ineab

olsty, W

Keywords:

d tthehan

useds and

moment the disposal in deep geological formations is the preferredform of disposal in Germany (AkEnd, 2002). Safety assessmentsneed to be done to ensure the long-term protection of the

biosphere (Kirchner, 2009; Pinedo et al., 2005). For Europe, thedevelopment of important climatic parameters such as tempera-ture and precipitation has been assessed for the time after disposalby analyzing paleoclimatic data (BIOCLIM, 2004). Accordingly,temperature changes of 10e15 C from the current conditions arepossible, accounting for warmer and colder periods. Changes inrainfall patterns may lead to drier or more humid conditions.

The inuence of climate on radioecological models was assessedin prior studies. The BIOMOSA project assessed the inuence of

* Corresponding author. Tel.: 49 (0)89 3187 2889; fax: 49 (0)89 3187 3363.E-mail addresses: christian.staudt@helmholtz-muenchen.de (C. Staudt), semi@

helmholtz-muenchen.de (N. Semiochkina), christian.kaiser@helmholtz-muenchen.de

Contents lists available at

Journal of Environm

journal homepage: www.els

Journal of Environmental Radioactivity 115 (2013) 214e223(J.C. Kaiser), g.proehl@iaea.org (G. Prhl).beginning of the 20th century. The resulting waste contains a widerange of different radionuclides, with different chemical andphysical properties inuencing their behavior in the environment.Radioactive waste is dened as materials contaminated withradioactive substances with activity concentrations above clear-ance levels from regulatory requirements for which no further useis foreseen (IAEA, 2007). Due to their radiotoxicity, these materialsneed to be isolated from the environment and the population insuch away that any subsequent radiological impacts are kept belowa critical level. One way to isolate the waste is the disposal in deepgeological repositories in appropriate geological formations. At the

disposal of long-lived radioactive waste poses notable challenges.In contrast to political and social changes, which are almostimpossible to predict with reliable accuracy, climatic and geologicalchanges may at least be roughly predicted and their impact on theexposure of a population to radionuclides from a deep geologicalrepository may be modeled.

Geological changes are considered in advanced planning andperformance assessments for repositories for the disposal ofradioactive waste in Sweden and Finland (Avila and Ekstrm, 2006;Hjerpe et al., 2009; Ikonen et al., 2008; Little et al., 2011; SKB, 2011).Climate change is the main driving factor for changes in theRadionuclidesFinal repositoryRadioactive wasteClimate reference regionsClimate changeBiosphere dose conversion factors

1. Introduction

Radioactive materials have beenmilitary and scientic technologie0265-931X/$ e see front matter 2012 Elsevier Ltd.doi:10.1016/j.jenvrad.2012.05.016the conditions at the reference climate regions. The model includes exposure pathways common to thosereference climate regions in a stylized biosphere and relevant to the exposure of a hypothetical self-sustaining population at the site of potential radionuclide contamination from a deep geologicalrepository. The end points of the model are Biosphere Dose Conversion factors (BDCF) for a range ofradionuclides and scenarios normalized for a constant radionuclide concentration in near-surfacegroundwater. Model results suggest an increased exposure of in dry climate regions with a highimpact of drinking water consumption rates and the amount of irrigation water used for agriculture.

2012 Elsevier Ltd. All rights reserved.

in industrial, medical,processes since the

population from negative effects of radionuclides from a deepgeological repository.

Since it is highly improbable that current political, social andeconomic entities will persist over long-time frames, the safeAvailable online 27 June 2012

developed according to the BIOMASS methodology of the IAEA and model parameters were adjusted toAccepted 14 May 2012dened for northern Germany according to model results from the BIOCLIM project. Nine present dayreference climate regions were dened to cover those future climate conditions. A biosphere model wasModeling the impact of climate changefor long-term safety assessment of nucl

C. Staudt a,*, N. Semiochkina a, J.C. Kaiser a, G. PrhlaHelmholtz-Zentrum Mnchen, German Research Center for Environmental Health, Ingb International Atomic Energy Agency, Division of Radiation, Transport and Waste Safet

a r t i c l e i n f o

Article history:Received 16 December 2011Received in revised form11 May 2012

a b s t r a c t

Biosphere models are usegeological repository. Sincepotential future climate cAll rights reserved.Germany with biosphere modelsr waste repositories

dter Landstrae 1, D-85764 Neuherberg, Germanyagramerstrae 5, 1400 Vienna, Austria

o evaluate the exposure of populations to radionuclides from a deeptime frame for assessments of long-time disposal safety is 1 million years,ges need to be accounted for. Potential future climate conditions were

SciVerse ScienceDirect

ental Radioactivity

evier .com/locate / jenvrad

AfsoonHighlight

AfsoonHighlight

climate on the exposure of a hypothetical population withmodeling approaches from different countries (Olyslaegers et al.,2005; Prhl et al., 2005). Other projects evaluated the inuenceof the evolution of geological, geographical or climatic conditionson the exposure of a population at a specic potential repositorysite (Posiva, 2010; SKB, 2011; Smith and Kozak, 2011).

