U.S. DEPARTMENT OF COMMERCE
                          National Technical Information Service
                          PB-280 118
Burning Waste  Chlorinated
Hydrocarbons in a
Cement Kiln
Environmental Protection Service, Montreal (Quebec)
Prepared for

Environmental Protection Agency, Washington, D C

Jan 78
                                          J

-------
                                      • *"
          Prapubltoati-on issue for EPA libraries
         and State Solid Waste Management Agenaiee
          BURNING WASTE CHLORINATED HYDROCARBONS

                     IN A CEMENT.KILN
      Shis report (SW-l47a) describes uork performed
for 1ha Office of Solid waste under aontraat no. 68-01-2966
     and is reproduced OB received from the contractor.
    the findings should be attributed to the contractor
           and not to One Office of Solid Waste.
     The reader is advised to utilize the information
        and data "herein witfc caution and judgement.

             Copies will be available from the
         National Technical Information Service
                U.S. Department of Commerce
               Springfield, Virginia  22161
           U.S. ENVIRONMENTAL PROTECTION AGENCY

                           1978

-------
This report has been reviewed by the U.S. Environmental Protection
Agency and approved for publication. Its publication does not signify
that the contents necessarily reflect the views and policies of the U.S.
Environmental Protection Agency, nor does mention of commercial products
constitute endorsement or recommendation for use by the U.S. Government.
An environmental protection publ cat1on (SW.(47c) in the sol 4 d waste
management series.

-------
NOTICE
THIS DOCUMENT HAS BEEN REPRODUCED
FROM T E BEST COPY FURNISHED US BY
THE SPONSORING AGENCY. ALTHOUGH IT
IS RECOGNIZED THAT CERTAIN PORTIONS
ARE ILLEGIBLE, IT IS BEING RELEASED
IN THE INTEREST OF MAKING AVAILABLE
AS MUCH INFORMATION AS POSSIBLE.

-------
heLIOGRAPHIC DATA 1. Report I Jd [ 2. -.
SHEET e flA/S3bfSW-/f’7 I
4. Title and Subtitle
R tA ss oo, o

5. Report Date
Burning Waste Chlorinated Hydrocarbons In a Cement
Kiln
1-78 Approval
6.
7. Author(s) LD. McDonald, p J• Skinner, F.J. Hopton and
G.H. Thomas
8. Performing Organization Repr.
No.
9. Performinl Orga izacioa Name and Address
Environmental Protection Service
Fisheries and Environment Canada
Montreal, Quebec
10. Project/Task/Work Unit No.
11. Contract/Grant No.
EPA 68-01-2966
12. Sponsoring Organization Name and Address
U.S. Environmental Protection Agency
401 M St., S.W.
Washington, D.C. 20460
13. Type of Report & Period
Covered
Final
14.
15. Supplementary Notes -.
Prepared for Fisheries and Environment Canada with partial
fundin from the United States Environment 1 Protection Agency
16. Abstr ets
An experimental program was carried out in 1975/76 at the St. Lawrence Cement Co.,
Mississauga, Ontario-in which waste chlorinated hydrocarbons, containing up to about
46 weight percent chlorine, were burned in a rotary cement kiln. Materials
burned included mixtures of ethylene d-ichloride, chlorotoluene and up to approximately
50 percent polychlorinated,biphenyls (PCB). - - -.
These materials were destroyed in -the cement kiln with at least 99.98 percent
efficiency in all cases. Emissions-of high molecular weight chlorinated hydrocarbons
were not detected. Three light chlorinated hydrocarbons, dichloromethane, chloroform
and carbon -tetrachlorjde, were found in the emissions in the part per billion or lower
range. The quantity of precipitator dust requiring disposal, as well as emissions of
particulate matter, increased during the test. - A reduction in fossil fuels used
while burning chlorinated hydrocarbons was noted.
17. Key Words and Document Analysis. ha. Descriptors -
Cement Kiln
Chlorinated Hydrocarbons
Cement Making -
Incineration
Polychiorinated Biphenyls Destruction
Fuel Reduction Cement Making Economics
Waste Utilization
lYb. Identifiers/Open-Ended Terms
1 7c. COSATI Field/Group -
I _________________________________________________________
18. Availability Statei enc V 19. Security Class (This - 21
o i ii i Report)
e ease ufli1m1 eu U 1 fCLA$S1FIEP
20. Securi y Class (This 22 . Price
Page -
- 1JNCLASS IPIED #4)0 — ‘1w I
FORM NTI5.35 IAEV. .73) ENDORSED BY ANs I AND UNESCO. THIS FORM MAY BE REPRODUCED U5COMM OC 5555. P74

-------
AB STRACT
An experimental program was carried out in 1975/76 at the St.
Lawrence Cement Co., Mississauga, Ontario In which waste chlorinated hydro-
carbons, containing up to about 46 weight percent chlorine, ware burned
in a rotary cement kiln. The chlorinated hydrocarbons were burned in three
distinct phases of increasing difficulty of combustion. Materials burned
Included mixtures of ethylene dichioride, chlorotoluene and up to approxima-
tely 50 percent polychlorlnated biphenyls (Pca).
These materials were destroyed in the cement kiln with at least
99.98 percent efficiency in all cases. Emissions of high molecular weight,
chlorinated hydrocarbons were not detected. Three light chlorinated hydro-
carbons, dlchloromethane, chloroform and carbon tetrachloride, were found
In the emIsslo. e In the part per billion or lower range. The quantity of
precipitator dust requiring disposal, as well as emissions of particulate
matter, increased during the test.
The chlorine input from the chlorinated hydrocarbon waste was up
to about 0.8 weIght percent relative to clinker and, this effectively reduced
the’alkàll concentration of the clinker in direct stoichiometric proportion.
A reduction iii fossil fuels used while burning chlorinated hydrocarbons was
noted.
iii
Preceding page blank

-------
TABLE OF CONTENTS
Page
ABSTRACT
TABLE OF CONTENTS ii i
List of Figures
List of Tables vii
EXECUTIVE OUTLINE xi
SUMMARY
RECOMMENDATIONS xv
- LIST OF ABBREVIATIONS xv
INTRODUCTION
2. CEMENT MANUFACTURE 2
2.1 General Principles 2-
2.2 Effect of Alkalies S
2.3 The St. Lawrence Cement Co. 6
2.3.1 Relevant unit processes - wet process kiln 6
2.3.2 Relevant unit processes - suspension preheater kiln 6
3. CONSIDERATIONS AT THE PROGRAM PLANNING STAGE 10
4. TRIAL ON THE SUSPENSION—PREHEATER KILN 12
5. TRIAL ON THE WET PROCESS KILN 15
5.1 Discussion 15
5.2 OrganIc Chloride Waste-Burned 17
5.3 Emissions 22
5.3.1 Free chlorine and hydrogen chloride 22
5.3.2 Gaseous organic compounds 22
5.3.2.1 Desorbed samples 23
5.3.2.2 Organic solvents extracted samples 25
5.3.2.3 Results obtained by the participating laboratories 26
5.3.2.4 Estimated minimum combustion efficiencies 27
5.3.3 Particulate emissions 28
5.4 Mass Balance on Wet Kiln 30
5.4.1 Significance of the mass balance 30
--

-------
TABLE OF CONTENTS (CONTSD )
5.4.2
6.
6. 1
6.2
6.3
6.4
6.5
6.6
34
34
34
37
38
38
39
42
43
46
Page
31
‘Chlorine and potassium retained
CONSIDERATIONS ON BURNING CHLORINATED HYDROCARBON
WASTES IN A CEMENT KILN
Effect on Production
Alkali Reduction While Burning Chlorinated Hydrocarbon
Wastes
Heat Recovery from Chlorinated Hydrocarbon Wastes
Cement Quality
Extrapolation to Other Kiln Types
Comparison of Cement Kiln Burning with Other Uses and
Disposal Methods for Waste Chlorinated Hydrocarbons
7. CONCLUSIONS
REFERENCES
ACKNOWLEDGEMENTS
APPENDIX A - QUANTIFYING, SAMPLINGAND ANALYSIS OF PROCESS
MATERIALS
APPENDIX B - ANALYTICAL DATA, CALCULATION AND DETAILS ‘OF
EXPERIMENT ON THE SUSPENSION PREHEATER KILN
APPENDIX C - RESULTS AND CALCULATIONS FOR WET PROCESS SYSTEM
APPENDIX D - QUALITY OF CEMENT PRODUCED
APPENDIX E - EQUIPMENT DESCRIPTION AND ASSOCIATED ECONOMICS
APPENDIX F - ONTARIO MINISTRY OF THE ENVIRONMENT EMISSION
GUIDELINES AND ANALYTICAL SUPPORT
APPENDIX G - LABORATORY ANALYSIS RESULTS FROM THE ST. LAWRENCE
CEMENT FACILITY TEST (TRW Systems Group)
APPENDIX H - DEVELOPMENT, CONSTRUCTION AND EVALUATION OF A
COLLECTION SYSTEM FOR LOW MOLECULAR WEIGHT
HYDROCARBONS (Ontario Research Foundation)
APPENDIX I - CC/MS/COMPUTER DETERMINATION OF CHLORINATED
HYDROCARBONS AND PCB’s (Air Pollution Control
Directorate, EPS, Environment Canada)
49’
93
1 07
129
137
149
165
207
221
V

-------
LIST OF FIGURES
Figure Page
Wet Process KIln 3
2 Dry Process Kiln
3 PrInciple of Fuller-Humboldt Suspension Preheater and
By-pass 7
4 Alkali By-pass 9
A.l Schematic of the Material Balance 50
A.2 Schematic of Port LocatIons 59
A.3 Gas Flow Distribution at Sampling Points 60
A.4 Grab Bag Sampling Equipment 61
A.5 Particulate Sampling Train 62
A.6 Gaseous Sampling Train 65
A.7 Gas Chromato raphi,c Profile, from-Flame Ioni ’za lon
Detector for Chlorinated ‘Aliphatics (WBA) Sample Feed 73
A.8 Gas Chromatographic Profile from Flame Ionization
Detector for-Chlorinated A’liphatics plus Aromatics
and Alicyclics (WBB) Sample Feed 74
A.9 Gas Chromatographic Profile from Flame Ionization
Detector for Chlorinated Aliphatics plus Aromatics,
Alicyclics and Polychiorinated Biphenyls (WBC) Sample
Feed 75
A.lO Gas Chromatographic Profile from’Electron Capture
Detector for WBB Sample Feed 76
A.ll Gas Chromatographic Prof!les from Electron Capture
Detector for WBC Sample Feed and Standard Arocolor 1242 77
A. 12 Gas Chromatographic Profiles from Electron ‘Capture
Detector for Low Molecular Weight Chlorinated Hydro-
carbons and for BLB and WBC Test Samples 79
A.l3 Gas Chromatographic Profiles from Flame Ionization
Detector for Low Molecular Weight Chlorinated
Hydrocarbons 80
A.14 Gas Chromatographic Profile from Electron Capture
Detector for ImpingerExtract from BLA Test 3 81
v i

-------
LIST OF FIGURES (CONT’D)
Figure Page
A.15 Gas Chromatographic Profile from Electron Capture
Detector for Impinger Extract from WBC Test 3 82
A.16 Gas Chromatographic Profile from Electron Capture
Detector for Impinger Extract from WBC Test 3 after
Cleanup and SeparatIon 83
B.l Chlorine Level in Stage IV,.June 3, 1975 102
B.2 Chlorine Level in Stage IV, June 10, 1975 103
E.l Schematic Diagram of Basic Chlorinated Hydrocarbon
Feed System 140
E.2 Schematic Diagram of Chlorinated Hydrocarbon FacilitIes 141
F.1 Representative Bar Chromatogram 156
F.2 Computer Reconstructed Bar Chromatograms for PCB Fuel
and Aromatic Fuel plus Arocolor 1242 158
F.3 Gas Chromatogram from CC/MS Analysis of Sample
PCB Fuel 59
G. 1 TRW Sample Coding System 169
G.2 Plan for the Combination of ORF Solvent Extracts :171
G.3 Desorption System for Chromosorb 102 Tubes (TRW) 171
H.1 ORF Test Duct Schematic 212
vii

-------
LIST OF TABLES
Table Page
Composition of Aliphatics (WBA) 19
2 Composition of Aromatics plus Complex (WBB) 20
3 CompositIon of Aromatics plus PCB’s (WBC) 21
L i Gas Sample Volumes and Sample Concentration Factors 2 e
5 Estimated Kiln Emission Concentrations (GC-EC) for
Specific Volatile Organochiorine Compounds 25
6 Estimated Minimum Combustion Efficiencies for Each
Waste Burn 28
7 Sununary of Particulate Test Data 29
8 Accumulated Mass Balance for Chlorine 32
9 Accumulated Mass Balance for 1(20 32
10 Average Reduction 1n.K 2 0 Content of Clinker 35
U Average DustDischarged 36
12 Recovery of Btu from Chlorinated Hydrocarbons 38
A.l Process Materials Studied and Approximate Normal
Production Quantities 49
A.2 Quantities of Aliphatic Mixture Burned Daily 54
A.3 Quantities of Aromatic plus Complex Mixtures Burned
Daily 55
A.4 Quantities of PCB Mixture Burned Daily 57
A.5 GC Analysis - System A 69
A.6 GC Analysis - System B 69
A.7 GC Analysis - System C 70
A.8 Gravimetric Dust Analyses 85
A.9 Results from Leco Induction Furnace Analyses 86
A.lO Results from Atomic Absorption Analyses 87
A.ll Least Squares Data for Calibration Lines 88
vii I

-------
LIST OF TABLES (cowr’D)
Table Page
8.1 Percent Bypass Gas Required to Maintain Chloride Levels 95
B.2 Results from Analyses of Dry Process Kiln Raw Feeds 98
B.3 Results from Analyses of Dry Process Kiln Clinker 99
B.4 Results from Analyses of Stage IV Dusts 100
C.l Results from Analyses of Clinker Samples 109
C.2 Results from Analyses of Slurry Feed Samples 111
C.3 Results from Analyses of Discard Dust 113
C. e Results from Analyses of Return Dust 115
C.5 Btu and Chlorine Content of Chlorinated Hydrocarbons 117
C.6 Btu, S and Cl Analyses from No. 6 Fuel 011 118
C.7 Daily Record of Production and Materials ConsumptIon 120
C.8 Material Balance for Chlorine 122
C.9 Material Balance for K 2 0 125
D.l Cements frcm Clinker Produced During Baseline Burn 131
D.2 Cements from Clinker Produced During Aromatic plus
Complex Chlorinated Hydrocarbon Burn 132
D.3 Cements from Clinker Produced During Polychiorinated
Biphenyl Burn 133
F.1 MOE Specifications Applied to Waste Burns 150
F.2 St. Lawrence Cement Waste Burn Experiments - Test 1
Chrcmosorb Adsorption Tube Analysis 153
F.3 St. Lawrence Cement Waste Burn Experiments - Test 2
Chrcmosorb Adsorption Analysis 153
St. Lawrence Cement Waste Burn Experiments — Test 3
Chromosorb Adsorption Analysis 153
F.5 Fuel Sample IdentificatIon 155
F.6 Gas Chromatograph Conditions 155
lx

-------
LIST OF TABLES (cowr’D )
Table Page
F.7 MS Data from GC Analysis of Sample PCB Fuel 160
F.8 Area Counts - 161
G.1 Summary of Samples Received from Each Waste Burn (TRW) 168
G.2 Organic Composition of Aromatic Waste by GC/MS 178
G.3 Trace Metals in the Chlorinated Aromatic Waste by SSMS 179
G.4 Organic Composition of PCB Waste by GC/MS 181
G.5 Trace Metals in the PCB Waste by SSMS 182
G.6 Summary of Organic Qualitative Survey Analysis of Sample
Extracts (TRW) 185
G.7 Approximate Constituent Levels of Trace Vapours Desorbed
from Sorbent Tube Samples by LRMS 188
G.8 Summary of the Interpretation of LRMS Spectra for Trace
Vapours Désorbed from Sorbent- Tube Sampl s 189
G.9 Results and Detection Limits from GC/FID Analysis of the
Concentrated Extracts 190
G.1O Results and Detection Limits from GC/ECD Analysis of the
Concentrated Extracts 192
G.11 Results and Detection Limits from GC/ECD Analysis of the
Unconcentrated Extracts 194
G.12 - Trace Metal Semi-Quantitative Survey of Filter Digests
- by ICPOES 195
G.13 Limits of Detection for Undetected Elements by ICPOES 196
G.14 Concentration of Trace Metals in Effluent Gas Particulate
Matter by MS 197
G. 15 Concentration of Trace Metals In Aqueous Samples by AAS 197
G.16 Results of Organic Survey Analysis on Clinker Product and
Disposal Dust Samples 199
G.l7 Results and Detection Limits from GC/FID Analysis 201
G.18 Results and Detection LIm!ts from GC/ECD Analysis 201
x

-------
LIST OF TABLES CONT’D)
Table Page
G. 19 Selected Trace Metals In SLC Clinker Product and
Discard Dust Samples by SSMS 202
xl

-------
EXECUTIVE OUTLINE
Experiments using chlorinated hydrocarbon wastes in the manu-
facture of cement appear to have transformed a difficult waste disposal
problem into a solution which is not only economically and environmentally
satisfactory, but has a beneficial effect on the quality of the cement.
Safe disposal of the large volumes of the chlorinated hydrocarbon
wastes which are generated in Canada each year is a difficult environmental
problem. It is estimated conservatively that Canada generates some 25-30
million pounds annually of these highly toxic end persistent wastes. This
figure does not include many toxic compounds such as insecticides and
polychlorinated biphenyls (PCB’s).
Environmentally safe disposal is difficult and expensive. Disposal
on land, or underground, requires special and expensive precautions to
prevent leaching into waterways. The favoured method of disposal, inci-
neration with recovery of hydrogen chloride, can be very costly. Without
recovery, scrubbing equipment is required to remove hydrogen chloride from
the products of combustion to control its emission. This, in turn, usually
necessitates a satisfactory disposal method for the scrubber liquid. Va-
luable fuel must be burned to maintain combustion while incinerating
chlorinated hydrocarbon wastes, as extremely high temperatures with long
residence times are required for thel.r thermal destruction.
A long high—temperature f lamé is required In a cement kiln to
achieve the desired product quality. During normal operations, the thermal
conditions that are necessary effectively consume the toxic materials. The
kiln also contains a considerable quantity of lime and thus has an intensive
“scrubbing” action.
Recovery and re-use of hydrocarbon wastes is not always felt to
be economic. The recycling of solvents Is practiced when it Is economically
and technically feasible. H vever, this entails the control of segregation,
storage, collection and ultimate treatment of the various hydrocarbon streams.
The report strongly urges that combustible liquid wastes, many of
which are persistant environmental contaminants, be combined by m. ans of a
xli

-------
single economical recovery system which would collect and later use the
wastes as supplementary fuel for cement manufacture.
The experiments described In this reportwere monitored by
the Department of Fisheries arid Environment, the Ontario Ministry of the
Envlronment the United States Environmental Protection Agency, and the
Ontario Research Foundation.
For these experiments, Industrial chlorinated hydrocarbon wastes
including polychiorinated biphenyls (PCB’s) were burned during the
commercial manufacture of cement.
This not only utilized the thermal value of the wastes, with a
reduction in oil consumption, but the results showed that almost all the
toxic wastes were completely destroyed In the kiln. Emissions of toxic
materials Into the atmosphere were negligible.
Calcium chloride Is often used In cement manufacture to reduce
the alkali content of the product... The only apparent effects that the
experimental burn had on the quality of the cement were the beneficial
effects due to the incidental addition of chloride-ion.
xiii

-------
SUMMARY
Chlorinated hydrocarbon wastes were burned in a carefully con-
trol led experimental trial as a partial fuel at the St. Lawrence Cement
Co., Mississauga, Ontario. The experiment was conducted to determine
whether chlorinated hydrocarbon wastes could be burned in a cement kiln
without adverse effectson air pollution levels.
The wastes used included a variety of chlorinated hydrocarbons
in the series of program phases designed to progress from easily combusted
chlorinated hydrocarbons to those which are combusted only with difficulty.
The last phase consisted malnlyof polychlorinated biphenyl wastes. These
materials were processed and formulated from chemical wastes as required
for the different phases and supplied by Chemtrol Pollution Services Inc.
Atmospheric emission measurements were made before, during and
after the burning of each blend of chlorinated waste. Two methods of emis-
sions sampling were used during each phase, the method normally used for
measuring emissions of particulate matter, and a sampling train designed
speciallyfor determining emissions-of organic material. Alisamples-
from both systems were analyzed for unburnedchlorinated hydrocarbons.
It was concluded that the combustion efficiency was at least 99.986
percent for the chlorinated hydrocarbons. Approximately 50 ppb of
volatile low molecular weight compbunds were found In the emission samples.
There were no detectable quantities of high molecular weight chlorinated
compounds in the stack gases.
A mass balance was carrledzoUt on áhlorine and potassium. This
showed that the chlorine input as chlorinated hydrocarbon was completely
reacted with the process solids.
The alkali content of the clinker showed a reduction which
corresponded exactly with the quantity of chlorine input to the system.
This agreement further confirms the data from emission testing and the mass
balance.
While burning chlorinated hydrocarbons with approximately 40
percent chlorine, a decrease, in oil consumption equivalent to 65 percent
of the Btu content of the chlorinated hydrocarbons was obtained.
The only differences in the quality of clinker produced while
burning chlorinated hydrocarbons were the beneficial effects which were
expected through the reduction in alkali content.
xi V

-------
- It was concluded that dii r1n ted dre erban te; ,
used in cement kllns, replacing other forms of chlorine used for -
reduction of alkali content. A small proportion of fossil fuel required
for cement manufacture is conserved through use of these materials.
Burning chlorinated hydrocarbon wastes Is considered a valuable means of
destroying persistent and toxic forms of pollutants while recovering useful
heat values.
xv

-------
RECOIIMENDATI ONS
The experiments have shown that there is virtually no adverse
effect on air pollution levels by burning chlorinated hydrocarbon wastes
In a cement kiln. These wastes Include polychiorinated biphenyls and other
materials which are difficult to destroy. When other methods, such as
Incineration, are used hydrogen chloride and chlorine may be emitted and,
if the incinerator operation Is poorly controlled, uncombusted material
may be released into the environment.
The following reconmiendations are the outcome of the present
report:
(1) BurnIng chlorinated hydrocarbon wastes In a cement kiln Is considered
a valuable means of destruction of persistent and toxic materials which
are members of this family of compounds. Since flame temperatures and
retention times are similar for all cement plants regardless of type
of kiln or fuel used, chlorinated hydrocarbons should be destroyed
in any cement kiln. The feasibility of doing so ina particular kiln
Installation can be determined bya technical and economic review.
(2) Due to lack of familiarity with organic chemicals, it Is considered
essential that Instructions on safe handling procedures be given to
cement industry personnel. -
(3) Since problems such as precipitation, solidification, heat or gas release
can arise through mixing Incompatible waste materials in storage tank,
it is considered advisable to obtain such materials from one reliable
source of supply at any given time.
xvi

-------
LIST OF ABBREVIATIONS
APCD Air Pollution Control Directorate (EPS)
BLA Baseline A, designates emission tests before waste
burns
BLB Baseline B, designates emission tests after waste
burns
DCM Dichioromethane
EC Electron capture (detector)
EPA Environmental Protection Agency (u.s.)
EPS Environmental Protection Service (Fisheries and Environment
Canada)
FID Flame ionization detector
CC Gas chromatography
MOE (Ontario) Ministry of the Environment
MS Mass spectrometry
ORF Ontario Researëh Foundation
PCB Polychiorinated biphenyl
SLC St. Lawrence Cement Co.
TRW TRW Systems Group
Ti, 12, 13 Test one, Test t , Test three
WBA Waste Burn A (chlorinated aliphatics)
WBB Waste Burn B (WBA plus chlorinated aromatics and alicyclics)
WBC Waste Burn C (WBB plus polychiorinated biphenyls)
XRF X-ray fluorescence
xvii

-------
1 INTRODUCTION
In Canada each year an estImated 25-30 millIon pounds of chlori-
nated hydrocarbon wastes require disposal or destruction, 17-20 millIon
pounds being generated In Ontario [ i]. These figures are based on 56%
of annual production and may be conservative. Experience In Europe
Indicates that 10% of productlon.Is a more reliable estimate of waste
material.
There are waste streams from plants manufacturing, processing,
or using chlorinated hydrocarbons which must be disposed of, the method
of disposal frequently being Incineration. Chlorinated wastes other than
those directly from chemical plants also present a serious disposal problem.
Among these latter chemicals are polychiorlnated biphenyls (PCB’s) and
insecticides which may require disposal.
Many of these wastes are toxic and persistent, and all pose a
serious disposal problem. Among the methods presently used are
Incineration and land disposal. incineration with recovery of hydrogen
chloride can be costly. - Without recovery, combustion gases must be scrubbed,
“thereby generating a liquid waste requiring disposal. Both incineration-me—
-thods require additional fuel. - Deep welling and other simIlar methods of
disposal are environmentally unsound because of the risk of water contamina
ti on.
cement manufacture, the kiln operates at higher temperatures
and for longer residence times then those used In incinerators for des-
truction of these waste materials. It is also coninon practice in the
cement Industry to add chlorides to the kiln to reduce the alkali concen-
tration of the final product. Use of chlorinated hydrocarbon wastes in
a cement kiln would provide useful recovery of chlorine and energy and,
at the same time, solve a serious disposal problem. -
The present research program was carried out to determine
whether waste chlorinated hydrocarbons can be burned in a rotary cement
kiln wtthout causing adverse air pollution. The approach taken was to
analyse stack emissions for uncombusted chlorinated hydrocarbon. A mate-
rial balance on chlorine was undertaken to confirm the emission findings.

-------
2
2 CEMENT MANUFACTURE
2.1 General Principles
While a variety of raw materials may be used In cement manufac-
ture, materials containing calcium, silicon, aluminum and iron without
an excess of certain other elements are required. These materials are
ground to a fine powder called raw meal, the chemical composition of which
is carefully controlled by proper blending of the various materials. Nor-
mally, blending is achieved by grinding all the raw materials together
(intergrinding). Raw meals required for wet and dry processes are similar
except that the raw meal for the wet process is in the form of a slurry
containing approximately 35% water, while raw meal for the dry process
contains less than 0.5% water.
The raw meal is fed Into the kiln (see Figures 1 and 2) and is
burned In the kiln to produce an intermediate product called clinker. - -
The kiln slopes towards the burning zone and rotates slowly,
causing the raw material to gradual ly move-Into the burning zone. React ions
which occur during gradual-heating In t the kiln are: evaporation of free
water, -evolution of combined water,, evolution of carbon dioxide, from’-
carbonates, and combination of lime with silica, alumina and Iron to form
the desired compounds in the clinker. These reactions require a final
material temperature of 1450°C (2650°F). Four main compounds are present
in Portland cement clinker:
Coninon Abbreviations
Used In The
Name Of Compound Chemical Formula Cement Industry
Tricalcium Silicate 3 CaO’S’l0 C 3 S
Dicalcium Silicate 2 CaO•S 10 2 C 2 S
Tricalcium Aluminate 3 Ca0’Al 2 O C 3 A
Tetracalcium Alumlnoferrlte 4 CaO•A1 2 0 3 ’Fe 2 0 3 C 4 AF
Minor compounds are also formed In clinker, coninbnly magnesia
(MgO), potassium sulfate (1 (2504) and sodium sulfate (Na 2 SO 4 ).
Traces of other elements present In either the raw materials or
fuel are also found in clinker. Upon leaving the kiln, the clinker is

-------
1- KILN
2-SLURRY FEED
4-PRECIPITATOR OUST SCREW
3- PRECIPITATOR
5DUST RETURN
6-FUEL
7- CUNKER COOLER
8- CLINKER
9- FILTER
WET PROCESS KILN
FIGURE 1.

-------
DRY PROCESS KILN
I RAW MEAL FEED
2 STAGE I.
3 STAGE II.
4 STAGE III .
5 STAGE IV.
6 KILN
7 CLINKER COOLER
8 CLINKER
9 FUEL
A FILTER
B PRECIPITATOR
C BY-PASS
-—-- - -
FIGURE 2.

-------
5
rapidly cooled to avoid undesirable crystal forms of the above compounds.
After cooling, the clinker Is ground and blended, normally by Intergrinding,
with gypsum to a fine powder. The final product, called Portland cement, is
the basic ingredient of concrete.
In the burning process, considerable CO 2 is driven from the raw
meal. Any elements not driven off are increased In the clinker in propor-
tion to the quantity of CO 2 evolved. Throughout most of this report, the
analyses have been reported on a natural basis, i.e. reported results are
actual concentrations present in samples as received. In some cases,
results have been reported on an Ignited basis, I.e. on CO 2 free basis,
and have been Indicated as such in the report.
2.2 . Effect of Alkalies
In cement manufacture, reference to alkalies Implies potassium
and sodium oxide (K 2 0 and Na 2 0). Both alkalies are frequently combined and
reported as equivalent Na 2 0 for purposes of specification. The raw materials
at St. Lawrence Cement are such that Na 2 Ois low and practically constant
(see Table A.lO). For this reason, only potassium oxide (K 2 0) Is consi-
dered in detail in thls,report.
The effect of alkalies on cement quality has been well documented
(2, 3, 4]. While some alkali may be desirable for early strength develop-
ment (3], an excess can be problematic. The most extensively reported of
these problems is the alkali-aggregate reaction (5, 6]. Certain glassy
silicates and some dolomites react slowly with alkalies and cause expansion
and disruption of concrete. It has been found by experience that cement
containing less than 0.60% total alkalies reported as Na 2 0 performs satis-
factorily with such aggregates. This specification is imposed with suff I-
dent frequency in the United States that it Is found as an optional spe-
cification under ASTM C-150 (7]. Strength attained and setting characte-
ristics are also related to the alkali concentration of the cement (3, 8].
it is common practice In the cement Industry to add chlorides
such as calcium chlorIde or waste hydrochloric acid to the raw meal to
reduce alkalies (9,- 10]. AlkalIes are normally present as suiphates which
at kiln operating temperatures are not readily volatilized, but are retained
in the clinker. Alkali chlorides are volatile at normal kiln operating

-------
6
temperatures. = These are evolved from the material in the kiln and carried
in the gas stream to the precipitator. The high alkali dust from the pre-
cipitator can then be discarded.
2.3 The St. Lawrence ‘Cement Co .
The St. Lawrence Cement Co., Mississauga Plant, has a nominal
production capacity of 1 75O,00O short tons per year. The company operates
two wet processes and one dry process suspension preheater kiln. General
information on the company and the plant has been published in two articles
(11, 12].
2.3.1 Relevant unit processes - wet process kiln
Each of the two wet process kilns are dumbell-shaped Allis
Chalmers kllns 402’ long with a diameter of ll’6”, having nominal capacity
of 1050 short tons per day (see Figure 1). The chain system in the drying
zone has 57 tons of loose hung carbon steel chains with a radiation curtain
at the front (flame end) of stainless steel chains. The chain system extends
-through 87’ of kiln length. The slurry’ feed system is a bucket wheel con-
veyor with a variablespeed drive taking slurry from-a constant level box.
Gases from each kiln (maximum capacIty 1,50,000 CFM at 450°F) pass through
a six—section electrostatIc precipitator. Gases from the precipitators
are exhausted via a coninon stack 554’ in height with 13’ exit inside dia-
meter. -No. 6 fuel oil Is burned in a single burner at the centre of the
burner pipe. For the test, chlorinated hydrocarbons were fed just above
and to one side of centre using different size nozzles for proper atomi-
zation at different flow rates. A detailed description of the chlorinated
hydrocarbon system is given In Appendix E.
2.3.2 Relevant unit processes - suspension preheater kiln
The kiln (FIgure 2) is a 17’ x 276’ Traylor unit normally fired
through three nozzles with No. 6 fuel oil. For the test, chlorinated
hydrocarbons were injected via a nozzle at the centre of the triangle
formed by the three oil nozzles.
The suspension preheater (Figure 3) consIsts of a system of
cyclones through which hot exit gases from the kiln are drawn by a fan.
The raw meal passes through the system in cóunterf low to the gas. Kiln
feed is introduced into the duct between the first and second stage cyclones.
It is swept with the hot exhaust gases into the uppermost (Stage I) cyclones

-------
I
1. RAW MEAL FEED
2STAGE I
3 STAGE II
6 STAGE III
5 STAGE IV
6
I
8
9
10
11
12
13
14
KILN
KILN EXH. FAN
PRECIPITATOR
DUST RETURN
CONDITIONING TOWER
BY-PASS PRECIPITATOR
OUST DISPOSAL
OUST DISPOSAL
WATER
MATERIAL FLOW
GAS FLOW
PRINCIPLE OF FULLER-HUMBOLDT
SUSPENSION PREHEATER
AND BY-PASS
FIGURE 3.
©

-------
8
where gas and material are separated. The raw feed from the cyclone drops
into the duct between the second and third stage cyclones and is again
suspended and separated . This procedure, being swept up with hot gases
and then being dropped into tt e stream entering the next lowest stage,
is repeated in Stages l l I and IV before the partially calcined feed enters
the kiln. The average retention time of the system is approximately
30 seconds. A detailed description of the Humboldt preheater has been
given by G.A. Schroth (13] from which article these notes have been pre-
pared. The raw material entering Stage I is preheated to approximately
300°C (600°F), while the gas temperature drops from 530°C (990°F) to 340°C
(650°F). At each stage, corresponding heat exhanges occur such that the
material enters the rotary kiln at approximately 800°C (1475°F) having
been partia lly decarbonated. The gas temperature at the point of exit
from the kiln into the preheater Is 10 40°c (1900°F) to 1090°C (2000°F).
The St. Lawrence Cement Co. preheater is a dual Fuller—Humboldt unit
with four stages in each.
ihe alkali bypass system-(Figure4) Is an important unit
in relation to this,’study. in convnon with other plants using suspension
preheater systems, special measures have to be taken to reduce the buildup
of chlorides and alkalies in the system. At the St. Lawrence Cement Co.
plant, a system is in use where a fraction of the kiln exhaust gases is
removed from the preheater and passed through a conditioning tower. In
the conditioning tower, water Is sprrayed into the gas stream to lower the
temperature and condition the gases for precipitation. Concurrently, the
gas velocity is reduced, since the cross section of the tower is greater than
that of the bypass duct. The net effect of cooling and velocity reduction
is to divide the particulate matter carried in the gas stream into two
fractions. One fraction., of lower alkali content, Is separated In a condi-
tioning tower and returned to the raw meal silos. The other fraction, of
higher alkali content, Is collected In an electrostatic precipitator, pel-
letized and discarded.

-------
1 PREHEATER TOWER BY-PASS HOOD
2 QUENCH AIR
3 CONDITIONU1G TOWER
4 PRECPFFATOR
5 EXHAUST FAN
6 WATER
7 AIR
8 650°F CAT OFF CONTROL
9 AIR QUENCH NOZZLE CONTROL (4 °F)
C CONTROLLER
PT PRESSURE TRANSMITTER
M MODULATOR
L___ _J ,

‘—1//I
-a®
• 1
I
ALKALI BY-PASS
FIGURE 4

-------
10
3 CONSIDERATIONS AT THE PROGRAM PLANNING STAGE
The history of suspension preheater kilns In the United States
indicates a trend to this system [ 14]. Thirteen preheater kilns were
installed in the years 1953-1955. Because of Inadequate knowledge about
the basic process and effects of raw-materials on this system, numerous
operating difficulties were encountered. These difficulties, possibly
combined with emphasis on alkali reduction, caused the shut down of six of
these units. From 1956 to 1969, only two suspension preheater kllns
were Installed, one of which has slnce’.been shut down. However, with
increasing fuel costs and-emphasis on energy conservatIon, 22 suspensIon
preheater kilns have been comm lss!oned since 1970. Another indication
of this trend is that five of the ‘eleven new.kl ins Installed in
1975 were, preheater—rotary kIlns: [ l5 ]. l,n addition, three existing
rotary kilns were converted to suspension preheater units. All five
kilns planned for completion In- 1976 wIll be preheater kllns. While the
situation in Canada is slmlier there’are fewer plants and such trends
are difficult to follow. --
Any study concerne with future use of the technology developed
must take into consideration the suspet sIon preheater kiln.’ Flame
characteristics-are Identical In a!l;cement kiln systems, hence,
demonstration that destruction of chlorinated hydrocarbons occurs In
one cement kiln implies destruction In-all cement kl lns. Reduction of
alkalies in a straight rotary kiln system by addItion of chlorides
(calcium chloride and hydrochloric acid) has been extensively demonstrated.
It was decided to conduct the experiment In the suspension
preheater kiln despite knowledge that the preheater Is prone to plugglnç
problems at high alkali chloride levels.
The primary purpose of the study was to determine whether
chlorinated hydrocarbons were destroyed In a cement kiln. It was
considered desirable to conduct the experiment ln stages with materials
of increasing difficulty of destruction. After each stage, ‘normal
operation was resumed, thus giving time to analyze samples in order to
determine that no waste chlorinated hydrocarbons were emitted. For
this study, it was agreed to use four different formulations of Industrial
chlorinated wastes:

-------
I)
1) aliphatics;
2) aromatlcs;
3) allphatics plus aromatics and allcycHcs; and,
4) allphatlcs, aromatics, alicyclics and large complex
molecules such as polychiorinated blphyenyls (PCB’s).
Generally these are filtered, processed and blended to obtain
specific formulations for control of energy value, chlorine content and
viscosity.

-------
12
ii TRIAL ON THE SUSPENSION PREHEATER KILN
The first phase of the trial was done in the suspension preheater
kiln to determine whether alkali reduction could be accomplished on
this unit by burning of chlorinated, hydrocarbons. Reduction of alkalies
by addition of chloride is achieved by -increasing the amount of alkali
volatilization. Alkali carbonates and chlorides are almost entirely
volatilized at the normal operating temperature of the kiln. Alkali
suiphates and alkalies which are complexed in the calcium silicate and
aluminate structures are preferentially retained In the clinker.
The addition of chloride Ion causes formation of the easily
volatilized alkali chloride. Intimate mixingof hot kiln gases and
countercurrent cooler raw meal causes alkali chlorides to condense on
the raw meal thereby being returned to the kiln. Alkali chlorides are
trapped between the burning zone of the kiln (volatilIzation) and the
bottom stage of the preheater (condensation), causing Increasing concentra-
tions-in the gas stream. At. high-concentrations ofalkali, chlorides 1
sufficient quantities condense in Stages III and IV of the preheaterto
cause plugging of this system. ro al ,lévlate’this situation, the bypass
s used to withdraw a fraction of the gases leaving the rotary kiln
and with these gases a fraction of the alkali chlorides. As greater
quantities of alkali chloride are volatilized In the kiln, high with-
drawal rates through the bypass are required.
Samples of the raw meal entering the kiln from the preheater
can be taken to monitor the alkali chloride concentration. Prior to
the burn, it was estimated that equilibrium alkali chloride concentration
at this point would be reached In approximately two hours.
Gases exhausted via the bypass have a temperature of approx-
imately 1000°C (1830°F). To cool these gases, they are mixed initially
with ambient air, then passed through a conditioning tower In which water
spray is used for further cooling and conditioning of the gases.
In the conditioning tower, the coarser fraction of the dust, which has the
lower alkali concentration, settles to the bottom of the conditIonI ig
tower and is returned to the process. The finer fraction, which has the
higher alkali concentration, Is carried with the gas stream to the bypass
precipitator. Dust collected from this precipitator Is discarded.

