FINAL TECHNICAL REPORT September 1, 1993 through August 31, 1994

Project Title: CHLORINE IN COAL AND ITS RELATIONSHIP WITH BOILER CORROSION

DOE Grant Number: DE-FC22-92PC92521 (Year 2) ICCI Project Number: 93-1/6.1B-lP Principal Investigator: M.-I.M. Chou, (217) 244-0312,

Illinois State Geological Survey (ISGS), 615 E. Peabody Drive, Champaign, IL 61820

Other Investigators: J.M. Lytle, R.R. Ruch, and C.-L. Chou, (ISGS) ; J.L. Margrave and V. Khabashesku, Rice University (RU); G.P. Huffman and F.E. Huggins, University of Kentucky (UK) ; W.P. Pan and L. Liu, Western Kentucky University (WKU) .

Pro] ect Manager: Ken K. Ho, ICCI

ABSTRACT

Literature and use-history data indicate that high-C1 Illinois coals do not cause boiler corrosion while extensive data developed by the British correlate corrosion with chlorine content and other parameters related to the coal and boiler. The differences in corrosivity in coals may be due to coal properties, to coal blending, or to boiler parameters. The goals of this study focused on coal properties. The goals were to determine the forms of C1, the characteristics of C1 evolution, and the ash composition of Illinois and British coals; to compare these parameters in Illinois and British coals; and to establish relationships, if any, between these parameters and the corrosivity of the coal.

The results of C1 analyses indicate that the C1 in coal exists in ionic forms, including small amounts of solid NaC1. When heated under air, Illinois coals released hydrogen chloride gas, with maximum evolution occurring between 425 to 475OC, whereas, for British coals, the maximum evolution occurred between 210 and 280OC. Ash analyses indicate that three of the four British coals contain a higher level of soluble sodium related to soluble potassium. A similar relationship was exhibited for the relative corrosion indices (Ic) calculated according to Borio et al. The I values for the British coals were 53.94, 15.05, 292.92, and 51.42; for the Illinois coals they were 17.04, 17.50, 11.62, 10.90, and 10.55. Caution should be exercised in interpreting corrosion indices reported in this study since they constitute an extrapolation outside their original correlation. This is because the soluble Na,O measured in this study was higher than the maximum values measured by Borio et al.

U.S. DOE Patent Clearance i s not required pr io r t o the publication of t h i s document.

blSTRIBUtlON OF THIS DOCUMENT IS UNLlMmo /ro’

EXECUTIVE SUMMARY

The question of whether or not chlorine in Illinois coals really causes corrosion has had an impact on the marketability of the coal. Literature and use history have suggested that some high-chlorine Illinois coals do not cause boiler corrosion. On the other hand, extensive data developed by the British correlate corrosion with chlorine content, other parameters of coals, and the type of boiler. Although methods such as blending coals of low and high chlorine content have been used in Britain to control corrosion, the presence of chlorine in Illinois coals, because of the extensive British data and other literature, has been a concern for the end users and boiler manufacturers. Market concerns should find relief and the use of Illinois coals should increase, if it can be shown scientifically that Illinois coals do not cause boiler corrosion. Such scientific study should also explain differences in corrosivity between British and Illinois coals. The differences in the corrosivity of coals may be due to properties of the coal, to coal blends, or to parameters of the boilers in which they are burned. This study focused on the effect of coal properties on corrosivity. The goals of this project were to determine the forms of C1, the evolution characteristics of C1, and the ash composition of Illinois and British coals; to compare these characteristics in Illinois and British coals; and to establish the relationship, if any, between these parameters and the corrosivity of the coal.

This was a cooperative effort among the ISGS; RU, Houston, TX; UK, Lexington, KY; and WKU, Bowling Green, KY.

Specific objectives were:

A.

B.

C.

D.

To determine the types of chlorine- and sulfur-containing compounds in coals by an X-ray Absorption Near Edge Spectroscopic (XANES) method and to determine the total chlorine, total sulfur, and inorganic forms of sulfur by ASTM methods.

To examine the chemical composition of coals and to predict the coal corrosion potential based on the ash composition of coals.

To compare, using XANES, the distribution and chemical state or states of chlorine in coals which have been reported to have different corrosion potentials.

To conduct thermogravimetry with Fourier transform infrared (TGA-FTIR) spectrometry and to examine the evolution characteristics of chlorine-containing compounds in coals under pyrolysis and oxidation conditions.

D IS CLAIMER

Portions of this document may be illegible in electronic image products. Images are produced from the best available original document.

E To conduct step-wise pyrolysis while using multi-surface matrix isolation-Fourier transform infrared (MI-FTIR) analyses and to examine the behavior during pyrolysis of heteroatoms, especially C1 compounds, that are present in coal.

F. To establish the relationships, if any, among the forms of chlorine in coals, the evolution characteristics of chlorine in coals, the chemical composition of coals, and the corrosion potential of the coals.

