You are on page 1of 60

Industrial Plant Control Student:

Technical University of Cluj-Napoca


Faculty of Automation and Computer Science
Department of Automation

Semester Project
discipline
Industrial Plant Control

Analysis and control of a


boiler – turbine – generator–system
electro-power radial

Coordonator:

Student:

Page 1
Industrial Plant Control Student:

Content

Chapter 1: Process presentation …………………..………….................... 3

Chapter 2: Data design

A. Thermal part ........................................................................................12

B. Electric part ......................................................................................15

Chapter 3: Simplified identification of the steam boiler near a functional steady


state
value…………………………………………………………………......17

Chapter 4: Representative transfer functions for the boiler sub-processes........18

Chapter 5: Controller tuning of the main control loops ..............................27

A. Processed steam pressure adjustment....................................................27


B. Controller computation for the control loop of the processed steam.....31

C. Control of the frequency and active power……....................................34


D. Reactive power and voltage control.......................................................47
Chapter 6: Conclusions..........................................................................................

Page 2
Industrial Plant Control Student:

1. Process presentation

Project Objective

The simplified design of the main control circuits for the following system:
boiler-turbine-generator (an electric-energetic radial system) and the analysis of
some important step responses.

Given Data:

Group Power P = 122 [MW]


Pressure of the living steam p = 90 [bar]
Thermal Efficiency (exact usage) 1/ = 2392 [Mcal/MWh]
Enthalpy of the living steam i2 = 780 [Mcal/to]

We will design in a simplified form the main automated control


circuits and we will analyze the significant responses to perturbations for a
Boiler group (c) – Turbine (t) – synchronous Generator (g) system and a
radial energetic system (s).
The details of the control circuit in a simplified form are :
EC1
u*f - uf
+ uf f
f

STATISM EC3
RUG + u*G
+ ∆UP ± ∆N -
u * + EC2 EEx
P RN SMH =
-
~
+uP V -I U

C T G S
f
qabv P+jQ

P UG
up
P

Page 3
Industrial Plant Control Student:

Figure 1: Radial energetic – electronic system(control circuit)

Notations: RN: active power regulator


SMH: hydraulic actuator drive of the partial intake valve V of the
steam flow
q
at the entry point of the turbine
ab
C: boiler
T: turbine
G: synchronous generator
S: radial power system
P/UP : transducers and power-voltage adaptors; unified;
U/f : frequency adaptors - unified voltage;
Statism : corrective block formed out of an active quadripol which
ensures a linear dependency with a certain predetermined
slope for slope f(P) ;
=/~ : voltage rectifier block
RUg : voltage regulator at the generators terminals
EEx : excitation element of the generator
Other notations :
- i * = pressure ref of processed steam;
apv
-i * = ref of O in burned gases ;
O2 2
-i * = source depressurization ref;
ΔPF
-i * = temperature ref for processed steam;
θav
-i* = water level ref from the tambour(cylinder) ;
h
-i = salinity ref.
S
The system is formed out of 4 subsystems: boiler of the group C; turbine T; sync
generator G ; radial power system S .
In the composition of the system there are control loops compared to active power
(P) and relative to the frequency (f). Also there are regulators for the steam flow
inside the turbine ensured by the hydraulic actuator drive (SMH) and valve (V)
which ensures partial air admission into the turbine. We also have a control
voltage system U at the generators terminals in association with current correction
offsetting effect. There are transducers and power-voltage adaptors (P/U ),
P
frequency-voltage adaptors (f/U), statism block, a voltage rectification block from
the voltage regulator at the generators terminals.

Page 4
Industrial Plant Control Student:

(~/=), power amplifier levels (EEx), representing the excitation circuit of the
generator and others. The block scheme of the system is the following:

Figure 2: Block scheme of the system


The components of the previous scheme are:
1. The boiler (C) for which the minimum number of input signals is 6. Under the
simplified form we consider only one output signal, qabv, which is adjusted by
strangulation with the valve (V), equivalent with the steam turbine inlet.
Cf
2. The turbine (T) drives the generator G. The live steam from the turbine
enters the condenser CO and due to the cooling process we get the same flow
of condensed water qabv.
3. The execution element SMH is composed of a rack driven by a hydraulic slide-
piston mechanism made of the slide S and the piston P. There also is a
mechanical negative feedback of positioning type using the lever L1, L2. By
using this negative feedback we get 2 advantages:
a) decreasing the equivalent mechanical inertia;
b) improving the linear behavior by decreasing the non-linear disturbances caused
by dry friction, mechanical wear and mechanical disturbances.
Because the oil is incompressible it cumulates numerous benefits: very small
mechanical inertia, undamped transient forces, high precision.

There are also functional blocks: RN – active power regulator;


Ap – power amplifier; SM - actuator
4. The generator (G) triggered coaxially by the T turbine delivers in the energetic
system in three phased regime RST, both active power, proportional with the
mechanical power taken from the turbine and reactive power needed for the
magnetization of the force circuits (transformers and electrical machines).
4. The generator rotor is powered in continuous excitation voltage, delivered by the Ex
exciter, having active power at most 1% of generator power. The exciter excitation is
powered in DC voltage at a current I exc resulted from the output terminals of the
controlled rectifier bridge formed by 6 thyristors.
The control of the medium rectified current is assured in the classical form
that is through command pulses on the grid having variable phase (a). The

Page 5
Industrial Plant Control Student:

modifying of a is done theoretically between 0˚ and 180˚, practically between 5˚


and 175˚.
The control of the active power is made through a parallel regulation
schema in relation to the active power P and in relation to the f frequency. The
decreasing linear characteristic f(P) assures through -k s the active power
regulation statism. There are also two adaptors: power–unified voltage and
frequency-unified voltage.
The voltage regulation and of the reactive power is assured through a triple
cascade having RUG (the voltage regulator at the generator terminals), RIE (the
excitation current regulator), RUE (the excitation voltage regulator), that commands
CCG block (the command complex on the grid). At the CCG output we obtain the
variable phase shift in wide limits for grid command of the 6 thyristors. This
equivalent loop in triple cascade has a positive current reaction with a much lower
weight, of a couple of percents (compound reaction).
The 6 control loops (in a very general form) for a steam boiler are presented
in the below figure:

Page 6
Industrial Plant Control Student:

Figure 3:Control loops (general form) for a steam boiler

Page 7
Industrial Plant Control Student:

Functional details

qabv - the steam flow controlled by a control active power classical circuit by comparing a
reference power Up* proportional to the actual active power given by the system.

RN – the active power controller which commands the partial inlet steam valve V, through
the hydraulic actuator SMH

The generator G is provided with adjustment reactions of the reactive power, terminal
voltage UG through RUG regulator and Eex excitation element. UG* is the reference
voltage generator.
Is deducted a comparison reaction against current i0. There is also ensured a statism
reaction against frequency f by comparing a reference frequency (usually 50 Hz) to the
voltage Uf proportional with the real frequency. The Uf*-Uf deviation is processed through a
proportional effect in the statism element providing an additional component ∆UP of the
active power.

The live steam pressure control is ensured through a scheme in cascade with two PI
controllers. On the main circuit we have a pressure-unified current converter and on the
internal circuit, a differential pressure –unified current converted, followed by a root
extractor. In general, on the flow’s feedback loop, a transducer with calibrated diaphragm,
the differential pressure ∆p being proportional to the square of the flow. It is required the use
of a root extractor, this way, the unified current being proportional to the flow. The external
feedback loop ensures the reaction against live steam pressure, and the inner loop ensures
reaction against fuel flow qc.

The second control loop ensures proper combustion by maintaining the oxygen
concentration resulted from the flue gas in acceptable limits. Has also a control loop in
cascade, in relation to the oxygen for external loop, respectively combustion air, qaer, for the
internal loop.

The third control loop is for depressure in ΔpF outbreak. It is necessary to maintain a
depressure of -2mmH2O (millimeters of water) inside the furnace. On the feedback loops
there are differential pressure – unified current transducers.

The water supply flow control is provided through a convergent setting with the following
reactions:
- In relation to the supply flow: qapa
- In relation to the water level in the reel: h
- In relation to the live steam pressure: pabv
- In relation to the live steam flow: qabv
The transducers and converters type in unified signal is the same as in the previous schemes.