-------
13
Initially, attempts to operate the bypass at the level required
to compensate for addition of chlorides resulted in high emission rates of
particulate matter. Weighings of materials showed that 50—60 tons per
day of dust were collected In the precipitator (normally 6-10 tons per
day), while k tons per day settled in the conditioning tower (normally
20 tons per day). The reversal of solids fraction removed In the two
units and the larger amount of material carried forward to the electrostatic
precipitator was attributed to the Increased quantity of ambient air used
for cooling the bypass stream. This increased gas flow gave higher
velocities in the conditioning tower, lowered Its settling efficiency
and resulted In particulate entrainment in the gas stream.
The burning of chlorinated aliphatic material was begun on
June 3, 1975 wIth a mixture having a specific gravity of approximately
1.2 and a chlorine content of about 55% w/w.
On June 5, the bypass duct-between the kiln and the conditioning
tower was plugged. The trial was suspended while the duct was cleaned.
A second attempt was begun on June10 using the same parameters- used in
the June 3 trial. On June 13, similar,buildup again caused the bypass
to fail. it was apparent that the addition of chlorides to the suspension
preheater kiln could not be done without major portions of the bypass system
being rebuilt.
Due to high particulate emissions, plugging of the bypass
system and the cost of the equipment changes envisaged to alleviate the
short comings, it was decided to continue the trial in one of the two
existing wet kilns.
The trial on the dry process kiln did show, however, that
chlorinated hydrocarbons were destroyed in this kiln. Chlorinated
hydrocarbons used are volatile at temperatures found In Stage IV (980°C),
and could not have condensed upon the material. The expected Increase in
chlorine concentration of Stage IV dusts was confirmed by analyses
(Table B.4). Equilibrium concentrations of chloride at Stage IV were
obtained in approximately four hours (Appendix B). Some slight alkali
reduction was apparent. During the five days on which chlorinated
hydrocarbons were burned, the raw meal feed 1(20 on ignited basis averaged
1.47%, while the average clinker 1(20 was 1.27%. With the same quantity

-------
14
of gases being removed bythe bypass, but without addition of any chlorinated
material, the raw meal feed contained on average l.4i % 1(20 and the
clinker contained on average 1.29% (20. WhIle the difference Is slight,
the duration of the experiment was short.
Complete details and analytical results from the experiment
on the suspension preheater kiln are given In Appendix B.

-------
15
5 TRIAL ON THE WET PROCESS KILN
5.1 DIscussion
The test program for this section of the study was designed to
determine if kiln emissions contained any chlorinated organic compounds
when chlorinated organic wastes were burned as supplemental fuel to reduce
the alkali content of the clinker.
From a knowledge of kiln zone temperatures and gaseous product
retention times In the kiln, It was anticipated that combustion of the
waste chloride fuel would produce predominantly C0 2 , H 2 0, MCI (hydrochloric
acid), C1 (chloride Ion) and a negligible amount of free Cl 2 (chlorine
gas).
H 2 0 + Cl 2 2HC1 + Oz
HC1 2 o
Kp- 2
0 Cl
2 2
15 at 1000°C(l832°F)
— 68 at 1500°C (2732°F)
—150 at 1500°C (3452°F) (28].
This shows that amount of free chlorine gas decreases with:
a) increase In temperature;
b) increase In water vapor content;
c) decrease in oxygen content; and,
d) decrease or removal of hydrochloric acid.
The hydrochloric acid will react and the chlorine will be
retained as-alkali chlorides In the process soUds. From a
knowledge of orthodox incineration systems, the combustion conditions
in a cement kiln were expected to provide a very favourable means for
the destruction of waste chlorinated hydrocarbons. in order to determine
the effect, If any, of burning waste chlorinated organic material on
air quality, emissions from the kiln were monitored before, during and
after three separate periods of burning the supplemental waste fuels.
Prior to each sampling period, emissions were checked for the presence of

-------
16
HC1 and Cl 2 with detector tubes at sensitivity limits of 2 ppm and 0.5
ppm, respectively.
A mass balance for chlorine was carried out by analyzing all
feed and process samples collected during each waste burn for this
element, to confirm the emission data.
Prior to the trial on the wet process kiln, light oil was put
into the chlorinated feed system for the purpose of calibrating and
testing the proportioning and metering devices. Some light oil remained
in the system when chlorinated aliphatic material was received, causing
the chloride content to Increase as the test progressed. While emissions
were tested, chlorinated aliphatic material was burned at different
rates while attempting to con ensate for the changing chloride content,
the rates being equivalent to:
October 28, 1975, 0.3l chlorine relative to clinker;
October 29, 1975, 0.32 chlorine relative to clinker; and
October 30, 1975,0.63% chlorine’ relative to clinker.
- Dueto excessive quantities of chlorine Input, a ring was
formed, which required akiln shutdown.for its removal. A ring is
caused by the buildup of material onthe Inside of the kiln to-such
an extent that it restricts the flow of process materials.
Due to production scheduling, the delay caused by the shutdown
made it necessary to drop the burning of aromatlcs alone and Imedlately
progress to the mixture of aromatic andcomplex molecules. The conduc-
tivity of this latter material was too low (0.3 x 1o6 mIios) for the
magnetic flow meter to function properly.
Control of flow rates was therefore achieved by measuring the
depth of liquid In the tank. While accurate over a long period of time,
short term control was difficult because one Inch of liquid In the tank
is 196 imperial gallons. Variable flow rates were encountered during
emission testing, the quantity added each day being:
December 10, 1975, 0.45 to 0.71% chlorIne relative to clinker;
December 11, 1975, 0.31 to0.5l% chlorine relative to clinker; and,
December 12, 1975, 0.79% chlorine relative tO clinker.

-------
17
A ring also formed during this burn but broke away after the
sampling sequence had been completed, thus avoiding another kiln shut-
down. The presence of this ring restricted the burning of the residual
quantity of this material (aromatic plus complexes) to intermittent burns.
Due to the above mentioned upset and requirements of the
production schedule, it became necessary toadd 13,000 gallons of the
polychlorinatecl biphenyl blend to the 12,000 gallons of material remaining
in the tank. The resulting mixture contained a significant quantity of
solids and also had very low conductivity, thus requiring tank measure-
ments as the method used for flow control.
The problem with solids separation and settling became more
severe as the test progressed. This caused burner nozzle restrictions
and pluggages which, in turn, caused Interruptions and Irregularities in
waste liquid flows.
For the last day of emission testing, the nozzle was removed
completely to permit uninterrupted flow rates of polychlorinated biphenyls.
Quantities added:duririó emission tests were:
January 7,1976, 0.06 to 0.14% chlorine relative to clinker;
January 8, 1976, 0.13 to 0.33% chlorine relative to clinker; and,
January 9, 1976, 0.61% chlorine relative to clinker.
5.2 Organic Chloride Waste Burned
The chemical compositions of the three waste materials burned,
the labelling used to identify each burn and the-time periods during
which they were fed to the kiln are given below:
Chlorinated aliphatics WBA October 23 - November 4/75
WBA plus chlorinated
aromatics and alicyclics - WBB December 5 - December 15/75
WBB plus poly-
chlorinated biphenyls
(PCB’s) WBC January 3 - January 9/76
It was anticipated that complete combustion of these materials
would occur in the kiln. However, a knowledge of the major constituents
of each of the fuels was necessary In order to determine which trace
components might be detected in the sampled emissions should incomplete
combustion occur. Previous studies on the incineration of waste

-------
18
chlorinated hydrocarbons (16] had shown that trace quantities of low
molecular weight compounds such as carbon tetrachioride (C d 4 ), chloroform
(CHC1 3 ),dichloranethane (CK 2 CI 2 ) were detected in the combustion product
gases. It was decided, therefore, to analyze collected emission samples
for the presence of the major components of each composite waste fraction,
and for the low molecular-weight organochloridesreferred to above.
Samples of the chlorinated wastes were taken from the line
between the tank and the kiln. Typical component analyses of the three
composite wastes burned during the program are recorded in Tables 1, 2
and 3.
The quantitative values given in the tables are based upon an
assumed equal response of all the components to a flame Ionization
detector (F1D). While this assumption is riot very accurate, It was
considered sufficient to provide the needed compositional data. identi-
fication of components at concentrations greater than one percent was
achieved by use of-gas chromatography—mess spectrometry (GC-MS). Identi-
fication of components at concentratIoris 1e’s’s thón one percent wasnot-.-
routinely performed. .Anaiysesof samplès collected on different days during-
a burn showed some variation in the concentration of individual components.
However, the overall chemical composition of the feeds remained the same.
Two of the waste burn sample feeds (WBB and Wad) were also
analyzed by the Ontario Ministry of the Environment (MOE), and the TRW
Systems Group. The results obtained by these agencies are reported in
detail in Appendices F andG, respectively. The TRW Systems Group analyzed
-composite samples of the daily feeds for both waste burns by GC—MS. The
MOE group analysed individual daily samples for both waste burns using
GC-FID operating conditions. The chromatographic parameters used by
the MOE group to analyse for aromatic and PCB components precluded the
Identification of aliphatic organochlorine compounds such as CHC1 3 and
cc’ 4 .
Some variation in the percent composition of the waste burn
samples was apparent between the TRW and ORF results, especially with
respect to the amounts of aliphatic organochiorine compounds present
In waste burn B (WBB). The discrepancies observed are due to the fact
that TRW received a composite sample of all daily feeds for each of the

-------
19
TABLE 1. COMPOSITION OF ALIPHATICS ( rnA)
Sample Feed - October 28, 1975
.
Approximate
Peak #
in
GC
Profile
Concentration
(cf
Figure
A.6)
Identification
1 17.4 Chioropropane s propene
2 1.4 ethy lchloride,
di chloromethane
3 0.8
4 1.5 Chlorobutane & butene
5 3.2
6 16.5 1, 2, Dlchioroethane
7 6.6 Carbon tetrachiorde
8 1.0
9 10.5 1, 1, 2—Trlchloroethane
10 0.7
11 0.3
12 1.9
13 1.4 D lchioropropanes
14 1.4
15 2.4 Tetrachloroethylene
16 7.3 Tetrachloroethane
17 15.6 Chiorobenzene
18 2.4
19 0.2
20 3.3
21 0.3 multichiorinated
22 1.1 butanes, butenes
23 0.6 hexanes, hexeries
24 0.5
25 1.6
26 0.6
27 0.7
Note: No identification of-compounds at concentrations of 1 or less was
attempted.

-------
20
TABLE 2. COMPOSITION OF AROMATICS PLUS COMPLEX (WBB)
7 0.1,
8 hi
11 0.2
12
‘3
14
15
16 52.2
17 0.2
18 0.1
19 0.1
Chloroform
1, 2-Dlchloroethane
Carbon tetrachioride
Hexach 1 orocyc 1 open tad I ene
Heptachiorocyclopentene
Pentach lorobenzene
Octachiorocyclopentene
Complex associated with
Hexachiorocyclopentad u ene
Note: No identification of
attempted.
compounds at concentrations of 1% or less was
Sample Feed - December 12,1975
Peak // in
GC
.
Profile
- Approximate
Concentration
(cf
Figure
A.7)
•
Identification
1 2.0
2 0.1
3 0.1
4 1.1,
5 1.5
6 3.5
Chioropropane
9
1.9
1, 1, 2—Trichioroethane
10.
0.2
0.6
-
0.5
Chlorobenzene
0.5
-
20
6.3
21
2.7
22
0.1
23
1.0
24
85
25
4 .1
26
6.9
27
1.3

-------
21
TABLE 3. COMPOSITION OF AROMATICS PLUS PCB’s (WBC)
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Sample Feed - January 8, 1976
3.4
0.3
1 .4
1.3
0.5
0.3
1.9
0.1
1 .9
0.5
28 5
< 0.1
< 0.1
1 .8
0.9
2.8
0.9.
2.2
0.3
2.1
0.9
6.7
3.4
5.9
1 .6
12.1
5.6
4.0
5.4
2.1
1 .6
Note: No identificatIon of compounds
at concentrations of 1% or less was
Peak I in
GC
Profile
Approximate
Concentration
(cf
Figure
A.8)
Identification
Ch loropropane
1, 1 , -Dlchloroethane
Carbon tetrachloride
1, 1, 2-Trichloroethane
Chiorotol uene
Hexachlorocyclopentad iene
.4 complex -
Dichiorob iphenyl
Tr lchlorobipheny l
Tetrachiorob I neny1
Pentachiorob i phenyl
I
I
}
attempted.

-------
22
waste burns, and different chromatographic cblumns and a different detector
system were used by the two laboratories.
Since the study was to evaluate the burning of waste chlorinated
materials of variable composition, minor differences between the
laboratories’ results do not affect the conclusions of the study.
The agreement among all groups was good with respect to the
identification of components present in the two waste fuels. The presence
of chlorinated aliphatics was observed in both waste fuels. WBB samples
were found to consist primarily of chlorinated aromatic compounds, in
particular o-cMorotoluene. WBC samples were found to consist of approx-
imately 50% polychiorinated biphenyls (Aroclor 1242) plus the chlorinated
aromatics and alicyclics found in WBB samples.
5.3 Emissions
Samples of the kiln emissions before (BL.A), during (WBA WBB, WBC),
and after (BLB). the burning of waste organicchlor ldes were taken,using.
-the equipment and methodsdesçribed in-Appendix A. -The analytical -
techniques used to determine ifany organic chloridecompounds-were
present in the emissions samples are.discussed with respect to both gaseous
and particulate emissions.
5.3.1 Free chlorine and hydrogen chloride
Prior to each sampling period, analyses ware made for the presence
of free chlorine (Cl 2 ) and hydrogen chloride (HCI) in the kiln emissions.
An MSA gas sampler and l iSA detector tubes were used to determine the
concentrations of these pollutants. In no instance, either during baseline
or waste burn sampling periods, were Cl 2 or HC1 detected. The sensitivity
limits using this analytical procedure are -0.5 ppm for Cl 2 and 2.0 ppm for
HC1.
5.3.2 Gaseous organic compounds
- Emission samples were collected using two sampling trains. The
first, a gaseous sampling train (Appendix H), was designed to collect any
volatile low molecular weight organic chlorides, e.g. Cd 4 , CHC I 3 , Ch 2 Cl 2 ,
by adsorption on an inert adsorbent. In this study Chromosorb 102 was
used. The second, a particulate sampling train (EPA Joy #5), was expected

-------
23
to collect higher molecular weight compounds, Including any noncombusted
starting materials in the impingers.
Components adsorbed on the sorbent were removed by thermal
desorptlon prior to analysis. All other collected fractions, from both
trains, were extracted with organic solvents prior to analysis. These
included probe rinse water solubles and insolubles, filters and impinger
solutions.
In considering the results obtained in this study It is
important to keep in mind the following:
- The collection efficiency of Chromosorb 102 sorbent for low
molecular weight chlorinated hydrocarbons was determined to
be better than 90%. The collection efficiencies of filters
and ice-water in the impingers of a Joy train for organo-
chloride components was not determined.
- The methods of collection, followed by thermal desorption or
solvent extraction, were designed to concentrate any organic
compounds which may have been present. Such techniques allow
detectlän and Identification of compounds at much lower levels
than would otherwise be possible. Concentration factors for
each type of sample have been calculated for each phase of the
burn and are given in Table 4. For the desorbed gas samples,
organic compounds from several cubic feet of kiln emission
gases were collected on Chromosorb 102 and then desorbed into
500 millilItres. The concentration factors are the ratios
of these two volumes. In the case of the solvent extracted
samples, molecular weights enter the calculation In the conver-
sion of volume to weight. A complete discussion of these
calculations is given in Section A.6.4.
5.3.2.1 Desorbed samples . The concentrations of volatile organic
chlorides calculated for the kiln emissions from gas chromatography -
electron capture (GC-EC) are recorded In Table 5.
The results reported were averaged for each test series.
Dichloromethane (0CM) was tentatively Identified as the major component
of these desorbed gas samples. Others tentatively identified were CHCI 3

-------
-24
TABLE 4. GAS SAMPLE VOLUMES AND SAMPLE CONCENTRATION FACTORS
(Gaseous Sampling Train)
Test
Volume*
Sample Concentration
Factors
.
Duration
Sampled
Desorbed So lvent**
Test # -
Date - -
(mins)
scf
- Gas Extracted
I BLA
Oct.
20 (1975)
317
2.80
159
2 BLA
Oct.
21
281
i:.86
141
510
3 BLA
Oct.
22
320
2.83
160
1 WBA
Oct.
28
362
-3.22
181
2 WBA
Oct.
29
286
2.53
143 500-1100
3 WBA
Oct.
30
285
2.53
143
1 WBB
Dec.
10
260
2.30
130
2 WBB
Dec.
11
317
2.80
159 600-2000
3 WBB
Dec.
12
‘250
2.21
125
1 WBC
2 WBC
- Jan.
Jan.
7 (1976)
8
268
-237
2.37
2;09
134
119 520-1500
3 WBC
Jan.
9
228
2.02
114
1 BLB
Jan.
19
170
1.50
85
2 BLB
Jan.
20
247
2. 18
124 380
3 BLB
Jan.
21
-- 240
2.12
120
.
*Volume sampled per adsorbent tube.
**These values are based on the lowest and highest molecular weights of
compounds found in the waste feeds. Figures given for-both baseline
series are based upon the molecular weight of diehloromethane (DCM).

-------
25
TABLE 5. ESTiMATED KILN EMISSION CONCEN1*MI N5 th -t )
FOR SPECIFIC VOLATILE ORGANOCHLORINE COMPOUNDS
Test
Series
Emission Concentrations
Dichioromethane
ppb ug/m 3
Chloroform
ppb ug/m 3
Carbon
Tetrachloride
ppb ug/m 3
BLA
4.1 14.5
0.004 0.020
0.0004 0.0026
WBA
5.4 27.3
0.015 0.080
0.0020 0.0128
WBB
18.0 19.1
0.038 0.190
0.0020 0.0128
WBC
7.7 102.7
0.069 0.345
0.0060 0.0385
BIB
29.0 63.7
0.018 0.090
0.0004 0.0160
and CC1 4 . It was found that GC-EC profiles of blank determinations
obtained with unused conditioned Chromosorb 102 were variable, and
the peaks that were present In the profiles possessed similar retention
timei tO the compounds of interást.’ “Therefore, reported results are’.
higher than actual concentrations In the emission gases. The differing
concentrations for 0CM quoted In Table 5, especially with respect to BLB
versus BLA, WBA or WBB, are probably a result of this background contamina-
tion rather than real differences. it was still noted, however, that
the maximum concentration calculated as 0CM In the kiln emissions was
no greater than 30 ppb even with this positive bias from background
contamination.
The MOE results confirm the ORF findings with respect to the
detection of low molecular weight chlorinated organlcs in the desorbed
gas samples. TRW did not find any trace of these volatile chlorinated
organics in their desorbed gas samples. However, TRW did not routinely
perform specific analyses for compounds present at concentrations below
0.1 mg/rn 3 .
5.3.2.2 Organic solvent extracted samples . Due to the much higher
gas flow rate through the particulate sampling train to maintain an
isokinetic sampling rate, the concentration factors for solvent extracted
samples for this train were greater than those collected with the gaseous
train. Calculated average concentration factors for each test series are

-------
26
given below:
BLA 7,000
WBA 6,000 - 13,000
-WBB 7,000 - 23,000
WBC 7,000 - 20,000
BLB 6,000
For the waste burns, the two values given represent the highest
and lowest molecular weight species present In the waste chlorinated hydro-
carbons. Calculations for the baselines are based on the molecular
weight of dlchloromethane. Subtraction of background interference based
on control blanks, (Appendix P.6.3) was performed on solvent extracted -
samples. Comparison was then made with the appropriate waste feed sample
chromatograms. It was concluded that no uncombusted waste fuel components
were present -in any of the organic solvent extract samples at the
detection limits of the analytical procedure used. Representative
chromatograms together with discussion are presented In Appendix A.6.5.
- TRW performed analyses onp rtlonsof all the organic solvent
-extracts obtained from BLA, WBB,WBC and BLB emission samples. The
samples were specifically searched for chlorinated compounds. Also,
the Chemistry Division, Air PollutloW,.Control Directorate (APCD),
Department of the Environment, Ottawa, Canada carried out analyses
on portions of the solvent extracts of WBC emission samples to search
for high molecular weight chlorinated hydrocarbons. Detailed reports of
the work undertaken by these groups are presented in Appendices G and I,
respectively.
The noteworthy result with respect to both reports was that In
none of the samples analyzed by either group were any chlorinated organic
compounds detected. Their results thus confirm the ORF conclusion that no
chlorinated organic residues were detectable in any of the organic solvent
extracts of the emission samples collected during the various waste
chlorinated hydrocarbon burns.
5.3.2.3 Results obtained by the particlpatin laboratories . In gas
chromatographic analyses, program parameters define the conditions used.
The choice of detector and column makes the analysis specific to certain

-------
27
groups ot compounds. Four laboratories analyzed the emissions samples
taken during this study. -
One of the laboratories, APCD, was requested to search for
high molecular weight chlorinated hydrocarbons. The TRW group were not
requested to identify low molecular weight chlorinated compounds present
at less than 0.1 mg/rn 3 in the stack gases. TRW did, however, search for
PCB’s using techniques designed to detect small quantities of these
compounds. ORF and MOE were assigned the task of detecting and identify-
ing all possible chlorinated organic species. Analytical design, and
hence results, reflect differences in the analyses requested of the
different laboratories.
Both ORF and MOE found low molecular weight hydrocarbons,
such as dichioromethane, to be present at microgram per cubic meter
(ug/m 3 ) levels in the emissions. That TRW detected no such compounds
is not contradictory. The levels in the emission gases were less than
their required detection limits. Similarly, it was not within the terms
of reference for the APCD work to determine these compounds.
Results of- the four laboratories can. b convenientl sunii arized.
While burning chlorinated hydrocarbons, low molecular weight chlorinated
compoundswere emitted at levels-of a few ig/m 3 (Table 5). None of the
participating laboratories detected any high molecular weight chlorinated
hydrocarbons from either air sampling train. At detection limits of
3 iig/rn in the stack gases, polychlorinated biphenylswere not found.
5.3.2.4 Estimated minimum combustion efficiencies . The maximum value
for total chlorinated hydrocarbon content in the kiln emissions was
obtained in test 2 of the WBC series. ignoring background subtraction
for the baseline samples and the Interferences from control blanks a
maximum value of 40 ppb was detirmined. - If a collection efficiency of
8o is assumed, a maximum level of 50 ppb in the kiln emissions is
obtained. Because of the high and uncertain background levels, the
estimate is higher than actual-levels In the emission gases.
‘An average molecular weight for each of the three composite
feeds can be obtained from a knowledge of the composition of the feed.
Using this information, together with the average fuel feed rates to
the kiln and average gas volume flow rates in the duct, the maximum

-------
28
chlorinated hydrocarbon content can be, determined. These values-are
presented InTable 6. Minimum combustion efficiencies were then calculated
using the estimated maximum value of 50 ppb derived above. These figures
in Table 6 are considered very conservative estimates because of the
method used to calculate them.
TABLE 6. ESTIMATED MINIMUM COMBUSTION EFFICIENCIES
FOR EACH WASTE BURN
Waste
Maximum
Content
Eat
in
Imated Organic
Kiln Emissions
Minimum
Combustion
Assuming no
-
Rrom..5an le
Composite
Combustion 3
ppm 9/rn
•
Chromatogram
ppb pg/m)*
Efficiency
v/v
%
WBA
-550 2.40
50 177.1
99.990
WBB
470 3.37
50 177.1
99.989
- WBC
350 3.02
-
.
50’ .177..l
99.986
*Based on the molecular weight ofdichloromethane.-
5.3.3 Particulate emissions
Sunvnary data for the particulatetests made during each test
period are provided in Table 7. For each series of tests, the particulate
loadings and emission rates are quite consistent except for the third
test in the WBA series. The very high.loading obtained probably resulted
from some temporary malfunction of the, precipitator. it is noticeable
that, when the chloride wastes were burned.in the kiln, the particulate
emission rate increased. For the WBA series, the emission rates were
about four times the rates obtained during the baseline tests and for
the WBB and WBC series the rates were twice those obtained during the
baseline tests.
Increased emission rates were not unexpected since combustion
of the chloride wastes produces HC1 a d Cl 2 , which react with the alkali
components in the raw feed to form volatile alkali chlorides. At tt
precipitator the particulate loading is, therefore, increased and, since
the condensed alkali chlorides are very fine and have a different
resistivity, the amount of material passing through the collector
increases. Mother factor influencing emissions of particulate matter

-------
TABLE 7. SUMMARY OF PARTICULATE TEST DATA
Probe
Rinse
Filter
Total
%
of
Volume
Emission
Gain
Test I Date (mg)
Gain
(mg)
Gain
(mg)
Total Gain
on Filter
Sampled
(std ft 3 )
Flow rate
ACFM
Concentration
(gralns/ft 3 )
Rate
(lb/hr)
I
BLA
Oct.
20/75
57.6
285.7
3’e3.3
83
l el.O e
157.000
0.0376
20.8
2
BLA
Oct.
21/75
58.8
366.0
421i.8
86
135.86
155,000
0.0483
25.4
3
BLA
Oct.
22/75
35.8
291.8
327.6
89
132.83
153,000
0.0367
$9.6
Average
0.0409
21.9
I
WBA
Oct.
28/75
99.0
1242.6
1341.6
93
142.81
157,000
0.1458
83.9
2
WBA
Oct.
29/75
94.8
1261.7
1356.5
93
137.34
153,000
0.15211
.
84.6
3
WBA
Oct.
30/75
237.5
2193.8
2431.3
90
309.88
165,000
0.3435
200.0
Average
0.2132
122.8
1
WBB
Dec.
10/75
143.0
424.1
567.1
75
106.64
148,000
0.0821
40.3
2
WBB
Dec.
11/75
86.1
468.4
554.5
84
117.06
362,000
0.0731
40.5
3
WBB
Dec.
12/75
145.6
621.6
767.2
81
116.211
161,000
0.3019
55.0
Average
0.0857
45.3
1
WBC
Jan.
7/76
62.2
546.3
608.5
90
119.62
172,000
0.0785
45.3
2
WBC
Jan.
8/76
43.1
436.5
479.6
91
113.52
160,000
0.0652
34.5
3
WBC
Jan.
9/76
87.8
600.0
687.8
87
118.99
167,000
0.0892
52.2
Average
0.0776
44.0
I
BIB
Jan.
1!/76
65.2
251.7
316.9
79
126.88
183,000
0.0385
23.9
2
BIB
Jar 1 .
20/76
53.4
240.1
291.5
82
116.60
163,000
0.0386
23.0
3
BIB
Jan.
21/76
11.9
183.5
195.4
94
116.45
$68,000
0.0259
14.6
Average
0.0343
19.8
p .,
0

-------
30
were the kiln rings formed while burning chlorinated hydrocarbons.
Kiln rings, by their effect on gas velocities through the kiln,
increase the amount of particulate material carried to the precipitator.
Extensive ring formation was noted during test WBA. The average
emission rate for WBA was about 3 lb/ton of clinker produced, compared to
the Canadian Federal Government objective of 0.9 lb/ton. The average
value for the other phases of the test was about 1.1 lb of particulate
emitted/ton of clinker produced and that of the baselines was 0.5. ThIs
apparent increase In the particulate emissions while burning chlorinated
hydrocarbons Is believed to be partially caused by the change in resis-
tivity of the dust entering the precipitator due to Its Increased
alkali content. A modification In the design of the precipitator could
compensate for this change.
The higher emission rates did not significantly add to the
suspended particulate in the ambient air In the vicinity of the plant.
Computed maximum ground level concentrations using standard dispersion
equations were, less than 2 pg/m 3 for the’basellne emission rates and,
during the burning of waste chlorIdes, did’ not exceed 10’ ig/m 3 . The
current Ontario standard Is 100 iig/m 3 .
5.k Mass Balanre on Wet Kiln
For the wet process kiln, the Input streams are slurry feed, No.
6 fuel oil and, when burned, the waste chlorinated hydrocarbons. Clinker
(the product) and a portion of the precipitator dust form the output
streams. The major portion of the precipitator dust Is ininediately
returned to the process. This stream, though not required for the mass
balance, was monitored for information on kiln operatIon while burning
chlorinated hydrocarbons. Balances of chlorine and potassium were
calculated for both baseline periods and for each type of chlorInated
waste burned. Methods of sampling and quantifying material streams are
detailed in Appendix A. Details of analytical results, calculations and
tables of daily mass balances are given In Appendix C.
5.4.1 Significance of the mass balance
In common with other material balance experiments on large
scale production systems, the material accounting in this experiment showed
apparent losses and gains when the data were expressed in percent retention.

-------
31
These should not be regarded as real system losses or gains. There are
random and systematic errors in weighing and quantifying very large
quantities of materials and In analytical results. In the case of
alkali chlorides, a cycle is developed within the kiln whereby alkali
chlorides volatilize In the hotter section of the kiln and condense In
the cooler section. Should a kiln upset occur, the cycle may be broken
by alkali chloride being retained In the clinker. Such an occurence may
be of sufficiently short duration that clinker samples taken may not be
indicative. This effect Is evident from the data of October 1114, 1975.
A power failure on the eleventh caused a three hour kiln shutdown. The
Indicated very low retention of both chlorine and potassium occurred
presumably while the cycle was being re—established. Since the equilibrium
changes when burning of chlorinated materials begins or ends, this feature
can also be seen with each phase of chlorinated waste burning. At the
start of any chlorinated burn, the mass balance for chlorine indicates
very low retention while the cycle Is established. At the end of the
chlorinated burn, a gain’ is indicated as chlorine is retained in ‘process
solids while the new equilibrium Is established. Formation of kiln rings
also caused an apparent loss of alkali chloride since a considerable quantity
was contained within the ring.
5.4.2 ChlorIne and potassium retained
The cumulative percentage of chlorine retained (Table 8) was low
in all cases throughout the study, ranging from 50.7 percent for the
chlorinated aliphatic burn to 92.2 percent for the final baseline. (20
(Table 9) closely followed the pattern of chlorine retention, although
at a different retention level, with 83.5 percent the lowest value found
during the chlorinated aliphatic burn and 97.8 percent the highest value
found during the final baseline. The reason Is clearly that potassium
chloride was being lost in ring formation and kiln cycle equilibrium
fluctuations.
The cumulative percentage 1(20 retained was less affected than
that of chlorine by such losses because the quantity of K 2 0 was f. m 3 to
20 times greater than the quantity of chlorine. Random and systematic
errors could also be expected to play a greater role in the chlorine
balance for the same reason.

-------
TABLE 8. ACCUMULATED MASS BALANCE FOR CHLORINE
p1
1 . . )
.
Accumulated Period
Chlorinated Accumulated - * of I
Hydrocarbon Input (lb) From Chlorinated Accumulated
Burned (Ti) Hydrocarbon Retention (lb)
‘
% Retention
7/10/75-22/10/75
23/10/75- 4/11/75
--— 28,249 0 20,768
Aliphatic 134,379 83.2 68,088
73.5
50.7
2/12/75-14/12/75
3/ 1/76- 9/ 1/76
Aromatic + 135705 85.0 112,640
Complex
PCB 58590 82.1. 41,701.
83.0
71.2
10/ 1/76-21/ 1/76
--- 18,351 - 0 16,927
TABLE 9. ACCUMULATED MASS BALANCE FOR K 2 0
,
92.2
.
Accumulation Period
Chlorinated
Hydrocarbon Accumulated Accumulated
Burned ln!ut (tons) Retention (tons)
%
Retention
7/10/75-22/10/75
--- 240.99 229.37
95.2
23/10/75- 4/11/75
A llphatk 199.67 166.72
83.5
2/12/75-14/12/75
Aromatic + Complex 194.59 186.68
95.9
3/ 1/76- 9/ 1/76
10/ 1/76-21/ 1/76
PCB 103.25 92.88
--- 280.40 274.36
90.0
97.8

-------
33
While the chlorine balance does show discrepancies, in all cases
while burning chlorinated hydrocarbons more chlorine was retained In process
solids than the quantity input with materials other than these wastes.
A major portion of the chlorine from the chlorinated hydro-
carbon materials is thus accounted for. Considering the uncertainties
involved with ring formation and kiln upsets, the mass balance for
chlorine confirms the finding of the emission measurements that all
chlorinated hydrocarbons are destroyed in the cement kiln.

-------
34
6 CONSIDERATIONS ON BURNING CHLORINATED HYDROCARBON WASTES IN A
CEMENT KILN
6.1 Effect on Production
Kiln ring formation normally decreases kiln production, apart
from any loss of production caused by downtime. However, the kiln rings
formed during this study were attributed to poor control of feed rates
of the chlorinated waste; Ring formation through chlorine addition rates
higher than desirable were previously encountered at St. Lawrence Cement
while adding waste hydrochloric acid. Adequate control would certainly
eliminate this problem and, since this Is a fairly simple proposition,
kiln ring formation should not be considerid a deterrent to use of this
waste in the kiln.
During the study, average production rates (Table 10) were
1038 tons per day while not burning chlorinated hydrocarbons and 1025 tons
per day while burning these materials.
There was also an Increase of-approximately 20 tons pir day in
the quantity of dust discarded (Table ll)whlle burning chlorinated
materials.
While these values may be partly due to ring formation, It is
probable that dust generated would Increase due-to Increased volatilization
of alkali chloride, and that this would have a corresponding effect on
clinker production. Thus, If.another form of chloride were being used for
alkali reduction, it is unlikely that any change would be detected by use
of chlorinated hydrocarbon wastes.
6.2 AlkalI Reduction While BurnlngChlorlnated Hydrocarbon Wastes
As indicated earlier in this report, Na 2 O Is not considered in
this study because it is low and practcally constant In the St. Lawrence
Cement Co. raw materials and products. -
To determine the efficiency of alkali reduction, the followIng
points require consideration. The percent I( O reported in the slurry
feed is on the natural or “as received” basis. To determine the quantity
which would be present if none were volatilized in the burning process,
the results must be calculated on the “Ignited basis”, that is, recalculated
for the CO 2 evolved from the raw materials In the burning process. The

-------
TABLE 10. AVERAGE REDUCTION IN 1(20 CONTENT OF CLINKER
Accumulation Period
Clinker
Prod.
Ct/day)
Chlorinated
Hydrocarbon
Cl Input With
Hydrocarbon
2 Relative to
Clinker Prod..,
,
Slurry
‘Natural
Basis
K 2 0 (2)
1(20
Clinker
(2)
1(20
Reduction
(2)
Ca lculated*
1(20
Reduction
(2)
Ignited
Basis
7/I0/75-22/IO/75
1056
-
0
0.92
1.42
1.21
0.21
0
23/10/75- 4/11/75
1Q50
Aliphatic
0.41
0.92
1.42
0.87
0.55
0.58
2/12/75-14/12/75
1020
AromatIc +
Complex
0.44
0.90
1.42
0.74
0.68
0.58
3/ 1/76- 9/ 1/76
1006
PCB
0.34
0.91
1.41
0.87
0.54
0.45
10/ 1/76-21/ 1/76
1020
-
0
0.9)
1.40
1.25
0.15
0
*Based on chlorine input.

-------
36
TABLE 11. AVERAGE DUST DiSCARDED
Period
.
Chlorinated
Hydrocarbon
Average Dust
Discarded
(Tons Per Day)
7/10175-22110/75
---
22.2
23/10/75- 4/11/75
AlIphat c
42.7
2/12/75-14/12/75
Aromatic +
Complex
74.0
31 1176- 9/ 1/76
PCB
62.5
10/ 1/76-21/ 1/76
—--
45.1
percent 1(20 In slurry feed on the Ignited basis minus the percent 1(20 In
the clinker yields the reduction through volatilization In. the kiln.
Since the 1(20 reduction -is achieved by formation of potassium chloride
(KC1), the reduction expected while bUrni g chlorinated hydrocarbons,
based on the assümp tion that-all chlorine is bound lntä potassium chloride,
can be readily calcu1ated Câmparlsàn of.actual and calculated values
(Table 10) are excellent. If ..ch1orlnè Is added to reduce alkal es, the
reduction Is stbichlometric. However, the total reduction Is relative to
alkali levels In the slurry and not to the baseline content in the clinker.
This result was confirmed both by this study and by previous experience
using hydrochorc acid as the source of chlorine. This may be due to
formation of potassium chloride from volatile compounds such as
potassium carbonate. On average, without chloride addition, a reduction
of 0.18 percent 1(20 Is apparent. While burning chlorinated hydrocarbons,
the average-reduction is that calculatedon the basis of potassium chloride
formation plus 0.05 percent.
The extremely good correlation between quantity of chlorine
input by burning of chlorinated hydrocarbons and alkali reduction In the
clinker gives better proof than the mass-balance that chlorinated
hydrocarbons are destroyed In a cement kiln. Otherwise, alkali chlorIde
could not be formed, and volatilizatlon.to the extent noted could not
occur. The high degree of correlatIon indicates that all of the chlorinated
hydrocarbon was destroyed.

-------
37
6.3 Heat Recovery From Chlorinated Hydrocarbon Wastes
Calculation of heat recovery from these materials, normally
difficult due to the low flow rates involved, was made more difficult
by the flow control problems encountered. During the aromatic plus
complex chlorinated hydrocarbon burn, the oil meter did not function
properly and assessment of heat recovery was made only over the first
two days of burning. As near as can be established, approximately 65%
of the heat of the chlorinated hydrocarbon was recovered (Table 12).
This can be expected since volatilization of potassium chloride
required energy. While the energy is recovered upon condensation, this
occurs too far along the kiln to be completely recuperated. Weber (18]
states “volatilization of alkalies consumes high grade heat In the
sintering zone and calcining zone above a material temperature of 800°c
and this heat Is subsequently released only at lower temperatures. Thus,
degradation of high-grade heat takes place”. -
When it i_s considered that-potassium chloride forms a cycle
within the kiln and maybe volatilized several times befoie escaping to
the precipitator, the heat consumed by this process becomes considerable.
In the suspension preheater kiln, an indication of the number
of cycles of volatilization of alkali chlorides is given by the relation-
ship between the Stage IV chloride level and quantity of chloride input.
Since in this study, the (20 at Stage IV was a factor of ten greater than
the (20 input, it follows that ten cycles of volatilization occur on average
in the suspension preheater kiln. The situation in the wet process kiln,
however, is dissimilar. There is no convenient method for determining
the number of alkali cycles Involved. That there will not be as many
cycles as In the suspension preheater kiln is certain when consideration
is given to the differences in the two processes.
It is assumed, for purpose of Illustration, that a chlorinated
hydrocarbon with 9300 Btu/lb and 1.2% chlorine is being burned with an
alkali cycle of three times. ‘For each pound of chlorinated material,
there are 0.42 pounds of chlorine which will produce 0.88 pounds of
potassium chloride. The heat of vaporization of potassium chloride is
38,840 cal/mo) [ 19] or 938.6 Btu/ lb. The 0.88 pounds of potassium

-------
38
TABLE 12. RECOVERY OF’ Btu FROM CHLORINATED HYDROCARBONS
Material
Average
Reduction
Input With
6
(10 x Btu/ton
Clinker)
.
Average
Input With
Chlorinated
Hydrocarbon
(106 x Btu/ton
Clinker)
•% Useful
Heat From
Chlorinated
Hydrocarbon
Aliphatic
Aromatic
Complex
+
0.125
0.128
0.205
0.217
61
59
PCB
0.157
0.228
69
chloride would require 2478 Btu for three cycles. On this basis we
might expect to recover:
9300-2478 x .100% 73% of the heat content of the chlorinated
hydrocarbon.
While the above 1s only anapproximation and ignores concurrent
heat exchange processes, it Indicates that the heat recovery of about
65% obtained in the present study Is reasonable.
it should be emphasized that the addition of calcium chloride
would require a similar quantity of heat to volatilize alkalies. In the
case of calcium chloride, an Increase in fuel quantity would be required.
6.4 Cement Quality
While the literature available (2, 3, 10] indicates that alkali
reduction Is beneficial with regard to cement quality, the possibility
of unburned organic material being retained In the cement and having
a deleterious effect was considered. Cements were ground In the laboratory
with clinkers produced while burning only No. 6 fuel oil as well as
clinkers produced during the aromatic plus complex and PCB burns.
Detailed results are given In Appendix.D. The results obtained indicated
the only effects to be those due-to alkali reduction.
6.5 Extrapolation to Other Kiln Types
Since all cement kllns have In coam on the requirement that
uniformly high temperatures be maintained, the authors believe that

-------
39
chlorinated hydrocarbons will be destroyed In all types of cement kilns -
regardless of fuel used. The only qualification to this statement is
that the wastes must be Injected Into the burning zone. While It is
advisable to atomize the chlorinated waste, there appeared to be no
difference during the PCB burn with or without a nozzle for atomization.
For fuel economy, cement kilns are operated at low excess oxygen.
This, combined with the high temperatures and hydrogen from the fuel,
will In all cases ensure that hydrogen chloride (HCI) is preferentially
formed rather than free chlorine. In all cement kilns, the lime will
rea Ily react with the hydrogen chloride. Based upon results of the
present study, no detectable quantities of either compound are expected
to be emitted from any type of cement kiln.
Since alkali chloride Is volatilized and condensed in the gas
stream as extremely fine particles, it would be expected that emission
of particulate matter would Increase by an amount depending on the
efficiency-of the precipitator. Thiá;wouldappiy to any method. of
chloride addition. To overcome the plugging problem caused by condensation
of these salts in suspension preheaters, design modificationsto-these
units need to be installed and demonstrated.
6.6 Comparison of Cement Kiln Burning with Other Uses and Disposal
Methods for Waste Chlorinated Hydrocarbons
information on disposal methods used and actual quantities of
chlorinated hydrocarbons requiring disposal is not readily available in
Canada. While a conservative estimate of 25-30 million pounds of
chlorinated hydrocarbon wastes was-obtained, it seems likely that this
quantity represents only wastes from plants manufacturing chlorinated
hydrocarbon products. The magnitude of the problem of disposal of these
wastes-in North America can be inferred by some of the methods used.
One method makes use of ships designed for burning of these wastes at
sea, this method being described in studies monitored by the United States
Environmenfal Protection Agency (20, 21]. Such a method of disposal
- is expensive, requires constant monitoring of temperatures within the
furnace, uses additional fuel for combustion and emits hydrogen chloride
(MCI) which Is dissolved inthe ocean.

