Lali Tchelidze and James Stovall

Transcription

1 Accelerator Division ESS AD Technical Note ESS/AD/009 Lali Tchelidze and James Stovall Estimation of Residual Dose Rates and Beam Loss Limits in the ESS Linac 0 April 01

2 Areportonthe Estimation)of)Residual)Dose)Rates)and)Beam)Loss)Limits)in)the)ESS) Linac) by LaliTchelidze*,JamesStovall

3 1.)Introduction) The ESS linac will accelerate protons from rest to. GeV through a variety of acceleratingstructures.oneoftheadvantagesofmodernlinearacceleratorscomparedwith their circular cousins is that they can typically provide very strong confinement for very high beam currents. While in the linac, the protons will be bunched and focused by electromagneticfieldstopreventthemfromstrikingthewallsoftheacceleratorandbeing lost.anyprotonsthatescapetheseconfiningfieldscannotberecapturedandwilleventually interactwiththeparentmaterialsfromwhichthelinacisconstructedviavariousprocesses. Thenuclearreactionsbetweentheerrantprotonsandtheatomsoftheparentmaterialsof theacceleratorwillresultin: inlsituproductionofionizingradiationfields(promptradiation)and theproductionofradioactivenucleiinsidetheparentmaterial(inducedactivity). Inthisreportweaddressonlytheinducedradioactivityinwhichradioactiveisotopes areproducedwithintheacceleratorstructures.theseisotopesdecaybyemittingneutrons, alphas,betasandgammasuntiltheyreachthe valleyofstability. SincethehalfLlivesofthe radioactiveisotopescanrangefromfractionsofsecondstoyearsandbeyond,theradiation fieldswillalwaysbepresentinthemachineonceitbecomesoperationalandwillbecome thesourceofresidualdoseratestowhichessworkersmaybeexposed. Thepurposeofthisreportistoevaluatetheresidualdoseratesresultingfromproton beamlossesintherfq(radiofrequencyquadrupole)anddtl(drifttubelinac)sectionsof theessaccelerator.basedontheseresultsweproposebeamlosscriteriathatcanbeused byacceleratordesignersthatwillensureresidualactivationlevelsintheearlypartofthe accelerator, which will be consistent with safe handslon maintenance as defined by statutoryregulations,essadministrativerequirementsandbestpracticeintheaccelerator community. ) )

4 .)Background)radiation)in)southern)Sweden) AccordingtotheInternationalAtomicEnergyAgency(IAEA),aresidentofsouthern Sweden,wouldbeexposedtobetween10and0mSvduringhislifetimefromnormal backgroundradiation 1.ThisistheequivalentofroughlyL.mSv/yearor0.L0.µSv/hr. Figure 1 shows the background radiation lifetime dose distribution for Western Europe compiledbytheiaea.othersources reportbackgroundradiationlevelsofonly1msv/year insouthernsweden.forthepurposeofcomparingdifferentradiationsafetycriteria,wewill assumeanambientbackgrounddoserateofmsv/yror~0.μsv/hr. ) ) Figure1.TypicallifetimedosesinWesternEurope(mSv). 1 ess.report.ref.10900v1.0l010

