Elemental analysis using natural gamma-ray spectroscopy

Nuclear Instruments and Methods in Physics Research A 353 (1994) 558-561

NUCLEAR INSTRUMENTS & METHODS IN PHYSICS RESEARCH Section A

ELSEVIER

Elemental analysis using natural gamma-ray spectroscopy A. AkSoy

a,*,

A.A. Nagvi

a,

F.Z. Khiari

a,

F. Abujarad

a,

M. Al-Ohali

a,

M. Sumani

b

a Energy Research Laboratory, King Fahd University of Petroleum andMinerals, Dhahran 31261, Saudi Arabia b Petroleum and Gas Technology Division, King Fahd University of Petroleum and Minerals, Dhahran 31261, Saudi Arabia

Abstract A gamma-ray spectroscopy setup has been recently established to measure the natural gamma-ray activity from potassium ( 4° K), uranium (238 U), and thorium (Z32Th) isotopes in rock samples of oil well-logs. The setup mainly consists of a shielded 135 cm3 Hyper Pure Germanium (HPGe) detector, a 5 in . X 5 in . NaT(TI) detector and a PC based data acquisition system . The core samples, with 70-100 g weight, have cylindrical geometry and are sealed such that radon gas from 238U decay would not escape from the sample. For room background subtraction, pure quartz samples identical to core samples were used . The sample is first counted with the HPGe detector to identify the elements through its characteristics gamma rays . Then the elemental concentration is determined by counting the sample with a Nat detector. In order to determine the absolute concentrations, the sample activity is compared with the activities of standards supplied by NIST and IAEA . The concentration of 238 U and Z32Th has been determined in ppm range with that of °K in wt .% . 1. Introduction Natural gamma-ray logs have been widely used to asses shale content in rocks [1]. These measurements provide means for determining the volume of shale and clay mineral type in conjunction with the other logs . However, calculations cannot be reliable unless the results are correlated with spectral gamma-ray analysis of rock samples in the laboratory [2,3]. Natural gamma radiation in the earth mainly comes from the decay of three radioactive isotopes : potassium-40 (4° K), uranium-238 (238U) and thorium-232 (Z32 Th). 4° K decays directly to stable argon-40 (4°Ar) with the emission of a 1460 keV gamma-ray, while uranium and thorium decay through long decay chains ending at the stable isotopes of lead . The 1765 keV gamma-ray from 2t4Bi (in the uranium decay chain) is very prominent and clear from all other naturally occurrm gamma-ray lines. Thus, it is used as the signature of 38 U. Similarly the 2615 keV gamma-line of 2°8 TI is commonly used as indication of Z32 Th . Since 238 U decay chain passes through radon (222Rn), which is a noble gas with a half-life of 3.8 days, one needs to have the sample air-tight sealed and kept for about 3 weeks to reach the equilibrium in the 238 U series [4]. The high energy resolution of Ge allows the identification of the element precisely while the higher efficiency of Nat detector allows the determination of the

* Corresponding author .

concentration of the element. Therefore a gamma ray setup consisting of a Ge and Nat detector can allow to the determination of the element concentration with good accuracy . A low level gamma-ray spectroscopy setup has been established at the Energy Research Laboratory of the Research Institute to measure the natural gamma-ray activity in various core samples. In this paper the setup is described and some results are presented. 2. Experimental The low level gamma-ray spectroscopy setup at ERL consists of a HPGe detector and a Nal(Tl) detector together with signal processing equipment and a data acquisition and analysis system . A schematic of the setup is shown in Fig . 1 . To identify the elements the sample was first counted with the HPGe detector with 2 keV resolution and a relative efficiency of 33%. Then a 5 in . X 5 in . Nal(TI) detector with 7 .3% resolution was used to determine the concentrations of the elements . After preamplification, the signal is processed through a Spectroscopy Amplifier and an ADCAM Multichannel Buffer Model 918A from EG&G ORTEC to an IBM-PC computer . The spectra are acquired in 4096 channels and calibrated with standard gamma-ray sources (60 Co and 237Cs). In the calibration, standard samples with high concentrations of K, U and Th are also used . The energy range of the spectrum covers the 1460 keV line from 4° K, the 1765 keV line from 214Bi and the 2615 keV line from 208 T1 .

