2001-01-3726

Life Cycle Assessment of Aluminum Casting Processes

Robert D. Stephens and Candace S. WheelerChemical and Environmental Sciences Laboratory, GM Research & Development and Planning

Maria PryorChemical Risk Management/Pollution Prevention, Worldwide Facilities Group

Copyright © 2001 Society of Automotive Engineers, Inc.

AbstractIn recent years, the environmental

impact of automotive products and processeshas become an issue of increasingcompetitive importance. Life cycle analysis(LCA) provides a tool that allows companiesto assess and compare the environmentalimpact of a variety of material and processchoices. This enables companies tomanufacture environmentally sound productsof exceptional value by environmentallyconscious processes. In this study, we usedLCA to compare the environmental burdensassociated with three aluminum castingprocesses: lost foam, semi-permanent mold,and precision sand. We obtained data fromone primary and one secondary facility foreach of the three processes studied. Thesedata included all of the environmental burdensassociated with raw material and energyconsumption, gaseous emissions, and wastegeneration. In addition, we modeled theenvironmental burdens associated with theproduction and transport of the materialsused during the manufacturing processes.Both capital and operating environmentalcosts were also considered. In general, theenvironmental burdens associated with thelost foam and semi-permanent moldprocesses were very similar while theburdens associated with the precision sandprocess were higher for each of theparameters studied. Overall, lost foam wasdetermined to be the most environmentallyfriendly way to cast aluminum heads andblocks.

IntroductionIn recent years, we have seen a

growing awareness and concern among thegeneral public about environmental issues.The automotive industry has been affected bythis increased awareness of environmentalissues. Industrial pollution and automotiveemissions have become subject to increasedgovernmental control. Now, automotivemanufacturers must consider not onlyimproving fuel economy and reducingemissions, but must place increasing effortinto making the vehicles more recyclable. Inaddition, not only the products but theprocesses used to make those productsneed to be environmentally friendly. This isaccomplished by reducing waste, consumingfewer natural resources, and reducing airpollutants released during the manufacturingprocess. To do this, the automotive industrymust have a way of measuring theenvironmental impact of both products andprocesses. Life Cycle Analysis (LCA) is onesuch tool. It allows engineers to assess theenvironmental impact that a material orprocess change will have on the overallenvironment. Life cycle assessment involvesthe analysis of the environmental impactsassociated with a process or product at everystage of its life cycle from the extraction ofraw materials from the earth to the finaldisposal.

Casting processes and castingtechnology are critical in the automotiveindustry. Several different casting processesare currently being used in the manufacture ofaluminum cylinder heads and blocks. This

has raised questions regarding the relativeenvironmental impacts of these technologies.To address these concerns, an LCA wasconducted on lost foam, semi-permanentmold (SPM) and precision sand casting ofaluminum engine heads and blocks. Thescope of the LCA included raw materialconsumption, energy consumption, CO2

emissions, air emissions (hydrocarbons,carbon monoxide, nitrogen oxides, sulfuroxides and particulate matter), solid wastesand environmental costs.

ExperimentalThis study has considered the

environmental burdens associated with themanufacturing of cast aluminum productsproduced by three different castingprocesses: lost foam, precision sand, andSPM. The products produced by theseprocesses include heads and blocks madevia lost foam and precision sand, and headsproduced by SPM. This study consideredthe following life-cycle inventory (LCI)categories:• Energy and raw material consumption• Air emissions of CO 2, CO, NMHC, CH4,

NOx, SOx, N2O, and PM (particulatematter)

• Solid waste generation, both land-filledand recycled

These burden categories were considered forthe production and transport of all rawmaterials used, as well as the burdens of themanufacturing phase. Use phase andproduct end-of-life were not considered. Alldata have been reported on a normalizedbasis, with the basis being burden(emissions, energy consumption, etc.) per1000 kg of degated “to-spec” product.

Two facilities for each of the threeprocesses were visited, with detailed dataacquired from one primary facility using eachprocess. A second facility using eachprocess was visited to determine that theprimary facility was typical for that process.We also collected production quantities,scrap rates, and part masses to enable thesedata to be reported on a normalized basis, asdefined above. All data were based on the1996 production year. Table I shows theparts produced, their masses – both beforeand after degating, the quantities produced,and the overall mass of aluminum poured andpresent as final degated product.

Table I. Production Data From Primary Facilities

Process ComponentMass Poured

(kg)Degated

Mass (kg) QuantityMass Poured

(kg)Degated

Mass (kg)Lost foam 2V Head 12.88 10.18 134,577 1,733,653 1,370,441Lost foam 4V Head 15.88 13.09 184,450 2,928,328 2,413,779Lost foam Block 29.98 21.27 316,465 9,488,557 6,732,426SPM 3.3 L head 14.68 9.89 400,000 5,871,398 3,957,751SPM 3.5L head 26.99 13.46 395,000 10,662,526 5,317,467SPM 2V head 26.39 18.12 536,000 14,142,849 9,713,514SPM 4V head 29.26 20.00 136,000 3,978,979 2,719,771SPM Quad 2 head 26.51 11.76 35,000 927,952 411,442SPM Quad 4 head 26.54 11.81 290,000 7,697,955 3,424,900SPM 1.5L 4V head 25.33 15.94 52,000 1,317,109 828,713Precision 1.25L head 19.20 11.25 248,100 4,763,520 2,791,125Precision 1.4L head 19.25 11.30 39,200 754,600 442,960Precision 1.25L block 30.55 16.15 247,400 7,558,070 3,995,510Precision 1.4L block 30.45 16.10 37,400 1,138,830 602,140Precision LVPP block 27.45 17.75 11,600 318,420 205,900

The lost foam process data were collectedfrom Saturn in Spring Hill, TN (primary) andthe GM foundry in Massina, NY (secondary).

The companies providing SPM and precisionsand data are confidential.

Of the three processes studied,speciated hydrocarbon emissionmeasurements were only available at Saturn.Most air emissions data from SPM werebased upon AP42 (2) estimates. Theprecision sand facility providedmeasurements on air emissions, but thesedid not include speciated hydrocarbons.

Within the material production stage, wehave considered the LCI categories for all materialsconsumed during the manufacturing stage as wellas the acquisition of fossil fuels and the generationof electricity used during the manufacturing stage.Table II shows the breakdown of factorsconsidered within each of the two life cycle stagesstudied.

A number of assumptions have beenused in determining the LCI. These

assumptions are listed in Table III. Forexample, all three casting facilities use 100%secondary aluminum to produce theirproducts. We have calculated the inventoryfor the collection and the melting of scrapaluminum only. No burdens for theproduction of primary aluminum have beencalculated. For electricity, we have used USaverage values for efficiency and energysources (hydro, coal, natural gas, etc.).

The lost foam and SPM facilitieswere located near the aluminum suppliers,which supplied aluminum in the molten state.The precision sand facility was not locatednear an aluminum supplier, hence theiraluminum was delivered in ingot

Table II. Assignment of Factors within Life Cycle StagesLife Cycle Stage Factors Included

Materials Production Stage • The burdens associated with the production of all materialsconsumed during the manufacturing stage.

