An Evaluation of a Cement Kiln's Emissions While Under Worst Case Operating Conditions

Presented at the December,1997 Rock Products Conference

Craig Cape
Operations Consultant - Gossman Consulting, Inc.
David Gossman
President - Gossman Consulting, Inc.
Henry Winders
Director of Environmental Affairs - Continental Cement Company

ABSTRACT

USEPA guidance suggests testing a cement kiln for emissions under "worst case" and "normal" operating conditions when conducting a trial burn. The Clean Air Act, under its MACT guidence, may also require testing of cement kiln emissions during "worst case" operating conditions. One of the basic reasons for these requirements is EPA's concern over the possible incomplete combustion of fuels or other organics in the combustion zone during less than optimum operating conditions. EPA's concern is in part for PIC's (Products of Incomplete Combustion) or organics that may be emitted and are hazardous. The model that prompts EPA's thinking is that of an incinerator operating under upset conditions. An evaluation of data from incinerator emission testing supports this model for incinerators. Does emission testing of cement kilns and evaluation of their combustion systems support the use of this same model for cement kilns? This paper provides an evaluation of recent test data and test conditions from a cement kiln that conducted emission testing under both "normal" (thermally efficient) and "worst case" conditions. The results of the evaluation of this data cast doubt as to the basis of EPA's guidance and trial burn requirements.

INTRODUCTION

When conducting emission testing for EPA regulatory programs such as MACT or the Boiler and Industrial Furnace regulations a number of emission constituents come under scrutiny. Virtually any combustion process results in emissions of incompletely burned or reformed organic compounds known as products of incomplete combustion (PICs). PICs can be unburned organic compounds coming from the combustion fuels or process feedstocks. They can be thermal decomposition products resulting from organic constituents in combustion fuels or process feedstocks. PICs can also result from the reformation of organic compounds during or immediately after combustion. Other emissions such as metals, and HCl/Cl2 also are a part of EPA's testing requirements. This paper will briefly review some of the concepts behind EPA's position to require testing under worst case operating conditions in order to evaluate these emissions from cement kilns. It will then look at real testing data and assess the consistency between EPA's concept and the data from an actual trial burn. Finally, it will review some of the reasons why the data doesn't support EPA's trial burn concept.

BACKGROUND

In the cement manufacturing spectrum there are a number of different types of kiln designs in use. There are long wet, long dry, pre-heater dry, pre-heater/pre-calciner dry etc. For the most part EPA regulates these different types on the same basis. However, the regulations can differ when taking into account different fuel types. When a cement kiln operation partially substitutes hazardous waste derived fuels for some of its traditional fossil fuel it comes under EPA's boiler and industrial furnace (BIF) regulations. Part of the RCRA/ BIF permitting process for these kilns involves air emission compliance testing during a trial burn. Other cement kilns that use traditional fuels and in some cases substitute non-hazardous waste fuels must normally meet certain air permit conditions that are often under state or local regulatory control. While all of the emissions in this paper may not necessarily be regulated by these local regulations they are of interest. In addition, cement kilns will be regulated by EPA's Clean Air Act MACT rules which seek to further refine regulations for the control of air emissions. To comply with the regulations, air emission testing during a specified operational condition is normally required to demonstrate compliance.

EPA has a concept in their guidance which states that the amount of toxic organic compounds emitted from a combustion source depends on the concentration of toxic compounds in the fuel and the combustion conditions. While this concept is acceptable on a broad basis, when applying it to cement kilns it also must properly take into account specific characteristics of cement kilns which differentiate this process from other combustion devices. Because of the much larger thermal mass and the much greater time, temperature, and turbulence associated with a cement kiln process it is unlike any other combustion device. In addition, the chemical nature and quantity of the raw material used in the process are not fully considered in EPA's concepts about cement kiln emissions.

