RECYCLING OF DUST FROM ELECTRIC ARC FURNACES--A CASE STUDY pP &-

E. Radha Krishnan, P.E., and William F. Remner PEI Associates, Inc. Cincinnati, Ohio

Dust generated from electric arc furnaces (EAF's) employed in steel- making plants is currently listed as a hazardous waste because of its toxic metgllic constituents, i.e., lead (Pb), cadmium (Cd), and hexavalent chromium (Cr 6). Disposal of the waste at controlled landfills is not only expensive, but also has associated long-term liability concerns. Recycling of pelletized electric arc furnace (EAF) dust from carbon and low-alloy steel production was evaluated through a comprehensive, experimental program at a participating steel plant. Equipment for pelletizing the dust was evaluated, and a 3-foot-diameter disc pelletizer with an 8-inch-diameter pin mixer was selected for the pilot trials. two separate blocks of testing. fresh pelletized EAF dust on power consumption and other important variables, whereas Block I1 focused on conducting a fate analysis of zinc and other heavy metals under continuous recycling conditions.

The experimental design was comprised of Block I focused on the effects of recycling

The tests demonstrated the feasibility of pelletizing and recycling EAF dust. recycling. increase in coke consumption was noted during recycling. operating conditions precluded the availability of high zinc-bearing scrap for the Block I1 tests, the zinc content of the dust was almost 50 percent lower than that in Block I. despite some degree of recycle. content of the dust, however, increased from 9 weight percent to 15 weight percent during the short continuous recycling period.

steel plants.

Power consumption increased approximately 4 percent during No significant change occurred in the tap-to-tap heat time.

Because plant An

The zinc

Preliminary economic .,analyses show that the recycling option may be a favorable one for many

The EAF dust recycling technology is immediately applicable; it re- quires only specific engineering design for a given steel plant installa- tion and the resolution of specific permitting requirements.

89

Introduction

Each year the electric arc steelmaking industry generates approxi- mately 500,000 tons of dusts containing valuable metallic resources such as iron. zinc, lead, and chr0mium.l becoming increasingly popular for the production of carbon and low-alloy steels and currently account for about 35 percent of total steel produc- tion. leachability of its toxfc constituents, i.e., lead (Pb), cadmium (Cd), and hexavalent chromium (Cr 6 ) . K061. remote from the point of origin. of $100 per ton of dust are quite common. standards of environmentally acceptable disposal, long-term liability is a concern. options for handling the dust.

-

These electric arc furnaces (EAF's) are

~

EAF dust is currently listed as a hazardous waste because of the

Its EPA-assigned hazardous waste number is -

Disposal is becoming costly as disposal sites become more scarce and Disposal costs (including transportation)

Furthermore, even with current

This has led the steelmaking j-dustry to look to other viable

Alternatives to the landfilling option for EAF dust management include chemical fixation, regional recovery processes, and onsite recycling. The major shortcomings of the various processes suggested for recovering metal- lic values from EXF dust have been the high capital cost of the equipment and the need for large quantities of dust for the processes to be economi- cal. Furthermore, the recovery processes require dusts with zinc contents greater than 20 percent to be economical. EXF dusts from carbon and low-alloy steel production typically range from 10 to 35 percent and 1 to 5 percent, respectively. FAF dust is generally not amenable to regional recovery processes because it is generated i n small quantities at a large number of locations, distant from industrial centers where regional recycling might occur.

The zinc and lead contents of

Pelletizing and recycling to the furnaces is a logical alternative for managing EAF dust. tion of a dust enriched in volatile elements such as zinc and lead; this blowdown dust stream could be removed for sale to zinc smelters. The main advantages of the recycling process are its low capital cost, relative insensitivity to EAF dust composition, and applicability to both mini-mills and large mills.

The recycling operations should result in the genera-

An experimental program was designed to obtain data under controlled conditions for investigation of several critical issues for the commercial application of recycling: 1) the partitioning (upon recycling) of heavy metals such as zinc, lead, and cadmium between the slag and the dust generate 2) the effect of recycling on energy consumption; 3) the effect on steel quality; and 4) the economics of recycling. The tests were conducted at the Green River Steel EAF shop in Owensboro, Kentucky, during the summer of 1985.

Green River's Owensboro site has two 65-ton electric arc furnaces rated at 24,000 kVA each (370 kVA/ton). specialty carbon steels. injected pneumatically. The furnaces are equipped with side-draft hoods, and canopy hoods in the roof vented to a 14-compartment baghouse that discharges into screw convey- OKS. charges into a flexible tube. The dust then falls about 6 feet through the tube into a large sealed rolloff box situated on the ground. filled, this box is hauled to a hazardous waste facility.