• The burdens associated with the generation of all electricity usedduring both material production and manufacturing stages.

• The burdens associated with drilling and delivery of natural gasused during the manufacturing stage.

• The transport of all materials to the manufacturing facilities.• The collection and melting of scrap secondary aluminum.

Manufacturing Stage • All air emissions generated directly by the manufacturingfacilities.

• All solid waste generated by the manufacturing facilities, bothlandfilled and recycled.

• Energy and raw material consumed by the manufacturingfacilities.

• The transport of landfilled wastes from the manufacturingfacility.

form. As a result, the precision sand facilityutilized additional energy to remelt thealuminum. In our analysis, we have assumedthat all facilities received aluminum in themolten state. To accomplish this, we havesubtracted from the precision sand facility’stotal energy consumption the amount ofenergy required to remelt the aluminumingots. This was done to provide a processdependent comparison, independent of thelocation of the casting facility relative toaluminum suppliers.

Except for plant specific data, mostLCI data were acquired from Boustead (1),except where otherwise noted. Air emissionsdata from the combustion of natural gas bythe manufacturing facilities were estimated byAP42 (2).

The transport of all materials to thecasting facilities, as well as the transport of alllandfilled wastes, has also been modeled.Where possible, specific data on truck size, loadcapacity, and frequency of deliveries have beenacquired and used in the transport model. Wehave assumed that all of our casting facilities

were located at fixed distances from suppliers soas to not preferentially reward or penalizemanufacturers for the location of their facilities.The burdens associated with transport wereacquired from Boustead.

Environmental costs associated withthe three casting processes were also

determined. Environmental costs werecategorized as capital and operational costsassociated with collecting, abating, anddisposing of waste and/or emissions. Thesecosts will be defined for each specificprocess within the Results section.

Table III. Major Assumptions Used in the LCIMaterial or Process AssumptionsElectricity We have used US average data for generation

efficiency and energy sources.Aluminum We have assumed that all aluminum is 100%

secondary and that it is provided to all three castingfacilities in a molten state.

Transportation of Manufacturing Materials andLandfilled Wastes

We have assumed a “generic” plant with supplierslocated at fixed distances from each facility. Thesedistances were modeled after the actual distances forthe lost foam and SPM facilities.

LOST FOAM The process flow diagram forthe production of aluminum castings via lostfoam is shown in Figure 1. In this diagram,the steps for generating an aluminum lostfoam casting move from left to right. Eachbox represents a process or material that isnecessary for the process. The flow diagramshown can be simplified and explained in thefollowing major steps:

1) Cluster making: A series of polystyrenemolds are produced and glued together toform a “cluster” that has the desiredshape of the final net part. The cluster isthen coated with a clay-like material andbaked in an oven to dry the coating.

2) Sand compaction: The coated cluster isplaced into a “flask”, and sand is addedand compacted around the cluster.

3) Melting/pouring: Molten aluminum ispoured into the cluster, burning away thepolystyrene and leaving the final shape ofthe part.

4) Cleaning: The casting is removed fromthe flask, and the casting is mechanicallyshaken to remove the sand from withinand around the casting. Subsequently,the casting is cleaned using plasticbeads as shot blast to remove soot andresidual sand from the casting.

5) Degating: Gates and risers, which arechannels for directing the molten metalinto the cluster, are removed from thecasting to produce a “degated” product.

6) Heat Treatment: The castings areheated to strengthen the metal.

All of the environmental burdensassociated with these steps were acquireddirectly from Saturn Corporation. Thematerials required for each of these processsteps, as well as the burdens associatedwith these steps, are listed in Table IV anddiscussed in detail below.

As shown in Figure 1, the processsteps leading to cluster making include theexpansion of polystyrene (EPS) beads.These beads originally contain 7% (byweight) pentane. After expansion, thesebeads contain approximately half thisamount. Therefore, pentane is a majorcompound emitted in the process. Zincstearate is added to the beads as a lubricantto prevent clumping during bead expansion.Polystyrene beads are then blown into moldsand further expanded to form polystyrenemolds. Several molds are glued together toform “clusters”. The clusters are then coatedwith a refractory material and oven dried.Finished clusters are inserted into flasks and

sand is compacted around the cluster.Plastic sandwich bags are used to cover thesprues during compaction to prevent sandfrom being entrapped within the casting.

Waste beads and scrap uncoated clusterswere, at the time of this study, sent to a

PolystyreneBead

ExpansionMold Making Cluster Making and

CoatingSand Fill

CompactionPouring Shakeout and

QuenchDegating and

CleaningHeat Treat

Lost Foam Process - Inputs/Outputs

Polyurea Production

Process Flow

Waste Polystyrene

AluminumScrap

Sand Fines/Shot Blast dust

Washer Sludge

HydrocarbonEmissions

AluminumOxides

Used FilterMedia

WasteGlue/Waste

Coating

Pentane Emissions

Water/Steam

Compressed Air

ParticulateEmissions

NaturalGas

Zinc StearateProduction

Glue Production

SecondaryAluminumProduction

Argon/Flux

Production

SandProduction

Coating/Baggies

Production

RefractoryMaterial

Production

PolystyreneProduction

Electricity

Waste Zn Stearate

Refractory Waste

Figure 1. Lost foam process flow diagram showing major inputs and outputs.

recycler. Coated clusters, waste glue, andwaste cluster coating are sent to a landfill.

Silica sand is used for compactionaround the clusters. After use in the castingprocess, the sand is contaminated withpolystyrene combustion products. Withoutreclamation, this sand is unfit for re-use. Toreclaim the sand for re-use, 10% of the usedsand is incinerated to remove the contaminants.This reclaimed sand is combined withunreclaimed used sand and size segregated.Sand fines are removed and delivered to arecycler for use in road construction. Theremainder of the sand is reused in the castingprocess. Newly purchased sand is used toreplenish the waste sand fines. The newly castheads and blocks are cleaned using plastic shotblast media to remove soot and residual sandfrom the castings. For blocks, this plastic ispolymethylmethacrylate (PMMA), and for heads,it is polyurea. Due to abrasion during use, theseplastic beads disintegrate into a dust that is nolonger useable as a shot blast medium. Thisplastic dust is returned to the supplier to berecycled.

Aluminum is supplied to Saturn in amolten state. Saturn maintains this aluminumin a molten state and also remelts their owninternal aluminum scrap. Argon gas is usedduring the process to degas the moltenaluminum and to minimize oxidation. Fluxesand alloys are also added to the moltenaluminum. Aluminum chips generated duringthe removal of gates, risers, and sprues arereturned to an aluminum recycler. The gates,risers, and sprues are remelted at Saturn.The degated blocks and heads continue on tobe machined. However, the machiningprocesses are not included in this analysis.

