TB5000 Useful Information Part 7

15.0 Rotor Blade Campbell Diagrams
 
Campbell diagrams are available for the first stage compressor rotor blade and for each of the four turbine stages. These follow on the next 5 pages.

TB5000 Useful Information Part 6

11.3 Gas generator ½X vibration

This can be a particularly troublesome and once again is linked to the critical speed at circa 4600 cpm. The problem shows itself as the gas generator approaches the IGV change over speed of about 9200 rpm and, although the speed falls a little and then rises to over 10000 rpm, the vibration usually locks to the speed as the load increases. Almost without fail it is an exact ½X component and if displacement probes are fitted it will show a forward rotation. It usually means there is a problem with the fit of the bearings giving the classical signs of looseness. It is mainly driven by the ct bearing but in some instances the inlet can also influence matters.
As well as non conformance of the housing diameters things to look out for are:-
• Lack of chamfer on the housing split line.
• Bearing location tang fouling on the housing.
• Incorrect nip; it should be about .004 ins. on the ct. It is a little less for the inlet bearing but it is very difficult to measure and get sensible figures.

• Correct bedding of the bearings within the housings.
• Housing discontinuity across the split line.
• Ovality of the bearing housings especially for the ct end.

Having identified how to address the 1/2X component it has to be said that in the vast majority of turbines the hardware does not experience any distress. It is known that many units without displacement probes do experience the problem but without the instrumentation the problem goes unnoticed. To the writers knowledge this issue has never led to components failures. This comment can also apply the power turbine vibration through whip / whirl.

11.4 CT2 Rotor Blade Fatigue Failures

There have been about 8 incidences of this problem in recent times but it is not seen as a generic problem because they have nearly all failed on turbines operated by one user, namely Conoco Phillips in Alaska. In all these cases CP themselves have taken the standard CT2 blades and applied various metallurgical treatments in an attempt to increase the service life. In principle the fatigue life of the blade should not have been affected to any significant degree.
However their experience has not backed this up and failures seem to occur at regular intervals. For some reason CP have not been deterred by the problems and to date are continuing to modify the standard blades supplied by Siemens.
 
There is an outside possibility that the cold ambient conditions are influencing the behaviour of the blades but at the moment this company does not believe that to be the case.

11.5 PT2 Rotor Blade Fatigue Failure

This problem is associated with mechanical drive units operating in very cold or arctic conditions. For mechanical drive units there is a 4th order interference with the fundamental blade frequency which can cause the alternating stresses to exceed the endurance limit if high powers generated in cold weather are utilised. A modified blade with a raised natural frequency is available in this situation. This increased frequency was achieved by thinning the aerofoil section. See section 15.0. TDN 89/063 provides additional background information.

12.0 Breathing

To state the obvious problems with the breather system usually means that there is more air than usual trying to pass through the various passageways there is a blockage in the system. Both of these result in high pressures;
usually within the oil tank.

Potential causes:-

• Failure of the compressor exit seal (bucket seal). This is a fairly
common problem and allows excessive amounts of high pressure air to enter the tank.
• Problems with the flame trap/coalescer – blockages etc.
• Overfilling of the oil tank which may restrict air flow.
• Blocked air holes in the drowned lube oil drain tubes which return the oil to the tank. This problem is often linked to over filling of the tank.
• Failure to fit the pt oil drain bypass pipe when 3 lobed bearings are installed.
Packages with depressed pressure in the enclosure may leak from the gearbox exit seal. In this situation a mod to pressurise to seal with air from the bleed band is available.

12.1 Nominal Breather Flows

Because of the relatively low pressures and temperatures for the TB the turbine does not have an external buffer air supply to the ct labyrinth seals. 

All flows are controlled by the internal pressures and clearances. The CT, PT and Tank breathers all exhaust through the low pressure discharge pipe via the coalescer and trap.   