In this study a model for the transport of radionuclides fromnear-surface groundwater to an exposed hypothetical self-sustaining population is presented for northern Germany. Themodel was developed according to the reference biosphere meth-odology of BIOMASS including the formulation of an assessmentcontext, a biosphere system description, the description ofa potentially exposed group, creation of a Feature Events andProcesses (FEP) list and model development (IAEA, 2003). Modelparameters were adjusted to mirror conditions at several referenceclimate regions. Those reference climate regions include a scenariofor current climate conditions in northern Germany and eightpotential future climate conditions according to the BIOCLIMproject (BIOCLIM, 2004). Only those conditions have been consid-ered which allow the survival of a small self-sustaining humanpopulation.

FEP and used for all the reference climate regions with minormodications for the tundra climate region. The model wasimplemented in the ECOLEGO software (FACILIA, Sweden). ECO-LEGO calculates the ux of radionuclides between user-denedcompartments based on the solution of coupled rst-order differ-ential equations.

2.1.1. Assessment context and conceptual modelPurpose of the assessment is the evaluation of the exposure to

a hypothetical self-sustaining population to radionuclides froma deep geological nal repository for high-level radioactivewaste inGermany, including an assessment of the effect of climate changeon the exposure. For this purpose a range of BDCF for relevantradionuclides are calculated. With the BDCFs and knowledge of theactivity concentrations of radionuclides in groundwater, annualindividual doses for a self-sustaining population at the repositoryarea may be estimated.

The conceptual model is a compromise between the use ofdifferent exposure pathways, the availability of parameters forthese pathways for the different reference climate regions and theimpact of those pathways on the result (Fig. 1). The model is

C. Staudt et al. / Journal of Environmental Radioactivity 115 (2013) 214e223 215The model end points are radionuclide-specic BDCF pertainingto each of the climate reference regions. The BDCF concept wasapplied in previous assessments of long-term disposal safety, suchas the Yucca mountain repository (EPRI, 2009; IAEA, 2001;Wasiolek, 2003). The BDCF depend on contributions of ingestion,inhalation and external exposure and are normalized to a constantradionuclide-specic activity of 1 Bq/m3 in groundwater for thecurrent work. When radionuclide activities in surface groundwater,used as drinking water and for the irrigation of crops are known,the BDCF can be used to calculate annual doses to the exposedgroup.

2. Material and methods

2.1. Model setup

The biosphere model was developed according to the method-ology established in the BIOMASS project (IAEA, 2003). The full FEPlist in addition to the equations and parameters used for themathematical model are given in the supplementary material ofthis paper. The conceptual model was developed according to theseFig. 1. Common conceptual model usedintended to cover all relevant factors in adequate detail. At the sametime the model should be clearly laid out and easy to handle, tofacilitate the application by third parties.

Themodel considers material uxes in the biosphere potentiallycontaminated by radionuclides in groundwater, especially agricul-tural land and open water bodies. For the calculation of the BDCFa constant normalized radionuclide activity of 1 Bq/m3 in near-surface groundwater is assumed. To calculate radionuclide activi-ties in soil either two scenarios are used. The rst scenario Welluses irrigation water from a well to balance climate-dependentwater decits of agricultural land. Radionuclides from the irrigationwater accumulate in the root zone of plants over time and migrateinto deeper soil layers and are subject to decay. The second scenarioRising Ground Water RGW assumes high groundwater tabledirectly inuencing root zone soil layers. Excess water is drainedfrom the soil in the RGW scenario to facilitate the use as agri-cultural land. In addition to this it is assumed, that the whole soil-water system is in a saturated, balanced condition for the RGWscenario (Prhl et al., 2004). Radionuclide activity in soil for theWell scenario increases over time until a balance betweenradionuclide inow by irrigation is balanced by radionuclidefor the reference climate regions.

AfsoonHighlight

well water for irrigation or as drinking water. The radionuclidescurrently described in the model are the long-lived ssion productsTc-99, Sn-126, Se-79, Zr-93, Cs-135, Pd-107 and I-129 in addition toCl-36, Ni-59, Nb-94, Ra-226, Th-230, Pa-231, U-238, Np-237, Am-243 and Pu-239 due to their long half-lives. In the result sectiononly results for selected radionuclides are shown. Other potentialexposure scenarios involving water, like bathing, are not includedin the assessment.

A soil to plant transfer factor describes the uptake of radionu-clides from soil into plants by plant roots. A distribution coefcientdescribes the ratio of radionuclides in the solid and watercompartments of agricultural soil and is used in the model to

Fig. 2. Potential future climate development for northern Germany assuming naturallychanging and elevated CO2 concentrations in the atmosphere due to the BIOCLIMproject. Gray depicts temperate, black subtropical and white boreal climate conditions.Between 100,000 years and 170,000 years in the future, temperatures oscillatebetween temperate and subtropical conditions for the elevated CO2 scenario witha short period of boreal conditions at 140,000 years.

C. Staudt et al. / Journal of Environmental Radioactivity 115 (2013) 214e223216outow due to migration and decay. The BDCF are calculated forthose balanced conditions.

