MINIPILOT SOLAR SYSTEM:

            Design/Operation of System and
          Results of Non-Solar Testing at MRI
                        by
                   Paul Gorman
                      Ed Ball
                    John Jones
                  Pam Murowchick

              Midwest Research Institute
              Kansas City, Missouri 64110
             EPA Contract No. 68-DO-0137
               Work Assignment No. 18
            EPA Technical Program Manager
                     C.C. Lee
Waste Minimization, Destruction & Disposal Research Division
         Risk Reduction Engineering Laboratory
                Cincinnati, Ohio 45268
     U.S. ENVIRONMENTAL PROTECTION AGENCY
      OFFICE OF RESEARCH AND DEVELOPMENT
     RISK REDUCTION ENGINEERING LABORATORY
              CINCINNATI, OHIO 45268

-------
                                DISCLAIMER
      This material has been funded wholly or in part by the United States
Environmental Protection Agency under Contract No. 68-DO-0137 to Midwest
Research Institute.  It has been subject to the Agency's review, and it has been
approved for publication as an EPA document.  Mention of trade names or commercial
products does not constitute endorsement or recommendation for use.

-------
                                 FOREWORD
      Today's rapidly developing and changing technologies and industrial products
and practices frequently carry with them the increased generation of materials that, if
improperly dealt with, can threaten both public health and the environment. The U.S.
Environmental Protection Agency is charged by Congress with protecting the nation's
land, air, and water resources. Under a mandate of national environmental laws, the
agency strives to formulate and implement actions leading to a compatible balance
between human activities and the ability of natural systems to support and nurture life.
These laws direct the EPA to perform research to define our environmental problems,
measure the impacts, and search for solutions.

      The Risk Reduction Engineering  Laboratory is responsible for planning,
implementing, and managing research, development, and demonstration programs to
provide an authoritative, defensible engineering basis in support  of the policies,
programs, and regulations of the EPA with respect to drinking water, wastewater,
pesticides, toxic substances, solid and  hazardous wastes, and Superfund-related
activities. This publication is one of the products of that research and provides a vital
communication link between the researcher and the  user community.

      This report was prepared to summarize work  performed by Midwest Research
Institute (MRI) for the United States Environmental Protection Agency (U.S. EPA),
Risk Reduction Engineering Laboratory (RREL). The work involved design,
construction, and operational testing of a Minipilot Solar System, followed by a series
of non-solar tests at MRI using EPA sampling and analysis methods. The work at MRI
was done to provide an operational system and non-solar test results in preparation
for transporting the system to the National Renewable Energy Laboratory (NREL) and
conducting tests with solar input.
                                           E. Timothy Oppelt, Director
                                           Risk Reduction Engineering Laboratory
                                       in

-------
                                   ABSTRACT

       Prior to this project, MRI had carried out work for the Environmental Protection
 Agency (EPA) on the conceptual design of a solar system for solid waste disposal1
 and a follow-on  project to study the feasibility of bench-scale testing of desorption of
 organics from soil with destruction in a solar furnace.2  The feasibility study involved a
 two-step process:  (1) thermal desorption of organics from soil (non-solar) with
 recovery of the organics in liquid form and (2) destruction of the organics in a solar
 furnace.

       Previous  laboratory work by the National Renewable Energy Laboratory (NREL)
 and others indicated efficient destruction of organics by solar input at high
 temperatures, but the experiments were limited to low concentrations of organic
 vapors in  air.3 The feasibility study conducted by MRI recommended the important
 difference that the organics should be fed to the solar furnace in liquid form, to provide
 destruction through combustion and solar photolytic effects.

       Based on the feasibility study, EPA directed MRI to proceed with design,
 construction, and testing of a Minipilot Solar Reactor System with liquid organic feed.
Testing of the system on this project was intended to be carried out in two parts.  The
first  part was to be non-solar testing at MRI using a synthetically prepared liquid
organic mixture.  The primary purpose of the  non-solar testing at MRI was to utilize
EPA sampling and analysis methods and determine if the organic emissions
quantitated by those methods would enable quantitation of the expected  significantly
lower emissions  when operating with solar input. A secondary purpose of the testing
at MRI was to conduct tests with a UV lamp to determine if  this provided any
                                        IV

-------
 significant reduction in emissions.  The second part of the project was to carry out
 subsequent testing with solar input at NREL.

       MR! proceeded with design and construction of the Minipilot System, and
 included complete instrumentation with computerized data logging, flame detectors,
 automatic shutdowns, and continuous monitors for CO, O2, and THC. NREL staff
 contributed valuable advice that aided in the development of the design.

       Extensive operational tests were conducted and resulted in some system
 design and operational changes that improved the system's overall utility.  Operational
 techniques were established for this system during the initial operating tests. The
 operational performance of the Minipilot System was demonstrated during a series of
 10 tests using the EPA-approved  sampling and analysis methods.  This report
 summarizes the results of those tests, as well as the design and construction work
 mentioned above.

       Results of the 10 tests showed that it should be possible to determine if there is
 a significant reduction in emissions (> 3X) when operating with solar input. Such
 reduction in the emissions should be determinable for two of the POHCs contained in
 the synthetic feed liquid (CCI4 and DCB),  for both the volatile and semivoiatile PICs,
 and for dioxins and furans. But reductions are probably not determinable for the other
 two POHCs (toluene and naphthalene).  Results also showed that a three-fold
 reduction in volatile PICs occurred in both single-chamber tests when an artificial UV
 light was utilized.

       The Minipilot System is now fully operational and has been designed to utilize
 the full solar capability of the NREL facility (10 kW) with temperature control via
 different aperture openings (to vary the radiation heat losses) and/or water injection.  It
 has already been operated in the  non-solar mode  over a temperature range of  1400°
'to 2000°F. Further, the baseline levels of the organic emissions have already been

-------
determined using EPA sampling and analysis methods for two different operating
modes (single- and dual-chamber) and two different flow rates. Thus the system is
ready for the next phase:  solar testing at NREL.

