Developments Towards an Intelligent Electric Arc Furnace

Developments towards an Intelligent Electric Arc Furnace at CMC Texas using Goodfellow EFSOP® Technology

Joseph Maiolo (contact author) Mohamed Boutazakhti

Cheng Wu Li

Tenova Goodfellow Inc. 170-7070 Mississauga Road, Mississauga, ON, L5N 7G2 (Canada)

Email: [email protected]

Chris Williams

CMC Steel Texas 1 Steel Mill Dr., Seguin, TX 78155 (USA)

Email: [email protected]

Key words: EFSOP, off-gas, energy, gas composition, energy cost, EAF optimization

ABSTRACT

Tenova Goodfellow Inc. (formerly Techint Goodfellow Technologies Inc.) has developed the Goodfellow Expert Furnace System Optimization Process (EFSOP®), which uses real-time analysis of EAF off-gases to optimize, dynamically, the chemical energy usage within the electric arc furnace. The benefits of the Goodfellow EFSOP® System are safety, increased process knowledge, lower conversion costs, and increased productivity.

In December 2005, Tenova Goodfellow installed and commissioned its Goodfellow EFSOP® system for CMC Steel at their Seguin, Texas melt-shop to optimize the operation of their 120 ton EAF. Subsequent to the initial EFSOP® installation and working towards continuous improvement in off-gas based optimization and control, Tenova Goodfellow has begun work at CMC Steel Texas to implement an intelligent control system for the EAF. In addition to off-gas composition and furnace operating parameters, the program uses continuous off-gas temperature and fourth-hole static pressure measurements to calculate dynamically a gas-phase mass and energy balance for the EAF. This balance is in turn used to elucidate important steel-making information such as: rate of air in-leakage into the furnace; rate of de-carburization from the bath; rate of oxidation; rate of water in-leakage into the off-gas; and rate of energy losses from the gas-phase. This paper will outline the path to EAF optimization and the benefits achieved at CMC Texas and present progress made towards the development of Tenova’s Intelligent Furnace (iEAF®) system.

mailto:[email protected]

mailto:[email protected]

OVERVIEW The CMC Texas EAF is a 22’ diameter EAF, 120 short ton eccentric bottom tapping, AC furnace powered by an 80 MVA transformer. Chemical energy is provided by three PTI JetBox modules, three 3.5 MW burners (for carbon, oxygen and methane) and one 1 MW EBT conventional burner.

The Goodfellow Expert Furnace System Optimization Process (Goodfellow EFSOP®) is a dynamic control and optimization system for electric arc steelmaking furnaces (EAF) based on real-time measurements of off-gas composition. The system uses state-of-the-art off-gas analysis combined with process data acquisition and real-time closed loop control to optimize the operation of the EAF. Typically, the objectives for optimizing the EAF operation are to reduce conversion costs (energy and material), increase production and improve safety. A schematic of the EFSOP® system is shown in Figure 1. It includes the following components:

• A water-cooled sampling probe • Gas analysis system • Supervisory Control and Data Acquisition (SCADA) system

Figure 1: Schematic of the EFSOP® System

A patented, water-cooled sampling probe, custom designed for use in the harsh environment of the EAF, is installed through a port in the water-cooled ductwork, at the entrance to the fixed duct and after the combustion gap. The location and positioning of the probe is such that the gases are sampled before introduction of combustion air from the gap and so are representative of the composition of gases in the freeboard of the furnace. The extracted gases are transported via a heated sample line to the EFSOP® gas analysis system where the continuous stream of sampled gases are analyzed for oxygen (O2), carbon dioxide (CO2), carbon monoxide (CO), and hydrogen (H2).

A secondary function of the analyzer unit is to periodically back-purge the sampling line and probe to remove accretions inside the probe and dust build-up from the various internal filters employed. The system is only purged during natural breaks in the process (e.g. during charging, tapping or other power-off times) so as to ensure reliable and continuous off-gas composition measurements.