 For the TB5400 the total breather flow is approximately 0.17 lb/sec For the TB5000 the total flow is marginally less at about 0.16 lb/sec. Of this, the CT accounts for about 0.07 lb/sec, the PT about 0.05 lb/sec and the remainder comes from the tank flow. These figures are very approximate

13.0 Rotor Overspeed

The question of whether to overspeed the rotating assemblies was reassessed in 2001 by Andrew Shepherd who concluded that:-
• All overhauled TB5000 rotors that are to be despatched without engine testing require overspeed.
• All new CT discs that are to be fitted to overhauled TB5000 rotors require overspeed. The compressor rotor does not require overspeed in this case.
• TB5000 compressor rotors that have any of the compressor shafts or discs replaced during overhaul do not require overspeed.
• TB5000 CT discs that have already operated in the field, but have had the CT blades replaced, do not require overspeed.

14.0 Performance

The four ratings are TB3000, TB4000, TB5000 and TB5200/5400. It is only with the most recent rating that the performance difference between liquid and gas fuel was formally recognised by quoting 5200 hp and 5400 hp zero loss respectively at the temperature limit. For the current range of turbines from SIT Lincoln the output power is a fixed quantity and the temperature to achieve this is set and limited during the works acceptance test. On the other hand, with the TB, the operating temperature is fixed and therefore the generated power can vary as a result of component efficiency variations. A performance degradation of 4% is allowed during the works acceptance test of overhauled turbines, although this is often not achieved and a concession may be needed to pass off the unit.

14.1 Performance curves
 
The performance group is the guardian of the performance graphs and it is impractical include all the available charts in this document in this document.
The short list below, particularly the 913 and 995 series should enable most simple gas related performance issues to be answered without problem. It is probably better to consult with the performance group if more specific information is required.
 
Chart Number                     Title
IC 913                                  series TB4900 Checking graphs
IC 995                                 Series TB5200/5400 Checking graphs
IC 857/7                               Compressor characteristic
IC 948                                    CT characteristic
IC 948/1                                    PT characteristic
IC 980/1t                               Test bay rating chart
IC 994/1                               Maximum operating temperature
IC 975                                  Nominal acceptance limits
Note that all generic performance graphs for the TB are prefixed by ‘IC’.
 
The major control parameter for the TB is the operating temperature Top, which is the difference between the exhaust stack temperature minus the compressor inlet temperature. In principle a constant Top value equates to a constant turbine inlet temperature although this is not strictly true at all ambient conditions. (see TDR 07/052)
 
Graph IC 975 shows the nominal temperature set points for the TB and includes a maximum continuous running temperature. However if proper provisions are made within the control system it is permissible to raise this maximum continuous value up to the warning set point to achieve extra power. There is a penalty in that hot component life is consumed at 04 times the normal rate.
 
14.2 Power shortfall

When new, most TB turbines easily generated the required power within the temperature limits but old or overhauled engines often struggle to achieve the rating. If there is a significant loss the most likely causes are:-

• Leakage past the ct bearing housing seal which abuts with the
upstream inner flanges of the quadrants. There is quite a high
differential pressure across this barrier and a gap of only a few
thousandths of an inch can have a significant affect. The actual air
leakage is less significant than the disruption it causes to the turbine
flow when it re-enters the gas stream.

• The piston ring seals on the downstream out flanges of the quadrant are another favoured leak path with similar consequences
• In general, the turbine section blades will be in good condition for a test, as it will have been critical to the integrity of the turbine to replace worn or damaged items. However the compressor blades are another matter.

These can become worn, corroded and damaged in so many ways that the compressor efficiency is compromised. Because of the diffusion activity within the stator the blades here can have a big influence on matters. In the past the writer has advised that the stator blades be replaced. After which the turbine performance was recovered.
 
14.3 Cold start performance
 
Due to changes in the clearances as the turbine heats up following a cold start there is a loss of almost 250 kw when the engine running at the nominal maximum operating temperature. Most of this fall off takes place during the first 5 minutes or so but degradation continues for perhaps 25 minutes until equilibrium is achieved.
14.4 Inlet and exhaust losses
 
These losses can significantly affect the generated power but the limits are really set by mechanical or flow problems within the turbine and in the extreme the pressure capabilities of the ducting.
The maximum inlet depression should be less than 15 inches water although levels up to 20 inches could just about be accommodated.
higher than this may have an adverse affect on cooling flows within the turbine.

TB5000 Useful Information Part 5

11.0 Vibration and Balance
 
Gas gen critical speeds: - 1st 4600 cpm - rigid body mode
                               2nd 5950 cpm - rigid body mode
                                         3rd 20860 cpm - bending mode
Gas generator balance limits are specified on drawing CT5020 and are specified as 9 gm inches at the CT journal and 3 or 7 gm inches depending on phase angles at the inlet journal.

