short rotation coppice for revaluation of contaminated land

Journal of

Environmental Radioactivity 56 (2001) 157–184

Short rotation coppice for revaluation ofcontaminated land

H. Vandenhovea,*, Y. Thirya, A. Gommersa, F. Goorb,J.M. Jossartb, E. Holmc, T. G.aufertc, J. Roedd, A. Grebenkove,

S. Timofeyevf

aSCK CEN, Boeretang 200, 2400 Mol, BelgiumbUCL, Place Croix du Sud 2-11, 1348 Louvain-la-Neuve, Belgium

cU.LUND, PO Box 117, 22100 Lund, SwedendRISOE, PO Box 49, 4000 Roskilde, Denmark

e IPEP, Sosny, Minsk, BelarusfRIR, Fedyuninsk 16, Gomel, Belarus

Received 8 March 1999; received in revised form 11 May 2000; accepted 12 July 2000

Abstract

When dealing with large-scale environmental contamination, as following the Chernobyl

accident, changed land use such that the products of the land are radiologically acceptable andsustain an economic return from the land is a potentially sustainable remediation option. Inthis paper, willow short rotation coppice (SRC) is evaluated on radiological, technical and

economic grounds for W. European and Belarus site conditions. Radiocaesium uptake wasstudied in a newly established and existing SRC. Only for light-texture soils with low soilpotassium should cultivation be restricted to soils with contamination levels below 100–370 kBqm�2 given the TFs on these soils (5� 10�4 and 2� 10�3m2 kg�1) and considering the

Belarus exemption limit for firewood (740Bqkg�1). In the case of high wood contaminationlevels (>1000Bqkg�1), power plant personnel working in the vicinity of ash conveyers shouldbe subjected to radiation protection measures. For appropriate soil conditions, potential SRC

yields are high. In Belarus, most soils are sandy with a low water retention, for which yieldestimates are too low to make production profitable without irrigation. The economic viabilityshould be thoroughly calculated for the prevailing conditions. In W. Europe, SRC production

or conversion is not profitable without price incentives. For Belarus, the profitability of SRCon the production side largely depends on crop yield and price of the delivered bio-fuel. Large-scale heat conversion systems seem the most profitable and revenue may be considerable.

*Corresponding author. Tel.: +32-14-335280; fax: +32-14-580523.

E-mail address: [email protected] (H. Vandenhove).

0265-931X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.

PII: S 0 2 6 5 - 9 3 1 X ( 0 1 ) 0 0 0 5 2 - 2

Electricity routes are usually unprofitable. It could be concluded that energy production from

SRC is potentially a radiologically and economically sustainable land use option forcontaminated agricultural land. # 2001 Elsevier Science Ltd. All rights reserved.

Keywords: Short rotation coppice; Contamination; Chernobyl; Remediation; Willow

1. Introduction

Following the Chernobyl accident, many thousands of square kilometres havebeen severely contaminated with radioactive fallout in the CIS. The application ofcorrective actions in contaminated areas remains a key issue in large territories ofRussia, Ukraine and Belarus. The application of simple clean-up operations leadingto a rapid return of such a large area to its normal state is unrealistic and would, inaddition, generate an enormous amount of radioactive waste. Consequently, animportant part of the contaminated agricultural land may be left deserted for manyyears, despite its agricultural potential remaining intact.When agricultural production must be abandoned in such territories because of

irremediably high activity levels in food products, an increasing interest arises indeveloping more integrated and ecologically based approaches. In this regard,industrial crops not used for food production, may be an alternative remediationoption.The research on the effectiveness of an alternative land use for the areas

contaminated by the Chernobyl accident is predominantly restricted to the screeningof varieties and species for differences in transfer of radionuclides to the edible parts.The reduction in activity concentrations was a factor 2–4 (Prister, Loshchilov,Perepelyatnikov, & Bondar, 1992; Alexakhin, 1993; Alexakhin et al., 1993).Information on the use of industrial crops for alternative land use in the event ofa contamination scenario is limited. If information is available, this is generallylimited to soil-to-plant transfer factors to primary and secondary products. Cropsfor sugar and oil production may be proposed since the contamination level in thefinal product is more than 10 times lower than in the harvested product (Alexakhinet al., 1993). For fibre crops, appropriate data are limited. For flax, reportedradiocaesium transfer factors (TFs) are 0.18� 10�3 and 0.23� 10�3m2 kg�1 forseeds and straw, respectively (GOPA, 1996). Recently, the transfer of radiocaesiumto the different plant compartments of flax and hemp (straw, fibres, leaves and seeds)was studied in an experiment using sandy soil (Vandenhove & Van Hees, 2000). Forflax, the transfer to fibres and straw was 0.06 (� 0.01)� 10�3 and 0.04(� 0.02)� 10�3m2 kg�1, respectively. For hemp, radiocaesium transfer to fibresand straw was 0.81 (� 0.23)� 10�3 and 0.44 (� 0.14)� 10�3m2 kg�1, respectively.Entry, Watrud, and Reeves (1999) found a soil-to-plant transfer of 137Cs to switchgrass of 10�2m2 kg�1. Except for flax, the fibre crops mentioned are also cultivatedas bioenergy crops (de Maeyer & Huisman, 1995; McLaughlin et al., 1997). Adetailed study on biodiesel sources in relation to soil decontamination in Belarus was

H. Vandenhove et al. / J. Environ. Radioactivity 56 (2001) 157–184158

conducted by GOPA (1996). They concluded that rape seed oil production forvegetable oil could be marginally valuable but use for biodiesel or lubricants isfinancially not viable. Radionuclide levels in rape seed oil are hence unlikely to be ofradiological concern. No information was found for other potential oil or liquidbiofuel crops.In SRC cultivation, fast-growing willows (Salix spp.) are intensively managed and

harvested for biomass in three to five year cutting cycles and a 22–25 year cropduration. Fast-growing willow species are among the fastest and highest biomassproducers when optimally provided with water and nutrients (Ledin, 1996). Theharvested biomass is converted into heat and power.Willow SRC can grow on a wide variety of soils. Harvest is in winter when farm

labour is available. Since SRC is a perennial crop, nutrients are partly recycledthrough litterfall and hence fertiliser requirements are low. Furthermore, erosion isdecreased. Water demands are, however, high. Specialised machinery for cropestablishment and harvest is not needed, though preferred. The cultivation is notlabour intensive, an advantage in contamination conditions. High energy efficiencies(energy output/energy input) between 20 and 30 can be reached if willow SRC isgrown under appropriate climate and soil conditions (Biewinga & van der Bijl, 1996;Goor, 1998).The emphasis in the present study is on the evaluation of short rotation coppice

(SRC) for energy purposes as an alternative usage of contaminated farm land withrestricted uses. The harvested biomass would be converted into heat or electricity.Two different regions are considered: most importantly Belarus, a large part ofwhich was severely contaminated as a result of the accident in the Chernobyl nuclearpower plant, and Western Europe, where energy crops have already been studiedextensively, for economic and ecological reasons.The flux of radiocaesium in SRC has been studied to obtain information on the

radionuclide levels in the wood (exploitable plant part) and in the ashes afterconversion. The radiation dose received during coppice culturing, subsequenthandling and conversion is assessed. The growth potential of SRC for the Belarusscenario is discussed. Finally, an economic cost benefit analysis of the productionand conversion of SRC is performed.

2. Radiocaesium levels in coppice wood

When proposing non-food crops as potential alternative land use for severelycontaminated land, information on the fluxes of radionuclides in the cultivation (andconversion) system proposed is required in order to conduct a robust assessment.Since SRC has not yet been considered as a possible re-valuation option forradioactively contaminated land, the flux of radionuclides should be studied toobtain information on the radionuclide levels in the wood (exploitable plant part)and in the ashes after combustion. Emphasis will be on radiocaesium, the majorradionuclide in the environment contaminated by the Chernobyl fallout.

H. Vandenhove et al. / J. Environ. Radioactivity 56 (2001) 157–184 159

Radiocaesium behaviour is studied in detail in a willow coppice ecosystem byinvestigating the soil-to-plant transfer and radiocaesium cycling on an experimentalscale (detailed study). Farmers’ coppice stands established in N. Sweden after theChernobyl accident were sampled to have a broader basis of transfer factors fordifferent soil types, coppice clones and age classes. Test sites in Belarus wereestablished to screen the radionuclide transfer and yield a potential of four coppicevarieties in the circumstances of elevated contamination, continental climate andpoor soil conditions.

