Correlation_of_properties_of_coal_with_radionuclides.pdf

Fuel Vol. 76, No. 10, pp. 951-955, 1997 © 1997 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0016-2361/97 $17.00+0.00

PIh S0016-2361(97)00082-3

ELSEVIER

Correlations of properties of Spanish coals with their natural radionuclides contents Pedro Fernandez, Ramona M. Diaz and Jorge Xiberta Department of Energy, University of Oviedo, E. T.S.I.M.O., Independencia 13, 33004 Oviedo, Spain (Received 12 December 1996; revised 25 February 1997) Proximate analyses, heating values and natural radionuclide contents (238U,23ZTh,40K)of 36 samples of bituminous coals and 4 samples of anthracite were determined. The proximate analysis data and higher heating value were correlated by multiple linear re~ression analysis with the radionuclide contents. For each coal rank, the correlations obtained (giving in most cases r > 0.9) are comparable with previous correlations based on measurements of natural gamma-ray activity of coal. Some improved correlations were obtained by considering separately the coals from each washing plant, with a degree of accuracy in accordance with the reproducibility tolerances of the Spanish and international standards UNE 32 004, 32 019 and ISO 1928-76. © 1997 Elsevier Science Ltd. (Keywords: coal properties; prediction; natural radionuclides)

Coals have a very important function in the energyproducing sector, so their good characterization is essential for the rational and efficient use of this energy resource. The characterization of coal by determining technological parameters such as ash yield, volatile matter and heating value is necessary since these parameters are required in many engineering calculations to evaluate the performance of existing combustion systems and/or to design new combustion systems. These parameters are generally determined by slow standardized analysis tests; but due to the fact that the level of natural gamma radiation emitted by coal increases linearly with ash yield, the use of the radiometric technique of measuring natural gamma radiation yields a rapid and accurate method for determining the ash yield of coals 1-3. Based on both the content of natural radionuclides 238U, 232Th and 4°K analysed by gamma-ray spectrometry and the existence of proved relations between ash yield and the above additional parameters 5, the development of correlations to predict the coal properties from composition of the natural radionuclides can be of great interest. The aim of this work was to study the relations between the properties of Spanish coals and the contents of their natural radionuclides 238U, 232Th and 4°K, as well as to develop correlations for estimating such properties in good agreement with the measured values, establishing both the scope and the accuracy of the correlations found. To this end, an experimental survey was carried out on a number of coals from several coal washing plants situated in the north of Spain (Asturias), used for electric power generation. EXPERIMENTAL

Sampling Samples of coal used were supplied by Lada power plant (Asturias), being representative of the whole range of

characteristics of the selected coals. Sampling was carried out at the power plant, the samples being separated by the station's own laboratory facilities. Most of the samples are classified as bituminous coal (samples 1-36); four are anthracites (samples 37-40).

Proximate analysis and higher heating value Routine proximate analyses (moisture, volatile matter, ash) together with heating value determinations were carried out at the Lada power plant laboratories. The determinations were performed for each coal as described below. A LECO MAC-400 thermogravimetric coal analyser was used for ash, volatile matter and moisture determinations according to ASTM standard D5142-90. These analyses were repeated at least three times for good accuracy. A LECO AC-300 adiabatic calorimeter was used to determine higher heating values of air-dried coal samples according to ASTM standard D2015-91. Calorific value determinations were also repeated at least three times for each coal, the results differing by < 50 kJ kg-l. The data are shown in Table 1. Natural radionuclide analysis The analysis of natural radionuclides in the coal, namely 238U, Z32Thand 4°K, w a s performed by both scintillation and semiconductor gamma-ray spectrometry. The results of the two analysis techniques agreed (see Table 1). The scintillation gamma-ray analyses were carried out in the Nuclear Energy Laboratories at the University of Oviedo by using a gamma-ray spectrometer equipped with an NaI(T1) gamma-ray detector and a multichannel analyser (Camberra Series 35 with 1024 channels) 2. The semiconductor gamma-ray analyses were carried out at Barcelona University by using a high-resolution semiconductor

Fuel 1997 Volume 76 Number 10

951

Correlations of properties of Spanish coals with their natural radionuclides contents: P. Fernandez et a l.

