Maximum Load, kW Heat Rate,
BTU/kWH
N1 133 3
N2 738 6
N3 12 1
Table 8 – End Packing Performance Improvements

Feedwater Heater Improvements


Improved feedwater heater performance was responsible for approximately 420 kW of the
increase in maximum load and 7 BTU/kWH improvement in gross turbine heat rate.
Final feedwater temperature increased from 472.6 F before the overhaul to 474.6 F after
the overhaul. Contributing factors are repairs to the leaking First- and Second-Point
heater high-level dump valves, and a tube leak (also found during the valve leakage
survey) in the Third-Point heater that was repaired during the overhaul.
Steam Flow Changes
Repair of the SPE damage in the stationary diaphragms following the first stage
decreased the flow-passing capability by approximately 0.5%, which, in turn reduced the
maximum load capability by 1,680 kW. Heat rate effects were negligible. This flow
restriction downstream of the first stage was verified by an expected increase to first-
stage pressure and a decrease in cold reheat pressure. Diaphragm nozzle area
measurements were omitted from the steam path closing audit. This omission was
responsible for the steam path audit not predicting the flow restriction.

Coons and Dimmick, Page 7 of 9


Comparison to Other Work
Branco and Stuckmeyer1 calculated loss recoveries of about 50 to 80 BTU/KWH and 150
BTU/KWH, respectively, after leakage surveys and repair of cycle isolation valves.
Branco performed unit heat rate tests before and after the repairs. Stuckmeyer
documented, with notes and photographs, the as-found condition of each leaker when it
was opened for inspection. He then calculated the effect of leakage of each valve by
taking into account the valve characteristics, the enthalpy loss and the extent of the
damage observed. When summed, the total effect amounted to about 150 BTU/KWH.
These figures are of the same order of magnitude as the present result.
As a check on the data from the on-line monitor, calculations were made for some of the
leakers at Coronado. The PEPSE system was used to determine the heat rate and
maximum load effects of a nominal leak rate of 10,000 lb/hr through each valve. The
results of those calculations are shown in Table 9, below.
10,000 LB/HR LEAK TO Load Heat Rate
CONDENSER Reduction, Penalty,
KW BTU/KWH
Main Steam 1,486 34
Aux Steam 1,242 24
Hot Reheat 1,078 24
Cold Reheat 1,077 30
1st Extraction 1,077 30
2nd Extraction 1,040 23
3rd Extraction 841 18
4th Extraction 612 13
5th Extraction 339 8
6th Extraction 213 5
7th Extraction 92 2
1st Heater 162 3
2nd Heater 104 2
3rd Heater 60 2
5th Heater 11 1
6th Heater 5 1
Table 9 – Estimated Individual Valve Leakage Effects
Hopson2 used a turbine cycle model to calculate the heat rate degradation and generation
loss due leakage of certain cycle isolation valves. He made his calculations on the basis
of assumed leaks of 5% and 50% of rated flow for each valve. His data are consistent
with Table 9, and with the later work done by Stuckmeyer.
® PEPSE is a registered trademark of the NUS Corporation.

1Branco,Maria, et al, "Heat Rate Recovery by Cycle Isolation Diagnosis and Correction," 1991 Heat Rate
Conference, EPRI
2Hopson,Warren H., et al, "The Best Buy in Heat Rate Recovery - Turbine Cycle Isolation Maintenance,"
EPRI Heat Rate Improvement Workshop, 1985

Coons and Dimmick, Page 8 of 9


Conclusions
The newly-installed on-line monitoring allowed CGS engineers to detect and partially
diagnose the degradation of both heat rate and maximum-load capability. Pre-outage
analysis of the weekly test data led to the theory that the degradation was principally in
unaccounted losses. This was confirmed by the post-outage measurements..
As a result of all the repairs undertaken during the outage, maximum-load capability and
heat rate both improved by about 3%.
• Valve repairs contributed 75% of the maximum-load improvement, and 71% of the
heat rate improvement.
The turbine repairs, packing repairs, and feedwater heater repairs combined were less
significant, but combined they contributed 25% of the maximum-load improvement and
29% of the heat rate improvement measured after the outage.
The cycle isolation valve database identified the 150 valves and traps that were important
to heat rate and maximum-load capability. The valve leakage survey pinpointed the 25
individual valves and traps which were the cause of most of the degradation, and
coincidently identified a leaking heater. Having this information prior to the outage
permitted the limited repair resources to be concentrated on the individual valves that
most affected the cycle degradation.

Coons and Dimmick, Page 9 of 9


Appendix A
Instrumentation and Test Procedures
Instrumentation
All instrumentation was provided by Leak Detection Services, Inc. The Acoustic Valve Leak
Analyzer (AVLA) is a portable instrument powered by rechargeable batteries. Originally
developed for nuclear submarines, it is in world-wide use in power plants, refineries, pipelines,
and U. S. Navy ships. From transducers temporarily attached to the outside of valves, it plots
signatures of acoustic amplitude versus frequency over the range of 10 kHz to 200 kHz. The
transducers detect ultrasonic acoustic emissions which are characteristic of internal leakage. Its
two-channel capability permits the leakage signals to be separated electronically from the
background noise even in cases where it is not possible to release the pressure to obtain a
separate background signature. High-Range transducers were required and used with the AVLA
for most of the signatures to keep the relatively high leak signal and background noise levels
from overranging the AVLA.
Test Procedures
Three basic test procedures can be used.
Where block valves can be closed to relieve the differential pressure across the valve being
tested, the Signature Comparison Method can be used. The pressure signature is recorded with
the valve closed against full pressure. The background signature is recorded with either the
upstream or the downstream block valve closed, while the valve from which the pressure
signature had been recorded is open. In the simplest case, if the pressure signature exceeds the
background signature, the valve is leaking.
In cases where there are no block valves, or the block valves cannot be closed, the Differential
Signature Method is used. Background signatures recorded from transducers about ten pipe
diameters up and downstream from the valve under test are automatically subtracted by the
AVLA from the valve signature, with the difference being plotted as a Differential Signature. In
general, if the Differential Signatures are positive, there is a stronger signal at the valve than on
the connected piping, which means that the valve leaks.
When the objective is to determine, with the least difficulty, which are the worst leakers among a
group of essentially identical valves at identical differential pressures, it is possible to record only
the Pressure Signatures, and then make comparisons among the similar valves. The drawback to
the Direct Comparison Method is that it is not possible to determine conclusively that any of the
tested valves are tight. This method is used only when testing time is restricted, as when the test
team is being exposed to radiation or heat.

Das könnte Ihnen auch gefallen