FIELD OF INVENTION
The present invention relates to a process of removal of impurities of hot metal
by oxygen in a basic oxygen furnace converter. More particularly, the invention
relates to an on line method in a BOF steelmaking process control system to
determine, monitor, correspondingly update the adaptive parameter for removal
of impurities in hot metal. The invention further relates to an on-line process
control system for carrying out the inventive method in a BOF plant.
BACKGROUND OF THE INVENTION
Hot Metal from the blast furnaces comprises many impurities which are
detrimental to the quality of final Steel products. These impurities need to be
removed for use in various Steel products. One of the known processes to
remove these impurities is the Basic Oxygen Steelmaking. The process is
implemented in a vessel, called Basic Oxygen Furnace (BOF). In this process,
most of the impurities for example, 'P', 'Si' & 'C' are removed by oxidizing these
with pure oxygen. Oxygen at supersonic speed is injected into the liquid hot
metal in the BOF converter. Most of the impurities get oxidized and float above
the metal as slag. Carbon forms CO gas and goes out of the metal. The Metal at
the end of treatment in BOF, comprises substantially pure Fe. It is important that
the impurities are removed at a desired level and correct temperature is achieved
at the end of the treatment. A high heat is generated during oxidation of these
impurities. A coolant such as iron ore and scrap are added to achieve the correct
temperature at the end of such treatment. A BOF converter is made of steel shell
in outer region. Inside of the BOF converter is relined with a basic refractory.
The oxides formed by oxidation of Si and Phosphorus, namely Si02 and P205 is

Acidic in nature and corrode the lining of BOF converter. Hence sufficient basic
material such as lime need to be added in the BOF so that overall slag is basic
and it doesn't corrode the BOF converter. Lime also helps in better removal of
Phosphorus from the metal.
Depending upon the application for which Steel is being produced, addition of
Ferro alloys are selected. To achieve a correct chemistry and temperature at the
end of the treatment, it is important that a correct amount of fluxes (such as
lime, limestone and dolomite etc.) and coolants (such as ore and scrap) are
added, and a right amount of oxygen is blown. Better control of chemistry and
temperature also help in optimization of the cost of steelmaking in BOF.
Prior art of BOF steel making mostly deals with measurement of carbon content
in BOF converter (e.g. WO/1997/016571), lance design for BOF steelmaking
(e.g. WO/2007/054957), addition of iron pellet and reduction in oxygen flow rate
(e.g. USPatent 5897684) and BOF steelmaking (e.g. US Patent 4529442).
The prior art do not disclose any process model which is enabled to predict
amount of fluxes and coolants including the oxygen blown for the metal being
treated in BOF converter.
OBJECTS OF THE INVENTION
It is therefore an object of the invention to propose an on-line method to
determine, monitor, correspondingly update the adaptive parameters for removal
of impurities from hot metal in a BOF Steelmaking process control system, which
eliminates the disadvantages of prior art.

Another object of the invention is to propose an on-line method to determine,
monitor, correspondingly update the adaptive parameters for removal of
impurities from hot metal in a BOF Steelmaking process control system, which is
enabled to accurately indicate the amount of fluxes and coolants including
oxygen blown for the treatment of the hot metal into the BOF Converter.
A still another object of the invention is to propose an on-line method to
determine, monitor, correspondingly update the adaptive parameters for removal
of impurities from hot metal in a BOF Steelmaking process control system, which
is capable to indicate the weight of addable lime, dolomite, and the iron ore in
the process including the volume and flow rate injectable oxygen, the weight of
produced steel and slag including their compositional analysis.
A further object of the invention is to propose an on-line method to determine,
monitor, correspondingly update the adaptive parameters for removal of
impurities from hot metal in a BOF Steelmaking process control system, which in
the preceding steps calculate the adaptive parameters for several batches in a
BFM model and generate a data base stored in a memory device of a general
purpose computer.
A still further object of the invention is to propose an on-line method to
determine, monitor, correspondingly update the adaptive parameters for removal
of impurities from hot metal in a BOF Steelmaking process control system, which
adapts a BSM model and implement the inventive method in the general purpose
computer and a server.