      Carrying out  the tests as planned at  NREL is crucial to the overall goal of the
project:  to determine if solar input provides a significant reduction in organic
emissions as compared to emissions in the non-solar mode at the same  operating
conditions (i.e., temperature and feed rates, etc.). Moreover, the tests at NREL are
intended to provide data on the effect of certain parameters that are very important to
the design of larger pilot- or full-scale systems.  Specifically, tests at NREL would:
(1) investigate effects of solar exposure time (over the range of 4 to 8 s), which is
critical to sizing of reactors; and (2) investigate operation at different solar flux levels,
up to 10 kW, to determine the optimum kilowatt input per unit of gas flow rate (scfm),
which is  critical to sizing the costly solar collector/concentrator units.  Considering the
important goals of the solar tests at NREL,  and  results of the non-solar tests that have
already been completed at MRI, it is  recommended that work proceed on transporting
the Minipiiot System to NREL and testing it  with solar input.
                                       VI

-------
                                 CONTENTS
Foreword	 .  iii
Abstract		iv
Figures . .	  viii
Tables	ix
Abbreviations	  x
Acknowledgments	xi

      1.     Introduction	  1
      2.     Design and Construction of the Minipilot System	  3
                  Minipilot System	  4
                  Reactors	;	  8
      3.     Initial Operation	  14
      4.     Preliminary Testing	  17
      5.     EPA Method Test Results	  23
                  Test Conditions	  24
                  EPA Method Test Results  	  29
      6.    Conclusions and Recommendations	  46
                  Conclusions	  46
                  Recommendations	  48
      7.    References	  50

Appendices

      A.    POHC inputs and VOST results with calculation of DREs for
                  CCI4 and toluene		  51
      B.    MM5 results and calculation of DREs for dichlorobenzene
                  and naphthalene	  58
      C.    PIC results for VOST and MM5 samples .  .	  62
      D.    Tabulation of analysis results for dioxins/furans	  77
      E.    Liquid organic feed analysis results for Cl, ash, and HHV  	  79
      F.    O2, CO2, and H2O analysis results for effluent gas in each
                  run	  81
      G.    Summary of quality assurance results	  83
                                      VII

-------
                                   FIGURES

Number

 1    Schematic diagram of the Minipilot Solar System	  5

 2    Solar detoxification reactor-section view	  9

 3    Minipilot-Scale Solar Reactor System (shown: overview of the system
      operating in the two-chamber-nonsolar mode)	  10

 4    Minipilot-Scale Reactor System (shown: close-up of the operating solar
      detoxification system with computerized data acquisition and control)	11
                                       VIII

-------
                                 TABLES



Number                                                              Page



 1     EPA test method results for Runs 17 and 18	  19



 2     Summary of EPA tests	  25



 3     Summary of operating data	  26



 4     Summary of results from Minipilot Solar Reactor Runs 31 through 40	  30



 5     Comparison of blanks with sample results	  32



 6     Comparison of replicate runs (single-chamber mode)	  35



 7     Comparison of non-UV results for different flows and operating modes ....  37



 8     Amount of volatile PICs found in a VOST trap pair sample (ng) .	  42



 9     Amount of semivolatile PICs found in the MM5 samples (jig) .	  44
                                    IX

-------
                              ABBREVIATIONS

NREL       National Renewable Energy Laboratory (Golden, Colorado)
VOST       Volatile Organic Sampling Train
VOA        Volatile Organic Analysis
MM5        Modified Method 5 sampling train
ORE        Destruction and Removal Efficiency (for POHCs)
THC        Total Hydrocarbons
OEMs       Continuous Emission Monitors
HHV        Higher Heating Value
SIM         Selective ion monitoring
UV         Ultraviolet
POHCs     Principal organic hazardous constituents
V-POHCs    Volatile POHCs
SV-POHCs  Semivolatile POHCs
PICs        Products of incomplete combustion
V-PICs      Volatile PICs
SV-PICs     Semivolatile PICs
QAP        Quality Assurance Plan
CCI4        Carbon tetrachloride
DCB        1,2-Dichlorobenzene
GC/ECD     Gas chromatography/electron capture detector
GC/MS      Gas chromatography/mass spectrometry
D/F         Dioxins/Furans
RRF        Relative Response Factor

-------
                            ACKNOWLEDGMENTS

      Dr. C. C. Lee and Mr. George Huffman of EPA/RREL contributed much to this
project.  Their help and support is gratefully acknowledged.

      Throughout the work at MRI, NREL staff (Drs. Greg Glatzmaier and Mark Bonn)
provided valuable assistance in the design of the reactors and the rest of the system,
as well as assisting in the preparation of work plans and many other aspects of the
project.  MRI and NREL staff also participated in quarterly technical review meetings
held under the Tri-Agency Agreement (EPA/DOD/DOE) on solar soil detoxification.
                                      XI

-------
                                   SECTION 1

                                 INTRODUCTION
       Prior to this project on the Minipilot Solar System, MRI carried out related work
 for EPA on "Conceptual Designs of Solar Systems for Solid Waste Destruction"1 and a
 "Feasibility Study for Bench-Scale Testing of Desorption of Organics from Soil with
 Destruction in a Solar Furnace."2 A major conclusion of that previous work was that
 the most feasible concept for solar soil detoxification would:

       1.     Use available conventional processes for low temperature thermal
             desorption of soil with recovery of desorbed organics in liquid form.

       2.     Feed the recovered organic liquid to a solar reactor for destruction of the
             organic components.

       The above concept was believed to provide the most feasible system for
 several reasons.  First, it separates the soil desorption process (non-solar) from the
 solar destruction system so that desorption can be done continuously, and is not
 dependent on solar availability or the need for  rapid shutdown capability if solar input
 is interrupted. It also avoids the  need for large amounts of solar energy that would be
 necessary for thermal treatment of large amounts  of soil. Second, the solar destructor
for the recovered organic liquid can be relatively small (approximately 1/1,400th of the
size needed if fed gaseous contaminants directly from the soil desorption process).
Design of the solar destructor would be based  on the availability of solar input, but
may or may not be based on the desorption rate of organics from the soil, depending
on whether the destructor is located on-site or at a separate off-site processing facility.
                                        1

-------
Third, the solar reactor combines destruction of the organics by both combustion and
photolytic reactions.  Fourth, a solar destructor, with liquid feed only, can be rapidly
shut down if solar input is interrupted.

      The above concept was the basis for the subsequent project to be carried out
by MRI for EPA, which consisted of two parts:

      Part I —     Design, construction,  and non-solar testing of a minipilot-scale
                   liquid feed solar reactor system at MRI.

      Part II —    Operation and testing of the minipilot-scale liquid feed solar
                   reactor system at NREL with  solar input.

      The objective of this report is to summarize the Part I work on the project,
which has been completed.  Associated with the work are three other documents:

      • Test Plan for Testing to  Be Done at MRI  (dated March 20, 1992)
      • Test Plan for Testing to  Be Done at NREL (dated April 2, 1992)
      • Quality Assurance Project Plan for Testing of the Minipilot-Scale Solar
        Reactor (dated March 20, 1992)

      Subsequent sections of this report are organized  as follows:

      Section 2 Design and Construction of the  Minipilot System
      Section 3 Initial Operation
      Section 4 Preliminary Testing
      Section 5 EPA Method Test Results
            •      Test conditions
            •      EPA method test results
      Section 6 Conclusions and Recommendations

-------
                                  SECTION 2

           DESIGN AND CONSTRUCTION OF THE MINIPILOT SYSTEM
       Design of the solar reactors and the rest of the Minipilot System was based on
the size of the solar collector/concentrator furnace at NREL and information provided
by NREL on their previous testing done with a small quartz reactor.