Composition measurements, as well as operational alarms and outputs from the analyzer are sent to the plant’s PLC network. The EFSOP® HMI is linked to the same network. The Human-Machine Interface (HMI) reads and logs the off-gas data, as well as all relevant process data at a frequency of 1 second. In total over 200 furnace parameters are sampled and logged. Both historical and real-time plots of the data are available. In addition to real time process data trending and historical data mining, the HMI allows regular data transfer to TGTI’s office in Mississauga. The EFSOP® SCADA system consists of a high-speed computer and GE’s iFIX SCADA software. Furthermore, the SCADA system performs control calculations, in response to furnace off-gas data and sends oxygen and methane flow setpoints PLC.

EAF Optimization

Furnace optimization via EFSOP®, typically, follows two parallel paths: 1) closed-loop control of chemical energy and post-combustion (oxygen, methane and carbon rates) in response to off-gas composition measurements and 2) adjustments to the EAF process from off-line analysis of furnace operation using a variety of proprietary tools such as advanced regression methods; and TGTI’s DECSIM electric arc furnace modeling software ([1] and [2]).

The first is the implementation dynamic control of post-combustion in the furnace freeboard in response to off-gas composition while working within the mechanical limitations of the existing hardware. During periods when the furnace off-gas is reducing, additional oxygen may be added or methane flow reduced; alternatively, during periods when the furnace off-gas is oxidizing, oxygen flow maybe reduced and/or methane increased. The challenge is to maintain a slightly reducing atmosphere in the furnace freeboard. In practice, the implementation of closed-loop control requires that constraints be placed flow set-points of oxygen/methane at different stages of the heat. For example: directly after charging the flows may be limited so that the flame does impinge on the scrap and rebound back onto the panel; during melting the flows are constrained so that the injectors operate around stoichiometric or slightly super-stoichiometric; and during refining, where supersonic oxygen flow is required, the constraints on oxygen flow are increased. The operating profiles for oxygen and methane delivery, definition of constraints and timing as well as the ability to save different programs are possible through customized screens developed within the EFSOP® HMI.

The second aspect of optimization via the EFSOP® system is in the form of process adjustments (e.g. timing and intensity of oxygen lancing, carbon injection, lime practice, fume system parameters, etc.). A holistic approach to optimization is taken and adjustments to the process are made according to what is observed in the off-gas. To aid in this process, the EFSOP® system includes proprietary software that forwards operating data directly to TGTI (on a daily basis). This data includes not only dynamic operational information from the process, but also alarms and operating status of the EFSOP® analyzer. Any operational issues with the EFSOP® analyzer are known directly and can be addressed quickly by feedback to the plant personnel. During the optimization stage of the EFSOP® implementation, TGTI engineers are able to follow plant performance remotely and communicate adjustments to the plant.

The EFSOP® system (probe, analyzer and SCADA system) was installed at CMC Texas in December, 2005. Preliminary observations of off-gas composition revealed that CMC’s operation resulted in a significantly reducing furnace off-gas with relatively high levels of hydrogen and carbon monoxide and that there was the opportunity to improve the practice by reducing the amount of chemical energy usage in the furnace. A different approach would have been taken if the process was determined to be oxidizing. An example off-gas profile before optimization is included (see Figure 3). The profile shows that carbon monoxide concentration (before optimization) was typically in the range of 20%-30%, during melting, and 30%-40% during refining.

The noted observation was typical during the preliminary data collections and monitoring period that extended into January and part of February of 2006. A strategy for optimization was developed over that period and implemented during the later part of February 2006. Using CMC’s typical burner and power programs as a starting point, the following adjustments were made to the process:

• Carbon injection was reduced during the melting stages of the process;

• Closed-loop control was implemented to automatically adjust the stoichiometric ratio of oxygen to methane in response to off-gas composition measurements provided by the EFSOP® system; and

• The timing and intensity of the fixed wall injectors was adjusted to control carbon evolution during the refining period.

The EFSOP® burner control program was customized over the course of the implementation period. Constraints on post-combustion, carbon injection and the timing and intensity of oxygen injection during refining were adjusted. To provide flexibility, the power-on-time period was divided into 12 stages that are cycled through according to a kWh/ton electrical energy clock. Each burner, during each of these stages, is able to operate in one of twelve modes (eight burner modes and four lance modes). Constraints on each mode included a minimum and maximum rate for oxygen and methane within which the control program is free to modulate. A default setting for each mode is also specified. The default is used in as a fail-safe in case the EFSOP® system is unable to read accurate off-gas composition or in case the plant chooses to disable closed-loop control.

