Title of Invention

A SYSTEM TO CONTROL CARBONISATION OF COKE BY PREDICTING READY OVEN TEMPERATURE OF COKE OVEN DURING COKING PROCESS

Abstract

The invention relates to a system to control the carbonization of coke by predicting the coke end temperature of coke oven at a very early stage of the coking process by generating matrix character for process parameters such as charge temperature , coking time, peaks and slopes of the oven temperature during the coking cycles by observing the thermocouples (TC) temperature embedded at the regenerators (4) beneath each oven. The temperature profile attends peaks and valleys during the coking cycle. The slope of the hill after attending the first peak after the coke mass is charged is found strongly correlated with the ready oven temperature ( the average temperature of the oven at the last hour of the coking cycle), based on which the coke end temperature is predicted and the coking cycle is controlled.

Full Text

FIELD OF THE INVENTION This invention relates to a system to control carbonization of coke by predicting ready oven temperature of coke oven during coking process. BACKGROUND OF THE INVENTION Coking is the process of heating coal in coke ovens to drive volatile matter from it. Ovens are generally heated by coke-oven gas, which burns in heating flues in an ovens side walls. Waste gases from this combustion pass out through a stack or chimney. At around 20 minute intervals, the flows of gas, air are reversed to maintain uniform temperature distribution across the wall. The design of heating systems varies from battery to battery. A significant proportion of the total coke gas produced from coking is returned to the heating system for burning. The process of filling the oven with coal is called charging. In Tata Steel the coals are stamped charged. It is a technology that involves formation of a stable coal cake with finely crushed coal by mechanically stamping outside the oven and pushing the cake thus formed inside the oven for carbonization. Once charged, coal is gradually carbonized to form "coke" in a period of 20-24 hours depending on the battery and the way it is operated. Times may vary even wider depending on the battery. The coking time can be changed depending on production requirements. During carbonization, coking coals undergo transformation into plastic state. It first swells and then re-solidifies to produce semi-coke and then finally the coke. In the initial stage of carbonization, plastic layers from adjacent to the heating walls, and with the progress of time, the plastic layers move towards the center of oven from either side and ultimately meet each other at the center. The quality and quantity of plastic layer determines the inherent strength of coke matrix. According to the invention there is provided a system to control carbonization of coke by predicting ready-oven temperature of coke oven during coking process by generating matrix character for process parameters such as charge temperature, coking-time, peaks and slopes of oven temperature patterns of coking cycles by observing continuously data through thermocouples located in the middle of twin regenerators, the thermocouple being connected and correlated with algorithm to filter out temperature drops by continuously evaluating the slope termed as K-slope of the filtered trend of the temperature pattern, normalizing the temperature trend by developing character to common charging temperature, correlating the same with developing character of oven temperature in the last hours of the coking cycle, coke end temperature being measured by pyrometer, computing graph of ready oven temperature on normalizing all the temperature trends against K-slopes, forming a set of a ready-oven temperature of the running trend, predicating ready-oven temperature of the running trend by shifting the average back to its original trend value and scheduling strategy on ready-oven temperature, the sequential control steps are being carried out by interfacing algorithm with a controller and operated through a software generated for this purpose. DESCRIPTION OF THE INVENTION The present invention will be best understood by the description with reference to the accompanying drawings in which Figure 1 represents a schematic diagram of coke-oven battery. Figure 2 represents a schematic diagram of operation of the coke oven according to the proposed invention. Figure 3 represents an oven temperature over the carbonization cycle of a typical oven. Figure 4 represents an oven temperature over the carbonization cycle for another typical oven. Figure 5 represents a graph showing ready-oven-temperature versus coke-end temperature (CET) (ROV and CET) Figure 6 represents a graph showing K-slope values Vs ready oven temperature [K-slope Vs ROV) where fixed time = 950)] Figure 7 represents percentage error (actual prediction) of ready-oven-temperature evaluated from K-slopes as distributed over number of data sets. Figure 8 represents a flow sheet illustrating the sequential steps of predicting ROV with correlating CET / CMT (coke mass temperature). Figure 1 relates to an existing general schematic diagram of coke-oven battery. Normally the coking time of a battery is a fixed time period of 20 hours. It is assumed that carbonization completes within the period. In case in an oven the carbonization completes before hand the oven would have to wait till the end of the coking period to discharge the ready coke. Additional energy is supplied to the oven in the last few hours or minutes of waiting time to maintain the coke temperature which could have been saved if the cokes are discharged just in time. In other cases an oven may not be ready in 20 hours. The present invention is related to development of a system which predicts the ready- coke temperature in an early stage of carbonization. This would help to control the coking process of a battery in two-folds: (a) adjusting the heat input during the coking process to attain the desired coke- mass temperature. (b) reduce the coking period if an oven is ready with the coke. Literature survey evidences that there is no prior art which is related to energy savings of coke-oven battery by adjusting heat input or coking time of ovens. Majority of the prior published or patent literature are related to productivity or design of batteries. The technologies that are published are like "production of soft coke", "design of bee-hive oven", "construction of bricks using fly-ash", "dust collection system of coke-oven" and similarly others. Underneath each oven (c/o) twin refractory regenerators (4) are located. The purpose of these regenerators are to preheat the fuel and air which enter the combustion chamber of the oven. The purpose of two regenerators is that when one acts as the source of preheating the fuel and the air the other acts as a sink to absorb heat from the flue gas of combustion which passes through it. After each reversal, the flue gas alters the path and