Elevated radionuclide concentrations in heavy mineral-rich beach sands in the Cox's Bazar region, Bangladesh and related possible radiological effects

Isotopes in Environmental and Health StudiesVol. 48, No. 4, December 2012, 512–525

Elevated radionuclide concentrations in heavy mineral-richbeach sands in the Cox’s Bazar region, Bangladesh and

related possible radiological effects

Mashrur Zamana*, Michael Schubertb and Sytle Antaoa

aDepartment of Geoscience, University of Calgary, Calgary, Alberta, Canada; bDepartment of AnalyticalChemistry, Helmholtz Centre for Environmental Research – UFZ, Leipzig, Germany

(Received 30 December 2011; final version received 1 May 2012)

The study focuses on elevated levels of environmental radioactivity present in heavy mineral depositslocated along a 120-km coastal section of Cox’s Bazar on the eastern panhandle of Bangladesh. Thedeposits are situated in or at sand dunes located on the recent beach (foredune area) or in attached paleo-beach areas (backdune area). This study investigates activity concentrations in bulk beach sands (sixrepresentative samples) and in five mineral fractions separated from the beach sands in order to assesspotential radio-ecological effects and the possible use of the mineral deposits as a source for uranium andthorium. The bulk beach sands and individual mineral fractions were analysed by gamma-ray spectroscopy.The activity concentrations of U-238, U-235, Th-232 and K-40 in the bulk beach sand samples were foundto be considerably high and positively correlated to the concentration of heavy minerals in the sand. Inthe mineral fractions, the highest activity concentrations were found in the zircon fraction followed bygarnet, rutile, ilmenite and magnetite. The determination of (i) the radium activity, (ii) several radiationhazard indices and (iii) adsorbed and effective gamma doses allowed to assess the related exposure ofthe environment and the local population to elevated radioactivity. It becomes evident from the presentdata that (1) if raw sands or mineral fractions mined in the study area are used for building purposesor industrial use, their activity concentrations have to be considered from a radio-ecological perspectiveand (2) if mining and processing of the minerals is being considered, uranium and thorium may becomestrategically significant by-products.

Keywords: Bangladesh; beach sand; heavy mineral deposit; isotope geology; natural radioactivity;potassium; radiological exposure; thorium; uranium

1. Introduction

Major accumulations of metallic heavy minerals (specific gravity above 2.85 g cm−3) have beendiscovered at many locations in the coastal areas of Bangladesh. The minerals were deposited asplacer deposits in association with the formation of sand dunes.

The deposits were first detected by the Pakistan Geological Survey in the early 1960s in theCox’s Bazar district, eastern Bangladesh [1], and the authors reported the occurrence of elevated

*Corresponding author. Email: [email protected]

ISSN 1025-6016 print/ISSN 1477-2639 online© 2012 Taylor & Francishttp://dx.doi.org/10.1080/10256016.2012.696542http://www.tandfonline.com

Dow

nloa

ded

by [

Hel

mho

ltz Z

entr

um F

Um

wel

tfor

schu

ng G

mbH

] at

05:

14 1

5 A

pril

2014

Isotopes in Environmental and Health Studies 513

levels of radioactivity. In the 1980s and 1990s, the Bangladesh Atomic Energy Commission(BAEC) carried out large-scale exploration activities along the whole coastline of Bangladeshand reported the presence of further economically relevant heavy mineral deposits in dunes andbeach sands [2–5]. Zaman et al. [6] discovered another major heavy mineral deposit in the Cox’sBazar region as part of a BAEC study. The metallic heavy minerals include magnetite (MT),ilmenite (IN), garnet (GN), zircon (ZC), rutile (RT), kyanite, leucoxene and monazite.

The activity concentrations reported by Schmidt and Asad [1] are specifically associated withthe mineral fractions of ZC and monazite with the highest activities being present in the MT-richdeposits. Alam et al. [7] analysed individual mineral fractions for their activity concentrationsof U-238, Th-232 and K-40 and found respective activities of 6439, 1324 and 472 Bq kg−1 forZC, 3951, 7903 and 213 Bq kg−1 for GN, 348, 388 and 60 Bq kg−1 for IN, 6643, 11670 and182 Bq kg−1 for RT, and 22, 43 and 293 Bq kg−1 for MT. Zaman et al. [8] measured activityconcentrations in sand samples taken in the Cox’s Bazar area at a 20 × 10 m grid pattern andobserved a strong correlation between activity concentration and heavy mineral occurrence.

Although elevated activity concentrations in beach sands and dunes have been reported andevaluated for a considerable number of sites along the Bangladesh coastline, no study was carriedout so far in the recent beach and paleo-beach areas that stretch between Moheskhali Islandand the Southern tip of the Bangladesh panhandle at Bodarmukam area (Figure 1(a)). However,related information is of relevance because, in particular, the paleo-beach areas south of Cox’sBazar are partly populated and used for agricultural purposes. Thus, the aim of this study is toinvestigate the activity concentration of U-238, U-235, Th-232 and K-40 in beach sands takenfrom this area. Bulk sand samples and separated mineral fractions of ZC, MT, IN, RT and GN wereanalysed. This study focuses on radiological effects on the coastal environment and discusses thepotential strategic significance of the heavy mineral deposits as natural resources for uranium andthorium [9].

2. Geological characteristics of the study area

The study area is located at the coastline of the south-eastern panhandle of Bangladesh with theBay of Bengal to the west. To the east, the area is bordered by hill ranges that rise to elevationsof about 100 m above sea level. To the north, it is bordered by the Kutubdia channel, and to thesouth by the Naf estuary (Figure 1(a)).

