70 Technical Article

Increasing EAF Efficiency at

Nucor Steel–Arkansas

Maintaining furnace efficiency

and adapting to different
circumstances is crucial in
N ucor Steel–Arkansas (NSA) is
always looking for opportuni-
ties to improve efficiency and cost
diameter and 150-ton tap capacity
while maintaining a 75-ton hot heel.
The capacity of the transformer is
order to attain sustainable at its electric arc furnaces (EAFs). 81.496 MVA. The EAFs are limited
results. This paper discusses In 2015 when the market slowed to a potential of 950 V and a cur-
the implementation of the down, NSA re-evaluated some of rent of 140 KA. Each furnace is
SmartFurnace system to increase its melting practices, including slag equipped with a top-feed system
electric arc furnace practices, voltage versus KA, direct with a capacity of 300 tons/hour,
efficiency. reduced iron (DRI) feeding and car- which is used for DRI, carbon, lime
bon usage. NSA already had incor- and dolomitic lime.
porated some newer systems, like The furnace is equipped with four
the RotoFeed for carbon injection PTI burners with a maximum flow
and the Zolo TDLAS offgas ana- capacity of 2,000 scfm of oxygen
lyzer; the focus was on maximizing and 300 scfm for natural gas; Fig. 1
their potential. Results indicate an shows the position of the burners
improvement in power, carbon and and the roof feeding.
electrode consumption.
Slag Practices — Increased slag foam-
ing in the EAF has not only lowered
Discussion refractory and electrode consump-
tion, but has also had a significant
Furnace Description — NSA has two impact on improving thermal effi-
DC furnaces that have a 26-foot ciency and decreasing melting time.

Figure 1
Authors
Ismael Valdez (top row, left)
Nucor Steel–Arkansas,
Blytheville, Ark., USA
[email protected]

Garrick Wilson (top row, right)

Nucor Steel–Arkansas,
Blytheville, Ark., USA
[email protected]

Aaron Brezinski
Nucor Steel–Arkansas,
Blytheville, Ark., USA
I IRON & STEEL TECHNOLOGY I AIST.ORG

[email protected]

Matthew Clayton
Nucor Steel–Arkansas,
Blytheville, Ark., USA
[email protected]

Jim Cole
Nucor Steel–Arkansas,
Blytheville, Ark., USA
[email protected]

Saul Gonzalez (bottom row)

NOV 2017

AMI GE, Monterrey N.L., Mexico

[email protected] Electric arc furnace (EAF) layout.
 71

The process of slag foaming can be easily explained changes, or the quantity of DRI used on a particular
as reactions that generate gas bubbles combined with heat, operators can make adjustments in the feed rate
proper slag chemistry sustain these bubbles.1 windows to maintain an optimized practice.
With respect to slag chemistry, most electric fur- In early 2013, NSA implemented a more defined
nace slagline refractories are basic, which requires slag practice. MgO supersaturation of 1–4%, FeO
MgO and/or CaO saturation to minimize the chemi- content of 25–35% and 1.4–1.7 B3 ratio aims were
cal attack of the refractories. Additionally, saturation defined.
of the slag with these oxides (the presence of second The primary focus of the slag practice is to operate
phase particles) is an essential requirement for good at a lower basicity at the beginning of the heat. After
foaming. Therefore, based on required foaming prop- arcing for approximately half the process time of the
erties and the desire to protect the refractories, dual heat, B3 ratio is increased by adding lime. A slag with
saturation (CaO and MgO), or at least MgO satura- a lower basicity will foam earlier than a slag with a
tion, is required.1 higher basicity; the higher the basicity, the higher the
NSA uses the slag model developed by Pretorius and FeO+MnO requirements to reach a good foamy slag.
Carlisle1 to calculate MgO saturation, which is used The foam height depends primarily on the slag
as a reference to make changes in lime and dolomitic volume, slag composition and rate of gas evolution.
lime additions. Maintaining the foam requires control over the slag
Both furnaces are equipped with flux feeding composition and continued production of CO gas.
systems. This greatly enhances the ability to control The main change in slag composition during the
the slag in the furnace during the heat as DRI is fed. refining stage is a change in FeO content. By control-
When the composition or metallization of the DRI ling the FeO content of the slag, the foaming proper-
ties of the slag are thus controlled.
The generation of gas comes from the decarburiza-
tion of the metal bath and due to reaction of injected
Figure 2 carbon and FeO in the slag. It is believed that bubbles
generated from the decarburization reaction in the
35
metal are fine and well dispersed, which is ideal for
30 foaming. The bubbles formed by the reaction of FeO
25
and injected C in the slag are equally important. As
the carbon content of the bath is continually reduced
%FeO+MnO