Table 1

Sources and properties of coals Proximate analyis (wt% db)

Sample no.

Washing plant

1 2

Modesta Modesta

3 4 5

Modesta Modesta Modesta

6 7c

Modesta Modesta

8 9 10d

Modesta b Modesta b Modesta b

11 12

Modesta b Modesta b

13 14 15 16d

Asturleonesa Asturleonesa Asturleonesa Asturleonesa

17

Asturleonesa

18 19 20 21 22 d

Asturleonesa Carrocera Carrocera Carrocera Carrocera

23 24 25 26 27 28 29 30 31 32 d

Carrocera Carrocera Tur6n Tur6n Tur6n Tur6n Tur6n Tur6n Candin Candin

33 34 c 35 36 d

Candin Candfn Candfn b Candfn b

37 38 39 40

Rank

Radionuclide analysis (ppmw db) 238U 232Th 40K

HHV ~ (MJ kg -l)

A

VM

FC

21.32 22.12

37.3 34.8

24.6 26.3

38.9

2.2 1.9

6.6 5.2

1.1

29.29 28.91 27.10

16.8 17.7 22.4

32.3 33.0 29.4

50.9 49.3 48.2

1.4 1.4 1.6

2.3 3.3 4.3

0.5 0.5 0.8

27.77 25.38 29.70 30.08 30.45 29.75 30.82 23.21

21.0 25.6 16.2 14.3 13.6 14.8 13.2 31.1

30.7 27.7 35.0 33.2 33.2 33.0 34.7 24.1

48.3 46.9 48.8 52.5 53.2 52.2 52.1 44.8

1.5 1.9 1.2 1.1 1.1 1.1 0.9 2.0

3.7 5.6 3.0 2.1 2.2 3.0 2.7 4.8

0.6 0.9 0.5 0.4 0.4 0.5 0.4 1.3

21.54 20.76

35.5 38.1

23.1 22.4

41.4 39.5

2.5 2.7

5.7 7.7

1.4 1.5

20.57 20.39 22.89 22.01

38.5 38.3 31.8 34.1

22.1 21.8 24.5 25.8

39.3 39.9 43.8 40.1

1.5 2.4 2.2 2.2

7.1 8.1 5.5 4.6

1.7 1.6 1.3 1.2

bituminous bituminous bituminous

23.55 27.12 23.83 26.93 21.15 20.62 23.17 29.22 27.32 22.33 23.16 26.70 26.99 28.26 27.98 33.12

30.5 22.1 30.2 21.9 37.9 40.4 33.4 17.3 22.0 36.2 32.8 21.9 20.8 18.4 17.4 5.2

26.3 28.5 25.9 28.8 23.6 14.2 17.7 34.0 30.7 16.2 17.5 29.9 29.8 30.8 31.0 36.7

43.2 49.4 43.9 49.3 38.5 45.4 48.9 48.7 47.2 47.6 49.7 48.2 49.4 50.8 51.5 58.1

2.1 1.4 1.7 1.6 2.3 2.9 2.6 1.2 1.4 2.5 1.9 1.2 1.1 1.1 1.3 0.3

4.6 4.3 5.8 3.5 7.3 7.0 5.6 2.8 3.6 6.3 6.1 2.9 3.6 3.4 4.5 0.1

1.0 0.8 1.2 0.7 1.4 1.5 1.0 0.5 0,7 1,3 1,1 0.6 0,8 0.6 0,6 0,1

bituminous anthracite anthracite anthracite anthracite

33.22 20.41 18.80 22.15 23.04

5.2 34.3 38.9 31.1 30.6

37.0 5.5 5.7 8.3 8.4

57.9 60.1 55.1 60.5 60.7

0.5 1.7 1.7 3.3 3.1

1.2 4.3 5.5 3.5 3.3

0.01 0.9 1.2 0.8 0.8

bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bituminous bitummous bituminous bituminous bituminous bituminous bituminous

38.1

1.2

Higher heating value bSpecial coal cAnalysed by only semiconductor gamma-ray spectrometry dAnalysed by both semiconductor and scintillation gamma-ray spectrometry

gamma-ray spectrophotometer equipped with an intrinsic germanium detector (Camberra model GR 2020 7500SL) a n d a m u l t i c h a n n e l a n a l y s e r ( C a m b e r r a Series 35 P L U S w i t h 4 0 9 6 c h a n n e l s ) . All s a m p l e s , c r u s h e d to < 2 0 0 / ~ m , w e r e s e a l e d a n d a l l o w e d to r e a c h r a d i o a c t i v e e q u i l i b r i u m before counting.