The coastal area between Cox’s Bazar and the Naf estuary is characterised by a continuoussandy beach, stretching for about 120 km. The width of the beach varies with time and place andranges from 100 m during the rainy season to 300 m during the dry season. The present beachslopes gradually towards the sea (from 4 to 6◦) and rises at its eastern boundary to about 4–5 mabove the mean sea level [10]. Elongated sand dunes parallel to the coastline rise east to it abovethe high tide mark. Due to wind activities, the dunes are dynamic in positions. They are composedof fine to medium-grained sands. The term ‘foredune area’ is used in this study to represent thearea of the present beach including the recent sand dunes (Figure 1(b) and (c)).

The area located between the foredune area and the eastern hills is characterised by sparselydistributed elongated paleo-sand dunes aligned parallel to the present coastline (Figure 1(b)). Forthat area, the term ‘backdune area’ is used in this study. The backdunes are not dynamic becausethey are not attached to the sea, somewhat sheltered from the wind and composed of more compactsand than the foredunes. Lithologically, the backdune sands are represented by fine to medium-grained sands with varying shares of silt and clay. The backdune area represented the actual seabeach about 5000 years ago [11]. The flat backdune lands are recently being used for agriculture.Human settlements, markets, schools, graveyards, cyclone centres, etc. are also abundant there.

Dow

nloa

ded

by [

Hel

mho

ltz Z

entr

um F

Um

wel

tfor

schu

ng G

mbH

] at

05:

14 1

5 A

pril

2014

514 Mashrur Zaman et al.

Figure 1. (a) Map with sampling locations, (b) the general surface distribution of heavy mineral deposits at foreduneand backdune areas at teknaf, Cox’s Bazar [2], and (c) a schematic cross-section of the coastal profile marked A-B on (b).

Dow

nloa

ded

by [

Hel

mho

ltz Z

entr

um F

Um

wel

tfor

schu

ng G

mbH

] at

05:

14 1

5 A

pril

2014

Isotopes in Environmental and Health Studies 515

In the study area, the dark brown to black coloured sands, which contain major shares of heavyminerals, are mainly deposited on the western lap and on top of most foredunes and backdunes [9].Figure 1(c) depicts schematically a typical cross-section illustrating the general locations of heavymineral deposits in the foredune and backdune areas. The thickness of the deposits varies from afew centimetres to some metres. The mineral deposits can be described as lens-type features withwidths (E–W) from a few metres to over 100 m and lengths (N–S) from 30 to over 500 m.

3. Material and methods

3.1. Sample collection

Sand samples were collected from three foredune and three backdune locations, where a com-paratively high concentration of heavy minerals was visual. About 1 kg of sand was collected ateach of the six locations (Figure 1(a)). At the sampling point KTF-3 (foredune), an additional 5 kgsample was collected for the sake of separating the individual heavy mineral fractions of ZC, GN,IN, RT and MT. All samples were dried at 100◦C. From each of the six samples, a small subsam-ple (about 100 g) was taken, and the bulk of heavy minerals was separated by gravity separation,using a heavy liquid (bromoform, ρ = 2.89). The light minerals were mainly quartz with a minoramount of feldspar. Table 1 summarises sample IDs, locations including geographical position,as well as the light and heavy mineral contents (HMs) in each sample.

3.2. Physical separation of heavy minerals

The 5 kg beach sand sample collected from the foredune area at sampling point KTF-3 was usedfor the separation of the five mineral fractions ZC, GN, IN, RT and MT applying a semi-industrialseparation technique based on the physical properties of each mineral fraction. A generalised flowchart is given in Figure 2. Using a wet gravity separator (shaking table), the sample material wasfirst fractionated into a light and a heavy fraction. The light fraction was discarded. The heavyfraction was first treated using an induced roll magnetic separator (IRMS) to separate the stronglymagnetic fraction from the moderately non-magnetic fraction. The strongly magnetic fractionwas predominantly MT with some IN, which was removed in a separate step by concentratingthe MT using a hand magnet. For separating the remaining moderately magnetic fraction fromthe non-magnetic fraction, the material was processed again with the IRMS after increasing themagnetic field. The moderately magnetic fraction was mainly composed of IN and GN, the non-magnetic fraction was mainly composed of RT and ZC. Thereafter, both fractions were heatedto about 200◦C for electric separation. Both fractions were processed using an electrostatic plate

Table 1. Sampling positions and mineral contents in the bulk beach sand sample.

Geographical positionHeavy mineral Light mineral

Sample ID Location Latitude Longitude content (wt.%) content (wt.%)

Foredune areaBDF-1 Bodarmukam 20◦44′43′′ N 92◦20′17′′ E 97 3MKF-2 Monkhali 21◦06′57′′ N 92◦06′14′′ E 88 12KTF-3 Kalatoli 21◦24′58′′ N 91◦58′57′′ E 91 9

Backdune areaLBB-4 Lomburi 20◦52′20′′ N 92◦15′59′′ E 60 40SPB-5 Shaplapur 21◦05′04′′ N 92◦08′02′′ E 70 30FHB-6 Fakirahata 21◦32′29′′ N 91◦56′14′′ E 54 46

Dow

nloa

ded

by [

Hel

mho

ltz Z

entr

um F

Um

wel

tfor

schu

ng G

mbH

] at

05:

14 1

5 A

pril

2014

516 Mashrur Zaman et al.

Figure 2. Flow chart showing the procedure of separation of five mineral fractions from a bulk beach sand sample.

separator (ESPS) to separate IN (conductor) and GN (non-conductor), and RT (conductor) andZC (non-conductor), respectively.

