Mathematical Modelling and Simulation of

Power Plants and CO2 Capture

The practical use of selected models of power

plant objects in various control systems of
pulverized coal-fired drum boilers.

University of Warwick
20-21 March 2012 Mariusz Lipiński
Institute of
Power Systems
Automation Ltd,
Wrocław, Poland
1
 Speech program

1. Inctroduction.
2. Thermal power unit models.
3. Model of a ring–ball coal mill.
4. The practical use of ring–ball coal mill model signals -
detection and prevention system of fire and ignition.
4.1. Identify the loss of coal in the coal mill system.
4.2. Identify ignitions in the coal mill system.

2
 1. IASE Ltd. - about us.

Founded in 1949, the Institute of Power Systems Automation,

IASE Ltd is an independent research and development
organization specializing in automated control and monitoring
systems for the electricity industry.
We supply reliable system for a whole range of applications,
from single operations to providing total integrated solutions.

3
 1. IASE Ltd. - about us.

Our service is comprehensive and covers:

- process control systems,
- monitoring systems,
- power plants and
combined heat and power generating plants,
- generation, transmissions and distribution,
- specialist advice and consultancy,
- hardware and software,
- design, production and installation,
- staff training.

We provide flexible solution

to meet individual requirements.
4
 1. IASE Ltd. - about us.

Our fields of activities:

- engineering and design of systems and equipment for

power engineering and industrial plants,
- management of deliveries,
- commissioning of systems and equipment in industrial
plants,
- providing training courses and maintenance services,
- implementing “turn key” projects.
Contact data:
ul. Wystawowa 1
51-618 Wrocław
www.iase.wroc.pl
[email protected]
Phone (071) 348 42 21
Fax (071) 348 21 83, 348 50 34
Secretariat (071) 347 72 94
5
 2. Thermal power unit models.

Models of thermal power unit describe static and dynamic

characteristic of power unit and collaborate with real automatic
block system (e.g. OVATION, PROCONTROL). They facilitate:

project,

research,

testing power unit control system.

6
 2. Thermal power unit models.

Parameters of models formulated by IASE Ltd:

live steam and re-superheated steam pressure,

unit active power,

rotational turbine set speed,

concentration of oxygen in flue gas,

furnace chamber vacuum pressure,

drum water level,

pulverized coal temperature,

coal mill engine power.

Using models cuts down the time of project realization and

reduces all the risks associated with the work on the real
object.
 2. Thermal power unit models.

Steps of making use of models on real power objects:

1. Performing model (e.g. in MATLAB/SIMULINK) based

on scientific literature and expert knowledge.
2. Model revision and optimization based on object data.
3. Model implementation and tuning on a real object in
different power plant operating condition (until obtaining
similar time courses between model and real signal).
4. Practical use of tuned model on real object.

8
 2. Thermal power unit models.

Models must:

be compact, not to use up space in modules in power plant

system (cost of implementation),

have the same logic blocks as in power plant system,

have fast time of reaction (counting process in power plant
system is very important).

That’s why it is important to simplify created

models to meet the above criteria.

9
 3. Model of a ring-ball coal mill

Grinding fuel process is very important in thermal and

electric power production at power units.
Essential qualities are static and dynamic characteristic of
coal mill, learned through object exploitation, research and
literature.
We also get knowledge about grinding process from models
research, based on data obtained from operating object.
Modeled signals can be used to identify a disturbance which
has influence on this process.

10
 3. Model of a ring-ball coal mill

Model perform is to reprocude transient flow-heat process

during operation of power unit and to obtain (in accordance
with reality) time course as typical disturbance responce.

We can take advantage modeled signal to:

get knownledge about phisical phenomenon proceed in
process, instalation and automatic control system,

disturbance indentification in process.

11
 3. Model of a ring-ball coal mill

Fig. 1. Block diagram of modelled pulverized coal temperature and coal mill engine power signals.
.Designations :
m pow -primary air flow signal, T pow-primary air temperature signal V pod - coal feeder rotational speed signal,
N sm − rz - coal mill engine power signal, Tmpp −rz - pulverized coal temperature signal, wc -humidity of fuel,
Tweg - coal temperature, Tmpp −m - pulverized coal temperature model signal
12
N sm−m -coal mill engine power model signal
 3. Model of a ring-ball coal mill

Heat flux, putting into coal mill with primary air, is

primary air fluid flow and hot air temperature
dependent.
Heat flux, putting with coal into coal mill, is coal feeder
constant temperature and rotational speed dependent.
Specific heats are selected based on phisical table and
then corrected during research.