In the summer of 1996 GCI completed a trial burn program for a cement operation applying for its final permit under the BIF rules. Discussions with EPA prior to beginning the test lead to incorporating the test conditions that EPA wanted for it's worst-case. These test conditions related primarily to the sampling of organics, metals and HCl/Cl2. The organics were tested under both high and low temperature worst-case conditions. Further, to address EPA's desire for emission data which could be used for an indirect exposure risk assessment additional organic, metal and HCl/Cl2 emission tests were performed. This additional testing was scheduled for normal (thermally efficient) kiln conditions. These different conditions represent a range of actual operational conditions of this kiln. The results of the testing will be examined in this paper with respect to these conditions and EPA's guidance.

EPA's MACT STANDARDS

Compliance with EPA's MACT standards may require cement kiln air emission testing under some type of worst-case operating conditions. Superficially reading the MACT rules would suggest that testing need not be done under worst-case conditions. However, considering how limits will be established causes a serious consideration for doing the testing under some type of worst-case conditions. Limits established under normal operations may be so restrictive to actually restrict normal operations more than is practical. The EPA concept is that emissions of certain regulated pollutants need to meet standards during the time, when a kiln operates, that will pose the greatest potential for significant risk from emissions.

EPA's BOILER AND INDUSTRIAL FURNACE RULES (BIF)

Cement operations utilizing hazardous waste as fuel must conduct a part of their trial burn under worst-case conditions for final permitting under the BIF rule. EPA specifies to a certain extent what those worst-case conditions should be. They are based on the incinerator model and calls for essentially the highest and lowest combustion temperatures normally operated at, along with maximum feed rates of certain kiln inputs including the hazardous waste. Further, the EPA theory poses that a kiln operated at what EPA has determined to be worst-case conditions will not have emissions representative of normal emissions. Part of the theory EPA bases this on is that PIC emissions of an incinerator during normal conditions do not pose a significant risk but that, during what EPA defines as worst-case conditions, PIC emissions will likely pose a significant risk. While this may be true for the incinerator model is it similarly true for cement kilns? Are the worst-case conditions actually conditions which generate PIC emission which will pose a significant risk? Are they the conditions in a cement kiln which actually generate the greatest emissions for other parameters (ie. metals, HCl/Cl2)?

THE TEST PLAN

A test plan for the cement kiln's trial burn compliance was drafted with three different phases with phase 1 (Day's 1 & 2) approximately 4 months apart from phase 2 (Day 3) and phase 3 (Day 4). The first phase was designed to coincide with the cement kilns normal maintenance turn-around. The objective was that this would be a time of maximum worn kiln components and would result in minimum kiln efficiency with maximum kiln exit gas temperature conditions, maximum hazardous waste fuel (HWF) feed rates, and spiking of additional metals. Therefore, testing would represent maximum high temperature worst-case combustion conditions. Phase two was after kiln turn-around maintenance and the kiln running with a permit maximum feed rate of hazardous waste fuel (HWF), combined with the spiking of difficult to destroy organics into the combustion zone. These conditions, along with new parts in such areas as the chain section, would then represent maximum kiln thermal efficiency for clinker production and minimum kiln exit gas temperatures. Testing would therefore represent a condition where minimal excess heat would be available for complete combustion and equate to minimum temperature worst-case combustion conditions. Phase 3 was run the day after phase 2 with the kiln running under normal (thermally efficient) conditions with respect to HWF feeds and all other process feeds and operating parameters.

THE TEST RESULTS

The test results demonstrated compliance with the regulatory standards (BIF) that needed to be met. In further assessing the data, some differences can be found from what the EPA expected for each of the conditions. While test results for some emissions obtained during worst-case combustion conditions did reflect small increases, those increases were attributable to other factors. The other emission test results, during worst-case conditions, reflected very little if any change when compared to normal conditions. The tables provide a comparison look at the results under the three conditions. An important comparison in the tables, in addition to the emission values, are the operating conditions which define the combustion conditions.

DIFFERENCES FROM EXPECTED

The emissions of organics, found in Tables 1 and 2 do not reflect an overall increase during the worst-case conditions. A review of the results for dioxins, PCBs, and semi-volatile organic emissions clearly demonstrate values that are close to the same for the three test conditions. The volatile organic emissions also show a closeness of values for the three conditions, although for some values they are slightly lower during the normal operational condition. The relatively low values and the relative consistency from condition to condition demonstrate that the worst-case combustion conditions had little effect on the organic emissions.