PT critical speeds: - 1st 4065 cpm - rigid body mode
                                  2nd 13760 cpm – bending mode 

The first gas generator critical speed at 4600 cpm is only one which has the potential have any real influence on the behaviour of the turbine. See paragraph 11.3.
 
Traditionally, the TB vibration limits applied solely to casing levels at prescribed points on the turbine and package. During factory engine testing the limit is 9 mm/s rms. In practice levels in excess of 7 mm/sec would be considered a bit on the high side and at least investigated to understand the reasons, even if nothing was done about it. Alarm and shut down are usually set to 10 and 13 mm/sec rms respectively.

Sometimes customers request acceleration set points but it is not really possible to give a single value as the relevance depends on the frequency of the vibration. In this case it is better to make a record of a benchmark spectrum and look for differences if the all pass level increases for any reason. It should then be possible to identify any offending frequency and take the appropriate corrective action.

Displacement vibration measurements are an option with the TB and the test acceptance limit at full load is now 50 μm pk to pk all pass; and 19 μm pk to pk for discrete non synchronous components. At all other loads there is a simple limit of 65 μm with no restriction on the frequency composition.

Installation warning and shut down levels are normally set to 65 and 90 μm pk to pk respectively.
The two types of vibration monitoring require different bearing types throughout the core turbine.

With displacement monitoring three lobed bearings should be used because they provide better control of rotor dynamics. It is very important that the displacement vibration protection equipment is fitted and operational when the 3 lobed bearings are installed. Failure to activate this equipment can lead to bearing failure in some situations; the two elements are mutually dependent upon each other.

For casing velocity vibration monitoring, offset half bearing shells are fitted, but it has to be accepted that the dynamic behaviour of the turbine may not be as good. There are several common vibration situations for the TB, either in form or frequency. These are:-

11.1 Rocking gearbox

Sometimes the gearbox and its supporting pedestal can be affected by a natural frequency vibration which occurs at somewhere between 1500 and 1800 cpm. Unfortunately, this coincides with the operating speed of an alternator driver and can lead to high levels of low speed synchronous vibration. It usually manifests itself as a side to side rocking of the gearbox on the pedestal. Only a few units become susceptible to this vibration but because it is driven by a natural resonance it is often difficult to deal with. It is essential that the drive train is in perfect balance to minimise the excitation and in some cases it has been necessary to provide added stiffness by welding extra support members into the base plate. Mechanical drive units are not affected by this problem.
 
11.2 Power turbine whip/whirl.
 
Although it is not a common occurrence the power turbine inlet bearing can experience a non synchronous vibration at about 3300 cpm. Again this is associated with a resonance and is triggered by dimensional discrepancies with the bearing / housing. It usually only shows itself with 3500 rpm mechanical drive gearboxes via displacement probes although in extreme cases it can be picked up with seismic measurement with offset half bearings.

TB5000 Useful Information Part 4

8.0 Combustion

There are two types of combustion arrangement for the TB. The original version requires that the gaps between the various elements (swirler, flame tube, extension and star ring) are carefully controlled to ensure the temperature spread and distribution around the annulus is within limits.

Temperature distribution within a can is limited to 80 Celsius and the maximum allowable difference between the can averages is 20 Celsius.
 
For flame monitoring the TB has 16 equally spaced thermocouples within the exhaust diffuser, starting with #1 at 12:30 and finishing with #16 at circa 11:30 when looking downstream. Because of the swirl within the gas path t/cs 16 to 3 monitor combustor 1; 4 to 7 look at combustor 2 etc.

Failure to achieve the required temperature spread can mean :-

• There is an offset or gap errors at the interface between the various components of the combustor.

• Liquid burners are partially blocked or coated with carbon which can cause flame distortion.

• Gas burners need balancing

If the problem is not linked to the above elements of the combustors can be exchanged 1and 4 and or 2 and 3. It is not usually possible to exchange components side to side because they are ‘handed’.

With the sealed assembly there is no option to adjust the fit of components because there is a tight sliding interface between the elements. If the burners themselves are ok then the only options are swap parts around. Again this is generally limited to left / right changes.