The six bulk beach sand samples and the separated five mineral fractions were finally exam-ined for their mineral content using a stereographic binocular microscope following the methoddescribed by Macdonald [12]. The mineral composition of each of the five mineral fractions ispresented in Table 2. In order to confirm that the chosen mineral separation approach allows theproper separation of IN, RT and ZC avoiding ‘contamination’ of the material with significantamounts of monazite, X-ray powder diffraction analysis (XRD) of the separated IN, RT and ZCsamples has been carried out in addition to the analyses discussed above. The resulting XRD spec-tra indicate no noticeable (in ZC and IN) or only minor fractions of monazite (≤2 % in RT) in the

Table 2. Mineral compositions of the mineral fractions won from thebulk beach sand sample collected at Kalatoli (KTF-3).

Fraction Mineral composition

ZC ZC 99 %, others 1 %GN GN 75 %, IN 20 %, monazite 4 %, others 1 %IN IN 95 %, leucoxene 4 %, others 1 %RT RT 60 %, IN 30 %, GN 5 %, monazite 2 %, others 5 %MT MT 99 %, others 1 %

Dow

nloa

ded

by [

Hel

mho

ltz Z

entr

um F

Um

wel

tfor

schu

ng G

mbH

] at

05:

14 1

5 A

pril

2014

Isotopes in Environmental and Health Studies 517

separated mineral materials (Table 2), thereby verifying the suitability of the applied separationtechnique. The weights of the bulk beach sand samples and the separated heavy mineral samplesranged from 205 to 289 g and from 258 to 326 g, respectively.

3.3. Gamma-ray spectroscopy

For gamma spectroscopy, each of the six beach sand samples and each of the five separatedheavy mineral fraction was put into a 108 cm3 cylindrical capsule, which was sealed tightly inorder to prevent the escape of radon (Rn-222, Rn-219 and Rn-220). Since the longest half-life ofthe three radon species is 3.82 days (Rn-222), the capsules were stored for 25 days in order toachieve the secular equilibrium between the radionuclides of interest and their short-lived progeny[13]. Subsequently, all samples were measured twice for about 24 h. Activity concentrations ofU-238, U-235 and Th-232 were detected based on the emission lines of their respective short-livedprogeny. For U-238, the gamma lines of Th-234, Pa-234m, Pb-214 and Bi-214, for U-235 thelines of U-235, Th-227, Ra-223 and Rn-219, and for Th-232 the lines of Ac-228, Pb-212, Bi-212and Tl-208 were used, respectively. Due to the lack of strong and distinct gamma emissionlines of the actual radionuclides U-238, U-235 and Th-232, the daughter nuclides listed aboveare generally used to derive the respective activity concentrations, assuming secular equilibriumthroughout the decay series [14–17]. For data evaluation, it was kept in mind that the assumption ofdecay equilibrium throughout the decay series may not be justified, since geochemical processeswith different mobilisation effects on the individual radionuclides may have occurred in formergeological times. As mentioned above, the gamma line of Pa-234m (at 1001.3 keV) was used forU-238 determination in addition to the conventionally used gamma lines of Th-234, Pb-214 andBi-214. That was done in order to minimise statistical uncertainties. Although the 1001.3 keVemission probability is low (0.84 %), it is suitable since it does not interfere with any otherpeaks, for example, in the presence of a high amount of thorium [18,19]. For the same reason, thetwo weak U-235 gamma lines at 163.4 and 205.5 keV were used in spite of their low emissionprobability of 0.22 % additionally to the ones of the short-lived progeny. K-40 was detected basedon its distinct gamma line at 1461 keV.

All of the gamma spectroscopy measurements were carried out at the Helmholtz Centre forEnvironmental Research (UFZ) in Leipzig, Germany, using two coaxial low-energy high-puritygermanium detectors, n-type (ORTEC) with an active volume of 39 cm3 and a 0.5 mm Be win-dow. Detectors and measuring geometry (108 cm3 cylindrical capsules) were calibrated using thecertified reference materials RGU, RGTh and RGK provided by the International Atomic EnergyAgency (IAEA).

For data evaluation, two different analytical approaches were applied: an automated approachand a ‘manual’ approach. For the automated approach, the efficiency calibration of the detectorsand the software package ‘GammaW®’ were used. The manual approach was based on the ruleof proportion as given in the following relationships:

ACS =[(

PASCTS

)− BGD

]× FC

WS, (1)

FC = 1

PAC/CTC/ACC, (2)

where ACS is the activity concentration of sample (Bq kg−1), WS the sample weight (g), CTSthe counting time of sample (s), PAS the peak area of sample (counts), FC the calibration fac-tor (Bq cps−1), BGD the detector background (cps), PAC the peak area of calibration standard

Dow

nloa

ded

by [

Hel

mho

ltz Z

entr

um F

Um

wel

tfor

schu

ng G

mbH

] at

05:

14 1

5 A

pril

2014

518 Mashrur Zaman et al.

(counts), CTC the counting time calibration standard (s) and ACC the activity concentration ofcalibration standard (Bq).