-------
40
Another method of disposal In North America Is incineration
with production of hydrochloric acid (221.
One report from Europe [ 23] indIcates that methods of disposal
are:
1) illegal dumping of small quantities in barrels or other
containers on uncontrolled refuse dumps;
2) deposition of larger quantities in barrels on refuse dumps
which are supposedly sanctioned-for this purpose;
3) combustion In simple facJlitles without hydrogen chloride
- scrubbing; -
4) combustion of piles of-barrels on remote beaches with an
offshore wind; -
5) dumping barrels on the open sea;
6) dumping liquids into the sea from moving vessels;
7) separation of waste materials and recovery of useful
-components;
8) combustIon with .recoveryof hydrochloric acid; and,
9) combust ion on the open ‘sea at temperatures guaranteeing
almost complete pyrolysis.
The same report states: “Only the last three procedures can
be considered to be not harmful for the environment. But proéedures 7)
and 8) are possibly very expensive and in special cases unsustainable for
the producer”. Also from the same report, It Is pointed out that combustion
at sea requires extensive observation of a variety of safety procedures.
A study from France [ 24] IndIcates that a variety of legal and
illegal means are used to dispose of chlorinated hydrocarbon wastes.
Of illegal means: “thedischarge of what in general are Insignificant
quantities is disposed of In drums, or by tankers, Into waterways, former
quarries now used for other purposes, or discharged with unsupervised
wastes that reason suggests should be retained”.
In comparing combustion in a cement kiln, with other methods
of disposal which are considered not harmful to the environment, the
following points become apparent:
- Incineration of these wastes Is normally done at a flame

-------
temperature of 1200°C to 1560°C (25] while cement kiln
flame temperatures are 2100°C (18] or higher.
- The retention time In a cement kiln flame envelope is consider-
ably longer then the 0.1 seconds normally found in an
- incinerator.
- For the production of cement clinker, the temperatures cited
are necessary (26], thus removing the necessity for constant
monitoring of temperatures as required when burning in an
Irici nerator.
- To prevent operating difficulties, such as kiln rings, in a- ..,
cement kiln, the amount of chlorine is restricted to
approximately O.k percent relative to clinker. Besides
maintaining constant temperature, this requirement ensures
that sufficient hydrogen is available to form hydrogen
chloride which is readily absorbed by-iii e.
- There Is always a high quantity of lime in the cement kiln
- to react with hydrogen chloride and thus prevent emission
of this compound to the atmosphere.
- Burning of these wastes in a cement kiln saves fossil fuels,
as opposed to the necessity of using fossil fuels to
ensure combustion of these wastes in an incinerator.
- Beneficial use is obtained In a cem nt kiln of persistent
and toxic waste materials which normally require disposal.

-------
7 CONCLUSIONS
The concept of using chlorinated hydrocarbon wastes In
cement processing derives f ra i n knowledge of kiln operating temperatures
and residence times in comparison with incinerators capable of destroying
these compounds. The action of cement kl lns as “dry-lime scrubbers” in
relation to halogens In the kiln gases was already known.
In all cases, in the present study, analyses of kiln emissions
indicated the efficiency of combustion of chlorinated hydrocarbon wastes
to have exceeded 99.98 percent. Traces of volatile low molecular weight
chlorinated hydrocarbons were found to be present at a maximum emission
concentration of 50 ppb above baseline. All other starting materials,
Including polychlorinated biphenyls, were-completely destroyed. There
were no detectable emissions of hydrogen chloride, free chlorine or
high molecular weight chlorinated hydrocarbons.
An increase in total particulate emissions was observed while
burning chlorinated hydrocarbon wastes. This’ is apparently related to
an increase in dust loading to the precip!tator-as indicated by thi
necessity ó f discarding more dust while burning chlorinated hydrocarbons.
Combustion in incinerators designed for destruction of chlorinated
hydrocarbon wastes caused the emission of most of the hydrogen chloride
except for those installations using expensive hydrochloric acid recovery
or scrubbing systems. These incinelators require continous monitoring
of temperature profiles and use fossil fuels to Initiate or maintain
combustion.
The present study Indicates that useful recovery of about 65
percent of the heat-value and approximately 100 percent of the chlorine
is attained by the burning of these wastes in a cement kiln.
A mass balance on chlorine and the effective alkali reduction
derived from the chlorine contained in the chlorinated hydrocarbon
wastes confirms the air emissions data.
Consideration of the data from this study and examination of
the general literature on cement manu acturing has led the authors to
conclude that all chlorinated hydrocarbon wastes may be used In cement
ki lns without adverse effect on air pollution levels.

-------
REFERENCES
1. Personal communication, January 15, 1975.
2. Nlel, E.M.M.G., “The Influence of Alkali-Carbonate on the Hydration
of Cement”, Proceedings of the Fifth International Symposium on the
Chemistry of Cements, Tokyo, 1968, PublIshed 1969.
3. McCoy, W.J. and 0.L. Eshenour, “Significance of Total and Water
Soluble Alkali Contents of Cement”, Proceedings of the Fifth
International Symposiu.an on the Chemistry of Cements, Tokyo, 1968,
Published 1969.
4. Steinour, H.H., “The Setting of Portland Cement, A Review of Theory,
Performance and Control”. Portland Cement Association, Research
Department Bulletin 98, 1958.
5. Lerch, W. “Studies of Some Methods of Avoiding the Expansion and
Pattern Cracking Associated With the Alkali Aggregate Reaction”.
Portland Cement Asiociatlon, Research Department Bulletin 31, 1950.
6. Powers, T.C. and H.H. Steinour, “An Interpretation of Published
Researches on the Alkali-Aggregate Reaction: Part 1 - The Chemical
Reactions and Mechanism of Expansion;Part 2 - A Hypothesis Concerning
Safe and Unsafe Reactions with Reactive Silica in Concrete”. Portland
Cement Association, Research Department Bulletin 55, 1955.
7. Annual Book of ASTM Standards, Part 13 (1975). American Society for
Testing and Materials.
8. Lerch, W. “The Influence of Gypsum on Hydration and Properties of
Portland Cement Pastes”, Portland Cement Association, Research
Department Bulletin 12, 1946.
9. Woods, H. “Reduction of Alkalies in Cement Manufacture”, Mill Session
Paper M-149, Portland Cement Association, Manufacturing Process
Department, 1956.
10. Woods, H., J.L. Gilliland, Jr., J.F. Weigel, B.E. (ester, and
l.A. Stevens, “Symposium on Alkali Removal and Problems”. Regional
Fall Meeting of General Technical Committee, PCA, Milwaukee, Wisconsin,

-------
44
Sept. 2l 24 1959. Mill Session Paper M—l58, Portland Cement Associa-
tion, Manufacturing Process Department, 1960.
11. Coles, C.W. and D.G. Dainton, “St. Lawrence Cement Co. Clarkson
Plant”, Cement Technology 1 (2), 113, 1970.
12. Herod, B.C. “St. Lawrence Cement Expands Clarkson Operation”,
Pit and Quarry , July 1968.
13. Schroth, G.A., “Suspension Preheater System Consumes Less Fuel”,
Rock Products 15 (5), 70, 1972.
14. Garrett, H.M., “The Potential Promise — Prospects and Pitfalls
in Energy Conservation by the U.S. Cement Industry”, Paper
presented at the Cement Chemists Seminar, Port1an Cement Associa-
tion, Lincolnwood, Iii., February, 1976.
15. Trauffer,W.E. “Portland Cement Outlook and Review”. Pit and Quarry ,
January 1976.
16. Campaan, K., Central Laboratory TNO Report No. CL 74/93. “On the
Occurrence of Organic Chlorides in The Combustion Products of an
EDC Tar Burnt by the Incinerator Ship ‘Vulcanus’; A Preliminary
investigation”. Oct. 1974.
17. Gagan, E.W. “Air Pollution Emission and Control Technology — Cement
Industry”, Environment Canada, Environmental Protection Service,
Economic and Technical Review Report EPS 3—AP-7 1 -3, 1974.
18. Weber, P. (TranslatIon), “Alkali Problems and Alkali Elimination in
Heat-Economising Dry-Process Rotary Ki lns”, Zement Kalk GIps , (8),
1964.
19. Perry, J.H., C.H. Chilton, and S.D. Kirkpatrick, Chemical. Engineers
Handbook , 4th Edition, McGraw-Hill.
20. Marine Environmental Monitoring of “Vulcanus”, Research Burn II,
December 2-10, 1974, Preliminary Report, U.S. Environmental
Protection Agency, December 10, 1974.
21. Badley, J.H., A. Telfer, and E.M. Fredericks, “At-Sea Incineration
of Shell Chemical Organic Chloride Waste”. Technical Progress
Report BRC - Corp. 13-75—F, Shell Development Company, April 1975.

-------
‘45
22. American Society of Mechanical Engineers, Research Coninittee on
Industrial Wastes Report. Incineration of Chlorinated Hydrocarbons
with Recovery of HC1 at E.I. du Pont de Nanours g Co., (Inc.)
Louisville, Ky. ASME Industry Survey, Present State of the Art,..
Disposal of Industrial Wastes by CombustIon, January 1971.
23. Grasshoff, K. (Translation) “Expertise Regarding the Effects of the
High Temperature Combustion of Chlorinated Hydrocarbon on Special
Vessels at High Seas”, 1973.
24. “Incineration of Industrial Chlorine Wastes on the High Seas”.,
Report from the Environmental Agency (Ministére chargé do
l’Environnement) of the Pollution and Nuisance Prevention
Administration (Direction do Ia Prevention des Pollutions et
Nuisances), France. 1974.
25. United States Environmental Protection Agency, Permit No. 730 DOO8C
(3) to Shell Chemical Company, Inc and.Ocean Combustion Ser icis,
B.V. December 12, 1974.
26. Peray, KE., and J.J. Waddell, The Rotary Cement Kiln , Chemical
Publishing Co. Inc. New York, 1972.
27. Reynolds, L.M., “Pesticide residue analysis in the presence of
Polychiorobiphenyls (PCB’s) Residue Reviews , 34, 27, 1971.
28. Santoleri, J.J. “Chlorinated Hydrocarbon Waste Recovery and Pollution
Abatement” Chem. Eng. Prog . 69 (1) 68, 1973.

-------
46
AC KNOWLEDGEMEPITS
The authors wish to thank and to acknowledge the participation
of thefol lowing people and organizations in this work.
Environment Canada
Mr. W.A. Neff (EPS/WPCD) for program coordination and Invaluable
assistance in eStablishing details of the program management.
Mr. J. Robert (EPS/APCD) for helpful advice in sampling proce-
dures, analyses performed and review of results.
Ontario Ministry of the Environment
Dr. F. Frantisak and his staff for aid in planning the program,
analytical assistance and review of results.
Energy Mines and Resources Canada
Dr. F.D. Friedrick for assistance-at program planning stage.
- St. Lawrence Cement Co .
Mr. L. Kraszewski for program coordination.
U.S. Environmental Protection Agency
A.W. Lindsey and J. Schaum (Hazardous Waste Management Division)
for program planning coordinationand review of results. TRW
Systems Group and Control Pollution Services Inc., who were
contracted by EPA.

-------
49
APPENDIX A
QUANTIFYING, SAMPLING AND ANALYSIS OF PROCESS MATERIALS
While analytical procedures used are identical for both wet
and dry process kilns, only the wet process kiln material streams are
considered in this Appendix. The dry process suspension preheater
kiln system is detailed in Appendix B.
The material flow through the wet kiln production system
Is illustrated in Figure A.1. To obtain the material balance, six
materials were quantified, sampled and analyzed. A seventh material, -
precipitator return dust, was monitored to obtain any additional operating
data this stream might yield. Table A.l lists the materials examined
with approximate relative quantities under normal production conditions.
TABLE A.l. PROCESS MATERIALS STUDIED AND APPROXIMATE NORMAL PRODUCTION
QUANTITIES
Material Approximate Quantity -
Slurry Feed 1540 dry tons/day
Clinker 1000 tons/day
Precipitator Discard Dust 0 - 140 tons/day
Precipitator Return Dust 350 tons/day
No. 6 Fuel Oil 20 gal/mm
Chlorinated Hydrocarbons 1 - 2 gal/mirt
Kiln Exhaust Gases 160,000 ACFM
A.l Quantifying and Sampling Slurry Feed and Clinker
There are four slurry basins, each of 6000 ton capacity,
including slurry water. The product from the slurry grinding mills is
pumped into these basins. The basins are supplied with air agitators-
for the purpose of blending the slurry and maintaining a uniform suspension
of solids in liquid. Slurry is pumped from one basin at a time to the
constant ‘level box of the slurry feed system. The material which over-
Page 47 and 48 blank

-------
AIR
FUEL (OIL CHLORINATED
DUST RETURN
CLINKER
PRECIPITATOR
STACK GAS
MATERIALS)
DUST WASTED
SCHEMATIC OF THE MATERIAL BALANCE
SLURRY FEED
U I
0
1
2
3
I.
5
6
1
A KILN
B PRECIPiTATOR
FIGURE A.1

-------
51
flows from the constant level box is returned to the slurry basins.
Slurry feed for the kiln is taken from the constant level box by an
Allis Chalmers “Ferris Wheel” bucket wheel feeder with variable speed
drive. From the bucket wheel feeder, the slurry flows into a small
basin with an outlet at the bottom,from which the slurry goes directly
into the kiln.
The slurry feed samples were taken every two hour-s from the
overflow of the constant level box, and blended into a daily composite.
A portion of the daily composite then was dried at 120°C for analysis.
Slurry feed quantities were obtained by multiplying the total
number of revolutions of the “Ferris Wheel” per day by a factor. Measure-
ment of the slurry basin levels while feeding the kiln but not allowing
slurry to be fed into the basin gives an accurate quantity of slurry
fed to the kiln. A second method, performed every two hours, is to
close a valve at the outlet to the kiln of the small basin fed by the
Ferris Wheel. By determining the time to fill this basin, the quantity.
fed - to - the kiln - I s known.
The clinker is fed to a bucket elevator by gravity through a
chute from the cooler. Clinkersamples are taken from this chute at
intervals of two hours to form a 24—hour composite. The clinker composite
sample is mixed, reduced by “cone and quartering” and a portion ground
for analysis.
In coimton with general practice in the industry, there is no
provision In the plant for continuous weighing of clinker. The water
content of the slurry feed Is determined thrice daily, from which the
following calculation is made:
• - ( ioo - H 2 0
SFD SF 100
where SFD — quantity of slurry feed on a dry basis
• SF — quantity of slurry feed including water
1120 — % water In the slurry feed. - -
To obtain the quantity of clinker produced,- the CO 2 which. Is
lost in the burning process (determined by loss on ignition of the dry
slurry feed) is deducted and a further correction is made for the quantity.

-------
52
of dust discarded from the system. A total material balance may then
be written as follows:
( 100— 101 (ioo -
Clinker SFD 100 SFD) - DD 100
where SFD quantityof slurry feed on dry basis
10 1 SFD loss on ignition on slurry feed on dry basis
DD =. quantityof dust discarded
LOIDD loss-on ignition on dust discarded.
A.2 Weighing and Samplinç Precipitator Dust
- Dust from the precipitator is separated Into two portions by a
system of screw, conveyers under the precipitator. The major -portion
is returned via a- conveyor.belt Into a small holding silo from which the
dust Is returned into the burning tone of the kiln (return dust). Samples
of the dust are taken at two—hour Intervals from the conveyor belt to
- - form a 2k-hour composite. The ,co posite is bléndedand a portion taken
‘fár añalysis. A minor portionáf the dust (dis ard dust) is fed intc a
second holding silo from whIch trucks are ,loaded’directly. A sample-
of this dust is taken frorr each truck load, and all samples for each day
are blended into a 2 1 -hour composite. Each load of dust s weighed on
the truck weigh scales before being disposed of.
A.3 Measurement and Sampiing No. 6 Fuel Oil
While it was considered unlikely that No. 6 fuel used in the
plant would contribute a substantial quantity of chlorine, samples of
oil were taken daily. Fuel quantities are continuously monitàred and
recorded hi the production data but are not given in this report. -
Chlorine content of the oil was determined, and the contribution to the
chlorine mass balarce due tc’ oil was included. -
A.4 Measurement and Sampling Waste Chlorinated Hydrocarbons
Sampling of.chicrirzate.d by4rocarbor.swaz carrJ - u. Lby withdrawing
material from the feed system twice daily and blending by vigorous mixing.
Samples were split into equal portions after blending.

-------
53
initially, it was intended to meter chlorinated hydrocarbon
flow continuously. For this purpose, a magnetic flow meter was installed
in the line to monitor quantities of these materials. For this type of
meter to function properly, it is required that the conductivity of the
material be greater than 2 x 106 mhos. While this was the case for the
chlorinated aliphatic wastes, the conductivities of the other materials
were less than 0.3 x 106 Tthos. As a result, chlorinated waste material
quantities were determined by taking measurements of the volume of
material in the storage tank. Tables A.2 through A.4 give detailed
information on calculated chlorinated waste material quantities.
A.5 Emission Sampling
Separate sampling t ains were used to collect representative
samples of kiln emissions for analysis to determine the emission rates
of particulate material and to determine if organic chloride compounds
were present in the gas stream passing to the stack. Particulate and
gaseous samples were coliected from a duct downstream f the preèipita—
tor controlling dust emissions, from kiln #1. .The location is described
below, followed by descriptions of samp!lng equipment and methodology.
A.5.I Sampling location
• Emissions from both #1 and #2 wet process kilns pass through
electrostatic precipitators (Joy Manufacturing Co.) and then through
sections of rectangular breeching before mixing in a comon section of
duct entering the stack. The stack, which is of height 554.0 feet and
has an exit diameter of 13.0 feet, vents the exhaust gases to the
atmosphere at a temperature of about 400°F and a velocity of almost
40 feet per second.
The ‘rectangular breech ing from the precipitator makes a 900
bend and tI;en angles at about 300 upwards from the horizontal for a
distance of approximately 45 feet to the coninon header which leads into
the stack. This 45 foot section of duct was considered the most suitable
for installation of--’sampl ing ports at a location which would meet both
Federal and Provincial source testing codes. it was subsequently
decided, therefore, to install five 4” diameter ports in the vertical
side of theduct at a position 12 feet upstream of the bend into the

-------
54
TABLE A.2. QUANTITIES OF ALIPHATIC MIXTURE BURNED DAILY
Time
.
.
Date From To Minutes Igpm Gallons
Gallons/Day
23/10/75 09:00 12:00 180 0.5 90 1170
12:00 06:00 1080 1 1080
24/10/75 06:00 06:00 1440 1 1440 1440
25/10/75 06:00 06:00 1440 1 1440 1440
26/10/75 06:00 06:00 1440 1 1440 1440
27/10/75 06:00 06:00 1440 1 1440 1440
28/10/75 06:00 06:00 1440 1 1440 1440
29/10/75 06:00 06:00 1440 1 1440 1440
30/10/75 06:00 09:30 210 1 210 2670
09:30 06:00 1 230 2 2460
31/10/75 06:00 06:00 - 1440 2 2880 2880
1/11/75 06: 00 06:00 1440 2 2880 2880
2/11/75 06:00 06:00 1440 2 2880 2880
3/11/75 06:00 06:00 1440 2 2880 2880
4/11/75 06:00 10:00 240 2 480 930
10:00 17:30 450 1 450
OFF at 17:30

-------
55
TABLE A.3. QUANTITIES OF AROMATIC PLUS COMPLEX MIXTURE BURNED DAILY
Date
Time
Tank Difference
Measurement gallons minutes
Gal/mm
Dec.
6
15:00
2 loads
•
in
12’ 0”
60930 lb + 40690 lb 1.299 SG a 7823
gal
Dec.
Dec.
7
8
12:15
15:00
1 1 4’ 6” (7823—5886 • 1937 gal)
1937 1275
14’ 0” 1177 1605
1.519
0.733
Dec.
9
15:30
13’ 6” 1177 1470
0.801
Dec.
10
12:00
12’ 6” 2354 1230
1.914
Dec.
11
15:00
11’ 8” 1962 1620
1.211
Dec.
12
09:00
11’ 4” 785 1080
0.727
Dec.
12’
15:30
11’ 0” 785 390
2.013
Dec.
13
13:30
10’ 3” 1373 1320
1.040
Dec.
Dec.
14
15.
13:20
14:00
.9’ 6” 1766- 1430
8’ 8” 1962 1480
1.235
1.326
Date
From
Gallons per Day
To Minutes Igpm Gallons
Total Gallons
Dec.
2
06:00
13:10
430
(1)
430
430
Dec.
5
10:00
12:00
120
1.519
182
1640
24:00
06:00
960
1.519
1458
Dec.
6
06:00
06:00
1440
1.519
2187
2187
Dec.
7
06:00
12:15
12:15
06:00
375
1065
1.519
0.733
570
781
1351
Dec.
8
06:00
15:00
15:00
06:00
540
900
0.733
0.801
396
721
1117
Dec.
9
06:00
15:30
15:30
06:00
570
870
0.801
1.914
456
1665
2121
Dec.
1O
ö6:0O
12:00
12:00
06:00
360
1080
1.914
1.211
689
1308
1997

-------
56
TABLE A.3.
Date
From
To
Minutes
lgpfl%
Gal
ions
Total Gallons
Dec.
11
06:00
15:00
540
1.211
654
1308
-
15:00
06:00
900
0.727
654
Dec.
12
06:00
09:00
180’
0.727
131
1821
09:00
15:30
390
2.013
785
15:30
06:00
870
1.040
905
Dec.
13
06:00
13:30
13:30
06:00
450.
990
1.040
1.235
468
1223
1691
Dec.
14
06:00
13:20
440
1.235
543
1869
‘
- :
13:20
06:00
1000
1.326
1326
Dec.
15
06:00
06:00
1440
1.326
1909
1909
(Cont’d)

-------
57
TABLE A.4. QUANTITIES OF PCB MIXTURE BURNED DAILY
Date
Time
Tank
Measurement
Difference
gallons minutes
Gal/mm
3/1/76
14:30
1)’ 4”
4/1/76
12:00
12’ 6”
1962
1290
1.521
5/1/76
09:15
91 8U
1962
1275
1.539
5/1/76
14:30
9’ 6”
392
315
1.244
6/1/76
09:10
8’ 7” 1766
1120
1.577
7/1/76
09:06
7’ll”
1962
1436
1.366
7/1/76
16:00
7’lO”
196
414
0.473
8/1/76
08:45
7’ 9”
196
1003
0.195
.
8/1/76
15:45
7’ 8”
196:
420
0.467
9/1/76
08:45
7’ 2”
1177
1177
1020
1.154
9/1/76
05:45
6’ 8”
540
2.180
Date
From
To
Gal lons
per Day
Gallons
Tptal
Gallons
Minutes
-Igpm
3/1/76’
07:45
14:30 -
.05
1.521
616
‘
2030
14:30
06:00
430
1.521
14 14
4/1/76
06:00
12:00
360
1.521
548
2210
‘12:00
06:00
1080
1.539
1662
5/1/76
06:00
09:15
195
1.539
300
2159
09:15
14:30
315
1.244
392
6/1/76
06:00
09:10
190
1.577
300
2008
09:10
06:00
1250
1.366
1708
7/1/76
06:00
09:06
186
1.366
254
614
09:06
16:00
16:00
06:00
414
840
0.473
0.195
196
164
8/1/76
06:00
08:45
165
0.195
32
1215
.
08:45
15:45
15:45
06:00
420
855
0.467
1.154
196
987
9/1/76
06:00
08:45
08:45
18:15
165
570
1.154
2.180
190
1243
1433

-------
58
header, and positioned on a line perpendicular to the gas flow. The
port locations; are shown schematically in Figure A.2. -
As agreed upon by the Ontario Ministry of Environment and
Environment Canada, sampling was carried out at eight points on each
of the five duct traverses for a minimum of five minutes per point. The
sampling points were located at the centres of equal rectangular areas
in the duct as shown in Figure A.3. The numbers given in each rectangle
are typical gas velocities measured during the study and indicate the
distribution •of gas flow at that location.
A.5.2 Sampling equipment
Priçr to setting up equipment for particulate and gaseous
sampling, grab samples of the kiln em ,ssions were taken using the
equipment depjcted in Figure A.4. A gas sample was pulled through an
in—stack 47 nm-glass fibre filter in a stainless steel holder to
remove particulate material and then through a heated teflon line into
a Tedlar bag. For each bag sample a clean bag was placed in the lung
and the lid tightened to effect ‘an-air-tightseat Valves A and C were
closed and B opened so that, with the pump on, the, sampling line was
purged Valve B was then àlÔsed nd valve C opened to evBcuaté the
sampling lung, making sure that the vacuum-did not exceed 5 H Hg. At a
vacuum of 4” Hg, valve A was opened and the sample bag filled with
emissions from the duct. Valve A was then closed and valve B opened.
The pump was shut off and the line to the pump disconnected. Valve C
was opened to bring the lung to atmospheric pressure and the bag was
removed from the drum and quickly capped. Each bag sample was transported
to the laboratory at ORF foranalysis on the°same day that the sample was
collected.
A.5.2.l Particulate train . Samples of particulate emissions were
collected using a Joy Manufacturing Company EPA (Model CU-2) train,
shown schematically in Figure A.5. This equipment conforms to that
reconinended by both the Ontario Source TestingCode [ A.l] and the
Environment Canada Code [ A.2]. Prior to each test a velocity traverse
was conducted across the duct through each part to determine an average
gas velocity. Gas temperatures were also recorded at each sample point.

-------
I
I- . ’
12’O’ TO BEND

321 OM TO BEND
— - GA
FLOW
• 3cr

-------
60
108.9
x
110.6
x
118.6
x
118.6.
x
117.6
x
114.7
x
107.8
x
103.2
x
92.5
x
‘107.8
x,
112.0
x
113.3
x
110.6
x
110.6
x
108.9
x
107.8
x
78.2
x
89.2
-x.
92.6
x
91.9
x
94.5
x
100.2
x
l02. 0
x
104.9
x
72.5
x
7k 2
x
78.2
x
78.2
x
82.0
x
91.2W
x
x
97.7
,c
42.8
,c
5l .9
x
66.4
x
73.4 -
x
.
78.2
x
t
80.5
x
.
74.8
x
80.5
x
Velocity in feet per.second
FIGURE A.3. GAS--FLOW DIS1RiBUTI0N -ATSN4 LING POINTS
An Orsat analysis was made to find the concentrations of CO C0 2 , °2 and
N 2 in the gas stream and a gas moisture determination was carried Out
using an established procedure (A.l]. With the preliminary data obtained,
the isoklnetic sampling rate was calculated using known standard
equations [ A.ln].
- For each particulate test, sampling was made for five minutes
at each of the, eight points across the- duct through the five ports.
Specified data (A.1] were recorded every 2.5im lnutes on a test data
sheet. The high flow rates encountered at the far wall of the duct
and across the top of the duct necessitated use of probe nozzles of
diameter less than 0.25 inches. Nozzles of dIameter 0.219 Inches and
0.188 Inches were used during certain tests.
At the conclusion of a test the glass fibre filter was
removed from its holder and placed In a labelled petri dish. The
volume of water in the Impingers was measured and the contents transferred
E
E

-------
61
FIGURE A.4. GRAB BAG SAMPLING EQUIPMENT
FILTER
PROBE
HEATED TEFLON LINE
C .
ROTAMETER
VACUUM
GAUGE
VALVES
BULKHEAD
FITTINGS
TEDLAR BAG
LUNG
PUMP

-------
62
HEATED AREA
ThERMOMETER
PITOT
MANOMETER
FILTER
HOLDER
FIGURE A.5.
PARTICULATE SAMPLING TRAIN

-------
63
to polyethylene bottles, which were then Pabelled. The probe and
nozzle were rinsed with distilled water as the inside of the probe was
carefully brushed. The rinsings were collected in a polyethylene bottle.
The glassware between the end of the probe and top of the filter holder
was washed with distilled water and these washings added to the probe
rinse. The bottom of the filter holder and glassware between filter
and Impingers was also washed and the contents added to the impiriger
catch.
After filtering the probe rinse, drying and weighing the
Insoluble particulate material, weighing the reconditioned filter, and
determining the soluble portion of particulates In the pçpbe rinse by
aliquot evaporation, the total weight of collected particulate material
was obtained. The moisture removed from the sampled gas was found by
adding the gain In weight of the silica gel to the volume of additional
water collected in the Impingers. The volume of gas sampled was corrected
to standard conditions and the particulate loading and ethissiôn rate
calculated by use of appropriate equations (A.-l].
A.5.2.2 Gaseous train . it was originally intendea to sampue ror
gaseous organic chloride, compounds using the EPA—type particulate train,
with specific absorbents replacing water in the impingers. Due to the
high flow rates (0.5 - 1.0 cfm) required for isokinetic sampling, however,
it was considered that the collection efficiency of any absorbent for
volatile organic compounds would be very low. An experimental program
was, therefore, devised and carried out at ORF to evaluate collection
methods and develop a suitable sampling train for gaseous’organic chlorides
Details of this study,- performed under contract to the Ontario Ministry
of the Environment, are given in Appendix H. Data obtained which are
relevant to this section are described below.
- A particulate train was set up in the laboratory and heated
air was pulled through the complete system at a flow rate of 0.5 cfm.
Water was placed in the first impinger and solvents such as xylene,
toluene and -decane, containing a few ppm of CHCI 3 , CH 2 C1 2 or CHC1 2 CH 2 C1,
were placed in the second impinger. The third impinger was left empty -
and the fourth contained sillcagei. After a period of three to four
hoàrs, the time anticipated for a particulate test, the impingers were -

-------
6 e
examined with respect to solvent volume and organic chloride concentra-
tion. In all cases, it was found that the -loss of chloride compounds
was considerable, For the polar solvents, such as toluene or xylene,
the chloride concentration waè retained but the total loss of solvent
plus chlorides was in the range of 50%-70%. For the non-polar solvents,
such as decane, the volume loss of solvent was negligible but the chloride
loss was in the range of 60%—90%. Even If efficient collection of
trace organic chlorides from the gas stream by specific solvents were
possible, these compounds would not be retained in solution at gas
flow rates of 0.5 cfm or higher.
Recent studies have shownthat collection and retention of
trace organic ompounds is possible using specific adsorbent materials
(A.3]. Inert materials, such as the Chromosorbs, are considered to
have certain a dvantages over activated carbons In that water vapour
present in the gas does not hinder organic compound adsorption, and
adsorbed compounds are readily removed by thermal desorption. Experi-
ments using the pa’rticulate train with Chromosorb 102 or activated
carbon in the third ’imp,inger were’carried out, passing heated air
çontalning;i fw ppm of CHCl through-thesystern.at áflówrate of 0.5
cfm. Neither adsorbent removed more than 70% of the CHC1 3 initially
and, after about thirty minutes, almost all of the CHC I 3 was passing
through the system. Itwas oncluded,therefore, that efficient
collection and retention of volatile organic compounds was not possible
with gas flow rates of 0.5 cfm or higher’ and the particulate train could
not be used for this purpose.
Afte; further studies In the laboratory with both Chromosorb
102 and activated carbon, th sampling train for gaseous organic compounds
depicted In Figure A.6 was cdnstructed. Tests made with an air Stream
containing 15 ppm of CHC1 3 showed that either adsorbent would remove
better than 95% of the CHC1 3 over a sampling period of four hours.
Chromosorb 102 was selected on the basis of the advantages previously
mentioned.
A 47 mm fibreglass filter was, placed directly behind the probe
nozzle to remove particulate material at the duct gas temperature. The
sample gas was passed througt( midget lmpngers containing water and

-------
65
PROBE
S.S TUBING
MIDGET
IMPINGERS
r
ROTAMETERS L
:1 .ICE BATH_S
SODIUM
I CHROMOSORS I4YDROXIO
1 ADSORBENT TUBES -
PUMP
FIGURE A.6. GASEOUS SAMPLING TRAIN

-------
66
Sample prepa rat Ion
noncombus ted
supplied
a r r i ved,
to make
actual ly
suppl led
analysis
Waste feed . Samples of the prospective WBA feed material were
In advance of the test burn for analyses. When the test period
however, insufficient quantlte s of some components were available
up the specific blend of waste feed. The composition of material
delivered to SLC was, therefore, different from the sample
to ORF In advance. In Order to obtain a true compositional
of the material being supplied to the kiln, and to determine
caustic soda (5 solution), respectively, to remove any HC1 and Cl 2
present in the kiln àmissions. After the impinger, the gas was filtered
through four Chromosorb adsorbent traps placed in parallel at a flow rate
controlled by rotameters. A flow rate of 250 ml per minute was selected
for each trap, giving a total flow of one litre per minute through the
train. The total flow over each test period was also monitored by a
wet test meter placed after the pump. Impinger solutions and adsorbent
traps were maintained at ice temperature to effect a better collection
of chloride cpmpounds.
Afte r each test the in—stack filter and impinger solutions were
stored in label led containers. The adsorbent traps were removed, capped
and labelled. The probe was rinsed and the rinsings placed in a
polyethylene bottle. All samples were then transferred directly to the
ORF laboratory for analysis. The gaseous sampling train was operated
simultaneously with the particulate sampling train for each test.
•A.6 Organic Chloride Analyses -
• Themethodology used to collect kiln .ernlssion samples, in
baseline or waste burn test perlods,.provLded four distinct types of
samples for analysis of chlorinated organic compounds. - These sanple
types and the analyses required were:
- chlorinated waste-feeds for compositional analysis;
- grab bag samples for any chlorinated organic species;
- Chromosorb adsoi bent samples fqi volatile low molecular
*eight organic compounds; and,
- solvent extracts of filters and solutions for
waste component .

-------
67
feed consistency, aliquots of the waste feed were collected on specific
sampling days during the burn.
Grab bag . Emission samples collected In Tedlar bags were
analyzed directly for gaseous chlorinated hydrocarbons by Injecting
syringe samples taken from the bag into a gas chromatograph. After direct
analysis, the contents of each bag were drawn through a glass tube
containing Chromosorb 102 adsorbent by means of a small pump. Any trapped
organic compounds were then thermally desorbed from the-adsorbent tube
into an evacuated gas jar. The gas jar contents were then subjected to
GC analysis.
Adsorbent tubes . Chromosorb 102, being readily available in
amounts-required for this study, was selected as a suitable adsorbent
for low molecular weight organohalides. Experiments were performed in
the laboratory to confirm the suitability and efficiency ? this adsorbent
for the collection of volatile organohalides from an air stream. Thermal
desorption of adsorbedcomponents into an evacuated gas jar was selected
as a method for preparing samples for GCanalysis. - Initial tubes were
prepared with fresh Chromosorb 102 from the bottle without conditioning.
It was determined after the first baseline test period that conditioning
of the adsorbent would be necessary to reduce the amount of bleed
produced on heat ing which tended to produce excessive background noise
during GC analysis.
Adsorbent tubes were made up by packing Chromosorb 102 in glass
containers 11.5 cm in length and 11.0 nm inside diameter, using silanized
glass wool plugs at either end for support. Tubes for the WBA and WB8
burns were preconditioned by heating to 200 0 C and passing a stream of-
nitrogen at 40 mi/mm through them for four hours. Chromosorb 102 used
in tubes for the WBC and BLB tests was extracted with acetone in a
Soxhlet for 18 hours and then treated at 240°C with nitrogen for 12
hours prior to use.-
- Adsorbed components on the Chromosorb 102 after each test were
removed by thermal desorption into an evacuated gas jar of 500 ml
capacity. The adsorbent tube and gas jar were connected by Teflon

-------
68
tubing and the tube heated to a fixed temperature of 170°C using a
heating tape control led by. a variable transformer. When the fixed -
temperature was reached the stopcock of the gas jar between the jar
and the tube was opened, and heating of the tube continued for a further
15 minutes. The.stopcock was then closed, the Jar removed from the
connecting tubing and taken outside the laboratory to fill up with clean
air. Previous ‘studies of thermal desorption using this removal
technique in th laboratory have sháwn that recoveries in excess of 90%
may be expected for adsorbents with adsorbed organohalides.
So.1.utbon and filter extracts . Pentane or hexane was used as an
•extractant for organic compounds from all filters and probe rinse
insoluble fractions, using a Soxhiet apparatus. The same solvents were
used to extract probe rinse solutions, and aqueous sodium hydroxide
and impinger solutions, using a liquid/liquid partitioning procedure.
The extracts were then’ dried over anhydrous Na 2 SO 4 , filtered, and
concentrated by evaporation using a crnbination of rotary and Ko’ntes
tube heaters. The concentrate- was made-up--to a -known small volume ‘with
solvent for GCanalysis. -
Pentane was used as extractant-for BL.A, WBA and WBB test
samples, since its high volatility would mInini tze sample loss of low
molecular weight compounds during evaporation. Hexane was used as
extractant for WBC and BLB test samples.
-A.6.2 Sample analysis
All samples were analyzed by-gas chromatography with flame
-ionization (Fi)-or electron capture (EC) detectors. Gas chromatography -
mass spectrometric (GC-MS) anaiysIs was-perforjned on waste feed samples
to confirm the identity of major components. The various conditions
and columns used are sumnarized In Tables A.5 to A.7. The mass
spectrometer used was an AEI MS—30 Instrument equipped with an electron
bombardment. Ion source. Between the GC and the MS the Interface Is
of all glass design with a silicone molecular membrane. GC columns similar
to those described in Tables A.5 to A.7 were used at approximately the
same instrument conditions.