Steam and water injection modules are available for both power and emissions control, but there is no DLE option for the TB turbine range.

Skid edge gas pressure should be in the range 180 – 200 psi.

8.1 Ignition.

Combustion light up for a TB is by small pilot flame torches and spark devices situated in each can. These torches are often fuelled with bottled propane gas but if suitable main fuel gas can be used. If a 50kg. propane bottle is used it is expected that at least 25 full starts could be achieved although in reality is probably nearer the mark.
The TB does not utilise the high energy ignition system available for the current turbine range from Lincoln.
 
9.0 Gearbox
 
The gearbox casing is an integral part of the structure of the TB engine and cannot be omitted. It is capable of accepting several gear ration options which at full speed are: - two epicyclic designs at 1500 and 1800 rpm for alternator drive, a star design for 3550 rpm pump duty and two parallel shaft designs at 11720 rpm and 13500 rpm for compressor duty. There is also the option for a straight through direct drive shaft without gears.
The epicyclic and straight through designs have in-line drive and rotation in the same direction as the turbine rotor; i.e. anticlockwise when viewed upstream.

The main engine driven lubricating oil pump and the mechanical overspeed trip are both driven from a power take off wheel on the gear assembly.
 
10.0 Lubricating Oil

The usual oil grade for the TB is VG46 which when operating at a
temperature of 60/65 Celsius will give acceptable behaviour. It is also permissible to use VG32 and VG68 in cold and hot environments. The very minimum oil temperature at start up must be 10 Celsius although most units are set to 15 Celsius. 


Originally three oil pump sizes were specified for the TB and the chosen one depended on the retirements of the gearbox and the driven unit (90 gpm, 120 gpm and 180 gpm). The smallest of these is no longer recommended and I believe almost all of them have been changed to the 120 gpm version. In most cases, the AC and engine driven pumps are identical.

The normal oil pressure for the TB is 50 psi with a low pressure shut down of 35 psi.

The emergency dc pump only supplies oil to the two hot bearings, namely the CT and PT shaft disc end locations. This supply is for cooling purposes only after a shut down if the AC system fails. It does not and cannot be used with an operating turbine.

With three lobed bearings the pressure from the dc pump may be as low as 15 psi. This is adequate for the provision of cooling flow.
Typical oil usage is in the range 3 – 5 gallons per week. Many turbines will be better than this, some not so good

TB5000 Useful Information Part 3

5.3 First Stage Rotor Blades.

Apart from the 3000 variant it is not recommended that TBs operate continuously at low loads, typically less than about 600 - 700 kw. The reason for this is that it may lead to fatigue failure of first stage compressor rotor blades just above the root platform. The problem is caused by poor incidence leading to flutter of the aerofoil. Over the years a dozen or so have occurred, although it doesn’t happen in all cases of low load operation. 

Ambient conditions and speed are the main influences although the exact trigger mechanism has never been fully identified. See section 15.0 - Blade Campbell diagrams.
 
One way round the problem is to fix the guide vanes as with the TB3000 but that will limit the available power to about 2100 kW. Alternatively the vanes could be set to +15 degrees for the start and change over to -5 degrees at the normal 1500 hp condition. This option requires changes to the control and fuel schedules. In principle this should be a sensible approach when the full power range of the turbine is required.

Both of the above options have been recommended to users. The fixed vane system has been used in the past but it is not known whether the 15 to -5 range has actually been implemented in the field.

The third stage rotor blade was redesigned with a thinner section circa 1980 to eliminate a failure inducing resonance. The current blade has a higher natural frequency and is not susceptible to the problem. It is believed that all these have been changed out but there may be units which are not under the control of Siemens and still contain the old design.

5.4 Third Stage Rotor Blade.
 
During the early 1980s problems were encountered with compressor surging after blade machining was brought in house and the aerofoils actually conformed to the drawing. The investigation resulted in small stagger angle changes to most stages to cure the problem. It is almost certain that all field units now have the current blade design.

6.0 CT Bearing Housing

The lower half of the housing is actually part of the centre casing with the top section being a casting. As the casing gets older it is not unusual for  distortionof the bearing bore to occur and this can have a significant influence on the performance of the bearing and the rotor vibration characteristics. The real problem centres on the fact that the bottom half is all part of the substantial lower centre casing member whereas the top cap is a small relatively flimsy component. There is a resultant mismatch in the thermal and stress characteristics between the two which contributes to the problem.

