Technical Report - Heavy Duty Gas Turbine

Technical Report

Tiger™ Diagnostics Analysis

GE Frame 7 Heavy Duty Gas Turbine

Notes on Tiger data between

6^th-17^th November 2000

Revision: 1

The contents of this report are confidential to Intelligent Applications Ltd and Sermatech

Jon Aylett and Robert Milne

Intelligent Applications Ltd

1 Michaelson Square, Livingston, West Lothian, Scotland, UK EH54 7DP

5^th January 2001

Tiger is a Trademark of Intelligent Applications Ltd

Introduction

This report summarises the occurrences of key problems seen by Tiger™ in the above period for the GE Frame 7 gas turbine. Some of these problems occur many times. In this document, we typically note the first few occurrences, but not all occurrences. For incidents that follow, some are merely indicated that they occurred, while others have been further diagnosed. All together, they convey a picture of the main activities on the turbine indicative of a range of problems. This illustrates the wide range of areas which are covered by Tiger’s fault detection and diagnosis modules. This unique capability also illustrates the value of Tiger as demonstrated here by the long list of problems which were detected and documented on a ‘healthy’ gas turbine. Each incident is uniquely numbered with an incident number (MBnn) within this file. We have also included some feedback comments from site engineers on some incidents, where appropriate.

Tiger was installed on the GE Frame 7 gas turbine, on the 9^th November 2000, although some data was collected earlier on a version running on a laptop computer on the 6^th November 2000. Note, when something is reported at ‘too high’ or ‘too low’, it is relative to a normal (average) set of operating ranges across GE turbines which Tiger has been used to monitor.

Tiger identified 26 key faults, problems or critical incidents which were then documented and presented in this report.  This includes:

Although the plant knew they had high exhaust temperature spread, they were not sure why. Tiger documented the effect of transferring to the Dry Low NOx  (DLN1) system on this and also documented a vibration problem which occurs when the turbine is switched to DLN.

Tiger identified major oscillations of the shaft, lasting about 1-hour after start up. This provided additional detail on a high vibration level. The plant used the information in Tiger extensively while they worked to resolve this problem.

Tiger identified an inlet air filter problem, which was then rectified.

Tiger identified and documented several faulty sensors. Many of these failures were intermittent, so the site had been unable to document them previously.

Tiger identified and documented several controller faults and problems, including a previously unknown, but potentially serious problem.

Tiger identified and documented various short spikes and disturbances. These include electrical grid problems, power output problems and short disturbances in the bearing vibration. These were previously unknown, but affect the wear and tear on the turbine.

The following table provides an overview of the incidents documented in this report from the first week V was installed.

           

INCIDENT REF.	DESCRIPTION
Mb4	Controller fault, could lead to unnecessary trip.
Mb5	Cooling fans running at the wrong time, digital input was reversed.
Mb6 Mb23 Mb25	Catalog the significant and frequent grid voltage and frequency changes and the effect on the gas turbine. These incidents let tiger check that everything is working in the gas turbine by how it reacts.
Mb8	High vibration levels on the generator. Bearing lubrication sensor not working.
Mb11	Intermittent faulty bearing overall vibration sensor. Tiger shows precisely when it works and when it fails.
Mb12	Faulty wheelspace thermocouple.  Replay of data shows it gets hotter, burn up and then fail.
Mb13	Bearing metal temperature sensor fault.
Mb14	Sensor failure on thrust bearing sensor.
Mb15	One unusually high exhaust thermocouple relative to neighbours.
Mb16	A second high exhaust thermocouple relative to neighbours, leading to high spread and unhealthy temperature differentials.
Mb21	Tiger detects unusual and rapid change in power factor.
Mb22	Tiger detects periodic oscillations of the VARS. Also an unusual spike in phase 1 and 2 collector end delta temperature.
Mb23	Unusual drop in grid voltage and frequency and oscillations in the VARS at the same time.
Mb25	Unusual oscillation in the power output (5 MW swing).
Mb38	Tiger monitors lube oil header temperature going too high. Faulty temperature control valve.
Mb40	Tiger detects unusual and significant transient on BB3 during the transfer to premix from lean-lean burning. Shows that the timing of the vibration increase is the same as the time of the burning transfer.
Mb41	Tiger showed that the two high exhaust temperature differentials occur when the fuel changes to lean- lean and documented the evolution of this.
Mb44	Tiger detects and documents divergence of compressor discharge temperatures.
Mb45	Tiger documents the effect of the external chiller and the large inlet temperature differential it causes.
Mb47	Mark V gives high inlet air differential. Tiger helps to show it is a problem with the inlet air filters.
Mb49	Documents how the turbine trips on reverse power during a run down.
Mb51	Tiger used to investigate exhaust cooling pressure low alarm, shown to be natural effect of a trip, so unnecessary alarm.
Mb52	Tiger detects 1-second jump in generator stator coupling end temperature.
Mb63	Tiger detects large oscillations of the shaft as seen by both the overall vibration and the Bentley Nevada sensors. This lasts for 45 minutes after start up. Tiger shows the vibrations are severe enough to be significant.
Mb64	Tiger detects and documents an unusual 5-second increase in Bentley Nevada radial position 8.
Mb65	Tiger detects that an unusual disturbance of the fuel valves and power output is taking place every day at 1pm. This was tracked down to a switch over of the gas supply every day at this time.

