Faster Cleanup of Contaminated Soil and Groundwater · Livermore are successfully cleaning up the groundwater beneath a utility pole yard in Visalia, California. Helped by Livermore

May 1998

Lawrence

Livermore

National

Laboratory

Also in this issue: • Computational Mechanics Move Ahead• DAT: Device Assembly Facility• Corsica: Simulations for Magnetic Fusion Energy

Faster Cleanup ofContaminated Soiland Groundwater

Faster Cleanup ofContaminated Soiland Groundwater

About the Review

• •

About the Cover

Lawrence Livermore National Laboratory is operated by the University of California for theDepartment of Energy. At Livermore, we focus science and technology on assuring our nation’s security.We also apply that expertise to solve other important national problems in energy, bioscience, and theenvironment. Science & Technology Review is published ten times a year to communicate, to a broadaudience, the Laboratory’s scientific and technological accomplishments in fulfilling its primary missions.The publication’s goal is to help readers understand these accomplishments and appreciate their value tothe individual citizen, the nation, and the world.

Please address any correspondence (including name and address changes) to S&TR, Mail Stop L-664,Lawrence Livermore National Laboratory, P.O. Box 808, Livermore, California 94551, or telephone (925) 422-8961. Our electronic mail address is [email protected].

Prepared by LLNL under contractNo. W-7405-Eng-48

S&TR is available on the Internet athttp://www.llnl.gov/str. As references becomeavailable on the Internet, they will be interactivelylinked to the footnote references at the end of eacharticle. If you desire more detailed informationabout an article, click on any reference that is incolor at the end of the article, and you will connectautomatically with the reference.

Electronic Access

We want to know what you think of ourpublication. Please use the enclosed survey formto give us your feedback.

What Do You Think?

Technologies developed at LawrenceLivermore are successfully cleaning up thegroundwater beneath a utility pole yard inVisalia, California. Helped by Livermoreresearchers as operational consultants, SouthernCalifornia Edison and SteamTech EnvironmentalServices are using the dynamic undergroundstripping and hydrous pyrolysis/oxidationtechniques to remove contaminants at anastounding rate—5,000 times that of conventionaltreatment.

SCIENTIFIC EDITOR

J. Smart

MANAGING EDITOR

Sam Hunter

PUBLICATION EDITOR

Sue Stull

WRITERS

Arnie Heller, Lori McElroy, Katie Walter,and Gloria Wilt

ART DIRECTOR AND DESIGNER

Ray Marazzi

INTERNET DESIGNER

Kitty Tinsley

COMPOSITOR

Louisa Cardoza

PROOFREADER

Al Miguel

S&TR is a Director’s Office publication,produced by the Technical InformationDepartment, under the direction of the Office of Policy, Planning, and SpecialStudies.

2 The Laboratory in the News

3 Commentary by Jay DavisChanging the Game

Features4 They All Like It Hot: Faster Cleanup of Contaminated Soil and

GroundwaterIn a hundredth the time of the original estimate, cleanup of a heavily contaminated Superfund site in California’s Central Valley progresses, using methods developed at Lawrence Livermore.

12 Computational Mechanics Moves AheadThe Engineering Directorate’s Methods Development Group has produced advanced modeling and simulation tools that solve an ever-widening range of engineering problems.

Research Highlights20 Corsica: Integrated Simulations for Magnetic Fusion Energy23 Device Assembly Facility: New Facilities for Handling Nuclear

Explosives

27 Patents and Awards

28 Abstracts

S&TR Staff May 1998

LawrenceLivermoreNationalLaboratory

Printed in the United States of America

Available fromNational Technical Information ServiceU.S. Department of Commerce5285 Port Royal RoadSpringfield, Virginia 22161

UCRL-52000-98-5Distribution Category UC-700May 1998

Page 4

Page 23

Page 12

2 The Laboratory in the News

Science & Technology Review May 1998

Lab help for BART high-tech trainingRepresentatives from BART (Bay Area Rapid Transit),

Lawrence Livermore National Laboratory, and Chabot–LasPositas Community College District met with U.S.Congresswoman Ellen Tauscher to sign a Memorandum ofAgreement to develop high-tech training tools for rapid-transittrain operators. The result will be performance-based trainingthat includes enhanced curricula, updated methodology, andhigh-tech simulations. The curricula will help focus training onthe areas of highest need in the shortest possible time.

Lawrence Livermore will provide scientific and technicalexpertise for identifying and recommending training deliverymethodology and equipment: computer simulation of complexsystems, advanced sensors and instrumentation, transportationstudies, and diverse engineering capabilities. The Chabot–LasPositas Community College District will design performance-outcome, competency-based curricula. BART brings extensiveexperience and expertise in all areas of public rapid transitoperations, workforce development, safety, and customer service.

Livermore Deputy Director for Science and Technology JeffWadsworth said, “This partnership represents a welcomeopportunity to apply the Laboratory’s state-of-the-art capabilitiesand scientific expertise to a real problem in the Bay Area.Improved training programs and facilities for rapid transitoperators benefit our community.”Contact: Elizabeth Rajs (925) 424-5806 ([email protected]).

Kidney disease gene foundYears invested in building a human genetic roadmap are

yielding big-time benefits. Lawrence Livermore NationalLaboratory scientists have teamed with internationalcollaborators to pinpoint the location of a gene responsible forcausing a kidney disease. This effort is described in a researchpaper in the March 1998 issue of Molecular Cell.

Scientists at Livermore’s Human Genome Centercollaborated with researchers from the Karolinska Institute inStockholm, Sweden, and the University of Oulu in Oulu,Finland. Together, the three institutions discovered the gene forcongenital nephrotic syndrome, a deadly disease that causesmassive amounts of proteins to be excreted from the body. Thegene is located on chromosome 19.

A progressive disease, congenital nephrotic syndromeusually leads to death by age 2. The only life-savingalternative is a kidney transplant. The disease is mostprevalent in Finland, striking about one in 10,000 Finnishchildren, but occurring with a significantly lower frequency inother nations. The discovery has already yielded immediateclinical assistance, enabling development of a diagnostic toolto identify carriers of the gene.

To date, Livermore’s Human Genome Center has mappedmost of chromosome 19. It has also sequenced, or ordered, the base pairs for about 3.5 million of the chromosome’s 65 million bases, in a process that is accelerating with thedevelopment of new technologies. The Center has assistedcollaborators in discovering the genes for a variety of diseases,including myotonic dystrophy, a disease that affects about onein 8,000 adults worldwide; the genetic cause of two forms ofdwarfism; a gene for migraine headaches; and the Peutz– Jeghers syndrome gene. Additionally, Lab researchers alsojoined in the effort to find a gene for cadasil, a type of stroke.Contact: Public Affairs Office (925) 423-3107 ([email protected]).

New associate director for EnergyDirector Bruce Tarter named Terry Surles the new Energy

Associate Director, commenting that Surles “brings to thisposition proven leadership in building large programs andcurrency with the issues and players in the broad energy andenvironmental arena. It is becoming increasingly clear to methat the development, assessment, and application of energytechnologies will play a stronger role in national policy andpractice. I look forward to working with Dr. Surles and theEnergy Directorate to make substantial contributions to such acrucial national issue.”

Surles most recently was the deputy secretary for scienceand technology for the California Environmental ProtectionAgency. Prior to that he served at Argonne NationalLaboratory in the environmental and energy areas and was thegeneral manager of Environmental Programs from 1993 to1997. He received his bachelor’s degree in chemistry from St.Lawrence University in 1966 and his doctorate in analyticalchemistry from Michigan State University in 1970. He joinedArgonne in 1978 after serving posts in academia and privateindustry. A member of the American Association for theAdvancement of Science and the American Chemical Society,he has published more than 30 peer-reviewed journal articlesand 70 technical reports.

(Continued on p. 26.)

HE Department of Energy’s cleanup problems are oftenrepresented as intractable. The cleanup budget, now the

largest single component in DOE’s budget, is seen asinadequate, unlikely to decrease, and detrimental to other DOEfunctions. Although unique problems with poorly containedand characterized radioactive wastes dominate DOE’s problems,the issue of spills and plumes of conventional materials such assolvents, fuels, and mixtures of organic materials is significant,too. Estimates have shown that DOE has a $10-billion legacy ofsuch plumes to clean up. The total cost of cleaning Departmentof Defense and industrial sites across the country is larger still.When Laboratory Director Bruce Tarter established the Earthand Environmental Sciences Directorate in 1994, he challengedus to change the game in environmental cleanup. The story ofthe Visalia pole yard in this issue of S&TR (beginning on p. 4)demonstrates the first success in response to that challenge.

Plumes of contaminants in groundwater pose severalproblems: hazards to public safety and health when the wateris used, cleanup cost burdens to property owners, and, becauseof the associated liabilities, transfer of property ownership.While transferring property is not a burden to most DOE sites(which are likely to remain in government possessionindefinitely), it is an obstacle to DOD site closure and reuse ofvaluable lands (such as California’s Mare Island or AlamedaNaval Air Station).

Remediation steps to date commonly rely on pump-and-treat.That method works well for near-term regulatory compliance(the site owner is doing something in good faith) to control thespread of plumes, but it rarely removes all the contaminants.Toxic compounds in spills have had decades to sink and bindto fine silts and clays. They are unlikely to be removed bysimply slowly flushing water through them.

Steam applied to the subsurface environment mobilizesdifficult materials such as heavy oils, which can be extractedat greatly accelerated rates. The addition of oxygen through airinjection further destroys contaminant materials. (Rememberthe elementary chemistry lesson that water at 100°C is a veryreactive substance.) As an added benefit, heating the subsurfacepromotes the growth of thermophylic bacteria that move intoa warmed region and actively bioremediate remaining tracesof contaminants.