Statistical methods T o derive correlations to predict the coal properties f r o m 238T 232 the natural r a d i o n u c l i d e c o n t e n t s e x p r e s s e d as [ U], [ Th] a n d [4°K], the f o l l o w i n g statistical m e t h o d s were used:

952

Fuel 1997 V o l u m e 76 N u m b e r 10

T h e i n f l a t i o n a r y v a r i a n c e f a c t o r m e t h o d 6, s t e p w i s e r e g r e s s i o n m e t h o d a n d D u r b i n W a t s o n a n a l y s i s 7 w e r e perf o r m e d w i t h the c o m m e r c i a l s o f t w a r e p a c k a g e S T A T V I E W (Abacus Concepts).

Principal components analysis. T h e d e t e r m i n a t i o n o f p r i n c i p a l c o m p o n e n t s c o n s i s t e d in the c a l c u l a t i o n o f the d e g r e e o f c o r r e l a t i o n b e t w e e n the i n d e p e n d e n t v a r i a b l e s ( n a t u r a l r a d i o n u c l i d e c o n c e n t r a t i o n s , Xi) b y u s i n g the inflat i o n a r y v a r i a n c e f a c t o r ( I V F ) for e a c h i n d e p e n d e n t v a r i a b l e . I f a set o f v a r i a b l e s are n o t c o r r e l a t e d , I V F = 1. F o r a h i g h

Correlations of properties of Spanish coals with their natural radionuclides contents: P. Fernandez et al. grade of correlation between variables, IVF -> 10. The independent variables showing multicollinearity (IVF > 10) were excluded 6. Multiple linear regression. The independent variables (radionuclide contents) were selected by the stepwise regression method, and the coefficients for multiple linear correlation were calculated for the equation Y = a + bXj + cX2 ÷ dX3, where Xi represent the independent variables (radionuclide contents) and Y the dependent variable (the measured values of a property). Residuals analysis. The autocorrelation errors in the model obtained were determined by analysis of residuals both by a graphic method consisting in plotting the measured values versus residuals and by the Durbin Watson method 7.

RESULTS AND DISCUSSION Proximate analyses, higher heating values and concentrations of trace components of the coals studied are given in Table 1. Analysis of Table 1 shows that: (1) Ash yield varies from 5.2 to 40.4 wt% for the bituminous coals and from 30.6 to 38.2 wt% for the anthracites. (2) Volatile matter varies from 14.2 to 37 wt% for the bituminous coals and from 5.5 to 8.4 wt% for the anthracites. (3) Higher heating value varies from 20.39 to 33.22MJkg -1 for the bituminous coals and from 18.80 to 23.04 MJ kg -~ for the anthracites. (4) The variables investigated as alternative correlating parameters for the above coal properties were the natural radionuclides 238U, 232Th and 4°K. For bituminous coals, the 238Ucontents are in the range 0.3-2.9 ppmw, with 232Th0.1-8.2 ppmw and 4°K 0.01-1.7 ppmw. For anthracites, the ranges of values are 1.7-3.3, 3.3-5.5 and 0.8-1.2ppmw respectively. These values are ,) within the ranges for Spanish coals , except for the low 4°K contents of samples 35 and 36, whose measured values are 0.1 and 0.01 ppmw respectively. This may be due to the low ash yield of these samples, since 4°K is associated with inorganic matter of coal . The dependence of the coal properties (ash yield, volatile matter and higher heating value) on the natural radionuclide

contents of the mineral matter may be expressed by the following general linear correlations: A = a + b[238U] ÷ c[232Th] ÷ d[a°K]

(1)

VM = e +f[238U] + g[232Th] ÷ h[4°K]

(2)

HHV = m + n[238U] +p[232Th] + q[4°K]