Note, European date formats are used, where applicable (dd/mm/yy).

Incident MB 4

There are many dropouts of the digital signal L3VOTE-T, which indicates that the T processor is not participating in the voting of the three triple redundant controllers. If the other two processors drop out at the same time, the turbine will trip.

There are also dropouts of other, seemingly unrelated channels, such as CTBH3.

Key List

DWATT	Generator load watts
L3VOTE_Q	Q is participating in the voting
L3VOTE_R	R is participating in the voting
L3VOTE_S	S is participating in the voting
L3VOTE_T	T is participating in the voting
TCQA_STAT_T	T's tcqa is on line. Data valid status

The above 6-day trend graph shows clearly how often L3VOTE-3 drops out. In contrast there are only a few changes to 0 of the Q, R, and S voting. The TCQA_STAT_T also shows the problem.

The Dwatt signal in the top window helps to show when the turbine is running.

The query results for the week of 9^th-16^th November 2000 shows how often L3VOTE-3 is dropping out compared to Q, R and S. 288 times compared to 7 times!

 

Key List

DWATT	Generator load watts
BTTQT_1	Bearing metal temp-torque conv radial brg
BTTQT_2	Bearing metal temp-torque conv radial brg
TGSASPRD	Stator collector end phase to phase max delta temp
L3VOTE_T	T is participating in the voting

The above 6-day trend graph shows a similar pattern on channels BTTQT-2 and TGSASPRD.

Key List

DWATT	Generator load watts
L3VOTE_T	T is participating in the voting
TCQA_STAT_T	T's tcqa is on line. Data valid status
CTBH3	Inlet air temperature thermocouple
BTTQT_2	Bearing metal temp-torque conv radial brg

The above 1 hour graph of once per second data shows a typical pattern of dropouts at the once per second resolution.

Key List

DWATT	Generator load watts
CTBH1	Inlet air temperature thermocouple
CTBH2	Inlet air temperature thermocouple
CTBH3	Inlet air temperature thermocouple
L3VOTE_T	T is participating in the voting
TCQA_STAT_T	T's tcqa is on line. Data valid status

Finally, the above 6-day trend graph shows the problem on channel CTBH3 compared to CTBH1 and CTBH2.

Action:

Intermittent turbine controller fault. Could potentially cause a trip at some point, or even total controller failure. Requires fix to controller card, or new card.

Status:

Fault – medium warning.
Incident MB 5

06/11/00 08:17:56

Fx149 - Cooling water fans failed to respond to signal.

L52FC follows L4FC1. L52FC indicates that the cooling water fan is running, but L4FC1 indicates that Fan 1 is off. There is no signal for fan2.

It was determined that the digital was backwards, making the fans run. The plant knew about this before and fixed it about 10:40 am on 16^th November (see L63WCL).

Key List

DWATT	Generator load watts
TNH	Turbine speed hp
L4FC1	Cooling water fan no. 1 control signal
L52FC	Cooling water fans running
L63WCL	Turbine water cooling pressure low signal
TTXM	Ex temp median corrected by average

The above 6-day trend graph shows the pattern of the turbine stopping and running. Note that L63WCL follows exactly L53FC, when it should be the opposite.

 

Key List

DWATT	Generator load watts
TNH	Turbine speed hp
L4FC1	Cooling water fan no. 1 control signal
L52FC	Cooling water fans running
L63WCL	Turbine water cooling pressure low signal
TTXM	Ex temp median corrected by average

The trend graph of 2 days shows how L63WCL was the same as L53FC, until the fault was corrected at 10:40 am on the 16^th. After that the signal is the opposite.

Action:

None. Now fixed.

Status:

Now OK.