2.1. Transfer and cycling of radiocaesium in an SRC culture at experimental scale

A study of radiocaesium cycling based on the concentration and biomass of thedifferent plant compartments (and especially in wood) during the growing seasonand over the years enables us further to compare willow SRC with traditionalforestry or other alternative non-food crops. Several authors mentioned somesimilarities between radiocaesium and potassium in plants (Bunzl & Kracke, 1987;Broadley & Willey, 1997). If similarities between caesium and potassium are foundfor SRC, an extension of radiocaesium behaviour to the following cropping cyclescould be made based on what is known for potassium. Recycling of radiocaesium tothe soil by litter fall and by throughfall is evaluated. Uptake and retranslocationpatterns for radiocaesium were compared to those of potassium.

2.1.1. Materials and methodsThe upper 30 cm of the original sandy soil of 8 experimental plots (2� 2m2) were

replaced with the upper soil layer of an orthic luvisol (loamy soil, 4 plots) or anorthic podzol (sandy soil, 4 remaining plots) (FAO, 1990). At the start of theexperiment, the soils were homogeneously contaminated with 134Cs by mixing theupper 25 cm (common ploughing depth) with a 134CsCl solution (about8� 106 Bqm�2). Final contamination levels were 24� 6 103 Bq kg�1 dry soil forthe loamy soil and 30� 4 103 Bq kg�1 dry soil for the sandy soil (differences in bulkdensity account for differences in contamination levels when expressed on weightbasis). Soil was sampled at the start of the experiment and at the end of each growingseason. Selected initial soil characteristics are summarised in Table 1. Soil analysisresults from other sampling occasions are mentioned in the text whereverappropriate. Details on soil analysis methods are given by Gommers et al. (2000).Exchangeable K and the concentration of K in the soil solution were measured eachyear at the start and at the end of the growing season.Two weeks after soil contamination (May 1996), plots were planted with willows

(Salix viminalis L. Var. Orm) with a density equivalent to 52 500 plants ha�1 (21plants per plot). This relatively high density allowed for destructive samplings at theend of each growing season. Plant densities during the second and the third growingseason were 42 500 and 35 000 plants ha�1, respectively. All results of biomassproduction or mineralomass are recalculated to the final plant density. One or tworows of plants were planted around and between the plots as a buffer zone.Fertilisers were added at the beginning of the second growing season (June 1997)


only, at a rate of 80 kgNha�1, 15 kgP ha�1 and 40 kgKha�1. Pests and diseaseswere controlled year-round and additional irrigation was provided intermittentlyduring summer.Plants were coppiced (cut-back to promote regrowth) at the end of the first

growing season (December 1996). At the end of each growing season (R1S1, R2S1and R3S2 plants: RxSy: Root x years old; Stem y year old), 3 or 4 entire plants perplot were sampled (stems, roots and cuttings).During autumn, the litter was collected on nets covering the entire surface of the

plot (0.1m above the soil). The collected litter was put back to the plot following thecollection and analysis. Additionally, stems and leaves of R2S1 and R3S2 plants (2 or3 shoots per plot) were sampled in April, June, August and October. On all samplingoccasions, standing biomass was estimated from length (R2S1 plants) or diameter(R3S2 plants) following a relation ‘‘weight ¼f (length or biomass)’’ established withthe stems sampled. All plant parts were separated and dry weight determined(1058C). Plants were analysed for 134Cs concentration (g-counting) and Kconcentration (calcination for minimum 24 h at 5008C followed by mineralisationwith HCl (38%), followed by atomic absorption spectrometry). All activitiesreported were corrected for radioactive decay with respect to the planting date.Radiocaesium Transfer Factors (TF, m2 kg�1) are calculated as follows:

TF ¼Cs½ �plantCs½ �soil

ð1Þ

with radiocaesium concentrations ([Cs]) in Bq kg�1 for plants and Bqm�2 for soil.Leaching of 134Cs and K from the foliage was studied from August 1997 tillDecember 1998. Rainwater below the plant canopy was continuously collected with

Table 1

Selected soil characteristics of the orthic luvisol and the orthic podzol at the start of the experiment [data

are means and standard deviations (in parentheses) of the 4 replicate plots]

Soil characteristics Orthic Luvisol}Loamy soil Orthic Podzol}Sandy soil

CEC (cmolc kg�1) 10.6 (0.7) 6.7 (0.9)

Texture (%) 100–50mm 10.7 (0.7) 90.6 (4.9)

50–20mm 40.6 (2.6) 5.8 (0.3)

20–10mm 31.7 (1.6) 0.6 (0.1)

10–2mm 2.0 (0.03) 0.7 (0.3)

52mm 15.0 (0.3) 2.3 (0.5)

Total C (%) 1.0 (0.4) 3.7 (0.7)

RIPa (cmolc kg�1) 332 (76) 45 (13)

pH (KCl) 6.9 (0.2) 4.6 (0.1)

Exchangeable cationsb (cmolc kg�1) Kþ 0.85 (0.05) 2.02 (0.02)

Ca2þ 1.43 (0.57) 3.65 (0.35)

Mg2þ 1.34 (0.03) 6.39 (0.17)

aRIP=radiocaesium interception potential.b Initial values (May 1996).


two gutters per plot and analysed for 134Cs (concentrated on resins) and Kconcentration (AAS).

2.1.2. ResultsSoil characteristics: The two soil types are characterised by a clearly different

texture (Table 1). The higher radiocaesium interception potential (RIP) value for theloamy soil type is a likely consequence of its finer texture. The sandy soil has a highercarbon content. For both soils, the exchangeable potassium content remains more orless constant during the period studied. The potassium concentration in the soilsolution, however, decreases each year during the growing season (from March toDecember) and increases again a little during winter. The decrease in Kconcentration is largest in the first growing season, when the biomass productionstarted from (nearly) zero. Also in the beginning of the second growing season theabove-ground biomass is zero and biomass production very high. However, the soilswere fertilised in June and are not depleted. In the third growing season, the Krecycled by foliar leaching and litter and root decomposition is sufficient to sustaincrop growth (results not shown, for detail, Vandenhove et al., 1999). Biomassproduction: Under the experimental conditions, woody biomass production wassimilar for both soil types. Plant growth was characterised by a rapid biomassproduction in the beginning of the growing season that slows down afterwards(Fig. 1). 1.0 and 1.4 t of dry wood per hectare were produced during the first year ofthe culture (R1S1 plants) on the loamy and the sandy soil, respectively. In December,the shoots were cut back and stools resorted the following growing season.R2S1 plants produced 8.5 t drywood ha�1 on the loamy soil and 10.0 t ha�1 on the

sandy soil. Woody biomass production during the last growing season accounted for12.1 t and 12.9 t ha�1, n the loamy and sandy soil, respectively.

Fig. 1. Biomass production in a willow SRC on experimental plots.


Evolution of radiocaesium accumulation in the wood with time: Radiocaesiumaccumulation in the willow wood depends both on biomass production and onevolution of radiocaesium concentration in the wood. Since the radiocaesiumconcentration in the wood was not constant during the three years studied, netaccumulation was not proportional to biomass production (Fig. 2).While the evolution of biomass production was similar for the two soil types

(Fig. 1), the evolution of radiocaesium accumulation differs due to differentevolutions of radiocaesium concentration. The large decrease in radiocaesiumconcentration from the first to the second year for plants grown on the loamy soilmay be due to a rapid fixation process of radiocaesium on the clay minerals aftercontamination. On the sandy soil, this process may have been less important (lower% of clay minerals, Table 1). Moreover, radiocaesium seemed to become more

Fig. 2. Evolution of radiocaesium concentration (diamonds and lines) and radiocaesium content (bars) in

the wood of willows (error bars are standard deviations, n ¼ 4).