The main principles in defining max and min post-combustion modes are based on combustion chemistry. Optimization of combustion during periods of reducing chemistry requires that more oxygen is added to provide additional energy for combustion. Methane may be reduced because there is already sufficient fuel in the off-gas and additional methane is not needed. Conversely, during oxidizing periods, an excess amount of oxygen is present and insufficient fuel. Therefore oxygen usage is reduced during these periods as it is not needed. There is overall less natural gas and higher oxygen to methane ratios in the EFSOP® burner mode for the optimized practice than the plant original practice because CMC Steel Tx’s furnace was observed to be highly reducing.

Figure 2: The EFSOP® Human-Machine Interface

Figure 3: Sample Heat Chemistry

Water-Leak Detection

In addition to the standard furnace optimization package, the installation of EFSOP® at CMC Steel Tx included a water-leak detection routine. TGTI engineers have noted that the presence of water in the EAF off-gas (through panel leaks, excessive electrode cooling flow, wet scrap, etc.) affect the level of hydrogen in the off-gas sample; and so even though water concentration is not measurable directly (gases are analyzed on a dry basis) the presence of water is detectible through hydrogen and carbon monoxide levels. Based on this, a water-leak alarming system was implemented that warns the operator of the possibility that water is leaking into the furnace. It has been determined that not only is the absolute value of hydrogen concentration important but so is the relationship between hydrogen and carbon monoxide. It has also been observed that absolute value of hydrogen concentration and the H2/CO ratio will vary over the course of the heat. This makes sense from a process point of view as things like the flashing of oils from the scrap and/or wet scrap will affect hydrogen and H2/CO ratio but do not constitute water-leaks. In order to account for these affects and avoid “false alarms” as much as possible, each charge was divided into four time segments. The thresholds for alarming are adjustable for each segment. These thresholds were defined by periodic adjustments with the goal of detecting water-leaks but minimizing false alarms. It is important to note that the system is heavily empirical and so must not be solely relied upon. It presents to the operator and alarm indicating the possibility of water-leak. The operator can then use this information, along with other indicators (e.g. panel temperatures) to determine whether or not a dangerous situation has developed. The water leak interface screen is shown if Figure 4.

Figure 4: Water Leak Detection Interface

The operator can set three kWh benchmarks allowing the division of a charge into four intervals (early melting, melting, early refining and refining). For each of these intervals two thresholds are set, one for absolute hydrogen concentration in the furnace and the second is for the relative concentration of hydrogen with respect to carbon monoxide. Functionally, this system provides the flexibility to assign different thresholds for alarming depending on the various stages of the process. For example, during refining hydrogen levels and the ratio of hydrogen to carbon monoxide are typically lower than during the initial stages of melting. This is due to the higher rate of methane usage during melting and also due to the flashing of oils from the scrap or the presence of water in the charge that is not present during refining.

EFSOP® RESULTS

Typically, the benefits due to EFSOP® optimization are determined by comparison to a representative historical baseline. Ideally, a long term baseline is used to account for seasonal affects. At CMC Texas, the injectors were changed at the end of 2005. Some performance benefits were noted during this period that was not attributable to EFSOP®. Therefore the baseline period was based only on the operation during the whole of January 2006. Figures 5, 6, 7 and 8 are, respectively, the average monthly specific consumptions of electrical energy, natural gas, oxygen, and carbon per inventory cast ton. The plots contain the averages for Oct-05, Nov-05, Dec-05, the baseline month (Jan-06) and the EFSOP® evaluation period (February 20th to march 19th, 2006).