(3)

where A, VM and HHV are the ash yield (wt%), volatile matter (wt%) and higher heatinAg value (kJ kg -1) respectively, and [238U], [23~Th] and [4~K] are the concentrations of the natural radionuclides in the coal (ppmw). All data are expressed on a dried-coal basis. The relations between the coal properties studied and the concentrations of natural radionuclides were determined for all the coals as one group, for each rank of coal, and for the coal from each coal washing plant. Tables 2 - 4 show the experimental ranges of values, the linear correlations obtained for each property studied, the squared linear correlation coefficients ( r 2) , the standard deviation and the maximum errors of prediction for the derived equations. To indicate the accuracy of the above equations, the terms 'satisfactory' and 'indicative' are used. The term 'satisfactory' is used to denote that the results obtained with a given correlation have the same degree of accuracy as the results obtained, taking account of the requirements of the experimental methods, and therefore the errors in the predicted values are in accordance with the reproducibility tolerances of the corresponding UNE and ISO reference methods. The term 'indicative' shows that the results are less accurate than those obtained with the above reference methods, although the correlations found are valid for design and other industrial purposes. According to UNE 32004 and 32019, for ash yields < 10 and > 10 wt%, the reproducibility tolerances are -< 0.3% and --- 3% for ash, and -< 0.5% and -< 5% for volatile matter, respectively. According to ISO 1928-76, the maximum admissible deviation is 300 kJ kg -~ for higher heating value. Ash correlations The linear correlations found are shown in Table 2. The experimental parameters fitted the above linear equations reasonably, yielding correlation coefficients > 0.9, except for the set of bituminous coal samples with ash yield > 30 wt%. For anthracites, the correlation obtained is 'satisfactory',

Table 2 Natural radionuclide correlations for coal ash yield (A) Samples

Ash range (wt%)

Correlation

Equation no.

Correlation coefficient (r2)

Standard deviation

Max. error (%)

A = 17.8 + 3.81232Th]

4

0.999

-< _+0.2

1.3

Anthracites All

38.9-30.6

Bituminous coals All

5.2-40.4

A = 2.8 + 5.21238U] + 15.814°K]

5

0.965

-< _+5.0

16.7

A < 25 wt%

5.2-22.4

A = 2.2 + 4.91238U] + 17.014°K]

6

0.945

-< +3.5

69.2

A > 30 wt%

30.2-40.4

A = 20.7 + 2.31232Th]

7

0.613

-< _+4.0

19.8

Candin

5.2-21.9

A = 1.81 + 6.71238U] + 15.314°K]

8

0.920

< _+3.0

15.4

Carrocera

21.9-37.9

A =

9

1.000

-< -+0.1

1.6

Modesta

16.8-37.3

A = 2.7 + 28.514°K]

10

0.960

-< -+3.2

12.8

- 0.8 + 7.81238U] + 14.914°K]

Modesta"

13.2-16.2

A = 1.5 + 12.81238U] + 2.2[23ZThl - 14.814°K]

11

0.984

-< _+0.1

3.5

Tur6n

17.3-40.4

A = 2.3 + 3.71238U] + 3.91232Th]

12

0.997

-< _+0.6

7.8

Asturleonesa

38.5-31.1

A =

13

0.978

-< +_0.5

4.8

- 9.4 + 4.41238U] - 0.71232Th] + 27.314°K]

Special coal

Fuel 1997 Volume 76 Number 10

953

Correlations of properties of Spanish coals with their natural radionuclides contents: P. Fernandez et al.

Table 3

Natural radionuclide correlations for coal volatile matter (VM)

Samples

Volatile matter range (wt%)

Anthracites All

5.5-8.4

B i t u m i n o u s coals All a

Equation no.

V M = 2.5 + 1.81238U]

Correlation coefficient (r 2)

Standard deviation

0.975

-< -+0.25

14

Max. error (%) 13.8

21.8-37.0

V M = 38.2 -

1.31238U] - 8.614°K]

15

0.966

-

+--2.5

12.1

21.8-27.7

V M = 38.1 -

lI.I[4°K]

16

0.834

<- ---2.5

11.1

A > 30 w t % ~

28.5-37.0

V M = 36.4 -

1.11238U] - 7.314°K]

17

0.954

-< ---0.7

4.8

Candfn

29.9-37.0

V M = 38.9 - 4.41238U] - 5.614°K]