5 .)Dose)limits)for)Radiation)Workers) We have two documents that specify ionizing radiation dose limits for radiation workersatess:theswedishradiationsafetyauthorityregulatorycode andgeneralsafety ObjectivesforESS.Swedishlawlimitstheannualeffectivedosetoradiationworkersto0 msv/year.amaximumrateof0msv/yrcouldbeallowedduringasingleyearprovided that the average over years does not exceed 0 msv/yr. This maximum exposure represents ~1 times the natural background which, by some estimates, represents an excessive exposure. The ESS Technical Advisory Committee (TAC) has established somewhatmoreconservativelimitsforessworkers.thesedoselimitsaresummarizedin Table 1. In this table incidents are defined as events that occur less than once per year. Unexpectedeventsoccurlessthanoncein10years. Table1.IonizationRadiationDoseLimitsforSwedishworkers. Event) Statutory)Dose) Administrative) Limits) Limits)for)ESS ) Normaloperation 0mSv/year * 10mSv/year Incidents 0mSv/event Unexpectedevents 0mSv/event DesignBasis Accident 100mSv/event *limitedtoatotalexposureof100msvinyears Ifweassumethatworkerswouldbeexposed000hoursperyearthentheirrateof exposurewouldbelimitedtoμsv/hrinnormaloperationsor10μsv/hrinthecaseofan incident. Swedish convention traditionally defines three levels of controlled radiation areasbasedontheambientdoserate.thezonehavingthelowestexposurerate(bluezone) limitsexposuretoμsv/hror~0timesthenaturalbackgroundrate.inthiscasethetotal workdurationwouldbelimitedto00hrs. CERN differentiates between supervised radiation areas and controlled radiation areas. For temporary work in a supervised area, no detailed job or dose planning is required. Before entry by workers, CERN carries out radiation surveys in the accelerator tunnelsfollowingamachineshutdownwheretheymeasurethedoseratesoncontactandat adistanceof0cmfromtheacceleratorstructures.theradiologicalareasarethenclassified asafunctionofthedoseratesmeasuredat0cmwhereitisconsideredtocorrespondto thelocalambientdoserate.tablesummarizestheclassificationofradiationareasatcern. 00L1E.pdf GeneralSafetyObjectivesforESS,EDMS11,revA011L11L0 CERNReportEDMSL1019,ReglesGeneralesd Exploitation,ConsignesGeneralesde Radioprotection,00L1L0

6 Table.EffectiveDoseLimitsforCERNradiationworkers. Area) Ambient)Dose) Personnel) Monitoring,) Dose)Limit) Classification) Rate) Access) Dosimetry) Permanent no 1mSv/yr <0.μSv/hr notrequired workplace restriction supervised passive, Supervised msv/yr <1μSv/hr radiation personal temporaryworkplace workers dosimeter Controlledradiation area 0mSv/yr <0μSv/hr supervised radiation workers active, personal dosimeter WeseeinthistablethattheintegratedannualdosetoworkersinanonLdesignated permanent workplace is limited to ~ times the natural background. However, in a lowl occupancy supervisedtemporaryworkplace,wheretheworkingtimewouldbelimitedto 00 hours per year, the exposure rate is limited to 1 μsv/hr or ~0 times the natural backgroundrate. Sinceweareonlyconcernedwithlimitedterm(lowoccupancy)exposureandtobe conservative we will assume for the purpose of this study the more conservative CERN ambientdoseratelimitof1μsv/hrfor 00hrs/yr.Wedefinethedesigngoalforlimiting uncontrolled beam loss to be that amount of beam that results in a local dose rate of 1 μsv/hr,0cmfromtheacceleratingstructure.applyingthecernradiationsafetycriteria wecanassumethat,whilealloftheesslinacwillbecontainedwithinasupervisedaccessl controlledarea,wecanassurethatanyacceleratorstructuremeetingthedesignguidelines inthisreportcanbeapproachedandmaintainedwithoutthenecessityofcreatingspecial doseandworkplans. ) )