the source regenerator becomes the sink and the vice versa. A thermocouple (TC) is installed in the middle of these two regenerators which may be treated as the representative oven temperature during coking cycle. The firing of flue gas arrangement is indicated by 3. The both way arrows indicate the movement of flue gas of combustion through the refractory walls (1) of the oven. Numerals 2 indicates emergence of flame and flue gas from the c/o. Figures 3 and 4 show the trend of oven-temperature (as measured by the thermocouple) observed over two different ovens of about 20 hours cycle. The trends show that the oven-temperature keeps rising for over an hour after the coal charging. The temperature rise is nominal of around 5-10° C after which the temperature descends down. It ascends again after it reaches the valley. The oven-temperature experiences multiple such ripples over the entire coking- cycle out of which the first peak and the slop are significantly large by magnitudes. This trend is similar to all ovens and statistically 95% of 700 randomly selected oven-temperature cycles of two different batteries follow the trend. 5% abnormal cycles which do not follow the trend are due to the facts like operational or instrumental abnormalities. Although oven-temperature cycles follow the same pattern, the process parameters like charge-temperature, coking-time, peaks and slopes of the trend are different for different ovens and different for different coking periods of the same oven as well. This proves that the oven-temperature pattern is an intrinsic property of the carbonization phenomenon and is not influenced by the oven- geometry, charging-schedule, reversal heating, energy input, refractory etc. With this, the logical conclusion can be made as that the final coke temperature of an oven is strongly co-related to the peaks and slopes of the trend. This invention relates to the system which predicts the ready coke-temperature at a very early stage of carbonization cycle using the peaks and gradients of the trend. The method is derived statistically using 700 and odd over-temperature cycles (randomly selected). The error percentage of predicted and actual temperature is around 0.4%. This invention also evolves a new strategy of sequencing the oven pushing based on the predetermined ready-oven-temperature which would help in energy savings in battery. An algorithm has been developed to trace the temperature-trend on real time. Since the thermocouple is located in the middle of twin regenerators the temperature drops during the phase-shift from one regenerator to the other are also reflected in the temperature trend. The algorithm continuously evaluates the slope of the filtered trend. The magnitude and sign of slopes of the past indicates the future pattern of the trend. The slope values and signs also indicate whether the trend at the current time has reached a peak or valley. Once the temperature trend reaches the first peak and starts descending towards the valley, the slope of the descending trend after the first peak has significant role in predicting the temperature of the oven in the last hour of the coking cycle. The slope is termed as K-slope in this study and the last-hour oven-temperature is termed as ready-oven temperature. A table termed as ROV-slope (Table 1.0) in page 17 is constructed by sets of ready-oven- temperature and the corresponding K-slope values. For an oven, this table is used to predict the ready-oven-temperature at an early stage of coking cycle. The detail of the table is presented in the following sections. Filtration and Normalization of trend As discussed in the previous section and as shown in Figures 3 and 4 the oven temperature trend progresses in a series of cyclic wavelets of rise and fall of temperature. As explained earlier, the temperature drops are due to phase-shift of regenerators. Control through the algorithm smoothens the trend by eliminating the valleys of the wavelets and reserving the peaks. The oven temperature at the time of charging varies significantly from one oven to another. The variation of charging temperature ranges from 1100° C to 1300° C. The first task of the algorithm is to normalize the temperature trends to a common charging temperature. It means that all the trends are shifted up or down in order to have a common charging temperature. The common charging temperature can be specified by the user and as default it is considered as 1200° C. As an illustration it can be stated that if an oven-temperature at the charging time is 1100° C then the temperature trend of that oven is shifted up by 100° C as if the oven is charged at common charging temperature of 1200° C. If another is charged at 1300° C the trend of that oven is shifted down by 100° as if it is charged at the common charging temperature of 1200° C. The temperature trends are normalized first and filtered as it progresses. Concurrently the slopes and K-slop factor are evaluated on the filtered trend for prediction of ready-oven-temperature. Ready oven temperature It is evident from the process that the oven-temperature in the last hour of the coking cycle has strong co-relationship with the coke-end-temperature (CET) or coke-mass temperature (CMT). The coke-end-temperature is measured by the pyrometer which is installed in the coke dump car. Figure 5 plots the average oven temperature of the last hour of the coking cycles and coke-end- temperatures (CET) of 100 different ovens which evidences the co-relationship. The ratio of these two temperatures is close to 0.8. For the convenience in terminology the oven which is in the state of last hour of coking cycle is called ready-oven and the average temperature of the last hour is called ready-oven- temperature. Rov-K-slope Seven hundered (700) coking cycles data of randomly selected ovens of two batteries of Tata Steel have been collected for this study. All the temperature trends of the cycles are normalized, filtered and K-slopes of all them are evaluated. The ready-oven-temperature of the cycles are plotted against the K- slopes. Figure 6 shows the plotted graph. The graph shape indicates that the ready-oven-temperature is a function of K-slope and is estimated with bounds from the K-slope value. The K-slope values ranges from - 0.004 - 0.03 with the decrement of 0.001. The table (Table 1.0) is the collection of sets of ready- oven-temperatures against different K-slopes values. The mean of each set is also included in the set as the most significant member of the set. The K-slope of a running trend, once computed can be searched as a key in the table. The corresponding set of the ready-oven-temperature is the list of probable ready-oven-temperature of the running trend. The average of the set is first de-normalized by shifting it back to its original trend values and then can be declared as the predicted ready-oven-temperature of the running trend. The sets of ready-oven temperature are continuously updated with each coking