EAF Dust Mar10

8/9/2019 EAF Dust Mar10

1/3

Steel Times International March 2010 43

Environment

Still no simple solution toprocessing EAF dust

IN THE USA, processing of electric arc fur-nace dust has been regulated since 1988. Although several facilities were in complianceat the beginning of this period, the regulations were structured to encourage investment innew metals recovery technologies and toavoid landfilling. In the past 22 years, virtually every new process has failed. Those that weresuccessful either used old technology or werebuilt by the original processor, Horsehead. This article will explain how such an unusually high failure rate occurred. The review is limit-ed to US facilities because of the unique busi-ness and regulatory conditions there, and thedust processing industry has had to succeedon the basis of economics. The primary remaining process is the Waelz kiln, with which the industry started.

Why an almost unbroken s tring of failures with the new processes? Why did the oldestablished process win out even given theregulatory preferences for the new processes?Reasons include finances, regulations, compe-tition, inadequate planning and technical fac-

tors. In the early years many new processes were preferred by regulators, but who almostalways demonstrated poor technical judgment.

Financial factors included difficulties infunding, a changing business environment andinadequate resources for proper developmentor to handle difficult startups. Regulatory issues included changes limiting entry andbyproduct regulations. Competition factorsincluded existing competitors, ignorance of steel company requirements and misjudgingmarket conditions. Project planning wasincomplete, overly optimistic and did notallow enough time for resolving operationalproblems. Each country had their own regula-tions, some focused on penalties and liabilitiesfor infractions (the US) and others were simi-larly restricted as to practices, but innovation was easier (as in Europe and Asia).

However, most failures were rooted in inad-equate attention to technical details and poor-ly conducted development, testing, design andstartup. A modicum of success during superfi-

For the past 22 years the steel industry inthe USA has attempted to process EAFdust in an environmentally friendly way,but generally without success. The ques-tion remains why this lack of success andwhy did older technology provide betterresults than new processes?By Larry M Southwick*

FASTMET plant at Kobe Steels Kakogawa Works to treat steelmaking dusts

*US environmental consultant, Cincinnati, Ohio

cial testing of the primary concept was auto-matically assumed to lead to financial success with commercial facilities. Science and eco-nomics were seemingly expected to meekly and dutifully conform to environmental goals.In short, the failures were mostly man-made.

Same problemsSuch development problems are not new. As was noted more than 100 years ago under the

heading Adoption of a New Process in the11 March, 1905 issue of the Daily Mining Record (Denver, Colorado):

When new metallurgical processes areintroduced on a commercial scale, it is impor-tant all the details be worked out by theinventor before making an expensive installa-tion; but this, unfortunately, is not alwaysdone. In such an event, the companys plantbecomes a school for experiment, not only forthe perfection of the chemistry of the process,but also for evolution of a satisfactory andeconomical means to apply it.

Often, the estimated expenseruns far in excess of the amountbelieved to be sufficient, and forthis reason ample funds shouldbe available, or all that that hasbeen done may be lost.

Sometimes it occurs that aftera plant has been installed itsdefects are realised, and it isseen that had a different coursebeen pursued, success mighthave resulted; but the funds areexhausted and the process iscondemned, when a more fortu-nate application of the ideasmight have brought success. In someinstances the chemistry of a new process hasbeen fully settled, but the details of themechanical portion have not been given suffi-cient consideration. The engineer engaged toreview the proposed installation fails to dis-

cover this important fact until too late.In short either:

the development programme was cut short; design and installation were faulty; startup and operations (and available funds)

could not overcome the first two; and examples of prior art were not properly

noted or evaluated.Usually, prior art proved to be negative

examples of the technology proof rather

than the process or equipment application would not work.Complicating the issues with EAF dust

were bureaucrats defining the science used todetermine hazardous waste definitions, char-acterisations, treatment and legal liability.Consequently, in-depth generation of testresults and their analysis, investigation of anomalies and operating limits or potentialbottlenecks, and hands-on familiarity withequipment were considered unnecessary.