4. Estimation of related radiation hazards

The beach sands in the study area are frequently used by the local people for construction pur-poses. Besides, the separated mineral fractions that are abundant in the area are used for differentindustrial purposes. For these reasons, the safe use of the sands and the mineral fractions wasevaluated from the radio-ecological perspective. Five parameters were used for estimating theradiation hazard related to the materials as follows:

The radium equivalent activity (Raeq) was defined by Beretka and Mathew [20]. It representsthe summed-up specific activities of Ra-226, Th-232 and K-40 by using just a single index. Raeq

(Bq kg−1) is defined as given in Equation (3), where ARa-226, ATh-232 and AK-40 are the activityconcentrations of Ra-226, Th-232 and K-40 (Bq kg−1), respectively:

Raeq = ARa-226 + 1.43ATh-232 + 0.077AK-40. (3)

The gamma dose rate (D) (nGy h−1) in air 1 m above the ground (in which Ra-226, Th-232 andK-40 are assumed to be uniformly distributed) is defined by the following relationship [21]:

D = 0.462ARa-226 + 0.621ATh-232 + 0.0417AK-40. (4)

The annual effective dose (DE) (mSv a−1) is determined by the indoor or outdoor occupancy factorof 0.8 or 0.2, respectively, and a conversion factor of CF = 0.7 Sv Gy−1 [21] using the followingrelationship:

DE(mSv a−1) = D(nGy h−1) × 8760(h a−1) × 0.7(Sv Gy−1) × occupancy factor. (5)

The Gamma hazard index (Iyr), which was defined by NEA-OECD [22], allows an estimationof the level of gamma radiation hazard associated with Ra-226, Th-232 and K-40 in specificmaterials. Iyr (Bq kg−1) is defined as

Iyr = 0.0067ARa-226 + 0.01ATh-232 + 0.00067AK-40. (6)

The external (Hex) and internal (Hin) hazard indices have been defined by Beretka and Mathew[20]. The equations for Hex and Hin are given in Equations (7) and (8). The recommended valueof both Hex and Hin is below or equal to 1.

Hex = 0.0027ARa-226 + 0.00386ATh-232 + 0.00021AK-40, (7)

Hin = 0.0054ARa-226 + 0.00386ATh-232 + 0.00021AK-40. (8)

5. Results and discussion

5.1. Activity concentrations

The activity concentrations of U-238, U-235, Th-232 and K-40 in all six bulk beach sand sam-ples from the foredune and backdune areas and in the five separated mineral fractions from theforedune sampling point KTF-3 are summarised in Table 3. The listed results represent the mean

Dow

nloa

ded

by [

Hel

mho

ltz Z

entr

um F

Um

wel

tfor

schu

ng G

mbH

] at

05:

14 1

5 A

pril

2014

Isotopes in Environmental and Health Studies 519

Table 3. Activity concentrations of U-238, U-235, Th-232 and K-40 in bulk beach sand samplesfrom the foredune and backdune areas and in separated mineral fractions.

Sample ID U-238 (Bq kg−1) U-235 (Bq kg−1) Th-232 (Bq kg−1) K-40 (Bq kg−1)

Foredune areaBDF-1 4245 ± 238 145 ± 18 5537 ± 202 372 ± 10MKF-2 2217 ± 186 87 ± 14 3217 ± 117 242 ± 10KTF-3 1651 ± 133 65 ± 13 2625 ± 96 162 ± 8

Backdune areaLBB-4 1082 ± 105 43 ± 12 1822 ± 67 126 ± 6SPB-5 1025 ± 88 39 ± 10 1440 ± 58 89 ± 6FHB-6 754 ± 85 30 ± 8 996 ± 42 110 ± 34

Mineral fractionsZC-1 14849 ± 751 124 ± 16 10405 ± 364 806 ± 35GN-2 3236 ± 233 12 ± 3 7258 ± 261 479 ± 7IN-4 419 ± 55 63 ± 11 529 ± 24 12 ± 2RT-3 1676 ± 146 3 ± 1 2011 ± 125 126 ± 10MT-5 61 ± 13 – 85 ± 5 10 ± 2

values of the manual and the automated approaches for gamma spectrometry data assessment asdescribed above.

The activity concentrations of U-238, U-235, Th-232 and K-40 in the bulk beach sand samplescollected in the foredune area (Bodarmukam, BDF-1; Monkhali, MKF-2 and Kalatoli, KTF-3),backdune area (Lomburi, LBB-4; Shaplapur, SPB-5 and Fakirahata, FHB-6) are illustrated inrelation to the HMs of the sands (Figure 3). In the fordune area, the highest activities weredetected at Bodarmukam (HM = 97 %), followed by the samples taken at Monkhali (HM =88 %) and Kalatoli (HM = 91 %). Although the heavy mineral concentration in the bulk sands

Figure 3. Activity concentrations of U-238, U-235, Th-232 and K-40 with respect to the heavy mineral concentrationin the bulk sands collected at the foredune (BDF-1, MKF-2 and KTF-3) and backdune (LBB-4, SPB-5 and FHB-6) areasof Cox’s Bazar district, Bangladesh. The parentheses in the legend show the regression values between the activity ofU-238, U-235, Th-232 and K-40, and heavy mineral concentrations in the samples.

Dow

nloa

ded

by [

Hel

mho

ltz Z

entr

um F

Um

wel

tfor

schu

ng G

mbH

] at

05:

14 1

5 A

pril

2014

520 Mashrur Zaman et al.

Figure 4. Correlation between U and Th concentrations in beach bulk sand samples collected from both foredune andbackdune areas.

of Monkhali is lower than that of Kalatoli, the activity concentrations of all radionuclides in theMonkhali sample are higher.

