Implementation of PyreJet Technology in DRI Based Electric Arc by rih47632

VIEWS: 9 PAGES: 9

									  Implementation of PyreJet Technology in DRI Based Electric Arc Furnace Process at OEMK (Russia)

                    N. A. Shlyahov, E. I. Gontaruk, V. P. Sidorov, Y. A. Zatakovoy, V. I. Fomin, V. G. Kobernik
                                                          OEMK
                                                     Stary Oskol – 15
                                                      Russia 309515
                                                 Tel.: +7 (0725) 374 499
                                                 Fax: +7 (0725) 329 429
                                             E-mail: techdep@oemk.oskol.ru

                                                          V. Shifrin
                                         ACI, division of Air Liquide America L.P.
                                              200 Chastain Ctr Blvd Ste 295
                                                   Kennesaw, GA 30144
                                                    Tel.: (678) 354 8232
                                                    Fax: (678) 354 8235
                                           E-mail: val.shifrin@airliquide.com


                                                         INTRODUCTION

Oskol Electrometallurgical Kombinat (OEMK) is one of the leading Russian manufacturers of high quality steels. Its annual capacity
is 2.2 millions of tonnes of steel using up to 100% of DRI as raw material. OEMK is planning to increase production to 3.0 million
tonnes while reducing number of Electric Arc Furnaces (EAFs) from four to three. One of the directions towards achieving this goal is
intensification of the EAF process due to introduction of additional chemical energy by means of modern technologies. After studying
the experience of several leading steel works of the world in this area, OEMK selected PyreJetTM technology from American
Combustion. OEMK decided to install one system on EAF #4 for the purpose of technology evaluation in light of medium to long
term steel production development. This article will discuss the principles of the technology, design, implementation and performance
results of first year of operation.

                                        DESCRIPTION OF OEMK AND ITS PRODUCTS

OEMK is the first stand alone Russian mini-mill that produced first steel in 1984. Currently the plant includes the following shops:

         DRI facility, that makes DRI pallets up to 94% metallization utilizing MIDREX process,
         melt shop producing 300 x 360 mm blooms,
         rolling mill 700 for large section long product production: rounds of 80 – 180 mm diameter, and squares 80 – 170 mm,
         bar mill 350 for 12 – 80 mm diameter bar production,

The melt shop of 2.2 tonnes annual capacity includes the following equipment:

         four EAFs of 150 tonnes capacity equipped with 90 MVA transformers,
         four argon bubbling stations by means of combination of top and bottom blowing. These stations are also equipped with wire
         feeders for microalloying and calcium treatment,
         two re-circulation type vacuum degassers,
         two ladle furnaces with 24 MVA transformers, that are also equipped with bottom argon blowing, top injection lances and
         wire feeders,



                                            AISTech 2004 Proceedings - Volume I                                               915
         four 4 – strand radial (12 m) continuous casting machines for 360 x 300 mm bloom production,
         four walking beam controlled cooling furnaces to cool the blooms to 300oC,
         two inspection/scarfing lines,

OEMK product mix consists of almost 2000 different steel grades. Major product applications include oil and boiler pipe, construction
steels, automotive steels (VW 4220, HS-41D20, HS 595, etc.), spring steel, ball bearing steel, as well as cord wire, rail and tool steels.

                                                         PYREJETTM TECHNOLOGY

PyreJetTM technology involves installation of multifunction sidewall injectors primarily in cold spots around the perimeter of the EAF.
This technology was developed and patented by American Combustion (ACI)1,2. The main principles of PyreJetTM technology are
shown in Figure 1.


                                                               Uniform Scrap                     Simultaneous
                                                                   Melting                        O2 Lancing
                                                                                              and Carbon injection

                          Increased Metallic Yield
                     (allow time to reduce slag before flushing)




                                                                   PyreJet™
                                                                   PyreJet™                  Quick Arc Stabilization /
                                                                    Technology
                                                                                             Enhanced Slag Foaming
                        Keep Slag Door Closed /
                        Keep Slag in the Furnace


                                                         Increased Active Power Input
                                                             Better Arc Protection
                                                       Increased Refractory Protection


                                                      Figure 1: PyreJetTM technology cycle
In the beginning of each charge the injectors are operating in burner mode in order to quickly heat and melt scrap in cold areas of the
furnace, thereby creating conditions for supersonic oxygen lancing and carbon injection. Strong and rigid flame (Figure 2) is achieved
due to proper mixing of gases in a water cooled combustion chamber (Figure 3) that also protects the injector exit openings from
plugging.




