chapter 6 conservation of strategic materials

CHAPTER 6

Conservation of Strategic Materials

ContentsPage

Introduction *e. **s* .*. o**. e#*s***, .+** *soO*** *0** **** *,o. e*. .0. *@*o 215Recycling * s * * * * * * .*e***s, * , * * * * * * * * * * * . * * * * * . * * e * * * . . * * * * . * . . * # . * 215Advanced Processing and Fabrication Technologies . ..................216Conservation Through Part Life Extension and Design. . ...............216Interrelationships of Conservation Approaches . . . e . . . . . . . . . . . . . . .. ...217

Case Studies of Strategic Materials Conservation Opportunities. . .........218Superalloy . . * * *,, * . . ,**, * * , . . * * . . * . e . . o.o*. * e . ,., . , o * , * . , * * * , * . , 219

Conservation of Manganese in the Steel Industry . . . . . . . . . . . . . . . . . . . . .227Automotive Catalysts 6., *.** .**, ,.*o, o*oe. .,o, .,. .,o*, **o*, *o*,,,,, 238Other Conservation Opportunities **** 9**m. * * * * * * * * * *0** . * * * * * . * * . * * * 245

Prospects for Conservation of Strategic Materials . . . . . . . . . . . . . . . . . . . . . . . 254Recycling Technology ● , * . , * . * * * * * * * , . * * * * * * . . * . * * . . * * * . * . * . , * , * * * * 257Nontechnical Impediments , * * * * . . * . * * * . * * * . . * * * * * . . * . . * * * . * , * * . * , * . 257Recycling and Conservation Data Base ● ***************.**.,.,,***,** 258

List of TablesTable No. Page6-1.6-2.6-3.

6-4.

6-5.6-6.

Cobalt Consumption in Superalloy, 1980 . . . . . . . . . . . . . . . . . . . . . . . . . 219Manganese Content for Major Grade of Steel.. . . . . . . . . . . . . . . . . . . . . 232Parameters Affecting Manganese Retention in theSteelmaking Process ● **0* *e** **0* **** +8 *****************e****** 238Current and projected Manganese Consumption inU.S. Steel Production ● . ,** * ,**, . .** * . . @, ********m.*. ,***,***,** 238Cobalt Consumption and Recycling in Catalysts, 1982 . . . . . . . . . . . . . . 246Selected opportunities for Increased Recycling of Chromium, Cobalt,and Platinum Group Metals ● * * . . . ., . * ., * * * * *, * * * * * * * * * * ******.*. 255

List of FiguresFigure No. Page6-i.6-2.6-3.

6-4.6-5.

Service Life Time of Turbine Blades . *, * * *, . * * * ., . . . . * *, ****,*.*, 226Historical Market Shares for Major Classes of Steel . . . . . . . . . . . . . . . . 231Flow Diagram Showing the Principal Process Steps Involved inConverting Raw Materials Into the Major Steel Product Forms,Excluding Coated Products . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234Current Pattern of Manganese Use in Steelmaking . . . . . . . . . . . . . . . . . 235PGM Demand by Vehicle Category ● .* . **e w*****. .*** .************ 239

CHAPTER 6

Conservation of Strategic Materials

Introduction

Past experience has shown conservation tobe an important response to shortages of ma-terials, and strategic materials are no excep-tion. In all of the post-World War II supply dis-turbances discussed in this report, industryresponded with increased recycling and, insome instances, adopted new practices whichused materials more efficiently, Conservationalso can be an ongoing strategy to reduce in-dustry dependency on insecure sources of sup-ply, In this chapter, several different conser-vation techniques that can conserve strategicmaterials are discussed, Included are recycling,product life extension, trimming of alloy ad-ditions to the low side of acceptable ranges,and improved processing techniques, Substi-tute materials and use of surface modificationtechnologies can also conserve strategicmaterials; these are addressed in chapter 7.

Recycling

Recycling differs from other conservationtechniques in that it directly adds to materialsupplies. Sources of recycled metals includescrap material generated by the primary pro-ducer, which is called “home” or “revert”scrap, and secondary (or purchased) scrap, Sec-ondary scrap generated by manufacturers andfabricators is called “prompt industrial scrap,”while scrap that comes from products that havebeen used and discarded is called “obsoletescrap. ” Another secondary source is processwaste that can be reclaimed from metal proc-essing plants.

Purchased scrap (comprised mostly of promptand obsolete scrap) accounted for 8 percent ofdomestic cobalt consumption, 15 percent ofplatinum consumption, and 12 percent of chro-mium demand in 1982.1 Actual levels of recy-

1 U, S. Department of Interior, Bureau of Mines, Mineral Com -modit~ Summaries 1983 (Washington, DC: U.S. Go\’ernmentPrinting Office, 1982), pp. 32, 36, and 116.

cling are significantly higher owing to internalrecycling by industry, toll refining arrange-ments, and incomplete reporting to the Bureauof Mines. In addition, a substantial amount ofmanganese is recycled from ferrous scrap, andthis may grow in the future.

Home scrap, which includes trimmings andends from castings and overflows from ingots,is generally recycled with great efficiency.Since it is clean and its composition is known,it can simply be returned to the melt for reproc-essing.

Recycling of prompt scrap is less uniformthan home scrap. In large manufacturing oper-ations, where substantial amounts of scrap aregenerated, efforts to separate scrap by type, tokeep it free from contaminants, and to main-tain collection and marketing operations areoften justified by the market value of the scrap.In smaller operations, or in processes whereoils or cutting fluids contaminate the scrap,materials may be discarded as waste. Evenwhen recycling offers economic advantages,firms may not want to divert their resourcesto the establishment of a new set of internalprocedures for segregation and collection ofscrap.

Collection of obsolete scrap presents evengreater problems. This scrap may be dissemi-nated across wide areas and mixed with a va-riety of other materials, Impurities from metal-lurgical contaminants must be eliminated tomake the scrap reusable. Owing to the expenseof collecting, identifying, and separating thisscrap, a substantial amount of metal is lost.Even when quantities of known materials arecollected, limitations set by the processingtechnology may result in the scrap being down-graded to a less demanding application. For ex-ample, superalloy scrap may be downgradedto make stainless steel in which the cobalt con-tent serves no purpose, and is essentially lost.

215

216 ● Strategic Materials: Technologies to Reduce U.S. Import Vulnerability

Significant quantities of strategic materialsare lost in process wastes, because the costsof treatment or handling for recovery exceedsthe value of the contained metals, Technologi-cal development, combined with imposition ofhigh costs on the disposal of wastes contain-ing hazardous materials, have given impetusto research on processing of such waste to easedisposal problems and, where possible, to ob-tain saleable products from the processing.

Advanced Processing and FabricationTechnologies

Industry constantly searches for new proc-essing and manufacturing technologies thatcan increase production, decrease costs, saveenergy and materials, or improve performance.Adoption of new technologies sometimes con-serves strategic materials. The rapid phase-inof the argon-oxygen-decarburization (AOD)process for making stainless steel during the1970s is a dramatic case in point. The AODprocess led to major improvements in chro-mium utilization in making stainless steel, in-cluding the ability to use more high-carbon fer-rochromium (which is cheaper, and requiresless energy to make than low-carbon ferro-chrome) as a raw material, and enhanced chro-mium recovery from scrap ferrochromiumcharges. Nearly all stainless steel producersnow use the AOD process, so that further im-provements in chromium utilization arisingfrom the process are likely to be modest.

Although it is rare for a new technology tobe adopted as pervasively and quickly as theAOD process, ongoing industrial trends in met-als processing and fabrication may lead to ap-preciable savings in some strategic materialson a per-part or per-unit basis in the years tocome. For the most part, industry investmentin advanced processing and fabrication tech-nologies is driven by concerns about productperformance, reducing costs, and minimizingreject parts. Material conservation is not oftena primary reason why industry adopts ad-vanced processing technologies, but it is oftenan important side benefit.

The need for U.S. steelmaker to remaincompetitive with foreign producers, for exam-ple, is leading to the replacement of older, lessefficient facilities with new processes whichare more cost effective and produce higherquality steel.2 Furthermore, major savings inmanganese can be achieved as steelmaker up-grade basic steelmaking processes. As newertechnologies and processes are introduced, im-portant reductions in the amount of manganeseneeded to process each ton of steel can be ex-pected. The pace of this phase-in will dependon industry resources committed to upgradingof domestic steelmaking capabilities.

Materials conservation is also a side benefitof advanced processing technologies that areincreasingly used in producing and fabricat-ing aerospace parts made from superalloys.Often referred to as “near-net-shape” proc-esses, these casting and powder metallurgytechniques are used primarily to reduce costsof machining and to improve control over man-ufacturing reliability. Because less metal is re-moved during processing, product yields areimproved significantly.

Conservation Through Part Life Extensionand Design

Life extension approaches to material con-servation are of primary relevance to very high-value parts. The overall value of a turbine bladeinstalled in a jet engine, for example, may be500 times the raw material value. Technologiesand maintenance strategies that extend the us-able life of a part, therefore, can result in sub-stantial savings in replacement part costs andmaterials over time.

Some life extension strategies are integrallyrelated to advances in nondestructive evalua-tion or testing (NDE or NDT) of products. Onthe inspection line of manufacturing facilities,NDE can be used to identify flawed products

‘The need for technological adjustments by the U.S. steel in-dustry is discussed in the OTA assessment, Technology and SteelIndustry Competitiveness, OTA-M-122, June 1980.

Ch. 6—Conservation of Strategic Materials ● 217

which can be immediately set aside for rework-ing or recycling.

NDE has also been used on the productionline to monitor machine tool conditions in or-der to avoid breakage or other damage. For ex-ample, tungsten carbide drill bits have oftenbeen retired prematurely because of concernabout breakage. A process developed by theNational Bureau of Standards for continuousmonitoring of the drill bit allows selectiveretirement of a bit when failure is detected tobe imminent, rather than relying on a statisti-cal average or waiting until the bit actuallybreaks, In addition, since cobalt is used to bindthe tungsten carbide, strategic material usemay be reduced.

NDE’s potential role in monitoring of in-service parts is expected to grow in the yearsto come as confidence in the reliability of in-spection devices increases. Formal retirementfor cause maintenance programs are beingevaluated by the Department of Defense andmay be implemented in the late 1980s.

It is also possible to conserve strategic ma-terials in the design of products, This some-times happens inadvertently. For example, U.S.automakers, in their effort to downsize cars,have substantially reduced chromium use incars even though strategic materials conserva-tion was not a primary objective. Design strat-egies specifically intended to minimize strate-gic materials use (e.g., industry’s effort toconserve cobalt in the late 1970s) are probablyrare. However, in many noncritical applica-tions, designers have a wide range of materialsthat could be used. Actual choice of a mate-rial may be governed not only by cost and per-formance factors, but also by the designers’ ex-perience and esthetic considerations. In theory,as computer-aided design technologies andcomputer-aided selection of materials advance,

designers may become more aware of alterna-tive materials that use less strategic materials.

Design of products so that they can be eas-ily recycled is another frequently proposedconservation technique. However, the trend insome industries is toward use of more complexmaterials and products that pose greaterdifficulties for recycling.

Interrelationships of Conservation Approaches

As discussed in subsequent sections of thischapter, the prospects for conserving strategicmaterials in many applications are quite good.However, the extent to which any given ap-proach will be adopted by industry will dependon how well it fits into the overall corporatestrategies of many different firms. Generallyspeaking, strategic materials conservation perse is seldom a primary motive in industrydecisions—although it is sometimes an ancil-lary benefit.

In the future, industry concerns about com-petitiveness, improving product performance,and cutting production costs will lead to adop-tion of new technologies, some of which maylead to conservation of strategic materials. Itis difficult, however, to predict overall impactsof conservation on U.S. import reliance: thesenew technologies will interact in complexways—sometimes simultaneously in a given ap-plication, For example, increased use of ma-terials-efficient fabrication processes in theaerospace industry will reduce the amount ofhome and prompt superalloy scrap that wouldotherwise be available for recycling, The over-all effect of this on strategic materials use, how-ever, is likely to be positive, since some fabrica-tion scrap is difficult to recycle. Life extensiontechniques would reduce demand for replace-ment parts containing strategic materials, butwould also reduce the quantity of obsoletescrap available for recycling.

38-844 0 - 85 - 8 : QL 3


Case Studies of Strategic Materials Conservation Opportunities

Conservation of some strategic materials in-creased during the late 1970s and early 1980s,encouraged in part by higher costs for rawmaterials and energy, Government waste dis-posal requirements, and industry investmentin new materials-efficient technology. How-ever, major opportunities to conserve strategicmaterials are still evident. These vary by ma-terial and application:

1. Despite advances in manufacturing tech-nologies, total cobalt losses in 1980 in unre-covered or downgraded scrap, or in process-ing wastes, amounted to 7.7 million pounds—nearly half of the estimated 16.5 million poundsof cobalt actually consumed domestically thatyears By far, the largest opportunity for in-creased cobalt recycling lies in the increasedrecovery of obsolete scrap. Of the 6.3 millionpounds of cobalt in products estimated to havebecome obsolete in 1980, nearly three-fourths(about 4.6 million pounds) was unrecovered ordowngraded. Potential sources for improvedrecovery include superalloy from obsolete jetengine parts and cobalt-bearing catalysts. Othercobalt recycling opportunities include reduc-tion of downgrading of obsolete scrap and im-proved recovery of processing waste, which to-gether entail cobalt losses of 3.1 million pounds.

2. The steel industry is the country’s domi-nant user of manganese, consuming nearly 90percent of total U.S. requirements. Manganeseconsumption is dependent not only on theamount of steel produced, but also on how effi-ciently manganese is used in steel production.Significant future reductions in the amount ofmanganese needed per ton of steel are ex-pected: greater attention to trimming is ex-pected to reduce the average amount of man-ganese in steel by 12 percent, from 13.8 to 12.2pounds per ton by 2000. This, together withadoption of improved steelmaking technolo-gies, such as external desulfurization, contin-

uous casting, and ladle and secondary refin-ing techniques are likely to decrease theamount of manganese required to produce aton of steel from 35.6 to 24.8 pounds by the year2000. The consumption of imported manga-nese ore and ferromanganese is estimated todecline from 17.8 pound per ton of steel to 9.5pound per ton, a reduction of over 45 percent.Most of these technologies will be implementedfor product quality and cost competitivenessreasons.

3. Widespread adoption of the AOD processby stainless steelmaker led to industrywideimprovements in chromium use that are notlikely to be repeated soon. However, chromiumconservation opportunities, including increasedrecycling, and reduced overspecification ofmaterials, could lead to major savings. In the1977-81 period, an average of 64,000 tons ofchromium contained in purchased stainlesssteel scrap was recycled each year.4 By con-trast, an estimated 62,000 tons of chromiumcontained in obsolete stainless steel scrap wentunrecovered or was downgraded in 1977 (thelast year for which detailed estimates have beenmade). The largest single opportunity for en-hanced recovery of obsolete scrap is in auto-motive recycling, where most chromium-bear-ing scrap is downgraded. Chromium recoveryfrom industrial processing wastes is increas-ing, owing to adoption of collection proceduresand commercial reprocessing systems,

4. The high price of platinum group metals(PGMs) encourages highly efficient use of thesematerials. Recycling probably accounts forover 40 percent of U.S. PGM consumptionwhen internal (toll refining) recycling by indus-try is taken into account. The key opportunityfor increased recycling over the next 10 to 15years will be recovery of PGMs from automo-tive catalytic converters which could add upto 500,000 troy ounces annually to U.S. PGM

aNational Materials Advisory Board, Cobalt ConservatbnThrough Techno]ogkal Alternatives, publication NMAB-406(Washington, DC: National Academy Press, 1983), fig. 3 (p. 23)and table 11 (p. 24).

4U. S. Department of Commerce, Critical Materials Require-ments of the U.S. Steel Industry (Washington, DC: U.S. Govern-ment Printing Office, 1983), p. 62,

Ch. 6—Conservation of Strategic Materials . 219

supplies by 1995. Obsolete electronic scrap isalso a promising source for secondary recov-ery of PGMs, even though such scrap is dis-seminated widely throughout electronic equip-ment which, in turn, is spread all across thecountry, making collection and separation ofmetals expensive. The Department of Defense,which is a large generator of electronic scrap,and the Bureau of Mines, have cooperated indeveloping improved means for processing thisscrap.

In the pages that follow, prospects for stra-tegic materials conservation in certain specificapplications are discussed.

Superalloy

Superalloy constitute the largest single cat-egory of cobalt use. They are also the principalusers of high-purity chromium metal and ma-jor users of low-carbon ferrochrome. Typicalcobalt-base superalloy contain up to 65 per-cent cobalt- and nickel-base, alloys may con-tain up to 20 percent cobalt. Chromium con-tent ranges between 10 and 25 percent.

Since 1978, when disturbances in Zaire ledto widespread concern about cobalt supplies,manufacturers of aircraft jet engines and en-gine components and several Federal agencieshave placed increased emphasis on reducingthe need for cobalt in jet engines. One ap-proach, covered in detail in chapter 7, is thedevelopment and use of substitute alloys thatcontain less cobalt. Other approaches empha-size conservation, through recycling of super-alloy scrap, use of near-net-shape manufactur-ing processes to reduce fabrication scrap, andextension of the useful life of components. Asindustry adopts these techniques, the need forcobalt, and for other strategic metals, such aschromium, nickel, and tantalum, will be re-duced on a per part basis.