In addition to the major sources ofemissions listed for the manufacturing phasein Table IV, air emissions from the combustion of natural gas at the lost foam casting facility and the emissions associated with the production of electricity used by the

Table IV. Lost Foam Process Steps and Major Environmental BurdensProcess Step Material and Energy Inputs Waste and emissionsCluster Making • Expandable polystyrene (EPS)

• Glue• Zinc stearate• Steam (water and energy)• Compressed air (energy)• Coating (consisting of muscovite,

kaolin, mica, quartz, Attapulgite,pyrophyllite, acrylic acidhomopolymer, and water)

• Energy for drying of coated clusters• Bags (low density polyethylene)

• Pentane emissions from EPSexpansion

• Waste polystyrene - beads, molds,and uncoated clusters (recycled)

• Coated clusters (landfilled)• Waste glue (landfilled)• Waste coating (landfilled)• Bags (low density polyethylene)

Sand Compactionand Reclamation

• New sand• Reclaimed sand• Bags (low density polyethylene)• Energy for sand reprocessing

• Sand fines (recycled)• Air emissions

Melting/Pouring • Molten aluminum• Flux/additives to aluminum• Inert gases (argon)• Energy for melting and holding

aluminum and incineration of airemissions

• Refractory materials

• Air emissions from EPScombustion

• Dross (recycled)• Scrap aluminum (recycled)• Waste refractory materials

(landfilled)• Inert gas

Shakeout andCleaning

• Plastic shot blast media (ureaformaldehyde and acryllic)

• Energy for shot blast process• Water

• Shot blast dust (recycled)• Washer sludge - shot blast dust

and sand fines (recycled)• Hydrocarbon emissions

Degating • Energy for removing gates and risers • Aluminum scrap (recycled)

facilities are also included in our results. Themajor areas of natural gas consumption forlost foam casting are: steam generation,ovens used for the processing of used sand,reactive thermal oxidizers (RTO) used forincineration of hydrocarbon emissionsgenerated during the casting process, themelting/holding furnaces for aluminum, andspace heating. Air emissions are abated bybaghouses and incinerators. Baghouse filtersare landfilled.

The environmental burdens forproducing most of the raw materialsassociated with lost foam casting wereacquired from Boustead (1). However,burdens for many materials were not availableand had to be simulated using surrogatematerials. Table V shows the assumptionsused to approximate the burdens for theproduction of materials where data were notavailable.

Environmental cost data have also been acquired for both capitalinvestments and annual operating costs. AtSaturn, capital costs were collected forbaghouses, incinerators for abatinghydrocarbon air emissions, and incineratorsfor processing used sand, as well as theductwork and fans specifically associatedwith each of these. Capital costs wereallocated equally over seven years to providean annualized capital cost. Operating costsincluded the natural gas used for theincinerators, the cost of filter cartridges forbaghouses, the cost of electricity for fansassociated with the baghouses andincinerators, and the cost of generalmaintenance of baghouses and incinerators.Operating costs also included the costsassociated with landfilling of solid wastes. All manpower costs were excluded from consideration.

Table V. Material Production Burdens for Lost FoamMaterial Approximations Ref.

Expandable polystyrene None 1Zinc stearate Sum of burdens for zinc chloride, stearic acid and 0.5 MJ

electricity to mix1,3

Cluster coating Mining, milling, and grinding of muscoviteand 0.5 MJ/kg of electricity to mix 1,3

Low density polyethylene(baggies)

None 1

Glue Sum of burdens for viscose, polypropylene,butene, and 0.5 MJ electricity to mix

1,3

Secondary aluminum Collection of aluminum cans and energy for melting andholding molten aluminum

1

Argon Twice the burdens of producing liquid oxygen 1,3Sand None 1Polyurea (shotblast) Sum of viscose, formaldehyde, urea, and 0.5 MJ/kg of

electricity to mix1,3

Polymethylmethacrylate(shotblast)

None 1

Natural gas delivery None 1Natural gas combustion None 1Electricity US Average 1Flux Omitted

SPM SPM is a casting process that utilizessand and resin to make sand cores. Thesecores are assembled into core packs thatrepresent the internal cavities of the castingand are placed in permanent steel dies whichare used to determine the outside structure ofthe casting. This process is shownschematically in Figure 2 and consists of thefollowing major process steps:

1) Coremaking: In this operation, sandand resin are mixed, formed into thecorrect shape, and cured using acatalyzing gas. The cores are thenassembled and placed in a metal mold.

2) Melting and Pouring: Molten aluminumis poured into the mold filling the spacesleft by the sand cores.

3) Degating: The gates and risers used inpouring the metal are removed from therest of the casting.

4) Shakeout: The casting is removed fromthe die and mechanically shaken toremove the internal sand cores.

5) Heat Treatment: The castings areheated to 500° C to strengthen themetal.

The materials used and the majorenvironmental burdens associated with eachof these steps are shown in Table VI anddiscussed in detail below.

During coremaking, a ratio of 5%virgin and 95% reclaimed sand along withBorden Sigma Cure Resin Part I and Part IIare mixed in a sand dryer and mixer. Thesand/resin mixture is poured into a core boxwhere it is formed and cured withtriethylamine (TEA) gas. Several cores areassembled into corepacks using smallamounts of glue and transferred to a carouselstation where they are placed inside a metalmold that has been pre-treated with a releaseagent. The major air emissions for thisoperation include: TEA, MDI (methyl-di-isocyanate), VOC, and PM emissions. Themajor solid and liquid waste include: spentscrubber solution, sand filter

Core Making Heat Treat

Semi-Permanent Mold Process - Inputs/Outputs

TEA Production

Core SandMixing

Core Drying/Assembly

Resins (Part I &II)Production

Process Flow

Metal Mold PreparationPouring and Solidifying Degating Shakeout

Sand Production

SecondaryAluminumProduction

Strontium/Flux

Production

Alloying Agents/Nitrogen

Production

TEAEmissions

Waste Resin

WasteSand

Particulate Emissions

HydrocarbonEmissions

AluminumOxides

ScrapAluminum

SandFines

Sulfuric Acid Production

Mold Release /Glue

Production

Water Production

Natural GasProduction

ElectricityProduction

Waste Sulfuric Acid

WasteFilterMedia

WasteCoating/

Glue

WasteRefractory

Figure 2. Process flow diagram showing major inputs and outputs to SPM process.

media, sand fines, and waste sand. Theexhausted TEA emissions are controlledusing a wet scrubber that captures the TEAand neutralizes it with sulfuric acid. Thescrubber solution is reclaimed by an outsidecontractor. The particulate emissions arecontrolled using dust collectors. The MDIand VOC’s are uncontrolled. The sand finesand sand filter media are landfilled. Thewaste sand is sent out for beneficial reuse.Defective cores and contaminated sand arereclaimed on site and are not consideredwastes.

Molten aluminum is received andmixed with strontium and flux and held innitrogen to prevent oxidation of the aluminum.The molten aluminum is then poured into themetal mold and corepack. In-house scrap ismelted in a natural-gas-fired melting furnaceand returned to a holding furnace.