    916                                         AISTech 2004 Proceedings - Volume I
                                                Figure 2: PyreJetTM in a burner mode

The body of the injector (Figure 3), including the supersonic oxygen nozzle, is made out of stainless steel and Inconel and is attached
to the combustion chamber and can be quickly detached, if necessary. The supersonic nozzle design provides for the supersonic jet
exit velocity of 2 – 2.1 Mach and the supersonic velocity is maintained for up to 1.8 m due to proper jet shrouding with concurrent
flame.




                                                                              Carbon Pipe




                                                                              Body with
                                                                              Nozzle




                                                                              Combustion
                                                                              Chamber




                                                Figure 3: Injector PyreJetTM assembly




                                            AISTech 2004 Proceedings - Volume I                                              917
The carbon injection pipe is located at the bottom of the combustion chamber. Such location allows utilization of the oxygen jet
energy for introduction of carbon to the slag – metal interface, at the same time protecting the carbon opening from plugging by
injector flame. Carbon and oxygen stream are located on different sides of the shrouding flame and, therefore, do not interfere with
each other until they come in contact with the metal.

Combination of supersonic oxygen and carbon injection leads to intensification of bath reactions, metal agitation and arc stability, that
in turn result in increased EAF efficiency and active power input achievable with the same transformer characteristics. At the same
time, the secondary voltage can be increased without an increase in heat losses or negative affect on refractory, provided that a proper
foamy slag is maintained throughout the flat bath operation. One of the noticeable advantages of PyreJetTM technology is the
capability of conducting almost entire heat with the closed slag door that results in better foamy slag retention in the furnace that often
leads to metallic yield improvement3.

                                                             DISCUSSION

Commissioning and Encountered Difficulties
PyreJetTM system was installed on EAF #4 in February of 2003. Prior to PyreJetTM system implementation EAF had water cooled
sidewall oxygen lance. Carbon was introduced via roof opening along with DRI and lime. Implemented PyreJetTM system consisted of
two sidewall injectors installed into copper panels as shown in Figure 4 & 5, oxygen and natural gas flow control valve trains, carbon
injection vessel with two injection lines, gas analyzer to determine off gas О2, СО, and СО2 concentration and computerized process
control system. The initial intention was to install one injector across from DRI feeding opening (between center and pit phases), but
this area is very difficult to access and, therefore, the injector was installed in the cold spot, between floor and center phases.




                                                       DRI




                                             Figure 4: PyreJetTM system layout (plan view)




    918                                      AISTech 2004 Proceedings - Volume I
                                           Figure 5: PyreJetTM system layout (cross-section)
Gas analyzer was installed out of safety considerations considering borderline baghouse capacity to alarm in case of explosive mix
formation. Explosive off gas mix was not detected after the system commissioning, however, due to the fact that oxygen consumption
doubled, the insufficient baghouse capacity created some additional difficulties during commissioning and further operations.

Another challenge in the course of commissioning and optimization was establishing consistent carbon flows to each injector. As was
mentioned above, one carbon vessel feeds both injectors. The fact that carbon lines are sufficiently different in length resulted in flow
variations between the injectors during automatic operation and prompted operators to periodically switch to manual carbon flow
control.

One of the characteristic features of the DRI operation is very low metal level at the beginning of oxygen lancing. In this case, by the
start of DRI feeding and oxygen injection, the distance from the supersonic nozzle to the metal was in excess of 2 m. Such large
distance was an obstacle to early oxygen injection, therefore, after a few months it was decided to lower the injectors in the sidewalls
by 200 mm for the purpose of increased oxygen efficiency and earlier onset of supersonic lancing.

Another characteristic features of OEMK melt shop are changing slag VB – ratio throughout the heat and almost complete deslagging
before tapping. The latter is necessitated by a combination of spout tapping and high cleanliness requirement of the end product. The
initial portion of fluxes is added during scrap melt down creating V – ratio in the 2.0 – 2.5 range. Further lime additions constitute 8%
of DRI amount and by the end of the heat the VB – ratio falls to 1.7 – 1.9. For the PyreJetTM operation lime addition profile was
slightly changed: for the first 60% of DRI the lime addition was 10% of DRI amount, for the remaining 40% it was 5% of DRI
amount. This lime addition practice provided more optimal slag composition and better foaming throughout the heat4 as well as the
ability to remove slag before tapping.