HCl/Cl2 results, found in Table 3, similar to the organics did not show an increase during the worst-case conditions. Although the emission values are similar in magnitude, the data suggests that the emission value obtained during normal conditions was higher. This despite higher chlorine input rates under the other condition. It should be further noted that the EPA sampling method for conducting HCl/Cl2 testing of cement kilns has come under criticism. The criticism is whether what is being analyzed from the emission sample train actually represents emissions or is it formed in the sample train or is it actually detecting chlorinated salts normally found in the emissions. Whatever form of chlorine is being measured the results do not meet EPA's expected values for the conditions.

Metals emissions found in Table 3, except for mercury, did reflect generally an overall increase during the worst-case conditions. This however was to be expected because the worst-case conditions for the metals test of Day 1 included the added spiking (Table 4) of five of the metals into the hazardous waste fuel in addition to the other worst-case operating parameters such as higher combustion zone temperature. The purpose of the spiking was to determine maximum metal input permit limits for those five metals which was a regulatory requirement (BIF) in conjunction with the emission testing. Therefore, this testing was reflective of more than a worst-case combustion condition but also of the worst-case metal feed rate condition. It should be noted that prior to beginning the emission tests with metal spiked HWF feeds, the procedure required that a metal equilibrium be established in the kiln system with the metals being spiked. This was done to compensate for any initial dilution effect of the process offsetting the spiking. Figure 1, demonstrates a comparison of the metal system removal efficiencies (SRE) for both metal test conditions. It demonstrates that SRE's were relatively unaffected by the conditions.

Gossman Consulting, Inc. (GCI) has looked at this issue many times in the past and has a paper looking at even more specific impacts of various other operational parameters on metal emissions. The paper is titled The Effects of Process Differences on System Removal Efficiencies (SRE) and the Fate of Metals in Cement Kilns. This paper can be found on the World Wide Web at the "gcisolutions.com" web site.

Particulate matter (PM) emissions were also tested. Comparing the results from the two conditions under which these emissions were tested does show a difference between worst-case and normal conditions. However, these results are not a function of the different combustion conditions but rather the differences of the air pollution control device (APCD). The different KVA's at which the electrostatic precipitator (ESP) was operated during the two test conditions is the key difference. With the lower power applied during the worst-case condition the APCD would not be expected to be as efficient as when under normal operations. It is interesting to note that despite the impact on particulate emissions, metal SRE's (system removal efficiencies) were apparently unaffected by the worst case conditions, including the lower ESP power. It is interesting because EPA contends that PM controls are a supplement to metal controls.

CONCLUSIONS

Certainly the organic emissions overall do not follow the EPA logic on combustion conditions and neither do the metals or HCl/Cl2 results. Worst-case combustion conditions on Day's 1 - 3 do not lead to significant results of higher emissions. The primary reason that it doesn't, is because the thermal mass, time, temperature and turbulence of a cement kiln process are very capable of continuing good combustion under those conditions. The variability of organic emissions are likely affected by the organic content of the raw material as it enters the process not the combustion conditions. Metals and particulate emissions were affected by other processes rather than combustion conditions to cause their trends toward lower emissions under normal conditions.

EPA's concept that there are great similarities between the operations of cement kilns and incinerators is once again questioned. EPA's belief that worst-case combustion conditions for incinerators are the same as in cement kilns is not supported by the data. It has long been argued by the cement industry that a cement kiln is simply a larger more stable combustion process than an incinerator. When it is operated with normal care during its operation it is much more able to withstand variability in operations than an incinerator. In essence what is a worst-case condition for an incinerator is easily accommodated by a cement kiln without any substantive increase in emissions. Based on this test data the only impacts on emissions observed were caused by higher metal input rates from metal spiking and by the ESP operating at a reduced KVA.

Table 1

Summary of Kiln Test Operating Conditions vs. Average Organic Emissions and Process Conditions


Unit

Test Day

Test Day

Test Day

Test Day



1

2

3

4

Test Mode

Max. Temp.

Max. Temp.

Min. Temp.