Incidents MB6 and MB23 and MB25

There are frequent reductions in the local grid frequency and voltage. These are external events, but do affect the gas turbine and require it to respond quickly. Each of these events causes more wear and tear on the gas turbine. They are noted here to help catalog these wear and tear causing incidents.

Key List

DWATT	Generator load watts
SVL	System line voltage
DV	Generator line volts
DF	Generator driven frequency
DVAR	Generator load vars scaled
SFL2	Bus_pt freq

The above 6-day trend graph shows the events on the grid with the local and grid voltages and the local and grid frequencies.

Tiger’s diagnostics capture these events when they are above a significant threshold. The above query shows how often these events have occurred in one week. The exact day and time for each can also be obtained from the Diagnostics Query. Note also that Tiger has detected 3 occurrences of reverse power.

Key List

DWATT	Generator load watts
FSG	Gas control valve lvdt position
FSGR	Speed ratio valve calibr pos
SVL	System line voltage
DV	Generator line volts
SVL	System line voltage
DF	Generator driven frequency

The above graph of 2 minutes of once per second data shows a typical incident which affects the local and grid voltage and frequency. Note that it does not seem to affect the gas turbine power level or cause the fuel valves to react in this case.

 

Action:

Abnormal intermittent grid transients. Could potentially cause more wear to the turbine. Should investigate if a problem or not.

Status:

Warning/query.

Incident MB8

The Tiger fault detection mechanisms frequently report that overall vibration levels on the generator, as measured at BB11 and BB12 are high.

The above snapshot of Tiger’s vibration screen shows these two channels at high levels. The fact that BB4 is not working is also visible. (reading is –8.00). See incident MB11 for more details on this.

Note that transducer BTGJ2 is now working, it was not working when Tiger was first installed. (see incident MB13).

Action:

Abnormal high vibration on the generator bearings. Could potentially cause damage, and/or a turbine. This is currently under investigation (see notes below).

Status:

Fault – high warning.

Site Notes:

There was an increasing vibration trend on bearing #5 at this time. Bearing #4 increased as well, but to a lesser extent. When the weather turned cool, found that the lube oil temp control valve didn't respond very well and the lube oil temp dropped to 125-130. This probably resulted oil whirl (oil whip?) in #4 bearing, causing the vibration levels to make a step change upward. Oil temp was adjusted back up to 140 or so, but #5 bearing continued to trend upward slowly, reaching .5 in per second and remaining at that level.

It appears that there is a 1/2x vibration orbiting in reverse rotation that doesn't change with load. That would appear to be mis-alignment except for the fact that the machine was aligned last spring. A balance shot will be carried out on 11/12/00, to attempt to improve the situation. As the shot will be a 1x imbalance, it is not certain how it will affect the 1/2x component.

Note, this could also account for the bearing vibration oscillations detected by Tiger (see MB 63).

Incident MB11

The bearing overall vibration transducer, BB4 provides a faulty reading (-8.00) on an intermittent basis. Normally it stops working when the turbine starts up and then starts working again some time after the turbine has stopped.

 

Key List

DWATT	Generator load watts
TNH	Turbine speed hp
BB3	Turbine vibration transducer # 3
BB4	Turbine vibration transducer # 4
BB5	Turbine vibration transducer # 5
L14HC	Auxiliary cranking speed relay

The above 6-day trend graph shows the overall pattern of when this transducer is providing a reading of –8.00 and when it is not. For comparison, other vibration transducers are shown. The final channel on the graph is to illustrate when the turbine start-ups take place.

Key List

BB1	Turbine vibration transducer # 1
BB2	Turbine vibration transducer # 2
BB3	Turbine vibration transducer # 3
BB4	Turbine vibration transducer # 4
BB10	Generator vibration transducer # 10
BB12	Generator vibration transducer # 12

The above graph shows 4 minutes of once per second data during a start-up. Note that BB4 has a ‘valid’ reading, then it drops out, recovers, then drops out again.

This illustrates the intermittent nature of this fault.

The above Diagnostics Query provides a summary of when Tiger has detected that BB4 has stopped providing a valid value.  This rule is triggered whenever the values of BB4 and B5 (which are on the same bearing) diverge. This provides a summary of the exact date and time of each drop out in the period of 9^th-16^th November 2000.

Action:

Faulty bearing transducers can sometimes generate very high spurious transient values. This could cause a turbine trip. Turbine fault requires repair.

Status:

Fault – high warning.

Incident MB12

06/11/00 08:17:56

Tiger detects excessive wheelspace temperature on second stage outer thermocouple no. 2 (TTWS2AO2).

