Back to Publications

Gossman Consulting, Inc.

METAL PRECOMPLIANCE FOR CEMENT KILNS

Gossman Consulting, Inc.

Jim Woodford Ronald Gossman David Gossman
Operations Consultant Senior Project Manager President
66 Sexton's Drive 35 Meigs Drive  45W962 Plank Road
Xenia, OH 45385 Shalimar, FL 32579 Hampshire, IL 60140

Presented at the AWMA International Specialty Conference on Waste Combustion in Boilers and Industrial Furnaces March, 1992

ABSTRACT

In February 1991, the US EPA promulgated the Burning of Hazardous Waste in Boilers and Industrial Furnaces (BIF) regulations. Whereas cement kilns using waste solvents as an alternative fuel source have previously been regulated by air permits, they are now also governed under the Resource Conservation and Recovery Act (RCRA).

These additional regulations require that specific metals going into the cement kiln (industrial furnace) be monitored so that metal emission limits will not be exceeded. The specific metals are antimony, arsenic, barium, beryllium, cadmium, chromium, lead, silver, mercury and thallium. Of particular interest are the metals arsenic, cadmium and chromium, for which the most stringent limits have been set.

Metals enter the cement manufacturing process in three ways; raw materials, primary fuels (i.e. coal) and waste fuels. The necessary precompliance testing and certification is examined with respect to available means for providing best engineering judgment of emission levels. The approach used will allow maximum operational flexibility while still allowing compliance with the BIF regulations.

INTRODUCTION

Antimony, arsenic, barium, beryllium, cadmium, chromium, lead, silver, mercury and thallium are ten metals that must be maintained in compliance with BIF regulations as of August 21, 1991. An operational approach to compliance results in realistic metals concentrations which will not restrict operations unrealistically. This paper explains how an operational approach to precompliance also benefits compliance testing.

BIF PRECOMPLIANCE

Lone Star Industries, Inc. (LSI) desired to get a jump on the complexities of compliance with BIF regulations. LSI, located near Greencastle, IN, operates a 5.1 million BTUs/hr coal fired industrial furnace (cement kiln) with the capacity to produce 90 tons of clinker/hr. It is a wet process kiln with a five stage Electro-static Precipitator (ESP). Gossman Consulting, Inc. was contracted by LSI to perform advance work culminating in the filing of their Precompliance Certification.

Gossman Consulting, Inc. approaches BIF metals compliance from an operational perspective. To keep operational aspects at the forefront of any precompliance considerations, it is important to get a handle on metals concentrations from all fuel sources and raw materials.

Samples of the three cement kiln feed streams were collected and analyzed for concentrations of the BIF metals: antimony (Sb), arsenic (As), barium (Ba), beryllium (Be), cadmium (Cd), chromium (Cr), lead (Pb), mercury (Hg), silver (Ag) and thallium (Tl). This sampling was performed in accordance with a sampling plan designed to eliminate potential problems in three critical sample taking areas; anomalies, contamination, and identification.

  Potential Problem Solution
sample anomalies: abnormally high or low analytical results composite several representative samples
sample contamination: unreliable analytical results proper sampling techniques and locations
sample identification: misidentification of samples logical alpha-numeric labels and chain of custody forms prepared in advance

Adherence to the sampling plan resulted in reliable analytical data. These results are presented in Table I for coal, Table II for raw materials (slurry) and Table III for hazardous waste fuel (HWF). This analysis provided metals data which accounted for all the feed streams going into the cement kiln for the cement manufacturing process.

From this data, distribution is determined (See Figure 1). The key is to develop a realistic picture of the concentrations of metals going into the cement kiln and the specific sources of those metals. It would be ideal to minimize compliance sampling and testing on a long term basis. Toward achieving those desired goals, a conservative starting point is helpful. A conservative starting point for each metal is determined by calculating the mean concentration, and then adding three standard deviations. (See Tables I, II and III) EPA guidance provides the reasoning for using three standard deviations. In effect, this reduces the chances of exceeding those conservative levels to <1 chance out of 1000. In a number of instances, even higher values were chosen either to provide easy analytical detection limits or to account for non-normal distribution.

