Thermal Load Effect on Vibration on a Reactor FWP

By:

Maki M. Onari MSI Principal Engineer

Manager of Turbomachinery Testing11 Apollo Drive

Whippany, NJ 07981Tel: (973) 326-9920

maki.onari@mechsol.com

George InchExelon Corp. Senior Staff Engineer

Nine Mile Point Nuclear Generating Station348 Lake Road

Oswego, NY 13126Tel: (315) 349-2441

george.inch@exeloncorp.com

Embassy SuitesSan Antonio, TX1.25 – 1.28.2016

2016 Feedwater SystemReliability Users Group (FSRUG)

PresentersMaki M. Onari:• Mechanical Solutions, Inc. – Manager of Turbomachinery Testing -

Principal Engineer: responsible for all MSI Turbomachinery Testing• B.S.M.E., Zulia University, Venezuela• Rotating Equipment Engineer, PDVSA: responsible for the predictive

maintenance of one of the largest petrochemical complex in Latin America• Co-Author Pump Vibration Chapter, McGraw-Hill Pump Handbook• Member of ASME and the ISO TC108/S2 Standards Committee for

Machinery Vibration• Presented tutorials, case studies, and lectured at the Texas A&M Symposia

since 2007 as lead instructor on vibration short courses.

George Inch:• Sr. Staff Engineer Mechanical Design Engineering at Nine Mile Point

Nuclear Station (Exelon Nuclear)• 26 years of experience at NMPNS

General Information

Constellation Energy – Exelon Corp.

Nine Mile Point Nuclear Station (NMPNS) Unit 2 – 1150 MW (Oswego, NY)

Application: Reactor Feed Water Pump (2FWS-P1A/B/C)

Manufacturer: Byron Jackson (Legacy Pump OEM)

Type: 20x20x22 HSB – 2 Stage (barrel type radial split)

Capacity: 8,000 gpm (min flow) to 22,870 gpm (max flow)

Normal operation steady slightly above BEP

Suction Press.: 550 psig

Discharge Press.: 1,200 psig

Speed: 3,339 rpm (55.7 Hz)

Temperature: 382 ºF

Normal Load: 10,500 to 11,500 BHP

General Information Cont.

Driver: Electric Machinery Induction Motor

(rated 16,500 HP @ 1785 rpm (29.8 Hz)

Gearbox: General Electric Speed Increaser (Gear ratio 1:1.87)

Modifications: Impeller vanes were changed from 5 to 7 vanes

New min-flow valve

Smoother suction passage

New bearing brackets and housings

New suction covers and bearings

General Information Cont.

Former 2FWS-P1A Cross Section

Byron Jackson 20x20x22 HSB – 2 Stage

Background• Mechanical seals have experienced repetitive failure since 1987 (MTBF of 6

months, mostly on the Alpha FWP). Possible axial shuttling of the pumprotor.

• Based on analysis performed by the OEM, it was concluded that the pumps

required a series of modifications in order to improve seal reliability and

also increase plant capacity by 15% (EPU).

– Redesigned mechanical seals

– New impellers

– Flow path to the first stage impeller

– Inboard (IB) suction cover

Background– Alpha FWP modified in spring 2010. High synchronous vibration was

detected only at the IB end of the pump shaft (over 4.2 mils pk-pk).

– FAT vibration on the order of 1.0 to 1.5 mils pk-pk.

– Vibration amplitude proportional to the load of the plant.

– It was recommended to modify the cold alignment between the pump

and the gearbox (on-line hot alignment monitoring).

General Information Cont.

Modified 2FWS-P1A Cross Section

Vibration Testing Approach

• Experimental Modal Analysis (EMA) test

• Continuous Monitoring (CM) testing

• Operating Deflection Shape (ODS) testing

Experimental Modal Analysis (EMA)

June 2010 Vibration DataAlpha FWP Initial Testing

2FWS-P1A Pump Typical FRF PlotsFrequency Respo nse(Signal 4,Signal 17) - M ark 1 (M agnitude)

Working : 01 h rec 6 : Input : Enhanced

0 40 80 120 160 200 240 280 320 360 400

1u

3u

10u

30u

100u

300u

1m

3m

10m

[Hz]

[g/lbf] Frequency Respo nse(Signal 4,Signal 17) - M ark 1 (M agnitude)

Working : 01 h rec 6 : Input : Enhanced

0 40 80 120 160 200 240 280 320 360 400

1u

3u

10u

30u

100u

300u

1m

3m

10m

[Hz]

[g/lbf]

44 Hz

89 Hz1x r

pm

(57 H

z) Horizontal

Direction

Axial

Direction

Frequency Respo nse(Signal 2,Signal 17) - M ark 1 (M agnitude)

Working : 01 a rec 7 : Input : Enhanced

0 40 80 120 160 200 240 280 320 360 400

1u

3u

10u

30u

100u

300u

1m

3m

10m

[Hz]

[g/lbf] Frequency Respo nse(Signal 2,Signal 17) - M ark 1 (M agnitude)

