See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/228429985

Control system simulator for steam boiler parameters

Article · January 2008

CITATIONS READS

3 2,190

4 authors, including:

Matei Vinatoru Gabriela Canureci

University of Craiova SC IPA SA CIFATT Craiova
48 PUBLICATIONS   65 CITATIONS    22 PUBLICATIONS   21 CITATIONS   

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Matei Vinatoru on 01 June 2014.

The user has requested enhancement of the downloaded file.

 Camelia Maican, Matei Vinatoru,
WSEAS TRANSACTIONS on SYSTEMS and CONTROL Gabriela Canureci, Eugen Iancu

Control System Simulator for Steam Boiler Parameters

CAMELIA MAICAN, MATEI VINATORU, GABRIELA CANURECI, EUGEN IANCU
Department of Automation
University of Craiova
Decebal Str., No. 107, 200440, Craiova
ROMANIA
camelia@automation.ucv.ro, vinatoru@automation.ucv.ro, http://www.automation.ucv.ro

Abstract: - The paper presents the research results related to control of the steam boiler parameters: drum level,
furnace pressure and steam pressure. It was developed the mathematical model of the water-steam system and
furnace part of the boiler. Also, the paper presents the algorithm for real time detection of model parameter
variations. In this context a correction method, validate through simulation was developed.

Key-Words: - steam boiler, steam pressure, furnace pressure.

1 Introduction
Related to the systemic point of view, the boiler is To simplify the procedures for determination of the
frame in the category of the MIMO system (more mathematical model will establish a block scheme
input more output). for the boiler as shown in Fig.1, specifying channels
Boiler
input output

DC DC
P
B B D ΜPt
A A ΜDp

Gp=Gp
B
ΜPA θav
A Dp Mθav
XVG XVG

Winj Winj

Dp
L
We Wa ΜL

B
CO2
A ΜC

Gp
Pf
xVG Ge ΜPf
VG

Fig. 1. The input-output variable for boiler

ISSN: 1991-8763 917 Issue 11, Volume 3, November 2008


WSEAS TRANSACTIONS on SYSTEMS and CONTROL
of input-output.
On the left side are passed the input variables
(variable which can be command directly with the
usual actuators), and in the right the measurable
output variables under which it will develop the
control structures with typical feedback.
B – the fuel flow
A – air
Ge – the evacuated gas
X VG - the command of the gas ventilator
Winj – the injection flow
L – the water level in the drum
Pf – the furnace pressure
VG – the gas ventilator
It is important to choose the adequate input-output
pair, which to assure an efficient control for the
output variables through the input variable, with
maxim effect and without main perturbation for the
other output variables [2].
We will choose as the output variables that
measurable variable how permit the obtaining a
complete information on the state of the boiler in
each moment of time (Fig. 1).
The dependence between input-output variables:
- The steam pressure p in the drum is a variable how
depend on the boiler steam flow Dp and the turbine
steam flow Dc.
- The furnace pressure depending on the resulted gas
flow Gp and the evacuated gas flow Ge.
- The water level in the drum depending on the
water flow, the boiler steam flow Dp and the turbine
steam flow Dc.
- The fuel flow B is an independent variable,
because it isn’t affected by the other variable and it
may be modified with the conveyor.
- The air flow, from the burning process, it must to
has a value how corresponding with fuel flow B.
- The oxygen, oxide and carbon dioxide in the
combustion gases depends on the air / fuel and
therefore may be acting as the flow of air and fuel.
- The boiler steam flow Dp depends on the quantity
of heat boiler developed in the furnace zone,
meaning depends on the fuel flow and the air flow
because the quantity of heat developed depends on
both.
- The turbine steam flow Dc is, in relation to the
boiler, a random variation of variable because the
loading turbine depends on the conditions of
operation of the energetically area.
We will choose the supply water flow for the control
of the level drum, and the fuel flow for the control
of the furnace pressure [3].
For the control of the steam pressure in the drum we
will choose the heat transfer coefficient through
ISSN: 1991-8763
Camelia Maican, Matei Vinatoru,
Gabriela Canureci, Eugen Iancu
radiation Kre, such as will be presents in chapter 2.
2 The automatic control structure of
the steam boiler
Writing the balance thermal equation for burning
process we obtain the next equation:
d (m g H gf )
= A ⋅ C pA ⋅ T A + K comb ⋅ B ⋅ H cn −
dt (1)
− Ge ⋅ C pg ⋅ Tg − K Re (Tm4 − Ts4 )
We make the next assumption [4]:
-The burning gases are treated like a perfect gas.
-The calorific power of fuel it is consider equal with
the net heat and deviation from the ideal burning is
represented by the burning coefficient Kcomb.
-The heat exchange to furnace zone (convection,
conduction and radiation) and the water from pipes,
at the saturation temperature Ts, is equivalent with a
direct heat exchange through radiation. Tm is the
middle temperature of burning gases.
From the low of perfect gases we can approximate:
VC vg Pf
mgH gf = (2)
R
where:
V- the volume used from the gases in the furnace
zone;
Cvg- the specific heat of gases at constant volume;
R- the perfect gases constant;
Pf- the furnace pressure.
Writing the equation (1) depends by the furnace
pressure Pf with Tg and Tm calculated, result:
VC vg dP f
= AC pA T A + K comb BH CN −
R dt (3)
− G e C pg T g − K Re ( T m4 − T s4 )
We consider that in the drum exist two phases
(steam and liquid in equilibrium) and the measured
liquid level don’t give as an exact information about
the existent liquid quantity from the drum and we
must to introduce the supplementary correction
about water flow at the input of the drum.
The balance material equation for water from the
drum is:
dm e
= We − Dc − D s (4)
dt
where:
918 Issue 11, Volume 3, November 2008
 Camelia Maican, Matei Vinatoru,
WSEAS TRANSACTIONS on SYSTEMS and CONTROL Gabriela Canureci, Eugen Iancu