Two different types of agricultural soils are modeled as pastureand arable land. They were distinguished by the depth of the soillayer: 10 cm for pasture and 25 cm for tilled agricultural land. It isassumed that pastures are used for the cultivation of grass and fruittrees, agricultural soils for cereals, potatoes, maize, fruit and leafyvegetables.

Contaminated soil particles enter the atmosphere as dust.Agricultural plants take up radionuclides from the soil by rootuptake in the Well and RGW scenarios and by foliar uptake fromirrigation water in the Well scenario. In addition, activity istransferred from the soil to the lower parts of the plants byresuspension. The resuspension-dependent uptake is estimated bythe amount of soil attached to the plant and an element-dependentenrichment factor. Radionuclides enter animals with drinkingwater and food plants or in the case of sh, by the surroundingfreshwater. Humans ingest drinking water, plants and animals,inhale contaminated air and work on contaminated soil.

Since the description of the processes in the nal repository andthe geosphere are not part of the model represented here, thisactivity concentration stays constant for the considered time frame.The model time frame is 1 million years, to cover the half-lives oflong-lived radionuclides. The ECOLEGO software used for theimplementation allows for the integration of a time-dependentsource term, as soon as such information becomes available fromother projects and assessment.

2.1.2. Biosphere system descriptionIn this work, we dene the biosphere as a set of environmental

components inuencing the exposure of a hypothetical self-sustaining human population. These components include organ-isms used for food production and environmental media like rootzone soil, surface water bodies and the atmosphere. The conceptualbiosphere model is the same for all reference climate regions, withsmall adjustment for the tundra climate.

Subsurface water is the main pathway of radionuclide transportfrom the repository to the biosphere. The water enters thebiosphere either from a shallow groundwater table or by the use ofTable 1Reference region climate data (Mller, 1996).

Reference station Kppen/Geigerclassication

Temperature [C],annual average

Temperature [C],growing period ave

Turku Dfb boreal 4.8 15.5Sodankyla Dfc boreal 0.4 11.1Vardo ET boreal 1.6 8.4Santander Cfb temperate 13.9 17.3Magdeburg Cfb temperate 9.2 16.3Rostov Dfa temperate 8.4 19.2Rome Csa subtropical 15.6 20.7Valladolid Csb subtropical 12.1 17.2Marrakesh BS subtropical 19.9 17calculate a migration factor. This factor describes the leaching ofradionuclides from the root zone to deep soil layers (Baes andSharp, 1983; IAEA, 2005). Those parameters are divided intoparameters for sand, loam, clay and organic soil, according to soiltexture and organic content (IAEA, 2009, 2010). When available,different transfer factors were used for boreal, temperate andsubtropical reference regions for the soil to plant transfer factors.

Depending on climate and agricultural practices, irrigation maynot be necessary in some of the climate reference regions. Never-theless even in regions where a low percentage of agricultural landis irrigated (Siebert et al., 2007), the Well scenario with irrigationwas used in addition to the RGW scenario for an estimation ofradionuclide activity in soil.

It is assumed, that the agricultural practice and food consump-tion rates of self-sustaining populations adapt to climate changes.All consumed food and drinking water is derived from thecontaminated area and the population spends thewhole year in thepotentially contaminated region. Current country-specicconsumption habits are assumed to mirror consumption habits ofself-sustaining populations at the corresponding reference climateregions. This approach was chosen, since consumption data sets forfully self-sustaining populations on this scale are not available. Forthe permafrost reference climate region Vardo it is assumed, thatthe population subsist on hunting and gathering food withoutpracticing agriculture.

Together with food consumption rates, changes of soil activityconcentration, due to changed irrigation patterns or soil properties,determine the activity concentration in plants, animals andhumans in this model. In addition to ingestion, internal exposuremay be caused by inhalation of air-borne particles. Air-borne gasessuch as C-14 are not included in the model. Finally, externalexposure from contaminated soil is integrated into the model.

2.1.3. Potentially exposed groupA hypothetical self-sustaining population exposed to radionu-

clides from a deep geological repository is assumed for thisassessment. The exposure of non-human biota is not included.Human habits are mostly assessed in the context of agriculturalpractice and diet. Resources used from the contaminated area

rageAnnual precipitation[mm]

Precipitation duringgrowing period [mm]

Average irrigation[l/m2 a]

576 184 92508 130 28544 130 0

1198 297 101513 266 229483 212 457874 289 599364 152 576

241 226 522

f Eu

ental Radioactivity 115 (2013) 214e223 217Fig. 3. (A) Temperature and (B) precipitation at the reference climate stations, (C) map oto balance the calculated water decit.

C. Staudt et al. / Journal of Environminclude drinking water, irrigation water and food produced fromplants and animals. For this purpose agriculturally grown food andgathered food like fresh water sh for all climate regions, as well asberries, mushrooms or reindeer in the tundra climate region areincluded. Since no large-scale agriculture is possible in the tundraclimate region, a self-sustaining population is assumed to use thesegathered food types in their diet. Since the population in the modelis unaware of the contamination, no countermeasures against theeffects of contamination are taken by this community.