available for the plants on the sandy soil, since radiocaesium concentration in woodincreased from the first to the second growing season. This increase may be explainedby a decrease in K concentration in the soil solution at the beginning of the secondgrowing season during rapid plant growth and before fertilisation. Because allabove-ground plant biomass was newly formed in 1997, the necessary K could not besupplied by what was already in the wood and K concentration in the soil solutionmay have decreased. The decomposition of the litter from the previous year mayhave contributed to the higher uptake on the sandy soil. Indeed, radiocaesiumreturned to the soil by litter fall during the first growing season accounted for3102 kBq ha�1 on the sandy soil, while this was only 340 kBq ha�1 for the loamy soil.Wood radiocaesium concentrations for the sandy soil decreased during the lastgrowing season. A reason may be that the SRC system became almost autosufficientin K. This stabilisation of the radiocaesium concentration in the wood for lighttextured soils is important for predictions of caesium levels in wood in more maturecultures. Since after each harvest the SRC is fertilised (NPK), SRC will not depletethe soil K and the TFs are not expected to be higher than the TFs recorded at thefirst rotation.Radiocaesium concentrations in the wood at the end of the third growing season

were 16 and 420Bq kg�1 for the plants grown on the loamy and the sandy soil,respectively (soil contamination 8� 106 Bqm�2). This is lower than the exemptionlimit for fuel wood suggested in the CIS of 740Bq kg�1 dry wood (Szekely, Amiro,Rasmussen, & Ford, 1994). Radiocaesium TFswood was 2� 10�6m2 kg�1 for willowson the loamy soil and 5� 10�5m2 kg�1 for willows on the sandy soil at the end of1998. Consequently, given the present results and under Belarus exemption limits,willow SRC can be established on a sandy soil (with similar soil properties as the onestudied) contaminated up to 14MBqm�2, or a loamy soil contaminated up to370MBqm�2. It should be mentioned that these results are for two-year-old stemsand that normally coppice is harvested after 3–5 years only. However, given thedecreasing trend in wood contamination level with time (Fig. 2) and since there was nosignificant effect of stand age found on the TF in the coppice sampled in Sweden (seefurther), the above statement on permissible soil contamination levels is a safe one.

Evolution of radiocaesium and potassium in willow SRC throughout the growingseasons: Only data for the willows grown on the sandy soil during the third growingseason (1998, R3S2) are discussed in detail. On the loamy soil, similar trends to thosefound on the sandy soil were obtained. The increase of radiocaesium accumulationin the biomass (Fig. 3) is comparable with the biomass increase during the first halfof the growing season (see earlier). However, while standing biomass remains moreor less constant in autumn, the total amount of radiocaesium accumulated in theabove-ground plant parts decreases. A part of the radiocaesium incorporated in theabove-ground plant parts is retranslocated to the below-ground parts (differencebetween maximal content during the growing season and the final amount in thewood, litter and throughfall water). Of the maximum amount of radiocaesiumpresent in the above-ground biomass (August 98), 44% is retranslocated to the roots.Only 0.01% (9.2MBqha�1) of soil radiocaesium was newly taken up by the plants

during the third growing season of which 39.5% is immobilised in stem and


branches and 0.5% in the roots. 60% is returned by litterfall and leaching. Theincrease in root radiocaesium content during 1 growing season is only 1% of the netamount taken up. On the other hand, 44% of the maximal total amountaccumulated in the above-ground biomass is retranslocated to the roots. Thisimplies that the roots have served as a source of radiocaesium at the start of thegrowing season.The radiocaesium balance at the end of 1998 for the total plant (above- and below-

ground plant parts) was as follows: the R3S2 plants (roots+wood) contained19.6� 106 Bq 134Cs ha�1, of which 9.7� 106 Bq (50%) was found in the wood andas much as 9.9� 106 Bq ha�1 in below ground plant parts (roots and cuttings). Sinceduring the third growing season only an extra 1% is accumulated in the roots, theamount of caesium in the below-ground plant parts is seemingly stabilising. The9.9� 106 Bq ha�1 allocated in the roots represents only 0.01% of the total soil

Fig. 3. Evolution of radiocaesium in above-ground plant parts (top) and cycling in terms of total uptake,

return and immobilisation (bottom) in a willow R3S2 SRC stand on a sandy soil.


radiocaesium. At the end of an SRC culture practice, the roots will be grubbed upand the Cs in the roots may be available for plant uptake by the subsequent crop.Since the Cs in the roots represent only 0.01% of the Cs-pool in the soil, this fractionis not considered significant.Trends in radiocaesium and potassium cycling are comparable but redistribution

patterns in the plants are not. For example, at the end of 1998 (third growingseason), potassium was principally incorporated in the wood (50.6 kgKha�1). Lessthan half of this was found in the below-ground plant parts (21.8 kg ha�1). Forradiocaesium, the largest part was found in roots and cuttings. This means that wecannot rely on the fate of K to predict the fate of radiocaesium in an SRC-system insubsequent rotation cycles.

2.2. Radiocaesium accumulation in willows of different ages and grown on different soiltypes at established farm-scale sites

The major objective of the extensive sampling at farm plots in Sweden was tohave a sufficiently large data set of coppice TF linked to different soil types, coppicevarieties and age categories to broaden the basis for estimating radiocaesium levels inthe coppice wood.

2.2.1. Materials and methodsExisting willow SRC plots were sampled once in central Sweden (north of

Uppsala) between November 1997 and April 1998. In this area, deposition fromChernobyl was amongst the highest in Sweden. The deposition at the plot sampleswas between 15 and 90 kBqm�2. The plots were of different ages (all established after1986), different clones and different soil types. Ten shoots per plot were cut, thewood was ashed to concentrate the radiocaesium activity and 137Cs in the ash wascounted in a HPGe-detector. The soils were analysed for the soils of the experimentalplots. For the analysis methods we refer to Gommers et al. (2000). The sampled plotsdiffered in their soil characteristics (Table 2). Few sandy soils were sampled (Tr .odje,Tierp 1, Tierp 2 1M). Most soils were very clayey differing, however, in organicmatter content and/or nutrient status.

2.2.2. Effect of stand maturity, clone and soil type on radiocaesium uptakeSoil-to-wood TFs ranged over three orders of magnitude: from 2.4� 10�6 to

1.4� 10�3m2 kg�1 (Table 2). There was no effect of stand maturity, biomass yield orclone on TF.Notwithstanding the very different values for the radiocaesium interception potential

(RIP) , the TF values between some plots did not differ much (e.g. Viksta 1 and Tierp 21G). Results from a simple regression analysis were consequently not satisfactory.The uptake of radiocaesium was evaluated as a combination of two processes: the

dynamic equilibrium between radiocaesium sorbed on soil particles and radio-caesium in the soil solution (KD, L kg�1) and the uptake of radiocaesium by the plantfrom the soil solution (Smolders, Van den Brande, & Merckx, 1997). Therefore, thetransfer factor (here defined as the radiocaesium in the plant (Bq kg�1) to the


Table 2

Plant variety, age, TF and soil characteristics for the different willow plots in Sweden

Site Variety Age

RxSy

TF

(m2 kg�1)

RIP

(cmolc kg�1)

CEC

(cmolc kg�1)

Exch. K

(cmolc kg�1)

mK

(mM)

pH

(KCl)

OM

(%)

Texture (%, mm)

52 2–20 20–200 >200

Viksta 1 Rapp R5S4 8.2� 10�5 688 20.1 0.34 0.12 6.19 3.9 30 22 38 9

183 R6S2 3.8� 10�5

Viksta 2 183 R6S2 3.1� 10�5 543 21.7 0.37 0.20 5.10 5.1 32 23 43 2

Viksta 3 183 R6S2 5.4� 10�5 545 25.0 0.36 0.09 5.02 4.8 33 25 39 3

Tierp 1 183 R6S2 3.2� 10�5 152 8.1 0.27 0.38 5.10 4.3 6 5 86 3

Tierp 2 1G Rapp R5S4 1.4� 10�5 406 9.5 0.24 0.19 5.79 2.3 15 13 70 2

Tierp 2 1M Rapp R5S4 1.3� 10�5 91 10.6 0.15 0.31 5.41 7.3 4 5 30 54

Bj .orkl. 2 183 R8S2 3.2� 10�5 523 14.4 0.23 0.15 6.20 3.1 22 33 38 7

Bj .orkl. OR Rapp R6S2 2.4� 10�6 493 22.4 0.58 0.16 5.01 3.2 34 30 34 2

Bj .orkl. 1 183 R7S2 9.0� 10�6 387 22.2 0.21 0.34 4.66 9.1 31 26 42 1.Osterf.ar. 183 R7S6 2.8� 10�5 343 8.6 0.24 0.32 4.24 3.0 14 16 48 22

Tr .odje 1 183 R4S4 8.9� 10�4 56 3.9 0.11 0.48 5.20 2.9 3 2 23 71

Tr .odje 2 183 R4S4 1.4� 10�3 56 4.5 0.08 0.29 6.59 2.9 3 2 23 71

Tr .odje 3 183 R4S4 1.0� 10�3 58 3.8 0.1 0.39 6.58 2.9 3 2 23 71

Tr .odje 4 183 R4S4 5.8� 10�4 52 3.1 0.18 0.95 5.25 2.9 3 2 23 71

H.