Working : 01 a rec 7 : Input : Enhanced

0 40 80 120 160 200 240 280 320 360 400

1u

3u

10u

30u

100u

300u

1m

3m

10m

[Hz]

[g/lbf]

1x r

pm

(57 H

z)

59.25 Hz

Frequency Respo nse(Signal 1,Signal 17) - M ark 1 (M agnitude)

Working : 01 v rec 8 : Input : Enhanced

0 40 80 120 160 200 240 280 320 360 400

1u

3u

10u

30u

100u

300u

1m

3m

10m

[Hz]

[g/lbf] Frequency Respo nse(Signal 1,Signal 17) - M ark 1 (M agnitude)

Working : 01 v rec 8 : Input : Enhanced

0 40 80 120 160 200 240 280 320 360 400

1u

3u

10u

30u

100u

300u

1m

3m

10m

[Hz]

[g/lbf]

395 Hz

1x r

pm

(57 H

z)

89 Hz

Vertical

Direction

2FWS-P1A Pump Main Mode Shapes

Lateral Rocking Mode @ 44 Hz Twist Mode @ 89 Hz

Axial Rocking Mode @ 59.3 Hz OBB Vertical Mode @ 395 Hz

Continuous Monitoring (CM) Testing

14

6600 gpm min-flow

9300 gpm min-flow

9300 gpm to Reactor

14000 gpm to Reactor

17000 gpm to Reactor

18000 gpm to Reactor

20000 gpm to Reactor

15400 gpm to Reactor

Pump Shaft Vibration

16

Pump Shaft Vibration

17

Pump IBB Housing VibrationAxial

Horizontal

Vertical

18

Pump OBB Housing VibrationAxial

Horizontal

Vertical

19

* View from pump end unless otherwise stated

Gearbox HS DE Orbit Plots

6600 GPM14:00, 6-12-2010

9300 GPM14:30, 6-12-2010

9300 GPM15:00, 6-12-2010

20000 GPM3:56, 6-13-2010

20

Pump IBB Orbit Plots

6600 GPM14:00, 6-12-2010

9300 GPM14:30, 6-12-2010

9300 GPM15:00, 6-12-2010

20000 GPM3:56, 6-13-2010

* View from pump end unless otherwise stated

21

Pump OBB Orbit Plots

6600 GPM14:00, 6-12-2010

9300 GPM14:30, 6-12-2010

9300 GPM15:00, 6-12-2010

20000 GPM3:56, 6-13-2010

* View from pump end unless otherwise stated

Shaft Center Line Position Relative to Original Cold Alignment Position

22

Operating Deflection Shape (ODS)

23

• Natural excitation signature of the rotating equipment

• Steady load operation

• Hundreds of vibration measurements with reference signal

• Data base of amplitude vs. frequency and phase angle

• 3-D CAD model assigning motion to each individual vibration data point

• Create animations of the equipment and related piping and foundation

Operating Deflection Shape (ODS)

24

ODS Animation at 1x rpm

25

• Misalignment: Severe offset misalignment between the HS

gearbox pinion shaft and the pump.

• Coupling Imbalance: Eccentricity of the coupling spacer from the

coupling mounting center line.

• Thermal growth between the pump shaft and the HS gearbox

pinion.

• Modify the cold alignment

• Verify the radial position of the coupling spacer for potential

eccentricity that could lead to imbalance forces.

Preliminary Conclusions & Recommendations (Possible Causes)

March 2011 Vibration DataAlpha FWP Testing – Root Cause Detection

• Radial bearings with tri-land bearings (7 mils diametral clearance) toforce the shaft operation closer to the center line of the pump.

• Reduced the torque on the OB pump foot bolts to facilitate thermalgrowth according to the OEM’s specifications.

• Adjusted the turn-buckles as per OEM’s specifications.

• Verified/ corrected the misalignment of the entire train.

• Installed temporary thermocouples at the IB and OB faces of the pumpcasing to measure transient temperature field.

• Installed temporary laser instrumentation to measure thermal growth ofthe pump casing in the axial and horizontal directions.

• Modified the warm-up procedure to alleviate rubbing during highthermal gradients within the pump (reduced seal cooling water flowrate from 10 to 5-7 GPM).

Modifications Performed in March 2011

GB HS DE - Shaft Vibration

Start-up Shut-down

Pump IBB - Shaft Vibration

Start-up Shut-down

Pump OBB - Shaft Vibration

Start-up Shut-down

GB HS DE - Shaft Movement

Warm-up Start-up Shut-down Cool-down

Pump IBB - Shaft Movement

Warm-up Start-up Shut-down Cool-down

Pump OBB - Shaft Movement

Warm-up Start-up Shut-down Cool-down

Shaft Orbits

N

View from East (pump OBB)

Shaft Orbits

N

View from East (pump OBB)

Shaft Centerline Plots

Through Cool-down

XY

XY

XY

N

View from East (pump OBB)

Preliminary Conclusions / Observations1. The root cause of the elevated vibration amplitude at the IBB of the

pump was due to thermally induced pump shaft bending.