       The solar collector/concentrator can provide about 10 kW of heat (34,000 Btu/h)
in a 4-in diameter beam. This thermal heat input is capable of heating about 10 scfm
of flue gas (mostly air) to 2000°F. The flow rate of flue gas (approximately 10 scfm  of
air) was considered to be a maximum value, and was used to estimate the reactor
volume necessary to provide a residence time of 2 to 3 s in the temperature range of
1400° to 2000°F. NREL suggested a length-to-diameter ratio of 3:1.  On this basis,
the reactor's internal dimensions are 10 1/2  in i.d. x 36 in long.

       The airflow rate of 10 scfm also was used to estimate the feed rate of a liquid
organic, assuming that one-half the oxygen available in an air feed of 10 scfm was
consumed  in the combustion of an organic liquid (i.e., 100% excess air).  Further,  it
was estimated that the liquid organic, recovered from soil desorption, would have a
heating value  on the order of 12,000 Btu/lb. On this basis, the system was designed
to provide a liquid organic feed rate of 1.5 Ib/h (11 g/min).

       The above air feed rate, liquid feed rate, and solar thermal input rate provided
the basic design criteria  from which other system components were specified.  These
design criteria also provided necessary information that served as input for heat

-------
transfer calculations used to determine the thickness of the reactor's internal and
external insulation. Design of the system  and the reactor are discussed in the
following two subsections.

MINIPILOT SYSTEM

      A schematic diagram of the solar reactor system is shown in Figure.1.
Important points about this process design are discussed below.
                                  V"                                 '
      1.     Dual Reactors—Some concern was expressed by NREL that the flame
             associated with combustion of the liquid organic might decrease the
             effectiveness of the solar input for photolytic destruction of organic
             components and/or products of incomplete combustion (PICs).  For that
             reason, a dual-chamber system was designed, so that the solar beam
             could be input to the second chamber, after combustion in the first
             chamber. However, the system could also be operated in a single-
             chamber mode, with combustion and solar input in the  same chamber.

      2.     Auxiliary Fuel—The system was designed to use propane as an auxiliary
             fuel to provide a continuous  pilot flame (for safety reasons) and to
             provide initial heatup of the reactors prior to introduction of any liquid
             organic feed.  It was also designed to fire fuel oil as an auxiliary fuel and
             for baseline testing with a liquid feed that does not contain any chlorine
             (Cl) or hazardous organic constituents.

      3.     Water Feed—When the liquid organic feed undergoes combustion, the
             heat released by this combustion,  combined with the solar thermal input,
             produces excessive temperature (> 2000°F). The temperature could be
             controlled by removal of external insulation from the reactor or by
             different size apertures on the quartz window to regulate  radiant heat

-------
o>
t
 i
o
03
 g.
'c

2

 
-------
      losses through the window.  However, these are physical changes to the
      system that are somewhat difficult to make and do not provide
      temperature "control" capability.  Therefore, the system was designed to
      provide water injection for temperature control.

      It was suggested that H2O injection might affect photolytic destruction
      reactions, but it was not known if this effect would cause an increase or
      decrease in effectiveness of destruction  (or no effect at all). Thus the
      system was designed to be able to provide H2O injection for temperature
      control, with the intent to perform tests to ascertain if it had any
      significant effect on destruction of the organics. (Subsequent non-solar
      tests at MRI indicated no detrimental effect from H2O injection, but the
      tests could be repeated later at NREL with solar input.)

4.     Compressed Air—Compressed air (- 90 psig) is supplied to the system
      for atomization of liquid feeds and serves as the source  of combustion
      air. The use of compressed air facilitates flow monitoring and control.
      Most of the compressed air (referred to as "secondary air") is  introduced
      tangentially into the reactor in order to swirl the gases and maximize
      turbulence.  "Primary air," a small fraction of the total air fed to the
      reactor, enters as air premixed with pilot fuel (propane).  A very small
      amount of air is also used for atomization of any liquid feeds.

5.     Reactor Exit Piping—Reactor exit piping was constructed of 40 ft of 2-in
      stainless steel tubing. Two 2-in tees were provided near the inlet to the
      final cooler for EPA-method sampling. Sections of removable insulation
      were installed, as needed, to maintain the temperature at the sampling
      points  in the range of 250° to 400°F.

-------
6.    Final Cooler—The final cooler was designed for removal of 30,000 Btu/h
      in order to cool the exit gases to < 100°F, which is necessary for
      condensation of H2O produced by combustion and any H2O  injected for
      temperature control.  It was designed to provide the heat transfer duty
      using coolant available at NREL, which has a flow limit of 8 gal/min. The
      material of construction for the final cooler was type 304 stainless steel.

  Although the final cooler has not failed (leaked), there is evidence of
  corrosion. Replacement with a heat exchanger constructed of Hasteloy C-22
  is being considered for better corrosion resistance (combustion gases contain
  up to 8,000 ppm HCI).

7.    Gas Cleanup—Final gas cleanup is accomplished in three 6-in Plexiglas
      vessels.  The first vessel is a water separator.  The second vessel
      contains a packed bed of AI2O3  pellets impregnated with NaOH for HCI
      removal.  Originally it was intended that the third vessel contain a bed of
      activated carbon for adsorption of any residual hydrocarbons. This
      carbon bed was replaced with the AI2O3 pellets to maximize  HCI
      removal, since the THC monitor  indicated low THC levels during
      operation  (i.e., usually below 1 ppm).  However, it is planned to install a
      carbon canister in the effluent gas discharge line from the vacuum pump.

8.    Vacuum Pump—The  final vacuum pump enables operation of the
      reactors below atmospheric pressure (-1 to -5  in of H2O).  Pressure
      drop through the system, especially the final effluent flowmeter, limits
      capacity of the system to near 10 scfm. Contributing to that pressure
      drop were cartridge filters that had to be installed to protect the pump
      from very  fine particulate matter  and HCI mist.

-------
       9.     CEMs—Continuous emission monitors were installed for CO, O2, and
             THC.  The sampling location for these extractive-type monitors was
             originally intended to be in the 2-in tubing near the inlet to the final
             cooler. However,  it quickly became obvious that the high HCI
             concentration at that location was severely detrimental to the monitors.
             Therefore, the sampling location for the CO and O2 monitors was
             changed to the exit from the first bed of AI2O3 pellets.  The sampling
             location for the THC monitors was moved to the exit of the H2O
             separator, since it  was not known if the pellets might be removing some
             of the hydrocarbons.