Vanden

hove

etal.

/J.

Enviro

n.

Radio

activity

56

(2001)

157–184

167

radiocaesium in the soil (Bq kg�1) can be written as

TF ¼ CF=KD; or logðTFÞ ¼ logðCFÞ � logðKDÞ ð2Þ

with CF (concentration factor) the ratio of radiocaesium in the plant (Bq kg�1) tothe radiocaesium in the soil solution (BqL�1). Radiocaesium concentrations in thesoil solution and thus CFs can be calculated by taking ageing into account accordingto Absalom et al. (1999). The log(CF) can be estimated based on the K concentrationin the soil solution ðmKÞ ½b0 � b1�logðmKÞ, Fig. 4]. The decreasing CF withincreasing mK is also found in different hydroculture experiments for different crops(Smolders, Kiebooms, Buysse, & Merckx, 1996; Buysse, Van den Brande, & Merckx,1996). Especially at very low K concentrations, the uptake of radiocaesium increasesvery sharply. The CF decreases more than proportionally to the reciprocal of the Kconcentration.From Fig. 4 it is clear that the Tr .odje plots differ from the other plots. Trendlines

(power functions) in the figure are best fits for both series. For the same Kconcentrations in the soil solutions, higher CFs are found for the plots in Tr .odje.Ageing, and hence the CF, may have been overestimated for the sandy Tr .odje plots.However, the soils of Tierp 1 and Tierp 2 1M also have low clay contents,comparable to the results in Tr .odje. Potentially (hypothesised) higher NH4

+

concentrations in the soil solution at Tr .odje may also have increased the CF atcomparable K levels (Minotti, Craig, & Jackson, 1965). A higher organic mattercontent could not explain the higher CF estimated for Tr .odje soils. Finally, thedifference may be due to a difference in clay mineralogy of the soils and thus differentselective retention properties. Indeed, Tr .odje is located more to the north than theother plots and may have different parent material.

Fig. 4. Radiocaesium CFs ([Bq kg�1 dry wood]/[BqL�1 soil solution]) in relation to the K concentration

in the soil solution (mM) for the plots in Tr .odje and for the other Swedish plots.


From the predicted CFs, TFs can then be calculated following Eq. (2) based on theKD value and mK according to the formula ½logðTFÞ ¼b0 � b1 logðmKÞ � logðKDÞ�.From a non-linear estimation (b0 ¼ 5:06� 0:33 and b1 ¼ 1:65� 0:21), this functionexplains 63.0% of the variation between the plots. However, when we leave out thedata for Tr .odje, less than 10% of the variation is explained.

2.3. Case study in Belarus

For two major reasons, field trials were established in Belarus on formeragricultural land, left uncultivated after the Chernobyl disaster. Firstly, yield andmost probably also radionuclide uptake and distribution are highly climate- and soil-dependent. Results obtained in other trials (Belgium, Sweden) would therefore notnecessarily befit the Belarus situation. Secondly, the contamination in Belarus was ofa different nature (more liable to the presence of hot particles than in Sweden) andmore aged (due to gradual fixation of radiocaesium by the soil particles) than in thetrial set-up in Belgium.

2.3.1. Materials and methodsAt Savichy, located on the edge of the exclusion zone, coppice trials were

established on a sandy and a peaty soil (May 1997). Four clones, selected for highyields and frost and pest resistance (Vandenhove et al., 1999) were planted in 4replicates at all four sites. The four clones selected were Rapp (Salix viminalis), Orm(S. viminalis), Jorr (S. viminalis), and Bj .orn (S. viminalis x S. schwerinii).Soil was analysed as in the other experiments and the results are presented in

Table 3. At the end of the 1997 growing season, the whole plot was harvested bycutting back the willows. At the end of each growing season (November 1998 and1999), the stems of 10 plants per plot were cut and homogenised. A subsample wasoven-dried, ground and analysed for 137Cs activity.During the growing season, measurements were performed for growth modelling.

2.3.2. Effect of soil characteristics and willow clone on radiocaesium uptakeThe soils are rather poor in potassium: only about 1% (peaty) to 5% (sandy) of

the regular cation exchange complex is occupied by potassium (Table 3). Soils with alow potassium status are rather liable to high radiocaesium transfer as is the low pHof both soils. Notwithstanding, exchangeable Cs-levels are low: 3% or less. Valuesfor exchangeable cations and exchangeable Cs hardly changed from one growingseason to the other.On the peaty soil, yield is comparable with yield obtained in W. Europe; on the

sandy soil, growth is limited due to extreme water stress. Yield figures are discussedin the section on biomass productivity.Radiocaesium TFs are up to a factor of 1000 higher than the TFs recorded on the

Belgian lysimeters and in Sweden (except for the Tr .odje sites) (Fig. 5). They rangebetween 0.24� 10�3 and 2.60� 10�3m2 kg�1 on the sandy soil and between0.49� 10�3 and 2.21� 10�3m2 kg�1 on the peaty soil at Savichy. It is clear from


Fig. 5 that due to the large variation there is no significant difference in TF betweenthe different willow clones.The high TFs obtained are comparable with TFs at the Swedish Tr .odje sites

(0.58� 10�3–1.4� 10�3m2 kg�1), where soils also showed low RIP values(52–58 cmolc kg

�1, Table 2) close to RIPs at Savichy (17.5–33.3 cmolc kg�1).

It is unclear why TFs in Belarus are a factor 100 higher than on the sandy soil inBelgium, in spite of the fact that in Belgium the contamination was rather recent(1996) compared to Belarus (1986) and the radiocaesium was assumed to be lessavailable in the latter scenario. Extreme potassium depletion may be at the basis ofthe observation. At the start of the experiment, the K in the soil solution wascomparable for the Belarus and Belgian sandy soil (around 0.25mM). Theexchangeable K was, however, only 0.043 cmolc kg

�1 for the Belarus sandy soiland 2.02 cmolc kg

�1 for the Belgian soil. The extremely low K-buffering capacity ofthe Belarus soil may have resulted in an extreme K-depletion during the growingseason, resulting in the high Cs-incorporation. The absence of Cs-fixing clayminerals, the lower soil and the likely presence of NH4

+ due to decomposition of theorganic matter in the peaty soil may provide additional explanations.Given the high TF but more importantly due to the extremely high deposition

levels at the test sites, the concentration of Cs in the wood (Table 4) was higher thanthe level acceptable for fuel wood which is set at 740Bq kg�1 in Belarus. If weconsider an average transfer factor of 2� 10�3m2 kg�1 during the second growingseason and assume that TF would not increase further, willows can only becultivated on sandy or peaty soil contaminated up to 370 kBqm�2.

Table 3

Soil characteristics of the Belarus soils ( standard deviations in parentheses)

‘‘Permanent’’ soil characteristics Savichy peaty Savichy sandy

137Cs (kBqm�2) 19 385 (4784) 1438 (308)137Cs (Bq g�1) 210 (53) 4.05 (970)

CEC (cmolc kg�1) 41.7 (1.2) 1.03 (0.42)

RIP (cmolc kg�1) 38.8 (13.6) 17.5 (5.5)

Total C (%) 48.60 (0.87) 1.47 (0.24)

Apparent density (kgL�1) 0.37 (0.01) 1.42 (0.02)

Texture 510 mm 6.5 (0.6)

10–50mm 8.8 (0.9)

50–250mm 72.6 (3.2)

250–500mm 11.4 (3.7)

>500mm 0.7 (0.1)

Exchangeable cations (cmolc kg�1) K+ 0.55 (0.09) 0.043 (0.006)

Ca2+ 33.35 (2.19) 0.68 (0.38)

Mg2+ 3.73 (0.58) 0.19 (0.03)

pH 4.5 (0.1) 3.9 (0.2)

K in soil solution (mM) 0.256 (0.134) 0.092 (0.046)


Table 4

Radiocaesium concentrations (Bqkg�1) in the willow wood

1997 1998

Savichy sandy 756 (� 563) 3015 (� 2363)

Savichy peaty 14 180 (� 8186) 36 657 (� 2363)

Fig. 5. Radiocaesium soil-to-wood transfer factors (R2S1 and R3S2) at Savichy in Belarus.