Page 8
Industrial Plant Control Student:

Process steam temperature control is a convergent regulation with respect to 2


reactions: the temperature of the process steam Өabv, and the temperature of the process
steam inside an intermediary steam over-heater Өabsi. On the feedback circuit a temperature
to unified current/voltage converter appears.
The salinity (S) of condensed water control from the cold funnel of the boiler.
The main loop is assured by the salinity, but there is also an additional correction loop
which is linearly dependent on the process steam flow. The main loop includes
salinity/voltage and a voltage/current transducers. The are also additional loops which
provide a corrective effect.
Notations:
i* – unified reference current
u* – unified reference voltage
t – time
Ө – temperature
qij – injection flow between 2 super-heater steps in order to regulate the process
steam temperature
ip*abv – reference in unified current for the process steam pressure
Qpj – purge flow
qabv – process steam flow that enters the turbine
iO2* – reference in unified current of oxygen for the fumes analyzer
Өabsi – steam temperature in the intermediary super-heater
iΔpF* – vacuum reference within the furnace
Өabv – process steam temperature at turbine input
ih * – water level reference in the drum
h – water level in the drum
iӨabv* – process steam reference
ΔpF – furnace vacuum
iS * – salinity reference
Cip,Cpi – current/pressure converter and vice versa
Q, q – refers to flow
P – refers to pressure
E – execution element
Δp – differential pressure (output from the metering orifice)
pabt – steam pressure in the drum
Er – square root extractor
SPR – pneumatic servomotor with tap
SPC – pneumatic servomotor with clack
S - salinity

Page 9
Industrial Plant Control Student:

Qc,qc -fuel flow


qaer, qapa, qga-combustion air flow, supply water flow and exhaust gas flow
EC -correction element
O2[%] -molecular oxygen percentage concentration
SHR -hydraulic servo valve
SHP -hydraulic damper actuator
R -controller

Explanation:

STATISM: correcting block composed out of an active quadripol, that ensures a


linear dependency, for the slope f(P), with a predetermined slope.
On the generator G terminals, the two power components (P, Q) are applied, obtaining
S – apparent power: S  P 2  Q 2 ; [ MVA ]

The frequency f of the radial electro-energetic system (unpaired with another system)
is introduced in the transducer (uf/f) and the unified voltage signal (-u ), which is linearly
f
dependent on the frequency, is introduced in the comparison element EC1, where u * is the
f
frequency reference value in unified voltage. The EC1 block output error is corrected in the
Statism element, representing a negative feedback signal which is introduced in EC2. The
second unified reference signal is u * - the active power reference which is compared with
p
the negative feedback signal –u , that is linearly dependent on the active power P, generated
p
by the generator G (the second feedback circuit).
The output error of EC2 is processed by the active power controller RN and the
command signal operates the SMH execution element. The SMH output signal represents the
axial displacement (high forces – hundreds of kg) that acts on the vent V, which is destined
to control the steam flow (q ).
abv
Control of the reactive power at the generator’s terminals is achieved through a
double reaction system that has a negative feedback for U and a positive feedback for I (with
a much smaller weight). The two signals (U and I) are rectified and the unified voltage signal
is applied as an equivalent feedback in EC3.

Page 10
Industrial Plant Control Student:

The output error of EC3 is processed in RU , and the command signal is applied on a
G
complex element called excitation circuit. The output of this element is the excitation current
of the generator G rotor, allowing it to produce the excitation flux.

Remarks:

the current references are denoted by i* and refer to slow pocesses: 2-10 or 4-20
-
mAcc; The natural zero (0 mA) is avoided in order to reduce the noise influence from
the process, because the slow processes have at least one time constant bigger than 10
sec.
the voltage references are denoted by u* and refer to fast processes 1-10V or  10Vcc;
-
the time constant is smaller than 10 sec.

All the liquid flows are denoted by q (Q), the temperatures by , the pressures with p,
the differential pressure by p and the salinity by S. Thus, we have:
1
Q 2  k * p; Q  k 2 * p ; p  p1  p 2

There are two different control circuits :


1) with respect to the active power for frequency control;
2) with respect to the voltage at the G terminals for voltage control and reactive power
control.

In the boiler control scheme we have the following control circuits:


the boiler charge, the charge controller, the pressure function acting both on the gas
1)
and the combustion air;
2) the combustion control, following the quality of the combustion, and the tracking
control in stoichiometric proportion of the combustion air in relation to the gas;
3) the differential pressure in the burner;
4) the water supply, depending on the quality of the combustion relative to the gas;

Page 11
Industrial Plant Control Student:

5) the output steam temperature, having a correction relative to the steam temperature
before the injection;
6) the continuous purge depending on the steam flow and the water salinity in the boiler.
The references for the six control circuits are:
iapv*, iO2*, iΔPF*, ih*, iqav* iS*.

2. Data design

A. Thermal part

PGnom  P  122[MW];
1) The main parameter: the nominal power of the generator:

2) The nominal pressure of the living steam in the boiler:


Pabvnom  p  90[bar ];

3) The thermic specific consumption:

1  Mcal 
 2392 ;
  MWh 
4) The processed steam enthalpy:

i =780 [Mcal/tona];
2

 MCal 
PCCH 4  8,5
 Nm 
3

5) The gas (methane gas) caloric power:

6) The nominal gas flow:

1  MCal 
 PGnom  MW 
  MWh  2335  360  Nm 3 
qCnom  qCH 4 nom    98894.1176  
 MCal  8.12
Page 5  h 
PC CH 4 
 Nm 
3
Industrial Plant Control Student:

7) Nominal value of processed steam:

 abnom  530 0 C

8) Values of the molecular oxygen contained in the burned gas:

O (%) = 5%
2

9) Nominal pressure of the steam in the drum:

P = 1,1 · P = 1.1·90 = 99 [bar]


abvtnom abvnom

Nominal temperature of the steam after the first superheater


10)
 ab sin om  480 0 C

11) Nominal combustion air flow nominal:

 Kcal 
PCCH 4 
 Nm   Nm   b)  359338.34  Nm 
3 3 3
q aernom  qcombnom      ( a  h 
 h   Kcal   
1000 3 
 Nm 

 - excess combustion air;  =1.1;


 a=1.09; si b=0.25; - coefficients that depend on the gas composition

12) Nominal thermal depressure:


p Fnom  2[ mm H 2 O ]
 2mm H 2 O

13) Burned gas flow:

q =q +q = 34332.235 [Nm3/h] + 359338.34 [Nm3/h] =393670.576 [Nm3/h]


ganom CH4nom aernom

Page 13
Industrial Plant Control Student:

14) Nominal processed steam flow:

1  MCal 
 PGnom  MW 
  MWh   tona 
q abvnom  q abviu   389.0986
 MCal 
(i2  i1 )   h 
 tona 

i1: Enthalpy of the feedwater when exiting the condenser at 300C;


- Thermal energy contained in the pressurized water

 MCal 
i1  30 
 Ton ă 

i2: Enthalpy of the processed steam after the superheader (the parameter limits of the
processed steam are taken into consideration and we determine the liniar
interpolation).
P = 100 [bar] => i2= 809 [Mcal/tonă]
abvvar
P = 210 [bar] => i2= 781 [Mcal/tonă]
abvvar

15) Feedwater flow:


q =q = 389.0986 [tona/h]
apanom abvnom

16) Nominal salinity:


S = 0,01 [mg/tonă] = 10-8 [kg/tona]
nom

17) Injection flow:


q = 10-2 * q = 0.01 * 389.0986= 3.8910 [tona/h]
ijnom abvnom

18) Purge flow:


q = 10-3 q = 0.001 * 389.0986=0.3891 [tona/h]
pjnom abvnom

Page 14
Industrial Plant Control Student:

B. Electric Part

The circuit parameters for the generator excitation:

Power Factor:
1.
P[W ]
cos    0.825
S [VA ]
The bigger the active power is, cosφ is higher by economical reason. A greater reserve
of stability is assured. Commonly cosφ = 0,91-0,92 is not surpassed to ensure a reactive
power big enough to force in short circuit regime the excitation of the generator.
Q[VAr ]
- required reactive power to magnetize the magnetic circuits of the
generator, transformers, electric engines. The compensation of reactive power is followed.