T

3


Commentary by Jay Davis

The technical success of Visalia resulted from finding acorporate sponsor, Southern California Edison, that waswilling to invest the funds up front for a significantdeployment of a new technology at the scale ($20 million)required for a substantive engineering and economic test.DOE sites tend to be constricted by budgets annuallycompeting for money to complete scheduled tasks, producinga context in which “experiments of scale” are perceived asactivities that divert funds from ongoing remediation or delaytasks required under Records of Decision. A corporationunderstands the value of accelerating cleanup and movingthe liability off its books, which DOD now finds attractiveas well.

Lawrence Livermore’s role in this project was to developand demonstrate relevant cleanup and diagnostic techniques onits own site and then transfer the technologies to a corporatepartner who owns a Superfund site and to a subcontractor withsteam-injection experience. We functioned as a supportingscience and technology team, bringing our experience insupercomputer modeling of subsurface flow and reactivetransport, rare-gas tracers and electrical tomographicdiagnostics, and isotopic analyses to validate results. Thesetools are not available in the commercial sector, and theprocesses are not readily done by the academic community,proving that there is a role for a national laboratory in solvingthis problem. Some of these tools, now in the hands of theindustrial participants, are available for use at other sites.

We now are exploring with DOE and other agencies thenext possible sites for testing these techniques. At each site,we will encounter a new combination of subsurface variablesand a new cocktail of contaminants. The remediation processwill have to be validated anew on each site—we have nosilver bullets for subsurface contamination.

As we work at these sites, more technology, understanding,and tricks of the trade will move into industrial hands.The rewards for us will be in helping to remove pressinghealth hazards cost effectively throughout the nation andgaining future support to move on to the next set of“intractable” problems.

■ Jay C. Davis is Associate Director, Earth and Environmental Sciences.

Changing the GameChanging the Game

4

LEAN up a greasy kitchen spill withcold water and the going is slow.

Use hot water instead, and progressimproves markedly. So it makes sensethat cleanup of greasy undergroundcontaminants such as gasoline might gofaster if hot water or steam weresomehow added to the process.

The Environmental ProtectionAgency named hundreds of sites to theSuperfund list—sites that have beencontaminated with petroleum productsor petroleum products or solvents.Elsewhere across the country, thousandsof properties not identified on federalcleanup lists are contaminated as well.Given that under current regulations,underground accumulations of solventand hydrocarbon contaminants (themost serious cause of groundwaterpollution) must be cleaned up, finding arapid and effective method of removingthem is imperative.

In the early 1990s, in collaborationwith the School of Engineering at theUniversity of California at Berkeley,Lawrence Livermore developeddynamic underground stripping. (For anexplanation of this method, see Energy& Technology Review, July 1992, pp. 1–7). This method for treatingunderground contaminants with heat ismuch faster and more effective than

C


They All Like It Hot: Faster Cleanup of ContaminatedSoil and Groundwater

They All Like It Hot: Faster Cleanup of Contaminated Soil and Groundwater

5


traditional treatment methods. Morerecently, Livermore scientists developedhydrous pyrolysis/oxidation, a processthat converts contaminants in the groundto such benign products as carbondioxide, chloride ions, and water. Byintroducing both heat and oxygen, thisprocess has effectively destroyed allpetroleum and solvent contaminants thathave been subjected to laboratory tests.

During the summer of 1997, bothprocesses were used for cleanup of afour-acre site in Visalia, California,owned by Southern California Edison(Figure 1). The utility company hadused the site for 80 years to treat utilitypoles by dipping them into creosote, a pentachlorophenol compound, orboth. By the 1970s, these highly toxic substances had seeped into thesubsurface to depths of approximately100 feet (30 meters). The Visalia pole yard bore the distinction of being one of the original Superfundsites.

Southern California Edison andSteamTech Environmental Services ofBakersfield, California (the firstcommercial site licensee of thedynamic underground strippingtechnology), are cleaning up the Visaliasite, with Livermore staff periodicallyon hand as operational consultants.

During the first six weeks of operation,between June and August 1997, theteam removed or destroyed in placeapproximately 300,000 pounds (135 metric tons) of contaminants, arate of about 46,000 pounds (22 metrictons) per week (Figures 2 and 3). Fornearly 20 years, Southern CaliforniaEdison had been removing contaminantsfrom the subsurface using the standardcleanup method, known as pump-and-treat, most recently at a rate of just 10 pounds (0.03 metric ton) per week.In contrast, the amount of hydrocarbonsremoved or destroyed in place in thosesix weeks was equivalent to 600 yearsof pump-and-treat, about 5,000 timesthe previous removal rate. Needless tosay, the Visalia cleanup using dynamicunderground stripping plus hydrouspyrolysis/oxidation is considered a wildsuccess by everyone involved.

Geophysicist Robin Newmark andgeochemist Roger Aines are LawrenceLivermore project leaders for the workat Visalia. Says Aines, “No one reallyknew what was underground at Visalia.Through the winter of 1998, SouthernCalifornia Edison and SteamTech have removed over 540,000 pounds(245 metric tons), and the job still isn’tfinished. However, contaminantconcentrations in recovered

groundwater continue to drop, so weknow the end is in sight.”

Finding a Better WayFor years, scientists in Livermore’s

Earth and Environmental SciencesDirectorate have been researching bettermethods to clean up soil and groundwatercontamination, in part because both theLivermore site and Livermore’s Site 300are also Superfund sites as a result ofU.S. Navy, Atomic Energy Commission,and DOE operations. Most contaminantsat the Livermore sites are eitherpetroleum distillates (e.g., gasoline, dieselfuel) or chlorinated hydrocarbons used assolvents. Existing methods to removethese compounds from soil andgroundwater have halted their migrationoff the site, but cleanup will still take adecade or more to complete.

For about 20 years, the traditionalmethod of cleaning up contaminatedgroundwater has been the pump-and-treat method. Water is pumped from thewater table to the ground surface, treatedto remove or destroy contaminants, andreturned underground. Huge amounts of water must be flushed through thecontaminated area for years or evendecades to clean it, and even then thecontamination may not be completelyremoved.

The original estimate to clean the Visalia Superfundsite was more than 100 years. Using technologiesdeveloped at Lawrence Livermore, cleanup ishappening in one to two years and at a much lower costthan with pump-and-treat methods.

of water because contaminants leach outvery slowly. When you try to clean themup with pump-and-treat, it’s like tryingto rinse a soapy sponge. You have to run

6


Visalia Cleanup

vast amounts of water through the spongebefore all the soap is finally out.”

Pump-and-treat systems are relativelyinexpensive to operate, but theyrepresent a long-term cost. They offercompliance in a regulatory sense, but theresults are not very satisfying becausethe site is unlikely to be completelycleaned up.

Boiling Off ContaminantsThe dynamic underground stripping

technology developed by Livermore andthe University of California was firstdemonstrated in the cleanup of anunderground gasoline spill at theLivermore site in 1993 (see Energy &Technology Review, May 1994, pp. 11–21). Dynamic undergroundstripping was so successful in thiscleanup that contaminants were removed50 times faster than with the pump-and-treat process. The cleanup, estimated totake 30 to 60 years with pump-and-treat,was completed in about one year. In1996, the Environmental ProtectionAgency and other regulators declared thatno further remedial action was required.

In this method, the area to be cleanedis ringed with wells for injecting steam attemperatures above 100°C. Extractionwells in the central area are used tovacuum out vaporized contaminants. Toensure that thick layers of less permeablesoils are heated sufficiently, electrodeassemblies are sunk into the ground andthe ground is heated, which forcestrapped liquids to vaporize and move tothe steam zone for removal by vacuumextraction. These combined processesachieve a hot, dry, contaminant-free zoneof earth surrounded by cool, damp,untreated areas. Steam injection andheating cycles are repeated as long asunderground imaging shows that cool(and therefore untreated) regions remain.

Although the initial capital outlay fordynamic underground stripping is higherthan for pump-and-treat systems, the

Says Newmark, “Some of thesolvents and other contaminants havevery low solubility. So very smallamounts can pollute millions of gallons

Groundwater flow direction

Steam plant

Cooling/condensingTreatment facility

29,400 lb vapor hydrocarbon burned in boilers

45,500 lb in situ destruction

(removed carbon dioxide)

300,000 lb free product

17,500 lb dissolved hydrocarbon

activated carbon filtrator

Figure 1. An aerial view of the Visalia site. Injection wells are shown in magenta, andextraction wells are shown in yellow.

Figure 2. During the first six weeks of operation in 1997, about 300,000 pounds (135 metric tons)of contaminated product was either brought to the surface or destroyed in situ at Visalia.Southern California Edison will treat the liquid “free product” on site and may use it as a lubricantin its operations.

7


Visalia Cleanup

process saves money in the long runbecause it is completed much morequickly. Most of the equipment, such asboilers for generating steam, can berented. Up-front costs include installingthe heating wells, renting theequipment, and operating the systemintensively for a short period of time.There are no long-term operation andmaintenance costs.

That 1993 field trial of dynamicunderground stripping cost about $110per cubic yard ($140 per cubic meter),although Livermore scientists believethey could repeat the project for about$65 per cubic yard ($85 per cubicmeter). Because contamination at thegasoline spill at the Livermore sitemigrated downward 40 meters, diggingup the contaminated soil and disposingof it would have cost almost $300 percubic yard ($400 per cubic meter). Soilremoval and disposal costs are moretypically in the range of $100 to $200per yard ($130 to $260 per cubicmeter); pump-and-treat method costsare as high as or higher than soilremoval costs.

Unexpected HelpThe Livermore team discovered an

unexpected benefit of dynamicunderground stripping: it encouragesbioremediation. Heating the soil at thegasoline spill site to temperatures above100°C was expected to sterilize it, withthe microorganisms that use petroleumproducts as food expected to returnslowly as the soil cooled. But soilsamples taken soon after completion ofthe cleanup revealed large numbers ofmicrobes that thrive in hightemperatures (known as thermophiles),apparently because predators andcompetition had been eliminated.