       10.    UV Light—A UV light was purchased and an air-cooling chamber built
             that could be installed in front of the quartz window whenever needed.
             This UV light is a Hanovia high  intensity, medium pressure mercury arc
             lamp operating at 300 W/in with a 6-in bulb. Approximately 300 W of the
             lamp's total output  is in the ultraviolet range (240 to 365 nm).

        Tests were subsequently carried out  at MRI to assess the effect of UV light
        on destruction of organics.  It should be noted that heat losses through the
        quartz window (with or without UV light) caused significant cooling of the
        gases (e.g., from 1400° to 1060°F).
REACTORS
      Two identical reactors were constructed so that testing could be carried out with
only one reactor, or with the two reactors connected in series (dual chamber mode),
for the reason discussed earlier. A diagram of the reactor is given in Figure 2 and is
shown in the photographs in  Figures 3 and 4.
                                       8

-------

-------

                                            0)
                                            TJ
                                            O

                                            £
                                            o
                                            CO

                                            o
                                            0
                                          o —
                                         co o)
2

g.
                                         W
                                         1-1
                                          CD
                                             0}

                                             o
                                             c


                                             o
                                            JZ
                                             03
10

-------

-------
      The reactors were constructed of 330 stainless steel for maximum temperature
capability.  The stainless steel shell is insulated with a nominal 4-in thickness of
alumina-silica ceramic fiber blanket (Zircar).  The internal insulation includes
approximately 3 in of the alumina-silica ceramic fiber blanket, with a solid 1/£-in-thick
alumina insulating cylinder forming the inner chamber. The alumina insulating cylinder
temperature rating is 3000°F. Dimensions of the chamber are 101/a in inside diameter
x 36 in long.

      A flange assembly was designed at one end of the reactor for installation of the
13.875-in  diameter x 0.50-in-thick quartz window for solar or UV input.  Graphite
gaskets (Grafoil) were used to seal the quartz window, as well as all the other reactor
flanges.  Design of the reactors included various size aperture openings (type 330
stainless steel with  1.0 in of insulation) inside the quartz window. These aperture
openings  can be  used to change the radiant heat losses through the quartz window,
thus providing some control on temperature of the exit gases.

      There is a  small opening in the top for installation of a "flame rod" for detection
of the propane  pilot flame.  A 1-in diameter pipe was also designed, aimed at the
center propane burner, for installation of a flickering infrared flame detector.

      As  mentioned earlier, a tangential air entry port was provided in the design.
Several other flange openings were designed into the system that are used for quartz
obseivation windows or can be  used for other purposes.  Also, a Vz-ln connection in
the reactor exit pipe provides access for a thermocouple to monitor and control exit
gas temperature.

      Each reactor was designed to provide insertion of up to five 1-in diameter,
water-cooled nozzles, as follows:
                                       12

-------
      • Propane burner (center)
      • Liquid organic nozzle
      • Fuel oil nozzle
      • Water nozzle
      • Spare nozzle port

      Only two types of water-cooled nozzles were designed and built for the
reactors.  The first type consists of an internal 0.5-in tube with a flame stabilizer on the
end.  Propane and air are fed to this nozzle to provide a pilot flame.  The second type
of water-cooled nozzle is  an MRI design that provides air atomization of the fluid
                                                           \
(liquid organic,  fuel oil, or water) with only a very small flow rate of atomizing  air.

      Originally, the 1-in  outer tube of the nozzles, which contains the cooling water,
was constructed of 316 stainless steel.  After several hours of operation, it was found
that the portion of the 1-in tubes inside the reactor's internal insulation was severely
corroded, apparently due  to condensation of H2O and associated absorption of HCI
from the combustion gases in contact with the cool metal surface. New nozzles were
built using Hasteloy C-22  tubing to better resist this corrosion.
                                        13

-------
                                  SECTION 3

                             INITIAL OPERATION
      Work proceeded on assembly of the reactor and the rest of the system, and
initial operation began on February 16, 1992, using propane and fuel oil only.  These
initial tests went well, with good performance of the air atomized nozzle for the fuel oil.

      Some changes were made in the system to improve operability.  It was during
this period that some problems were found with the flame  rod flame detection system
and the high-voltage spark igniter.  The flame rod detector was found to be
temperamental. It activated the automatic flame failure shutdown, even though flame
was present. For that reason, it was later supplemented with a flickering infrared
flame detector system.  The high-voltage  spark igniter was also a problem, because it
required careful adjustment and also sometimes caused electrical interferences and
shutdown of the computer control system; Therefore, the  electric igniter system was
eliminated.  Instead, the propane is manually lighted  with a small H2/air torch,  which
provides a reliable ignition flame, eliminating uncertainties associated with the  spark
igniter.

      Part of the initial operating tests included determination of the maximum
temperature achievable when operating in the single-chamber mode.  The maximum
temperature achieved in this mode at  a "high-flow rate" condition (8.8 to 10.0 scfm of
air input) was 2000°F.  However, after the quartz window was installed  in the
chamber, this maximum was reduced to 1690°F.  Likewise, the maximum temperature
with the quartz window at "low flow" (5 scfm of air input) was 1505°F. These
                                      14

-------
maximum temperatures were produced by increasing the feed rate of auxiliary fuel
until the O2 concentration in the effluent gas was reduced to 3.0% to 4.0%, since
3.0% was considered to be a reasonable minimum operating level (shutdown occurs
at 2.0%).

      Another part of the initial operating tests was directed to assessing the
accuracy of the thermocouple (T/C) located in the reactor effluent pipe. This T/C is
the primary measure of the reactor temperature and was initially based on a !4-in
diameter shielded T/C.  At the suggestion of NREL, an aspirating T/C was also used
to measure the temperature, which indicated a reading 50° to 80°F higher than the
shielded T/C.  Also, an exposed bead T/C (i.e., unshielded) was tested and showed
good agreement with the aspirating T/C. The exposed bead  T/C was used in
subsequent testing to avoid the problems associated with the aspirating T/C system,
caused by the high HCI concentrations.

       During the initial operating period, a Tri-Agency meeting was held that included
selection of the composition of the synthetic liquid organic that would be used
throughout the non-solar tests  at MRI and the first phases of subsequent testing at
NREL.  The composition selected by the Tri-Agency group was:

                         Component          Wt (%)
                     No. 2 fuel oil                 40
                     o-Dichlorobenzene           40
                     Carbon tetrachloride          10
                     Naphthalene                   5
                     Toluene           "        	5
                                          Total  100

       MRI purchased the above components and blended them together to prepare a
 single 150-gal batch of the mixture, which should be sufficient for all the tests, using
this synthetic liquid organic at  MRI and at NREL. As discussed  in the next section,
                                       15

-------
this mixture was used as feed to the Minipilot System after the initial tests with
propane and fuel oil.
                                         16

-------
                                 SECTION 4
                           PRELIMINARY TESTING
      Preliminary testing of the Minipilot Solar System emissions was carried out
using an on-line GC/ECD, but also included two tests using EPA methods (VOST per
SW-846 Method 0030 and MM5 per SW-846 Method 0010 with GC/MS analysis).
The on-line GC/ECD involved continuously extracting a small flow of the reactor
effluent gas (- 50 mL/min) through a heated Teflon line and through the autosampling
valve of the GC/ECD. About every 30 to 40 min the autosampling valve was activated
to inject a  1-mL gas sample into the GC system. Output from the GC/ECD was
intended to provide data on the concentration of CCI4 and dichlorobenzene (DCB) in
the reactor effluent gas, for calculation of DREs.