Buying into the pseudo-science and bureau-cratic basis of many regulations could reflect a

mindset uncompatible with scien-tific methods. Also missing wereaspects such as thorough execu-tion, engineering and metallurgi-cal skills, acuity and integrity,actual equipment behaviour andfinancial reality. Unfortunately,university training also seemed toencourage these attitudes. Suchdeficiencies can be remedied by developers and designers havingactual operating experience.Sadly, many of the attempts withdust treating technologiesinvolved companies with little

commercial experience in processing plants. The following review covers commercial-

sized examples of dust treatment projects. A few successes are noted, but most havenot been commercially viable.

The primary technology areas explored here

AbbreviationsRHF Rotary Hearth FurnaceEPA Environmental Protection AgencyHTMR High Temperature Metals RecoverySAF Submerged Arc FurnaceCZO - Crude Zinc-OxideDRI Direct Reduced Iron

Complicating the issues with EAF dust were bureaucrats defining the science used to determine hazardous waste definitions, character- isations, treatment and legal liability.

43-45_STI_Mar10_R-Larry 7/4/10 09:59 Page 43


2/3

44 March 2010 Steel Times International

Environment

Terra Gaia and Cashman/American MetalsRecovery, have been unable to attract fullscale financing. The problems are the cost topressurise a relatively low value dust up tooperating conditions, the purity of precipitat-ed oxide or the complications in electrowin-ning zinc metal. Insufficient financing has notallowed adequate development of these steps,though it has also prevented premature con-struction of speculative facilities.

Pyrometallurgy The primary successful process is the Waelz

kiln, commercialised 80 yearsago to process zinc oxide ores.In the USA HorseheadIndustries (three plants and oneunder construction), ZincNacional (ZN, in Mexico) andSteel Dust Recycling (now owned by ZN) use this technol-ogy. The initial EnvironmentalProtection Agency (EPA) regula-tions emphasised high tempera-ture metals recovery (HTMR)technologies, but asked for new processes to be developed. Those that were, and subse-

quent efforts, have led to failures, except inone case. Non-HTMR processes that the EPA directives flushed out have already been dis-cussed.

The single continuing success with new processes was the flame reactor developed by Horsehead and installed at North Star Steel inBeaumont, TX. It fumes heavy metals, burningthem to the oxides and produces an irony slag. While its economics may not be as compellingas Horseheads Waelz kiln operations, it fitstheir markets and helps compete with ZN.

Another of the new plants, by ZiaIndustries (Caldwell, TX), uses a two-diame-ter rotary kiln. Operation was not satisfactory until petroleum coke was tried as a reductant.

But by then a competitor had won away theirclients. Philip Environmental also built arotary kiln plant in Hamilton, Ontario,Canada which operated satisfactorily but wastoo small to be commercial. A tandem two-kiln plant was planned for Cardiff, Wales, UK but this was not a workable concept (oxidisingkiln followed by reduction) and after a techni-cal review was not built.

Three plants have used a plasma furnace -two with Tetronics (Nucor Blytheville, AR,and Florida Steel - Jackson, TN), and the otherat Plasmet, Houston. These suffered becausethey fumed zinc directly into splash condensersfor metal recovery. Such operations wereknown to be problematic because of chloridesand feed dusts arriving with the zinc fumes,fogs and mists. Also, splash condensers arenotoriously inefficient, usually requiring two ormore in series. These negative examples wereinadequately reviewed by design engineers andmetallurgists. These furnaces also had seriousslag bath containment problems and the DCplasmas were difficult to control.

Two proposed plants, Oxide Recycle atLone Star Steel, Texas and Philip atHamilton, ON with the Zinc Iron PlasmaProcess, were to use the plasma cupola manu-factured by Westinghouse. In both cases any zinc fumed would have been condensed andcollected within the cupola and recycled backdown the shaft. The result was the formationof scaffolds of agglomerated feed, con-densed zinc, sticky zinc oxide and a unit

plugged, much like an iron blast furnace witha zinc feed. These projects were halted priorto construction.

There are ways to deal with the zinc conden-sation problem. These include feeding low-zincdust (less than 2%), using large feed briquettes(as the Imperial Smelting process, and evenlarger ones of EAF dust in the OxiCup processin Germany), high top temperatures to keepthe fumes above zinc condensation tempera-ture (again, used by ISP), and providing gas withdrawal ducts to remove zinc fumes prior tocondensation (as was tried in zinc industry retorts 100 years ago). Without these modifica-tions, cupola or blast furnaces will not operate with EAF dust. Applying these changes unfor-tunately results in expensive operations. A new blast furnace outside the US is in the processof re-proving these axioms.