Bioremediation is an important finalstep in soil and groundwater cleanupsbecause the microorganisms destroy

residual contaminants missed during theinitial cleanup process.

Oxygen Is Key to ApproachWith dynamic underground

stripping, the contaminants arevaporized and vacuumed out of theground, leaving them still to bedestroyed elsewhere. In fact, about halfthe cost of a typical cleanup is intreating the recovered groundwater andhauling away and disposing of thecontaminated material that is brought tothe surface.

“Livermore’s hydrous pyrolysis/oxidation technology takes the cleanupprocess one step further,” explainsAines, “by eliminating the treatment,handling, and disposal requirements anddestroying the contamination in theground.” The Visalia pole yard cleanupis the only application of this method todate, but indications are that large-scalecleanups with hydrous pyrolysis/oxidation could cost as little as $25 percubic yard ($33 per cubic meter), anenormous savings over current methods.Best of all, the end product of a hydrouspyrolysis/oxidation cleanup withbioremediation as a final step is expectedto be a truly clean site.

The hydrous pyrolysis/oxidationprocess builds on the team’s experiencewith heating large amounts of soil thatwas gained during earlier work withdynamic underground stripping. Toprovide the oxygen, steam and air areinjected in parallel pipes, building aheated, oxygenated zone in thesubsurface. When injection is halted,the steam condenses and contaminatedgroundwater returns to the heated zone.The groundwater then mixes with thecondensed steam and oxygen, whichdestroys dissolved contaminants. Thisprocess avoids many of the mixingproblems encountered in other in situoxidation schemes. In such processes,an oxidizing reagent is typicallyinjected into the subsurface, resulting inthe displacement of the contaminant.Without a return process such as steamcondensation, the contaminant andoxidant never mix or mix poorly at best.

During the heating process inhydrous pyrolysis/oxidation, the dense,nonaqueous-phase liquids and dissolvedcontaminants are destroyed in placewithout surface treatment. Thetechnique improves the rate andefficiency of remediation by renderingthe hazardous materials benign by a

Figure 3. Contaminantfloating on water in thedissolved-air-flotationtank (at lower right).Dissolved air formsbubbles that captureand lift free-productcontaminant to thesurface of theseparator.

8


Visalia Cleanup

completely in situ process. Hydrouspyrolysis/oxidation also takes advantageof the large increase in mobility thatoccurs when the subsurface is heatedand makes contaminants more availablefor destruction. Many remediationprocesses are limited by the access ofthe reactants to the contaminant, makingthe lack of mobility the bane ofremediation efforts in low-permeabilitymaterials such as clays.

Most early Livermore experimentson the hydrous pyrolysis/oxidationprocess, funded by DOE, were withtrichloroethylene (TCE), a solvent thatwas widely used in degreasing and otherindustrial processes. TCE is the mostcommon groundwater contaminant inthe DOE complex and in most industrialareas. Unlike gasoline, TCE and similarsolvents are heavier than water, whichmeans that they can sink below thewater table, making cleanup extremelydifficult, if not impossible, withconventional methods.

“The oxidation process occursnaturally, but without heat it is veryslow,” explains Kevin Knauss, theLivermore geochemist who leads theeffort in the laboratory, “so we neededto know how hot the soil had to be.”The team learned that with TCE, just afew degrees can make an enormousdifference in how quickly thebreakdown occurs. At 90°C, it takes afew weeks; at 100°C, it takes a fewdays; and at 120°C, it occurs in just afew hours. Laboratory results indicatedthat the contaminants at Visalia wouldreact at similar rates.

Figure 4. Three-dimensional images fromelectrical resistance tomography datashow how resistivity increases whensteam injection is under way. After a shortshutdown for hardware modifications,steam injection began on October 31, 1997.Steam injected into central wells (arrows)preferentially fills an old alluvial channel(river). Even when a well is not in activeuse for steam injection, it is kept open witha very small amount of steam; hence, thesmall “puddles” near inactive injectionwells. (Images are courtesy ofSteamTech, Bakersfield, California.)

November 1, 1997

November 4, 1997

Steam injectionturned on

Steam injectionturned on

9


Visalia Cleanup

properties vary with temperature, soiltype, and fluid saturation. For example,higher electrical resistivity is found inmore permeable sand and gravel soils.Conversely, less permeable clay soils show lower resistivity (higherconductivity). Baseline measurementswith electrical resistance tomographyare used to characterize a site and topredict steam pathways.

During treatment at Visalia, dailyresistivity measurements supplied apicture of the progress of the steamfront and the heated zones. Monitoringthe progress of the heating frontsensured that all soil was treated.Temperature measurements made inmonitoring wells revealed details of thecomplex heating phenomena in theindividual soil layers.

To evaluate the progress of thechemical destruction of contaminantsin situ, the team also developed fieldmethods for sampling and analyzinghot water for contaminants, oxygen,intermediate products, and products of reaction. Because hydrouspyrolysis/oxidation is an aqueous-phase reaction, capturing and

Experiments with dissolved oxygenshowed that the oxygen in air issufficient to degrade the contaminants.Because oxygen is corrosive, pumpingpure oxygen into the ground could veryquickly damage the piping system, butthe less concentrated oxygen from theair is less corrosive and easy tointroduce.

The demonstration of hydrouspyrolysis/oxidation at Visalia confirmedthe effectiveness of this technology fordense, heavier-than-water groundwaterpollutants such as creosote andpentachlorophenol. The method has alsobeen tested successfully in thelaboratory on contaminants resistant tocleanup in the past—for example, carbontetrachloride, a chemical used as arefrigerant and a dry-cleaning solvent,and polychlorinated biphenyls (PCBs), achemical used in electrical transformersand capacitors.

Project co-leader Aines notes, “Thisnew technology could also be used tomop up methyl tert butyl ether(MTBE), a gasoline additive that hasbegun showing up in Californiagroundwater.”

The method can also be used to clean up groundwater and soils to almostany depth.

Controlling the Process Several geophysical techniques were

used at Visalia to monitor theunderground movement of steam andthe progress of heating. One techniquewas electrical resistance tomography, atechnology developed at Livermore,applied during the 1993 gasolinecleanup, and now availablecommercially. Electrical resistancetomography is an imaging method likea CAT scan that provides near-real-time images of the undergroundprocesses between pairs of monitoringwells (Figure 4). Soil electrical

evaluating the fluid in that phase isessential. At elevated temperatures,many of the key constituents aresufficiently volatile that traditionalsampling techniques are not suitable.The Livermore team developed high-temperature systems that can deliver apressurized, isolated fluid stream tothe surface, where in-line analysis canbe performed.

Building on Livermore’s experiencein using noble-gas tracers to track watermovement (see S&TR, November 1997,pp. 12–17), Bryant Hudson designedtracer experiments to help verifyhydrous pyrolysis/oxidation in the field.Noble-gas tracers—including helium,neon, krypton, and xenon—were addedto injected water and steam to track themovement of the steam (and subsequentcondensation to liquid water) and themovement of other gases initiallypresent in the steam (Figure 5).Naturally occurring dissolved gases(nitrogen and argon) providedmeasurements of atmospheric andnative groundwater interaction. Once a“packet” of water had been tagged withgas tracers, it could later be identified

Figure 5. Livermorephysicist BryantHudson (right), whohas developedseveral methods formonitoringgroundwater withnoble-gas tracers,adjusts the gas flowwith mechanicaltechnician AllenElsholz. Boilers arein the background.


pyrolysis/oxidation (Figure 6) wasfound in a number of sources, includingthe disappearance of dissolved oxygen(consumed through the hydrouspyrolysis/oxidation reactions), theappearance of oxidized intermediateproducts, and the production of carbondioxide (the final product of thisprocess). Important information on theisotopic content of the carbon in thecarbon dioxide was obtained fromLivermore’s Center for AcceleratorMass Spectrometry. The ratios of 14C/ 12C and of 13C/12C in carbondioxide from the subsurface were moresimilar to those of the petroleum-basedcontaminants than to those ofgroundwater in the area, indicating thatthe contamination was being destroyedand converted to carbon dioxide.

Modeling to Predict/EvaluateThe team used NUFT, a widely used

three-dimensional groundwatermodeling code developed several yearsago at Livermore, to model suchimportant process parameters as mixingof steam, air, and groundwater. Ingeneral, simulations with models suchas NUFT provide a diagnostic meansfor anticipating the results of adecontamination scheme in complicatedsoil environments and, later, forevaluating field results.

In the case of a cleanup such as theone at Visalia, where much of thedecontamination occurs in situ and istherefore not directly observable, thenoble-gas tracers provide data criticalfor validating initial modeling results.

“Modeling proved invaluable atVisalia and remarkably accurate aswell, compared with results frommonitoring wells,” explains Newmark.“Livermore models predicted steam andtracer movement to within an hour ortwo in most instances.” (See Figure 7.)

• Correlating the intermediate hydrouspyrolysis/oxidation destructionproducts with temperature and oxygen.• Identifying the overall isotopiccontent of the extracted carbon with regard to carbon-14 (14C) andcarbon-13 (13C).

In soil-gas and water samples,evidence of the progress of hydrous

by the types and amounts of tracers in it.The tracers thus assisted with manytasks, including:• Following the injected steam–water–oxygen pattern from each injection well. • Determining how much mixingoccurred. • Determining oxygen consumption,carbon dioxide production, and transport.

Visalia Cleanup10

Figure 6. Gene Kumamotoand Robin Newmark takemeasurements in Livermore’smini-laboratory, which housesa mass spectrometer, gaschromatograph, and otherequipment.