Substantial amounts of cobalt, chromium,and other strategic metals used in the produc-tion and fabrication of superalloys do not endup in the final part. Table 6-1, adapted froma 1983 National Materials Advisory Board(NMAB) report on cobalt conservation oppor-

Table 6-1 .–Cobalt Consumption in Superalloys, 1980

Wrought Cast TotalInput:Primary cobalt . . . . , 3.47 1.73 5.20Prompt scrap. . . . . . . 0.26 1.50 1.76Obsolete scrap ., . . . 0,87 0.32 1.19

Total production . . 4.60 3.55 8.15

output:Final parts ., . . . . . . . 2.34 1.24 3.58Total lost ., . . . . . . . . 2.00 0.82 2.82

(Waste) . . . . . . . . . . 0.21 0.28 0.49)(Downgraded) . . . . 1.79 0.54 2.33)

Prompt scrap. . . . . . . 0.26 1.50 1.76SOURCE Adapted from National Materials Advisory Board, Coba/t Conservation

Through Technological Alternatives, Nat!onal Academy of Sciences,1983, pp. 147, 148

tunities,5 estimates sources and end results ofthe cobalt used in the production of superalloyin 1980, the most recent year for which detailedinformation is available. Several importantpoints can be drawn from the table:

●

●

●

●

Of nearly 8.2 million pounds of cobalt usedin superalloys in 1980, only 3.6 millionpounds, less than 45 percent, were con-tained in final parts.Almost 55 percent of the consumption ofprimary metal was lost through down-grading or waste in the production of su-peralloy parts in 1980.Almost 44 percent of the initial amount ofwrought alloys were lost or downgraded;for cast alloys the fraction is lower, about23 percent.The use of obsolete superalloy scrap wasquite low, accounting for less than 15 per-cent of the total cobalt used in processingsuperalloys. NMAB also estimated thatonly half of the cobalt in superalloy scrapthat became obsolete in 1980 was recov-ered for recycling by the superalloy indus-try (1.19 million out of 2.4 million pounds),

hCobait Conservation Through Technological Alternut i~~es, op.cit. Some uncertainty surrounds the estimates sho~vn in table6-1, owing to the inadequacy of the available data on scrap useand new cobalt requirements of the superallo} industr}. As aresult, various assumptions had to be employed b}’ Nhl AB toapproximate conditions in 1980. While NMAB believed its co-balt material flow model to be “reasonable approximation of re-ality, ’ it noted that further work would be needed to confirmor reject its assumptions (pp. 25-26).

220 Ž Strategic Materials: Technologies to Reduce U.S. Import Vulnerability

These statistics suggest the tremendous po-tential of conservation techniques to reducerequirements for primary cobalt. In 1980, 2,8million pounds of cobalt were lost throughdowngrading and to waste in fabrication ofsuperalloy parts. In the future, much more ofthis material may be recoverable as superalloyrecycling technologies advance. Alternatively,the amount of material lost could also be re-duced through the increased use of near-net-shape manufacturing processes, particularlywith respect to the wrought alloys.

Obsolete scrap is a large potential source ofcobalt. Most of the 3.58 million pounds of co-balt used in finished superalloy parts in 1980will eventually be available for recycling, butmuch of this material may be downgraded un-less jet engine manufacturers are confident thatscrap can be safely recycled for jet engines, Theraw materials for superalloys, whether it is co-balt, nickel, chromium, or ferrochromium,must be of the highest purity in order to avoidfaults in the final part that could cause it to fail,Thus, manufacturers of aircraft engines oftenprohibit use of obsolete scrap unless certifiedprocesses for collecting and processing thescrap have been approved, Rates of obsoletescrap generation in the future may be affectedby techniques that extend the service life ofparts.

Near-Net-Shape Processes

The use of high-temperature alloys in a jetengine is very inefficient in terms of the amountof material consumed to produce a finishedpart. Ratios of 8 to 1 for input material to finalpart (also referred to as the “buy-to-fry” ratio)are common, and ratios of 20 to 1 are not un-heard of. This ratio can be significantly re-duced by near-net-shape processes.6 Three gen-eral types of near-net-shape processes are inuse at this time: precision casting, advancedforging, and powder processing.

8These processes are discussed in detail in New Metals Proc-essing Technologies, a report prepared for OTA by Charles RiverAssociates, Inc.

PRECISION CASTING

Casting, the solidification of molten metal inmolds, is a near-net-shape process that hasbeen used for jet engine components since thebirth of the industry. Casting is very efficientin its use of material. Although it is necessaryto use gates and risers to control the flow ofmetal into the mold, resulting in as little as 40percent of the metal actually ending in the fi-nal casting, large castings are very near to theirfinal shape (only 5 to 10 percent of the mate-rial needs to be removed during the final ma-chining and finishing operations). Further,most of the scrap generated in casting is “im-mediate revert’’—that is, it can be recycledwithout further processing. Since it is recycledin the same shop in which it is generated,segregation and identification are simple mat-ters. Therefore, most material poured from themelting furnace is eventually used.

Recent trends in casting technology have re-sulted in precision castings of up to 500 pounds(typically 10-pound units were produced in thepast). This results in two advantages. First, alarger portion of the engine can be producedby the material-efficient casting method, andsecond, the elimination of numerous steps ofcomponent assembly reduce the cost of assem-bling the engine.

New casting technologies have also resultedin improved material properties. Through theuse of directional solidification, components,principally blades and vanes, can be cast sothat grain boundaries are very long and arealigned with the axis of the component. It isalso possible to cast components as a singlecrystal, eliminating grain boundaries altogether,This allows the elimination of elements that areoften added to strengthen grain boundaries,such as zirconium and hafnium.

ADVANCED FORGING METHODS

Closed die forging is used extensively to fab-ricate gas turbine components ranging fromsmall turbine buckets to large fan blades andvery large disks. In this process, the startingstock is heated to various temperatures depend-ing on the alloy being forged and is placed be-


tween two dies that are forced together, enclos-ing the stock and deforming it to the shape ofthe die. Usually the starting stock is wrought—i.e., it consists of a bar or billet that has beenhot-worked from an ingot. The hot working en-sures that the original casting achieves full den-sity and chemical homogeneity.

The major disadvantage of this process, asconventionally practiced, is that a large amountof the material must be removed to reach fi-nal shape. It is common for the final part tocontain only one-tenth of the original materials.This large material loss due to forging maybeconsiderably reduced through the use of twoadvanced technologies.

Isothermal Forging.—Isothermal forging is dis-tinguished from conventional forging opera-tions in that the temperature of the forging dieis the same as the piece being worked, approx-imately 1,7500 to 1,9500 F. Through reductionof heat loss from stock to die, the plastic defor-mation of metal is significantly improved. Theoperation is conducted in an inert gas atmos-phere. Parts can be processed to close to thefinal specifications, although the shape mustbe configured to allow nondestructive inspec-tion, usually by acoustic methods (hence theterm “sonic shape” for the form as it leaves thedie). Further machining is required to reducethe shape to its final dimensions. Pratt & Whit-ney Aircraft has patented and licensed a varia-tion of the isothermal forging process whichit claims can reduce the input material by 50percent, thereby making a large reduction inthe “buy-to-fly” ratio. Isothermal forging is nowroutinely used for the forging of turbine disks..

Near-Isothermal Forging.—Also referred to as“hot-die forging, ” near-isothermal forgingdiffers from isothermal forging in that the dieis not heated to the temperature of the stock,but to 1500 to 3000 F below the stock temper-ature. The operation is conducted in air andcan be adapted to conventional presses, mak-ing it more widely available at a lower cost thanisothermal forging. Although the improve-ments are not as great as those for isothermalforging, this process does reduce the amount

of input material required by conventionalclosed die forging,

POWDER PROCESSING

A third process for the manufacture of super-alloy components is based on the processingof ultrafine powder into billets or parts. Thepowder may be of a single composition or amixture of materials. By the latter method, ox-ide dispersion-strengthened alloys, such asMA-754, are produced.

Powder metallurgy is attractive, not just be-cause it can produce new materials, but be-cause it can produce substantial reductions inthe amount of scrap generated during formingoperations. Powder “pre-forms” are producedby techniques similar to casting processes. Thepre-forms are then compressed at high tem-peratures and pressures to eliminate voids andinclusions. The components are then ready forinspection and final machining.

The combination of powder pre-forms andhot isothermal forging has allowed the devel-opment and use of some of the most advancednickel-based alloys in turbine disks, includingIN-100, MERL-76, and Rene-95, none of whichcould be reliably formed into components byconventional means.

The advantages of powder metallurgy are notlimited to conservation of strategic metals inthe more exotic applications such as jet enginecomponents. Gears and camshafts, parts whichnormally require considerable machining, canbe manufactured to near-net shape by powderpre-forms, thereby reducing scrap productionand machining, The need for resulfurized steelin these applications, normally used for its im-proved machining characteristics, can be elim-inated as well, reducing the requirements formanganese in this type of steel.

Recycling

Until cobalt supplies were seen as threatenedby the 1978 Katangese rebellion in Zaire, re-cycling in the superalloy industry was largelylimited to home scrap. Prompt industrial scrapand obsolete scrap were either downgraded for


stainless steel production or exported out of thecountry. As prices rose and supplies becameuncertain, the importance of recycling of super-alloy scrap became apparent. Alloy producers,parts fabricators, and turbine manufacturersbegan to segregate fabrication (prompt indus-trial] scrap by alloy type so that it could beremelted and reused. Some turbine manufac-turers modified earlier specifications to allowgreater use of prompt industrial scrap and evensome obsolete scrap, provided that the sourceand type of the scrap was known. The exactamount of the savings provided by recyclingduring the period of high cobalt prices ($25 perpound at their producer price high point in1979-80) is not known, but domestic consump-tion of primary cobalt in superalloy was prob-ably 10 to 25 percent less than might otherwisehave been the case.7

Identification and separation of scrap is timeconsuming and expensive. With 1983-84 cobaltprices ranging from $6 to $12 per pound, it isnot clear whether the gains made in the late1970s will continue. Some firms have contin-ued their recycling programs, particularly forprompt scrap, by encouraging suppliers andcustomers to separate scrap by alloy type. Spe-cialized scrap processing firms now have great-er experience with this kind of recycling, whichmay encourage jet engine manufacturers to up-date specifications to allow greater use of scrapin the production of superalloys. Once suchspecifications are established, recycling ofsuperalloy scrap can be increased quickly inresponse to increased metal prices or supplyuncertainties.

PROCESSING OF SUPERALLOY SCRAP

Commercially used technologies for recy-cling superalloy scrap are based on the sameprocesses used in the original production andrefining of the alloy, including vacuum induc-tion melting and vacuum arc remelting, butstrict attention is paid to segregation of scrapby type.

For production of superalloys, uncontami-nated home scrap of known composition canbe and is reverted directly to the charge. Ad-ditional processing sometimes involving sev-eral steps may be needed if prompt industrialscrap or obsolete scrap is to be recycled.

Vacuum and air furnace technologies neededto remelt high-quality superalloy scrap intousable master alloys are already in place in theUnited States. In-house scrap is now carefullyseparated and routinely added to the alloy meltby many producers of superalloys, and a spe-cialized industry has developed to producemaster alloys from fabrication scrap. Somefirms report that they now recover virtually allin-house scrap. Obsolete scrap is also recycledto some extent, although generally not for themost demanding application—rotating parts ofjet engines. As is discussed in box 6-A, instru-ments able to identify complex alloys and con-taminants are now being marketed and mayalleviate some technical problems associatedwith scrap sorting. According to recycling in-dustry sources, several million pounds of high-temperature alloy scrap have been shippedfrom scrap processors to melters in the last 10years.8

Air-melting processes, involving electric fur-nace melting of the scrap charge and subse-quent refining in an argon-oxygen-decarburi-zation vessel, can produce alloys from scrapthat are suitable for many noncritical superal-loy and super stainless steel applications, butare not appropriate in critical applications suchas gas turbines used in aviation. Nickel and co-balt recovery is high, but readily oxidizable al-loys are lost in the slag. Air melting of scrapcan also be used to produce master melt alloyssuitable for further processing through vacuummelting into superalloys.9

Vacuum melting processes are used to pro-duce superalloy suitable for the most demand-ing applications or superalloys containinghighly reactive elements. Only the highest qual-

7As discussed in Charles River Associates, Inc., E~fects of the1978 Katangese Rebellion on the World Cobah Market, contractreport prepared for OTA, December 1982, pp. 1-6.

‘Information provided by the National Association of Recy-cling Industries.

‘J. J. deBarbadillo, “Nickel-Base Superalloy: Physical Metal-lurgy of Recycling, ” Metallurgical Transactions A., vol. 14A,March 1983, p. 332.

Ch. 6—Conservation of Strategic Materials • 223

Box 8-A.-Scrap Sorting and Identification Instruments

Recycling of complex materials requires de-tailed knowledge of scrap constituents. Nick-el- or cobalt-based superalloy scrap may con-tain 8 or even 10 separate alloying elements,as well as potentially troublesome tramp ele-ments and contaminants that need to be re-moved during processing. Concerns aboutscrap quality are not limited to superalloys:in stainless steel production, for example,presence of phosphorus in scrap needs closemonitoring, due to its adverse effect on work-ing, forming, and ductility of stainless steel.Very small amounts of titanium in scrap chargesare believed to be harmful in tool steels andhigh strength steels. Failure to correctly iden-tify and sort scrap is a major reason for theextensive downgrading of superalloy andstainless steel scrap that is still prevalent.

Certification requirements placed on scrapprocessors by their customers have becomeincreasingly stringent. In some premium uses,scrap processors may have to identify scrapconstituents to an accuracy of a few parts permillion. Under these circumstances, conven-tional scrap sorting techniques—entailing vis-ual inspection, magnetic separation, or sparktesting—are not sufficient in themselves.

Growth in recycling of superalloy and spe-cialty materials during the last decade has ledto increased use of mobile and inexpensive($4,000 to $50,000) sorting instruments. Thesedevices were not developed primarily for usein scrap processing, but most can be used byunskilled operators (or in some cases trainedtechnicians) to identify complex scrap com-ponents more accurately.16 Some of these (e.g.,chemical spot testing) are used to confirm thepresence of alloys after preliminary hand sort-ing is carried out. These techniques are qual-itative or semiquantitative in nature. Morecomplex readings can be obtained through X-ray spectrometers, thermoelectric instru-ments and eddy current devices—all of which

I@Xavier Spiege] and E. Horowitz, Instruments for the sortingand Identification oj Scrap Metal, Sponsored by the Institute ofScrap Iron and Steel (Baltimore, MD: The Johns Hopkins Univer-sity Center for Materials Research, Oct. 15, 1981), p. 1.

are now available as portable units which canbe used in scrap yards.17

A recent recycling advance has been linkageof scrap identification devices to user friendlycomputers, which provide workers with step-by-step procedures for classifying scraps.Commercially available programs specificallygeared to the needs of small firms in the re-cycling industry have been marketed for sev-eral years. One such program, used in con-junction with a portable X-ray spectrometerdevice, reportedly allows rapid identificationof up to 16 elements in a sample.18 Initial anal-ysis can be completed in 10 to 50 seconds; amore complete quantitative analysis requiresseveral minutes, if needed. If properly pro-grammed, the computer will identify the tradename of the alloy or materials which the scrapmost closely approximates. This is especiallyuseful in obsolete scrap identification, wherethe origin of the materials may not be known.

Identification devices are increasingly usedby the scrap industry–with the encourage-ment of the National Association of RecyclingIndustries, the Institute of Scrap Iron andSteel, and the Bureau of Mines. Most of theseinstruments were designed for laboratory orscientific use, not the difficult operating envi-ronment of the scrap yard. Portable deviceshave the ability to test a large number of sam-ples rapidly on site, but are not capable ofsuch detailed analysis as laboratory or station-ary instruments. As instrument manufactur-ers become more aware of the scrap industry,instruments specifically designed for scrap in-dustry use may be developed.

Likely trends in materials processing sug-gest that scrap classification techniques willneed continuing refinement. Use of morecomplex alloys, and parts made from multi-ple alloys will increase recycling difficulties.

~zNeWell, R. E. Brown, I). M. Soboroff, and H. V. Makar, A Re-view of Methods for ldentijjing Scmp Metals, Bureau of MinesInformation Circular 8902 (Washington, DC: U.S. GovernmentPrinting Office, 1982).

I$AS discuss~ in: “Phi]ips PB9500 X-Ray Spectrometer SpeedsRecycling of Critical Materials,” Norelco Reporter, vol. 28, No.1,May 1981, pp. 40-42.


The Bureau of Mines is identifying alloys forwhich current identification and sortingmethods are not adequate, as a preliminarystep towards development of new methods.19

Another area of government R&D (undertakenby both the Bureau of Mines and the NationalBureau of Standards) is instrumented sparktesting techniques, which could reduce sub-jectivity in spark analysis by workers,

Ultimately, effective recycling requires well-developed linkages among processors, meltshops, fabricators, and manufacturers. Dur-

IOU.S. ~wtient of tie htarior, Bureau of Mktm AVOn~OJeResearch Center Research Program Fiscal Year IW3, (n.p., n.d.)

ing the last decade, many firms in the super-alloy industry established recycling programs,entailing close coordination of requirementsalong the entire chain of the industry. A prop-erly controlled program to ensure scrap integ-rity entails accurate separation and identifica-tion of scrap, and adequate measures to re-duce risk of contamination.20

~A detailed discussion of superalloy scrap processing require-ments can be found in R. S. Cremisio and L. M. Wasserman,“Superalloy Scrap Processing and Traae Elemeht Consideration,”Vacuum Metallurgy: Pmced@a of the 1977 Vacuum MetallurgyConferenm, R. S. Krutenat (ad.) (Princeton, Nl: Science Presa,1977), pp. 353-3S8.

of known composition and history can be vac-uum melted without extensive preprocessing,

In the long term, the potential for recyclingwill depend more heavily on obsolete scrap asthe use of near-net-shape technologies increases.Where traditional machining of superalloyparts can typically require up to 10 times asmuch material as contained in the final part,new processes can reduce this figure to as lowas 2 or 3 to 1. This reduces the amount ofprompt scrap generated and thus increases theimportance of obsolete scrap.

The technical feasibility of using obsolete su-peralloy scrap to produce investment cast jetengine turbine blades has been demonstratedby Certified Alloy Products. The companyreportedly has produced a master alloy fromused superalloy turbine blades.10 The processentails oxygen lancing in an electric arc fur-nace, followed by refining in a vacuum induc-tion furnace. Testing of blades made from therecycled alloy showed comparable quality toturbine blades manufactured from virgin ma-terials. The firm developed the process in theaftermath of the 1978 cobalt price spike andhas subsequently discontinued its use.11 Al-

ity of scrap is suitable for vacuum melting.Such scraps are certified by alloy content andmust be free of contaminants and tramp ele-ments. With careful separation and identifica-tion, some in-house scrap and fabrication scrap

———IOThis process is described in Michael J. Woulds, “Recycling

of Engine Serviced Superalloy s,” Supera]]oys 1980: Proceedingsof the Fourth International Symposium on Superafloys, John K.Tien, et al. (eds.) [Metals Park, OH: American Society for Metals,1980), pp. 31-41.

~lprivate communication between personnel of Certified AIIoYProducts and OTA staff in May 1983.