me- the effective liquid mass from the drum; will be modify on the time of functioning of the
Dc- the turbine steam flow from the drum; boiler the next variables:
Ds- the water outflow (through the broken pipes). -the necessary water flow We, controlled by the
measured level;
⎡ dρls dp ⎤ -the heat transfer coefficient through radiation Kre;
⎢[Ve + Vtv + VV (1 − α )] dp dt ⎥ - the burning coefficient Kcomb.
⎢ ⎥ The structures of the program who witch realize the
dme dVe ⎢ dρvs dp ⎥
= (ρls − ρvs ) + ⎢+ (Vlm − Vl + VVα ) ⋅ −⎥ (5) simulation is presented in the Fig. 2 .
dt dt ⎢ dp dt ⎥
We compare the measured values from the real
⎢ dα ⎥ process: the level in the drum, the water flow We,
⎢− VV (ρls − ρvs ) ⎥
⎣ dt ⎦ the pressure in the drum PT and the relative pressure
in furnace, with the measured values from the model
where: of the boiler [1].
Vl- the liquid volume used in the drum; This results permit as to estimate the correction
Vtv- the descending pipes volume; which be done in the model of the boiler to assure
Vlm- the liquid volume measured with a transducer the tuning parameters correction’s of the controllers.
level;
Ve- the volume used from steam in the liquid space;
The correction
Vv- the volume used from steam in the drum;
α - the fraction used from steam in Vv; We Kre Kcomb
ρls , ρvs - the density at saturation for steam and
liquid L
p- the steam pressure in the drum. PI
B + + M B
The balance thermal for the water and steam from -
O - O O
the drum and the boiler pipes: p
I PID D I
⎧ dTs dH dH ⎫ L + E L
+ M1 ls + Mv vs − (Vlm + Vv + Vtv ) +⎪ -
⎪MmCm E L E
⎪ dp dp dp ⎪ Pf
R PI R
⎪⎪ dρls ⎪⎪ dp
⎨ + ( Hvs − H ls ) ⋅ [Vl + Vtv + Vv (1 − α )] + ⎬ = +
⎪ dp ⎪ dt -
⎪ dρvs ⎪
⎪+ (Vlm − Vl + Vvα) ⎪
⎩⎪ dp ⎭⎪
= −Wl (Hvs − Hl ) − Dp (Hvs − Hls ) + (1 + β ) * (6)
* (Hvs − Hls )(Wl − Dc − Dp ) + KRe(Tm4 − Ts4 ) Fig. 2 The simulated model of the boiler