0168-9002/94/$07 .00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0168-9002(94)00861-2

A. Aksoy et al . /Nucl. Instr. and Meth. in Phys . Res. A 353 (1994) 558-561

559

LEAD SHIELDING

MUMM rZ/

« ,M~ MO, r /% /r, -z:

WO/

VIIIIell ZU~ "

High

~~RO i äääää-

Voltage

Preamplifier

Multichannel Buffer

Pulser

IBM-PC GAMMASPECTRUM

DATA ANALYSIS REPORTS

Amplifier

Fig. 1. Schematic of the experimental setup used for natural gamma rays measurements in core samples.

The detector is well shielded, thus the room back-

measured several times for the same period as the sample

negligibly small. In order to have better statistics, the

in the data analysis . During each counting interval a pulser

hours and the acquired spectra were stored . Samples of

The pulse position indicates if any gain shift happened

ground is very low and the dead time of the system is gamma-ray activity of each sample was measured for 24 pure quartz (silicon dioxide, Si02) were used for room

background measurements [8,9]. Thus, several forms of quartz samples were measured to make sure that they are free from K, U, Th. Then, the room background was

and the average value from the background runs was used

was used to give a known pulse rate at a certain channel.

during the counting period . The pulser was brought back to the same channel before each counting period .

Fig. 2 shows a typical background-subtracted gamma-

ray spectrum of a sample taken at a depth greater than

400

z

200

2000 0

2500 .0

3000 0

ENERGY (keV ) Fig. 2. Net gamma ray spectrum of a core sample at the depth above 5000 ft from a sandstone reservoir. VIIIb. ENVIRONMENTAL APPLICATIONS

560

A. Aksoy et al. /Nucl. Instr. and Meth. in Phys. Res. A 353 (1994) 558-561

5000 ft from a sandstone reservoir. The collected spectra were then analyzed off-line by putting the region of inter-

est (ROI) around the 1460 keV peak of 40 K, 1765 keV 214Bi 238U) 208 peak of (from and the 2615 keV peak of T1 Z32 Th). (from The width of these gates were taken large enough to cover the whole peak. The intervals of these

gates were 230 channels for K, 300 channels for U, and 390 channels for Th. The HPGe spectra did not show any

interferences in the ROIs . The integrated counts under each peak were determined . The background subtraction

and dead time correction were also done . The stripping factors, which are the effects of scattering of high energy

gamma rays to lower energy peaks, were measured and the corrections were incorporated [10]. Later on, the sample weight normalization corrections were also done . Similar measurements were done for standards of rock and soil samples. These standards were provided by Inter-

national Atomic Energy Agency (IAEA), Austria and National Institute of Standards and Technology (NIST), USA. They have certified values of K in percentage by weight (wt.%), Th, and U in parts per million (ppm) range, respectively . Finally, from the comparison of corrected net

counts from samples and standards the concentrations of K in wt .%, U and Th in ppm in the core samples were determined . The total gamma rays emitted by core

samples is

widely being used in well-logs for the interpretations of the type of the rock samples [10] . The total gamma rays is usually given in

American

Petroleum

Institute

Units

(APIU) . The APIU is defined such that the activity of an

average shale considered to contain 6 ppm of U, 12 ppm of Th and 2% of K will read in the vicinity of 100 APIU on

where Al, A2, A3 are the total integrated counts/day under the peak in the Th, U, and K region of interest (ROI), respectively, Fh is the calibration factor for the Z32Th series in the Th ROT, i.e . the net counts/day in the Th ROT of the standard of 100 g divided by the concentration of the actual Th in the standard, Fu is the calibration factor for the 238U series in the U ROT, Fk is the calibration factor for the 4° K series in the K ROI, Sth is the stripping factor for the Z32Th series over the U ROT, (i.e . ratio of the Z32Th series counting over the U ROI), S~;, is the stripping factor for the Z32Th series over the K ROT, S;;

is the stripping factor for the 238U series over the K ROT, Bt, B2, B3 are the background counts over the Th, U and K ROT, respectively measured on pure quartz (Si0 2) samples of the same weight and over the same period of time, Z32Th, 238U and Ath , A., Ak are the net counts for and 40 K over the Th, U, and K ROT and normalized to the sample weight of 100 g.