Both destructive temperature-programmed thermogravimetry with fourier transform infrared (TGA-FTIR) and non-destructive X- ray absorption near-edge structure (XANES) techniques were used to examine the forms and the evolution profiles of chlorine-containing compounds in coals. The results from XANES analysis indicate that chlorine in coal exists in ionic forms including small amounts of solid sodium chloride. The H C 1 evolution profiles obtained from oxidation of the four British coal samples indicate that British coals release HC1 gas at temperatures from 150 to 45OOC with the maximum HC1 evolution temperature between 21OOC and 280OC. On the other hand, Illinois coals release HC1 gas at temperatures from 150 to 6OOOC with the maximum HC1 evolution temperature between 41OOC and 450OC.

The results of ash analyses indicate that British coals contain a higher level of soluble sodium with respect to soluble potassium. The soluble Na-to-K ratios calculated as oxides on a dry coal basis (Na20/K20) for the British coals were 9.53, 1 . 8 6 , 56.67, and 9.00, whereas, the ratios for the Illinois coals were 2.20, 2.31, 1.16, 1.02, and 0.92. A similar relationship was exhibited for the relative corrosion indices (I,) calculated according to Borio et al. The I values for the British coals were 53.94 15.05, 292.92, and 51.42; for the Illinois coals, the values were 17.04, 17.50, 11.62, 10.90 and 10.55. However, the soluble Na20 measured in this study were higher than the maximum values measured in the study of Borio et al. and caution should be exercised in interpreting those results because they constitute an extrapolation outside their original correlation.

Overall, the results indicate that the C1 ions in the Illinois coals are more strongly bound than those in the British coals, and suggests that the way in which C1 ions are associated in Illinois coals is different from that in the British coals. This difference, plus the difference in the ratios of the acid-soluble alkali metals, may contribute to the corrosivity differences of the two coal types.

DISCLAIMER

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsi- bility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Refer- ence herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recom- mendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

L

1

GOALS AND OBJECTIVES

The goals of this study were (1) to determine the forms of C1, the evolution characteristics of C1, and the ash composition of Illinois and British coals; ( 2 ) to compare these characteristics in Illinois and British coals; and (3) to establish the relationships, if any, between these characteristics and the corrosivity of the coal.

Specific objectives were

A. To determine the types of chlorine- and sulfur-containing compounds in coals by an X-ray Absorption Near Edge Spectroscopic (XANES) method and to determine the total chlorine, total sulfur, and inorganic forms of sulfur by ASTM methods.

B. To examine the chemical composition of coals and to predict the coal corrosion potential based on the ash composition of coals.

C. To compare, using XANES, the distribution and chemical state or states of chlorine in coals which have been reported to have different corrosion potentials.

D.

E

To conduct thermogravimetry with Fourier transform infrared (TGA-FTIR) spectrometry and to examine the evolution characteristics of chlorine-containing compounds in coals under pyrolysis and oxidation conditions.

To conduct step-wise pyrolysis while using multi-surface matrix isolation-Fourier transform infrared (MI-FTIR) analyses and to examine the behavior during pyrolysis of heteroatoms, especially C 1 compounds, that are present in coal.

F. To establish the relationships, if any, among the forms of chlorine in coals, the evolution characteristics of chlorine in coals, the chemical composition of coals, and the corrosion potential of the coals.

INTRODUCTION AND BACKGROUND

Literature and use-history data have suggested that high-C1 Illinois coal does not cause boiler corrosion (Hamling and Kaegi, 1984, Wright et al, 1994, Abbott et al, 1994) while extensive data have been developed by the British which

2

attribute the rate of corrosion to the C1 content as well as other parameters of coal, such as alkaline content, and coal boiler materials (Reid, 1971, Gibb, 1983, Meadowcroft and manning, 1983, Raask, 1985, Latham et al, 1991, Oakey et al, 1991, and Bakker et al, 1990). Although methods such as blending coals of low and high chlorine content have been used in Britain to control corrosion, the presence of chlorine in Illinois coals, because of the extensive British data and other literature, has been a concern for the end users and boiler manufacturers. Market concerns should find relief and the use of Illinois coals should increase, if it can be shown scientifically that Illinois coals do not cause boiler corrosion. Such scientific study should also explain differences in corrosivity between British and Illinois coals.

A study of the materials for future fossil fuel fired steam generators was conducted by Bakker et a1 (1990). They indicated that superheater corrosion, in general, can be attributed to five factors. These factors are (1) the amount of liquid sulfate deposited, ( 2 ) the sulfur content of the coal, (3) the C1 content of the coal, (4) the temperature of the superheater tube wall, and ( 5 ) the chemical composition of the alloy. They also indicatedthat superheater corrosion was most common, in the U . K . , where it was attributed to the high level of C1 in the British coals. The total C1 content in coal has been used in Great Britain for predicting the corrosivity of a coal. It has been indicated, however, that high-C1 Illinois coals do not create the serious corrosion problems which the British coals have shown during combustion. Since these two studies involve two different coals and two different boilers, it is not clear whether the differences in observed corrosion were due to the properties of the coals, or to boiler parameters. For example, the chemical forms of C1, their behavior during combustion, and their interaction with sulfur and other components of coal may play roles in the corrosivity of a coal with high C1 content. Also, boiler-tube corrosion may be affected by the temperatures of the tubes, the materials from which the boiler tubes were constructed, properties of the coal, and whether a coal that produced corrosion was blended with one that did not.