The furnaces produce low-alloy and Oxygen is injected with hand-held pipes, lime is Three to four charges of scrap are used per heat.

These conveyors ultimately join into one cross conveyor that dis-

After being

90

Evaluation of Pelletizing Equipment

Prior to plant selection, fresh dust was obtained from two operating EAF shops representing low-alloy medium carbon steels (Plant A) and plain carbon steels (Plant B). Manufacturers of pelletizing equipment were contacted for conducting pelletizing tests. with water is often called “greenballing.” In this paper, however, we refer to the product as pellets and the process a s pelletizing.) the bench-scale test results a 3-ft.-diameter disc pelletizer and hopper with a 8-in.-diameter pin mixer and a 3-ft” live bin feeder was selected for conducting the pilot-plant studies.

Characterization of EAF Dusts

(The agglomerating of EAF dust

Based on

PEI performed Bahco analyses for analyzing partir’e size distribution of two EAF dusts. fine particles and that approximately 50 percent of the dust was below 5 microns.

The results showed that the EAF dusis were composed of

The variability of important metallic constituents in the dust is illustrated in Table I . during a heat, depending on the materials charged, stage of the heat, and other factors. insensitive to changes in both the chemical and physical characteristics of the dust.

The EAF dust analysis varies from heat to heat and

Consequently, the recycling technique must be relatively

Pelletizing Operation

The pelletizing process flow diagram for the Green River installation is illustrated i n Figure 1. The temporary storage hopper, which is com- pletely enclosed, receives the EAF dust from the baghouse. The pelletizing operations were enclosed in a temporary shelter. The pellets produced were stored either in cardboard drums or wooden boxes lined with polyethylene, All the pellets produced were transferred and stored in a covered warehouse before charging into the furnaces. regulatory requirements.

These arrangements met the necessary

Dust composition appears to be the most important parameter affecting pellet quality. Fortunately, variations in size ( < 1 / 4 in. to > 3 / 4 in.) did not affect the efficacy of recycling.

The initial arrangement of the system included the pin mixer, which was believed to offer the advantage of premixing water with the dust to start the slaking of the lime in the dust before it entered the pelletizer. After several trials, however, the pin mixer became clogged with wetted dust, which hindered the process; therefore the mixer was removed. Even though the Green River EAF dust varies from 10 to 20 percent lime, the pellet quality was not noticeably different without the pin mixer.

The moisture content of the pellets was approximately 15 percent. Pellet strength tests were mostly qualitative. i.e., the pellets were squeezed by hand. the pellets on a scale, and the pellets showed breaking strength of 9 to 10 lb. (green strength). Pellets from the pelletizer as made were strong enough to withstand shoveling and/or dropping into the container. aging for 24 hours, the pellets increased in strength. Tests on three different batches (of different sizes) showed a crush strength of 1 8

In some cases, however, a flat strip was used to press

After

91

to 25 lb. 48 hours of aging. and charging to the furnaces.

Recycling Operation

The same batches showed a crush strength of 35 to >80 1b after The strength of the pellets was adequate for conveying

Transportation of pellets to the furnace and the timing of their charging are important for a successful recycling program. Preferably, the pellets should be charged into the furnace through the use of a separate box or skip pan. the molten metal from the meltdown of scrap charged earlier. tates quick assimilation of the pellets into the bath and reduces flash off as dust. The scrap backcharge can take place immediately after the pellets are charged. pellets between the scrap in the charging bucket.

They should be charged while a slag layer I s present on This facili-

Another variation in charging consists of "sandwiching" the

At Green River, the recycling operation involved moving the drummed or crated pellets by a forklift to a warehouse area and then to the melt shop floor, where they were dumped into a skip pan. and emptied into the furnace by crane. the second charge of scrap (first backcharge).

Results

The skip pan was weighed The pellets were charged prior to

Results of pellet recycling were analyzed by comparison with collected baseline data to determine the effects on heat time, power consumption, yield, steel quality, carbon and electrode consumption, and lime and ferro- alloy consumption.