There are two versions of the bearing cap; the original design and a later variant which is thicker with increased stability. It is possible to retrofit the more substantial cap but it does involve modifying the seal assembly to accommodate the larger part.

7.0 Bearings

Both offset half and three lobed bearings are utilised on the TB with the latter only being used if displacement vibration equipment is installed. See the section on vibration.

7.1 Bearing Clearances (inches):-

                                   Offset half              Three lobed
Compressor inlet      0.002/0.004             0.002/0.004
Compressor turbine  0.0047/0.0067         0.0075/0.0095
Power turbine inlet   0.0047/0.0067         0.0047/0.0067
Power turbine exit    0.0025/0.0045          0.0025/0.0045
 
Note the difference at the ct bearing location where the three lobed clearance is particularly large and therefore greater than would normally be expected.

This is necessary because the shaft expands rapidly during start and rapid load changes.

7.2 Bearing loads and oil flow requirements
The flow figures in the table below are not theoretical values but are based on
some field measurements taken circa 1980 plus some extrapolations and
estimates made by the writer. The units are imperial gallons per minute. The
compressor inlet and PT exit locations include the thrust bearing flows.
Location           Comp inlet  CT bearing  PT disc end  PT exit end
 
Engine driven
Offset Half              9.0              3.5                4.0                4.5
 
Engine driven
3Lobe                    10.5             6.5                  7.5               5.5
 
AC 3 Lobed          5.0               2.5                  3.5                 2.5
 
DC 3 Lobed             0              1.5                   1.5                 0
 
Static bearing
loads (lbs)           620                  860                 500              -50**
 
** This static load figure is negative because the journal naturally sits at the top of the bearing because of the disc overhung moment.
 
7.3 Bearing options

Several oversize bearing shells have been defined for special fault
circumstances. These are not necessarily formally issued or available but the following list dated 12/03/01 may act as a starting point if any activity is required. This would normally within the responsibility of the repair group. Some SGT200 bearing options are included in this list.

TB5000 Useful Information Part 2

3.0 Inlet volute
 
There no real problems with this component except for the condition of the trash screen which should be monitored to check for break up of the mesh.
There are advisory service bulletins on this matter. Three locations on the volute are used for seismic vibration monitoring.
 
Unfortunately the casing can act as a vibration amplifier and
disproportionately indicate that the turbine is in distress. It is not unknown to record velocities in excess of 20 mm/sec rms. See the vibration section for more on this subject.
 
The casing carries two RTDs for inlet air temperature measurement. The outputs from these are utilised for control purposes. If required compressor washing injection nozzles or spray bars are also fitted to this component.

4.0 Inlet Bearing Housing

This component is cast from aluminium alloy but is vulnerable to distortion and oil leaks. There are numerous A forms and service bulletins that have attempted to address the problems.

The current design of the TB inlet bearing housing is vulnerable to oil leaks from the joint faces and/or the threaded fasteners. Various modifications have been introduced over the years in an attempt to deal with the problem.

5.0 Compressor

5.1 General.

For the TB5400 at full load ISO conditions, the twelve stage compressor absorbs roughly equal work per stage delivers about 21 kg/sec of air. This is achieved with a pressure ratio of just over 7:1 at a discharge temperature of
about 275 C.

In terms of its mechanical construction the rotor is a multi-disc assembly with the rotor blades located between the discs in dovetails and held together with a central tension bolt. The LP blade stages and disc air washed faces are normally coated to provide corrosion resistance.

The rotor journals are prepared to accommodate displacement vibration equipment even though perhaps only about 20% of TB turbines are so fitted.

It is not permissible to repair the CT journal using a re-chroming process because this procedure has resulted in service failures. These incidents have always occurred during significant transient events and it is believed that it is as a result of rapid changes to the temperature of the cooling air that passes through the unprotected centre of the journal.

5.2 Inlet Guide Vanes.

For the TB3000 the IGVs were fixed at +15 degrees, but for the other ratings the vanes change between +28 and -5 degrees as the load passes through about 1500 hp. These angles are calibrated by the use of a setting block which fixes distances between the mechanism and the casing. There is no modulation; it is just a flip flop mechanism that moves quickly between the two set points and is driven by P2 air pressure. 