The net counts under each peak were obtained by subtracting the background from the measured counts in the same ROT of the spectra. The stripping factors from Th and U peaks were obtained from the measurements of pure thorium and uranium. The calibration factors were deter-

mined from the measurements of the standards. The weights of the samples were determined with 1 mg uncertainty.

Therefore, inserting these measured values into the above equations, the concentrations of U and Th in ppm and K in wt .% can be calculated. The uncertainties of the measure-

ments are the statistical uncertainties of the measurements with those of the background counts folded in.

total gamma rays logs [ll] . Therefore by measuring the

3. Results

the APIU .

Table 1 lists the results of the measurements for 10 studied core samples taken from various depths. Using pure quartz samples, five background spectra were ac-

standards the total counts in each sample was converted to

2.1 . Sample preparation Rock samples with 70-100 g/sample were pulverized, carefully weighted with a balance having a sensitivity down to 1 mg range. They were also packed inside plastic

containers of about 3 in. diameter and 0.5 in. height. The sample containers were in turn sealed with silicon rubber

and placed in heat sealed thin plastic bags . To prevent the

radon gas from escaping, the samples were left for about 3 weeks (4 half-lives) to reach equilibrium due to the radon gas decay. Similarly, several quartz samples were prepared with a geometry identical to that of the core samples. 2.2 . Measurement method To determine the concentration of Th, U and K in the

sample the following equations are used [12] : Th = (1/Fth)(AI -B1) = (1/Fh)Ath1

U = (1/F~)(A2 - SthAth -B2) = (1/F~)A., K= ( I /Fk)(A3 - SIh Ath S" Au -B 3) = (1/Fk)Ak,

quired . The average of the background spectra corrected

Table 1 Concentrations of natural gamma rays of potassium, uranium, and thorium in core samples from a sandstone reservoir at depths above 5000 ft Sample number

K

(wt.%)

U (ppm)

Th (ppm)

Total GR (APIU)

1 2 3 4 5 6 7 8 9 10

0.2±0 .01 1.2+0 .01 0.7±0 .01 0.3±0 .01 1.7±0 .01 0.3+_0.01 0.1+_0.01 1.3+0 .01 0.9+_0.01 0.7+0 .01

0.2±0 .03 1 .6+0 .04 1 .2±0 .03 0.7+0 .03 1 .8±0 .03 0.4+_0.04 0.2+_0.03 1 .4+0.03 2.4+_0.03 2.4+0.03

1.4+0 .15 5.2+_0.17 4.8±0 .14 2.9±0 .14 6.9±016 1 .8+_0.17 1.4+_0.15 5.6+0 .15 10 .5_+0.16 6.7+0 .15

7.1±1 .5 48 .1+2 .0 30 .6±1 .3 16 .2±1 .1 63 .8±2 .0 11 .1_+1 .5 5 .8+1 .2 45 .8+1 .6 47.4_+1.1 39.2+1 .2