Precompliance with an eye on Compliance Testing

Although metals precompliance is the subject of this paper, the bottom line is the ultimate need for realistic metals concentrations under normal kiln operating conditions. Actual operating conditions will be examined during the compliance test and precompliance numbers will be adjusted accordingly. So it is important to keep compliance testing conditions in mind throughout this process.

BIF metals will require sufficient monitoring in the fuels and raw materials to insure that the precompliance emission limits will not be exceeded. Compliance test conditions become the maximum operating limits for the kiln. Consequently, optimal test conditions would reflect desired maximum waste fuels substitutions and metals input.

An EPA approved Industrial Source Complex, Short Term (ISC/ST) air dispersion model was used to determine the maximum annual average off-site ground level concentration (MEI), consistent with precompliance requirements. A maximum allowable emission rate was then determined by using the MEI value in conjunction with maximum allowable reference air concentrations (RACs), provided in BIF Appendices IV and V.(1)

Table I. Coal Concentrations (ppm)

Element

CL1

CL2

CL3

CL4

CL5

CL6

CL7

CL8

CL9

CL10

CL11

CL12

CL13

CL14

CL15

CL16

Silver

<0.25

<0.25

<0.25

<0.25

<0.25

<0.25

<0.25

<0.25

<0.25

0.453

<0.25

1.34

<0.25

1.25

<0.25

<0.25

Arsenic

2.73

8.63

9.57

9.38

9.43

6.89

6.57

8

1.99

9.69

12.2

1.62

4.37

1.81

3.35

4.83

Barium

75.3

75

27.6

68.2

79.1

77.3

70.4

77.7

56.8

54.1

59.1

51.6

57.7

233

56.8

73.2

Beryllium

1.08

2.32

2.97

2.67

2.62

2.45

2.26

2.45

0.736

2.44

2.64

1.63

2.52

12

2.68

2.26

Cadmium

<0.25

<0.25

0.265

0.567

0.796

0.802

0.438

1.1

0.307

0.34

0.556

0.428

0.396

1.68

0.58

<0.25

Chromium

9.15

24

24.5

19.4

24.3

22.6

22.9

25.5

7.11

14.1

17

8.45

20.8

94.3

19.9

22.4

Mercury

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

Lead

10.1

23.1

12.8

28.2

35.7

20.9

18.3

28

11.5

22.9

28.5

26.2

16.3

20.7

17.6

20.6

Antimony

<2.8

<2.8

<2.8

<2.8

<2.8

<2.8

<2.8

<2.8

<2.8

<2.8

<2.8

<1.4

<1.4

2.39

1.77

2.27

Thallium

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

0.763

0.785

Element

Average

Standard Deviation

Average +3SD

Maximum

Count

Silver

0.393

0.345

1.427

1.34

16

Arsenic

6.316

3.288

16.18

12.2

16

Barium

74.56

42.94

203.4

233

16

Beryllium

2.858

2.431

10.15

12

16

Cadmium

0.563

0.371

1.677

1.68

16

Chromium

23.53

19.19

81.08

94.3

16

Mercury

<0.1

0

<0.1

<0.1

16

Lead

21.34

6.726

41.52

35.7

16

Antimony

2.502

0.5

4.003

2.8

16

Thallium

0.841

0.025

0.917

0.85

16

Table II. Raw Materials Concentrations (ppm)