7 .)Computational)Method) The radiological studies discussed below are based on MonteLCarlo simulations of protonreactionsandradiationtransportintherfqanddtlusingthemarsprogram.for thepurposeofcalculatingtheactivationwehaveconstructedsimplemodelsofboththerfq anddtlthatrepresentthegeometryandparentmaterialsusedintheirfabrication.inall cases we have assumed as a source of lost beam a 1Lwatt beam of protons, having a trajectoryperpendiculartothesurface,incidentatasinglepointontheinnersurfaceofthe acceleratingstructure.weusethemarscodetocalculatetheresidualequivalentdoserate ontheoutersurfaceofthemodelandatvariousdistancesfromtheirsurface.theresidual doseratesaredirectlyproportionaltothepoweroftheincidentprotonbeamsofromthese calculationswecanderivethebeampowerrequiredtomeetthedoseratelimitdiscussed above. Wehaveattemptedtoduplicatetheactivationexpectedduringamaintenanceperiod followingatypicalruncycle.inallexampleswehaveassumeda100ldayruninwhichthe RFQacceleratorisexposedtoa1LWpointsourceofprotonsfollowedbyacoolLdownperiod. Figureshowstheeffectofwaiting1,andhoursfortheRFQstructuretocooldown beforecharacterizingitseffectivesurfacedoserates.wecanseethatinthisenergyrange,l 10MeV,bywaitinghours we would expect the residual radiation level to decreaseby approximatelyafactorofthree. Residual Activity vs. Cool Down Time Residual Surface Activity (µsv/hr) day exposure, cm Cu thickness cool-down time 1 hr hr 1 day (MeV) Figure.RFQresidualsurfaceactivityasafunctionofprotonenergyandcoolLdowntime. Figureshowsthestoppingrangeforprotonsincopperandstainlesssteelalloy0 (SSL0),thepredominantparentmaterialsinbothRFQsandDTLs.Wecanseeinthisplot that the range of 10LMeVprotons in both Cu and SS is very short, less than 0 μm. One might think that because the wall of the accelerator is very thick relative to the stopping range that it would serve as an effective radiation shield. In fact, we find that exactly the opposite is true. When a proton is captured by a Cu nuclei it typically decays emitting a neutron. Because Cu has a relatively high atomic mass (A ) it is a poor moderator for neutrons.asaresult,theneutronsproducedattheendoftheprotonrangemultiplyscatter throughoutthewalloftheaccelerator,creatingmultipleisotopes.

8 Figureshowsaspectrumoftheelements(isotopes)createdby10LMeVprotonsin Cuwhilefigureshowsthattheexpectedsurfaceactivationincreasessignificantlywithwall thickness.inthisexampleweirradiatedthecufor100dayswitha10mevprotonbeamand calculatedtheresidualactivationfollowing1dayofcooldown.wecanseethattheresulting activationapproachesanasymptoticlimitforawallthicknessof~cm Stopping Range (m) Cu SS (MeV) Figure.StoppingrangeforprotonsasafunctionofenergyinCuandSSL0. 1 For 10 MeV incident protons Relative amount of elements in Cu E- 1E- 1E- 1E- 1E- H He C N O K Ca Sc Ti V Cr Mn Fe Co Ni Cu Zn Element Figure.Spectrumofelementscreatedby10LMeVprotonsinCu.

9 Residual Surface Activity (µsv/hr) Cu Thickness (cm) ) Figure.ResidualsurfaceactivityasafunctionofCuthicknessfora10LMeVincidentproton beam. ) ) ) Residual Activity vs. Cu Thickness 100 day exposure, 1 day cool down =10 MeV

10 .)RFQ) TheESSRFQisdesignedtoaccelerateprotons,havinganaveragebeamcurrentof ma, from kev to MeV. RFQs for higher current applications have been built to accelerate100lmaprotonbeamstoenergiesashighas.mev.modernhighperformance RFQsaretypicallyfabricatedfrombrazedOxygenLFreeElectronicgrade(OFE)Copperwith possible stainless steel flanges. Figure shows the typical RFQ cross section of a high performancerfq. Figure.TypicalcrosssectionofamodernRFQ. BeamlossinRFQstypicallyoccurswhentheinjectedprotonbeamismismatchedto thefocusinglattice.when protonsfalloutofsynchronism withtherfandarenolonger acceleratedtheywilltypicallybetransportedwithoutfurtheraccelerationtotheendofthe structure.protonsthatescapetransverseconfinementwilltypicallystrikeeithertheouter wallorendwallofthestructure. To maintain the precise geometry required and assure good thermal stability, the wallsoftherfqaretypically~cmthick.therfqmodelforourmarscalculationsisa simple Cu cylinder 0 cm in diameter with a Lcm wall thickness. Figure shows the expected surface activation for an RFQ as a function of incident proton energy. In this calculationwehaveassumeda1lwpointsourceofprotonsfor100daysfollowedbyalhr coolldownperiod.