In the backdune area, the highest activities were detected in the sample taken at Lomburi(HM = 60 %) followed by the samples taken at Shaplapur (HM = 70 %) and Fakirahata (HM =54 %). Compared to the foredune samples, the activities of all radionuclides are relatively low.The Fakirahata sample, which shows the lowest HM of the three, revealed the lowest activityconcentration.

Although a general positive correlation between bulk HM and bulk radioactivity becomes obvi-ous, the correlation cannot be quantified reasonably. The observed discrepancies are probably dueto the varying concentrations of the respective mineral fractions that contain elevated radionuclideconcentrations in the bulk sands. However, a very good correlation (R2 = 0.99) was found betweenU-238 and Th-232 concentrations in the six bulk sand samples (Figure 4). This correlation indi-cates that there is a consistency of the presence of the habitually uranium- and thorium-bearingminerals ZC (ZrSiO4) and, occurring in minor amounts, monazite ([Ce,La,Nd,Th]PO4) in thebeach sands, as reported in other studies [23–25].

The highest activity concentrations in the five examined mineral fractions were found in the ZCfraction (Figure 5). GN and RT show significantly less activity, with GN showing a considerablyhigher activity than RT. The activity concentrations in IN and MT are insignificant.

The ZC fraction analysed in this study was composed of 99 % ZC crystals as confirmed by thebinocular microscopic observation (Table 2). The activity concentrations of U-238 and Th-232 inthe ZC found in this study are higher than the concentrations found in the study conducted byAlamet al. [7], who analysed ZCs from comparable materials. The activity values are also higher thanthe concentrations reported in other studies on ZC sands collected worldwide [23,26,27] (Table 4).

Figure 5 illustrates the element concentrations of U and Th in the five mineral fractions applyingthe conversion factors 1 mg kg−1 U = 12.25 Bq kg−1 U-238 and 1 mg kg−1 Th = 4.07 Bq kg−1 Th-232 [28]. The figure shows U and Th concentrations in the ZC of 1212 mg kg−1 and 2557 mg kg−1,respectively. Commercially used ZCs generally show U and Th concentrations of about 250–350 mg kg−1 and 100–200 mg kg−1, respectively, that is, concentrations that are significantlybelow those found in the analysed material [29]. Even though high radionuclide concentrations

Dow

nloa

ded

by [

Hel

mho

ltz Z

entr

um F

Um

wel

tfor

schu

ng G

mbH

] at

05:

14 1

5 A

pril

2014

Isotopes in Environmental and Health Studies 521

Concentrates

Zircon Garnet Ilmenite Rutile Magnetite

Act

ivity

con

cent

ratio

n (B

q/kg

)

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

U-238U-235Th-232K-40

ConcentratesZircon Garnet Ilmenite Rutile Magnetite

Con

cent

ratio

n (m

g/kg

)

0

500

1000

1500

2000

2500

3000

UTh

Figure 5. Activity concentrations and element concentrations in the five mineral fractions separated from the bulk beachsand collected at Kalatoli (KTF-3).

in ZCs may cause health-related problems in the manufacture of zirconia, the stability of the ZCstructure generally impedes the release of radionuclides from the mineral matrix. The presenceof U and Th in the analysed GN and RT fractions results from the presence of minor amounts ofmonazite in the separated fractions (Table 2).

5.2. Potential radiation hazard related to the bulk beach sands

The potential radiation hazard as revealed by the bulk sand materials and separated mineralfractions was estimated and quantified by applying the parameters Raeq (Bq kg−1), D (nGy h−1),

Dow

nloa

ded

by [

Hel

mho

ltz Z

entr

um F

Um

wel

tfor

schu

ng G

mbH

] at

05:

14 1

5 A

pril

2014

522 Mashrur Zaman et al.

Table 4. Activity concentrations of U-238 and Th-232 in ZC sand obtained in this study and some selective locationsin the world.

Description/locations U-238 (Bq kg−1) Th-232 (Bq kg−1) Sources

Processed ZC, Kalatoli, Cox’s Bazar, Bangladesh 14849 ± 751 10405 ± 364 This studyProcessed ZC sand, Cox’s Bazar, Bangladesh 6438 ± 326 1324 ± 96 [10]ZC sand, Chhatrapur, Orissa, India 3450 ± 150 1850 ± 180 [24]ZC sand, Erasama, Orissa, India 3500 ± 100 1750 ± 100 [23]Australian ZC sand 2400 ± 200 520 ± 40 [27]Australian ZC sand 2200 ± 200 480 ± 40 [27]South African ZC sand 3200 ± 300 520 ± 40 [27]South African ZC sand 2900 ± 200 450 ± 40 [27]Ukrainian ZC sand 1830 ± 150 370 ± 30 [27]Ukrainian ZC sand 1860 ± 160 380 ± 30 [27]

Table 5. Equivalent radium concentration (Raeq), gamma radiation dose (D) and radiation hazard indices Iyr , Hex andHin in bulk beach sands of both foredune and backdune areas and in separated heavy mineral fractions.