This could be thermocouple or seal fault, but looks more like a thermocouple problem. This results in MKV alarm messages, as the temperature reaches the alarm limit.

The following screen indicates that Tiger has detected this problem.

Key List

TTWS1AO1	Turb wheelspace temp 1st stg aft outer
TTWS1AO2	Turb wheelspace temp 1st stg aft outer
TTWS2AO1	Turb wheelspace temp 2nd stg aft outer
TTWS2AO2	Turb wheelspace temp 2nd stg aft outer
TTWS3AO1	Turb wheelspace temp 3rd stg aft outer
TTWS3AO2	Turb wheelspace temp 3rd stg aft outer
TNH	Turbine speed hp
DWATT	Generator load watts
FSG	Gas control valve lvdt position

The graph above indicates the excessive level of the wheelspace thermocouple, and shows some of the other thermocouples for comparison.

The second message in the list shown above, indicates that Tiger is warning of the thermocouple problem

The fifth diagnostic message in the list shown above indicates that a MkV alarm relating to this thermocouple problem has also been detected.

Action:

Thermocouple could reach trip limit and cause a turbine trip. Turbine fault, requires repair.

Status:

Fault – high warning.

Incidents MB13, MB14, MB15 and MB16

All of these incidents were detected in the same data sequence, and the Tiger screens relating to these are shown together after MB16.

MB13

06/11/00 08:17:56

Tiger detects bearing metal temperature generator journal #2 sensor fault.

Looks broken.

Action:

Turbine fault, requires repair.

Status.

Fault – low warning.

MB14

6/11/00 08:17:56

Tiger detects sensor failure on bearing metal thermocouple #1 - thrust active (BTTA1-1).

Looks broken.

Action:

Turbine fault, requires repair.

Status:

Fault – low warning.

MB15

06/11/00 08:17:56

Exhaust temperature thermocouple 1 (TTXD1-1) is detected high by Tiger w.r.t adjacent thermocouples 2 and 18.

Possible thermocouple or combustor problem.

Action:

Check for possible fault developing.

Status:

Warning – possible fault developing.

MB16

06/11/00 08:17:56

Exhaust temperature thermocouple 10 (TTXD1-10) is detected high by Tiger w.r.t adjacent thermocouples 9 and 11.

Possible thermocouple or combustor problem.

Action:

Check for possible fault developing.

Status:

Warning – possible fault developing.

Key List

BTTA1_1	Brg metal thermocouple #1 - thrust active
BTGJ2	Bearing metal temp. Generator journal #2
TTWS2AO1	Turb wheelspace temp 2nd stg aft outer
TTWS2AO2	Turb wheelspace temp 2nd stg aft outer
TTWS3AO1	Turb wheelspace temp 3rd stg aft outer
TTWS3AO2	Turb wheelspace temp 3rd stg aft outer
TNH	Turbine speed hp
DWATT	Generator load watts
FSG	Gas control valve lvdt position

The previous two screens indicate the problems with BTGJ2 (MB13), BTTA1-1 (MB14) and TTXD1-1 (MB14) and TTXD1-10 (MB15).

The second and third message lines above indicate the problems with BTGJ2 (MB13), BTTA1-1 (MB14).

Key List

BTGJ2	Bearing metal temp. Generator journal #2
TTWS2AO1	Turb wheelspace temp 2nd stg aft outer
TTWS2AO2	Turb wheelspace temp 2nd stg aft outer
BTTA1_1	Brg metal thermocouple #1 - thrust active
BTTA1_2	Brg metal thermocouple #2 - thrust active
TNH	Turbine speed hp
DWATT	Generator load watts
FSG	Gas control valve lvdt position

The previous screen indicate the problems with BTGJ2 (MB13), BTTA1-1 (MB14)

The next two screens indicate the problems with TTXD1-1 and TTXD1-10 (MB14 & MB 15).

Key List

TTXD1_1	Exhaust temperature thermocouple 1
TTXD1_2	Exhaust temperature thermocouple 2
TTXD1_18	Exhaust temperature thermocouple 18
TTXD1_10	Exhaust temperature thermocouple 10
TTXD1_11	Exhaust temperature thermocouple 11
TTXD1_9	Exhaust temperature thermocouple 9
TTXSP1	Combustion monitor actual spread 1
TTXSP2	Combustion monitor actual spread 2
TTXSPL	Combustion monitor allowable spread
TTXSP3	Combustion monitor spread 3
TNH	Turbine speed hp
DWATT	Generator load watts
FSG	Gas control valve lvdt position

The graph above indicates the problems with TTXD1-1 and TTXD1-10 (MB14 & MB 15).