Apparent Power: P = UIcosφ; Q = UIsin φ


2. S
Q
S = PGnom [MW] = 147.8788 [MVA]
Gnom cosnom φ
P
Nominal reactive power:
3.

Q ²= S ²[MVA] - P ²[MW] = 6984.1359 [MVAr] => Q = 83.5711 [MVAr]


Gnom Gnom Gnom Gnom

Generator terminals tension: U = 15 [kV]


4. Gnom

Nominal excitation tension: U = 300 [V]


5. Enom

Exciter nominal Power: P = 0,0025 * P = 0.3[MW]


6. Enom Gnom

Controlled bridge nominal power: P = 0,01 * P = 0.003 [kW]


7. PCnom Enom

PEnom [MW]
Excitation nominal current: i = = 0.001 [kA]
8. Enom UEnom [V]

Controlled bridge nominal tension: U nom = 270 [V]


9. PC

Page 15
Industrial Plant Control Student:

Exciter excitation time constant: T = 0.1 [sec]


10. EE

The time constant of the generator’s excitation: T = 0.5 [sec]


11. E

12. The time constant of a three phased generator: TG = 4 [sec]

13. The current of the bridge rectifier:


PPCnom [MW]
iPcnom= UPCnom [V] = 0.00001 [A]

The nominal values of U ,U ,U , T , T , T have been chosen according to the


Gnom Enom Pcnom E EE G
tables below:
(U nom – the voltage drop of the bridge rectifier)
PC

P [MW]
Gnom 100÷200 201÷300 301÷400 401÷>401
(aici =420)
U [kV] 15 20 25 30
Gnom
U [V] 300 400 500 600
Enom
U [V] 270 360
Pcnom

PGnom [MW] 100 – 300 301 – 500 >500


TEE [sec] 0.1 0.15 0.2
TE [sec] 0.5 0.75 1
TG [sec] 4 5 6

UGnom – the voltage drop at the generator’s terminals


UEnom – the nominal voltage drop of the excitation
UPCnom – the voltage drop of the bridge rectifier

Page 16
Industrial Plant Control Student:

TEE – the time constant of the exciter’s excitation

TE – the time constant of the generator’s excitation

TG – the time constant of the three phased generator

3.Simplified identification of the steam boiler near a functional


steady state value

It is considered that the boiler has 6 input signals and 9 output signals. Theoretically,
there is a transfer function between each input and each output. Practically, most of these
transfer functions have negligible weight, resulting only 19 important transfer function.
The graphical approach is used in order to establish the dependence between signals.

H11 pabv
1
qCH4 H
H1221 O2
1 2
H22
H23 pF
qaer 3
2
H33 h
H44
qga 4
3 H15
H25 pabt
H27 5
qapa H16
4 H26 qabv
H46 6
H28 H17
qij absi
5 H57 7 S
H18 abv
qpj H49
6 H69 8
Figure 4. The connections between the main signals
9

Page 17
Industrial Plant Control Student:

We will go through the significant step responses between these signals to


approximate the transfer functions. Not all these transfer functions have a meaningful
phenomenological interpretation or sufficient weight to be taken into account.
Only the ones having a sufficiently large magnitude will be taken into consideration.

4. Representative transfer functions for the boiler sub-processes

Taking into consideration the low-pass filter with phase delay characteristic of the majority
of sub-processes included in the transfer functions, the approximations can be done through
step response identification. Despite the fact that the approximation of transfer functions
using proportional forms to first order functions can be somewhat rough, it is generally used
in design, while the second order ones are rarely used.
We have two cases:
A.
Command
i
Nominal value
i0

0 t
e 1 3
Output
est

0 Tm t
T

In all the cases, we have a time delay Tm. (e=output, i=command)

Le k
H1    Tm s ;
Li 1  Ts
For curve 1 we have:

Page 18
Industrial Plant Control Student:

For curve 2 we have:


For curve 3 we have:

Le k
H2    Tm s ;
Le Li 1  (T
k 1  T2 ) s  T T
1 2 s 2

H3    Tm s ;  p   n 1   2 ;
Li 2 1
1 s  2 s2
n n

Command
B.
i
Nominal value
i0

0 t
e Output
0 Tm t
1
2
est

3
T
In this case, the transfer functions corresponding to each curve will be opposed to the ones at
point A.
The following notations were used:
est- the stationary value of the output
i0-the stationary value of the input
T- time constant of the sub-process
Tm-time delay constant

Calculus of the main transfer functions*:

1. The transfer function of the steam’s pressure with regard to the fuel flow:
L{ p abv } k11
H 11 ( s )    Tm11 s ;
L{q c } 1  T11 s

p abvnom [bar ] 90 h * bar


k11  3
  2.6 * 10 3 [ ]
Nm 34332.2352 N * m3
q cnom [ ]
h

Page 19
Industrial Plant Control Student:

Tm11  10[min]  600[sec]; T11  100[min]  6000[sec];

2.6 10 3  600 s


H11 ( s )  1 6000 s
;

*k – steady state gain; T - delay time constant; T – time constant


m

2. Transfer function of O2 content from exhaust in relation to the fuel flow:

L{O2 } k 21
H 12 ( s )    Tm12 s ;
L{qc } 1  T12 s O2 [%] [%  h]
k12   h  %3   1.4563 *10 -4 [ Nm 3 ]

q cnomNm 
Nm
h
3

 
Tm12  10[sec] ; T12  60[sec];

1.456310 4  10 s
H 12 ( s )  1 60 s
;

3. Transfer function of the processed steam in the drum, in relation to the fuel flow:
L{ p abt } k15
H 15 ( s )    Tm15 s ;
L{q c } 1  T15 s
Tm15  9[min]  540[sec];
T15  90[min]  5400[sec];
(1.1)· pabvnom[bar ] 99[bar ]  h  bar 
k15    2.8  10 3 
 Nm   Nm   Nm 
3 3 3
qcnom   34332.2352 h 
 h   
3 540 s
H 15 ( s)  2.9110 5400

s
;

4. Transfer function of the processed steam flow in relation to the fuel flow:
L{q abv } k16
H 16 ( s )    Tm16 s ;
L{q c } 1  T16 s
Tm16  12[min]  720[sec];
T16  120[min]  7200[sec];

 tonă   tonă 
q abvnom   389.0987 
k16   h   h   11 .3  10 3  tona 
 Nm 3   Nm 3   Nm 3 
qcnom   34332.2352  h 
 h   

Page 20
Industrial Plant Control Student:

3 720 s
H 16 ( s)  11.3110 7200 s ;

5. Transfer function of the steam temperature in the intermediary over-heater in relation


to the fuel flow:

L{ abvsi } k17


H 17 ( s )    Tm17 s ;
L{q c } 1  T17 s
Tm17  10[min]  600[sec];
T17  100[min]  6000[sec];
 ab sin om [C ] 480[C ]  3  h C 

k17    14  10  3 
 Nm 3   Nm 3   Nm 
q cnom   34332.2352 
 h   h 
3 600 s
H 17 ( s)  1410 
1 6000 s
;

6. Transfer function of the processed steam temperature in relation to the fuel flow:
L{ abv } k18
H 18 ( s )    Tm18s ;
L{q c } 1  T18 s
Tm18  12[min]  720[sec];
T18  120[min]  7200[sec];
 abvnom [C ] 530[  C ]  3  h C 

k18    15. 4  10  3 
 Nm 3   Nm 3   Nm 
qcnom   34332.2352  
 h   h 
3 720 s
H 18 ( s)  15.4110 7200 s ;
7. Transfer function of the processed steam pressure in the tank in relation to the combustion
air flow :

L{ p abv } k 21
H 21 ( s )    Tm 21s ;
L{q aer } 1  T21 s
Tm 21  10[min]  600[sec];
T21  100[min]  6000[sec];

p abvnom [bar ] 90[bar ]  h  bar 


k 21    2.5046  10 4  3 
 Nm 
3
 Nm 
3
 Nm 
q aernom   359338.3407  
 h   h 

4 600 s
H 21 ( s)  2.50461106000s ;

Page 21
Industrial Plant Control Student:

8. The transfer function of the oxigen percentage contained in the exhaust gases with respect to
the combustion air flow:
L{O2 } k 22
H 22 ( s )    Tm 22s ;
L{q aer } 1  T22 s
Tm 22  10[sec];
T22  60[sec];
O2 nom [%] 5[%] %  h
k 22    1.3914  10 5 
 Nm   Nm   Nm 
3 3 3

q aernom   359338.3407  
 h   h 
5 10 s
H 22 ( s)  1.391411060 s ;

9. The transfer function of the furnace vacuum with respect to the combustion air flow:
L{p F } k 23
H 23 ( s )    Tm 23s ;
L{q aer } 1  T23 s
Tm 23  10[sec];
T23  5[sec];
p Fnom [mmH 2 O ]  2[ mmH 2 O]  6  h  mmH 2 O 
k 23    -5.5658  10  Nm 3 
 Nm 3   Nm 3   
q aernom   359338.3407  
 h   h 
6 10 s
H 23 ( s )  -5.5658110 5 s  ;
When the air flow increases, the furnace vacuum decreases to negative values.