This normally occurs at 52 psi rising and 48 psi falling. For turbines operating at high altitude where the pressures will be lower the mechanism is fitted with a weaker spring and the transition takes place at circa 44 psi. Apart from a sudden change of compressor speed the changeover happens quite smoothly.It is not unknown for surging to be a problem on dirty or old turbines.

 It normally occurs around the IGV change over point which may be the result of a set up error. Surging at other times, especially during run up, may be indicative of errors in the fuel schedule. Correct operation of the inter stage blow off valves may also cause surge problems. 

TB5000 Useful Information Part 1

1.0 Introduction

This document identifies some of the writer’s experiences with the TB and is based on operating experience, field tests and assessments. The information has not necessarily been formally approved by calculations or development activity.

It is not an exhaustive list because many of the answers provided to
questioners are not necessarily based on hard evidence. It is often a case of linking various remembered facts that may not documented anywhere but which enable a solution to be provided.

This document is part of three lever arch files of TB useful information which has been filed within the library system

Compiled By:

R W Kemp
Principal Engineer

2.0 Design Concept

The mathematical tools and material properties available during the design of the TB turbine were not as well advanced as they are today and therefore a conservative approach was made to stress and safety margins. Because of this many of the components within the turbine are quite tolerant of any deviations from the design intent.

The original design of the TB was made in the late 1960’s and it was conceived as a 3000 hp twin shaft turbine for both mechanical and electrical drive duties. In reality the first turbines built did not have an interduct and there were so many internal leak paths that only 2000 hp was produced.

 An interduct was introduced between the ct and pt and the pt1 stator was reduced. This improved the flow characteristics and after most of the leak paths were eliminated the 3000 hp was achieved. Since that time the power has been increased three times to 4000 hp, 5000 hp and finally 5400 hp.

After the initial development modifications, there has been very little change in the mechanical arrangement since its conception.
The gas generator consists of a simple 13 stage compressor with C4 profiles and a two stage compressor turbine with shrouded rotor blades. The simple reverse flow combustion chambers are external to the pressure casing giving a shortened overall length for the engine and easy maintenance of the combustors. Following a departure with the SGT100 and 200 turbines this concept was resurrected for the 300 and 400 machines.

As a twin shaft machine the gas generator speed varies with load. It operates between about 7200 rpm when idling to about 10400 rpm at full load under ISO conditions.

 Under some conditions the speed can be as high as 10900 rpm but this should be considered the maximum permissible.

The unshrouded two stage free power turbine is supported from the gear box casing which is effectively the fixed support for the turbine. All axial expansion is accommodated by the A frame supporting the compressor inlet. 

The design speed for the power turbine is 7950 rpm but it can operate as low as 4000 rpm and up to a maximum permissible continuous speed of 9000 rpm. This top speed is inclusive of any matching allowance but the overspeed trip can be set to 9900 rpm; 10% above the max continuous.

TB5000 LIGHT UP AND STARTING GUIDE.

Terms Used.

UV Flame Detectors. These detect ultraviolet radiation from a flame.

‘Inference’ Flame Detection. This is a software program to detect flame from PT Exit temperature rise.

1. IGNITION.

For ignition of the four pilot flames two three things are needed.

a.    Air supplied by the turbine gas generator.

b.    A correct flow of fuel into the igniter blocks.

c.    A good spark to ignite the igniter gas.

2.0 Air.

This is not adjustable on an electric motor start. The control system will not switch on the igniters until a gas generator speed of greater than 1800 rpm has been reached and the engine has had enough air through to be purged of any fuel gas. For gas motor starting or hydraulic, at least 1800 rpm is necessary by the time igniters are switched on (40 seconds norbit) or to start the core purge timer. If the starter reaches more than 2800 rpm before main fuel is on the governor can start ramping the throttle open causing stalling and high temperature shut down.

Pilot fuel will usually light OK.

3.0 Igniter fuel.

3.1  

Propane has been found to be the most reliable pilot fuel if good quality propane is available, it also has a standard set up. An alternative is using main gas fuel but this will need adjustments to be made to the igniter fuel system to increase flow. Some natural gas fuel is low in UV output so cannot easily be detected by UV detectors. Low calorific value gas fuel makes it difficult to detect pilot flames by temperature rise. 

