EAF Energy Optimization at Nucor YamatoOmar Quintanilla Regulation Engineer Nucor-Yamato Steel Company 5929 Highway 18 East Armorel Arkansas 72310 870-762 5500 ext 146
[email protected] Guillermo Fernandez Technology Manager AMIGE International Blvd Gustavo Diaz Ordaz 402 Piso 2 Monterrey NL, México 64650 (52) (81) 1001-4076
[email protected] Introduction Nucor Yamato Steel is continuously working to improve the operations all time. The energy input optimization has always been at a good performance level, but there is always been turnarounds on the way to achieve better results. The most difficult part of the adjustment process is that the control system is divided into several control sub-systems which interact between each other. These sub-systems do not take into account the effects of changes made to another sub-systems, and are developed by different vendors in their own knowledge area. As an example, reducing the exhaust gas negative pressure may produce less energy losses, but it may cause the furnace to be more prone to cross-arcing and so to damages and loss of time. In order to optimize the EAF process as a whole, new SmartFurnace modules were added to the SmartArc system. The SmartArc control system has been in place since 1997 for controlling the electrical energy input at NYS. The new SmartFurnace modules were added in order to take on-line control of the Burner System and the Exhaust Gas system. In this paper we will first present a description of the EAF system, then we’ll talk about the main issues in the pursue of high performance, then we will present a brief description of the new control modules and their added flexibility in order to cope with the main variations in the other subsystems. Finally we’ll show some of the new features tested and the evolution of some of the adjustments and the results. System description NYS has an AC EAF with a 90MVA transformer with a two position online reactor. Using a 0.609 meter electrode diameter in a in a 1.12 meter pitch circle diameter. Figure 3. The Table has the transformer low voltage amps and volts in all the tap positions. Xfrmr tap Sec volts HV current A LV cuerrent kA MVA 16 929 1,506 55.94 90.0 17 940 1,506 55.30 90.0 18 947 1,506 54.87 90.0 19 958 1,506 54.22 90.0 20 966 1,506 53.79 90.0 21 978 1,506 53.14 90.0 22 986 1,506 52.71 90.0 23 994 1,506 52.28 90.0 24 1,002 1,506 51.85 90.0 25 1,015 1,506 51.21 90.0 26 1,023 1,506 50.78 90.0 27 1,036 1,506 50.13 90.0 28 1,045 1,506 49.70 90.0 29 1,055 1,506 49.27 90.0 30 1,064 1,506 48.84 90.0 31 1,078 1,506 48.20 90.0 32 1,088 1,506 47.77 90.0 33 1,098 1,506 47.34 90.0 Electrical System Burner System Four burners are located around the shell generating 6.9 MW each. The three port burners are used in Burner Mode creating combustion between low flow oxygen and natural gas, the Lance Mode uses supersonic oxygen flow and injected carbon. The nominal flows are: 10.5 m3/min Natural Gas 42 m3/min Oxygen Gas 40 kilogram/min Carbon Figure 4. The four burners modules in the furnace Shell The shell was designed to tap 122 metric tons and handle 10% of hot heal. Exhaust Gas System The exhaust gas system control keeps a constant pressure inside the furnace during regulation. This is possible by a pressure transmitter inside the furnace shell and a variable damper position. A step table is used to change the pressure set-point during the heat. System Control The old AMIGE regulator ARCMeter was replaced by the DigitARC Plus in March 2007, also the new SmartArc version was installed. The electrical system was improved with this change, regulation (electrode movement), and electrical profiles were adjusted to have a better performance. Figure 5. Power-on time since 2000 in both furnaces. The furnaces have the same characteristics. Still some improvement was needed in the chemical side, NYS decided to implement a project to optimize the EAF Operation. As part of this project new SmartFurnace modules where added to the SmartARC. These modules control the burners, Carbon Injection, the Furnace pressure ad a new furnace balance. Main issues affecting the performance NYS has had three remarkable difficulties in order to achieve a better performance in the EAF, Hot Spots, Cross Arc, and skulls on the roof and electrodes. The cross-arc has been an obstacle in NYS progress, not only able to produce damage in the furnace shell or roof, also increases the Power On and Power Off time. The hot spots