Magnus Genrup GT Workshop Gotland 2017

Gas Turbine Seminar -17
Lund University
Farligaste djuret?
Lund University / LTH / Energy Sciences / TPE / Magnus Genrup / 2017-10-02 Page 2
Farligaste djuret?
Additive Manufacturing - AM
Almost 100% renewable in Germany W616
The European issue…
Cost of Electricity (CoE) vs. OH
CC
oo
E E6
t 0
t0
0
.
7
6
1

0
.
9
6
1

 
6
0
0
  
 
0
0
×4
×2
Design
Courtesy of VGB Powertech
The ”Duck-Curve”…
Sweden
California
4.3 GW/h
Grid codes
Courtesy of Siemens Flexibility
Load
Shutdown
Spinning reserve SS
Steady state
Start
Active generation off-peak

control turndown
LF
Load
Flexibility
• Start-up
– Air attemperation
– Sky venting
– Cascaded steam bypass with attemperation
• Ramping
• Turn-down
– MECL
• Lock-out
GT start-up
Emissions vs. firing level
Flame temperature vs air/fuel Temperature influence on NOx and CO
Stoichiometric
Tmax
NOx
CO
Increasing
Increasing
Temperature
Rich Lean
NOx formation rate
0 10 20 30 40 50 60 1100 1300 1500 1700

AFR = FAR-1 Combustion temp, °C
Typical optimum for design at AFR ≈ 30 (i.e. λ ≈ 2)
Min Emission Compliance Load – MECL
Maximum EGT
III II I
Emissions, mass/unit time

EGT
Exhaust temperature, °C
CO
NOx
MECL
10 30 50 70 90
GT load, percent
Staging to prevent from:
• Lean blow out (LBO)
• Combustion dynamics
Ansaldo sequential (SEV)
Ansaldo (Alstom) GT36
Exhaust loss
N.B. The figure is based on a certain exhaust size – a general
figure should be based on the annulus velocity (VAN=Q/A) since:
• Loss ~ rating or flow
Δha • Loss ~ 1/pressure
• Loss ~ 1/area ~ 1/ stress
60
50
Optimum
Turn-up design Super-
40
region sonic
30 w2 c2
c 22
20 c2 Δha 
u 2
10
0

V m3 s 
Fatigue – low or high?
Wöhler- or σ,N curve
Stress
Stress amplitude, [σa]
σa
σa
Time
Failure
Fatigue limit
No failure – no crack initiation
Low-cycle fatigue High-cycle fatigue
105…7
Life cycles, [n]
Rotor stress – temperature gradients
Temperature induced stresses (radial and hoop):
T αE  r 2  r 2 ry r 
σ r r    2
i
2 
T r dr   T r dr 
TSteam 1  μ r 2  ry  ri ri ri 
TSurface
αE  r 2  r 2 ry r 
2
ΔT σ θ r    2
i
2 
T r dr   T r dr  Tr 
TAverage 1  μ  r 2  ry  ri ri ri 
TCenter
Simplified equations for a case without a bore:
r  1 rrim
1
r
 α E ΔT  r 
σ r r   α E  2 0    
r 2 0
T r dr  T r dr  1  
 rrim  3  r rim 
Example
 1 rrim
1
r
 α E ΔT  r 
σ θ r   α E  2 0    
r 2 0
α= 1.8∙105 °C-1 T r dr  T r dr  T  1  2 
 rrim  3  r rim 
E = 6.9∙104 MPa
ΔT = 55 °C
σ@r=0 = 46 MPa
Compressor blade failure
Siemens SGT5-8000H
Turbine flow path – typical 3-spool aero
Rolls-Royce: The Jet Engine
State-of-the-art combined cycle – GTCC
~ 1.4…1.5 kg/MWs ~ 600°C (+) steam admission

~ 20%
SCR
~ 0.02 bar(a)
Combustor ~ 625…680°C
100%
~ 40%
Compressor Turbine
COT ~ 1,500°C (+) ηCC  ηGT  ηSC  ηHRSG  1 ηGT 

PR ~ 19…25
Siemens HL-class
• 42% efficiency
• 85 MW/min
• 1000 kg/s
• 680 °C EGT
Courtesy of Siemens
Siemens HL-class rotor #1
Convective heat transfer hL 1

Nu    f Re, Pr,... h ~
Conductive heat transfer k L Nu
Courtesy of Siemens
Siemens SGT-800
GE LM9000
Simple Cycle CC & CHP Mechanical Drive
• Free power turbine (2400…3780 rpm) 4 stage HPT
ISO Performance Maintenance Emissions