The air emissions from the melting andpouring operations include NOx, CO,CO2, VOC’s, SO2 and particulate matter.Controls on air emissions consist only ofbaghouses for the collection of particulatematter. Solid wastes include dross withsand, dross, flashing, and refractorymaterial. The dross, dross with sand,and flashing are recycled offsite. Scrapaluminum is recycled onsite and is notconsidered waste. Refractory material islandfilled. The cast part is cooled andthe sand is mechanically removed via ashakeout operation. The major outputsinclude: NOx, CO, VOC’s, SO2,particulate emissions, sand fines, andsand filter media. These emissions areuncontrolled except for the particulateswhich are captured in a baghouse. Thesand fines and sand filter media arelandfilled. The parts are heat treated

Process Step Material and Energy Inputs Waste and EmissionsCoremaking • Sand

• Resin I• Resin II• TEA• Mold release agent• Glue• Sulfuric acid• Energy• Water• Coating

• Waste sand (recycled)• Sand fines (landfilled)• Filter media (landfilled)• Waste resin and glue (landfilled)• Waste sulfuric acid (recycled)• TEA emissions• Air emissions

Melting/Pouring • Aluminum• Strontium• Flux• Nitrogen• Release agent• Alloying agents• Energy

• Air emissions• Dross (recycled)• Dross with sand (recycled)• Flashing (recycled)• Refractory material (landfilled)

Degating • Energy • Aluminum scrap (recycled)Shakeout • Energy • Air emissionsHeat Treatment • Energy

• Water• Air emissions

in furnaces at 500o

C. The expectedemissions from this operation include: NOx,CO, VOC’s, SO2, and particulate. Theseemissions are all uncontrolled.

The environmental burdens for thematerials used in the SPM process wereacquired directly from databases or by usingsurrogate materials. These materials, orsubstitute materials, and the databases usedto supply the data are listed in Table VII.Sand, glue, secondary aluminum, flux,natural gas delivery and combustion, andelectricity generation and distribution weremodeled as listed for lost foam. Theenvironmental costs for SPM included thecapital and operating costs of baghouses, thesand reclaim furnace, and the aminescrubber. The SPM facility did not haveincinerators for hydrocarbon emissionabatement. Operating costs include theenergy to operate the abatement equipmentas well as the costs associated withlandfilling solid wastes. All manpower costswere excluded from consideration.

PRECISION SAND Precision sand castingis referred to as an all core process. It is acasting process that utilizes sand and resinto make sand cores for both the internalcores as well as the external surfaces of thecasting. These cores are assembled intocomplete core packs that determine both theinternal cavities as well as the outsidestructure of the casting. For aluminumcasting, the precision sand process consistsprimarily of the following steps:

1) Coremaking: In this operation, sandand resin are mixed, formed into thecorrect shape, and cured using acatalyzing gas. The cores are thenassembled to form complete core packs.

2) Melting and Pouring: Molten aluminumis poured into the mold filling the spacesleft by the sand cores.

3) Heat Treatment/Shakeout: Thecastings are heated to high temperatureswhich serve both to strengthen the metaland to loosen the sand from the castings.The castings

Table VI. SPM Process Steps: Inputs and Outputs

Table VII. Material Production Burdens for SPMMaterial Approximation Ref.

Phenol formaldehyde resin Sum of phenol and formaldehyde and 0.5 MJ/kgelectricity to mix 1,3

Isocyanate resin Sum of methyldiisocyanate and polyol and0.475 MJ/kg electricity, 0.1 MJ/kg kerosine, and 1.584 MJ/kgnatural gas to process

1,3

Core coating Mining and processing of mica and koalin plus 0.5 MJ/kgelectricity to mix

1,3

Carbon dioxide None 1Nitrogen None 1Sulfuric acid None 1Hydrochloric acid None 1TEA Simulated as aniline 1

are vibrated to assist in the removal of thesand from the castings. In the process, thesand is reclaimed for future use.4) Degating: The gates and risers used inpouring the metal are removed from the restof the casting.

These operations are outlined inFigure 3 and described in more detail below.Table VIII lists the major materials andenvironmental burdens associated with eachof these steps.

During coremaking, a ratio of 2%virgin and 98% reclaimed sand is mixed withAshland Isocure Resin Part I and Part II andpoured into a core box machine where thecores are formed and cured withdimethylethylamine (DMEA) gas. The coresare assembled into a core pack using a smallamount of glue. In some cases, blocks arefitted with steel cylinder liners. These linersare subjected to a shotblast process prior toinsertion into the core assembly. The corepacks are then transferred to the pouringoperation.

The major air emissions fromcoremaking include DMEA, MDI, phenol,formaldehyde, hydrocarbons, and particulateemissions. Waste streams include spentscrubber solution, defective cores, sand filtermedia, sand fines, iron dust, and waste sand.The exhausted DMEA emissions arecontrolled using a wet scrubber that capturesand neutralizes the DMEA gas with sulfuricacid. The scrubber solution is reclaimedoffsite. The particulate emissions arecontrolled using dust collectors. During the

core assembly process, a cross draft of air isused to capture some of the residual DMEA,MDI, phenol, formaldehyde, andhydrocarbons being emitted from the cores.This air is routed into an incinerator. Thesand fines and sand filter media are landfilled.The waste sand and defective cores arerecycled.

Aluminum is received in ingots andmelted in a furnace for use in the melting andpouring operations. The molten aluminum ismixed with alloying agents, strontium andflux. An inert gas, argon, is used to preventoxidation of metal exposed to air. The moltenaluminum is poured into the assembledcorepacks.

Precision sand is the only process ofthe three studied that receives aluminum inan ingot form. The SPM and lost foamfacilities receive aluminum in a molten state.To maintain a consistent comparisonbetween the three processes, we haveperformed an analysis that assumes that theprecision sand facility also receives aluminumin the molten state. To accomplish this, wehave subtracted the energy necessary to meltthe aluminum (except the internal scrap) fromthe manufacturing energy consumed.

The outputs from the pouringoperation include: NOx, CO, CO2,hydrocarbons, and particulate emissions.The air emissions of hydrocarbons andparticulate matter are abated via incineratorsand baghouses, respectively. Other wastesinclude refractory material, sand/Al mix, Al dust, flashing, and dross. The dross, dross with sand, and flashing are recycled offsite.

Core Making Core Assembly/DryingPouring andSolidifying

Heat Treat/ Shakeout Degating

Precision Sand Casting Process - Inputs/Outputs

Core SandMixing

Cleaning

Resin (Part I & II)Production

Process Flow

DMEA Production

Sand Production

Nitrogen/Argon

Production

RefractoryProduction

Steel ShotProduction

Waste Sand/Sand

Fines

DMEA Emissions

Waste Resin

AluminumOxides

WasteGlue

HydrocarbonEmissions

ParticulateEmissions

Scrap Aluminum

Waste Sand/Steel Shot mixture

Mold Release/Glue

Production

WaterProduction

Strontium/Flux

Production

SecondaryAluminumProduction

WasteFilterMedia

Sulfuric Acid Production

Natural GasElectricityProduction

Sulfuric AcidEmissions

WasteRefractoryMaterial

Figure 3. Precision sand process flow diagram showing major inputs and outputs.

For the most part, scrap aluminum isrecycled onsite and not consideredwaste.Scrap blocks, which already have steelliners installed, are recycled offsite.