-------
69
TABLE A.5. GC ANALYSIS SYSTEM A
Column
Column Temperature
Injector Temperature
Detector Temperature
Detector
Flow Rates
Range and Attenuation
Chart Speed
- Chromosorb 102 (80/lao mesh)
6’ x 1/8” SS
- 180°c
- 2 )5°C
- 215°C
- FID and EC
- N 2 at 40 mI/mm
Air and H 2 adjusted for maximum
sensitivity
- As r quired
- As required
TABLE A.6. GC ANALYSIS - SYSTEM B
Column
Column Temperature
Injector Temperature
Detector Temperature
Detector
Flow Rates
Range and Attenuation
Chart Speed
- 15% SE 30 on Chromosorb W (AW; HDNS;
60/80 mesh) 12’ x 1/4” SS
0
- 60 C 1 isothermal for 20 minutes then
programed at 10°C/mm to maximum
temperature
- 190°C
- 230°C
— FID
- N 2 at 40 mI/mm
Air and H 2 adjust i for maximum
sensitivity
- As required
- As required

-------
70
Co I umn
TABLE A.7.
Column Temperature
Injector Temperature
Detector Temperature -
Detector
Flow Rates
Range and Attenuation
Chart speed
GC:ANALYSIS - SYSTEM C
4% SE 30/6% QF I on Chromosorb W (HP;
60/80 mesh) 6’ x 1/8” 55
- ‘ 200°C
- 250°C
- 23Ô°C
- Linearized EC
- N 2 at 25 mi/mm
- As required
- As required - -
ii) Gaseous train:
(a) Particulate filter + probe rinse
filter + Soxhlet thimble.
(b) Waterj
(a) In-stack filter + probe rinse
filter + Soxhiet thimble
‘Water
Aqueous NOH
Pentane (100 ml concentrated to
1 ml)
(b) Hexane (100 ml concentrated to
These conditions are used for routine PCB analyses. The same parameters
were used for! WBB waste feed except that an isothermal temperature of
155 C was maintained.
A.6.3 Blanks and standards -
For trace ana)yti a-lstudiés it ks necessary to avoid -
contamination of samples -at -all stages of samp)e collection and
preparation. Though extreme care wá takenduringthé study it was
not possible to pre-extract all hardware and chemical reagents used.
Due to the high sensitivity of the EC detector to many compounds, some
of the blank extracts, therefore, gave complex GC-EC profiles, which had
to be subtracted as background from the sample chromatograms. The
following blanks were obtained and analyzed for this purpose.
i) Particulate train:
(b)
(c)
iii) Solvents: (a)
1 ml)

-------
71
lv) Polyethylene containers: (a) Solvent extracts (100 ml concentra-
ted to 1 ml)
Standard gas samples of various components of interest, such as
Cd 4 , CHC1 3 , CH 2 CI 2 and 1,2-dichioroethane, were made by Injecting 50
ul allquots of these compounds Into an evacuated 500 ml gas Jar through
a silicone septum. Clean air was then allowed to enter the Jar to
attain atmospheric pressure. The sealed gas jar, therefore, represented
nominally a 100 ppm standard. Gas standards of lower co%ceritration were
made by dilution of-the 100 ppm standard using a slmi1ar procedure.
These standards were used inunediately after preparation and were not
retelned for use on the following day.
Solution standards ware made by weighing accurately known
amounts of the components of Interest and dissolving them In a known
volume of pentane.
A.6.4 Concentration factors
- - Desorbed compounds-from the ad sorbent tubes were àoncentraied
-into a 500 ii gas jar. For each particular test, therefore, the concentra-
tion factor would be the volume of’ emission passed through the tube
divided by 500 ml. The volumes sampled for each tube varied from 1.5
cubic feet to 3.22 cubic feet during the complete test piogram giving
a range of concentration factors from 85 to 181. Thus, a concentration
for a particular component of I ppm in the gas jar sample would mean a
concentration of about 10 ppb in the kiln emission, assigning a 100
percent collection efficiency, and subsequent desorptlon of the adsorbent.
- Most of the extracted samples were concentrated into a 2 ml
volume of solvent. The following example illustrates how the concentration
factors ware -determined for various components detected. Assume that
CH 2 CI 2 was detected at a concentration of 1 ppm in the solvent extract.
This corresponds approximately to a weight of 2 ug of the compound in
2 ml of solvent. The molecular weight of CH 2 CI 2 Is 85. Therefore, 2 ug
of CH 2 CI 2 at 70°C, the temperature at which the emission sample was
measured, occupies a volume of:
2 6 x 24 lItres, or 6 x 24 cubic feet.
85 x 10 85 x 10 28.3

-------
72
For test WBA 1, for example, a volume of 142.81 scf was sampled. Thus,
the concentration of CH 2 C1 2 In the kiln emission is: -
2 24 106
x x ppm or 0.14 ppb.
85 x lOu 28.3 142.81
The concentration factor is therefore ) Ø3 — 7140.
0.14
Factors were calculated for feed compounds of lowest and
highest molecular weight.
A.6.5 Sam le. chromatograms
for lOw molecular weight organohalides and noncombusted
chlorinated oi .ganic compounds present in any collected samples was made
using GC, techniques only, by selecting appropriate GC parameters and
using retention time data. trio specific cleanup or separation procedures
were performed in order to segregate components of interest from possible
interfering compounds because, althoughGC profiles obtained were quite -
complex, very low concentrations äf organic compounds were evident from
thepeak heights obtainedfor the atteriuatioris used; Estimates of specific-
compound concentrations in waste burn samples were made by subtracting
blank and baseline levels and comparing with a standard for that compound.
Waste feeds . GC—hD profiles for the three waste feed materials
burned in this study are gi #en in Figures A.7 to A.9. Identification
of the main components in the respective waste feeds are presented in
Tables 1, 2 and 3 in Section 5 of the report.
It was ant cipat.d that the levels, if any, of uncombusted
components l i the stack emissions would be very low. Therefore GC-CC
analysis was used to evaluate the organic solvent extracts of the
various impinger samples frRm the sampling trains for the presence of
any uncoflibusted components. Figures A.l0 and A.ll represent GC-EC profiles
for diluted waste feed material for WBB -and WBC, respectively. WBB
also contained approximately 50% o-chlorotoluene which has a weak
response to the EC detector. Therefore, in this instance the solvent
-extracts were also analyzed by GC-FID in order to determine whether any
uncombusted o-chlorotoluene was present.

-------
FIGURE A.7
GAS CHROMATOGRAPHIC PROFILE
FOR CHLORINATED ALIPHATICS
FROM FLAME IONIZATION
DETECTOR
8
SAMPLE FEED -OCT.
OC-SYSTEM B
1
I—
(I)
z
Lu
I-
z
15
17
6
12
2
7
0
9
5
8
10
20
25
30
35
40
TIME (mm)
WBA) SAMPLE FEED

-------
FIGURE A.8 GAS CHROMATOORApHIC PROFILE FROM FLAME IONIZATION DETECTOR
FOR CHLORINATED ALIPHATICS PLUS AROMATICS AND ALICYCLICS
(WBB) SAMPLE FEED
SAMPLE FEED - DEC. 12 t1’) 1975
OC- SYSTEM B
6
16
9
>-
F-
U)
z7
w
F-
z
5
3
d l
1’
5
7
26
5
27
35
60
45
TIME (mm)

-------
16
FIGURE A.9
GAS CHROHAT0GRAPHIC PROFILE FROM FLAME 1ONIZAT ON
DETECTOR FOR CHLORINATED ALIPHATICS PLUS AROMATICS,
ALICYCLICS AND POLYCHIORINATED BIPHENYLS (WBC)
SAMPLE FEED
>-
U)
z
w
1—
z
SAMPLE FEED-JAN. 8 th 1976
GC-SYSTEM B
I
11
2
9
3
1•
7
22
0
5
10
15 20 25 30 35 40 1.5 50 55
TIME (mm)

-------
SAMPLE FEED- DEC.l2th 1975
OC-SYSTEM C
irig/ra.L
5.Jouquot
(Temp. 155C)
FIGURE
A.1O
9.
8
7
5
4
3
2
1
0
10-
9.
B
GAS CHROMATOGRAPHIC PROFILES FROM ELECTRON CAPTURE
DETECTOR FOR WB8 SAMPLE FEED
0.7 ng/ u(
5g11’t aliquot
I 32
(Temp.155C)
o 2 4 6 8 10
TIME (mm)
I
7.
S
S
3.
2
1-
0.7ng / M. 1
5 i t aUquot
1 64
(T.vnp.190’ C)
0
TIME Imhr)
10
>-
I .-
z
w
I-
z
0246
10
12
t

-------
10
STANDARD AROCLOR 1242
OC SYSTEM C
l1*1O-3g/ 4 *..e)
WBC-AROMATICS • PCBs
SAMPLE FEED JAN. 8 th 1976
GC SYSTEM C
1 att( 2,73x 1O itg/ . e)
1x128
1x256
0 2 4
TIME (mm)
0’
2
4
12 14
TIME (mm)
FIGURE All
GAS CHROMATOGRAPHIC PROFILES FROM ELECTRON CAPTURE
DETECTOR FOR STANDARD AROCLOR 1242 AND SAMPLE FEED
6
>-
I—
U)
z
Lii
I-
z
I -,
1•
WBC

-------
78
Standards . Figure A. 12 shows the GC’EC prof Ie for the low
molecular weight chlorinated hydrocarbons, e.g. CH 2 C1 2 , CHCI 3 , CC1 4 and
1,2—dichiorethene. Figure A.13 shows the GC-FID profile for the same
compounds to Indicate the greater sensitivity of the EC detector.
A GC—EC profile for 5 ng of standard Aroclor 121e2 injected
into the column Is shown In Figuri AJI.
Waste Burn Samples . Figure A.12 shows GC—EC profiles for
desorbed gas samples for BLB and WBC test samples. The main difference
between the samples is in the Intensity of the peaks of interest.
The GC-EC profiles for all the solvent extract samples showed
the presence c f small peaks. When the profiles were compared to those
obtained from the BLA samples and the various control blanks (cf. Section
A.6.3), no pe s pecul iar to the waste burn samples were detected.
BLB samples became contaminated with WBC feed material and were discarded.
in most Instances, the EPA Impinger extracts showed the largest
response to the EC detector. Figures A. lie - and A. 15 show GC—EC profiles
for the organic solvent extracts of the EPA lmpinger samples obtained
for BLA-T3 andWBC-T3, respectively. Profiles obtained for the two
samples are quite similar. There isa large EC response occurring as an
unresolvedpeak in the early portion of the chromatographic profile.
This limited the sample size that could be used for injection, and
restricted the use of high sensitivity settings on the gas chromatograph.
It is normal practice when analysing for PCB’s to perform
cleanup and separation procedures In order o segregate the PCB’s from
interfering components and thus facilitate their analysis. Such a
cleanup and eparatlon procedure was performed on WBC-T3 using Florisil
adsorbent (27) In order to determine whether the early unresolved peak
(Figure A.15) could be removed. After cleanup, it was possible to use
a larger aliquot for Injection and higher sensitivity settings on the
chromatograph. The GC—EC profile of the cleaned up WBC-T3 sample is shown
In Figure A.16..
The GC—EC profile shown in figure A.l6 was obtained on on
organic solvent extract of a sample that should contain PCB’s if any were
present In the kiln emissions. The profile shows the presence of a few
small peaks. When compared with the Standard Aroclor profile (Figure A.ll),

-------
10
GAS CHROMATOGRAPHIC PROFILES
DETECTOR FOR LOW MOLECULAR
AND FOR BIB AND WBC TEST SAMPLES
FROM ELECTRON CAPTURE
WEIGHT CHLORINATED HYDROCARBONS
cc ’ 4
OCE
CHCI 3
cc ’ 4
(4ppb)
10.05 ml)
BLB-2B
10.5 ml aliquot)
9
8
7
6
4
WBC—3A
>-
1-
U)
z
w
I—
z
GC-SYSTEM -A
10.2 ml aLiquot)
GRAB SAMPLE
WBC adsorbed /desorbed
on Chromosorb 102
(imi oliquot)
2
1
0
6
802
128
0
FIGURE A.12
2
1K128
4 6 0
6
10

-------
8o
STAN DARD
(1 ppm)
GC - SYSTEM-B
1, 2- Dithtoro thane
: 9
0
0
C
0
>-
In
2
w
2
FIGURE A.13 GAS CHROMATOGRAPHIC PROFILE FROM FLAME IONIZATION
DETECTOR FOR LOW MOLECULAF WEIGHT HYDROCARBONS
g
8
7
6
5
C
V
£
E
0
U
15
TIME (mm)

-------
81
GC SYSTEM C
i J Inj•ction
>-
U,
z
LU
1 -5
z
5
4
3
2
FIGURE A.14 GAS CHROMATOGRAPHIC PROFILE FROM ELECTRON CAPTURE
DETECTOR FOR IMPINGER EXTRACT FROM BLA TEST 3
10
9
8
1
1* 64
5
10
15
20
TIME (mm)

-------
10
82
GC- SYSTEM-C
I *LL injection
9
8
7
6
5
>-
eM
z
w
z
4
3
2
d l
lic6L
0
15
TIME (mm)
FIGURE A.15 GAS CHROMATOGRAPHtC PROFILE FROM ELECTRON AP1URE
DETECTOR FOR IMPINGER EXTRACT FROM WBCTEST 3
20

-------
83
10
GC-SYSTEM C
•5 ’ INJECTION
8
I —
— 7,
(f
z
w
‘ —6
z
5
4.
3.
TIME (mm)
FIGURE A.16 GAS CHROMATOGRAPHIC PROFILE FROM ELECTRON CAPTUR
DETECTOR FOR IMPINGER EXTRACT FROM WBC TEST 3 AFTER
CLEANUP AND SEPARATION

-------
81.
and taking Into -account that the extract represents a >7000 concentratIon
factor, If any of the peaks In Figure -A. 16 represent PCB components the
levels must be extremely low. -
A.7 Analysis of Process Solids
For the purpose -of obtaining a mass balance during the
experimental c’hlorinated hydrocarbon burn, it was required to carry out
chemical analyses of clinker, slurry feed and discard dust for chlorine
and potassium. •Sulphur was determined-on process solids and the return
dust samples were analyzed for changes which may have affected kiln
operation.
X-ray fluorescence (XRF) analysis has been extensively used
for determlnaeion of sulphur, potassium and chlorine in cement, cement
raw materials and related materIals,- and was selected for use In this
study.
A.7.1 Analytical procedures
Sam leswerepr epared by grinding 10 grams ofsample’ for 90
seconds ln,a : PEX Shatterbox with tungsten -:carbidecontair ers and pucks.
From this material a 1.25 inch dlametèr ’pel let at 8 tons pressure was
made In SPEX aluminum sample caps. -
For the X-ray fluorescencemethod to be applied, there must
be a linear relationship between elemental concentration (over the full
range of concentration required) and-the measured fluorescence intensity.
The slope of the line representing this-relationship then can ‘be used
directly as a calibration constant, usually In the form of the ratio, counts
per second:pe çcent element. When a calibration line is generated by
known additions of the test element to’a material being analyzed (“spiking”),
the line intercept gives the percent of the element present In the
original sample prior to any additions.:
In ddltlon to linearity, ràproduclblllty of the calibration
constant Is a prerequisite If it is to be applied across a range of
materials having some degree of matrix variabllity. Both conditions
were examined prior to the final development of the analytical procedure.
The calibration curves for chlorine content of clinker and
raw meal—slurry feed were prepared by addItion of standard lithium

-------
85
chloride in alcohol solution to samples of these materials. Dust samples
were analyzed by the standard Volhard method for calibration of the
chlorine curve and the standard gravimetric method of precipitation with
barium for calibration of the sulphur curve (Table A.8). For calibration
of curves for sulphur in clinker and raw meal/slurry feed, determinations
were made on a Leco Induction Furnace Model 523CS with Leco Automatic
Titrator Model 517 (Table A.9). For calibration of the potassium curve,
all potassium results, as well as those for sodium, were obtained from
analyses on a Perkin-Elmer Model 460 Atomic Absorption Spectrophotaneter
in the flame emission mode (Table A.lO).’
Examination of these samples yielded linear calibration
relationships between X-ray fluorescence count rates and percent element.
The least squares computations for each set of data gave the slopes,
intercepts and regression (correlation) coefficients listed in Table A.1l.
TABLE A.8. GRAVIMETRIC DUST ANALYSES
Dustlype
Date
%Cl
%S0 3
Discard
19/10/75
2.00
—
1
11/11/75
2.08
3.17
U
31/11/75
5.97
2. 6
1
4/12/75
2.89
6.53
“
5/12/75
7.11
12.12
IS
6/12/75
4.69
7.27
“
7/12/75
6.18
6.90
“
5/ 1/76
4.71
5.08
“
6/ 1/76
3.72
4.23
U
9/ 1/76
4.43
4.18
U
14/ 1/76
0.46
4.65
“
16/ 1/76
0.90
4.32
Return
5/12/75
-
2.11
7.35
U
13/12/75
4.34
.22
“
1/ 1/76
2.10
4.99
I’
•
18/ 1/76
•
1.00
5.66

-------
86
TABLEA.9. RESULTS.FROM LECO INDUCTION FURNACE ANALYSES
Material . Date % $03 (Total S as)
Clinker 19/10/75. 0.93
21/10/75 1.34
23/10/75 0.75
I ’ 31/10/75 0.30
I’ 1/11/75 0.16
5/11/75 0.66
.8/12/75 0.93
I
‘I 31/12/75 1.39
1/ 1/76 1.43
11/ 1/76 1.04
20/ 1/76 1.42
Slurry Feed 13/10/75 0.45
20/10/75 0.42
I’ 27/10/75 0.41
II 9/11/75 0.45
6/12/75 052
‘I 9/12/75 0.48
13/12/75 0.60
II 14/12/75, 0.55
31/12/75 0.43
IS 3/1/76 0.47
ii 8/ 1/76 0.46
SI
21/ 1/76
0.47

-------
87
TABLE A.10. RESULTS FROM ATOMIC ABSORPTION A jALYSES
Material Date % K 2 0 % Na 0
Discard Dust 11/11/75 6.36 0.32
31/11/75 10.30 0.39
4/12/75 7.14 0.36
6/12/75 8.23 0.44
7/12/75 9.96 0.45
1 5/1/76 812 043
6/ 1/76 6.43 0.34
I ’ 9/ 1/76 7.94 0.39
14/ 1/76 4.84 0.35
16/ 1/76 4.18 0.34
Return Dust 5/12/75 7.79 0.42
IS 13/12/75 5.54 0.41
I’ 1/ 1/76 6.02 0.41
Clinker 19/10/75 1.32 0.23
I ’ 21/10/75 1.61 0.25
23/10/75 1.00 0.20
II 31/10/75 0.54 0.21,
1/11/75 0 40 0.21
51 5/11/75 1.01 0.21
8/12/75 0.64 0.20
31/12/75 1.18 0.21
H 1/1/76 1.31 0.24
11/ 1/76 1.10 0.22
‘S 20/ 1/76 1.45 0.26
Slurry Feed 13/10/75 0.93 0.19
20/10/75 0.93 0.22
27/10/75 0.89 0.18
9/11/75 0.88 0.20
6/12/75 0.91 0.19
‘S 9/12/75 0.90 0.22
SI 31/12/75 0.90 0.18
I’ 3/ 1/76 0.91 0.22
8/ 1/76 0.93 0.19
“ 21/ 1/76 0.94 0.22

-------
- 88
TABLE A.Il. - LEAST SQUARES-DATA FOR CALIBRATION LINES
I
Material
-
-
Element
Correlatlän
Coefficient
-
Slope
intercept
Raw Meal +
Slurry Feed
Potassium (as 1(20)
Chlorine
Sulphur (as $03) -
0.9172
- .0.9995
0!9913
27128 -0.01
5260 0.060
739 0.27
Clinker
Potassium (as 1(20) 0.9829
Chlorine 0.9999
Suiphur (as SO 3 ) - 0.9917
24980
4781
400
-0.01
0.011
—0.23
Dust
Potassium (as 1(20)
Chlorine
Sulphur (as SO 3 )
0.9893
0.9548
3166
1927
278
-0.12
-0.14
-0.10
A.8 Determination of Heat Value, Chlorine Content and Sulphur
Content in No. 6 Fuel
The heat value was determined using standard methods on a Parr
Calorimeter Model 1241 equIpped with Parr oxygen bomb Model 1108. Sulphur
was determined using the standard barium precipitation method on the
washings from the bomb. Where chlorine-content was required, the
washings were analyzed for chlorine by the standard Volhard method.
A.9 DetermInat Ion of Heat Value end Chlorine Content In Chlorinated
Hydr carbons
The heat value was determined on a Parr calorimeter model l2 l
equipped with Parr oxygen bomb model 1108 following standard operating
procedures for determining hea.t value of oil. Due to the corrosive nature
of the combustion products, It Is recoimiended that the chlorinated
materials be diluted with standard oil. it has been found that, while
corrosion of the interior of the bomb is extensive, burning these materials
without dilution gives similar results. Time between weighing and
igniting Is extremely important due to the volatile nature of some of
these compounds.

-------
89
After igniting and determining the heat value of the chlorinated
material, the interior of the bomb iS thoroughly washed with distilled
water -into a volumetric flask. Aliquots of the proper size for the
chlorine content expected ware taken. Nitric acid is added to allow a
more distinct end point, and the chlorine Is determined by the standard
Voihard titration.
REFERENCES
A.l Ontario Ministry of the Environment, Source Testing Code.
January, 1973.
A.2 Standard Reference Methods for Source Testing: Measurement of
Emissions of Particulates from Stationary Sources. Environmental
Protection Service-Report EPS. l—AP-74-l, Air Pollution Control
Directorate, Environment Canada, February, 1974.
A.3 Pellizan, E.D., J.E. Bunch, and B.H. Carpenter. Env. Sd .
- -
- Technology 9 (b),.552-560 (1975).

-------
93
APPENDIX B
ANALYTICAL DATA, CALCULATION AND DETAILS OF
EXPERIMENT ON THE SUSPENSION PREHEATER KILN
B.l Theoretical
The excellent article by Weber cited in the body-of the report
describes in detail the problem of alkali reduction In a su pension
preheater kiln, and was used in preparing the following notes.
At material temperatures above 800°C in the rotary kiln.
alkalies (K 2 O and Na 2 0) are volatilized from the material being burned
and are carried along with the exit gases to- the kiln inlet. These
condense at gas temperatures below 900°C. In part, they are recaptured
by the material, particularly in the preheater, and are thus carried
back to the sintering zone, so that an Internal alkali cycle is established.
The alkalies are in partalso precipitated at the transition from the
preheater to the kiln. -
- Coatings thus formed Interfere with material and gas flows
and, in the most severe cases, completely plug the system.
In preheater kilns, the raw meal Is heated in the finely divided
condition in the gas stream and these particles act as condensation nuclei
for the alkalies. Between 81 and 97% of the (20 volatilized in the kiln
is trapped in the preheater.
The greater the alkali cycle, and therefore the concentration
of alkali in the gas, the more alkali will condense on the bottom
ducts and cyclone stages. If the alkali is very volatile, accretions
are likely to occur in the preheater. However, the alkali cycle can be
reduced by the provision of a bypass, that Is by drawing off a portion
of the gas at the bottom of the preheater. (See Figure 3, Section 2 of
the report.) -
- Of alkali which enters the kiln system with the raw meal, a
proportion (c 1 ) will be volatilized and the remaining c ) will be
discharged with the clinker. If the proportion of gas withdrawn through
the bypass is V then of the volatiljzed alkalies, the proportion c 1 •V
will be removed by the bypass. The remaining portion l (l-V) will return
with the feed, thereby giving rise to the internal cycle in the kiln.
- Pages 90, 91, and 92 blank

-------
94
The alkali in the internal cycle has a different volatility c 2 than
alkalT from the raw meal which has not yet been volatilized. By defining
total alkali including raw meal alkali into the kiln by K, then the quantity
of alkali from the cycle alone is K—l. Of this quantity, (K-l)•c 2 •V is
removed by the bypass.
Equilibrium is attained when the total quantity of alkali input
with raw meal and fuel is equal to the quantity output by the bypass and
retained with, the clinker. This is most easily determined by considering
only alkali in the cycle. Entering the cycle is the volatile raw meal
alkali — Cl; leaving the cycle are a proportion in the clinker (K-l)•( l-c 2 )
andaproporç’lon through the bypass— [ + (K-i) c 2 )].V.
- By setting the quantities i nput and output from the cycle as
equal, the equation -may be solved to determine the amount of gas the
bypass must r emove to keep the alkali. cycle from becoming excessive.
B.2 Experience at St. Lawrence cement
At it.’ Lawrence Cement, san p’les are taken of the material
between the-Stage IV. ;cyclone-and- th kiln inlet (se Figure 3), this
material being called, not quite appropriately, “Stage IV dust”. From
experience at St. Lawrence Cement,-it was known that a chloride content
of 2.5 chlor’ine in this material (natural basis) is excessive. The
quantity of g as required to be withdrawn by the bypass to maintain lower
chloride levels was calculated on the assurnption -that the chloride
volatilities were — 0.99 and 1.0 (completely volatile). These
quantities are given in Table B.l. -
While the bypass was designed to remove the quantities of
gas required° For the program, attempts to reach this level resulted
in excessive emissions of particulate matter from the bypass precipitator.
These were attributed to excessive gas F lows. To alleviate this problem,
measures to increase cooling water and decrease the quantity of ambient
air were undertaken, with limited success.

-------
95
TABLE 8.1. PERCENT BYPASS GAS REQUIRED TO MAINTAIN CHLORIDE LEVELS
-s-
Chloride
addition
relative
to clinker
% (20
in
clinker
Bypass
percent
Stage
required to maintain
IV Cl at:
1.56%
nat.
or 2.0%
ign.
2.0%
nat.
or 2.56%
ign.
2.5%
nat.
or
3.2%
ign.
0
1.31
2.1
1.6
1.3
0.05
1.24
4.6
3.6
2.9
0.10
1.18
7.1
5.5
4.4
0.15
1.11
9.6
7.5
6.0
0.20
1.04
12.1
9.4
7.0
0.30
0.91
17.1
11.7
10.6
0.40
0.77
22.2
17.2
13.8
B.3 Equation of Approximated Time for Equilibrium
It was essential to the success of the program to determine
the time required to attain the equilibrium state. To obtain an approxi-
mation of the time required, and realizing that it is only an approxima-
tion since retention time In kiln and absorption phenomena influence
this time, the following approach was taken:
— Cl —Cl . —Cl
dt RN clinker bypass
where:
= change In quantity with tirñe (pounds/minute)
CIRM chlorine/time In with raw meal (pounds/minute)
Cl . chlorine/time out with clinker (pounds/minute)
clinker
chlorine/time out through bypass (pounds/minute)
C 1 RM and Cll;k are approximately constant, hence they can be
combined as:
I = Cl —Cl
ci RN clinker

-------
96
and Cib — VQ Ct)
where:
Q(t) quantity in transition chamber at time (t) in pounds
V — bypass valve proportion.
Then:
— 1 cl - VQ (t)
Solving for time gives:
in (I . VQ (t) ] Q 0
t = C minutes.
V
Since In 0 — must be taken-as a very small distance from
the true equilibrium.
The sampling sequence at Stage.IV was-planned from this model.
8.4 Sampling and Analyt!cal Methods -
Analytical methods were those described in Appendix A. Samples
of clinker, raw meal feed, bypass precipitator dust and conditioning
tower dust were taken at two hour intervals.. These were then blended
into 24-hour composite samples. On the-basis of the calculation in
Section B.3, samples were taken froniStage IV at 3 to 10 minute
intervals when attempting to start the-chlorinated burn. These 1 atter
samples were analyzed Individually.
B.5 Discussion
In both cases described, the bypass was operated to withdraw
9 percent of the gases from the preheater tower. This was the maximum
attainable due to the emission rate of particulate matter. On the first
attempt to burn chlorinated hydrocarbons, the quantity to yield 0.24
percent chlorine relative to clinker was fed starting at 09:38, June 3,
1975. Plugging of the duct between the kiln and the conditioning wwer
caused a bypass shutdown on June 5, 1975. It was not possible to
maintain feed of chlorinated material to the kiln unless the bypass was
functioning. On June 10, 1975 a second and similar attempt was started

-------
97
at 11:20. The bypass again failed due to accretion on the duct on
June 13, 1975. On the second attempt, sampling of Stage IV material
was continued for a longer period on the basis of the first set of
resul ts.
It was realized with the failure that major revisions to the
bypass system would be required to continue this experiment on the
suspension preheater kiln.
8.6 Analytical Results
Analytical results are given in Tables 3.2 through 0.4. The
raw meal feeds showed a gradual increase in both the content of chlorine.
and (20. The results from the clinker analysis (Table B.3) showed that
the burning of chlorinated hydrocarbons resulted in a reduction in the
I( O content (June 3-5, 1975 and June 10-12, 1975).
The results of the analyses from Stage IV dusts (Table 0.4)
show that the chlorinated hydrocarbons were destroyed in the burning
‘irocess. Chloride content of these samples Increased while burning
chlorinated wastes. This finding could not have resultedwithout,
breakdown of the hydrocarbons since the raw material is at too high
temperature at this position in the preheater for the chlorinated
hydrocarbons to condense upon it. By the methods described in Section
B.3, the approximate time to equilibrium was calculated to be about
90 minutes. It was also expected that an additional timeof approximately
20 minutes would be required to travel through the kiln. The actual
curve is less steep and equilibrium is reached In about four hours.
For the June 3, 1975 burn, the curve was extrapolated to the calculated
equilibrium (Figure 8.1). On June 10, 1975 samples were taken for a
longer period. Although there are points above the calculated curve
(Figure 3.2), the data from June 11, 1975 show that equilibrium was
reached; the higher values found indicate random sampling or
analyticaL errors. -

-------
98
TABLE B.2. RESULTS FROM ANALYSES OF DRY PROCESS KILN RAW FEEDS
Date
..
% Cl
%S0
(Total S 3 as)
% K 2 0
30/5/75
0.067
0.60
0.92
31/5/75
0.069
0.60
•
0.95
1/6/75
0.073
0.60
0.93
2/6/75
0.071
0.60
0.93
3/6/75
0.074
0.60
0.93
4/6/75
0.082
0.60
0.95
5/6/75
0.079
0.60
0.96
6/6/75
0.094.
0.62
0.98
7/6/75
0.087
•
0.60
0.97
8/6/75
0.093
0.60
0.97
9/6/75
0.088
0.58
0.97
10/6/75
0.084
0.60
0.95
11/6/75
12/6/75
.
0.092
— -
L
NO
0.60
SAMPLE
0.98
-
-
13/6/75
-o. 084
0.60
0.96
14/6/75
0.104
0.60
1.00
15/6/75
0.102
•
0.62
1.00

-------
99
TABLE B.3. RESULTS FROM ANALYSES OF DRY PROCESS KILN CLINKER
Date Time % Cl % SO 2 % (20
30/5/75
Average
0.03
1.05
1.27
31/5/75
‘I
0.03
1.15
1.29
1/6/75
‘I
0.03
1.15
1.31
2/6/75
U
0.21
1.16
1.30
3/6/75
“
0.03
1.10
1.24
4/6/75
1
0.05
1.19
1.30
5/6/75
“
0.04
1.23
1.33
6/6/75
U
0.04
1.34
1.44
7/6/75
U
0.07
1.33
1.48
8/6/75
“
0.05
1.31
1.43
9/6/75
U
0.06
1.50
1.63
10/6/75
U
0.04
1.22
1.29
10/6/75
12:00
0.06
1.38
1.25
10/6/75
14:00
0.04
1.00
0.97
10/6/75
16:00
0.04
1.07
0.87
10/6/75
18:00
0.05
1.30
1.10
10/6/75
, .10/6/75.
11/6/75
1I/6/75
20:00
.‘22: 0O
Average
00:00
0.05
0.05
0.06
0.04
1.03
.4.11-
1.29
1.18
0.96
1.05
1.24
1.27
11/6/75.
02:00
0.06
1.36
1.25
11/6/75
04:00
0.07
0.98
1.00
11/6/75
06:00
0.06
0.96
0.99
11/6/75
08:00
0.06
1.10
1.10
11/6/75
10:00
0.05
1.08
1.08
11/6/75
14:00
0.05
1.18
1.15
11/6/75
16:00
0.05
1.08
1.13
11/6/75
18:00
0.06
1.12
1.14
11/6/75
20:00
0.05
1.23
1.21
11/6/75
22:00
0.05
1.01
1.06
12/6/75
Average
0.07
1.28
1.29
12/6/75
00:00
0.07
1.07
1.15
12/6/75
02:00
0.05
1.12
1.16
12/6/75
04:00
0.05
1.11
1.15
12/6/75
:
06:00
-
0.05
1.10
1.16
12/6/75-
08:00
0.05
1.11
1.18
12/6/75
10:00
0.10
1.25
1.36
12/6/75
12:00
0.05
1.11
1.15
12/6/75
14:00
0.06
1.23
1.28
12/6/75
16:00 .
0.05
1.11
1.17
12/6/75
18:00
0.05
1.14
1.18
13/6/75
Average
0.05
1.29
!.ko
14/6/75
f l
0.05
1.40
1.52
15/6/75
“
0.04
1.35
1.41

-------
00
30/ 5/76 0.67 1.14 1.77
TABLE B.4. RESULTS
FROM ANALYSES OF STAGE IV DUSTS
- ,
Date - Time -
%S0 3
C l : ota1 ,,S as) % 1(20
2/ 6/75
‘-
0.82
1.30
2.26
3/ 6/75
09:00
0.86
1.23
2.11
3/ 6/75
09:06
0.84
1.25
2.03
3/ 6/75
09:12
0.81
1.19
2.03
3/6/75
09:18,
0.85
1.22
2.14
.

3/ 6/75
3/ 6/75
09:24
09:30
0.82
0.86
1.18
1.30
2.06
2.17
3/ 6/75
09:38
0.90
1.34
2.15
3/ 6/75
09:44
0.97
1.32
2.21
3/ 6/75
09:47
0.96
1.35
2.38
3/ 6/75
09:50
1.02 -
1.38
2.39
3/6/75
09:53
1.20
1.29
2.37
3/ 6/75
09:56
1.21
1.31
2.42
3/ 6/75
09:59
1.21
1.34
2.51
.
3/ 6/75
10:02 ‘
1.22
1.30
,
2.55
3/ 6/75.
‘10:0
,
1.34 - ‘
1.46’
2.68
3/:6/75
31-6/75
3/6/75
10:08.
10:1,1.
10:14
1.44.
i.51
1.44
‘1 ’38 ’

‘1.29
2.66’
2.95
2.91
3/.6/75
10:23
1.56
1.41
2.94
3/ 6/75
10:32
1.79
.
1.37
3.38
3/ 6/75
10:41
1.83
1.34
3 :43
3/ 6/75
10:54
1.85 ‘
1.32
•
3.35
3/ 6/75
10:59
1.95
1.46
3.74
3/ 6/75
.
11:04
2.08
1.59
3.85
3/ 6/75
11:09
2.03
1.35
3.72
3/ 6/75
11:14
2.00
1.39
3.66
3/ 6/75
11:19
2.03
1.41
3.70
3/ 6/75
11:24
2.30
1.47
4.15
3/ 6/75
11:29
2.17
1.36
3.93
3/ 6/75
.
11:34’
2.21
1.37
4.04
3/ 6/75
11:39
2.22
1.43
4.00
3/ 6/75
11:44
2.29
1.51
4.18
3/ 6/75
11:49
2.29
1.57
4.11
3/ 6/75
11:54
2.32
1.66
4.24
3/ 6/75
11:59
2.38
‘
1.47
4.36
3/ 6/75
12:04
2.49
1.51
4.45
3/ 6/75
12:09
2.28
1.1,6
4.12
3/ 6/75
12:14
2.34
1.44
4.21
3/ 6/75
14:00
2.47
1.45
4.44
3/ 6/75
15:30
2.84
1.58
5.07
4/ 6/75
09:00
1.92
1.142
4.17
4/ 6/75
14:00
1.59
1.28
3.48
5/ 6/75
-
0.97
1.4i.
2.54
6/ 6/75
—
1.11.
-
1.17
2.71

-------
101
TABLE B.4. (Cont’d)
t
Date Time % Cl (Total
so 3
S as) % 1(20
9/ 6/75
-
1.66
1.20
3.79
10/ 6/75
09:00
1.09
1.18
2.64
10/ 6/75
09:10
1.17
1.22
2.93
10/ 6/75
09:20
1.19
1.36
3.07
10/ 6/75
09:30
1.15
1.39
2.92
10/ 6/75
09:40
1.16
1.29
3.01
10/ 6/75
09:50
1.16
1.27
2.99
10/ 6/75
10:00
1.14
1.29
2.95
10/ 6/75
11:15
1.11
1.29
2.91
10/ 6/75
11:25
1.20
1.34
3.04
10/ 6/75
11:35
1.46
1.41
3.22
10/ 6/75
12:05
1.91
1.49
4.02
10/ 6/75
12:25
1.99
1.46
4.07
10/ 6/75
12:35
2.15
1.64
4.50
10/ 6/75
12:45
2.25
1.59
4.54
10/ 6/75
12:55
2.31
1.66
4.65
10/ 6/75
10/ 6/75
13:05
13:15
2.45
2.37
1.79
1.71.:
4.98
10/ 6/75
13:25
2.32
1.47
4.48
10/ 6/75
13:35
2.62
1.67
5.10
10/ 6/75
13:45
2.51
1.54
4.83
10/ 6/75
13:55
2.63
1.69
5.12
10/ 6/75
14:15
2.57
1.69
5.02
10/ 6/75
14:25
2.87
1.81
5.53
10/ 6/75
14:35
2.74
1.67
5.18
10/ 6/75
14:45
2.85
1.70
5.35
10/ 6/75
14:55
2.77
1.77
5.23
10/ 6/75
15:05
2.79
1.52
5.23
10/ 6/75
15:25
3.02
2.41
5.66
10/ 6/75
15:35
3.28
1.59
6.07
10/ 6/75
15:45
3.09
1.62
5.78
11/ 6/75
-
2.58
1.51
5.23
12/ 6/75
-
2.42
1.49
5.13

-------
w
z
0
1
0
CHLORINE
LEVEL
iN STAGE
IV
3 /611975
2,80
a
0
0’
0-
0
0
180
00
Q - CALCULATED
0
0.80
b - ACTUAL
10:00 11:00 12:00 13:00 14:00 15:00 TIME
FIGURE B.1

-------
3.00
w
z
O 2.00
-J
I
C-)
1.50
1.00
9:00 11:00
CHLORINE LEVEL IN STAGE
IV .10161 1975
2.50
a
b
0
Q— CALCULATED
b—ACTUAL
13:00 15:00 TIME
FIGURE B.2

-------
107
- APPENDIX C
RESULTS AND CALCULATIONS FOR WET PROCESS SYSTEM
C. 1 Mass Balance Experimentation
C.1.1 Results of analyses of process materials
Analyses of process materials for the mess balance were
carried out at the St. Lawrence Cement Co. plant. Analytical results
are given in Tables C.l to C.6.
C.l.2 Calculation of material balances
A dafly record of production and material consumption is given
in Table C.7 for the period October 7 1975 to January 21, !976. Produc-
tion was disrupted during this period by a kiln shutdown for- the removal
of a kiln ring.
The program was further disrupted by formation of a ring,
not requiring kiln shutdown, but time to reduce the quantity of aromatic
phis complex chlorinated hydroëarbori in-the tank and-difficulty in -
scheduling deliveries. From the daily composite analysis.of each material
for eaàh element, and the total daily quantity of each material, daily
elemental quantities (as pounds or tons-per 25 hour period) were determin-
ed by the relationship:
X.. N.
Ii iJ 100
where -
— total daily weight of element x in material I on day j
C 1 concentration of element x in the composite sample material
I on day j.
These data were then used to form the separate Individual
elemental balance accounts given in Tables C.8 and C.9. Due to process
fluctuations and the very low concentrations In the case of chlorine,
little significance should be attached to individual daily balance results,
except as they indicate the responsiveness of the system to major changes.-
in input of chlorinated hydrocarbons. Significant assessment of the
Pages 104, 105, and 106 blank

-------
I 08
balance is available by considering prolonged periods of plant operation.
Summary balances for this purpose are given in the body of this report
(see Section5.4, Tables 8 and 9).
Materials entering the miss balance calculation s (see Figure 1,
Section 2) were slurry feed, No. 6 fuel oil and chlorinated hydrocarbon
materials as Input streams with clinker and discard dust forming the
output streams. The. return’ dust Is returned almost immediately to the -
system and does not, form a art of the mass balance.

