PrepubliaatJon issue for EPA
         and State Solid Waste Management Agenoiee
          BURNING WASTE CHLORINATED HYDROCARBONS

                     IN A CEMENT .KILN
      This report (SU-l4?a) describee work performed
for the Office of Solid waste under contract no. 68-01-2966
     and ie reproduced OB received from the contractor.
    The findings should be attributed to the contractor
           and not to the Office of Solid Waste.
     The reader ie advised to utilize the information
        and data herein wiih 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 publication  (SW-l47c) in the sol^d waste
management series.

-------
          Propublioation issue for EPA librcari.«s
         and State Solid Waste Manag0m*nt Ag»nai»8
          BURNING WASTE CHLORINATED HYDROCARBONS

                     IN A CEMENT .KILN
      This report (SW-l47a) describee work performed
for -the Office of Solid waste wider contract no. 68-01-2966
     and is reproduced as received from the contractor.
    The findings should be attributed to the contractor
           and not to the Office of Solid Waste.
     The reader is advised to utilize the information
        and data herein wii;h 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 publication (SW-W7c) in the sol^d waste
management series.

-------
                              ABSTRACT

          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, were 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 dichloride, chlorotoluene and up to approxima-
tely 50 percent polychlorinated biphenyls (PCS).
          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 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.
          The chlorine  input from the chlorinated hydrocarbon waste was up
to about 0.8 weight percent relative  to clinker and  this effectively reduced
the alkali  concentration of the  clinker  in  direct stoichiometric proportion,
A  reduction  in  fossil  fuels used while burning  chlorinated  hydrocarbons was
noted.
                                111

-------
                           TABLE OF  CONTENTS
                                                                     Page

 ABSTRACT                                                                '
.TABLE OF CONTENTS                                                     Mi
           List of Figures                                               v
           List of Tables                                              vii
 EXECUTIVE OUTLINE                                                      xi
 SUMMARY                                                              xi i i
 RECOMMENDATIONS                                                        xv
 LIST OF ABBREVIATIONS                                                 xvi
 1.         INTRODUCTION                                                 '
 2.         CEMENT MANUFACTURE                                           2
 2.1        General Principles                                           2
 2.2        Effect of  Alkalies                                           5
 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                                                  ]5
 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.k       Mass Balance on Wet Kiln                                     30
  5.4.1     Significance of the mass  balance                            30
                               Vi

-------
                         TABLE  OF  CONTENTS  (CONT'D)

                                                                    Page

5-^.2      Chlorine and  potassium  retained                             31

6.         CONSIDERATIONS ON BURNING  CHLORINATED HYDROCARBON
           WASTES  IN  A CEMENT KILN                                    3k

6.1        Effect  on  Production                                       34

6.2        Alkali  Reduction  While  Burning Chlorinated Hydrocarbon
           Wastes                                                      34

6.3        Heat  Recovery from Chlorinated Hydrocarbon Wastes           37

6.k        Cement  Quality                                             38

6.5        Extrapolation to  Other  Kiln  Types                           38

6.6        Comparison of Cement Kiln  Burning with Other Uses and
           Disposal Methods  for Waste Chlorinated Hydrocarbons         39

7.         CONCLUSIONS                                                42

REFERENCES                                                           43

ACKNOWLEDGEMENTS                                                      46

APPENDIX A -  QUANTIFYING, SAMPLING AND ANALYSIS OF PROCESS
              MATERIALS                                                49

APPENDIX B -  ANALYTICAL DATA,  CALCULATION  AND DETAILS OF
              EXPERIMENT ON  THE SUSPENSION  PREHEATER KILN              93

APPENDIX C -  RESULTS AND CALCULATIONS  FOR  WET PROCESS SYSTEM         107

APPENDIX D -  QUALITY OF CEMENT PRODUCED                             129

APPENDIX E -  EQUIPMENT  DESCRIPTION AND ASSOCIATED ECONOMICS          137

APPENDIX F -  ONTARIO MINISTRY  OF  THE ENVIRONMENT EMISSION
              GUIDELINES  AND ANALYTICAL SUPPORT                      149

APPENDIX G -  LABORATORY  ANALYSIS  RESULTS  FROM THE ST. LAWRENCE
              CEMENT  FACILITY TEST (TRW Systems Group)                165

APPENDIX H  -  DEVELOPMENT, CONSTRUCTION AND EVALUATION OF A
              COLLECTION  SYSTEM FOR LOW MOLECULAR WEIGHT
              HYDROCARBONS (Ontario Research  Foundation)              207

APPENDIX I   -  GC/MS/COMPUTER DETERMINATION  OF  CHLORINATED
              HYDROCARBONS AND  PCB's  (Air Pollution Control
              Directorate, EPS,  Environment Canada)                   221

-------
                            LIST OF FIGURES

Figure                                                               Page

   1       Wet Process Kiln                                              3

   2      Dry Process Kiln                                              k

   3      Principle of Fuller-Humboldt Suspension Preheater and
          By-pass                                                       7

   k      Alkali By-pass                                                9

  A.I     Schematic of the Material Balance                            50

  A.2     Schematic of Port Locations                                  59

  A.3     Gas Flow Distribution at Sampling Points                     60

  A.k     Grab Bag Sampling Equipment                                  61

  A. 5     Particulate Sampling Train                                   62

  A.6     Gaseous Sampling Train                                       65

  A.7     Gas Chromatographic Profile  from Flame  lonization
          Detector for Chlorinated Aliphatics  (WBA)  Sample Feed        73

  A.8     Gas Chromatographic Profile  from Flame  lonization
          Detector for Chlorinated Aliphatics  plus  Aromatics
          and Alicyclics  (WBB) Sample  Feed                             /A

  A.9     Gas Chromatographic Profile  from Flame  lonization
          Detector for Chlorinated Aliphatics  plus  Aromatics,
          Alicyclics and  Polychlorinated  Biphenyls  (WBC)  Sample
          Feed                                                         75

  A. 10    Gas Chromatographic Profile  from Electron Capture
          Detector for WBB Sample  Feed                                76

  A.11    Gas Chromatographic Profiles 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. 13    Gas Chromatographic Profiles from  Flame lonization
          Detector for Low Molecular Weight  Chlorinated
          Hydrocarbons                                                 80

  A.\k    Gas Chromatographic Profile  from Electron Capture
          Detector for Impinger  Extract from BLA  Test 3                81
                             Vi

-------
                        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.I     Chlorine Level  in Stage  IV,  June 3,  1975                    102

  B.2     Chlorine Level  in Stage  IV,  June 10,  1975                   103

  E.I     Schematic  Diagram of  Basic Chlorinated Hydrocarbon
          Feed  System                                                 HO

  E.2     Schematic  Diagram of  Chlorinated Hydrocarbon Facilities     \k\

  F.I     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 GC/HS Analysis of Sample
          PCB Fuel                                                    159

  G.I     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.I     ORF Test Duct Schematic                                    212
                               V11

-------
                            LIST OF TABLES

Table

  1       Composition of AUphatics (WBA)                                19

  2       Composition of Aromatics plus Complex  (WBB)                    20

  3       Composition of Aromatics plus PCB's  (WBC)                      21

  k       Gas Sample Volumes and Sample Concentration  Factors            2k

  5       Estimated  Kiln Emission Concentrations  (GC-EC)  for
          Specific Volatile Organochlorine  Compounds                     25

  6       Estimated  Minimum Combustion  Efficiencies  for  Each
          Waste  Burn                                                    2&

  7       Summary of Particulate Test  Data                               29

  8       Accumulated Mass  Balance  for  Chlorine                          32

  9       Accumulated Mass  Balance  for  K.O                               32

  10       Average  Reduction in  K-0  Content  of Clinker                   35

  11       Average  Dust  Discharged                                        36

  12       Recovery  of  Btu  from  Chlorinated  Hydrocarbons                 38

 A.I       Process  Materials Studied and Approximate Normal
          Production Quantities                                         ^9

 A.2       Quantities of Aliphatic Mixture Burned Daily                  51*

 A.3       Quantities of Aromatic plus Complex Mixtures Burned
           Daily                                                         55

 A.**        Quantities of PCS 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.10      Results from Atomic Absorption Analyses                       87

 A.11      Least Squares Data for Calibration  Lines                      88


                             Vi 11

-------
                         LIST OF TABLES  (CONT'D)

Table                                                                Page

B.I       Percent Bypass Gas Required  to Maintain Chloride Levels      95

8.2       Results from Analyses of  Dry Process  Kiln Raw Feeds          98

B.3       Results from Analyses of  Dry Process  Kiln Clinker            99

B.l»       Results from Analyses of  Stage IV  Dusts                     100

C.I       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.k       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  Mo.  6 Fuel Oil                  118

C-7       Daily Record of  Production and Materials  Consumption        120

C.8       Material  Balance for Chlorine                              122

C.9       Material  Balance for 1^0                                    125

D.I       Cements from 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 Polychlorinated
          Biphenyl  Burn                                                133

F.I       MOE  Specifications Applied to Waste Burns                   150

F.2       St.  Lawrence Cement Waste Burn Experiments - Test 1
          Chromosorb  Adsorption Tube Analysis                         153

F.3       St.  Lawrence Cement Waste Burn Experiments - Test 2
          Chromosorb  Adsorption Analysis                              153

F.I*       St.  Lawrence Cement Waste Burn Experiments - Test 3
          Chromosorb  Adsorption Analysis                              153

F.5       Fuel  Sample Identification                                  155

F.6       Gas  Chromatograph Conditions                                155

                                  ix

-------
                        LIST OF TABLES (CONT'D)

Table                                                                 Page

F.7       MS Data from GC Analysis of Sample PCB Fuel                  160

F. 8       Area Counts                                                  161

G.I       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 Desorbed from Sorbent Tube  Samples                  189

G.9       Results and Detection Limits from GC/FID Analysis of  the
          Concentrated Extracts                                       190

G.10     Results and Detection Limits from GC/ECD Analysis of  the
          Concentrated Extracts                                       192

G.ll      Results and Detection Limits from GC/ECD Analysis of  the
          Unconcentrated  Extracts                                      19^

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 AAS                                                 197

G.15     Concentration  of Trace Metals  in Aqueous  Samples by AAS       197

G.I6      Results of Organic Survey Analysis  on Clinker  Product anH
           Disposal  Dust  Samples                                        199

G.17      Results and  Detection Limits  from GC/FID  Analysis           201

G.18     Results and  Detection Limits  from GC/ECD  Analysis           201

-------
                         UST OF TABLES  (CONT'D)

Table

G.19      Selected Trace Metals  In  SLC  Clinker  Product and
          Discard Dust  Samples by SSMS                                202
                                X1

-------
                          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 and 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  their thermal  destruction.
          A  long high-temperature flame 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.   However,  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  r.^ans of a
                               Xi i

-------
single economical recovery system which would collect and later use the
wastes as supplementary fuel for cement manufacture.
          The experiments described  In this report were monitored by
the Department of Fisheries and Environment, the Ontario Ministry of the
Environment, the United States Environmental Protection Agency, and the
Ontario Research Foundation.
          For these experiments, industrial chlorinated hydrocarbon wastes
including polychlorinated 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.
                              X111

-------
                                SUMMARY

          Chlorinated hydrocarbon wastes were burned in a carefully con-
trolled 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 effects on 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 mainly of 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
specially for determining emissions of organic material.  All samples
from both systems were analyzed for unburned chlorinated 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  compounds 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 carried  out  on chlorine 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.
                               xiv

-------
           It MAS concluded  that chlorinated hydrocarbon vtastes
used in cement kilns, 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

-------
                               RECOMMENDATIONS

          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 polychlorInated 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 recommendations 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  in a  particular kiln
      installation can  be determined  by  a 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           Dichloromethane

EC            Electron capture  (detector)

EPA           Environmental Protection Agency (U.S.)

EPS           Environmental Protection Service  (Fisheries and Environment
              Canada)

FID           Flame  ionization  detector

GC            Gas  chromatography

MOE            (Ontario)  Ministry of the  Environment

MS            Mass spectrometry

ORF           Ontario  Research  Foundation

PCB           Poiychlorinated btphenyl

SLC           St.  Lawrence  Cement Co.

TRW           TRW  Systems Group

Tl,  T2,  T3   Test one,  Test  two, Test three

WBA           Waste Burn A  (chlorinated aliphatlcs)

WBB           Waste Burn B  (WBA plus chlorinated aromatics and alicyclics)

WBC           Waste Burn C  (WBB plus poly chlorinated biphenyls)

XRF           X-ray fluorescence
                              XVH

-------
I          INTRODUCTION
          In Canada  each year  an estimated 2J-30 million pounds of chlori-
nated hydrocarbon wastes require disposal  or destruction, 17-20 million
pounds being generated  in Ontario [:].   These figures are based on 5-6fc
of annual  production and may be conservative.  Experience in Europe
indicates  that 10% of production 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 polychlorinated 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-
tion.
           In cement  manufacture, the kiln operates at higher  temperatures
and  for longer  residence times  than  those used  in  incinerators  for des-
truction of these waste  materials.   It  is also  common 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 without 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          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 gradually move into the burning zone.  Reactions
 which  occur during gradual heating in 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 H»50°C (2650°F).   Four main compounds are present
 in Portland cement clinker:
                                                     Common Abbreviations
                                                      Used  In The
 Name Of Compound               Chemical  Formula     Cement  Industry	
 Tricalcium Silicate            3 CaO*SiO_               C,S
 Dicalcium  Silicate             2 CaO-SiO.               C?S
Tricalcium  Aluminate            3 CaO'Al.O.              C.A
Tetracalcium Aluminoferri te     4 CaO-Al-OyFe 0.        C.AF

          Minor  compounds  are also formed In  clinker, commonly magnesia
 (MgO),  potassium sulfate  (K-SO.)  and sodium sulfate (Na SO.).
          Traces of other  elements present in either the raw materials or
fuel are also found  in  clinker.   Upon leaving the kiln, the clinker is

-------
www
L*
      VX^y,'"'.^ "'^' •_/-
                                                                YYY
                                                    •TT
                   1-KILN
                   2-SLURRY FEED
                   4-PRECIPITATOR DUST SCREW
                   3-PRECIPITATOR
                   5-DUST RETURN
6- FUEL
7- CLINKER COOLER
8- CLINKER
9- FILTER
                             WET  PROCESS  KILN
                                  FIGURE 1.

-------
        YYYr
        •-   -   i
1  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
DRY  PROCESS KILN

-------
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-  is driven from the raw
meal.  Any elements not driven off are  increased  in the clinker in propor-
tion to the quantity of CO- 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  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 (K20 and Na20).  Both  alkalies  are frequently combined and
reported as equivalent Na-0 for  purposes of specification.  The raw materials
at St. Lawrence Cement are such  that Na«0  is  low and practically constant
(see Table A.10).  For this reason, only potassium oxide  (K.O) is consi-
dered in detail in this report.
          The effect of alkalies on cement quality  has  been well documented
[2, 3, *0.  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 a 1ka11-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.O performs  satis-
factorily with such aggregates.  This specification  is  imposed with  suffi-
cient frequency  in ths United States that  it  is  found as an optional  spe-
cification  under ASTM C-150  [?].  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 sulphates  which
at kiln operating temperatures are not  readily volatilized, but are  retained
in the clinker.  Alkali chlorides are volatile at normal kiln operating

-------
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  1,750,000 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 dumbel 1 -shaped All is
Chalmers  kilns J»02'  long with a diameter of  11'6", 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 8?'  of kiln  length.  The slurry feed  system  is a bucket wheel con-
veyor with a variable speed  drive taking  slurry from a constant  level box.
Gases from each  kiln  (maximum capacity  150,000 CFM at  450 F)  pass through
a six-section electrostatic  precipitator.   Gases  from  the precipitators
are exhausted via a  common  stack 55^'  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  1?" 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  counterflow 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

-------
                      MEAL FEED
                        II
                        III
                        IV
     2 STAGE
     3 STAGE
      STAGE
     5 STAGE
      KILN
      KILN EXH.FAN
     8 PRECIPITATOR
     9 DUST  RETURN
    10 CONDITIONING TOWER
    11 BY-PASC PRECIPITATOR
     12  DUST  DISPOSAL
     13  DUST  DISPOSAL
     U  WATER

       MATERIAL FLOW
           	[>  GAS  FLOW
PRINCIPLE  OF  FULLER-HUMBOLDT
   SUSPENSION   PREHEATER
     AND
BY- PASS
FIGURE 3.

-------
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  the stream entering the next lowest stage,
is repeated  in  Stages  III 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  3*»0°C
 (650°F).   At each stage,  corresponding heat exhanges occur  such  that  the
material enters the rotary  kiln at approximately  800°C  (1*475°F)  having
been partially  decarbonated.  The gas temperature at  the point of exit
from the  kiln  into the preheater is 10l»0°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.
           The alkali bypass system (Figure 4) fs  an important unit
 in  relation  to  this study.   In common 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 sprayed 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 pr«cip!tator, pel-
let ized and  discarded.

-------
  I
 c
  1
  I	
          T71
         T
                  T70
                       T73
ALKALI BY-PASS
    FIGURE 4
PREHEATER TOWER BY-PASS HOOD   7
QUENCH AIR                     8
CONDITIONING TOWER              9
PREOPfTATOR                   C
EXHAUST FAN                   PT
WATER                         M
AIR
650°F CAT OFF CONTROL
AIR QUENCH NOZZLE CONTROL (4®>°F)
CONTROLLER
PRESSURE TRANSMITTER
MODULATOR

-------
                                    10
3         CONSIDERATIONS AT  THE  PROGRAM  PLANNING STAGE
          The  history of suspension  preheater  kilns  In the United States
indicates a  trend  to this  system [!*»]•   Thirteen preheater kiins 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 kilns
were installed, one of  which has since been  shut down.  However, with
increasing fuel costs and  emphasis on  energy conservation, 22 suspension
preheater kilns have been  commissioned since 1970.  Another  indication
of this  trend  is that five of  the eleven new kilns Installed  in
1975 were preheater-rotary kilns [15].   In addition, three existing
rotary kilns were  converted  to suspension preheater  units.  All five
kilns  planned  for  completion in  1976 will be preheater kilns.  While the
situation in Canada is  similar,  there  are fewer plants and such trends
are  difficult  to follow.
          Any  study concerned  with future use of the technology developed
must take into consideration the suspension  preheater kiln.   Flame
characteristics are identical  In all cement  kiln systems, hence,
demonstration  that destruction of chlorinated  hydrocarbons occurs  in
one  cement kiln implies destruction  in all cement  kilns.  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 plugging
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 in  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:

-------
                                    1]
          I)  aliphatics;
          2)  aromatics;
          3)  aliphatics plus aromatics and alicyclics; and,
          k)  aliphatics, aromatics, alicyclics and large complex
              molecules such as polychlorinated biphyenyls (PCB's).
          Generally these are filtered, processed and blended to obtain
specific formulations for control of energy value, chlorine content and
viscosity.

-------
                                    12
 k          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
 sulphates 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  mixing of 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 of alkali  chlorides,
 sufficient quantities condense in Stages III  and IV of  the preheater to
 cause plugging of this system.   To alleviate this situation,  the bypass
 is 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 alkal!  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 conditioning
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
participate 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 June 10 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 rebuiIt.
          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.A).  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 K_0 on  ignited basis averaged
      , while the average clinker K-0 was 1.27%.  With the same quantity

-------
of gases being removed by the bypass, but without addition of any chlorinated
material, the raw meal feed contained on average l.Mfc K_0 and the
clinker contained on average 1.291 K.O.  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.

-------
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 C02> H20, HC1 (hydrochloric
acid), Cl"  (chloride  ion) and a negligible amount of free C12 (chlorine
gas).

          H2° + C12   *   2HC1  + i02
           Kp -
                         
              = 15 at 1000°C (1832°F)
              • 68 at 1500°C (2732°F)
              -150 at 1500°C (3*52°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 solids.  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
 eir 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  C12 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  compensate for the  changing chloride content,
the rates being equivalent to:
October  28,  1975, 0.31* chlorine relative  to  clinker;
October  29,  1975, 0.32* chlorine relative  to  clinker; and
October  30,  1975, 0.63* chlorine relative  to  clinker.
           Due to  excessive quantities of chlorine  input, a  ring was
formed,  which required  a kiln shutdown for its  removal.  A  ring Is
caused by the buildup of material on the Inside of  the kiln to such
an  extent that it restricts the  flow of  process materials.
                                                                  y
           Due to  production scheduling,  the delay caused by the shutdown
made  it  necessary to drop  the burning of aromatics  alone and  immediately
progress to the mixture of aromatic and  complex molecules.  The conduc-
tivity of this latter material was  too low (0.3 x  10   mhos)  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 baing:
December  10,  1975, 0.^5  to 0.71* chlorine  relative  to clinker;
December  11,  1975, 0.31  to 0.51* 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 to add 13,000 gallons of the
polychlorinated 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 1iquid flows.
          For the  last day of emission  testing,  the nozzle was  removed
completely  to permit uninterrupted  flow rates of polychlorinated biphenyls.
Quantities  added during 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 V75
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 tetrachloride (CO.), chloroform
(CHClJ, dichloromethane  (CH.C1.) 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 organochlorldes referred 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 ion i ration
detector (FID).  While this assumption  is not 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-mass spectrometry (GC-MS).   Identi-
fication of components at concentrations less than one percent was not
routinely performed.  Analyses of samples 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 WBC) 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 and G, 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. and
          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 organochlorine 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 (WBA)
                   Sample Feed - October 28, 1975
Peak # in GC Profile
(cf Figure A. 6)
1
2

3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
2k
25
26
27
Approximate
Concentration
% Identification
17.4
!.4

0.8 |
1.5
3-2
Chloropropane & propene
ethyl chloride,
dichloromethane

Chlorobutane & butene

16.5 1» 2, Dichloroethane
6.6 Carbon tetrachloride
1.0
10.5 1,1, 2-Trichloroethane
0.7
0.3
1.9
1.4
1.4

Dichloropropanes

2.4 Tetrachloroethylene
7.3 Tetrachloroethane
15.6 Chlorobenzene
2.4 -v
0.2
3.3
0.3
1.1
0.6
0.5
1.6
0.6


mul ti chlorinated
butanes, butenes
hexanes , hexenes



0.7 '
Note:   No identification of compounds at concentrations of \% or less was
       attempted.

-------
                                   20
         TABLE  2.   COMPOSITION OF  AROMATICS  PLUS  COMPLEX  (WBB)
                    Sample Feed -  December  12,  1975
Approximate
Peak # In GC Profile Concentration
(cf Figure A.7) %
1 2.0
2 0.1
3 0.1
If 1.4
5 1.5
6 3.5
7 0.4
8 I.I
9 1-9
10 0.2
11 0.2
12 0.4
13 0.6
14 0.5
15 0.5
16 52.2
17 0.2
18 0.1
19 0.1
20 6.3
21 2.7
22 0.1
23 1.0
24 8.5
25 4.4
26 6.9
27 1-3
Identification
Chloropropane
-
-
Chloroform
1, 2-Dlchloroethane
Carbon tetrachloride
-
-
1, 1, 2-Trichloroethane
-
-
-
-
Chlorobenzene
-
Chloro toluene
-
-
-
Hexach 1 orocyc 1 open tad i ene
Heptachlorocyclopentene
•
Pentachlorobenzene
Oc tach 1 orocyc 1 opentene

Complex associated with
Hexach 1 orocyc 1 open t ad i ene
Note:  No identification of compounds at concentrations of 1% or less was
       attempted.