391.20

387.36

385.19

376.75

367.52

355.00

360.00

365.00

370.00

375.00

380.00

385.00

390.00

395.00

Oct-05 Nov-05 Dec-05 Jan-06 EFSOP Evaluation Period, Feb 20- Mar 19, 2006

Figure 5: Electrical Energy Consumption, KWH / inventory cast ton

265.91263.22

259.71

236.98

227.83

200.00

210.00

220.00

230.00

240.00

250.00

260.00

270.00


Figure 6: Natural Gas Consumption, scf / inventory cast ton

1109.23 1108.821108.02

1131.88

1124.00

1095.00

1100.00

1105.00

1110.00

1115.00

1120.00

1125.00

1130.00

1135.00


Figure 7: Oxygen Consumption, scf / inventory cast ton

26.68 26.77

27.18

28.21

30.57

24.00

25.00

26.00

27.00

28.00

29.00

30.00

31.00

Oct-05 Nov-05 Dec-05 Jan-06 EFSOP Evaluation Period, Feb 20 -Mar 19, 2006

Figure 8: Injected Carbon (Carbon Jets and Door Lance) Consumption, lb / inventory cast ton

Average consumptions of oxygen, methane, electricity and carbon have been calculated by taking the sum of consumption over the period under investigation and dividing through by the total inventory cast tons of steel produced over the same period. This analysis reveals that the average electrical energy, natural gas and oxygen specific consumption were reduced, by 9.23 kWh, 9.15 scf and 7.87 scf per ton inventory cast, respectively, during the EFSOP® evaluation period compared to January 2006.

Carbon specific consumptions reported in the plant heat summaries show an increase in the specific carbon consumption. It was expected that average carbon consumption should have decreased as setpoints and timing were adjusted specifically to reduce overall carbon usage. Instead, the specific carbon consumption was noted to increase during the evaluation period as indicated in Figure 8. The reason for the increase was due to operator override for adjustment of slag quality.

91.08

90.97 91.00

91.83

92.33

90.00

90.50

91.00

91.50

92.00

92.50


Figure 9: Yield, inventory cast ton / total metallic charge, %

2.95 2.962.97

3.19

3.22

2.80

2.85

2.90

2.95

3.00

3.05

3.10

3.15

3.20

3.25


Figure 10: Productivity, inventory cast ton / power on time, ton/min

40.56 40.49 40.41

37.94

37.45

35.50

36.00

36.50

37.00

37.50

38.00

38.50

39.00

39.50

40.00

40.50

41.00


Figure 11: Power On Time (POT), min

Yield (inventory cast ton per total metallic charge) increased by 0.5% during the evaluation period in comparison to January 2006. This increase is evident in Figure 9, where the average yield increased from 91.83% during Jan. 2006 to 92.33% during the evaluation period.

As EFSOP® only affects the operation during power-on-time productivity is calculated based on power-on-time and not on tap-to-tap time, as is typically done. The productivity (Inventory cast ton per minute power on time) is shown in Figure 10. It has increased during the evaluation period by 0.03 % compared to Jan-06, increasing from 3.19 ton/min to 3.22 ton/min. The increase in productivity comes from the increase in yield as noted above and from a reduction in power-on-time. In fact, POT was decreased by 0.49 minutes compared to Jan. 2005 as indicated in Figure 11.

The economics for oxygen, methane, carbon and electrical usage have been evaluated in the usual straight-forward manner. Specifically, benefits are calculated as the product of the unit cost and the difference in average consumption during the baseline period compared to the evaluation period. Figure 12 is the average dollar cost per ton inventory cast contribution from electricity, natural gas, oxygen, and injected carbon. These costs dropped by 0.29 $/InvCastTon. As in the increase in carbon usage was suspect, an analysis was also considered without including carbon and found to be 0.52 $/InvCastTon.

24.65

24.46

24.37

23.91

23.62

23.00

23.20

23.40

23.60

23.80

24.00

24.20

24.40

24.60

24.80


Figure 12: Dollar Cost of Electricity, Natural Gas, Oxygen and Injected Carbon, $ / inventory cast ton

The formula for evaluation of the benefits due to increases in production is presented below:

)1(*)/($ −=POT

POTNETInvCastTontyproductivi REF

Where NET is the net earnings per ton billet sold and POTREF and POT are the power on time averages for the reference (Jan-06) and the evaluation periods, respectively. A value of $50/ton good billet was assumed. It is a measure of the additional profit made by being able to produce more (sell more) as result of improved productivity (lower POT and increased yield) using the same fixed costs. Figure 10 is the plot of the productivity improvement average dollar equivalent. The reference being Jan-06, therefore the productivity dollar equivalent is zero for January. The improvement in productivity during the evaluation period resulted in additional profit of 0.66 $ per ton inventory cast.