      The original purpose of the preliminary testing with the GC/ECD was to carry
out tests over a range of temperatures for each  operating mode to identify the
temperature (or other operating conditions) that provided a DRE at or near 99.99% for
one or both of the volatile POHCs.  Once that was accomplished,  subsequent testing
would be done at the 99.99% DRE conditions, using EPA methods, to ascertain DRE
for all POHCs and to semiquantitate the concentration of volatile and semivolatile PICs
in the reactor effluent gas. Thereafter, it was planned to conduct tests at NREL, at
those same conditions but with solar input, to assess the effect of solar input on the
DREs and PIC emissions.

      A very important aspect of the non-solar tests at MRI  was to evaluate the
POHC and PIC concentrations relative to the  detection limits of the EPA sampling and
                                     17

-------
analysis methods and relative to their amounts in blank samples.  That is, the
sampling and analysis methods needed to be capable of quantitating a significant
decrease in POHC and/or PIC emissions that might be provided by solar input.  If the
results of non-solar tests at MRI exhibited POHC/PIC concentrations at or near the
detection limits of the sampling and analysis methods, or concentrations at or near
those in sample blanks, it would not be possible, using those methods, to show any
decrease that might be provided by solar input.

      Operation of the Minipilot System at a ORE near 99.99%, as opposed to higher
DREs, would be well above the detection limits of the  EPA methods, thus enabling the
determination of any significant improvement in ORE with solar input.  Also, it would
seem to be likely that the same would be true for PIC  emissions, assuming that PIC
emissions associated with a ORE of 99.99% would be greater than those at higher
DREs. Partly for that reason, the preliminary testing with the GC/ECD was directed to
identifying the temperature, or other operating conditions, that provided ORE near
99.99%.

      MRI proceeded with the preliminary testing with the GC/ECD, yielding results
that indicated relatively low emissions of CCI4 (>  99.99%) but higher levels for DCB,
near the desired DRE of 99.99. However, to confirm those results, two tests were
carried out using the EPA methods (referred to as Runs 17 and 18).  Results from the
two EPA tests did not agree with the GC/ECD data. In fact, the EPA results showed
much lower levels of DCB.

      Data for the EPA tests from Runs 17 and  18  are summarized in Table 1.  As
can be seen in this table, the lowest DRE for DCB (Run 18) was > 99.99983%  at a
concentration of < 23 |ig/dscm as determined by the EPA method using the MM5
train.  In contrast, the concentration  result for the GC/ECD was much higher, at near
3,000 u.g/dscm. Results for CCI4 were considerably better in comparison (i.e.,
10.9 ng/L for VOST versus 30 ng/L by GC/ECD), but showed DREs well above
99.99% (i.e., 99.99969%).
                                      18

-------
        TABLE 1.  EPA TEST METHOD RESULTS FOR RUNS 17 AND 18
CCI4

Toluene


D ichlorobenzene

Naphthalene
                        Run 17 (low flow)
                  Single chamber with quartz
                   window
                  Low gas flow rate (3.6 scfm)
                  Low waste feed rate (4.7
                  g/min)
                  Chamber exit temperature =
                   1321°-1365°F
                      Cone.
                   ORE
   6.4 ng/L

   24 ng/L
99.99986

99.9989
< 5.5 u.g/dscm   > 99.999970

< 5.5 jj.g/dscm   > 99.99976


    Cone.
Total volatile PICs      283 ng/L

Total SV-PICs       1,181 u.g/dscm
                                   Run 18 (high flow)
                              Single chamber with quartz
                               window
                              High gas flow rate (8.3 scfm)
                              High waste feed rate (14
                              g/min)
                              Chamber exit temperature =
                               1612°-1616°F
                 Cone.
10.9 ng/L

7.6 ng/L
                                 153 ng/L

                              6,886 ng/dscm
                 DRE
99.99969

99.99957
              < 23 ng/dscm    > 99.99983

              < 23 ug/dscm    > 99.9987


                 Cone.
                                     19

-------
      The large difference in results for DCB (i.e., 23 u.g/dscm for the EPA method
versus 3000 (ig/dscm by GC/ECD) was of great concern and required considerable
investigation. This investigation led to the conclusion that the GC/ECD results for
DCB were incorrect and, further, that valid results for DCB could not be provided by
the on-line GC/ECD sampling and analysis system (even with the replacement
GC/ECD instrument that was found to be necessary). As a consequence, all
subsequent  preliminary testing with the replacement GC/ECD relied on CCI4 results
rather than DCB results, since these were in much better agreement (i.e., 10.9 ng/L
for VOST versus 30 ng/L by GC/ECD).

       Additional preliminary tests were carried out at different temperatures in an
attempt to identify operating conditions that provided DRE near 99.99% for CCI4.
Temperatures were varied over a range of 1330° to  1900°F, but all indicated DREs
were above 99.999% (near the detection limit of the GC/ECD).

       Further tests were carried  out at other conditions in an attempt to "degrade"
performance of the Minipilot System to DRE near 99.99%, including the following:

       1.     Lower atomization air pressure
       2.     Atomization using N2 instead of air
       3.     Non-tangential entry of secondary air
       4.     Change in composition of liquid organic feed
       5.     H2O injection

       1.     Lowering of the atomization air pressure, at a temperature near 1300°F,
             did  decrease  DRE (to 99.9927%), but it also resulted in carbon particles
             (soot) that caused plugging of the sampling system and concern for
             downstream equipment, especially the effluent pump.  It also caused CO
             levels of near 300 ppm, which is not a realistic operating condition.
             Surprisingly, this did not seem to affect THC levels.
                                       20

-------
2.     Atomization of the liquid organic feed using N2 instead of air at
      temperatures near 1400°F had no significant effect on ORE for CCI4.
      Again, lowering the atomization pressure resulted in soot formation.

3.     Testing of the Minipilot System was normally carried out using the
      tangential entry of secondary air.  Tangential air entry was designed into
      the system to maximize turbulence, since the Reynolds Number (NRe)
      calculated for this system was low (230).  In an effort to decrease ORE
      by reducing turbulence, the tangential air entry was changed to introduce
      combustion air through one of the spare nozzle connections (i.e.,
      horizontal air entry). Results indicated that ORE for CCI4 was still near
      99.999%.