cycle. Frequency of occurrence of each data in each is monitored continuously. The data of very low or no re-occurrence can be eliminated from the set. The frequency of occurrence is used as a weight factor of averaging the set members. Introduction of a new member may occur any time. Validation Predicted ready-oven-temperature of 500 ovens are validated with actual data. The percentage of error obtained is 0.4% which is within industrial acceptance limits. Scheduling strategy on ready-oven-temperature: Coke-oven battery follows a sequence of charging the ovens. A typical sequence that is commonly followed is 1-5-8-12-16 which means that the fourth oven from the current one will be charged after charging the current one. The proposed scheduling strategy will follow the sequence of sorted ready-oven- temperature in descending order. In the existing practice the coke cycle is 20 hours long. It is only after 20 hours (after coke discharge) the coke mass temperature (CMT) is measured. If CMT is found lower or higher than the desired temperature then there is no means to correct the coke-mass temperature other than accept it and use it in blast furnace. But in the present invention CMT is evaluated in the very early time of coking cycle (within 2 hours after coal-cake charging) by using algorithm, which algorithm evaluates ROV. Since ROV is distinctively related to CMT so the algorithm is indirectly evaluating CMT and heat input to the battery can be controlled accordingly to get the coke discharged at desired CMT. The present invention as herein described and illustrated should not be read in a restrictive manner as many variations, modifications, adaptation and changes are possible within the limit and scope of the invention as encampused within the appended claims. WE CLAIM 1. A system to control the carbonization of coke by predicting the ready-oven temperature in a coke oven system, the coke oven system comprises of: - a coal carrrying car charging the coal into each coke oven of the coke- oven battery at the charging temperature ranging from 1100 °C - 1300°C; - means to supply air and fuel through alternate side-wall flues of the coke oven and heating up the oven from both the sides for the carbonization of coal; - a twin regenerator disposed beneath each coke oven for alternately preheating the fuel and air in one regenerator, and absorbing the heat as a sink from the flue gas in the second regenerator; - a thermo-couple (TC) placed in the middle of each twin regenerator; - a program logic set up in a computer to record the regenerator temperature profile over the entire carbonization time (Temperature- time flow-chart) and evaluate the slope of the temperature profile in particular, peak and valley after normalizing the charging temperature of each profile at 1200 C, the system is configured to: - record the temperature of coke oven through the regenerator thermo couple (TC) in a process computers; - evaluate the time to ascend the peak of the temperature profile and slope of of the same descending down the hill using a computer program and the slope is known as K-slope; - measure actual coke end temperature at the discharge end of the coke oven by a pyrometer; - correlate a ratio of the actual (measured) coke end temp with temperature trend of the ready oven temperature which is 0.8; characterized in that the system is enabled to correlate of ready oven temperature profile with the evaluated K-slope of the normalized regenerative temperature profile is made at a very early stage of the carbonization process and accordingly predict the coke end temperature after the completion of the carbonization process. 2. A system as claimed in claim 1, wherein the predicted of ready-coke temperature in an early stage of carbonization is made by control of adjustment of heat input during coking process to attain desired coke-mass temperature and reducing the coking period when an oven is ready with the coke. 3. A system as claimed in claim 1, wherein the control of carbonization of coke is made through observing the oven temperature pattern of a coke oven during carbonization cycle through development of an algorithm which reflect also temperature drops during phase-shift from one generator to the other. 4. A system as claimed in claim 1, wherein a ROV-slope is constructed by sets of ready-oven temperature and the corresponding K-slopes. 5. A system as claimed in claim 1, wherein the oven temperature trend is smoothened through elimination of the valleys of the wavelets and reserving the peaks. 6. A system as claimed in the preceding claims wherein all the trends are shifted up or down to have a common charging temperature. 7. A system as claimed in the preceding claims wherein the normalized temperature is filtered as the carbonization progresses and concurrently the slopes and K-slopes are evaluated on the filtered trend for prediction of ready-oven temperature. 8. A system as claimed in the preceding claims wherein the ratio between average oven temperature and coke-end temperature of the last hour of coking cycle is maintained close to 0.8. 9. A system as claimed in the preceding claims wherein all the trends are shifted up or down to have a common charging temperature. 10. A system as claimed in the preceding claims wherein the normalized temperature is filtered as the carbonization progresses and concurrently the slopes and K-slopes are evaluated on the filtered trend for prediction of ready-oven temperature. 11. A system as claimed in the preceding claims wherein the ratio between average oven temperature and coke-end temperature of the last hour of coking cycle is maintained close to 0.8. 9. A system as claimed in the preceding claims wherein computed K-slope of a running trend is searched as a key in the evaluated table of corresponding ready oven temperature and the average of the set is first de-normalized by shifting it back to its original trend values and then selected as the predicted ready-oven temperature of the running trend and scheduling sequence of charging the ovens. ABSTRACT A SYSTEM TO CONTROL CARBONISATION OF COKE BY PREDICTING READY OVEN TEMPERATURE OF COKE OVEN DURING COKING PROCESS The invention relates to a system to control the carbonization of coke by predicting the coke end temperature of coke oven at a very early stage of the coking process by generating matrix character for process parameters such as charge temperature , coking time, peaks and slopes of the oven temperature during the coking cycles by observing the thermocouples (TC) temperature embedded at the regenerators (4) beneath each oven. The temperature profile attends peaks and valleys during the coking cycle. The slope of the hill after attending the first peak after the coke mass is charged is found strongly correlated with the ready oven temperature ( the average temperature of the oven at the last hour of the coking cycle), based on which the coke end temperature is predicted and the coking cycle is controlled.