— -—


though such techniques may enhance the po-tential for reuse of obsolete superalloy scrap,further work to establish process reproducibil-ity and results may be needed.

Both vacuum and air melting furnace tech-nologies continue to evolve. New processes,such as Electron-Beam Refining, Vacuum ArcDouble Electrode Remelting (VADER), andPlasma Arc processes may have considerablerelevance to superalloy recycling. Some ofthese technologies are in the early stages of usein the United States.

Several laboratory-scale processes have beendeveloped which separate superalloy scrap intoits constituent metals, but none have been com-mercialized. Such techniques overcome mostconcerns about contaminants. In a laboratoryproject sponsored by the U.S. Bureau of Mines,Inco developed a process that utilized a com-bination of hydro- and pyre-metallurgical proc-esses to recover 93 percent of the chromium,99 percent of the nickel, 96 percent of the co-balt, and 92 percent of the molybdenum fromclean, solid superalloy scrap.12 Other Bureauof Mines research projects on superalloy scrapare underway. In one experiment, the super-alloy scrap is melted with aluminum or zincto form an intermetallic compound which canbe crushed into small particles. The particlesare then dissolved by acid; the resulting solu-tion would then be treated hydrometallurgi-cally to recover the metals. The process isthought to be especially promising as a meansfor recycling mixed combinations of bulksuperalloy scrap which have not been sorted.13

GTE Sylvania Corp. has a commercial facil-ity for recovering high-purity cobalt from sec-ondary sources that include some types of su-peralloy scrap, including grindings. GTE’s

—’21. 1. cie Elarbi~illO, J. K. Pargetter, and H. V. Makar, Process

for Recoloring Chromium and Other Metals From SuperalloJrScrap, U.S. Department of the Interior, Bureau of Mines Reportof Investigation 8570 (Washington, DC: U.S. Government Print-ing Office, 1981),

lsJera]d R, Pederson, compiler, Bureau (~~ Mines Research ~ ~~.~(Washington, DC: U.S. Government Printing Office, n.rl.), p. 103:a n d (;. B. A t k i n s o n , I n c r e a s i n g the l,eaching Rote a.f HuJkSuperallo~r Scrap b}’ Nfelting \lrith Aluminum, U.S. Departmentof the 1 nterior, Bureau of Mines Report of Investigations 8833(Washington, DC: LJ.S. Government Printing Office, 1983).

proprietary manufacturing process uses chem-ical means to recover highly purified cobaltpowder that is now used in cemented carbideproducts. The firm is considering a larger scalerecovery process which would produce com-pacted briquettes of cobalt that maybe suitablefor superalloy production.14

Commercialization of advanced superalloyreclamation processes by industry is impededby the currently low price of cobalt. Plans toundertake a pilot plant to test the Inco/Bureauof Mines laboratory process discussed abovewere discontinued when cobalt prices fell inthe early 1980s. Feasibility studies showed thatthe pilot plant process, and a small-scale com-mercial facility would have been profitable at1980 metal prices, but would operate at a defi-cit at cobalt prices in the 1981-84 period. Esti-mated costs of a pilot plant project able to han-dle 100 pounds of scrap feed per hour wereestimated to be about $5 million in 1980 dollars,assuming no donation of equipment by indus-try. Operating costs for such a facility wouldbe in the neighborhood of $2 million for an 18-month period.15

RETIREMENT FOR CAUSE AND PRODUCTLIFE EXTENSION PROGRAMS

currently, critical components of jet engines,such as turbine disks, are retired from serviceon a predetermined schedule, based on thenumber of hours a part has been in operation—not on the basis of actual detection of flaws ordefects in the part, This has been necessary be-cause of difficulties in detecting the early stagesof flaws that could lead to a disastrous failureof a part. While the current retirement proce-dures minimize risk of failure, most parts areremoved from service long before actual flawsoccur.

Turbine disks used in the F-100 jet engine,for example, are only kept in service for thenumber of hours in which there is a statisticalprobability of a fatigue crack developing in oneout of a thousand disks. As a result, 999 partscan be expected to be removed from service

141 nformation pro~.ided by GTE Products Corp.1s1 nformation provi(]e(~ by the U.S. Bureau Of Mines.


prematurely for every single part that is defec-tive. Subsequent testing has shown that the use-ful life of many of these retired parts could be10 times longer than the disk in which the ini-tial crack occurs.21

With the substantial technical advances thathave been made in NDE in recent years, it maysoon be possible to evaluate individual partsfrom operating jet engines at prescribed inter-vals in order to determine whether the initialstages of a flaw have appeared. Parts whichhave such flaws could then be retired for cause(RFC); those without flaws could be kept inservice.

Such a strategy underlies the Air Force’sWright-Patterson Aeronautical Laboratory eval-uation of methodologies for a maintenance pro-gram based on retirement for cause. If imple-mented, the program will permit parts fromsome jet engines to be used until there is ac-tual evidence of an incipient flaw. An actualdecision about implementing the program willprobably not be made until after 1985. How-ever, early indications are that the program hasthe potential to reduce spare-part requirements,and therefore to conserve strategic materials.

The initial Air Force evaluation centers onthe F-100 jet engine. If initiated in 1985, theprogram’s savings in spare parts over the 15-year average engine service life left in the F-100 fleet could be $249 million. By 2000, a 70percent material savings from requirements forreduced spare parts would also be achieved,according to the Air Force.

A related Air Force initiative is aimed at“stockpiling” retired jet engine parts until theycan be returned to service rather than simplydisposing of them as scrap, as is ordinarilydone.22 Many of these parts have been retiredsimply because they have reached the end ofthe predetermined service life discussed above,and could be returned to service if retirement-for-cause (RFC) methodologies are implemented.

“J. A, Harris, et al,, Concept Definition: Retirement for Causeof FIOO Rotor Components. Technical Report AFWALTR804118prepared for U.S. Air Force Materials Laboratory at Wright-Pat-terson Air Force Base by Pratt & Whitney Aircraft Group (Wash-ington, DC: U.S. Government Printng Office, 1981).

ZZInformation provided by the U.S. Air Force.

These parts are also being used as test speci-mens for the evaluation of RFC methodologies.Other parts in these “holding accounts” maybe capable of rejuvenation through current andprospective technologies discussed in the pre-vious section.

It is possible to extend the lifetime of com-ponents of the jet engine if flaws and defectscan be caught at an early time. Researchers inWest Germany, for example, have reported thatif an engine is dismantled at three-quarters ofits normal life and the turbine blades are sub-jected to hot isostatic pressing to eliminate in-cipient cracks, the overall life of the blades willextend to 50 percent beyond their normal lifein service, as illustrated in figure 6-1.

Hot isostatic pressing (HIP) has also beenshown to extend the service life of vanes andblades if it is used as a post-casting treatmentto eliminate all porosity remaining in the ma-terial after casting. The post-casting HIP treat-ment has been shown to increase the averagestress-rupture life by 100 percent and the fa-tigue life by 800 percent in two nickel-base al-loys, B-1900 and IN-792, both modified by theaddition of hafnium.

Figure 6-1 .—Service Life Time of Turbine Blades

Dismantle

For For

SOURCE” H. Huff, R. Grater, and R Frohling, “Measures for Materials Conserva.tion in Aero.Engine Construct ion,” paper delivered to the NATO Ad-visory Group for Aerospace Research and Development Conference,Vimelro, Portugal, October 1983


A quite different approach toward materialconservation is to design parts for the life ofthe aircraft. Such approaches may requiremore strategic materials to be used in the ini-tial part, but because these parts have a longerlife expectancy, the end result may be to con-serve materials. Such strategies are beingevaluated through the engine structural in-tegrity program of the Air Force.

OUTLOOK FOR CONSERVATION OF STRATEGIC MATERIALSIN SUPERALLOYS

All of the factors discussed above—materials-efficient processing, recycling, and life exten-sion—will play a role in determining conserva-tion levels achievable in superalloy productionand parts manufacturing. More efficient proc-essing and fabrication could reduce input rawmaterials needed for each jet engine by 10 per-cent by 1990 and 25 percent in the early yearsof the next century, compared to current levels.This is a judgmental estimate, which assumesthat many currently used technologies (whichare less efficient in conserving materials) willcontinue to be used throughout much of theperiod. Advanced technologies are likely to beintroduced primarily when new alloys and suc-cessive generations of jet engines are broughtinto production, typically at 5- to 10-year in-tervals.

As fabrication becomes more efficient, homeand prompt scrap—now preferred for recy-cling—will comprise a less important portionof the materials available for recycling. If proc-essing problems are overcome, obsolete scrapwill probably ascend in importance, althoughto what extent is difficult to foresee. Part lifeextension technologies and RFC maintenancephilosophies could extend the material life cy-cle, thus temporarily reducing obsolete scrapproduction levels, even if advanced recyclingtechnologies are implemented.

Development of new alloys and manufactur-ing techniques by superalloy producers and jetengine makers can have unpredictable effectson recycling. Increased use of powder metal-lurgy, coatings, multiple alloy components, andphase-in of advanced ceramics and compositesmay make the job of recycling more difficult.

For example, coated or clad superalloys mayreduce the value of some superalloy scrapowing to added costs of identifying and recov-ering the surface layer. continuing evolutionin scrap processing technologies to keep pacewith rapid changes by producers and compo-nent manufacturers will be required if thepromise of recycling is to be realized.

Conservation of Manganesein the Steel lndustry23

Nearly 90 percent of the manganese con-sumed in the United States is used in theproduction of steel, either as an alloying ele-ment or as a processing agent.24 Consumptionof manganese in this industry is dependent notonly on the level of steel production, but alsoon the mix of steel grades produced and proc-essing technologies used. This section dis-cusses the function of manganese in both steeland steelmaking, and assesses the effects ofproduct and processing trends on the patternof manganese use in the steel industry.

Manganese as an Alloying Element in Steel

Manganese is second only to carbon in im-portance as an alloying element in steel. Throughits influence on steel chemistry and microstruc-ture, manganese influences an assortment ofphysical and mechanical properties. As dis-cussed in box 6-B, manganese provides sulfurcontrol, enhances hardenability, wear resist-ance, and solid solution strengthening, andretards recrystallization. The amount of man-ganese used as an alloying element depends onthe specific property requirements of the par-ticular steel product. Some grades make greateruse of it than do others.

ZaMuch of the analysis in this section was based on G. R. St.Pierre, et al., Use of Manganese in Steelmaking and Steel Prod-ucts and Trends in the Use of Manganese as an AJloying Elementin Steels, OTA contractor report, February 1983.

zqBased on data from Minera]s Yearbook Voiume 1 (Washing-ton, DC: U.S. Bureau of Mines, Department of the Interior, 1980,1981, 1982). Includes manganese consumed in ores with greaterthan 5 percent Mn, ferroalloys, and metal, but not the manganesecontained in most iron ores

228 ● Strategic Materials: Technologies to F/educe U.S. Import Vulnerability

Box 6-B.—The Role of Manganese as an Alloying Element25

Sulfur Control.--In nearly all grades of steel, sulfur is an unwelcome impurity. It is a tramp elementthat is picked up from the coking coal, fuel oils, and ferrous scrap used in the production of ironand steel. Iron and sulfur form compounds (FeS) at the grain boundaries within steel, causing adetrimental condition known as “hot shortness.” Problems arise because FeS has very little struc-tural integrity, being either liquid or extremely plastic at the temperatures encountered during manyfinishing (hot working) operations. The presence of this strengthless phase as a film along the net-work of grain boundaries severely weakens steel at high temperatures and causes cracking duringfabricating operations, Furthermore, in ferritic grades, FeS causes problems with low-temperaturemechanical properties. To minimize these effects, the sulfur content is typically held below 0.05 per-cent by weight. Manganese is added to react with the remaining sulfur, which would otherwise com-bine with iron. The globular MnS precipitates formed upon the addition of manganese are solid attypical hot working temperatures and are structurally amenable to hot forming operations. Althoughit is not a remedy for all sulfur-related problems, manganese generally improves both the high- andlow-temperature properties of steel.

Hardenability.--Heat treatment is the most potent method for optimizing the strength and ductilityof steel. The cooling rate required to form a given martensite profile during the heat treatment quenchis a very important parameter, known as “hardenability,” of steel. Additions of the common alloyingelements, namely molybdenum, manganese, chromium, silicon, and nickel, facilitate the formationof martensite, thereby reducing the cooling rate requirement and increasing the hardenability of steel.Manganese is a very cost-effective hardenability agent and consequently is one of the most vital alloyingelements for influencing the response of steel to heat treatment.

An allied effect of some alloying elements concerns their influence on the shape and distributionof iron carbides in steel. Weld embrittlement can occur if the iron carbides form a continuous filmat the grain boundaries during the thermal excursion. Manganese tends to enhance weldability ofsteel by inhibiting the development of such film.

Wear Resistance.--Manganese, like nickel, lowers the austenite-to-ferrite transformation temperatureof steel and is termed an austenite stabilizer. High-carbon steels that remain austenitic at room tem-perature can be produced by adding manganese to levels greater than 10 percent by weight. Austeniticsteels as a class are attractive because their strength increases markedly as they are deformed andthey have excellent resistance to fracturing under impact conditions, two properties important forwear resistance. The austenitic manganese grades, called Hadfield steels, display good abrasion andimpact resistance and serve in construction, mining, quarrying, oil-well drilling, steelmaking, cementand clay manufacturing, railroading, dredging, and lumbering applications. Other austenitic grades(containing 15 to 25 percent manganese) have been developed for their nonmagnetic properties andhigh toughness. These high-manganese austenitic steels are suitable for structural uses in strong mag-netic fields and at cryogenic temperatures and may replace the stainless steels commonly used inthese applications.

Solid Solution Strengthening.--Steel generally becomes stronger when an alloying element is dissolvedin the iron phases. This effect is known as “solid solution strengthening.” It is a very important meansof strengthening ferritic steels, which cannot be strengthened by heat treatment. Alloying elementsvary markedly in their potency as solid solution strengtheners. In ferritic steels the order of decreas-ing effectiveness appears to be: silicon, manganese, nickel, molybdenum, vanadium, tungsten, andchromium. While silicon is highest on this list, it is not used extensively because of an accompanyingloss of ductility. Thus, manganese has great utility as a solid solution strengthener.

Recrystallization.–When steel is deformed or heavily worked, its microstructure is altered in a waythat is detrimental to the properties of steel. To alleviate problems, steel in this condition is usuallyannealed, whereupon the microstructure is recrystallized. Manganese slows the rate of recrystalliza-tion considerably. This is not a desirable effect and can have negative impacts on the productivityof various steel-treating processes, such as continuous annealing operations.

~~he terms ferrite, austenite, and martensite that appear throughout this discussion refer to the various microstructural phases of steel.


MAJOR CLASSES OF STEEL

Steel is produced in a myriad of composi-tions, and the relative popularity of each gradechanges as consumer demands shift. Trendsin manganese use vary among the four broadcategories of steel: carbon, full alloy, high-strength low-alloy (HSLA), and stainless. Allfour classes contain manganese, but in differ-ing proportions.

The most widely used systems for designat-ing steels in terms of their chemical composi-tions are those of the American Iron and SteelInstitute (AISI) and the Society of AutomotiveEngineers (SAE). These cataloging systems arevery similar and are carefully coordinated bythe two groups. In addition to the commongrades listed by these organizations, steel-maker produce proprietary grades, but themanganese contents of these are comparable.The chemical specifications for a particulargrade of steel vary slightly according to theproduct nature. Specifically, the ranges inplates are a little wider than those for bars, sim-ply because compositional variations due tosegregation are greater in the large rectangu-lar ingots used in plate production than in thesmaller ingots used for bars.

Carbon steel.—The main constituents of carbonsteels are iron, carbon, and manganese, withmanganese present up to a maximum of 1,65percent. The majority of carbon steels havemanganese specifications whose midpoints fallin the range 0.40 to 0.85 percent. The averagemanganese content of the entire carbon steelcategory is estimated to be about 0.65 percent,or 13. o pounds per ton of steel.

Manganese is used in carbon steels primar-ily for its sulfur control and solid solution-strengthening characteristics. It also improvesthe heat treating characteristics (hardenability)of these grades, though this is not a major con-sideration for many uses of carbon steel.

Two types of carbon steel have especiallyhigh manganese and sulfur contents in orderto improve machinability, These resulfurizedand resulfurized-rephosphorized grades typi-cally contain 0.7 to 1.65 percent manganeseand find service in applications, such as gears

and cams, where considerable machining isneeded. The large number of manganese sul-fide particles in these steels improve surfacefinish, enhance metal removal rates, and in-crease the tool life by facilitating the chip for-mation process.

Full Alloy Steel.—Alloy steels usually containmore manganese and silicon than do carbonsteels and, in addition, may contain other ele-ments such as chromium and molybdenum.The extra alloying additions enhance mechan-ical properties, fabrication characteristics, orother attributes of the steel. In the more com-mon alloy steel grades, manganese is added pri-marily for hardenability and less importantlyfor sulfur control. The average manganese con-tent of alloy steels is estimated to be 0.75 per-cent, or 15. o pounds per ton of steel.

There are experimental alloy steels whichhave not yet been assigned regular AISI-SAEdesignations. In general, these experimentalgrades have higher manganese specificationsthan the standard grades. This suggests thatmanganese is becoming more popular as an al-loying element and that the average manganesecontent of alloy steels should be expected toincrease.

Although most alloy steels contain under 2percent manganese, several grades containing10 to 14 percent manganese are used. Theseare the Hadfield grades, which have excellentabrasion and impact resistance and are usedin construction, mining, and other high-wearapplications. There are also alloy steels withmanganese contents as high as 25 percent thathave been developed for use in strong magneticfields and at cryogenic temperatures.

HSLA Steel.—HSLA steels have been devel-oped as a compromise, having the convenientfabrication characteristics and low cost of car-bon steels and the high strengths of heat-treatedalloy steels. HSLA steels are generally used inapplications such as energy production and au-tomobiles, where savings in weight can bemade as a result of their greater strength andatmospheric corrosion resistance and wherebetter durability is desired.


HSLA’s derive their superior properties froma variety of alloying elements in relatively lowconcentrations of 0.2 to 1.6 percent, coupledwith controlled rolling and heat treatments.Common alloy additions include manganese,nickel, chromium, columbium, molybdenum,titanium, and vanadium. In comparison withthe majority of carbon and alloy steels, theHSLA steels contain greater manganese con-tents. The limits in the 14 alloys designated bythe SAE range from 1 to 1.65 percent manga-nese, with an average of about 1.1 percent, oran average of 22.0 pounds per ton of steel.