11 Residual Surface Activity (µsv/hr) Residual Activity vs. Proton Energy 100 day exposure, hr cool down cm thick Cu (MeV) Figure.ResidualsurfaceactivityinanRFQasafunctionofenergy. ThethresholdforneutronproductioninCuhasbeenmeasuredtobe.1MeV,well belowthemevfinalenergyoftheessrfq.howeverthecrosssectionforthisreactionis so small that the resulting activation is insignificant. The expected residual activation predictedbythemarscodeisinsignificantforlostprotonshavingenergiesbelowmev. To calculate the ambient dose rate at 0 cm from the surface of the RFQ the MARS codemusttransporttheradiationfluxthrough0cmofair.becausethesurfaceactivation belowmevissolow,wellbelowtheessrfqfinalenergy,wehavenotcalculatedtheflux at0cm. Ifthesurfaceactivation abovemevcouldberepresentedas apointradiation sourcewecouldsimplyapplyaninversesquarescaling(1/d )toderivethedoserateat0 cm.aswewillseeinthenextsection,apointsourceisnotanaccuraterepresentationofthe surfaceactivation.becauseoftherandomnatureofthecalculation,thechanginggeometry and different incident proton energies, it is difficult to find a general scaling law that describestheradiationattenuationoutsidetheacceleratorstructure.byapplyingalawthat scalestheambientdoserateinverselywiththedistance(1/d)fromtheacceleratorsurface however, we can conservatively estimate the dose rates as a function of energy at 0 cm fromthestructureasweshowinfigure. W.E.Shoupet.al.,ThresholdfortheProtonLNeutronReactioninCopper,Phys.Rev..1L (19)

12 Ambiant Dose Rate (µsv/hr) Residual Activity vs. Proton Energy 100 day exposure, hr cool down Dose Rate=a/(D+) D (cm) Figure.Expectedambientdoseratescaledas1/DfromtheRFQsurface. Figure9showsthecalculatedsurfaceactivationasafunctionofprotonenergy.Italso showstheambientdoserateat0cmfromthestructurederivedbyscalingtheradiation inversely with the distance from the structure. In both figures we have applied the relationshipdose=a/(d+),whereweusethecoefficienta,tofitthecurvethroughthedose Rates the surface & 0 cm. vs. Proton Energy rateatthesurfaceandtheoffset,,toavoidthesingularityatd=0. Dose Rate (µsv/hr) day exposure, hr cool down surface activation at 0 cm scaled as a/(d+) (MeV) Figure9.Expectedsurfacedoseratesandat0cmfromtheRFQ. Weknowthatwithallotherparametersfixedthedoserateisdirectlyproportionalto thepowerofthelostprotonsthatareincidentontheinnersurfaceoftheaccelerator.ifwe apply our ambient dose rate limit of 1 μsv/hr to the radiation field at 0 cm from the surfaceoftherfqwecanderivethedesignlimitforbeamlossasafunctionoftheincident proton energy. Figure 10 shows both the beam loss limit in units of average power and beamcurrentthatwouldresultinanambientdoserateof1μsv,0cmfromthesurfaceof 10 cm Cu thickness 9 10 (MeV)