A. Aksoy et al. /Nucl. Instr. and Meth . in Phys. Res. A 353 (1994) 558-561

for the sample weight was used to calculate the background-subtracted net counts under the three peaks coming from K, U and Th . Fig. 2 shows a typical background-subtracted gamma-ray spectrum of the sample extracted at the depth above 5000 ft from a sandstone reservoir. The spectra were stored and then analyzed offline. The integrated counts under the three peaks were obtained . The background-subtracted net counts under these three peaks were then normalized to the sample weights to get the net counts per 100 g sample . The K, U and Th relative concentrations versus the depths at which the core samples were taken are determined. The uncertainties of the measurements were mainly statistical. These uncertainties varied depending on the amount of activity in the sample . They ranged between 0.6% and 11% for potassium, 1.3% to about 27% for uranium and between 1 .5% and 23% for thorium. The overall uncertainties were derived from the statistical errors of the measurements and the errors in the background counts. In total, 13 standards of U, Th and K were measured ; 7 of them from IAEA, and 6 from NIST . Each standard was measured for 24 h, with the same geometry as the sample . The effect of the stripping factors were also measured using pure standards of Th and U. The stripping factor of K on U was 5.6% while those of Th and U on the K peak were 3.4% each . Using the calibration curves for K, U and Th, the final results of the concentrations were converted into K in wt .%, U, and Th in ppm. The total counts equivalent of one APIU was determined and the results of total gamma rays versus depths for the samples were obtained . The results of 10 samples are shown in Table 1. The results show that the activities were varying from 0.1 to 1.7 wt .% for K, from 0.2 to 2.4 ppm for U and 1.4 to 10 .5 ppm for Th. The total gamma rays were ranging from 7.1 to 64 APIU for these samples. The results show that there are some zones radioactively rich which might suggest the presence of clay type shale. 4. Summary An experimental setup has been installed for radiometric measurements of core samples in the laboratory for Rock and Mineral Characterization . The setup consists of a 5 in . X 5 in . Nal(TI) detector and a HPGe detector with signal processing electronics and a PC based data acquisition and analysis system . Core samples taken from a

561

sandstone reservoir at different depths above 5000 ft were prepared and their natural radioactivities measured . The (integrated) background-subtracted net counts were obtained which correspond to the relative contents of potassium, thorium and uranium in the samples. Similar measurements were done for the standards from NIST and IAEA which have certified values of the radioisotopes concerned. By comparing the counts from core samples and the standards the concentrations of K in wt .%, U and Th in ppm were determined . The total gamma rays emitted by the samples in ÀPIU were also determined . Acknowledgements The authors wish to acknowledge the Saudi-Aramco for the support to this work on the KFUPM, RI Project No . 21110. The Research Institute of King Fahd University of Petroleum and Minerals is acknowledged for the support provided to establish the setup. References

[2] [3] [4] [5] [6]

[8] [9] [10] [11] [12]

O. Serra, J. Baldwin and J. Quirein, Theory, interpretation and practical applications of natural gamma-ray spectroscopy, SPWLA, 21st Annual Logging Symp . 1980, p. 1. W.H . Ferti, J. Petroleum Technol. (February 1987) 175. J.T . Dewan, Essentials of Modern Open-Hole Logs Interpretation (PennWell, Oklahoma, 1983) pp . 50-109. R.J . Budnitz, AN . Nero, D.J . Murphy and R. Graven, Instrumentation for Environmental Monitoring, Vol. I, (1983) p. 440. A. Aksoy, J. Radioanal. Nucl . Chem. Articles 169/2 (1993) 463. A. Aksoy, A.A. Naqvi, F.Z. Khiari, M. Raashid, A. Coban, R.E . Abdel-Aal and H. Al-Juwair, Nucl. Instr. and Meth . A 332 (1993) 506. A. Aksoy, Abstract Book 12th Int. Conf. on the Application of Accelerators in Research and Industry, Denton, Texas, USA, p. 116. G.I . Khalil and Cs .M . Buczko, J. Radioanal. Nucl. Chem . Lett . 95 (2) (1985) 101. N.K. Mumba, L. Vas and Cs . Buczko, J. Radioanal. Nucl . Chem . Lett . 95 (5) (1985) 311. Radiometric Reporting Methods and Calibration in Uranium Exploration, IAEA Technical Reports series No . 174 (1976). J. Dewan, Essentials of Modern Open-Hole Logs Interpretation, (PennWell, Okalahoma, 1993) p. 51 . P.G . Killeen and C.M . Carmichael, Can. J. Earth Sci. 7 (1970) 1093 .

VIIIb. ENVIRONMENTAL APPLICATIONS