The heat treatment operation servesthree purposes: heat treatment, core removal,and sand reclamation. The expectedemissions from this operation include NOx,CO, hydrocarbons, and particulate matter.The air emissions from this operation are alsoabated via incinerators and baghouses.

The environmental burdens for theproduction of sand, glue, argon, andsecondary aluminum, as well as the deliveryand combustion of natural gas, and thegeneration and distribution of electricity weremodeled as listed for lost foam. Core resinsand nitrogen were modeled as listed for SPM.DMEA burdens were acquired from Boustead(1) using aniline as a surrogate. The burdensfor sodium hydroxide were acquired directlyfrom Boustead (1).

Hence, direct comparisons ofenvironmental cost data are not valid.

Environmental costs for precision sand included the capital andoperating costs associated with incinerators,baghouses, filtration equipment for sand finesat the heat treat and sand silo areas, and theamine scrubber. It should be noted that sandreclaim is done simultaneous to heat treat,hence these costs are associated with dualfunctions in the process. Capital costs wereallocated over seven years to determineannual capital costs. Operating costsincluded the cost of energy to run the airabatement equipment, the costs of baghousefilters, and landfill costs. The precision sandfacility operating costs were provided asestimates and (unlike data for SPM and lostfoam) include the costs associated with laborassociated with maintaining the abatementequipment.

Table VIII. Precision Sand Casting Process Steps: Inputs and OutputsProcess Step Material and Energy Inputs Waste and EmissionsCoremaking • Sand

• Resin I• Resin II• DMEA• Μοld release agent• Glue• Sulfuric acid• Energy• Water

• Waste Sand (recycled)• Sand fines (landfilled)• Filter media (landfilled)• Waste sulfuric acid (recycled)• DMEA emissions• Air emissions

Melting/Pouring • Aluminum• Strontium• Flux• Nitrogen• Argon• Refractory materials• Energy

• Air emissions• Dross (recycled)• Dross with sand (recycled)• Flashing (recycled)• Refractory material (landfilled)

Heat Treatment andShakeout

• Energy • Air emissions

Cleaning and Degating • Steel shotblast media• Energy

• Shotblast dust (recycled)• Aluminum scrap (internally

recycled)

ResultsThe results will be presented in five

burden categories: 1) material consumption,2) energy consumption, 3) air emissions, 4)solid waste, both recycled and landfilled, and5) environmental costs. Environmental costshave only been determined for themanufacturing stage. These results will bepresented in terms of burden (consumption,emissions, costs, etc.) per 1000 kg ofdegated product.

MATERIAL CONSUMPTION Thematerials consumed by each castingprocess, and their quantities, are listed inTable IX. In addition to the burdensassociated with mining or producing thesematerials, we have included within thematerial production stage, the burdensassociated with transporting these materialsto the casting facilities. The transport ofliquid and solid wastes from themanufacturing facility to the landfill have beenincluded within the manufacturing stage.Transportation of waste materials to recyclershas not been included.

The major material consumptionissues surrounding aluminum castingprocesses are related to the use of sand andaluminum. The results for consumption ofthese materials are shown in Figure 4. Mostother materials are used in lesser quantitiesand are not common between the threeprocesses. The fossil fuel consumption dueto producing hydrocarbon-based materialsshows up in the energy consumptioncategory of Table IX. The remaining non-hydrocarbon based compounds utilized ineach of the processes is different and uniqueto each process.

The use of aluminum is comparablebetween each casting process and variesonly by the degree that chips and scrap arenot internally remelted. The numbersreported in Table IX, represent the quantityof aluminum required to produce 1000 kg ofdegated product. Much of the wastealuminum generated during the process isinternally remelted and used at each facility.Hence internal scrap eventually becomesfinished product and does not contribute tothe values reported in Table IX. However,

Table IX. Materials and Energy Consumed (per 1000 kg of degated product)a

Material Lost Foam SPM PrecisionExpandable Polystyrene 15.3Cluster Coating 64.3Plastic Shot (PMMA) 15.0Plastic Shot (polyurea) 34.7Plastic Baggies 0.09Zinc Stearate 0.03Liquid Argon 4.7 2.37Glue 0.32 0.03 0.08Natural Gas (MJ) 15,983 22,814 24,983Electricity (MJ) 6,838 5,679 12,425Molten Aluminum 1110.3 1019.0 1112.1Sand 235.4 381.3 976.5Hydrochloric Acid 0.005Core Coating 0.37CO2 21.84Nitrogen 22.82 55.1Phenolic Resin 9.19 51.38Isocyanate 9.24 51.51TEA/DMEA 4.01 10.08Sulfuric Acid 1.63 25.31Sodium Hydroxide 1.62Steel Shot 8.34

a All values are in kg, except energy, which is in MJ.

since considerable energy is required toremelt aluminum, this internal scrapcontributes to energy consumption. It isuseful, therefore, to compare the amount ofaluminum that must be poured to generate agiven mass of finished product. Thesequantities for lost foam, SPM, and precisionsand are 1431, 1913, and 1765, respectively,per thousand kilograms of degated product.The differences between these masses andthose listed in Table IX are the masses ofinternally remelted scrap, e.g. sprues, gates,and risers, which contribute to the inefficientuse of molten aluminum. These masses, inconjunction with other process differencescontribute to the difference in energyconsumption by each type of castingprocess.

Significant differences exist in theconsumption of sand by the three castingprocesses. The least amount of sand isconsumed by lost foam, followed by SPM,then precision sand. These differences resultfrom how sand is used and handled in theprocesses. The numbers listed for sandconsumption in Table IX represent only the

sand required to replace the waste sandgenerated. Each facility studied reclaimsused sand thereby reducing the consumptionof new sand. Sand consumption is dictatednot only by the total amount of sand neededper casting, but by the rate at which sandbecomes no longer re-useable. Processdifferences dictate each of these. Lost foamuses sand (without resin) to compact aroundpolystyrene clusters. SPM and precision useresin-based binders to form sand cores.Also, lost foam reclaims only 10% of thesand that is re-used during each castingcycle, whereas SPM and precision reclaim100% of used sand. Hence, sand used in theSPM and precision sand processes isexposed to conditions that probably increasethe rate at which degradation occurs. As thesand degrades to smaller particle sizes, itbecomes un-useable. These “sand fines” areusually either landfilled or used in roadconstruction projects.

SPM clearly uses less sand thanprecision sand, due to the mass of coresrequired for each process. This is a directconsequence of SPM’s use of a permanent

mold rather than a sand core to determineexternal casting dimensions. Several othermaterials are also used by both SPM andprecision sand. For all materials utilized byboth processes, precision sand has thelargest consumption. The consumption ofmore raw materials by the precision sandprocess means that energy consumption inthe production of these raw materials is alsohigher for precision sand.