-------
109
TABLE C.1.
RESULTS FROM
ANALYSES OF CLINKER SAMPLES
Date % Cl % SO 3 % K 2 0
7/10/75
8/10/75
9/10/75
10/10/75
11/10/75
12/10/75
13/10/75
14/10/75
15/10/75
16/10/75
17/10/75
18/10/75
19/10/75
20/10/75
2 1/10/75
22/10/75
23/10/75
24/10/75
25/10/75
26/10/75
27/10/75
28/10/75
29/10/75
30/10/75
3 1/10/75
1/11/75
2/11/75
3/11/75
4/11/75
5/11/75
6/11/75
7/11/75
8/11/75
9/11/75
0.033
0.032
0.034
0.042
0.032
0.029
0.030
0.031
0.031
0.031
0.035
0.029
0.032
0.042
0.030
0.029
0.030
0.030
- 0.029
•0.031
0.031
0.034
0.032
,0.039
0.034
0.039
0.032
0.028
0.031
0.031
0.037
0.029
0.035
0.88
0.79
0.86
0.62
0.93.
0.80
0.80
0.86
1.27
0.90
1.11
0.99
0.97
1 .03
1 . 32
0.99
0.77
0.92.
0.59
1.06’
0.82
1.16
1.27
0.88
0.34
0.18
NO SAMPLE
0.26
0.30
0.64
0.70
0.71
0.68
0.72
1.13
1.12
1.14
0.88
1 .27
1.08
1 .09
1.15
1 .47
1.16
1.31
1 .24
1.21
1 .28
1.51
1.28
1 .06
1.20
0.83
1.24
1 .00
1 .31
1 .37
0.96
0.49
0.36
0. 48
0.56
1 .00
1.18
1 .20
1.15
1.19
2/12/75
3/12/75
4/12/75
5/12/75
6/12/75
7/12/75
8/12/75
9/12/75
10/12/75
11/12/75
12/12/75
13/12/75
14/12/75
0.031
0.029
0.034
0.033
0.029
0.028
0.038
0.040
0.034
0.035
0.032
0.032
0 ..03 1
1.13
1 .39
1 .02
1 .23
1.18
1.18
0.91
0.71
0.47
0.70
0.92
0.75
0.73
0.94
1 .28
1.01
1.16
0.90
0.84
0.74
0.56
0. 36
0.48
0.54
0.39
0.48

-------
110
TABLE C.1 (CONTD)
- Date Cl SO 3 % K 2 0
• 31/12/75 0.029 1.40 1.20
- 1/ 1/76 0.029 1.48 1.34
2/ 1/76 0.029 1.49 1.37
3/ 1/76 0.033 1.12 0.77
4/ 1/76 0.030 1.68 1.42
5/ 1/76 0.035 1.14 0.78
6/ 1/76 0.031 0.70 0.58
7/ 1/76 0.029 1.14 1.26
8/ 1/76 0.030 0.88 0.83
9/ 1/76 0.028 0.46 0.46
10/ 1/76 0.028 0.91 1.05
11/ 1/76 0.030 0.92 1.12
12/ 1/76 0.030 0.94 1.15
13/ 1/76 0.029 1.13 1.35
14/ 1/76 0.030. 1.14 1.28
15/ 1/76 0.029 1.12 1.22
16/ 1/76 0.030 1.38 .1.45
17/-1/76 0.037 1.36 1.42
18/ 1/76 0.028 1.20 .1.27
19/ 1/76. -0.028’. 1.06 1.15,
‘20/ 1/76 0.029 1.42 1.46
21/1/76 0.029 0.94 -1.12

-------
111
TABLE C.2 RESULTS FROM ANALYSES OF SLURRY FEED SAMPLES
SO.
Date % Cl (Total as) % K 2 0
7/10/75 0.080 0.48 0.92
8/10/75 0.056 0.48 0.91
1/10/75 0.052 0.45 0.91
10/10/75 0.058 0.46 0.91
11/10/75 0.060 0.46 0.92
12/10/75 0.068 0.45 0.92
13/10/75 0.050 0.50 0.93
14/10/75 0.044 0.44 0.92
15/10/75 0.046 0.44 0.93
16/10/75 0.043 0.43 0.93
17/10/75 0.051 0.43 0.93
18/10/75 0.055 0.43 0.93
19/10/75 0.046 0.44 0.93
20/10/75 0.043 0.42 0.93
21/10/75 .0.038 o;41 0 92
22/10/75 0.040 - 0.40 0.93
23/10/75 0.040 o.4o 0.93
24/10/75 0.050 0.40 - 0.93
25/10/75 0.043 0.41 0.92
26/10/75 0.055 0.42 0.93
27/10/75 0.042 0.43 0.91
28/10/75 0.047 0.41 0.92
29/10/75 0.055 0.43 0.93
30/10/75 — NO SAMPLE -
31/10/75 0.050 0.42 0.91
1/11/75 0.047 0.42 0.91
2/11/75 0.043 0.42 0.92
3/11/75 0.043 0.43 0.93
4/11/75 0.044 0.44 0.92
5/11/75 0.042 0.41 0.93
6/11/75 0.042 0.43 0.92
7/11/75 0.044 0.41 0.92
8/11/75 0.042 0.40 0.91
9/11/75 0.050 0.43 0.94
2/12/75 0.046 0.48 0.90
3/12/75 0.042 0.48 0.90
4/12/75 0.041 0.47 0.89
5/12/75 0.040 0.52 0.91
6/12/75 0.040 0.51 0.90
7/12/75 0.052 0.53 0.91
8/12/75 0.041 0.51 0.91
9/12/75 0.040 0.51 0.89

-------
112
TABLE C.2 (C0NT’D)
Date
% Cl
CT
% SO 1
otal  as)
% K O
10/12/75
11/12/75
12/12/75
13/12/75
14/12/75
0.041
0.039
0.050
0.042
0.048
0.54
0.52
0.55
0.58
0.54
0.91
0.91
0.91
0.90
0.91
31/12175
1/ 1/76
2/ 1/76
3/ 1/76
4/ 1/76
5/ 1/76
6/ 1/76
7/ 1/76
8/ 1/76
9/ 1/76
10/-1/76
11/ 1/76
12/ 1/76
13/ 1/76
14/ 1/76
15/ 1/76
16/ 1/76
17/ 1/76
18/ 1/76
19/ 1/76
20/ 1/76
21/ 1/76
0.059
0.042
0.045
0.043
0.046
0.038
0.038
0.048
0.039
0.048
o.o’e8
0.048
0.040
‘.0.041
0.044
0.042
0.048
0.041
0.054
0.039
0.040
o.o4
0.48
0.46
0.46
0.46
0.46
0.46
0.46
0.47
0.46
0.45
0.46
0.44,
-:0.44
0.44
0.46
0.47
0.50
0.47
0.46
0.44
0.43
0.43
0.90
0.91
0.92
0.91
0.92
0.91
0.91
0.91
0.89
0.92
0.9 1
0.91
0.91
0.92
0.91
0.91
0.91
0.90
0.91
0.91
0.91
0.92

-------
113
TABLE C.3. RESULTS FROM ANALYSES OF DISCARD DUST
%S03
Date % CL (Total S
as) K O
7/10/75
8/1 0/75
9/10/75
10/10/75
11 / 10/75
-12/10/75
13/10/75
14/10/75
15/10/75
16/10/75
17/10/75
18/10/75
19/10/75
20/1 0/75
21/10/75
22/10/75.
23/10/75
24/1 0/75
25/10/75
26/10/75
27/10/75
28/10/75
29/10/75
30/10/75
3 1/10/75
1/11/75
2/11/75
3/11/75
4/11/75
5/11/75
6/11/75
7/11/75
8/11/75
9/11/75
2/12/75
3/12/75
4/12/75
5/12/75
6/12/75
7/12/75
8/12/75
-9/1 2/75
0.99
1.01
0.89
0.96
0.96
1.35
2.51
1 . 64
2.40
1 .88
1 .86
1.77
2.08
1 .41
3.08
5.11
4.28
4.01
1 .39
3.89
6.79
6.62
5.85
5.82
3.09
1.35
1 .07
0.75
0.83
5.33
4.14
2.64
7.14
4.76
5.75
6.24
4.76
7.09
7.22
7.28
7.48
7.39
7.59
10.34
6.65
NO DUST DISCARDED
7.35
6.00
6.02
6.12
NO SAMPLE
NO DUST DISCARDED
6.23
5.65
6.05
5.56
5.49
5.22
5.15
4.84
NO SAMPLE
NO SAMPLE
3.62
3.27
3.10
2.81
4.08
4.12
4.11
3.76
4.13
5.84
6.37
6.51
12.66
6.28
6.41
6.97
5.41
7.37
7.45
7.34
7.66
7.48
8.26
11 .27
7.65
9.16
7.32
7.22
7.13
- 7.83
6.70
8.60
10.07
9.28
8.63
8.87
8.21
9.62
9.38
8.57
8.34
7.46
5.83
5.47
4.67
5.42
9.36
8.78
7.81
16.28
9.10
9.88
10.70
8.24

-------
114
TABLE C.3
(CONT’ D)
10/12/75
11/12/75
12/1 2/75
13/12/75
14/12/75
31/12/75
1/ 1/76
2/ 1/76
3/ 1/76
4/ 1/76
5/ 1/76
6/ 1/76
7/ 1/76
8/ 1/76
9/ 1/76
10/ 1/76
•i1/ 1/76
12/ 1/76
13/ 1/76
14/ 1/76
15/ 1/76
16/ 1/76
17/ 1/76
18/ 1/76
19/ 1/76
20/ 1/76
21/ 1/76
5.35
5.52
5.53
5.68
1 .98
4.55
4.58
4.03
3.06
4.50
1.20
1.11
.01
1 .20
0.76
0.92
0.80
0.78
0.69
0.72
0.76
0.86
5.51
4.90
5.62
NO SAMPLE
5.54
NO SAMPLE
6.19
NO SAMPLE
N0 SAMPLE
4.62
4.62
4.70
NO SAMPLE
5.31
4.51
5 .43;
5.27
5.16
5.37
4.94
4.89
4.41
4.50
4.28
5.01
4.81
4.28
8.27
8.20
8.46
8.65
7.37
7.38
6.89
7.25
7.60
6.07
5.79
5.54
6.10
5.16
5.15
4.62
4.68
4.49
5.32
5.19
4.84
- %SO
Date Cl (Total as) % 1 (20
6.88

-------
115
TABLE C.4. RESULTS FROM ANALYSES OF RETURN DUST
7/10/75
8/10/75
9/10/75
10/10/75
11/10/75
12/10/75
13/10/75
14/10/75
15/10/75
16/10/75
17/10/75
18/10/75
19/10/75
20/10/75
2 1/10/75
22/10/75
23/1 0/75
24/10/75
- 25/10/75 ..
26/10/75
27/10/75
28/10/75
29/10/75
30/10/75
3 1/10/75
1/11/75
2/11/75
3/11/75
4/11/75
5/11/75
6/11/75
7/11/75
8/11/75
9/11/75
10/11/75
2/12/75
3/12/75
4/12/75
5/12/75
6/12/75
7/1 2/75
8/12/75
9/12/75
0.80
0.73
0. 54
0.73
1 .59
1 . 36
1 .29
2.46
2.29
2.15
2.33
1.81
1.86
1 .76
2.56
1 .53
1;52
‘3.82
3.41
3.87
3.48
3.02
3.51
4.22
4.66
6.37
3.96
4.00
1 .27
1 .00
0.79
0.54
0.44
4.00
2.52
2.49
2.58
1 .00
0.7
0.58
3.68
6.25
6.28
5.36
7.26
7.05
7.79
6.21
8.16
6.91
6.80
7.20
5.61
6.51
5.86
6.54
6.00
5. 3 7•
5.36
4.87
4.84
4.48
4.75
4.24
NO SAMPU
2.58
2.84
3.22
2.47
2.70
3.60
3.67
3.77
3.54
3.30
NO SAMPLE
5.65
5.97
6. .51
6.96
3.69
3.81
3.57
4.68
6.22
6.25
5.15
7.54
8.20
8.90
7.02
9.83
8.54
8.32
8.73
6.71
7.73
7.08
8.44
.6.92
-6.60
8.63
.7.75
8.21
7.47
6.91
7.23
6.49
7.07
9.05
6.28
6.69
5.20
4.91
4.92
4.50
3.89
10.70
9.63
8.58
8.33
4.94
4.93
4.61
6.59
$03
Date % Cl (Total S as) (20

-------
116
TABLE C. 1 e (coWr’D)
- %S0 3
Date %C1 (Total S as) % K 2 0
10/12/75 2.51 3.68 4.60
11/12/75 2.36 3. 1*5 4.22
12/12/75 3.69 4.22 5.94
13/12/75 4.011 4.07 6.05’
14/12/75 3.02 4.06 5.16
31/12/75 1.56 5.09 5.60
1/ 1/76 1.78 5.14 5.99
2/ 1/76 1.77 4.70 5.65
3/ 1/76 2.42 4.62 5.82
4/ 1/76 3.55 4.48 6.33
5/ 1/76 2.37 3.36 4.29
6/ 1/76 2.56 3.43 4.64
7/ 1/76 1.35 3.23 3.58
8/ 1/76 1.76 4.91 5.56
9/ 1/76 2.30 3.71 4.91
10/ 1/76 1.06 4.711 5.05
I’l/ 1/76 0.93.. 5.05 5.46
12/ 1/76 0.80 ‘5.14 5.37
13/ 1/76 0.79 .4;82 5.10
14/ 1/76 0.78 4.85 4.93
15/ 1/76 0.78 4.41 ‘4.59
16/ 1/76 0.65 4.27 4.38
17/ 1/76 0.58 4.11 4.10
18/ 1/76 0.69 4.19 4.22
19/ 1/76 0.58 4.56 4.70
20/ 1/76 0.72 4.47 4.62
21/ 1/76 0.87 4.24 4.83

-------
117
TABLE C.5. Btu AND CHLORINE CONTENT OF CHLORINATED HYDROCARBONS (SAMPLES
FROM LIME TO KILN)
Material
Date
Btu/lb
%
Chlorine
Viscosity
(Centipoise
23°C)
Specific
Gravity
Aliphatic
24/10/75
27/10/75
28/10/75
3/11/75
4/11/75
13/11/75
13400
11390
10750
8900
8590
8340
21.52
32.76
39.40
42.23
43.52
42.38
,
-
18/11/75
Average
8410
9970
43.38
37.88
1.13
Aromatic +
10/12/75
9530
40.56
Complex
11/12/75
12/12/75
13/13/75
15/12/75
Average
9500
8820
9320
9310
9300
45.91
44.16
41.80
40.48
42.58
40.8
1.27
Polychiorinated
Biphenyl
3/ 1/76
4/ 1/76
-5/ 1/76
6/ 1/76
7/ 1/76
8/ 1/76
9/ 1/76
Average
ll38O
l159O
11880
12170
12070
12050
12000
11880
36.16
37.75
34.90
33.90
33.19
34.93
33.90
34.97
17.0
1.18

-------
118
TABLE C.6. Btu, S AND Ci ANALYSES FROM NO. 6 FUEL OIL
Date Btu/ lb - % S Cl
7/10/75 17809 2.1 .0
8/10/75 17222 ‘2.42
9/10/75 17978 2.34
10/10/75 17942 2.26 - Average
14/10/75 18106 2.26 7/10/75 - 22/10/75
15/10/75 18008 2.28 % Chlorine 0.028
16/10/75 17920 2.31
17/10/75 17988. 2.30
20/10/75 17924 2.24
21/10/75 17877 2.24
22/10/75 17984 2.09
23/10/75 18189 2.23
24/10/75 17701 222
27/10/75 17901 2.21.
28/10/75 17862 2.22
29/10/75 17970 1.95
30/10/75 17917 2.29
31/10/75 17889 1.77
3/11/75 18135 ‘1.62.
4/11/75 18077 1.57
5/11/75 ‘ 18047 1.55
6/11/75 18079 1.51 Average
7/11/75 18036 1.56 23/10/75-12/11/75
10/11/75 17979 1.26, % Chlorine = 0.064
11/11/75 17782 1.89
12/11/75 17719 2.06
2/12/75 17700 2.51.
3/12/75 17676 2.28
4/12/75 17714 2.56
5/12/75 17648 2.58
8/12/75 17679 2.55 Average
9/12/75 17732 2.52 2/12/75-15/12/75
10/12/75 17695 2.58 % Chlorine 0.047
11/12/75 17596 2.50
12/12/75 17655 2.56
15/12/75 17504, 2.64
30/12/75 17678 2.59
2/ 1/76 17565 2.41
5/ 1/76 17498 1.61 Average
6/ 1/76 171.33 2.07 30/12/75-9/1/76
7/ 1/76 17630 2.28 % Chlorine • 0.030
8/ 1/76 17900 2.47
9/1/76 17719 2.42
12/ 1/76 17990 1.96
13/ 1/76 17896 1.95

-------
TABLE C.6
(CONT’ D)
Date
Btu/lb -
% S
% C l
14/
1/76
17935
1.54
15/
1/76
17952
2.00
16/
1/76
18025
1.98
Average
19/
1/76
17762
1.92
12/1/76—22/1/76
20/
1/76
17846
1.89
% Chlorine — 0.038
21/
1/76
18005
1.54
22/
1/76
18160
1.42
119

-------
120
TABLE C.7. DAILY RECORD OF PRODUCTION AND MATERIALS CONSUMPTION
SLURRY FEED -
CHLORIN.
KC
CLINKER
DISCARD
DUST
Date (tons/2 1 + hr)
(gal/2 1 +
hr)
(tons/24 hr) (tons/2 4 i
hr)
7/10/75 1653 0 1043 40.44
8/10/75 1676 0 1041 63.44
9/10/75 1681 0 - 1073 24.84
10/10/75 1658 0 1027 65.66
11/10/75 1585 0 1029 0
12/10/75 1325 0 81+5 -18.94
13/10/75 1632 0 . 1036 6.96
14/10/75 1604 0 1047 23.00
15/10/75 1639 0 1064 0
16/10/75 1646 0 1058 12.90
17/10/75 1661 0 1065 17.86
18/10/75 1682 0 . 1081 15.32
19/10/75 1669 0 1079 4.93
20/10/75 1669 0 1071 15.97
21/10/75 1661 0 1078 0
22/10/75 1664 0 1046 45.36
23/1Q/75 1681 1170 1082. 11.96
24/10/75 1686 1440 1079 19.72
25/10/75 1669 1440; 1067 20.96
26/10/75 1732 - 1440 1112 16.42
27/10/75 1656 1440 1061 18.32
28/10/75 1652 1440 - 1040 42.58
29/10/75 1621 1440 1021 1+1.36
30/10/75 1633 2670 1032 37.17
31/10/75 1647 2880 1027 56.52
1/11/75 1658 2880 1029 63.32
2/11/75 1673 2880. 1057 38.92
3/11/75 1679 2880 1030 80.66
4/11/75 1679 930 1010 107.72
5/11/75 1692 0 1041+ 72.82
6/11/75 1693 0 1054 60.46
7/11/75 1698 0 1058 59.90
8/11/75 1670 0 1030 73.50
9/11/75 1701 0 1056 64.26
2/12/75 1622 430 1038 20.58
3/12/75 1627 0 1042 18.30
4/12/75 1644 0 1039 37.72
5/12/75 1653 1640 1028 61.40
6/12/75 1670 2187 . 1056 38.26
7/12/75 1655 1351 1074 0
8/12/75 1666 1117 1043 51.36
9/12/75 1664 2121 981+ 129.78

-------
121
TABLE C.7. (cowr’D)
SLURRY FEED
CHLORIPI.
IC
CLINKER
DISCARD
DUST
Date (tons/24 hr)
(gal/V.
hr)
(tons/V. hr) (tons/2 4
hr)
10/12/75 1676 1997 994 127.66
11/12/75 1669 1308 999 113.76
12/12/75 1669 1821 1024 79.12
13/12/75 1661 1691 985 125.96
14/12/75 1653 1869 956 157.68
15/12/75 1658 1909 980 130.30
31/12/75 1606 0 1024 24.72
1/ 1/76 1595 0 1035 0
2/ 1/76 1608 0 1039 6.11
3/ 1/76 1630 2029 1017 55.43
4/ 1/76 1629 2210 1005 70.90
5/ 1/76 1627 2168 970 116.98
6/ 1/76 1616 1985 1034 19.86
7/ .!/76 1612 604 -. 944 1-37.39
8/ 1/76 1615 1230 10145 4.54
9/ 1/76 1619 1367 1027 32.10
10/ 1/76 1621 0’ 1020- 43.92
11/ 1/76 1627 0 1011 61.20
12/ 1/76 1636 0 1032 40.52
13/ 1/76 1639 0 1040 32.36
114/ 1/76 1618 0 1028 142.80
15/ 1/76 1633 0 1046 19.10
16/ 1/76 1596 0 1026 13.38
17/ 1/76 1606 0 979 85.76
18/ 1/76 1635 - 0 1026 47.62
19/ 1/76 1608 0 1020 32.22
20/ 1/76 1602 0 991 66.56
21/ 1/76 1629 0 1016 55.75

-------
TABLE C.8. MATERIAL BALANCE FOR CHLORINE
‘
CHLORINE
INPUT
(lb/2 1 . hr)
•
CHLORINE
RETAINED
(lb/24
hr)
TOTAL %
RETAINED
SLURRY
16
CHLORINATED
.
DISCARD
DATE
FEED
OIL
HYDROCARBON TOTAL CLINKER
DUST
TOTAL
71.10/75
8/10/75
2,645
1,877
8)
81
O’
0
2,726
1,958’
688
666
801
1,28)
1,489
1,947
.
54.6
99.4
9/10/75
1,748
82
0
1,830
730
442
1,172
64.0
10/10/75
1,923
81
0
2,004
.836
1,261
2,097
104.6
11/10/75
1,902.
78
0
1,980
658
0
658
33.2
12/10/75
1,801
67
0
1,868
490
51)
1,001
53.6
13/10/75
1,632
81
0
1,713
622
349
97)
56.7
14/10/75
15/10/75
1,412
1,508
80
81
0
0
1,492
1,589
649
660
754
0
1,403
660
94.0
41.5
16/10/75
1,4 )6
81
0
1,497
656
6)9
1,275
85.2
17/10/75
18/10/75
1,691$
1,850
81
81
0
0
1,775
1,931
‘
71.6
627.
672
570
1,4)8
1,197
79.9
62.0
19/10/75
1,535
8)
0
1,616
690
174
864
53.5
20/10/75
1,1 .35
81
0
1,516
900
575
1,475
97.3
21/10/75
1,262
80
0
1,342
6’47
0
647
‘.8.2
22/10/75
23/10/75
1,331
1,345
81
185
0
2,845
,
1,412
4,375
•
607
61.9
,
1,887
337
2,494
986
176.6
22.5
24/10/75
25/10/75
.
1,686
1,435
183
183
3,502
3,959
5,371
5,577
.
61.7
6)9
1,215
2,142
1,862
2,76)
34.7
49.5
26/10/75
1,905
190
4,4)6
6,511
689
1,406
2,095
32.2
27/10/75
1,39)
182
4,873
6,446
658
1,469
2,127
33.0
28/10/75
1,553
182
6,411
8,146
707
3,738
4,445
54.6
29/10/75
1,783
183
6,642
8,608
653
3,2)8
3,871
45.0
30/10/75
1,502*
18)
12,316
13,999
805
3,970*
4,775
34.)
31/10/75
1,61.7
186
13,284
15,117
698
6,036*
6,734
44.5
1/11/75
1,558
.
179
13,284
15,021
803
8,599
9,402
62.6
2,’11/75
1,439
179
13,284
14,902
676
5,153
5,829
39.1
3/1)/75
1,441.
178
13,284
14,906
659
9,437
10,096
67.7
4/11/75
1,478
179
13,743
15,400
566
12,539
13,105
85.1
5/11/75
1,421
184
4,574
6,179
647
4,500
5,147
83.3
6/11/75
1,422
182
0
1,604
653
1,632
2,285
142.4
p .3
p .3
* Calculated from average data.

-------
TABLE c.8 (coNT’D)
CHLORINE
INPUT
(lIi/24 hr)
CHLORINE
RETAINED
(lb/24
hr)
TOTAL %
RETAINED
DATE
SLURRY
FEED
#6
OIL
CHLORINATED
HYDROCARBON
CLINKER
DISCARD
DUST
TOTAL
TOTAL
7/11/75
1,’ .94
183
0
1,677
•
783
1,282
2,065
123.1
8/11/75
1,403
181
0
1,584
597
1,102
1,699
107.3
9/11/75
1,701
182
0
1,883
739
1,067
1,806
95.9
2/12/75
1,492
124
2,325
3,941
644
2,194
2,838
72.0
3/12/75
1,367
133
0
1,500
604
1,515
2,119
141.3
4/12/75
1,348
133
0
i,48 1
708
1,992
2,700
182.3
5/12/75
1,322
132
8,868
10,322
678
8,768
9,446
91.5
6/12/75
1,336
131
11,826
13,293
612
3,642
4,254
32.0
7/12/75
1,721
131
7,306
9,158
601
0
601
6.6
8/12/75
1,366
142
6,040
7,548
793
6,410
7,203
95.4
9/12/75
1,331
132
11,470
12,933
787
12,355
13,142
101.6
10/12/75
1,374
130
10,287
11,791
676
13,660
14,336
121.6
11/12/75
1.302
134
7,626
9,062
699
12,559
$3,258
$46.3
12/12/75
1,669
134
10,213
.12,016
655
8,751
9,1.06
78.3
13/12/75
1,395
$33
8,977 -
.10,505-
630
14208
14,838
141.2
14/12/75
1,587
131
10,107
11,825-
593
17,906
18,499
156.4
31/12/75
1,895
85
0
1,980
594
979
1,573
79.4
,
Il 1/76
2/1/76
1,340
I,447
87
87
0
0
1,427
1,534
600-
603
0
242
600
845
42.0
55.1
.3/ 1/76
4/ 1/76
1,402
1,499
83
82
8,657
9,844
10,142-
11,425
671
.603
4,431k
6,452
5,105
7,055
50.3
61.8
5/.1/76
6/ 1/76
7/ 1/76
1,236
1,228
-1,548
83
83
85
8,928
7,950
2,366
.
10,247
.9,261
3,999
679.
641
548
10,715
1,601
10,991
$1,394
2,242
$1,539
1ll.,2
24.2
288.5
8/ $/76
1,260
83
5,070
6,413
627
278
905
$4.1
9,’ 1/76
10/ 1/76
1,554
1,556
81
$09
5,1.68
0
7,103
1,665
.575
571
2,889
1,054
3,464
1625
48.8
97.6
11/ 1/76
1,562
$09
0
1,671
607
1,359
1,966
117.6
12/ 1/76
-13/ 1/76
1,309
1,344
109
108
0
- 0
1,418
1,452
6 19
603
8 18
777
.
1,437
1 38O
101.3
95.0
p ..)

-------
TABLE c.8 (coNT’D)
CHLORINE
INPUT (1
b/2 1 . hr)
CHLORINE
RETAINED
.
(lb/24 hr)
TOTAL %
RETAINED
DISCARD
SLURRY
#6 C
ULORINATED
DATE
FEED
OIL HYDROCARBON
TOTAL
CLINKER
DUST
TOTAL
14/
1/76
I,1 121i
108
0
1,532
617
650
1,267
82.7
15/
16/
l/7
1/76
1,372
1,532
108
107
0
0
1,480
1,639
607
616
351
21 14
958
830
6 1i.7
50.6
17/
1/76
1,3)7
108
0
I,I 25
72’.
1,338
2,062
1 1 .4.7
18/
1/76
1,776
108
0
1,871.
575
657
1,232
65.7
19/
1/76
1,254
107
0
1,361
571
‘ .6 1 4
1,035
76.0
20/
21/
1/76
1/76
1,282
1,336
107
109
0
0
1,389
1,14 1 ,5
575
589
1,012
959
1,587
1,548
1 )4.2
107.1
—
Pd

-------
125
TABLE C.9. MATERIAL BALANCE FOR 1(20
KiO INPUT (tons/24 hr)
I( O RETAINED (tons/24hr)
TOTAL %
RETAINED
DATE SLURRY FEED
CLINKER
DUST
TOTAL
7/10/75 15.21 11.78 2.98 14.76 97.0
8/10/75 15.25 11.66 4.73 1’6.39 1O7.5
9/10/75 15.30 12.23 1.82 14.05 91.8
10/10/75 15.09 9.04 5.03 14.07 93.2
11/10/75 1 1 4.58 13.07 0 13.07 89.6
12/10/75 12.18 9.13 1.56 0.69 87.8
13/10/75 15.18 11.29 0.78 2.07 79.5
14/10/75 14.76 12.04 1.76 13.80 93.5
15/10/75 15.24 15.64 0 15.64 102.6
16/10/75 15.31 12.27 1.18 13.45 87.8
17/10/75 15.145 13.95 1.31 15.26 98.8
18/10/75 15.64 13.40 1.11 14.51 92.8
19/10/75 15.52 13.06 0.35 13.41 86.4
20/10/75 15.52 13.71 1.27* 14.98 96.5
21/10/75 15.28 16.28 0 16.28 . 106.5
22/10/75. .15.48 13 39 3.55, 16.94 109.4
23/10/75: ‘I5.63 11.47 0.80 12.27 78.5
24/10/75 15.68 2.95 1.70 14.65 93.4
25/10/75 15.35 8.86 2.11 10.97 71.5
26/10/75 16.11 3.79 1.52 15.31 95.0
27/10/75 15.07 10.61 1.58 12.19 80.9
28/10/75 15.20 13.62 3.78 17.40 114.5
29/10/75 15.08 13.99 3.40 17.39 115.3
30/10/75 15.02* 9.91 2.91* 12.82 85.4
31/10/75 14.99 - 5.03 4.42* 9.45 63.0
1/11/75 15.09 3.70 6.09 9.79 64.9
2/11/75 15.39 4•44* 3.65 8.09 52.6
3/11/75 15.61 4.94 6.91 11.85 75.9
4/11/75 15.45 5.66 8.88 14.54 94.1
5/11/75 15.74 10.44 5.43 15.87 100.8
6/11/75 15.58 12.44 3.52 15.96 102.4
7/11/75 15.62 12.70 3.28 15.98 102.3
8/11/75 15.20 11.84 3.43 15.27 100.5
9/11/75 15.99 12.57 3. 148 16.05 100.4
2/12/75 14.60 9.76 1.93 11.69 80.1
3/12/75 114.64 13.34 1.61 14.95 102.1
4/12/75 14.63 10.49 2.94 13.43 91.8
5/12/75 15.04 11.92 10.00 21.92 145.7
6/12/75 15.03 9.50 3.48 12.98 86.4
7/12/75 15.06 9.02 . 0 9.02 59.9
8/12/75 15.16 7.72 5.50 3.22 87.2
9/12/75 14.81 5.51 0.69 16.20 109.4
10/12/75 15.25 3.58 10.56 14.14 92.7 .
11/12/75 15.19 4.80 9.33 14.13 93.0
* Calculated from average data.

-------
126
TABLE C.9 (CONT’D)
K INPtJT (tonst24 Pu)
K 2 0 RETAINED (tons/ 4 hr)
TOTAL %
RETAINED
DATE
SLURRY FEED
CLINKER
DUST
TOTAL
12/12/75
15.19
5.53
6.69
12.22
:
80.4
13/12/75
14.95
3.94
10.71*
14.55
97.3
14/12/75
15.04
4.59
13.64
18.23
121.2
31/12/75
14.45
12.29
1.70*
13.99
96.8
1/ 1/76
14.51
13.87
0
13.87
95.6
2/ 1/76
14.79
14.23
0.42*
14.65
99.0
3/1/76
14.83
7.83
4.01
11.84
79.8
4/ 1/76
14.99
14.27
5.22
19.49
130.0
5/ 1/76
14.80
7.57
8.63
16.20
109.4
6/ 1/76
14.70
6.00
1.37
7.37
50.1
7/ 1/76
-
14.67
11.89
0.93*
21.82
148.7
8/ 1/76
14.37
8.67
0.33
9.00
62.6
9/ 1/76
14.89
4.72
2.44
7.16
48.1
10/ 1/76
14.75
10.71
2.66
13.37
90.6
11/ 1/76
14.80
11.32
3.54
14.87
.100.4
12/ 1/76
14.89
11.87
2.24
14.11
94.8
13/ 1/76
14/ 1/76
15.08
14.72
14.04
13.16
1.97
2.21
16.01
15.37
106.2
104.4
15/ 1/76
14.86
.12.76.
0.98
13.74
92.5
16/ 1/76
14.52
14.88
0.62
15.50
106.7
17/ 1/76
14.45
13.90
4.01
17.91
123.9
18/ 1/76
14.88
13.03
2.14
15.17
101.9
19/ 1/76
14.63
11.73
1.71
13.44
91.9
20/ 1/76
14.58
14.47
3.45
17.92
122.9
21/ 1/76
14.99
11.38
2.70
14.08
“
* Calculated from average data.

-------
129
APPENDIX D
QUALITY OF CEMENT PRODUCED
While remote, the possibility of non-volatile chlorinated
hydrocarbons not being destroyed and remaining with the clinker was
considered. It is conmion knowledge that excessive quantities of organic
materials can detrimentally affect setting, air entraining and compressive
strength characteristics of cement.
After preparing the daily composite clinker sample for analysis;
the quantity of sample which had not undergone size reduction was
retained. These were combined into a composite of several days, three
such composites being prepared for each of the aromatic plUs complex
chlorinated hydrocarbon burn, the PCB burn and the final baseline burn.
Each composite, after being crushed to approximately one—half inch, was
blended with the correct --amount--of-the-nôi,nal’production gypsum used at
St. Lawrence Cement. The clinker-gypsum blends were ground in a laboratory
ball mill. - The. grinding-was timed and samples were taken at suitable
intervals during grinding to determine surface area.
The cements were produced to conform to the requirements of
Symbol 10 cement a designated in the Canadian Standards Association
(CSA) Standard A5 “Portland Cements”, and Type I cement as defined in
the American Society for Testing and Materials (ASTM) designation C150,
“Standard Specification for Portland Cement”.
Each cement was tested in conformance to the following ASTM
methods:
C185, Test for Air Content of Hydraulic Cement Mortar;
Cl09, Test for Compressive Strength of Hydraulic Cement Mortars
(using 2—in, cube-specimens);
C359, Test for False Set of Portland Cement (Mortar Method);
C204, Test for Fineness of Portland Cement by Air Permeability -
Apparatus; -
Cl51, Test for Autoclave Expansion of Portland Cement;
C19l, Test for Time of Setting of Hydraulic Cement by Vicat
Need I e. -
Pages 127 and 128 blank

-------
130
To ensure that differences In quality, were not due to unexpected
differences in cement composition, a. chemical analysis was performed on
each cement tested. Procedures followed were those normally used for
control purposes at St. Lawrence Cement.
The technique used for determination of the concentration of
each oxide was fusion with lithium tetraborete followed by x-ray fluorescence
analysis. This method is coninonly used for cement analyses. With the
exception of lower 1(20 content in the, cements from clinkers produced while
burning chlorinated hydrocarbons, there were no significant differences
In their chemical compositions (tables D.l, 0.2 and D.3). -
Differences In physical characteristics found by testing these
cements were:
Setting TIme - The water required to obtain normal consistency
was greater and setting time shorter for baseline clinker cements.
Shorter time of set and higher water requirement were attributed
to the higher alkal l ’content of the ba e1Ine ‘clinker....
False Set. - False set was severe on the. cements from clinker
produced durtng baseline ‘burning. ‘There was ‘no’ indication of
false set in.cements from clinker produced while burning chlorinated
hydrocarbons. The improvement was due to lower alkali content of
the cement. In this regard, the determination of false set by
ASTP4Method Cls5l, sliest for False Set of Portland Cement
(Paste Method) ’ would not have shown as dramatic an Improvement.
This latter test takes into account the water requirement of
cements, and therefore compensates for the greater quantity of
water required by the hlgher ’alkall cements.
Compressive Strength - Higher Initial and lower ultlmite
strengths are characteristics of higher alkali content cements.
The only apparent differences. In cements from clinkers produced
with and wIthout burning of chlorinated hydrocarbons were those
expected because of the differences In alkali content.