An Elkem submerged arc furnace (SAF) was built at Laclede Steel in Alton, IL. Theplant had several problems (splash condenser,furnace cooling panel failures and burnout)and was taken over by Laclede and modified.Eliminating the splash condenser resulted information of high-quality crude zinc-oxide(CZO), but eventually there was a furnacefailure and fire, terminating operation.

The pig iron zinc oxide process, fumingCZO and making iron in a circular inductionfurnace, is trying to establish operability attheir plant at Nucor in Blytheville, AK. Littleinformation has come out of this operation,now almost two years old, so its success orlack of is unknown though the dearth of information is a bad sign. Pilot plant opera-tions noted refractory problems and, basedon the other examples cited herein, mustsurely be the case with the full scale plant.Induction furnaces are successfully applied tomelting metal (as in casting or galvanizingoperations), but these are relatively quietbaths compared to the cyclone of splashingand bubbling with dust fuming.

Dust process A rotary hearth plant built by AllMet atNucor, Blytheville, AR, made a CZO and ahighly metallised (~90%) direct reduced iron. The collected oxide was to be re-fumed in anelectric furnace and zinc metal collected in asplash condenser. Feed was to be EAF dustand other mill iron wastes. Unfortunately, thefeed and rotary hearth operation was trouble-prone, DRI was of poor quality and not highly metallised, collecting fumed CZO incurredoperating problems, and the splash condenser was never operated. Poor development anddesign, inadequate startup staff and anunworkable concept were to blame.

Stainless no comparison This plant is just one example from many which have based their plans on the financially successful operation of the Inmetco stainlesssteel dust plant at Elmwood City, PA. Thatplant sends dust pelletised with reductant anda binder to a RHF, CZO is fumed off and sentto Horsehead and DRI goes to a SAF. In theSAF ferro-nickel and ferro-chrome are pro-duced. Unfortunately, other developers missedthe financial factors that Inmetco is able tocharge five times or more for processing andits ferroalloy products bring five times whatiron would bring and the iron is also beingsold. Inmetcos SAF operates at 1450C, 150-200C cooler with ferroalloys than when mak-ing iron. So Inmetco still makes money,

are glassification, hydrometallurgy and pyromet-allurgy. The full presentation at the August 2009MetSoc meeting in Sudbury, Ontario, Canadaprovides extensive details on these plants andothers around the world. The criteria for suc-cess here emphasises commercial viability, which helps focus and complete the analysis.

Glassification Two plants glassifying dust were built in theearly 1990s: Inorganic Recycling (Nucor,Hickman, AK) and Glassification International(Oregon Steel, Portland, OR). In general they produced an abrasive frit. Theformer used a typical glass fur-nace, which did not provide themixing required to complete thechemistry and thoroughly fix leadin the feed materials (primarily dust and sand). The latter usedinduction heated crucibles, whichalso did not provide good mixing.Both suffered from the sensitivechemistry of glassification and variable dust compositions. Thesecond also contaminated a fritusers site and personnel.

A recent example isInternational Melting and Manufacturing(La Porte, IN). This started up in 2008, usinga new furnace design with improved mixing.However, there has been little informationpublished on this plant so its success remainsunknown. It would still have to deal with vari-able dust chemistry, as well as plant designand startup issues. Another example, RichlandMolded Brick (Mansfield, OH), operated suc-cessfully making architectural bricks from thedust. But its normal brick was highly prof-itable, so they stopped processing dust.

Hydrometallurgy Much of the zinc in EAF dust is tied up asferrites, Fe2O 3. ZnO. Ferrites do not break

down with normal solvents, so an initialreduction roast is common practice.Unfortunately, most hydromet technology developers did not do sufficient homework tolearn this. Metals Recycling, using ammoniumchloride as a solvent and precipitating highpurity zinc oxide fell into this trap, their firstplant at Nucor - Darlington, SC failing. Theirsecond plant at Florida Steel, Jackson, TNadded a rotary hearth furnace (RHF) to deal with ferrites, but then the back end zinc oxidepurification and precipitation step did notperform as required. Its shortcomings hadbeen hidden by the lack of a roaster.