This study focused on the fundamental effects of C1 and other elements in coals with different corrosion potentials.

3

EXPERIMENTAL

Task 1. Prepare whole coal samples and standard samples (ISGS).

C1-containing

Five Illinois coals and four British coals were collected for this project. Five pounds of coal, ground to -200 mesh (C- 33068) was obtained from Dr. Anthony G. Fonseca of the Consolidated Coal Company. Four additional Illinois Basin coals (C-33873, C-33874, C33875, and C-33876) were obtained from Dr. I. Demir. These 60-mesh sample were approximately 300 g each. All samples were split to amounts needed for analysis and preserved in the facilities of the Coal Sample Bank Program.

Four high-C1 coals were obtained from Great Britain. Selection was based on availability and upon the most frequently used and studied coals. Sampling was accomplished with the assistance of the Coal Research Establishment Technical Services in Great Britain. Two coal sample (C-33499 and C-33500) were obtained from mine. The other two samples (C-33501 and C-33502) were purchased from the CRE Technical Service Coal Bank. These samples were ground to 60 mesh, split to sizes needed for analysis and preserved in the facilities of the Coal Sample Bank Program.

Task 2. Determine the forms of C1 and sulfur in coals by XANES and total C1, total sulfur, and inorganic forms of sulfur by ASTM analyses. (ISGS/UK)

Total C1 in coal was determined with a Leco model C1-350 chlorine analyzer. Carbon, hydrogen, and nitrogen were determined with a Leco model CHN-600 carbon, hydrogen, nitrogen analyzer. Proximate analysis, total sulfur, and inorganic forms of sulfur were completed using ASTM procedures (D-3172, D-4239, and D-2492, ASTM 1991). The forms of C1 was determined by XANES analysis.

Task 3. Conduct a complete chemical analysis and examine mineral matter, major, minor, and trace elements in coals with different corrosion potential ( I S G S ) .

Ashing of coals and determination of coal ash composition. Approximately 2 grams of coal was dried at llO°C for two hours to determine moisture content. Then, a separate coal sample was ashed at 75OOC to determine the ash content. About 0 . 6 grams of coal ash was then blended with about 5.4 grams of 50% lithium metaborate/50% lithium tetraborate flux in a 95% Pt-5% Au crucible and fused in a furnace at 1000°C for 15 minutes.

4

This was followed by a short-cycle fusion in a Claisse Fluxer- Bis using a propane burner. After fusion, the fluxer automatically poured the molten mixture into a 30-mm diameter 95% Pt-5% Au mold in which a glass disk specimen was formed upon cooling. The specimen was analyzed using a Rigaku 3371 WDXRF spectrometer with an end-window rhodium X-ray tube. Concentrations of components in the coal ash were calculated by the computer which control the spectrometer, and which used calibration curves based on natural and artificial standards plus Lachance-Trail matrix correction coefficients.

Determination of acid-soluble sodium and potassium in coal. Two grams of the coal sample, weighed to the nearest milligram, was placed in a 250-ml Erlenmeyer flask which contained 40 ml of 5% HC1 solution. The flask was then fitted with an air-cooled condenser and the sample was digested for 16 hours at a gentle boil. The sample was then cooled, filtered through Whatman #40 filter paper, and washed at least four times with small amounts of distilled water. Next, the filtrate was diluted to 100 ml in a volumetric flask. Sodium and potassium were determined using an atomic absorption flame spectrophotometer.

Task 4. Compare, using XANES, the distribution and chemical state or states of C1 in coals which have been reported to behave differently with respect to corrosion problems during combustion (ISGS/UK).

The XANES results obtained in Task 2 were examined, and the results from the Illinois and British coals were compared.

Task 5. Analyze the evolution profile of C1-containing compounds in coals under pyrolysis and oxidation conditions (ISGS/WKU) .

Temperature-programmed thermogravimetry, in conjunction with Fourier transform infrared spectrometer (TGA-FTIR), was used to obtain the characteristics of the thermal evolution of HC1 gas from the coals under both oxidation and pyrolysis conditions. Simultaneous TGA/FTIR techniques were based on a combined analytical method in which thermogravimetry (TGA) was used in conjunction with Fourier transform infrared spectroscopy (FTIR) . By using the combined TGA/FTIR instrumental system, the volatile species produced in a Dupont-951 TGA during pyrolysis or combustion of a coal sample were analyzed by a Perkin Elmer-1650 FTIR spectrometer. The system was able to continuously measure the mass change of the coal sample with temperature increasing at a heating rate of 10°C/min. The system was also able to identify qualitatively and determine quantitatively the individual

6

RESULTS AND DISCUSSION

CHEMICAL ANALYSES

The results on forms of sulfur analyses for two Illinois and three British coal samples by XANES method are shown in Table 1. The data from the proximate, ultimate, and sulfur forms for all nine coals by ASTM method are listed in Table 2. The C1 content and calorific value are also presented. The ash composition and acid-soluble alkali metals analyses on four British and five Illinois coals are also listed in Table 2.