-------
131
TA8LE D.l. CEMENTS FROM CLINKER PRODUCED DURING BASELINE BURN
False Set
Temperature F
Penetration (mm)
3 minutes
5 minutes
8 minutes
11 minutes
Rem i x
Compressive Strength
I day (psi)
3 day (psi) -
7 day (psi)
28 day (psi)
Chemical Tests
Sample Number 7 8 - 9 Aver.
Loss on Ignition ( )
Si0 2 (%)
1.52
20.69
1.51
20.59
1.43
20.53
1.49
20.60
A1 2 0, “
Fe2O H
CaO “
5.93
2.20
63.42
5.90
2.22
63.39
6.00
2.24
63.35
5.91
2.22
63.39
MgO “
2.52
2.52
2.52
2.52
SO 3 “
(20 “
Free CaO (%)
2.74
1.15
0.44
2.79
1.23
0.52
2.89
1.22
0.54
2.81
1.20
0.50
C3S ( )
C 2 S “
50.1
21.6
50.8
20.8
50.0
21.2
50.3
21.2
C 3 A “
C 4 AF “
12.0
6.7
11.9
6.7
12.1
6.8
12.0
6.7
Physical Tests
Fineness 2
Blame (cm /g)
.
3406
.3579
3607
Passing 200 mesh
(*)
96.4
97.6
.98.8
Setting Time
N.C. Penetration
(mm)
10.0
9.5
9.0
N.C. Water (%)
22.5
23.0
23.0
Initial (minutes)
115
106
106
Final (minutes)
225
216
221
73 73
3530
97.6
22.8
109
221
72
40
22
l0
8
50
1990
3040
3660
4380
73.0
84. 5
7.7
70
50
50
21
16
50
2050
2910
3480
4130
73.0
80.6
7.2
37
7
4
4
50
2010
2990
3610
4470
73.0
86.8-
7;’.’6
Air Content
Water ( )
Flow (%)
Air (%)
34
8
4
3
50
1910
3210
3880
4530
73.0
86.0
8.3
Autoclave Expansion ( )
0.07 . 0.09
0.05 0:07

-------
132
TABLE D.2.
CEMENTS FROM CLINKER PRODUCED DURING AROMATIC PLUS
COMPLEX CHLOR I NATED HYDROCARBON BURN
Sample Number 4 5 6, Aver.
Chemical Tests
Loss on Ignition (%)
1.63
1.86
1.63
1.71
Si0 2 (%) ‘
20.37
20.51
20.83
20.57
A1 2 0 3 (%)
Fe 2 0 (%)
CaO 1%)
5.85
2.26
63.19
5.79
2.23
62.99
5.89
2.22
63.56
5.84
2.24
63.25
MgO (%)
2.51
2.49
2.52
2.51
SO (t)
K2O (%)
2.81
0.66
2.69
0.53
2.89
0.69
2.80
0.63
Free CaO (%)
0.50
0.59
0.48
0.52
C 3 S (%)
C 2 S (%)
C 3 A (%)
C4AF (%)
51.8
19.4
11.7
6.9
50.8
20.6
11.6
6.8
49.4
22.5
11.9
6.7
50.7
20.8
11.7
6.8
Physical Tests
Fineness 2’
Blame (cm /g)
Passing 200rnesh (%)
-
3626
98.0
3562
97.4
3561
97.2
3583
97.5
,
Setting Time
N.C. Penetration (mm)
10.0
10.0
9.5
9.8
N.C. Water (%)
22.0
-22.0
22.0
22.0
Initial (minutes)
118
135
135
129
Final (minutes)
238
250
250
246
False Set
- Temperature °F
70
72
73
72
Penetration (mm)
3 minutes.
- 50
50
50
50
5 mInutes
50
50
50
50
8 minutes
50’
50
50
50
11 minutes
50
50
50
50
Remix
50
50
50
50
Com ress lye Strength
I day (psi)
1900
1640
1890
1810
3 day (psI)
3520
3200
3180
3300
7 day (psi)
4590
4360
4160
4370
28 day (p5 I)
5900
6050
5460
5800
Air Content
Water (%)
71.0
72.0
72.0
71.6
Flow ( )
80.0
90.5
81.3
83.9
Air (%)
9.2
9.0
6.7
8.3
Autoclave Expansion (%)
0.03 o .o4
6
0.02 0.03

-------
133
TABLE D.3. CEMENTS FROM CLINKER PRODUCED DURING POLYCHLORINATED
BIPHENYL. BURN
Sample Number 1 2 3 Aver.
Chem!cal Tests
Loss on Ignition (%) 1.46 1.58 1.48 1.51
Si0 2 (%) 20.51 20.72 20.87 20.70
A1 2 O 3 (%) 5.91 5.86 5.89 5.89
Fe 2 0 1 (%) 2.16 2.18 2.16 2.17
CaO (%) 63.24 63.27 63.65 63.39
MgO (%) 2.53 2.56 2.58 2.56-
S03 (%) 2.94 2.64 2.61 2.73
1 (20 (%) 1.01 0.74 0.96 0.90
Free CaO ( ) 0.53 0.57 0.69 0.60
C 3 S (%) 50.4 50.0 ‘50.3 50.2
C 2 S (%) 20.9 21.8 22.0 21.6
C 3 A ( ) 12.0 11.9 12.0 12.0
C4AF (%) 6.6 6.6 6.6 6.6
Physical Tests
FIneness 2
Blame (cm /g) 3388 3579 3593. 3520
-PassIng 200 mesh (%) 96.0 97.0 :96.8 96.6
Setting Time
N:C. Penetration (mm) 10.0 11.0 9. 10.0
-N.C. Water ( ) 22.0 22.0 22.0 22.0
Initial (minutes) 153 125 112 130
FInal (minutes) 273 250 235 253
False Set
0
Temperature F 70 71 71 71
Penetration (mm)
3 mInutes 50 50 50 50
5 mInutes 50 50 50 50
8 mInutes 50 50 50 50
11 mInutes 50 50 50 50
Remix 50 50 50 50
Compressive Strength -
1 day (psi) - 1980 1800 1990 1920
3 day (psi) 3080 3000 3260 3110.
7 day (psi) 3770 4000 4090 3950
28 day (psi) 4620 5110 5080 4940
Air Content
Water ( 71.0 71.0 71.0 71.0
Flow ( ) 88.0 84.5 82.2 84.9
Air (%) 7.7 7.9 9.2 8.3
Autoclave Expansion (%)
0.05 0.06
0.08 0.06

-------
137
APPENDIX E
EQUIPMENT DESCRIPTION AND ASSOCIATED ECONOMICS
In order to burn waste chlorinated hydrocarbons in a cement kiln,
they must be blended to be compatible with the handling and storage system
and to achieve uniform feed to the kiln. A storage tank and feed system
with associated control devices must be installed at the kiln site.
For the program to be economically viable, the price structure
must be-favourable to the originator of the wastes and to the cement
producer. The waste disposal company must be able to generate sufficient
revenue to cover shipping, handling, blending, disposal of non-fuel
fractions and receive some profit. The economics from the point of view
of the cement producer are discussed in Section E.2.
Equipment Description
The —system- Installed-at the cement plant can vary: A simple
inflexible initallation with manual controls and a used rail car for
storage would cost approximately-$25,000. At the otherextreme, a
sophisticated system with large storage capacity, corrosion resistant
construction, well Instrumented could cost $200,000. -
Although there are general similarities between cement kilns.
the flow rate of waste chlorinated hydrocarbon depends on the percentage
of chlorine in the waste, the decrease desirec In the potassium and
sodium levels of the finished product and the production capacity of
the kiln. Thus, for the optimum system, the design, engineering,
installation and operation-must be compatible with these parameters.
Two systems are described herein; one Is the actual system
installed at the St. Lawrence Cement -Co. kiln and the other has been
somewhat arbitrarily selected as a basic system. This basic system
assumes that a quality controlled blend of chemical wastes will be used.
The fuel product would be a low viscosity, single—phase liquid, non-
corrosive or mildly corrosive to carbon steel, not highly volatll€.
having an approximate composition of:-- 60,000 Btu/gallon, 4O Cl,
and specific gravity of 1.2. Fuels with higher or lower Btu and
chlorine can be prepared for those applications where other compositions
are optimum. -
Pages 134, 135, and 1-36 1 1ank

-------
138
At these compositional values,, a feed rate of l lgpm into a
1000 ton per day kiln would provide an alkali reduction of 0.45% as 1(20.
The Btu’s provided would be a small percentage of the total heat requirement
of’ the kiln. With waste streams with a lower chlorine level, larger volumes
can be used to make up a greater fraction of the heat requirements, in
practice up ‘to 10-15% of total Btu input.
Other than capacity and theprovlslon that the kiln be rotary,
it makes little difference what the primary fuel or fuels are. Wet or
dry fuels can burn waste chlorinated hydrocarbons, with consideration given
to the increase In’ dust loading to thë dust handling equipment.
E. 1. 1
‘entail:
Cost of basic system
The basic system with minimum component cost estimates would
Equipment Cost:
Tank, 10,000 gal carbon steel
eed Pump, CentrIfuga1 , 7.5 hp
Flame Arrestor
Vent Scrubber, Actlvate4 Car pp . .
Tank Level Indicator
Grounding.
Tank Berming
Unloading Piping
Piping, Tank to KIln, 100’
Electrical—Combination Starter
Concrete Pad, Site Preparation
Miscellaneous, Hardware
Painting and NFPA Code Markl ng,
Instrumentation -
Nozzle Assembly
Installation:
Total System Cost, Installed
$8,000
1 ,200
:100
1,000
2,500
400
1 ,000
500
1 ,000
1 ,500
2,500
2,500
400
.200
$22,800
4 - ,o00
$26,800

-------
139
A schematic representation of the system is shown in Figure E.I.
Operating Cost:
Maintenance, 4% of investment/year $1,080
Operating Labour, Based on 330 days/year
Operating 24 hr/day
1 hr per shift @ $8.00 8,000
Electricity, 7.5 hp, 8000 hr
@ 3c/kwh 1 ,800
Startup of the system should be achieved with minimum effort
once the system is installed and tested for leaks. Allowance of $1,500
for initial startup should be adequate. Safety equipment for personnel
protection and a fire exinguisher at the unloading/pumping station can
be provided for about $250.
E.2.2 System at the St. Lawrence Cement Co. plant
-The actual system used for the test (Figure E.2) was designed
- to contain corrosive materials as well as those with low flash points.
Sophisticated equipment is used to’control the feed to the kiln and
the system includes traps to avoid vapours being emitted to the atmosph&re.
The storage tank received two liners of Furogiass” membrane and one
liner of 2k” acid proof brick.
Trucks were unloaded via a centrifugal pump through filters
into a 55,000 gal storage tank. Fumes exhausted from the storage tank
while unloading trucks were returned to the truck by a return vent line.
Fumes emanating from the tanks by normal vapour diffusion or expansion
due to temperature were removed by sodium hydroxide and charcoal filters.
Either of two centrifugal pumps was used to feed the kilns at
a pressure.of approximately 120 psi, the pressure maintained by returning
an amount of material through a suitable orifice to the storage tank.
Each kiln was controlled separately using an electro—pneumatic control
valve, magnetic F low.meter and transmitter.
The tubing inserted inside the pipe sleeve located on top of
the burners was made from tantalum. The tubing was threaded so that a
titanium nozzle, of correct size to atomize the quantity of liquid input
to the kiln, up to a maximum capacity of 4 gallons per minute, could be
attached.

-------
CHEMTROL STORAGE TANK
2 PUMP
3 TO KILN
£ RETURN LINE
S VENT PIPE
6 LEVEL GAUGE
1 FILLER PIPE
FIGURE E.1
SCHEMATIC DIAGRAM OF BASIC CHLORINATED
HYDROCARBON FEED SYSTEM
tIfl
0

-------
13 CAUSTIC SCRUBBER
14 CHARCOAL FILTER
15 TRUCK
FIGURE E.2 SCHEMATIC DIAGRAM OF CHLORINATED HYDROCARBON
FACILITIES
©
I STORAGE TANK
2 KILN FEED PUMPS
3 FILTERS
4 FLOW METER
S CONTROL VALVE
6 To KILN III
7 FLOW RECORDING CONTROLLER
8 TO KILNII
9 TO KILN I
-10 FILTER
11 TO ODOR CONTROL SYSTEM
12 BACK TRAP

-------
142
Equipment and Installation:
Tank
Acid Brick Tank Lining
Teflon* Lined Pumps and Filters
Teflon* Lined Piping
Odour Contro i Equipment
Foundations
Di ke
Pun p House
Site Clearing and Leveling
Instrumentation
Painting
Sales Tax,
Engineering
Consul tent
Travel Expense -
Operating Cost:
Maintenance and Operating labour
cost per year)
Electricity (23 /hour for 8,000
hours)
TOTAL
$25,000
35,000
12,000
45,000
5,000
10,000
1 , 000
10,000
3,000
20,000
5,000
5,000
18,000
8,000
101000
$212,000
Due to the potentially dangerous nature of these fuels, both
with regard ‘to personnel safety and environmental considerations, an
extensive training and instruction program was carried out.
E.2 Economics of Burning Waste Chlorinated Hydrocarbons
It is not possible to give a definite statement on the economics
of burning chlorinated wastes for all circumstances. Some cement p!ants
comprisi
The total installed cost of this system was over $200,000,
ng the following items:
TOTAL
$25, 450
1 , 840
$27,290
*Trademark E.I. du Pont de Neffiours.

-------
143
require a reduction-in the alkali oxideS (1(20 + Na 2 0) content of the
finished product. In this case, chlorine must be added to the process
materials, and Is available from different sources In different geographical
areas. Waste hydrochloric acid from the steel industry may be used where
available; In other areas cement plants purchase calcium chloride. If
reduction of alkali is not required, then addition of chloride would
cause the generation of greater quantities of dust, thereby Increasing
dust disposal costs. - -
Another factor making economic comparisons difficult is the
different waste streams available. In some cases, the chlorinated waste
streams may be suitable for purifying and further use. Other waste
streams, such as PCB’s and insecticides banned for environmental reasons,
present a serious disposal problem. Obviously, these different materials
would not all cost the same for use In a cement kiln. However, to present
some models, the price structure used by Chemtrol Pollution Services, Inc.
for blended noncotrosive wastes suitable for use in.thé’baslc system’ -.
described-above has beenadopted; At present, Chemtrol plans to charge
80 :percent of the, fuel value ‘and, where applicable, 30 percent of -
the chlorine value for these blended ufuelsll. In the examples, the
following assumptions have been made: The kiln is a wet process type
producing 1 ,000 tone per day with a fuel requirement equivalent to
5,150,000 Btu per ton of production.
E.2.l Process requiring addition of chlorine
Assume in this case that the fuel cost Is $1 per million Btu and
that it is necessary to reduce the 1(20 content of the clinker by 0.45 percent.
This would require approximately 6,800 pounds of chlorine per day. A 32
percent solution of calcium chloride weighing 13 pounds and costing $0.14
per imperial gal ion ($0.053 per pound of chlorine) is used. In this case,
2,565 gallons would be required at a total cost of $359.
This could be replaced by ‘9’rol Fuel*” containing a nominal 70
percent chlorine and ‘65,000 Btu per imperial gallon. At a weight of 13
pounds-per gallon, 750 gallons would be required each day. Using the
*Trademark, Chemtrol Pollution Services Inc.

-------
144
above mentoned price structure, the cost per gallon would be:
9.1 lb Cl 2 x x — $0. 145
5,000 Btu 1,000,000 Btu X
TOTAL $0.20
Thus, the 750 gallons would cost $150 and would supply 48.8
million Btu.
The total daily tangible-savings then become:
Cost of Calcium Chloride $359
Cost of Normal Fuel Replaced $ 49
Less Cost of “rrol Fuel” $150
Net Savings Per Day $258
E.2.2 Process not requiring add itlon of chlorine
The 1,000 ton per kiln-and a specificheat consumption of
5,150,000 Btu per. ton of clinker produced and:a fuel cost equivalent
to $1.00 permillion Btu is again taken as the model. - -In this case,
however, no chlorine is required and the waste stream contains less than
5 percent chlorine. This-material Is available at $0.80 per million Btu
and Is used at a replacement rate of 10 percent of the Btu requirement
or 515,000 Btu per tonof clinker.
The fuel normally used,on the kiln but replaced by this waste
stream would cost $O.515 per ton of clinker or $515 per day. The daily
savings can’be calculated as:
CO st of Normal Fuel Replaced $515
Cost of “Trol FueP’ 412
Net Daily Savings $103
This saving can be negated by the cost of discarding the
additional dust collected in the precipitators (See Section 6.1), a factor
which will vary with each cement plant.
E.2.3 Total economic consderations
The above models are not intended to be all inclusive. Due
to the many variables involved, each possible application must be assessed
based upon economic factors existing In the locale. At the time the system

-------
145
s installed at the St. Lawrence Cement Co., it was necessary to use
chlorine In two wet kilns each having a capacity of 1,000 tons per year.
For the period January through September 1976, maintenance on the system
amounted to $18,760. Thus, the figure of $27,290 for yearly operating
and maintenance cost is considered accurate. Under these circumstances,
the economic picture over the period of a year becomes:
Total Savings Using “Trol Fuel”
(600 kIln operatIng days @
$258 per day) $154,800
Maintenance and Operating Cost 27,290
interest on $212,000 @ 11%/year 23,320
Net $104,190
This estimated saving, which would have given excellent payback’
on the system, has not been realized. Law alkali requirements have been
falling Into disfavour because such requirements are not compatible with
the goal of energy conservation. With the requirement for chlorine
addition, savings must be based only, on fuel,.costs which. If-all factors
are favourable, would-total for the two kllns:
TotalSavingsUslng-”Trol :Fue l”
(600 kiln operating days @
$103/day) - $61,800
Maintenance and Operating Costs 27,290
interest on $212,000 11%/yr 23,320
Net $11,190
Even this modest sum has not been realized, for two reasons:
the cost of discarding additional dust, and the unsteady supply of these
wastes.
It is obvious from this discussion that plants- requiring the
addition of chlorine are in a better position to make burning of
chlorinated wastes attractive than those plants not requiring chlorine.
Also, a less. sophisticated system than that Installed at St. Lawrence -
Cement Co. is necessary if the only consideration is the fuel value of
these wastes.

-------
149
APPENDIX F
ONTARIO MINISTR 6 Y OF THE ENVIRONMENT EMISSION
GUIDELINES AND ANALYTICAL SUPPORT
F.l Emission Guidelines
All industrial sources of potentially significant emissions to
the atmosphere In Ontario come under the jurisdiction of the Province’s
Environmental Protection Act. The Act requires such sources to operate
under a Certificate of Approval which specifies the acceptable operating
conditions and emission rates for potential air contaminants. The
Ministry of the Environment required that the experimental waste chlorinated
hydrocarbon burn at St. Lawrence Cement Co., Mississauga meet certain
conditions specified in a Certificate of Approval. Conditional approval
was granted for each of the three stages of the waste burn (A, B, C) under
the terms of the tentative guidelines for the allowable emission rates;
of organic chloride.
- - Pasquill—Gifford dispersion éa,lculatlons based on the guideline
emission rates from the wet kiln were used to predict the worst-case ground
level concentration of residual organic chloride (4-hour average).
Table F.l shows the specifications which were applied to each
stage of the waste burn, along with the measured emission rates.
The less stringent guideline for the first two waste burns was
met easily. Because of the large percentage of highly persistent poly-
chlorinated biphenyls anticipated in the fuel for the third waste burn,
a much more stringent guideline was given.
In the most critical case, the third burn, the calculated
approximate flue gas concentration of residual chloride, 50 pg/rn 3 (as
Cl), based on the stated guideline, corresponds to roughly 5 ppb (v/v)
of Aroclor 1242 or 17 ppb (v/v) of dichlorcmethane (at ambient temperature),
the predominant individually identified component of the flue gas.
These values compare with the detection limit for PCB’s of 3 pg/rn 3 or 0.3
ppb In the flue gas (ambient temperature).
- - Since no PCB was found In the flue gas samples and the methy—
lene chloride concentration only slightly exceeded the very stringent
guideline for organic chloride in this stage, the guideline should be
considered to have been met without question.
Pages lii. 6 ,lLi.7, and u i. 8 blank

-------
150
TABLE F.1. MOE SPECIFICATIONS APPLIED TO WASTE BURNS
Waste
Burn
1
2
3
4
5*
6
7
A
B
C
0.5
0.5
0.005
99.5
99.5
99.995
99.990
99.989
99.986
5000
5000
50
23-150
16-150
88-150
O.2i
0.2
0.002
0.001-0.006
0.001-0.006
O;004-O.006
Column Identification :
1. Guideline emission rate for residual organic chloride (g Cl/sec).
2. Estimated required percent destruction or retention of chlorinated fuel.
3. EstImated minimum percent combustion of chlorinated fuel from test data
(Table 6, Section 5.3).
ii. Approximate calculated flue gas residual chloride concentration based
on guideline emission rate (pg.C1 /m Y.
5. - Approximate measured flue gas organic chloride concentration based on
dichlorométhane and .data:’fromTable 5, Se,ction5.3 (pg Cl/rn 3 ). -The
upper limit of the range corresponds to 50 ppb organic Cl as CH 2 C1 2 .
6. Calculated point of impingement i—hour average chloride concentration
based on guideline emission rate (pg Cl/rn 3 ).
7. Approximate calculated point of Impingement i—hour average chloride
concentration based on dichloromethane and data from Tables 5 and 6,
Section 5.3 (pg Cl/rn 3 ).
*Since no PCB’s were detected in the flue gas samples, none of the
residual organic chloride values in column 5 should be attributed to
PCB’s in any of these stages.

-------
151
If the concentration of residual organic chloride in the
flue gas (column 7, Table F.l) were taken to represent PCB’s in a “worst
case” (recall that no PCB’s were detected), the calculated point of
impingement concentration by the Pasquill-Gifford method is approximately
equal to typical measured ambient air concentrations of PCB’s in urban
areas. There would appear to be little cause for concern when the
emission guidelines are met in this process configuration.
The above discussion relates only to the wet-kiln tests at
St. Lawrence Cement. if the same guidelines were applied to another kiln,
for example the ‘ower-level, multi-stack dry kiln at St. Lawrence Cement,
the same emission rate as measured In the wet kiln tests would lead to a
significant ground level concentration. For example, even if the
stringent guideline for PCB emissions were met In this hypothetical case,
the maximum ground level concentration Is calculated by virtual source
methods (applied to the dry kiln stack configuration) to be 0.07 izg/m 3
:(asAroclbr 121+2) whiãh Is 7tä 70 tImes thà typical measured ambient’
air concentration of 0.001 t O’.Ol g/m 3 The point is that some caution.
should be exercised in applying the results of these tests to another
stack configuration.
In the case of the third waste burn, it was estimated that the
total organic vapour concentration in the flue gas during the tests was
about 50 ppb, of which about 30 ppb was found to be dlchloromethane.
Since none of the components of the original fuel mixture were detected,
however, the maximum residual concentration of Aroclor 1242, for example,
could be only about 1 ppb or less (noting the approximate detection limit
of 0.3 ppb). For speculation purposes only, in this worst case, I ppb
Aroclor 1242 in the flue gas (at ambient temperature) would correspond
to about 11 ug/m 3 PCB (or about 5 ug/m 3 as chloride) and about 0.1 kg
of Aroclor 1242 emitted per day (24 hours of continuous operation). In
reality, this quantity Is an overestimate of the total emissions and is
probably not significant.
F.2 Gas Chromatographic Analysis of Process and Emission Test Samples
F.2.l Chrcmosorb 102 adsorption tube analysIs
Duplicate Chromosorb adsorption tubes for each of the waste
burns were analyzed by the Air Quality Laboratory, Laboratory Services

-------
152
Branch, Ontario Ministry of the Environment for support and cross-check
of analytical results obtained by the Ontario Research Foundation. Each
adsorption tube sample represented 60 litres of emission gas. Analysis
was done after desorpti on at 120°C Into an evacuated glass vessel by
injecting a 1 ml gas sample Into a gas chromatograph. This instrument
was equipped with a 9 foot x’ 1/8 inch Chromosorb 102 column held at a
I temperature of 180°C anda Sc 3 H electron capture detector.
The results from these analyses are suninarized in Tables F.2
to F.LI. All Chromosorb tube samples, Including those from “Baseline B”,
were found -to contain small -but measurable amounts of six or seven compounds
and traces of a further six to eight compounds. These “measurable”
compounds were almost certainly chlorinated, low—molecular weight
compounds which, in the tables, have been designated “major” and”minor”,
according to the size of their gas-chromatographic peaks. Of these
compounds, only chloroform had been tentatively identified and quantitatively
determined. All other compounds separated from each.sample were estimated
together b ’y-relatlng the sum of thegaschromatógraphic peak areas with
those obtai:ned from a chloroform standard. ’ Their concentra ions are.,
listed under “Other Compounds”.
The results shown in Tables F.2 to F.4 indicate that the total
of all chlorinated compounds in the Chromosorb adsorption tube samples
corresponds to concentrations only in the low parts per billion range
in the stack emission gases.
The ,agreement between these results and those of the Ontario
Research Foundation and TRW is only qualitative. The generally small
amounts (ppb range) of organic chloride- vapours found by ORF are
confirmed but ‘trends from stage to stage are not reproduced. Such
disagreement is merely indicative that the limits of sensitivity of
the combined sampling and analytical methods are being approached.
F.2.2 Fuel feedstock analysis
Samples of the waste chlorinated hydrocarbon fuels used in
waste burns B and C were analyzed In the laboratory of Professor F.W.
Karasek, Department of Chemistry, University of Waterloo under-a
research grant from the Air Resources Branch, Ontario Ministry of the
Environment.

-------
153
TABLE F.2. ST. LAWRENCE CEMENT WASTE BURN EXPERIMENTS - TEST 1
CHROMOSORB ADSORPTION TUBE ANALYSIS
Waste Burn
Experiment
“Major” “Minor” Other Total
Compounds Compounds Chloroform Compounds Compounds
ppb ppb ppb
A
3 4 1.192 7.121% 8.313
B
1 5 0.058 0.646 0.704
C
2 4 0.329 2.909 3.238
Baseline B
2 5 0.055 2.487 2.542
Blank
(Chromosorb)
0 0 - -
TABLE F.3.
ST. LAWRENCE CEMENT WASTE BURN EXPERIMENTS - TEST 2
-
CHROMOSORB ADSORPTION TUBE ANALYSIS -
Waste Burn
Experiment
,
“Major”.. “Minor”- Other Total
Compounds Compounds Chloroform Compounds - Compounds
- ppb ppb - ppb-
A
4 3 1.462 8.827 10.289
B
4 1 1.464 5.270 6.734
C
3 4 0.924 12.089 13.013
Baseline B
0 7 0.022 0.443 0.465
Blank
(Chromosorb)
0 0 - -
TABLE F.4.
ST. LAWRENCE CEMENT WASTE BURN EXPERIMENTS — TEST 3
CHROMOSORB ADSORPTION TUBE ANALYSIS
Waste Burn
“Major” “Minor” Other Total
Experiment
Compounds Compounds Chloroform Compounds Compounds
- ppb ppb ppb
A -
- 5 0 2.356 15.773 18.129
B
3 3 0.481 1.916 2. 97
C
3 8 0.560 2.089 2.649
Baseline B
3 4 0.030 5.077 5.107
Blank
(Chromosorb)
0 0 - - -

-------
154
F.2.2.l Preparation of fuel samples . All samples were diluted ten
times with Burdick and Jackson “Distilled in Glass” methylene chloride.
A 0.1 ml aliquot of the fuel was made up to 1.0 ml with the methylene
chloride.
Methylene chloride was chosen for the selvent based on previous
fuel samples analyzed in this laboratory.
F.2.2.2 Analysis of fuel samples . All samples were surveyed by gas
chromatography using a 10’ x 2 nm’i l.d. glass column with a specially
prepared, high resolution column packing, referred to as Ap-20M. A
similarly prepared solution of Aroclor 1242 was also chromatographed for
comparison.
The CC conditions used are shown in Table F.6 and were closely
monitored by the use of a HP 5830 A Digital Gas Chromatograph using a
flame ionization detector. The area of the solvent peak was rejected
to facilitate comparison of CC data.
Mass’spectrél data were obtained using a Perkin-Elmer 900-CC
interfaced to a HITACHI RMU-6 magnetic mass spectrometer via a Biemann-—
‘Watson effusion separator The spictra were Initiated and counted manually.
All spectra were obtained with 70 eV ionizing voltage at 250°C.
Samples 1 and 9 were analyzed using GC/MS techniques as above.
F.2.3 Results and Discussion
Samples 8 to 10, the arOmatic chlorinated fuels, displayed only
seven compounds with individual integrated area percentages greeter than
1% and only nine with similar area percentages greater than 0.5%. There
appear to be only three major components: one with Ca. area, one
with Ca. 4.5% , and one with Ca. 6.8%. This appeared to be the case for
all the aromatjc fuels. A representative bar chromatogram is shown In
Figure F.l. The peak at four minutes in the aromatic fuel was Ca.
Samples 1 to 7, the PCB fuels, displayed several peaks In
addition to those of the aromatic fuels, samples 8 to 12. Comparison of
the additional peaks with a chromatogram of Aroclor 1242 indicates That
these peaks are due almost entirely to PCB compounds (see Figure F.l).
To illustrate this observation, the CC data from the aromatic
fuel was added to the GC data from Aroclor 1242 and the computer program

-------
155
TABLE F.5. FUEL SAMPLE IDENTIFICATION
appeared to have water in It.
had reacted with liner of sample vial cap.
TABLE F.6. GAS CHROMATOGRAPH CONDITIONS
HP 5830-A GC
500
4 mm
iO
‘. 1mm
2400
280°
3000
28 mi/mm
H EL I UM
28 256
Sample
No.
ColTvnents
1
Chemtrol
Line
Sample
(PCB’s), Jan.
3/76, 10:00
AM
2
“
“
“
, Jan.
4/76, 12:00
3
1
I
Jan.
,
5/76
4
“
“
H
, Jan.
6/76
5
“
‘I
Jan.
,
7/76, 9:00
AM
6
“
“
I
, Jan.
8/76, 8:45
7
“
I’
U
Jan.
,
9/76*
8
Chemtrol
Line
Sample
(Aromatics),
Dec. l0/75**
9
“
“
“
“ ,
Dec. ll/75**
10
1
“
,
Dec. 12/75**
11
U
H
1
,
Dec. 13/75**
12
•,
‘I -
SI
,
Dec. 15/75**
-
*Samp le
* Samp.l e
TEMP I
TIME
RATE
TEMP 2
TIME 2
INJ TEMP
FID TEMP
FLOW RATE
CARRIER GAS
50 mm
ATTENUATI ON

-------
0
25
20
15
10
5
25
20
15
10
5
25
20
15
10
5
AROM FUEL
I. p
1.
I,
ZAREA 529693
CLOR 1242
I
iii lit1 1
ZAREA 706087
I,
PCB FUEL
I
I
0
10
ZAREA 608448
t SI
20
30
40
50
RETENTION TIME (MIN)
60
70
FIGURE Fl
REPRESENTATIVE BAR CHROMATOGRAM

-------
157
treated the data as though it originated from a single sample. This
computer-produced bar chromatogram, titled “AROM + CLOR”, was then plotted
in comparison with the bar chromatogram of the PCO fuel sample as shown
in Figure F.2. The data were not subjected to any manipulation other than
straight combination of the data. Inspection of the two plots shows such
a close correspondence of GC peaks between the computer reconstructed
mixture and the actual PCB fuel mixture that there is a reasonable assurance
they are the same. -
A GC/MS analysis was conducted on sample 1, one of the PCB
fuels. A representative gas chromatogram with numerically identified
peaks is shown in Figure F.3. Mass spectra were obtained for these
numerically identified peaks. The probable identity, molecular weight
and/or- degree of chlorination is shown in Table F.7.
Since all of the peaks are found in the PCB fuel and only some
,ofthe peaks are found in the aromatic fuel.. an estimate of the dilution’
or mixing ófthese compoUnds.can be derived. During the gas chromatograph
analysis,- allvariables were ‘held constant; thus, absolute integrated
area counts are indicative of changes in relative concentrations. These
results are shown In Table F.8. The units are area counts and the values
shown are l/lO of output values.
Peak #3 is absent in the aromatic fuel and in Aroclor 1242 but
present in the PCB fuel. The mass spectrum of peak #3 suggests
tr ichlorobenzene. -
Peak #19 is present In small amounts in the chlorinated
aromatic fuels, absent in Aroclor 1242 but present in increased concen-
trations in the PCB fuels. A definitive mass spectrum of this peak was -
not obtained but was suggestive-of lack of halogenat ion and is most
likely an unsaturated hydrocarbon of high molecular weight.
P.2.4 Conc’usions - -
The aromatic fuel is primarily o-chlorotoluene and the PCB fuel
is a mixture of Aroclor 1242 and the aromatic fuel. Area data for peak
//1 indIcated an initial dilution of approximately one-third (by comparison
of area counts for aromatic fuels to area counts for sample 1) which
increased to approximately four-fifths by the end of sampling data.

-------
25
PCB FUEL ZAREA 608448
20
15 -
10
5
I I. t IIIII.I .1 I III .111111 . iii..,.
25
AROM + cLOR zAREA 1235780
e 20
15
10
RETENTION TIME (MIN)
FIGURE F2 COMPUTER RECONSTRUCTED BAR CHROMATOGRAMS FOR PCB FUEL AND
AROMATIC FUEL PLUS AROCHLOR 1242

-------
19
18
I
16
17
15
9
8
‘6.
5
PCI FUEL
3/1/76
4
52 48 44 40 36 32 28 24 20 16
TIME (MIN)
12
8
FIGURE F3 GAS CHROMATOGRAM FROM GC/MS ANALYSIS OF SAMPLE PCB FUEL
1
U’
4
0

-------
60
The results of the University of Waterloo group with respect
to identification of components of the waste burn B and waste burn C
fuels are in agreement with those of ORF and TRW.
TABLE F.7. MS DATA FROM GC ANALYSIS OF SAMPLE PCB FUEL
(Reference to Figure F.3.)
Peak No. Comments
1 MW 126, 1 chlorine — spectra suggest o-chlorotoluene
2 MW 160, 2 ch!orlnes - spectra suggest arcmatic compound -
dichlorotoluene
3 MW 180, 3 chlornes - spectra suggest trichlorobenzene
14 MW appears to be 216 with 4 chiorines - possible identity
C 6 H 14 C 1 14
5 MW appears to be 216with le chiorines -spectra very
- - similar to peak #14
6-8 MW 222 2 chlorines - spectra suggest dichiorobiphenyls -
PCB’s
9-12 MW 256 3 chlorines — PCB’s
13-16 MW 290 1. chiorlnes - PCB’s
17 & 18 MW 3214 5 chlorlnes —PCB’s
19 MW uncertain - no suggestion of halogenation

-------
TABLE F.8. AREA COUNTS (x 10k)
10
8
9
I I
12:
I
2
3
1*
5
6
7
Peak No. 10/12/75
11/12/75
12/12/75
13/12/75
15/12/75
3/1/76
4/1/76
5/1/76
6/1/76
7/1/76
8/1/76
9/1/76
1
3860
3976
3769
3991*
3842
1200
1 11 2
913.4
770.6
766.8
753.0
702.4
2
80.5
82.5
78.8
82.5
80.4
24.9
23.1
19.0
16.1
$6.1
15.7
14.6
3
-
-
-
-
-
20.4
23.8
38.0
48.2
50.9
49.7
57.8
1*
243.9
247.6
21*0.6
251.4
245.1
74.2
72.3
64.5
60.6
61.7
58.3
60.4
5
363.1
362.1
360.5’
373.3
363.4

154.1
$48.1
$55.0
$50.2
153.8
152.0
$57.6
6
-
-
-
-
158.2
154.8
160.6
156.5
160.2
161.1
167.8
7
8
-
-
-
-
-
-
‘
-
-
-
-
76.9
340.4
75.7
333.7
77.0
348.4
86.5
328.8
88.6
335.2
89.4
336.7
95.1
351.3
9
-
-
-
—
—
465.6
454.8
451.2
431.3
439.3
1 40.6
1*57.2
0
-
-
-
-
-
82.7
80.6
84.9
84.3
85.6
85.2
89.1
I I
—
-
—
—
—
217.7
211.8
212.4
203.3
207.9
208.9
217.1
12
-
-
-
,
-
-
-
51*3.5
518.0
503.6
‘e8i.8
1*89.1
491.2
507.6
13
-
-
-
-
265.0
271.3
275.5
258.3
265.9
268.6
282.)
$4
‘
-
-
-
-
-.
$97.0
$83.8
179.7
177.0
178.6
$78.3
$86.7
15
-
-
-
-
-
287.0
292.4
302.5
284.3
292.8
294.5
309.8
16
-
-
-
-
-
$91.3
191.1
$86.8
180.2
$78.6
188.4
190.5
17
-
-
-
-
-
102.0
107.5
134.8
143.2
149.1
151.4
$67.5
18
•
-
52.1
51.5
54.9
50.)
51.3
49.5
50.2
19
7.8
8.6
8.4
8.2
8.6
177.5
168.9
143.8
$20.5
$20.5
120.9
115.9

-------
“3
APPENDIX G
LABORATORY ANALYSIS RESULTS FROM THE
ST. LAWRENCE CEMENT FACILITY TEST
by
D.G. Ackerman, J.F. Clausen and.C.A. Zee
TRW Systems Group
Preceding page lilank

-------
165
APPENDIX C
LABORATORY ANALYSIS RESULTS FROM THE ST. LAWRENCE CEMENT FACILITY TEST
C. 1 Sumary
Analysis of the flue gas samples for organic composition gave
the following results:
- Hydrocarbons were not detected in the samples of flue gas with
detection limits ranging from 0.6 mg/rn 3 to 0.009 mg/rn 3 .
- Low molecular weight chlorinated organic comp unds such as
methylene chloride, chloroform, and carbon tetrachioride were
specifically searched for by gas chromatography with electron
capture detection (ECD) and were generally not detected in most
of the samples. Four samples indicated, the possible presence
of some of these compounds, but all at levels of less than
0.1 mg/rn 3 of flue ’gas.
- Polychlor.inatedbiphenyls were searched ‘forby GC/ECD and GC/MS
and were not found in any samples at the CC/MS detection limit
of 3 i g/m 3 of flue gas.
Trace metal concentrations in the flue gases, as determined in
the samples taken with the EPA sampling train, were all less than 10 pg/rn 3 ,
with the exception of lead In the WBC tests. The emission of lead during•
the WBC tests averaged 0.12 mg/rn 3 .
Analysis of the clinker product and discard dust samples for
organic composition gave the following results:
- Hydrocarbons were not detected in solvent extracts of any of
the clinker product or discard dust samples. The level of
detection by GC/ECD was 5 pg/g of sample or lower.
- Polychiorinated biphenyls were not detected in any of the
clinker or dust samples by GC/ECD. Detection limits for PCB’s
in the samples were typically 0.04 pg/g or better.
The inorganic character of the clinker products and electrostatic
precipitator discard dusts was changed very little if at all by the addition
rêc 111 ‘page blank’

-------
166
of the chlorinated wastes to the SLC wet kiln process. Trace metal
levels In the WBB and WBC test samples did not differ significantly
from the baseline samples.
Analytical techniques selected for analysis of flue gas
constituents which include vapours, condensables and particulate matter
were gravimetric, IR,.LRMS, GC, GC/MS, SSMS, ICPOES, and AAS. These
were selected on the basis of sensitivity and selectivity criteria to
enable identification of species in the flue gas at concentrations on
the order of 0.1 mg/rn 3 , representing the threshold level of the most
toxic species as defined by OSHA and other occupational health and
safety organizations; Detection limits for many of the techniques
extend to lag/rn 3 levels. However, specific analyses to identify compounds
below the level of interest, 0.1 mg/rn 3 , were not routinely performed.
The techniques used were both qualitative and quantitative in nature, with
an intended accuracy range of plus o minus a factor of two to three.
G.2 Introduction
In cooperatIon,with Env ronment Canada and theMinistry f
Environment of Ontario, the United States Environmental Protection Agency
supplemented the Canadian studies through participation in analyses of
samples acquired from various streams during test burns of chlorinated
hydrocarbons In the St. Lawrence Cement Co. wet process kiln. These
analyses were performed by TRW Defense and Space Systems Group, Redondo
Beach, California, under contract to the U.S. EPA. Samples were provided
to TRW through the courtesy of the Ontario Research Foundation and
St. Lawrence Cement Co., of Ontario, Canada.
Emphasis in the EPA sponsored work was directed toward analy , ing
for input waste residual compounds and by-products In the clinker (product),
dust from the electrostatic precipitators, and the various co Wponent
samples from the EPA Method 5 and ORF sorbent trap trains. These analyses
were limited to the two test burns performed using chlorinated aromatic
and polychlorinated biphenyl waste blends. In addition, background anal-
ytical work was accomplished on sampiesacquired during the two baseline
tests in which the kiln was fired using only residual oil as a fuel.
Analyses were also performed on the. two waste blends. This work was done
to supplement the Canadian laboratory results as well as to acculre

-------
167
additional evaluation data in accordance with U.S.. EPA protocol used
for ongoing chemical waste incineration test programs in the United States.
A brief economic analysis was made by St. Lawrence Cement Co.
using data from both their own operations and from Chemtrol Pollution
Services Inc. (Appendix E).
G.3 Analysis Techniques
The purpose of the analyses performed on the samples from SLC
was to identify and quantify:
- known hazardous species present from the waste as determined
by pretest analysis of a sample of the waste material,
- secondary decomposition products (e.g., incomplete combustion
products which are predictable), and
- other species found to be present but which are not predicted
or otherwise expected.
A complete list of the sampleS received by TRW for analysis is
given in.Table G.l. This table shows what portiorrof the total sample
collected by the trains was sent to TRW as well as the respective burns
and test numbers for the received samples. The coding system used to
uniquely Identify each sample is shown in Figure G.l. These codes are
used in subsequent tabulations of data and results.
The first step in the analytical approach involved various
extraction and preparation procedures to separate the organic and inorganic
constituents and/or to concentrate the samples In a suitable form for
analysis. The extracts and concentrates were then analyzed by the
techniques to be described in’ this section.
G.3. 1 Extractions and sample preparatIon
G.3.l.l Solvent extracts received from ORF . With a few exceptions, solvent
extractions for organic species were performed by ORF using pentane or
hexane. Subsequently, aliquots were received by TRW and combined
according ‘to the plan shown in Figure G.2. These combinations were
performed because: -
1) the aliquots represent relatively small gas sample
volumes, thus making constituent concentrations very low
and difficult to measure, and