Normal

PARAMETERS






Chlorine

tons/hr

0.56

0.39

0.20

0.07

Coal

tons/hr

6.7

7.4

5.5

10.6

Total HWF

tons/hr

15.6

16.1

15.7

9.6

Kiln Feed

tons/hr

140.2

140.0

130.6

130.3

Kiln Excess Oxygen

%

2.2

2.6

2.1

2.6

Combustion Zone Temp.

F

2391.0

2662.7

2213.0

2391.7

Kiln Exit Gas Temp.

F

605.6

602.7

558.3

559.4

ID Fan

rpm

587.3

586.1

579.7

580.7

ESP (Primary)

KVA

230.0

257.5

421.4

409.0

CO ( 7 % O2 corrected)

ppm

13.6

18.5

21.2

17.5

ORGANIC EMISSIONS




Dioxins and Furans (TEQ)

ng/dscm


1.4

0.576

1.04

PCBs (Total)

ug/dscm


0.321

0.108

1.58

Semi-volatile organics




Naphthalene

ug/dscm


22.3

23.5

20.1

Phenol

ug/dscm


15.7

16.1

18.2

2 - Methylnapthalene

ug/dscm


11.1

2.09

2.03

2,4,6 - Trichlorophenol

ug/dscm


9.05

3.49

3.95

Dibenzofuran

ug/dscm


5.28

6.60

6.24

2,4 - Dichlorophenol

ug/dscm


4.71

1.30

2.78

Acenaphthylene

ug/dscm


3.77

2.76

2.54

3&4 - Methylphenol

ug/dscm


3.32

1.33

2.05

2 - Chlorophenol

ug/dscm


3.21

1.40

1.68

Benzyl alcohol

ug/dscm


2.77

1.61

2.25

1,3 - Dichlorobenzene

ug/dscm


1.52

0.435

1.67

1,2,4 - Trichlorobenzene

ug/dscm


1.43

0.485

0.805

Fluorene

ug/dscm


1.35

0.0861

0.0964

Pyrene

ug/dscm


1.22

0.705

0.600

Fluoranthene

ug/dscm


1.05

1.15

1.01

2 - Methylphenol

ug/dscm


0.979

0.592

0.915

Anthracene

ug/dscm


0.789

0.476

0.432

1,2 - Dichlorobenzene

ug/dscm


0.720

0.366

0.472

2,4,5 - Trichlorophenol

ug/dscm


0.560

0.228

0.255

Chrysene

ug/dscm


0.428

0.655

0.548

1,4 -Dichlorobenzene

ug/dscm


0.347

0.237

0.327

Benzo(a)anthracene

ug/dscm


0.260

0.208

0.228

Benzo(b)fluoranthene

ug/dscm


0.229

0.240

0.219

Indeno(1,2,3-cd)pyrene

ug/dscm


0.0391

0.0419

0.0388

Table 2

Summary of Kiln Test Operating Conditions vs. Average Volatile Organic Emissions and Process Conditions


Unit

Test Day

Test Day

Test Day

Test Day



1

2

3

4

Test Mode

Max. Temp.

Max. Temp.

Min. Temp.

Normal

PARAMETERS






Chlorine

tons/hr

0.56

0.39

0.20

0.07

Coal

tons/hr

6.7

7.4

5.5

10.6

Total HWF

tons/hr

15.6

16.1

15.7

9.6

Kiln Feed

tons/hr

140.2

140.0

130.6

130.3

Kiln Excess Oxygen

%

2.2

2.6

2.1

2.6

Combustion Zone Temp.

F

2391.0

2662.7

2213.0

2391.7

Kiln Exit Gas Temp.