ENERGY CONSUMPTION As seen inTable IX, material production for lost foamrequires the least amount of energy andprecision sand the most. The majorcontributor to energy consumption in thematerial production stage of all three castingprocesses is due to the generation of theelectricity used by each facility. Anothermajor contributor to energy consumption inthe material production stage is associatedwith melting aluminum. The pie charts,Figures 5, 6, and 7, show the relative

Raw Mate r i a l Consumpt ion

Aluminum Sand Total RawMaterials

0

500

1000

1500

2000

2500

kg/1

000

kg P

rodu

ct

Aluminum Sand Total RawMaterials

Lost FoamSPMPrecision

Figure 4. Consumption of raw materials used in manufacturing phase.

Energy Consumption Contributors - Lost Foam

Electricity53%

Urea Formaldehyde7%

Other6%

Natural Gas3%

PMMA Beads5%

Furnace/Casting26%

Figure 5. Distribution of the energy consumed during the production of materials usedin the Lost Foam process.

Energy Consumption Contributors - SPM

Natural Gas6%

Other5%

Isocyanate3%

Phenol Formaldeyde2%

Furnace/Casting30%

Electricity55%

Figure 6. Distribution of the energy consumed during the production of materials usedby the SPM process.

Energy Consumption Contributors - Precision

Electricity61%

Furnace/casting16%

Phenol Formaldeyde6%

Isocyanate8%

Other6%

Natural Gas3%

Figure 7. Distribution of the energy consumed during the production of materials usedfor precision sand casting.

Energy Consumption

Upstream Gate-to-Gate

Total0

10,00020,00030,00040,00050,00060,00070,00080,00090,000

100,000

kWh/

100

0 kg

Pro

duct

Upstream Gate-to-Gate

Total

Lost FoamSPMPrecision

Figure 8. Comparison of the energy comsumed by each of the three processes studied.

contributions to energy consumed to producethe materials used by each of the castingprocesses. The production of electricity isresponsible for 52.8% of the total for lostfoam (19,412.1 MJ), 54.7% for SPM(16,124.7 MJ), and 60.9% for precision sand(35,271.7 MJ). After electricity generation,the molten aluminum that is delivered to eachcasting facility is the next major source ofenergy consumption. Production of moltenaluminum consumes 16.4% of the totalmaterial producion energy for precision sand(9,518.1 MJ), 25.9% for lost foam (9,502.7MJ), and 29.6% for SPM (8,721.3 MJ). Thenext largest consumer of energy during thematerial production stage for each process isthe production of shot blast at 11.7% for lostfoam (4,299.6 MJ), 13.7% for production ofresin used in the precision sand process(7,954.6 MJ), and 4.8% for production of resinused in the SPM process (1,424.9 MJ).Hence, the use of electricity and moltenaluminum clearly dominate the energyconsumption issues related to materialproduction.

During the manufacturing stage, lostfoam utilizes the least amount of energy andprecision sand the most. The distribution ofenergy consumption within each castingfacility is not available since these data(except that provided by Saturn) wereprovided on a plant-wide basis. Figure 8shows the energy used by each processduring the material production stage, themanufacturing stage, and the overall energyconsumed.

AIR EMISSIONS Emissions of CO2 arelargely dictated by the consumption ofenergy, due to its dependence on combustionof fossil fuels. For the material productionstage, SPM generates the least CO2 andprecision sand the most. This is primarilydue to the low electricity consumptionrequired to produce the materials used in theSPM process and the much higher electricalenergy required to produce the materialsused in the precision sand process. Duringthe manufacturing stage, the vast majority of

CO2 emissions are due to the combustion ofnatural gas. The primary areas of natural gasconsumption are furnaces for melting andholding aluminum, incinerators for processingused sand, furnaces for heat treating finishedcastings, and in the case of lost foam andprecision sand, the incinerators for abatementof hydrocarbon emissions. During themanufacturing stage, lost foam generates theleast amount of CO2. Overall, the CO2emissions from the SPM process arecomparable to that of lost foam, and theemissions from precision sand greatlyexceed those of either lost foam or SPM.These emissions are shown in Figure 9.

The emissions of HC, CO, NOx,SOx, and PM (particulate matter) are shownin Figure 10. The data shown representemissions (after abatement) released to theenvironment. For each of these air emissions(except HC), SPM generates the least andprecision sand the most. For HC, lost foamgenerates the least and precision sand themost. It can also be seen from Figure 10that the manufacturing stage dominates overthe material production stage in thegeneration of HC and CO, whereas thematerial production stage is dominant forgeneration of NOx, SOx, and PM.

For precision sand, the HC dataincludes total HC, whereas the other twoprocesses include only non-methane HC.Also, the SPM HC emissions are largelybased on AP42 estimates, rather than actualmeasurements. Hence a valid comparisonbetween the magnitude of HC emissions isnot available for the three processes.

SOLID AND LIQUID WASTES Solid andliquid wastes have been categorized into twogeneral types: landfilled and recycled. Thesedata are shown in Figure 11. Precision sandgenerates the most landfilled and recycledwastes. SPM generates the least landfilledwaste and lost foam generates the leastrecycled waste. In considering the totalwaste generated (landfilled and recycled), lostfoam generates the least and precision sandthe most.

CO2 Emiss ions

Upstream Gate-to-Gate

Total0

1,000

2,000

3,000

4,000

5,000

6,000

kg/1

000

kg P

rod

uct

Upstream Gate-to-Gate

Total

Lost FoamSPMPrecision

Figure 9. Analysis of CO2 generated by each of the three processes.

ENVIRONMENTAL COSTS Data forenvironmental costs are shown in Figure 12.As with air emissions and solid/liquid waste,the environmental costs are reported on a perthousand kilogram of product basis.However, these costs are reported for themanufacturing stage only. For capital costs,this was accomplished by allocating costsover seven years, at 1996 production rates.For both operating and capital costs, SPMcosts are lowest and precision sand costsare the highest. SPM costs are lowest dueto the absence of incinerators to abate HCemissions and less reliance upon baghousesfor reduction of PM emissions. Precisionsand operating costs appear higher due tothe inclusion of manpower costs formaintenance of the incinerators, baghouses,and amine scrubber. These manpower costswere not included in data provided by theSPM or lost foam manufacturers.

DiscussionSENSITIVITY ANALYSIS The sensitivity ofenvironmental burdens associated with thematerial production phase was evaluated by

reducing, in a stepwise fashion, the quantityof each material consumed. For thisanalysis, the decrease in overallenvironmental burdens in thematerial production phase was calculated fora 20% reduction in material consumption.This sensitivity analysis can be considered ameasure of the dependence of the overallenvironmental burdens due to theconsumption of an individual material.Alternatively, this analysis shows thesensitivity of the overall upstream burdens toa change in the burdens of an individualmaterial. Consequently, this analysis alsoindicates the dependence of the results uponthe accuracy of the material production data.These results are shown first in Table X.This table shows which material change hasthe largest impact upon an environmentalburden category and the degree of impactwhich results from a 20% change in thatmaterial’s consumption. Table X does notconsider the impact of electricity

Air Emissions

HC

CO

NO

x

SO

x

PM

0

5

1 0

1 5

2 0

2 5

3 0

3 5

kg/1

000

kg P

rod

uct

HC

CO

NO

x

SO

x

PM

L o s t F o a mS P MP r e c i s i o n

HC

CO

NO

x

SO

x

PM

0

10

20

30

40

50

60

70

80

90

kg/1

000

kg P

rodu

ct

HC

CO

NO

x

SO

x

PM

Lost FoamS P MPrecision

HC

CO

NO

x

SO

x

PM

01 02 03 04 05 06 07 08 09 0

1 0 0

kg/1

000

kg P

rod

uct

HC

CO

NO

x

SO

x

PM

L o s t F o a mS P MPrec is ion

Upstream Emissions

Manufacturing Emissions

Total Emissions

A

B

C

Figure 10. Comparision of specific air emissions released during the manufacturing(A) and upstream (B) phases for each of the three processes..