-------
                                  21
         TABLE 3.  COMPOSITION OF AROMATICS PLUS PCB's (WBC)
                    Sample Feed - January 8, 1976
Peak # In GC Profile
(cf Figure A. 8)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
2k
25
26
27
28
29
30
31
Approximate
Concentration
% Identification
3.4 Chloropropane
0.3
1.4 1, 1 ,-Dichloroethane
1.3 Carbon tetrachloride
0.5
0.3
1.9 1,1, 2-Trichloroethane
0.1
1.9
0.5
28.5 Chlorotoluene
< 0.1
< 0.1
1.8
0.9
2.8
0.9
2.2
0.3
2.1
0.9
6.7
3.*
5.9
1.6
12.1
5.6
4.0
5.4
2.1
1.6

Hexachlorocycl open tad iene
+ complex


Dichlorobiphenyl



Trichlorobiphenyl


Tetrachlorobipnenyl



Pentachlorobiphenyl

Note:   No identification of compounds at concentrations of 1% or less was
       attempted.

-------
                                   22
waste burns, and different chromatographIc columns 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.  VBB samples
were  found to  consist primarily of chlorinated aromatic compounds, in
particular o-chlorotoluene.  WBC samples were found to consist of approx-
imately  50% polychlorinated  biphenyls (Aroclor  12
-------
                                    23
to collect higher molecular wefght compounds, Including any noneombusted
starting materials in the impingers.
          Components adsorbed on the sorbent were removed by thermal
desorption 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
             detection 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 k.  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 cf 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  (DCM)  was  tentatively  identified as the  major  component
of  these  desorbed gas  samples.   Others  tentatively  identified  were CHCK

-------
    TABLE 4.   GAS  SAMPLE VOLUMES AND SAMPLE CONCENTRATION  FACTORS
                       (Gaseous Sampling Train)
Test #
1 BLA
2 BLA
3 BLA
1 WBA
2 WBA
3 WBA
1 WBB
2 WBB
3 WBB
1 WBC
2 WBC
3 WBC
1 BLB
2 BLB
3 BLB
Date
Oct.
Oct.
Oct.
Oct.
Oct.
Oct.
Dec.
Dec.
Dec.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
20 (1975)
21
22
28
29
30
10
11
12
7 (1976)
8
9
19
20
21
Test
Duration
(mins)
317
281
320
362
286
285
260
317
250
268
237
228
170
247
240
Volume*
Sampled
scf
2.
1.
2.
3.
2.
2.
2.
2.
2.
2.
2.
2.
1.
2.
2.
80
86
83
22
53
53
30
80
21
37
09
02
50
18
12
Sample Concentration
Factors
Desorbed Solvent**
Gas Extracted
159
Hi
160
181
1*3
143
130
159
125 .
134
119
114
85
124
120

510


500-1100


600-2000


520-1500


380

 -Volume sampled per adsorbent tube.
-'-These values are based on the  lowest and highest molecular weights of
  compounds found in the waste1 feeds.  Figures given for both baseline
  series are based upon the molecular weight of dichloromethane (DCM).

-------
                                   25
       TABLE  5.   ESTIMATED KILN EMISSION CfMlCENTftATllikltt
                 FOR  SPECIFIC  VOLATILE ORGANOCHLORINE COMPOUNDS
Emission Concentrations
Test Series Dichloromethane
ppb yg/m*
BLA
WBA
WBB
WBC
BLB
4.1
5.4
18.0
7.7
29.0
14.5
27.3
19.1
102.7
63.7
Chloroform
ppb pg/m'
0.004
0.015
0.038
0.069
0.018
0.020
0.080
0.190
0.345
0.090
Carbon
Tetrachlor ide
ppb yg/m3
0.0004
0.0020
0.0020
0.0060
0.0004
0.0026
0.0128
0.0128
0.0385
0.0160
and CC1,.   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
times to the compounds of interest.  Therefore, reported results are
higher than actual concentrations in the emission gases.  The differing
concentrations for DCM quoted  in Table S, 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 DCM 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 organics  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/m .
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  dichloromethane. Subtraction of background  interference based
on control blanks  (Appendix A.6.3) was performed on  solvent extracted
samples.   Comparison  was then made with the appropriate waste  feed sample
chroma tog rams.   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  on portions of 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 Pollution 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 participating 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  of  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/m   in the stack gases.  TRW did, however, search for
PCB's using  techniques designed to detect small quantities of these
compounds.   ORF and MOF 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 dichloromethane,  to be present at microgram per  cubic meter
(yg/m ) 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  be conveniently  summarized.
While burning chlorinated hydrocarbons,  low  molecular weight  chlorinated
                                               7
compounds were emitted at levels  of a  few yg/m  (Table 5).   None of the
participating  laboratories detected any  high molecular weight  chlorinated
hydrocarbons  from either  air  sampling  train.   At  detection  limits  of
3 ug/m3 in the stack gases,  polychlorinated  biphenyls-were  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 kO ppb  was  determined.   If  a collection efficiency of
80%  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 in Table 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
Compos \ te
WBA
WBB
WBC
Maximum
Content
Assuming no
Combustion.
ppm g/m
550 2.40
*»70 3-37
350 3.02
Estimated Organic
in Kiln Emissions
From Sample
Chroma tog rams
ppb yg/nr*
50 177-1
50 177.1
50 177.1
Minimum
Combustion
Efficiency %
v/v
99.990
99.989
99.986
*Based on  the  molecular weight of dichloromethane.

5.3.3      Particulate emissions
           Summary  data for  the particulate  tests 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 and C1-, which react with the alkali
components in  the  raw feed  to form  volatile alkali  chlorides.  At ths
precip!tator 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.  Another  factor  influencing emissions of particulate  matter

-------
TABLE 7.  SUMMARY OF PARTICULATE TEST DATA
Test #
1 BLA
2 BLA
3 BLA

1 WBA
2 WBA
3 WBA

1 WBB
2 WBB
3 WBB

I WBC
2 WBC
3 WBC

i 8LB
2 BLB
3 BLB

Date
Oct.
Oct.
Oct.

Oct.
Oct.
Oct.

Dec.
Dec.
Dec.

Jan.
Jan.
Jan .

Jan .
Jan .
Jan.

20/75
21/75
22/75

28/75
29/75
30/75

10/75
11/75
12/75

7/76
8/76
9/76

15/76
20/76
21/76

Probe
Rinse
Gain
(mg)
57.6
58.8
35.8

99.0
94.8
237.5

.143.0
86.1
145*6

62,2
43 1
8?,8

65.2
5'. .4
11.9

Fi Iter
Gain
(rag)
285.7
366.0
291.8

1242.6
1261.7
2193.8

424.1
468.4
621.6

546,3
436.5
600,0

251.7
240.1
183.5

Total
Gain
(mg)
343.3
424.8
327-6

1341.6
1356.5
2431.3

567.1
554.5
767.2

608,5
479-6
687-8

316.9
291-5
195-4

% of
Total Gain
on Filter
83
86
89

93
93
90

75
84
81

90
9i
87

79
82
94

Vo 1 ume
Sampled
(std ft3)
141.04
135.86
132.83

142.81
137.34
109.88

106.64
117-06
116.24

119.62
113-52
118.99

126.88
116.60
116,45

Flow rate
ACFM
157,000
155,000
153,000
Average
157,000
153,000
165,000
Average
148,000
162,000
161,000
Average
172,000
1 60 , 000
167,000
Average
183,000
163,000
163,000
Average
Concentration
(grains/ft3)
0.0376
0.0483
0.0367
0.0409
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
1458
1524
3415
2132
0821
0731
1019
0857
0785
0652
0892
0776
0385
0386
0259
0343
Emission
Rate
(Ib/hr)
20.8
25.4
19-6
21.9
83.9
84.6
200.0
122.8
40.3
40.5
55.0
45.3
45-3
54.5
52.2
44.0
23-9
21.0
14.6
19.8

-------
                                    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 Ib/ton  of  clinker produced, compared to
the Canadian Federal Government objective of 0.9  Ib/ton.  The average
value  for  the  other  phases of the test  was about  1.1 Ib 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 yg/m  for the  baseline emission rates  and,
during the burning of waste  chlorides,  did not exceed  10 yg/m .   The
current Ontario standard  is  100 yg/m .
$.k        Mass Balance  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  immediately
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.
5-1*.!      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  11-14, 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.A.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.  K20
 (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 K20 retained was  less affected than
 that of  chlorine by such losses because the quantity of K20 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
Accumulated Period
7/10/75-22/10/75
23/10/75- VI 1/75
2/12/75-1V12/75
3/ 1/76- 9/ 1/76
10/ 1/76-2 I/ 1/76


Accumulation Period
7/10/75-22/10/75
23/10/75- VI 1/75
2/12/75-1V12/75
3/ 1/76- 9/ 1/76
10/ 1/76-2 1/ 1/76
Chlorinated Accumulated % of T(
Hydrocarbon Input (1b) From Chlorinated Accumulated
Burned (T|) Hydrocarbon Retention (Ib) % Retention
28,249 0
Aliphatic 134,379 83.2
Aromatic + 135,705 85.0
Complex
PCS 58,590 82.4
18,351 0
TABLE 9. ACCUMULATED MASS BALANCE
Chlorinated
Hydrocarbon Accumulated
Burned Input (tons)
240.99
Aliphatic 199.67
Aromatic + Complex 194.59
PCS 103.25
280.40
20,768
68,088
112,640
41,704
16,927
FOR K20

Accumulated
Retention (tons)
229-37
166.72
186.68
92.88
274.36
73-5
50.7
83.0
71-2
92.2


% Retention
95.2
83.5
95.9
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.

-------
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 considered a deterrent to use of this
waste in  the kiIn.
          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 per day in
the quantity of dust discarded (Table 11) while 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 Burning Chlorinated  Hydrocarbon Wastes
          As indicated earlier in this report, Na_0 is not considered in
this study because it  is  low and practically 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 K.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.  evolved  from  the raw materials in the burning process.  The

-------
                                        TABLE 10.
AVERAGE REDUCTION IN 1^0 CONTENT OF CLINKER
Accumulation Period
7/10/75-22/10/75
23/10/75- 4/11/75
2/12/75-14/12/75

3/ 1/76- 9/ 1/76
10/ 1/76-21/ 1/76
C linker
Prod.
(t/day)
1056
1050
1020

1006
1020
Cl Input With
Hydrocarbon
Chlorinated % Relative to
Hydrocarbon Clinker Prod.
-
Al iphatic
Aromatic +
Complex
PCB
-
C
0.41
0.44
-
0.34
0
Slurry
Na t u ra I
Basis
0.92
0.92
0.90

0.91
0.91
Kzp (?)
Igni ted
Bas i s
1.42
1.42
1.42

1.41
1.40
K20
Cl inker
(*)
1.21
0.87
0.74

0.87
1.25
K20
Reduction
(*)
0.21
0.55
0.68

0.54
0.15
Calculated*
K20
Reduction (*)
0
0.58
0.58

0.45
0
*Based on chlorine input.

-------
                                   36
                   TABLE 11.  AVERAGE DUST DISCARDED
Period
7/10/75-22/10/75
23/10/75- Vll/75
2/12/75-1V12/75
3/ 1/76- 9/ 1/76
10/ 1/76-21 / 1/76
Chlorinated
Hydrocarbon
—
Aliphatic
Aromatic +
Complex
PCB
—
Average Dust
Discarded
(Tons Per Day)
22.2
42.7
74.0
62.5
45.1
percent  K.O  In  slurry feed on  the  ignited basis minus the percent 1^0 in
the clinker  yields the  reduction through volatilization  in the kiln.
Since  the  K.O reduction  is achieved by  formation of potassium chloride
(KC1), the reduction expected  while burning  chlorinated  hydrocarbons,
based  on the assumption  that all chlorine is bound  Into  potassium chloride,
can be readily  calculated.  Comparison  of actual and calculated values
(Table 10) are  excellent.   If  chlorine  is added to  reduce alkalies, the
reduction  is stolchiometric.   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  hydrochoric 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 K.O is  apparent.   While burning chlorinated hydrocarbons,
the average  reduction is that  calculated on  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 volatilization 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  is considered that potassium chloride  forms a cycle
within the kiln and may be  volatilized  several  times before 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  l<20 at  Stage  IV was a  factor  of ten  greater than
the K 0  input,  it  follows that  ten  cycles of volatilization occur on average
in  the suspension  preheater kiln.   The situation in the wet process kMn,
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 k2% chlorine is being burned with an
alkali cycle of three times.  For each pound of chlorinated material,
 there are O.J»2  pounds of chlorine which will produce 0.88 pounds of
potassium chloride.  The heat  of vaporization of potassium chloride is
 38,840 cal/mol  [19] or 938.6 Btu/lb.  The 0.88 pounds of potassi
                             ium

-------
                                    38

       TABLE  12.   RECOVERY  OF Btu  FROM CHLORINATED HYDROCARBONS
Material
Aliphatic
Aromatic +
Comp 1 ex
PCB
Average
Reduction
Input With
(10 x Btu/ton
Clinker)
0.125
0.128
0.157
Average
Input With
Chlorinated
Hydrocarbon
(10^ x Btu/ton
Clinker)
0.205
0.217
0.228
% Useful
Heat From
Chlorinated
Hydrocarbon
61
59
69
 chloride would require 2^78 Btu for three cycles.  On this basis we
 might  expect  to recover:

           93°°     7
               9300     X 10°* " 73* of  the  heat content of the chlorinated
 hydrocarbon.
           While the above is  only an approximation 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 kilns have in  common 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  (HC1)  is preferentially
formed rather than free chlorine.   In all cement kilns, the lime will
readily 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.  This would apply to any method of
chloride addition.  To  overcome  the  plugging problem caused by condensation
of these salts in suspension preheaters,  design  modifications to  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 read!ly 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
Environmental  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
 (HC1) which  is dissolved  in  the  ocean.

-------
          Another method of disposal  in  North  America  is  Incineration
with production of hydrochloric acid  [22],
          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  facilities  without  hydrogen chloride
               scrubbing;
           k)   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  recovery of hydrochloric acid; and,
           9)   combustion 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 procedures 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  [2k]  indicates  that  a variety of  legal and
illegal means  are used to  dispose of  chlorinated hydrocarbon  wastes.
Of  illegal  means:  "the discharge 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 than the O.I seconds normally found in an
i nci nerator.
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
i ncinerator.
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 lime.
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  cement kiln of persistent
and toxic waste materials which  normally require disposal.

-------
7         CONCLUSIONS
          The concept of using  chlorinated  hydrocarbon wastes  in
cement processing derives  from  knowledge  of kiln  operating temperatures
and residence times  in comparison  with  incinerators  capable of destroying
these compounds.  The action  of cement  kilns 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 precipitator as indicated by the
necessity  of 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  incinerators 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 manufacturing has led  the authors to
conclude that all chlorinated hydrocarbon wastes  may be  used  in cement
kilns without adverse effect  on air pollution levels.

-------
                              REFERENCES
 1.   Personal  communication,  January 15,  1975-
 2.   Niel,  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  O.L. Eshenour, "Significance  of Total  and  Water
     Soluble Alkali  Contents  of  Cement",  Proceedings of the Fifth
     International  Symposium  on  the Chemistry of Cements, Tokyo,  1968,
     Published 1969.
 A.   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  Association,  Research Department Bulletin  3',  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-1^9, Portland Cement  Association, Manufacturing Process
     Department,  1956.
10.   Woods,  H., J.L.  Gilliland,  Jr., J.F. Weigel, B.E. Kester,  and
     H.A. Stevens, "Symposium on Alkali Removal and Problems".   Regional
     Fall Meeting of General  Technical  Committee, PCA, Milwaukee,  Wisconsin,

-------
                                    44
      Sept.  21-24 1959.   Mill  Session Paper M-158,  Portland Cement  Associa-
      tion,  Manufacturing Process Department,  I960.
 11.   Coles,  C.W. and D.G.  Dainton,  "St.  Lawrence Cement Co.  Clarkson
      Plant",  Cement Technology 1 (2), 43,  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 75 (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, Portland Cement  Associa-
      tion,  Lincolnwood,  111.,  February,  1976.
 15.   Trauffer, W.E. "Portland  Cement Outlook  and Review".  Pit  and Quarry,
      January 1976.
 16.   Campaan,  H.,  Central  Laboratory TNO Report  No.  CL 74/93.   "On the
      Occurrence of Organic  Chlorides in  The Combustion Products of an
      EOC  Tar  Burnt by  the  Incinerator Ship 'Vulcanus1; 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-74-3,  1974.
 18.  Weber, P.  (Translation),  "Alkali  Problems and Alkali Elimination in
     Heat-Economising  Dry-Process Rotary Kilns",  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  "Vulcanus11,  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 Committee on
     Industrial Wastes Report.  Incineration of Chlorinated Hydrocarbons
     with Recovery of HC1 at E.I.  du Pont de Nemours & 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-
2k.  "Incineration of Industrial Chlorine Wastes on the High Seas".,
     Report from the Environmental Agency  (Ministere chargfe de
     1'Environnement) of the Pollution and Nuisance Prevention
     Administration  (Direction de  la Prevention des Pollutions et
     Nuisances), France.   1974.
25.  United States  Environmental Protection Agency, Permit No. 730 DOOSC
     (3) to Shell Chemical  Company,  Inc. and Ocean  Combustion Services,
     B.V. December  12,  1974.
26.  Peray, K.E., and J.J.  Waddel1, The  Rotary Cement  Kiln, Chemical
     Publishing Co.  Inc. New York,  1972.
27.  Reynolds,  L.M., "Pesticide residue  analysis  in the presence  of
     Polychlorobiphenyls  (PCB's) Residue Reviews,  34,  27,  1971-
28.  Santoleri, J.J.  "Chlorinated  Hydrocarbon Waste  Recovery and Pollution
     Abatement" Chem. Eng.  Prog. 69  (0  68,  1973-

-------
                          ACKNOWLEDGEMENTS

          The authors wish to thank and  to acknowledge  the participation
of the following 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. Friedrlck 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 coordination and review of results.  TRW
          Systems Group and  Control  Pollution Services  Inc.,  who were
          contracted by EPA.

-------
                       APPENDIX A
QUANTIFYING, SAMPLING AND ANALYSIS OF PROCESS MATERIALS

-------
                              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.I.  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.I lists the materials examined
with approximate relative quantities under normal production conditions.
TABLE A.I.  PROCESS MATERIALS STUDIED AND APPROXIMATE NORMAL PRODUCTION
            QUANTITIES
      Material                             Approximate  Quantity
      Slurry Feed                          15^0  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/min
      Chlorinated Hydrocarbons             1  -  2 gal/min
      Kiln Exhaust  Gases                   160,000 ACFM
A.1       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-

-------
                     1  SLURRY FEED
                     2  AIR
                     3  FUEL (OIL,CHLORINATED  MATERIALS)
                     4  DUST RETURN
                     5  CLINKER
                     6  PRECIPITATOR DUST WASTED
                     7  STACK GAS
                                                                 vn
                                                                 o
A KILN
B PRECIPITATOR
SCHEMATIC  OF THE  MATERIAL  BALANCE
                 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 hours 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 is known.
          The clinker is fed to a bucket elevator by gravity through a
chute from the cooler.  Clinker samples are taken from this chute at
intervals of two hours to form a 2A-hour composite.  The clinker composite
sample is mixed, reduced by "cone and quartering" and a portion ground
for analysis.
          In common 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:

                      100 - HO
          SFD  - SF  '
                         100
where     SFD  «  quantity of slurry feed on a dry basis
          SF   »  quantity of slurry feed including water
          HO  =  % water in the slurry feed.
          To obtain the quantity of clinker produced, the C02 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 -  L01    \        /100- LOI
          "inker  -  SFD ^   ]QQ  SFDj  - DD   ^   )00

where     SFD      -  quantity of slurry feed on dry basis
          LOI __   =  %  loss on  ignition on slurry  feed on dry basis
          DD       *  quantity of dust  discarded
          LOI _    *  %  loss on  ignition on dust discarded.
A.2       Weighing and Sampling  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 zone  of  the kiln  (return dust).  Samples
 of the dust  are taken at two-hour intervals from the conveyor belt to
 form a 2^-hour  composite.  The  composite is blended and a  portion taken
 for  analysis.   A minor portion  of the dust (discard dust)  is fed  into a
 second holding  silo  from which  trucks are  loaded directly.  A sample
 of this dust is  taken from each truck load, and all samples for each day
 are blended  into a  24-hour composite.  Each  load of dust  is weighed on
 the truck weigh scales before  being disposed  of.
 A.3      Measurement and  Sampling  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 monitored and
 recorded in  the production data but are not  given  in  this report.
 Chlorine content of  the  oil  was determined,  and the contribution to  the
 chlorine mass  balance due  to oil was  included.
 A.k      Measurement and  Sampling  Waste Chlorinated Hydrocarbons
          Sampling of chlorinated hydrocarbons was carried out by 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 10   mhos.  While this was the case for the
chlorinated aliphatic wastes, the conductivities of the other materials
were less than 0.3 x 10   mhos.  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 Sampli ng
          Separate sampling trains 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 collected from a duct downstream of the precipita-
tor controlling dust emissions from  kiln  #1.  The  location  is described
below,  followed by  descriptions of sampling equipment and methodology.
A.5.1     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 common  section of
duct entering  the stack.   The  stack,  which  is of  height  55^-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
1»0  feet per  second.
          The  rectangular  breeching  from  the  precipitator makes a  90
bend and  ti.en  angles at about  30  upwards from  the horizontal  for  a
distance of  approximately  45  feet  to the  common header which  leads  into
the stack.   This  k$ foot  section of  dgct  was  considered  the most suitable
for installation  of sampling  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 the  duct  at a  position  12  feet upstream of the  bend into the

-------
                            54
TABLE A.2.  QUANTITIES OF ALIPHATIC MIXTURE BURNED DAILY
Date
23/10/75

24/10/75
25/10/75
26/JO/75
27/10/75
28/10/75
29/10/75
30/10/75

31/10/75
1/11/75
2/11/75
3/H/75
V 11/75


Time
From To
09:00
12:00
06:00
06:00
06:00
06:00
06:00
06:00
06:00
09:30
06:00
06:00
06:00
06:00
06:00
10:00
OFF at
12:00
06:00
06:00
06:00
06:00
06:00
06:00
06:00
09:30
06:00
06:00
06:00
06:00
06:00
10:00
17:30
17:30
Minutes
180
1080
1440
1440
1440
1440
1440
1440
210
1230
1440
1440
1440
1440
240
450

Igpm
0.5
1
1
1
1
1
1
1
1
2
2
2
2
2
2
1

Gallons
90
1080
1440
1440
1440
1440
1440
1440
210
2460
2880
2880
2880
2880
480
450

Gal Ions/Day
1170

1440
1440
1440
1440
1440
1440
2670

2880
2880
2880
2880
930



-------
                                  55





TABLE A. 3.  QUANTITIES OF AROMATIC PLUS COMPLEX MIXTURE BURNED DAILY
Date
Dec.

Dec.
Dec.
Dec.
Dec.
Dec.
Dec.
Dec.
Dec.
Dec.
Dec.

6

7
8
9
10
11
12
12
13
14
15
Time
15:00
2 loads
12:15
15:00
15:30
12:00
15:00
09:00
15:30
13:30
13:20
1*4:00
Tank Difference
Measurement gallons minutes
12'
in 60930
14'
IV
13'
12'
ir
11 '
ir
10'
9'
8'
0"
Ib + 40690

Ib @ 1.
6" (7823-5886
1937
0"
6"
6"
8"
4"
0"
3"
6"
8"
1177
1177
2354
1962
785
785
1373
1766
1962

299 SG - 7823
= 1937 gal)
1275
1605
1470
1230
1620
1080
390
1320
1430
1480
Ga 1 /m i n

gal
1.519
0.733
0.801
1.914
1.211
0.727
2.013
1.040
1.235
1.326
Gal Ions per Day
Date
Dec.
Dec.

Dec.
Dec.

Dec.

Dec.

Dec.