-------
TABLE G.1. SUMMARY OF SAMPLES RECEIVED FROM EACH WASTE BURN
a ..
S leDescription
Portion Received
outofTotal
S I.
Rum and Test No. of S les Origin
BLA
I SA
NR C
SR
— ——
.1
2
3
I
2
3
1
2
3
I
2
EPA Train
‘
• Probe rinse
F ll t eredinsolubles
All
I
I
I
I
I
I
I
I
Solventextract
1/4
2
1
1
1
1
1
1
I
I
Aqumods solution after filtration and extraction
All
- -
1
3
• Filter
.
Solvantextract
1/4
I
1
1
I
I
I
I
1
1
Filterwit liparticulats
AU
I
1
.1
I
I
I
I
I
• I in9srs
Solvantextract
1/4
1
1
1
1
1
I
I
I
1
- solution after ontrectios
All
I
I
O i l Trsii
• In-stack filter
Solventextract
1/4
I
1
II
I
I
I
I
I
FIlt.rw ltkpart lculate
All
3
3
3
3
3
• Probe rinse
Filtarodi.selubles
All
I
I
I
I
I
I
I
I
Solvuntextract
1/4
I
1
I
1
1
1
1
I
I
a solution after filtration and extraction
All
I
I
• U Ii I tngsr
-
Sedventextract
1/4-
1
2
2
1
1
1
1
X
1
Aqueous solution after extraction
All
I
I
• Distilled uster i lngsr
Solventextract
1/4
1
I
1
I
I
1
1
1
1
Aqueous solution after extraction
All
1
1
•Clwosovb lO lsorbs nttubes
1/4
I
I
I
3
3
3
3
3

-------
SAMPLE CODES CONSIST OF 3- 5 SECTIONS
Burn
Sampling Train
Train Component Extracted
Any Special
Preparation -
Test No.
First Baseline on
Standard EPA
Filter/Insolubles -
F
Extracts of Acidified
11,12, or 13 If
Primary Fuel - BLA
Method 5 Train -
EPA
Probe Rinse -
Caustic Impinger -
.
PR
NAOH
Solutions
- AE
sample Is from
only one test
Waste Burn on
Chlorinated
Water Impinger -
Combined Impingers -
H 2 0
I
Aromatics -
-
Blank if sample
Is a combination
Waste Burn on
of all three
PCB Blend - USC’
ORF Designed
tests
NSo,J fltN Train-
01W
Second Baseline on’
Primary Fuel - 3D
.
FOR EXAMPLE • ALL OF THE EXTRACTED SAMPLES FROM ONE WASTE BURN WOULD BE COOED AS FOLLOWS:
WBB-ORF-FE = Waste Burn B, ORF Train,Combined Filter Extracts from
all three tests
WBB-ORF-PRE & NAOHE Waste Burn B, 01W TraIn, Combined Probe Rinse and Caustic
impinger Extracts from all three tests
WBB-ORF-NAOH-A(-T3 = Waste Burn B, 01W TraIn, Extract of the acidified caustic
linpinger solution from Test 3
WBB- ORF-H 2 OE - Waste Burn B, ORF Train, Combined Water Impinger Extracts
from all three tests -
WBB-EPA-FE & PRE Waste Burn B, EPA Train, Combined Filter and Probe Rinse
Extracts from all three tests
WBB-EPA-IE Waste Burn B. EPA Train, Combined Impinger Extracts from
all three tests
FIGURE G.l. TRW SAMPLE CODING SYSTEM

-------
170
2) all three tests of each waste were performed at one,.
nominal operating condition.
In general, following this plan, the organic concentrate obtained for
each extracted sample type (e.g., filters, impingers, solids, etc.) from
each of the two sampling trains were the combination of’ail three tests
of each waste. Exceptions to this plan were the BLA and aqueous samples
for which only one of the three test samples were received (see Table G.l).
At no time were EPA train samples combined with ORF train samples.
Small aliquots (2-5 ml) were first taken of the “as received”
extracts and set aside for analysls of volatile compounds that would
be lost in the next step which was to concentrate the remaining solvent
sample using Kuderna-Danish evaporators and a steam bath.
The solvent extracts of the probe rinses, in addition to being
combined as in the plan shown in Figure G.2, were also combined with the -
sample (EPA train-filter, ORF train—first implriger) whose juxtaposition
in’ the sampling trainand similar physical characteristics permit. the
combination.. :The rinses of the ‘EPA train were-made of the glass’ probe
liner infront of the .fllter’. However, ‘with’the’ 0RFt rain which only
has a short nozzle in front,of the In-stack filter, rinses were made of
the 14 feet of probe and tubing between the filter and the first liquid
impinger. The resulting solutions were filtered through standard filter
-paper and then extractedby ORF. -
G.3.L2 Solids, aqueous solutions, and filters . The solid samples
were prepared for organic analyses by extraction in a Soxhiet apparatus
for 24 hours with distilled-in-glass grade pentane. These extracts were
concentrated with Kuderna—Danish evaporators to a 10 ml volume. Prepara-
tion of the solid samples for Inorganic analyses consisted of a low-
temperature plasma ashing to remove possible organic interferences.
Other samples for inorganic analysis also required some degree
of preparation. Aliquots of the aqueous Impinger and probe rinse samples
were taken and acidified with nitric acid to stabilize any metals present.
Each of the insoluble residues obtained by filtering the probe rinses was
combined by ORF with its matching particulate f liter for treatment as one
sample. These insolubles/f liter samples were plasma ashed and then

-------
171
TO ANALYSIS OF
ORGANIC CONCENTRATES
FIGURE G.2. PLAN, FOR THE COMBINATION OF ORF SOLVENT EXTRACTS
VACUUM GAUGE
SORBENT TUBE
V
V 3
VACUUM
FIGURE G.3. DESORPTION SYST

-------
172
extracted by refluxing constant boiling aqua regla over the each sample
for two hours. The resulting acid digests were made to 50 ml for
analysis.
G.3.l.3 Sorbent tubes . The Chramosorb 102 sorbent tubes from the
ORF sampling train were prepared for analysis by a quantitative desorption
technique. -The apparatus used for the .desorptlon Is shown schematically
In Figure G.3. The sorbent tube Is attached with a minimum length of
Tygon tubing to a 20 ml glass sample bulb and is then heated in a small
furnace to 185° ±10 0 C.
With valve A closed and valves B and C open, the sample bulb
was evacuated and Immersed in liquid nitrogen. Valve A was then opened
for 30 minutes to allow the contents of the sorbent tube to transfer to
the sample bulb. Valves A and C were then closed, the LN 2 flask removed,
the sample bulb allowed to equilibrate to room temperature, and the
pressure recorded.
The volume of the entlre’manLfold system, including each sample
bulb, was carefully calibrated and on the average was 68.8 cc. With this
Information and the recorded pressure (assuming the temperature to be
constant), the volume of the material desorbed was calculated according
to the ideal gas law:
Ply 1
where: P 1 the pressure measured, 1n nwn Hg
V 1 the volume of the sample bulb and manifold, ‘ 68.8 cc
P 2 760 mm Hg
V 2 — the-calculated volume of desorbed material at 1 atm.
The desorbed material was recovered from the manifold by
reimmersing the sample bulb in LN 2 . Valve B was then closed and the
sample bulb removed from the desorption system for analysis while still
at negative pressure.
G.3.2 Analytical methods
The extracted and prepared samples were analyzed by various
methods. The specific analyses selected depended to some extent on the
samples’ forms, which were:

-------
173
For Inorganics
- Solids
- Aqueous and acid solutions
For Organlcs
- Aliquotsof the neat solvent extracts
- Concentrates of the solvent extracts
- Desorbed materials from the sorbent tubes
G.3.2.l Analyses for inorganics . The inorganic composition of the solid
samples was determined by spark source mass spectrophotography (SSMS)
with electronic detection, which will detect elements present down to
a concentration of 1 ppm. This SSI4S technique has an accuracy from 100-500%
and will survey the sample for all the elements except gases (i.e.,
H, N, 0, He, Ne, Ar, Kr, Xe, and Rn) and Hg because of its high volatility.
Other elements with appreciable volatilities, such as Be and Se, have far
reduced accuracies in this analysis. - -
The aqueous and acid solutions were’first surveyed for metals by
Inductively coupled argon plasma optical emission spectroscopy (ICPOES).
The ICPOES analysis determines 32 elements, Including most of the toxic
elements of interest In the program, down to ppb levels, with an accuracy
of 100-200%. The purpose of this survey was primarily to check that the
metals in the test samples were present in approximately the same amounts
relative to each other as they were in the waste material. Those elements
which, from the results either of the ICPOES survey or of the analysis of
the waste material, seemed to be present at potentially toxic levels, were
determined quantitatively by atomic absorption spectrometry (AAS). The-
sensitivity of this method varies :k 0m approximately 1.0-0.001 ppm for
the elements which were determined ,viith-an accuracy between 10-50%.
G.3.2.2 Analyses for organics.’ ’ Organicconstituents of the test samples
were determined by a combination of the following techniques:
- infrared spectrometry (IR); -
— gas chromatog phy (GC) with either an electron capture (EC)
or flame ioni tion detector (FID);
- low solutioi iiiass spectrometry (LRI4S);
— combined gas chromatography/mass spectrometry (GC/MS).

-------
1 7i
She aliquots of the neat solvent extracts were analyzed only
for volatile organochiorine compounds. EC/GC was used for this analysis
and the samples were compared to standards of CH 2 C1 2 , CHC1 3 , and CCl as
well as the original waste material. The column was 183 cmx 0.635 cm o.d.,
glass, l.5 DV 17 and i.95 QF-l on Chromosorb WHP.
The concentrates of the solvent extracts were analyzed to identify
the less volatile compounds that would not be lost by evaporation of the
extract. An aliquot of each sample was first taken and evaporated in a
clean, 7 ml weighing bottle at ambient conditions to remove the solvent.
The residue was then weighed on a micro balance to 0.001 mg and scanned
by IR as an evaporated thin film. The sample was returned to the weighing
bottle and stored for use In the LRMS analysis. To obtain an LRMS spectra,
the sample was rinsed Into the glass vessel for the solids-probe inlet system
and the solvent evaporated again at ambient conditions. The IR and LRMS
analyses yield qualitative information about the classes or types of
compounds (e.g., hydrocarbons, ‘phenols, POM’s, etc.), present as well as
anidep of the complexity and toxic nature of the concentrated sample.
The sensitivity of the LRMS solIds- probe .techn’ique’for specific
compounds in an organic matrix can be estimated even though the mass
spectrometer Is not strictly considered a quantitative instrument. At
typical mass spectrometer operating-pressures of l&6 torr, the relative
volatility of organic materials at normal solids probe temperatures
(100 -250°c) is not an important factor. That is, the volatilities of
materials of: interest are sufficiently high for adequate detection.
The solids probe of the mass spectrometer Is In such close proximity to
the ionizer that sample diffusion does not significantly reduce sensitivity.
Previous experience and contacts with other laboratories using similar
equipment indicate that If ‘ 0 micrograms of an organic compound is
present in a solids probe along with other material, it will be detected
to the extent that the ten or more strongest fragment peaks will be
recorded. Since the typical weight of sample placed in the solids probe
is 1 milligram, the required weight percentage of a given organic compound
to be detected in an organic matrix is approximately one percent. Therefore,
any compounds not detected by LRMS were assumed to be present at less than
one percent levels. -

-------
175
The organic compounds in the extract concentrates were separated
and quantified by CC using both flame ionization and electron capture
detection. Columns and CC conditions are given below:
Flame ionization detection
- Varian 1860, dual differential FID
— Columns: dual, 183 cm x 2 c l i ii i.d., stainless steel, 3.5%
OV-l7 on 100/120 mesh Chromosorb WHP
- Temperatures: column, ambient for 5 minutes, then ambient
to 275°C at 10°C/mm; detector, 275°C; injectàr, 250°C.
— Flow rate: helium carrier at 30 ml/min; hydrogen at
30 mi/mm.; air, 300 mi/mm.
- Attenuation: 1 x io10 a/mv
Electron capture detection
- Tracor MT-1 50, • 63 Ni slng:le ECD
— Column: 183 cmx 0.4 cmi.d., glass, 1.5% OV—17 and 1.95%
QF-l on 80/100 mesh
— Temperatures: column, 200 0 C; detector, 225°C; injector 225°C.
- Flow rates: pre-purified N 2 carrier through column at 60 mI/mm.;
detectcr purge at 40 ml/mln.
-8.
- Polarizing voltage: 14V; bucking range -2 x 10 ; input
attenuation, 102; output attenuation, X2 to X64.
G.4 Analytical Data
Data obtained from the various analyses performed will be
presented in this section in the following order:
Chlorinated Hydrocarbon Wastes Tested
- Chlorinated Aromatics
- Polychlorinated Biphenyls (PCBs)
Samples Obtained from the EPA and ORF Sampling Trains
- Organic Constituents
- Inorganic Characterization
Solid Residues and Effluents from the SLC Kiln Process
- Clinker Products
— Electrostatic Precipitator Discard Dusts

-------
1 76
G.AI.1 Chlorinated hydrocarbon wastes tested
- Samplesof the chlorinated wastes burned at St. Lawrence Cement
were taken by ORF from the liquid waste feed tank on each day of testing.
Composites of these were sent to TRW. The analyses used to characterize
the wastes and determine the expected compounds of interest in the test
burn samples were:
- - thermal content — gross heat of combustion;
— viscosity;
- specific gravity;
— loss on ignition (LOl);
- C, H, N, S, and halogens;
- infrared spectroscopy (IR);
- gas chromatography/mass spectroscopy (GC/MS); and,
- spark source mass spectroscopy (SSMS).
G.4.l.l Primary fuel . In addition to the two wastes, a sample of the
,primary fuel, bunker ‘C” oil, was-also received.1 Of the-analyses ‘listed
above only the gross elemental ‘determination (C, H, N, S, and halogens)
and a trace metal scan by X-ray fluorescence (XRF) were performed on the
oil. Physical properties were not determined since they are fairly standard,
and accurate compound identification was not necessary since the background
of compounds contributed by burning the bunker “C” oil as fuel can best be
determined by analysis of the combustion test samples from the baseline
burn. -
The elemental determination on the oil gave the following
results: -
- 86.87% C
- 9 .89% H
- C?36% N
- 2.62% S
- 0.05% Cl (total halogens as chlorine).

-------
177
The XRF scan detected six elements at the levels listed below:
Element Approximate Concentration (ppm )
SI >500
P 50-500
Ni 5—50
5-50
Ca <25
Ii <5
G.k.l.2 Chlorinated aromatics . The aromatic waste was a dark brown,
low viscosity liquid that was visually free of sediment. The measured
physical characteristics were:
- thermal content - 5170 kcal/kg (9310 Etu/ib);
- viscosity — 1.09 centlstokes at 38°C (100°F);
- specific gravity - 1.281 at 16°c (60°F);
- LOI - 99.98%.
Elemental analyses performed showed the following composition:
44.00% C
- 3.48% H
- 0.028% N
- 0.019% S
- 49.10% C I (total halogens as chlorine).
Analytical techniques used to determine the organic composition
included IR and GC/MS. -
IR
The JR spectrum indicated the waste to be composed primarily of
aromatic hydrocarbons as wall as aliphatic alkanes and alkenes with a strong
response in the 600-800 cm region which can correspond to C-Cl bonding.
No indication of any other functional groups such as phenols, ethers, or
amines was found. Comparison of the waste spectrum with spectra from
the Sadtler indices for the compounds •found by GC/MS, showed the principle
*El ent is potentially toxic - OSHA TLV of <1 mg/rn 3 for an eight—hour
exposure. -

-------
178
constituent to be o-chlorotoluene, the most characteristic peak of which
is a strong, sharp peak at 750 cm . -
G C/MS
- Chromatographic separation was carried out on a Finnigan GC/MS
system using two columns. One was Chromosorb 101, temperature programmed
from 300 - 220°C at 10°C/mm, and the other was OV-17 temperature programmed
from 30° - 275°C at 10°C/mm. The compounds shown in Table G.2 were
identified and their concentrations calculated based on relative peak
areas found from both columns and assuming equal response to the total
ion monitor of the Finnigan instrument. These results reflect the fact
that, in the course of performing the tests at SLC, the waste feed tank
was not emptied between wastes. Thus, the- material for the “aromatic”
burns contains ‘ i33% chlorinated aliphatics carried over from the previous
burn (as a cost savings to the program, TRW did not plan to analyze
samples from. the chlorinated aliphatics burn). The primary constituent
of the aromatlcwaste is chlorotoluene (52.5%) wIth the remainder made up
of three other chlorinated aromatics: dichlorotoluene, octachiorocyclopentene,
and octach loronaphtha lene.
TABLE G.2. ORGANIC COMPOS 111014 OF AROMATIC WASTE BY GC/MS
Estimated Concentration
Compound (% w/w)
Acetone 1.6
Methylacetate 0.3
Dichloromethane 0.4
Chloroform 11.1
Carbon Tetrachloride 13.9
Dichioroethane 3. 1
Trichloroethane 2.5
Tetrachloroethane 0.4
Trichioroethylene 1.0
Toluene 1.0
Chlorotoluene 52.5
Dichlorotoluene 4.6
Dimethyl Benzene (Xylene) 0.9
Octachlorocyclopentene 3.7
C 10 C1 8 (Octachloronaphthalene) 3.0

-------
179
SSMS
Trace metals in the chlorinated aromatic waste were determined
by spark source mass spectroscopy (SSMS). The waste was first ashed by
a suiphated dry method. A very small amount of ash, 0.014%, was obtained
of which the major elements were Ba, Fe, Na, P, S, Si, and Ti (the
sulphur most likely represents the contribution from the H 2 S0 4 added during
ash ing). Other elements detected by SSMS down to 1 ppb are shown in
Table G.3. This analysis was performed to determine what potentially toxic
elements might have been present in the waste at high enough concentrations
to warrant quantitative examinationof the flue gas test samples. Given
the semi-quantitative nature of the data, the SSMS results should not be
construed as a quantitative characterization of the waste material.
TABLE G.3. TRACE METALS IN THE CHLORINATED AROMATIC WASTE BY SSMS
Approximatea Approximatea
Concentration. - Concentration
Element - (ppm) . Element (ppm)
Al 0.7-1.5 Sn 0.008
Ca 0.7-1.5 Zr 0.008
Cu 0.7-1.5 Vb 0.007
Mg 0.7-1.5 A 5 b 0.006
Zn 0.7-1.5 Ag 0.003
K 0.6 Cdb o.ooz
Crb 0.4 Ce 0.002
Pbb 0.3 La 0.002
Mn 0.07 Sbb 0.002
0.07 Beb 0.00)
Sr 0.06 Bi 0.00)
Pt 0.05 Ge 0.001
B 0.05 Li 0.001
Cob 0.05 Rb 0.001
Ni 0.03 - Seb 0.001
U 0.01 Sm 0.001
Mo 0.008 Yb 0.001
aThe accuracy of this technique ranges from 100 to 500 percent.
bPotentially toxic metals — OSHA TLVof cl mg/rn 3 for an eight-hour
exposure. - -
cHg was determined by a highly quantitative atomic fluorescence
technique.

-------
180
G.Le.l.3 Polychiorinated bi.phenyls (PCB’s) . The PCB waste was a medium
brown, low viscosity liquid with fine suspended particulate that tended
to float to the top and cake out on the sides of the container. The
measured characteristics were:
- thermal Content - 6,710 kcal/kg (12,083 Btu/lb);
- viscosity - 5.87 centistpkes at 38°C (100°F);
0 0
— specific gravity - 1.196 at 16 C (60 F);
- LOl - 99.90%
Elemental analyses performed showed the following composition:
— 57.94% C
— 5.23%H
- 0.018% N
- 0.12% S
- 33.60% Cl (total haloqens as chlorine)..
The analytical techniques usedto determine the organic .con posi-
tion of the -waste were IR and GC/MS.
IR
The IR scan showed strong peaks at 1100 cm 1 and 1450 - 1480 cm
which are characteristic of biphenyls, and a peak at 750 cm 1 indicative
of remaining o-chlorotoluene In the waste. Other small peaks in the
spectrum matched those In the aromatic waste IR scan which substantiates
-the GC/MS results discussed below.
GC/MS
Analysis of the PC waste was performed by the same GC/MS
methods and cánditions as for the aromatic waste, and the compounds that
were identified are shown In Table G.le. The PCB waste composition showed
the effect of previous waste material left in the feed tank. It was composed
of 12% chlorInated aliphatlcs, 33% chlorInated aromatics, and 45% PCBs.
There were 22 separate PCB compounds Identified including several isomers
of each of the multiple chlorinated biphenyls.
SSMS
Trace metals in the PCB waste were determined by SSMS. The waste
was first ashed by a sulphated dry method. A small amount of ash, 0.15%,

-------
181
TABLE G.4. ORGANIC COMPOSITION OF PCB WASTE BY GC/MS
Estimated Concentration
Compound (% w/w)
Water 0.7
Acetone 1.3
Methylacetate 0.5
Methanol 0.4
Chloroform 1.4
Carbon Tetrachioride 6.8
Ethanol 5.2
Dichloroethane 0.6
Trich loroethane - 0.7
Hexachloroethane 0.2
Trichloroethylene 1.6
Tetrach loroethy lene 0.2
Xylenes 2.3
Toluene 1.0
Chloroto luene 17.2
Dich loroto luenes 2.4
.Trich lorotoluenes 3.3
Octachlorocyclopentene 4.0
Chiorobiphenyl . 0.5.
Dichlorobiphenyls 8.4
Trichlorobiphenyls 15.5
Tetrachlorobiphenyls 11-15
Pentachlorobiphenyls 2-6
Hexachlorobiphenyls 1-2
Heptachlorobiphenyl s 0. 5
Octachlorobi phenyls 0.2
Di-n-octyl or Di-ethyl Hexyl Phthalate 4.0
C 10 C1 8 - Octach loronaphthalene 2.5
was obtained of which the major. elements were Fe, Na, P, Pt, and Si.
Other elements delected by SSMS down to 1 ppb are shown in Table G.5.
This analysis was performed to determine what potentially toxic elements
might have been present in the waste at.. high enough concentrations to
warrent quantitative examination of the flue gas test samples. Given the
semi-quantitative nature of the data,thó SSMS results should not be
construed as a quantitative characterizatIon of the waste material.
G.4.2 Samples obtained from the EPA and ORF sampling trains
Samples obtained by the two trains from the WBB, WBC, and BLB
tests were received for analysis. A list of the samples derived by ORF

-------
182
TABLE G.5. TRACE METALS IN THE PCB WASTE BY SSMS
.
...
-
Approximatea
4proximatea
Concentration
Concentration
Element
(ppm)
Element
(ppm)
K
-
7—15
Seb
0.1
Zn
Cu
-S
-
7-15
7
5
Srb
HgbPC
Cd
.
0.07
0.06
0.05
Mg
5
Th
0.02
Al
4
La
0.01
Ni
4
Nb
0.01
Mob
- Ba
3
2
BI
Li
0.009
0.009
Ti
2
Rb
0.006
B
1
Ta
0.004
Ca
1
Au
0.003
WL
CrU
Pbb
Vb
1
0.8
0.8
b.7
Nd
Hf
,Sm
Beb -
0.003
0.002
0.002
‘O.OOl
Ag
0.6
Dy
0.001
- Mn
- Sbb
0.5 -
0.5 -.
Ga-
Ge
0.00)
0.001
SnL
Asu
0.5
0.2
Pr
Sc
0.001
0.001
Zr
c 0 b
0.2
- 0.1
U
0.001
aThe accuracy of this technique ranges from 100 to 500 percent.
bpotentlally toxic metals - OSHA TLV of cl mg/rn 3 for en eight-hour exposure.
.CH 9 was determined by a highly quantitative atomic fluorescence
technique.
from each trainwas presented In Table G.1. All of these samples were
scheduled to be analyzed until, In early April, ORF’s review of their
analytical data brought them to conclude that the solvent extracts of the
BLB samples had been contaminated with PCB waste material [ G.l] which was
confirmed by the analyses at TRW. The contamination appeared to have
occurred before the split of the samples Into aliquots. In view of
this development, analysis of the BLB solvent extract samples was
discontinued and extracts from a BLA test were sent as replacements.
However, there was no evidence of a contamination problem with
the remaining BLB samples. Thus, the filters, probe rinse Insolubles,

-------
183
aqueous solutions, and Chromosorb 102 sorbent tubes from the BLB
tests were analyzed for a background reference to the corresponding
WB8 and WBC samples.
All of the test samples were coded for ease of reference.
The coding used was presented In Figure G.1, which also showed the
combination of the probe rinse samples with other appropriate samples.
The reasons and methods for combining the probe rinses with either the
filter (for the EPA train) or impinger (for the ORF train) samples,
as well as the preparation and analytical techniques for all the samples
were described In Section G.3.
Where constituents found In the test samples are reported as
concentrations in the effluent gas, the sample gas volumes needed for •the
calculations were taken from Section 5.3 of the report. Volumes for. the
ORF train were taken from Table 4, and volumes for the EPA train from
Table 7.,
G.4.2.l . Organic-constituènts . SamØles for the analysis of organic
cor iposition were in the form of:
- aliquotsof the neat solvent extracts;
- concentrates of the solvent extracts;
- residues from the evaporation of the extract concentrates; and
- desorbed materials from the sorbent tubes.
These samples were first surveyed for their qualitative nature by gravi-
metric, infrared spectrometry (IR), and low resolution mass spectrometry
(LRMS) techniques. Quantitative determinations and analysis for specific
compounds, such as chlorinated hydrocarbons and PCB’s, were then performed
by gas chromatography (GC) and combined gas chromatography/mass spectrometry
(GC/MS). The results obtained from all of these analyses are presented
in the following paragraphs. The methods themselves are described in
Section- G. 3.
Qualitative data . The purpose of the qualitative or survey analysis
was to Identify any material in the samples which were not present
in the original weste, and therefore were not expected. in addition,
this survey analysis searched for secondary or incomplete combustion
products which would result from the waste being converted to compounds

-------
1 8’
other than C0 2 , H 2 0, and HC1. T ie qualitative data will be described
in two groups: 1) Extract concentrate residues, and 2) material desorbed
from the sorbent tubes.
1) Extract concentrate residues - The results of these qualitative
analyses on the residues, obtained by evaporating aliquots of the solvent
extract concentrates, are sununarized and presented in Table G.6. This
table Includes: -
- The Identification of the extract samples according to the
coding shown In Figure C.). This coding shows how certain
related samples have been combined.
- The Sample/Al iquot Ratio shows the total volume of extract
• prepared by ORF In the numerator and the volume received by TRW
in the denominator. One-fourth of each extract was sent to
TRW.
- The amount of residue found in TRW’s portion of the total
extract is expressed as milligrams (mg). Values have been’
- corrected for appropriate blanks. -
- The Total Residue Is that amount (in mg) of residue found in
TRW ’s portion multiplied by the Sample/Al iquot Ratio. This
value is indicative of what may be found ln.the total extract
if analyzed by the same procedures.
Inspection of the IR and LRMS data revealed that the only
materials detected in the evaporated residues of the various extracts
(except for the contaminated BLB samples) were silicones, hydrocarbon
olls,fatty acids or fatty acid esters, and phthal lc acid esters.
These same materials were also present in the blank and .ontrol
samples; their concentrations vary from sample to sample as estimated by
LRMS. These materials, as classes of compounds, are not toxic and would
not be considered an emissions hazard even If they were present in much
higher concentrations. These materials are not believed to have come from
the combustion gas. Samples from the EPA train did not show increased
levels of these materials consistent with more than ten—fold sample volumes.
In the analysis of trace organics, It becomes very difficult to
completely avoid these ubiquitous oils, plasticisers, lubricants, antioxidants,

-------
185
TABLE G.6. SUMMARY OF ORGANIC QUALITATIVE SURVEY ANALYSES
OF SAMPLE EXTRACTS
Sample Identification
m tt1Ct
R
W Aliquot
Survey Residue
Aliquot (eg)
Found In TRW
;Total Residue
(mg, corrected
Factor)
for Aliquot
Baseline Burn A
BLA-ORF-FE-T3
8LA- ORF-PRE+NAOHE-T3
BLA- ORF-N,OE-T3
BLA-EPA-PPE+FE—T3
BLA-EPA-IE—T3
200/50
200/50
100/25
300/75
100/25
3
N/D
N/&’
0.12
2.09
12.92
MID 
-------
1 B6
etc., that can enter the test matrix in very small amounts despite -
the most-careful and extensive preparation of equipment and selection
of reagents. - One cannot be sure whether the source of these materials -
is the hardware, the reagents, the samples themselves, or combination
of these. A large number of blank and control samples might have
distinguished the sourcd, but due to the noncritical nature of the materials,
such an examFnation was clearly not-warranted.
It was pointed out that the survey analysis of the BLB samples- -
known to contain PCB’s [ G.1) did, In fact, reveal their presence. Evidence
of PCB,s was clearly shown In the LRNS data for the following BLB
samples: -
- ORF train,, combined probe rinse extracts and caustic
impingerextracts; -
- ORF train,-water impinger extracts;
- --EPA train, combined probe rinseextract and filter extract.
Since these PCB-’swerefound only as estimated minor constituents in samples’
whose total wáights were -lessthan one-milligram, the ability of the ma s -
spectrometer ‘to detect -tnese small quantities was dearly established.’ ‘The
LRMS procedures can usually, detect 10 micrograms, or about one percentof
a typical l.mi.lligram sample. -The presence of a specific compound in an
organic matrix is not normally detected by IR-with adequate certainty
unless present,at 10 percent or higher. The sensitivity and detection
limits of these techniques were discussed in Section G.3.2.
It is stressed that chlorinated species were specifically
searched for in all of these residue samples and none were found except
for In the contaminated BLB samples described above.
2) Materigls desorbed.from the sorbent tubes - The Chromosorb 102 sorbent
tubes from the ORF sampling train were thermally desorbed to recover the
collected sample accordin9 to the procedure described in Section G.3.l.
The desorbed gases were then analyzed qualitatively by LRP4S using a gas
inlet system. The conclusions drawn from reduction and interpretation
of the LRMS data were that all of the sorbent tubes contained essentially
the same types of materials; however, the relative amounts of these
constituents did appear to vary from sample to sample.

-------
- 187
- Although the LRIIS Is a qualitative technique, estimates can
be made through inspection of the-data to provide an indication as to
whether a species is present as a trace, minor, moderate, or major component.
Such a summary for the six selected sample sorbent traps that were analyzed
is presented in Table G.7. An unused sorbent trap was also analyzed by the
same procedure to estimate background contributions by the Chromosorb 102.
Unfortunately a leak in the system resulted in the loss of sample and,
therefore, usable data. Only carbon dioxide was seen in the mass spec
data at greater than trace levels.
The volatile compo inds removed from the sorbent traps and
collected in the gas sample bulbs were largely the oxygen, nitrogen,
carbon dioxide, and water that are the usual species found in combustion
gases. The remainder of the compounds detected were those that could
indicate incomplete combustion, or possibly the formation of other
compounds. The levels at which these remaining compounds could be
present were estimated on the basis of LRMS instrument response for key
peaks attributed-to these trace residual organics. These estimates
indicated that the C 1 to C 5 hydrocarbons present as a major component
of these trace residual organics could be present in the sampled
combustion gases in the I to 10 ppm range. The minor or trace components-
of these trace organics were believed to be present at very much lower
levels.
A summary of the LRMS peak patterns observed in the spectra
of the desorbed gases and the compounds assigned to these patterns is
given in Table G.8.
Quantitative data . Quantitation of the compounds detected and identified
in the test samples was performed by GC and GC/MS techniques.
1) Extract concentrates - The chromatography with flame ionization
detection was performed chiefly to analyze organic-species. It should
be noted that the FID is quite sensitive.to-chlorinated hydrocarbons
as long as chlorination is not complete. At the electrometer settings
used, the sensitivity,- or the minimum detectable quantity, was 0.002
as benzene, naphthalene, or Aroclor 1232.
The chief result of the analyses using the FID was that none
of the concentrated samples showed peaks other than those found in the

-------
TABLE G.7. APPROXIMATE CONSTITUENT LEVELS OF TRACE VAPOURS
DESORBED FROM SORBENT TUBE SAMPLES BY LRMS
Sorbent
Tube
Sample
a
Constituent level
1-5
Hydrocarbons
Nttrornethane
NO, NO 2 ,
and
possibly:
Ethanol
Benzene
Ilethyl
Chloride
,
.
Substituted
Benzene
,
Methyl
Siloxane
WBB-T1
WBB-T2
WBC-T1
WBC-T2
BIB-Ti
BLB-T2
Minor
Major
Major
Major
Major
Major
Trace
Minor
Minor
Minor
Moderate
Minor
Trace
Major
Major
Major
Moderate
Moderate
Trace
Moderate
Minor
Moderate
Major
Moderate
Nob
Moderate
ND
‘ND
ND
ND
ND
Trace
Trace
Trace
Trace
Trace
Major
Minor
Minor
Trace
Trace
Trace
aT range of trace to major Ieve7s represents 0.1-10 ppm concentrations In the fluegas.
bND - Not Detected.

-------
189
TABLE G.8. SUMMARY OF ThE INTERPRETATION OF LRMS SPECTRA FOR
TRACE VAPOURS DESORBED FROM SORBENT TUBE SAMPLES
41, 1 .3,
28. 32
144
30, 46, 61
30, 46
78 , 5o,
50, 52,
91
147, 207, 281
solvent controls. Consequently, a “less than” value was assigned in
terms of benzene for the analysis of hydrocarbons, or Aroclor 1232 for
- the analysis of chlorinated organics. The results are given In Table G.9.
The values presented InTable G.9 were derived in the following
manner. The sample labeled WBB-EPA—FE+PRE, for example, totaled 225 ml
as received. A 213 ml portion of this sample was concentrated to 10 ml
ma Kude,rna-Danish evaporator. - The chromatogram of the concentrated
sample showed no peaks other than solvent. Since the aromatic waste
contained considerable chlorinated hydrocarbons, quantitat ion as Aroclor,
1232 was considered appropriate.
Thus, WBB-EPA-FE+PRE contained 
-------
TABLE G.9. RESULTS AND DETECTION LIMITS FROM GC/FID ANALYSIS OF TIlE CONCENTRATED EXTRACTS
Voltmue of s 1e received by TRW.
bA liq,Jot Factor: for ex 1e 67 ml Of the 75 .1 BLA-EPA-Fë+PRE sample received was concentrated to
CFractGon of S.ui,le: *11 s ,les received by TRW were 251 of the total sample.
dihe instninoet was calibrated with Aroclor 1232. However, th Aroclor peak pattern
was not fo md in any saiple chromatogra.. High sensitivity GC/NS (3vg/m 3 of flue
gas) did not detect any chlorinated species In s 1es WBC-ORF-FE and
W BC-061-N20C.
e ’_ Not detected. Values In parentheses indicate the detection limits.
,
S le ID
/ •
H?ocarb ofldL
Volome
S 1e
of
( 1 )a
Atiquot
Factorb
Fraction of
SaIIIP1eC
Volume of Sa 1e
Gas, Std.
Concentration
in Flue Gas
(m g i . 3 )
DLA-EPA—FE4PRE-T3
-1f-T3
0.052
0.26
‘ 75
25
67/20
21/10
1/4
1/4
3.76
3.76
0.001
0.003
-ORF-FE-T3
0.26
50
46/10-
1/4
0.32
- 0.04
- 1 120 1- 73
0.061
25
21/10
1/4
0.32
0.09
-PRE+ IIA O II I-T3
0.21
50
42/20
1/4
0.32
0.06
WRB-EPA-FE PRC
-IC
ND( o.0038)e
0.89
225
75
213/10
63/10
1/4
1/4
9.62
9.62
,
ND(c O.00002)
0.00 1
- ORF4E
0.12
150
138/10’
1/4
‘0.83
0.006
l!i M0,I 1
0.26
0.053
50
100
42/10
42/10 -
1/4
1/4
0.83
0.83
0.02
0.003
C-CPA-FfPR1
—if
-ORF-FE
: MflE
0.019
.--- (lost)
1 1*0.0038)
ND(0.0038)
0.26
225
75
150
75
, 150
‘


213/10.,
63/10’
138/10
- 63/10
138/10
1/4
1/4
114
1/4
1/4
,
9.91
9.97
0.73
0.73’
0.73
0.00008
(lost)
ND(c O.00002)
N0(cO.00002)
0.015
tD
0
20 ml for analysis.

-------
191
of the whole sample was divided by the volume, in rn 3 of gas sampled.
A concentration value of <0.009 mg/rn 3 in the flue gas was thus derived
for the sample WBB-EPA-FE+PRE.
Since the minimum detectable quantity is constant and since both
the volume of gas sampled and the volume of extract vary, this method of
calculation gives different values of mg/rn 3 . It must be stressed that
these analyses showed no species other than solvent.
Since the ECD is much more sensitive to polychlorinated species
than the FID, the analyses using electron capture detectlbn were performed
to determine po lychlorlnated compounds below the level that could be
detected by the FID analyses. At the electrometer setting used on the
electron capture detector, the minimum detectable quantities for typical
species of interest were:
- Aroclor 1232: 0.0038 ng/ul
-, Naphthalene: 0.066 ug/ul -
- Benzene:. O. II ug/ijl.
The most noteworthy result of the GC analyses using the ECD
was’that none of the concentrated extracts showed peak patterns corresponding
to those given by Aroclor 1232 (the PCB mixture most similar to that
burned in waste burn C) or Aroclor 1242. The quantified ECD results are
given in Table G. 10.
Some of the concentrated sample extracts, e.g., WBB-EPA-FE+PRE,
showed no peaks other than solvent. Most of the extracts showed several
very small peaks. After quantitation, all but one of the samples were
below the level of interest, 0.1 mg/rn 3 , and no further effort was expended.
One sample, WBC-ORF-FE, showed a number of peaks, some of which were large.
Another sample, WBC-ORF-H 2 OE, showed a large number of small peaks. Both
of these samples were analyzed by GC/MS. The GC/MS analyses of the filter
extract indicated one vanishing small hydrocarbon peak, a small phthalate
ester peak, and 10 peaks which were various trimethylsilyl compounds
(TMS). The GC/MS analyses of the water extract showed extremely small
hydrocarbon peaks and 16 peaks that were trimethylsilyl compounds.
Because these TMS compounds were most likely the result of some contamina-
tion, the peaks appearing in the chromatograms were considered spurious.

-------
TABLE G.lO. RESULTS AND DETECTION LIMITS FROM GC/ECD ANAISIS OF THE CONCENTRATED EXTRACTS
Detectable
timits of
- Vohmie of Concentration
pg/u i as pg)’pl as Yolimie of M1quo Fraction of S 1e Gs, in Flue Gas.
Sample ID Benzened Ifa1ocavbon 1 Sample. ml ’ Factor Samplec Std. rn’ .gFm 3
BLA-EPA-FE4PRE-T3 N0( 4.002)° - 75 67/20 1/4 3.76 1 10(4.05)
-IE-T3 ND 4.002 - 25 21/10 1/4 3.76 NI) (0.03)
- ORF-Ft-T3 ND 4.002 - 10 46/10 1/4 0.32 ND 4.3
-H20(-T3 MD 4.00? - 25 21/10 1F4 0.32 110 €0.3
-PNLiIINIIIE-13 MD 0.002 - 50 42/20 1/4 0.32 ND 4.6
ND8-IPA-FEP 00(4.002 225 213/10 1/4 9.62 ND ‘0.009)
-11 00(cO.002 75 63/10 1/4 9.62 ND 4.01)
-ORF-F1 110(4.002 150 138/10 1/4 - 0.83 ND €0.1
- 1 1201 110(4.002 50 - 42/10 1/4 0.83 ND €0.1 —
-P ’IS IHE Im(c0.002 100 92/10 1/4 0.93 80 (0.1
VDC-EPAF(.P01’ 00(4.002 225 213/10 1F4 9.97 110(4.009)
-11 - 1 1 0( 0.002 75 6)/10 1/4 9.97 804.01)
-O -FE 110(4.002 150 138110 1/4 0.73 ND 0.1
-11201 110(4.002 75 63/10 1/4 0.73 ND 4.1
-P 4fS Jf Np(c0.0 02 150 138/10 1/4 0.73 ND 4.1
aVolune of sample received by TRW.
b*iiqunt Factor: for example 67 ml of the 75 ml NLA-EPA-F.+PRE s le received was concentrated to 20 ml for snelysis.
C 1 1 of s 1e. All samples received by TRW were 25% of the total sample.
instru.ent was calibrated with benzene or Aroclor 1232. However, the Aroclor peali pattern was not found In any sample
chrunatogram. 111gb sensItivity CC/MS (3pg/m 3 of flue gas) did not detect any chlorinated species in samples NIC-ORF-FE
and ISC-ORF-1120(.
e 110 Not detected, values in parentheses indicate the detection limits.