4.     NREL suggested that a preliminary test be conducted using a different
      composition of liquid organic feed having higher concentrations of two
      POHCs. At this suggestion, MRI prepared a special feed mixture
      containing 80% DCB and 20% CCl4.

      Even though this special feed mixture had a  heating value of only about
      7,000 Btu/lb, compared to the normal feed of near 12,000 Btu/ib, it
      appeared to burn well.  Test results did not indicate any significant
      increase in CCI4  emissions (i.e., decrease in ORE) for this special waste
      feed.

5.    As discussed previously, the Minipilot System was designed with H2O
      injection capability for temperature control during solar operation.  This
      raised some concern that  H2O injection might affect ORE.  In order to
      address that concern, at least in part, one of the preliminary tests
      included water injection to assess  its effect on ORE (in the non-solar
      mode).
                                 21

-------
            During one preliminary test when the temperature of the reactor effluent
            was at 1897°F, water was injected to lower the temperature to 1504°F.
            GC/ECD results for CCI4 clearly showed that the H2O injection did not
            decrease the ORE.  It may have actually increased the ORE from
            99.9995% to 99.9999%, but this is somewhat uncertain considering the
            low concentration level (i.e., 16 ng/L versus 4 ng/L).

      None of the above changes in operating conditions  identified any practical
means of achieving ORE near the desired level of 99.99%. It was therefore decided
to suspend preliminary testing and proceed with tests using the EPA methods,
carrying out all tests at the relatively low temperature of 1400°F to produce the lowest
achievable ORE (but likely not as low as 99.99%).
                                      22

-------
                                 SECTION 5

                        EPA METHOD TEST RESULTS
      The primary purpose of conducting non-solar tests on the Minipilot System,
using EPA sampling and analysis methods, was to determine if the POHC and PIC
emission levels were sufficiently above the detection limits of the methods, and
sufficiently above blank levels, to enable determination of significantly lower  levels in
subsequent testing with solar input.  This purpose became even more critical in view
of the fact that the non-solar tests would necessarily have to be done at DREs above
99.99%.                                                          '

      Non-solar testing consisted of 10 runs conducted over the period of May 29 to
June 16, 1992. Each run covered 2 h of sampling by VOST and MM5 per EPA
Methods 0030 and 0010 in SW-846.  The only exception to the methods was that the
MM5 was limited to single-point sampling due to the small 2-in diameter sampling
location (near the inlet to the final gas cooler).  All analysis was by GC/MS.  Full scan
analysis was used for quantitation of POHCs and  PICs in the samples.  MM5 extracts
were also analyzed by SIM in order to provide  lower detection limits for the two  SV-
POHCs.  Dioxin/furan analysis was also done on some MM5 samples.  Samples of
each batch of liquid organic feed used in the tests were taken to confirm the expected
POHC concentrations and were also analyzed  for HHV and Cl content.

      The following  subsection summarizes the operating conditions during  each of
the 10 runs.  Then, the next subsection  summarizes the analysis results for the
10 runs, based on the more detailed analysis results contained in Appendices A
                                     23

-------
through F, along with the summary of quality assurance results contained in
Appendix G.

TEST CONDITIONS

      The series of 10 tests involved operation in both single-chamber and dual
chamber modes at two different waste feed and secondary airflow rates (referred to as
"low flow" and "high flow"). Some tests were done with UV light, and some tests
included analysis for dioxins/furans (D/F).

      Ail tests were done at essentially the same temperature of 1400°F.  For the
dual-chamber tests, the preliminary testing had shown that the gas temperature fell
about 300°F in the second chamber, even if preheated. Although  later solar tests
would be done without any flame in the second chamber, it was assumed that it would
be best to carry them out at the same temperature as the  first chamber, without the
300°F temperature drop (i.e.,  1400°F). Therefore, except for the last test (Run 40),
the dual-chamber tests at MRI were conducted while firing propane in the second
chamber to maintain the gas at 1400°F.  However, the last dual chamber test was
done without firing  propane, allowing the 300°F temperature decrease across the
second chamber.

      The basic conditions for the 10 tests, in the order in which they were done, are
shown in Table 2.  Notations indicate duplicate runs and those runs that were selected
to include analysis of MM5 samples for D/F.  A summary of the process operating
conditions monitored during each run is provided in Table  3.
                                      24

-------
TABLE 2. SUMMARY OF EPA TESTS
Run No. Conditions
Notations
: : Single chamber tesis
31
32
33
34
35
36


37
38
39
40
Low flow3

High flow3
High flow
Low flow
High flow


with UV light
Duplicate of Run 34
Duplicate of Run 33
Duplicate of Run 32,
D/Fb
Duplicate of Run 31,
D/Fb

Low flow with UV light

(Firing propane
High flow
Low flow
High flow
High flow
Dual chamber tests

in second chamber, except Run 40)


with UV light
with UV light
D/Fb

D/Fb
D/Fb
a Based on waste feed and secondary airflow rates.
b D/F indicates runs which also included analysis for dioxins and furans.
              25

-------

























<
^^
Q
CO
p
tr
LJJ
a.
O
UL


DC
 f-- *—
in CM CD
o o o
o" o o"

0 0 CM

CO p CD
cn cn cn

CO CM Tf
o o" o




o 10 in
CM T- CM
000




CM 0 TT
TT TT Tf



^ 0 CO
TT Tf TT





o o o
co co" CD"






cn CD o
f— • F"^ CO




Tf r— - co
Tf § ^




-z.~z.~z.


CO *- IO
CM CO CM


" " I
all.
T~ Q) *^ ;J) f\|
"•$22
3 3
cr cr

Tf CD CD
co in CD
o o o
000

0 0 CM

CO CO i-
CM — CO

co in in
r*- co cn
CO "- TT




— CO CO
CM i- CO
CM CM CM
000




cn T- CM
r-. in CM
1^ t-- 00



SS8
I-- l~- 00





o o o
—. — CM






in co cn
CO CO OO




i- CD CM
— O CM




-z.-z.-z.


g§fe
V T °


„ " "
al|
2 E|
111!
3
cr

•- 1>- r.





°. 0 °.







in T- co
OO CO 00




in o cn
o •» •»




-z.-z.~z.