13 therfq.alsoplottedinthisfigureistheaveragepowerandbeamcurrentthatwouldresult insurfaceactivationof1μsv/hr. Proton Beam Power Required to Activate Cu to 1 µsv/hr 100 day exposure, hr cool down Proton Beam Power (W) Power Power 1 µsv/hr at 0 cm 1 µsv/hr on the surface Current Current (MeV) Figure10.PowerandcurrentoflostprotonsrequiredtoactivatetheRFQsurfaceto1 μsv/hrandproduceanambientdoserateof1μsv/hrat0cm. Applyingeithercriteria,surfaceactivationortheambientdoserateat0cm,wecan seethatthelimitingcurrentforbeamlostat MeVisessentiallyunlimited.Applyingthe ambient dose rate limit, the average beam loss at MeV would be limited to ~00 W or ~10 μa at a single point. Applying the much more conservative criteria of limiting the surfaceactivation,theaveragebeamlossatmevwouldbelimitedto~0wor~μaata singlepoint. WhileOFECuistypicallyverypureitsmetallurgicalspecificationallowsforCoorNi contamination at the level of 10 ppm. We found that the residual surface activation is increasedby0%forlmevincidentprotonsandby0.001%for10lmevincidentprotonsin thecaseofcoimpurity.inthecaseofniimpurity,thesurfaceactivationisincreasedby0% forlmevincidentprotonsandby %for10lmevincidentprotons. ) ) ) Proton Beam Current (µa)

14 .)DTL) IntheESSDTL,allofthedrifttubes seen bytheprotonbeamwillbemadeofofecu. Half of the drift tubes will contain permanent magnet quadrupoles (PMQs) containing Samarium (Sm) and Cobalt (Co). Assuming that the drift tubes will be solid Cu to reduce theirnaturalmechanicalfrequencythepmqswillbewellshadowedbythecoppernosesof thedrifttubeasshowninfigure11.withincreasingenergy,therangeofprotonsincopper increases,butsowillthethicknessofthecopperinthenosesofthedrifttubessoprotons neverreachthemagnets.theotherhalfofthedrifttubesinthedtlwilleitherbeempty, containdiagnosticdevicesorelectromagneticdipoles.inbothcasesweexpectthesedevices tobewellshadowedbythecunosesofthedrifttubeasinthecaseofthepmqs. Figure11.CrosssectionofatypicaldrifttubeinwhichthePMQisshadowedbyCopper. As we found in the RFQ study, the Cu noses on the drift tubes will not reduce the inducedactivationbutmayamplifyit.tomoreaccuratelysimulatetheactivationinthedtl wehaveconstructedamorecomplicatedmodelthatscaleswithprotonenergy.figure1 showsthecomputationalmodelusedinourmarscalculationsforthedtl.thedriftltube bodyscalesinlengthwiththevelocityofthebeamwhilethepmqremainsaconstantlength. The PMQ is comprised of segments of sintered permanentlmagnet material enclosed in a stainlesssteelholder.eachdrifttubehasawatercoolingchannel.thedriftltubediameteris typically90mmwhiletheinnerdiameterofthetankis0mm.thetankwallwillbemade oflcmthickssl0.

15 Figure1.ComputationalmodelfortheESSDTL Because the PMQ is well shadowed by the copper drift tube we expect any errant protons to be lost on the leading edge or nose of the drift tube. There is a possibility howeveroflowenergyprotonsbeinglostintheinnerboreofthedrifttubeatitscenter. Because the geometry is considerably more complicated, containing a variety of materialsthespectrumofelements(isotopes)createdbytheprotonsismuchlarger.table liststhecompositionofssl0byelement.tableliststhecompositionofthepermanent magnetmaterialbyelement. Table.CompositionofSSL0. Element) Composition)(%)) Fe balance C <0.0 Cr 1. 0 Ni 11 Mn < Si <1 P <0.0 S <0.0 Table.CompositionofthePMQ. Element) Composition)(%)) Sm 0L0 Fe 10L0 Cu L1 Zr L Co balance Figure1showsthespectrumofisotopescreatedasaresultofirradiatingthefinal drifttubeinthe ESS DTL with 0LMeVprotons.Theresultingradiationfieldmaybevery complicated.