Pulverized coal temperature model signal is formed by

dynamical elements signal conversion:

heat flux loaded to coal-mill (coal flow, primary air flow,
hot air temperature),

heat flux to water evaporation in coal,

coal and air fluid flow.
13
 3. Model of a ring-ball coal mill

Fig. 2. Block diagram of modelled pulverized coal temperature signals in MATLAB/SIMULINK.

14
 3. Model of a ring-ball coal mill

t [s]

Fig.3. Results of coal mill engine power signals in MATLAB/SIMULINK (sedimentation

of mill complex at the end): pulverized coal temperature – green, pulverized coal
temperature (with idle running) – blue. 15
 3. Model of a ring-ball coal mill

t [s]

Fig.4. Results of coal mill engine power signals in MATLAB/SIMULINK (starting up of

mill complex at the beginnig and dig in of mill complex at the end): pulverized coal
temperature – green, pulverized coal temperature model signal – blue. 16
 3. Model of a ring-ball coal mill

To compensate all the inside disturbance, imposisible

to model (e.g. homogeneous fuel fluid flow put into coal
mill and fuel phisical characteristic – especially
humidity), the correction controler was inserted.

To model pulverized coal temperature the inertial

dynamical elements were used in order to reflect
natural processes and inertial behavior of meter circuit
of the pulverized coal temperature.

17
 3. Model of a ring-ball coal mill

Fig. 5. Block diagram of modelled coal mill engine power signals in MATLAB/SIMULINK.

Coal mill engine power model signal is created by

conversion of coal feeder rotational speed in inertial
dynamical elements, which is the answer of coal mill engine
power to the coal feeder rotational speed changes. 18
 3. Model of a ring-ball coal mill

t [s]

Fig.6. Results of coal mill engine power signals in MATLAB/SIMULINK

(sedimentation of mill complex at the end): engine power signal – green, coal mill
engine power model signal (with idle running) – blue. 19
 3. Model of a ring-ball coal mill

t [s]

Fig.7. Results of coal mill engine power signals in MATLAB/SIMULINK (starting up

of mill complex at the beginnig and dig in of mill complex at the end): engine power
signal – green, coal mill engine power model signal (with idle running) – blue.20
 3. Model of a ring-ball coal mill

In this case correction controler was used to adapt the

gain of the above mentioned inertial dynamical
elements.
The input signals of this controller are:

modelled coal mill engine power,

actual coal mill engine power.

21
4. Practical use of the modelled mill motor power signal
and the pulverized coal temperature signal

• Fire in coal mills is most undesired, and the likelihood of this

to occur is increased with biomass co-firing. The
phenomenom of biomass clustering in mill chamber occurs
frequently, which may casue ignition and resulting fire in
the mill.

• During expoiltation of coal mills in systems with biomass co-

firing it can be observed that there is greater number of fuel-
air mixture ignitions during mill’s operation, shutdown and
start up stages than for coal only systems.

22
4. Practical use of the modelled mill motor power signal
and the pulverized coal temperature signal

Taking into account before mentioned issues, IASE Ltd

created the system to detect and prevent ignition and
fire in coal mills.

The system relies on:

•Algorithm to detect loss of coal in a coal mill –prevents
hot air flow into the coal bunker.

•Algorithm to detect fire in a coal mill, based on the coal

mill model.

23
4. Practical use of the modelled mill motor power signal
and the pulverized coal temperature signal

The output signal of heat balance controller (used in the

mill model) may provide means to measure diversity of
fuel stream supplied to a coal mill and its varying physical
properties (mainly humidity).

24
4.1. Loss of coal detection system based on the
coal-mill model.

• In the system the difference is created between

modelled mill motor power and the measured motor
power which is then compared with its limiting set value
(in the overrun indicator with set hysteresis).

• When the limiting value is exceeded the logic signal is

generated informing about loss of coal.

• The signal is stored and displayed, as well as

practically used in the process of disturbance
elimination.

25
4.1. Loss of coal detection system based on the coal-
mill model.

Fig 8.The block logic of the algorithm enabling detection loss of coal in coal mill. Designations
as in Fig.1, SP – overrun idicator with set histeresis and boundary value, SBW – logical signal
of loss of coal in coal mill.

The system was implemented on 14 power units with

the total power generating capacity of about 2800 MW.
26
 4.1. Loss of coal detection system based on the
coal-mill model.