The third and fourth message in the message list above indicates the problems with TTXD1-1 and TTXD1-10 (MB14 & MB15).

Incident MB18

06/11/00 08:17:56

Tiger detects following MKV alarms set:

Explanation:

L20CBZ	Auxiliary signal to l20cbx (time delay)
L30BN_F	Bentley Nevada fault
L30BN_F_ALM	Bentley Nevada fault
L30DIAG_C	Diagnostic alarm c-q
L30WSA1	Wheelspace temp differential high
L39VF_4	Vibration fault - transducer input #4 - caused by BB4 failure
L39VS_A	Bentley Nevada vibration alert
L39VS_A_ALM	Bentley Nevada vibration level high
L4Z	Starting device trip
L5051CRA_ALM	Starting motor undercurrent
L60SP6	Exhaust temperature spread - adjacent tc's -Properly due to TTXD problems
L63CA3L_ALM	Gen air filter cleaner pressure low
L63HQ1LY	T.d. hydraulic supply press low
L63WCL	Turbine water cooling pressure low signal
L86CH2_ALM	Inlet chiller trouble - chiller trip

Action:

These require checking to confirm as alarms, or if tags need removing from modbus or alarm list.

Status:

Warning/query.

Incident MB21

06/11/00 08:17:56, 11:46:53- 11:50:53

Tiger detects that generator power factor (DPF) is outside of normal range.

             

Key List

DVAR	Generator load vars scaled
DPF	Generator power factor
DV	Generator line volts
SVL	System line voltage
TNH	Turbine speed hp
DWATT	Generator load watts
SFL2	Bus_pt freq
DF	Generator driven frequency

Tags DVAR & DPF show some unexplained oscillations and changes. The Level of DVAR and DPF is unusual, but due to the characteristics of the local electrical grid. However, there are some unexplained grid disturbances occasionally, which should be checked (also see incidents MB6 and MB25).

The first message in the message list above indicates that Tiger has detected abnormal levels for DVAR.

Action:

Check reason for grid disturbances.

Status:

Warning/query

Incident MB22

06/11/00 08:29:34 – 08:31:59, 09:18:20. 16/11/00 09:30:35.

Key List

DVAR	Generator load vars scaled
DPF	Generator power factor
TGSDIF3	Phase 3 collector end to coupling end delta t
TGSDIF1	Phase 1 collector end to coupling end delta t
TGSDIF2	Phase 2 collector end to coupling end delta t
SFL2	Bus_pt freq

The graph above indicates some unusual oscillations occur in phase 1 and 2 collector end to coupling end delta temperatures (TGSDIF1, TGSDIF2), and oscillations in DVAR. Maybe ok, but could be symptomatic of a problem.

Action:

Check if this is ok.

Status:

Warning/query.

Incident MB23

06/11/00 11:20:49

Key List

DVAR	Generator load vars scaled
DPF	Generator power factor
TNH	Turbine speed hp
DWATT		Generator load watts
DV		Generator line volts
SVL		System line voltage
SFL2		Bus_pt freq
DF		Generator driven frequency

Tiger detects unusual drop in generator voltage (DV) and grid frequency (SFL2). May be ok, but could be symptomatic of a problem. There is also an oscillation in DVAR at this time. (Also see  MB6 & MB25).

The first message listed above indicates that DVAR is at an abnormal level. The last message indicates a problem with the grid frequency.

The above message indicates a possible abnormal transient has occurred in the grid frequency.

Action:

Check if this is ok.

Status:

Warning/query.

Incident MB25

09/11/00 11:30:14

Key List

DVAR	Generator load vars scaled
DPF	Generator power factor
FPRGOUT	Gas ratio valve servo command
FPG2	Gas fuel inner valve pressure
DWATT	Generator load watts
DV	Generator line volts
SVL	System line voltage
SFL2	Bus_pt freq
DF	Generator driven frequency

The graph above indicates a grid disturbance has occurred. Tiger detects an unexplained voltage spike (DV, SVL). (Also see MB6& MB23). May be ok, but could be symptomatic of a problem.

Action:

Check if this is ok.

Status:

Warning/query.

Incident MB26

09/11/00 11:32:06

Tiger detects unexplained power spike – 76.5 -> 80 Mw. May be ok, but could be symptomatic of a problem.

The message above indicates Tiger has detected a power disturbance.

Action:

Check if this is ok.

Status:

Warning/query.

Incident MB38

12/11/00 07:19:52