The tests were conducted in two blocks. The main objective of the Block I tests was to determine the effect of recycling fresh once-through pellets on power consumption and other furnace variables, whereas the primary objective of the Block I1 tests was to determine the fate of heavy metals under continuous recycling conditions. Table I1 summarizes the data for the heats with pellet charging In Blocks I and 11. operation was not continuous during the experiments, some heats were made after the furnace had sat idle for an entire turn, whereas others were made Immediately after the previous heat. This situation affects apparent power use in kWh/ton and heat time regardless of pellet use. The heats are therefore divlded into these two categories ("first heat", i.e., cold furnace versus "not first heat", i.e., continuous operation), for more accurate analysis. Table I11 summarizes similar data for baseline heats, i.e., heats without pellet charging. obtained from the plant's heat log sheets.

Because the shop

The data for these heats were largely

During the Block I test period, both furnaces operated 16 hours per day. During the night turn, the furnace roof was kept open and the furnaces were allowed to cool down. I1 tests include a variety of heats representing the normal product mix.

The data collected during the Block I and Block

The Block I1 tests were limited to 4 days of continuous recycling because of the plant operating schedule. The duration of the Block I1 tests was not long enough to determine blowdown dust requirements. the Block I1 test period, pellets were made directly from the baghouse catch and recycled to the furnace. during this period. fate of zinc. Block I1 tests.

During

Two furnaces operated 16 hours per day Slag and dust samples were collected to evaluate the

Table IV presents the EAF dust analysis for the Block I and

92

The composition of the dust generated was fairly consistent during Block I due to the intermittent nature of these tests. In the Block I tests, the zinc content of the dust generated varied between 20 and 28 percent based on composite samples.

A long shutdown occurred at the plant prior to the Block I1 experi- ments. Because of Uncertain business conditions, inventory was reduced by consuming all available scrap in the plant during the Block I1 tests with- out regard for a balanced scrap mix. The zinc content of this cleanup scrap was significantly lower than that of the scrap usually used. No high-zinc scrap was observed during the Block I1 testing; consequently, the zinc content of the recycled dust was lower than that previously observed in Block I despite the recycling. An increase in the zinc content of the dust from 9 weight percent to approximately 15 weight percent, however, was observed during the continuous recycling period. No concomitant increase, however, was observed for the other volatile metals (e.g., Pb. Cd) in the dust. Green River EAF dust. 0.01 percent in all samples for both the Block I and Block I1 tests.

This is probably due to the very low levels of Pb and Cd in the The zinc and lead content of the slag was less than

Results of pellet charging on heat time are inconclusive. Heats with pellets on a cold furnace took longer while heats with pellets during continuous operation were shorter. In neither case, however, are these differences statistically significant.

Power consumption for heats on a cold furnace was about ten percent higher than during continuous operation whether pellets were used or not. The use of pellets caused an increase in power use of about 4 percent; this was consistent for both cold furnace heats and continuous operation but the test of statistical significance is conclusive only for continuous opera- tion. This increase in power consumption is consistent with results re- ported in previous research.2

Average total yield for the heats to which pellets were added was not significantly different than that for the baseline heats (no pellets added). It should be noted, however, that several factors other than pellet addition affect yield.

To assess the effect of recycling on steel quality, the plant kept a special vigil and followed all the heats charged with pellets from the ingot stage through finishing. Recycling had no deleterious effects on the steel quality. This is based on the results of extensive quality control tests conducted by the plant to determine hardness, internal cleanliness, and content of residual metals in the steel.

Because the plant operators were concerned about carbon loss with pellets, they compensated for this by making carbon additions exceeding the theoretical requirements for the pellet recycling trials. Coke consumption during continuous operation was about 33 percent (260 lbs) higher for heats with pellet charging although the difference does not meet the test of

~ statistical significance (at 5 percent level). additional carbon was recovered in the bath. Electrode consumption data were collected for the period when pellets were used in the Block I experi- ments and compared with the electrode consumption of the previous month. Electrode consumption increased slightly during pellet recycling (by about 0.26 lb/net ton of steel produced).

A large portion of this

93

i

. . . :

In the Green River operations, pea size lime is blown into the furnace pneumatically near the top of the furnace, above the bath line. An average of 1.2 lb of extra lime was consumed per ton of steel produced. Comparison of ferroalloy consumption data did not indicate any significant differences between heats with and without pellets after adjusting for differences in aim specifications. Observations did not reveal any noticeable change in dust or slag generation rates due to the addition of pellets.

Economics

Table V presents the capital and annualized costs developed for three hypothetical EAF units of different sizes. small, medium, and large plants were postulated to be 1,000, 3,000, and 9,000 tons per year, respectively. The economic analyses presented herein are preliminary in nature, and must be refined for each specific plant site. analysis indicates recycling to be attractive, especially for larger plants.