F

605.6

602.7

558.3

559.4

ID Fan

rpm

587.3

586.1

579.7

580.7

ESP (Primary)

KVA

230.0

257.5

421.4

409.0

CO ( 7 % O2 corrected)

ppm

13.6

18.5

21.2

17.5

ORGANIC EMISSIONS




Volatile organics




Chlorobenzene

ug/dscm


11.8

9.63

9.60

1,2,4 - Trimethylbenzene

ug/dscm


5.16

6.47

3.30

Bromochloromethane

ug/dscm


5.03

1.88

1.29

Ethylbenzene

ug/dscm


4.26

4.48

3.85

o - xylene

ug/dscm


4.22

8.64

7.85

n - butylbenzene

ug/dscm


2.44

1.35

0.609

Chloroform

ug/dscm


1.50

1.74

1.67

n - propylbenzene

ug/dscm


1.40

1.29

0.693

1,3,5 - Trimethylbenzene

ug/dscm


1.10

8.68

1.33

Table 3

Summary of Kiln Test Operating Conditions vs. Other Average Emissions and Process Conditions


Unit

Test Day

Test Day

Test Day

Test Day



1

2

3

4

Test Mode

Max. Temp.

Max. Temp.

Min. Temp.

Normal

PARAMETERS






Chlorine

tons/hr

0.56

0.39

0.20

0.07

Coal

tons/hr

6.7

7.4

5.5

10.6

Total HWF

tons/hr

15.6

16.1

15.7

9.6

Kiln Feed

tons/hr

140.2

140.0

130.6

130.3

Kiln Excess Oxygen

%

2.2

2.6

2.1

2.6

Combustion Zone Temp.

F

2391.0

2662.7

2213.0

2391.7

Kiln Exit Gas Temp.

F

605.6

602.7

558.3

559.4

ID Fan

rpm

587.3

586.1

579.7

580.7

ESP (Primary)

KVA

230.0

257.5

421.4

409.0

CO ( 7 % O2 corrected)

ppm

13.6

18.5

21.2

17.5

OTHER EMISSIONS




Particulate

lb/hr

29.1



13.0

HCl/Cl2

lb/hr

39.154



48.171

Metals




Mercury

lb/hr

6.11E-03



1.25E-02

Silver

lb/hr

2.16E-03



2.14E-04

Arsenic

lb/hr

< 2.80E-03



< 8.70E-05

Barium

lb/hr

2.53E-03



1.24E-03

Beryllium

lb/hr

1.11E-04



< 4.35E-05

Cadmium

lb/hr

9.02E-02



2.57E-03

Lead

lb/hr

5.34E-01



9.44E-02

Chromium

lb/hr

2.09E-03



4.93E-04

Nickel

lb/hr

< 3.78E-04



4.25E-04

Antimony

lb/hr

< 2.67E-03



3.04E-04

Selenium

lb/hr

< 5.60E-03



4.76E-03

Thallium

lb/hr

< 1.57E-02



1.51E-03







Table 4

Summary of Kiln Test Constituent Spike Inputs

Parameter

Unit

Test Day

Test Day

Test Day

Test Day



1

2

3

4


Mode

Max. Temp.

Max. Temp.

Min. Temp.

Normal







Chlorine

lb/hr

1128




Arsenic

lb/hr

11.3




Beryllium

lb/hr

0.755




Cadmium

lb/hr

13.5




Lead

lb/hr

73




Chromium

lb/hr

71










Sulfur Hexafluoride

lb/hr



7.21


1,2,4 - Trichlorobenzene

lb/hr



49.9


Tetrachloroethylene

lb/hr



108.2








Figure 1

REFERENCES

1. Trial Burn Report, Continental Cement Company, Hannibal, Missouri, Prepared by Gossman Consulting, Inc., August 1996

2. Environmental Protection Agency, "Burning of Hazardous Waste in Boilers and Industrial Furnaces - Final Rule", Federal Register, Volume 56, No. 35, February 21, 1991.

3. Environmental Protection Agency, "Hazardous Air Pollutants: Regulations Governing Equivalent Emissions Limitations by Permit - Final Rule", Federal Register, Volume 59, No. 97, May 20, 1994

4. Woodford, Jim; Gossman, David; Jameson, Rex; Gossman, Sue; "The Effect of Process Differences on System Removal Efficiencies and the Fate of Metals in Cement Kilns, in The Proceedings of the 1995 A&WMA Conference on Waste Combustion in Boilers and Industrial Furnaces

5. Environmental Protection Agency, "Guidance on Trial Burns", "Draft" June 2, 1994 attachment to the "Exposure Assessment Guidance for RCRA Hazardous Waste Combustion Facilities", "Draft" April 1994