• 65 MW power output
• 43% simple-cycle efficiency
• 99% availability
• >80% cogeneration efficiency
• 33:1 pressure ratio
• 12,000 hours: inspections 1 stage HPT
• 36,000 hours: hot section replacement 9 stage HPC
2 stage HPT
• 72,000 hours: overhaul 4 stage LPC
• Package design allows engine swap in 24 hours
• 15 ppm DLE NOx & 25 ppm CO
• DLE 1.5 dual fuel (natural gas and liquid fuel)
Fuel flex
• 30…54 MWI gas
SGT-A45 TR
P&W FT4000
Performance
• 140 MW nominal output in twin-engine configuration
• Wet compression for improved performance above
ISO conditions
• Single or dual engine operation (common alternator)
• 50 or 60 Hz performance with no penalty
• 41% (+) thermal efficiency without external cooling
Operational
• Less than 10 minutes start-up time
• 30 MW/min ramp rate
• Synchronous condensation with spinning PT, a FT has
a windage loss of 500…1,000 kW
• NO maintenance penalty for start/stop!
• Fleet has 17,500 OH’s and 1400 cycles
>900 engines
P&W has substantial experience in synchronous >20 years
>40 MOH
condensation without SSS-clutch!
Ventilation work – spinning PT
• The spinning power turbine will feed work into the ”entrapped” air
by increasing the angular momentum
• It is “standard” to assess the work with the equation from the
book by Traupel:
PV    C  1       D2  l2  u 23
PV @ 500
 ~ 23  8 !!!
PV @ 250
• The preceding equation shows why only non-geared PTs can be

operated at nominal speed in ventilation
Firing level
Firing temperature – the misery factor…
• Creep follows an Arrhenius type of expression…

– The Larson-Miller parameter:
PL  M  TM 20  logt 10 3
PL  M  K1  K 2 ln  c   K 3 ln  c 
2
 c   KTaper AN 2 1010 

1.08
~ AN 2
– A more convenient and practical approach is to introduce a

maintenance factor – MF (consumed life per hour of operation)
Firing Lifing MF
850 40000 1
860 20000 2 MF  e 0.0693COT
870 10000 4
880 5000 8
Creep – Larson-Miller parameter
975 1056 1138 1219 @40000

C.R.S
Lifing
• It may be argued that creep-induced damage will reduce the

fatigue life of a metal, and that fatigue-induced damage will
reduce the metals creep life…
• One approach is to use the linear damage summation (DS) model

(also called the linear life fraction, or linear cumulative damage)
for assessing creep-fatigue-life: Dfatigue + Dcreep = Dtotal
• By combining the Robinson Rule (1952) for creep and Miner for
fatigue, one gets the cumulative damage index (failure at unity):
N t
N t D

f r
This is the rationale behind the concept of ”equivalent

operating hours”
The expensive way – equivalent hours…
OH nstarts ntrips
EOH  1 F fuel  F firing  1 Fstarts  Fload rate  1 FTrips
1 1 1
 
     
creep and oxidation LCF
Where:
EOH equivalent operating hours
OH actual operating hours
Ffuel factor depending on fuel
Ffuel factor depending on firing level
nstarts number of “fired” starts
Fstarts number of hours per start
Fload rate load rate factor
Lifing
Miner-Palmgren
N t
N N D
f r
Robinson
Operational hours
24000
Maintenance interval  hours
Factored Hours Actual Hours
Factored Hours = (K + M  I)  (G + 1.5  D + Af  H + 10  P)
Actual Hours = (G + D + H + P)
G = Annual Base Load Operating hours on Gas Fuel

D = Annual Base Load Operating hours on Distillate Fuel
H = Annual Operating Hours on Heavy Fuel
Af = Heavy Fuel Severity Factor (Residual = 3 to 4, Crude = 2 to 3)
P = Annual Peak Load Operating Hours
I = Percent Water/Steam Injection Referenced to Inlet Air Flow
M&K = Water/Steam Injection Constants (see GE documentation)
Noff cycles
900
Maintenance interval  Starts
Factored Starts Actual Starts
η
Factored Starts  0.5  NA  NB  1.6  NP  20  E  2  F   aTi  1Ti
i1
• Actual Starts = (NA + NB + NP)