Ezinex, also using ammonium chloride, usedelectrowinning to recover zinc metal. Theirplant in Italy benefited from a dust apparently low in ferrites (because it was collected in a wet scrubber, not a hot system), so recovery rates were good. Initial plans to build in theUS included furnace pretreatment, but financ-ing failed. Initial plans in Spain by Zincex for asulphuric acid leach followed by solventextraction did not include a furnace, nor didplans using the same technology begun by ZincOx for Ohio in the US. Current ZincOxplans for the US plant, now with an RHF,have collapsed since delays caused their part-ner to sell the dust supply contracts to a majorcompetitor. They have however purchased theidled RHF at Rouge Steel in Michigan, which will provide a platform for future plans.

A couple of pressurised hydrochloric acidprocesses (which will break down the ferrites),

Buying into the pseudo-science and bureaucratic basis of many regulations could reflect a mindset incompatible with scientific methods.



3/3

Steel Times International March 2010 45

Environment

tional operating data on the ITmk3 Processunder US conditions. ITmk3 does not processEAF dust, using iron ore instead, but that is

not critical. What is critical is theunits economics, energy require-ments, product quality, yields andraw material consumption, as well as the mechanical design andoperating features under USoperating philosophies.

Homework Established technology (Waelzkilns) succeeds as do some estab-lished zinc companies. None of

the new process developers and the EPA didtheir homework, they did not have sufficientindustry experience to determinethe viability of their technology,they did not allow a long andextensive enough developmentprogramme to resolve operationaland design issues and that thenew plants did not make sufficientallowance for startup problemsand delays.

Why would such amateurs: plan to enter a complex and

demanding market they are notfamiliar with;

propose a technology about which they do not complete ly understand;

build and begin operation without adequatepreparation via testing and without suffi-cient staffing to deal with startup problems;

have such complete disasters; and

be so completely clueless about their failuresthat they still try to build other similar plants,other developers choose the same processingconcepts, and financial firms continue to sup-port such questionable concepts?One possible reason is that regulations for

landfilling encouraged the steel industry tosupport new technology, even those with nohope of success. Following the initial string of failures in the mid 1990s, the EPA allowedrelief from the processing monopoly, whichtheir regulations and technology designs hadinduced.

Yet the root problem appears to be a devel-opers conviction that the environmental ben-efit from their technology trumps metallurgi-cal science and good engineering. Even

stranger, a study published by the Swedish research organisa-tion Mefos in 2000 determinedthat recycling of EAF dust wasless environmentally friendly than landfilling.

In the US, for a period upuntil the last three years, land-filling held its own on an eco-nomic basis as well as provideda simple option to monopoly.Now, economics has begun tofavour processing, which hasbecome virtually the sole dispos-al method of choice. It meansEPAs 22-year-goal of ending

landfilling of EAF dust is nearing realisationbut it is not through their own contributionor on the part of their science and technology expertise.n

though its DRI quality is poor, resulting inconsiderable carbon boil. The SAF, while highmaintenance, is not as bad as when makingiron from poor DRI since it oper-ates at lower temperatures.Carbon steel dust cannot pro-duce the same financials, whethermaking iron or not.Interestingly, RHF technology has found wide application in Japan (and Korea plans to buildunits) to manufacture DRI fromiron and steel waste oxides.Developers include Demag(RedSmelt), Maumee and Kobe(FASTMET), as well as improvements basedon units operated by Nippon Steel. Feedsinclude up to 10-20% EAF dust.

However, economics, regulations, operatingphilosophies and disposal options differ in Japan from the US, so comparisons aredifficult.

Five FASTMET plants are currently operat-ing in Japan. The first to be built was at KobeSteel Kakogawa works and can treat 14kt of dusts a year and the remaining four each havecapacities of 190kt/y. Kobe Steel's develop-ment work on Fastmet led to the ITmk3Process. The world's first commercial ITmk3plant which began production in January inHoyt Lakes, Minnesota in a joint venturebetween Steel Dynamics and Kobe Steel. While ITmk3 is a higher temperature processthan Fastmet, forming iron nuggets ratherthan DRI, it uses most of the same mechani-cal design and operational features.

The commercial plant should provide addi-

Poor development and design,

inadequate startup staff and an unworkable concept were to blame.

A study published by the Swedish research organisa- tion Mefos in 2000 determined

that recycling of EAF dust was less environmentally friendly than landfilling.


Documents

EAF Dust Mar10