Stainless Steal .—Stainless steels are ferrous al-loys that are rich in chromium and sometimesnickel. They are used in a variety of applica-tions requiring good corrosion and oxidationresistance, as well as good strength. Composi-tional specifications for stainless steels usuallydo not prescribe minimum manganese levels,only maximum concentrations. Stainless alloy304, which constitutes roughly 42 percent ofthe stainless steel produced in the United States,has a manganese specification of 2.0 percentmaximum. On the average, stainless steels con-tain 1.7 percent manganese, or 34.0 pounds perton of steel.

Several stainless steels, however, contain farmore manganese than the average. In part,these alloys reflect the fact that an increase inmanganese levels permits a reduction in nickelcontent. Tenelon steel, containing 14.6 to 16.5percent manganese, was developed by U.S.Steel Corp. following concern over nickelshortages that occurred during World War IIand the Korean conflict.

TRENDS IN THE STEEL PRODUCT MIX AND MANGANESE USE

The average amount of manganese consumedstrictly for alloying purposes is being influ-enced by two major trends. The first is a chang-ing product mix. Steel consumers are using in-creasingly greater proportions of full alloy,stainless, and HSLA steels, which usually con-tain more manganese than do carbon steels.Figure 6-2 shows the historical product sharesfor each of the major steel classes. The classescontaining the most manganese, namely, fullalloy, HSLA, and stainless steels, are expected

to increase their portions of the market. By2000, the product shares for the various prod-ucts are expected to be approximately: carbon80 percent; full alloy 9 percent; HSLA 9 per-cent; and stainless 2 percent.

The second contrasting trend is the decreas-ing manganese composition of carbon steels.This is a result of improvements in the steel-making process, namely, better sulfur removaltechniques and more reliable composition con-trol, This effort is driven by economic pres-sures to conserve raw materials and increaseproduct fabricability. New technologies allowsteelmaker to retain the same quality productwith less manganese. They can now afford toreduce manganese content to save on rawmaterials costs and to avoid some of the moreunpleasant attributes of manganese, such as therecrystallization problem alluded to in box 6-B. Two manganese reduction techniques arelikely to be administered: trimming to the lowerend of the specified manganese range for eachof the grades and reduction of minimum man-ganese compositions because less is needed forsulfur control.

Trimming.—Most manganese specifications forsteels are listed as ranges. To be on the safeside, steelmaker generally aim for the mid-points of these ranges. Trimming involves aim-ing for the lower end of the specified range.Improved ladle practice is the key to effectivetrimming. Better manganese composition con-trol is attainable with improved melt protec-tion, alloy additions packaging, and sensing,analytical, and control instrumentation. Trim-ming practices could be implemented veryquickly in an emergency, especially if opera-tors become better aware of the statistical na-ture of quality control. The manganese contentof steel could be reduced by as much as 0,05to 0.10 percent with the use of trimming.26

Reduction in Specified Manganese Range.—Sulfurspecifications of most steels, except for theresulfurized grades, are usually about 0.05 per-cent maximum. In practice, steels are pro-duced with average sulfur contents of about

ZeJohn R. stubbles, “Conservation Could Cut Steelmaking Re-quirements,” American Metal Market, Oct. 20, 1983.

Ch. 6—Conservation of Strategic Materials Ž 231

Figure 6-2. —Historical Market Shares for Major Classes of Steel

stainless

SOURCE Office of Technology Assessment, based on shipments data from the American Iron and Steel Inst!tute, Washington DC


0.025 percent. Manganese specifications couldbe reduced to reflect this lower sulfur level.

Technologies such as external desulfuriza-tion of blast furnace iron and ladle metallurgytechniques allow steelmaker to reduce the sul-fur content of the finished product further still.External desulfurization involves injecting cal-cium carbide or magnesium metal into the hotmetal prior to its introduction to the basic oxy-gen steelmaking units. Lime is sometimes in-jected as an auxiliary agent with these desul-furization materials. Implementation of thistechnique will become almost universal in thedomestic industry by 1990. This will reduce theneed for desulfurizing during the steelmakingstage and lead to much lower residual sulfurlevels in new steel and, eventually, in recycledscrap. Furthermore, ladle injections of desul-furizing agents such as lime, calcium, or rareearths just prior to primary casting reduces thesulfur content below 0.01 percent, and oftenbelow 0.005 percent. Ladle metallurgy technol-ogies may not be installed primarily for sulfurreasons, but once they are in place they willno doubt be used to reduce sulfur levels in ad-dition to providing the other benefits.

By 2000, the average sulfur content of steelmay be as low as 0.01 or 0,02 percent. Thismeans that less manganese is required in thefinished product solely for sulfur control pur-poses. Even though proportionally higher man-ganese-to-sulfur ratios are required at lowerlevels of sulfur, a net reduction in the actual

amount of manganese needed is realized by re-ducing the sulfur content. The manganese con-tent of each grade of steel could probably bereduced by 0.2 percent with the widespread im-plementation of external desulfurization andladle technologies. For carbon steels, this rep-resents a 30-percent reduction in the manga-nese content.

SUMMARY OF TRENDS IN THE MANGANESECOMPOSITION OF STEELS

The effects of the product mix and manga-nese composition trends on the average man-ganese content of steel is illustrated by table6-2. Note that even though the high-manganesealloy products become more pervasive, theaverage manganese composition is likely todrop from 0.69 to 0.61 percent. That is a re-duction of 1,6 pounds (from 13,8 to 12.2 pounds)of manganese for each ton of steel.

Manganese in the Steelmaking Process

On an industry average, almost three timesmore manganese is added to steel during itsprocessing than ends up in the finished prod-uct. When the 11 percent of the manganese ad-ditions that comes from in-house recycling isdiscounted, the overall manganese recoveryrate comes to about 39 percent. This sectionoutlines the major trends in steelmaking prac-tice that affect the efficiency of manganese usein the production process.

Table 6-2.—Manganese Content for Major Grade of Steel

Current (1982) Forecasted 2000a –

Market Manganese Market Manganese –

share content share content

Carbon b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87.6 0.65 80 0.5 -

Full alloy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1 0.75 9 0.8HSLA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.8 1.1 9 1.2Stainless . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.5 1.7 2 1.7

Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100.0 0.69—

100 0.61aDra~tiC ~h~”~~~ in the bala”c~ of economic activl!y Could shift the market shares in 2000 to: 90 percent carbon, 9 PerCent combined full alloY and HSLA, and 1 Percen~

stainless In the low case; or 70 percent carbon, 27 percent combined full alloy and HSLA, and 3 percent stainless in the high case An average manganese contentof O 67 percent (high case) might arise from a shift In favor of alloy steels caused by a severe reduction in new construction and manufacture of machinery, vehicles,and containers with a relative gain in the construction of new chemical, petrochemical, coal conversion, and powerplants. Similarly, a substantial shift (n favor ofcarbon steels caused by extremely high demand for new buildings, highways, vehicles, and primary machinery might result In a manganese content of 0,56 percent(low case)

bTrimming, together with reductions in the specified manganese range, could cut the manganese content of carbon steels by 0.25 to O.so p3rCent. A COrl SerVatiVe reduc-

tton of O 15 percent is shown here to account for less than full implementation of these practices.

SOURCE: Office of Technology Assessment based on data from the American Iron and Steel Institute, Washington, DC, and G R St Pierre, et al , Use of Manganesein Stee/nraking and Stee/ Products and Trends in fhe Use of Manganese as an A//eying E/ernent In Stee/s, OTA contractor report, February 1983


PATTERNS OF MANGANESE USE IN STEELMAKING

Figure 6-3 shows the conventional steelmak-ing sequence, from raw materials preparationto casting and finishing. Manganese enters theprocess flowstream from five distinct sources:iron ore, fines and waste materials, manganeseore, iron and steel scrap, and ferromanganese.

On an industry average, iron ore and finesand waste materials contribute approximately36 percent of the manganese (disregarding thatpresent in ferrous home scrap] ultimately addedto the process stream. Iron ores contain ap-proximately 0.16 percent manganese.27 Al-though this is a small concentration (and hasbeen declining in recent years), large quantitiesof iron ore are added to the blast furnace. Finesand waste materials generated by the steel pro-duction process contain large amounts of man-ganese. Though currently there are no satisfac-tory methods for extracting manganese fromthe iron and steelmaking slags, dusts, sludges,scales, and other wastes, some plants operatesinter strands to recycle some of these mate-rials, The ore sinter (consisting of partiallyfused limestone, fine iron ore particles, andsteelmaking dust and other waste products)produced in sinter strands can be charged(added) to blast furnaces for hot metal produc-tion. Steelmaking slags and dusts consume alarge proportion (approximately 42 percent) ofthe manganese added to the process stream.Slags from both electric and basic oxygen steel-making furnaces can be charged into blast fur-naces, however, basic oxygen furnace (BOF)slags are more commonly recycled in this man-ner. This is because of the proximity of BOFsto blast furnaces—BOFs are always part of in-tegrated steelmaking facilities, while electricarc furnaces (EAF) are not, According to a 1976NMAB study, about 40 percent of the BOF slag(containing on average 4.5 percent manganese)produced is recycled to the blast furnace: therest is discarded. 28 No blast furnace slag is

ZpA~~~rding t. the American Iron Ore Association (Cleveland,OH), in 1982, 7(I percent of U.S. iron ore consumption was fromU.S. sources, while 20 percent was from Canadian sources, Onaverage, the U.S. ores contained 0.12 percent manganese andthe Canadian ores contained 0.31 percent manganese.

Z8LN ~tiona] Materials Advisory Board, I?langanesc ~e[:otr[?r}’

Tcchnolog}, publication NMAB-323 (Washington, DC: NationalAcademy Press, 1976].

recycled in this manner, but it does find othercommercial uses in the brick, aggregate, glass,and bedding industries. There are no reliabledata on the overall extent to which manganeseis recovered from steelmaking waste materials,but most likely less than 20 percent of themanganese contained in iron and steel indus-try waste materials is currently recovered forreuse.

Manganese ore charged to the blast furnaceaccounts for 5 percent (industry average) of thetotal input of manganese to the processingstream. Charging this ore is a relatively inex-pensive way of increasing manganese concen-trations to levels beneficial for the processingcharacteristics of ironmaking and steelmaking.

Iron and steel scrap is charged to the steel-making furnaces and, to a lesser extent; theblast furnace. The manganese content of homescrap depends on the stage in the steelmakingprocess at which it is generated. That createdat the steelmaking stage contains roughly 0.2percent manganese, while that coming fromcasting and finishing reflects the manganesespecifications of the product mix of the par-ticular mill in question. Manganese contribu-tions in purchased scrap depend on the trendsin the manganese content of steel products, dis-cussed in the preceding section.

Ferromanganese is added to the processstream during the ladle refining stage. Approx-imately 45 percent (industry average) of the to-tal manganese input is added as ferromanga-nese. This manganese is not intended to besacrificed for the sake of processing improve-ments. Coming late in the steelmaking process,it is intended to bring the level of manganesein the steel up to final product specifications.

Much of the manganese contributed by thesefive sources does not make it into the finishedproduct, but ends up instead in steel scrap andprocessing waste products. The percentagethat does end up in the finished product is afunction of the manganese recovery factors forthe various process stages. These recovery fac-tors are dependent on process yield rates andmanganese migration characteristics. Pooryield rates result in both large quantities of


Figure 6-3.— Flow Diagram Showing the Principal Process Steps Involved in Converting Raw Materials Into theMajor Steel Product Forms, Excluding Coated Products

1coal Coal

rmines

+

1 1 L J

v ~

Manganeseore

●

Alloyingelements

andadditionagents

‘ea.. FeMnr -, m

e.g. Ferromanganese

I billet frills)

~ —Shows manganese

SOURCE The Making, Shaping, andmanganese flows.

Sheets Coils

additions.

Treating of Stee/, the United States Steel Corp., Pittsburgh, PA, 1971, adapted by the Office of Technology Assessment to show

-. . - . -. . . . . ---Ch. 6—Conservation of Strategic Materials • 235

wastes that cannot be recycled, and sizableamounts of home scrap from which additionalwastes are produced during recycling. Man-ganese losses are exacerbated because wastematerials such as slags and dusts commonlycontain manganese in concentrations greaterthan those in the steel, Manganese, as well asmany other alloying elements, often migratesto the process wastes. Sometimes this segre-gation of elements between the steel and thewastes is intentional, as during the steelmak-ing stage when manganese is used to react withdissolved oxygen and carry it into the slag.29

In a sense, these are “sacrificial losses. ” Inother instances the segregation is an unwantedconsequence of the process chemistry, Regard-less of the intent, this manganese migration

2Q~xygen disso]tred in mo]ten steel reacts with carbon to Pro-duce gaseous carbon monoxide, which, when trapped duringingot solidification, causes defects known as blowholes. Thisproblem can be avoided by reducing the oxygen content of thesteel with the addition of elements that react with the oxygenand carry it off to the slag. Manganese, aluminum, and siliconare the most important of the deoxidizing agents. Though a weakdeoxidizer by itself, manganese enhances the deoxidizing ca-pacity of silicon. Moreover, the presence of manganese oxidein the slag helps to develop proper slag characteristics, in partthrough more rapid dissolution of lime, If the manganese con-tent of the steelmaking charge is too low, additional slag fluxescontaining fluorspar, which is considerably more expensive thanferromanganese, must be used.

causes the recovery factors to be lower thanthe yield rates.

TRENDS IN STEELMAKING TECHNOLOGY

Figure 6-4 shows the current pattern of man-ganese use in the production of one ton of steelproduct. Of the total manganese consumed inthe steelmaking process, 39 percent is embod-ied in the finished steel product, and 61 per-cent is lost in slags, dusts, and wastes. Note that13 percent of the manganese inputs are recy-cled as home scrap, generated principally dur-ing the casting and finishing steps, The flowsin the figure are industry averages and do notreflect operations in any particular steel plant,They are aggregates based on the current mixof equipment and operating characteristics forthe industry. Shifts in this mix of steelmakingtechnologies are expected to change the man-ganese flow pattern.

Ironmaking.—From 50 to 80 percent of themanganese charged to blast furnaces is re-tained in the molten pig iron, Typically, theaverage recovery rate is 75 percent. The re-mainder is lost to blast furnace slag, which isnot recycled or processed for its manganesecontent. Although construction and operationof direct reduction ironmaking plants will con-

Figure 6-4.—Current Pattern of Manganese Use in Steelmaking (pounds of manganese of finished steel)

Home scrapPurchased

4.97 I I 0.47

W a s t e

2.86 I

1 F i n i s h e d

‘Recovery with respect to ferromanganese additions IS assumed to be 82 percent, while complete recovery of Mn IS assumed for raw steel

SOURCE Office of Technology Assessment; and G R, St Pierre, et al , Use of Manganese in Sfeelmak(ng and Sfee/ Products and Trends in fhe Use of Manganeseas an A//eying E/ernent in Sfee/s, OTA contractor report, Februa~ 1983.


tinue, particularly in certain parts of the world,e.g., the Middle East, Mexico, South America,and Soviet Union, where natural gas is abun-dant, the principal source of domestic crudeiron will continue to be the blast furnace,

Over the next few years, the trend to shutdown the oldest and least efficient blast fur-naces and to upgrade the better furnaces willcontinue. Furnaces are upgraded in a numberof ways; including modifications of the charg-ing systems and relining configurations, adop-tion of improved hot blast, additive, pressuri-zation, and tapping procedures, and installationof better instrumentation and control equip-ment. All of these developments taken togetherwill increase the operating performance ofblast furnaces but are unlikely to have a majoreffect on manganese use.

Manganese retention in the blast furnace canbe expected to decrease somewhat from its cur-rent value of 75 percent because of the use ofmore acidic slags allowed by external desul-furization, However, this should be offset bythe opportunity to include more manganese-bearing steelmaking slag (primarily for its iron,lime, and flux content, but also for manganese)in the blast furnace charge.

Steelmaking.—Currently, manganese recoveryin BOF steelmaking ranges from 20 to 50 per-cent, depending on the final carbon contentand temperature of the melt (manganese re-covery improves with increases in both), Re-covery in electric steelmaking is substantiallygreater, The proportion of steel produced inEAFs is increasing relative to BOFs. The cur-rent ratio of EAF steel to BOF steel, roughly30:70, is expected to increase to 40:60 by 2000.This not only improves the average manganeserecovery rate, but also increases the averageferrous scrap charge rate. Scrap is typically 100percent of the charge to EAFs while only 30percent to BOFs. Thus, greater use of EAFs in-creases the recycling of manganese,

Operating practices for both EAFs and BOFsare changing. The basic oxygen processes areundergoing dramatic changes with the incor-poration of bottom blowing and inert gas mix-ing techniques, Manganese recovery will prob-

ably increase to the range of 40 to 60 percentover the next few years as a result of these andother changes, such as separate desiliconiza-tion and better control of the charge and slag.Manganese recovery in EAFs is expected to im-prove somewhat over the next few years frombetter charge materials, reduced oxidation dur-ing high-speed meltdown, improved end-pointcontrol, and greater use of secondary and la-dle refining.

Based on current operating practices andmix of BOF and EAF steelmaking, the averagemanganese recovery of the steelmaking stageis estimated to be 26 percent. The increasinguse of EAF steelmaking and improvements inoperating characteristics are expected to im-prove this average recovery rate to 42 percentby 2000.

Ladle and Secondary Refining.—The role of ladleand secondary refining has grown rapidly dur-ing the past few years, In stainless steel produc-tion, adoption of AOD processing, which is asecondary refining process, is almost universal.For low-alloy and carbon grades, deoxidationand desulfurization procedures have beengreatly improved with the increased use of ar-gon shrouding of steel surfaces and pouringstreams. In addition to the trimming benefitsmentioned in the previous section, these proc-esses capture the manganese content of ferro-alloy additions to a much higher degree andreduce reoxidation loss during subsequentpouring. These improvements are expected toraise the recovery of manganese from the fer-roalloys from 82 to 90 percent which raises theoverall manganese recovery rate of the proc-ess from 87 to 95 percent.