Logic signals informing about loss of coal together with

the differential signal showing the dynamic
characteristic and strength of this disturbance are used
to decrease the adverse inpact of the disturbance on a
power unit control systems, e.g. in control systems of
the load and the amount of total air.

27
4.1. Loss of coal detection system based on the coal-
mill model.

This information is vital, because distrubances in the

fuel flow to the boiler tend to destabilize work of most
control systems, leading to creation of exagerated,
harmfull deviations in technological parameters and
their nominal values.

This further leads to:

•components damage,
•shortened life time of the equipment,
•overheat of the coal mill,
•lowering of the work comfort,
•potential life threats to the power plant personnel
working in proximity of coal mills.
28
4.1. Loss of coal detection system based on the coal-
mill model.

It is important to prevent clustering of the coal in the

coal bunkers or in the coal feed pipes.
In the case such a disturbance occurs it is vital to
receive the notification.

The system is characterized by:

• very good reliability of the signal informing about the
disturbance in a wide band of the mill load,
•high sensitivity,
• ease of application to the systems already being
exploited (usually equiped with power transducers).

29
 4.1. Loss of coal detection system based on the coal-
mill model.

Fig.9. Work presentation of loss of coal in coal mill detection: correction controller of power
engine coal mill output signal (green), coal mill engine power signal (blue), coal mill engine power
inertial signal – without idle running (light green), modelled coal mill engine power inertial signal –
without idle running (light blue), logical signal of loss of coal (pink), loss of coal power (yellow),
powet of unit (red), gradient of drum steam pressure (black). 30
4.2. Fire detection system based on the coal-mill model.

The system enabling ignition and fire detection in coal-mills is

based on comparison of the pulverized coal temperature with
the modelled pulverized coal temperature.

Fig. 10. The block logic of the algorithm enabling detection of overheat and fire in mill.
Designations as in Fig.1, SP – exceed signaling device with histeresis and boundary value,
SPT – logical signal of overheat or fire in coal mill.

The system was implemented on 14 power units with the total

power generating capacity of about 2800 MW.
31
4. 2. Fire detection system based on the coal-mill model.

The binary signal informing about overheat of the mill is

created when the set point (in the overrun idicator with set
hysteresis) is exceeded by the signal relating the
difference in temperatures between the pulverized coal
and its model (obtained from the heat equilibrium
deviations due to disturbances not included in the created
mill model).

32
4.2. Fire detection system based on the coal-mill model.

It is then successfully used in the system protecting the mill

against the propagation of a threat (fire), e.g. through
introduction of chemically neutral gas or steam.

Implementation of the above system is espessially effective

in systems with inertial temperature and the pulverized coal
measurement devices.

33
4.2. Fire detection system based on the coal-mill model.

Fig. 11. Results of modelled and actual pulverized coal temperature subject to natural
inputs: coal feeder rotational speed (red); corrected airflow to the coal mill (blue);
pulverized coal temperature (green); pulverized coal temperature model signal (light
blue); output of heat balance correction controller (violet); logic signal informing about
overheat or fire in the mill (yellow). 34
4.2. Fire detection system based on the coal-mill model.

Fig. 12. Presentation of power unit’s outputs of the fire detection system in the coal-mill based on the coal-mill
model (1 tick=1 minute): pulverized coal temperature in the neighboring mill (red); pulverized coal temperature
model signal in the neighboring mill (blue); logic signal informing about overheat or fire in the mill (green); fire
detection logic signal (light blue); pulverized coal temperature in the mill (violet); pulverized coal temperature in
the mill model signal (orange); temperature differential signal of pulverized coal in the mill (brown), output of
heat balance correction controller (black). 35
 4.2. Fire detection system based on the coal-mill model.

Fig. 13. Presentation of power unit’s outputs of the fire detection system in the coal-mill based on the coal-mill
model (1 tick=1 minute) : pulverized coal temperature in the neighboring mill (red); pulverized coal temperature
model signal in the neighboring mill (blue); logic signal informing about overheat or fire in the mill (brown); fire
detection logic signal (black); pulverized coal temperature in the mill (violet); pulverized coal temperature model
signal in the mill (orange); temperature differential signal of pulverized coal in the mill (green), output of heat
balance correction controller (light blue). 36
Ministry of Economy Cup in 2011 for product: „The system to
detect and prevent ignition and fire in ball-ring coal mills”37
 Mathematical Modelling and Simulation of
Power Plants and CO2 Capture

Thank you for attention!

University of Warwick
20-21 March 2012
Institute of
Power Systems
Automation Ltd
Wrocław, Poland

38