00.0

0.2

0.4

0.6

0.8

1.0

1.2

5 10 15 20Hours since steam turned on

Fra

ctio

n of

initi

al x

enon

con

cent

ratio

n, %

25 30 35 40 45

Tracer measurements NUFT simulations

Figure 7. Modeled results using the NUFT code indicate when xenon gas would appear atmonitoring wells after having been injected into the subsurface. Curves represent simulatedxenon breakthrough concentrations assuming different initial conditions. Observed xenonconcentrations (dots) reveal an initial breakthrough, then slight decrease during a drop ininjection pressure.


Modeling also effectively predicted thetime of thermal breakthrough, whichoccurs when sufficient heat has built upin the subsurface for vaporization ofcontaminants to begin, and steamcollapse, which is the oppositephenomenon.

The team found that the ratio oftracer gas to natural air mixed intowater was much greater than predictedby the model’s initial assumption of nomixing of the atmosphere and steamzone. These data demonstrate thatmixing is important and the process ismore efficient than envisioned.

From Liability to AssetIn short order, just months after

laboratory experiments werecompleted, the new hydrous pyrolysis/oxidation method succeeded at theVisalia site. The project team hadbrought together Livermore’s expertisein underground imaging, noble-gas-tracer monitoring, supercomputermodeling, and accelerator massspectrometry to create and verify thefield results of a technology totransform the groundwater and soilcleanup process. Far faster than othertechniques, the technology provides arelatively inexpensive way to clean updifficult contaminants that plaguedozens of sites across the country. Fortheir efforts, the team was recognizedwith the Laboratory Director’sPerformance Award in December 1997.

The project team was mindful of theneed to make the techniques simple forothers to operate and maintain.Integrated Water Technologies of SantaBarbara, California, recently becamethe first nationwide licensee ofLivermore’s new cleanup technologies.The company plans to begin using themthis year to clean up several Superfundsites.

Work at Visalia is not yet complete.The best estimates today are thatcleanup will be completed in a year,with another four years of monitoringthe site. Southern California Edisonhad expected to meet EnvironmentalProtection Agency requirements inabout 120 years with traditional pump-and-treat technology combined withenhanced bioremediation. Instead, apiece of real estate that had been amajor liability will soon become avaluable asset.

— Katie Walter

Key Words: dynamic undergroundstripping, electrical resistance tomography,groundwater contamination, hydrouspyrolysis/oxidation, modeling, noble-gastracers, NUFT code, remediation, soilcontamination.

Visalia Cleanup 11

ROBIN L. NEWMARK, leader of the Applied Geology andGeophysics Group in the Earth and Environmental SciencesDirectorate, has been at Livermore since 1985. Her early workfocused primarily on borehole geophysics applications. Since1990, she has been involved in the development of thermalremediation methods and subsurface detection and imagingtechniques essential for monitoring and process control of in situ

environmental remediation. She earned a B.S. in earth and planetary sciences fromMassachusetts Institute of Technology in 1978, an M.S. in marine geophysics fromthe University of California at Santa Cruz in 1980, and a Ph.D. in marine geophysicsfrom Columbia University in 1985. Author of many papers, reports, and patents,Newmark is associate editor of Geophysics, the journal of the Society of ExplorationGeophysicists.

ROGER D. AINES came to the Laboratory in 1984 and iscurrently leader of the Geochemistry Group in the Earth andEnvironmental Sciences Directorate. Since 1990, he has beenworking on the development of thermal remediation methods.Earlier, he gained experience centered on geochemical researchand modeling for the Yucca Mountain nuclear waste repositoryproject. Roger received a B.A. in chemistry from Carleton College

in 1978 and a Ph.D. in geochemistry in 1984 from California Institute ofTechnology. He is an author of many papers, patents, and reports related to disposalof nuclear waste and thermal remediation of contaminated groundwater.

About the Scientists

For further information contact Robin Newmark (925) 423-3644([email protected]) or Roger Aines(925) 423-7184 ([email protected]).

ARADYN, a breakthroughcomputer code for parallel

machines, was ready for action—thereal action that only the most powerfulof supercomputers could provide. Actionbegan late last year when the IBM SP2arrived, the first of a series of massivelyparallel supercomputers for LawrenceLivermore. Within days, ParaDyn wasloaded on the machine and used tosimulate, for the first time, a one-million-element model of a groundshock—a very large problem. The firstsimulation immediately demonstratedthe machine’s great power and potential,as well as the efficacy of ParaDyn.

New codes like ParaDyn are animportant component of Livermore’seffort to ramp up to 100-teraflops(100 trillion floating-point operationsper second) power, requisite to solvingproblems for DOE’s AcceleratedStrategic Computing Initiative. ParaDynis the parallel-computing version ofDYNA3D, one of Livermore’s premierecodes for modeling and predictingthermomechanical behavior.

The development of ParaDyn,ongoing for the past six years, wasperformed not just in anticipation ofparallel computing needs, but becauseof the forward momentum provided bycomputational science in the MethodsDevelopment Group of LawrenceLivermore’s Engineering Directorate.The group was formed in the 1970s todevelop modeling tools that were

P

ComputationalMechanics Moves AheadComputationalMechanics Moves Ahead

Powerful computer codes, capable

of performing complex structural

and thermal analyses, are changing

engineering work. New capabilities

enable problem solving at an

unprecedented scale.

Powerful computer codes, capable

of performing complex structural

and thermal analyses, are changing

engineering work. New capabilities

enable problem solving at an

unprecedented scale.

12

13


critically needed by Laboratory nuclearweapons projects but were commerciallyunavailable (see box, p. 17). The broadcapabilities of these tools—for analyzingthe complex system deformations andnonlinear material interactions—madethem valuable commercial products.Over the next 20 years, the grouppursued cutting-edge code developmentand expanded the modeling technologybase. They produced an entire family ofstate-of-the-art software for nonlinearanalysis of thermomechanical systems.

And the work continues. As long asthe Laboratory’s scientific knowledge isneeded for weapon systems and otherextremely large systems advancements,the group must continue to develop andimprove codes. The tools they producemust satisfy the Laboratory’s analysisand computational performancerequirements. Peter Raboin, leader ofthe group, puts it this way: “If we endup merely duplicating the work ofothers and, worse yet, they are solvingthe same problems faster and better,then we might as well close up the shop.”So the group forges ahead, adding to thecode family and making existing codesmore robust, more accurate, faster, andcapable of ever more applications.

Two Classes of CodesThe Methods Development Group’s

most prominent mechanical codes areDYNA3D and NIKE3D, which addressthe behavior of structures as they deform

and (possibly) fail, and TOPAZ3D,which addresses the thermal behavior ofmaterials undergoing heating or cooling.

Finite-element codes solve problemsby marching them forward step by step.There are two methods for calculatingthe solution at each new step. Onemethod, called explicit, relies only ondata from previous steps. Its equationsare not coupled to other variables,thereby keeping to a minimum theamount of computer memory and workrequired for a solution. DYNA3D, anexplicit code, has the ability to handle alarge number of mesh elements becauseof smaller memory requirements thanthe counterpart implicit method.DYNA3D is suitable for analyzingrapidly changing events—a car collision,for example. However, because explicitmethods require very small time steps,DYNA3D is suitable for calculatingonly brief phenomena (typically lessthan one second in duration).

NIKE3D and TOPAZ3D are implicitcodes. At each step, the unknown futurevalues of all variables must be solved.The equations thus become coupled in aset of simultaneous equations, which arestored in matrix form and must besolved to reach the next step. Solvingthese sets of equations requiresconsiderable computer memory, therebylimiting the number of elements that canbe solved by this method. However,very large steps in time can be taken bythe implicit method compared with

those taken by the explicit method,allowing events of very long duration tobe modeled (typically seconds to yearsin duration). In addition to dynamicanalysis, implicit codes such as NIKE3Dand TOPAZ3D can be used to simulatequasistatic events, such as the stresses ina bridge as a car drives across, and staticevents, such as stress in a bridge underits own weight.

Over time, DYNA3D, NIKE3D, andTOPAZ3D have been improved andapplied to an ever-widening range ofengineering analyses. DYNA3D is alsowidely known and used throughoutindustry, as a result of its disseminationthrough a collaborator’s program (seeEnergy & Technology Review,September–October, 1993, pp. 1–5).New applications for all of these codes are developed nearly as fast as thecomputers available to handle theiranalyses.

Enhancing DYNA3DDYNA3D, suitable for solving

problems involving rapid change, hashad many applications in safety analysis.Laboratory analysts have used DYNA3Dto study crashworthiness in a number ofvehicle safety studies, where models ofcomplex vehicles impact roadside safetystructures and other vehicles, deformingunder the impact. Now, on a project ledby physicist Rich Couch, in collaborationwith the Federal Aviation Administrationand Boeing, AlliedSignal, and Pratt &

The model required for simulating afan-blade failure is a complex one. Thespinning rotor body inside the jet enginemust first be simulated in its steady-state operation. The engine model thenundergoes rapid change as a fan bladebreaks off, hits other blades, and throwsthe engine rotor off balance, causingpossible critical vibrations in the entireaircraft. The simulation required manyimprovements in DYNA3D, which weredesigned and implemented by EdwardZywicz, a computational mechanicsspecialist in the Engineering Directorate.

To construct the model, computationalanalyst Greg Kay started with an enginemesh that had been used previously tosimulate blade breakoff but had obtainedthe wrong results. The model’s simulationhad disagreed with observed test data by

14


Computational Mechanics

overestimating the consequences of fan-blade failure and predicting the loss of asubstantial portion of the fan-bladeassembly. The errors helped Kaydetermine what improvements wereneeded for the mesh and how DYNA3Dshould be modified and enhanced forthis problem.