•G oo in
co CM in
— CM O


Average =
Minimum =
Maximum =
n 34
3
cr

oo r^- o
in in CD
o o o
0 0 O

o o co

o cn oo
co i-~ cn

CM o •":
*~ ° CM




r»- co CM
— O CM
000




•* CD CM
— r«- co
rr CO -t



CO Tt CO
CM OJ CM






o o o
CD" CD" CD"






O 00 CO
CO f1- OO




r~- in r»-
5 ° ^




-z.-z.-z.


in cn h-
*"-f oo r^*
coco v


Average =
Minimum =
Maximum =
n35
3
cr

v- 
CM CM CM
000




in in •^>
•q- co i-~
r-.r-.t--



coioS






p o p
T-" T^ 1-^






CM O •*




in co eo
— O CM




-z.-z.-z.


CM S -
CM CM TJ-


Average =
Minimum =
Maximum =
n36
3


r— Tt ^f
O CM CO
o o o
o o o

— o »-

CM CO Cl
CO CM CO

m co CM
— o ^i-




— oo m
CM — CM
000




OO CD CO
CM i- -t
•» •* "*



CM CM CO
•* Tt ^r





o o o
co" CD" CD"






CO ^- CD
CO 00 OO




CO I-- ^f
Sfei?




-z.~z.~z.


•r- o in
CO CO T-


Average =
Minimum =
Maximum =
n37
3
cr

CO t-- O)
••- 0 •—
o o o

O O i-

r-~ •- co
00 00 Ol

in co CD
o o o




CO O O
T— T~ CM
CM CM CM
000

r


SO CO
TJ- r-.
i-r- t-



SSS
^ ^ Is-





p p p
O O -r-






01 in co
I-.I-.OO




CO T-.CO
*- O CM




O 00 CO
CM O CO


in CM o
— Tf O
CM CM Y


Average =
Minimum =
Maximum =
n38
3
cr

in o TT

o o o

O o T-

CM CO CO
Tf CO Tf

•*- 0 CM
o o o




Ol Tf Tf
T- i- CM
000




858
Tf Tf Tf



CO CMTf
Tf Tf Tf





GO OO 00
ui in" in"






CO O t--




t-- CD CO
i— O CM




co co r-*
•— o CM


— p op
in r-. r-.
CO CO rp


Average =
Minimum =
Maximum =
n39
3
cr

co in <-
Tf o in
0 O 0

o o CM

I-- CO CD
CD in f-.

Tf co r— •
'-'-'-




co cn xf
o cn co
CM T- CM
000




88S
t-- r- co



fcS§
t- f~ I--





0 O O
o o o






co cn oo
cn co cn




I-- cn in
O CO CM
Tf CO Tf




co in i-
— — CM


CO CD —
CM Tf ^
CM CM —


Average =
Minimum =
Maximum =
n40
3
DC

Ol O CO

o o o

o o 2j

oo in T-
— t- CM"

cn co in
co CM in




cn TT o

o o o




oo cn CD
CO *— f-.
i-. r-. t^



00 00 f--
co in f^
i-.i-.i-.





0 O 0
o o o






Tf O 00
cn cn cn




Tf CO CO
CO CO CM
O O CM




co o cn
Tf CO Tf


CD f-. CD
CM IO i-
cvi CM Y


Average =
Minimum =
Maximum =









CD"
•o
o
CD
i
.C
O
01
01 1-:
.£ o
""§
CD ^
.£=""
', - "O
"o>8
-^ CD

rs CD
_£- £
CO CD
"" 3
CD CD
CO CL
£ I
fi
• 3 "-5
c5"o 2

™ £ S.
"Si-2 °

2 o »3
£§• =
5«fl
CO — C
CO Ol'to
cr ^ o
1 °>
1 CLO>
H> o co
r- CD C -
§•§ c"
CO C C
0 2 oi^
S-g-co
S ^ CO c
S.i1'§£r
0 « CO g
O ra g E
2 a, cr 5

~ i- ~U C
0 g CD 0
-O ._ U
SJt?.£

2 CDr^-S=
2 :=•-=: co
S CD CO
^ CD 3_ >
— "o
CD C — CM
.9 'I 0 CD
< "o "o
j -z.-z.~z.
26

-------
      Residence time was calculated for each run based on the temperature of the
gas at the exit of each chamber, air feed rate, and liquid organic feed rate, as well as
other fixed input rates. These calculated residence times are as follows:
                          Residence Time (seconds)
:'•'•' .-'.-; '•• ';'v:%:, - Single chimiberfriode ': •••;.•• ::/:. ' :" %>:
. • ' ' • . ' ••.•.•-'.-'- .-.•••• , '. :*ryv '.. • .•;•*' •:'•'•;• . • ,•••.-•: • - ' ' :

Run 31
7.1
Low flow
Run 34 Run 36
7.0 6.8
High flow
Run 32
4.0
Run 33
4.2
Run 35
4.0
Dual chamber mode
(total residence time for both chambers}
Low flow
Run
38
13.3
High flow
Run 37 Run 39
7.8 7.8
Run 40
8.9
      Some important items to note about the operating conditions shown in Table 3
are as follows:

      1.    All runs were carried out at essentially the same temperature of 1400°F.
            Thus the heat loss is relatively constant for each mode (except in tests
            with UV light where there was increased heat loss through the quartz
            window). As a consequence,  the low-flow conditions required a
            somewhat higher ratio of heat input to combustion air feed than the ratio
            at the high-flow conditions, in  order to maintain the same temperature.
            Therefore, the O2 levels were  lower at the low-flow condition than the
            high-flow condition for each operating mode.

      2.    Heat loss was, of course, greater in the dual-chamber mode, and
            propane was fired in the second chamber to maintain the 1400°F
            operating temperature. As a consequence of this, the O2 levels were
                                      27

-------
      lower in the dual-chamber mode than the single-chamber mode at each
      of the two flow conditions.  Thus the lowest average O2 level of 4.2%
      was associated with the dual-chamber, low-flow test (Run 38) (excluding
      tests with the UV light).

3.     In the single-chamber tests with UV light, the higher heat loss (through
      the quartz window) was compensated for by firing of fuel oil (in a
      separate nozzle) to maintain the 1400°F temperature.  This reduced the
      O2 content of the effluent gas for each flow condition.

      Because of the above, the lowest average O2 level  (3.2%) occurred in
      the low-flow, single-chamber test with UV light (Run 36).  Since this was
      considered the minimum allowable O2 level, it was not possible to
      conduct any low-flow, dual-chamber tests with UV light, at the  1400°F
      -temperature.

4.    As noted above, fuel oil was fired in the single-chamber tests with UV
      light in order to maintain the 1400°F temperature. However, no fuel oil
      was fired in the dual-chamber tests.  Instead, propane was fired in the
      second chamber to maintain the temperature at 1400°F.