2. Excessive pump shaft vibration. Mostly at 1x rpm, representing 57% ofthe total bearing clearance of 7 mils. There was a reduction in theamplitude by approximately 2.0 mils pk-pk as compared to the start-upperformed in December 2010 (~6.0 mils pk-pk).

3. At the start-up, the shaft bow effect quickly moved from the OB towardsthe IB end, then progressively worsened before leveling off at about 3.9mils pk-pk.

4. During the cool-down process, the pump shaft shifted approximately 2.5mils at the IB end while it shifted less than 1.0 mil at the OB end. Thegearbox HS shaft shifted approximately 0.5 mils at the DE.

5. Overall vibration throughout the pump train was less than 0.14 in/s rms.

Preliminary Recommendations

1. MSI recommended performing a Total Indicator Run-Out (TIR) map ofthe pump rotor mounted on V-blocks at room temperature. This shouldbe repeated after heating the entire pump rotor between bearings toconfirm the presence, location, and severity of any kinks in the shaft.

2. Conduct Finite Element Analysis (FEA) to analyze the thermal behavior ofthe pump casing and rotor at different temperature gradients measuredwith the thermocouples.

June 2012 Vibration DataAlpha, Bravo & Charlie FWP

Final Follow-Up Testing

Modifications (EPU Project)• New rotating elements on all three pumps.

• Improved inlet geometry for the 1st stage impeller (smoother suction stream passage or flow path towards the first stage impeller) like 2FWS-P1A pump.

• Balance drum re-dimensioned to provide unidirectional thrust load at any flow rate (rotor at times had been under compression towards the outboard (OB) end).

• Higher thrust load capacity (22,000 lbf) Kingsbury bearing with leading edge grooves (LEG) and RTDs.

• Thicker inboard (IB) suction cover due to resonance at 5x rpm and due to suspected acoustic resonances at 5x and 10x in the suction pipe (concluded by others).

Modifications (EPU Project)• Min-flow valves FV2B and FV2C were also replaced (like FV2A in 2010)

• 360 drip-pockets like 2FWS-P1A pump in 2010.

• Tri-land bearings like 2FWS-P1A pump in 2011.

• New gear sets on all three pumps to increase the output speed from 3,339 rpm to 3,400 rpm along with a new mechanical lube-oil pump.

• 11 GPM seal injection replaced by an API Plan 23 (thermo-syphon closed loop with a pumping ring). This was causing deformation of the inboard end cover of the pump.

Modifications (EPU Project)• Previous seal injection design allowed accumulation of condensate seal

water at the bottom of the pump casing with lower temperature than the reactor feed water. This was causing distortion of the inboard cover plate and therefore internal misalignment at the IBB.

2FWS-P1A Gearbox Low-Speed Shaft - NDE

44

Low and acceptable vibration amplitude of the gearbox LS shaft

2FWS-P1A Pump Shaft Vibration - Inboard

45

Low and acceptable vibration amplitude of the pump shaft

2.3 mils pk-pk

2FWS-P1A Pump Shaft Vibration - Outboard

46

Low and acceptable vibration amplitude of the pump shaft

2FWS-P1A Pump Inboard Bearing Vibration

47

Acceptable overall vibration

Vibration amplitude change

0.29 in/s rms

0.20 in/s rms

2FWS-P1A Pump Outboard Bearing Vibration

48

Acceptable overall vibration

Vibration amplitude change

0.30 in/s rms

0.20 in/s rms

2FWS-P1A Shaft Orbits @1am 6-12-2013

49

Pump Inboard Bearing GB HS Driven Bearing GB LS Driven Bearing Motor Driven Bearing

Pump Outboard Bearing GB HS Blind Bearing GB LS Blind Bearing

Acceptable orbit sizes

Final Conclusions1. Modifications implemented on these pumps during the outage (i.e. seal

injection API plan 23) eliminated the thermal distortion effect that wascausing deformation of the inboard end cover of the pump.

2. Previous seal injection design allowed accumulation of condensate sealwater at the bottom of the pump casing with lower temperature thanthe reactor feed water. This was causing distortion of the inboard coverplate and therefore internal misalignment at the IBB.

3. The shaft vibration at the IBB used to be approximately 4.5 mils pk-pk,while the vibration at the OBB was approximately 1.0 mil pk-pk.Vibration data gathered after the outage indicated significantimprovement at the IBB (3x reduction).

4. Radial proximity probe GAP voltage readings were not changing overtime during the warm-up or cool-down processes, versus such changesthat had been identified in previous tests.

Final Recommendations1. Pre-installation FEA analysis (structural and thermal) would have been

able to reduce risk and can spot problems before they occur(PREVENTION).

2. ODS / EMA / Continuous Monitoring Tests coupled with appropriateanalysis is a powerful troubleshooting tool to identify and visuallyunderstand the most difficult vibration problems in turbomachinery andpumping systems.