where:
Mm– the metal mass of the drum and pipes; 2.1. The design of the automatic control structure
Cm – the specific heat of the metal; of the drum pressure
Ml – the total liquid mass from drum and pipes;
Mv – the total steam mass from drum; The drum Dc
Hvs, Hls – the steam and liquid content heat at
saturation;
β -the steam fraction who condense in the drum;
The boiler
We can express the measured liquid level depending p
temperature
on the liquid volume measured Vlm. Dp
Өf
γ ⋅l ⋅ R2 1 γ
V lm = − ( R − L ) R sin
2 2 2 Өf
γ L (7)
cos =1− ;
2 R
V v = π ⋅ R 2 ⋅ l − V lm

In the equation (1) to (6) we don’t know exactly and Fig. 3. The drum of the boiler

ISSN: 1991-8763 919 Issue 11, Volume 3, November 2008

 Camelia Maican, Matei Vinatoru,
WSEAS TRANSACTIONS on SYSTEMS and CONTROL Gabriela Canureci, Eugen Iancu

2.1.1.The mathematical model of the drum D P0 = DC 0 = D 0 (10)

pressure
The variation speed of the steam pressure in the For dynamic regime:
drum is proportional with the difference between the
boiler steam flow and the turbine steam flow and dΔp
inverse proportional with the thermal accumulation C a (D 0 ) = ΔD P − ΔDC (11)
dt
capacity of the boiler.
So, we can write the equation: After integration we obtain:
dp DP − DC
=
∫ [ΔD P (τ ) − Δ DC (τ )]dτ
t
Δ p (t ) = Δ p (t0 ) +
(8) 1
(12)
C a (D0 ) t 0
dt Ca

Δ p (t ) = Δ p1 (t ) + Δ p 2 (t )
where Ca is the thermal accumulation capacity.
The numerical value of the coefficient Ca depending (13)
on the next variable:
- The steam flow supplied in the boiler in stationary
Δ p (t ) = (s ) − Δ D C (s ) (14)
K ΔP K ΔP
ΔD P
regime; sT Δ P sT Δ P
- The metal mass of the vaporization system;
- The rotation m ∈ (0,1÷0,8) where:

wa
m = (9) K ΔP =
pN C p
, T Δ P = a 0 N , C a 0 = C a (D 0 )
D max D max D max

where wa is water quantity from the boiler and Dmax According to equation (14) result the bock scheme
is the maximum steam flow. from Fig. 4:
Conform to this specification, result:
ΔDP(s)
dp ( t )
C a (D 0 ) KΔP/sTΔP +
= D P (t ) − D C (t )
dt
Δp(s)
p(t)=p0+Δp(t) -
ΔDC(s)
Dp(t)=DP0+ΔDP(t) KΔP/sTΔP

DC(t)=DC0+ΔDP(t)
Fig. 4. The block scheme for the control of p
For stationary regime:

WL DC
KW
Wp
Tw

P
+ 1 DP
CW EE KB Boiler
CP
Ti ⋅ s
+ Dp

P*
TP

Fig. 5. The block scheme for the automatic control circuit of the drum pressure

ISSN: 1991-8763 920 Issue 11, Volume 3, November 2008

 Camelia Maican, Matei Vinatoru,
WSEAS TRANSACTIONS on SYSTEMS and CONTROL Gabriela Canureci, Eugen Iancu

H'
2.2. The design of the automatic control structure Ot = C ' + + S ' − O' = 0,0183 moli / gram
of the furnace pressure 2

A Ge The necesar theoretical air:

MF
B Ot
M Pf Pf αT = = 0,087
0, 21
X VG M VG The amount of excess air to be introduced:
Gp
Fig. 6. The block scheme for the automatic control α = a ⋅ α t = 2,5 m 3 / kg
circuit of the drum pressure
a=1,3 is the coefficient of excess air.
The total quantity of gas resulting from the
2.2.1. The mathematical model of the furnace combustion complete a gram of fuel is g p .
zone relative to the balance material
The oxidation chemical process of the fossil fuel
g p = 22,4 ⋅ (mco 2 + mh2 o + mn 2 + mo 2 + mso 2 )
(carbon, hydrogen and sulf) consist in the next
relations:
GP = g p ⋅ B
C + O2 = CO2
GP = k1 ⋅ B + k 2 ⋅ A
Q1 = 8140Kcal / kg
C A=α ⋅B
C' = 16,8 ⋅ 10 −3
12 ⋅ 100
k1 = 22,4 ⋅ (C + H + W + N + O + S ) + 5 ⋅ Z inj
2C + O2 = 2CO
k1 = 1,33 ; k 2 = 0,838
Q2 = 2473Kcal / kg
B = 50,83kg / s ; A = 127 ,007 ; G P = 174,008
H
H' = = 9,5 ⋅ 10 −3
2 ⋅ 100 The furnace zone may be treated as a gas container
at constant volume and having behavior such with a
2 H + O2 = 2 H 2 O perfect gas [5].

Pf ⋅ V f = m ⋅ R ⋅ T f
O
O =
'
= 3,5 ⋅ 10 −3
32 ⋅ 100 The material balance equation for the mixture of gas
volume will be:
S + O2 = SO2
dm
S = G p − Ge
S' = = 0,25 ⋅ 10 −3 dt
32 ⋅ 100
where Ge represents the evacuate gas flow.
N
N' = = 0,14 ⋅ 10 −3
28 ⋅ 100 Vf dP f
⋅ = G p − Ge
R ⋅T dt
The theoretical oxygen quantity:
The evacuate gas flow Ge depends on the furnace

ISSN: 1991-8763 921 Issue 11, Volume 3, November 2008

 Camelia Maican, Matei Vinatoru,
WSEAS TRANSACTIONS on SYSTEMS and CONTROL Gabriela Canureci, Eugen Iancu

pressure and the fluidic resistance appearing on the

channel of the evacuate gas and the position C api , Tapi , Fai - the specific heat, the temperature
adjustment flap of the gas ventilator.
and the water flow at the entrance
V – The interior volume of the superheater
Vf dΔPf
⋅ = ΔG p − ΔGe X ab - the steam concentrantion from composition
R ⋅T dt Trough linearization, in stationary regime, we obtain
the next equation:
Ge = k 3 ⋅ Pf + k 4
dΔT f
Vf T3 ⋅ = k 6 ⋅ ΔB2 + k 7 ⋅ ΔTai + k8 ⋅ ΔA − Δ ⋅ T f −
T2 = = 1,05s dt
R ⋅ T f ⋅ k3 − k 9 ⋅ ΔGe + k10 ⋅ ΔTat
dΔTa
T4 ⋅ = k11 ⋅ ΔT f + k12 ⋅ ΔTapi − ΔTat − Δ ⋅ T f −
2.2.2 The mathematical model of the furnace dt
zone in terms of heat balance − k13 ⋅ ΔFat
The thermal balance equation for gas:
The equation balance of flow:
dT f
ρ ⋅V f ⋅ C f ⋅ = Pci ⋅ B2 + ρ ai ⋅ C ai ⋅ Tai ⋅ A −
dt ΔFa = ΔFai
⎛⎛ T f ⎞4 ⎛ T ⎞4 ⎞
− Pf ⋅ C f ⋅ G p − k r ⎜ ⎜⎜ ⎟ −⎜ a ⎟ ⎟ The coefficients T and k in stationary regime are
⎜ ⎝ 100 ⎟⎠ ⎝ 100 ⎠ ⎟ determined by the following amounts:
⎝ ⎠
Pci = 1652 Kcal / Kg
B2 = k 5 ⋅ B1 the fuel flow of entrance and burned
B2 = 0,965x86,112
k 5 = 0,965 ρ ai = 0,672 Kg / m 3
C ai = 1094,6 J / KgK
The water-steam balance equation: Tai = 603K
A = 140,4m 3 / s
(Cab ⋅ X ab ⋅ ρab + Cab (1− X ab )⋅ krap )⋅V ⋅ dTa = Pf = 0,404Kg / m 3
dt
⎛⎛ Tf ⎞ ⎛ T ⎞ ⎞
4 4 T f = 1373K
= kr ⋅ S ⋅ ⎜⎜⎜ ⎟⎟ − ⎜ a ⎟ ⎟ + Capi ⋅Tapi ⋅ Fai − C f = 3146,6 J / KgK
⎜ 100 ⎝100⎠ ⎟
⎝⎝ ⎠ ⎠
G P = 189,2m 3 / s
− (Cap ⋅ X ab + Cab ⋅ (1− X ab ))⋅Ta ⋅ Fa
k r = 8,64
S = 2384 m 2
T f the gas temperature at output of the furnace
Ta = 589,6 K
zone
Pci - the inferior calorifical power of the fuel Tapi = 562,4 K
C ai , Tai , A - the specific heat, the temperature, the X ab = 0,82
entrance air fuel C api = 2136,2 J / KgK
k r - the heat transfer coeficient trough radiation Fapi = 164,1Kg / s
S – the surface of the shield C ab = 4986,6 J / KgK
Ta - the composition temperature from superheater
C P = 3142,5K / KgK
C ab - the specific heat of steam from superheater
The control scheme for furnace pressure is presented
C ap - the specific heat of water from superheater in Fig. 7