For the spinning rotor-body model,Kay performed a series of analyses todetermine how stress was distributedover the fan-blade assembly and howit was affected by rotation, bladebreakoff, and an unbalanced shaft.Before constructing a full spinning-body model, Kay first experimentedwith a simplified shaft-and-bladeassembly with the same radialdimensions and material properties asthe more detailed engine model. Hesimulated the spinning of thissimplified, eight-blade engine for tenrevolutions; this told him whetherDYNA3D programming predicted thecorrect load stresses in the spinningblades. The DYNA3D simulation of thefull-blade assembly (Figure 1) showsthat the modeled circumferentialstresses of steady-state engineoperation are consistent for a curvedfan-blade assembly.

The depiction of the blade fragmentsinteracting with each other, as well aswith following blades and the enginecontainment casing, posed specialchallenges. These “contact” problemsrequired accurate mesh details at theinteraction points, to more specificallycalculate the material interactions there.Automatic contact algorithms wereused for this large simulation. For hismodel, Kay explicitly defined thecontact surfaces, but recentimprovements in the code’s ability toautomatically define interacting partswould make that task easier today.

Whitney, DYNA3D is being applied toa different kind of safety evaluation.Couch and his team are studying thedynamics of fan-blade failure in a jetengine and its effect on the airframe.

When an engine fan blade breaksoff, it generates fragments and debristhat could further damage the engine orthe airframe unless they are confinedby the engine’s containment system.Although such breakage is rare, theFederal Aviation Administrationnevertheless will not certify a new jet-engine design unless engine tests, inwhich a fan blade is intentionallybroken during operation, show that the engine casing can contain thefragments. Modeling with DYNA3Dreduces the number of expensive testsneeded for certifying a new design.

Effective stress,megapascals

207

0

Figure 1. The distributionof stresses in the rotatingblades of a jet engine operating in asteady state is shown as modeled by DYNA3D.

15



Figure 2 shows the engine approximatelyone-third of a revolution after the fanblade had broken off.

The simulations made by Kaydemonstrate that the DYNA3D codecan provide high-confidence simulationsof blade failure. Based on thosesimulations, additional tasks, with thehelp of partners AlliedSignal and Pratt& Whitney, were begun in late 1997 tosimulate the response of enginecontainment casings to the blade breakageand subsequent debris collisions.

NIKE3D for BiomechanicsJoint-replacement surgery, an

increasingly common medicalprocedure, has been hampered by theinadequacies of today’s prosthetic jointimplants. Replacement prosthesessometimes do not reproduce normaljoint movement and, because they havehigh failure rates, often cause theirrecipients to undergo additional costly,painful surgery. Attempts to improvethe prosthetic designs have been basedon limited information garnered fromtesting and failed implants. To obtainbetter information, implant designershave needed a way to quantitativelyassess the stresses and loads on animplanted prosthesis and the wear andtear on its fabrication materials.

Nonlinear, three-dimensional, finite-element modeling is a powerful tool forperforming such quantitativeassessment. NIKE3D, developed forstudying dynamic, finite deformations,can model the behavior of joint tissuesand bones subjected to different loadsand joint movement with and withoutprosthetic implants. It is being used byKarin Hollerbach, Scott Perfect, andElaine Ashby, a team of scientists fromLivermore’s Engineering andComputations directorates, to model

both prosthetic implants and humanjoints. Data from the prostheticsimulations are providing usefulinformation for prosthetic design and,when integrated with data fromsimulations of human joints, willprovide information for evaluating theperformance of implanted devices.

Models of implants are based oncomputer-aided design (CAD) drawingsof the prosthesis, from which the finite-element meshes are made. The modelsinclude numerical data that define theproperties of the implant materials—

including polyethylene and severalmetal alloys in the case of the prostheticdevices being studied—and the forcesacting on the joints. For the thumb-jointmodels that the researchers haveconstructed (Figure 3), the loadsproduced during commonly used grasps(the key pinch, screwdriver grasp, andtip pinch) were used as the basis forassessing the performance of threethumb-joint designs. (See S&TR,September 1996, pp. 19–21, for moreinformation on this project.) The stresspredictions from the finite-element

Figure 2. Simulation of a jet-engine blade breaking off. The engine is shown one-third of arevolution after the initial blade release.

16



tomography (CT) to acquire geometricdata and bone surface definitions. Tothese data they added magneticresonance imaging (MRI) data fromcollaborators that delineate soft jointtissues. Together, CT and MRI datadefine how bones and tissues attach toeach other. Their resulting images wereconverted into three-dimensionalsurfaces using specially developedprocessing software, and those three-dimensional surfaces were then meshedfor finite-element analysis (Figure 4).

The researchers are currentlyworking to integrate the data from the

human and prosthetic joint models.These integrated models will be used tocalculate the bone stresses resultingfrom the loads exerted by the implantsand make bone–implant interfaceanalyses possible.

TOPAZ3D for Laser HeatingBefore scientists at the National

Ignition Facility (NIF) begin performinglaser experiments, they must fullyunderstand how the various NIFcomponents will behave in operationso they can design their experimentsaccordingly. One bit of crucialinformation they must know is how theNIF target chamber will expand andmove as it heats up over the course ofrepeated laser shots. Attached allaround the chamber are series of lensesthat focus laser beams onto the target.

Design analyses predicted thesteady warming of the chamber and theequilibrium temperatures it would reachduring various operating scenarios; thedesign of the lens assembly around thechamber allowed for these temperatureincreases. However, other transienttemperature effects also must be takeninto account, in particular, the periodic

simulations were correlated againstclinical findings. The agreementbetween simulations and clinicalfindings obtained by the researchersvalidates the modeling approach andpinpoints the most failure-pronedesigns. Thus, resulting modeling datacan be used for selecting availableimplants as well as designing new ones.

For the human joint models, theresearchers had to develop a process foracquiring human data and converting itinto a form suitable for finite-elementanalysis. They started by scanning a human joint, using computed

Figure 3. Livermore-developed NIKE3D is used forfinite-element analyses of thumb carpo–metacarpaljoint implants. Colored areas show reaction forcesexperienced by the implants as they would grasp ahouse key.

Contact stress,megapascals

40

30

20

10

0

Asymmetric saddle design

Symmetric saddle design

17



temperature spikes caused by individuallaser shots.

These temperature spikes cause arapid thermal expansion of the targetchamber. The chamber then slowlycontracts, back to its equilibrium shapein time for the next laser shot. Theconcern is whether contractions affectthe focus of the lenses. Just before alaser shot, these lenses must be in theirpositions to within tolerances of only afew millionths of a meter.

Wayne Miller, a thermal analyst, wascalled on to model and simulate thetemperature changes of an operatingtarget chamber. He used the heattransfer code TOPAZ3D, developed byArt Shapiro of the Thermo-FluidsGroup, in conjunction with NIKE3D tocreate an analysis model.

The target-chamber model consistsof an aluminum inner shell, a concreteouter shell, and aluminum portsthrough which laser beams aredelivered. First, the finite-elementgeometry was created using TrueGRID,a commercial mesh-generation code.Then TOPAZ3D was used to predictthe transient thermal behavior of themodel, primarily due to laser energydeposited on the inner wall surfaces,which is then conducted through bothshells and convected to the outside air.The thermal results were then given toNIKE3D, which predicted the thermalexpansions of the chamber andprescribed the desired motion of thelaser lenses for the next shot (Figure 5).

The composite model was used tosimulate thermal changes occurring inthe chamber when laser shots are firedonce every 4 or 8 hours. The 4-houroperating sequence, more challengingto the thermal stability of the chamber,was simulated under various conditions:

Structural Problems, Computer Solutions

A Parallel DYNA3D (ParaDyn) simulation of weapons is only one example offinite-element analysis of structural behavior. Other finite-element problems includesimulating car crashes and train accidents, falling nuclear waste containers, ground-shock propagation, aircraft-engine interaction with foreign debris, metal forming,biomedical interactions, and component designs for cars and aircraft.

The physics of structural behavior can be expressed in mathematical equationsthat can be translated into problems practicable for solving on the computer. Of themany computational techniques developed and used for solving structural mechanicalproblems, the one that became the most widely adopted by the 1970s—and the onethat Laboratory computational scientists have advanced—has been the finite-elementmethod.

In the finite-element method, a complex, solid object is divided into anassemblage of simple elements that become the basic units on which the computercalculates the structural behavior. Each element can be made as small and asirregularly shaped as needed to model the object being analyzed. Visually, thecollection of elements resembles a wire mesh; the element boundaries are defined bylines that intersect at junctions called nodes. The nodes at each element corner definethe movement and deformation of the elements.

The powerful versatility of finite-element codes is best exemplified in thenumerous nonlinear material behaviors they can model: elasticity, plasticity,viscoplasticity, viscoelasticity, hyperelasticity, viscous flow, granular flow,thermomechanics, thermodynamic equations of state, composite behaviors, nonlinearfoams and solids, void growth, and evolutionary damage accumulation. Simulatingall of these behaviors is possible because the finite-element method uses constitutiveequations to predict the element stress state based upon changing elementdeformations.

Figure 4. NIKE3D simulated theflexing movement of an indexfinger by using information frommagnetic resonance imaging andcomputed tomography.


with laser energies deposited uniformlyand nonuniformly, with differentoutside air temperatures, and withdifferent values for heat transfer fromthe chamber to the environment.

Because one major goal of the overallanalysis was to evaluate two ways ofcooling down the chamber, simulationswere performed with boundaryconditions for passive cooling, in whichheat transfers out through radiation andconvection, and active cooling, anddepends on the installation of a coolingsystem (Figure 6).

The effects of the target chamberpedestal were also studied. A separatemodel provided data on how thispedestal grew taller and shrank backdown and how this change affected thechamber and ultimately the focusingangle of the lenses.

The results from the simulations forthe different operating scenarios, cooling-down options, and pedestal influenceswere evaluated together to provide detailson lens displacements—tangential,radial, or rotational—under differenttime and temperature conditions.