2
5

6
7

8

9

10

From
06:00
10:00
14:00
06:00
06:00
12:15
06:00
15:00
06:00
15:30
06:00
12:00
To
13:10
12:00
06:00
06:00
12:15
06:00
15:00
06:00
15:30
06:00
12:00
06:00
Minutes
430
120
960
1440
375
1065
$1*0
900
570
870
360
1080
Igpm
(1)
1.519
1.519
1-519
1.519
0.733
0.733
0.801
0.801
1.914
1.9H
1.211
Gal Ions
430
182
1458
2187
570
781
396
721
456
1665
689
1308
Total Gallons
430
1640

2187
1351

1117

2121

1997


-------
                                   56
TABLE A.3.  (Cont'd)
Date
Dec.

Dec.


Dec.

Dec.

Dec.
11

12


13

14

15
From
06:
15:
06:
09:
15:
06:
13:
06:
13:
06:
00
00
00
00
30
00
30
00
20
00
To
15:
06:
09:
15:
06:
13:
06:
13:
06:
06:
Minutes
00
00
00
30
00
30
00
20
00
00
540
900
180
390
370
450
990
440
1000
1440
Igpm Gallons Total Gallons
1.
0.
0.
2.
1.
1.
1.
1.
1.
1.
211
727
727
013
040
040
235
235
326
326
654
654
131
785
905
468
1223
543
1326
1909
1308

1821


1691

1869

1909

-------
                         57
TABLE A.4.  QUANTITIES OF PCB MIXTURE BURNED DAILY
Date
3/1/76
4/1/76
5/1/76
5/1/76
6/1/76
7/1/76
7/1/76
8/1/76
8/1/76
9/1/76
9/1/76

Date
3/1/76

VI /76

5/1/76

6/1/76

7/1/76


8/1/76


9/1/76

Time
14:30
12:00
09:15
14:30
09: 10
09:06
16:00
08:45
15:45
08:45
05:45

From
07:45
14:30
06:00
12:00
06:00
09:15
06:00
09: 10
06:00
09:06
16:00
06:00
08:45
15:45
06:00
08:45
Tank
Measurement












To
14:
06:
12:
06:
09 =
14:
09:
06:
09:
16:
06:
08:
15:
06:
08:
18:













30
00
00
00
15
30
10
00
06
00
00
45
45
00
:45
:15
11 ' 4"
12' 6"
9' 8"
9' 6"
8' 7"
7'H"
7'10"
7' 9"
7' 8"
7' 2"
6' 8"
Gal Ions
Minutes
405
430
360
1080
195
315
190
1250
186
414
840
165
420
855
165
570
Di f ference
gal Ions minutes

1962
1962
392
1766
1962
196
196
196
1177
1177
per Day
Igpm
1 .521
1.521
1 .521
1.539
1.539
1.244
1.577
1.366
1.366
0.473
0.195
0.195
0.467
1.154
1.154
2.180

1290
1275
315
1120
1436
414
1003
420
1020
540

Gal Ions
616
1414
548
1662
300
392
300
1708
254
196
164
32
196
987
190
1243
Gal An in

1.521
1-539
1.244
1-577
1.366
0.473
0.195
0.467
1.154
2.180

Total Gallons
2030

2210

2159

2008

614


1215


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
           Prior to setting up equipment for particulate and gaseous
sampling,  grab samples of the kiln emissions were  taken using  the
equipment  depicted  in Figure A.**.  A gas sample was pulled through an
 in-stack kl mm 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-tight seal.  Valves A  and  C were
closed and B opened so that, with  the  pump on, the sampling  line was
purged.  Valve B was  then closed and valve C opened to  evacuate the
sampling  lung, making sure that the vacuum did not exceed  5"  Hg.  At a
vacuum of  V 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 for analysis  on  the  same day  that the sample  was
collected.
A.5.2.1    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
recommended by both the  Ontario Source Testing Code [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.

-------
                             59
DUCT  DIMENSIONS
(AS MEASURED) >
 56-5" WIDE x 78"HIGH
               SAMPLING  PLATFORM.
FIGURE A.2.   SCHEMATIC OF PORT LOCATIONS

-------
                                  60
[
c
c
c
108.9
X
92.5
X
78.2
X
72.5
X
42.8
X
110.6
X
107.8
X
89.2
X
74.2
X
51.9
X
118.6
X
112.0
X
92.6
X
78.2
X
66.4
X
118.6
X
113.3
X
91-9
X
78.2
X
73.4
X
117.6
X
110.6
X
94.5
X
82.0
X
78.2
X
114.7
X
110.6
X
100.2
91.2
X
80.5
X
107.8
X
108.9
X
102.0
X
97.7
X
74.8
X
103.2
X
107.8
X
104.9
X
97.7
X
80.5
X
                       Velocity in feet per second
         FIGURE A.3.  GAS FLOW DISTRIBUTION AT SAMPLING POINTS
An Orsat analysis was made to find the concentrations of CO, C0_, 0_ and
N- in the gas stream and a gas moisture determination was carried out
using an established procedure [A.)].  With the preliminary data obtained,
the isokinetic sampling rate was calculated using known standard
equations [A.I].
          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.I] were recorded every 2.5 minutes 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

-------
                        6)
VACUUM
GAUGE
                                      IN-STACK  FILTER

                                      PROBE
                        HEATED  TEFLON LINE
                        ROTAMETER
  TEDLAR BAG
LUNG
PUMP
 FIGURE A.A.  GRAB BAG SAMPLING EQUIPMENT

-------
                                       62
                    HEATED AREA -
r FILTER
 HOLDER
                                     THERMOMETER
                                                                        CHECK
                                                                        VALVE
         r
REVERSE-TYPE
  PITOT TUBE
       PITOT
MANOMETER

                                 [CYCLONE_
                                 (OPTIONAL)
          IMPINGERS_
          "ICE BATH
                                                   	l
                   THERMOMETERS^  )
        BY-PASS
         VALVE
                                             VACUUM
                                              0AUGE
                                                 \=^
                                                                              »'
                                    C>RYTEST
                                    METER
                  FIGURE A.5.  PARTICULATE SAMPLING TRAIN

-------
to polyethylene bottles, which were then labelled.  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 impinger
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 probe 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 emission  rate
calculated by  use of appropriate equations  [A.I].
A.5.2.2   Gaseous train.   It was originally  intended to sample for
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 CHCl^, CH2C12 °r CHC12CH2C1'
were  placed  in the  second impinger.   The third impinger was left empty
and the  fourth contained silica gel.   After a period of three to four
 hours,  the time anticipated for a particulate test, the impingers were

-------
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 was 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%-3Q%.  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 shown that collection and retention of
trace organic compounds is possible using specific adsorbent materials
[A.3].   Inert materials, such as the Chromosorbs, are considered to
have certain advantages 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  particulate train with Chromosorb 102 or activated
carbon in the third impinger were carried out, passing heated air
containing a few ppm of CHC1. through the system at a flow rate of 0.5
cfm.  Neither adsorbent removed more than 70$ of the CHC1, initially
and, after about thirty minutes, almost all of the CHC1  was passing
through the system,  it was concluded, 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.
          After  further studies in the laboratory with both Chromosorb
102 and activated carbon,  the sampling train for gaseous organic compounds
depicted in Figure A.6 was constructed.  Tests made with an air stream
containing 15 ppm of CHC1. snowed that either adsorbent would remove
better than 95% of the CHC1. 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 through midget impingers containing water and

-------
                                     65
                PROBE
                                       S.S TUBING
              IN- STACK FILTER
                        — ROTA METERS
r

                              .ICE BATH
                          ^	 CHROMOSORB
                          1    ADSORBENT TUBES
                                                           MIDGET
                                                         /IMPINGERS
       WATER
  SODIUM
  HYDROXIDE


  PRESSURE
  GAUGES
                                 PUMP









TEST
METER
             FIGURE A.6.  GASEOUS  SAMPLING  TRAIN

-------
                                   66
caustic soda (5$ solution), respectively, to remove any HC1 and Cl»
present in the kiln emissions.  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 compounds.
          After  each  test  the in-stack filter and impinger solutions were
stored  in labelled 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
          The methodology  used to collect kiln emission samples,  in
baseline or waste burn test periods, provided four distinct  types of
samples for analysis  of chlorinated organic compounds.  These  sample
types  and the analyses required were:
          -  chlorinated waste feeds for compositional analysis;
          -  grab bag samples for any chlorinated organic species;
          -  Chromosorb adsorbent samples for volatile  low molecular
             weight  organic compounds; and,
          -  solvent  extracts of filters and solutions  for noncombusted
             waste components.
A.6.1     Sample preparation
          Waste  feed.  Samples of the prospective  WBA  feed material were
supplied  in advance  of the test  burn for analyses.   When  the test period
arrived, however,  insufficient quantities of some  components were available
to make up  the specific  blend of waste feed.  The  composition of material
actually delivered to SLC was, therefore, different  from the sample
supplied  to ORF  in advance.   In  order  to obtain  a  true compositional
analysis of the  material  being supplied  to  the  kiln, and to determine

-------
                                   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 of this adsorbent
for the collection of volatile organohalides from an air stream.  Thermal
desorption of adsorbed components  into an evacuated gas jar was selected
as a method for preparing samples  for  GC analysis.   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 heatjng which tended  to  produce excessive  background noise
during GC analysis.
          Adsorbent tubes were  made  up by  packing Chromosorb  102  in glass
containers  H.5 cm  in length  and  11.0  mm inside diameter,  using  silanized
glass wool  plugs at either end  for support.  Tubes  for the  WBA and WBB
burns were  preconditioned by  heating to  200 C  and passing  a  stream of
nitrogen at kO ml/min 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 2^0°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 controlled 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 the  laboratory have shown that recoveries in excess of 30%
may be expected for adsorbents with adsorbed organohalides.
           Solution and filter extracts.  Pentane or hexane was used as an
extractant for organic compounds from all  filters and probe rinse
insoluble  fractions,  using a Soxhlet 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.SO., filtered, and
concentrated by evaporation using a combination of rotary and Kontes
tube  heaters.  The concentrate was made up to a known small volume with
solvent for GC analysis.
           Pentane was  used as extractant for BLA, WBA and WBB test
samples, since  its high volatility would minimize 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 (Fl) or electron capture  (EC) detectors.   Gas  chromatography  -
mass  spectrometric  (GC-MS) analysis was performed on  waste feed samples
to confirm the  identity of major components.  The various conditions
and columns used are  summarized  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
Co Iumn

Column Temperature
Injector Temperature
Detector Temperature
Detector
Flow Rates
Range and Attenuation
Chart Speed
-  Chromosorb 102 (80/100 mesh)
   6' x 1/8" SS
-  180°C
-  215°C
-  215°C
-  FID and EC
-  N  at 40 ml/min
   Air and H. adjusted for maximum
   sens!tivi ty
   As required
   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; HDMS;
    60/80 mesh)  12' x  1/V SS
-   60°C  isothermal for  20 minutes then
    programmed  at  10°C/min to  maximum
    temperature
-   190°C
-   230°C
-   FID
-   N2 at  AO ml/min
    Air  and H?  adjusted  for  maximum
    sens i tivity
    As requi red
    As required

-------
                                   70
                  TABLE A.7.   GC ANALYSIS - SYSTEM C
Column
Column Temperature
Injector Temperature
Detector Temperature
Detector
Flow Rates
Range and Attenuation
Chart speed
                         -   k% SE 30/6* QF 1  on Chromosorb W (HP;
                            60/80 mesh)   6'  x 1/8"  SS
                         -   200°C
                         -   250°C
                         -   230°C
                         -   Linearized EC
                         -   N  at 25 ml/min
                         -   As required
                         -   As required
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 analytical studies it is necessary to avoid
contamination of samples at all stages of sample collection and
preparation.  Though extreme care was taken during the 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:
 i!)  Gaseous train:
iii)  Solvents:
                         (a)   Particulate filter + probe rinse
                              filter + Soxhlet thjmble.
                         (b)   Water
                         (a)   In-stack filter •»• probe rinse
                              filter + Soxhlet thimble
                         (b)   Water
                         (c)   Aqueous NaOH
                         (a)   Pentane (100 ml concentrated to
                              1  ml)
                         (b)   Hexane (100 ml concentrated to
                              1  ml)

-------
                                   71
iv)  Polyethylene containers:       (a)  Solvent extracts (100 ml  concentra-
                                        ted to 1  ml)
          Standard gas samples of various components of interest, such as
CC1,, CHC1,, CH Cl- and 1,2-dichloroethane, were made by injecting 50
yl aliquots 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 concentration were
made by dilution of the  100 ppm standard using a similar procedure.
These standards were used immediately after preparation and were not
retained for use on the  following day.
          Solution standards were made by weighing accurately known
amounts of the components of  interest and dissolving them  in a known
volume of pentane.
A.6.**     Concentration  factors
          Desorbed compounds  from the adsorbent  tubes were  concentrated
 into a 500 ml  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  program giving
 a range of  concentration factors from 85 to  181. Thus, a  concentration
 for  a particular  component  of 1  ppm  in  the gas jar  sample  would  mean  a
 concentration  of  about 10  ppb in  the  kiln emission,  assuming  a  100
 percent collection efficiency,  and  subsequent desorption 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 were  determined  for various  components detected.   Assume that
 CH Cl  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 CH2C12 is 85-  Therefore, 2 pg
 of CH Cl   at  70°C,  the temperature at which the emission sample  was
 measured,  occupies a volume of:

           	7-  x 2k litres, or	r  x —— cubic feet.
           85 x 106                  85 x 10°    28.3

-------
                                   72
For test WBA 1, for example, a volume of 142.8l scf was sampled.   Thus,
                             n
                             6
the concentration of CH Cl- In the kiln emission is:
                x 	x	  ppm or 0.14 ppb.
       85 x 1(T   28.3   U2.81
The concentration factor is therefore  10  _ • 7140.
                                      0.14
          Factors were calculated for feed compounds of lowest and
highest molecular weight.
A.6.5     Samp1e chroma tog rams
          Analysis for low molecular weight organohalides and noncombusted
chlorinated organic compounds present in any collected samples was made
using GC techniques only, by selecting appropriate GC parameters and
using retention  time  data.  No specific cleanup or separation procedures
were performed  in order  to segregate components of interest from possible
interfering compounds because, although GC profiles obtained were quite
complex, very low concentrations of organic compounds were evident from
the peak heights obtained for the attenuations 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-FID profiles for the three waste feed materials
burned  in this  study  are given  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 anticipated that the levels, if any, of uncombusted
components  in the stack  emissions would be very low.  Therefore GC-CC
analysis was used to  evaluate the organic solvent  extracts of the
various impinger samples from the sampling trains  for the presence of
any uncombusted components.  Figures A.10 and A.11  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.

-------
  8
>-
h-

i/v
z
LU
H-
Z
                         SAMPLE FEED -OCT. 28th1975

                         GC-SYSTEM  B
                                       12
                                                                                         25
               -10
              10  x 256
10
                                      15
20
25
30
35          40
  TIME (min)
           FIGURE A. 7     GAS CHROMATOGRAPHIC PROFILE  FROM FLAME IONIZATION  DETECTOR

                           FOR CHLORINATED  ALIPHATICS   (WBA) SAMPLE  FEED

-------
          SAMPLE FEED - DEC. 12th 1975
          GC-SYSTEM B
                                     25
30
35
                                                                     26
40      45
  TIME(min)
FIGURE A.a GAS CHROMATOGRAPHIC  PROFILE FROM FLAME IONIZATION  DETECTOR
          FOR CHLORINATED ALIPHATICS  PLUS  AROMATICS AND  ALICYCLICS
          (WBB) SAMPLE  FEED

-------
   10-

   9

> 8
                                                             16
       SAMPLE FEED-JAN. 8th1976
       GC-SYSTEM B
            10"11x256
            —r—
            5
                          10"1(i256
10^.128
10      15      20     25      30      35      40      45
                   50      55
                     TIME (min)
                FIGURE  A. 9  GAS CHROMATOGRAPHIC  PROFILE FROM  FLAME  IONIZATION
                              DETECTOR FOR CHLORINATED ALIPHATICS  PLUS  AROMATICS,
                              ALICYCLICS  AND  POLYCHLORINATED  BIPHENYLS  (WBC)
                              SAMPLE FEED

-------
to
til
o-
9-

8"
7-
6-
5-
4-
3-
2-
1-
0:









I



1

py












10
9
SAMPLE FEED-DEC.12th1975
GC-SYSTEM C 8.





/
0246
*
/ 6.
5 .«./ aliquot
(Temp.155°C)
4
3-
I
I
2-
1-
0-









r
8 10 12 14 0
TIME(min)
lo-
g-

s'





A
7
07 / i 6"
5^^ aliquot
1x32
(Temp.155°C)
*•
3-
2-
/ ^"

1







/



0.7 ng /<>«•<•
5 <>i/ aliquot
1x64
(Temp.190°C)



I-
r
2468 10 02468
TIME(min) TIME (mm)
                 FIGURE  A.10   GAS CHROMATOGRAPHIC  PROFILES  FROM  ELECTRON CAPTURE
                                 DETECTOR  FOR WBB SAMPLE  FEED

-------
  10-
   9
   8-
   7-
V- 6-
                                               10
   5-
LU
h-
Z
   3-
   o_
   1-
     1x128
                STANDARD AROCLOR 1242
                GC  SYSTEM  C
5-
3-
                                                2-
   1x256
                              8     10     12     0
                                      TIME (min)
                 WBC-AROMATICS + PCBs
                 SAMPLE FEED JAN. 8th 1976
                 GC  SYSTEM C
                            9
                           8     10
12     14
  TIME (min)
                      FIGURE  A. 11    GAS CHROMATOGRAPHIC PROFILES  FROM ELECTRON CAPTURE
                                      DETECTOR  FOR STANDARD  AROCLOR 12^2  AND  SAMPLE FEED WBC

-------
                                   78
          Standards.  Figure A.12 shows the GC-IC profile for the low
molecular weight chlorinated hydrocarbons, e.g. CH.Cl ,  CHC1., CC1,  and
1,2-dichlorethene.  Figure A.13 snows 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 1242 injected
into the column  is shown in Figure A.11.
          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 of  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 peaks peculiar 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.\k and A. 15 show GC-EC profiles
for the organic  solvent extracts of the EPA impinger samples obtained
for BLA-T3 and WBC-T3, respectively.  Profiles obtained for the two
samples are quite similar.  There is a large EC response occurring as an
unresolved peak  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 to segregate the PCB's from
interfering components and thus facilitate their analysis.  Such a
cleanup and separation procedure was performed on WBC-T3 using Florisil
adsorbent (2?) 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.16 was obtained on ?n
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.11),

-------
   10-
LU
                                                                                    GC-SYSTEM-A
                                                                                    GRAB SAMPLE
                                                                                    WBC adsorbed/desorbed
                                                                                    on Chromosorb 102
                                                                                      (1 ml aliquot )
BLB-2B
(0,5ml aliquot )
WBC-3 A
(0.2 ml aliquot
                                                 2   <.   6     02
                                                  TIME   (min)
                                                                                                       u>
                                             8   10
             FIGURE  A.12   GAS CHROMATOGRAPHIC PROFILES FROM  ELECTRON  CAPTURE
                             DETECTOR  FOR  LOW MOLECULAR WEIGHT  CHLORINATED  HYDROCARBONS
                             AND  FOR  BLB  AND WBC  TEST SAMPLES

-------
                               80
   9-




   8-


>
h-
-  7^
       LU
          6'-
                                STANDARD

                                  { 1 ppm )

                                GC-SYSTEM-B
                                   1,2-Dichloroethane
              10"1x256
FIGURE A.13
      )             5             10             15
                                       TIME  (min)

      GAS CHROMATOGRAPHIC PROFILE FROM  FLAME  IONIZATION

      DETECTOR FOR LOW MOLECULAR WEIGHT  HYDROCARBONS

-------
8
5-
3-
2-
1-
                                81
                           GC  SYSTEM  C
injection
                              —i	1	1	1	1	1	1	1—i	1	1—
                               10             15             20
                                                 TIME  (min)
FIGURE A.U   GAS  CHROMATOGRAPHIC PROFILE  FROM  ELECTRON CAPTURE
             DETECTOR  FOR IMPINGER  EXTRACT  FROM  BLA  TEST  3

-------
                                   82
  101
   9-
   8-
   7-
tn
z
HI
   6-
   5-
   4-
   3-
   2-
   1-
         1x64
                             GC-SYSTEM-C
                                  injection
                                 10
15            20
    TIME (min)
    FIGURE  A.15  GAS CHROMATOGRAPHIC PROFILE  FROM ELECTRON  CAPTURE
                 DETECTOR FOR IMPINGER  EXTRACT  FROM WBC-TEST  3

-------
                                  83
 10 -{
  9-
  8 -I
w-  *7
to
z
HI
  5-
  3-
   2-
   1-
        1x32
    o
                                   GC-SYSTEM  C
                                       INJECTION
10
15            20
   TIME  (min)
     FIGURE A. 16  GAS CHROMATOGRAPHIC  PROFILE FROM ELECTRON CAPTURE
                   DETECTOR FOR IMPINGER EXTRACT  FROM  WBC  TEST 3 AFTER
                   CLEANUP AND SEPARATION

-------
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 chlorinated 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 determination 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
          Samples were  prepared  by grinding  10  grams of sample for 90
seconds  in  a  SPEX Shatterbox with tungsten  carbide  containers and  pucks.
From  this material  a  1.25  inch  diameter pellet  at  8 tons  pressure  was
made  in  SPEX  aluminum  sample caps.
           For  the  X-ray fluorescence method  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:percent  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  addition  to linearity,  reproducibi1 Ity of  the calibration
 constant is a  prerequisite if it is  to be applied across  a range of
materials  having  some  degree of matrix variability.  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  Spectrophotometer
in the flame emission  mode  (Table  A.10).
          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.11

                 TABLE A.8.  GRAVIMETRIC  DUST  ANALYSES
Dust Type Date
Discard 19/10/75
11/11/75
31/11/75
4/12/75
5/12/75
6/12/75
7/12/75
5/ 1/76
6/ 1/76
9/ 1/76
14/ 1/76
16/ 1/76
Return 5/12/75
13/12/75
I/ 1/76
18/ 1/76
% Cl
2.00
2.08
5.97
2.89
7.11
4.69
6.18
4.71
3-72
4.43
0.46
0.90
2.11
4.34
2.10
1.00
% 503
-
3-17
2.36
6.53
12.12
7.27
6.90
5.08
4.23
4.18
4.65
4.32
7.35
4.22
4.99
5.66

-------
                            86
TABLE A.9.  RESULTS FROM LECO  INDUCTION  FURNACE ANALYSES
Material Date
Clinker 19/10/75
" 21/10/75
23/10/75
31/10/75
1/11/75
5/H/75
8/12/75
" 31/12/75
11 I/ 1/76
ll/ 1/76
20/ 1/76
Slurry Feed 13/10/75
20/10/75
27/10/75
9/11/75
6/12/75
9/12/75
13/12/75
14/12/75
31/12/75
3/ 1/76
8/ 1/76
21/ 1/76
% SO (Total S as)
0-93
1.34
0.75
0.30
0.16
0.66
0.93
1.39
1.43
1.04
1.42
0.45
0.42
0.41
0.45
0.52
0.48
0.60
0.55
0.43
0.47
0.46
0.47

-------
                         87
TABLE A.10.  RESULTS FROM ATOMIC ABSORPTION ANALYSES

D!