Sample ID Raeq(Bq kg−1) D (nGy h−1) Iyr(Bq kg−1) Hex (Bq kg−1) Hin (Bq kg−1)

Foredune areaBDF-1 12192 ± 528 5415 ± 236 84 ± 4 32.91 ± 1.42 44.37 ± 2.07MKF-2 6836 ± 354 3032 ± 159 47 ± 2 18.45 ± 0.96 24.44 ± 1.46KTF-3 5417 ± 271 2400 ± 121 37 ± 2 14.62 ± 0.73 19.08 ± 1.09

Backdune areaLBB-4 3697 ± 201 1637 ± 90 26 ± 1 9.98 ± 0.54 12.90 ± 0.83SPB-5 3091 ± 171 1372 ± 77 21 ± 1 8.34 ± 0.46 11.11 ± 0.70FHB-6 2187 ± 148 971 ± 67 15 ± 1 5.90 ± 0.40 7.94 ± 0.63

Heavy mineral fractionsZC 29790 ± 959 13355 ± 574 204 ± 9 80.42 ± 3.44 120.52 ± 5.47GN 13652 ± 475 6022 ± 270 95 ± 4 36.85 ± 1.64 45.59 ± 2.27IN 1176 ± 56 523 ± 40 8 ± 1 3.18 ± 0.24 4.31 ± 0.39RT 4561 ± 216 2028 ± 145 31 ± 2 12.31 ± 0.88 16.84 ± 1.27MT 183 ± 13 81 ± 9 1 ± 0 0.49 ± 0.05 0.66 ± 0.09

Iyr (Bq kg−1), Hex (Bq kg−1) and Hin (Bq kg−1) as defined by Equations (3) and (6)–(8). Theresults are summarised in Table 5.

The Raeq activity in the bulk beach sands of the sampled foredune and backdune locations werefound to be 33, 18 and 15 times, and 10, 8 and 6 times, respectively, higher than the related worldstandard value, which amounts to 370 Bq kg−1 [30]. The gamma hazard index, Iyr, which shouldbe less than or equal to 1 Bq kg−1 [7] does also show considerably higher values in both forduneand backdune areas. The external and internal hazard indices (Hex and Hin) in the foredune andbackdune beach sands should be less than 1 [31,32]. By evaluating the values found for Raeq, Iyr,Hex and Hin, it becomes obvious that, due to their elevated radioactivity levels, the bulk beachsands from both foredune and backdune areas should not be used for any kind of constructionpurposes without previous assessment and, if necessary, processing. The elevated radioactivitylevels correlate with the high amounts of heavy minerals present in the sand.

The gamma doses (D) related to the bulk beach sands (using Equation (4)) are also given inTable 5. The world average level of D is 55 nGy h−1 [33]. In the sampled foredune and backdunelocations, D was found to be 98, 55 and 44 times and 30, 25 and 18 times as high, respectively.Assuming an outdoor occupancy factor of 0.2, the outdoor annual effective doses (DE) at theforedune and backdune areas can be estimated using Equation (5). According to the IAEA [30],the outdoor annual effective dose for the public should be below 1 mSv a−1 above the backgroundwhen the dose is a result of a ‘practice’. Even though the doses found in our study are naturally

Dow

nloa

ded

by [

Hel

mho

ltz Z

entr

um F

Um

wel

tfor

schu

ng G

mbH

] at

05:

14 1

5 A

pril

2014

Isotopes in Environmental and Health Studies 523

Sample ID

BDF-1 MKF-2 KTF-3 LBB-4 SPB-5 FHB-6 ZC-1 GN-2 IN-3 RT-4 MN-5

DE (

mSv

/a)

0

2

4

6

8

10

12

14

16

(c) Mineral concentrates(b) Backdune area(a) Foredune area

Figure 6. Annual effective doses in the heavy mineral-rich (HM more than 54 wt.%) bulk beach sands collected at sixdifferent locations of the foredune and backdune areas and in separated mineral fractions.

occurring (and thus ‘background’) and not the result of any intervention by humans, it shall bepointed out that in the sampled foredune and backdune areas, the outdoor annual effective doseswere found to be significantly elevated (Figure 6). A comparison of the doses detected in theforedune and backdune areas reveals that the heavy mineral-rich foredune locations are morehazardous for general public use than the backdune area. The foredune area is free of humansettlements and mainly used by fishermen. However, the backdune area, where dwellings arepresent (in some places, on top of the heavy mineral deposits), has to be considered generallyunsuitable for human settlement from the radio-ecological point of view. The outdoor occupancyfactor for estimating DE in the backdune area was assumed to be 0.2. However, since people areactually living in some of the places where the heavy mineral concentration is comparatively highand spend more than 20 % of their time there, these estimates are rather underestimating the actualhuman exposure.

5.3. Potential radiation hazard related to the separated mineral fractions

The Raeq, Iyr, Hex and Hin values for the ZC, GN, IN, RT and MT fractions are summarised inTable 5. The Raeq values for ZC, GN, IN and RT are 81, 37, 3 and 12 times higher than theworld standard value (370 Bq kg−1), respectively; the Raeq value found for MT is below the worldstandard value. The Iyr values for ZC, GN, IN and RT are found to be significantly higher than1 Bq kg−1, the value for MT equals 1. Similarly, the Hex and Hin values for ZC, GN, IN and RT arehigher than 1, again MT is an exception. From these results, it becomes evident that the ZC, GN,IN and RT fractions have to be considered as radio-ecologically hazardous and unsafe for any kindof industrial use. In this regard, it has to be kept in mind that the technical staff who are involved inthe mineral processing in the Beach Sand Minerals Exploitation Centre pilot plant are exposed tothe radiation being emitted from bulk sand and separated mineral fractions. The absorbed gammadoses 1 m above the processed ZC, GN, IN, RT and MT fractions that are stored in open drums inthe mineral processing pilot plant are 243, 109, 10, 37 and 1 times higher than the recommended