-------
193
These GC/MS analyses did not detect any chlorinated species in
either sample (typical Sensitivity of 3 1zg/m 3 of the flue gas).
2) Aliquots of neat extracts - There was concern that the process of
concentrating the sample extracts, as received In pentane and/or hexane,
might cause the loss of relatively low boiling halocarbons such as
chloroform which might be expected in the samples. Therefore, the retained
portions of the as-received extract samples were chromatographed on the
Tracor Instrument. The column was operated Isothermal ly at 70°C to increase
resolution of low boiling compounds. The results are given in Table G.l1.
Most of the samples showed no peaks other than solvent, Of those
samples showing peaks other than solvent, all quantified below the level
of interest (0.1 mg/rn 3 ), and no compound identification by GC/MS was
performed.
3) MaterIals desorbed from sorbent tubes - The sorbent traps were desorbed
as discussed inSection G.3.l. -Portions of the desorbed vapours contained
in the sample bulbs were chramatographed Isothermally at 70°C with electron
capture detection. The samples were qualitatively similar to the sorbent
trap blank (unused chromosorb 102) and the baseline B (fuel oil burn)
samples. The waste burn B samples each contained a peak which might be
CH 3 C1: WBB Test 1, 0.0001 mg/rn 3 ; and WB8 Test 2, 0.0002 mg/rn 3 . The
waste burn C samples were essentially the same, qualitatively and
quantitatively as the blank and baseline B samples.
G.k.2.2 Inorganic characterization . InorganIc elemental concentrations
in the flue gas were determined by analysis of the particulate filters
and the aqueous probe rinse and impinger samples. The filters (includ:ng
the filtered probe rinse Insolubles) were all acid digested. Out of these
20 digests (18 test samples and two blanks), three samples were prepared
for a trace element survey analysis by combining equal aliquots from each
of the three tests of WBB, WBC and BIB using the filter samples from only
the EPA sampling train. The EPA train samples were selected on the basis
of 1) representIng larger sample gas volumes, and 2) having been taken by
a method specific for accurate and representative particulate sampling.
The results of the survey analysis are shown in Table G.l2. The blank

-------
TABLE G.I1. RESULTS AND DETECTION LIMITS FROM GC/ECD ANALYSIS OF THE UNCONCENTRATED EXTRACTS
ng/ul as
Halocarbon
Volume
Sample
of
.
- -
Fraction of -
Sample
Volume
of Sample 3
Gas, Std. m
Concentration
in Flue 3 Gas,
mg m
,
,
BIA-EPA-FE+PRE-T3
ND(<0.0038)d
75
3.76
ND(cO.0003)
-IE-T3
ND(cO.0038)
25
?
3.76
ND(cO.000 ))
-ORF-FE-T3
ND(cO.0038)
50
A
0.32
ND(cO.002)
-H 2 OE-T3
-PRE-NAOHE-T3
0.0051
ND(cO.OO38)
25
50
A
A
0.32
0.32
o.b 016
ND(.cO.002 1i)
WBB-EPA-FE+PRE
ND(cO.0038)
225
?
9.62
ND(
-------
195
TABLE G.l2. TRACE METAL SEMI_QUANTITATIVEa SURVEY OF FILTER DIGESTS BY ICPOES
S
Element
Average
Concentrationb (mg/rn 3 )
WBB
WBC
BLB
Al
0.078
0.51
0.26
Ba
0.011
0.014
0.012
B
0.018
0.084
0.038
Ca
2.2
11.
6.5
Cd
0.0006
0.002
0.001
Cr
0.003
0.006
0.004
Co
MDC
0.001
NDC
Cu
0.003
0.007
0.003
Fe
0.18
1.1
0.73
Pb
0.047
0.12
0.044
Mg
Mn
0.076
0.003
0.38
0.018
0.21
0.013
Ni
0.003
0.006
0.002
P
0.014
0.055
0.044
K,
S 1
.14.
0.082
.
.32.- .
‘O’.05 1
6.7
0.093.
Ag
Na
0. 0003
1.8
0.003
5.5
0.0004
2. 1 .
Sr
0.004
0.013
0.0008
ti
0.002
0.008
0.004
V
0.0006
0.002
0.001
Zn
0.014
0.042
0.027
aAcCra estimated to be a
factor of
±2
or better.
bCalculated based on average sample gas volumes of:
WBB - 3.2 m 3
WBC - 3.3
BLB - 3.1+ m
Not corrected for filter contribution
CNot detected (<0.0003 ppm)
(an acid digest of unused filters) sample was not surveyed, so the values
reported in this table are uncorrected for the filter background. -
Of the 32 elements that can be determined by the ICPOES analysis,
ten were not detected in the filter digest samples. These ten elements
with their lower detection limits are listed in Table G.13, along with
a calculation of the average detectable limit for each of these elements
in the flue gas.

-------
196
TABLE G.13. LIMITS OF DETECTION FOR UNDETECTED ELEMENTS BY ICPOES
Element
ICPOES Detection
Limit (ppb)
Average Detectablea
Limit in Flue -Gas
- (mg/rn 3 )
Au
5
0.00008
As
40
-
0.0006
•
Be -
1
0.00002
Eu
15
•
0.0002
Mo
11
0.0002
Se
60
,
0.0009
Te
65
0.0010
Sn
50
0.0008
W
90
0.0014
u
.
80
0.0012
aBased on an average sample gas volume of 3.3 cubic meters.
- The results of thIs survey indicated onlyone element, lead, was
present at potentially toxic levels in the’stack. - While nàtaproblem
with the. stack configuration In the present study, the concentration and
-the effect of thermal dispersion should be considered when extrapolating
these results to other kilns. To be sure of an accurate measurement of
the levels of lead and certain other toxic metals which the waste analysis
indicated might be present at levels of Interest, a quantitative determina-
tion by MS was performed on five elements. The results from this analysis
are presented In Table G.l4.
The MS results were corrected for background levels as determined
by analysis of the filter blank sample. It should be noted that back-
ground contrt butions were not only from trace contaminants in the filter
materials but also in some cases from matrix effects In the mixed acid
solutions.
MS was also used to analyze for selected elements In the
aqueous probe rinse and impinger samples. Table G.l5 lists the results
from these analyses and shows that, in general, nothing significant was
found. No background or blank solutions were available so the BLB results
should be used to correct for background effects. The results for these
aqueous samples were left in ppm because the total sample volumes were
unknown.

-------
197
TABLE G.14. CONCENTRATION OF TRACE METALS IN EFFLUENT GAS
PARTICULATE MATTER BY MS
Waste
Burn
Train
Test
Concentration
of Element
(mg/rn 3 )
Ba
Cd
Co
Cr
Pb
WBB
EPA
I
2
3
0008 a
<0.008
<0.008
<0.001
<0.001
0.001
<0.002
<0.001
<0.002
<0.009
<0.002
<0.004
0.028
0.012
0.079
WBC
EPA
1
2
3
0.013
<0.008
0.027
0.001
0.002
0.003
<0.002
<0.002
<0.002
<0.008
<0.008
0.003
0.096
0.103
0.153
BLB
EPA
‘
. 1
2
3
<0.007
<0.008
<0.008
0.001
0.002
0.001
<0.002
<0.003
<0.002
0.001
<0.007
<0.002
0.040
0.062
0.015
a,< a lesstiian or equal to sign, in
detected but not significantly above
TABLE G.15. CONCENTRATION OF TRACE M
dicates those elements whi
background levels.
ETALS IN AQUEOUS SAMPLES B
ch were
Y MS
Waste
Burn
vs.
Train/Component Baseline
Concentration of Element
(ppm)
Pb
Cr Cd
Co
EPA-Probe Rinse WBB
ND
1.9 ND
10.2
WBC
ND
0.03 0.01
ND
BLB
Impingers WBB
ND
ND ND
ND
ND
ND ND
0.05
WBC
MD
0.02 ND
ND
BLB
ND
ND ND
ND

-------
1 98
Analysis of the filter acid digest, probe rinse, and impinger
samples from the ORF train-by AAS confirmed the presence of those elements
found in the EPA train samples. However, since the ORF brain was not
operated with the intent of collecting an accurate particulate sample
(i.e., isoki,Setic, traverses, etc.) the MS results for this train were
not calculated out to concentrations in the flue gas.
G.4.3 Solid residues. and effluents from the SLC kiln process -
- Of the four types of solid samples collected during the test
program:
- clinker-product;
- clinker fines;
— electrostatic precipitator discard dust; and
- - cement mixes, - - -
analyses -were performed only on the cli iker product and discard dusts for
the following reasons. The clinker fines were recovered from air blown
through the clinker productand were then returned to the clinker storage.
The cement mixes were simply clinker product with gypsum added. Thus,
neither of these materials represented unique samples. The’following
paragraphs describe the results of anal ’ses performed on the clinker
products (CR) and discard dust (DD’) samples. If significant amounts of
toxic materlalFs were found ln the analyzed samples, further tests on the
remaining samples would have beenconducted. However, this did not prove
to be necessary.
G. 4.3. 1. Organi c consti tuents . Portions of the clinker product and
discard dust samples were extracted with pentane using a Soxhlet extractor
and preextracted paper thimbles. A blank sample was prepared by running
a pentane extraction on an empty, precl-eaned thimble. In addition, a doped
control sample consisting of ‘ .3O grams .of clinker, to which 0.9 mg
of a known PCB mixture had been added, was also extracted for analysis.
This doped sample represented a PCB concentration ‘ ‘3O ppm in the clinker
product. These samples were all analyzed both qualitatively and quantita-
tively by.the same techniques described in Section G.4.2.l.
Qualitative data . Aliquots of the pentane extracts were evaporated and
the residue was weighed and scanned by both IR and LRMS. These results

-------
199
TABLE G.16. RESULTS OF ORGANIC SURVEY ANALYSIS ON CLINKER PRODUCT AND
DISPOSABLE DUST SAMPLES
Sample
Identification
Weight of
Extracted
Sample (g)
Weight
Residue
Extract
of
in
(mg)
Qualitative Nature of Residue
BLB-CP
30.270
<1.16*
Hydrocarbon oils, phthalate
- BLB-DD
24.185
1.31
-
esters,- traces of fatty acids
and silicones
WBB-CP
29.445
1.30
WBB-DD -
18.948
- 1.16
WBC-CP
21.816
<1.16*
WBC—DD
33.112

-------
200
survey technique of about 1 ppb In the clinker or dust. The IR survey
data was not useful in detecting PCB’s at this level.
Quantitative data . The clinker and dust extracts were ana1yz d by CC with
both FID and LCD detection. The results are given In Tables G.17 and G.18,
respectively. Using the FID, no peaks-other than solvent were found, and
the results are expressed as minimum detectable quantities. Three of the
samples showed no peaks other than solvent when analyzed with the LCD.
The three samples which had peaks. other than solvent using the ECD were
quantitated-at well below the level of-Interest.
To ascertain the recovery of PCB’s from the clinker and dust
samples, a 30 gram sample ofcllnker was-doped wIth 0.9 mg of Aroclor
1232, extracted and concentrated In thesame fashion as the clinker and dust
samples. The ECD chromatogram of this artificial sample matched the pattern
of a neat sample of Aroclor 1232, and the-recovery of PCB’s from the doped
sample was calculated as 25%. This recovery factor has been applied to
the clinker and dust samples in TablesG.ll and G.l8. (The recovery
factor is the reason why the values for the samples in Table G.17 exceed
several ppm. . f the minimum detectable 4iant-ity using the FID is 0.002
ug/ul, then after applying the recover fa tor; the concentration-in the
sample cannot exceed 0.Q p8 ug/ul. For a 10-mi extract volume and a 30 g
clinker sample, the-compositton Is thus -2.7 ug/g.)
G.k.3.2 Inorganic characterization . The CP and’DD samples were also
examined for trace metal content by spark source mass spectrometry (SSMS).
The major constituents -of both types of- .-mater lal were determined to be
aluminum, calcium, iron,.magnesium, sodium, sulphur, and silicon. Other
‘trace inorganic elements which were detected down to a 1 ppm concentration -
are listed in Table G.19. It should be .remembered that SSMS data are
semi-quantitative. The values reported should be considered as being
accurate within 500% of the true value.
The trace metal character of- the clinkers and dusts did not
change significantly from one-waste burn to the other. The relatively
low feed rate of the wastes, compared to that of the precursor material
and the primary fuel, appeared to make-any contribution of trace metals
from the wastes negligible.

-------
201
TABLE G.17. RESULTS AND DETECTION LIMITS FROM GC/FID ANALYSIS
Sample ID
ug/ul asa
Halocarbon
Volume
Sample
of
(ml)
ug/g of
Sample Material
BLB—CP
ND(cO.OO 2 )b
10
•
ND(c3.O)
WBB—CP
ND(cO.002)
10
ND(c3.3)
WBC-CP
ND(cO.002)
•
10
ND(c4.5)
BLB—DD
ND(
-------
202 -
TABLE G.19. SELECTED TRACE METALS IN SLC CLINKER PRODUCT
AND DI SCARD DUST-. SAMPLES BY SSMS
.
-
Element
-
Concentration
(ppm)a
.
WBB
WBC. -
, .
BLB
CP
DD
CP
DD
-
-
CP DD
‘
As
<1.5
.3
<1.5
3
‘
<2.9’
1
•
B
30
30
‘
30
10
30
30
Ba.
Be
70 .
•ND’
70
ND
300
1-
100
ND
70
7
30
ND
Ce
10
30
70
30
.10
10
Co
ND
ND
ND’
3’
-1
ND
Cr
7
10
30’
- 10
30
10
Cu
1
10’
3
7
7
3
Cs
ND
10
ND
30
ND
10
Dy
1
ND
3
1
1
ND
Ga
1
1.
3
1
3
10
Ge
.- ND
ND
ND
ND.
. - 1
3’
-
La
Li ,
7
ND
‘7
1000
30
30’
10
100.
10
100’.
.7
100
:
Mn
1,00
‘.100
3,00
. .300
300
100
.
Mo
I
: 1
1
1
‘3
1
-
Nb
-.3
.3
7.
3
3
3
Nd,
-10
10
10
10
10
‘
7
Ni
7
-3
7
7
.)0
3
Pb
ND
70
.3
- 100
-ND
30
Pr
.3
3
10
3
.
3
1
Rb
7
100
10 -
300
10
70
Sc
ND
1
ND’
ND
7
3
Se
ND
10
Nb’
3
1
1
Sm
ND
ND
1’
1
1
ND
Sr
‘
700
300
700
700
700
100
Th’
1
7’
7’.
7
3
1
.Tl
700
700
700 .
700
700
700
U
V
lD
30
1
30
3
30
3
30
1
70
‘ND
30
Y
3
3
10
3
3
Zn
30
30
70
100
30
30
Zr
30
30
100
70
70
10
a 551 5 data generally ranges within 500% accuracy.
bND - Not Detected ( cl ppm).

-------
203
REFERENCES
G.1 Comunicatlon, GordonThomas, Ontario Research Foundation to
Arnold Grant, TRW Systems, 7 Apr11, 1976

-------
APPENDIX H
DEVELOPMENT, CONSTRUCTION AND EVALUATION OF A
COLLECTION SYSTEM FOR LOW MOLECULAR
WEIGHT HALOCARBONS
by
F.J. Hopton and G.H. Thomas
OntarIo ResearchFoundation
ifisslssauga, Ontario
preceding page blanK

-------
207
APPENDIX H
DEVELOPMENT, CONSTRUCTION AND EVALUATION OF A COLLECTION SYSTEM
FOR LOW MOLECULAR WEIGHT HALOCARBOIIIS
H. I Sunmiary
Numerous experiments have been performed to determine the
effectiveness of absorbents and adsorbents to remove and retain low
molecular weight halocarbons from a flowing gas stream. Initial studies
using.a particulate sampling train, requiring a high gas flow rate of at
least 0.5 cfm, indicated that neither absorbents nor adsorbents would
efficiently collect compounds such as chloroform (CHCI 3 ) present in the
gas stream at low concentrations of a few parts per billion. An inert
adsorbent, e.g. Chromosorb 102, was found, however, to efficiently remove
and retain CHC1 3 at much lower gas flow rates of 0.2 - 1.0 litres/minute.
A collection system for low molecular weight halocarbons was constructed
and evaluated at low flow rates over sampling periods of up to four hours
duration. Results obtained from several tests indicated that the system
had a collection efficiency of better than 90* over the longest time
period. Desorption of adsorbed halocarbons was readily achieved by thermal
treatment of the adsorbent.
H.2 Introduction -
Late in 1975, an experimental program to burn waste chlorinated
organic compounds in a rotary cement kiln was carried out at the St.
Lawrence Cement Company (SLC) plant in Mississauga, Ontario. The program,
sponsored by Environment Canada, was designed to obtain information on
- the use of waste chlorinated hydrocarbons as a supplemental
fuel to the kiln;
- the reduction of alkali concentration of the clinker by the
chlorine contained In the waste materials added; and,
— the effect of emissions from the kiln on air quality.
In order to collect the required information, process and emission samples
were collected during periods when different composite wastes were burned
in the kiln. The emission samples were analyzed for trace chlorinated
organ i c compounds.
Pricedin page blank

-------
208
During preliminary discussions between members of the Ontario
Ministry of the Environment (MOE), the Ontario Research Foundation (ORF)
and SLC prior to initiating the study, some concern was expressed about
the sampling methodology for trace chlorinated organic compounds. Since
a major program requirement was to obtain the particulate emission rate
during each burn, it was hoped that the EPA Method 5 sampling train, used
to collect particulate material, could also be used for the collection
of chlorinated organic compounds. However, it was not known whether
absorbents or adsorbents were more efficient for collection of the organic
compounds, or whether either collection medium would be efficient at the
high gas flow rates required for particulate sampling.
MOE, therefore, requested that ORF complete a laboratory
investigation to evaluate suggested procedures, and to develop an optimum
method for the sampling and collection of trace chlorinated organic
compounds which might be present In the kiln emissions. This report
describes the test program carried out in the laboratory and presents the
data obtained using specific chlorinat’ed compounds as representative
pollutants.
H.3 Background Information
Previous studies involving the incineration of chlorinated
organic compounds (H. I] have indicated that, if complete combustion to
CO 2 H 2 0 and Cl 2 or HC1 is not realized, then trace amounts of volatile
compounds such as Cd 4 , CHC1 3 and CH 2 CI 2 may be present in the combustion
gases. (Though it was unlikely that any HC1 or Cl 2 produced would pass
through a cement kiln without reacting-to form alkali chlorides, it was
necessary to consider these gases as possible pollutants for analysis
also.)
The EPA Method 5 train for particulate sampling is also used to
collect organic compounds in the impingers by condensation and scrubbing.
The efficiency of collection for a particular substance is dependent on
gas flow rate and volatility of the organic compound. For compounds more
volatile than water, the collection efficiency is probably close to zero.
To collect these volatile compounds, specific organic solvents have been
used in the impingers, replacing the water normally present. There is,
however, little data available on the collection efficiencies obtained.

-------
209
In recant years, Organic compounds In the ambient air have been
collected using inert adsorbents [ H.2]. These materials have tended to
replace activated carbon for most applications since they are unaffec!ted
by water vapour, and recovery of adsorbed species Is considered, In
general, to be easier and more efficient. Recovery of adsorbed species
maybe accomplished either by thermal desorptlon or solvent extraction.
An application of this technique has been to collect polycyclic organic
compounds from combustion effluents (H.3].
Bearing In mind that, In the study to be carried Out at SLC
only trace amounts of more volatile organic compounds were expected to be
in the kiln emissions, it was decided to evaluate both adsorbents and
adsorbents in an EPA Method 5 sampling train for collection, retention
and subsequent recovery of compounds such as CHCI 3 and CH 2 C1 2 . If a
collection method proved to be effective for capture of these compounds,
then it would almost certainly be effective for the collection of higher.
-mólécular: weight species, such s those èhlorlnated compounds present in
the waste feed material.
H.k - Test Methodolog y
The normal gas flow rate through a particulate sampling train
to maintain isokinetic sampling for most sources is 0.5 to 1.0 cfm. A
flow rate of 0.5 to 0.7 cfm was, therefore, chosen for initial experiments
using the EPA Method 5 traIn, which was illustrated in schematic formas
Figure A.5 in Appendix A.
Retention and collection studies
H. 1 e.1.l Absorbents . Initially, tests were conducted to find a suitable
high boiling solvent for use in an impinger system, in order to trap low
molecular weight organochlorIne compounds, e.g. CHCI 3 , CCI , DCE and TCE.
-The following solvents were tested for their suitability as an
impinger solution for use in an EPA stack sampling train: - -
- reagent grade toluene; -
- reagent grade xy lene; and,
- - reagent grade decane. -
All solvents were redistilled prior to use in order to remove
or-limit to usable level: interfering peaks in the GC—EC chromatOgraphic

-------
210
profiles which tended to obscure measurement of the components of
Interest. -
Tests were made to see if trace quantities of volatile
chlorinated compounds could be retained in a solvent such as toluene
with a high gas flow rate passing through the train for a period of four
hours, a time anticipated for tests at SLC. A sampling train was,
therefore, assembled in the laboratory with probe and oven temperature
set above 250°F to warm incoming gas.
Toluene (100 ml) containing known amounts of 1,2-dichloroethane
(EDC) and l,l,2—trichloroethane (TCE) (niàln constituents of the first
waste feed in the waste chloride study at SLC) was placed in !mpinger A’
(see Figure A.5). 100 ml of pure toluene was placed in impinger B,
impinger C was empty and impinger D contained silica gel. Air was
pulled through the sampling train after cooflng the impingers to ice-
water temperature, at a flow rate of 0.5 cfm. After running for four
hours, th’e contents oUeach lmpinger were nated:and analyses made for
EDC and TCE. Ats,pecific tlmes durin each test, the parameters normally
recorded during source sathplingtest , s uchas’ Impinger inlet and oLitiet
temperatures and orifice Ap-, were recorded. At the conclusion of each
test, the total volume of gas sampled was determined.
This type of experiment was repeated, with 100 ml of water in
impinger A, 100 ml of toluene containing EDt and TCE in impinger B,
impinger C empty and impinger D con ainIflg silica gel. Both experiments
were then repeated using xylenè and décàne as absorbents. Additional
experiments were made using decane with known quantities of C d 4 , CHCI 3
and CHC1 2 in the standard Impinger solution.
H. 1 i.l.2 Adsorbents . Though a large number of adsorbents are capable of
adsorbing a wide variety of organic câinpoUnds, Chromosorb 102 was used for
the tests described below because it was readily avaflable, and could be
purchased without delay from various companies supplying chromatographic
materials.
- In order to determine fairly rapidly if Chrdmosorb 102 would
adsorb volatile chlorinated compounds, the following experiment was
performed. 100 ml of decane containing known quantities of Cd 4 , CKC1 3
and TCE was placed in impinger B of the sampling train, and 25 gm of

-------
211
Chromosorb 102 In impinger D. Impinger A contained 200 ml of distilled
water and impinger C was empty. Room air at 0.5 cfm was then pulled
through the system for one hour and at intervals of 10 minutes the decane
solution was analyzed for chlorinated organic concentration. After one
hour the Chrcmosorb was removed from the imp Inger, placed in a septum
sealed container and heated to 100°C. Head—space samples were then
analyzed for the respective chlorinated organic compounds. All three
halocarbons present in the spike had been retained by the resin.
Following this experiment, a series of tests were carried out
using the ORF test duct system, depicted in Figure 11.1. This system
allows dust or gaseous pollutants to be added at a controlled rate into
an air stream flowing through a one—foot diameter duct at flow rates
of 500 to 2500 cfm. Sampling ports are located at ideal and non-ideal
positions in the system, to allow sampling of the air stream with an
EPA type sampling train or other collection equipment. A low flow of air
wàs allowed tb.bubble through CHCl 3 lnamidget impinger and bleed into
the main air f 1ow in the duct. By varying-the carrier flow through the
CHCI 3 , concentratiâns of 10 to 140 ppb of CHCI 3 were obtained in the duct
air stream. These concentrations-were determined both from weight loss
of CHC1 3 with time and direct analysis of the air flowing in the duct.
Tests were made with Chromosorb in impingers C and 0 but considerable
carry over of the adsorbent occurred at flow rates of 0.5 - 0.7 cfm. This
was minimized in further tests by placing a fine mesh screen on top of the
Chromosorb layer in the impinger. At a CHCI 3 concentration of 13 ppb,
breakthrough of this pollutant with 25 gm of Chromosorb in the third
impinger occurred within 30 minutes.
It was’recognised that high volume flow rates through the
adsorbent caused not only contaminant problems with the material in the
sampling train but saturated the bed fairly rapidly. This latter occurrence
was proved by repeating the experiment with the Chromosorb packed tightly
in a collector. A final series of tests were, therefore, completed,
evaluating both Chromosorb and activated-carbon In tube collectors at
much lower air flow rates of 1-2 lItres/minute. The CHC1 3 concentration
in the air-stream was analyzed before and after the collector at selected
time intervals over a period of four hours. For both adsorbents collection
of CHC1 3 was more than 90% efficient for the duration of each test. Since

-------
212
0
j
©
1 DUST FEEDER
2 BLOWER
3 SAMPLING PORTS
4 SAMPLING-PORTS
5 SAMPLING PORTS
6 TO ATMOSPHERE
NON IDEAL LOCATION
IDEAL LOCATION
0
0
)
I
w, V fW YW
TEST
DUCT SCHEMATIC
FIGURE H.1

-------
213
desorption of adsorbed species from Chromosorb was readily accomplished
thermally and since, unlike activated carbon, the adsorption characteris-
tics of Chromosorb are unaffected by moisture in the gas stream, Chromosorb
102 was selected as a suitable adsorbent for the SLC waste chloride program.
H.5 Collection System Evaluation
This section outlines the experimental protocol undertaken to--
verify the performance of Chromosorb 102 as a collection media for the
concentration and analysis (by adsorption/desorption) of low molecular
weight chlorinated hydrocarbons from an air stream under conditions
relevant to field sampling. The concentration and analysis procedure is
essentially an extension of the method used to concentrate organohalogen
compounds present in potable water (H. 1 e].
- Adsorptive capacity values were determined. Adsorptive capacity
is defined as the amount of solute vapour retained by a given weight of
sorbent. -Inthis study, only low levels of chlorinated hydrocarbons
(1- 10 ppb) were examined’since anticipated levels of organohalides in
the sampling streams were expected to be low (I.e. ng/m 3 ).
The sampling cartridges consisted of glass tubes (11 mm i.d. x
various lengths) containing various depths of sorbent bed supported
by plugs of silanized glass wool. The tubes were conditioned by heating
to 200°C and passing a stream of N 2 through them (40 mI/mm) for four
hours. -
Known concentrations of air-solute vapour mixtures were prepared.
The solutes examined were CHC I 3 , CC1 4 and dlchloroethane (DCE). The
solutes were examined both Individually and as mixtures. The air—solute
vapour mixtures were prepared as follows.
Microlitre quantities of organic compounds were injected into
gas jars of known volume. The gas jar was heated to 50°C and, after
cooling, aliquots were removed and added to Tedlar bags. The bags
initially were evacvated and then half filled with air, at which stage
the aliquot from the gas jar was Injected Into the bag. The bag s
finally filled to the desired capacity (6, 20 and 40 litres).

-------
214
A schematic diagram of the ‘component parts of the test
assembly is shown below.
TEDLAR ___
- The various parts of the assemb1ywere connected to each other
by means of Teflon tubing. Differing air-solute mixture concentrations of
CKC1 3 were used to determine the. adsorptive capacity of the Chromosorb
102. The CHCI 3 concentrations used were 5 ppb and 10 ppb. A flow rate
of 250 ml/min was maintained throughout the run. For these conditions
the collection efficiency of Chromosorb 102 for CHC1 3 was 100 percent.
At various time intervals the pump was stopped and the gas jar disconnected.
Aliquots (1 rn!) were removed from the gas jar and analyzed for CHC1 3 by
gas chromatography (cc). The first detection of CHCI 3 in the gas jar
was taken as the ste ye in the test for the determination of breakthrough
volume i . e. the volume of a.i necessary to purge the adsorbed vapour
through the cartridge.
From the breakthrough vOlume of the air-solute mixture of known
concentration and the weight of sorbent in the cartridge it was possible
to calculate the adsorptive capacity of Chromosorb 102 for CHC1 3 . -‘der the
given experimental conditions.
From the data determined with, respect to the adsorptive capacity
of Chromosorb 102 for CHC1 3 and the knowledge that a four hour sampling
period at a flow rate of 250 ml/miri would be required in the field tests,

-------
215
sampling cartridges of the following dimensions were constructed: 11 mm
i.d. x 11.5 cm long containing 6 g of Chromosorb 102. These cartridges
were used to determine desorption efficiency data.
quantity CHC1, from Chromosorb 102
Desorption efficiency —
CHC1 concentration Volume of air
in ir sampled X sampled
The same test assembly as described previously was used to collect
CHC1 3 on the cartridge. The adsorbed CHC1 3 was thermally desorbed into
an evacuated gas jar. The adsorbent tube and the gas Jar were connected by
Teflon tubing. The tube was heated to a fixed temperature by wrapping
with heating tape controlled by a variable transformer (170°C). After
reaching maximum temperature, the stopcock of the gas jar between the jar
and the tube was opened. Heating ‘of the sample tube was continued for a
further 15 minutes, after which time the stopcock of the gas jar was
closed,’ the gas Jar removed from the-sample-tube, taken outside the
laboratory and allowed to fill up with the cleanest possible air. The
contents of’ the gas jar’ were then subjected to GC analysis. Aliquots
(1 ml) were removed from the gas jar using an air—tight syringe and
injected into a gas chromatograph.
Gas chromatography was conducted on a Varian 1200 series
chromatograph equipped with an electron capture detector. The GC
parameters used are shown below.
Column - 2 m x 2 mm SS containing Chromosorb
- 102 (80/100 mesh)
Column temperature - 180°C
Injector temperature - 215°C
Detector temperature - 215°C
Detector — EC
Flow rate - N 2 “a 31 mi/mm
Range and attenuation - as required
Chart speed - as required
- The concentration of solutes present in the gas jars following
desorption was determined by measuring peak areas of the vapours and
comparing them with those of prepared standards. The initial concentration

-------
216
In the Tedlar bag was known, and, from the volume of air drawn through
the cartridge, the actual concentration contacting the cartridge was
calculated. The desorptlon efficiency was then calculated.
H.6 Results and Discussion
Experiments using toluene and .xylene as absorbents showed that a
considerable loss of both solvent- and-chlorinated organic compounds Occurred
after a four hour test. The solvent was found in all Impingers following
the ãne In which the standard solution was placed. Estimated solvent
losses were 40-60%. Losses of CC1 , CHCI 2 and TCE ranged from 50 to 90%.
With decane as the adsorbent, the solvent -loss was less than 10%.
However, the loss of chlorinated compounds was 60 to 100%. Due to time
restraints and the obvious problem of solvent or pollutant loss at the
high gas flow rate, this method of collection using absorbents was rejected.
As discussed in Section H.4.l.2, retention of volatile halocarbons
by the adsorbent at-high gas-volume flow:rates was poor-and the break-
through rapid. The results for the collection system finally selected
(Section H.5) are sunvnarlzed below.
Adsorptive Capacity.
Component il/g (sorbent )
CHC1 3 0.1
Desorptlon Efficiency
Tedlar bag concentration Desorptlon Efficiency
of CHC1 3 .
1.25 ppb 90
- 96
104
5.0 ppb 92
110
10.0 ppb 94
102
(The higher concentration samples required a second thermal desorpt ion
in order to remove all the solute from the sorbent. However, the first
desorption was always better than 75%.)

-------
217
Two additional experiments were performed. They were:
a) An air-solute mixture of CC1 4 and DCE (5.00 ppb level;
40 2. Tedlar bag) was run through the system using a back-up
gas jar in order to test whether these solutes were retained
by the Chromosorb 102. Regular checks of the gas jar using
CC to analyze for the presence of CC1 4 and/or DCE failed to
show any trace. Thermal desorptlon of the tube followed by
CC analysis gave good recoveries for CC1 4 and DCE (>90%).
b) An air—solute mixture of CHC1 3 , CC1 4 and DCE (2.5 ppb level;
40 & Tedlar bag) was prepared. This mixture was drawn
through the test system which contained two additional traps
prior to the Chromosorb 102 cartridge. One trap contained
water (50 ml) and the other 5% NaOH solution (80 ml). These
traps were Included in the ORF train in order to remove
-residual Cl 2 and/orHCI. Thermal desorptIon of the tube
contents gave good recoveries of the solutes (>90%).
H.7 - Conclusions ’ -
The laboratory study undertaken showed that low molecular weight
halocarbons cannot be efficiently collected by absorbents or adsorbents
from a gas stream at high flow rates (“i 0.5 cfm).
There are numerous literature references to the technique for
collecting trace quantities of component vapours by drawing air through
cartridges containing sorbent media such as polymer beads, activated carbons
or stationary liquid phases chemically bonded to solid supports [ H.5].
This report has demonstrated such a procedure for the collection
of low molecular weight chlorinated hydrocarbons on Chromosorb 102. It
has been shown to give quantitative collection together with quantitative
recovery of the trapped species using a thermal desorption technique.
The sampling tube technique appears to give better results for
sampl ing traces of low molecular weight organochioririe pollutants in air
than impinger methods.

-------
218.
REFERENCES
H. Marine Environmental Monitoring of Vulcanus Research Burn II,
December 2-10, 1974, Preliminary report, U.S. Environmental
Protection Agency, December 10, 1974.
H.2 Pellizarri, E.D. , Bunch, J.E., and Carpenter, B.H., Env.. Sci.
Technoloqy .2. 552 (1975).
H.3 Jones, P.W., Glamar, R.D., Strup, P.E., and Stanford, T.B., Env.
ScI. Tech . 10608 (1976).
H.4 Bellar, l.A., and Llchtenberg, J.J., J.M..Water Works Assoc. ,
66739 (1974).
H.5 Pellizarri, E.D., EPA-600/2-75 076 Nov. 1975.

-------
APPENDIX I
CC/MS/COMPUTER DETERMINATION OF CHLORINATED HYDROCARBONS AND PCB’S
Chemistry Division
Air Pollution Control Directorate
Environmental Protection Service
Envi ronrnent Canada
Submitted by
Dr. R.C. Lao

-------
221
APPENDIX I
CC/MS/COMPUTER DETERIIINATIOA OF CHLORINATED HYDROCARBONS AND PCB’s
In February 1970, 30 extracted samples were received from
Dr. C. Thomas of the Ontario Research Foundation. These samples were
taken from the burning of chlorinated hydrocarbon wastes in a cement kiln..
Requests were made to determine the nature and quantity of any heavy
chlorinated hydrocarbons, In particular PCB’s, present in the samples.
Experiments were completed and the procedures are briefly
described as follows:
Material : All apparatus and reagents used in the experiment were examined
by gas chromatograph-f lame ionization detector (GC—FID) analysis of a pure
hexane extract to ensure freedom from organochlorine residue contamination.
The chlorinated isomers (PCB’s) and Aroclor mixtures were obtained from
Anal-abs, North Haven, Conn., U.S.A.
GC-FID : A Perkin-Elmer 990 model GC-FID with a datasystem PGP—l was used.
The procedures for standard calibrations were given in a previous paper
(to be published as a chapter in the book “Advances in Dynamic Mass
Spectrometry” 1976). Chromatograph operating parameters were as follows:
Column 12’ x 1/8” 0.0. stainless steel
Column packing 6% Aplezon L on chromosorb U 80/100 mesh
Column temperature Initial lOO 0 C programmed at 4°C/mm.
to final temp. 200°C and hold
injection temperature 250°C
Manifold temperature 250°C
Carrier gas Helium 40 ml/mln.
CC/MS/Computer : A Finnigan 1015 0CC/MS system,was. used with a data system
6000 series. its performance has been studied and documented in the same
paperas mentioned above. It also includes standard PCB mass spectra and
computed reconstructed Aroclor chromatograms. The instrumental data
are as follows:
PrecediNg page blank

-------
222
Finnigan lOl5D CC-MS Instrumental Data
A
Instrumental Data GLC
Column 12’ x *“ all glass
Column packing 6% Aplezon L 80/lao
mesh chràmosorb W
Column temperature 225°C -
Injection temperature 250°C
Carrier gas ‘helium 40 mi/mm.
Sample size 5 to 7 mlcrolitres per injection
B
instrumental Data Ms
Filament current 100 mlcroamperes
Electron energy 70 or 20 eV
Operating pressure 6.68 x 10 N/rn 2
(5x io 6 torr)
Scanning speed 4 seconds
Standard deviation of
spectra maxImum 570
After the quadruple MS operating parameters were adjusted the
sample was Injected Into the CC. The Ion abundance chromatogram of the
CC effluent was acquired by scanning. the mass range (40 to 400). The
dialogue required for mass spectrometer control, data acquisition and
obtaining the plot are given by software programs. At the end of the
CC run the computer plots a reconstructed gas chromatogram (ion abundance
chromatogram) of total ion amplitude versus the spectrum number.
Identification of these chramatographic peaks can be accomplished by
plotting the mass spectrum of a specified peak or by a limited mass
plot chrornatogram which is obtained under computer control by searching
through the collected spectra and identifying spectra containing ions
with a specific rn/e value.

-------
223
Result and Discussion
(All chromatograms and computer reconstructed chromatograms are
kept on file at the Air Pollution Technology Centre, Department of the
Environment, Ottawa, Canada.)
Computer reconstructed gas chromatograms of Aroclor l2 e2, l25 i,
and 1260 were made. By focusing on specific mass/charge ratio (m/e)
peaks such as 290 (tetra-chloro biphenyls), 32L, (penta-chioro) or 358
(hexa—chloro), an ion current plot for a particular PCB isomer versus
the Aroclor spectrum is obtained.
Gas chromatograms on Perkin Elmer 990 GC were done for some
blank solutions and concentrated samples (all samples were concentrated
from about 25 ml to 1 ml). No chlorinated hydrocarbons nor PCS’s were
found in the samples. If they are present the concentrations are below
the sensitivity limit of GC—FID of I ng or less.
Mass spectra for the GC peaks of spectrum numbers 365 and L.8 7
from the reconstructed chromatogram- plot of sample (WBC-T3, EPA train,
Impinger) reveals that only hydrocarbons are present in the sample. No
evidence for chlorinated compounds was found. A limit mass searching
technique has failed to give any indication of the presence of PCB’s.
The computer library search of standard Aroclors did not match those
obtained from the sample.
It was concluded, therefore, that there were no heavy chlorinated
hydrocarbons or PCB’s in the samples. If they were present the weights were
less than I ng.
aa 1545
SW—i 47c

-------