SRC is not yet cultivated for energy production in Belarus. Afforestation isanother possible alternative allocation of contaminated arable land. Initialinvestments are comparable with SRC, but returns show only after 60 years ormore, except for the limited revenue from the thinning. At present, no information isavailable on transfer factors or radiocaesium fluxes in forests that were planted afterthe Chernobyl accident. For the comparison of TFs and fluxes between willowvegetation system and forests we have to rely on data from forests in place at thetime of the accident. Soil-to-wood TFs to coniferous and deciduous wood rangebetween 6.7� 10�3 and 2.8� 10�3m2 kg�1 (Ponomarev, 1997; Zadbudko, Petrov,Kozmin, & Svetove, 1995). They are comparable with the high TFs observed inwillows in Tr .odje and at the Belarus plots and a factor 10–100 higher than the TFsobserved in the other Swedish plots and at the Belgian test sites. Potential biomass isalso an important parameter for system profitability. Since biomass production bywillows is higher than for forest (12 t for SRC compared to 6 t ha�1 for forests),wood production on contaminated arable land in an SRC system seems to be a goodalternative.

3. Fate of radiocaesium during conversion

To reliably estimate the transfer from biofuels to ash it is important to know theash content in the fuel and the proportion of radiocaesium escaping with flue gasesand condensation water.From the information collected under the RECOVER project (Vandenhove et al.,

1999), it could be concluded that the main part of the radiocaesium can be found inthe ash in a modern biofuel-fired heating plant. The enrichment factor is generallyhigher for fly ash than for bottom ash. Mainly, fly ash is produced but the ratio ofproduced fly ash/produced bottom ash has a large variation. With an appropriatefiltering system, more than 99% of the fly ash can be separated from the flue gases.The separation of radiocaesium can be a few percent lower due to higher enrichmentof radiocaesium and lower collection efficiency for fine ash particles. For theradiocaesium cycling tests in a pilot gasification plant, Cs recoveries were about95%. An average ash percentage for wood of 2% was taken (based on experimentsto determine ash content for willow wood, results not shown. This is also the averagewood ash content as proposed by Hedvall, 1997).

3.1. Results

If the high TFs for coppice in Belarus apply, use of wood for combustion inhousehold wood stoves should be considered with care. It should be emphasised thatthe TFs obtained for Belarus were recorded on very poor soils, which were,moreover, out of production for more than 10 years when the coppice trials wereestablished. Their K-status was very low and it is known that at low K Cs-TFincreases exponentially (Smolders, Van den Brande, & Merckx, 1997, Nisbet,Woodma, & Haylock 1999). The mean TF for Sweden, of 10�5, and of


5� 10�5m2 kBq�1 recorded for the Belgian sandy soil, are probably more realisticTFs for Western European soil conditions.Cs-levels in the exhaust gas were calculated from the Cs emissions from a pilot

gasification unit (Vandenhove et al., 1999). These calculations are conservative inthat the filtering system used was not optimal. When the low-end TF is applied, Csemissions are always below the exemption limit for gaseous effluents. With thecoppice TFs recorded in Belarus, permissible levels for the public are exceeded whenwillows are cultivated in areas with contamination levels exceeding 620 kBqm�2.Dose limits for workers are not exceeded. The values mentioned in Table 5 (see alsoTable 6) are, moreover, conservative since the exhaust emissions always mix with thesurrounding air. This assumption is corroborated by the few mSv dose contributionfrom atmospheric release of contaminated fly ash from wood with 3000Bq kg�1137Cs (see under 4).A comment on the exemption level of 740Bq kg�1 for firewood is in order. The

level of 740 Bq kg�1 is for the use of fuelwood by the general public. Limits forfuelwood levels for combustion at industrial scale with adequate filtering followed byappropriate ash-handling could be considerably higher.Following the remediation activities in the Chernobyl-contaminated territories, a

large volume of radioactive waste was generated (called ‘‘Conventionally Radio-active Waste’’, CRW) with activity levels below the low level waste (LLW, 103–105 Bq kg�1). Nearly all the bulk of ash from contaminated biomass-fired facilities issupposed to belong to CRW and the lowest range of LLW. The State Standard ofthe Russian Federation, No. GOST(R) 22.8.02 states that radioactive waste from

Table 5

The TFs of coppice wood and ash for 3 contamination scenarios. Values in bold indicate that the

exemption limits presented in Table 6 are exceeded

Deposition, kBqm�2

185 555 1480

TF (103m2 kg�1) Contamination level (Bq kg�1)

Coppice, moderate-high RIP and K

Wood 0.01 1.9 5.6 15

Ash (2%)a 0.5 92 277 740

Cs in exhaustb 0.06 0.18 0.48

Coppice, low RIP and K

Wood 2 370 1110c 2960

Ash (2%) 100 18 500 55 500 148 000

Cs in exhaustb 11.2 35.7 95

aAsh content of coppice after calcination. Also mean ash content of wood according to Hedvall (1997).bData from the pilot gasification unit were used: with a measured dust loading of 0.87 gm�3 and a dust/

wood concentration factor of 37 (Vandenhove et al., 1999).cA component is only considered as waste if the exemption limits presented in Table 6 are exceeded. Ash

is only considered as waste if the 137Cs concentration is higher than 1000Bqkg�1, the lowest limit for

LLW.


agriculture is to be disposed of or used to extract economic values by energyproduction or secondary valuable products (cellulose, pulp, chemicals, etc.) otherthan food. Incineration is considered as an appropriate technology to convert thecombustible agricultural waste into ash. Incineration can be applied if the activity ofthe initial bio-fuel does not exceed 3.7MBqkg�1. This is higher than the acceptablelevel for fuel wood but IAEA (1992) notes that when a large area is contaminated, asis the case after a nuclear accident, nuclear waste handling may have to be performedoutside the IAEA regulations. The ashes should be confined to prevent resuspensionand be disposed of outside any settlement border. A simple repository (infiltrationcoefficient 510�5 cm s�1 and dose rate at 1m 528 mSv h�1) with an expected cost of8Eurom�3 waste would comply with the requirements (Grebenkov, 1997, seeVandenhove et al., 1999).Since biomass containing up to 3.7MBqkg�1 may be incinerated under specified

controlled conditions, and considering the coppice TF for Belarus(2� 10�3m2 kg�1), this would imply that coppice can be cultivated in areascontaminated up to 1850MBqm�2. But, in such a scenario, cultivation would resultin an exceedingly high external exposure (>1Sv a�1).

3.2. Conclusions

For soils with a medium to high radiocaesium interception potential (RIP) andsufficient soil potassium, transfer factors range between 0.0024� 10�3 and0.08� 10�3m2 kg�1. Only 0.01% of the total soil radiocaesium is immobilised inthe wood. For soils with a low RIP and low soil K, TFs are between 0.5� 10�3 and2� 10�3m2 kg�1. In the latter scenario, SRC cultivation should be restricted to areaswith a deposition of 370 kBqm�2 or lower. However, if materials with contamina-tion levels up to 3.7MBqkg�1 may be burned in controlled electricity or heat plants,willow SRC cultivation may be advised in regions with much higher contaminationlevels. Adequate exhaust filtering systems should be installed in these circumstancesand appropriate disposal of the ash should follow. In the case that the soil-to-woodTFs apply for soil with a high RIP and sufficient K-status, levels in the wood will bebelow the exemption limit for fuel wood (where deposition is less than100 000 kBqm�2).

Table 6

Exemption limits for radiocaesium in fuelwood, low-level waste (both for Belarus) and gaseous effluents

(Belgian KB, 1963 Art. 36)

Item Limit

Fuelwood (general: public, industry) 740Bqkg�1

Incineration of contaminated biomass

(special regulations, authorised power plants only)

3.7MBqkg�1

Low-level waste 1000–100 000Bqkg�1

Gaseous effluents Public 40Bqm�3

Radiation workers 2000Bqm�3


We are not yet in a position to estimate the radiocaesium concentration in thewood from soil parameters. K seems to play an important role but cannot explaindifferences between TFs obtained for willows in different regions. Clay mineralogyand hence radiocaesium fixation potential (ageing) may play an important role. InK-poor soils, fertilisation may decrease the radiocaesium TF considerably. Since Csand K have different distribution patterns among the plant compartments, Cs levelsin the wood in subsequent rotation cycles cannot be predicted from what is knownfor potassium.There was no clear effect of stand maturity or willow clone on the TF.The TFs for willow wood are generally smaller than for common forests in place

at the time of the accident. Since the biomass production in an SRC is higher (ifwater supply and soil fertility are adequate) and the revenue faster and more regularthan in common forestry, willow SRC seems a valuable alternative compared toforests.