Casting and Finishing .—The benefits of continu-ous casting have been fully established. Con-tinuous casting leads to overall cost savings aswell as improved product quality (cleanliness,homogeneity), reduction of energy require-ments and decreased metal loss (scaling) dur-ing ingot reheating. In addition, it gives greateryields of steel products from raw molten steel.30

30 The in g o t. to-s]ab Yie]d of c o n v e n t i o n a l i n g o t c a s t i n g o p e r a -

tions varies greatly according to the nature of the products.Owing to large shrinkage cavities, the yields for killed grades

Ch. 6—Conservation of Strategic Materials

Direct rolling of hot slabs and billets

Photo credtt Amertcan Iron and Steel lnstttute

Continuous casting of steel is one of a variety ofimproved steelmaking technologies

that conserve manganese

From the manganese standpoint, increasedyield means that less home scrap must be recy-cled through processes where there are ir-recoverable losses. In 1982, about 29 percentof domestic steel was continuously cast. Thisrose to about 36 percent in the first half of 1984,and is likely to increase to 70 percent by 2000.Manganese losses are not a major problem atthe casting stage because there is very littlemigration of manganese to the waste products.— —can be quite poor, even as low as 70 percent. The yields forrimmed products are substantially better. On an industrywidebasis, the average yield of ingot casting operations is estimatedto be approximately 87 percent. Continuous casting, the moremodern and more desirable process, has a yield of approximately96 percent.

. 237

fromcontinuous casters has been demonstrated inseveral plants throughout the world. By elimi-nating the cooling and reheating steps, thistechnology offers potential for further improve-ments in metal recovery as well as energysavings.

Based on the trends toward greater use ofcontinuous casting and direct rolling, the yieldrate during casting and finishing operations isexpected to increase from its 1982 value of 75to 85 percent by 2000.

Summary of the Effects of Processing Trends

The current and expected yield rates andmanganese recovery factors for each stage ofthe steelmaking process are shown in table 6-3. These embody the shifts in technology dis-cussed above. The overall manganese recov-ery rate increases from 39 to 49 percent.

Outlook for Manganese Use in the Steel Industry

The expected impacts of steel product andprocessing trends on the manganese use pat-tern in 2000 are outlined in table 6-4. High andlow cases are included in the table to illustratethe effects of different product mixes and var-ious schedules of investment in new equipment.

A 12-percent net reduction in the averagemanganese content of steel (from 13.8 to 12.2pounds per ton) is expected to accompany thecontrasting shifts towards the greater use ofhigh-manganese products and the increased at-tention to trimming. Trends in steelmakingtechnology are expected to cut average man-ganese consumption losses from 21.8 to 12,6pounds per ton of steel, a reduction of 42 per-cent. In addition, these technology changes in-crease the role of purchased ferrous scrap(available domestically) as a source of man-ganese.

Most importantly, these product and proc-ess trends decrease the per-ton-of-steel require-ment for manganese in the form of ferroalloysand manganese ore by 47 percent, from 17.8to about 9.5 pounds. Depending on the prod-uct mix and the rate of adoption of the proc-ess improvements, the requirement could be


Table 6-3.—Parameters Affecting Manganese Retention in the Steelmaking Process

Current (1982) Expected 2000

Manganese ManganeseYield rate recovery factor Yield rate recovery factor(percent) a (percent) b (percent) a (percent) b

Ironmaking (blast furnace) . . . . . . . . . . . . . . 100 75 100 75Steelmaking (BOF and EAF). . . . . . . . . . . . . 85 26 90 42Ladle and secondary refiningc . . . . . . . . . . . 100 87 100 95Casting and finishing . . . . . . . . . . . . . . . . . . 75 75 85 85ayleld ~at~ = tlbs of iron in primary Output product) + (Ibs Of iron in inPut materials),bRecoveW factor = (Ibs of manganese in primary output product) + (Ibs of man9anese in inPut materials).CRecoveV factors for ladle and seconda~ refining are based on Ir)() percent recovey of manganese in molten steel combined with 82 percent (current) and W pOrCent

(2000) recovery of the manganese in ferroalloy additions.

SOURCE: Office of Technology Assessment; and G R. St Pierre, et al , Use of Manganese in Sfee/making and Sfee/ Products and Trends in the Use of A4arrganeseas an A//eying E/emerrt in Stee/s, OTA contractor report, February 1983

Table 6-4.—Current and Projected Manganese Consumption in U.S. Steel Production(all figures in pounds of manganese per ton of steel product, except where noted)

Current Projected, Year-2000

(1982) Expected Low High

Total manganese use (excluding home scrap) . . . . . . . .outputs:

Retained in steel products . . . . . . . . . . . . . . . . . . . . . . .Losses (slag, dust, waste) . . . . . . . . . . . . . . . . . . . . . . .

Inputs:Manganese ore. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Manganese ferroalloys . . . . . . . . . . . . . . . . . . . . . . . . . .Iron ore and sinter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Purchased ferrous scrap. . . . . . . . . . . . . . . . . . . . . . . . .

Overall manganese recovery rate(ignoring home scrap) (percent). . . . . . . . . . . . . . . . . . .

Assumptions:Manganese content of steel (percent) . . . . . . . . . . . . .Percentage of molten steel produced in

EAF processes (percent) . . . . . . . . . . . . . . . . . . . . . .Manganese recovery rates and yield rates:

Ironmaking:Recovery (percent) . . . . . . . . . . . . . . . . . . . . . . . . . . . .Yield (percent) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Steelmaking:Recovery (percent) . . . . . . . . . . . . . . . . . . . . . . . . . . . .Yield (percent) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Ladle and secondary refining:Recovery (percent) . . . . . . . . . . . . . . . . . . . . . . . . . . . .Yield (percent) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Casting and finishing:Recovery (percent) . . . . . . . . . . . . . . . . . . . . . . . . . . . .Yield (percent) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

35.6 24.8 20.6 30.4

13.417,0

13.821.8

12.212,6

11.29.4

1.915.912.95.0

1.28.38.56.8

1.05.57.16.9

1.512.210.16.6

39 49 54 44

0.69 0.61 0.56 0.67

3828 43 48

75100

75100

75100

75100

2685

4290

5093

3487

87100

95100

98100

90100

7575

8585

9090

8080

SOURCE Office of Technoloav Assessment: and G R. St. Pierre, et al., Use of Manganese in Steelmaking and Steel Products and Trends in fhe Use of Manganeseasan A//eying E/efient in Stee/s, OTA contractor report, February 1983 -

as low as 6.5 pounds per ton or as high as 13.7 United States, accounting for over 30 percentpounds per ton. of all PGM consumption. Platinum, palladium,

and, since 1981, rhodium are the active ingre-Automotive Catalysts clients used in converters to control hydrocar-

bon (HC), carbon monoxide (CO), and oxidesSince 1975, the catalytic converter (now used of nitrogen (NOX) emission levels. In addition,

on most cars to control air pollution) has be- about 10,000 tons of chromium are used annu-come the single largest use for PGM in the ally in the stainless steel shell of the catalytic


converter—about 1.5 percent of U.S. chromiumconsumption, and 3 percent of chromium con-sumption in stainless and alloy steels.

As a result of nearly a decade of using cata-lytic converters, a very large “above-the-groundmine” of PGM has accumulated in the Nation’sautomotive fleet-some portion of which canbe retrieved each year through collection andrecycling of converters from scrapped cars. Asthe fleet of converter-equipped cars ages, theinventory of PGM available from scrapped carswill grow very rapidly through the 1980s andinto the 1990s. If effective retrieval and recy-cling of catalytic converters from these scrappedcars occurs, U.S. PGM imports can be reducedappreciably below what they would be other-wise,

The current level of PGM recovery from cata-lytic converters is low, in part because con-verter-equipped cars are only now appearingat scrapyards in large numbers. This maychange in the near future: the average life ofa car is about 9 to 10 years, so post-1975 modelcars (the first year in which production modelcars were equipped with a converter on a na-

tionwide basis) will comprise an increasingproportion of scrapped vehicles.

Figure 6-5, based on projections prepared forOTA by Sierra Research and Energy and Envi-ronmental Analysis, Inc., compares the maxi-mum theoretical potential for catalyst recyclingfrom scrapped cars with projected PGM de-mand for new vehicles through 1995.31 A sshown, the theoretical potential for recyclingwill grow from about 115,000 troy ounces in1983 to over 800,000 troy ounces in 1995. As-suming a 50 to 60 percent recovery rate, cata-lytic converter recycling could add 400,000 to500,000 troy ounces annually to domestic PGMsupplies by the mid-1990s. This compares withnew vehicle demand of 770,000 troy ounces in1983, which is projected to grow to 1.4 milliontroy ounces in 1995. Less economic incentiveexists to recycle chromium values from thestainless steel shell of catalytic converters.

Slsierra Research and Energy and Environment Analysis, Inc.,Critical Metal Consumption in Automotive Catalysts, Trends andAlternatives, report prepared for OTA, December 1983, Muchof the analysis and data in this section are derived from thisreport.

Figure 6-5.— PGM Demand by Vehicle Category

Heavy trucks and buses

75 76 77

SOURCE: Sierra Research

90 91

—

~

\ ,,1I

II

I I

I I

J

92

— —

,,

I II

II I

II

‘ 1 ’ ;

I

‘1~’l I ~~• ,

III I

93 94

Model year


However, if the above recovery rates are as-sumed, 5,000 to 6,000 tons of chromium couldbe provided annually.

It should be noted that considerable uncer-tainty surrounds the projections in figure 6-5—especially those related to PGM requirementsfor new vehicles. The projections for new ve-hicle PGM requirements are for the total U.S.market for vehicles, without assumption aboutprospective import levels. The projected de-mand is based on medium growth assumptionsabout vehicle sales. The projections also reflectanticipated new emission control standards ex-pected to be applied to vehicles in the 1985-90period. Changes in the timing and nature ofthese standards would change the projections.Readers are referred to chapter 3 for furtherdiscussion of the projected growth in PGM usefor vehicle emission control.

Somewhat less uncertainty surrounds theprojections of PGM potentially available forrecycling through 1995. Cars already on theroad will constitute the great majority of ve-hicles scrapped during the projection period.Actual recycling levels will be considerablylower than the theoretical potential, due tolosses of PGM while cars are in service, andat each of the several steps entailed in recy-cling. Estimates of PGM losses from catalyticconverters while cars are in service range fromunder to well over 10 percent, and this is notreflected in figure 6-5. According to the Au-tomotive Dismantles and Recyclers Associa-tion (ADRA), only 70 to 80 percent of the 10million cars that are “detitled” in a normalscrappage year (1979 was the last such year)actually reach dismantling yards where theconverter can be removed for recycling. More-over, not all dismantles remove the converterfrom the car before it is compacted into a baleor shredded. Some PGM shipped to a refiner(perhaps 10 percent) is lost in refining. Hencepotential addition of PGM to U.S. suppliesfrom catalytic converter recycling probablywill only be 50 to 60 percent of the levels shownin figure 6-5, Actual levels or recycling will de-pend on economic incentives and the effective-ness of the collection networks that are estab-lished.

Structure of the Recycling Industry

Among the types of firms that have an inter-est in the growing market in PGM recyclingare primary and secondary PGM producers,manufacturers of emission control devices,scrap processors and dealers, automotive dis-mantles, muffler shops, collectors and distrib-utors of used automotive parts, and decan-ners—a new specialized business that has aris-en around the catalytic converter. There are nomajor technological barriers to recycling of ei-ther PGM or chromium values from catalyticconverters, and the number of firms involvedin catalyst collection networks is expected toincrease gradually as the market continues togrow. In addition, the industry appears capa-ble of responding quickly to price increases,so that recycling could be greatly expanded ifthere were a supply disruption.

The industrial refining capacity needed torecycle PGM from automotive catalysts takenfrom scrapped cars is already in place. Majordomestic PGM refining facilities operated byGemini Industries and Johnson Matthey areavailable for converter recycling. These facil-ities are able to recover PGM values both fromnew catalysts rejected by automakers and fromscrapped converters. Domestic scrap proces-sors are also able to recycle the stainless steelshell of the catalytic converter as a scrap feed.Steps entailed in recycling spent catalysts aredescribed in box 6-C.

Institutional factors could impede recyclinglevels, however. Three categories of catalyticconverters are potential sources of PGM forrecycling: unused catalysts that fail to meetproduct specifications, rejected and damagedcatalysts from operational vehicles, and spentcatalysts from scrapped vehicles. Of these, thelast category represents the largest future sup-ply of PGM, but also presents the greatest col-lection difficulties. Reasons for this includefluctuating PGM prices, the fragmentation ofthe automotive dismantling industry, and alack of effective linkages between refiners anddismantles.

The availability of automobile catalysts forrecycling is affected by several factors. Some


Box 6-C.—How Catalytic Converters Are RecycledWith the advent of the catalytic converter, a new industry has evolved to recover PGM values

from scrapped automobile catalysts. From a technical standpoint, there are few barriers to refin-ing PGM from spent catalysts. In the mid- and late 1970s, facilities were set up to recycle unusedcatalysts that failed to meet product specifications. This gave refiners the experience needed totake advantage of the far greater amount of PGM that can be recovered from scrapped cars. How-ever, the business of organizing effective collection systems has proven surprisingly complex.

Part of the reason for this is that catalytic converters are not uniform products whose valuecan be easily identified. In 1982 model cars, for example, there is a 10-to-l range in PGM contentamong different converter models—().33 troy ounces on heavily loaded converters down to 0.03troy ounces on lightly loaded converters. (This wide range in catalyst loadings is probably not ameasure of the efficiency of the converter in reducing emissions; instead the heavily loaded catalystsare probably installed in car models with engines that produce higher than average emission levelsprior to catalytic treatment.) Many of the converters, especially on older model cars, have catalystpellets (some of which maybe lost in the course of vehicle operation), while others are monolithic,with the catalyst deposited on a secure substrate. Chromium content in the stainless steel shellof converters also varies from 1.5 to 3 pounds. Finally, early model converters contained just plati-num and palladium, but more recent “three-way converters” contain rhodium as well.

Before lots of spent catalysts actually arrive at a refiner for reprocessing, several separate firms—including automotive dismantles, collectors, decanners, and scrap processors—can be involved.Some of these firms specialize in collection of certain types of converters and do not purchaseother types. Some work under an arrangement with refiners. Others are independent.

Preliminary sorting is done by the automotive dismantle after the converter is removed fromthe car by torching or cutting or, in the case of larger shredding operations, popped off with aforklift. Some converter containers contain little or no catalyst. Spent catalysts commonly containat least some lead, which is particularly undesirable to refiners, and moisture, which can also af-fect chemical analyses and reprocess smelting. If the scrapped converter is sold to a collector, itmay be visually inspected by the collector for catalyst content. However, the dismantle may choosenot to sell the converter to a collector and may instead stockpile it, waiting for a price increase,or keep it for sale to retail customers.

The decanner (who maybe the collector or the the auto dismantle) removes the catalyst sub-strate from the stainless steel container. After it has been removed, the catalyst is placed in drumsto prevent further moisture accumulation, and is sold to a refiner. The stainless steel shell of theconverter may be sold to a scrap processor, some of whom also may serve as collectors or decanners.

Both the quality and quantity of the catalyst play major roles in determining the price paidto decanners by refiners. Some require a minimum shipment of 3,000 pounds of catalyst substrate.Bonuses may be paid for shipments that exceed average PGM content, while penalties may becharged for below-average shipments. A partial payment for the catalyst maybe paid upon receipt,with the balance reserved until actual assay, often 6 to 10 weeks later.

Once the spent catalysts are accepted by the refinery, several commercialized processes canbe used to recover the PGM. Two major domestic refiners—Gemini and Johnson Matthey--useproprietary hydrometallurgical processes to separate the substrate from the PGM through an acidor alkaline leach. Final removal of the PGM from the resulting leach liquor or sludge is complex,involving as many as 18 separate steps, but can recover more than 90 percent or more of the PGM.Scrapped catalysts can also be run through copper or nickel pyrometallurgical refining circuitsto recover about 90 percent of the PGM. This approach was taken by Inco, a large Canadian nickeland copper producer, in 1980 and 1981, when PGM prices were high, but has been subsequentlydiscontinued.

In addition to recycling PGM, interest is growing in the possibility of “regenerating” spentcatalysts. This would involve leaching out lead and other impurities and washing additional PGMsas required, while keeping the substrate and shell intact.


cars are simply abandoned; some are takendirectly to a scrap processor for baling orshredding, where the value of the PGM (andchromium in the converter shell) is lost; othersare exported. As a result, about 20 to 30 per-cent of all cars are never taken to a dismantle.

Even those cars that reach dismantling yardswill not necessarily have their converter re-moved. According to the ADRA, the majorityof automotive dismantling firms have less than10 employees. With a 1983-84 scrap price ofonly about $5 to $7 per converter, not all dis-mantles feel it is worth their employees’ timeto remove the converters, while others maybestockpiling the catalysts until prices rise. A1982 survey of some ADRA members and someunaffiliated dismantles by Arthur D. Little,Inc., found that 12 percent of those firms re-sponding to the survey leave the majority ofconverters on the car.32 However, ADRA mem-bers and other respondents may not be repre-sentative of the industry as a whole. Smallfirms that lack the resources to participate inthe trade association may be more likely toleave converters on cars.

Providing refineries with a steady supply ofscrapped converters has also been a problem.The auto dismantling industry is highly decen-tralized, having 11,000 separate firms, accord-ing to ADRA, while only a few large firms areinvolved in refining the catalyst. As more ac-tors have become involved, intermediary firmsspecializing in collection and decanning havearisen either independently or in affiliationwith refiners. Because refiners are trying toassemble economic volumes for recovery, apremium is paid for larger volumes of mate-rial. When a sufficient number of convertershas been collected by a dismantle they are soldto collectors and/or decanners. Usually a min-imum of 50 to 100 converters is required to dealwith a large collector.

Collectors accumulate small lots of wholeconverters within a particular region and selllarger lots at higher prices to decanners. Somecollectors are also decanners. Often, collectors

have regular truck routes to purchase con-verters from dismantles and, in some cases,from shredding yards.

Decanners purchase large lots of whole con-verters, and separate the catalyst’s substratefrom the housing or shell, usually by shearingthe canister in half. The pellets and honey-combs are packed separately in 55 gallondrums. Owing to the low nickel content andaccumulated impurities, the 409 stainless steelused in the converter shell has a low marketvalue ($1 10 per ton in late 1983). Nevertheless,the shell represents an additional source ofvalue to the decanner.