• S = Maximum Starts-Based Maintenance Interval (Model Size Dependent)
• NA = Annual Number of Part Load Start/Stop Cycles (<60% Load)
• NB = Annual Number of Base Load Start/Stop Cycles
• NP = Annual Number of Peak Load Start/Stop Cycles (>100% Load)
• E = Annual Number of Emergency Starts
• F = Annual Number of Fast Load Starts
• T = Annual Number of Trips
• aTi = Trip Severity Factor = fcn(Load, Trip during accel. = 2, Peak = 10)
• η = Number of Trip Categories (i.e. Full Load, Part Load, etc.)
Turbine governor – twin-shaft
EGTcontrol  f COTNom , PR, n,...
T 4 P η p  γ 1  n 1
42 41 4
T4x  p4  γ p  n
    4 
T5  p5   p5 
5 P5
s
Courtesy of Siemens
Performance modelling
HPC
3.6
5
0.8
6
0.8
3.2 4
0.8
2.8
P3q25 HPC Pressure Ratio
N=1.05
0
0.8
N=1
2.4
N=0.95
2
0.8
N=0.9
2
N=0.8
5
0.7
N=0.8
5
1.6
N=0.7
N=0.7
5
0.70
1.2
N=
N=
0.6
0 .5
Not converged points are marked with a red symbol.

.8
4 8 12 16 20 24 28 32 36 40
Mass Flow W25RSTD [kg/s]
LPT
2.5
N=0 N=1.1
P48q5 Power Turbine Pressure Ratio
1
N=.2
1
.9
2
N=
4
0.8
0.9
93
N=
0.
2
0 0.9 7
0.9 0.
.88 N=
0 .85 0
1.5 0.80
N=0.6
Not converged points are marked with a red symbol.
44 48 52 56 60 64 68 72 76
W48*sqrt(T48)/(P48/Pstd) [Kg/s]
Some results…
Burner Temperature ZT4 = 850 ... 880 [C]
1.012
112
19.5
14
3
*10
110
1.01
12
1.008
108
10
19
1.006
106
2.71828^(0.0693*(ZT4-850))
8
Rel. HP Spool Speed
Relative Torque [%]
Electric Power [kW]

1.004
104
18.5
6
1.002
102
4
100
18
1
2
.998
98
848 856 864 872 880

Burner Exit Temperature T4 [C]
Impact on output from firing and ambient
Burner Temperature ZT4 = 850 ... 880 [C] Inlet Temperature T2 [C] = -28...32
Ambient Temperature Ts0 = -30 ... 30 [C]
28
*10 3 -28
-24
-20
-16
26 -12
-8
-4
0
4
8
24 -28 12
16
-24 20
24
-20 -12 28
22 -16 -8
Electric Power [kW]
-4
20 0 8
4
12
18 16
20
ZT4 = 850
24
ZT4 = 855
28
ZT4 = 860
16 ZT4 = 865
ZT4 = 870
ZT4 = 875
ZT4 = 880
14
848 852 856 860 864 868 872 876 880 884
Flat rate @18 MW
Ambient Temperature Ts0 = -30 ... 30 [C]
925
.315
32
900
28
875
24
.31
850
20
2.7183^(0.0693*(ZT4-850))

825
.305
16
Thermal Efficiency
800
12
775
8
.3
750
4
725
.295
0
700
-4
-40 -20 0 20 40
Inlet Temperature T2 [C]
Emulsionseldning

Magnus Genrup GT Workshop Gotland 2017

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Magnus Genrup GT Workshop Gotland 2017

Uploaded by

Copyright:

Available Formats

Gas Turbine Seminar -17

Courtesy of VGB Powertech

Active generation off-peak

Flame temperature vs air/fuel Temperature influence on NOx and CO

NOx formation rate

0 10 20 30 40 50 60 1100 1300 1500 1700

Typical optimum for design at AFR ≈ 30 (i.e. λ ≈ 2)

Emissions, mass/unit time

Wöhler- or σ,N curve

Simplified equations for a case without a bore:

Rolls-Royce: The Jet Engine

~ 1.4…1.5 kg/MWs ~ 600°C (+) steam admission

COT ~ 1,500°C (+) ηCC  ηGT  ηSC  ηHRSG  1 ηGT 

Convective heat transfer hL 1

ISO Performance Maintenance Emissions

• The preceding equation shows why only non-geared PTs can be

• Creep follows an Arrhenius type of expression…

 c   KTaper AN 2 1010 

– A more convenient and practical approach is to introduce a

975 1056 1138 1219 @40000

• It may be argued that creep-induced damage will reduce the

• One approach is to use the linear damage summation (DS) model

This is the rationale behind the concept of ”equivalent

Factored Hours = (K + M  I)  (G + 1.5  D + Af  H + 10  P)

G = Annual Base Load Operating hours on Gas Fuel

• Actual Starts = (NA + NB + NP)

P3q25 HPC Pressure Ratio

Not converged points are marked with a red symbol.

Not converged points are marked with a red symbol.

Relative Torque [%]

Electric Power [kW]

848 856 864 872 880

Burner Exit Temperature T4 [C]

You might also like