Results
It was expected that reduction in electrical consumption and power on time would result from the following:

         Increase in oxygen consumption by 13 – 15 nm3/T and proportionate increase in carbon consumption,
         Increase in EAF efficiency due to accelerated DRI melting and higher active power input,

The energy effect of additional carbon and oxygen should contribute approximately 30 kWh/T but the expected (and guaranteed)
reduction in electrical consumption was 60 kWh/T, thus remaining 30 kWh/T savings must be accomplished as a result of increased
EAF efficiency. It should be mentioned, that typical OEMK charge consists of over 50% of DRI and the scrap must fall no closer than
0.5 m to the sidewall, therefore, it was not expected to achieve a noticeable energy benefit from the natural gas combustion. The main
goal of the burner mode was to quickly preheat and melt scrap in front of the injectors and create conditions for early supersonic
oxygen and carbon injection. The main energy and power on time benefit was expected to come from increased DRI melting rate as a



                                            AISTech 2004 Proceedings - Volume I                                                919
result of more intensive input of electrical and chemical energy as well as more intensive bath agitation. Active power input and DRI
feeding profiles before and after system implementation are illustrated in Figures 6 & 7 respectively.



                                              75
                                                                                                                                                 Before
                                              70
                                                                                                                                                 After
                                              65
                           Active Power, MW




                                              60

                                              55

                                              50

                                              45

                                              40

                                              35

                                              30
                                                   1

                                                       6

                                                           11

                                                                16

                                                                     21

                                                                          26

                                                                               31

                                                                                      36

                                                                                            41

                                                                                                 46

                                                                                                      51

                                                                                                            56

                                                                                                                    61

                                                                                                                         66

                                                                                                                              71

                                                                                                                                   76

                                                                                                                                         81

                                                                                                                                                 86

                                                                                                                                                      91
                                                                                           Power on, min

                                                                                    Figure 6: Power input profile




                                   2500
                                                                                                                                        Before

                                                                                                                                        After
        DRI Feeding Rate, kg/min




                                   2000


                                   1500


                                   1000


                                        500


                                               0
                                                           11

                                                                16

                                                                     21

                                                                          26

                                                                               31

                                                                                      36

                                                                                            41

                                                                                                 46

                                                                                                      51

                                                                                                           56

                                                                                                                 61

                                                                                                                         66

                                                                                                                              71

                                                                                                                                   76

                                                                                                                                        81

                                                                                                                                                86

                                                                                                                                                      91
                                                   1

                                                       6




                                                                                           Power on, min
                                                                                Figure 7: DRI addition profile

Figure 6 indicates that maximum active power was raised from 65 MW to 68 – 69 MW (by 5%) due to an increase in secondary
voltage that became possible as a result of improved foamy slag practice. Active bath mixing, owing to supersonic oxygen injection in
two points, helped avoiding accumulation of unmolten DRI in the slag, that was somewhat anticipated. Thereby, the maximum DRI
feeding rate was increased from 1800 kg/min to 2200 – 2300 kg/min (approximately by 25%) that is unproportionately higher than


    920                                                                   AISTech 2004 Proceedings - Volume I
increase in active power input and constitutes an indirect proof of improved EAF efficiency. The combination of these technological
changes allowed to successfully achieve guaranteed reduction in electrical consumption of 60 kWh/T and exceed guaranteed reduction
in power on time of 12 min by over 1 min during the performance test.

OEMK EAFs have a spout tapping arrangement and, as was mentioned previously, the slag carry over is strictly limited. This
necessitates maximum deslagging during the refining period, that in turn results in significant reduction in power input because of
poor arc shielding at low slag volumes. Therefore, last 10 minutes of the heat (or 12%) are not suitable for chemical energy input at
this point in time and remain a potential for the future optimization.

Despite positive results in electrical consumption and power on time, the first month of operation after the commissioning resulted in a
considerable increase in electrode consumption (Figure 9) and a reduction in delta life. After the injectors were lowered by 200 mm in
April, the electrode consumption dropped below its pre-installation level.

The dynamics of monthly averages for 2003 in electrical consumption, electrode consumption and power on time are shown in Figures
8, 9 & 10 respectively.