2.1.4. Model developmentThe following parameters were varied between the reference

climate regions: dust concentration in air, feed of domestic animals,food consumption rates of the population, harvests of leafy vege-tables per year, irrigation amount, soil to plant transfer factors,resuspension factors, agricultural yields and growth periods ofplants. Climate data were derived from a handbook of climatestations (Mller, 1996) and the necessary irrigation amounts tobalance an agricultural water decit were calculated from this data(Achtnich, 1980). Food consumption rates and agricultural yieldsfor the reference climate regions were derived from a Food andAgriculture Organization of the United Nations (FAO) database(FAO, 1996, 2007, 2010) and from national guidelines (StrlSchV,2001). It is assumed that the exposed population derives all itsfood and drinking water from the area around the repository.

The parameters and equations used in the model are given inthe supplementary material of this paper. Element-specic distri-bution coefcients and transfer factors in agriculturally used soilswere extracted from literature for the different boreal, temperateand subtropical reference regions (IAEA, 2009, 2010; Paasikallioet al., 1994; Rosen, 1996; Rosen et al., 1996). Due to the scarcity ofdata for boreal climate regions, transfer factors for Chernobyl-derived radionuclides were used for the model.rope and North Africa with the climate reference stations, (D) average irrigation neededRadionuclides enter the model biosphere by near-surfacegroundwater. The radionuclide activity concentrations ingroundwater for the RGW scenario are the same for all refer-ence stations where this scenario is used. The soil is saturatedwith groundwater and excess water is drained. Because of this,no additional irrigation is needed. The activity concentration insoil is only dened by the distribution coefcients of theradionuclides in soil and the radionuclide activity concentrationin groundwater for the RGW scenario. Since the radionuclideconcentration in the groundwater is standardized to 1 Bq/m3 theradionuclide concentration in soil for the RGW scenario doesnot change over time.

Fig. 4. Soil activity concentrations for Cs-135 at different climate reference regions forthe Well and the RGW scenario during the rst 10,000 years of the model for sandsoils. Irrigation amounts are 92 l/m2 a for Turku, 229 l/m2 a for Magdeburg and599 l/m2 a for Rome. The groundwater used for irrigation has a standard radionuclideactivity concentration of 1 Bq/m3.

In this study a steppe reference climate region was added as anextreme case, since it cannot be excluded for the next 1 millionyears. A tundra reference climate region was added to simulateshort permafrost conditions. Glacial conditions were not includedin the model, since no self-sustaining human population cansubsist in the region during a glacial period.

A maximum variation of 15 C in temperature was assumed tocover the range of possible temperature changes during the culti-vation months of agricultural crops. Eight present day climatestations, changing gradually in temperature compared to a ninthclimate station in northern Germany, were selected to reproducethis temperature range (Table 1, Fig. 3A). For the selection of theclimate reference regions, the Kppen/Geiger classication wasused (Peel et al., 2007). According to this classication the presentday climate reference stations cover a wide range of climate zonesfrom tundra to steppe climate. They can be subdivided into threetemperature groups: boreal, temperate and subtropical.

C. Staudt et al. / Journal of Environmental Radioactivity 115 (2013) 214e2232182.2. Selection of reference climate regions

Fig. 5. Activities for selected radionuclides and different soil types at the Magdeburgreference climate region (A) for the Well scenario and (B) for the RGW scenario and1 Bq/m3 radionuclide concentration in groundwater.In the BIOCLIM project possible changes of climate in northernGermany were derived from paleoclimatic data and models forpotential future climate developments (BIOCLIM, 2004). Twopotential model scenarios were used in the BIOCLIM project toevaluate future climate conditions in northern Germany. Onescenario assumes elevated, the other scenario natural variations ofthe CO2 content of the atmosphere. In scenarios where naturallychanging CO2 concentrations in the atmosphere are assumed,mostly temperate and short boreal conditions will be predominantin northern Germany for the next 200,000 years. In elevated CO2scenarios temperate and subtropical conditions with short periodsof boreal conditions can be expected. Short-term permafrostconditions cannot be excluded (Fig. 2).

The soil radionuclide activity concentration for the Well

Fig. 6. Radionuclide activity concentrations in Bq/kg for the different climate reference regiogroundwater used for irrigation in the Well scenario and directly inuencing the top so1 Bq/m3.scenario is primarily dened by the irrigation amount, since theradionuclide concentration in groundwater is normalized to 1 Bq/m3 for all climate regions. A possible inuence of a dilution bydifferent precipitation amounts at the reference climate regions isalready inherent in the model, since higher precipitation results inless irrigation with contaminated groundwater. Although radio-nuclides can be directly absorbed by plant leaves during sprinklingirrigation and contribute to the activity concentration in plants bythe foliar uptake pathway, most radionuclides are accumulated inthe soil and subsequently absorbed by plant roots. Due to thisaccumulation the root uptake dominates total activity uptake in

ns in the equilibrium state for sand soils. The activity concentrations of radionuclides inIn every temperature group, a humid, normal and dry presentday climate station was selected according to the precipitation andresulting water decit during the growth periods of agriculturalplants (Table 1, Fig. 3B). The selected reference stations are pre-sented in a 3 3 matrix with the German reference station asa central point (Fig. 3C). To these stations climate regions areassigned for which irrigation rates are estimated to balance agri-cultural water decits during the cultivation months (Fig. 3D).These reference climate regions are described by sets of realisticallypossible climate-related parameters which may evolve from thecurrent condition in northern Germany (Fig. 2).