-3.23-3.15

-3.05

0.00

0.66

-3.50

-3.00

-2.50

-2.00

-1.50

-1.00

-0.50

0.00

0.50

1.00


Figure 13: Productivity Increase Induced Additional Profit, $ / inventory cast ton

The mathematical formula for the calculation of benefits due to increased yield is presented below:

)11(**100)/($YieldYield

ScrapCostInvCastTonyieldREF

−=

Scrap Cost is the current nominal market value of scrap ($ per ton metallic charge) and has been assumed to be $210/ton ([3]). YieldREF and Yield are the average yield values for the reference and the evaluation periods, respectively (in %). It is a measure of the amount of scrap saved by improving yield. Figure 14 is the plot of the average savings resulting from yield improvement. The savings in scrap for the evaluation period are $1.25 per ton inventory cast compared to Jan-06.

-3.12

-3.59-3.49

0.00

1.25

-4.00

-3.00

-2.00

-1.00

0.00

1.00

2.00


Figure 14: Yield Improvement Induced Savings, $ / inventory cast ton

In total, considering oxygen, methane, carbon, electricity, productivity and yield, benefits of $2.21/ton per ton good billet have been realized during the evaluation period of the Goodfellow EFSOP® system at CMC Steel Tx. The largest benefit comes from the increase in yield and the corresponding savings in metallic scrap. Further benefits come from electrical energy savings and increased productivity resulting from a reduction of 0.5 minutes of POT, from 38 minutes to 37.5 minutes. Decreases in oxygen and methane usage also contribute positively to the overall savings. Overall, the only negative comes from an increase in carbon consumption of 2.36 lb/ton, from 28.21 lb/ton to 30.57 lb/ton.

TOWARDS AN INTELLIGENT CONTROL SYSTEM FOR THE EAF

Automation and control of the electric arc furnace (EAF) steel-making process is limited, in part, due to the many challenges of implementing reliable, low-maintenance process sensors in such a harsh environment. The process, by nature, is extremely erratic due to such things as scrap mix and operator input and therefore presents an opportunity for advanced automation. Two common control systems that do exist on EAFs are electrical energy and chemical energy delivery to the EAF. System for controlling electrical energy and chemical energy are typically supplied independently. Off-gas analysis provides more information than simply a measure of the extent of combustion in the EAF freeboard. TGTI engineers, when optimizing the operation of the EAF, use off-gas analysis to gauge the dynamics of the steel-making process and are able to relate steel making parameters to what is observed in the off-gas. For example, sudden increases in carbon evolution (CO + CO2) have been related to carbon boils, hydrogen spikes can be related to water-leaks into the freeboard, etc. Off-gas analysis is one of the few dynamic measurements available in the operation of the EAF. A key benefit of the Goodfellow EFSOP® system is the ability to provide off-gas composition dynamically and in real-time over the course of the steel-making process. Without such dynamic information, steel-makers must rely on static information (e.g. sampling of the slag/bath at the end of the heat, initial scrap composition, amount of charged carbon, etc.) and make assumptions regarding the dynamics of the process.

MODEL INPUTS AND OUTPUTS Initial work towards a dynamic control system for the EAF has focused on building a mathematical model for calculating a dynamic mass and energy balance applied to the gas-phase of the EAF. The idea of calculating a mass/energy balance of the gas-phase of the EAF is not new. What differentiates the current approach from that of others is that here the balance is calculated dynamically using actual measurements of off-gas composition and temperature. Figure 15 shows the general structure of the modeling effort:

OFFGAS

TemperatureStatic PressureComposition

O2, CH4, CAIR INLEAKAGE

WaterInleakage

Figure 15: EAF gas-phase mass and energy balance. In the past, measurement of furnace off-gas temperature has been attempted with traditional sheathed thermo-couples. Although thermocouples have been able to provide temperature measurements for brief periods of time, they have been found to be un-reliable long-term in the harsh EAF environment. TGTI has identified an infra-red (IR) gas-phase pyrometric technology based on the emissivity of carbon dioxide and is working towards adopting the technology to the EAF environment. Initial trials at CMC Texas have shown promise. The model takes as its inputs the instantaneous rate of flow of methane, oxygen and injected carbon into the EAF freeboard. In addition to process parameters, real-time values of off-gas composition and temperature are provided by the EFSOP® system. An additional piece of information that is required to close the mass balance is an estimate of the rate of off-gas leaving the EAF. From a modeling point of view, this value scales the model and forms the basis of the calculation. The initial approach is to relate the off-gas rate to a measure of static pressure at the fourth hole. The empirical relationship can be tuned to an overall carbon balance. To this end, TGTI has developed a static pressure probe that is loosely based on the same design of the EFSOP® extractive probe for off-gas analysis. Simply, it is a water-cooled probe, fitted with a pressure transducer and positioned in the elbow of the EAF. Given these inputs, the model uses elemental (carbon, hydrogen, oxygen and nitrogen) and energy balances to calculate:

a) Carbon entering the gas-phase (as CO gas) through decarburization and combustion of carbon – from carbon balance b) Hydrogen entering the gas-phase (as H2O) from water-cooling, leaks, wet scrap, etc. – from hydrogen balance c) Oxygen leaving the gas-phase due to oxidation (carbon, iron, silicon, etc.) – from oxygen balance. d) Air in-leakage entering the gas-phase – from nitrogen balance. e) Heat losses from the gas phase – enthalpy balance given composition and temperature.

Figure 16 is a plot of a partial energy balance for a typical heat at CMC Texas, calculated by the off-gas model. For comparison, the off-gas composition is super-imposed on the power profile and shows the concentrations of the off-gas components. Due to a relatively constant flow of off-gas leaving the EAF, the rate of energy leaving the furnace as sensible energy fairly constant at about 8 MW. Chemical energy, being the energy that could have been recovered in the furnace assuming complete combustion, varies over the course of the heat but peaks at around 32 MW. The energy loss curve is the energy that is lost from the off-gas and contributes to the heating and melting of scrap. Note that this term is most significant, reaching about 10 MW, at the start of the charge where there is cold scrap to absorb the energy.

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 300 600 900 1200 1500 1800

time (s)

off-g

as c

ompo

sitio

n(m

ole

frac

tion)

0

10

20

30

40

50

60

Off-

gas

loss

es (M

W)

Sensible Chemical EnergyLossyCO yCO2 yH2yH2O yO2 yN2

Figure 16: Plot of partial energy balance for a typical heat.

CONCLUSIONS The furnace performance at CMC Texas, during the EFSOP® evaluation period (February 20th to march 19th, 2006) compared to January 2006 is summarized in Table 1. The reduction in the power on time and the electrical energy consumption and the increase in yield in particular resulted in an overall saving of 3.02 dollars per cast ton.

Table 1: EFSOP® Performance Summary based on cast short tons.

Performance Parameters

Baseline

EFSOP®

Optimization

Difference

Savings

($/t)