The quality of catalyst material sold to re-finers, as well as the volume, plays a major rolein determining the price paid to decanners.Decanners are cautioned not to include con-taminated units. Pure white catalyst substratesindicate burnout and loss of the PGM coating,Catalysts coated with lead and oil are also un-desirable in the refining process.

Auto dismantles would prefer to have morethan one purchaser (collector or core buyer) forused converters to ensure the best price andmost reliable pickup, while refiners prefer topurchase large quantities of catalysts from onecollector. During periods of low platinumprices, both dismantles and collectors maystockpile catalysts, waiting for prices to rise.In the long term, many of the problems appar-ent in the early stages of the industry’s evolu-tion may be overcome, due to the increasednumbers of scrapped cars equipped with con-verters available for recycling and possible in-creases in PGM prices.

USED AND RECONDITIONED CATALYTIC CONVERTERS

A recent development in the automotive dis-mantling industry has been the emergence ofa market for secondhand catalytic convertersbought as replacement parts. The exact mag-nitude of this market is not known, but a re-cent survey in California found that about80,000 secondhand converters were used an-nually in the State.33 Most of these secondhand

sZAutomotive Dismant]ers and Recyclers Association, profile:Automotive Dismantling Industry, published in cooperation withArthur D. Little, Inc. (Washington, DC: Automotive Dismantlesand Recyclers Association, 1983), p. 12,

ttMobile Source Control Division, Used Catalytic Converter Sur-vey (El Monte, CA: State of California Air Resources Board, July1983), memo.

Ch. 6—Conservation of Strategic Materials ● 2 4 3—

converters were obtained from parts distrib-utors who purchased them from auto salvagedealers at a somewhat higher price (averaging$7 per unit) than is typically paid by refiners.Auto dismantles also sell converters directlyto retail customers as used parts, often at aprice of $45 to $700

The apparently growing market for second-hand converters concerns some refiners sinceit adds an additional layer of competition intheir efforts to secure a supply of catalysts forrecycling. The competition is great because thismarket is an important source of income to dis-mantles, who typically obtain three-fourths oftheir income from sales of used parts, whilescrap sales form only a small part of theirbusiness.

An emerging issue associated with the appar-ently growing market for secondhand catalyticconverters concerns possible reuse of defectivecatalysts in cars. While secondhand convertersare substantially cheaper to the motorist thannew replacement converters, which range inprice from $170 to $320, many secondhandconverters are defective, and will therefore ad-versely affect air quality if they are installedin cars.

It is not possible to determine whether a usedconverter is operating properly without emis-sions inspection and testing. A gradual deterio-ration in catalyst efficiency occurs over time,and most automakers guarantee the converterfor only 50,000 miles. In addition, leaded gas,which destroys the emissions control capabil-ity and the platinum metal values in the cata-lysts, is often used in converter-equipped carsand trucks. Occasionally this happens inad-vertently, if fuel supplies in a filling station be-come contaminated or are switched. Moreoften, motorists deliberately and illegally mis-fuel because of lower price and perceived im-proved performance of leaded gasoline.

Currently, overall procedures for testing andcertifying secondhand catalytic converters arenot in place. Although the U.S. EnvironmentalProtection Agency has issued voluntary after-market part self-certification regulations, theserelate primarily to new replacement convert-

ers. However, methods to test used catalyticconverters are under development by a majormuffler firm, using the voluntary self-certifi-cation regulations. The entire issue associatedwith use of secondhand converters is highlycomplex, and is likely to be an increasingly im-portant concern in both emission control pro-grams and recycling efforts in the years tocomer

ALTERNATIVES TO PGM USE IN THE CATALYTIC CONVERTER

At present, PGM use in the catalytic con-verter is the most effective means for meetingemissions control standards under the FederalClean Air Act. Near-term (before 1995) alter-natives to PGM use in the catalytic converterare not promising, Most currently developedalternatives would entail major loss of fueleconomy and/or a need to relax automotiveemissions standards. These problems could beovercome, given several years lead time, butthere is no persuasive reason at this time forauto makers to pursue these alternatives.

Increased use of diesel engines could also re-duce the need for PGM catalysts. However,while diesels do not currently require con-verters to reduce carbon monoxide and hydro-carbons, they do produce substantial amountsof particulate. A more stringent particulate na-tional emission standard for diesel-fired light-duty vehicles (cars and light trucks) is expectedto take effect in 1987. Some manufacturers areexpected to employ a platinum-catalyzed trapoxidizer to meet the standard. The other mostlikely alternative–a trapping system whichuses onboard dispersing of fuel additives—would not entail PGM use. A particulate emis-sions standard for heavy duty vehicles, ifadopted as expected in 1990, most likely willrequire PGM catalyzed trap oxidizers, becauseadditives are unlikely to prove practical overthe long lifetime of these engines. (In the pro-jections shown in figure 6-5, it was assumedthat half of the light duty diesel vehicles willuse a PGM particulate trap when the standardstake effect—1986 in California and 1987 nation-wide—but that all California light trucks anddiesels would use particulate traps to meet themore stringent State standard expected in 1989.


It was also assumed that all heavy duty dieselengines will require PGM particulate traps tomeet the anticipated 1990 national standard.)

Even in the longer term (beyond 1995), PGMuse for emission control is likely to continue,although use of substitute materials, more ef-ficient loading of PGM, and basic changes inthe overall design of automobiles, might reduceper vehicle requirements. At present, the ma-jor auto makers foresee no major change intotal loadings of PGMs used in converters, al-though ratios of platinum, palladium, andrhodium may change, Alternative base metalcatalysts are not likely to substitute for PGMcatalysts, unless breakthroughs overcome ma-jor technical problems that make them less ef-fective in emissions control.

Downsizing of engines to meet federallymandated fuel efficiency standards, coupledwith improved catalyst formulations, may re-duce the amount of PGM needed to treat emis-sions effectively. With long-term increases ingasoline prices, downsizing may occur regard-less of whether Federal standards are extendedbeyond 1987. Alternative trends in downsizingto the year 2000 are discussed in detail in the1982 OTA assessment, Increased AutomobileFuel Efficiency and Synthetic Fuels.34

Considerable speculation exists about possi-ble future use of alternative automotive enginesthat are radically different from those usedtoday, such as the automotive gas turbine(AGT) and stirling engines. These engineswould have very low emissions levels, thus re-ducing emission treatment needs. As is dis-cussed in chapter 7, technical problems (in-cluding development of reliable ceramic parts)make it unlikely that the AGT will be widelyused until after the year 2000, if then. Alterna-tive fuels, such as methanol, are also underconsideration as a principal fuel for some au-tomobiles, Emissions from cars retrofitted toburn methanol do not meet current standardswithout catalytic treatment. Moreover, emis-sion tests on methanol-burning cars show rela-

aq~ncreased Automobile Fuel Efficiency and Synthetic Fuels(Washington, DC: U.S. Congress, Office of Technology Assess-ment, September 1982), p, 17.

tively higher levels of aldehyde emissions thanfor gasoline-powered cars. Aldehyde emissionscan be reduced by catalytic after treatment. Al-dehydes are not covered in current emissionstandards, so new standards could be requiredif methanol is widely used.

STRATEGIES FOR EMISSION CONTROL IN A SUPPLY DISRUPTION

PGM recovered from catalytic converterscan be reprocessed into virtually any formneeded by industry. In a major supply disrup-tion, automotive catalyst production could betemporarily curtailed, thus permitting PGMthat is recovered from scrapped catalytic con-verters to be diverted to more critical economicand military uses. PGM use in catalytic con-verters is critical to controlling air pollution,but is not otherwise a critical component of theU.S. economy. In a national emergency, a tem-porary discontinuance of PGM use in the man-ufacture of new catalytic converters, togetherwith intensified recycling of PGM from scrappedcars, would insulate the most critical economicuses from the effects of a supply disruption.

As a result, U.S. ability to contend with a sup-ply cut-off has theoretically improved, owingto reliance on PGMs for emission control, eventhough PGM imports have increased becauseof it. Use of such an emergency strategy, how-ever, would have major impacts on the emis-sions control program established under theFederal Clean Air Act.

Alternative strategies for auto makers andGovernment officials for dealing with a near-term curtailment of PGM supplies were ad-dressed in the report prepared for OTA bySierra Research and Energy and EnvironmentalAnalysis, Inc. This analysis demonstrates thatpractical alternatives to the catalytic converterare not currently available that could meet bothcurrent emissions and fuel economy standards.In a near-term supply disruption in which noPGM was available for emissions control, automakers would probably have little alternativeexcept to build cars without catalytic convert-ers—although this could be accompanied witha requirement for retrofit when the shortageended.

.—

Ch. 6—Conservation of Strategic Materials ● 2 4 5

In a supply disruption lasting 1 to 3 years,recalibration of new vehicles for lower emis-sions without catalyst use would be feasible.This could entail programming of fuel meter-ing systems for leaner operation to reducehydrocarbon and carbon monoxide emissions.In addition, retarding spark advance to furtherreduce HC and NOX emissions could be under-taken. These changes would not entail substan-tial design changes or development costs. Al-though some fuel economy penalties would beincurred, average new car emission levels ap-proximately equal to 1977-79 standards couldbe achieved in this way. Since computer con-trolled systems are now used in many new ve-hicles, retrofitting after the disruption wouldbe comparatively simple, entailing replacementof a computer chip or (more likely) the entireelectronic control unit of the vehicle.

With an intensive program of research anddevelopment (R&D), according to the analysis,alternative emission control systems that donot use PGM but would approach current emis-sions and fuel economy standards could prob-ably be developed. However, this would re-quire at least 5 years and is unlikely to beundertaken by auto makers simply out of con-cern about future supply disruptions.

Development of such technology could benecessary in the event of a “worst case” supplydisruption of several years’ duration in whichnational emergency requirements could not besatisfied from recycled PGM alone. Under suchdire circumstances, pressures would mount torecover PGM from operating vehicles.

The total current in-service vehicle fleet con-tains roughly twice as much PGM as was con-sumed for all uses domestically in 1982, Thiswill more than double by 1995—to a total of 9million troy ounces of PGM, some portion ofwhich could be recovered in a dire emergencythrough Government programs aimed at re-moving some catalytic converters from the in-service automotive fleet.

Ultimately, both the objectives of nationalpreparedness in the event of a supply emer-gency and national air quality objectives couldbest be served if contingency planning were

undertaken to determine under what circum-stances, and in what way, allocation of PGMwould occur in a supply emergency. Such plan-ning could also be used to identify emissioncontrol strategies available to auto makers inthe event that supplies of PGM are not avail-able for manufacture of catalytic converters,

Other Conservation Opportunities

Petroleum and Chemical Catalysts

Both PGMs and cobalt are extensively usedin the petroleum and chemical industries ascatalysts. PGM use in these catalysts is alreadyhighly efficient so that only limited improve-ments in current patterns of use can be madein the foreseeable future. The PGM containedin industrial catalysts is generally retained inownership by the user firm and is refined orregenerated on a toll basis by a specializedrecycling firm. This system is highly efficient:Of the 235,000 troy ounces of new PGMs thatwere purchased for petroleum and chemicalindustry catalyst applications in 1982, onlyabout 10 percent was to make up metal lossesduring processing and toll refining; the restwas to meet new production needs. The onlycatalysts that are not now recycled contain lessthan 0.05 percent PGMs. Recovery of PGMvalues from these catalysts is technically fea-sible, but the refining costs (including trans-portation costs to and from the refinery) exceedthe contained metal value.

In the case of cobalt-containing catalysts,however, major opportunities exist to increaserecycling levels, particularly in the case of thecobalt contained in spent petroleum refiningcatalysts. Although some other metals are nowreclaimed from these spent catalysts, cobaltvalues are not.

Overall national trends in cobalt catalyst use(exclusive of paint dryers) were extensivelyanalyzed for OTA by the Inco Research & De-velopment Center, Inc.,35 and are summarized

35F. j. Hennion, 1. J. deBarbadiIlo, j. H. Weber, and G. AlexMills, Use of Critica] Materials in Catalysts for the Petroleum andChemical Industries, Cobcdt and Platinum Group MetaLs (preparedfor OTA by Inco Research & Development Center, Inc.]. Muchof the information in this section is derived from this report.


in table 6-5. This information, obtained fromestimates provided by industrial sources in late1983, is believed to be the most up-to-date anddetailed information about these trends that ispublicly available.

As shown in table 6-5, three major processesdominate cobalt catalyst consumption: petro-leum hydroprocessing, used to remove con-taminants from sour and heavy crude oil; al-cohol production through hydroformulation;and production of terepthalic acids and deriva-tives, used to produce polyester fibers andfilms. Cobalt is also used in catalysts to pro-duce synthetic fibers, solvents, pharmaceuti-cals, and industrial chemicals.

Significant recycling of the cobalt used inchemical catalysts is occurring. According toInco, about 1.5 million pounds of cobalt wereconsumed in both chemical and petroleum cata-lysts in 1982. Of this, 1.1 million pounds—about72 percent—was contributed through recy-cling. Fresh cobalt needed to makeup the dif-ference was only about 426,000 pounds. Thisis less than some other estimates which mayunderestimate the extent of recycling by thechemical industry.

However, cobalt from spent petroleum hy-droprocessing catalysts is not now recycled.Since none is recovered, all of the cobalt con-sumed in hydroprocessing was fresh cobalt.Inco estimated that about 270,000 pounds ofcobalt was consumed in hydroprocessing in1982. This is less than some estimates for prioryears, due to lower refinery utilization. The

Table 6-5.—Cobalt Consumption andRecycling in Catalysts, 1982

Cobalt New cobaltconsumption for makeup

(thousand Percent (thousandProcess pounds) recycled pounds)

270Hydroformulation . . 700 90 70Terephthalic acid 500 90 50Organic acid . . . . . . 35 0 35Alkyl amines . . . . . . 30 95 1.5

Total . . . . . . . . . . . 1,535 72 426.5SOURCE” F. J Hennion, J. J. de Barbadillo, J. H. Weber, and G. Alex Mills, Use

of Critical Metals in Catalysts for the Petroleum and Chernica/ /n-dustries (Cobalt and Platinum Group Metals), 1983. Report preparedfor OTA by the Inco Research & Development Center, Inc., December1983.

Inco estimate is consistent with a NationalMaterials Advisory Board finding that between200,000 to 300,000 pounds of cobalt were con-sumed in this use in 1982.

Expanded use of hydroprocessing catalystsin the petroleum industry is likely in the yearsto come. However, growth in demand willprobably be less than was anticipated in themid and late 1970s, when refineries were rap-idly installing hydroprocessing capacity toprocess heavy source crude oils obtained fromnon-OPEC sources. The combined effects ofenergy conservation and the 1981-82 recessionhave resulted in scaling back projected require-ments. While hydroprocessing capacity can beexpected to grow appreciably from the de-pressed 1982 levels, “fresh” cobalt require-ments will not necessarily increase proportion-ately. According to the National MaterialsAdvisory Board, future cobalt make up require-ments could be in the range of 200,000 to400,000 pounds a year, assuming that effectiverecycling takes place. Substitution of nickel-molybdenum catalysts could reduce future re-quirements by another 100,000 pounds a year,according to NMAB.36

In current practice, hydroprocessing cata-lysts may be regenerated one or more times inthe course of their life cycle (2 to 3 years). Atsome point, however, the catalyst becomes socontaminated that subsequent regeneration isimpossible. These spent (unregenerated) cata-lysts may be stored temporarily on site byrefiners, landfilled, or sent to a processor forrecovery of non-cobalt metal values or for ex-port, The typical spent catalyst will containabout 8 percent molybdenum, 1.5 percent co-balt, 0.5 to 1.5 percent nickel, 1 to 6 percentvanadium, 10 to 15 percent carbon, 10 percentsulfur, 0,2 percent phosphorus, and the bal-ance, alumina.

During the late 1970s, when molybdenumand cobalt prices were very high, considerableeffort was made to recover metal values fromthese spent catalysts. However, most of theseefforts were discontinued as prices fell. By

Wobalt Conservation Through Technological Alternatives, op.cit., p. 119.


1982, Gulf Chemical & Metallurgical Co., ofFreeport, TX, was believed to be the only do-mestic firm recovering metal values from spenthydroprocessing catalysts. Gulf C&M, whichhas been in the catalyst recycling business forover two decades, uses a combination of pyro-and hydro-metallurgical techniques to recovermolybdenum and vanadium values from thespent catalysts but does not currently recovercobalt, nickel, or other metals.

From years of reprocessing, some 100,000tons of processing residues are reported to haveaccumulated at the Gulf C&M facility, As muchas 7 million to 8 million pounds of cobalt andnickel may be present in this residue, some por-tion of which could be recovered with adop-tion of adequate reprocessing technology. Thecompany has reportedly developed a pilot-scale

process to recover nickel, cobalt, and aluminavalues from this residue, but economic re-covery of these metal values at current pricesis not feasible. Because Gulf C&M reprocessesspent catalysts from both foreign and domes-tic sources, more cobalt residues are accumu-lated at the facility each year than are gener-ated from domestic hydroprocessing activitiesalone.

At least two other domestic firms–HallChemical of Wickliffe, OH, and CRI Metals ofBaltimore, MD—have announced plans to con-struct facilities to reclaim cobalt and other met-als from spent hydroprocessing catalysts. HallChemical, which currently produces 1 millionpounds of cobalt salts annually, is construct-ing a facility (at Arab, AL) said to be able toreclaim individual components of spent cata-

Molybdenum and vanadium are recovered from petroleum refinery catalysts at this plant in Texas. Several millionpounds of potentially recoverable nickel and cobalt are left in the waste and slag that is stored in

containment ponds surrounding the plant


lysts, including cobalt, nickel, molybdenum,vanadium, sulfur, and alumina. Hall intendsto purchase spent catalysts from generatorsaround the world.

CRI Metals, a unit of Catalyst Recovery Inc.,has announced plans to establish a hydroproc-essing catalyst reclamation facility in theSouth. The CRI process has been evaluatedthrough pilot plant trials, but CRI has yet tocommence construction, owing to depressedmetal prices. Like the Hall process, the CRIplan is to recover several contained metals—cobalt, nickel, vanadium and alumina—fromthe spent catalysts for reuse by the catalyst in-dustry. CRI apparently plans the recycling ven-ture to complement its current catalyst regen-eration activities.