16 10 0 For 0 MeV incident protons 10-1 Amount in Cu Amount in Steel Amount in SmCo Amount in Water Relative Amount of Elements H He C N O K Ca Sc Ti V Cr Mn Fe Co Ni Cu Y Nd PmSm Eu Element Figure1.SpectrumofelementscreatedbyirradiatingthefinaldrifttubeintheESSDTL with0lmevprotons. Figure1showstheexpectedsurfaceactivationoftheDTLasafunctionoftheenergy ofprotonslostonthedrifttube.inthisfigureweshowtheactivationexpectedfrombeam lossontheleadingcunoseofthedrifttubes(blue)andinthecenterofthedrifttubeonthe SSPMQholder(red).Wecanseethattheactivationisjustslightlyhigherwhenirradiating thenose.therefore,weonlyfollowtherestoftheanalysisforprotonslostonthedriftltube nose.thesurfaceactivationforenergiesbelow0mevissolowastobeirrelevantforthis Residual Surface Activity, DTL study. Residual Surface Activity (µsv/hr) W point source on drift tube 100 day exposure, hr cool down W (MeV) Figure1.ResidualsurfaceactivityintheDTLasafunctionofincidentprotonenergyfor lossonthenoseandinthecenterofthedrifttube. center nose 9 9

17 WeusetheMARScodetotransporttheradiationfromtheDTLsurfaceandcalculate theambientdoserateasafunctionofdistance(d)fromthetankwall.becauseofthenoise in the data and the inconsistency of the curves fit to the data we were unable to find a general rule to accurately describe the rate of attenuation. We fit the data using the following curve where the coefficient a determines the magnitude of the fit and the exponent b describes the rate of attenuation from the tank wall. The constant is appliedtoavoidthesingularityatd=0. #$ = ( ), = , = Wefindthatingeneralthatthefluxgeneratedbyhigherincidentprotonenergiesis Ambiant Dose rate, DTL Nose attenuatedonlyslightlymorerapidlythanthefluxgeneratedbylowerenergies. Ambiant Dose Rate (µsv/hr) day exposure, hr cool down 1 W point source on the nose of the drift tube Dose=a/(D+) b (MeV) Distance from DTL (cm) Figure1.Expectedambientdoserateasafunctionofincidentprotonenergyscalesas a/(d+) b. Figure1showstheexpectedambientdoserateatthesurfaceand0cmfromthe surfaceofthedtlasafunctionofprotonenergyfora1lwpointsourceincidentonthenose ofthedrifttube.

18 Residual Activativity (µsv/hr) Residual Activity, DTL 100 day exposure, hr cool down protons are lost on the leading nose of the drift tube surface at 0 cm (MeV) Figure1.Residualdoserateatthesurfaceand0cmfromthesurfaceoftheDTLasa functionofprotonenergy. Allofthepreviouscalculationshavebeencarriedoutassuminga1LWprotonsource. Becausetheresidualsurfaceactivityisdirectlyproportionaltothepoweroftheprotonsofa givenenergylostatapointontheinnersurfaceoftheacceleratorwecan,bynormalizing theresidualactivityto1μsv/hr,derivethecorrespondingpowerofthelostprotons,asa function of energy, required to activate the DTL to this level. Figure 1 shows the beam power and average current required to activate the surface of the DTL to 1 μsv/hr and produce an ambient dose rate of 1 μsv/hr at 0 cm from the surface. It also shows the correspondingaveragebeamcurrentlimits. Proton Beam Power required to Activate DTL to 1 usv/hr, DTL Proton Beam Power (W) power power (MeV) Figure1.BeampowerandcurrentofprotonslostatapointrequiredtoactivatetheRFQ surfaceto1μsv/hrandproduceanambientdoserateof1μsv/hrat0cm. ) ) ) 1 µsv/hr on the surface 1 µsv/hr at 0 cm Current Current Proton Beam Current (µa)