Dow

nloa

ded

by [

Hel

mho

ltz Z

entr

um F

Um

wel

tfor

schu

ng G

mbH

] at

05:

14 1

5 A

pril

2014

524 Mashrur Zaman et al.

value, respectively. Considering an indoor occupancy factor at the processing pilot plant of 0.8and the reported yearly working hours (1736 h), the DE related to the stored ZC, GN, IN, RT andMT fractions are 12.98 ± 0.56, 5.85 ± 0.26, 0.51 ± 0.04, 1.97 ± 0.14 and 0.08 ± 0.01 mSv a−1,respectively. The average annual effective dose per year for workers over five consecutive yearsis recommended to be less than 20 mSv a−1 [30]. Therefore, it can be concluded that the workersand technical staff at the processing pilot plant are nearly receiving the maximum annual effectivedose from the mineral concentrates during their work in the mineral processing pilot plant.

6. Conclusion

This study indicates that several locations of the recent and paleo-beach areas along the coastlinesouth of Cox’s Bazar, Bangladesh, where the beach sand contains considerable amounts of heavyminerals (more than 54 wt.%), show high concentration of uranium, thorium and potassium,resulting in an elevated exposure of the coastal environment and the people living in those specificlocations to radioactivity. The main conclusion arising from this study is the suggestion that if theraw sands or the mineral fractions are used for building purposes or industrial use, their elevatedactivity concentrations have to be considered from a radio-ecological point of view.

Acknowledgements

The gamma spectroscopic analysis was conducted during a training fellowship under a Technical Cooperation project(BGD/7/006) of the IAEA at the Department ofAnalytical Chemistry of the Helmholtz Centre for Environmental Research(UFZ) in Leipzig, Germany. We are thankful to the IAEA for the technical and financial support during the trainingfellowship. We are grateful to the Department of Analytical Chemistry of the UFZ for the help in gamma spectroscopicanalysis. We thank the scientific and technical staff of Beach Sand Minerals Exploitation Centre, BAEC, for their helpduring sampling and sample preparation.

References

[1] R.G. Schmidt and S.A. Asad, A Reconnaissance Survey of Radioactive Beach Sand at Cox’s Bazar, Vol. 3 (InterimGeological Report, The Geological Survey of Pakistan, East Pakistan, 1963), pp. 1–15.

[2] BSEC, Beach Sand Exploitation Centre Brochure, Cox’s Bazar, Scientific Information Division (Bangladesh AtomicEnergy Commission BAEC, Dhaka, 1994), pp. 1–21.

[3] M.A.B. Biswas, Lithological Criteria of Localization of Titano-Zirconium Placers in the Inshore Sediments of TeknafCoast, Bangladesh, J. Sci. Res. 2, 41 (1979).

[4] M.A.B. Biswas and S.K. Kunda, Mode of Occurrence and Origin of Heavy Mineral Deposits of Nijhum Dwip,Hatia, Nucl. Sci. Appl. 16 & 17, 69 (1985).

[5] M.H. Rahman, M.A. Islam, F.M.M. Shine, and F. Ahmed, Heavy-Mineral Studies of the Silkhali-Teknaf Beach,Dune and Cliff Sands, Cox’s Bazar, Bangladesh, Indian Min. 48, 167 (1994).

[6] M.M. Zaman, F. Deeba, M.Z. Kabir, M. Rajib, and S.M. Rana, Finding a New Heavy Mineral Deposit and ItsMineralogical Composition at Sonarpara, Cox’s Bazar, J. NOAMI 26, 17 (2009).

[7] M.N. Alam, M.I. Chowdhury, M. Kamal, S. Ghose, M.N. Islam, M.N. Mustafa, M.M.H. Miah, and M.M. Ansary,The 226Ra, 232Th and 40KActivities in Beach Sand Minerals and Beach Soils of Cox’s Bazar, Bangladesh, J. Environ.Radioact. 46, 243 (1999).

[8] M.M. Zaman, M.Z. Kabir, F. Deeba, M. Rajib, and S.M. Rana, Anomalous Radioactivity Due to the Presence ofHeavy Minerals in Beach Sands of Cox’s Bazar, Abstract 52, International Conference on Geoscience for GlobalDevelopment, Dhaka, Bangladesh, 2009.

[9] J.F. Donoghue and M.B. Greenfield, Radioactivity of Heavy Mineral Sands as an Indicator of Coastal Sand TransportProcesses, J. Coast. Res. 7, 189 (1989).

[10] M.S. Alam, N.E. Huq, and M.S. Rashid, Morphology and Sediments of the Cox’s Bazar Coastal Plain, South-EastBangladesh, J. Coast. Res. 15, 902 (1999).

[11] M.H. Monsur and A.S.M.M. Kamal, Holocene Sea-Level Changes Along the Maiskhali and Cox’s Bazar-TeknafCoast of the Bay of Bengal, J. NOAMI 1, 15 (1994).

[12] E.H. Macdonald, Manual of Beach Mining Practice – Exploration and Evaluation (Department of Foreign Affairs,Canberra, Australia, 1972).

[13] R.D. Evans, The Atomic Nucleus (McGraw-Hill, New York, 1969).