10. The transfer function of the steam pressure in the drum with respect to the combustion air
flow:
L{ p abt } k 25
H 25 ( s )    Tm 25 s ;
L{q aer } 1  T25 s
Tm 25  9[min]  540[sec];
T25  90[min]  5400[sec];

(1.1)· pabvnom [bar ] 99[bar]  h  bar 


k 25    2.7551 104 
 Nm   Nm   Nm 
3 3 3
qaernom   359338.3407  h 
 h   
4 540 s
H 25 ( s )  2.75511105400s ;

11. The transfer function of the processed steam flow with respect to the combustion air
flow:

Page 22
Industrial Plant Control Student:

L{q abv } k 26
H 26 ( s )    Tm 26 s ;
L{q aer } 1  T26 s
Tm 26  12[min]  720[sec];
T26  120[min]  7200[sec];
 tona   tona 
q abvnom   3.5934 
 h    h   tona 
k 26   1.1  10 -3  ;
 Nm   Nm   Nm 
3 3 3

q aernom   359338.3407  
 h   h 
3 720 s
H 26 ( s)  1.1110 7200

s
;

12. The transfer function of the steam temperature in the intermediary superheate with
respect to the combustion air flow:
L{ absi } k 27
H 27 ( s )    Tm 27 s ;
L{q aer } 1  T27 s
Tm 27  10[min]  600[sec];
T27  100[min]  6000[sec];
 ab sin om [C ] 480[C ]  3  h C 

k 27    1.3  10  ;
 Nm 3   Nm 3   Nm 3 
q aernom   359338.3407  
 h   h 
3 600 s
H 27 ( s)  1.3110 6000

s
;

13. The transfer function of the process steam temperature with respect to the combustion air
flow:
L{ abv } k 28
H 28 ( s )    Tm 28s ;
L{q aer } 1  T28 s
Tm 28  12[min]  720[sec];
T28  120[min]  7200[sec];

 abvnom  C  530 C   3  h C 

k 28    1. 5  10  
 Nm 3   Nm 3   Nm 3 
q aernom   359338.3407  
 h   h 
H 28 ( s)  1.51107200
3 720 s

s ;
14. Fdt of the depressure in the burner, in relation with the flow of burned gases :
L{p F } k 33
H 33 ( s )    Tm 33 s ;
L{q ga } 1  T33 s
Tm 33  5[sec];
T33  10[sec];

Page 23
Industrial Plant Control Student:

p Fnom [mmH 2 O]  2[mmH 2 O]  h  mmH 2 O 


k 33    -5.0804  10 6  
 Nm 
3
 Nm 
3
 Nm
3

q ganom   393670.576 
 h   h 
6 5 s
H 33 ( s)   5.080411010 s  ;
In the event of qair = 0 and qwater = 0 it is considered that the pressure in the burner is equal
with the outside pressure. If the exhausters are turned on at q ganom maintaining qaer = 0 in the burner,
results a depressure which in stationary regime is –2 [mmH2O].

15. Transfer function of the water level in the tambour in relation with the flow of the supply
water:
L{h} k 44
H 44 ( s )    Tm 44 s ;
L{q apa } 1  T44 s
Tm 44  20[sec];
T44  20[sec]; hnom=100[mm]=0.1 [m];
hnom [ m] 0.1[ m] h  m
k 44    2.57 * 10  4  
 tona   tona   tona 
q apanom   389.0987  h 
 h   
4 20 s
H 44 ( s)  2.57*110 20s ;

16. Transfer function of the processed steam flow in relation with the flow of the supply water:
L{q abv } k 46
H 46 ( s )    Tm 46 s ;
L{q apa } 1  T46 s

Tm 46  12[min]  720[sec]; T46  120[min]  7200[sec];

 tona   tona 
q abvnom   389.0987 
k 46   h    h   1;
 tona   tona 
q apanom   389.0987  
 h   h 

720 s
H 46 ( s )  1 7200 s ;
17. Transfer function of the saltiness in relation with the flow of the supply water:
L{S } k 49
H 49 ( s )    Tm 49 s ;
L{q apa } 1  T49 s
Tm 49  60[min]  3600[sec];
T49  600[min]  36000[sec];
S nom  % 1[%] h  %
k 49    2.57  10 5  
 tona   tona   tona 
q apanom   389. 0987 h 
 h   

Page 24
Industrial Plant Control Student:

4 3600 s
H 49 ( s)  2.57110 
 36000 s
;

18. Transfer function of the temperature of the processed steam in the intermediary
overheater in relation with the injection flow (water or steam)

L{ absi } k 57
H 57 ( s )    Tm 57 s ;
L{qij } 1  T57 s
Tm 57  10[sec];
T57  10[sec];
 ab sin om  100C ;
 ab sin om [C ] 100[  C ]  h C 
k 57     25.7004 
 tona   tona   tona 
q ijnom  3.890986
 h   h 
10 s
H 57 ( s )   25.7004 
110 s
;

19. Transfer function of the flow saltiness of purge :

L{S } k 69
H 69 ( s )    Tm 69 s ;
L{q pj } 1  T69 s

Tm 69  60[sec];
T69  60[sec];
S nom [%] 1[%] h  %
k 69     0.0257 
 tona   tona   tona 
q pjnom  0.3890986
 h   h 

60 s
H 69 ( s)   0.0257 
1 60 s
;

The transfer matrix of the boiler having the dimension (9x6) with 19 transfer functions
is now presented. The other 35 elements have the value of zero:

qc qaer qga qapa qij qpj


pabv H11 H21 0 0 0 0

O2 H12 H22 0 0 0 0

ΔpF 0 H23 H33 0 0 0

Page 25
Industrial Plant Control Student:

h 0 0 0 H44 0 0

pabt H15 H25 0 0 0 0

qabv H16 H26 0 H46 0 0

θabsi H17 H27 0 0 H57 0

θabv H18 H28 0 0 0 0

S 0 0 0 H49 0 H69

This process is a typical example of a large process, having the following


important characteristics: in theory, from one major input signal (qc) to a major output
signal (qabv) there are complicated relations, but by interpreting some typical responses these can
be approximated using transfer functions that only have a few dominant poles and a large
number of residual poles which can be neglected; in this case with up to 2 dominant poles and
at least 6 to 10 residual poles these approximations of the transfer functions are justified.
This observation is viable especially in the case of inertial processes, where the first
two derivatives are significant, while the other derivatives are heavily damped, thus justifying
that they can be neglected.
(q – flow rate;  - temperature; h –level; PF – differential pressure; O2 – concentration;
P – pressure
H26 – transfer function of the boiler having as input qaer and as output qabv)

5. Controller tuning of the main control loops

R11
A. Processed Cip
steam pressure adjustment
R1 R12 12 E11
Pabv
*
i pabv a11 ci11 a12 ci12 c qCH4 k11  sTm11
1
PI PI i
The control loop is a tracking system with respectSPR
p12
to the referenced )  1T11having
H11 ( scommand, s
a fast behaviour because of the low inertia ofΔp
- the pneumatic servomotor and the valve(SPR).
CΔpi12
- iqCH4 Er11
i
r11 Δp
A
KM12 P
Page 26 i
p
Cpi Figure 6
Industrial Plant Control Student:

cip12 — unified current converter for differential pressure


SPR —Pneumatic Servomotor with a Valve
cpi12 — unified current-pressure convertor which ensures the external reaction
(the main on in the steady state)
i*pabv — unified reference current for processed steam stabilization
Er12 — radical extractor

The internal control loop has a converter Δp i12 and because the differential
pressure used as a output signal from the calibrated diaphragm is proportional to the
square or the flow, we need a square-root extractor.
Calibrated diaphrahm :

p1 > p2 (Δp=p1-p2); q2=k·Δp;

Note: - c i11 and ci12 are unified current control and cp12 is a control signal in
unified pressure.