19 .)Distribution)of)Particle)Loss) Thepurposeofthisstudyhasbeentoestablishaguidelineforacceleratordesigners by defining acceptable limits for beam loss in RFQs and DTLs. From our models we have calculatedthesurfaceactivationfortherfqandthedtlresultingfromapointlossof1lw of protons incident perpendicular to the inside surface of these structures. From these valueswehavederivedthelimitingbeampowerthatwouldjustmeetthemaximumdose rateof1μsv/hrtomeetcern scriteriafortemporaryworkinasupervisedworkplace. Figure1a(leftLhandordinate)showsinredthelimitingaveragepowerofprotonsstriking asinglepointrequiredtoactivatethesurfaceto1μsv/hrasafunctionoftheenergyofthe lostprotons. Proton Beam Current required to Activate Linac to 1 usv/hr, DTL Lost Proton Power (W) day exposure, hr cool down DTL surface DTL at 0 cm RFQ at surface RFQ at 0 cm (MeV) Figure1a(leftLhandordinate).Powerofprotonslostatasinglepointrequiredtoactivate thelinacsurfaceto1μsv/hrandthepowerrequiredtoproduceanambientdoserateof1 μsv/hr,0cmfromthesurface. Figure1b(rightLhandordinate).Powerdensityoflostprotonsrequiredtoactivatethe linacsurfaceto1μsv/hrandthepowerrequiredtoproduceanambientdoserateof1 μsv/hr,0cmfromthesurface. Wehavealsocalculatedtheexpectedambientdoserates0cmfromtheaccelerator surfacewherethelimitingrateof1μsv/hrisalsoapplied.therateofattenuationfromthe surfaceisdependentonboththegeometryoftheacceleratingstructureandtheenergyof thelostprotons.forthedtlwefindthat,tofirstorder,theattenuationintheambientdose ratefollowsaninversescalinglaw,1/d,orfaster.fortherfqwehavenotcalculatedthe attenuationforenergiesabovemevbutwehaveconservativelyassumedforcompleteness thattheradiationalsofallsoffas1/d.fromthesevalueswehavederivedthelimitingbeam powerforprotonslostatasinglepointthatwouldjustmeetthemaximumdoserateof1 μsv/hr at 0 cm to meet CERN s criteria for temporary work in a supervised workplace. Figure1a(leftLhandordinate)showsinbluethelimitingaveragepowerofprotonsstriking asinglepointrequiredtocreateanambientdoserateof1μsv/hrat0cmasafunctionof theenergyofthelostprotons. Inrealityprotonsdonotstrikethewalloftheacceleratorperpendiculartothebeam axisandatypical hotspot ontheacceleratorsurfacewillnotbeapointsource,butwillbe Lost Proton Power Density (W/m)

20 muchmorediffuse.inadditionwecanexpecttheretobemultiplehotspotsduetobeam lossatmultiplepointsinternally.inthedtlforexamplewemightexpectthefirstfewdrift tubesineachtanktointercepterrantprotonsresultingfromminormismatchesorsteering errors between tanks. In evaluating beam loss, most modern multilparticle codes identify themagnitudeoflossescellbycellbutitiscommontoquotethelimitingvalueforbeamloss intermsofpowerperunitlengthorw/m. Toevaluatethedoserateoutsidetheacceleratorresultingfrommultiplepointsources oflostprotonsinside,wesumtheradiationfromthesurfaceactivationatmultiplepoints over1mofacceleratorlengthatadistancedfromtheaccelerator.figure19showsinlight bluethenormalizedambientdoseratefromasingle1lwpointsourceoflostprotonsasa functionofdistance(d)fromtheaccelerator.thelightredcurveshowsthecumulativedose rateresultingfrompointsourcesofbeamlosseachrepresenting1/lwofprotons.the dark blue curve shows the cumulative dose rate resulting from 9 point sources of 1/9LW each.thepinkcurveshowscumulativedoseratefrom1pointsourcesof1/1lweach. In the limit, the beam loss from multiple sources approaches a line source of beam loss.thecumulativedoserateatadistancedfromtheacceleratorduetosuchalinesource oflostprotonsisshownasthedarkredlineinfigure19.itisinterestingtonotethatat~0 cm the ambient dose rate resulting from 1LW of distributed proton loss is nominally independent of the number of points over 1 m that the losses are distributed. To be conservativewehaveassumedthatat0cmtheworstcasescenario,inwhichweseethe leastattenuation,isrepresentedbyasinglepointloss. Attenuation Attenuation Point Sources line D (mm) Figure19.Ambientdoserate(normalized)asafunctionofdistancefromtheaccelerator resultingfrommultiplepointprotonlosses. ByassumingasinglepointLsourcelostprotonpowerlimit,butdistributingitovera meter, we can be assured that using the uniform power density will yield a more conservative guideline than is indicated by a multilpoint loss distributed over a meter. Therefore,wecanredefinethelimitingpowerlossintermsoflinearpowerdensity(W/m) andthecorrespondingcurrentlossintermsoflinearcurrentdensity(μa/m)andbesure that the ambient dose rate at 0 cm will be 1 μsv/hr resulting from a linelsource of protonloss.figure1bshowsontherightlhandordinatethelimitingpowerdensityfor