Table 1: Forms of sulfur (wt%) in coals by XANES analysis.

British coals Illinois coals

Forms C-33499 (2-33500 C-33502 of Sulfur

Pyritic S 1.00

Sulfide 0.24

Thioph. S 0.75

Sulfoxide ND

Sulfate 0.04

Sulfone ND

0.38

0.14

0.72

ND

0.01

ND

0.09

0.09

0.73

ND

0.01

ND

0.71

1.28

2.03

0.09

0.09

ND

0.84

0.46

1.22

0.03

0.03

Trace

ND: Not detectable

7

Table 2: Chemical analysis of coal and coal ash and calculations of corrosion potential

c-33499 C-33500 C-33501 C-33502 C-33068 C-33873 C-33874 C-33875 C-33876

Proximate and other standard analysis of coal samples

Mois 4.55 4.16 8.84 3.21 3.52 14.97 7.86 9.31 6.61 Vol 32.61 29.10 38.10 35.54 35.17 40.19 38.50 31.89 36.06 Fx C 52.61 48.66 58.68 61.33 59.37 50.23 50.75 52.27 55.98 Ash 14.79 22.24 3.23 3.13 5.46 9.58 10.74 15.83 7.96

1

H 4.64 4.15 4.90 5.09 4.97 4.96 4.83 4.59 5.01 C 70.07 64.56 79.57 82.36 77.70 70.16 71.37 68.60 74.49 N 1.51 1.43 1.62 1.75 1.71 1.26 1.24 1.38 1.47 0 6.96 6.36 9.81 6.76 9.20 9.63 8.17 8.60 8.12 Tots 2.04 1.26 0.86 0.91 0.97 4.42 3.66 1.00 2.90 OrgS 1.13 0.82 0.71 0.66 0.58 2.96 2.29 0.53 1.76 sus 0.05 0.03 0.04 0.02 0.03 0.12 0.14 0.10 0.14 PYRS 0.86 0.41 0.11 0.23 0.36 1.33 1.24 0.42 0.99 TCL 0.95 0.61 1.54 1.15 0.54 0.22 0.12 0.44 0.45 BTU 12466 11319 13893 14383 13603 12653 12685 12120 13238

Major oxides in coal ash (dry coal basis)

Si02 7.58 12.14 0.61 1.03 2.79 4.42 4.74 9.52 4.05 A2°3 3.65 5.97 0.50 0.79 1.26 1.76 1.89 4.03 1.74 Fe203 1.74 1.50 0.31 0.86 0.60 2.16 2.05 1.00 1.61

MgO 0.22 0.40 0.06 0.08 0.06 0.10 0.09 0.19 0.08

Na20 0.31 0.31 0.34 0.10 0.12 0.18 0.06 0.17 0.07 TiOz 0.14 0.21 0.03 0.02 0.07 0.01 0.09 0.20 0.09

G O 0.59 0.40 0.68 0.31 0.12 0.31 0.43 0.37 0.21

K2O 0.48 0.89 <0.01 0.02 0.12 0.21 0.19 0.43 0.20

PZOS 0.02 0.05 0.07 0.04 0.01 0.01 0.01 0.02 0.01 MnO 0.01 0.02 <0.01 0.01 <0.01 <0.01 <0.01 <0.01 <0.01 so3 0.48 0.25 0.75 0.10 co.01 0.11 0.28 0.29 0.08

Acid soluble alkali metals in coal (dry coal basis) and coal relative corrosion index

(NazO),, 0.162 0.197 0.340 0.072 0.112 0.180 0.050 0.130 0.062 (KzO),, 0.017 0.106 0.006 0.008 0.051 0.078 0.043 0.128 0.067 NaK (mole) 14.45 2.82 85.91 13.64 33.3 3.50 1.76 1.54 1.40 Na2O/K2O(as) 9.53 1.86 56.61 9.00 2.20 2.31 1.16 1.02 0.92 IC 53.94 15.05 292.92 51.42 17.04 17.50 11.62 10.90 10.55 as: acid-soluble

a FORMS OF CHLORINE IN COALS

- by X-ray Absorption Near-Edge Structure (XANES) The XANES spectra of the British and Illinois coals are shown in Figure 1. Based on the XANES spectra of standard compounds (Huggins and Huffman, 19941, a strong broad peak at about 1 eV followed by a second, much weaker, broad peak at about 20 ev is indicative of an ionic form of C1. XANES analysis is presently not able to differentiate between the types of positive counter ions that are associated with ionic chloride. However, the appearance of a sharp peak at 12.5 eV is indicative of solid sodium chloride. As seen in these XANES spectra, all of the coal samples contained ionic C1. A small peak shown at 12.5 eV is indicative of the appearance of solid sodium chloride. Sodium chloride was observed in only two of the British coals and none of the Illinois coals.