Element

RM1

RM2

RM3

RM4

RM5

RM6

RM7

RM8

RM9

RM10

RM11

RM12

RM13

RM14

RM15

RM16

Silver

<0.25

<0.25

<0.25

<0.25

<0.25

<0.25

<0.25

<0.25

<0.25

0.453

<0.25

1.34

<0.25

1.25

<0.25

<0.25

Arsenic

<0.85

<0.85

<0.85

<0.85

<0.85

6.89

0.987

0.871

<0.85

1.32

1.21

<0.85

1.04

1.01

1.07

1.46

Barium

78.8

49.1

77.5

76.6

77.7

67.3

70.8

75.3

46.7

61.4

59.2

71.7

69

63.3

90.6

81.5

Beryllium

<0.15

<0.15

<0.15

<0.15

<0.15

<0.15

<0.15

<0.15

<0.15

0.173

0.163

0.379

0.206

0.404

0.508

0.21

Cadmium

<0.25

<0.25

0.25

0.476

0.399

0.668

<0.25

<0.25

<0.25

<0.25

<0.25

<0.25

<0.25

1.25

<0.25

<0.25

Chromium

33.9

21.3

31.4

33.4

39

31.7

34.1

35.3

20.3

27.1

24.6

33.1

25.4

29.3

34.6

29.6

Mercury

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

<0.1

Lead

2.98

2.29

2.68

3

3.14

2.65

3.01

3.07

2.09

2.99

3.23

24

1.86

2.83

4.21

4.06

Antimony

<2.8

<2.8

<2.8

<2.8

<2.8

<2.8

<2.8

<2.8

<2.8

<2.8

<1.4

<1.4

<1.4

<1.4

0.733

0.589

Thallium

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

<0.85

0.742

0.596

Element

Average

Standard Deviation

Average +3SD

Maximum

Count

Silver

0.313

0.242

1.039

1.25

16

Arsenic

1.363

1.439

5.679

6.89

16

Barium

69.78

11.31

103.7

90.6

16

Beryllium

0.212

0.109

0.54

0.508

16

Cadmium

0.362

0.256

1.13

1.25

16

Chromium

30.26

5.101

45.56

39.

16

Mercury

<0.1

0

<0.1

<0.1

16

Lead

4.256

5.132

19.65

24

16

Antimony

2.183

0.825

4.659

2.8

16

Thallium

0.827

0.065

1.023

0.85

16

Table III. HWF Concentrations

Element

WFL1

WFL2

WFL3

WFD4

WFD5

WFH1

WFH2

WFH3

Average

Standard Deviation

Maximum

Count

Silver

<0.25

0.611

<0.25

0.27

0.46

<0.25

<0.25

0.372

0.339

0.126

0.611

8

Arsenic

5.26

2.23

1.02

2

1.5

4.85

<0.85

0.883

2.324

1.649

5.26

8

Barium

753

517

1130

176

269

1330

802

548

690.6

372.7

1330

8

Beryllium

<0.15

<0.15

<0.15

<0.5

<0.5

<0.15

<0.15

<0.15

0.238

0.152

0.5

8

Cadmium

4.18

6.1

4.58

5.5

7.7

7.4

4.34

3.07

5.359

1.52

7.7

8

Chromium

93.3

305

412

87

205

59.3

325

110

199.6

124.3

412

8

Mercury

<0.1

<0.1

0.429

<.5

<0.5

<0.1

0.425

<0.1

0.282

0.184

0.5

8

Lead

485

560

1900

193

392

368

1320

127

668.1

578

1900

8

Antimony

23.4

17

35.7

11.2

17.8

11.1

13.3

8.3

17.23

8.282

35.7

8

Thallium

<0.85

<0.85

<0.85

<0.38

<0.85

<0.85

<0.85

<0.85

0.733

0.204

0.85

8

Note: In the above tables, the calculations for Average, Standard Deviation, and Maximum did not take into account the "less than" signs. Please remember to make adjustments for this fact.

Maximum Allowable

Maximum Raw Material Feed

250 tons/hr

Maximum Coal Feed Rate

20 tons/hr

Maximum HWF Feed Rate

12.5 tons/hr

Feed Rate

Avg+3 SD

Worst

Avg+3 SD

Worst

Limits

Calculated

Elements

(lb/hr)

(ppm)

(lb/hr)

(ppm)

(lb/hr)

(ppm)

(lb/hr)

(ppm)

(lb/hr)

(ppm)

(lb/hr)

(ppm)

(lb/hr)