18

0.985

<-- _+0.4

4.5

Carrocera

23.6-26.3

V M = 36.0 - 2.71238U] - 0.91232Th]

19

0.970

-< -+0.4

4.3

Modesta

24.6-33.0

V M = 37.9 -

10.814°K]

20

0.977

--< _+0.7

6.2

Turrn

14.2-30.7

V M = 4 7 . 9 - 5.01232Th]

21

0.974

<- _+2.2

28.4

Asturleonesa

21.8-24.5

V M = 35.4 -

22

0.956

-< _+0.3

3.2

A < 25 w t % a

a

Correlation

1.01238U] - 7.014°I(]

E x c l u d i n g coals f r o m T u r 6 n

since it reproduces experimental results with a maximum error of 1.3%, the standard deviation being -< ---0.15 (see Equation 4 in Table 2). The development of correlations for the whole set of bituminous coals did not give good results. Although the inclusion of all samples of bituminous coals results in the largest range of ash values, the correlation derived is not good for prediction, since the maximum error and standard deviation are 16.7% and -< +--5.0 respectively. The calculated ash yields are compared with the experimental data in Figure 1. In an attempt to improve the results, two sets of samples were analysed, with ash < 25 and > 30 wt%. However, the correlations found were no better, indicating that the ash yield cannot be predicted accurately from them (see Equations 5 and 6 in Table 2). Smaller errors were obtained by using reduced sets of data that corresponded to the samples from each coal washing plant. As shown in Table 2, the relation between ash yield and natural radionuclides varies for the raw materials from the five coal washing plants (Equations 8-13). The best correlations were obtained for Carrocera (showing satisfactory results), special coal from Modesta, Tur6n and Asturleonesa. The correlations derived for Candfn and Modesta were rather poor, meaning that ash cannot be predicted accurately from them. The results obtained using the correlations from each coal washing plant are shown in

Higher heating value correlations

Natural radionuclide correlations for coal higher heating value (HHV)

Samples

Anthracites All Bituminous coals All A < 25 wt% A > 30wt% Candin Carrocera Modesta Modesta" Tur6n Asturleonesa

Higher heating value range (kJ kg - i)

Correlation

Equation no.

Correlation Standard Max. coefficient deviation error(%)

(r E)

18799-23040

HHV = 27429.5 - 3490.21232Th]+ 8815.114°K]

23

0.994

-< -+120

2.3

2O 3 9 0 - 3 3 218 2 6 7 0 3 - 3 3 218 20 3 9 0 - 2 3 801 26 7 0 3 - 3 3 2 1 8 21 1 5 2 - 2 7 118 21 3 1 5 - 2 9 2 9 1 29 6 9 7 - 3 0 819 20 6 2 0 - 2 9 215 20 3 9 0 - 2 3 212

HHV HHV HHV HHV HHV HHV HHV HHV HHV

24 25 26 27 28 29 30 31 32

0.964 0.908 0.672 0.962 0.999 0.961 0.792 0.996 0.995

-< __.800 -< 4-1720 -< _+1000 _< _+600 ------+90 -- -+1200 -< -+175 -- -+250 -< _+90

4.4 6.2 9.8 7.1 1.2 4.6 2.2 2.9 1.3

= = = = = = = = =

34220.1 33 694.2 28 366.4 35 152.0 35 588.1 34856.7 34988.1 34669.7 37 788.5 -

Special coal

954

The linear correlations found are shown in Table 4. In general, good correlations were obtained for volatile matter; the worst correlation (r 2 = 0.834) was obtained for samples with ash yield < 25 wt%. For anthracites, the results obtained can be classified as 'indicative'. The corresponding equation (14) in Table 3 reproduces experimental volatile matter values with a maximum error of 13.8%, the standard deviation being -< ___0.25. For the complete set of bituminous coals, the correlations derived are not acceptable, since the data from Turrn are not fitted to the corresponding equations. By excluding the Turrn coals, a very reasonable linear equation was found (r 2 = 0.966), the standard deviation and the maximum error being -< ---0.25 and 12.1% respectively (see Equation 15 in Table 3). By considering the two sets of samples with ash yields < 25 and > 30 wt%, the correlation found for A > 30 wt% is similar to the general Equation (15), but showing a more efficient behaviour than this. In most cases, the volatile matter correlations obtained for coals from each coal washing plant are considered as 'indicative'. The best correlation is found for Asturleonesa, whose results are 'satisfactory'. In contrast, the correlations obtained for Turrn and the special coal from Modesta are not acceptable.