21 protonlossredefinedinunitsofw/m.figure0showthecorrespondinglimitingaverage currentdensityforprotonlossinunitsofμa/m. Proton Beam Current required to Activate Linac to 1 usv/hr, DTL Lost Proton Current Density(µA/m) (MeV) Figure0.Averagelinearcurrentdensityoflostprotonsrequiredtoactivatethelinac surfaceto1μsv/hrandthecurrentdensityrequiredtoproduceanambientdoserateof 1μSv/hr,0cmfromthesurface. WhileweinitiallyassumedtheCERNradiationexposurelimitof1μSv/hrmeasured 0cmfromtheacceleratorsurfacewefindthatitisreasonabletoexpectlinacdesignersto achieveexpectedbeamlosses(100l100w/mor1l00μa/m)thatwouldyieldasurface activation of 1 μsv/hr. For the purpose of deriving a beam loss guideline we therefore ignorethe0cmdataandfocusonlyonthesurfaceactivation. ItisimportanttonotethatintheRFQthesurfaceactivationresultingfrombeamlost atenergiesbelowmevisinsignificant.inthedtlbeamlossatenergiesbelow0mevare likewiseinsignificant.byfittingacurve(red)tothedtlsurfaceactivationdataforbeam lostatenergiesabove0mevwecanderiveasimplebeamlossguidelinewhichwillassure usthatthesurfaceactivationratewillnotexceed1μsv/hr; ## 100. where ## isthemaximumallowablebeamlossinunitsofμa/masafunctionofenergy,w, inunitsofmev. 100 day exposure, hr cool down DTL surface DTL at 0 cm RFQ at surface RFQ at 0 cm

22 .)Conclusion) WeconcludefromthisstudythatbeamlossintheESSRFQisexpectedtobeofno consequence.likewisebeamlossinthedtlatenergiesbelow0mevisexpectedtobeof no consequence. By setting the surface activation limit to 1 μsv/hr we find that the allowable beam loss above 0 MeV can be defined by a simple exponential function in energy. Followingthisdesignguidelineassuresusthattheradiationenvironmentinthelinac tunnel due to activated accelerator components will meet the CERN criteria for limited ( 00hrs/yr)accesstoradiationworkersforroutinemaintenancewithoutthenecessityof detailedjobordoseplanning.wefurtherobservehoweverinfigure0thattheambient doserate0cmfromthesurfaceisexpectedtobeaboutonetenthofthesurfaceactivation (~1. μsv/hr) at all energies. This guideline therefore meets even the most conservative ESSexposurelimitfor normaloperations of10msv/yrorμsv/hrfor000workhoursif measuredat0cmfromthesurface. It is important to note that this guideline applies only to exposure from activated components that radiation workers might receive during routine but infrequent maintenance periods after the accelerator has been shut down for hours. Longer interventions may require longer coolldown periods. It does not apply to exposure from prompt radiation sources nor does it apply to uncontrolled areas accessible to untrained workers.