', British Coal

11-33499 i+ (1.95 ut% c1

c

l-l

i! C-33502

1.15 u t X Cl

s 1-

0 J I -20 20 4'0 f

Energy. eV

Illinois Coal

3

I in C-33876

-u 0 40

Energy. eV I

Figure 1: XANES spectra of three British coals and three Illinois coals

9

ASH COMPOSITION OF COALS

The study by Borio et al. (1969) showed that alkali metals, alkaline earths, iron, and sulfur were the most significant constituents relative to high temperature corrosion (Borio et al, 1969 and Meadowcroft, 1989). Because alkali metals generally comprise small portion of the coal, they are usually the most sensitive corrosion indicators. Alkali metal concentration, measured by an acid leaching process, was also found to give a better estimation of alkali metal concentration for reaction in forming molten alkali-iron- trisulfate, a corrosive complex. Certain relationships were developed by Borio et a1 (1969), in which the amount of acid soluble sodium and potassium (calculated as oxides) in the coal and the amount of calcium oxide and magnesium oxide obtained as ash in coal were used to calculate the relative corrosion indices of coals (Eq. 1).

Na, 0 I, = 5 .96 + 5.07- - 0.42(CaO + MgO) (Borio et a l . , 1969)

K20

In this study, the ratio on the amount of acid-soluble sodium to the amount of acid-soluble potassium, (Na,O/K,O) ,,, as well as the relative corrosion indices, I,, were obtained according to the relationships developed by Borio et a1 (1969). The results (Table 2) indicate that, with an exception of one coal ( C - 3 3 5 0 0 ) , the weight ratios of (NaZO/K20),, are higher for the British coals than for the Illinois coals. A similar relationship is indicated for the relative corrosion indices (I ) . The 1,s are higher for those British coals than for the Illinois coals a

When applying current analysis data to the formula established by Borio et a1 (1969) for relative corrosion index, one should be very careful about interpreting the resulting values. This is because none of the coals analyzed by Borio et a1 (1969) had soluble sodium oxide in amounts greater than 0.1% on the dry coal basis, whereas, as indicated in Table 2, six of the nine coals analyzed in this investigation had soluble sodium oxide greater than 0.1%. The absolute values of I, for the four coals should not be used in grading coal corrosion potential according to the framework defined by Borio et a1 (1969). The values, however, may give a relative measure on boiler corrosion that could derive from alkali metals and alkali-earth metals in coal.

10

THE EVOLUTION CHARACTERISTICS OF CHLORINE IN COALS - by Temperature-Programmed Thermogravimetrywith Fourier Transform Infrared (TGA-FTIR)

HC1 evolution profiles during combustion-The HC1 evolution profiles for the British coals and four Illinois coals determined by TGA-FTIR are presented in Figure 2. The results indicate that the coals release their C1 as HC1 gas. British coals release this gas beginning at 15OOC and ending at approximately 450OC. The maximum HC1 evolution temperature occurs in a range from 21OOC to 280OC. Illinois coals release HC1 gas beginning at 15OOC and ending at approximately 6 0 0 O C . The maximum H C 1 evolution temperature occurs in a range from 410 to 450OC.

0 0 Temperature (C )

--0-.C-33499 --+-4!-33500 +C-33873 e C-33068 --A-- C-33502 - -X--C-33501 *C-33875 *C-33876

Figure 2: The HC1 evolution profiles for British Coals and Illinois coals during combustion.

A comparison of the HC1 evolution profiles during combustion of Illinois and British coals indicates that the British coal release gas at lower temperature than Illinois coal. This is true for all Illinois and all British coals based on analyses of nine samples. This difference may be due to the differences in forms of C1 in coals or the differences in pore size distribution of the coals, which needs further investigation.

Carbon dioxide evolution profiles during combustion-The carbon dioxide evolution profiles, as determined by TGA-FTIR, for the three British coals are shown in Figure 3, and for three of the Illinois coals are shown in Figure 4. The profiles for British and Illinois coals are very similar, and all coals show a maximumn rate of CO, release around 41OOC. The data

11

Temperature [C ]

-C-33499 +C-33501 + C-33502 Figure 3: CO, evolution profiles of three British coals.

6 0

4 0 M D

a D 4

b7 A 4

30

: 20

10

200 400 600 800 1000 Temperature (C)

+C-33068 +C-33873 +C-33876

Figure 4: CO, evolution profiles of three Illinois coals.

indicate that C1 , for British coals, is released when CO, evolution is low. In other words, C 1 in the British coals is released before the coal is combusted. C1 in the Illinois coals is released after the C02 maximum and in the range of high CO, evolution. In other words, C 1 in the

On the other hand,

12

1 nn

BU-

6U-

40-

7n-

Illinois coals is released during coal combustion.

F i r s t Dcrivativo

-_.- _I..

Weight loss profile during combustion-TGA was used to determine the mass chanqes of coal samples durinq temperature - programmed oxidation. B r i t i s h a n d Illinois coals show a similarity of three-stage mass variations. A typical weight-loss profile is shown in in Figure 5. At the initial stage of heating, from ambient to 15OoC, the sample weight i s r e d u c e d slightly. This is attributed to the loss of moisture. Further heating from 150 to 3OO0C,

The results indicate that both

n j I I I 0 200 400 6aa SA0

resulting in a Figure 5 : The TGA curve and its first slight increase in derivative for a typical coal. sample weight (1.5 - 2.2 % ) . This weight gain may be attributed to oxygen being absorbed during mild oxidation of the coal. The sample weight is substantially reduced (80 to 95%) by further heating from 3 0 0 to 6OOOC. At this stage, an extensive degradation and combustion of the coal sample have taken place.