Logic modifications to bring main fuel on four seconds after pilot fuel help the Ultra Violet or inference flame detection systems because main flames will be easier to detect than pilot flames. In this case even if one combustion chamber is not lit sufficient heat from the others will cause flame to be detected but the will shut down on deviation.

3.2  

Containers of propane need to be large enough and full enough to generate  sufficient gas at the necessary pressure during the light up period and for rapid restarts. Propane bottles will need to be joined in parallel when small ones are used or where they are located a distance away from the turbine or where there are additional pressure drops in the system.

3.3

 To provide the correct flow of igniter gas is done either by a PCV (Pressure Control Valve ) or an orifice normally located after the filter (See diagram 1) in the fuel valve module.

In the case of diagram 1, as well as a PCV in the fuel module, extra PCV’s are fitted on the bottles and an orifice is fitted in the main gas supply line. In this case the on skid PCV was wound full open and the pressure was regulated by the PCV’s on the propane bottles or by the orifice in the main gas supply line.

3.4

 Many sites will need the propane bottles to be heated to maintain sufficient pressure. Others in cold climates where the turbine is outside will need the pipe to the turbine trace heated and lagged to prevent the propane condensing.

3.5

 Tests have shown that for optimum recognition of a propane pilot flame by UV detectors, a pressure of 45-48 psig before the four 1/16^th inch orifices is required. Lower pressure can lead to late recognition on one pilot burner and higher pressures can lead to an unstable flame and the UV detectors losing sight of it. When inference detection is used, higher pressures will give more temperature rise and faster detection.

If a gauge is fitted it can be seen quickly whether the propane supply is sufficient and corrected as necessary.

NOTE Turbines can vary in the optimum pressures for light up. The object is to set pressures so that all four flames light quickly and stay stable.

 3.6  

Pressures for optimum main gas pilot fuel ignition will be higher than for propane. Tests on one onshore supply showed that approximately 60 psig before the 1.16^th orifices was needed for UV detection.  Other sites may need higher pressures.

High pilot fuel flow has been found to cause the gas generator to accelerate above 28%, 2800 rpm and ramp the throttle open before main fuel is switched on. This can lead to high temperatures or stall due to high heat input. (Affects turbines more when gas start and 4 flames detected needed for main fuel).

4.0 Spark igniters.

4.1  

The method of lighting the pilot fuel is by a spark plug. The spark plugs are connected to units(polarity sensitive) that input nominal 24 volts DC and output high voltage pulses sufficient to generate sparking at the plug.

Where there are long distances between the 24 volt dc supply and the igniter units, voltage drop on the supply cables can a problem leading to weak or intermittent sparks. Either an extra battery cell to increase the supply voltage or greater cross section cables can cure the problem.

4.2  

Gaps of 0.025 inch between the outer electrodes and the inner of the spark plug are recommended and should produce the optimum spark .

 4.3

It is possible for the spark units can fail or produce a weak or intermittent spark so reducing start reliability and for the flexible leads from the igniter unit to be damaged.

Replacement of these parts are the cure.

4.4

TB4000 and TB5000 igniter leads are not interchangeable, retrofit units have had problems when fitting TB4000 leads directly to the igniter unit. 

5.0 Ignition Fault Finding.

5.1

Find which pilot flame(s) are not being lit.

Use the control system to indicate which combustion chamber either did not

light or lit late. Late ignition indication on the control system may mean that

the pilot flame did not light or is receiving insufficient gas.

 Observing for presence of flame through the sight glasses on the combustion

chamber can be useful.

5.2

Check that there is plenty of pilot fuel. Propane bottles may be low in

 pressure, if so fit full ones.

5.3

Check that the 1/16 th orifices are clear on the suspect combustion chamber.

Blow out with air.

5.4

Check that the non return valves after the orifices are working. Replace if

necessary.

5.5

Check that the correct pressure is being achieved before the 1/16 th orifices.

If pressure incorrect investigate why.

5.6

 Check for leaks. Repair as necessary.

5.7

Check the spark plug, check gaps or for damage and replace if necessary.

5.8

Check for sparks. Replace spark plug or spark unit or leads.