The linear correlations found (Equations 2 3 - 3 2 ) are

Figure 2.

Table 4

Volatile matter correlations

Fuel 1997 V o l u m e 76 N u m b e r 10

1513.5123SU] - 6689.014°K] 8861.414°K] 414.31232Th] - 2915.514°K] 4348.2123SU] - 4362.714°K] 3108.9123SU] + 308.81232Th] - 6841.414°K] 11 197.714°K] 3713.1123SU] - 323.71232Th] 1057.91238U] - 1559.91232Th] 1595.4123SU] + 235.21232Th] - 9693.014°K]

Correlations of properties of Spanish coals with their natural radionuclides contents: P. Fernandez et a l.

50.00

'

40.00 -

I

'

I

'

-I-

Anthracites (Eq. 4)

0

Bituminous Coals (Eq. 5)

50.00

'///

I

-

Oj ~"

--

#..

30.00

'

X

40.00

Diagonal line

-

'

/

A ._¢ >J: m < "O

I

A

I

'

Candin (Eq, 8)

I

'

Zl

Carrocera (Eq. 9)

0

Modesta (Eq. 10)

Jr-

Special Modesta (Eq. 11)

[]

I

I

J / /

',

/

/

. "

J ~r

Turon (Eq, 12)

f.~

Asturleonesa (Eq. 13)

._Q 30.00 >J: m

'

-

,~

"O

20.00

20.00 _o

ca

ca

O

10.00

0.00 0.00

10.00

i

I 10.00

,

I 20.00

=

I 30.00

,

I 40.00

,

0.00 0.00

50.00

,

Measured Ash Yield (%)

I 10.00

,

I 20.00

,

I 30.00

J

I 40.00

=

I 50.00

Measured Ash Yield (%)

Figure 1 Ash yield of coals using the linear correlations from coal rank

Figure 2 Ash yield of coals using the linear correlations from coal washing plant

shown in Table 4. For anthracites, the results obtained are 'satisfactory'. The corresponding equation in Table 4 (23) reproduces the experimental higher heating value with a maximum error of 2.3%, the standard deviation being -< +120. For the whole set of bituminous coals, the correlation obtained is not acceptable. For the two sets of samples with ash yields < 25 and > 30 wt%, the correlations found did not improve the results (see Table 4). The correlations obtained for the higher heating value for each coal washing plant are quite satisfactory. The best correlations are found for Carrocera, Asturleonesa and the special coal from Modesta, while the correlation obtained for Tur6n can be classified as 'indicative'. On the other hand, the correlation obtained for Modesta is not acceptable.

acceptable correlations than those developed for blends of coals. Some of the equations obtained can be used in process research and design, having a similar accuracy to that obtained by the corresponding UNE and ISO test methods, and are valid for application in industrial practice.

REFERENCES 1 2

Alvarez, M. C. and Dopico, M. T., Nuclear Geophysics, 1991, 5, 507. Alvarez, M. C. and Dopico, M. T., International Journal of Environmental Issues in Minerals and Energy Industry,

3

CONCLUSIONS

4

Linear correlations were found between technological parameters of coal (ash yield, volatile matter and higher heating value) and natural radionuclide contents (2~U, 232Th and 4°K). Coals from a single washing plant have low variation in properties and produce generally more

5 6 7

1993, 63. Alvarez, M. C., Dopico, M. T. and Gonz~ilez, J., International Journal of Environmental Issues in Minerals and Energy Industry 1993, 95. Alvarez, M. C., Garz6n, L. and P6rez, J. M., Industria Minera, 1988, (280), 35. Urkan,M. K. and Arikol, M., Fuel, 1989, 68, 527. Berenson, M. I. and Levine, D. D., Basic Business Statistics Concepts and Applications. Prentice-Hall, Mexico, 1989. Canavos,G. C., Probability and Statistics. Applications and Methods. McGraw-Hill, Mexico, 1989.

Fuel 1997 Volume 76 Number 10

955