SUMMARY OF THE INVENTION
The invention proposes a method for on line control of a basic oxygen furnace
(BOF) process for removal of impurities in hot metal produced in the blast
furnaces. The invention is enabled to indicate weight of lime, dolomite and iron
ore; volume of oxygen, weight and analysis of steel and slag before the start of
BOF process using the process of this invention.
The system of invention relies on instrumentation that automatically senses the
condition of the Hot Metal. While many parameters are measured, the following
lists the parameters that are required to be measured for this invention for
example, Temperature, Chemistry and Weight of Hot Metal; Weight and
Chemistry of scrap; Target Temperature and Chemistry; Weight of Lime,
Dolomite and Iron ore; Volume of Oxygen, Chemistry of steel and slag. The data
is polled at second intervals, and is stored in a computer data base, from where
the information is extracted and used for the parametric measurement. The
quantity of the lime, dolomite and iron ore; volume of oxygen is determined by
this invention, as described hereinbelow, and is downloaded to a set of actuators
that automatically regulate the flow of the fluxes and coolants, including the flow
rate of oxygen. The flow regulation implies both the total amount as well the
rate of flow, as both the parameters have to be controlled to achieve an effective
regime.
A BSM model is developed using the principle of "Mass Balance and Heat
Balance" which takes into consideration as temperature, analysis; and Adaptative
coefficients as inputs and predict the weight of lime, dolomite and iron ore;
volume of oxygen, weight and analysis of steel and slag as the outputs.

Relationship and data in respect of Phosphorus partition, Sulphur partition, and
Mn partition has been established using past several hundred data and
multivariate regression analysis. They depend on the initial Target Temperature,
CaO(%), and Slag Fe(%).
After receiving the actual analysis and data at the end of the treatment, a second
Adaptative Model named BOF Feedback Model (BFM) linked with the BSM, which
runs when all batch data of the BOF, is available and changes Adaptative
parameters (such as Heat Loss, Oxygen efficiency, Slag Fe efficiency,
Phosphorus partition efficiency, Sulphur partition efficiency and Mn partition
efficiency) based on actual conditions of BOF converter and other operating
systems linked with it. It takes care of any drift in the instruments. There is a
statistical filter to remove erroneous values of adaptative parameters. The
filtered adaptative parameters are stored in the database for all the good
batches. The filtered adaptative parameters of last few batches are multiplied by
adaptative weightage matrix to calculate the adaptative parameters, which shall
be used for the next batch. This is repeated after processing of each batch.
Hence, under identical condition, predictions of the system is different as time
progresses and ensures that the targetted results are always achieved.
BRIEF DESCRIPTION OF THE ACCOMPANYING DRAWINGS
Figure 1 - Shows a flow chart depicting the process steps for implementing a
basic oxygen furnace static module (BSM) according to the
invention.
Figure 2 - Shows a flow chart of operating a Basic Oxygen furnace feed back
module (BFM) linked with the BSM according to the invention.

DETAILED DESCRIPTION OF THE INVENTION
The invention is an on line process control system to predict the weight of fluxes
and coolants, volume of oxygen using Basic Oxygen Furnace (BOF) Static Model
(BSM).
The BSM works on the principle of mass balance and heat balance. The inputs to
BSM are:
1. Hot Metal Weight
2. Hot Metal Temperature
3. Hot Metal Analysis
4. Scrap weights and its analysis
5. Target temperature and analysis
6. Adaptative coefficients (Heat Loss, Oxygen efficiency, Slag Fe efficiency,
Phosphorus partition efficiency, Sulphur partition efficiency and Mn
partition efficiency)
The outputs from the BSM are:
1. Weight of Lime
2. Weight of Dolomite
3. Weight of Iron Ore
4. Volume of Oxygen
5. Steel and slag weight
6. Steel and slag analysis

The system also automatically controls the predicted parameters as indicated
hereinabove. The Complete control of the BOF process is implemented via a
software program, which runs on a general purpose computer and a server.
The system and BSM has been developed using Microsoft Dot Net C# and
PL/SQL. Database of the system is Oracle. Oracle database runs on the server
and the system runs on the client computer. The system communicates with
level-I (PLC and Scada system) using OPC (OLE for process control).
Figure 1:
As shown in Figure 1, the BSM Model Starts at process step 100 after occurrence
of following :
1. Receive of blow Start event
2. Change of any input parameters as mentioned above
At process step 101, the BSM reads batch data (weight, temperature and
analysis of Hot Metal, scrap weight, target chemistry and temperature) and
master data (Limits of input and output parameters, analysis of scraps, iron ore,
lime, limestone, dolomite , Adaptative parameters such as heat loss, oxygen
efficiency, dephosphorization efficiency, Slag Fe Efficiency, and Sulphur
efficiency.). Limits of all input parameters are checked in step 102. If these are in
the given limit, the BSM calculation moves to step 103.