Arsenic

12.0115

5.67943

2.83972

15

7.5

16.1806

0.64722

50

2

100

2.5

100.461

2.51153

Beryllium

5.61549

0.53978

0.26989

5

2.5

10.151

0.40604

15

0.6

10

0.25

100.62

2.51549

Cadmium

8.32291

1.12951

0.56476

5

2.5

1.67715

0.06709

20

0.8

200

5

200.916

5.02291

Chromium

88.7782

45.5606

22.7803

50

25

81.0836

3.24334

100

4

2000

50

2391.13

59.7782

Mercury

28.0775

<0.1

0.05

20

10

<0.1

0.004

20

0.8

100

2.5

691.098

17.2775

Lead

336.929

19.6503

9.82515

25

12.5

41.5168

1.66067

50

2

1500

37.5

12897.2

322.429

Thallium

46.7958

1.02294

0.51147

20

10

0.9768

0.03667

20

0.8

100

2.5

1439.83

35.9958

Knowing the maximum allowable particulate emission rate for the cement kiln and the conservative emission rate for each metal, can any of these metals be eliminated from part of this process? The EPA Methods Manual for Compliance with the BIF Regulations states on page 10-8; "Determine which metals need to be monitored (i.e. all hazardous metals for which Tier III emission limits are lower than PM emission limits.....)".(2) Consequently, when the conservative emission rate for each metal is compared with the maximum allowable kiln particulate emission rate, it is readily determined that the allowable emissions for barium and silver exceed the allowable particulate emissions for the kiln. Since Ba and Ag compounds would be nonvolatile at kiln exit gas temperatures and the kiln must remain in compliance with overall particulate emissions, then Ba and Ag emissions will be in compliance also. This effectively eliminates Ba and Ag from all testing requirements. (Comments made by Bob Holloway of the US EPA at the recent 1992 Air & Waste Management conference indicate that in order to use this option, the full alternative methodology for implementing metals controls, Section 10 of the BIF guidance document, must be followed.) (3)

More in depth analysis of metals concentrations in the fuels and raw materials reveals that antimony, mercury and thallium can be addressed under Tier IA. This approach assumes that whatever amount of a given metal goes into the kiln, goes out the stack. A true worse case assumption. These metals will still need sufficient monitoring in the coal, raw materials, and waste fuel to demonstrate that there is no significant change in concentration, which might affect Tier IA consideration. Tier IA treatment does however, effectively remove them from spiking consideration for the compliance test. This results in significant savings. There is also the additional benefit of fewer variables during compliance test kiln spiking.

There is another critical compliance consideration for these metals. Antimony, mercury and thallium, may present leachate problems in the dust. So even though these metals are effectively removed from compliance test spiking considerations, monitoring and controlling them to levels that may be well below levels allowed by Tier IA is important towards maintaining Bevilled status for the kiln dust.

Metals for Compliance Testing

Thus far in the process, five metals have been eliminated from compliance test spiking consideration. Now the five remaining metals; lead, arsenic, beryllium, cadmium, and chromium must be addressed.

There is extensive data available on Pb,(4) ,(5) ,(6) ,(7) so when this is factored into Best Engineering Judgment considerations, a reasonable practical limit for Pb is readily determined. A critical factor with Pb is that most of it will end up in the dust. Several studies report that over 60% of the Pb concentrations going into the kiln ends up in the dust(8) ,(9) ,(10) ,(11) . Therefore, in terms of spiking concentrations, it is imperative that those levels going into the kiln do not result in dust concentrations that will leach out in excess of allowable TCLP limits during the compliance test. This would result in the loss of Bevilled status for the dust.

It is our experience that Pb levels exceeding 2000 ppm in the dust can result in leachable concentrations exceeding TCLP allowable levels. On the other hand, there have also been instances where Pb levels in dust have reached as high as 7000 ppm yet not exceeded TCLP levels, so the range to be considered is quite broad. Ultimately, historical data from the given kiln or perhaps data available from similar kilns can and should be used in final Pb spiking determinations.

Carcinogenic Metals

The remaining four metals are the carcinogenic metals. Tier III BIF treatment for carcinogenic metals cited under 40 CFR 266.106(d)(3) requires that: '...the sum of the ratios of the predicted maximum annual average off-site ground level concentrations...to the risk specific dose (RSD) for all carcinogenic metals emitted shall not exceed 1.0 ...." The RSD for a metal is the acceptable ambient level for that metal provided that only one of the four carcinogenic metals is emitted. Best Engineering Judgment and process capture efficiencies are also taken into account.