Material
scard Dust
ii
ii
ii
it
ii
1 1
ii
ti
ii
Return Dust


Cl










SI









ii
ii
inker
ii
ii
ii
ii
1 1
ii
ii
ii
1 1
1 1
urry Feed
ii
ii
n
ii
1 1
1 1
1 1
1 1
1 1
Date
11/11/75
31/11/75
VI 2/75
6/12/75
7/12/75
5/ 1/76
6/ 1/76
9/ 1/76
IV 1/76
16/ 1/76
5/12/75
13/12/75
I/ 1/76
19/10/75
21/10/75
23/10/75
31/10/75
1/11/75
5/11/75
8/12/75
31/12/75
I/ 1/76
ll/ 1/76
20/ 1/76
13/10/75
20/10/75
27/10/75
9/H/75
6/12/75
9/12/75
31/12/75
3/ 1/76
8/ 1/76
21/ 1/76
% K20
6.36
10.30
7.14
8.23
9.96
8.12
6.43
7.9*
4.84
4.18
7.79
5.54
6.02
1.32
1.61
1.00
0.54
0.40
1.01
0.64
1.18
1.31
1.10
1.45
0.93
0.93
0.89
0.88
0.91
0.90
0.90
0.91
0.93
0.94
% Na20
0.32
0.39
0.36
0.44
0.45
0.43
0.34
0.39
0.35
0.34
0.42
0.41
0.41
0.23
0.25
0.20
0.21
0.21
0.21
0.20
0.21
0.24
0.22
0.26
0.19
0.22
0.18
0.20
0.19
0.22
0.18
0.22
0.19
0.22

-------
         TABLE  A.)].   LEAST SQUARES  DATA  FOR  CALIBRATION  LINES
Material
Raw Meal H
Element
h
Slurry Feed Potassium (as K20)
Chlorine

Clinker

Dust

A. 8
Sulphur (as SO,)
Potassium (as ICO)
Chlorine
Sulphur (as SO,)
Potassium (as K.O)
Chlorine
Sulphur (as SO,)
Correlation
Coefficient

0.9172
0.9995
0.9913
0.9829
0.9999
0.9917
0.9893
0.9548
Determination of Heat Value, Chlorine
Content in No. 6 Fuel

Slope

27128
5260
739
24980
4781
400
3166
1927
278
Content and

Intercept

-0.01
0.060
0.27
-0.01
0.011
-0.23
-0.12
-0.14
-0.10
Sulphur

          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       Determination of Heat Value and Chlorine Content in Chlorinated
          Hydrocarbons
          The heat value was determined on a Parr calorimeter model 1241
equipped with Parr oxygen bomb model 1108 following standard operating
procedures for determining heat value of oil.  Due to the corrosive nature
of the combustion products, it is recommended 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 were taken.  Nitric acid is added to allow a
more distinct end point, and the chlorine  is determined by the standard
Volhard titration.

                              REFERENCES
A.I   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 1-AP-7A-1, Air Pollution Control
      Directorate, Environment Canada, February,  197^.
A.3   Pellizan, E.D., J.E.  Bunch,  and B.H. Carpenter.  Env. Sci.
      Technology 9  (6), 552-560  (1975).

-------
                APPENDIX B

ANALYTICAL DATA, CALCULATION AND DETAILS OF
EXPERIMENT ON THE SUSPENSION PREHEATER KILN

-------
                                   93
                              APPENDIX B
              ANALYTICAL DATA, CALCULATION AND DETAILS OF
              EXPERIMENT ON THE SUSPENSION PREHEATER KILN
B.I        Theoretical
          The excellent article by Weber cited in the body of the report
describes in detail the problem of alkali reduction in a suspension
preheater kiln, and was used in preparing the following notes.
          At material temperatures above 800 C in the rotary kiln,
alkalies (K 0 and Na.O) 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 part also 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 37% of the  K20 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  (e  ) will  be volatilized and  the remaining (1-  EJ)  will  be
discharged  with the  clinker.   If  the  proportion  of  gas withdrawn  through
the bypass  is  V,  then of  the  volatilized alkalies,  the proportion  eyV
will be  removed by the bypass.  The  remaining portion e.  (l-V) will  return
with the  feed, thereby giving  rise  to the internal  cycle in the  kiln.

-------
The alkali in the internal cycle has a different volatility e,, than
alkali 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)-e -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 «  e.; leaving  the cycle are a proportion in the clinker = (K-l)-(l-e  )
and a proportion  through the bypass -  [e. +  (K-l) e_)]-V.
          By setting  the quantities input and  output from the cycle as
equal, the equation may be solved to determine the amount of gas the
bypass must  remove to  keep the alkali  cycle from becoming excessive.
B.2       Experience at St. Lawrence Cement
          At St.  Lawrence Cement, samples are  taken of the material
between  the  Stage  IV  cyclone and  the kiln  inlet (see 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% chlorine  in this material  (natural basis)  is excessive.  The
quantity of gas required to be withdrawn by the bypass to maintain  lower
chloride  levels was calculated on the  assumption that  the chloride
volatilities were e. * 0.99 and e  *  1.0  (completely volatile).  These
quantities are given  in Table B.1.
          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 flows.   To alleviate this problem,
measures to  increase cooling water and decrease the quantity of ambient
air were undertaken, with limited success.

-------
                                  95
 TABLE B.I.  PERCENT BYPASS GAS REQUIRED TO MAINTAIN CHLORIDE LEVELS
Chloride
addi tion
relati ve
to cl inker
0
0.05
0.10
0.15
0.20
0.30
0.^0
B.3
% K20
i n
cl i nker
1.31
1.2A
1.18
1.11
1 .Ok
0.91
0.77
Equation of
Bypass
1 .56% or 2,Q%
nat. ign.
2.1
1».6
7.1
9.6
12.1
17-1
22.2
Approximated T
percent required to
Stage IV Cl at:
2.0% or 2.561
nat. ign.
1.6
3.6
5.5
7.5
9.1»
11.7
17.2
ime for Equi 1 ibr ium
ma intai n
2.5% or 3-2%
nat. ign.
1-3
2.9
k.k
6.0
7-0
10.6
13-8

          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:
     dt

where:

     dQ
     dt


     Cl
                  C1RM  " C'clinker " C1bypass
                   change  in quantity with  time
                                                     (pounds/minute)
        RM
                   chlorine/time  in  with raw meal      (pounds/minute)
      d         =   chlorine/time out  with clinker      (pounds/minute)
        clinker
       bypass


ClRMand C'clinker
combined as:
                   chlorine/time out  through bypass    (pounds/minute)


                    are approximately constant,  hence they can  be
       cl
               -  r i   - r i
               ~  U RM     clinker

-------
                                   96
and  C1U        =  V0_ (t)
       bypass
where:
     Q(t)       «  quantity in transition chamber at time (t) in pounds
     V          »  bypass valve proportion.
Then:

     $         -  ' , - VQ (t)
     dt             cl
Solving  for  time gives:
                   In  [I   - Vd  (t)]    QQ
      t          =  	—	          minutes.
                           V            Q

           Since  In 0 » °°,  QF must  be taken  as a  very small distance  from
the  true equilibrium.
           The sampling sequence  at Stage IV was  planned  from this  model.
Q.k        Sampling and Analytical  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  from Stage  IV  at 3 to  10 minute
 intervals when attempting to  start the chlorinated burn.  These latter
samples  were analyzed  individually.
B. 5        Pi scussion
           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 0.2^
 percent  chlorine relative to  clinker was fed starting at 09:38, June 3,
 1975.   Plugging  of the duct between the kiln and the conditioning tower
 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
results.
          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.
B.6       Analytical Results
          Analytical results are given in Tables B.2 through B.*».  The
raw meal feeds showed a gradual increase  in both the content of chlorine
and K_0.  The results from the clinker analysis  (Table B.3) showed that
the burning of chlorinated hydrocarbons resulted  in a reduction  in the
K20 content  (June 3-5, 1975 and June  10-12, 1975).
          The results of  the analyses  from Stage  IV dusts  (Table B.^)
show that the chlorinated hydrocarbons were destroyed in  the burning
process.  Chloride content of  these samples increased while burning
chlorinated wastes.  This finding  could not have  resulted  without
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  time of 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  B.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 FROH ANALYSES OP DRY PROCESS KILN RAW
Date
30/5/75
31/5/75
1/6/75
2/6/75
3/6/75
4/6/75
5/6/75
6/6/75
7/6/75
8/6/75
9/6/75
10/6/75
11/6/75
12/6/75
13/6/75
14/6/75
15/6/75
% Cl
0.067
0.069
0.073
0.071
0.07**
0.082
0.079
0.094
0.087
0.093
0.088
0.084
0.092
-
0.084
0.104
0.102
% SO
(Total S-'as)
0.60
0.60
0.60
0.60
0.60
0.60
0.60
0.62
0.60
0.60
0.58
0.60
0.60
NO SAMPLE
0.60
0.60
0.62
% K20
0.92
0.95
0.93
0.93
0.93
0-95
0.96
0.98
0-97
0.97
0.97
0.95
0.98
-
0.96
1.00
1.00

-------
                              99
TABLE B.3.  RESULTS FROM ANALYSES OF DRY PROCESS  KILN CLINKER
Date
30/5/75
31/5/75
1/6/75
2/6/75
3/6/75
4/6/75
5/6/75
6/6/75
7/6/75
8/6/75
9/6/75
10/6/75
10/6/75
10/6/75
10/6/75
10/6/75
10/6/75
10/6/75
11/6/75
11/6/75
11/6/75
11/6/75
11/6/75
11/6/75
11/6/75
11/6/75
11/6/75
11/6/75
11/6/75
11/6/75
12/6/75
12/6/75
12/6/75
1 2/6/75
12/6/75
12/6/75
12/6/75
12/6/75
12/6/75
12/6/75
12/6/75
13/6/75
14/6/75
15/6/75
Time
Average
1 1
1 1
1 !
II
II
1 1
1 1
1 )
1 1
1 1
1 1
1 2 : 00
14:00
16:00
18:00
20:00
22:00
Average
00:00
02:00
04 : 00
06:00
08:00
10:00
14:00
16:00
18:00
20:00
22:00
Average
00:00
02:00
04:00
06:00
08:00
10:00
12:00
14:00
16:00
18:00
Average
1 1
1 1
% Cl
0.03
0.03
0.03
0.21
0.03
0.05
0.04
0.04
0.07
0.05
0.06
0.04
0.06
0.04
0.0^
0.05
0.05
0.05
0.06
0.04
0.06
0.07
0.06
0.06
0.05
0.05
0.05
0.06
0.05
0.05
0.07
0.07
0.05
0.05
0.05
0.05
0.10
0.05
0.06
0.05
0.05
0.05
0.05
0.04
% so2
1.05
1.15
1.15
1.16
1.10
1.19
1.23
1.34
1.33
1.31
1.50
1 .22
1.38
1.00
1.07
1.30
1.03
1.11
1.29
1.18
1.36
0.98
0.96
1.10
1.08
1 .18
1.08
1.12
1.23
1 .01
1 .28
1.07
1. 12
1.11
1.10
. 1 1
.25
. 1 1
.23
. 1 1
* 1
. 14
.29
1 <•!
1 .40
1 .35
% K20
.27
1.29
1-31
1.30
1.24
1.30
-33
.44
.48
-43
.63
.29
.25
0.97
0.87
1 . 10
0.96
1.05
1 .24
1.27
1-25
1 .00
0.99
1 .10
_ _ f\
1 .08
.15
1 .13
, i
. 14
.21
f\f
.06
.29
-15
i /
. !6
- 15
* f
. 16
i O
. IO
-> £
• 36
1 - 15
t o O
i . 2o
?t -»7
- 17
i 1 O
1 . 1 0
i li f\
1 . 40
IP" f\
.52
1 f. 1
1 . 41

-------
                        100
TABLE B.4.  RESULTS FROM ANALYSES OF STAGE  IV DUSTS
Date
30/ 5/76
2/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
3/ 6/75
4/ 6/75
V 6/75
5/ 6/75
6/ 6/75
Time
-
-
09:00
09:06
09:12
09:18
09:24
09:30
09:38
09: 44
09: 47
09:50
09:53
09:56
09:59
10:02
10:05
10:08
10:11
10:14
10:23
10:32
10:41
10:54
10:59
:04
:09
:14
:19
:24
:29
:34
:35
:44
:49
:54
:59
12:04
12:09
12:14
14:00
15:30 .
09:00
14:00
-
-
% so,
% Cl (Total S as)
0.67 1.14
0.82
0.86
0.84
0.81
0.85
0.82
0.86
0.90
0.97
0.96
.02
.20
.21
.21
.22
.34
.44
• 51
.44
• 56
.79
.83
.85
1.95
2.08
2.03
2.00
2.03
2.30
2.17
2.21
2.22
2.29
2.29
2.32
2.38
2.49
2.28
2.34
2.47
2.84
1.92
1.59
0.97
1.14
.30
.23
.25
.19
.22
.18
• 30
.34
.32
.35
.38
.29
.31
.34
.30
.46
.38
.52
.29
.41
.37
.34
• 32
.46
• 59
• 35
.39
.41
.47
.36
.37
.43
.51
-57
.66
.47
.51
.46
.44
.45
.58
.42
.28
.44
.17
% K20
1.77
2.26
2.11
2.03
2.03
2.14
2.06
2.17
2.15
2.21
2.38
2.39
2.37
2.42
2.51
2.55
2.68
2.66
2.95
2.91
2.94
3.38
3.43
3.35
3.74
3.85
3.72
3.66
3-70
4.15
3.93
4.04
4.00
4.18
4.11
4.24
4.36
4.45
4.12
4.21
4.44
5-07
4.17
3.48
2.54
2.71

-------
                                    101
TABLE B.4.  (Cont'd)
Date
9/ 6/75
10/ 6/75
10/ 6/75
JO/ 6/75
10/ 6/75
10/ 6/75
107 6/75
10/ 6/75
107 6/75
107 6/75
10/ 6/75
10/ 6/75
107 6/75
107 6/75
10/ 6/75
107 6/75
107 6/75
10/ 6/75
107 6/75
107 6/75
107 6/75
10/ 6/75
107 6/75
10/ 6/75
107 6/75
107 6/75
107 6/75
JO/ 6/75
107 6/75
10/ 6/75
107 6/75
117 6/75
127 6/75
% so3
Time % Cl (Total S as)
-
09:00
09: 10
09:20
09:30
09:40
09:50
10:00
11:15
11:25
11:35
12:05
12:25
.66 1.20
.09 1.18
.17
.19
.15
.16
.16
.14
.11
.20
.46
.91
• 99
12:35 2.15
12:45 2.25
12:55 2.31
13:05 2.45
13:15 2.37
13:25 2.32
13:35 2.62
13:45 2.51
13:55 2.63
1^:15 2.57
14:25 2.87
14:35 2.74
14:45 2.85
14:55 2.77
15:05 2.79
15:25 3-02
15:35 3.28
15:45 3-09
2.58
2.42
.22
• 36
• 39
.29
.27
.29
.29
.34
.41
.49
.46
.64
.59
.66
.79
.71
.47
.67
.54
.69
.69
.81
.67
.70
.77
.52
2.41
.59
.62
• 51
.49
% K20
3-79
2.64
2-93
3-07
2.92
3-01
2-99
2-95
2.91
3-04
3-22
4.02
4.07
4.50
4.54
4.65
4.98
4.75
4.48
5.10
4.83
5.12
5.02
5.53
5.18
5-35
5.23
5.23
5.66
6.07
5.78
5.23
5-13

-------
  2.80 -
  2.30
UJ
z
cr
o
_i
x
o
  1.80-
  1,30
  0.80
                                                                    O
                               Q - CALCULATED


                                 - ACTUAL
              10:00
11:00
12:00
13:00
U:00
                                                              15:00    TIME
                                                         o
                                                         ro
                       CHLORINE  LEVEL  IN STAGE IV   3/6/1975
                                     FIGURE B.1

-------
  3,00
   2.50-
LU
2
o:
O 2.00
_J
I
O
   1,50
   1.00
                                                       Q-CALCULATED

                                                       t)-ACTUAL
     9:00
11:00
13:00
15:00   TIME
                     CHLORINE  LEVEL IN STAGE IV 10,5/^5

                                    FIGURE  B. 2

-------
                   APPENDIX C
RESULTS AND CALCULATIONS FOR WET PROCESS SYSTEM

-------
                                   107
                              APPENDIX C
            RESULTS AND CALCULATIONS FOR WET PROCESS SYSTEM

C. 1        Mass Balance Experimentation
C.I.I      Results of analyses of process materials
          Analyses of process materials for the mass balance were
carried out at the St. Lawrence Cement Co. plant.  Analytical results
are given in Tables C.I to C.6.
C.I. 2     Calculation of material balances
          A daily record of production and material consumption is given
in Table C.7 for the period October 7, 1975 to January 21, 1976.  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
plus complex chlorinated hydrocarbon  in the tank and difficulty in
scheduling deliveries.  From the daily composite analysis of each material
for each 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:
                        C.
where
          y    a  M   x
          *ij       ij    100
     X..  »  total daily weight of  element  x in  material  i on day j
     C    «  concentration of  element  x in  the composite  sample material
       ij
             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

-------
                                    108
balance is available by considering prolonged periods of plant operation.
Summary balances for this purpose are given  in the body of this report
(see Section 5-A, Tables 8 and 9).
          Materials entering the mass balance calculations (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 part of the mass  balance.

-------
                         109
TABLE C.
RESULTS FROM ANALYSES OF CLINKER SAMPLES
Date
7/10/75
8/10/75
9/10/75
10/10/75
11/10/75
12/10/75
13/10/75
1 VI 0/75
15/10/75
16/10/75
17/10/75
18/10/75
19/10/75
20/10/75
21/10/75
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
31/10/75
1/11/75
2/11/75
3/11/75
4/11/75
5/H/75
6/11/75
7/11/75
8/11/75
9/H/75
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
% Cl
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
0.031
0.029
0.034
0.033
0.029
0.028
0.038
0.040
0.034
0.035
0.032
0.032
0.031
% $03
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
1.39
1.02
1.23
1.18
1.18
0.91
0.71
0.47
0.70
0.92
0.75
0.73
% K20
1.13
1.12
1.14
0.88
1.27
1.08
1.09
1.15
1.47
1.16
.31
.24
.21
.28
.51
.28
.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
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

-------
TABLE C.I  (CONT'D)
                                   110
Date
31/12/75
I/ 1/76
2/ 1/76
3/ 1/76
4/ 1/76
5/ 1/76
6/ 1/76
11 1/76
8/ 1/76
9/ 1/76
10/ 1/76
ll/ 1/76
12/ 1/76
13/ 1/76
14/ 1/76
J5/ 1/76
16/ 1/76
17/ 1/76
18/ 1/76
19/ 1/76
20/ 1/76
21 / 1/76
% Cl
0.029
0.029
0.029
0.033
0.030
0.035
0.031
0.029
0.030
0.028
0.028
0.030
0.030
0.029
0.030
0.029
0.030
0.037
0.028
0.028
0.029
0.029
% S03 2
.1*0 1
.48 1
.49 1
.12 c
.68 1
.14 C
0.70 (
1.14 1
0.88 (
0.46 C
0.91
0.92
0.94
1.13
.14
.12
• 38
.36
.20
.06
1.42
0.94
; K2o
.20
.34
.37
).77
.42
).78
K58
.26
).83
).46
.05
.12
.15
-35
.28
.22
.45
.42
.27
.15
.46
.12

-------
                         11
TABLE C.2  RESULTS FROM ANALYSES OF SLURRY FEED SAMPLES
Date
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
21/10/75
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
31/10/75
1/11/75
2/11/75
3/H/75
A/11/75
5/H/75
6/11/75
7/11/75
8/11/75
9/H/75
2/12/75
3/12/75
VI 2/75
5/12/75
6/12/75
7/12/75
8/12/75
9/12/75
% Cl
0.080
0.056
0.052
0.058
0.060
0.068
0.050
0.044
0.046
0.043
0.051
0.055
0.046
0.043
0.038
0.040
0.040
0.050
0.043
0.055
0.042
0.047
0.055
-
0.050
0.047
0.043
0.043
0.044
0.042
0.042
0.044
0.042
0.050
0.046
0.042
0.041
0.040
0.040
0.052
0.041
0.040
% S03
(Total b as)
0.48
0.48
0.45
0.46
0.46
0.45
0.50
0.44
0.44
0.43
0.43
0.43
0.44
0.42
0.41
0.40
0.40
0.40
0.41
0.42
0.43
0.41
0.43
NO SAMPLE
0.42
0.42
0.42
0.43
0.44
0.41
0.43
0.41
0.40
.0.43
0.48
0.48
0.47
0.52
0.51
0.53
0.51
0.51
% K20
0.92
0.91
0.91
0.91
0.92
0-92
0.93
0.92
0.93
0.93
0.93
0.93
0.93
0.93
0.92
0.93
0.93
0.93
0.92
0.93
0.91
0.92
0.93
-
0.91
0.91
0.92
0.93
0.92
0.Q3
0.92
0.92
0.91
0.94
0.90
0.90
0.89
0.91
0.90
0.91
0.91
0.89

-------
TABLE C.2  (CONT'D)
Date
10/12/75
11/12/75
12/12/75
13/12/75
14/12/75
31/12/75
I/ 1/76
2/ 1/76
3/ 1/76
4/ 1/76
5/ 1/76
6/ 1/76
11 1/76
8/ 1/76
9/ 1/76
10/ 1/76
\\l 1/76
12/ 1/76
13/ 1/76
IV 1/76
15/ 1/76
16/ 1/76
17/ 1/76
18/ 1/76
19/ 1/76
20/ 1/76
21/ 1/76
% Cl
0.041
0.039
0.050
0.042
0.048
0.059
0.042
0.045
0.043
0.046
0.038
0.038
0.048
0.039
0.048
0.048
0.048
0.040
0.041
0.044
0.042
0.048
0.041
0.054
0.039
0.040
0.041
% SO,
(Total § as)
0.54
0.52
0.55
0.58
0.54
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
% K20
0.91
0.91
0.91
0.90
0.91
0.90
0.91
0.92
0.91
0.92
0.91
0.91
0.91
0.89
0.92
0.91
0.31
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
Date
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
21/10/75
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
31/10/75
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
9/12/75
% CL
0.99
1.01
0.89
0.96
0.96
J.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
4.39
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
% S03
(Total S as)
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
% K20
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

-------
TABLE C.3  (CONT'D)
                                   114
Date
10/12/75
11/12/75
12/12/75
13/12/75
14/12/75
31/12/75
I/ 1/76
2/ 1/76
3/ 1/76
4/ 1/76
5/ 1/76
6/ 1/76
If 1/76
8/ 1/76
9/ 1/76
10/ 1/76
ll/ 1/76
?2/ 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
% Cl
5.35
5.52
5.53
-
5.68
-
1.98
-
-
A. 55
4.58
4.03
-
3.06
4.50
1.20
1.11
J.01
1.20
0.76
0.92
0.80
0.78
0.69
0.72
0.76
0.86
% so3
(Total S as)
5-51
4.90
5.62
NO SAMPLE
5.54
NO SAMPLE
6.19
NO SAMPLE
NO 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
% K20
8.27
8.20
8.46
-
8.65
_
6.88
_
-
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

-------
                         15
TABLE C.4.  RESULTS FROM ANALYSES OF RETURN DUST
Date
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
21/10/75
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
31/10/75
1/11/75
2/11/75
3/11/75
VII /75
5/H/75
6/11/75
7/H/75
8/11/75
9/11/75
10/11/75
2/12/75
3/12/75
VI 2/75
5/12/75
6/12/75
7/12/75
8/12/75
9/12/75
% Cl
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
0.79
0.58
3.68
% so3
(Total S as)
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
5.37
5-36
4.87
4.84
4.48
4.75
4.24
NO SAMPLE
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
% K20
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
i f* «
4.91
I t** f\
4.92
1 r™ /^
4.50
_ f\ f\.
3.89