Solid and Liquid Waste

TotalLandfilled

TotalRecycled

Total Waste0

200

400

600

800

1000

1200

1400

kg/1

000

kg P

rodu

ct

TotalLandfilled

TotalRecycled

Total Waste

Lost FoamSPMPrecision

Landfilled Recycled Total Waste0

1 02 03 04 05 06 07 08 09 0

kg/1

000

kg P

rod

uct

Landfilled Recycled Total Waste

Lost FoamSPMPrecision

Landf i l led Recyc led Total Waste0

200

400

600

800

1000

1200

kg/1

000

kg P

rod

uct

Landf i l led Recyc led Total Waste

L o s t F o a mS P MPrecision

Upstream Waste

Manufacturing Waste

Total Waste

A

B

C

Figure 11. Solid and liquid wastes (both landfilled and recycled) were compared foreach of the three processes for the manufacturing phase (A) and the upstream phase(B) and total (C).

Environmental Cost

Opera t ing Annua l i zedCapital**

T o t a lAnnual ized

C o s t

0

100

200

300

400

500

600$/

1000

kg

Pro

du

ct

Opera t ing Annua l i zedCapital**

T o t a lAnnual ized

C o s t

L o s t F o a mS P MPrecis ion

* Precis ion operating costs include labor** Capital costs are annual ized over 7 years

*

Figure 12. Comparison of environmentally related capital and operating costs.

consumption. From this table, it can beobserved, for example, that when aluminumconsumption decreases by 20%, thegeneration of mineral waste from the lostfoam process changes by 3.9%. Forprecision sand and SPM, mineral waste ismost impacted (changing by 5.2% and 4.4%,respectively) by a 20% change in sandconsumption. It should be noted that as thepercent of change in the overall environmentalburdens approach 20%, the burdens are morenearly tied directly to that material. That is, ifthe consumption of a material is decreasedby 20% and this leads to a 20% reduction inan environmental burden, then that burden isdue solely to that material.

Some of the environmental burdensare impacted more by changes in theconsumption of electricity at the castingfacility than by changes in materialconsumption. Table XI shows analogousdata to that shown in Table X but includesconsideration of the changes in theenvironmental impacts due to a 20% changein electricity consumption as well as changesin material consumption.

These data show that many of thematerial production phase burdens are mostdependent upon either the consumption ofaluminum or electricity. Opportunities forreducing consumption of aluminum can befocused on the design of gates, risers, andsprues, i.e. the portion of the casting thatdoes not contribute to the net shape of thefinished part.

Although this study did not includethe burdens associated with machining, theamount of aluminum machined away from thefinished casting also has the potential tocontribute significantly to those burdencategories heavily influenced by aluminumconsumption. For lost foam, the ability tocast in many features that reduce requiredmachining is potentially a significantenvironmental advantage.

For the lost foam process, theconsumptions of polyurea and PMMA are themajor contributors to several

Table X. Sensitivity of Environmental Burdens duringMaterial Production to Changes in Material Consumptiona

Lost Foam Precision Sand SPMEnvironmentalBurden

%Changein Burden

Due to 20%Change inMaterial

%ChangeinBurden

Due to 20%Change inMaterial

%ChangeinBurden

Due to 20%Change inMaterial

EnergyConsumption

3.2 Aluminum 2.0 Aluminum 3.0 Aluminum

Particulate Matter 17.9 Aluminum 16.4 Aluminum 18.1 AluminumCarbon Monoxide 15.5 Polyurea 3.8 Aluminum 6.8 AluminumCarbon Dioxide 4.9 Aluminum 2.9 Aluminum 5.2 AluminumSulfur Dioxide 4.4 Aluminum 2.9 Aluminum 5.2 AluminumNitrogen Oxides 4.9 Aluminum 3.1 Aluminum 5.4 AluminumNitrous Oxide 10.9 PMMA 15.8 Isocyanate 12.6 IsocyanateHydrocarbons 5.4 EPS 3.6 Isocyanate 5.5 AluminumMethane 2.1 Aluminum 1.4 Aluminum 1.6 AluminumMineral Waste 3.9 Aluminum 5.2 Sand 4.4 SandMixed IndustrialWaste

5.8 PMMA 13.8 Isocyanate 10.8 Isocyanate

Slags/Ash 4.7 Aluminum 2.9 Aluminum 5.3 AluminumInert Chemical 13.4 Polyurea 18.5 Isocyanate 17.1 IsocyanateRegulated Chemical 19.2 Polyurea 16.8 Isocyanate 14.6 Isocyanatea Only the material with the greatest impact on the environmental burden category is listed.

environmental burden categories. Thesematerials are used for cleaning heads andblocks, respectively. Reduction inenvironmental burdens might be possibleusing alternative materials or processes.

For precision sand and SPM,consumption of isocyanate and sand aremajor contributors to several environmentalburden categories. These materials are usedin the creation of cores. Reduction of theseburdens are possible either through the use ofalternative binders or reduction in thegeneration of sand fines, which is responsiblefor sand consumption.

For all three casting processes, theconsumption of electricity at the castingfacility is a major contributor to mostenvironmental burden categories. For mostenvironmental burden categories for whichelectricity is the largest contributor, morethan half the burden is due to electricity.Hence, improved efficiency of electricityconsumption will have direct environmentalbenefits.

IMPLICATIONS Life cycle assessmentstudies often go beyond the acquisition ofinventory data and estimate environmentalimpacts. The impacts that are oftenconsidered include tropospheric ozone andsmog formation, surface water eutrophicationand/or acidification, stratospheric ozonedepletion, greenhouse warming potential, andtoxicity to human and/or terrestrial life.However, it is often technically difficult, if notimpossible to calculate these. In somecases, it is inappropriate to use life cycledata to calculate the impacts. For example,smog formation occurs over short time scalesand limited geographical regions. But theemissions of smog precursors such as NOx,HC, and CO that result from the life cycle ofcast aluminum products occurs over a widegeographical and temporal scale. Therefore,it is inappropriate to sum these life cycleemissions and use them to estimate smogformation as an impact. This is often alsotrue for surface water eutrophication and/oracidification.