Spreadsheets were used to calculate maximum allowable feed rates for the BIF metals. These calculations take into account such factors as RACs, dispersion coefficients, APCS removal efficiencies, etcetera. An artificial distribution fraction was introduced in calculating the maximum allowable emission and feed rates for the carcinogenic metals. This produced maximum allowable HWF feed rate limits. Setting actual HWF feed rate limits less than the calculated maximums, results in a sum of ratios for the carcinogenic metals which does not exceed 1.0.

These calculations may result in emission levels that are high based upon realistic worst case input concentrations for a given metal. At this point, it may be desirable to increase the allowable input of one of the other metals. The distribution fraction could be adjusted to incorporate those desired changes.

As an example, the allowable input of chromium might be 3,000 ppm and 100 ppm for cadmium. 3,000 ppm Cr exceeds what can reasonably be expected to be found in the feedstreams, while 100 ppm cadmium is close to maximum concentrations. Adjusting the distribution factor allows the increase of cadmium to a more desirable level of 200 ppm while reducing chromium to 2,000 ppm, both of which are now more operationally realistic.

Practical Operational Limits

Tables I, II and III list the ten BIF metals concentrations for individual samples, the average concentration for each of the ten metals, and the average concentration plus three standard deviations; in raw materials, coal and waste fuels respectively. This average concentration plus three standard deviations is also presented in Table IV.

Using a statistical approach effectively reduces the monitoring frequency for coal and raw materials. Samples are taken once each shift. Daily composites are retained until results from a randomly chosen composite for a given month confirms that concentrations are indeed in compliance. In the event that concentrations are out of compliance, the retained samples can be used to help isolate the problem. HWF variability dictates a batch by batch analytical approach as compared to the statistical approach used for coal and raw materials.

Maximum metals concentrations for monitoring purposes are also presented in Table IV. As an example, lead levels should not exceed 1500 ppm in the blended waste fuels. As previously stated, these metals concentrations are based upon actual data from the materials used at this facility. These are realistic values determined from actual operating conditions. Although sufficient monitoring to insure continued compliance with BIF is not only required but a good idea in general, these numbers should not result in unrealistic operational restrictions for the Lone Star plant.

References

1. United States Environmental Protection Agency. Burning of Hazardous Waste in Boilers and Industrial Furnaces, 40 CFR Part 266, Government Printing Office (NTIS), 1990.

2. United States Environmental Protection Agency, Methods Manual for Compliance with the BIF Regulations, Government Printing Office (NTIS), 1990, p. 10-8.

3. Robert Holloway, Discussed during Panel Discussion 1-"BIF-What the Regulators Say Now", in Proceedings of the New RCRA Regulations for Industrial Boilers, Furnaces, and Incinerators Conference,Air & Waste management Association, Orlando, 1992

4. Myron Black, David Gossman and Mark Ward, "The Fate of Trace Metals in the Wet Process Cement Kiln", in Proceedings of the Waste Combustion in Boilers and Industrial Furnaces Conference, SP-73, Air & Waste Management Association, Kansas City, 1990, pp 70-93.

5. J. Bruce Tompkins and Michael Von Seebach, "The Behavior of Metals in Cement Kilns", in Proceedings of the 26th International Cement Seminar, Rock Products, New Orleans, 1990.

6. John Chadbourne Ph.D., "Behavior of Toxic Metals in Cement Kilns", in Proceedings of the 1990 Emerging Technologies in Resource Recovery and Emission Reduction in the Cement Industry Conference, Portland Cement Association, Dallas, 1990.

7. Standard Handbook of Hazardous Waste Treatment and Disposal, "Cement Kilns", Harry M. Freeman Ed., McGraw Hill, New York, 1989, pp 8.57-8.75.

8. Black, Gossman and Ward.

9. Tompkins and Von Seebach.

10. Chadbourne

11. Handbook.