10.70
9.63
8_ rt
.58
8«-* **
.33
i f^\
4.94
i _ _
4.93
I x i
4.61
6m
.59

-------
                                   116
TABLE C.4   (CONT'D)
Date
10/12/75
11/12/75
12/12/75
13/12/75
14/12/75
31/12/75
I/ 1/76
2/ 1/76
3/ 1/76
4/ 1/76
5/ 1/76
6/ 1/76
11 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
% Cl
2.51
2.36
3.69
4.04
3.02
1.56
1.78
1.77
2.42
3.55
2.37
2.56
1.35
1.76
2.30
1.06
0.93
0.80
0.79
0.78
0.78
0.65
0.58
0.69
0.58
0.72
0.87
% S03
(Total S as)
3.68
3.45
4.22
4.07
4.06
5.09
5.14
A. 70
4.62
4.48
3.36
3.43
3.23
4.91
3.71
4.74
5.05
5.14
4.82
4.85
4.41
4.27
4.11
4.19
4.56
4.47
4.24
% K20
4.60
4.22
5.94
6.05
5.16
5.60
5-99
5.65
5.82
6.33
4.29
4.64
3.58
5.56
4.91
5.05
5.46
5.37
5.10
4.93
4.59
4.38
4.10
4.22
4.70
4.62
4.83

-------
                                   117
TABLE  C.5.   Btu AND CHLORINE CONTENT OF CHLORINATED HYDROCARBONS (SAMPLES
            FROM LIME TO KILN)
Material
Aliphatic







Aromatic +
Comp 1 ex




Polychlor inated
Biphenyl






Date
24/10/75
27/10/75
28/10/75
3/11/75
A/11/75
13/11/75
18/11/75
Average
10/12/75
11/12/75
12/12/75
13/13/75
15/12/75
Average
3/ 1/76
4/ 1/76
5/ 1/76
6/ 1/76
7/ 1/76
8/ 1/76
9/ 1/76
Average
Btu/!b
13400
H390
10750
8900
8590
8340
8410
9970
9530
9500
8820
9320
9310
9300
11380
11590
11880
12170
12070
12050
12000
11880
Vi scosi ty
(Centipoise Specific
% Chlorine 3 23°C) Gravity
21 .52
32.76
39.40
42.23
43.52
42.38
43.38
37.88 - 1.13
40.56
45-91
44.16
41.80
40.48
42.58 40.8 1.27
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 Cl ANALYSES FROM NO. 6 FUEL OIL
Date
7/10/75
8/10/75
9/10/75
10/10/75
14/10/75
15/10/75
16/10/75
17/10/75
20/10/75
21/10/75
22/10/75
23/10/75
2 VI 0/75
27/10/75
28/10/75
29/10/75
30/10/75
31/10/75
3/11/75
4/11/75
5/11/75
6/11/75
7/11/75
10/11/75
11/11/75
12/11/75
2/12/75
3/12/75
VI 2/75
5/12/75
8/12/75
9/12/75
JO/12/75
11/12/75
12/12/75
15/12/75
30/12/75
2/ 1/76
5/ 1/76
6/ 1/76
7/ 1/76
8/ 1/76
9/ 1/76
12/ 1/76
13/ 1/76
Btu/lb
17809
17222
17978
17942
18106
18008
17920
17988
17924
17877
17984
18189
17701
17901
17862
17970
17917
17889
18135
18077
18047
18079
18036
17979
17782
17719
17700
17676
17714
17648
17679
17732
17695
17596
17655
17504
17678
17565
17498
17433
17630
17900
17719
17990
17896
% S
2.40
2.42
2.34
2.26
2.26
2.28
2.31
2.30
2.24
2.24
2.09
2.23
2.22
2.24
2.22
1.95
2.29
.77
.62
.57
.55
.51
.56
.26
.89
2.06
2.54
2.28
2.56
2.58
2.55
2.52
2.58
2.50
2.56
2.64
2.59
2.41
1.61
2.07
2.28
2.47
2.42
1.96
1.95
% Cl



Average
7/10/75 - 22/10/75
% Chlorine = 0.028%















Average
23/10/75-12/11/75
% Chlorine - 0.064






Average
2/12/75-15/12/75
% Chlorine - 0.047





Average
30/12/75-9/1/76
% Chlorine - 0.030





-------
                                   119
TABLE C.6  (CONT'D)
Date
IV 1/76
15/ 1/76
16/ 1/76
197 1/76
20/ 1/76
21/ 1/76
22/ 1/76
Btu/lb
17935
17952
18025
17762
17846
18005
18160
% S
1.54
2.00
1.98
1.92
1.89
1.54
J.42
% Cl


Average
12/1/76-22/1/76
% Chlorine = 0.







038



-------
                                120
TABLE C.7.  DAILY RECORD OF PRODUCTION AND MATERIALS CONSUMPTION
Date
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
21/10/75
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
31/10/75
1/11/75
2/11/75
3/11/75
4/11/75
5/11/75
6/11/75
7/H/75
8/11/75
9/11/75
2/12/75
3/12/75
A/12/75
5/12/75
6/12/75
7/12/75
8/12/75
9/12/75
SLURRY FEED
(tons/24 hr)
1653
1676
1681
1658
1585
1325
1632
1604
1639
1646
1661
1682
1669
1669
1661
1664
1681
1686
1669
1732
1656
1652
1621
1633
1647
1658
1673
1679
1679
1692
1693
1698
1670
1701
1622
1627
1644
1653
1670
1655
1666
1664
CHLORIN. HC
(gal/24 hr)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1170
1440
1440
1440
1440
1440
1440
2670
2880
2880
2880
2880
930
0
0
0
0
0
430
0
0
1640
2187
1351
1117
2121
CLINKER
(tons/24 hr)
1043
1041
1073
1027
1029
845
1036
1047
1064
1058
1065
1081
1079
1071
1078
1046
1082
1079
1067
1112
1061
1040
1021
1032
1027
1029
1057
1030
1010
1044
1054
1058
1030
1056
1038
1042
1039
1028
1056
1074
1043
984
DISCARD DUST
(tons/24 hr)
40.44
63.44
24.84
65-66
0
18.94
6.96
23.00
0
12.90
17.86
15.32
4.93
15-97
0
45-36
11.96
19.72
20.96
16.42
18.32
42.58
41.36
37.17
56.52
63-32
38.92
80.66
107.72
72.82
60.46
59-90
73.50
64.26
20.58
18.30
37.72
61.40
38.26
0
51.36
129-78

-------
                                   121
TABLE C.7-   (CONT'D)
Date
10/12/75
11/12/75
12/12/75
13/12/75
1V12/75
15/12/75
31/12/75
I/ 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
IV 1/76
15/ 1/76
16/ 1/76
17/ 1/76
18/ 1/76
19/ 1/76
20/ 1/76
21/ 1/76
SLURRY FEED
(tons/24 hr)
1676
1669
1669
1661
1653
1658
1606
1595
1608
1630
1629
1627
1616
1612
1615
1619
1621
1627
1636
1639
1618
1633
1596
1606
1635
1608
1602
1629
CHLORIN. HC
(gal/24 hr)
1997
1308
1821
1691
1869
1909
0
0
0
2029
2210
2168
1985
604
1230
1367
0
0
0
0
0
0
0
0
0
0
0
0
CLINKER
(tons/24 hr)
994
999
1024
985
956
980
1024
1035
1039
1017
1005
970
1034
944
1045
1027
1020
1011
1032
1040
1028
1046
1026
979
1026
1020
991
1016
DISCARD DUST
(tons/24 hr)
127.66
113.76
79.12
125.96
157-68
130.30
24.72
0
6.11
55.43
70.90
116.98
19.86
137.39
4.54
32.10
43-92
61.20
40.52
32.36
42.80
19-10
13.38
85.76
47.62
32.22
66.56
55.75

-------
                               TABLE C.8.  MATERIAL BALANCE FOR CHLORINE

DATE 1
7/10/75 ;
8/10/75
9/10/75
10/10/75
11/10/75
12/10/75
13/10/75
1 VI 0/75
15/10/75
16/10/75
17/10/75
18/10/75
19/10/75
20/10/75
21/10/75
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
31/10/75
1/11/75
2/11/75
3/11/75
4/11/75
5/11/75
6/11/75
CHLORINE
SLURRY
•EED
2,645
,877
,748
,923
,902
,801
,632
,412
,508
,416
,694
,850
,535
,435
,262
,331
,345
,686
,435
,905
,391
,553
,783
,502*
,647
,558
,439
,444
,478
,42)
1,422
INPUT
16
OIL
81
81
82
81
78
67
81
80
81
81
81
81
81
81
80
81
185
183
183
190
182
182
183
181
186
179
179
178
179
184
182
(lb/24 hr)
CHLORINATED
HYDROCARBON TOTAL
0 2,726
0 ,958
0 ,830
0 2,004
0 ,980
0 ,868
0 ,713
0 ,492
0 ,589
0 ,497
0 ,775
0 ,931
0 ,616
0 ,516
0 ,342
0 . ,412
2,845 4,375
3,502 5,371
3,959 5,577
4,416 6,511
4,873 6,446
6,411 8,146
6,642 8,608
12,316 13,999
13,284 15,117
13,284 15,021
13,284 14,902
13,284 14,906
13,743 15,400
4,574 6,179
0 1,604
CHLORINE
CLINKER
688
666
730
836
658
490
622
649
660
656
746
627
690
900
647
607
649
647
619
689
658
707
653
805
698
803
676
659
566
647
653
RETAINED
DISCARD
DUST
801
1,281
442
1,261
0
511
349
754
0
619
672
570
174
575
0
1,887
337
1,215
2,142
1,406
1,469
3,738
3,218
3,970*
6,036*
8,599
5,153
9,437
12,539
4,500
1,632
(lb/24 hr)
TOTAL
1,489
1,947
1,172
2,097
658
1,001
971
1,403
660
1,275
1,418
1,197
864
1,475
647
2,494
986
1,862
2,761
2,095
2,127
4,445
3,871
4,775
6,734
9,402
5,829
10,096
13,105
5,147
2,285
TOTAL %
RETAINED

54.6
99-4
64.0
104.6
33-2
53-6
56.7
94.0
41.5
85-2
79-9
62.0
53.5
97.3
48.2
176.6
22.5
34.7
49-5
32.2
33.0
54.6
45.0
34.1
44.5
62.6
39-1
67.7
85.1
83.3
142.4
* Calculated from average  data.

-------
TABLE 'C. 0  (CONT* 0)

<
DATE
7/H/75
8/11/75
9/H/75
2/12/75
3/12/75
VI 2/75
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
31/12/75
I/ 1/76
2/ 1/76
3/ 1/76
V 1/76
5/ 1/76
6/ 1/76
11 1/76
8/ 1/76
9/ 1/76
107 1/76
ll/ 1/76
12/ 1/76
13/ 1/76
CHLOR
SLURRY
rEED
,494
,403
,701
,492
,367
,348
,322
,336
,721
,366
,331
,374
,302
,669
,395
,587
,895
,340
,447
,402
,499
,236
,228
,548
,260
,554
,556
,562
,309
,344
INE INPUT
#6
OIL
183
181
182
124
133
133
132
131
131
142
132
130
134
134
133
131
85
87
87
83
82
83
83
85
83
81
109
109
109
108
(lb/24 hr)
CHLORINATED
HYDROCARBON
0
0
0
2,325
0
0
8,868
11,826
7,306
6,040
11,470
10,287
7,626
10,213
8,977
10,107
0
0
0
8,657
9,844
8,928
7,950
2,366
5,070
5,468
0
0
0
0

TOTAL
1,677
1,584
1,883
3,941
1,500
1,48?
10,322
13,293
9,158
7,548
12,933
11,791
9,062
12,016
10,505
11,825
1,980
1,427
1,534
10,142
11,425
10,247
9,261
3,999
6,413
7,103
1,665
1,671
1,418
1,452
CHLOR IN
CLINKER
783
597
739
644
604
708
678
612
601
793
787
676
699
655
630
593
594
600
603
671
603
679
641
548
627
575
571
607
619
603
IE RETAINED (l
DISCARD
DUST
1,282
1,102
1,067
2,194
1,515
1,992
8,768
3,642
0
6,410
12,355
13,660
12,559
8,751
14,208
17,906
979
0
242
4,434
6,452
10,715
1,601
10,991
278
2,889
1,054
1,359
818
111
b/24 hr)
TOTAL
2,065
1,699
1,806
2,838
2,119
2,700
9,446
4,254
601
7,203
13,142
14,336
13,258
9,406
14,838
18,499
1,573
600
845
5,105
7,055
11,394
2,242
11,539
905
3,464
1,625
1,966
1,437
1,380
TOTAL %
RETAINED

123-1
107-3
95.9
72.0
141.3
182.3
91.5
32.0
6.6
95.4
101.6
121.6
146.3
78.3
141.2
156.4
79.4
42.0
55-1
50.3
61.8
111.2
24.2
288.5
14.1
48.8
97.6
117.6
101.3
95.0

-------
TABLE C.8  (CONT'D)
CHLORINE INPUT (lb/24 hr)

DATE
I A/
15/
16/
17/
18/
19/
20/
21/
SLURRY
FEED
1/76
1/76
1/76
1/76
1/76
1/76
1/76
1/76
,424
,372
,532
,317
,776
,254
,282
,336
#6
OIL
108
108
107
108
108
107
107
109
CHLORINATED
HYDROCARBON
0
0
0
0
0
0
0
0

TOTAL
1,532
1,480
1,639
1,425
1,874
1.361
1,389
1,445
CHLORINE

CLINKER
617
607
616
724
575
571
575
589
RETAINED (lb/24 hr)
DISCARD
DUST
650
351
214
1,338
657
464
1,012
959

TOTAL
1,267
958
830
2,062
1,232
1,035
1,587
1,548
TOTAL %
RETAINED


82.7
64.7
50.6
144.7
65.7
76.0
114.2
107.1

-------
                                  125
TABLE C.9.   MATERIAL BALANCE FOR K 0
K20 INPUT (tons/24 hr) K?0 RETAINED (tons/24 hr)
DATE
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
21/10/75
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
31/10/75
1/11/75
2/11/75
3/11/75
4/11/75
5/H/75
6/11/75
7/H/75
8/11/75
9/H/75
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
SLURRY FEED
15-21
15-25
15-30
15-09
14.58
12.18
15.18
14.76
15.24
15.31
15.45
15.64
15.52
15.52
15.28
15.48
15.63
15.68
15.35
16.11
15.07
15.20
15.08
1 5 . 02*
14.99
15.09
15.39
15.61
15.45
15.74
15.58
15.62
15.20
15.99
14.60
14.64
14.63
15.04
15.03
15-06
15.16
14.81
15.25
15.19
CLINKER
11.78
11.66
12.23
9-04
13.07
9.13
11.29
12.04
15.64
12.27
13-95
13.40
13-06
13.71
16.28
13-39
11.47
12.95
8.86
3-79
10.61
13.62
13-99
9-91
5.03
3-70
4.44*
4.94
5.66
10.44
12.44
12.70
11.84
12.57
9-76
13-34
10.49
11.92
9-50
9.02
7-72
5.51
3.58
4.80
DUST
2.98
4.73
1.82
5.03
0
1.56
0.78
1.76
0
1.18
1.31
1.11
0.35
1.27*
0
3.55
0.80
1.70
2.11
1.52
1.58
3.78
3.40
2.91*
4.42*
6.09
3.65
6.91
8.88
5.43
3.52
3.28
3.43
3.48
1.93
1.61
2.94
10.00
3.48
0
5.50
0.69
10.56
9.33
TOTAL
14.76
16.39
14.05
14.07
13-07
0.69
2.07
13.80
15.64
13.45
15.26
14.51
13-41
14.98
16.28
16.94
12.27
14.65
10.97
15-31
12.19
17-40
17-39
12.82
9.45
9.79
8.09
11.85
14.54
15.87
15.96
15.98
15.27
16.05
11.69
14.95
13.43
21.92
12.98
9.02
3.22
16.20
14.14
14.13
TOTAL %
RETAINED
97.0
107.5
91.8
93.2
89-6
87.8
79.5
93.5
102.6
87-8
98.8
92.8
86.4
96.5
106.5
109.4
78.5
93-4
71.5
95.0
80.9
114.5
115.3
85.4
63-0
64.9
52.6
75-9
94.1
100.8
102.4
102.3
100.5
100.4
80.1
102.1
91.8
145-7
86.4
59.9
87.2
109.4
92.7
93.0
  * Calculated  from average data.

-------
                                  126
TABLE C.9  (CONT'D)
KjO INPUT (tons/24 hr) K?0 RETAINED (tons/24 hr)
DATE
12/12/75
13/12/75
1 VI 2/75
31/12/75
I/ 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
ll/ 1/76
12/ 1/76
13/ 1/76
IV 1/76
15/ 1/76
16/ 1/76
17/ 1/76
18/ 1/76
19/ 1/76
20/ 1/76
21 / 1/76
SLURRY FEED
15.19
14.95
15-04
14.45
14.51
14.79
14.83
14.99
14.30
14.70
14.67
14.37
14.89
14.75
14.80
14.89
15.08
14.72
14.86
14.52
14.45
14.88
14.63
14.58
14.99
CLINKER
5.53
3-94
4.59
12.29
13.87
14.23
7.83
14.27
7.57
6.00
11.89
8.67
4.72
10.71
11.32
11.87
14.04
13.16
12.76
14.88
13-90
13.03
11.73
14.47
11.38
DUST
6.69
10.71*
13-64
1.70*
0
0.42*
4.01
5.22
8.63
1.37
0.93*
0.33
2.44
2.66
3.54
2.24
1.97
2.21
0.98
0.62
4.01
2.14
1.71
3.45
2.70
TOTAL
12.22
14.55
18.23
13-99
13.87
14.65
11.84
19-49
16.20
7.37
21.82
9.00
7.16
13.37
14.87
14.11
16.01
15.37
13.74
15-50
17.91
15.17
13.44
17.92
14.08
TOTAL %
RETAINED
80.4
97.3
121.2
96.8
95-6
99-0
79.8
130.0
109.4
50.1
148.7
62.6
48.1
90.6
100.4
94.8
106.2
104.4
92.5
106.7
123-9
101.9
91.9
122.9
93-9
* Calculated from average data.

-------
        APPENDIX D
QUALITY OF CEMENT PRODUCED

-------
                                   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 common 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 normal 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;
           C109, 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;
           C151, Test  for Autoclave Expansion of  Portland  Cement;
           C191, Test  for  Time of  Setting of Hydraulic  Cement by Vicat
                Needle.

-------
                                   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 tetraborate followed by x-ray fluorescence
analysis.  This method is commonly used for cement analyses.  With the
exception of lower K~0 content in the cements from clinkers produced while
burning chlorinated hydrocarbons, there were no significant differences
in their chemical compositions (Tables D.I, 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 alkali content of the baseline clinker.
          False Set - False set was severe on the cements from clinker
          produced during 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
          ASTM Method C*»51, 'Test 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 higher alkali cements.
          Compress Ive Strength - Higher initial and lower ultimate
          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
TABLE D.I.  CEMENTS FROM CLINKER PRODUCED  DURING  BASELINE BURN
Sample Number
Chemical Tests
Loss on Igni t ion (%)
Si02 (*)
Al^O- "
Fe20o "
CaO "
MgO "
$03
K20 "
Free CaO (%}
C3S (%)
C2S "
C3A "
Ci,AF "
Physical Tests
Fineness ?
Blaine (cm /g)
Passing 200 mesh (%)
Setting Time
N.C. Penetration (mm)
N.C. Water (%)
Initial (minutes)
Final (minutes)
False Set Q
Temperature F
Penetration (mm)
3 minutes
S minutes
8 minutes
11 minutes
Remix
Compress ive Strength
1 day (psi)
3 day (psi)
7 day (psi)
28 day (psi)
Air Content
Water (%)
Flow (%)
Air (%)
Autoclave Expansion (%)
7

1.52
20.69
5.93
2.20
63.42
2.52
2.74
1.15
0.44
50.1
21.6
12.0
6.7


3406
96.4

10.0
22.5
115
225

70

50
50
21
16
50
2050
2910
3480
4130

73-0
80.6
7.2
0.07
8

1.51
20.59
5.90
2.22
63.39
2.52
2.79
1.23
0.52
50.8
20.8
11.9
6.7


3579
97.6

9.5
23.0
106
216

73

37
7
4
4
50
2010
2990
3610

73.0
86.8
7.6
0.09
9

1.43
20.53
6.00
2.24
63.35
2.52
2.89
1.22
0.54
50.0
21 .2
12.1
6.8


3607
98.8

9.0
23.0
106
221

73

34
8
4
3
50
1910
3210
3880
4530

73.0
86.0
8.3
0.05
Aver .

1.49
20.60
5.94
2.22
63-39
2.52
2.81
1.20
0.50
50.3
21.2
12.0
6.7


3530
97-6

9-5
22.8
109
221

72

40
22
10
8
50
1990
3040
3660
4380

73-0
84.5
7-7
0.07

-------
                              132
TABLE D.2.  CEMENTS FROM CLINKER PRODUCED DURING AROMATIC PLUS
             COMPLEX CHLORINATED HYDROCARBON BURN
Sample Number
Chemical Tests
Loss on Igni tion (%)
Si02 (%)
A1203 (%)
CaO \%)
MgO (*)
so3 U)
K20 (%)
Free CaO (%)
C3S (%)
C2S (%)
C3A (fc)
C^AF (%)
Physical Tests
Fineness -
T
Blaine (cm /g)
Passing 200 mesh (%)
Setting Time
N.C. Penetration (mm)
N.C. Water (%)
Initial (minutes)
Final (minutes)
False Set
o
Temperature F
Penetration (mm)
3 minutes
5 minutes
8 minutes
11 minutes
Remix
Compress ive Strength
1 day (psi)
3 day (psi)
7 day (psi)
28 day (psi)
Air Content
Water (%)
Flow (%)
Air (%)
Autoclave Expansion (%)
4

1.63
20.37
5.85
2.26
63.19
2.51
2.81
0.66
0.50
51.8
19.4
11.7
6.9


3626
98.0

10.0
22.0
118
238

70

50
50
50
50
50
1900
3520
4590
5900

71.0
80.0
9-2
0.03
5

1.86
20.51
5.79
2.23
62.99
2.49
2.69
0.53
0.59
50.8
20.6
11.6
6.8


3562
97.4

10.0
22.0
135
250

72

50
50
50
50
50
1640
3200
4360
6050

72.0
90.5
9-0
0.04
6

1.63
20.83
5-89
2.22
63.56
2.52
2.89
0.69
0.48
49.4
22.5
11.9
6.7


3561
97-2

9-5
22.0
135
250

73

50
50
50
50
50
1890
3180
4160
5460

72.0
81.3
6.7
0.02
Aver.

1.71
20.57
5.84
2.24
63-25
2.51
2.80
0.63
0.52
50.7
20.8
11.7
6.8


3583
97.5

9.8
22.0
129
246

72

50
50
50
50
50
1810
3300
4370
5800

71.6
83-9
8.3
0.03

-------
                                133
TABLE D.3.   CEMENTS FROM CLINKER PRODUCED DURING POLYCHLORINATED
             BIPHENYL BURN
Sample Number
Chemical Tests
Loss on Ignition (%)
Si02 (fc)
A1203 (%)
Fe20o (%)
CaO 1%)
MgO (%)
S03 U)
K20 (%)
Free CaO (%)
c3s (%)
C2S (*)
C3A ($)
^ i A CT i Of 1
V* /i** * \ ^ /
Physical Tests
Fineness 2
Blaine (cm /g)
Passing 200 mesh (%)
Setting Time
N.C. Penetration (mm)
N.C. Water (%)
Initial (minutes)
Final (minutes)
False Set
Temperature F
Penetration (mm)
3 minutes
5 minutes
8 minutes
1 1 minutes
Remix
Compress ive Strength
1 day (psi)
3 day (psi)
7 day (psi)
28 day (psi)
Air Content
Water (%)
Flow (%)
Air (*)
Autoclave Expansion (%)
1

1.46
20.51
5.91
2.16
63.24
2.53
2.94
1.01
0.53
50.4
20.9
12.0
6.6


3388
96.0

10.0
22.0
153
273

70

50
50
50
50
50

1980
3080
3770
4620

71.0
88.0
7-7
0.05
2

1.58
20.72
5.86
2.18
63.27
2.56
2.64
0.74
0.57
50.0
21.8
11.9
6.6


3579
97.0

11.0
22.0
125
250

71

50
50
50
50
50

1800
3000
4000
5HO

71.0
84.5
7.9
0.06
3

1.48
20.87
5.89
2.16
63.65
2.58
2.61
0.96
0.69
50.3
22.0
12.0
6.6


3593
96.8

9-0
22.0
112
235

71

50
50
50
50
50

1990
3260
4090
5080

71.0
82.2
9.2
0.08
Aver.