The composing transfer functions of the control loop above


a) Current-pressure converter C ip12

- is considered non-inertial, with the proportionality factor resulting from the static
characteristic of the generator
p[bar]

1 (1  0.2) bar   bar   bar 


k Cip12   0.1   H Cip12 ( s)  k Cip12  0.1 
(10  2) mA  mA   mA 

0,2

2 10 i[mA] Page 27

Figure 7
Industrial Plant Control Student:

b) SPR pneumatic actuator

- is considered to have a liniar characteristic, but the inertia revealed by the1s time
constant can't be neglected
cp12[bar]
qCH4[Nm3/h] 0.8
qCH4nom

qc[Nm3/h] t[sec]
qCH4nom TE11

cp11[bar]
t[sec]
0,2 1 [ ] 126
Figure 8 Figura e9: Step response of SPR

 Nm 3   Nm 3 
qcnom   34332.2352
 h   h   Nm 3 
k E12    42915.294 ;
(1  0.2) bar  0.8[bar ]  h  bar 
TE12  1[sec];
L{qc } k E12 42915.294
H E12 ( s )   
L{C p12 } 1  TE12 s 1 s
c) The internal feedback circuit

- includes differential pressure-flow transducer, differential pressure-current converter,


the radical extractor which cancels the quadratic effect of the transducer. All these
iqCH4considered
elements are [mA] to be overall non-inertial.
- the coefficient of proportionality in the figure bellow:
10

(10  2) mA 8[mA]  A  h 


k M 12    2.3301 10  4  3 
 Nm 
3
 Nm 
3
 Nm 
2 q cnom   34332.2352 
 h   h 
Page 28
qCH4nom qCH4[Nm3/h]
Figure 10
Industrial Plant Control Student:

The internal control loop has the following form:

R12 Cip12 E12


ci11 a12 ci12 cp12 K E12 qc
HR12 Kcip11 H E12 ( s) 
1  TE12  s
-
iqc

KM12

Er12 + CΔpi12
This system being stabilizer and of small inertia, the controller R 12 will be tuned using
the modulus criterion, with the dead time constant neglected (no dead time Tm).
1 1
H R12 ( s )   ;
2T12 s (1  T12 s ) H EX 12 ( s )
K cip12  K E12  K M 12 1 K Ex12
H EX 12 ( s )   ;
1  TE12 s 1  T12 s (1  TE12 s )(1  T12 s )

TΣ – sum of time constants that can't be compensated, comparable in magnitude with


TR12 in HR12 (here);

TΣ12 = 10-2 [sec] = 0.01 [sec] si TE12 = 1 [sec]; KEX12=1;

1 (1  T12 s )(1  TE12 s ) 1  TE12 s


H R12 ( s )    ;
2T12 s (1  T12 s ) 1 2  T12 s
TE12  s 1 1 1 1 1
H R12 ( s )      50   VR12  ;
2T12 s 2T12 s 0.02 0.02s 0.02s Ti12  s

VR=50; Ti=0.02 [sec]; → reg. P.I.

d) External feedback circuit

Page 29
Industrial Plant Control Student:

R11
11

i* ++ a c 1 q K11 K Tm11s p
pabv 11
PI 11 CH4 H1111(s()s)1 T s  11  Tm11abvs
H
-- 2T2 s 2  2T s  1 111 T s
11

r
11
i
i
p
p
Cpi11
Cpi11
p

Page 30
Industrial Plant Control Student:

Page 31
Industrial Plant Control Student:

The external loop of the cascade control structure has a quite big time constant: Tm11=10
min. So, the controller tuning R 11 will be made according to slow processes using Ziegler-
Nichols method.

1 / K M 12 K11 K K
H EX 11 ( s )   Tm11s K cpi11  11 cpi11  Tm11 s
1  2T s  2T s 1  T11s
2 2
1  T11s
P1 i[mA]
The proportionality constant KCpi11
can be deduced from the following
10
figure:

T = 100 min (T << T );


11 Σ12 11
T = 10 min (T << T );
m11 Σ12 m11
2

(10  2)[mA] 8[mA]  mA  Pabvnom Pabv [bar]


K cpi11    0.08889 
pabvnom [bar ] 90[bar ]  bar 

Convertor design:
i0=2mA
i
P k + +

(10  2)[mA] 8[ mA]  mA 


K    0.08889 ;
pabvnom [bar ] 90[bar ]  bar 
i  i0  K  p  2[ mA]  8[ mA]  10[ mA];

Using the Ziegler-Nichols criterion we obtain: VR11  0.8 T11  0.8 6000[sec]  8;
Tm11 600[sec]
T = 3T = 30 [min] = 1800 [sec];
i11 m11
1
H R11 ( s )  8  ;  reg .P.I .
1800 s

B.Controller computation for the control loop of the


processed steam

Page 32
Industrial Plant Control Student:

ih*  unified reference current of the water level in the drum(h)

The condensed water at boiling temperature in the drum

a) Current-pressure converter Cip43

- it is considered non-inertial, with the proportionality factor resulting from


the characteristic shown in the figure below:
p[bar]

(1  0.2)[bar ]  bar 
K Cpi 43   0.1 
(10  2)[mA]  mA 
 bar 
H Cpi 43 ( s )  K Cpi 43  0.1 ;
0,2  mA 

2 10 i[mA]
b) Pneumatic servomotor with SPR control valve
Figure
- it is considered that it has a linear characteristic, with the proportionality coefficient
resulting from the figure below:

qapa[to/h] cp43[bar]

Page 33
qapa[to/h] t[sec]
qapanom t[sec]
Industrial Plant Control Student:

qapanom
0
TE43

cp43[bar]
0 0
0,2 1

 tona   tona 
q apanom   389.0987
K E 43   h   h   1945.493 tona ;
(1  0.2)[bar ] 0.8[bar[  h  bar 
TE 43  1[sec];
L{q apa } K E 43 1.945493
H E 43 ( s )    ;
L{C p 43 } 1  TE 43  s 1 s

Page 34
Industrial Plant Control Student:

c) Internal reaction circuit

- it contains the flow-differential pressure transducer, the differential pressure-current


converter and the square root extractor – whose effect compensates the squared effect
of the transducer; all these elements are considered non-inertial, having the
proportionality coefficient:

(10  2)[mA] 8[mA]  mA  h 


K M 43    205.603 10-4  ;
 tona   tona   tona 
qapanom  389.0987
 h   h 
The coefficient resulted from the figure below:
P1 iqapa[mA]

10

d) External reaction circuit

- it contains the same elements as the


internal loop, having:
2

(10  2)[mA]  4  mAh 


qapanom qapa[to/h]
K M 41   K M 43  205.60310  ;
 tona   tona 
qabvnom  
h
Tm46  720[sec];
The process is slow, with a time constant
T  7200[sec]; K 46  1;
T46=7200
46 [sec] =2 [h] and it has a considerable time
delay of Tm46=720 [sec] =12 [min]. Because of the high delay
introduced
1 Tm46s 1 720s L{qabv }
in the process, the controller will act on the water flow
fixed part is: H 46 (s)     
1  T46  s  720035s
1Page L{qapa }
Industrial Plant Control Student:

through the internal reaction loop, at its tuning the effect of the external reaction
loop being neglected.
Because the internal loop has no time delay, the controller can be tuned using
the modulus criterion as follows:

43 ( s )  kCip 43  H E 43  k M 43  kCip 43  1TE 43s  k M 43 ;


' k E 43
H Ext
4
43 ( s )  0.1  1 s  0.0205603 
'
H Ext 1.945493
;
1 s
It is added to this a time constant T Σ43=0.01 [sec], considered to be the sum of the
essentially uncompensated time constants.

1 1 1
H Ext 43 ( s )  H ' Ext 43 ( s )    ;
1  T 43 s 1  s 1  0.01s
1 1
H R 43 ( s )  H Ext 43 ( s)   
2T s(1  T s) 0.02 s (1  0.01s)
1 (1  s )(1  0.01s) 1 s 1 s
 H R 43 ( s)     50 ;
0.02 s (1  0.01s) 1 0.02  s s

Thus, it is obtained a PI controller with VR=50 and Ti=1[sec].

C. Control of the frequency and active power

Figure : Control system

There are two main control loops:


- control relative to the active power
- control which provides correction statism

Page 36
Industrial Plant Control Student:

Remarks:

RP - active power regulator


SMH - hydraulic actuator
V - partial inlet valve steam turbine
T - turbine
G - synchronous generator with a pair of poles under triphase regime
U/f - frequency-voltage converter
Uf - unified voltage (OV OM) proportional to frequency
Uf * - frequency reference voltage
KS - element of statism

Since this control system enter in the category of fast processes, unified
current signals are replaced by unified voltage signals. Unified current is used to
slow processes for which distances are large and therefore suitable for induction
noise.
Since the unified power generators have high output resistance which is
associated to negative feedback of strong current, there is almost a complete
desensitization to noise inductions and to changes of equivalent load resistance.
The presence of a false null desensitizes more noise influence.
Unified voltage is suitable for fast processes for which circuit’s length is short.
It is suitable generating constant voltage supply with low output resistance
because they are cheaper and more beneficial.

Uf *, Up * - frequency reference voltage, active power


Uf - voltage proportional to the frequency response uf
Us - output voltage of statism element Ks
a - deviation control between Up and Up* which is voltage proportional to the
active power
C
qabv - live steam flow control

m
– actuator signal of live steam flow: is the angular displacement that develops
qabv
hydraulic camshaft and which opens gradually by turning the n partial steam inlet
valves.
The deviation between UP and UP* is considered an input signal for the RP
controller, resulting in the command signal Cqabv for the hydraulic actuator SMH,
which through the valve operates the turbine T and the generator G.