In step 103 effect of time delay ( period in BOF vessel was empty) is calculated
in terms of million calorie heat loss. Here radiation and convective heat loss has
been considered. Radiation heat loss has been calculated using Boltzman's
equation and convective heat loss is calculated using enthalpy. Output from step
103, is total heat loss due to time delay.
In step 104, elemental inputs are calculated. Total materials are converted in C,
Si, Mn, Fe, P, S, CaO, Si02, FeO, Fe203 etc. In next modules separate materials
are not considered. They are taken as elements. In next step 105, CaO and Slag
Fe percentage in Slag is calculated. CaO is calculated based on target
temperature, Si/P ratio in input materials and target phosphorus.
CaO(%) = kl+al*target_temp+ bl*SLP_ratio+cl*target_Phosphorus
Slag Fe is calculated based on target temperature and CaO(%)
Slag Fe(%) = k2 + a2* target_temp+ b2*CaO(%)
Constant kl, al, bl, cl, k2, a2, and b2 is calculated based on past data of the
converter and optimization is done to best possible removal of phosphorus from
metal. These parameters need to be tuned when model is being used in BOF
converter.
In step 106, steel and slag analysis is calculated. It was found out that sum of
four slag components CaO(%), Si02(%), Fe(%) and P205(%) is almost
constant. Moreover sum is Adaptative parameter and changes based on past
converter conditions. CaO(%) and Fe(%) are known from step 105. Hence sum
of Si02(%) and P205(%) is known. All Si coming from metal is get oxidized in
BOF converter. Weight of Si02 is calculated using mass balance.

Phosphorus partition depend on target temperature, CaO and slag Fe
percentage as defined below:
Phosphorus Partition between Slag and metal = k3 + a3*target_temp+
b3*CaO(%) + c3*Fe(%)
Constant k3, a3, b3 and c3 are calculated based on past data of the converter.
Total phosphorus coming in the BOF converter is known. Steel weight and slag
weight is assumed. Using steel weight, slag weight, total phosphorus and
phosphorus partition, weight of phosphorus in steel and P205 in slag is
calculated using mass balance.
Si02(%)/P205(%)=weight of Si02/ weight of P205
Si02(%)+P205(%)= known
Above equation is solved to get Si02(%) and P205(%).
Slag weight = weight of Si02/ Si02(%);
Sulphur in steel and slag is calculated as follows:
Sulphur partition between steel and slag depends on target temperature,
CaO(%) and Slag Fe(%)
Sulphur Partition between Slag and metal = k4 + a4*target_temp+ b4*CaO(%)
+ c4*Fe(%)

Constant k4, a4, b4 and c4 are calculated based on past data of the converter.
Total Sulphur coming in the BOF converter is known. Steel weight and slag
weight is assumed. Using steel weight, slag weight, total Sulphur and Sulphur
partition, weight of Sulphur in steel and Sulphur in slag is calculated using mass
balance.
Mn in steel and slag is calculated as follows:
Mn partition between steel and slag depends on target temperature, CaO(%)
and Slag Fe(%)
Mn Partition between Slag and metal = k4 + a4*target_temp+ b4*CaO(%) +
c4*Fe(%)
Constant k4, a4, b4 and c4 are calculated based on past data of the converter.
Total Mn coming in the BOF converter is known. Steel weight and slag weight is
assumed. Using steel weight, slag weight, total Mn and Mn partition, weight of
Mn in steel and MnO in slag is calculated using mass balance.
All Ti, Cr and Al get oxidized and form Ti02, Cr203 and AI203.
Weight of Ti02, Cr203, AI203, S and MnO is divided by slag weight to get
Ti02(%), Cr203(%), AI203(%), S(%) and MnO(%) in slag.