1-51
20.70
5.89
2.17
63.39
2.56
2.73
0.90
0.60
50.2
21.6
12.0
6.6


3520
96.6

10.0
22.0
130
253

71

50
50
50
50
50

1920
3110
3950
4940

71.0
84.9
8.3
0.06

-------
                  APPENDIX E




EQUIPMENT DESCRIPTION AND ASSOCIATED ECONOMICS

-------
                                  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.
E. 1       Equipment  Description
          The system installed at  the cement plant can vary.  A  simple
 inflexible  installation with manual controls and a used  rail  car for
storage would cost approximately  $25,000.  At  the other  extreme, 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  desired 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 tow viscosity, single-phase liquid,  non-
 corrosive or mildly corrosive to carbon steel, not highly volatile,
 having an approximate composition of:   60,000 Btu/gallon, kQ% 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.

-------
                                  138
          At these compositional values, a feed rate of 1  Igpm into a
1000 ton per day kiln would provide an alkali reduction of 0.^5% as K 0.
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 the provision 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  the  dust  handling equipment.
E.I.I     Cost of basic system
          The basic system with minimum component cost estimates would
entai1:
          Equipment Cost:
          Tank,  10,000 gal carbon  steel          $ 8,000
          Feed Pump,  Centrifugal,  7.5 hp           1,200
          Flame  Arrester                             100
          Vent Scrubber, Activated  Carbon          1,000
          Tank  Level  Indicator                     2,500
          Grounding                                  400
          Tank Berming                             1,000
          Unloading Piping                           500
          Piping, Tank  to  Kiln, 100'               1,000
          Electrical-Combination  Starter           1,500
          Concrete Pad, Site  Preparation           2,500
          Miscellaneous, Hardware
            Painting  and NFPA Code  Marking         2,500
           Instrumentation                            AGO
          Nozzle Assembly                            200
                                                 $22,800
           Installation:                            ^,000
          Total  System  Cost,  Installed           $26.800

-------
                                  139

         A  schematic  representation of the system is shown  in  Figure E.I.
         Operating  Cost:

         Maintenance, b% of investment/year        $1,080
         Operating  Labour, Based on 330 days/year
            Operating  2k hr/day
            1  hr per shift @ $8.00                   8,000
          Electricity, 7-5 hp, 8000 hr
            @ 3
-------
O
6-
    IL
                                  1  CHEMTROL STORAGE TANK
                                  2  PUMP
                                  3  TO  KILN
                                  A  RETURN LINE
                                  5  VENT PIPE
                                  6  LEVEL GAUGE
                                  7  FILLER PIPE
     FIGURE  E.1
SCHEMATIC DIAGRAM  OF  BASIC CHLORINATED
HYDROCARBON FEED  SYSTEM

-------
     1 STORAGE TANK
     2 KILN FEED PUMPS
     3 FILTERS
     4 FLOW METER
     5 CONTROL VALVE
     6 TO KILN  III
 7 FLOW RECORDING CONTROLLER
 8 TO KILN II
 9 TO KILN I
10 FILTER
11 TO ODOR CONTROL SYSTEM
12 BACK TRAP
13  CAUSTIC SCRUBBER
U  CHARCOAL FILTER
15  TRUCK
F.GURE £.2  SCHEMATIC  DIAGRAM  OF CHLORINATED  HYDROCARBON
                                   FACILITIES

-------
          The total installed cost of this system was over $200,000,
comprising the following items:
          Equipment and Installation:
          Tank '                                   $25,000
          Acid Brick Tank Lining                   35,000
          Teflon* Lined Pumps and Filters          12,000
          Teflon* Lined Piping                     45,000
          Odour Control Equipment                    5,000
          Foundations                              10,000
          Dike                                       1,000
          Pump House                               10,000
          Site Clearing and  Leveling                 3,000
           Instrumentation                          20,000
          Painting                                   5,000
          Sales Tax                                  5,000
          Engineering                              18,000
          Consultant                                 8,000
          Travel  Expense                            10.000
                                    TOTAL         $212,000
          Operating  Cost:
          Maintenance  and Operating Labour
             (cost per  year)                        $25,^50
           Electricity  (23
-------
require a reduction in the alkali oxide'  (K 0 + Na_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 noncorrosive wastes  suitable for  use in the basic  system
described above has been adopted.  At present, Chemtrol plans  to  charge
80 percent of the  fuel value  and,  where  applicable, 30 percent of
the chlorine value for these  blended "fuels".   In  the examples, the
following assumptions have been  made:  The  kiln  is  a wet  process  type
producing 1,000  ton?  per day  with  a  fuel  requirement equivalent to
5,150,000 Btu per  ton of production.
E.2.1     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 ICO  content  of  the  clinker  by  0.*5  percent
This  would require approximately 6,800  pounds of chlorine per  day.  A  32
percent  solution of  calcium chloride weighing 13 pounds  and costing  $0.1*
per  Imperial gallon  ($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 "Trol  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.

-------
above mentioned price structure,  the cost per gallon would be:
          '•"""a'TET^  "if                -$0-'*5
                 Btu *           B.u
                                   TOTAL              $0.20
          Thus, the 750 gallons would  cost  $150 and would supply 48.8
mi i I ion Btu.
          The total daily tangible  savings  then become:
          Cost of Calcium Chloride                 $359
          Cost of Normal Fuel  Replaced            $ kS
          Less Cost of "Trol Fuel"                 $150
            Net Savings Per Day                   $258
E.2.2     Process not requiring addition  of chlorine
          The  1,000 ton per kiln and a specific heat  consumption of
5,150,000 Btu  per ton of clinker produced and a fuel  cost equivalent
to $1.00 per million 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 ton of clinker.
          The  fuel normally used on the kiln but  replaced by this waste
stream would cost $0.5)5 per ton of clinker or $515 per day.  The daily
savings can be calculated as:
          Cost of Normal Fuel Replaced            $515
          Cost of 'Trol Fuel"                      412
            Net Daily Savings                     $103
          This saving can be negated by the cost  of discarding the
additional  dust collected in the precipi tators (See Section 6.1), a factor
which will  vary with each cement plant.
E.2.3     Total economic considerations
          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
was 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 @  1 U/year            23,320
            Net                                     $104,190
          This estimated saving,  which would  have given excellent payback
on the system, has not  been realized.   Low  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  kilns:
          Total Savings Using "Trol Fuel"
           (600  kiln operating days @
          $103/day)                                  $61,800
          Maintenance and  Operating Costs             27,290
           Interest on  $212,000  @ 1 U/yr              23,320
             Net                                      $'1,190
           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.

-------
                 APPENDIX F
ONTARIO MINISTRY OF THE ENVIRONMENT EMISSION
      GUIDELINES AND ANALYTICAL SUPPORT

-------
                               APPENDIX F
              ONTARIO MINISTRY OF THE ENVIRONMENT EMISSION
                    GUIDELINES AND ANALYTICAL SUPPORT

F.1        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.
          Pasqui11-Gifford dispersion calculations based on  the guideline
emission  rates  from the wet kiln were used to  predict  the worst-case ground
 level  concentration of residual  organic chloride (i-hour average).
          Table F.I 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 ug/m   (as
 Cl), based  on the  stated guideline,  corresponds to  roughly 5 ppb (v/v)
 of Aroclor  12^2  or  17 ppb  (v/v)  of dichloromethane  (at  ambient temperature),
 the predominant  individually  identified component of  the flue gas^
 These values compare with  the detection limit for PCB's  of 3 vg/m   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.

-------
                                  150

        TABLE  F.I.  MOE  SPECIFICATIONS APPLIED  TO WASTE  BURNS
Waste
Burn
A
B
C
Column
1
0.5
0.5
0.005
2
99.
99.
99-

5
5
995
3
99
99
99

.990
.989
.986
k
5000
5000
50
5*

23-150
16-
88-
150
150
6
0.2
0.2
0.002
7
0.001-0.
0.001-0.
0.004-0.

006
006
006
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).
*4.  Approximate calculated flue gas residual chloride concentration based
    on guideline emission rate (yg Cl/m ).
5.  Approximate measured flue gas organic chloride concentration based on
    dichloromethane and data from Table 5, Section 5-3  (yg Cl/m ).  The
    upper limit of the range corresponds to 50 ppb organic Cl as CH C12-
6.  Calculated point of impingement i-hour average chloride concentration
    based on guideline emission rate  (yg Cl/m ).
7.  Approximate calculated point of impingement i-hour average chloride
    concentration based on dichloromethane and data from Tables 5 and 6,
    Section 5-3 (ug Cl/m3).
*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.I) 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 Pasqui11-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 ug/m
 (as  Aroclor  1242)  which is  7 to  70 times the  typical  measured  ambient
air  concentration of  0.001  to 0.01  yg/m3.  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 dichloromethane.
 Since none of the components of the original  fuel mixture were detected,
 however, the maximum residual concentration of Aroclor 12lt2,  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, 1 ppb
 Aroclor }2k2 in the flue gas (at ambient temperature) would correspond
 to about 11 ug/m3 PCB  (or  about  5  ug/m3 as chloride)  and  about  0.1 kg
 of Aroclor  \2k2 emitted per  day  (2k hours of  continuous  operation).  In
 reality, this quantity  is  an overestimate of  the total emissions and is
 probably not significant.
 P. 2       Gas Chromatographic Analysis of  Process  and Emission  Test Samples.
 p.2.1     Chromosorb  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 desorption 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
temperature of 180°C, and a Sc H electron capture detector.
          The results from these analyses are summarized in  Tables F.2
to F.*».  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,  iow-molecular weight
compounds which,  in  the tables, have been designated 'toajor" and  'toinor",
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 by  relating  the  sum of  the  gas  chromatographic  peak areas with
those  obtained  from  a chloroform  standard.   Their concentrations  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.

-------
TABLE F.2.
                              153
ST. LAWRENCE CEMENT WASTE BURN EXPERIMENTS - TEST 1
CHROMOSORB ADSORPTION TUBE ANALYSIS
Waste Burn
Experiment

A
B
C
Basel ine B
Blank
(Chromosorb)
"Major"
Compounds

3
1
2
2

0
"Minor"
Compounds

1»
5
4
5

0

Chloroform
ppb
1.192
0.058
0.329
0.055

-
Other
Compounds
ppb
7.121
0.646
2.909
2.487

-
Total
Compounds
ppb
8.313
0.704
3-238
2.542

-
TABLE  F.3.   ST.  LAWRENCE CEMENT WASTE BURN EXPERIMENTS -  TEST  2
             CHROMOSORB ADSORPTION TUBE ANALYSIS
Waste Burn
Experiment

A
B
c
Basel ine B
Blank
(Chromosorb)
"Major"
Compounds

4
4
3
0

0
"Minor"
Compounds

3
1
4
7

0

Chloroform
ppb
1.462
1.464
0.924
0.022

•*
Other
Compounds
ppb
8.827
5.270
12.089
0.443


Total
Compounds
ppb
10.289
6.734
13-013
0.465


 TABLE F.4.  ST. LAWRENCE CEMENT WASTE BURN EXPERIMENTS - TEST 3
             CHROMOSORB ADSORPTION TUBE ANALYSIS
Waste Burn
Experiment
A

r
Basel ine B
Blank
(Chromosorb)
"Major"
Compounds
5
3
3
3

0
"Minor"
Compounds
0
3
8
4

0
Chloroform
ppb
2.356
0.481
0.560
0.030


Other
Compounds
ppb
15-773
1.916
2.089
5.077


Total
Compounds
ppb
18.129
2.397
2.649
5.107



-------
F.2.2.1   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 solvent 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 mm  i.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 GC 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 GC data.
          Mass spectral data were  obtained using  a Perkin-Elmer 900 GC
interfaced  to a  HITACHI RMU-6 magnetic mass spectrometer via a  Biemann  -
Watson effusion  separator.   The spectra  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  greater  than
]%  and only nine with  similar area percentages greater  than  0.5%-   There
appear to be only  three major components:   one with  ca.  73%  area, one
with ca. 4.51, and one with ca. 6.8%.   This appeared  to  be  the  case for
all  the aromatic fuels.   A  representative  bar  chromatogram  is  shown in
Figure F.I.  The peak  at  four minutes in the  aromatic  fuel was  ca_.  73%-
           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.I).
           To illustrate this observation,  the GC 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
Sample No.
1
2
3
k
5
6
7
8
9
10
11
12

Chemtrol
u
u
1 1
1 1
u
1 1
Chemtrol
n
n
M
it
Comments
Line Sample (PCB's),
n u n
9
II II II
1
It II II
II II II
II U II
II II II
Line Sample (Aromati
M u n
M u n
n n n
n n M

Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
Jan.
cs),
9
9
9
9

3/76
V76
5/76
6/76
7/76
8/76

, 10:00 AM
, 12:00


, 9:00 AM
, 8:1*5
9/76*
Dec.
Dec.
Dec.
Dec.
Dec.
10/75**
11/75**
12/75**
13/75**
15/75**
 *Sample appeared to have water in it.
**Sample had reacted with liner of sample vial cap.
                TABLE F.6.  GAS CHROMATOGRAPH CONDITIONS
                            HP 5830-A GC
TEMP 1
TIME
RATE
TEMP 2
TIME 2
                                            50
                                           2kO
                                            50  min
 INJ TEMP
 FID TEMP
 FLOW  RATE
 CARRIER GAS
 ATTENUATION
                                           280°
                                           300°
                                            28 ml /min
                                           HELIUM
                                              g
                                             2  256

-------
25
20
15
10
5

25
20
5

25
20
15
10
5

-
-
-
-
-
u,





AROM FUEL



ii ,
—
CLOR
•»
	 i 	






1242

,




r












i i





_


,








r

PCB FUEL
-
-
_
, .


, i i



1



	



I i



. i



,,

ZAREA 529693



f •



i
ZAREA 706087
ill
III 	 	

ZAREA 608448



,
1 i
lll.lli.l 	 	 I
                                                                         U!
          10
20
30
40
50
60
70
                          RETENTION TIME (MIN)
FIGURE Fl  REPRESENTATIVE BAR CHROMATOGRAM

-------
                                  157
treated the data as though it originated from a single sample.   This
computer-produced bar chromatog ram, titled "AROM + CLOR", was then plotted
in comparison with the bar chromatogram of the PCB 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
of the peaks are found in the aromatic  fuel, an estimate of the dilution
or mixing of  these compounds can  be derived.   During  the gas chromatograph
analysis, all variables 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
              4
shown  are  1/10  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  halogenation  and is  most
 likely an  unsaturated hydrocarbon of  high molecular weight.
 F.2.4     Conclusions
           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

20
15
10
< 5
Ul
$
f 25
|
^ 20
15
10
5

"*
PCB FUEL ZAREA 608448

-



, ,


1
.,.11 , 	 lll.ll.l .1 .11 II .III. It Ml 	 	 J






I i i

AROM + CLOR ZAREA 1235780



III III, ,1, 	
10 20 30 40 50 60 7G
                                                                       \n
                                                                       OB
                            RETENTION TIME (MIN)
FIGURE F2 COMPUTER RECONSTRUCTED BAR CHROMATOGRAMS FOR PCB FUEL AND

         AROMATIC FUEL PLUS AROCHLOR 1242

-------
     19
                      15
                         13
                    16
14
                   17
               18
                           11
                            10
                                                      PCB FUEL

                                                      3/1/76
                                                                               vn
                                                                               10
52    48     44    40    36    32
           28    24


           TIME (MIN)
20    16     12
8
    FIGURE F3 GAS CHROMATOGRAM FROM GC/MS ANALYSIS OF SAMPLE PCB FUEL

-------
                                  160
          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-chlorotoiuene
    2          MW 160,  2 chlorines - spectra suggest aromatic compound -
              dichlorotoluene
    3          MW 180,  3 chlorines - spectra suggest trichlorobenzene
    k          MW appears  to  be  216 with  4 chlorines -  possible  identity
              W1*,
    5          MW appears  to  be  216 with  4 chlorines -  spectra very
              similar  to  peak #4
   6-8         MW 222 2 chlorines  -  spectra  suggest dichlorobiphenyls  -
              PCB's
   9-12       MW 256 3 chlorines  -  PCB's
  13-16       MW 290 4 chlorines  -  PCB's
  17 &  18     MW 324 5 chlorines  -  PCB's
   19         MW uncertain - no suggestion  of halogenation

-------
TABLE F.8.  AREA COUNTS  (x  10k)
Peak No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
8 9 10 11 12 1
10/12/75 11/12/75 12/12/75 13/12/75 15/12/75 3/1/76
3860 3976 3769 3994 3842 1200
80.5 82.5 78.8 82.5 80.4 24
20
243.9 247.6 240.6 251.4 245.1 74
363.1 362.1 360.5 373.3 363.4 154
158
76
340
465
82
- - - - 217
543
265
197
287
191
102
52
7.8 8.6 8.4 8.2 8.6 177
.9
.4
.2
.]
.2
.9
.4
.6
.7
.7
.5
.0
.0
.0
.3
.0
.1
.5
2
4/1/76
1142
23-
23.
72.
148.
154.
75.
333.
454.
80.
211.
518.
271.
183.
292.
191.
107.
51.
168.
1
8
3
1
8
7
7
8
6
8
0
3
8
4
1
5
5
9
3
5/1/76
913.4
19-0
38.0
64.5
155.0
160.6
77.0
348.4
451.2
84.9
212.4
503.6
275.5
179.7
302.5
186.8
134.8
54.9
143.8
4
6/1/76
770.6
16.1
48.2
60.6
150.2
156.5
86.5
328.8
431.3
84.3
203-3
481.8
258.3
177.0
284.3
180.2
143.2
50.1
120.5
5
7/1/76
766.
16.
50.
61.
153.
160.
88.
335.
439.
85.
207.
489.
265.
178.
292.
178.
149.
51.
120.
,8
6
8/1/76
753.0
1 15.7
9
7
8
2
6
2
3
6
9
1
9
6
8
6
1
3
5
49.
58.
152.
161.
89.
336.
440.
85.
208.
491.
268.
178.
294.
188.
151.
49-
120.
7
3
0
1
4
7
6
2
9
2
6
3
5
4
4
5
9
7
9/1/76
702.4
14.6
57.8
60.4
157.6
167.8
95.1
351.3
457.2
89. 1
217.1
507.6
282.1
186.7
309.8
190.5
167.5
50.2
H5.9

-------
               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

-------
                                  165
                              APPENDIX G
LABORATORY ANALYSIS RESULTS FROM THE ST. LAWRENCE CEMENT FACILITY TEST

G.1        Summary
          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/m   to 0.009 mg/m .
          - Low molecular weight chlorinated organic compounds such as
            methylene chloride, chloroform, and  carbon tetrachloride 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/m  of flue  gas.
          - Polychlorinated biphenyls  were searched  for  by GC/ECD and GC/MS
            and were  not found  in any  samples at the GC/MS detection  limit
            of 3  ug/m   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  yg/m  ,
 with  the  exception of  lead in  the WBC  tests.   The emission of  lead  during
 the WBC tests  averaged  0.12 mg/m .
          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  dis.card dust samples.  The level of
            detection by  GC/ECD was 5 yg/g of  sample or lower.
           - Polychlorinated 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 0.0^ yg/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

-------
                                  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/m  ,  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 vg/m^  levels.   However,  specific analyses to  identify compounds
below the  level of  interest,  0.1  mg/m , were  not routinely performed.
The  techniques  used were both qualitative and quantitative  in nature, with
an  intended accuracy  range of plus or minus a factor  of two  to three.
G.2        Introduction
           In cooperation with Environment Canada and  the Ministry  of
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 analyzing
 for  input  waste residual compounds and by-products in the  clinker (product)
dust from  the  electrostatic precipitators,  and  the various component
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 samples acquired 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 acquire

-------
                                   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.I.   This table shows what portion of 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.I.  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.3.1.1   Solvent extrarts received  from  ORF.  With a few  exceptions,  solvent
 extractions for organic  species were  performed by ORF using pentane or
 hexane.  Subsequently, i aliquots were received  by  TRW and combined
 according to the plan  shown  in  Figure G.2.  These combinations were
 performed because:
           1)  the  i aliquots  represent  relatively small  gas  sample
               volumes,  thus  making  constituent concentrations very  low
               and  difficult  to  measure,  and

-------
TABLE G.I.  SUMMARY OF SAMPLES RECEIVED FROM EACH WASTE  BURN
Staple Description
iPA Train
• Probe rinse
Filtered Insoluble*
Solvent mtr*ct
Aqueous solution aft*r ftltr«tton Md txtr«ctlon
• Filter
Solvent extract
Filter with perttculate
• Io»1ngtrs
Solvent extract
AqHOOttS Solution *fUr extraction
Otf Train
• In-iuck filter
Solvent extract
Filter with particuUte
t Prate rinse
Filtered tntolubles
Solvent extract
ftqeeoui solution after filtration and extraction
e NaOM Uvtnger
Solvent extract
, Aqueous solution after extraction
e Distilled water Inptnger
Solvent extract
Aqueous solution after extraction
e ChrOBOsorb 102 sorbent tubes
Portion Recelvid
out of Total
Sanple

All
1/4
All
1/4
All
1/4
AH
1/4
All
All
1/4
All
1/4
All
1/4
All
1/4
torn and Test Ho. of Seattle* Origin
IU
1







2







3

X






I



MM
1

X
X









X


X
2

X
X









X


X
3

I
X
X



X




X

I

X
X
UK
1

X
X



X




X

X

X
X
2

















X
3

















X
MJ
1













^



X
2











X




I

1



X



X



X
X

X

X
X
                                                                                         o>
                                                                                         CO

-------
                                               SAMPLE CODES CONSIST OF 3-5 SECTIONS
        Burn
   Sampling Train
Train Component Extracted
Any Special Preparation
     Test No.
First Baseline on
Primary Fuel - BLA
Waste Burn on
Chlorinated
Aromatics      - WBB

Waste Burn on
PCB Blend      - WBC

Second Baseline on'
Primary Fuel - BL>B
Standard EPA
Method 5 Train - EPA
Filter/Insolubles -     F

Probe Rinse -          PR
Caustic Impinger -   NAOH
Water Impinger -      h^O
Combined Impingers -    I
 Extracts of Acidified
 Solutions - AE
ORF Designed
"Sorbent" Train- ORF
Tl. T2. or T3 if
sample Is from
only one test
                                                           Blank if sample
                                                           is a combination
                                                           of all  three
                                                           tests
                    •FOR EXAMPLE, ALL OF THE EXTRACTED SAMPLES FROM ONE  WASTE BURN WOULD BE CODED AS  FOLLOWS:

                    WBB-ORF-FE
                    WBB-ORF-PRE & NAOHE


                    WBB-ORF-NAOH-AE-T3
                    WBB-ORF-H2OE
                    WBB-EPA-FE & PRE


                    WBB-EPA-IE
                       = Waste Burn B, ORF Train, Combined Filter Extracts from
                         all three tests

                       = Waste Burn B, ORF Train, Combined Probe Rinse and Caustic
                         Impinger Extracts from all three tests

                       = Waste Burn B, ORF Train, Extract of the acidified caustic
                         impinger solution from Test 3

                       = Waste Burn B, ORF Train, Combined Water Impinger Extracts
                         from all three tests

                       = Waste Burn B, EPA Train, Combined Filter and Probe Rinse
                         Extracts from all three tests

                       = Waste Burn B, EPA Train, Combined Impinger Extracts from
                         all three tests
                                                                                                                 vx>
                                               FIGURE G.I.   TRW  SAMPLE CODING  SYSTLi!