20
during refining, the evolution of gas from the bath
15 also declines and the bubbles generated in the slag
10 from carbon injection become more important to
maintain foaming.
5
In the mass transfer–controlled region of decar-
0
1 1.5 2 2.5 3 3.5
burization, not only does the gas evolution from the
B3 Ratio metal decline, but the FeO content of the slag can also
increase. This increase in FeO content will lead to a
FeO+MnO required for dual saturation. more fluid slag and the potential decrease in foam
stability by decreasing slag viscosity (sever-
ity will depend on the conditions on the
Figure 3 isothermal stability diagram (ISD); see
Fig. 3). By matching the carbon injection
18 to the excess oxygen and thus FeO genera-
Good Foamy Slag
16 tion, the FeO content is maintained and
14 Mediocre Foamy Slag gas generation increased. The reduction
NOV 2017

12 of FeO by injected C is thus important to

maintain the foaming slag, both by con-
%MgO

10
I IRON & STEEL TECHNOLOGY I AIST.ORG

8 trolling slag composition and generating

6
Poor Foamy Slag
the required gas bubbles. Fig. 3 shows the
4 effect of excess FeO on foaming.
2 At the beginning of 2015, NSA started
0 to focus on decreasing B3 ratio on the
0 10 20 30 40 50 60 slags to maximize foaming properties in
%FeO + %MnO the slag.

Isothermal stability diagram (ISD) for good slag foaming.

 72 Technical Article
Figure 4
between heats. During that time,
50.0 adjustments to the K As were
2.0

made in order to adjust to the

45.0 1.8 slow market. Productivity was not
a priority; efficiency was. At the
40.0 1.6 beginning, a decrease in electrode
consumption was not observed
FeO%

since the time in between heats

B3
35.0 1.4

affected kWh/ton and power-on

30.0 1.2 times. It was not until later in the
year when results were observed.
25.0 1.0
With better slag control, arc cover-
age was achieved and an increase
20.0 0.8
in average voltage was possible.
Jan
Mar
May
Jul
Sep
Nov

Jan
Mar
May
Jul
Sep
Nov
Jan
Mar
May
Jul
Sep
Nov
Jan
Mar
May
Jul
Sep
Nov

Jan
Mar
May
Jul
Sep
Nov
Jan
Both EAFs operate with the
2012 2013 2014 2015 2016 2017
EAF1 Avg. Fe2O3 EAF2 Avg. Fe2O3 EAF1 Avg. B3_Ratio EAF2 Avg. B3_Ratio
SmartFurnace system from
AMI GE. This system measures
Average FeO and B3 ratio trend. arc stability, which provides an
advantage in helping to identify
the heat stages and slag condi-
tions during the heat. This mea-
surement is used to control volt-
Effect of Good Slag Practices — Good foaming practices age (arc length) during the heat and maximize power
can decrease heat losses to the sidewalls and achieve input. As arc stability is achieved, the voltage can be
better heat transfer and less energy losses, increased increased.
arc stability, and the ability to increase arc length and In DC furnaces it is critical to take into account
power input. the sidewall heating events. The main concern is the
Electrode consumption is directly proportional to natural hot spot caused by arc deflection. The system
time of the heat (power-on time) and average current. takes into account the temperature changes on the
With better arc coverage, an increase in average volt- water-cooled shell and its rate. This is an extra protec-
age is possible, which helps to reduce power-on times. tion in order to maximize arc length without causing
If there is not a need for lower power-on times, the damage to the shell.
current (KAs) can be decreased to reduce electrode
consumption and maintain the same average power How To Achieve Good Slag Practices — Earlier in this
input. paper, the importance of FeO control during the
At the beginning of 2015, the slow market affected heat was discussed. FeO has a direct impact on slag
EAF efficiency, since longer breaks were needed