Weight of Phosphorus , Sulphur and Mn is divided by steel weight to get
Phosphorus(%), Sulphur(%) and Mn(%) in steel.
Once calculation of slag analysis, steel analysis and slag weight is completed,
BSM moves to step 107.
In step 107, heat of reaction is calculated using standard data available from
standard literature . Heat of reaction is calculated for following reactions:
1. C+1/2O2->CO
2. Si + O2->SiO2
3. 2P + 5/2O2-> P2O5
4. Fe + 1/2O2-> FeO
5. 2Fe + 3/2O2-> Fe2O3
6. Mn + 1/2O2 -> MnO
7. 2Cr + 3/2O2-> Cr2O3
Heat of enthalpy is calculated to heat each element from input temperature to
target temperature taking into account phase change, heat of solution, melting
etc. Coefficient of specific heat and heat of phase change and melting have been
taken from standard literature.
In step 108, heat balance is done to find out surplus or deficient heat. Surplus
heat is divided by heat of cooling of cooling agent to calculate weight of cooling
agent.

In step 109, oxygen balance is done. Total oxygen required for oxidation of C, Si,
P, Fe, Mn and Cr is calculated. Oxygen coming from Iron ore is also calculated.
Difference of oxygen required for oxidation and coming from Iron ore gives total
theoretical oxygen. It is multiplied by oxygen efficiency to get oxygen required
for the batch.
In step 110, limits of output of BSM are checked. If BSM calculation fails at any
step, then error message is generated. If it is OK, output is written in Oracle
database and downloaded in level-1 system for execution. It is also displayed to
operator.
Prediction of height of oxygen lance and its flow rate; timing and rate of addition
of fluxes and coolants; and flow rate of purging gas is done using decision
matrix, which depends on BOF converter life (number of batches since relining of
the converter), Input Hot Metal chemistry and Target chemistry.
The BFM, linked with the BSM is implemented after all the measurement results
of BOF process is available. Flow diagram of BFM is shown in figure2.
The Model starts at process step 200 after receiving of slag analysis data.
At process step 201, the BFM reads batch data (weight, temperature and
analysis of Hot Metal, scrap weight, steel temperature and its analysis, weight of
lime, dolomite, limestone and iron ore added, slag analysis) and master data
(Limits of all parameters, analysis of scraps, iron ore, lime, limestone, dolomite).

Limits of all parameters are checked in step 202. If these are in the given limit,
the BFM calculation moves to step 203.
In step 203, effect of time delay is calculated in terms of million calorie heat loss
as described in step 103 of BSM.
In step 204, elemental inputs are calculated as described in step 103 of BSM
except that target analysis and temperature is replaced by actual steel analysis
and temperature. In next step 205, HM and slag analysis are checked for Si/P
ratio, as Si and P remaining in metal is known. Rest is going in slag. If it is within
limits, then calculation moves to step 205.
In step 205, weight of steel and slag is calculated using mass balance. In step
206, heat of reaction and enthalpy is calculated same way as described step 107
of BSM.
In step 208, heat balance is done to calculate heat loss in BOF process. Mass
balance is done to find theoretical oxygen required. Ratio of actual and
theoretical oxygen gives oxygen efficiency. Phosphorus, Sulphur and Mn partition
is calculated based on actual values of turn down temperature, CaO (%) and Fe
(%). Actual values of Phosphorus, Sulphur and Mn partition is also known. Ratio
of actual and calculated values gives Phosphorus, Sulphur and Mn efficiency.
Lime efficiency is calculated as ratio of CaO in slag and total CaO added. Ratio of
actual and calculated slag Fe(%)( depends on turn down temperature, CaO (%),
Si and P ratio) gives slag Fe efficiency.
Statistical filter is used to check whether calculated Adaptative parameter could
be used for next batch prediction. If the parameters are OK, then they are used
to calculated value of Adaptative coefficient to be used for next batch. Actual
values of last good batches are multiplied with a matrix to give Adaptative
parameters for next batch. Calculation moves to step 209 and exit.