Power on time (min.) 37.94 37.45 0.49

Electricity, Kwh/t 376.75 367.52 9.23 $ 0.40

Gas, scf/t 236.98 227.83 9.15 $ 0.10

Carbon, lb/t 28.21 30.57 -2.36 - $ 0.22

Oxygen, scf/t 1131.88 1124.00 7.87 $ 0.02

Yield, % 91.83 92.33 0.5 $ 1.25

Productivity

Cast ton/POT 3.19 (t/min) 3.22 (t/min) 0.03 (t/min) $ 0.66

Overall performance, $/ cast ton $ 2.21

Many of the steel-making parameters in the EAF, due to limitations in sensor technology, are measured sporadically and possibly only once or twice per heat. For example, bath carbon and temperature may be sampled once or twice per heat or sometimes not at all; slag and bath residuals are sampled only once (if at all) and analysis is done off-line. Very little dynamic information exists for the steel-maker to gauge the progress of the process from start to finish. The advantage provided to steel-makers by the EFSOP® system is a dynamic measure of chemical energy usage over the course of the heat and the corresponding link to the process. The goal of the modeling effort explained above is to provide a tool that relates dynamic off-gas measurements to important steel-making considerations and the operation of the furnace in general. The following modules are being developed to extend the capabilities of the Goodfellow EFSOP® system. Enhanced water-leak detection: Above was described an empirical method determining the possible incidence of water-leaks in the EAF. The gas-phase model, through a hydrogen balance is able to calculate water in-leakage into the EAF directly. Some tuning will still be necessary to account for wet scrap, flashing of oils, etc. but it is expected that the method will be more accurate than the empirical approach presented. Cost-based post combustion: Presently, the control of post-combustion in the EAF is done through post-combustion ratios or other similar measures of the extent of combustion in a fee-back loop to oxygen and methane rates. These simple methods do not take into account the fact that the efficiency of post-combustion varies over the course of the heat. To account for this, constraints on the maximum and minimum allowable rates of oxygen and methane are varied over the course of the heat. For example, if post-combustion is deemed inefficient, then the maximum allowable oxygen rate is set lower to minimize oxygen usage during that particular interval. The mass/energy model may be used to estimate the efficiency of combustion within the EAF directly. This parameter, in turn, along with costs for oxygen, methane and energy may be used to calculate the marginal benefit per unit of oxygen or methane. Feedback to the burners can then be based on the cost/benefit of oxygen or methane usage accordingly. Fume system control: In many EAF shops, the threat of explosions within the de-dusting system and other safety concerns have resulted in over-drafting of the EAF. A larger than necessary ballast of air entering the EAF takes with it valuable energy and represents a considerable in-efficiency in the operation of the EAF. A dynamic measure of off-gas composition, along with a mass/energy balance makes it possible to calculate the amount of in-leakage air and the heat-load on the EAF fume system. This information may be used to minimize (within safety constraints) the amount of air-in leakage into the EAF.

Efficiency of burners and fixed-wall injectors: Oxygen injected into the EAF will either facilitate combustion within the freeboard or participate in decarburization/oxidation reactions within the slag or steel bath. The efficiency with which an injector is able to decarburize the steel bath (or alternatively the efficiency of a burner to provide post-combustion) is dependent on many parameters (e.g. angle of injection, location in the freeboard, jet velocity, flame shrouding affects, etc). Furthermore, the efficiency of an injector changes over the course of the heat depending on things like bath height, slag depth, scrap loading and type. The mass/energy balance makes it possible to evaluate the true efficiency of the burner over the course of the heat and provide feedback to the closed-loop control of oxygen and methane. Furthermore, the information may be used by operators when selecting, evaluating and tuning injectors for use within their EAFs. Dynamic Melting Profile: EAFs are paced according to either electrical kWh or kWh/ton clock. Typically, the melting and refining process is divided into stages by the electrical kWh clock and burners, injectors and electrode regulation systems are stepped through pre-defined setpoints for each stage. As steel-makers rely more and more on chemical energy within their EAF, pacing of the furnace solely on electrical energy introduces inconsistencies and inefficiencies in the operation. There have been many attempts to pace the furnace on the total energy usage (chemical and electrical) using nominal values for heats of reaction within the furnace freeboard. This method has limitations in that the efficiency of combustion changes over the course of the heat and even from one heat to the next. The mass/energy balance calculates directly the amount of heat transfer from the gas-phase and so the EAF can be paced according to actual energy (electrical and chemical) that has been transferred to the scrap and steel bath.

REFERENCES 1. Marshall Khan, Howard D. Goodfellow, Joe Maiolo, “Optimization of EAF Practices Based on Real Time Off-Gas Chemistry

Analysis Using Goodfellow EFSOP®”, SOHN Symposium, San Diego, CA USA, August 2006. 2. Howard D. Goodfellow, Luis Ferro, Paolo Galbiata, “ Operating Results at EAF Steelplants Using Goodfellow EFSOP®

Technology”, 8th EESC Conference, Birmingham, UK, May 2005. 3. http://www.steelonthenet.com/commodity_prices.html

Documents

Developments Towards an Intelligent Electric Arc Furnace