Although the industrial capability to reclaimcobalt from spent catalysts may soon be inplace, prospects for improved recovery will de-pend on an increase in cobalt prices. Sizablequantities of spent catalysts may continue tobe landfilled, with cobalt values effectively lostin the interim. From the refinery’s point ofview the fastest method of disposal of spentcatalysts is preferred. At current prices, thisoften means paying the freight to ship spentcatalysts to reprocessors, storing the catalystat a temporary site off the premises, or land-filling. While the technology exists to reclaimlandfills, the high cost of such activities meansthat recovery of cobalt values from landfills isunlikely.

An important issue associated with bothrecycling or landfilling of spent hydroprocess-ing catalysts concerns hazardous waste dis-posal requirements. Although spent hydro-processing catalysts are not a listed hazardouswaste, some spent catalysts may contain traceamounts of lead, arsenic, and other compoundsthat are toxic. For hazardous wastes, refinersmust decide within 90 days about whether thewaste is to be landfilled in an approved site orrecycled. With current depressed metal prices,there is little economic incentive to recyclethese spent catalysts, so that many refiners maychoose to landfill them.

Stainless Steel and Steel Process Wastes

Chromium used in stainless steelmaking isobtained from ferrochromium and from scrap.The scrap charge includes home scrap gatheredwithin the plant during various stages of thesteelmaking process, prompt scrap obtainedfrom mills and fabricators, and obsolete scrapfrom postconsumer products and salvage. Withthe exception of steelmaking wastes, homescrap is generally recycled with a high degreeof efficiency. Stainless steel producers rou-tinely recycle most of the solid scrap generatedin making stainless steel and in forming pri-mary products, such as sheets, bars, and wires.An estimated 35 to 45 percent of stainless steelproduction is derived from such home scrap,but since the home scrap cycle is within themill, it is not considered to be a secondary sup-ply source.

Owing to adoption of the AOD process,which cut losses of chromium to slag by morethan half, purchased scrap is an increasinglyimportant source of chromium in stainless steelproducts. Prompt industrial scrap and obsoletescrap together form the purchased componentof scrap supplies. An estimated one-fourth ofthe chromium needed to make stainless steeland heat-resisting steels in the 1977-81 periodcame from purchased scrap.37 Chromium ispresent in other alloy steels and in cast ironscrap, but in low concentrations that do notfavor recycling for chromium content. Thesematerials are usually recycled for their ironcontent, and the value of the alloying metalsis lost (this practice is referred to as “down-grading”).

More chromium may be lost each year todowngrading or to failure to recover obsoleteproducts than is recovered from purchasedscrap. According to estimates prepared for theBureau of Mines,38 about 135,000 tons of stain-

STU,S, Department of Commerce, Critical Materials Require-ment of the U.S. Steel Industry (Washington, DC: U.S. Govern-ment Printing Office, 1983), p. 32.

Snchar]es L. Kusik, et a]., Availability of Critica] Scrap MetakContaining Chromium in the United States: Wrought StainlessSteel and Heat Resisting Alloys, U.S. Department of the Interior,Bureau of Mines Information Circular 8822 (Washington, DC:U.S. Government Printing Office, 1980).


less steel scrap containing about 24,000 tonsof chromium were included in the carbon steelscrap used in nonstainless steelmaking in 1977.Another 38,000 tons of chromium was lost inobsolete scrap that was not recovered. (Therange of uncertainty of these estimates is highowing to a variety of factors, including impre-cise export and import data and long lifetimesof stainless steel products,) Of course, not allof the stainless steel that was downgradedcould have been recovered, since it may havebeen present only in small amounts or havebeen inseparable from other materials, but thevolume of downgraded scrap makes it clearthat there are major opportunities to increasethe recovery of chromium from scrap.

Large quantities of scrap are generated whenstainless steel is made into fabricated products,but most firms collect and separate the betterquality scrap for sale either to scrap collectorsor to primary producers for recycling. Thecomparatively low value of chromium in promptindustrial scrap may slow the return of somescrap for recycling and lead to some down-grading. The nickel-chromium stainless steels(AISI 300 series) are usually recycled becauseof the high value of nickel and other alloyingagents, Recycling of the 400 series scrap, whichhas 12 to 30 percent chromium but little nickelor other alloying agents, is affected by the rela-tively low value of chromium.

Recovery of obsolete stainless steel scrap isimpeded by dispersed location, the uncertainquality of materials in the scrap, and the longlifetime (5 to 25 years or more) of stainlessproducts. Only a small proportion of stainlesssteel used in appliances, cutlery, and other con-sumer items is ever likely to be recovered fromscrap, although some may end up in munici-pal solid waste collection systems from whichmetals recovery is possible. Fuller recovery ofstainless steel scrap from used automobiles,machinery, and other industrial applicationstakes place, but seldom in excess of 60 percent.

Automobiles provide the single greatest op-portunity for reduced downgrading of obsoletechromium scrap. About one-third of the obso-lete stainless steel recycled in 1977 came from

scrapped automobiles, but this amounted toonly about 30 to 40 percent of the potentialquantities of chromium that could have beenrecovered from scrapped cars, Baling, whichis used for 10 to 20 percent of scrapped cars,prevents recovery of most stainless steel excepteasy-to-remove items, such as hubcaps, Shred-ding of vehicles permits better recovery sincenonmagnetic stainless steel can be separatedfrom the carbon steel scrap and sold separately,However, magnetic stainless steel, includingthe steel used for converter shells, is not eas-ily separated. This downgrading of stainlesssteel is not likely to be overcome because ofthe low value of the chromium and the highcost of identifying and separating it from othersteels.

Prospects for improved recovery of chro-mium from scrapped automobiles will be closelytied to efforts to recycle PGMs in catalytic con-verters, since the converter shells are made ofstainless steel, The average 1976 U.S.-built car(now approaching obsolescence) contained about4.8 pounds of chromium, of which 2.6 poundswas contained in the shell of the catalytic con-verter. Since 1975, chromium used in catalyticconverters has averaged about 10,000 tons peryear. Assuming that 70 percent of the cars thatannually become obsolete reach dismantlingyards, 7,000 tons of chromium could theoreti-cally be recovered from converter shells alone.Since additional losses are likely, recovery of5,000 to 6,000 tons per year is probably themaximum practical level of recycling. Mostcars now manufactured have monolithic-typeconverters (containing about 1.5 pounds ofchromium), so that per vehicle chromium re-covery rates in the mid-1990s will decline com-pared to today. At the same time, the numberof cars scrapped each year may grow, hence,the overall amount of chromium that is poten-tially available for recycling will increase.

The value of chromium in the converter islow, however, and the recovery of the shell forrecycling will depend largely on the value ofthe PGMs contained in the catalyst itself. Mostdismantlers–even those who remove the con-verter to collect PGM catalysts—do not segre-

38-844 0 - 85 - 9 : QL 3


gate the shell for recycling as stainless steelscrap. Instead, the shell is shredded or baledwith the rest of the car.

While small in relation to overall chromiumdemand, appreciable quantities of chromiumas well as other critical metals are also lost influe dusts, slags, sludges, and other steelmak-ing and chemical wastes each year. Thesewastes are not suitable for direct recycling be-cause of contaminants such as lead or zinc, butthe chromium and other metals from some ofthese wastes are increasingly recovered throughspecial processing,

According to Bureau of Mines researchers,about 17,400 tons of chromium were containedin metallurgical wastes generated in 1974 (themost recent year which comprehensive datawere collected).39 Until recently, these low-grade wastes were almost entirely unrecov-ered. However, some firms are now commer-cially recovering chromium containing wastes,and several steel producers have establishedtheir own in-house waste recovery programs.A partial impetus for this has been environ-mental regulations restricting the disposal ofsolid wastes.

Since 1978, Inmetco, a subsidiary of Inco,has been recovering chromium and nickel fromlow-grade steelmaking wastes. This facility hasthe capacity to process 47,000 tons of waste peryear to produce 25,000 tons of pig metal con-taining 4,500 tons of chromium and 2,000 tonsof nickel. The pig metal, which is 18 percentchromium and 8 percent nickel, is suitable foruse in AISI-type 304 stainless steel or, if it con-tains molybdenum, AISI 316. Inmetco current-ly processes about 75 to 80 percent of the fluedust, mill scale, and grinding swarf of U.S.stainless steelmaker. The company estimatesthat 3,100 tons of chromium was contained inwastes entering their plant in 1982, of which2,800 tons were recovered. The Inmetco facil-ity also has the capacity to recover chromium

WW. M, Dressel, L. C. George, and M. M. Fine, “Chromiumand Nickel Wastes: A Survey and Approval of Recycling Tech-nologies, ” Proceedings of the Fijlh Mineral Waste Symposium,Eugene Aleskin (cd.), cosponsored by the U.S. Bureau of Minesand the I IT Research Institute (Chicago, IL: I IT Research In-stitute, 1976), p. 269.

and other materials from catalysts, superalloys,and other products. An Inmetco subsidiary,Pittsburgh Pacific Processing Corp., reproc-esses oil-contaminated scrap.

Economic processes for recovery of chro-mium and other metals from specialty steel-making slags have yet to be developed, Esti-mates made in the 1970s of nationwide lossesof chromium to slag in stainless steelmakingvary widely—from 3,000 to 30,000 tons peryear. The lower figure is from the previouslycited 1974 estimate of total chromium lossesin metallurgical wastes, and appears to be quitelow. The latter figure, based on 1965-75 data,is from a study by the National Materials Advi-sory Board, which also projected that losseswould be reduced to about 10,000 to 15,000

——Ch. 6—Conservation of Strategic Materials . 251

tons per year as stainless steelmaker switchentirely to the AOD process,

Revised estimates do not appear to have beenmade since the NMAB study, which was pub-lished in 1978, but the AOD process is nowused almost universally. Nonetheless, the lossesremain appreciable. Chromium content insteelmaking slags analyzed by the Bureau ofMines Rolla (MO) Research Center range be-tween 1 and 12 percent, with most samples inthe range of 6 to 7 percent. Experiments be-ing conducted at the Center are aimed at up-grading chromium and nickel from crushedand ground slag. Concentrations of more than20 percent chromium have been achieved. Fu-ture research will assess ways to recycle theconcentrates in steelmaking furnaces.40

Over 3,000 tons of chromium were also lostannually in processing wastes from the chem-ical industry, metal plating and etching shops,and leather tanning firms, according to the1974 data mentioned previously. Recovery ofthe chromium in such wastes is expensive, anduntil recently wastes were generally disposedof in municipal wastewater systems or land-filled without significant preprocessing. Envi-ronmental requirements under Subtitle C of theResource Conservation and Recovery Act(RCRA) and the Federal Clean Water Act haveresulted in the need to process such wastesprior to disposal. As a result, the opportunityto recover chromium from the waste has arisen.

Many metal finishing wastes are classifiedas hazardous wastes by the Environmental Pro-tection Agency and are subject to the disposalrequirements of RCRA. Pollutants, includingchromium, must be reduced to acceptablelevels before disposal in Subtitle C landfills.Disposal fees at landfills may run on the orderof $25 to $50 per drum, which does not includecosts of transportation and preprocessing.41 Re-covery of metal values may be incorporated

‘As discussed in Bureau of Mines Research 83, op. cit., p, 104.41 ~’ ,s. ~nvironmenta] protection Agency, Environ mcmtal Po~-

Iution Control Alternatives: Sludge Handling, Dewatering and Dis-posa~ Alternatives for the Metal Finishing Industry (Washington,DC: U.S. Government Printing Office, 1982), p, 1, EPA 62515-82-018.

into waste processing, and the value of the met-als can then help reduce the cost of disposal.

Several processes have been developed or areunder development by public agencies and in-dustry to recover or recycle metals from proc-essing wastes. Corning Glass reportedly has de-veloped a closed-loop process which permitsreuse of chromic acid etching solutions, thusreducing the amount of waste and reducing theneed for new chromium-bearing chemicals.Another process to recover chromic acid etch-ing solutions, developed by the Bureau ofMines, has been demonstrated to be economi-cally competitive with the purchase of newacid, when savings of disposal costs are in-cluded in the analysis. This process, which ex-tends the useful life of the acid a hundred fold,results in major materials savings and also per-mits the recovery of other metals, such as cop-per, which can be sold.42

Electronic Scrap

The key constraint in recycling of PGMsfrom electronic scrap is the very small amountof metal used in each unit. The two largest con-sumers of PGMs in the electrical and electronicsector are the telecommunications industry,which uses them in electromechanical con-tacts, and the ceramic capacitor industry. Intelephone relays, palladium and platinum arefound in the contact points, which are brazedto base-metal alloy arms. In a single hand-dismantled relay module taken from a largeunit, the Bureau of Mines found 18 large and36 small PGM contacts, weighing a little morethan 0.06 of 1 percent of the weight of thewhole module.43 The National Association ofRecycling Industries estimates that 1,000 tonsof telephone relay scrap are generated eachyear. However, this scrap contains only a mi-nute proportion of PGMs.

aZDeborah A. spotts, Economic Evaluation af a i$fethod to Re-generate \$’aste Chromic Acid-Sulfuric Acid Etchants, U.S, De-partment of Interior, Bureau of Mines Information Circu]ar 8931(Washington, DC: US, Government Printing Office, 1983].

atFre~ Ambrose and B. Wr. Dumay, Jr,, ‘‘Mechanical process-ing of Electronic Scrap to Recover Precious Metal Bearing Con-centrations, ” paper presented at the International Precious Met-als Institute Conference, Vicksburg, VA, May 1983.


Another potential barrier to recycling ofPGMs from electronic scrap is the fact that pal-ladium (which, with an average 1982 dealerprice of $100 per troy ounce is much lower inprice than platinum or gold) predominates inthis sector. Palladium is used in ceramic capac-itors and is increasingly substituted for goldin low-voltage, low-current electrical contacts,such as telephone relay contacts. Higher reli-ability electronics, including those for militaryuse, require greater amounts of platinum andgold in contacts, capacitors, and other elec-tronic components.

Despite these potential barriers, there is agrowing electronic scrap recycling industry inthe United States. An industry observer44 esti-mates that 102,000 troy ounces of platinum, ormore than the total yearly platinum purchasesby the industry, were recovered from elec-tronic scrap in 1982. The growth in recyclingof this type of scrap is reflected in the locationof the secondary precious metals industry.Twenty years ago, most precious metals re-finers and semi-refiners were clustered aroundthe jewelry industry on the east coast. Today,many semi-refiners and refiners are located inCalifornia’s “Silicon Valley,” Texas, and otherareas where the electronics industry has de-veloped.

As in other sectors, PGM recovery is great-est from manufacturing scrap. Large refinerssuch as Johnson Matthey deal in both primarymetals and scrap, supplying the electroniccomponent industry with palladium metal,powders, and pastes, and recycling the metalfrom their reject components. In addition,some large electronics companies have set uptheir own in-house recycling processes torecapture the precious metals from both theirmanufacturing scrap and obsolete scrap. How-ever, the percentage of PGMs recycled fromobsolete electronic scrap is generally lower. Be-cause of the high labor costs involved in manu-ally processing electronic scrap prior to finalrefining, some electronics firms and scrapdealers find that foreign dealers offer the best

44Mark Stringfel]ow, “Platinum: Its Present and Future, ” pa-per presented to the National Association of Recycling Indus-tries, Dec. 15, 1982.

price for their scrap. Material is shipped bothto Europe, where such firms as SGM in Bel-gium have a long history of expertise and so-phisticated technology for precious metal refin-ing, and to Taiwan and other Asian countries,where lack of raw materials and low labor costsmake low-grade scrap more attractive.

Industry assessments of the quantity of elec-tronic scrap exported by the United States vary.Total sales of PGMs to the electrical and elec-tronic sector were 438,000 troy ounces in 1982.By comparison, exports of all types of PGM-bearing scrap that year were 388,437 troyounces. 45

Future trends in domestic recovery of PGMsfrom electronics are difficult to predict. Overthe next decade, the scrap market is expectedto increase as technological advances makeequipment obsolete. Computer manufacturers,concerned about possible resale of their obso-lete components, may crush the parts and sellthem for recycling. However, whether the plat-inum group metals found in this scrap are re-covered for U.S. consumption depends on a va-riety of factors. While some of the materialsmay be exported for overseas refining, this maybe done on a toll-refined basis. For example,

A5u.s. Department of Interior, Bureau of Mines, Mineral in-dustry Surveys, Platinum Quarterly, June 7, 1983, table 5,

Photo credit’ GTE Corp

Precious metal scrap produced during the manufactureof electronic components is segregated to enhance

recycling of gold, silver, and platinum group metals


a recently opened Engelhard plant in Cinder-ford, England, will refine metal from scrappedcomponents and circuitry from U.S. sourceson a toll basis, providing an equivalent amountof PGMs to those sources from Engelhard’sU.S. refineries.46 In addition, a number of Euro-pean firms, such as De Gussa, a German firm,have begun to operate PGM recycling plantsin the United States. It is too early to tell wheth-er these trends will have a major impact on theamount of scrap that is exported.

As in other platinum-consuming industries,the long-term potential for recycling of strate-gic materials from electronics will depend onthe quantities of materials used in production,The rapid evolution of the electronics indus-try has led to continual improvements in effi-ciency of use of the precious metals. For ex-ample, palladium-silver alloys are increasinglyreplacing pure palladium in electrical contacts,dampening both palladium use and potentialfuture levels of palladium recycling. In addi-tion, composites show promise for replacingcontacts made entirely of PGMs.47 A fiber com-posite made of a copper matrix with embed-ded palladium fibers and an overall palladiumcontent of 30 percent has the same electricaland thermal conductivity as a palladium-copper alloy having a composition of 60 to 85percent palladium. One of the greatest impactson PGM use will come from the Bell System’spresent conversion to solid-state switching,thus reducing electromechanical switchingand platinum and palladium demand.

Despite these trends in increased efficiencyof use, opposing trends are causing overall de-mand for PGMs in electrical and electronicequipment to increase. Because of its low cost,palladium is increasingly being substituted forgold in electrical contacts and printed circuits,and new types of components, such as ceramiccapacitors, are making use of PGMs. In addi-tion, ruthenium, which is already used in thelarge quantities of reed switches used in elec-tronics, is finding new applications in the in-

—— — —.4eJudith Crown, “Engelhard Revamping Precious Scrap Set-

ups, ” American Metals Market, July 8, 1983, p. 1.4TStockel, “composites for Electrical Contact Applications, ”

Zeitschrift fuer Werkstofj’echnik, July 1979.

dustry. Overall, it appears that consumption ofpalladium and ruthenium will increase, andthat there will be large quantities of thesematerials available for recycling beyond 1990.