Page 37
Industrial Plant Control Student:

The deviation between UP and UP* represents the deviation from the
regulation without the statism correction.
It is the main loop the one for which the weight of the statism correction is
much smaller (about 10%) assuring the load of active power group, proportional
with the rated power of the group, in the hypothesis of a deviation of the 50Hz
frequency, in fault regime appears.
The statism correction is done through the deviation between UF and UF* and
of the statism element Ks.
We are interested in determining and interpreting the transfer functions of the
elements’ components.

a) Hydraulic Actuator SMH

The SMH sketch is presented in the figure below:

The valve (drawer) S fed at Pal pressure (between 6-100 atm) seals completely the
two inlets, towards the P piston, only if the valve (small drawer) S is in a central
position. As a consequence, the P piston is not moved and the output signal (axial
displacement ‘u’) is zero.
The opposing spring R is calibrated according to the position of a tension screw. We
have to assure an axial position of the hydraulic reaction element formed of the case
C and interior piston F. There is no axial connection between the points O3 and O,
only by intermediary action, by varying the oil pressure above and beyond the
piston P.
The input signal is the deviation (error) ‘a’, and the output signal is the
command signal ‘u’ which acts directly on the valve. If the deviation ’a’ increases,
the L1L2 lever starts to rotate around O3 , ensuring the displacement ’b↑’ and also
Page 38
Industrial Plant Control Student:

the oil intake in the upper part of the piston P through the valve (small drawer) S,
which will be driven downward( ’u’ decreases). There will appear a reaction through
L3L4, which starts to rotate round point O and finally resulting a downward
displacement of L3L4 and ’v’. Thus descend the housing C too, resulting a small
overpressure at the top of the piston F. This overpressure causes a downward
movement of O3 through 'w↓'. Thus descends O2, resulting 'r↓'.
Since ‘b↑' and '↓r' are antagonistic, there appears a negative hydraulic
reaction, which results some advantages:
- decreases the mechanical equivalent inertia
- improves linearity, by damping the attenuation of some mechanical disturbances
- increases the stability, by damping the attenuation of some possible oscillations.

Since the oil is incompressible, and taking into consideration the advantages
of the negative hydraulic reaction, will result a low inertial valve-piston system,
which ensures very high forces and torques. All these should present very
complicated and inefficient solutions in the case of the electric motors.
There can be written the following equations:
L2
(1) b  K1  a where K1  ;
L1  L2
( 2) c  b  r;
du
(3)  K  c;
dt
L3
( 4) v  K2 u where K2  ;
L3  L4
d
(5) F (v  w)  R  w;
dt
L1
(6) r  K3  w where K3  ;
L1  L2

Equation (3) represents the speed of the piston P, which is proportional both
to the intake surface of the valve (small drawer) and to the feeding pressure Pal.
Equation (5) represents the balance between force developed by the
hydraulic damper F and the antagonistic force caused by spring R.

Finally after applying for the equations (l)÷(6) the Laplace transform will
result:

Page 39
Industrial Plant Control Student:

(1) L{b}  K1  L{a}

( 2) c  L{b  r};

(3) s  L{u}  K  L{c}

( 4) v  K2  u

(5) F  s  L{v  w}  R  L{w}

(6) L{r}  K 3  L{w}

Eliminating the intermediary variables L{b}, L{c}, L{u}, L{w}, L{r} we get
the transfer function of the valve-piston system:

1
Kp 
L{u} Ti  s
 unde
L{a} 1 T  s
F  K  K1 R  K  K2  K3  F F
Kp  ; Ti  ; T ;
R  K  K2  K3  F F  K  K1 R  K  K2  K3  F

F - the viscosity coefficient of the hydraulic damper


Because K is very big, the T constant usually has values that are a lot
L{u}
smaller than 1 second, so it can be neglected. We will get then L{a} with
PI behavior, without self inertia.
1
 H SMH ( s )  K SMH  ;
TiSMH  s
m qabvnom 0.1[ m] m
K SMH    0.01 
c qabvnom 10[V ] V  Cqabv[V]

where mq abv nom – the displacement of the SMH rod at extremes.


Cqabvnom
TiSMH  20[sec] 
1
 H SMH ( s )  0.01  ;
20  s
0 t[sec]

mqabv[m] mqabv[m]

mqabvnom
Page 40

Cqabvnom Cqabv[V] 0 t[sec]


Industrial Plant Control Student:

b) Valve V
 tona   tona 
q abvnom   389.0987 
L{q abv }  h    h   3890.987  tona 
H V ( s)   KV   m  h 
L{m qabv } m qabvnom[ m] 0.1[ m]
c) Turbine + generator T+G

L{PG } K T G
H T G ( s)   ;
L{q abv } 1  TT G  s
TT G  10[sec];
PGnom [ MW ] 122[ MW ]  MW  h 
K T G    0.3135 
 tona 
q abvnom  389 .0987
 tona   tona 
 h 
 h 
0.3135
 H T G ( s ) 
1  10  s
d) Active power feedback loop and statism correction

1. For the active power-voltage transducer

L{U P } 10[V ]  V 
Kp    0.0820  ;
L{P} 122[ MW ]  MW 

Page 41
2. The transfer coefficient on the statism correction circuit

L{U f } 10[V ] V 
KF    0.2  ;
L{ f } 50[ Hz ]  Hz 

3. The coefficient of the statism reaction – KS


The KS coefficient results from the desired behavior in stationary regime.
Generally f(P) is considered to be linearized. If an accidentally
decrease of the energy system’s frequency with (-∆f) happens, the result is
expected to be a certain rising contribution of the active power (+∆P),
proportional with the nominal power of the assembly, thus resulting an effect of
partial bringing back of the frequency to the nominal value.
df
The statism S is defined as: S = ;
dP
Let PS be the active power consumed in system. If for example the
accidentally decrease (-∆f) appears, it results the automatic loading through the
statism correction loop of the assembly from PSO at PS with effect of frequency’s
bringing back at the nominal frequency, as it can be seen in the following figure:
f[Hz]

f0 f= f0 - Δf
∆f
f

Ps    U f   ( f 0  f ) 
*

U f  f 0  f
*

P
S0 ΔP
0 PS0 ∆P PS P[W]
The next relations can be written:
U S  K S  (U *f  U f )  K S  (U *f  K f  f )
PS  K SP  U S  K SP  K S  (U *f  K f  f )  Ps  statism correction power
where KSP is the statism correction power related to the
reaction
f  f 0  f  PS    U *f   ( f 0  f )    U *f    f 0    f  PS 0  P

where   K SP  K S  K F si   K SP  K S
So for a decrease of (-Δf) corresponds from the power group a contribution
∆P = μΔf.
It requires this statism under the form:
f[Hz]
df f 1 f
S    S 
dP   f  P
0.4[ Hz ]  Hz  f0
S   3.3  10 -3   -∆f
Pnom [ MW ]  MW 
f

0 PS0 ∆P PS
where:
P[W]
-0.4 is the maximum frequency deviation around nominal frequency;
-S is the ratio between the power system frequency variation at the variation of
the group power from "0" to Pnom;

d) The equivalent control of frequency and active power


If the power supplied by the group is zero (PG = 0), the power system frequency
is f0. Under the same consumption conditions, by progressively increasing PG
shows an approximately linear increase in frequency. If the group is charged to
a power PG, but the load shows fluctuations due to statism, , the group will
auto-charge / discharges as the frequency decreases or increases.
f = f + ·P where = 3.3 10-3[Hz/MW]
0 G
As a result, a power delivered in the system PG = 122 [MW] will attract an
increase in the system 3.3 10-3 [Hz]

uf* -uf 0,2

Uf*- Uf f/Uf
fo
KS
SMH T+G
uS V
*
uP 0,01 
1 0,336 PG + f
10703.93 α=1.11 10-3
RP 20 s 1  10 s
- af Cqabv mqabv qabv

Up 0.0277

F/Uf

Giving the controller active power RP is made after the module criterion, the equivalent
system is the closest to a fast system, being imposed a frequency stabilization behaviour.