-------
                                  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 all 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.I).
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 analysis of volatile compounds that would
be  lost  in the next step which was to concentrate the remaining solvent
sample using  Kuderna-Oanish 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  impinger) whose juxtaposition
 in  the sampling  train  and similar physical  characteristics permit the
combination.   The  rinses of  the EPA  train  were  made  of  the glass  probe
 liner  in  front of  the  filter.   However, with  the ORF  train which only
has  a  short  nozzle  in  front  of the  in-stack filter,  rinses were  made  of
 the U feet  of probe  and tubing between the filter  and  the first liquid
 impinger.   The resulting solutions  were filtered through standard filter
 paper  and  then extracted by  ORF.
 G.3.1.2    Solids,  aqueous solutions, and  filters.   The  solid samples
 were prepared for  organic analyses  by extraction in a SoxMet apparatus
 for 2k hours with  disti1led-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 filter for treatment as one
 sample.   These insolubles/fiIter samples were plasma ashed and  then

-------
                                  171
                      TAKE AND
                      COMBINE
                      ALIQUOTS FOR
                      ANALYSIS OF
                      VO LATHES
                         GC
                         GC/MS
                                              SOLVENT
                                              EXTRACT T.
                          COMBINE, CONCENTRATE,
                          AND DILUTE TO 10ml
                                  T
                          TO ANALYSIS OF
                          ORGANIC CONCENTRATES
       FIGURE G.2.   PLAN FOR THE COMBINATION OF ORF  SOLVENT EXTRACTS
                                       VACUUM GAUGE
SORBENT TUBE
d t=



t
FURNACE

=T~W



A
/^
DEWAR
FLASK
FORLK
&=



<

2
^
•V
^


k



•>


~s


vl
,/c
— Q




* C VACU

PLJ 	 Qy— •




          FIGURE G.3.  DESORPTION  SYSTEM FOR CHROMOSORB 102 TUBES

-------
                                 172
extracted by refluxing constant boiling aqua regia over the each sample
for two hours.  The resulting acid digests were made to 50 ml for
analysis.
G.3.1.3   Sorbent tubes.  The Chromosorb 102 sorbent tubes from the
ORF sampling train were prepared for analysis by a quantitative desorption
technique.  The apparatus used for the desorption 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°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_ flask removed,
the sample bulb allowed to equilibrate to room temperature, and the
pressure recorded.
          The volume of the entire manifold 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:
          P V  =  P V
          pri    V2
where:  Pj = the  pressure measured, in mm Hg
        Vj = the volume of the sample bulb and manifold, ^68.8 cc
        ?2 = 760 mm Hg
        V2 = the calculated volume of desorbed material at 1 atm.
          The desorbed material was recovered from the manifold by
reimmersing the sample bulb in LN2«  Valve B was then closed and the
sample bulb removed from the desorption system for analysts 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 Organics
          - Aliquots of the neat solvent extracts
          - Concentrates of the solvent extracts
          - Desorbed materials from the sorbent tubes
G.3.2.1   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 SSMS 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 from  approximately  1.0-0.001  ppm  for
the elements which were  determined,  with an accuracy  between 10-50%.
G.3.2.2    Analyses for organics.  Organic  constituents of the test  samples
were determined by a  combination  of  the  following techniques:
           -  infrared  spectrometry (IR);
           -  gas chroma tography  (GC)  with either an electron  capture (EC)
            or  flame  ionization detector  (FID);
           -  low resolution  mass spectrometry  (LRMS);
           -  combined  gas chromatography/mass spectrometry  (GC/MS).

-------
          The aliquots of the neat solvent extracts were analyzed only
for volatile organochlorine compounds.  EC/GC was used for this analysis
and the samples were compared to standards of CH-C^. CHC1_, and CCl^ as
well as the original waste material.  The column was 183 cm x 0.635 cm o.d.,
glass, 1.5* 0V 17 and 1.95* OJ-1 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 O.OOJ 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,  ROM's,  etc.)  present  as well  as
an idea  of  the complexity and  toxic nature of the  concentrated sample.
           The sensitivity of the LRMS solids probe technique  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 10   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 10  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 GC using both flame ionization and electron capture
detection.  Columns and GC conditions are given below:
          Flame ionization detection
          - Varian  i860, dual differential FID
          - Columns:  dual,  183 cm x 2 mm  i.d., stainless steel,  3-5%
            OV-17 on  100/120 mesh Chromosorb WHP
          - Temperatures:  column, ambient for  5 minutes, then ambient
            to 275°C  at  10°C/min; detector, 275°C;  injector, 250°C.
          - Flow  rate:   helium  carrier at  30 ml/min;  hydrogen at
            30 ml/min.;  air,  300  ml/min.
          - Attenuation:   1  x  10     a/mv
          Electron  capture detection
          - Tracer  MT-150, 63Ni  single ECD
          -  Column:  183 cm  x 0.4 cm i.d., glass,  1.5* OV-17 and  1.95*
             QF-1  on 80/100 mesh Chromosorb WHP
           -  Temperatures:   column,  200°C;  detector, 225°C;  injector 225 C.
          -  Flow rates:   pre-purified N2 carrier through column  at 60 ml/min.;
             detector purge at ^0 ml/min.
           -  Polarizing voltage:  H»V; bucking range -2 x 10  ;  input
             attenuation,  102; output attenuation, X2 to X64.
 G.*»       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  frnm  the  SLC Kiln  Process

           -  Clinker  Products
           -  Electrostatic Precipitator  Discard Dusts

-------
                                   176
G.4.1     Chlorinated hydrocarbon wastes tested
          Samples of 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.4.1.1   Primary fuel.  In addition to the two wastes, a sample of the
primary fuel, bunker "C" oil, was also received.  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 standaid
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:
     E 1 emen t                          Approximate Concentration (ppm)
       Si                                          >500
       P                                            50-500
       Ni                                             5-50
       V*                                            5-50
       Ca                                          <25
       Ti                                            <5
G.A.I. 2   Chlorinated aroma tics.  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  Btu/lb);
          - viscosity -  1.09 centistokes  at 38°C  (100°F);
          - specific  gravity -  1.281 at 16°C  (60°F) ;
          - LOi  - 99.982.
          Elemental analyses performed  showed the following  composition:
                    C
                    H
          -   0.028% N
          -   0.019% S
          -  49.10% Cl (total  halogens as chlorine).
          Analytical  techniques used to determine the organic composition
 included  IR  and  GC/MS.

 15.
          The 1R spectrum indicated the waste to be composed primarily of
 aromatic  hydrocarbons as well as aliphatic alkanes and alkenes wi th a strong
 response  in  the  600-800 cm"'  region which can correspond to C-C1 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
                                                    3
Element is potentially toxic - OSHA TLV of <1 mg/m^ 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
GC/MS
          Chromatographic separation was carried out on a Finnigan GC/MS
system using two columns.  One was Chromosorb 101, temperature programmed
from 30° -  220°C at  10°C/min, and the other was OV-17 temperature programmed
from 30° -  275°C at  10°C/min.  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  ^33% 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 aromatic waste  is chlorotoluene (52.5$) with  the  remainder  made  up
of  three other  chlorinated  aromatics:   dichlorotoluene, octachlorocyclopentene,
and octachloronaphthalene.

      TABLE G.2.   ORGANIC COMPOSITION OF AROMATIC WASTE  BY GC/MS

                                           Estimated  Concentration
               Compound                            (% w/w)
Acetone                                               '-6
Methylacetate                                         °-3
Dichloromethane                                      O.H
Chloroform                                          n-'
Carbon Tetrachloride                                13-9
Dichloroethane                                        3-1
Trichloroethane                                      2-5
Tetrachloroethane                                     °-^
Trichloroethylene                                     1>0
Toluene                                               K0
 Chlorotoluene                                       52.5
 Dichlorotoluene                                      ^-6
 Dimethyl  Benzene (Xylene)                            0-9
 Octachlorocyclopentene                               3-7
 C,rtClQ (Octachloronaphthalene)                        3-0
  10  o

-------
                                   179
SSMS
          Trace metals in the chlorinated aromatic waste were determined
by spark source mass spectroscopy  (SSMS).  The waste was first ashed by
a sulphated 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-SO, added during
ashing).  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 examination of 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
Element
Al
Ca
Cu
Mq
«7
Zn
K
Crb
Pbb
Mn,
Hgb'°
j
Sr
Pt
B
Cob
Ni
W
Mo
a
Approximate
Concentration
(ppm)
0.7-1.5
0.7-1-5
0.7-1.5
0.7-1.5
0.7-1-5
0.6
0.4
0.3
0.07
0.07
0.06
0.05
0.05
0.05
0.03
0.01
0.008
Element
Sn
Zr
Vb
Asb
Ag.
Cdb
Ce
La
Sbb
Beb
Bi
Ge
Li
Rb
Seb
Sm
Yb
Approximate
Concentration
(ppm)
0.008
0.008
0.007
0.006
0.003
0.002
0.002
0.002
0.002
0.001
0.001
0.001
0.001
0.001
0.001
0.001
0.001
 aThe accuracy of this technique ranges from  100 to 500 percent.
 Potentially toxic metals - OSHA TLV of <1 mg/m3 for an eight-hour
  exposure.
 CHg was determined by a highly quantitative  atomic fluorescence
  technique.

-------
                                   180
G.4.1.3   Polychlorinated biphenyls (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 cent!stokes at 38°C (100°F);
          - specific gravity - 1.196 at 16°C (60°F);
          - LOI - 99-90%
          Elemental analyses performed showed the following composition:
          - 57.9**% C
          -  5.23* H
          -  0.018$ N
          -  0.12% S
          - 33.601 Cl  (total  halogens as chlorine).
          The analytical  techniques used to determine the organic composi-
tion of the waste were  IR and GC/MS.
 IR
          The  IR scan showed strong peaks at  1100 cm   and 1^50 - 1^80 cm
which are characteristic of biphenyls, and a peak at 750 cm   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  PCB waste was performed by the same GC/MS
methods and conditions as  for the aromatic waste, and the compounds that
were identified are shown  in Table G.*t.  The PCB waste composition showed
the effect of previous waste material  left in the feed tank.  It was composed
of 12% chlorinated aliphatics, 33% chlorinated aromatics, and k$% 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 Tetrachloride                                 6.8
Ethanol                                              5.2
Dichloroethane                                       0.6
Trichloroethane                                      0.7
Hexachloroethane                                     0.2
Trichloroethylene                                    1.6
Tetrachloroethylene                                  0.2
Xylenes                                              2.3
Toluene                                              1.0
Chlorotoluene                                        17-2
Dichlorotoluenes                                     2.4
Trichlorotoluenes                                    3-3
Octachlorocyclopentene                               4.0
Chlorobiphenyl                                       0.5
Dichlorobiphenyls                                    8.4
Trichlorobiphenyls                                   15-5
Tetrachlorobiphenyls                              11-15
Pentachlorobipheny1s                               2-6
Hexachlorobipheny1s                                1-2
Heptachlorobiphenyls                                 0.5
Octachlorobipheny1s                                  0.2
Di-n-octyl or Di-ethyl  Hexyl  Phthalate               4.0
CI«C'Q " Octachloronaphthalene                       2.5
  10  0
was obtained of which  the  major  elements  were  Fe,  Na,  P,  Pt, and Si.
Other elements detected  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, the 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

                 Approximate3                             Approximate
                Concentration                            Concentration
Element             (ppm)               Element               (ppm)
K
Zn
Cu
S
Mg
Al
Ni
Mo.
Bab
Ti
B
Ca
W .
CrJ
Pbb
vb
Ag
Mn.
Sbb
Sn.
« b
As
Zr.
Cob
7-15
7-15
7
5
5
k
k
3
2
2
1
1
J
0.8
0.8
0.7
0.6
0.5
0.5
0.5
0.2
0.2
0.1
Seb
Srb c
Hgb»C
Cdb
Th
La
Nb
Bi
Li
Rb
Ta
Au
Nd
Hf
Sm
Beb
Dy
Ga
Ge
Pr
Sc
U

0. 1
0.07
0.06
0.05
0.02
0.01
0.01
0.009
0.009
0.006
0.00k
0.003
0.003
0.002
0.002
0.001
0.001
0.001
0.001
0.001
0.001
0.001

aThe  accuracy  of  this  technique  ranges  from 100  to 500  percent.
Potentially toxic metals  -  OSHA TLV  of <1  mg/m3 for  an eight-hour  exposure
cHg was determined by  a  highly quantitative atomic fluorescence
  technique.

from  each train was presented in Table  G.I.  Al 1 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.I]  which was
confirmed by the  analyses at TRW.  The  contamination appeared to have
occurred before  the split of the samples into i 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
WBB and WBC samples.
          All of the test samples were coded for ease of reference.
The coding used was presented  in Figure G.I, 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  A,  and volumes for the EPA train from
Table  7.
           t
G.It.2.1   Organic constituents.  Samples for  the analysis of organic
composition were  in the form of:
          - aliquots of 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 waste,  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

-------
other than CO., HO, and HC1.  The qualitative data will be described
in two groups:  l) 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 summarized and presented  in Table G.6.  This
table includes:
          - The identification of  the extract  samples according  to the
            coding shown  in  Figure G.I.  This  coding shows how certain
            related  samples  have been combined.
          - The Sample/Aliquot 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/Aliquot  Ratio.   This
             value is indicative  of what  may be found  in 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
 oils, fatty acids or fatty acid esters,  and phthalic acid esters.
           These same materials were also present in the blank and control
 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
Baseline Burn A
8LA-ORF-FE-T3
BLA-ORF-PRE+NAOHE-T3
BLA-ORF-H2OE-T3
BLA-EPA-PRE+FE-T3
BLA-EPA-IE-T3
Baseline Burn 8
BLB-ORF-FE
BLB-ORF-PRE+NAOHE
BLB-ORF-H20E
BLB-EPA-PRE+FE
BLB-EPA-IE
Waste Burn 6
WBB-ORF-FE
WBB-ORF-PRE+NAOHE
WBB-ORF-HZOE
WBB-ORF-NAOH-AE-T3
WBB-EPA-PRE+FE
WBB-EPA-IE
Waste Burn C
WBC-ORF-FE
WBC-ORF-PRE+NAOHE
WBC-ORF-H20E
WBC-ORF-NAOH-AE-T1
WBC-EPA-PRE+FE
WBC-EPA-IE
Total Extract -^-^^^
(nin^"^
^--"tRW Aliquot (ml)

200/50
200/50
100/25
300/75
100/25

600/150
600/150
300/75
900/225
300/75

600/150
400/100
200/50
N/AC
900/225
300/75

600/150
600/150
300/75
N/AC
900/225
300/75
Survey Residue
Found 1n TRW
Aliquot (mg)

3.23
N/D*
N/ir
0.12
2.09

0.34
N/D6
0.50
0.48
0.04

0.07
N/Db
N/Db
0.28
1.50
11.18

0.58
N/Db
14.83
0.23
1.12
6.25
Total Residue
(mg, corrected
for Aliquot
Factor)

12.92
N/D (<1.00)
N/D (<2.68)
0.48
8.36

1.36
N/D (<1.00)
2.00
1.92
0.16

0.28
N/D (<1.00)
N/D (<2.68)
1.12
6.00
44.72

2.32
N/D (<1.00)
59.32
OQO
.yd
4.48
25.00
"There  1s no significant difference  In the results between test  runs.  The materials  in the
 residues are not believed to have come from the combustion gas.

bN/D •  Not detected at  levels higher than the blank.

°N/A •  Not applicable;  the total  caustic Implnger contents from  one test were used.

-------
                                  186
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 a combination
of these.  A  large number of blank and control samples might have
distinguished the source, but due to the noncritical nature of the materials,
such an examination was clearly not warranted.
           It was pointed out that the survey analysis of the BLB samples
known to  contain PCB's  IG.1] did, In fact, reveal their presence.  Evidence
of PCB.s  was  clearly shown  in the LRMS data for the following BLB
samples:
           - ORF  train,  combined probe  rinse extracts  and caustic
             impinger extracts;
           - ORF  train,  water  impinger  extracts;
           - EPA  train,  combined probe  rinse extract and filter extract.
Since these PCB's were  found only as estimated minor  constituents  in  samples
whose total weights were  less than  one milligram,  the ability of  the  mass
spectrometer  to  detect  these  small  quantities was  clearly  established.  The
LRMS procedures  can usually detect  10  micrograms,  or  about one percent  of
a  typical  1 milligram 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)  Materials desorbed  from the sorbent  tubes -  The Chromosorb  102 sorbent
tubes from the ORF  sampling train were thermally desorbed to recover the
collected sample according  to the procedure  described in  Section  G.3-1-
The desorbed  gases were then analyzed  qualitatively by LRMS 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 LRMS  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  compounds  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, to C5 hydrocarbons present as a major component
 of these  trace  residual organics could be  present in the sampled
 combustion  gases   in the  1 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,   duantitat ion of the compounds  detected and  identified
 in the test  samples  was performed  by GC and GC/MS techniques.
 1)  P*tract  concentrates -  The  chrbmatography 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
 yg/yl 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

                                   DESORBEO FROM SORBENT TUBE SAMPLES BY LRMS
Sorbent
Tube
Sample
WBB-T1
WBB-T2
WBC-T1
WBC-T2
BLB-T1
BLB-T2
a
Constituent Level
Cl-5
Hydrocarbons
Minor
Major
Major
Major
Major
Major
NUrome thane
Trace
Minor
Minor
Minor
Moderate
Minor
NO » N0« ,
and £
Possibly
Ethanol
Trace
Major
Major
Major
Moderate
Moderate
Benzene
Trace
Moderate
Minor
Moderate
Major
Moderate
Methyl
Chloride
NO"
Moderate
NO
NO
NO
NO
Substituted
Benzene
NO
Trace
Trace
Trace
Trace
Trace
Methyl
Siloxane
Major
Minor
Minor
Trace
Trace
Trace
                                                                                                               QD
                                                                                                               CD
aThe range of trace to major levels represents 0.1-10 ppm concentrations in the flue gas.


bNO - Not Detected.

-------
                                  189
    TABLE  G.8.   SUMMARY  OF THE INTERPRETATION OF LRMS  SPECTRA  FOR
                 TRACE VAPOURS DESORBED FROM SORBENT TUBE  SAMPLES
  Peak Pattern
(Peaks at  Atomic
   Mass Units
     (AMU))                             Assignment
15
16
17,
27,
55,
14,
44
30,
30,
78,
50,
91
147


18
29, 41, 43,
57
16, 28, 32

46, 61
46
50, 51, 52
52, 15

, 207, 281
Methyl group, CH
Methane, CH^
Water
C -C,. hydrocarbons
£. 5
Ni trogen/oxygen
co2
Possibly ni tromethane CH, NO--
NO , NO * and/or ethanol
Benzene
Methyl chloride CH.C1
Substituted benzene ring (e.g., toluene)
Methyl siloxane
*The 30 AMU peak was very  large  and  it  is  believed  that nitric oxide,
 nitrogen dioxide, ethanol  and nitromethane may  have all contributed to it.

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  in Table 6.9 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
in a Kuderna-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 <0.002  yg/yl  of  species as  Aroclor
1232   The  amount  of material in the entire received  sample  volume of  225
ml was determined.   This  value  was  multiplied by four, since  only one-quarter

-------
  TABLE G.9-   RESULTS AND DETECTION LIMITS  FROM  GC/FID ANALYSIS OF THE  CONCENTRATED  EXTRACTS
• Swple ID
ILA-EPA-FE+PRE-T3
-IE-T3
-ORF-FE-T3
-H20C-T3
-PRE+HAOHE-T3
WBB-EPA-FE+PRE
-IE
-ORF-FE
-H90E
-PftE+WWHE
MBC-EPA-FE*rHE
-IE
-ORF-FE
-IbOC
-PRE+NMHE
«9/wl as .
Halocarbon*
0.052
0.26
0.26
0.061
0.21
*D(<0.0038)«
0.89
0.12
0.26
0.053
0.019
(lost)
ND(<0.0038)
ND(<0.0038)
0.26
Volume of .
Sample (ml)0
75
25
50
25
50
225
75
150
50
100
225
75
150
75
150
Aliquot
Factor*
67/20
21/10
46/10
21/10
42/20
213/10
63/10
138/10
42/10
42/10
213/10
63/10
138/10
63/10
138/10
Friction of
SampleC
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
Volume of Sample
Gas. Std. mJ
3.76
3.76
0.32
0.32
0.32
9.62
9.62
0.83
0.83
0.83
9.97
9.97
0.73
0.73
0.73
i
Concentration
In Flue Gas
(•a/-3)
0.001
0.003
0.04
0.09
0.06
ND(<0. 00002)
0.004
0.006
0.02
0.003
0.00008
(lost)
N0(<0.00002)
ND(<0. 00002)
0.015
*Vo1uM of sample received by TRW.

b Aliquot  Factor:  for example 67 ml of th* 75 »1 BLA-EPA-Fe+PRE sample received was concentrated to 20 ml for analysis.

°Fract1on of Sample:   All samples received by TRU were 251 of  the total sample.
     Instrument was calibrated with Aroclor 1232.  However,  the A rod or peak pattern
 was not found 1n any sample  chromatogram.   High sensitivity GC/MS (3pg/m3 of flue
 gas) did not detect any chlorinated species In samples WBC-ORF-FE and
 UBC-ORF-H20E.

"»'• Not detected.  Values 1n parentheses  Indicate the detection limits.

-------
                                    191
of the whole sample was divided by  the volume,  in m  , of gas sampled.
A concentration value of <0.009 mg/m   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/m .   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 detection were performed
to determine polychlorinated 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/yl
          - Naphthalene:   0.066 ug/vl
          - Benzene:   0.11 yg/yl.
          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/m3,  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-H2OE, 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  trimethylsi lyl compounds
 (IMS).   The GC/MS analyses  of the water  extract showed extremely small
 hydrocarbon peaks and 16  peaks that  were trimethylsi lyl 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.10.   RESULTS  AND  DETECTION  LIMITS  FROM  GC/ECD ANALSIS OF  THE  CONCENTRATED  EXTRACTS
Sample ID
M.A-EPA-FE+PRE-T3
-IE-T3
-ORF-FE-T3
-M20E-T3
-PRE+NMHE-T3
UBB-EPA-FE+PRE
-IE
-OKF-FE
-H20E
-PRE+NAOHE
WBC-EPA-FE+PRE
-IE
-ORF-FE
-H20E
-PRMMOHE
M9/u1 as
Benzene"
ND(<0.002)e
N0<
NO
NO
ND










<0.002
<0.002
<0.002
<0.002
.
-
-
-
-
_
-
-
-
"*














ufl/iil as .
Halocarbon0

.
.
.
-
ND
ND
ND
NO
ND
NO
NO
NO
ND
NO
<0.002
<0.002
<0.002
<0.002
<0.002
<0.002
<0.002
<0.002
<0.002









<0.002)
Volume of
Sample, ml4
75
25
50
25
50
225
75
150
50
100
225
75
150
75
150
Aliquot
Factor"
67/20
21/10
46/10
21/10
42/20
213/10
63/10
138/10
42/10
92/10
213/10
63/10
138/10
63/10
138/10
Fraction of
SampleC
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
1/4
Volume of
Sample Gas.
Std. m3
3.76
3.76
0.32
0.32
0.32
9.62
9.62
0.83
0.83
0.93
9.97
9.97
0.73
0.73
0.73
Detectable
limits of
Concentration
In Flue Gas,
mg/m3
ND(<0.05)
NO
ND
ND
ND
ND
ND
ND
ND
NO
NO
NO
ND
NO
NO
<0.03)
<0.3)
<0.3)
<0.6)
<0.009)
<0.01)


<0.1)
<0.009)
<0.01)
<0.1)

 ND • Not detected, values  1n parentheses Indicate the detection limits.