DYNA3D Becomes ParaDynThe availability of massively parallel

supercomputers, such as the IBM SP2,means that analysts now can solveproblems larger by one or two orders ofmagnitude than was possible with otherexisting supercomputers. It is nowpossible and practical for engineeringanalysis codes to contain meshes withup to ten million elements. Now thelimiting factor for problem size andcomplexity is the amount of time thatanalysts need to design the mesh ratherthan the raw speed of the computers.

That boost in computing powerdepends on capable, efficient parallelcodes. Because of the work of a code-

Computational Mechanics18

Figure 5. TheNIKE3D andTOPAZ3D codesanalyze the expansion ofthe National IgnitionFacility’s laser target chamberresulting from the heat of a laser shotand its contraction to equilibrium in the cool-downperiod.

Figure 6. (a) Passive and (b) active coolingmethods for the target chamber wereanalyzed with a combination of the NIKE3Dand TOPAZ3D codes.

Aluminum

Gunite(b)

(a)


developer team led by Carol Hoover,such a code, ParaDyn, is now available.As a result, parallel computing in solidand structural mechanics is becominga powerful tool in computationalengineering.

To take advantage of the power ofparallel computers, the mesh for a largeproblem must be partitioned so that thecalculations for each part may beperformed in parallel by the processors.The more processors working on theproblem, the faster the solution isobtained. Multiprocessor calculationsdepend on the computer program toappropriately divide a problem, tologically sequence its calculations, andto allow communication among theprocessors so that their solutions can beintegrated into a final correct answer.

For computing efficiency, theworkload must be evenly distributedamong processors so that no processorsare waiting on others. Furthermore,calculation time and processorcommunication time must be balanced;the latter must be limited to a smallfraction of the overall computing time.Thus, for optimal efficiency andperformance, a problem must bedivided using a method that minimizescommunication among the processors.

The task of partitioning a problemmay be needed more than once at thebeginning of a parallel calculation.Because the problems are dynamic, themesh may change significantly duringthe evolution of the deformation,affecting the parallel efficiency. Whenthis happens, repartitioning the meshand boundary conditions is often

necessary. Developing methodologiesfor these repartitioning tasks ischallenging, and research is still inprogress in this area.

ParaDyn’s success has spun offseveral benefits to the weaponsengineering programs at Livermore.Calculations that previously tookseveral weeks are now performed in aday or less. New models are beinggenerated for mesh sizes between onemillion and ten million elements—anorder of magnitude larger than thelargest models possible in the past.

The Next GoalsWith the completion of ParaDyn,

the development of parallel algorithmsfor NIKE3D and TOPAZ3D hasbegun. “This effort,” Raboin explains,“is an even greater challenge.Developing efficient procedures forsolving implicit matrix equations onsingle-processor computers is alreadydifficult, so imagine the immensity of

doing so for parallel processors. That’snot to say that projects to couple theDYNA3D, NIKE3D, and TOPAZ3Dcodes to solve more complex physicsproblems will be any easier. Those arehard problems, too.”

With their experienced researchand development, the MethodsDevelopment Group is helping tochange the nature of engineering workand, in the long term, the very natureof problem solving.

—Gloria Wilt

Key Words: computational mechanics,computer modeling and simulation,DYNA3D, finite-element method,nonlinear behavior, NIKE3D, ParaDyn,parallel computing, solid and structuralanalysis, TOPAZ3D.

Computational Mechanics 19

PETER J. RABOIN, joined Lawrence Livermore in 1989, whenhe became a member of Mechanical Engineering’s AppliedMechanics Group. He has been serving as the group leader forMethods Development since 1995. Raboin received degrees inmechanical engineering from Georgia Institute of Technology(B.S., 1982) and Massachusetts Institute of Technology (M.S.M.E.,1985; Sc.D., 1989). He has published papers on thermomechanicsand finite-element method.

About the Scientist

For more information contact Peter Raboin (925) 422-1583([email protected]).

20

Corsica: Integrated Simulationsfor Magnetic Fusion EnergyCorsica: Integrated Simulationsfor Magnetic Fusion Energy

OR years, magnetic fusion energy (MFE) scientists havedreamed of an integrated, easy-to-use, and comprehensive

family of computer codes that would simultaneously simulateall of the important physics processes that take place in amagnetic fusion reactor. Such a package would be valuable forenhancing the understanding of the extremely complexphenomena observed in experiments. It would also provide atool to optimally design future MFE experiments.

Although this goal of virtual experiments is still yearsaway, researchers at Lawrence Livermore have madeimportant advances in developing such a comprehensivesimulation package. In designing such a code, called Corsica,they have developed techniques to efficiently couple separatephysics processes. These techniques plus continuing advancesin high-performance computer hardware and software offerthe prospect of achieving the goal. Corsica is only one part ofa widespread effort, evident throughout Livermore researchprograms, to simulate to unprecedented levels of accuracy thephysical phenomena taking place on scales ranging fromatomic particle interaction to global weather patterns.

In fusion, two light nuclei (such as hydrogen) combine intoone new nucleus (such as helium) and release enormous energyin the process. One approach to fusion uses a powerfulmagnetic field to confine a plasma (a gas consisting of chargedions and electrons), generating energy in a controlled manner.To date, the most successful approach for achieving controlledfusion is in a donut-shape configuration called a tokamak.

Future experimental facilities will be much larger thantoday’s research tokamaks and much more expensive, costing asmuch as several billion dollars each. Although advancedsimulations will never replace experimental work, they areneeded to optimize the design of future tokamaks andexperiments, which will save millions—maybe billions—ofdollars. Simulations are also needed to analyze and optimizealternative concepts to the tokamak, such as the small spheromakdevice that Livermore is now constructing. Scientists considersuch simulations essential to resolving several important physicsissues, such as how to structure the magnetic fields to producethe maximum pressure and what processes drive electric currentsand magnetic fields.

F The Need for Integration“Simulations of individual phenomena—physics

‘packages’—now exist as essential tools for analyzing fusionexperiments,” says Livermore’s Keith Thomassen, DeputyAssociate Director for MFE. Phenomena such as theequilibrium of the plasma, turbulent transport, stability, andheating, are examples of such processes and are interdependent.Thus, codes that simultaneously describe all these phenomenahave these packages “hard wired” together, and the codes areextremely complex. A contributor to this complexity is thedisparate time and spatial scales of these phenomena.

MFE processes span a wide range of time scales, fromturbulent fluctuations on the microsecond scale to transportprocesses with scales of seconds to hours. For example,particle velocities parallel to the magnetic field are orders ofmagnitude greater than those moving perpendicular to themagnetic field. Additionally, MFE models must reflect arange of spatial scales that extend, for example, from thespiral orbit size of an electron to the several-meter-widetokamak device.

David Baldwin, vice president for fusion research atGeneral Atomics (GA) in San Diego, was Associate Directorfor Energy Programs at Livermore in the early 1990s whenCorsica development began. “The MFE community had donea very good job in developing discreet physics packages buthad little experience integrating them. We wanted to providethat capability for the first time but in modular units,” he says,“so that, as individual physics packages improved, they couldbe exchanged. But the entire code would not have to berewritten.”

The inspiration for Corsica, says Baldwin, was Livermore’slong-standing LASNEX integrated code for laser fusion thatsimulates interactions between laser light and its targets. Healso notes that Corsica’s objective of integrating all thephysics phenomena is analogous to that of the AcceleratedStrategic Computing Initiative (ASCI), the Department ofEnergy’s effort in the Stockpile Stewardship Program todevelop full-system simulations running on new generationsof high-performance computers. Indeed, several LivermoreCorsica developers are also contributing to ASCI.

Research Highlights


21


Corsica

CORSICA CODE

Coretransport

1.05

1.00

.95

.90

.85

.80

.75

.70

.65

.60

.55

.50

.450 5 10 15 20 25 30 35 40 45 50 55 60

2200 cycles

200 cycles

Initial temperature

1

3

Tem

per

atu

re

Position

2

Simulation code required over 10,000 time stepsto approach this solution

core

edge

00 10 20 30

Minor radius, meters40 50 60

200

600

1000

1400

1800

2200

2600

3000

Ele

ctro

n t

emp

erat

ure

, kilo

elec

tro

n-v

olt

s

Thomsonscattering

After

Dots = experimental dataLines = simulation

Before

Heating etc...

Current drive Engineering

Fueling Synthetic diagnostics

Macroscopicstability

-1.00

-1.10

-1.20

-1.30

1.30 1.40 1.50 1.60Radius, meters

From experiment

Hei

gh

t, m

eter

s

-1.40

-1.20

-1.00

-0.80

-0.60

1.00 1.20 1.40Radius, meters

UEDGE simulationH

eig

ht,

met

ers

1.60 1.80

Edge transportUEDGE

Turbulencesimulation

Numerical tokamak

III

III

Axisymmetricmagnetic

configuration

The Corsica magnetic fusion simulation code is a prototype for an integrated simulation that would solve models for all aspects of tokamakoperations. Roman numerals in the center graphic indicate successive Corsica code releases; images demonstrate their capabilities.

(1) Initial and (2 and 3) converged plasmashapes with self-consistent turbulence.

High-resolution visualization oftokamak turbulence.

Temperature before and after transitionfrom low- to high-confinement regime.

Two-dimensional electron–temperature data.

Grid for coupledcore–edge transportcode.

22


Corsica I, released in 1994, flexibly coupled one-dimensional calculations of particle, energy, and magnetic-flux transport in the core, or confined region, of the plasma toa calculation of a two-dimensional magnetic configuration.Corsica II, released for testing by sophisticated users in 1995,coupled the one-dimensional core transport calculation to asimulation of the two-dimensional “edge” where magnetic-field lines intersect material surfaces.

A still more advanced version, Corsica III, is beingdeveloped as funding permits. Its ambitious goal is to couplethe evolution of the fusion process to plasma turbulence modelsfor a more comprehensive simulation. The turbulence couplingeffort benefits from the sheer computing power of the newestsupercomputers as well as from the experience in turbulence ofLivermore’s MFE theory group. The group participates in theDepartment of Energy’s Numerical Tokamak TurbulenceProject, designated a DOE grand challenge.