Figure 5 Figure 6
3.0 900 130 40
ElectCons lb/ton Avg Volts Avg KA Avg Pwr On min
2.8
38
850
2.6 125 36

2.4
I IRON & STEEL TECHNOLOGY I AIST.ORG

800 34
ElectCons lb/ton

Avg Pwr On min

Avg Voltage

2.2 120 32
Avg KA

2.0 750
30

1.8 115 28
700
1.6
26

1.4 110
650 24

1.2
22

1.0 600 105 20

Jan-14
Feb-14
Mar-14
Apr-14
May-14
Jun-14
Jul-14
Aug-14
Sep-14
Oct-14
Nov-14
Dec-14
Jan-15
Feb-15
Mar-15
Apr-15
May-15
Jun-15
Jul-15
Aug-15
Sep-15
Oct-15
Nov-15
Dec-15
Jan-16
Feb-16
Mar-16
Apr-16
May-16
Jun-16
Jul-16
Aug-16
Sep-16
Oct-16
Nov-16
Dec-16

Jan-14
Feb-14
Mar-14
Apr-14
May-14
Jun-14
Jul-14
Aug-14
Sep-14
Oct-14
Nov-14
Dec-14
Jan-15
Feb-15
Mar-15
Apr-15
May-15
Jun-15
Jul-15
Aug-15
Sep-15
Oct-15
Nov-15
Dec-15
Jan-16
Feb-16
Mar-16
Apr-16
May-16
Jun-16
Jul-16
Aug-16
Sep-16
Oct-16
Nov-16
Dec-16
NOV 2017

Electrode consumption and average voltage, EAF1. Power-on time and average KAs, EAF1.
 73

Figure 7 Figure 8
3.0 900 130 40
ElectCons lb/ton Avg Volts Avg KA Avg Pwr On min
2.8 38
850
2.6 125 36

2.4 34

Avg Pwr On min

800
ElectCons lb/ton

Avg Voltage
2.2 120 32

Avg KA
2.0 750 30

1.8 115 28
700
1.6 26

1.4 110 24
650
1.2
22
1.0 600 105 20
Jan-14
Feb-14
Mar-14
Apr-14
May-14
Jun-14
Jul-14
Aug-14
Sep-14
Oct-14
Nov-14
Dec-14
Jan-15
Feb-15
Mar-15
Apr-15
May-15
Jun-15
Jul-15
Aug-15
Sep-15
Oct-15
Nov-15
Dec-15
Jan-16
Feb-16
Mar-16
Apr-16
May-16
Jun-16
Jul-16
Aug-16
Sep-16
Oct-16
Nov-16
Dec-16

Jan-14
Feb-14
Mar-14
Apr-14
May-14
Jun-14
Jul-14
Aug-14
Sep-14
Oct-14
Nov-14
Dec-14
Jan-15
Feb-15
Mar-15
Apr-15
May-15
Jun-15
Jul-15
Aug-15
Sep-15
Oct-15
Nov-15
Dec-15
Jan-16
Feb-16
Mar-16
Apr-16
May-16
Jun-16
Jul-16
Aug-16
Sep-16
Oct-16
Nov-16
Dec-16
Electrode consumption and average voltage, EAF2. Power-on time and average KAs, EAF2.

foaminess. Having a high concentration on the slag

will be as detrimental as not having enough. Figure 9
The FeO level in the slag is controlled by the follow-
ing two key reactions:
The formation of FeO from oxygen injection:

Fe + 1/2 O2(g) = FeO

The reduction of FeO by carbon injection:

FeO + C = Fe + CO(g)
Zolo system probes, EAF1.
The injection of carbon to reduce FeO and form
CO gas is also the main foaming reaction in the slag.
The better the slag foam, the better the control of the
FeO levels in the slag. The amount of carbon and oxy- The single multiplexed light path is pitched across
gen required to meet a specific steel and slag target is the combustion space via a SensAlign Head. A second
very important but not easy to predict. Scrap and DRI SensAlign Head is used as a catch head to receive
chemistry variation can make this task a challenge. the light and send it back to the control rack for de-
When the composition or metallization of the DRI multiplexing and signal processing.
changes, or the quantity of DRI used on a particular The Zolo TDLAS offgas analyzer provides real-time,
heat, the contribution of carbon into the bath may in-situ, simultaneous measurement of temperature,
change considerably and is very dependent of the DRI CO, CO2 and H2O at an agreed location near the
rate used and the heat stage. These variations of DRI fourth hole gap.
will directly change the emissions of CO during the
process. Carbon Injection Control — The installation of the off-
NSA installed the Zolo offgas system in order to gas system and the integration into the SmartFurnace
NOV 2017

integrate the information into the SmartFurnace system required a precise control for carbon and
system to control oxygen and carbon usage during oxygen flows to achieve full benefits. NSA decided to
I IRON & STEEL TECHNOLOGY I AIST.ORG

the heat. Also upgraded was the carbon injection install the system offered by Schenck Process.
system to a RotoFeed system, which provides better The old carbon system operated with a butterfly
flow control. valve and flow control was not good enough. Figs. 10
and 11 show the comparison between the old carbon
Zolo TDLAS Offgas Analyzer Description — ZoloSCAN system and the RotoFeed.
is a tunable diode laser absorption spectroscopy
(TDLAS) technology with proprietary multiplexing Implementation Into the SmartFurnace — The offgas
technology. Multiplexing enables many lasers to be module control takes the information from the off-
combined onto one single light beam (or light path). gas analyzer and validates all input variables such as
 74 Technical Article
Figure 10
operational limits, quality of the
signal and process the variables
in order to have a robust control.
During the process, it is a
challenge to estimate what the
amount of CO emissions emit-
ted by the steel process will be
and to determine an acceptable
range considering the needs and
natural behavior of the process.
The offgas control module esti-
mates the COCO2 ratio = (CO2%/
(CO%+CO2 %))·100 in the gas
emission as a function of the
Old carbon injection system control. DRI rate flow and other variables
like the percentage of the total
heat weight in the furnace, car-
Figure 11 bon injection flow, natural gas
and oxygen flows, energy con-
sumption, and temperature of the
offgas.
Using the estimated COCO2
ratio, the offgas control module
will take action optimizing the
carbon usage in coordination
with the other control modules of
the SmartFurnace to guarantee
the best performance in the EAF.
The offgas control module
individually modifies each car-
bon injector to ensure the best
efficiency. Fig. 13 concerns the
operation on the carbon injec-
tor in burner No. 4 and shows
RotoFeed carbon system control. the estimated COCO2 ratio
(EstCOCO2Ratio), the COCO2
ratio, the total DRI flow and car-
Figure 12 bon reference offgas corrected
(B4CCarbRefOff).
Arc Stability Fig. 13 shows a clear example of the
(add more if response of the CO to CO2 ratio from the
stability is bad)
offgas probe when the DRI flow setpoint
is changing.
FeO From Slag Flames Zolo
Base Carbon Sample Camera (monitors The system compensates up to a cer-
Profile (add or reduce (reduce setpoint CO/CO2 ratio to tain point for variations in DRI quality.
(kWh base) depending on if high flames, maximize
The main variation in DRI quality is the
I IRON & STEEL TECHNOLOGY I AIST.ORG

aim FeO%) increase draft) efficiency)

carbon content, and when less carbon is
present in the DRI, there are also less CO
DRI Input
(reduce setpoint
emissions. The system compensates for
depending on these variations based on the offgas probe
carbon input) feedback.

Logic to control injection carbon.