Book References
1. Metallurgical Thermodynamics - Fifth Edition- Kubaschewski & Alcock
WE CLAIM :
1. An online method in a basic oxygen Steelmaking process for removal of
impurities from hot metal to determine adaptive parameters based on a
comparison between an analytical data set from a static model in respect
of weight of lime, dolomite, and iron ore, volume of oxygen, weight and
compositional analysis of steel and slag, the method comprising the steps
of :-
- reading batch data (201) in respect of weight, temperature, analytical
data of hot metal including scrap, weight of lime, dolomite, limestone,
added iron ore, and data on slag analysis;
- reading master data representing limits of all said parameters such as
analysis of scrap, iron ore, lime, limestone, and dolomite;
- checking (202) whether all the parameters are within the given limits;
- calculating (203) the heat loss due to time delay representing the time
BOF converter remains empty;
- converting (204) total materials in terms of C, Si, Mn, Fe, P, S, CaO, SiO2,
FeO, Fe2O3, each being considered as an element;
- calculating (205) the elemental inputs based on actual steel analysis and
temperature;
- calculating CaO and slag Fe percentage in slag, the CaO being calculated
based on actual temperature, Si/P ratio in input elements, and
phosphorus in steel and slag, wherein the slag Fe percentage being
calculated based on actual temperature and percentage of CaO;
- calculating weight of steel and slag adapting the technique of mass
balance;
- determining heat of reaction and enthalpy, the heat reaction calculated
for a plurality of reactions, wherein heat of enthalpy being calculated for
each element from input temperature to actual temperature including
phase change, heat of solution, melting;
- implementing heat balance to determine whether the heat is deficient or
surplus, weight of cooling agent being calculated by dividing the surplus
heat that with the heat for cooling the cooling agent;
- determining the phosphorous, sulphur, and manganese partition based on
actual values of turndown temperatures including cao%, and Fe%; and
- calculating efficiency of P, S, and Mn based on actual and calculated
values, wherein the lime efficiency being determined as a ratio of CaO in
the slag, and correspondingly the CaO added;
- determining slag Fe efficiency as a ratio of actual and calculated slag Fe
(%), and
- verifying the calculated adaptive parameters using a statistical filter, and
the best values are considered to calculate the value of adaptive
coefficient.
2. The method as claimed in claim 1, wherein the adaptive parameters
comprise heat loss, oxygen efficiency, slag Fe efficiency, phosphorus
partition efficiency, and Mn partition efficiency.
3. The method as claimed in claim 1, wherein the process operates on the
principle of Mass balance and heat balance which considers the data
relating to temperature, analysis and weight of hot metal; weight and
analysis of Scrap; target temperature and analysis; and adaptive co-
efficients as the inputs, and wherein the output provide data respecting
to weight of lime, dolomite and iron ore; volume of oxygen, weight and
analysis of steel including slag.
4. The method as claimed any of the preceding claims, wherein the
outputted data is downloaded to a set of actuators of the BOF converter
which automatically regulate the flow of fluxes and coolants including the
flow rate of oxygen into the converter during implementation of the
impurities removal process from the hot metal.
5. An online method in a basic oxygen Steelmaking process for removal of
impurities from hot metal to determine adaptive parameters based on a
comparison between an analytical data set from a static model in respect
of weight of lime, dolomite, and iron ore, volume of oxygen, weight and
compositional analysis of steel and slag, as substantially described and
illustrated herein with reference to the accompanying drawings.

The invention relates to hot Metal from the blast furnaces comprise many
impurities which are detrimental to the final quality of the final Steel products.
These impurities need to be removed for use in various Steel products. One of
the processes to remove these impurities is Basic Oxygen Steelmaking. Oxygen
with supersonic speed is injected into the liquid hot metal in BOF converter. Most
of impurities get oxidized and the metal at the end of treatment is almost pure
Fe. Fluxes and cooling agent are added during BOF treatment. A basic oxygen
furnace static model (BSM) has been developed using first principles i.e. Mass
Balance and Heat Balance which takes temperature, analysis and weight of Hot
Metal; weight and analysis of scrap; target temperature and analysis; and the
Adaptative coefficients as the inputs and determines the possible weight of lime,
dolomite and iron ore; volume of oxygen, weight and analysis of steel and slag
as the outputs. After receiving the actual analysis and data at the end of
treatment, another Adaptative Model named BOF Feedback Model (BFM) linked
with the BSM, is implemented when all batch data of BOF, is available, and the
Adaptative parameters (such as Heat Loss, Oxygen efficiency, Slag Fe efficiency,
Phosphorus partition efficiency, Sulphur partition efficiency and Mn partition
efficiency) are updated based on actual conditions of the BOF converter and
other systems linked with it. The calculated adaptive parameters are filtered
using a statistical filter to remove the erroneous data. The Complete control of
the BOF process is implemented using a software program, which runs on a
general purpose computer and a server.