create weak areas inside the shell that can lead the system to produce damage on the shell. Figure1. The cross-arc creates an unbalance in the secondary side voltage. The arc during a cross arc is capable to make big holes in the furnace shell. Units in the picture are Volts. The skulls on electrodes, arc and flash outside of the furnace shell being a problem to the operators and the roof structure, burned pipes and hoses are the typical result of skulls on the electrode. Figure 2. The figure shows a portion of a skull in the electrode, with a common size of 1m x 0.4m is formed by steel. When the skull falls from the electrode during regulation it arcs over the furnace roof. Units in the picture are Centimeters. The mentioned obstacles are the most remarkable, because they create higher power off time and are the ones that are at NYS aim. New Furnace Control The control system capabilities are as important as the equipment being used, since the control system provides the means to fully utilize the equipment. As a starting point to optimize the EAF control system, it was divided in three main control subsystems; Electrical Energy Input (EAF Transformer), Chemical Energy Input (Burners) and Exhaust Gas System (Bag House). All of them are very critical in the EAF process and may and the interactions between eachother may greatly affect their operation and requirements. The Smartarc control system has controlled the electrical energy since 1997, and now two new SmartFurnace modules were added to the SmartARC in December 2007 to control the Burners System and the Exhaust Gas System. The SmartFurnace was designed based on Master Programs, this function groups all the sub-programs included in the system. One-Change per trial is the one reason of this function, this permit the user to mix new profiles with some already tested profiles, users will be able to make a change in one of the sub-system profile, and keep the others sub-systems profiles the same, this is very helpful when trying to make improvements while keeping a base operation. This allows the operators to go back to previous operation setting with the change of a single selector. At NYS, the electrical system is controlled by the SmarARC. This control uses a Power Program that sends a Current Set-Point to the DigitARC Plus, this Current SP is a dynamic number affected by a couple of variables, could be Electrical Stability, Charge Weight, Energy Used, Temperature Protection, or the combination of all them. The simplest behavior is the combination of two variables, charge weight and the used energy (KWH perTon), making a different electrical profile for each heat, related to what each heat needs. The use of KWH per Ton, is simplified to the users by SmartArc creating a new variable called OperationalKWHperTon, this makes a consistent profile to each heat and easy to understand in system with two or more charges. Adding a third variable to the Power Profile, the Arc Stability, makes the profile more dynamic and accurate, not only reacts to the scarp weight, also to the density and the way the scarp fits in the furnace shell. The starting line of any Power Program inside SmartARC will be the use of the last three variables, where it changes the current set point (arc length) to fit to the scarp weight, density, and mix. This type of programs allows the users to shorten the arc gradually and follow the scrap behavior. A long period time in long arc or a high scrap density produces almost the same effect in the furnace shell (shell damage), electrical stability will protect the shell and add the maximum energy when needed, as soon as the furnace is in flat bath. One improvement made to NYS was to achieve the energy level in the furnace when it was required, this allow the furnace to have a lower power on time. The Power Program has different additions related to the furnace needs, in NYS the temperature protection is modifying the Power Program SP. Resistance Temperature Detectors (RTD) are located around the furnace shell, the logic, inside SmarARC, by a fuzzy logic make the arc shorter when the temperature is high. The Balance Program modifies the Current SP to distribute the power as needed inside the furnace, moving the melting pattern inside the shell. Since we are regulating current the pattern stays for long periods of time in the same way. Electrical System Control Burner System Control NYS burners used a ten step control with MWH as