H Ex1 (s)  H SMH (s)  H V  H T G (s)


1 0.3135 (0.2 s  1)
Deci, H Ex1 (s)  (0.01  )  3890.987   61 
20 s 1  10 s s (10 s  1)

The direct path circuit has the transfer function without regulator HEX, the
whole loop together with the active power-voltage converter will have transfer
function:

1  0.2 s 0.2 s  1
H d ( s )  H Ex1 ( s )  K P  61  0.0820  744
s (1  10s ) s  (1  10s )
1 744(0.2 s  1)
H Ex ( s )  H d ( s )   ;
1  T  s s (10s  1)(0.01s  1)
where TΣ = 0.01 [sec] and represents the sum of the mainly irretrievable time
constants.
1 1
H RP ( s )   
2  T  s  (1  T  s ) H Ex ( s )
1 1 s (1  10s )(1  0.02s ) 10s  1
    0.0672
2  0.01s (1  0.01s ) 744 1  0.2 s 0.2 s  1

e) Analogical modeling of the controller


The transfer function of the regulator RP can be modeled analog with 2
operational amplifier connected in series, as it is shown in the figure below:

R2 C2 R4

R1 C1
- R3 -
af A01
+ A02
+ af
R R

HA01 HA02

The properties of operational amplifiers:


- in open circuit the amplification factor is very high (at least 10 4)
- it has 2 input terminals: one is inverting (usually used) and another one is non
inverting, that is associated with positive feedback
- the input impedance is very high (at least 1 MΩ)
- the output impedance is very low (1÷10Ω) – ideal generator at output
- the equivalent time constant is negligible (Max 10 -4)
L{Vo }  Z R
H AO ( s )  
L{Vi } Zi
R is calibrated in the way that the equivalent resistive bridge has to be
balanced in steady state.
Consider that C1  C 2  C
1
R2 
C2  s R 1  T2  s
H AOech ( s )    (1)  4  K
1 R3 1  T1  s
R1 
C1  s
R4
unde K ; T1  R1C ; T2  R2 C ;
R3

The first amplifier brings in a pole and a zero. The second one is an
inverting amplifier. Using identification: ( HRP(S) = HAOech(s)) is obtained.

For AO1:
The value of capacitor C is chosen from the catalog:

Resistance R1 can be obtained by connecting in series two resistances of 10 kΩ or by a


10 kΩ resistance and a potentiometer of 1 - 100 kΩ.

For AO2:

The values of resistances R3 and R4 are chosen from the catalog:

f) Computetion of the statism correction loop:

Coefficient KS is calculated from the condition that the frequency range ±0.2Hz, i.e.
Δf=0.4Hz, around the nominal frequency of 50 Hz, can be ensured by the nominal power of
the group.

f [ Hz ] 0.4  Hz 
S   3.278  10-3  
PGnom [ MW ] 122  MW 

The increase of nominal power PGnom has the effect of decreasing the frequency f, as
shown in the figure below:
f[Hz]
50,2

50

49,8

0
0.5P P P [MW]
Gnom Gnom G

Power transducer has the coefficient 0.2[V/Hz].


In steady state regime:
PG 
1
10
 PGnom  (U P*  U S ) 
10
1

 PGnom  U P*  K S  (U *f  0.2  f ) 
1
PG   PGnom  (U P*  K S  U *f )  0.02  PGnom  K S  f
10
1 1 
 f    PGnom  (U P*  K S  U *f )  PG 
0.02  PGnom  K S 10 
df 1 1 1
S   KS     125.026  125
dPG 0.02  PGnom  K S 0.02  PGnom  S 0.02  122  3.278  10 3
Remark: (1/10)*PGnom shows the reserved power adjustable with frequency,
imposed by the dispatcher based on the group performance.

From the above modification, it results the following figure:

U* - 0,2
f

KS
U
S
+ a 36 P + f f
f G 0
U * +
0,02s  1 1.1 10
-3
P

D. Reactive power and voltage control


It implements a triple-cascade control, with respect to the voltage across the generator’s
terminals, excitation current, excitation voltage.

u * a c a c a c α
G UG RUG iE iE RiE UE UE RUE CCG P
_ _ _ C Upc
Ex1
UUEx = KUEUE Mu UE

Ex2
UiEr=KiE·iE
Mi IE
UG
G
- IG
On the above figure, the following notations are used:
UGr=KUG·UG-KiG·iG = +

RU – generator terminal voltage regulator
G
RI – excitation current regulator
E
RU – excitation voltage regulator
E
M , M – measurement transducers+ current, voltage adaptors
u i
c – excitation current’s command
iE
c – excitation voltage’s command
uE

We have a cascade of three controllers, with respect to the signals: u , i si u .


E E G
Generator’s voltage deviation, a is the difference between the reference voltage u *
uG G
and reaction voltage u =K · u - K · i , where K and K are the weighting
Gr uG G iG G uG iG
coefficients of voltage and current reactions. The deviation a is applied to the
uG
controller at generator’s terminals resulting the command signal: c . The deviation
iE
between this signal and the reaction voltage u , is proportional with the excitation
iEr
current and it is applied to the current excitation controller RI . Deviation between c
E uE
and the voltage proportional with the excitation voltage u , is applied to the excitation
uEr
voltage controller, RU . The complex control on the grid CCG, generates an impulse at
E
the output, having the phase α which controls the bridge’s power, PC resulting the
voltage U .
PC

a)
Main component elements

1. Synchronous generator

Apparent nominal power S = 147.8788 [MWA]


Gnom

Power factor cos=0.825


Voltage at generator terminals UGnom=15 [KV]

S Gnom 147.8788 10 6


I Gnom    5.6919[kA]  5691.9[ A]
3  U Gnom 3 15 10 3
PGnom  S G  cos   147.878  0.825  122[ MW ]
2 2
QGnom  S Gnom  PGnom  21868.26  14884  83.5711[ MVAr ]

2. EE Exciter

iPC IE
R S T
Ex=EE

LEE
REE UPC UE IE
GEE G EX1 EX2
UPC ~
UE

n=cst.

PEnom  0.25%  PGnom  0.0025 122  0.3[ MW ]


U Enom  300[V ]
PEnom [ MW ] 0.3 10 6
I Enom    1000[ A]
U Enom[V ] 300
3. PC Controlled bridge
UPC[V]
It supplies operating winding of the exciter with the voltage U , current i and
PC PC
power P : 270
PC

0 90 α[˚]
PPCnom  1%  PEnom  0.01  0.3  10 6  3[kW ]

It is considered U = 270 [V] for which


PCnom

PPCnom [W ] 3 10 3
iPCnom    11 .11[ A]
U PCnom[V ] 270

It is considered: U PC  270  270   .


90

This way we obtained, through approximation, a linear dependence between U and α.


PC
The α angle, through an adapter, belongs to 0˚ and 90˚.

4. Complex grid control

Cf

UEE = UPC

REE, LEE

In circuit analysis, the dead time of this block will be neglected: T ≈0;
mCCG
R S T
The voltage-pahse conversion factor is defined: K CCG     90  11 .25  
α [˚] c 8V V 
This ratio corresponds to a maximum variation of the phase shift control pulses of
180variation of the voltage into CCG achieved by
current to the grid which for a 8V
designing a maximum phase shift of 900.
90

-8 0 8 c[V]
The control voltage of the regulator denoted by „c” is between –8V and 8V.
The condition is ensured:
c=0 => α =90˚

90
  90   c  90  11 .25  c
8V
U PC  270  3    270  3  (90  11 .25  c )  33.75  c
In general :
U PCnom U
U PC  U PCnom     U PCnom  PCnom  (90  11 .25  c )
90 90
U PCnom
 U PC   11 .25  c
90
For U PCnom  270[V ]  U PC  33.75  c

L{U*G}The
L{a }
control L{ciE}looks
scheme L{aiElike
} L{cUE} below:
the figure L{ aUE } L{c} L{ UPC }
UG
HRIE
K EE
HRUG RUE 33.75
1  TEE s
- - - KUE LUE
L{UUEr}
KE
1  TE s
L{UiEr}
KiE
LIE
KEG
LUG
KG LIG
1  TE s
+ -
~
=
L{UGr}
b) The calculation of the transfer functions of the component elements

1. The transfer function of the excitation of the excitation circuit EE

L{U E } K EE
H EE ( s )  
L{U PC } 1  TEE s

 diPC
U
 PC EE L   REE  iPC
 dt
U E  K EE  iPC

We apply the Laplace transform for zero initial conditions:

U PC (s)  (LEE  s  REE )  L{iPC }



U E (s)  K EE '  L{iPC }
LEE
TEE   0.1[sec];
REE
U Enom[V ] 300[V ] 1.11
K EE    1.11  H EE ( s ) 
U PCnom[V ] 270[V ] 1  0.1  s
2. The transfer function of generator excitation

L{i E } KE
H E (s)  
L{u E } 1  TE s
i Enom [ A] 1000[ A]  A
KE    3.33 
U Enom [V ] 300[V ] V 
3.33
TE ( s)  0.5[sec]  H E ( s) 
1  0.5s

3.The transfer function of generator


L{iG }
H EG ( s )   K EG
L{uG }
U Gnom [V ] 15  10 3 [V ] V 
K EG    15 
iEnom [ A] 1000[ A]  A
iGnom [ A] 5.6919  10 3 [ A]  A
KG    0.3795 
U Gnom [V ] 15  10 [V ]
3
V 
TG ( s )  4[sec]
0.3795
 H G (s) 
1 4 s

4. Transfer function of compoundation reaction

- KUG – is chosen in hypothesis iG=0 for which:

U Gr  KUG U G

 * 10 3
U G  10[V ]  KUG   0. 66 
10
 15  10 3

U Gr  U G*
0.5[V ] 0.5[V ] V 
- KiG – is chosen in hypothesis: K iG    8.78  10 5  
iGnom [ A] 5691.9[ A]  A

- KiE – is chosen in hypothesis:


U iEr [V ] 10[V ] V
U iEr  10[V ]; si K iE    0.01 
L{U G}
*
L{aUG} L{ciE} L{ai Enom
} [V ] 1000
L{cUE[}V ]L{ aUE }  A  L{c} L{ UPC }
iE
HRIE
K EE
HRUG R 33.75
- KUE – is chosen in hypothesis:
UE
1  TEE s
- - U UEr [V ] -
10[V ] V
K =0.033 LUE
U UEr  10[V ]; si K UE   L{U  0.033
} 
UE
 A
U Enom [V ] 300[V ] UEr V
KE
The following block diagram is resulted: 1  TE s
L{UiEr} KiE =0.01
LIE
B
KEG
LUG
KG LIG
1  TE s

KUG =0.66*10-3 C

+
L{UGr}
KiG =8.75*10-5
-
Regulators adjustment will be made based on the module criterium, because of
the perturbations which occur and are of the step type and the desired effect is the
stabilization.

a) Internal loop A
K EE 1.11 1.2362
H ' Ex ( s )  K CCG PC   KUE  33.75   0.033 
1  TEE  s 1  0.1  s 0.1  s  1
1
H ExA ( s )  H ' Ex ( s )  where T  10 2 [sec];
1  T  s
1 1 1 (1  0.1  s )  (1  0.1  s )
 H RUE ( s )    
2T s (1  T  s ) H ExA ( s ) 0.02  s  (1  0.01  s ) 1.2362
(0.1  s + 1) 1 1
H RUE ( s )   4.0453   4.045 
0.02472 s 0.02475  s 0.025  s

VRUE  4.045  4
A PI controller such that: 
TiRUE  0.025[sec]

a) Internal loop B
The transfer function of the interior loop is:
L{U E } H RUE ( s )  K CCG PC  H EE ( s )
H1 ( s)  
L{c IE } 1  H RUE ( s )  K CCG PC  H EE ( s )  KUE
1  0.1s 1.11s 37.46
 33.75 
H1 ( s)  0.03296s 1  0.1s  0.03296s
1  0.1s 1.11s 33.75  1.11  0.033
1  33.75   0.033 1 
0.03296s 1  0.1s 0.03296s
27.3014
H1 ( s) 
0.026s  1

L{U*G} L{aUG} L{ciE} L{aiE} L{cUE} L{ UE } L{IE} L{ uG } L{ iG }

27.3 3.33 0.3795


HRUG HRIE 15
1  0.026s 1  0.5s 1  4s
- -

B
L{UiEr} KiE =0.01

C
KUG =0.66*10-3

+
L{UGr} KiG =8.75*10-5
-

KE 27.3014 3.33 0.9091


H '' Ex ( s )  H1 ( s )   K IE    0.01 
1  TE  s 1  0.026 s 1  0.5s (1  0.026 s )  (1  0.5s )
1
H ExB ( s )  H '' Ex ( s )  unde T  10 2 [sec];
1  T  s
0.9091
H ExB ( s ) 
(1  0.026 s)  (1  0.5s )  (1  0.01s )
L{cUE } 1 1
 H RIE ( s )   
L{aiE } 2T s (1  T  s ) H ExB ( s )
1 (1  0.02s )(1  0.01s )(1  0.5s )
H RIE ( s )  
0.02s (1  0.01s ) 0.9091
(1  0.02s )(1  0.5s ) 1
H RIE ( s )    28.6  0.55s
0.01818s 0.01818s
 1 
H RIE ( s )  28.6  0.019 s 
 0.519s 
which represents a PID controller with the following parameters:

VRIE  29

TiRIE  0.519[sec]  0.52[sec]
T  0.019[sec]  0.02[sec]
 dRIE

c) Internal loop C

The transfer function of the internal loop H2(s) from the simplified block scheme from
the figure is:

L{iE } H RIE ( s )  H 1 ( s )  H E ( s )
H 2 (s)  
L{cIE } 1  H RIE ( s )  H 1  H E ( s )  K IE
(1  0.02 s )(1  0.5s ) 27.3014 3.33
 
H 2 (s)  0.01818s 0.026 s  1 1  0.5s 
(1  0.02 s )(1  0.5s ) 27.3014 3.33
1    0.01
0.01818s 0.026 s  1 1  0.5s
100
H 2 (s)  `
1  0.0199  s

The block scheme will have the following structure:

L{U*G} L{aUG} L{ciE} 1.11 L{u G} 0.3795 L{iG}


HRUG 1  0.1s
15
1  4s
-
KUG =0.66*10-3 C

+
L{UGr}
KiG =8.75*10-5
-

The scheme is transfigured:


0.66  10 3  3.32  10 5  2.64  10 3  s 0.62683 (1  4.21s )
H reactie ( s )  
1  4s 1  4s

100 1500
H direct ( s )   15 
1  0.02  s 1  0.02  s

L{U*G} L{aUG} L{ciE} K EE L{uG}


HRUG
- 1  TEE s

L{UGr} 1  4.21s
0.6268  10 3 
1  4s

1500  0.63  103  (1  4.21  s) 1


H Ex ( s )   with T  0,01[sec]
(1  4s )  (1  0,02  s ) (1  T  s )
0.945  (1  4.21  s )
 H Ex ( s ) 
(1  4 s )  (1  0,02  s )  (1  0,01  s )

L{ciE} 1 1
HRUG s)(    
(1  4 s )(1  0.02s )(1  0.01s )

1

L{aUG} 2T s (T s 1) HEx s)(


0.945  (1  4.21s ) 0.02s (1  0.01s )

52.91  (1  4 s )  (1  0,02 s )
 
s (1  4.21s )
1. If it is considered 1  4 s  1  4.21s  a PI controller :
(1  0,02  s) 1
H RU G ( s)  52.91   1.05(1  )
s 0,02  s
 VR RU G  1.05

Ti RU G  0,02[sec]

2. Thus, results a PID controller with a first order filter:


1 1
4.02   0,08  s 1  0.02 s
1  4.02 s  0.08s2
s 4.02 s
H RU G ( s )  52.91   52.91   212.7 
s (1  4.21s ) 1  4.21s 1  4.21s
 VR RUG  212.7
T  4.02[sec]

  i RUG
Td RUG  0,02[sec]
 TN  4.21[sec]

The model of the RU G controller:

Fig. 45
1
R2  1
C2  s R4 
H RU G ( s )  H AO1 ( s )  H AO2 ( s ) , where H AO1 ( s )   şi H ( s )   C4  s
1
R1  AO 2
R3
C1  s
 1   1 
 R2    R4  
 C2  s   C4  s  C1 ( R  C  s  1)  ( R4  C 4  s  1)
 H RU G ( s )     2 2
 1   R3  C 2  C 4  R3 R1  C1  s  1
 R1    
 C1  s   

By identification methods 
4.21
T1  R1  C1  4.21[sec] ; we choose R1  100 K  C1   42F
105
0,02
T2  R2  C 2  0,02[sec] ; we choose R2  100k  C 2   0,2 F
10 5
T4  R4  C4  4.02[sec] ; we choose R4  100 K  C4  40 F

We choose the condensers from the catalog, E6 condensers, with 6 types, between 1 and
10 and which have standardized values, multiple of 1, 2.2, 3.3, 4.7, 6.8, 10.
0.4
In this way we choose C1=4.7µF  R1   85.106 K , but we don’t have a
4.7  10 6
resistance with that value in the catalog so we connect in series R1’= 82KΩ with a
0.02
variable resistance R1”= 4.7KΩ, we choose C2=0.22µF  R2   90.909 K
0.22  10 6
and we obtain R2’=85kΩ connected in series with a variable resistance R2”=10kΩ, we
0.52
choose C4=4.7µF  R4  6
 110 .638 K and we get R4’=110k Ω connected in
4.7  10
series with a variable resistance R4”= 820Ω.
C1 4.7 1
VR   6
 40.22  R3   10 6  R3  4.54k which we
C 2  C 4  R3 0,22  4.7  10  R3 0.22
obtain from R3’=4.3kΩ connected in series with the variable resistance R3”=390Ω.

' "; R ' " ' "


Therefore,
R R R
1 1 1
2  100k ;
R R R R R R
;
3 3 3 4 4 4
C1  4.7 F , C 2  0,22F ; C 4  4.7 F ;

You might also like