3. Results and discussion

3.1. Activity concentrations in soilil layer by water saturation limited only by drainage in the RGW scenario is set to

Fig. 7. Radionuclide activity concentration in cereals at the different reference climate regions, assuming sand soil and a radionuclide activity concentration in irrigation water of1 Bq/m3 for the (A) Well and (B) RGW scenario.

Fig. 8. Contribution of the different food types to the ingestionpathway BDCFof selected radionuclides at the reference climate regionMagdeburg for the Well scenario and sand soil.

C. Staudt et al. / Journal of Environmental Radioactivity 115 (2013) 214e223 219

BDC

entaplants. This is a simplication of the actual processes, since radio-nuclides bound to soil may not be available for the uptake by plants,depending on their speciation or binding to clay particles overlonger time periods.

The activity concentrations of radionuclides in the soil

Fig. 9. Contributions of the ingestion, inhalation and external exposure to the total

C. Staudt et al. / Journal of Environm220compartments for pasture and arable land are calculated in annualtime increments for the Well scenario, with a radionuclide inowdue to irrigation and an outow due to decay and migration intolower soil layers. For the calculation of activity concentrations insoil, the irrigation amounts are averaged for the crop types used,since different crops will be cultivated in the same soil over a long-time period.

The radionuclide inow and outow reach a steady state aftera few hundred to thousand years depending on soil type andradionuclide properties. Stable climate conditions are assumed inthe Well scenario during those accumulationperiods, even thoughit may take several thousand years until radionuclide inow due toirrigation and outow due to decay and migration to lower soillayers achieve a balance. The activity concentration in soil is stronglydependent on the amount of contaminated groundwater used forirrigation. Arable land and pasture soils reach the same steady stateradionuclide activity concentrations, but attain themwith differentkinetics. In the RGW scenario, a balance between soil and water isassumed to be reached in the rst year due to a large surplus ofgroundwater and drainage of the soil. The activity for a radionuclideis the same for all reference regions in the RGW scenario (Fig. 4).

Since a surplus of groundwater is present in the RGW scenarioand is constantly drained from the soil, the radionuclide concen-tration in soil is only limited by the distribution coefcient. Thisalso implies that radionuclide concentrations above the limit set bythe distribution coefcient would be washed out of the soil withthe drained groundwater in the RGW scenario. The radionuclideactivity concentrations for sand soil and the RGW scenario arecomparable to the Well scenario steady state activity concentra-tions for sand soil at the Turku reference climate region with anirrigation amount of 92 l/m2 year (Fig. 4).Since the specic soil types in the reference climate region arenot known, the radionuclide activities for several tilled arable soiltypes can be calculated using the corresponding distributioncoefcients and transfer factor parameters for sand, loam, clay andorganic soils. Soil activity concentrations for several radionuclides

Fs in the (A) Well and (B) RGW scenario of selected radionuclides for sand soil.

l Radioactivity 115 (2013) 214e223and different soil types at the Magdeburg reference climate regionare shown in Fig. 5. The following results are given for sand soils(Figs. 6e10). The steady state activity concentrations of the radio-nuclides in agricultural soil are shown in Fig. 6 for the referenceclimate regions in the Well scenario and the uniform concen-tration in the RGW scenario.

3.2. Activity concentrations in food

With the calculated steady state radionuclide activityconcentrations in soil, the corresponding radionuclide activityconcentrations in crops, livestock and animal products can becalculated for a groundwater contamination of 1 Bq/m3 (Figs. 7and 8).

In the Well scenario, a combination of irrigation amounts anddifferent transfer factors inuences the radionuclide activityconcentration in plants (Fig. 7A). To allow for different radionuclidespeciation in soils from different reference climate regions differenttransfer factors for boreal, temperate and subtropical climateregions were used when available.

For the RGW scenario the activity concentration in soil isassumed to be the same at all reference regions. Because of this,the different radionuclide concentrations in plants are mainly theresult of different soil to plant transfer factors for radionuclidesat subtropical, temperate and boreal climate regions. In additionto this, the resuspension factor has an inuence on the radio-nuclide concentration in plants. This inuence can be seen forPu-239 concentration in cereals for the RGW scenario at theMarrakesh and Rostov reference climate regions, where theresuspension factors are slightly higher due to the dry climate(Fig. 7B).

or sa

ental Radioactivity 115 (2013) 214e223 221For the subtropical climate regions, soil to plant transfer factorsfor Cs, Tc, I and Pu differ from the temperate transfer factors (IAEA,

Fig. 10. Total BDCF of selected radionuclides for the (A) Well and (B) RGW scenario freference climate region.

C. Staudt et al. / Journal of Environm2010). For the boreal climate regions only separate soil to planttransfer factors for Cs are available (Paasikallio et al., 1994; Rosen,1996; Rosen et al., 1996). For the other transfer factors, theparameter values for temperate climate regions are used for allreference regions (IAEA, 2010).