Conclusions

Dust generation rates for the

When compared to landfilling costs of $100/ton dust, the economic

This research has proved the concept of pelletizing EAF dust with water only and recycling it to the furnace. tives underlying PEI's philosophy for EAF dust management have largely been met:

The following original objec-

' Use of a simple process design

Use of existing, proven, low-cost equipment

Use of EAF dust only, with no additives other than water * Use of a process with no residual pollution streams

Use of a process with a high turndown ratio and capable of processing widely varying dust compositions

References

1. Center for Metals Production, Pittsburgh, Pennsylvania. Electric Arc Furnace Dust - Disposal, Recycle and Recovery. Report No. 85-2. Project No. RP-2570-1-2. (1985).

2. J. D. Lynn. Electric Furnace Fume Greenball Recycling--A Technical and Economic Evaluation. Pollution Abatement Technology for 1983, Chicago, Illinois. (October 18-20, 1983).

Presented at the Symposium on Iron and Steel

94

Table I. Elemental concentrations in electric arc furnace dusts (wt. percent)

Steel plant ~

Element A B C D

Caa Fea

b Zn

Pbb

9 . 1 7 3 . 1 8 2 .81 14.71

16 .7 3 2 . 1 32 .9 28 .2

2 6 . 9 20 .9 20 .6 12.9

2 .24 3 .85 5 . 0 8 0 . 8 3

a Analyzed by ASTM Reference Method 3682.

Analyzed by ASTM Reference Method 3683.

95

Table 11. Comparison of data for heats with pellet charging in Blocks I and 11.

TOTAL OBSKRVATIONS: 14

TAPTAP -ED TOTYIELD NETYIELD OPBNCRBN APAURBN RRS ?MI/" x x x x

N OF CASES 14 14 14 14 14 14 MINIMU4 4.830 536.950 85.550 79.500 0.070 0.150 MAXIWIM 9.670 670.740 93.800 93.600 0.471 0.480 MBAN 7.319 616.919 89.909 86.693 0.270 0.352 STANDARD DBV 1.217 40.442 2.752 3.734 0.120 0.113

TOTAL OBSERVATIONS: 24

CoRBClwd LBS

14 0.

3200.000 1132.143 799.425

TAPTAP PURUSW NETYIBLD OF'KNCRBN FNAURBN CoHlcIlRO RRS "N x x x x LBS

N OF CASES 24 24 24 24 24 24 24 MINIMU4 4.500 493.830 82.500 65.500 0.050 0.160 1oo.oO0 MAXIMlEl 8.000 672.500 98.900 93.900 0.860 0.470 2000.000 MEAN 5.872 560.903 91.231 85.788 0.349 0.338 1050.000 STANDARD DBV 0.953 37.725 3.656 5.494 0.203 0.112 521.703

Variable names are as follows:

TAPTAP = Tap-to-tap heat time (in hours). PWRUSEO = Power used (in kilowatt-hours/ton), TOTYIELD = Yield of steel as tappped as percentage of total metallic charge

NETYIELD = Yield of prime steel as poured (in percentage). GPENCRBN = Opening carbon in bath (in percent). FNALCRBN = Ftnal tap carbon in bath (in percent). CGKECHRG = Coke charged (in pounds). PLLTCHRG * Pellets charaed (in oounds).

including scrap, ferroalloys, and pellets.

PLLTcBllG LBS

14 91o.oO0

2180.000 1129.286 322 .OW

PLLTCmG LBS

24 100.000

2155.000 1083.542 328.019

Table 111. Comparisons of dqta for baseline heats.

N OF CASES MINIMUM CUXIERM MEAN STANDARD DBV

WTAL OBSRRVATIONS: 35

TAPTAI' PWRUSBD T0TYIBI.D NEMIBLD OPBNCRBN FNALCRBN HRS IwH/" x x x %

28 28 28 28 27 27

9.083 766.090 97.100 96.600 1.170 1.070 6.777 592.898 90.118 87.172 0.348 0.381 1.211 74.942 3.438 4.799 0.295 0.271

4.670 483.220 03.400 73.000 0.060 0.110

TAPTAP PWRVSBD BRS IR1L?/m

W OF CbSBS 36 35 MINIMIM 4.083 466.440 I(AxIMIw 9.250 634.440 HEAN 6.093 540.882 STANDARD DBV 1.204 37.959

ToTllIBLD x

35 85.000 96.600 91.180 2.712

NBTYIBLD %

35 71.000 95.600 87.397 4.290

Var iab le names are as fol lows:

TAPTAP = Tap-to-tap heat t ime ( i n hours). PWRUSEO - Power used ( i n k i lowat t -hours l ton) , TDTYIELD - Y ie ld of s tee l as tappped a5 percentage of total m e t a l l i c charge

inc lud ing scrap, fer roa l loys. and pe l l e t s . NETYIELD = Y ie ld o f prime s tee l as poured ( i n percentage). OPENCRBN = Opening carbon i n bath ( i n percent). FtlALCRBN = Flr.al tap carbon In bath ( i n percent). COKECHRG - Coke charged ( i n pounds). PLLTCHRG - P e l l e t s charged ( i n pounds).

OPBWCRBN x

35 0.090 1.040 0.336 0.257

FNNCRBN %

35 0.100 1.060 0.350 0.203

coI(BcHRG LBS

28 0.

2500.000 1078.57 1 654.532

co1[BcIIRD LBS

35 0.

2300. 000 788.571 546.447

PLLTCmG LBS

28 0 . 0 . 0 . 0 .

PLLTCERG LES

I 0 . 0. 0 . 0 .

I.

Table IV. balysis of EAF dust generated in Block I and Block I1 experiments. (wt. percent)

Date (1985)

Constituents

Phosphorus

Sulfur

Cadmium

Calcium oxide

Chromium (+3)

Chromium (+6)

Copper

Iron

Lead

Magnesium

Manganese

Molybdenum

Nickel

Potassium

Silica (SiO,)

Sodium

Vanadium

Zinc

a Not detected.

6/12 - 712

Block I

Range of daily composite averages

NDa

0.44-1.00

0.01-0.02

13.4-17.5

0.23-1.80

NDa

0.07-0.10

16.7-21.1

0.47-0.70

1.76-2.02

2.60-3.30

0.05-0.80

0.12-0.17

0.70-1.00

1.50-3.20

0.50-0.60

CO.01

20.1-27.9

Arithmetic mean

ND'

0.57

0.02

15.3

0.53

N D ~

0.09

18.3

0.66

1.88

3.10

0.35

0.14

0.80

2.00

0.58

<0.01

24.8

8/26 I 8/27 I 8/28 I 8/29

Block I1

7

0.03

0.50

0.01

19.0

0.58

0.00

0.09

24.4

0.66

3.63

5.0

0.14

0.29

1.0

3.1

0.5

co.01

9.0

Daily composite averages

0.03

0.47

0.02

15.1

0.44

0.006

0.08

21.2

0.58

2.29

3.5

0.16

0.24

1.0

2.8

0.5

<0.01

12.4

- 0.04

0.63

0.02

18.1

0.57

0.003

0.13

26.7

0.65

2.64

4.5

0.37

0.27

1.2

3.0

0.7

co.01

12.3

0.02

0.48

0.02

16.1

0.62

0.002

0.10

28.5

0.65

2.77

4.4

0.33

0.27

1.1

3.1

0.7

c0.01

14.7

98

. . . - , . .. , . ,. . I C * , .

Table V . Capital and annualized costs f o r selected EAF recycling units.

I Costs per Operating unit, $

I i Other -

Fixed capital costs

Total capital

Credit for sale of high

Manufacturing costs

Raw material CLW labor (includes onsite supervision & clerical) (2060 hlylperson @ 16/h

I

burdened)

0 33,28G

6.000 75c

I o 33,280

0 66.560

38,OOC 2.25G

Uti 1 i t ies Electricity 2.000 Uater 1 25C

T o t a l Manufacturing costs

Fixed charges

06M supplies (10: of FCl) 7.000 Lab charges (Illton) I 1 . m - .___

43.500

Equipment payments (5 yr loan @ 14'1) includes working capital Local taxes 1 insurance (3% of FCl)

22,500

2.100

HOm€ Office Management

(2% of mfg. costs) and Administration

41,700

3.900

8,400

14.000

64,220

6.000

17.370

28,950

Freight

Total Annualized Cost

Unit Cost for Dust processed, Siton

Medium

130.000 200.000 19.500

10.000 30.000 90.000

124.000 232.500

41.33 25.83

85.500

85.50

13.000 20,000 3,000 9,000

I I I

99

BAGHOUSE

FLEXIBLE - TUBE

TEMPORARY SHELTER 1 ....................... I I

EX IS T I NG ROLLOFF

BOX

Figure 1. Dust pelletizing process f l o w schematic for Green River Steel installation.