                                                640
                                                                                                          EAF #1-3
                                                                                                          EAF #4
                Electrical Consumption, kWh/T




                                                620


                                                600


                                                580


                                                560


                                                540


                                                520
                                                      Jan Feb Mar   Apr May Jun          Jul   Aug Sep Oct Nov Dec
                                                                                   Month

                                                                Figure 8: Electrical energy consumption




                                                              AISTech 2004 Proceedings - Volume I                             921
                                    3.2
                                                                                            EAF #1-3
                                                                                            EAF #4
                                     3
      ELectorde Consumption, kg/T

                                    2.8


                                    2.6


                                    2.4


                                    2.2


                                     2
                                          Jan Feb Mar   Apr May Jun       Jul    Aug Sep Oct Nov Dec
                                                                     Month
                                                         Figure 9: Electrode consumption


                                    110
                                                                                            EAF #1-3

                                                                                            EAF #4
                                    105
      Power on time, min




                                    100



                                     95



                                     90



                                     85
                                          Jan Feb Mar   Apr May Jun        Jul   Aug Sep Oct Nov Dec
                                                                     Month


                                                            Figure 10: Power on time




922                                                AISTech 2004 Proceedings - Volume I
Table 1 illustrates comparison of main operating parameters of EAF #4 before and after PyreJetTM installation.

                                                            Before                After                  Delta
                            Parameter
                                                          (3 months)        (Jul – Dec 2003)      Absolute        %
                Electrical consumption, kWh/T                 610                  553             (57.0)        (9.3)
                Power on time, min                            96                   88               (8.0)        (8.3)
                Oxygen consumption, nm3/T                     14                   27               13.0         92.9
                Natural gas consumption, nm3/T                 0                   2.7               2.7
                Electrode consumption, kg/T                   2.4                  2.3              (0.1)        (4.2)


                                                    Table 1: EAF #4 performance
Comparative operating analysis for the period from July to December of 2003 between EAF #4 equipped with PyreJetTM system and
EAFs #1 – 3 equipped with the water cooled sidewall oxygen lances is presented in Table 2.

                                                                                                      Delta
                             Parameter                   EAF #1 – 3            EAF #4
                                                                                               Absolute      %
                  Electrical consumption, kWh/T             599.2               553.7           (45.5)      (7.6)
                  Power on time, min                        95.4                87.8             (7.7)      (8.1)
                  Electrode consumption, kg/T                2.5                 2.3             (0.1)      (5.9)

                                  Table 2: Comparison of EAFs’ performance at OEMK melt shop
One can see from the graphs and tables that average power on time on EAF #4 was reduced by 8 min as compared with 13 min during
the performance test. The main reasons were restraining influence of the insufficient exhaust system that prevents operators from full
transformer utilization, and lack of a dedicated ladle car at the EAF, leading to the reduction in power input prior to tapping due to
frequent wait on ladle. Currently a dedicated ladle car is being installed at the EAF that will allow either considerable reduction or
complete elimination of ladle delays and, thereby, further decrease in power on time.

Unfortunately, insufficient exhaust system capacity is also negatively affecting electrical consumption. Higher then allowed off gas
temperatures before the baghouse often force the operators to reduce or completely terminate injection of oxygen and carbon. Relative
to January 2003 electrical consumption on EAF #4 is reduced by 60 – 70 kWh/T, however, relative to other EAFs it is lower only by
45 kWh/T. Another negative influence stemming from the exhaust system is delta life. It was reduced from 150 to 100 heats per delta
and up to present day delta life remains an unresolved issue.

The electrode consumption on EAF #4 is consistently lower as compared to the pre-installation level as well as to the other furnaces in
the shop. The monthly fluctuations are associated with usage of different electrodes on different EAFs. The trend of lower electrode
consumption continued into 2004 dropping below 2.0 kg/T level by May.

                                                           CONCLUSION

Installation of the PyreJetTM technology on EAF #4 of OEMK melt shop proved good applicability of this technology to a high percent
DRI operation. It resulted in significantly reduced electrical consumption and power on time, increased active power input and DRI
feeding rate, and moderate reduction in electrode consumption. At the same time, as was anticipated, the increase in oxygen usage by
a factor of 2, additional carbon consumption, and increase in transformer active power as a result of improved foamy slag, exceeded
the capabilities of the existing exhaust system that somewhat constrains current EAF performance. However, positive results of the
first year of operation prompted OEMK to consider installation of similar systems on two other EAFs.

                                                           REFERENCES

1. US Patent 4,622,007 Variable Heat generating method and apparatus.
2. US Patent 5,599,375 Method for electric steelmaking.
3. G. Moraes, “Implementation of PyreJetTM technology in Electric Arc Furnaces at Siderurgica Barra Mansa” 60 EAF Conference,
San Antonio, USA, November 2002.
4. E. Pretorius and C. Crlisle, “Foamy slag fundamentals and their practical application to Electric Furnace steelmaking”, 1998
Electric Furnace Conference Proceedins, pp. 275 – 291.




                                            AISTech 2004 Proceedings - Volume I                                              923

								
To top