4. Radiological aspects of SRC production and conversion

Producing coppice fuel wood on contaminated land may lead to enhancedradiation doses to workers involved in production and conversion. Establishing,maintaining and harvesting the crop results in external exposure from radiocaesiumdeposited on the ground. Further, there is the exposure during transport of the fuelto the combustion site. Workers may be exposed at the power plant. Wood-firing of137Cs-contaminated fuel causes an enrichment of 137Cs in the ash produced. Theconcentration of 137Cs in the ash is usually 40 to 80 times higher than in the woodfuel. This makes deposits and containers where ash is accumulated critical sites froma radiological point of view. The following is a summary of dose calculations fordifferent situations relevant to production of energy from contaminated willow.

4.1. External effective doses at a radiocaesium-contaminated cultivation field

One important parameter from a radiological point of view is the time spent at acontaminated area. In SRC production, time is spent at the field for establishment,management and harvesting. A high mechanisation level will result in higherworking efficiency and better shielding, both entailing lower doses. The methodologyused for the calculation of the radiation doses is described in Finck (1992).For a Western European situation, coppice is not a labour-intensive culture

compared with a normal agricultural cultivation system (1.14 h ha�1 a�1 comparedto 4 h ha�1 a�1). For a deposition level of 1480 kBqm�2 and without ploughing,this would result in a yearly dose of 3.3 mSv ha�1 (no shielding from machinery isconsidered). For a ploughing depth of 25 cm, the exposure would reduce by a factorof 5. If less adapted planting or harvesting machinery is used (e.g. adapted machinefor planting vegetables or for harvesting corn) as may be the case in Belarus, theannual dose averaged over the whole rotation cycle (25 years) could amount to3.8 mSv ha�1. If all planting and harvesting would be done manually, the doses in the


year of planting and harvesting would be 116 and 333.5 mSv ha�1. Under theseconditions and considering a dose limit of 1mSv a�1, a farmer should not plant orharvest more than 8 or 5 ha a�1.Estimations of doses from transportation of contaminated wood showed that the

dose contribution from this pathway is negligible (0.2 mSv ha�1 a�1) even in worse-case conditions.

4.2. Doses to workers at a power plant firing 137Cs contaminated wood fuel

The calculations were based on the conceptual design of the operating M(abjergplant (I/S Vestkraft and Elsaproekt A/S, 1994) in northern Jutland, Denmark. Doserates were calculated using the Monte Carlo model MCNP (Briesmeister, 1993). Sixlocations of significant ash concentration were identified (described in Vandenhoveet al., 1999) and doses calculated for two distances (0.5 and 5m) from these locations(Table 7).For the calculations, the following example scenario was taken: the maximal soil-

to-wood TF obtained at the Belarus trials (2� 10�3m2 kg�1) and a deposition of1480 kBqm�2 (resulting 137Cs level in willow wood: 3000Bq kg�1). With the averagesoil to wood TF obtained for the Swedish and Belgian trials (105m2 kg�1) the 137Cscontamination level would only be 60Bq kg�1. For bottom and fly ash, densities of,respectively 800 and 400 kgm�3 were taken (Jensen & J�rgensen, 1998).The concentration of ash in the fly-ash silo, in the bottom-ash containers and in

the ‘big bags’ (fly-ash-collection bags) give the highest contributions to the extraexternal dose rate to workers in the power plant. If personnel are working very closeto these locations throughout an entire working year, dose contributions of about 15mSv may be expected, if Belarus soil-to-wood TFs and a deposition of 1480 kBqm�2

apply. In the W. European scenario, this value would be reduced by a factor of150due to the lower soil-to-wood TFs. Since workers are not expected to be positionedat a distance of just 50 cm from the ash concentrations throughout the wholeworking year, this estimate is probably highly conservative. At a distance of 5m, thedose rate would be reduced by a factor of at least 5, depending on the characteristicsof the ash-container. Inhalation doses received through routine operation of thepower plant were considered to be negligible.

Table 7

Calculated annual doses (mSv a�1) received by workers (2000h) at a power plant firing willow wood with

3000Bqkg�1 137Cs

D (m) Boiler Bag house Fly-ash silo Big bags Bottom-ash

conveyor

Spiral bottom-ash

conveyor

Bottom-ash

containers

0.5 1.5 0.9a 16.4 13.5 1.6 0.9 14.1

5 0.7 3.5 0.4 0.03 0.02 1.0

aDoses to people standing directly under the bag house filter.


4.3. External effective doses at ash deposits

Ash is generally not returned to the cultivation field as a fertiliser given its oftencaustic character. Instead the ash is deposited close to the power plant or atmunicipal refuse dumps. Disposal of radioactive contaminated ash should becontrolled in order not to result in secondary contamination. Workforce exposure atthe ash deposits was calculated using Microshield 4.10 software (MicroShield1,1993). Wood ash composition is taken from Steenari (1998). A wood contaminationlevel of 3000Bq kg�1 (Belarus TF: 2� 10�3m2 kg�1; deposition: 1480 kBqm�2) willgive approximately 150 000Bq kg�1 of 137Cs in the ash. This leads to a dose rate of10 mSv h�1 or an annual dose of 20mSv (2000 h per year) at 1m. At 5m, the dose rateis 2.25 mSv h�1 or 4.5mSv annually.

4.4. Assessment of the effective inhalation and external dose due to atmosphericdischarge of 137Cs-contaminated fly ash from a biofuel-powered heating plant

Atmospheric release of contaminated fly ash will lead to an elevated concentrationof radiocaesium in the air in the vicinity of the power plant. To assess the maximumeffective internal and external dose from atmospheric releases of fly ash, a Gaussianplume model (Whicker & Schultz, 1982) has been applied in combination with doseconversion factors found in the literature (ICRP, 1991; Kocher, 1983; Lindborg,1988; Finck, 1992). The fly-ash-collection efficiency was set to 90% in order not tounderestimate the doses. It was assumed that no depletion of the plume occurred dueto wet or dry deposition or by gravitational settling.A person standing continuously in the dose maximum area near a 100 MW

heating plant firing wood with a 137Cs concentration of 3000Bq kg�1 would onlyreceive an effective dose of a few mSv, due to inhalation of ash. The dose contributionfrom atmospheric release of contaminated fly ash can hence be neglected for heatingplant with a modern filtering system.

4.5. Conclusions

Dose calculations show that doses received during willow cultivation are lowerthan in intensive agriculture since crop maintenance is limited. External exposure ofsomeone constantly working close to the ash collectors in the conversion unit or atthe ash deposit may exceed the limit for the public if wood contamination levelsexceed 3000Bq kg�1. Under these conditions, doses received by the personnel shouldbe controlled and workers may even be regarded as radiation workers. Doses duringtransport or due to inhalation are negligible.

5. Modelling of SRC biomass yield

Willow, poplar and eucalyptus are the most popular tree species cultivated in shortrotation coppice (Jossaert, 1994). When adequately provided with water and


nutrients and depending on soil and climate conditions, willow yield ranges onaverage between 8 and 12 t ha�1 (extremes between 1 and 30 t ha�1).Apart from the yield figures from the Belarus test sites, no information is available

on SRC production in Belarus. A growth model was therefore developed to estimatethe potential yield of SRC (Vandenhove et al., 1998; Goor, Ledent, Timofeyev, &Vandenhove, 2001). The growth model was developed to see how potential andactual biomass yields compare and to assess the potential yield for different standmaturity and soil types other than those presently studied. The willow yield isestimated for the R2S2 growing year (2-year-old roots and shoots, no cut-back afterthe first year) and the maximum value is hence not yet attained (about 20% yieldincrease during the third year). Calculations were performed with Belarusmeteorological data. Minimum and maximum temperature (8C), global dailyincident irradiance (Jm�2) and precipitation (mm) were recorded on a daily basis forthe duration of the growing season.The SRC biomass yield was estimated to be 5.1 t ha�1 for sandy and 10.5 t ha�1 for

peaty soils. Yield estimates for the peaty soil are comparable with estimates forBelgian climate conditions (11.7 t ha�1).A peaty soil is characterised by a high soil water reserve; the yields obtained in this

case may therefore be considered as potential (non-water limited) values. Waterreserve is much lower for the sandy soil than for the peaty soil and plants will sufferfrom water shortage. The lack of water and the generally lower nutrient status on asandy soil are the main reasons for the 50–60% lower yield on a sandy soil comparedto a peaty soil in Belarus.Fig. 6 clearly shows that the willow biomass production is extremely low on the

sandy soil. Yield in 1997 (cutback yield, R1S1) was about 0.25 t ha�1 for the sandy