In addition to this, other parameters like plant yields, inu-encing the foliar uptake, are different in the different referenceregions. This result in slightly higher radionuclide concentrations,for example for cereals, at the Valladolid reference regioncompared to the Rome reference region (Fig. 7A), despite the lowerradionuclide concentrations in soil (Fig. 6).

3.3. BDCF

If the radionuclide activity concentrations in food and drinkingwater (Table 2) and the consumption habits of the population areknown, the transport of radionuclides through the food chain toa human population and the contributions of the different foodtypes on the ingestion pathway BDCF can be calculated (Fig. 8).Depending on the radionuclide, the contribution of drinking waterto the ingestion ranges from 3% for Cs-135 to 83% for Pu-239 in theexample shown.

For most radionuclides, the ingestion pathway accounts formore than 90% of the total BDCF. Exceptions are Am-243, Pa-231,Pu-239 and Th-230 with 61%, 75%, 49% and 47% contribution of theinhalation pathway and Nb-94, Ra-226 and Sn-126 with 97%, 24%and 90% contribution of the external exposure to the total BDCF inthe Well scenario. The contributions of the single pathways to theBDCF in the Well scenario (Fig. 9A) are comparable to thecontributions in the RGW scenario (Fig. 9B).

By adding the ingestion, inhalation and external exposure BDCF,the total BDCF can now be compared between the differentreference climate regions (Fig. 10). The main factors inuencing theBDCF in the RGW scenario are the food consumption of the

nd soil. There is no BDCF for a well scenario in Vardo since no irrigation is done in thispopulation, the irrigation amounts for agriculture and the transferfactors for radionuclides from soil to plant. There are other factorschanging between the climate reference regions, for examplelength of the cultivation period or the feed type for domesticatedanimals (grass or maize), but those have a negligible inuence onthe BDCF.

When the same transfer factors and radionuclide concentrationsin soil in the RGW scenario are used for the reference climateregions, the calculated BDCF are almost exclusively inuenced bythe consumption habits and are very similar between the referenceclimate regions. This is the case for U-238 and Cs-135were 93% andmore than 99% of the total exposure are due to the ingestionpathway (Fig. 9B). In contrast to U-238 were 58% of the exposuredue to ingestion pathway are contributed by radionuclides in

Table 2Radionuclide activity concentration [Bq/kg] for different food types at the referenceclimate region Magdeburg, assuming sand soil and a radionuclide activity concen-tration in irrigation water of 1 Bq/m3 for the Well scenario.

[Bq/kg] Tc-99 I-129 Cs-135 U-238 Pu-239

Drinking water 1.0Ee03 1.0Ee03 1.0Ee03 1.0Ee03 1.0Ee03Milk 4.8Ee05 7.4Ee04 1.1Ee02 4.6Ee05 6.5Ee07Freshwater sh 4.0Ee02 2.5Ee02 4.3Ee01 2.6Ee04 4.7Ee04Beef 3.6Ee05 9.9Ee04 2.7Ee02 1.4Ee05 6.8Ee06Pork 1.2Ee05 3.4Ee04 7.2Ee02 5.1Ee06 3.1Ee06Poultry 5.0Ee05 3.0Ee06 1.5Ee02 3.9Ee04 6.4Ee09Mutton 1.7Ee04 2.8Ee04 7.4Ee02 3.1Ee05 4.5Ee05Egg 1.1Ee03 8.3Ee04 2.2Ee03 5.7Ee04 3.8Ee07Cereal 1.1Ee03 8.0Ee04 4.3Ee02 2.2Ee03 5.7Ee04Fruit 8.4Ee04 7.5Ee04 2.7Ee01 3.2Ee03 5.2Ee04Potato 1.9Ee04 3.7Ee04 2.5Ee02 1.1Ee03 4.4Ee04Leafy vegetable 3.8Ee03 3.4Ee04 2.7Ee03 5.5Ee03 1.5Ee03Fruit vegetable 3.2Ee04 4.1Ee04 5.5Ee03 6.4Ee04 4.6Ee04

entadrinking water, only 3% are contributed in the case of Cs-135 at theMagdeburg reference climate region (Fig. 8). The dust concentra-tion in air is only an important parameter for radionuclides witha high contribution of the inhalation pathway to the BDCF, like Pu-239, Am-243, Pa-231 and Th-230 (Fig. 9).

These results indicate, that the soil to plant transfer factor hasa large inuence on the resulting BDCF for different referenceregions. Preliminary sensitivity analysis conrms this result (datanot shown). Because of this, the evaluation of the inuence ofradionuclide speciation and other soil parameters on the transferbetween soil and plants is the most critical step in the model.

In the Well scenario in addition to the soil to plant transferfactors and consumption habits, the inuence of different irrigationamounts can be observed. As expected, more irrigation withcontaminated groundwater generally leads to higher radionuclideactivity concentrations in soil, plants and animal products and thusto higher BDCFs since the radionuclide concentration in soil directlyinuences the ingestion, inhalation and external exposure path-ways (Fig. 10A). Since irrigation is not a factor in the RGWscenario, the differences in the calculated BDCF for the referenceclimate regions are smaller (Fig. 10B).