When heated under air, both Illinois and British coals exhibited similar burning characteristics. They all exhibited maximum rate of mass loss at temperatures around 470OC. This is also the temperature at which CO, has its maximum evolution. A comparison between these mass loss profiles and those previously described for CO, and for C 1 evolution shows an interesting phenomenon. The results indicate that C 1 in the British coals is released before optimum volatile matter evolution and coal combustion, whereas, C1 in the Illinois coals is released during optimum volatile matter evolution and coal combustion.

Profiles of HC1 evolution during pyrolysis-Figure 6 shows the HC1 evolution profile determined by TGA-FTIR for both Illinois and British coals during pyrolysis. In general, the evolution of HC1 gas occurs at the temperature from 3 0 0 to 41OOC. These data indicate that the temperature of maximum release of chlorine was lower for the British coals than for the Illinois

13

coals. However, the difference in the evolution of chlorine observed under pyrolysis condition is not as obvious as those observed under oxidation condition. This may suggest that Illinois coals are more easily oxidized, and C 1 in the oxidized Illinois coals is more strongly bound than C1 in the less oxidized British coals.

12 3 =4Gx e

i Y 10 - # %

w 8 - I P

Z 6 -

2 4 -

Y

M w d* +#--I '& "J', ',

.3k c1 i d J I \ I

bl u

k 0 II)

2 -

0 I I I I I I I I 150 200 250 300 350 400 4 5 0 500 550 600

Temperature (C)

- -e - -C-33499 --+--C-33501 -W-C-33068 --0-C-33875

Figure 6: H C 1 evolution profiles of British and Illinois coals during pyrolysis by TGA-FTIR.

EVOLUTION OF CHLORINE AND VOLATILES FROM COALS - by Step-Wise Pyrolysis with Multi-Surface Matrix Isolation-Fourier Transform Infrared

Profiles of volatiles evolution from coals during pyrolysis- The profiles of volatiles evolution from one Illinois coal and one British coal under nitrogen were examined by step-wise pyrolysis withmulti-surfacematrix isolation-Fouriertransform infrared spectrometry. The results are shown in Figures 7 and Figure 8, respectively. The three-dimensional graphs indicate that Illinois coal produced more tar (2700 to 3000 cm-I) than British coal.

Profiles of HC1 evolution from coals during pyrolysis-The data collected by step-wise pyrolysis with multi-surface matrix

14

4 Wavenumber md

Figure 7 : The 3-D graph of volatiles evolved during step- wise pyrolysis of an Illinois coal (C-33875).

. - k - -

J

4 d 0 0

Figure

3500

8 :

I I

* -387-425 -345-387 .

-3 0 5- 3 4 5 4270-305

a d 2 4 0 -270 - 1207-240

F I . 4162-207 3000 2 5 0 0 2000 l5bO 1000 500

Wavenumber cm-1 The 3-D graph of volatiles evolved during wise pyrolysis of a British coal (C-33499

step- ) .

isolation-Fouriertransforminfrared spectrometrywere further examined for HC1 evolution. The HC1 evolution profiles of the 11 linois

15

/

348-407 307-348 288-307 239-288

2820 2805 Wavenumber cm-1

Figure 9 : The HC1 evolution (left) and one Bri wise pyrolysis.

prof .tish

345-387 305-345 270-305

2 8-2 Q 2805 Wavenumber c m l

iles of one Illinois coal coal (right) during step

coal and the British coal are shown in Figure 9. The profiles indicate that most of the HC1 gas evolves at the temperatures from 3 0 0 to 4OOOC for the Illinois coal and from 240° to 636OC for the British coal. It is difficulty to make a direct comparation between these results and those obtained from the TGA-FTIR data for evolution of chlorine from the coal under pyrolysis conditions. The later (Fig. 7 ) showed a slight difference in the temperature of maximum release of C1 from the Illinois coals and British coals. The temperature of maximum release of chlorine was a slightly lower for the British coals than for the Illinois coals.

SUMMARY AND CONCLUSIONS

At elevated temperatures and under oxidizing conditions, both British and Illinois coals released hydrogen chloride gas. The Illinois coals, however, released the gas at higher temperatures, with maximum evolution around 425 to 475OC. The British coals released HC1 gas at lower temperature with maximum evolution temperatures occurring between 210 and 28OOC. C1 exists in ionic form in the coals from both countries. Three of the British coal samples contained higher concentrations of soluble sodium in comparison to soluble potassium than the Illinois coal samples. The ratio of soluble sodium to soluble potassium, calculated as oxides (NazO-to-K20) for the British coals were 9 . 5 3 , 1.86, 5 6 . 6 7 , and

1 6

9 . 0 , whereas, the ratios for the Illinois coals were 2.20, 1 . 1 6 , 1.02 and 0 . 9 2 . A similar relationship was indicated for the relative corrosion indices (I,) as defined by Borio et a1 ( 1 9 6 9 ) . The 1,s for the British coals were 5 3 . 9 4 , 1 5 . 0 5 , 2 9 2 . 9 2 , and 5 1 . 4 2 ; for the Illinois coals, the values were 1 7 . 0 4 , 1 7 . 5 0 , 1 1 . 6 2 , 1 0 . 9 0 and 1 0 . 5 5 .