-------
                                    193
These GC/MS analyses did not detect any  chlorinated  species  in
either sample  (typical sensitivity of  3  yg/m  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
Tracer instrument.  The column was operated  isothermally at  70°C to increase
resolution of  low boiling compounds.   The  results are given  in Table G.ll.
          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/m  ), and no  compound identification by GC/MS was
performed.
3)  Materials  desorbed from sorbent tubes  - The sorbent traps were desorbed
as discussed in Section G.3-1.   Portions of  the desorbed vapours contained
in  the sample  bulbs were chromatographed 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 Cl:  WBB Test 1, 0.0001 mg/m3; and  WBB  Test  2, 0.0002 mg/m3.  The
waste burn C samples were essentially  the  same, qualitatively and
quantitatively as the  blank and  baseline B samples.
G.4.2.2   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 (including
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  BLB 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.12.  The blank

-------
       TABLE G.I1.  RESULTS AND DETECTION LIMITS FROM GC/ECD ANALYSIS OF THE UNCONCENTRATED EXTRACTS

BLA-EPA-FE+PRE-T3
-IE-T3
-ORF-FE-T3
-H2OE-T3
-PRE-NAOHE-T3
WBB-EPA-FE+PRE
-IE
-ORF-FE
-H2OE
-PRE+NAOHE
WBC-EPA-FE+PRE
-IE
-ORF-FE
-H2OE
-PRE+NAOHE
ng/yl as
Halocarbonc
ND(<0.0038)d
ND(<0.0038)
ND(<0.0038)
0.0051
N0(<0.0038)
ND(<0.0038)
0.40
ND(<0.0038)
ND(<0.0038)
ND(<0.0038)
ND(<0.0038)
0.10
ND(<0.0038)
0.15
ND(<0.0038)
Volume of
Sample
75
25
50
25
50
225
75
150
50
100
225
75
150
75
150
Fraction, of
Samp 1 e
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
Vo 1 ume
of Sample,
Gas, Std. m
3.76
3-76
0.32
0.32
0.32
9.62
9.62
0.83
0.83
0.83
9-97
9.97
0.73
0.73
0.73
Concentration
in Flue-Gas,
mg/m
ND(<0.0003)
ND(<0.0001)
ND(<0.002)
0.0016
ND(<0.0024)
ND(<0. 000*0
0.012
ND(<0.003)
ND(<0.0009)
ND(<0.0018)
ND(<0.0003)
0.0030
ND(<0.0030
0.061
ND(<0.0031)
Volume of sample received by TRW.

Fraction of sample.  AH samples received by TRW were 25% of the total sample.

The instrument was calibrated with Aroclor 1232.  However, the Aroclor peak pattern was not found
in any sample chromatogram.  High sensitivity GC/MS (3 yg/m^ of flue gas) did not detect any chlorinated
species in samples WBC-ORF-FE and WBC-ORF-H20E.
NO
Not detected.  Values in parentheses indicate detection limits.

-------
                                   195
TABLE G.12.  TRACE METAL SEMI-QUANTITATIVE3 SURVEY OF FILTER DIGESTS BY ICPOES
                                Average Concentratiorr (mg/m^)
Element
Al
Ba
B
Ca
Cd
Cr
Co
Cu
Fe
Pb
Mg
Mn
Ni
P
K
Si
Ag
Na
HO
Sr
Ti
v
Zn
WBB
0.078
0.011
0.018
2.2
0.0006
0.003
NDC
0.003
0.18
0.01*7
0.076
0.003
0.003
0.014
14.
0.082
0.0003
1.8
0.004
0.002
0.0006
0.014
WBC
0.51
0.014
0.084
11.
0.002
0.006
0.001
0.007
1.1
0.12
0.38
0.018
0.006
0.055
32.
0.051
0.003
5-5
0.013
0.008
0.002
0.042
BLB
0.26
0.012
0.038
6.5
0.001
0.004
NDC
0.003
0.73
0.044
0.21
0.013
0.002
0.044
6.7
0.093
0.0004
2.4
0.0008
0.004
0.001
0.027
 aAccuracy estimated to be a factor of ±2 or better.
 Calculated based on average sample gas volumes of:
           WBB - 3-2 mi*
           WBC - 3-3 nC
           BLB - 3-4 nr
     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
                                                   Average Detectable3
                    ICPOES Detection                Limit in Flue Gas
Element               Limit (ppb)                        (mg/m^)
Au
As
Be
Eu
Mo
Se
Te
Sn
W
U
5
40
1
15
11
60
65
50
90
80
0 . 00008
0.0006
0.00002
0.0002
0.0002
0.0009
0.0010
0.0008
0.0014
0.0012
 Based on an average sample gas volume of 3-3 cubic meters.
          The results of this survey  indicated only one element, lead, was
present at potentially toxic levels in the stack.  While not a problem
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 AAS was performed on five elements.  The results from this analysis
are presented in Table G.I4*.
          The AAS results were corrected for background levels as determined
by analysis of the filter blank sample.  It should be noted that back-
ground contributions were not only from trace contaminants  in the filter
materials but also in some cases from matrix effects  in the mixed acid
solutions.
          AAS was also used to analyze for selected elements  in the
aqueous probe rinse and  impinger samples.  Table G.15 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.ll*.  CONCENTRATION  OF  TRACE METALS  IN  EFFLUENT GAS
                  PARTICULATE MATTER  BY AAS
Waste
Burn
WBB


WBC


BLB


Train Test
EPA ]
2
3
EPA ]
2
3
EPA 1
2
3
Concentration
Ba
<0.008a
£0.008
£0.008
0.013
£0.008
0.027
£0.007
£0.008
£0.008
Cd
£0.001
£0.001
0.001
0.001
0.002
0.003
0.001
0.002
0.001
of Element (mg/m )
Co
£0.002
<0.001
£0.002
£0.002
£0.002
£0.002
£0.002
£0.003
£0.002
Cr
<0.009
<0.002
£0.00**
£0.008
£0.008
0.003
0.001
£0.007
£0.002
Pb
0.028
0.012
0.079
0.096
0.103
0.153
0.0*40
0.062
0.015
"£", a less than or equal  to  sign,  indicates  those  elements which were
detected but not significantly above  background  levels.
TABLE G.I5.  CONCENTRATION OF TRACE METALS  IN AQUEOUS SAMPLES BY AAS


Tra i n/Component
EPA-Probe Rinse


Impingers


Waste
Burn
vs.
Basel ine
WBB
WBC
BLB
WBB
WBC
BLB




Concentration of Element (ppm)
Pb
ND
ND
ND
ND
ND
ND
Cr
1.9
0.03
ND
ND
0.02
ND
Cd
ND
0.01
ND
ND
ND
ND
Co
10.2
ND
ND
0.05
ND
ND

-------
                                  198
          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 train was not
operated with the intent of collecting an accurate particulate sample
(i.e., isokinetic, traverses, etc.) the AAS 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  clinker product and discard  dusts for
the  following  reasons.   The  clinker  fines  were recovered from air blown
through  the  clinker  product  and  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 analyses performed  on the clinker
products (CP)  and discard dust  (DD)  samples.   If  significant amounts of
toxic materials were found fn the analyzed samples,  further tests on  the
 remaining samples would have been conducted.   However,  this did not prove
 to be necessary.
 G.4.3.1.   Organic constituents.   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, precleaned  thimble.   In  addition,  a doped
 control  sample consisting of ^30 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 ^30 ppm  in the clinker
 product.  These samples were all analyzed  both qualitatively and quantita-
 tively by the same techniques described in Section G.4.2.1.
 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
Sampl e
Ident i f icat ion
BLB-CP
BLB-DD
WBB-CP
WBB-DD
WBC-CP
WBC-DD
PCB Doped
Control
Weight of
Extracted
Sample (g)
30.270
24.185
29.445
18.948
21 .816
33.112
29.998
Weight of
Residue in
Extract (mg)
< 1 . 1 6*
1.31
1.30
1.16
< 1 . 1 6*
<1. 16"
< 1 . 1 6*
Qualitative Nature of Residue
Hydrocarbon oils, phthalate
esters, traces of fatty acids
and s i 1 icones



Al 1 of the above compounds
plus polychlor i nated biphenyls
 The residue obtained from  the blank  thimble and pentane sample was 1.16
 mg.  Sample values varied  with  some  being  less than 1.16 mg due to the
 variability of the background.  Over  this  range, the differences in
 residue weights are not significant.
are summarized in Table G.16.   In most  cases, the amount of weighable residue
found in the extracts did not exceed  that found in the blank.  The
materials identified by IR  and LRMS are indicative of the low level
contamination by greases, oils,  soaps,  etc., that often accompanies
trace organic analysis.  Hydrocarbon  oils,  phthalate esters, fatty acids,
and silicone compounds were found in  all the extract residues including
the blank.   No indication of the chlorinated species was found in any
of the test samples.
          Po'ychlorinated biphenyls were clearly detected in the PCB
doped control sample.  The  LRMS  spectrum contained peaks at 220, 222,
290, 292, and 294 AMU.  The PCB's present at 30 ppm  in the doped clinker
were easily found and it is certain that much lower  levels could have
been detected.  It  is estimated  that  the mass spectrometer can detect
the strongest peaks from a  specifically searched for compound when
present at the 10 ng level  in a  1 milligram organic  residue from the
clinker.  This results in a lower limit of  detection by this qualitative

-------
                                  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 analyzed by GC with
both FID and ECD 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 ECD.
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 of clinker was doped with 0.9 mg of Aroclor
1232, extracted and concentrated  in the same 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 Tables  G.17 and  G.18.   (The recovery
factor  is the reason  why the values for the samples  in  Table G.17 exceed
several ppm.   If  the  minimum detectable quantity using  the FID  is 0.002
ug/yl,  then  after  applying  the  recovery factor,  the concentration  in  the
sample  cannot exceed  0.008  ug/yl-   For  a 10 ml extract volume and a  30 g
clinker sample,  the  composition is thus <2.7  yg/g-)
G.i».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 material 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
BLB-CP
WBB-CP
WBC-CP
BLB-DD
WBB-DD
WBC-DD
yg/yl as
Halocarbon
ND(<0.002)b
ND(<0.002)
ND(<0.002)
ND(<0.002)
ND(<0.002)
ND(<0.002)
Volume of
Sample (ml)
10
10
10
10
10
10
yg/g of
Sample Material
ND(<3-0)
ND(<3-3)
ND(<1».5)
ND(<3.9)
ND(<5.1)
ND(<2.9)
 The instrument was calibrated with Aroclor  1232.  However, the Aroclor
 peak pattern was not found  in any sample chromatogram.
 ND
Not detected.  Values in parentheses  indicate detection limits.
    TABLE G.18.  RESULTS AND DETECTION  LIMITS  FROM GC/ECD ANALYSIS
Sample ID
BLB-CP
WBB-CP
WBC-CP
BLB-DD
WBB-DD
WBC-DD
CP+PCB
ng/yl asa
Halocarbon
0.059
0.11
ND(
-------
                                 202
      TABLE  G.19.   SELECTED TRACE  METALS  IN  SLC  CLINKER PRODUCT
                   AND DISCARD  DUST  SAMPLES  BY SSMS
Concentration (ppm)
WBB
Element
As
B
Ba
Be
Ce
Co
Cr
Cu
Cs
Dy
Ga
Ge
La
Li
Mn
Mo
Nb
Nd
Ni
Pb
Pr
Rb
Sc
Se
Sm
Sr
Th
Tf
U
V
Y
Zn
Zr
CP
<1.5
30
70b
NDD
10
ND
7
1
ND
1
1
ND
7
ND
100
1
3
10
7
ND
3
7
ND
ND
ND
700
1
700
ND
30
3
30
30
DD
3
30
70
ND
30
ND
10
10
10
ND
1
ND
7
^1000
100
1
3
10
3
70
3
100
1
10
ND
300
7
700
1
30
3
30
30
WBC
CP
<1.5 '
30
300
1
70
ND
30
3
ND
3
3
ND
30
30
300
1
7
10
7
3
10
10
ND
ND
1
700
7
700
3
30
7
70
100
DD
3
10
100
ND
30
3
10
7
30
1
1
ND
10
100
300
1
3
10
7
100
3
300
ND
3
1
700
7
700
3
30
10
100
70
BLB
CP
<2.9
30
70
7
10
1
30
7
ND
1
3
1
10
100
300
3
3
10
10
ND
3
10
7
1
1
700
3
700
1
70
3
30
70
DD
1
30
30
ND
10
ND
10
3
10
ND
10
3
7
100
100
1
3
7
3
30
1
70
3
1
ND
100
1
700
ND
30
3
30
10
SSMS data generally ranges within 500% accuracy.
ND - Not Detected
                      ppm).

-------
                                   203
                              REFERENCES
G.I  Communication, Gordon Thomas, Ontario Research Foundation to
     Arnold Grant, TRW Systems,  7 April,  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 Research Foundation
            Mississauga, Ontario

-------
                                   207
                              APPENDIX  H
    DEVELOPMENT, CONSTRUCTION AND  EVALUATION  OF A  COLLECTION SYSTEM
                 FOR  LOW MOLECULAR WEIGHT  HALOCARBONS

H.1       Summary
          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  (CHC1  )  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, 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  kiIn;
          - the reduction  of a 1kali  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
organic compounds.

-------
                                   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 chlorinated 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  , HO and  Cl  or  HCl  is not  realized, then  trace amounts of volatile
compounds such  as  CC1,,  CHC1. and  CH-Cl. may be present  in  the combustion
gases.   (Though it was  unlikely  that  any HCl or C12 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 recent 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 unaffected
by water vapour, and recovery of adsorbed species  is considered, in
general, to be easier and more efficient.  Recovery of adsorbed species
may be accomplished either by thermal desorption 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  CHC1, and CH CJ .  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
molecular weight species, such as those chlorinated compounds present in
the waste feed material.
H-i*       Test Methodology
          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 form as
Figure A.5 in Appendix  A.
H.A.I     Retention  and collection  studies
H.A.1.1   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.  CHC1_,  CCl^, 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 xylene; 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 chroroatographic

-------
                                   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  1,1,2-trichloroethane  (TCE)  (main constituents of  the  first
waste  feed  in  the waste chloride  study at  SLC) was  placed  in iropinger A
 (see  Figure A.5).   100 ml of pure  toluene  was  placed  in impinger B,
 impinger C  was empty and  impinger  0 contained  silica  gel.  Air was
 pulled through the  sampling  train after  cooling  the impingers  to ice-
 water temperature,  at  a  flow rate of 0.5 cfm.   After  running for four
 hours, the contents of each  impinger were noted  and analyses made for
 EDC and TCE.  At specific times during each test, the parameters normally
 recorded during a source sampling test,  such as  impinger inlet and outlet
 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 EDC and TCE in impinger B,
 impinger C empty and impinger D containing silica gel.  Both experiments
 were then repeated using xylene and decane as absorbents.   Additional
 experiments were made using decane with known quantities of CCl^, CHCl^
 and CHC1, in the standard impinger solution.
 H.l».1.2   Adsorbents.   Though a large number of adsorbents are  capable of
 adsorbing a wide variety of organic compounds, Chromosorb 102 was used for
 the tests described below because  it was readily available, and could be
 purchased without delay from various companies supplying chromatographic
 materials.
            In order  to determine fairly rapidly if Chromosorb  102 would
 adsorb  volatile chlorinated compounds, the following  experiment was
 performed.  100 ml of decane containing known quantities of CCl^,  CHCl^
 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 Chromosorb was removed  from the impinger,  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 H.I.   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
was allowed  to bubble through CHC1- in a midget  impinger and  bleed  into
the main  air  flow  in the duct.  By varying the carrier flow through the
CHC1V  concentrations of 10 to  H*0 ppb of  CHC13 were obtained in  the duct
air stream.   These concentrations were determined both from weight  loss
of  CHC1   with time and direct analysis of  the air flowing in the  duct.
Tests  were made with Chromosorb  in  impingers C and D 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 CHC13  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  CHC13 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   was  more  than 30% efficient  for  the duration of  each  test.   S.nce

-------
      A
                        212
                 1  DUST FEEDER
                 2 BLOWER
                 3 SAMPLING PORTS NON IDEAL LOCATION
                 6  SAMPLING PORTS IDEAL LOCATION
                 5  SAMPLING PORTS
                 6  TO ATMOSPHERE
FIGURE  H.I   TEST   DUCT   SCHEMATIC

-------
                                   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       Col 1ect i on System Eva 1uat i on
          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.4].
          Adsorptlve  capacity values were determined.   Adsorptive  capacity
is  defined as the amount  of  solute vapour retained  by  a given weight of
sorbent.   In this 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 ).
          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  NZ through them (l»0 ml/min)  for four
hours.
           Known concentrations of air-solute vapour mixtures were prepared.
The solutes  examined were CHClj,  CCl^ and dichloroethane (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 evacuated  and then half filled  with  air, at which stage
 the aliquot  from the gas  jar was  injected  into the  bag.  The bag .as
 finally filled to the desired  capacity (6,  20  and 40  litres).

-------
          A schematic diagram of the component parts of the test
assembly  is shown below.
     TEDLAR
      BAG
Gas Jar
          The various parts of the assembly were connected to each other
by means of Teflon  tubing.  Differing air-solute mixture concentrations of
CHC1. were used to  determine the adsorptive capacity of the Chromosorb
102.  The CHC1, 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- was 100 percent.
At various time intervals the pump was stopped and  the gas jar disconnected.
Aliquots (1 ml) were removed from the gas jar and analyzed for CHC1, by
gas chromatography  (GC).  The first detection of CHC1. in the gas jar
was taken as the stare  in the test for the determination of breakthrough
volume, i.e. the volume of air 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 CHCl^ u-»der the
given experimental  conditions.
          From the  data determined with respect to  the adsorptive capacity
of Chromosorb 102 for CHC1- and the knowledge that  a four hour sampling
period at a flow rate of 250 ml/min 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 = -i--  , - • - - — : - rr-i - Z — : —
          K               7   CHC1_ concentration   Volume of air
                               in air sampled         sampled
          The same  test  assembly as described  previously was used to collect
CHC1  on the cartridge.   The adsorbed CHCl^ 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.
           Co]un1n                    - 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                 -  "2  "31 ml/mln
           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 desorption 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 one in which the standard solution was placed.  Estimated solvent
losses were kO-6Q%.  Losses of CC1, , CHC1- 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.4.1.2, 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 summarized below.
Adsorptive Capacity
          Component                            yl/g  (sorbent)
          CHC1.                                    0.1
Desorption Efficiency
  Tedlar bag concentration                 Desorption  Efficiency
  _ of CHC13 _                _ | _
           1.25 ppb                                   90
                                                     96
          5.0 ppb                                    92
                                                    HO
          10.0 ppb                                    9^
                                                    102
 (The  higher  concentration  samples  required a second thermal  desorption
 in order  to  remove all  the solute  from the sorbent.   However,  the first
 desorption was  always  better  than

-------
                                  217
          Two additional experiments were performed.  They were:
          a)  An air-solute mixture of CC1, and DCE (5.00 ppb level;
              40 I 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
              GC to analyze for  the presence of CC1, and/or DCE failed to
              show any trace.  Thermal desorption of the tube followed by
              GC analysis gave good recoveries for CC1, and DCE (>90%).
          b)  An air-solute mixture of CHC1,, CC1, and DCE  (2.5 ppb level;
              ^»0 H 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? and/or HC1.  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  (^ 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
sampling traces of  low  molecular weight  organochlorine pollutants  in  air
than  impinger methods.

-------
                                   218
                              REFERENCES
H.I  Marine Environmental Monitoring of Vulcanus Research Burn II,
     December 2-10, 197*»» Preliminary report, U.S. Environmental
     Protection Agency, December 10,
H.2  Pelllzarri, E.D., Bunch, J.E., and Carpenter, B.H., Env. Sci.
     Technology 
-------
                            APPENDIX I
GC/MS/COMPUTER DETERMINATION OF CHLORINATED HYDROCARBONS AND PCB'S
                        Chemistry Division
                 Air Pollution Control Directorate
                 Environmental Protection Service
                        Environment Canada
                            Submitted by
                            Dr. R.C. Lao

-------
                                   221
                              APPENDIX I
  GC/MS/COMPUTER DETERMINATION OF CHLORINATED HYDROCARBONS AND PCB's

          In February 19705 30 extracted samples were received from
Dr. G.  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-flame  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
Analabs, North Haven,  Conn., U.S.A.
GC-FID;  A Perkin-Elmer  990 model GC-FID with  a  datasystem  PGP-1 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" O.D.   stainless  steel
           Column packing           6fc Apiezon L on chromosorb W 80/100 mesh
           Column temperature       initial 100°C programmed at  *»°C/min.
                                     to final  temp.  200°C and hold
           Injection temperature    250 C
           Manifold temperature     250 C
           Carrier gas              Helium *»0 ml/min.
 GC/MS/Computer:   A Finnigan 1015 D GC/MS system was used with a data system
 6000 series.  Its performance has been studied and  documented in the same
 paper  as mentioned above.    It also  includes standard PCB mass spectra and
 computed reconstructed Aroclor  chromatograms.  The  instrumental data
 are as follows:

-------
                                 222
               Finnlgan 1015D GC-MS Instrumental Data
                                  A
                        Instrumental Data GLC
         Column                         12' x i" all glass
         Column packing                 6% Apiezon L  80/100
                                         mesh chromosorb W
         Column temperature             225 C
          Injection temperature          250 C
         Carrier gas                    helium  kO ml/min.
         Sample size                    5  to 7  microlitres per  injection

                                   B
                         Instrumental  Data  Ms
          Filament  current               100 microamperes
          Electron  energy                 70 or  20 eV
                                                  "3    2
          Operating pressure             6.68 x 10   N/m
                                          (5 x  10"6 torr)
          Scanning  speed                 4  seconds
          Standard  deviation of
           spectra                        maximum 570

          After the quadruple HS operating  parameters were adjusted the
sample was injected Into the GC.  The ion abundance chromatogram of the
GC effluent was acquired by scanning the mass  range (40 to kQQ).  The
dialogue required for mass spectrometer control, data acquisition and
obtaining the plot  are given by software programs.  At the end of the
GC run the computer plots a reconstructed gas chromatogram (ion abundance
chromatogram) of total  ion amplitude versus the spectrum number.
Identification of these chromatographic peaks can be accomplished by
plotting the mass spectrum of a specified peak  or by a  limited mass
plot chromatogram which is obtained under computer control by searching
through the collected spectra and  identifying spectra containing  ions
with a specific m/e_ value.

-------
                                   223
Result and Discussion
          (All chroma tograms 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 12^2, 125**,
and 1260 were made.  By focusing on  specific mass/charge  ratio  (m/e)
peaks such as 290  (tetra-chloro biphenyls),  32A  (penta-chloro)  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 PCB's were
found  in  the samples.   If they are present the concentrations are  below
the sensitivity  limit  of  GC-FID of 1 ng or less.
          Mass  spectra for the GC peaks of spectrum numbers 3&5 and ^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 1 ng.

  yo 1545
  SVJ-147c

-------