4. Conclusions

This study presents an assessment of the impact of climatechange on the exposure of a population in northern Germany toradionuclides from a geological repository. By using one dedicatedmodel instead of several site-specic models, the effects of varyingparameters in different climatic situations at the reference regionscan be separated from model-specic properties. Based on thendings of the BIOCLIM study the selected climatic conditions areconsidered to provide a comprehensive coverage of those situa-tions that allow the survival of a self-sustaining population.

The reaction of the BDCF to the conditions at the differentreference climate region is a complex interplay between thedifferent model parameters. For example, higher radionuclideconcentrations in soil due to increased irrigation rates in subtrop-ical climate regions can be balanced by lower soil to plant transferfactors compared to temperate climates in the case of Cs-135. Thestudy highlights the critical importance of the effects of radionu-clide speciation on the model result, since it directly inuencesradionuclide accumulation in soil and transfer to plants. For mostanalyzed radionuclides, the amount of consumed drinking water isa very important parameter, since the activity from drinking waterhas a large contribution to the ingestion BDCF and the ingestionBDCF has a large contribution to the total BDCF. Parameters oflesser, but not negligible, importance are climate-dependent irri-gation amounts and food consumption habits of the population aswell as dust concentrations in the air for a limited range ofradionuclides.

The calculation of activity concentrations in the top 10 cm ofpasture and 25 cm of tilled agricultural soil is an important part ofthe model. At the moment two different equations for risinggroundwater and irrigation are used for the root zone soil layers.The capillary rise of ground water from near-surface aquifers to theroot zone of the soil was not modeled in this study. The simpleexponential removal of radionuclides in the soil model has well-known weaknesses. Future work should be aimed to developmodels with an improved description of transport processes in theroot zone soil layer.

The parametrization of models for different generic regions isoften obtained from the same databases such as the IAEA Tec-Doc1616. This may lead to an articial convergence of model resultsand an underestimation of the spread between BDCF pertaining to

C. Staudt et al. / Journal of Environm222different biospheres. These biases cannot be avoided with genericmodels, when a clearly dened location for a nal repository is notknown, and should be kept in mind, when future developments areassessed.

Acknowledgment

Thework has been supported by the German Federal Ministry ofEnvironment, Nature Protection and Reactor Safety under contractnumber 3609S50005. The modeling approaches and exposurepathways were discussed with agency staff. We thank the partici-pants of EMRASII WG 3, of the BIOPROTA initiative and especially G.Smith and G. Kirchner for stimulating discussions.

Appendix A. Supplementary material

Supplementary material associated with this article can befound, in the online version, at doi:10.1016/j.jenvrad.2012.05.016

References

Achtnich, W., 1980. Bewsserungslandbau. Agrotechnische Grundlagen derBewsserungswirtschaft. Eugen Ulmer, Stuttgart.

AkEnd, 2002. Auswahlverfahren fr Endlagerstandorte Empfehlungen des AkEnd eArbeitskreis Auswahlverfahren Endlagerstandorte (AkEnd), Abschlussbericht,Langfassung. Kln.

Avila, R., Ekstrm, P., 2006. Development of Landscape Dose Factors for DoseAssessments in SR-Can TR-06-15. Svensk Krnbrnslehantering AB.

Baes, C., Sharp, R., 1983. A proposal for estimation of soil leaching and leachingconstants for use in assessment models. Journal of Environmental Quality 12,17e28.

BIOCLIM, 2004. Modeling Sequential BIOsphere Systems under CLIMate Change forRadioactive Waste Disposal. EC-Contract FIKW-CT-2000-00024.

EPRI, 2009. Yucca Mountain Total System Performance Assessment Code (IMARC)Version 10: Model Description and Analyses. EPRI, Palo Alto, CA. 2009. 1018712.

FAO, 1996. Global Climate Change and Agricultural Production. Direct and IndirectEffects of Changing Hydrological, Pedological and Plant Physiological Processes.FAO, Rome, Italy.

FAO, 2007. FAOSTAT. http://faostat.fao.org.FAO, 2010. FAO Statistical Yearbook 2010.Hjerpe, T., Ikonen, A., Broed, R., 2010. Biosphere Assessment Report 2009. POSIVA

2010e03.IAEA, 2001. An International Peer Review of the Biosphere Modelling Programme of

the US Department of Energys Yucca Mountain Site Characterization Project.IAEA, 2003. Reference Biospheres for Solid Radioactive Waste Disposal: Report of

BIOMASS Theme 1 of the BIOsphere Modelling and ASSessment (BIOMASS)Programme.

IAEA, 2005. Derivation of Activity Concentration Values for Exclusion, Exemptionand Clearance. Safety Report Series No. 44, Vienna.

IAEA, 2007. IAEA Safety Glossary Terminology Used in Nuclear Safety and RadiationProtection. IAEA, Vienna.

IAEA, 2009. Quantication of Radionuclide Transfer in Terrestrial and FreshwaterEnvironments for Radiological Assessments TECDOC-1616.

IAEA, 2010. Handbook of Parameter Values for the Prediction of RadionuclideTransfer in Terrestrial and Freshwater Environments. Technical Reports SeriesNo. 472, Vienna.