Fig. 6. Willow yield (R2S1) on the sandy and peaty soils at Savichy.


soil and about 1.4 t ha�1 for the Savichy peaty soil. In 1998, yield ranges between 3(Orm) and 6 (Bjorn and Rapp) t ha�1 on the Savichy peaty soil, the higher valuebeing comparable with the yield obtained in Western Europe for the first year aftercut-back. Only the peaty soil in Savichy gave sufficiently high yields. The sandy soil,where the willow died due to lack of water and nutrients, could better be reforestedwith low-water-demanding indigenous Scots pine.Potential yield estimates (model simulations) are for adequate nutrient conditions

and crop management. The difference between the actual and the potential yieldobtained for the Belarus peaty soil may be explained by the lack of partial nutrientand poor management practices (fertilising, lack of weeding and no fencing: someanimal intrusion). These factors could be easily improved in the case of extendedand/or intensive plantations.Potential yields estimated for a R2S2 crop are comparable for Western European

and Belarus conditions when Salix is grown on a peaty soil. On a sandy soil, the yieldpotential is about 30% lower in Belarus.

6. Economic sustainability of cultivation and conversion of SRC

The economic sustainability of cultivation and conversion of SRC was evaluatedfor W. European and Belarus conditions. For Belarus, the SRC is considered forland currently removed from agricultural production. Therefore, this land has noeconomic productivity at present. If this land would be reused it is only aftergovernmental decision/motivation.Economic analysis was performed using the RECAP-model (Renewable Energy

Crop Analysis Programme) for Western European (United Kingdom and Belgium)and Belarus conditions. Different types and scales of conversion systems forproduction of heat and electricity were studied. Table 8 gives a general list of inputparameters for the economic modelling. The conversion scenarios modelled arepresented in Table 9. Results of the economic analysis are discussed in detail inVandenhove et al. (1999) and Vandenhove, Goor, O’Brien, Grebenkov, & Timofeyer(2001). Only the major conclusions are mentioned here.Important differences in system parameters between W. Europe and Belarus are

that labour costs are a factor of at least 10 lower in Belarus, farm machinery pricesare lower in Belarus by up to a factor of five, domestic heat and electricity prices aremuch lower in Belarus and, finally, boilers for heat production are cheaper by afactor of three to five in Belarus.In Western Europe, electricity production from SRC is currently only viable with

both production incentives and electricity price support at the conversion side.The SRC production cost in Belarus is lower than in Western Europe, due to the

lower labour and machinery costs. To maintain this advantage, the yield of SRC inBelarus must approach that in W. Europe. A yield decrease by 50% renders theproduction unprofitable. This implies that SRC cultivation for energy production onsandy soils under the Belarus climatic conditions will hardly be profitable since thepotential yield is only about 5 t ha�1. Since about 60% of the soils in Belarus are of a


sandy nature, only a small percentage of soils is suitable for SRC cultivation. Thesesoils should better be aforested or used for crops adapted to low water and nutrientconditions.Another important parameter for SRC-production profitability is the price of the

bio-fuel delivered at the plant. To make production profitable in Belarus, about 40EUR t�1 is required (in W. Europe it is even higher). At this bio-fuel price, which is afactor of 4 higher than the amount paid for waste wood as biofuel, a profit can bemade both at the production and the conversion site. The harvesting technique alsoaffects the profitability at the production site. Harvesting in chips is preferred to

Table 8

General list of input parameters for the economic modelling

Cultivation General Land cost, subsidies, interest rates

Scale (number of hectares)

Rotation characteristics (length, spreading, cut back)

Harvest (yield, losses)

Biofuel price

Material Cost and number of planting materials

Cost and amount of fertilisers, pesticides

Machinery Tractor, trailer

Cost of fuel, maintenance, depreciation

Planting and harvesting options

Workforce Salaries, time requirement

Transport Machinery, capacity, time required

Storage Cost for barn, losses during storage

Conversion Power plant Scale (capacity and % availability),

Type (combustion, gasification; production

of heat or electricity)

Capital cost, subsidies, interest rate, depreciation

Prices Biofuel price

Electricity/heat price, price subsidy

Workforce Salaries, time requirement

Table 9

Conversion scenarios modelled

Conversion technology Belarus UK Belgium

Heat

Large scale, 28 MW(t) X

Medium scale, 3 MW(t) X

Small scale, 330 kW(t) X X

Electricity

Large scale, 30 MW(e) X X

Medium scale, 8 MW(e) X X

Small scale, 150 kW(e) X X


harvesting in sticks and separate chipping. Transport distance only affectsproduction profitability to a limited extent.For conversion, the most significant parameter was the price that could be

achieved for the electricity or heat. In Belarus, domestic tariffs (0.0032EURkWhe�1

for electricity and 0.001EURkWh�1 for heat) are much lower than industrial tariffs(0.034 EURkWhe�1 and 0.026EURkWh�1), and none of the schemes consideredwas economic for domestic electricity or heat production.For industrial tariffs, the heat schemes considered were all economic, on condition

that they are run at a higher availability (80%) than the quoted 16%, which is forseasonal heat production. Heat schemes already exist in Belarus, using forestresidues as fuel. Our work has shown that SRC could be used as the fuel, paying40EURodt�1 to the producer for the delivered chipped fuel. Even when the costs ofthe conversion system are increased by a factor 3–4 compared to the costs quoted forBelarus, these conversion routes may still be viable.For electricity production in Belarus, results are more speculative, since no

conversion plant currently exists. In our original calculations, we assumed a lowercapital cost of conversion plant in Belarus, which led to large- and small-scaleschemes being marginally profitable, where industrial electricity tariffs were achievedand availability was high. However, if part of the power plant components have tobe imported, capital cost may become more likely 50–100% of that in WesternEurope. Using these figures, electricity production is not economic in Belaruswithout price support for the electricity or capital grants for plant construction.Only a small percentage of the revenue from energy production is to be dedicated

to waste disposal in the case of the wood bio-fuel pathway. This will not render theprofitable systems uneconomic.An important disadvantage of growing a perennial crop is the higher risk for the

farmer in comparison with cultivation of an annual crop. After having invested inSRC, farmers can only change to other crops at high costs. Perennials can hardly beincluded in rotation-set-aside schemes. Compared to normal forestry, revenues arisefrom the third year onwards in SRC and only after 20–80 year for forests. If theforested area is sufficiently large, in the longer run, returns will come regularly. Onlyin the short-medium-term perspective are revenues from forestry small.Three other potential energy crops were also evaluated as an alternative land use

for contaminated arable land (Vandenhove et al., 1999): oil seed rape (OSR), winterwheat (WW) and sugar beet (SB). The conversion routes considered are for OSResterification to produce rape methyl esters and for WW and SB fermentation toproduce ethanol. These crops were selected since they are already grown in Belarusand their biofuel conversion routes are well known in Western Europe. Technicaland economic feasibility is discussed for Western European and Belarus conditions.The OSR, WW and SB can be grown successfully in Belarus if the potentialyields (appropriate soil conditions) are attained. In Western Europe, OSRproduction is only profitable with price support. Both in Belarus and WesternEurope, the cost of liquid bio-fuels is about 3–4 times the cost of fossil fuels andhence price subsidy is needed to compete with fossil fuels. It is hence very unlikelythat these crops can be advocated as potential alternative crops for a contamination


scenario since the production-conversion schemes are unprofitable even inuncontaminated conditions.

7. Conclusions

Energy production from SRC is radiologically and economically a potentiallysustainable landuse option. However, the feasibility of the biofuel chain depends ona number of factors: legislation, crop performance and economics, socialacceptability and market.When advocating or installing a new culture and related conversion route, an

important aspect to consider is also the infrastructure needed and the availability ofa market to sell the product (heat and electricity) to. The viability of the system willalso depend on the macroeconomics [price evolution of fossil fuels, electricity andheat (as already partly investigated), transport costs, etc] and on socio-politicalperceptions. In case of a severe contamination, public perception and boundaryeconomic conditions will certainly be affected. Assessment of the impact ofmacroeconomics and socio-political aspects, though extremely important, was,however, outside the scope of this project.

References

Absalom, J. P., Young, S. D., Crout, N. M. J., Nisbet, A. F., Woodman, R. F. M., Smolders, E., & Gillett,

A. G. (1999). Predicting soil to plant transfer of radiocaesium using soil characteristics. Environmental

Science and Technology, 33, 1218–1223.