Dow

nloa

ded

by [

Hel

mho

ltz Z

entr

um F

Um

wel

tfor

schu

ng G

mbH

] at

05:

14 1

5 A

pril

2014

Isotopes in Environmental and Health Studies 525

[14] A. Malanca, R. de Pieri, and A. Gazzola, Radiogenic Heavy Minerals in Brazilian Beach Sand, J. Radioanal. Nucl.Chem. 230, 557 (1998).

[15] R.J.D. Meijer, I.R. James, P.J. Jennings, and J.E. Koeyers, Cluster Analysis of Radionuclide Concentrations in BeachSand, Appl. Radiat. Isot. 54, 535 (2001).

[16] A.S. Alencar and A.C. Freitas, Reference Levels of Natural Radioactivity for the Beach Sands in a BrazilianSoutheastern Coastal Region, Radiat. Meas. 40, 76 (2005).

[17] S.U. El-Kameesy, S.A.B.D. El-Ghany, S.M. El-Minyawi, Z. Miligy, and E.M. El-Mabrouk, Natural Radioactivityof Beach Sand Samples in the Tripoli Region, Northwest Libya, Turkish J. Eng. Environ. Sci. 32, 245 (2008).

[18] H. Yucel, M.A. Cetiner, and H. Demirel, Use of the 1001 keV Peak of 234mPa Daughter of 238U in Measurement ofUranium Concentration by HPGe Gamma-Ray Spectrometry, Nucl. Instrum. Methods Phys. Res. A413, 74 (1998).

[19] H. Yucel, H. Karadeniz, M.A. Cetiner, H. Demirel, and S. Turhan, Measurement of Absolute Intensity of 1001 keVGamma-Ray of 234mPa, J. Radioanal. Nucl. Chem. 258, 445 (2003).

[20] J. Beretka and P.Y. Mathew, Natural Radioactivity of Australian Building Materials, Industrial Wastes andBy-Products, Health Phys. 48, 87 (1985).

[21] UNSCEAR, Exposures from Natural Sources of Radiation. Sources, Effects and Risks of Ionizing Radiation, Reportto the General Assembly, United Nations, New York, USA, 1988.

[22] NEA-OECD, Nuclear Energy Agency, Exposure to Radiation from Natural Radioactivity in Building Materials,Reported by Nuclear Energy Agency group of Experts, OECD, Paris, 1979.

[23] A.K. Mohanty, S.K. Das, K.V. Van, D. Sengupta, and S.K. Saha, Radiogenic Heavy Minerals in Chhatrapur BeachPlacer Deposit of Orissa, Southeastern Coast of India, J. Radioanal. Nucl. Chem. 258, 383 (2003).

[24] A.K. Mohanty, D. Sengupta, S.K. Das, V. Vijayan, and S.K. Sahab, Natural Radioactivity in the Newly DiscoveredHigh Background Radiation Area on the Eastern Coast of Orissa, India, Radiat. Meas. 38, 153 (2004).

[25] V. Ramasamy, M. Dheenathayalu, R. Ravisankar, V. Ponnusamy, G. Victor Rajamanickkam, D.K. Sahayam,V. Meenakshisundram, andV. Gajendran, Natural Radioactivity Measurements in Beach-Rock Samples of South-EastCoast of Tamil Nadu, India, Radiat. Prot. Dosim. 111, 229 (2004).

[26] A.K. Mohanty, D. Sengupta, S.K. Das, S.K. Saha, and K.V. Van, Natural Radioactivity and Radiation Exposure inthe High Background Area at Chhatrapur Beach Placer Deposit of Orissa, India, J. Environ. Radioact. 75, 15 (2004).

[27] S. Righi, M. Andretta, and L. Bruzzi, Assessment of the Radiological Impacts of a Zircon Sand Processing Plant, J.Environ. Radioact. 82, 237 (2005).

[28] P.H.G.M. Hendriks, J. Limburg, and R.J.D. Meijer, Full-Spectrum Analysis of Natural γ -Ray Spectra, J. Environ.Radioact. 53, 365 (2001).

[29] J.H. Selby, The Industrial Uses of Zircon and Zirconia and the Radiological Consequences of These Uses, inProceedings Series, 5th International Symposium on Naturally Occurring Radioactive Material, Seville, Spain.IAEA, 95 (2007).

[30] IAEA, International Basic Safety Standards for Protection against Ionizing Radiation and for the Safety of RadiationSources, Safety Series no. 115, Schedule II, Vienna, Austria, 91 (1996).

[31] Y. Orgun, N. Altinsoy, S.Y. Sahin,Y. Gungor, A.H. Gultekin, G. Karahan, and Z. Karacik, Natural and AnthropogenicRadionuclides in Rocks and Beach Sands from Ezine Region (Canakkale), Western Anatolia, Turkey, Appl. Radiat.Isot. 65, 739 (2007).

[32] S. Murugesan, S. Mullainathan, V. Ramasamy, and V. Meenakshisundaram, Radioactivity and Radiation HazardAssessment of Cauvery River, Tamil Nadu, India, Iran. J. Radiat. Res. 8, 211 (2011).

[33] E.M. El Afifi, M.A. Hilal, S.M. Khalifa, and H.F. Aly, Evaluation of U, Th, K and Emanated Radon in Some NORMand TENORM Samples, Radiat. Meas. 41, 627 (2006).

Dow

nloa

ded

by [

Hel

mho

ltz Z

entr

um F

Um

wel

tfor

schu

ng G

mbH

] at

05:

14 1

5 A

pril

2014