Ikonen, A., Aro, L., Leppnen, V., 2008. Forecasts of future terrain and vegetationtypes at Olkiluoto and implications for spatial and temporal aspects ofbiosphere modelling. Applied Radiation and Isotopes 66, 1754e1758.

Kirchner, G., 2009. Use of reference biospheres for proving the long-term safety ofradioactive waste repositories. Journal of Environmental Radioactivity 100,435e437.

Little, R., Avis, J., Garisto, N., Humphreys, P., King, F., Metcalfe, R., Penfold, J.,Ramlakan, A., Suckling, P., Towler, G., Walke, R., Walsh, R., West, A., Wilson, J.,2011. Postclosure Safety Assessment NWMO DGR-TR-2011-25. Quintessa Ltd.,Georma Engineering Ltd. and SENES Consultants Ltd.

Mller, M., 1996. In: Handbuch ausgewhlter Klimastationen der Erde, fth ed.Gerold Richter.

Olyslaegers, G., Zeevaert, T., Pinedo, P., Simon, I., Proehl, G., Kowe, R., Chen, Q.,Mobbs, S., Bergstroem, U., Hallberg, B., Katona, T., Eged, K., Kanyar, B., 2005.A comparative radiological assessment of ve European biosphere systems inthe context of potential contamination of well water from the hypotheticaldisposal of radioactive waste. Journal of Radiological Protection 25, 375e391.

Paasikallio, A., Rantavaara, A., Sippola, J., 1994. The transfer of cesium-137 andstrontium-90 from soil to food crops after the Chernobyl accident. The Scienceof the Total Environment 155, 109e124.

Peel, M., Finlayson, B., McMahon, T., 2007. Updated world map of the Kop-

l Radioactivity 115 (2013) 214e223peneGeiger climate classication. Hydrology and Earth System Sciences 11,1633e1644.

Pinedo, P., Thorne, M., Egan, M., Calvez, M., Kautsky, U., 2005. Consideration ofenvironmental change in performance assessments. Journal of EnvironmentalRadioactivity 84, 185e209.

Posiva, 2010. Interim Summary Report of the Safety Case 2009. Posiva Oy, Olkiluoto.Prhl, G., Chen, Q., Pinedo, P., Olyslaegers, G., Simon, I., Bergstrom, U., Hallberg, B.,

Kanyar, B., Kowe, R., Mobbs, S., Zeevaert, T., 2004. Biosphere models for safetyassessment of radioactive waste disposal based on the application of thereference biosphere methodology BioMoSA Final Report GSF-Bericht 06/04.GSF-Forschungszentrum.

Prhl, G., Olyslaegers, G., Kanyar, B., Pinedo, P., Bergstrm, U., Mobbs, S., Eged, K.,Katona, T., Simon, I., Hallberg, U., Chen, Q., Kowe, Zeevaert, T., 2005. Developmentand comparison of ve site-specic biosphere models for safety assessment ofradioactive waste disposal. Journal of Radiological Protection 25, 343e373.

Rosen, K., 1996. Transfer of radiocaesium in sensitive agricultural environmentsafter the Chernobyl fallout in Sweden. II. Marginal and seminatural areas in thecounty of Jmtland. The Science of the Total Environment 182, 135e145.

Rosen, K., Eriksson, A., Haak, A., 1996. Transfer of radiocaesium in sensitive agri-cultural environments after the Chernobyl fallout in Sweden. I. County ofGgvleborg. The Science of the Total Environment 182, 117e133.

Siebert, S., Dll, P., Feick, S., Hoogeveen, J., Frenken, K., 2007. The Digital Global Mapof Irrigation Areas. FAO.

SKB, 2011. Long-term Safety for the Final Repository for Spent Nuclear Fuel atForsmark: Main Report of the SR-Site Project, TR-11e01. Technical Report.Svensk Krnbrnslehantering AB.

Smith, G., Kozak, M., 2011. Historical development and evolution of EPRIs post-closure dose assessment of potential releases to the biosphere from theproposed HLW repository at Yucca Mountain. Health Physics 101 (6), 709e721.

StrlSchV, 2001. Strahlenschutzverordnung vom 20. Juli 2001 (BGBl. I S. 1714; 2002 IS. 1459), die zuletzt durch Artikel 2 des Gesetzes vom 29. August 2008 (BGBl. I S.1793) gendert worden ist", in: Justiz, B.d. (Ed.).

Wasiolek, M., 2003. Biosphere Model Report MDL-MGR-MD-000001. YuccaMountain Project, Las Vegas, Nevada (USA).

C. Staudt et al. / Journal of Environmental Radioactivity 115 (2013) 214e223 223

Modeling the impact of climate change in Germany with biosphere models for long-term safety assessment of nuclear waste rep ...1. Introduction2. Material and methods2.1. Model setup2.1.1. Assessment context and conceptual model2.1.2. Biosphere system description2.1.3. Potentially exposed group2.1.4. Model development

2.2. Selection of reference climate regions

3. Results and discussion3.1. Activity concentrations in soil3.2. Activity concentrations in food3.3. BDCF

4. ConclusionsAcknowledgmentAppendix A. Supplementary materialReferences