a step variable, with the option to receive the burner set-points from a Level 2 system. The addition of the new burner control from AMIGE was the upgrade of the SmartARC logic on the same computer we were using, was introduced as a trial version and ended as our new control system. The burners control from SmartFurnace was included to the SmartARC logic, this new nodes have a Burner Program designed to work with different variables, MWH, KWH/Ton, Stability Factor, O2PPM and Cycles. The new logic has a twenty step window per charge and can be modified to have as many as the user would like. The Burner Program has is capable to handle hundred programs per burner and also has a Burner Program Selector that allows the user to make quick changes in some burners, keep others the same and use them in a new program keeping the old program as the back-up. Figure 6. Burner Program Selector MWH is the simple way to control the steps, without considering that the scrap varies all the time, weight, density, type. As we mentioned earlier variables like OperationalKWH/Ton and Stability Factor will make a more accurate profile. OperationalKWH/Ton is useful to apply the same chemical energy to one ton of steel, helping the furnace operators to become constant in the steel chemical composition and slag properties at the end of the heat. Stability, in the other hand, could make your profile to be opposed to some of the OperationalKWH/Ton benefits, just if your electrical profile is not also reacting to electrical stability, if the scarp density is too high or it didn’t fall evenly during the melting down process, waiting on stability will remove some chemical energy from the furnace but help the user to avoid blow-backs. A blow-back is produced any time the supersonic oxygen lances in the burners are run in lance mode with a thick or big piece of scrap in the front of the lances. Normally this happens too early in the heat, this cause the oxygen to be redirected to the furnace walls, creating a damage during long periods of time. The Carbon Jet Cycle (CJ_Cycle) and the Oxygen Jet Cycle (OJ_Cycle) were created by AMIGE to be used during the burner mode, this are time based repetitive cycles. The cycles modify the burner set-points to create different patters in the melted scrap in front of the burner to prepare the area for the supersonic oxygen lance to avoid blow-backs and have a more efficient lance going to the bath. This allows the users to introduce more chemical energy and earlier in the heat. The injected oxygen reacts with some scarp components like Carbon (C), Silicon (Si) and Manganese (Mn), as well as the injected Carbon (C), and Natural Gas (Methane CH4, as primary component). Figure 7. The chart show the energy generated by some combustibles involved in the EAF process. The manufacturers of the burners intended the supersonic oxygen ports to work at a nominal flow, if the working flow is increased above the nominal, the lance become less efficient, but still add extra oxygen to bath. This extra oxygen needs more components to react with, carbon as an example, the extra oxygen will take more carbon from the scrap and the injected carbon, generating more energy and raising the temperature in the bath faster. That could sound like using less KWH/Ton in the heat, but also changes the carbon content in the melted steel, the carbon amount to react with FeO in the slag, the amount on needed minerals at the end of the heat. The use of the OperationalKWH/Ton help the users to add this extra oxygen in an ‘efficient’ way. The increase of the oxygen flow generates an increase in the lances oxidation rate, reacting faster with the present components, since the C, Si and Mn are a limited source the higher oxidation rate will have more impact in low weight heats and scrap with low components value. The use of this variable equalizes the impact to different weight heats and helps the users to have a benefit of that extra oxygen used in the heat. Figure 8. The figure shows the step variables and cycles used for one burner at the second charge. Variables like MW, KWH/Ton, Sf (Stability Factor), O2PPM are used in the same step. The logic also has a panel temperature protection, which turns of the lances during long periods of blow-backs. This function extend the panel life. The logic is also included in the SmartArc. Figure 9. Temperature protection logic. Different variables can be use to jump from one step to