Stratospheric ozone depletion andgreenhouse warming are phenomena that arecaused by gas phase species that have

Table XI. Sensitivity of Environmental Burdens during Material Production to Changes in Materialand/or Electricity Consumption

Lost Foam Precision Sand SPMEnvironmentalBurden

%Changein Burden

Due to 20%Change inMaterial

%Changein Burden

Due to 20%Change inMaterial

%Changein Burden

Due to 20%Change inMaterial

EnergyConsumption

8.8 Electricity 10.0 Electricity 7.5 Electricity

Particulate Matter 17.9 Aluminum 16.4 Aluminum 18.1 AluminumCarbon Monoxide 15.5 Polyurea 10.7 Electricity 9.7 ElectricityCarbon Dioxide 13.6 Electricity 14.8 Electricity 13.3 ElectricitySulfur Dioxide 13.2 Electricity 15.9 Electricity 14.1 ElectricityNitrogen Oxides 11.7 Electricity 13.6 Electricity 11.8 ElectricityNitrous Oxide 10.9 PMMA 15.8 Isocyanate 12.6 IsocyanateHydrocarbons 6.4 Electricity 7.3 Electricity 7.9 ElectricityMethane 4.8 Electricity 5.6 Electricity 3.4 ElectricityMineral Waste 11.5 Electricity 11.0 Electricity 11.0 ElectricityMixed IndustrialWaste

5.8 PMMA 13.8 Isocyanate 10.8 Isocyanate

Slags/Ash 14.1 Electricity 2.9 Aluminum 14.3 ElectricityInert Chemical 13.4 Polyurea 18.5 Isocyanate 17.1 IsocyanateRegulated Chemical 19.2 Polyurea 16.8 Isocyanate 14.6 Isocyanate

long atmospheric lifetimes relative to the lifecycle of an automobile. Hence, thesephenomena are more relevant to life cycleassessment studies. Stratospheric ozonedepletion is a phenomenon associatedprimarily with chlorofluorocarbon (CFC)emissions. In the case of this study, CFCemissions are insignificant, hencestratospheric ozone depletion is not animpact category of concern. Greenhousewarming is a phenomenon associatedprimarily with species such as CO2, CFC’s,methane, and N2O. Casting processes andthe production of materials used duringcasting do generate CO2, N2O, and methaneemissions, so this is an impact category ofconcern.

For the manufacturing phase, we donot have measurements of CO2 and N2Oemissions from any of the casting processesand only have measurements of methanefrom the lost foam process. To estimate themethane emissions from SPM and precisionsand, and the CO2 and N2O emissions fromall three casting processes, we have used theAP42 estimates of these emissions from the

combustion of natural gas. The lack of datafor emissions from sources other than naturalgas combustion means that the CO2equivalent greenhouse gas emission valuesare probably underestimates. For thematerial production phase, Boustead datawere used (1). The life cycle greenhouse gasemissions of the three casting processes aresummarized in Table XII. The totalemissions are adjusted for the global warmingpotentials (GWP) of methane and N2Orelative to CO2. These data show that theGWP of all casting processes are dominatedby the emissions of CO2. Since the majorityof CO2 emissions are a consequence offossil fuel combustion, either directly fromnatural gas combustion at the casting facilityor from the generation of electricity, thesedata indicate the importance of energyconservation as a means to reduce GWP.

OVERALL ENVIRONMENTALPERFORMANCE The data for materialand energy consumption, emissions of eachpollutant to the atmosphere, wastes (bothrecycled and landfilled), and environmental

costs (both capital and operating) have beensummarized in a spider chart shown asFigure 15. This figure shows, at a glance,

Table XII. Greenhouse Warming Potentials of the Casting Process Life CyclesCompound GWP

(relative toCO2)

c

SPMa

PrecisionSand

aLost Foam

a

CO2 1.0 3495.5b 5874.8b 3689.8b

Methane 21 22.52b 29.82b 18.98N2O 310 0.0061b 0.0067b 0.0043

Total (mg x GWP) 3970.3b 6503.1d 4089.7a Values are kilograms emitted per 1000 kilograms of degated product.

b Values from AP42 estimates of emissions from natural gas combustion. Other process emissions are

unknown.c Greenhouse warming potentials are based upon a 100 year time frame (4).

d Contribution from methane emissions only include methane emitted from natural gas combustion, not from

other sources in the casting process.

.

SummaryTotal Energy

CO2

HC

PM

SOx

NOx

CO

Al

Sand

Other Materials

Operating Cost

Capital Cost

Landfilled Waste

Recycled Waste

Lost FoamSPMPrecision

Figure 15. Summary of environmental burdens analyzed in this study.

the overall values for each environmentalburden summed over material production andmanufacturing, except in the case of rawmaterial consumption and environmentalcosts, which were calculated for themanufacturing stage only. On this figure, an

optimal value would be zero, located at thecenter of the chart.

We have evaluated these data andhave identified the parameters that wouldmost influence an environmentally-basedchoice of one of these three casting

processes. These data are summarized in aspider chart shown as Figure 16. Theseburden categories are raw material andenergy consumption, solid and liquid wastes(summed over recycled and landfilledwastes), HC, CO2, and PM emissions, andtotal environmental costs.

ConclusionsA thorough life cycle analysis of

aluminum lost foam, precision sand, andsemi-permanent mold casting has beencompleted. Inventories of material andenergy consumption, air emissions, andliquid and solid wastes have been performed.We have mass-normalized the results on thebasis of 1000 kg of degated product. Ouranalysis omits the use-phase and end-of-lifephase burdens of the products, which shouldbe approximately equal for all processes onthe basis of mass of product. Also, we havenot considered the burdens associated withmachining. In addition to the inventoryanalysis, a sensitivity analysis has beenperformed on the material production burdens

and the life cycle CO2 equivalent greenhousegas emissions have been calculated.

For all processes, the manufacturingphase electricity and aluminum consumptionare major contributors to the materialproduction stage burdens. Other majorcontributors are shot blast media andpolystyrene for the lost foam process andresins and sand for the precision sand andSPM processes.

This analysis indicates that, for mostcategories, lost foam and SPM burdens arecomparable. For all categories, the burdensfrom precision sand are the highest.Although lost foam and SPM burdens arenearly equal, SPM has slightly lowerenvironmental costs, while lost foam haslower HC and PM emissions, and less totalwaste. SPM and lost foam have nearly equalmaterial and energy consumption as well asCO2 equivalent greenhouse gas emissions.However, the data probably underestimate themethane emissions from the SPM process.Overall, our analysis indicates that lost foamcasting of aluminum heads and blocks hasless environmental impact than either SPM orprecision sand casting.

SummaryRaw Materials

Energy

Waste

Environmental CostCO2

HC

PM

Lost FoamSPMPrecision

Figure 16. Summary of the major environmental burdens analyzed in this study.

References1. The Boustead Model, Version 3, January,

1997.2. AP42, Volume I: Stationary Sources, US-

EPA, Office of Mobile Sources, AnnArbor, MI, 1996.

3. Boustead, Ian, personal communication,February, 1998.

4. Houghton, J. T., Filho, M., Bruce, J., Lee,H., Callander, B. A., Haites, E., Harris,N., and Maskell, K., Climate Change1994: Radiative Forcing of ClimateChange and An Evaluation of the IPCCIS92 Emission Scenarios, CambridgeUniversity Press, 1995.