Technologies for recycling PGMs from elec-tronic scrap require further development.While several hydrometallurgical and pyro-metallurgical processes are available for finalrefining of PGMs, preprocessing technology isnot yet adequate. Currently, there are numer-ous steps in the refining process, often carriedout by different companies, with each step add-ing to recycling costs. The initial manual dis-assembly of the most valuable componentsfrom the aluminum and copper frames (e.g.,snipping off contacts with wire cutters) maybe done either by the electronics company orby a specialized preprocessing firm. Thesefirms sell some components for reuse andothers for scrap to both domestic and foreignbusinesses.

Semi-refiners of electronic scrap collect cir-cuit board scrap after initial disassembly andprocess it further. This may involve simplystripping off the plastics with chemicals, fol-lowed by shredding and chopping the compo-nents, or more complex steps such as ball millcrushing and pulverizing of capacitors, whichleave a “sweep” or pulp containing preciousmetals. Other firms combine these processingsteps with smelting out the PGMs as crude bul-lion, which may be either sold to electroniccomponent manufacturers or shipped to re-finers. Final refining to high specifications isgenerally done by one of a handful of special-ized refiners using pyrometallurgical tech-niques.

Using obsolete electronic scrap provided bythe Defense Property Disposal Service, the Bu-reau of Mines has developed a process for me-chanically upgrading electronic scrap, Theprocess is made up of a series of unit opera-tions using technology already widely availablein the recycling industry. These operations in-clude shredding, eddy current separation, andhigh-tension separation. The process can breakdown mixed scrap into an iron-base fraction,an aluminum base fraction, a wire fraction,and a high and a low precious metal concen-

254 . Strategic Materials: Technologies to Reduce U.S. Import Vulnerability

trate, greatly reducing the amount of materialthat must be processed for precious metal re-covery and also reducing shipping and tollrefining costs.

The Bureau of Mines process could reducecosts of refining electronic scrap by replacingcostly, slow, and unpredictable hand strippingwith reliable automated techniques. While itcan be used on mixed scrap, it is most effec-tive when incoming scrap is first separated intolots. For example, connector assembly plating,printed circuit cards, wiring, and other com-ponents high in precious metals should still bemanually separated from their outer box cov-ers. While all segments of the precious metalsrecycling industry have expressed interest inthe process, economic factors may preventwidespread adoption of fully automated proc-essing. Some in the recycling industry view thehigh capital costs of the equipment and thehigh energy costs of operating it as formida-ble barriers to profitable domestic processingof low-grade mixed scrap.

Cemented Carbides

About 8 percent of the cobalt consumed do-mestically each year is used as a binder incemented carbides, which are used in cuttingtools or other high-wear applications. Mostscrap produced during fabrication of cementedcarbides is now recovered, and in the last fewyears, producers of cemented carbides have be-gun to recycle substantial quantities of obso-lete cemented carbide scrap. This improved re-covery became pronounced in the late 1970sbut the improvement was less a function of in-dustry response to the cobalt price increasethan of the ongoing adoption of technologicaladvances by tungsten carbide producers.

According to the NMAB, up to 0.5 millionpounds of cobalt from obsolete products arenow recovered by tungsten carbide producerseach year. In early 1984, installed capacity toreprocess such scrap is estimated to be 36 per-cent of domestic consumption. Obstacles to in-creased recycling of obsolete carbide scrapinclude widespread geographic diffusion ofproducts and difficulties in recovering wornor broken parts,

Tungsten carbide scraps constitute a majorproportion of current tungsten recycling. GTECorp., the largest domestic producer of tung-sten, established a patented process in the early1970s for recycling tungsten scraps and in1978, a proprietary process for producingextra-fine, highly purified cobalt powder fromrecycled scrap on a toll and nontoll basis. Withsintering, the powder can be turned into ametallurgical cobalt material suitable for usein superalloy, GTE has also established a pi-lot plant project to reclaim cobalt from super-alloy scrap. The recycled cobalt is equivalentto highly refined cobalt produced from virginores. Although the cobalt is produced as powderfor cemented carbide use, it may be suitablefor superalloy if compacted and pelletized.48

Other cemented carbide producers, such asTeledyne, employ a zinc process to recoverboth cobalt and tungsten carbide as powdersfrom tungsten carbide scrap. The process wasdeveloped by a British firm in 1946 and fur-thered by the U.S. Bureau of Mines,

qBInformation provided to OTA by GTE, Aug. 10, 1983.

Prospects for Conservation of Strategic Materials

Table 6-6 summarizes key opportunities for quantities of materials are involved. Underincreased recycling of chromium, cobalt, and tight supply conditions, appreciable increasesPGMs. Data limitations prevent close estima- in recycling can be anticipated as scrap col-tion of prospective recovery opportunities in lectors, processors, and consumers respond tomany applications, but it is clear that very large increased prices of primary metals.

Table 6-6.—Selected Opportunities for Increased Recycling of Chromium, Cobalt, and Platinum Group Metals

Current levelof recycling

Superalloy (Co, Cr):An estimated 2.8 million Ibs of

cobalt-bearing superalloy proc-essing scrap was lost or down-graded in 1980; 1.2 million Ibsof cobalt in obsolete scrap wasnot recovered. However, thesefigures are based on 1976 scrapuse rates.

Key recoveryopportunities

Increased recovery of obsoletescrap, reduced downgradingof high-quality industrial scrapand recovery of wastes, usingadvanced recycling tech-nologies.

Petroleum hydroprocessing catalysts (Co):No cobalt recovery at this-time;

270,000 Ibs of cobalt was notrecovered from spent catalystsin 1982.

Cemented carbides (Co):10 to 30°/0 of cobalt used in do-

mestic cemented carbide ship-ments came from recycling.

Cobalt consumption in catalystsis expected to grow to 675,000Ibs in 1990, 900/. of whichcould be recovered.

Accumulated spent catalystresidues may include 8 millionIbs of cobalt and nickel.

Obsolete scrap (a high propor-tion of industrial scrap is al-ready recovered.)

Barriers to increased recycling

Technical Economic

Concern about contaminantslimits recycling in superalloyproduction to high-qualityscrap and processes that havebeen certified.

Experimental processes to re-claim elements separatelyhave not been commercialized.

Not significant; various proprie-tary processes are purportedto recover Cr, Co, Ni, Mo, Va,and other metals for catalyticor chemical purposes.

Some technical problems; pilot-scale recovery process is un-der investigation by a privatefirm.

Not significant: several firmsuse a Bureau of Mines proc-ess to recover cobalt binders.One firm produces highly re-fined, pure cobalt powdersfrom scrap that could poten-tially be used in” otherpremium uses.

Current prices encourage down-grading of superalloy scrapfor use in stainless steel ornickel alloy production inwhich the cobalt is notneeded.

Certifying recycled materials forsuperalloy use is expensiveand time-consuming.

Current prices may discouragerecovery of cobalt in prefer-ence for molybdenum. How-ever, at least two firms haveproprietary processes for re-covery of all elements incatalysts.

Reprocessing of residue is notprofitable

Recovery ofdispersedeconomic

at current prices.

obsolete scrap fromuses may not beat current prices.

Institutional

Not significant once manufactur-ing specifications for use ofrecycled materials have beenestablished; however, this cantake several years.

Landfilling of spent catalysts byrefineries prevents possiblefuture recovery of somecatalysts.

Not significant; residues arestored on site of majorprocessor.

Lack of effective scrap collec-tion programs by industrialusers. Practical limits forpostconsumer recovery varyby industry. However, someusers (such as the oil indus-try) return 80% of their scrapfor recycling; the coal miningindustry, by contrast, onlyreturns 150/0 (out of a practi-cal limit of 500/0).

Table 6.6.–Selected Opportunities for Increased Recycling of Chromium, Cobalt, and Platinum Group Metals (Continued)

Current level Key recovery Barriers to increased recyclingof recycling opportunities Technical Economic Institutional

Stainless steal (Cr):Home scrap is efficiently recov-

ered; purchased scrap for stain-less steel accounted for about20°/0 of total demand for Cr in1983.

Catalytic converters (PGM, Cr):Up to 30°/0 of scrapped converters

may now be recovered forrecycling.

Electronic scrap (PGM):High for industrial scrap, low for

obsolete scrap.

Improved recovery, reduceddowngrading of obsoletescrap, especially automotivecatalytic converters.

Increased recycling of industrialprocessing wastes (slags, fluedust, low-quality fabricationscrap, etc.)

Retrieval of PGM from scrappedautomobiles, muffler shops,etc. Maximum theoreticalPGM recovery will grow to800,000 troy oz by 1995. Maxi-mum practical recovery ratewill probably be about 500,000troy oz.

Obsolete scrap.

Not significant, but current in- Comparatively low price of chro-dustrial practices encourage mium and labor costs asso-downgrading. ciated with removal of stain-

less steel components fromautomotive and other obsoletescrap encourages down-grading.

Recovery of chromium from Resource recovery from indus-some wastes (e.g., slags) istechnically difficult.

trial wastes may not be eco-nomic: however. commercialrecovery of low-grade fluedusts, grindings, turnings, bor-ings is now being undertakencommercially.

Several companies have proprie- Dependent on PGM prices, buttary processes to recycle appears favorable.catalysts.

Lack of technology for Depends on PGM prices; handpreprocessing of scrap to a labor disassembly encouragesform usable by a refinery. export of scrap.

Recovery of obsolete consumerproducts containing stainlesssteel catalytic converter shellsis a primary candidate for re-covery, but will depend oneffective linkage of PGM andCr recovery efforts.

Environment and solid wastedisposal requirements are pro-viding increased impetus torecovery where economic.

Some dismantles do not re-move catalytic converter fromcars prior to scrapping.

Collection networks have notbeen established in manyareas.

Dismantles want more for thecatalysts than reclaimers are

(i.e., households) is unlikelywithout incentives.

Value of scrap is not widely rec-ognized; monitoring of trendsis inadequate.

willing to pay.

Collection of obsolete scrapfrom consumer electronics

SOURCE: Office of Technology Assessment.


Evidence from past “crises’ ’-most notablythe Canadian nickel strike of 1969 and the co-balt price spike in 1978—suggests that U.S. in-dustry can quickly increase recycling whenconcerns about supplies are paramount orwhen prices rise rapidly. In 1969, nickel scrappurchases exceeded 1968 levels by over 60 per-cent despite the fact that the Canadian strikedid not occur until mid-1969. Nickel scrap asa proportion of domestic consumption ex-panded rapidly over the following 3 years—accounting for 22 percent of domestic suppliesin 1972 as compared to 8 percent in 1968, (Thecurrent level is about 23 percent,) Cobalt scrappurchases doubled in 1978 and remained at orabove 1 million pounds per year through 1981,a growth from 4 percent of domestic consump-tion to 8 percent.

It is more difficult to quantify savings of stra-tegic materials likely to arise from adoption ofnew manufacturing and processing technol-ogies by U.S. industry, but they are also large—particularly in the cases of manganese used insteelmaking and strategic materials used insuperalloy. Upgrading of basic steelmakingprocesses could lead to important reductionsin the amount of manganese needed to produceeach ton of steel (see table 6-6). Use of advancedmanufacturing technologies by the aerospaceindustry will reduce chromium, cobalt, andother strategic material input requirements forsuperalloys on a perpart basis. Part life exten-sion programs could also conserve materialsover the long run. As discussed in chapter 7,increased use of coatings and other surfacetreatment technologies and use of advanced ce-ramic and composite materials also could re-duce per part strategic material requirements,especially in the long term.

Recycling Technology

With some notable exceptions, there appearto be few significant technological barriers toincreased recycling of strategic materials.High-quality scrap generated in the productionand manufacture of products is recognized tobe a valuable source of raw materials by indus-try and is usually effectively recovered and re-used. Ongoing industrial trends, including

establishment of specialized scrap processingfirms and improved industrial waste treatmentprocesses, have expanded the range of wastematerials considered suitable for recycling, sothat continued incremental improvement inrecycling is likely in the coming decade.Difficulties in identifying and sorting complexscrap containing several materials have re-cently been eased by the commercial availabil-ity of portable and relatively inexpensive in-struments suitable for use in scrapyards, asdiscussed in box 6-A.

In the area of superalloy recycling, consid-erable technological advances have been madeover the last decade. However, additional in-novations may be needed if full use of obso-lete scrap and other less preferred scrap is tooccur. In the most demanding applications,even small trace amounts of impurities may beunacceptable. Because of concern that con-taminants acquired from scrap material couldresult in failure of superalloy parts, use ofsuperalloy scrap in jet engines has been limitedto scrap of the highest quality—generally, homeand prompt industrial scrap, although limiteduse has been made of obsolete scrap obtainedfrom commercial aircraft, where parts can beidentified by alloy type. Much of the remain-ing scrap is downgraded to use in stainlesssteel, where it is used for its nickel and chro-mium content. When this occurs, cobalt is lost,and the previously high-quality chromium usedin superalloys becomes a replacement for chro-mium usually supplied from high-carbon fer-rochrome or stainless steel scrap.

Nontechnical Impediments

In most applications, institutional and eco-nomic constraints on recycling are more for-midable than the technical constraints. Fluc-tuating commodity prices, costs associatedwith segregation and processing of scrap,transportation costs, and industry structure allaffect the ability of recycled materials to com-pete with primary metals.

ECONOMICS OF RECYCLING AND CONSERVATION

The economic incentive for recycling high-priced materials, such as PGMs, is very high.


However, other strategic metals are less expen-sive and often account for only a very smallproportion of the scrap. For many applications,raw materials comprise only a small fractionof the total value of the product. A 6,000-poundjet engine, for example, may cost $3.5 million,while the value of the materials it contains isonly $60,000. Even though an alloy may con-tain metals that are relatively rare and highpriced, the cost of separation of the alloy intoits constituents may be so high that it discour-ages any attempt to recover the metals; so, thealloy is downgraded to recover its nickel andchromium content.

Recycling opportunities will change alongwith changes in manufacturing processes andproduct designs. More efficient manufacturingprocesses will require less raw material, sothere will be less prompt industrial scrap avail-able for recycling, New product designs thatuse combinations of materials to achieve desir-able properties will present difficult and expen-sive problems for the separation of materialsfor recycling. Often, products can be designedfor ease of recycling, but since raw materialcosts are generally a small component of over-all cost, this is seldom done,

Materials conservation is an important sidebenefit of the advanced manufacturing tech-nologies widely used by the aerospace indus-try. These technologies may gradually filterdown to other industrial users. Diffusion ofcomputer-aided design and nondestructiveevaluation techniques may be impeded to someextent by lack of experience of engineers, de-signers, and others with these new approaches.This is not a major problem in the aerospaceindustry or in other critical uses, however.

INFRASTRUCTURE

The complex structure of recycling indus-tries, often entailing primary metals proces-sors, collectors and dealers, scrap processors,and parts manufacturers, can inhibit initiationof recycling programs, even when the pro-grams may be economically advantageous.Generally, the secondary metals industry com-prises a large number of small firms that re-spond to prices offered by the primary metals

industry, This system works well for materialsfor which there is a relatively constant demandfor a standard product. Although recyclingfirms can respond rapidly to changes in mate-rials demand due to shortages of primary ma-terials, exacting standards for superalloy andother premium materials can only be metthrough the use of sophisticated equipment andexperienced personnel who are familiar withthe needs of consumers. This can delay recy-cling response time in a shortage situation,

Recycling and Conservation Data Base

Accurate information is needed to have arealistic picture of the potential for recyclingto reduce U.S. materials import vulnerability.The existing data are weak, with many uncer-tainties about their accuracy. Key areas whereimprovements can be made are: upgrading ofannual information about scrap purchases byconsuming industries, development and use ofmaterials flow models on a periodic basis, andimproved tracking of the ultimate dispositionof obsolete scrap and industrial wastes thatcontain appreciable amounts of strategic ma-terials,

The U.S. Bureau of Mines prepares annualestimates of scrap purchases by domestic in-dustry. In most instances, these estimates arebased on voluntary reporting by private firms.Not all firms are canvassed, and of those thatare, not all respond. Consumers of more thanone strategic material may be surveyed aboutonly one material. Although adjustments canbe made to compensate for incomplete report-ing, the resulting estimates are subject to con-siderable uncertainty. In the course of thisstudy, OTA found evidence that substantiallymore recycling may be taking place than is cur-rently reported in Bureau of Mines’ data.49

Scrap purchases alone provide only a limitedpicture of recycling trends, Substantial recy-cling occurs internally and through toll refin-ing arrangements that do not involve purchases,

goother reports have reached a similar conclusion. See, for ex-ample, the NMAB report, Cobalt Conservation Through Tech-nological Alternatives, op. cit., p. 25.


Also, scrap purchases provide no insight intothe ultimate fate of scrap that is not recovered.Periodically, Federal agencies such as the Bu-reau of Mines and the NMAB have publisheddetailed and relatively comprehensive mate-rials flow models for individual metals. Thesemodels have appreciably increased under-standing of overall recycling trends, but theyhave been prepared infrequently and there hasbeen little or no consistency in procedures usedin developing the models.

A notable impediment—from the perspectiveof identifying possible responses to disruptionsin materials supply—is the absence of informa-tion about accumulations of unrecovered ob-solete scrap containing strategic materials.Only a portion of the scrap that becomes ob-solete each year is recycled domestically. Some

is exported, while the rest is unaccounted forin national statistics. Much of this lost mate-rial may now be effectively unrecoverableowing to disposal in landfills, dissipative cor-rosion, contamination, or abandonment inunknown locations. Some, however, may beretrievable from industrial disposal sites, com-mercial scrap yards, and military storagedepots for use during supply disruptions. Ex-tensive efforts have been made over the yearsto estimate the accumulated “inventory” of po-tentially recoverable ferrous scrap, but a simi-lar effort to monitor the disposition of strate-gic materials in obsolete superalloy parts,catalysts, and other applications has not beenmade. As a result, plausible estimates of theabove-ground mine of strategic materials re-coverable in an emergency are not available.

chapter 6 conservation of strategic materials

Documents