the other. N YS is using two more variables to manage the programs, Free02 and O2PPM (Oxygen Parts Per Million) . These variables can be used for the same purpose the difference is the units, both are estimating the amount of oxygen present in the molten steel. The FreeO2 variable generated by the SmartArc is estimating the oxidation level in the bath, the O2PPM variable is generated from the Free02 for easy use. The O2PPM variable estimates the oxygen part per million present in the molten steel. This variable is not only useful to use the adequate amount of oxygen during the heat, also allows the users to end the heat in a certain range of O2Ppm. This kills power of time at the end of the heat (Decarburization Time). Figure 10. The horizontal line is based on time, with forty heat data. The dark dots have the Oxygen Ppm value from the furnace sampler, the shady dots represent the estimated Ppm value from SmartArc when the sample was taken. Thermal Balance Adjustment Another achievment was to allow the system to determine the total thermal balance of the system. The oxygen jets, the carbon injectors and the burner cycles were adjusted not only to provide the total amount of energy and carbon content in the bath, but also to maintain a better balance of the total enrgy balance including the electrical and chemical energy in the furnace. Some adjustments were made only to modify this balance. Also thanks to the automatic cross-arc detection it was possible to modify the operation in order to minimize the cross-arc occurances. Every change in the operation was kept as a new Master program so all the steps of the practices optimization is also kept for the records. Exhaust gas system was not working in an optimum condition, changes in the PID control, pressure feedback signal inside the PLC were made to improve it regulation. The changes in the exhaust gas system contributed to eliminate the skulls on the electrodes and roof. Exhaust Gas System Summary In the following charts the results of all the improvements are shown. Averag e P ower O n T ime by Week 31.0 Ave P ower on time (min) 30.0 P ower On T ime 29.0 28.0 27.0 26.0 25.0 Wk 11 Wk 14 Wk 18 Wk 20 Wk 22 Wk 24 Wk 26 Wk 28 Wk 30 Wk 34 Wk 36 Wk 41 Wk 43 Wk 45 Wk 47 Wk 49 Wk 51 Wk 11 Wk 13 Wk 15 Wk 1 Wk 3 Wk 5 Wk 7 Wk 9 Wk 1 Wk 3 Wk 5 Wk 16 Wk 32 Wk 39 Wk 7 Wk 9 Week # Figure 11. Average Power On Time by week. Wk1 in the horizontal axis at the left, means the week number one in 2007 (Sunday 31st, on December 2006 to Saturday 6th, on January 2007). E AF Averag e Meg Watts 92.0 Ave E lectrical P ower Input (MW ) 90.0 A verag e Meg aWatts 88.0 86.0 84.0 82.0 80.0 78.0 Wk 11 Wk 14 Wk 16 Wk 18 Wk 20 Wk 22 Wk 24 Wk 26 Wk 28 Wk 30 Wk 32 Wk 34 Wk 36 Wk 39 Wk 41 Wk 43 Wk 45 Wk 47 Wk 49 Wk 51 Wk 11 Wk 13 Wk 15 Wk 1 Wk 3 Wk 5 Wk 7 Wk 9 Wk 1 Wk 3 Wk 5 Wk 7 Wk 9 Week # Figure 12. Average Megawatts by week. Wk1 in the horizontal axis at the left, means the week number one in 2007 (Sunday 31st, on December 2006 to Saturday 6th, on January 2007). O xyg en Us ag e / T on vs . P ower O n T ime Ave P ower on time (min) Ave T otal O xyg en C ons umption (S C F /ton) 31.0 1440.0 1420.0 30.0 1400.0 1380.0 1360.0 28.0 1340.0 1320.0 27.0 1300.0 1280.0 1260.0 26.0 25.0 1240.0 Wk 1 Wk 3 Wk 5 Wk 7 Wk 9 Wk 11 Wk 14 Wk 16 Wk 18 Wk 20 Wk 22 Wk 24 Wk 26 Wk 28 Wk 30 Wk 32 Wk 34 Wk 36 Wk 39 Wk 41 Wk 43 Wk 45 Wk 47 Wk 49 Wk 51 Wk 1 Wk 3 Wk 5 Wk 7 Wk 9 Wk 11 Wk 13 Week # Figure 13.Oxygen usage per ton and Power on time. Wk1 in the horizontal axis at the left, means the week number one in 2007 (Sunday 31st, on December 2006 to Saturday 6th, on January 2007). Wk 15 O xyg en S C F / T on 29.0 P ower O n T ime Natural G as Us ag e / T on vs . P ower O n T ime Ave P ower on time (min) Ave Natural G as C ons umption (S C F /ton) 31.0 280.0 30.0 270.0 260.0 29.0 250.0 28.0 240.0 27.0 230.0 26.0 220.0 25.0 210.0 Wk 1 Wk 3 Wk 5 Wk 7 Wk 9 Wk 1 Wk 3 Wk 5 Wk 7 Wk 11 Wk 14 Wk 16 Wk 18 Wk 20 Wk 22 Wk 24 Wk 26 Wk 28 Wk 30 Wk 32 Wk 34 Wk 36 Wk 39 Wk 41 Wk 43 Wk 45 Wk 47 Wk 49 Wk 51 Wk 9 Wk 11 Wk 13 Week # Figure 13.Natural Gas usage per ton and Power on time. Wk1 in the horizontal axis at the left, means the week number one in 2007 (Sunday 31st, on December 2006 to Saturday 6th, on January 2007). 4. References. Richard J. Fruehan, “The Making, Shaping and Treating of Steel,” 11th. Edition. The AISE Steel Foundation., Pittsburg, PA, 1998, pp. 311-395 Wk 15 C H4 / S C F / T on P ower O n T ime