By handling a wide range of physics with disparate timeand spatial scales, Corsica permits more complete modeling oftoroidal (donut-shaped) plasmas and other magnetic fusionconcepts than previously possible. As a result, the softwarehas been adopted at both national and international researchsites. For example, Corsica is used to model plasmaconfinement in the DIII-D experimental tokamak at GA,where a team of Lawrence Livermore scientists is working.Corsica was used on simulations of the Tokamak Fusion TestReactor that operated at the Princeton Plasma PhysicsLaboratory. Corsica has also been used to model the designfor the International Thermonuclear Experimental Reactor andLivermore’s Sustained Spheromak Physics Experiment.

Corsica Keys on FlexibilityLivermore physicist Tom Casper uses Corsica to replicate

experiments conducted on GA’s DIII-D tokamak. He saysCorsica helps him obtain a better understanding of pastexperiments and plan future experiments. He points to thecode’s flexibility as one of its strong suits. “I can modify thecode and manipulate variables, capabilities we don’t find inother codes.”

In addition, Casper notes that budgets and space constraintslimit the number of diagnostic instruments that can be used onthe DIII-D tokamak. With Corsica running on a powerfulworkstation, Casper can add as many so-called syntheticdiagnostics as he desires and thereby gain a more completepicture of an experiment.

Corsica’s flexibility is provided in part by BASIS, a codesteering system developed by Livermore computational

scientist Paul Dubois in the 1980s. BASIS also served as theunderlying system for LASNEX. Leaders of both the Corsicaand ASCI programs have recently chosen Python as thesuccessor to BASIS because of its greater flexibility and itsability to run on many different computers.

Baldwin says Corsica today is “darn good” but still farfrom the complete simulation code that he and other leaders ofthe MFE community envision. For that to happen, he says,other major MFE centers need to join forces with Livermoreand GA and integrate their own specialized codes, as well astap the experience of other comprehensive code efforts suchas ASCI.

“We want to create a national project that benefits from ourexperience here,” says Ron Cohen, leader of Livermore’sMFE theory and computation program. Cohen and others haveproposed a “national transport code,” a modern code thatwould draw upon the concepts in Corsica and the vast wealthof physics simulation “building blocks” available nationally.Cohen notes that because the family of MFE simulation codeswas created at different research centers in different softwarelanguages and on different computers, the effort to modifythem and then combine them into a seamless, integratedpackage will demand strong cooperation among researchers.

In the meantime, Livermore scientists are collaboratingwith colleagues at GA to combine the best features of GA’stransport code with those of Corsica. They hope to carry out their project within the context of the national transportcode project.

If the national effort is successful, users will be able to selectfrom many different physics “plug-in” modules to build customsimulations of various—or all—aspects of a magnetic-fusionexperiment (tokamak or other configuration). The integratedcode would also allow a user to modify experimental conditionsand theory parameters or even whole models in real time,without having to use elaborate commands.

“We want to have an integrated, ‘living’ code that keepschanging and growing as computer capabilities and ideasadvance,” says Cohen. While it could never serve as asubstitute for an actual experiment, such an integratedsimulation would undoubtedly save money by enabling better-designed and more aggressive experiments. “By combiningdata from this code with hardware experiments, we’d get themost out of the research dollar,” he says.

— Arnie Heller

Corsica

For further information contact Ronald H. Cohen(925) 422-9831 ([email protected]).

23

Device Assembly Facility: New Facilitiesfor Handling Nuclear ExplosivesDevice Assembly Facility: New Facilitiesfor Handling Nuclear Explosives

S the 18-wheeler slowly eased up to the loading dockunder a heavily armed escort, massive steel doors opened

from the inside of a partially buried structure. Anannouncement was broadcast over the public address systemalerting specially trained personnel of the arrival of a safe andsecure transport vehicle. This special big rig is part of a fleetused by the U.S. government to transport nuclear weapons orweapon components from one secure site to another. Technicalspecialists stood ready to unload the shipment.

The scene of this activity at the Nevada Test Site is theDepartment of Energy’s newest facility, located 90 milesnorthwest of the famous Las Vegas strip in a remote part ofthe Nevada desert. The recently opened Device AssemblyFacility, or DAF, offers one of the safest, most securelocations anywhere in the U.S. weapons complex to conductnuclear explosive operations. Other than the Pantex Plant inAmarillo, Texas, the Nevada Test Site is the only location inthe country where special nuclear material such as plutoniumcan be mated with high explosives.

Under a unique arrangement with theDepartment of Energy, LawrenceLivermore and Los Alamos nationallaboratories are designated jointoperators and users of the facility. Sincethe facility’s inception, the laboratorieshave collaborated on every aspect of thefacility—including its design,construction, certification, management,and use. Management responsibility isrotated between the laboratories,nominally every two years. “This is trulya joint-lab facility,” explains JamesPage, DAF startup team leader forLivermore. “We share commonprocedures and integrated hazardsanalyses and operate with a managementteam composed of personnel from bothlaboratories.” In contrast, nuclearexplosive operations at the Nevada Test

A Site in the past were conducted in two separate facilitiesaccording to Laboratory affiliation.

Adapting to ChallengesDAF dates from the mid-1980s, when the weapons

laboratories were engaged in active nuclear testing. As withmany aspects of the Department of Energy’s weaponsprogram, DAF has been adapted to a changing environmentbrought about by changes in national nuclear testing policy.Designed and built for the purpose of assembling the twolaboratories’ nuclear test devices prior to placing themunderground for testing, the new facility retains its originalname. Its mission, however, has evolved since the nucleartesting moratorium began in October 1992.

Instead of the underground test assembly work for whichthe facility was originally intended, DAF will accommodateother hands-on activities involving high explosives, special

Research Highlights


��

��

��

��

��

��

��

0 20 40 60 ft

Assembly cells

Radiography bays

Assembly bays and high bays

Staging bunkers Shipping/Receiving docks

Laboratories

Vaults Decontamination rooms

Administration

��

��

��

��

��

��

��

��

��

��

��

Compacted earth

Schematic of the main floor of the Device Assembly Facility at the Nevada Test Site.

24


Device Assembly Facility

The bunker-style Device Assembly Facility has about 10,000 squaremeters of floor space in its 30 steel-reinforced concrete buildings.

nuclear material, nuclear weapon components, and nucleardevices. These projects, an integral part of DOE’s StockpileStewardship Program, will include assembly work to supportnew subcritical (no-nuclear-yield) experiments beingconducted nearby as well as stockpile maintenance activities,such as enhanced surveillance technology development andpersonnel training. These activities will provide necessary datato help maintain the nation’s nuclear deterrent. Future usecould also include DOE activities to maintain a nuclearemergency response capability.

Anatomy of the FacilityIn a bunker-style arrangement, DAF is a collection of

30 individual steel-reinforced concrete buildings connected bya rectangular racetrack corridor. The entire complex, coveredby compacted earth, spans an area the size of eleven footballfields (see schematic drawing on p. 23). Because the buildingswere designed to accommodate potentially hazardousoperations, they meet the most stringent set of safetyregulations.

Explains Page, “This facility was intended to provide thesafest possible environment for conducting hands-on operationsinvolving special nuclear material and high explosives. TheDAF design incorporates the most modern safety and securityfeatures in the weapons complex.”

The DAF multistructure occupies about 10,000 square metersof usable floor space and consists of operational buildings for

work involving high explosives and special nuclear material andsupport buildings for laboratories and offices.

Most isolated of the operational buildings are five assembly“cells” for activities involving uncased conventional highexplosives and special nuclear material. Four high bays andthree assembly bays provide facilities for less hazardousoperations, such as those involving uncased insensitive highexplosives. When radiography is required to verify theintegrity and spatial relationships of objects, a 9-megaelectron-volt, movable-beam, linear accelerator is available in one oftwo radiography bays. Five staging bunkers provide amplespace for interim storage of nuclear components or highexplosives. Finally, all materials packages arrive or departDAF through either of two shipping and receiving bays.

The support buildings include three small vaults for storingsmall quantities of high explosives or special nuclear material;two decontamination areas; and an administration areacontaining office space, a conference area, personnel changingand shower rooms, and a machine shop. In addition, twobuildings provide laboratory space: one for conductingcomponent, instrumentation, and environmental testing andthe other for controlling remote operations of an adjacentassembly cell.

Blast Protection and ContainmentTo provide the utmost protection for personnel working

within the facility and for the environment, DAF employsseveral state-of-the-art safety features. Chief among them areblast doors on all operational buildings to mitigate thepropagation of an accidental explosion, blast-actuated valveson the ventilation system to prevent the spread ofcontamination, special ventilation features such as zoned air-supply systems and high-efficiency particulate air filters, andthe unique design of the assembly cells.

Indeed, the assembly cells are whimsically called “gravelgerties,” after a 1950s Dick Tracy comic-strip characterbecause the roof is overlaid with nearly 7 meters of gravel,said to resemble the original Gravel Gertie’s thick curly grayhair. Modeled after the structures at Pantex, where hands-onassembly and disassembly of U.S. nuclear weapons takeplace, they provide the maximum environmental andpersonnel protection in the event of an inadvertent high-explosive detonation. The cells are designed to absorb theblast pressure from a detonation of up to 192 kilograms ofplastic-based explosives (equivalent to 250 kilograms of TNT,or approximately one-fifth of the explosive energy released inthe World Trade Center bombing). Should a detonation occur,

25


Device Assembly Facility

the gravel gertie would minimize environmental release ofnuclear material and propagation of the event to other areas inthe facility.

“Although it is extremely unlikely that the cells would everbe required to perform to their full potential,” says Page, “theirdesign provides the extra assurance necessary for moderatelyhazardous activities, such as weapon assembly anddismantlement and some processes for monitoring changes dueto aging. In addition, the assembly cells could be used to disablenonstandard explosives, such as a clandestine nuclear device.”