5.9

Check for 24 volt supply to the turbine skid with igniters switched on.

If there is a large volt drop with the igniters turned on the cable from the supply will need increasing in size, or the voltage increased to compensate.

5.10

Check for condensing pilot fuel. Cold weather can cause this. Pipe from the propane bottles should be heated and lagged.

 5.11

Check for correct operation of the solenoid valve and blockages in the propane line. Replace or repair components as applicable.

6.0 Main Fuel ON.

6.1 When electric 4 to 2 pole motors are used the main fuel should cause the change from 4 to 2 pole by reducing the current through the motor. An undercurrent relay should be set so that the motor cannot change to 2 pole until main fuel is on and lit. Changing to 2 pole without main fuel on and lit can burn out motors and starters because the motor has insufficient torque to accelerate the gas generator and will stay on starting current which is at least 6 times full load current. 

6.2 Once the pilot fuel is lighting up reliably it may be that the turbine trips on deviation. A likely cause is dirty burners or low fuel flow at the point of  main fuel on. If flow is very low one combustion chamber may get less than the others and as the turbine accelerates and deviation limits close a deviation shutdown can happen. A simple solution is to increase the light up flow of fuel. Indications of this problem is that it is more likely when the turbine is cold. Bad deviation is an indication that burners are not clean and will also show as poor spread.

6.3 Stalling.

This is where the temperature rises without the gas generator accelerating.

Stalling can happen at main fuel on but the starter motor will continue to accelerate the gas generator. The stall is likely to sound unusual. 

If happening at light up the light up fuel demand can be reduced although this can lead to deviation shut down.

If the stall happens after light up there are two fuel rates that can be reduced to reduce the chance of stall. This may be necessary in the case of a turbine that uses solids cleaning when running just to achieve a start. In such cases the turbine is more likely to start cold.

A cold wash on a dirty engine can increase air flow and reduce chances of a stall.

7.0 STARTER OFF.

7.1 The starter dog is disconnected at a CT speed of 4800 rpm and the starter motor switched off. Also the fuel input changes to a slower ramp rate. A STALL can then be detected with slow or no acceleration of the gas generator and increase of engine exhaust temperatures. This can be due to over fuel which is due to a dirty or worn gas generator supplying insufficient air.

If fuel input is correct (ramp rate 1) and turbine is in good condition the amount of fuel at starter off should be sufficient to maintain the CT speed and continue acceleration smoothly but at a slower rate. 

7.2 IGV’s in open position. It should not be possible for the turbine to start with IGV’s open, check micro switch operation. IGV change over should be at above 9000 rpm CT speed.

7.3  Interstage BOV’s closed. This can cause stalling if stuck shut. They should be nearly closed at generator no load or very low load and close tight with more than 200-300 kW. On some turbines the interstage BOV’s close at no load. The purpose of the interstage BOV’s is to prevent surging by dumping excess air from the middle stages of the CT which cannot be coped with by the hp stages during the start. Also when load is put on the turbine from zero if surging takes place, later closing BOV’s can help prevent it.

7.4 Electric centre casing BOV’s open or leaking. If these leak or fail to close the result is likely to be high PT exit temperatures during this period of the start or during no load running.

Causes could be the diaphragm damaged, the valve seat damaged, blocked orifices, pipe leaks or leaking ASCO solenoid valves.

 Note: 4 off ASCO solenoid valves are fitted to TB5000 or up-rated TB4000 turbines.

8.0 Poor governing at no load.

8.1 More fuel than demanded. At no load the fuel should be in the region of 14-15 degrees for electronic actuator (approx. 3800kw for STAR systems). If the fuel is at 11 degrees or sitting on the blow out limits for long periods at no load then the fuel regulation is incorrect and it will be difficult to keep speed low. The control system will not allow less than the above fuel demands to prevent the flames blowing out. High regulated gas pressure for electronic actuator systems can cause this problem.

         

   For STAR actuator systems the normal no load heat demand must be higher than the blow out flow setting.

If more is being put into the turbine than asked for the speed can rise and the centre casing blow of valves will open. Usually set to open on 4% above demand speed.

Generators would have difficulty synchronising with others with this problem.

ADJUSTMENTS TO THE CONTROL SYSTEM SHOULD ONLY BE DONE  BY COMPETENT PERSONNEL.