ISSN: 1991-8763 922 Issue 11, Volume 3, November 2008

 Camelia Maican, Matei Vinatoru,
WSEAS TRANSACTIONS on SYSTEMS and CONTROL Gabriela Canureci, Eugen Iancu

ΔB
k1
+
ΔA 1/ k3 ΔPf
+ + 1
k2
Δα -
k 3 (1 + T2 s )
k3 / k 4
ΔGe k3
+

k4 +

Fig. 7. The control scheme for furnace pressure

2.3. The design of the automatic control structure We- the command variable for control of the water
of the water level in the drum level in drum (at boilers with drum) or control
vaporization zone at boilers without drum

steam Dp=We , so L= constant,Dp is not dictated by We

L

DP
DC
MDp L
B
ML
W We

Fig. 9. The block scheme for the water level in the

Pomp drum
Fig. 8. The drum of the boiler

Fig. 10. The block scheme for control of the water level

ISSN: 1991-8763 923 Issue 11, Volume 3, November 2008

 Camelia Maican, Matei Vinatoru,
WSEAS TRANSACTIONS on SYSTEMS and CONTROL Gabriela Canureci, Eugen Iancu

2.4. Adjusting the fuel volume
No matter we talk about a natural circulation boiler
or about a forced crossing one, generally the task
adjustment loop follows to maintain constant the
steam pressure at the boiler's output, modifying the
fuel quantity and the air for burning volume.
For the boilers which work on fuel oil, we use a
cascade loop control, which contains a main task
regulator (PID type) and a fuel regulator (PI type) in
following loop.
The stability of the loop may be increased by
introducing in the adjustment the steam volume
signal as a perturbation. It is observed that in this
case, the quantity of coal is represented by the sum
of revolutions of the redler belts.
to
D. steam TURBINE
P - the transfer function for the transporting belt:
The execution elements are revolution variators of
the belts (which may be frequency converters or
mechanical variators for revolution) and the shutter
for the gas volume adjuster.
These are commanded through some proportional
regulators. The main task regulator (PID steam
pressure regulator) and the volume of steam took
over from the turbine establish a reference value for
the fuel quantity (gas + coal), which is then
compared to the sum of gas and coal volumes, and
the obtained signal is introduced in the fuel
regulator which will determine increase or decrease
of the fire power of the boiler (Fig.11).
2.4.1. The mathematical model of the automated
process
The following equations are deduced from the
material evaluation and thermo evaluation:

BM l
Reference
H BT = e −τ b s = , τb = B
pressure
BB vB

PID - the equation of the coal mass:

VM ⋅ C B
Adjustment revolution mB (t ) = ⋅ B2 (t )
air BRn g2

- - + revolution Applying the Laplace transformation, it results

Revolutions bands sum
+
BR2 through simple transformations the transfer function
PI
of the coal mill:

B (s )
la BR2, 3,....
1
HM = =
TM s + 1 BM (s )
P

3 The experimental results

n The experimental results are obtained by simulation
under MATLAB Simulink.
BR1 Variation
The air flow A, from the burning process, it must to
revolution has a value how corresponding with fuel flow B:
band A=k·B, k=2,5.
If we modify the coefficient k, using the control
gaz structure from Fig. 2 we can determinate the
correction which be done in the model of the boiler
D. gaz to assure the tuning parameters correction’s of the
coal
M1
controllers.
In Fig. 13 and Fig. 14 we have the correction’s
Fig.11. The control loop in cascade from boiler with values Kre, Kcomb for k=2,6, respective for k=2,4.
gases and coal

ISSN: 1991-8763 924 Issue 11, Volume 3, November 2008

 Camelia Maican, Matei Vinatoru,
WSEAS TRANSACTIONS on SYSTEMS and CONTROL Gabriela Canureci, Eugen Iancu

1 600

0.5 Kcomb 400

Kre; Kcomb; Wa

H[cm]
0 200

Kre We
-0.5 0

-1 -200
0 5 10 15 0 100 200 300 400
Time[s] Time[s]

Fig. 12. The corrections at A=2,5·B Fig. 15. The water level in the drum

2 3000
Furnac e pres s ure v ariations [m m m H2O]
2500
1.5

2000
Kre; Kcomb

1
Kcomb
1500
0.5
Kre 1000

0
500

-0.5 0
0 5 10 15 0 10 20 30 40 50
Time[s] Time[s]
Fig. 16. The furnace pressure variations
Fig. 13. The corrections at A=2,6·B

0.5 1000
Pressure variations p[m m H2O]

800
0
600
Kre
Kre; Kcomb

-0.5
Kcomb 400
-1
200

-1.5
0

-2 -200
0 5 10 15 0 100 200 300 400 500
Time[s] Time[s]

Fig. 14. The corrections at A=2,4·B Fig. 17. The steam pressure variations in the drum

ISSN: 1991-8763 925 Issue 11, Volume 3, November 2008


WSEAS TRANSACTIONS on SYSTEMS and CONTROL
In Fig. 15, Fig. 16 and Fig.17 is represent the level
in the drum, the furnace pressure variations and
respective the steam pressure variation in the drum.
4 Conclusion
We have realized the simulated model of the boiler
and the control structure for the boiler.
This structure of simulation permit as to estimate the
correction which be done in the model of the boiler
to assure the tuning parameters correction’s of the
controllers.
We intend to realize a program, which can be offer
to users the dedicated software for fault detection
and the solution for the value corrections which be
done in the model of the boiler.
ISSN: 1991-8763
View publication stats
Camelia Maican, Matei Vinatoru,
Gabriela Canureci, Eugen Iancu
References:
[1] G. DiPierro, An on-line method of detecting
parametric faults, M.S. Thesis, George Mason
[2] J. Gertler, Fault detection and diagnosis in
engineering systems, Ed. Marcel Dekker, 1998, New
York.
[3] E. Iancu, M.Vinatoru, Fault detection and
isolation in dynamic systems, Ed. SITECH. Craiova,
1999.
[4] E. Iancu, M.Vinatoru, Analytical method for
fault detection and isolation in dynamic systems
study case”, Ed. Universitaria Craiova, 2003.
[5] C. Vinatoru, M. Vinatoru, E. Iancu, Modal
Control of Distributed Parameters Heat Transfer
Process, International Symposium on System
Theory, Robotics, Computers ans Proces
Informatics, SINTES 10, 25-26 May, pag. A118.
926 Issue 11, Volume 3, November 2008