Documents:

01295-kol-2006 abstract.pdf

01295-kol-2006 claims.pdf

01295-kol-2006 correspondence others.pdf

01295-kol-2006 description(complete).pdf

01295-kol-2006 drawings.pdf

01295-kol-2006 form-1.pdf

01295-kol-2006 form-2.pdf

01295-kol-2006 form-3.pdf

01295-kol-2006 general power of attorney.pdf

1295-KOL-2006-(06-07-2012)-CORRESPONDENCE.pdf

1295-KOL-2006-(21-12-2011)-EXAMINATION REPORT REPLY RECIEVED.PDF

1295-KOL-2006-(21-12-2011)-FORM-13.pdf

1295-KOL-2006-(22-12-2011)-ABSTRACT.pdf

1295-KOL-2006-(22-12-2011)-AMANDED CLAIMS.pdf

1295-KOL-2006-(22-12-2011)-CORRESPONDENCE.pdf

1295-KOL-2006-(22-12-2011)-DESCRIPTION (COMPLETE).pdf

1295-KOL-2006-(22-12-2011)-DRAWINGS.pdf

1295-KOL-2006-(22-12-2011)-FORM-1.pdf

1295-KOL-2006-(22-12-2011)-FORM-2.pdf

1295-KOL-2006-(22-12-2011)-OTHERS.pdf

1295-KOL-2006-(29-08-2012)-AMANDED CLAIMS.pdf

1295-KOL-2006-(29-08-2012)-CORRESPONDENCE.pdf

1295-KOL-2006-CANCELLED PAGES.pdf

1295-KOL-2006-CORRESPONDENCE.pdf

1295-KOL-2006-EXAMINATION REPORT.pdf

1295-KOL-2006-FORM 1.pdf

1295-KOL-2006-FORM 13.pdf

1295-KOL-2006-FORM 18.pdf

1295-KOL-2006-GPA.pdf

1295-KOL-2006-GRANTED-ABSTRACT.pdf

1295-KOL-2006-GRANTED-CLAIMS.pdf

1295-KOL-2006-GRANTED-DESCRIPTION (COMPLETE).pdf

1295-KOL-2006-GRANTED-DRAWINGS.pdf

1295-KOL-2006-GRANTED-FORM 1.pdf

1295-KOL-2006-GRANTED-FORM 2.pdf

1295-KOL-2006-GRANTED-FORM 3.pdf

1295-KOL-2006-GRANTED-SPECIFICATION-COMPLETE.pdf

1295-KOL-2006-REPLY TO EXAMINATION REPORT.pdf

abstract-01295-kol-2006.jpg