The difference in the HC1 evolution temperature between the Illinois and British coals suggests that the way in which C1 ions are associated in Illinois coals is different from that in the British coals. This difference, plus the difference in the amounts of alkali metals, may contribute to the corrosivity differences of the two coal types. Further study is recommended to examine if, or how, the interaction of C1 with other species, such as mineral matter, in the coal could affect the corrosion potential of a coal.

REFERENCES

Abbott, M.F., W.C. Corder, J.A. Campbell, and E.P. Doane, 1 9 9 4 Chlorine in coal: Utility experience burning Illinois coals, American Boiler Manufacture Association Clean Power For the g o ' s , Arlington VA, May 11-13 .

American Society for Testing Materials, 1 9 9 1 , Annual Book of ASTM Standards, Volume 5.

Bakker, W. T., J. L. Blough, and S. Kihara, 1 9 9 0 , Materials for future fossil fuel fired steam generators, Proceedings of the American Power Conference in Chicago, IL, 1 9 9 0 .

Borio, R.W. , R.P. Hensel, R. C. Ulmer, H.A. Grabowski, E.B. Wilson and J. W. Leonard, 1 9 6 9 , The control of high- temperature fire-side corrosion in utility coal-fired boilers, Office of Coal Research R&D Report No. 4 1 , April.

Brinsmead, K. H., and R. W. Kear, 1 9 5 6 , Behavior of sodium chloride during the combustion of carbon, Fuel, V o l . 35, 1 9 5 6 , p . 8 4 - 9 3

Gibb, W.H., 1 9 8 3 , The nature of chlorine in coal and its behavior during combustion, D.B. in Meadowcroft and M.I. Manning (editors), Corrosion Resistant Materials for Coal Conversion Systems, Applied Science Publishers, New York, p. 2 5 - 4 6 .

Hamling , J., and D . Kaegi, 1 9 8 4 , Characterization and removal of chlorides in Illinois coal, Proceedings, Annual Pittsburgh Coal Conference (DE85012243) September.

17

Huggins, F.E. and G.P. Huffman, 1995, Chlorine in coal: An XAFS spectroscopic investigation, Accepted for publication in Fuel

Latham, E, D.B. Meadowcroft, and L. Pinder, 1991, The effects of chlorine on fireside corrosion, Coal Science and Technology 17: J. Stringer and D.D. Banerjee (eds.): Elsevier: EPRI, p. 225.

Meadowcroft, D.B., 1989, Corrosion control in power plants, in Surface Stability, T.N. Rhys-Jones (ed.) , The Institute of Metals, London, 1989, p.52-94.

Meadowcroft, D.B. and M.I. Manning (editors), 1983, Corrosion resistant mterials for coal cnversion sstems, Applied Science Publishers, New York.

Raask, E., 1985, Mineral impurities in cal cmbustion: Behavior, problems and remedial measures, Hemisphere Publishing Corporation, New York.

Reid, W.T., 1971, External corrosion and deposits: boiler and gas turbines: Elsevier: New York.

Oakey, J.E., A. J. Minchener, and N. J. Hodges, 1991, The use of high chlorine coals in industrial boilers, Coal Science and Technology 17: J. Stringer and D.D. Banerjee (eds. ) : Elsevier: EPRI, p. 251.

Wright, I.G., K.K. Ho, and A.K. Mehta, 1994, Survey of the effects of coal chlorine on fireside corrosion in pulverized coal-fired boilers, in Proceeding of the EPRI Symposium "The Effects of Coal Quality on Power Plants: Fourth International Conference" August 17-19, Charleston, SC.

PROJECT MANAGEMENT REPORT June 1, 1994 through August 31, 1994

Project Title: CHLORINE IN COAL AND ITS RELATIONSHIP WITH BOILER CORROSION

DOE Grant Number: DE-FC22-92PC92521 ICCI Prin

Other Investigators:

Pro] ect Manager:

Project No. : 93-1/6.1B-lP .cipal Investigator: M.-I. M. Chou, (217) 244-0312,

Illinois State Geological Survey (ISGS), 615 E. Peabody Drive, Champaign, IL 61820 J.M. Lytle, R.R. Ruch, and C.-L. Chou, (ISGS) ; J. L. Margrave and V Khabashesku, Rice University (RU) ; G.P. Huffman and F.E. Huggins, University of Kentucky (UK); W.P. Pan and L. Liu, Western Kentucky University (WKU) . Ken K. Ho, Illinois Clean Coal Institute

2

SCHEDULE OF PROJECT MILESTONES . X

S 0 N D J F M A M J J A

Mi 1 estones:

A. B. C . D. E.

F. G. H.

Coal sample preparation (Task 1) XANES and ASTM analyses of coal samples (Task 2) A complete chemical analysis of coal samples (Task 3) XANES data interpretation (Task 4) Characterization of the evolution profiles of C1-containing compounds in coal (Task 5) Step-wise pyrolysis MI-FTIR analysis Data interpretation (Task 7) Technical project management reports prepared and submitted

(Task 6)

P r o j e c t T i t l e :

EQUIPMENT LIST September 1, 1993 through August 31, 1994

CHLORINE I N COAL AND I T S RELATIONSHIP WITH BOILER CORROSION

Other Inves t i ga to rs :

DOE Grant Number: DE-FC22-92PC9252 1 (Year 2) I C C I P r o j e c t Number: 93-1/6.1B-lP P r i n c i p a l I nves t i ga to r : M.-I.M. Chou, (217) 244-0312, I l l i n o i s S ta te

Geological Survey (ISGS) , 615 E. Peabody Drive, Champaign, I L 61820 J.M. Ly t l e , R.R. Ruch, and C.-L. Chou, ( ISGS); J . L. Margrave and V . Khabashes ku, R i ce Uni ve rs i t y (RU); G.P. Huffrnan and F.E. Huggins, U n i v e r s i t y o f Kentucky (UK); W.P. Pan and L. Liu, Western Kentucky U n i v e r s i t y (WKU) .

Pro j e c t Manager: Ken K. Ho, I l l i n o i s Clean Coal I n s t i t u t e

No new equipment was purchased under t h i s p r o j e c t .

PUBLICATION LIST September 1, 1993 through August 31, 1994

P r o j e c t T i t l e : CHLORINE I N COAL AND I T S RELATIONSHIP WITH BOILER CORROSION

DOE Grant Number: DE-FC22-92PC92521 (Year 2) I C C I P r o j e c t Number: 93-1/6.1B-lP P r i n c i p a l I n v e s t i g a t o r : M.-I.M. Chou, (217) 244-0312, I l l i n o i s State

Geological Survey (ISGS), 615 E. Peabody Drive, Champaign, I L 61820

Other Inves t iga tors : J.M. Ly t le , R.R. Ruch, and C.-L. Chou, (ISGS); J.L. Margrave and V. Khabashesku, Rice U n i v e r s i t y (RU); G.P. Huffman and F.E. Huggins, U n i v e r s i t y o f Kentucky (UK); W.P. Pan and L. L iu, Western Kentucky U n i v e r s i t y (WKU) .

P r o j e c t Manager: Ken K. Ho, I l l i n o i s Clean Coal I n s t i t u t e

Chou, M.-I. , J.M. Ly t le , and K.K. Ho. 1994. "Ch lor ine and Ash Composition i n Two I l l i n o i s Coals and Two B r i t i s h Coals." Proceeding paper presented t o the Second I n t e r n a t i o n a l Conference on "The E f f e c t o f Coal Q u a l i t y on Coal Power Plants" EPRI i n August 17- 19, 1994.

Chou, M.-I. , W.P. Pan, F.E. Huggins, J.M. Ly t le , L. L iu, and K.K. Ho. 1994. "Chlor ine i n Coal and B o i l e r Corrosion." Paper t o be presented t o the P i t tsburgh Coal Conference i n Septemberl2-16, 1994.

Chou, M.I.M., J.M. L y t l e , R.R. Ruch, C.-L. Chou, J.L. Margrave, C.J.Chu, B.H. Hauge, G.P. Huffman, F.E. Huggins, W.P. Pan, and L. L iu . 1994. "Chlor ine i n Coal and i t s Re la t ionsh ip w i t h B o i l e r Corrosion," Abst ract and Poster Paper, The I 1 1 i n o i s Coal Research Board c o n t r a c t o r ' s techn ica l meeting, Champaign, IL, August 2-4, 1994.

HAZARDOUS SUBSTANCE PLAN

Sol i c i t a t i on No. : ICCI/RFP93-1 Program Area: Coal Characterization Proposer: I1 1 inoi s C1 ean Coal Ins t i tu te Principal Investigator: Mei-In M . Chou Project Ti t le : Chlorine in Coal and Its Relationship With Boiler Corrosi on Project Duration: September 1, 1993 - August 31, 1994 Total Estimated Cost: $60,591

No sol id and l iquid waste were produced by t h i s project.

Hazardous Waste Transporter: The Office of Environmental Safety & Health, University of I l l i no i s , would make necessary determination and dispose of the waste in the manner approved by the IL EPA.

Office of Health & Safety, University of I l l i no i s : 101 S. Gregory, Urbana, I1 1 inois 61801 (Van Anderson) (217)244-9278

Hazardous Waste Disposal Faci 1 i t y , Contractor, and Location: All col lect ion, treatment and disposal, i f necessary, was accordance with a1 1 RCRA and TSCA rul es/regul a t i ons. Incineration or treatment was performed a t cer t i f ied f a c i l i t i e s . The current University of I l l i no i s contract (annually bid) i s with Rollins, Inc., Which has f a c i l i t i e s in Baton Rouge, LA and Deer Park, TX.

Hazardous Waste Treatment Method: Incineration or treatment.