Detection of Water Leaks — An added
benefit from the SmartFurnace system
integration is the ability to detect an
NOV 2017

abnormal quantity of water vapor inside

the EAF in the different stages of the heat.
 75

Figure 13
The system estimates the actual
amount of water vapor inside the
furnace using mainly the H2O%
measured by the offgas analyzer
near to the fourth hole gap.
The offgas control module esti-
mates the amount of water vapor
and then compares the estimated
value with the calculated value
from the offgas analyzer.
The number of variables used
to estimate the water vapor inside
the EAF is quite large. The offgas
control estimates the water vapor
in function of total natural gas
flow, total electrode spray water,
electrode position, offgas temper-
ature, air intake, furnace pressure Offgas control module controlling the COCO2 ratio.
and others.
Fig. 14 shows the real water
vapor calculated and estimated
during a heat where no abnormal Figure 14
situations were reported.
Fig. 14 shows an accurate
estimation of the water vapor
(EstimH2OPct) and the calculat-
ed values from the offgas analyzer.
Fig. 15 shows the same variables
but with an abnormal water vapor
event.
A major difference between the
real amount of water vapor going
outside the EAF and the estimat-
ed value can be observed. During
this heat an important water leak
from a burner box was detected. Estimated and real water vapor by offgas control module.

Results — Tables 1 and 2 show

the results of the integration of
the offgas system, the closed loop Figure 15
with the SmartFurnace and the
improved slag practices.
The better slag practices and
FeO control allowed for increased
voltage and reduced KAs with-
out a big difference in power-on
NOV 2017

times. This helped relieve some of

the stress to the bottom electrode
I IRON & STEEL TECHNOLOGY I AIST.ORG

by running longer arcs. This in

turn allowed for the number of
heats per EAF campaign to be
increased.

Estimated and real water vapor with an abnormal water vapor detection.
 76 Technical Article
Table 1 Table 2
EAF1 Results With the SmartFurnace Optimization EAF2 Results With the SmartFurnace Optimization
No. of heats 2,020 2,020 No. of heats 605 617
DRI 15–35% Off On Difference % difference DRI 17–23% Off On Difference % difference
USD/ton 27.30 26.40 –0.90 –3.30 USD/ton 27.36 25.71 –1.65 –6.03
Power-on 34.78 24.10 –0.68 –2 Power-on 34.87 34.04 –0.83 –2.83
Cinj/ton 36.71 30.45 –6.26 –17.10 Cinj/ton 33.70 28.50 –5.20 –15.40
Oxy/ton 1,216 1,135 –81 –6.67 Oxy/ton 1,153 1,103 –50 –4.30
Ngas/ton 116.10 115.50 –0.60 –0.50 Ngas/ton 117 109.40 –7.60 –6.40
kWh/ton 333.40 327.90 –5.50 –1.60 kWh/ton 329.70 330.60 –0.90 –0.27

Conclusions

Furnaces represent the largest cost and energy usage

in steel mills. Maintaining efficiency and adapting to Figure 16
different circumstances is crucial in order to attain 1600
sustainable results. 1400
Teamwork between production and maintenance
1200
teams has been the biggest contributor to the success
of this project. NSA operators are highly skilled and 1000

were comfortable controlling the voltage, carbon and 800

oxygen in the furnace by looking at different process
600
variables and listening to the sounds produced. The
implementation of the SmartFurnace system took into 400

consideration the different ideas of each operator in 200

order to create the right logic to control the furnace. 0
Today, a standard practice has been implemented for
Jan-13
Mar-13
May-13
Jul-13
Sep-13
Nov-13
Jan-14
Mar-14
May-14
Jul-14
Sep-14
Nov-14
Jan-15
Mar-15
May-15
Jul-15
Sep-15
Nov-15
Jan-16
Mar-16
May-16
Jul-16
Sep-16
Nov-16
Jan-17
all operators yielding the results shown in this paper.
The capability to adapt to all changes caused by
scrap and DRI usage has been enhanced by making Heats on bottom electrode per EAF campaign.
the process safer and more efficient.

References

1. E. Pretorius and R. Carlisle, “Foamy Slag Fundamentals and Their

Practical Application to Electric Furnace Steelmaking,” Iron and
Steelmaking, No. 10, 1999, pp. 79–88. F

This paper was presented at AISTech 2017 — The Iron & Steel Technology
Conference and Exposition, Nashville, Tenn., USA, and published in the
Conference Proceedings.
I IRON & STEEL TECHNOLOGY I AIST.ORG
NOV 2017