A National ResourceThe design of the facility, its remote location, and its safety

features make DAF well suited to address new national

challenges—both predicted and as yet unforeseen—inmaintaining the nuclear stockpile. Like Livermore’s LaserPrograms, started in the 1960s to research electrical power butnow studying fundamental physics issues as well, DAF is avaluable national resource whose future applications willlikely extend far beyond its original mission.

“DAF is the first facility of this complexity,” sums upLivermore physicist Willy Cooper, Nevada Experiments andOperations Program Leader, “for handling unique andpotentially hazardous nuclear weapons components andoperations that DOE has brought online in several decades.With its 21st-century focus on environmental, safety, andhealth considerations, DAF is a functional testimony to thevision of the former generation of weapons scientists, validtoday even with a shift in strategic national policy.”

— Lori McElroy

Key Words: Device Assembly Facility (DAF), DOE, highexplosives, Nevada Test Site, nuclear-explosive operations, nuclearweapon assembly and disassembly, stockpile stewardship, subcriticalexperiments.

A “gravel gertie” assembly cell during and afterconstruction. The assembly cells weredesigned to provide the utmost protection tothe public, nearby workers, and theenvironment in the event of an inadvertentexplosion.

For further information contact James Page (925) 423-1195 ([email protected]).

26


Lunar ice discovery confirmedNASA has confirmed the 1994 discovery of lunar ice

made during lunar mapping by the Clementine I satellite.Lawrence Livermore physicist Stewart Nozette proposedthe novel idea of using Clementine’s communicationstransmitter as a radar to test a theory that ice crystals mightbe trapped on the moon.

Scientific instruments, including a neutron spectrometeraboard NASA’s Lunar Prospector, launched in January1998, confirm that at both lunar poles moon dirt is mixedwith ice, enough to support 2,000 people for a century.This means the moon could be used as a springboard tospace travel.Contact: Stewart Nozette (925) 424-4964 ([email protected]).

Reno addresses computer security workshop“Protecting critical infrastructures and assuring their

continued operation” is the central mission of the President’sCommission on Critical Infrastructure Protection (PCCIP).Recent workshops put on by Livermore’s Center for GlobalSecurity Research and Stanford University’s Center forInternational Security and Arms Control (CISAC) developedpriorities for the commission’s recommendations.

Speaking in Livermore at the last of three workshops,Attorney General Janet Reno gave the keynote address andin it unveiled the National Infrastructure Protection Center,an initiative with a mission to detect, protect, and respond tocyber attacks on the infrastructure. The Center will be anoutgrowth of the FBI’s Computer Investigations andInfrastructure Threat Assessment Center. The initiative callsfor a partnership of federal agencies, private industry,academia, and national laboratories.

(Continued from p. 2.)

Also attending the conference were former DefenseSecretary William Perry, General Tom Marsh, chair of thePCCIP, and Michael May, former Lab director and codirectorof CISAC.Contact: Ronald Lehman (925) 423-3711 ([email protected]).

Radio system becomes modelIn acceptance for use of the Lab’s Multi-Site Trunked

Radio Project, beginning in 1994 and ending in September1997, DOE recently recognized the project for being the firstnarrowband digital trunked radio system in the federalgovernment. The network expands radio communications butuses one-half the bandwidth of previous systems and retainsencryption. Because the communications are digital,conversations can no longer be monitored by scanners.Maximum spectrum efficiency is achieved by combiningreduced bandwidth and wide-area trunking technology. Thetrunking process manages channel resources by dynamicallyallocating radio frequencies to users as needed instead ofallowing channels to sit idle during periods of low activity.Greater system management and control are also madepossible because of the centralized computer network. Thenew system is now a model for the DOE complex and otherfederal facilities.

Project manager Clarence “Pete” Davis likens thetrunking process to that of waiting in line at a bank for thenext available teller—a computer quickly links the radiooperator with the first available radio channel. “We’ve metthe new federal regulations for bandwidth reduction, whileeliminating channel congestion and increasing privacy andcontrol,” says Davis.Contact: Pete Davis (925) 422-0370 ([email protected]).

The Laboratory in the News

27


Each month in this space we report on the patents issued to and/orthe awards received by Laboratory employees. Our goal is toshowcase the distinguished scientific and technical achievements ofour employees as well as to indicate the scale and scope of thework done at the Laboratory.

Patents and Awards

Patent issued to

Fred MitlitskyJoel B. TruherJames L. KaschmitterNicholas J. Colella

P. Henrik Wallman George E. Vogtlin

John Thomas FeoDavid Carlton HanksThomas Arthur Kraay

Patent title, number, and date of issue

Fabrication of Polycrystalline ThinFilms by Pulsed Laser Processing

U.S. Patent 5,714,404February 3, 1998

Diesel NOX Reduction by Plasma-Regenerated Absorbent Beds


Method for Data Compression byAssociating Complex Numbers withFiles of Data Values


Summary of disclosure

A method for fabricating polycrystalline thin films on low- or high-temperature substrates. Processing temperatures are low enough to avoiddamaging the substrate. Selected layers of the thin films are transientlyheated with at least one pulse of a laser or other homogenized beamsource, allowing recrystallization and/or dopant activation. This method isparticularly applicable for fabricating solar cells.

A process for reducing NOX and particulates from diesel engine exhaust byusing plasma-regenerated absorbent beds. NOX is alternately absorbedonto one of two solid-absorbent beds that act as physical traps forparticulates. Periodically, a plasma is discharged across the bed forregenerating the absorbing material, followed by air oxidation of thetrapped particulate matter. The process can be used for mobile andstationary diesel engines as well as for lean-burn gasoline engines.

A method for compressing data for storage or transmission with few lossesand high compression ratios. Given a complex polynomial and a valueassigned to each root, a root-generated data file is created one entry at atime. Each entry is mapped to a point in a complex plane. An iterative root-finding technique is used to map the coordinates of the point to thecoordinates of one polynomial root whose value is assigned to the entry.An equational data compression (EDC) method reverses the procedure.The EDC method uses a search algorithm to calculate a set of complexnumbers and a value map that will generate the target data file. The errorrate between a simple target data file and generated data file is typicallyless than 10%.

Patents

Former editor of Science & Technology Review Ravi Upadhye hasbeen elected a Fellow of the American Institute of ChemicalEngineers. Upadhye has been active in the organization, serving aschairman of local chapters and recipient of the society’s ProfessionalProgress Award for his work in molten salt destruction of highexplosives.

For innovative use of the Internet to securely exchange financialinformation and pay its bills electronically, the Laboratory has won theprestigious 1998 George Mitchell Payment System Excellence Awardfrom the National Automated Clearing House Association. Innovatorsbehind the system are John Rhodes and Chip Hatfield. Last year, theLaboratory also won an award for these efforts from the InternationalInstitute of Business Technologies.

Awards

28


They All Like It Hot: Faster Cleanup ofContaminated Soil and Groundwater

Two complementary technologies developed byLawrence Livermore are being used to clean upcontaminated groundwater and soil at a utility pole treatingfacility in Visalia, California, in 5 years instead of a planned120 years. Southern California Edison, the site’s owner, andSteamTech, the first site licensee of Livermore’s dynamicunderground stripping, began cleanup in June 1997.Livermore is also field testing hydrous pyrolysis/oxidation,a new technique for destroying contaminants in situ. Byinjecting steam and oxygen and vacuuming out vaporizedcontaminants, about 300,000 pounds of contaminants wereeither brought to the surface or destroyed in situ during thefirst six weeks of operation. That figure contrasts sharplywith the 10 pounds per week that Southern CaliforniaEdison had been removing with conventional cleanupmethods. Electrical resistance tomography (anotherLivermore development), noble-gas tracers, and analysesof extracted gases are being used to monitor steam injection,underground temperatures, fluid movement, and in situdestruction of contaminants.Contact: Robin Newmark (925) 423-3644 ([email protected])or Roger Aines (925) 423-7184 ([email protected]).

Computational Mechanics Moves Ahead

For over 20 years, the Methods Development Group,consisting of computational specialists in the Laboratory’sEngineering Directorate, has been expanding the computermodeling technology base to develop tools needed foranalyzing complex, nonlinear material phenomena. Theadvanced computer codes developed by this group supportthe highly specialized research needs of Lawrence Livermorescientists. As the codes are continually enhanced to keep pacewith the Laboratory’s scientific advances, their applicationrange has widened. The article describes some currentproblems that DYNA3D, NIKE3D, and TOPAZ3D—thepredominant methods development codes—are being usedto solve.Contact: Peter Raboin, (925) 422-1583 ([email protected]).

Abstracts

© 1997. The Regents of the University of California. All rights reserved. This document has been authored by the The Regents of the University of California under Contract No. W-7405-Eng-48 with the U.S. Government. To request permission to use any material contained in this document, please submit your request in writing to the Technical Information Department,Publication Services Group, Lawrence Livermore National Laboratory, P.O. Box 808, Livermore, California 94551, or to our electronic mail address [email protected].

This document was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor the University of Californianor any of their employees makes any warranty, expressed or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information,apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service bytrade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or theUniversity of California. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or the University of California andshall not be used for advertising or product endorsement purposes.

U.S. Government Printing Office: 1998/683-051-60044

Printed on recycled paper.

Nonprofit O

rg.U

. S. P

ostage

PAID

Albuquerque, N

MP

ermit N

o. 853

University of C

aliforniaLaw

rence Livermore N

ational LaboratoryS

cience & Technology R

eviewP.O

. Box 808, L-664

Livermore, C

alifornia 94551

Documents

Faster Cleanup of Contaminated Soil and Groundwater · Livermore are successfully cleaning up the groundwater beneath a utility pole yard in Visalia, California. Helped by Livermore