Alexakhin, R. M. (1993,). Countermeasures in agricultural production as an effective means of mitigating

the radiological consequences of the Chernobyl accident. The Science of the Total Environment, 137,

9–20.

Alexakhin, R. M., Frissel, M. J., Schulte, E. H., Prister, B. S., Vetrov, V. A., & Wilkins, B. T. (1993).

Change in land use and crop selection. The Science of the Total Environment, 137, 169–172.

Biewinga, E. E., & van der Bijl, G. (1996). Sustainability of energy crops in Europe: A methodology

developed and applied. Centre for Agriculture and Environment, Utrecht, CLM-234.

Briesmeister, J. (Ed.). (1993). MCNP}A general Monte Carlo N-particle transport code, Users Manual,

Los Alamos, New Mexico, USA: Los Alamos National Laboratory, LA-12625.

Broadley, M. R., & Willey, N. J. (1997). Differences in root uptake of radiocaesium by 30 plant taxa.

Environmental Pollution, 97, 11–15.

Bunzl, K., & Kracke, W. (1987). Seasonal variation of soil-to-plant transfer of K and fallout 134,137Cs in

peatland vegetation. Health Physics, 57, 593–600.

Buysse, J., Van den Brande, K., & Merckx, R. (1996). Genotypic differences in the uptake and distribution

of radiocaesium in plants. Plant and Soil, 178, 265–271.

de Maeyer, E. A. A., & Huisman, W. (1995). Techniques for harvesting and storage of fibre hemp. In: Ph.

Chartier et al. (Eds.), Biomass for energy, environment, agriculture and energy, pp. (731–739). Oxford:

Pergamon Press.

Entry, J. A., Watrud, L. S., & Reeves, M. (1999). Accumulation of 137Cs and 90Sr from contaminated soil

by tree grass species inoculated with mycorrhizal fungi. Environmental Pollution, 104, 449–457.

FAO. (1990). Revised legend to the FAO-UNESCO soil map of the world. Soils Bulletin 60. FAO, Rome

(119pp.).


Finck, R. (1992). High resolution field gamma spectrometry and its application to problems in environmental

radiology. Ph.D. thesis, Departments of Radiation Physics, Malm .o and Lund, Lund University, ISBN

91-628-0739-0.

Gommers, A., Thiry, Y., Vandenhove, H., Vandecasteele, C. M., Smolders, E., & Merckx, R. (2000).

Radiocaesium uptake by willows planted as valorasition tool for contaminated waste farmland.

Journal of Environmental Quality, 29, 1384–1390.

Goor, F. (1998). RECOVER: Relevancy of short rotation coppice vegetation for the remediation of

contaminated areas: Literature review of short rotation coppice management and comparison with other

energy crops. Deliverable to Project FI4-CT95-0021c, SCK CEN, Mol, Belgium.

Goor, F., Ledent, J.-F., Timofeyev, S., & Vandenhove, H. (2001). Willow short-rotation coppice for

energy: an alternative to food crops on radio-contaminated farmlands? Agriculture, Ecosystems and

Environment, submitted.

GOPA. (1996). Belarus: Study on alternative biodiesel sources in relation with soil decontamination. Project

TACIS/Reg93, Contract WW 9204/0401/BOB, GOPA-Consultants, Bad Homburg, Germany.

Grebenkov, A. (1997). Internal deliverables for EC-RECOVER Project.

Hedvall, R. (1997). Activity concentrations of radionuclides in energy production from peat, wood chips and

straw. Ph.D. thesis, Lund University, Sweden, ISBN 91-7170-788-3.

IAEA. (1992). Radioactive waste management: An IAEA source book. Vienna: IAEA.

ICRP. (1991). Annual limits on intake of radionuclides by workers based on the 1990 recommendations. Ann

ICRP 21, No 4.

Jensen, J. M., & J�rgensen, B. H. (1998). Elsamprojekt A/S, Kraftværksvej 53, DK-7000 Fredericia,

Denmark. Personal communication.

Jossaert, J. M. (1994). Le taillis "a courtes rotations: Alternative agricole. Laboratoire d’Ecologie de

Grandes Cultures, Catholic University of Louvain-la-Neuve, Belgium.

Kocher, D. C. (1983). Dose-rate conversion factors for external exposure to photons and electrons. Health

Physics, 45(3), 665–686.

Ledin, S. (1996). Willow wood properties, production and economy. Biomass and Bioenergy, 11, 75–83.

Lindborg, L. (1988). Storheter vid str (alskyddsarbete}extern bestr (alning, SSI-rapport 88-17.

McLaughlin, Bouton, S. J., Beransby, D., Conger, R., Ocumpaugh, W., Parrish, D., Taliaferro, C., Vogel,

K., & Wullschelger, S. (1997). In Making a business from biomass in energy, environment, chemicals,

fibers and materials. R.P. Overend, & E. Chornet (Eds.), Proceedings of the third biomass conference of

the Americas. pp. (137–144).

MicroShield1 version 4.10, (1993). Users manual. Grove Engineering Inc.

Minotti, P. L., Craig, D., & Jackson, W. A. (1965). High cesium uptake in wheat seedlings cultured with

ammonium. Soil Science Society Proceedings, 29, 220–221.

Nisbet, A. F., Woodman, R. F. M., & Haylock, R. G. E. (1999). Recommended soil-to-plant transfer

factors for radiocaesium and radiostrontium for use in arable systems. NRPB-R304.

Panomarev, A. V. (1997). Caesium migration in forest ecosystems, unpublished.

Prister, B. S., Loshchilov, N., Perepelyatnikov, L., & Bondar, P. (1992,). Efficiency of measures aimed at

decreasing the contamination of agricultural products in areas contaminated by the Chernobyl NPP

accident. The Science of the Total Environment, 112, 79–87.

Smolders, E., Kiebooms, L., Buysse, J., & Merckx, R. 137 (1996). Cs uptake in spring wheat (Triticum

aestivum L. cv Tic) at varying K supply. I. The effect in solution culture. Plant & Soil, 181, 205–209.

Smolders, E., Van den Brande, K., & Merckx, R. (1997). Concentrations of 137Cs and K in soil solution

predict the plant availability of 137Cs in soils. Environmental Science and Technology, 31, 3432–3438.

Steenari, B.-M. (1998). Chemical properties of FBC ashes. Ph.D. thesis, Chalmers University of

Technology (Sweden), ISBN 91-7197-618-3.

Szekely, J. G., Amiro, B. D., Rasmussen, L. R., & Ford, B. (1994). Environmental assessment of

radiological consequences for forestry in contaminated areas of the republic of Belarus. World Bank

Report (57 pp.).

Vandenhove, H., Thiry, Y., Gommers, A., Goor, F., Jossart, J. M., G.avfert, T., Holm, E., & Roed, J.

(1998). RECOVER: Relevancy of short rotation coppice vegetation for the remediation of contaminated

areas. Project FI4-CT95-0021c. Mid-Term Report, SCK .CEN, Mol, Belgium, BLG 762.


Vandenhove, H., Thiry, Y., Gommers, A., Goor, F., Jossart, J. M., G.avfert, T., Holm, E., & Roed, J.

(1999). RECOVER: Relevancy of short rotation coppice vegetation for the remediation of contaminated

areas. Project FI4-CT95-0021c. Final Report, SCK CEN, Mol, Belgium, BLG 826.

Vandenhove, H., Goor, F., O’Brien, S., Grebenkov, A. & Timofeyev, S. (2001). Economics of short

rotation coppice for re-valuation of contaminated land in Belarus. Biomass and Bioenergy, submitted.

Vandenhove, H., & Van Hees, M. (2001). Flax and fibre hemp as alternative land use in areas

contaminated by a nuclear accident. The International Journal of Phytoremediation, submitted.

Whicker, F. W., Schultz, V. (1982). Radioecology: Nuclear energy and environment, vol. 2. Bora Raton,

FL: CRC Press.

Zabudko, A. N., Petrov, E. E., Kozmin, G. V., & Svetove, V. A. (1995). Comprehensive radiological

investigations of forests in Kaluga, Tula and Or’ol regions. IAEA Technical Meeting on ‘‘Clean-up

Criteria for Forests and Forestry Products Following a Nuclear Accident’’ 30 October–3 November

1995, unpublished.


short rotation coppice for revaluation of contaminated land

Documents