5.    Average THC levels were low in all tests, near 1  ppm. CO levels were
      also low (0 to 3 ppm) in most  tests, with the exception of the two high-
      flow tests in the single-chamber mode (Runs 32 and 33), which averaged
      37 ppm and 11 ppm, respectively. These higher CO levels did not occur
      in the high-flow, single-chamber test with UV  light (Run 35), which
      included firing of fuel oil with reduced O2 level.
                                28

-------
EPA METHOD TEST RESULTS

      A summary of all the analysis results for the 10 EPA method tests is presented
in Table 4.  This table contains a large amount of data, since it includes the following:

            V-POHCs  (CCI4 and toluene)
            V-PICs
      •     SV-POHCs (dichlorobenzene and naphthalene)
            SV-PICs
            D/F
      •     All the above for blank samples

      As expected, the DREs were above 99.999% for all four POHCs in all runs.
Several interesting conclusions can be drawn from the data and are discussed in this
report. However, as noted previously, the primary purpose of these non-solar tests
was to determine if the  amounts of POHCs and/or PICs were sufficiently above the
detection limits, and above the blank levels, to allow determination of significantly
lower amounts in later tests with solar input. This is the first item discussed below in
terms of each of the POHCs and V- and SV-PICs.
                                      29

-------
o
CC
X
CO
z

DC

o:
O
CC

£C

5
o
CO


§
0.
1
u.


§
co
111
£C

fe
CO
    1
    Q
       1?
                                                            II
                                                      S
                                                         -
                                                            8S8
                        11
                        ii
            cSco-cM--:
           oil-It-;  "-
                                                             ;  < t
                                                                    •*-«•
                                                                     V V
                                   o
-------
Ability to Determine Significantly Lower PQHC/PIC Emissions

      The test data were evaluated in terms of the ability to determine lower levels of
POHCs and PICs in later solar tests, based on the results given in Table 5.  In this
evaluation, the test data for runs that included UV light were not considered, since
they were intended for other purposes as will be discussed  later. Also, the evaluation
is based on the amounts in samples, rather than concentration, because it is the
amount in the samples that is critical.

      Ability to determine lower levels may be limited by detection limits of the
methods or by levels  present in blank samples.  It was concluded from these tests that
the main factor was the levels present in blank samples. Also, the results show that
the lowest sample value  always occurred in the low flow tests for all POHCs, PICs,
and D/F (but the lowest sample value was not always associated with either the
single-chamber mode or dual-chamber mode).  Conversely, the highest sample value
did not always occur  in the high-flow tests.  The fact that the lowest sample value
always occurred in the low-flow tests means that a larger decrease in sample amounts
could be determined  for high flow rate conditions than for low flow rate conditions.
Bearing this in mind,  the minimum decrease in amount that would be possible to
determine was calculated for each POHC and total V/SV-PICs for both the low-flow
and high-flow conditions.  The equation used for this calculation was:
           Minimum  possible decrease (MPD)  _  Lowest sample amount
                that  could be determined        Highest blank amount


      If the lowest sample amount found in the  non-solar tests were the same as the
highest blank amount (i.e., MPD = 1), there would  be considerable uncertainty that
any solar test sample amounts  would be found below the blank amount. Conversely,
if the lowest sample amount found in the non-solar tests were much larger than the
blank amount (e.g., MPD = 10), it would indicate that  a significant decrease in sample
amount could be determined in solar tests, and still be well above the blank  amount.

                                    .   31

-------
TABLE 5. COMPARISON OF BLANKS WITH SAMPLE RESULTS
Item
CCI4
Toluene
V-PICs
DCB
Naphthalene
SV-PICs
Dioxins
Furans
Highest blank
6.5 ng
7.7 ng
85 ng
0.06 ng
4.32 ng
225 ng
0.000618 fig
0.0000776 ng
Sample results (excluding UV tests)
Mode
Single chamber
Dual chamber
Single chamber
Dual chamber
Single chamber
Dual chamber
Single chamber
Dual chamber
Single chamber
Dual chamber
Single chamber
Dual chamber
Single chamber
Dual chamber
Single chamber
Dual chamber
Low flow
21 .9 & 37.2 ng
13.6 ng
5.6 &61.2 ng
15.1 ng
4,685 & 6,862 ng
779 ng
0.22 & 0.36 ng
0.21 ng
3.70 & 7.72 ng
5.55 ng
1 ,560 & 3,248 ng
2,148 ng
O.O0161 ng
0.00871 ng
High flow
46.9 & 124.9 ng
40.9 ng
23.5&41.6ng
22.0 ng
4,824 & 5,827 ng
895 ng
0.68 & 0.86 ng
0.45 ng
7.72 & 8.00 ng
9.09 ng
2,01 4 & 4,01 8 ng
3,579 ng
0.00316 ng
0.00655 ng
0.021 92 ng
0.04470 ng
                        32

-------
      Results of calculations using the MPD equation and the data in Table 5, are as
follows:
    CCI4
    Toluene
    V-PICs
    DCB
    Naphthalene
    SV-PICs
    Dioxins
    Furans
                        Minimum possible decrease that could be determined
                             Low flow
     13.6/6.5 = 2.1
      5.6/7.7 = 0.7
      779/85 = 9.1

     0.21/0.06 = 3.5
     3.70/4.32 = 0.8
     1,560/225 = 6.9

 0.00161/0.000618 =  2.6
0.00871/0.0000776 = 112
                                      High flow
     40.9/6.5 = 6.3
     22.0/7.7 = 2.8
     895/85 = 10.5

     0.45/0.06 = 7.5
     7.72/4.32 = 1.8
     2,014/225 = 8.9

 0.00316/0.000618 = 5.1
0.02192/0.0000776 = 282
       In simplest terms, the above results mean that it would be possible to
 determine CCI4 emission levels that are a factor of 2.1 less than the levels found in
 the low-flow, non-solar tests at MRI, or a factor of 6.3 less than the levels found in the
 high-flow tests.  These minimum possible decreases are somewhat conservative,
 because they are based on the highest blank value and the lowest sample value found
 in single- or dual-chamber modes. The main conclusion from this information is that
 a significant decrease (> 3 times) is determinable for both V- and SV-PICs.

       A significant decrease (> 3 times) is also determinable for two of the POHCs
 (CCI4 and DCB) at the high-flow condition  and for DCB at the low-flow condition. On
 the other hand, the results indicate that it will not be possible to determine a decrease
 for toluene  or naphthalene at the low-flow  condition, and only a limited decrease at the
 high-flow condition.
                                        33

-------
      Like CCI4, a significant decrease can be determined for dioxin at the high-flow
condition, but not the low-flow condition. A large decrease can be determined for
furans at both low- and high-flow conditions.

      As indicated previously, the decrease that could be determined is always
greater for the high-flow condition.  This suggests that future solar tests might best be
carried out at the high-flow condition.  However, the effect of solar input may be a
function of residence time (i.e., time the gas is exposed to solar input) so at least
some testing at the low-flow condition  may be warranted, or could be done optionally,
dependent on results from high-flow tests.  Hopefully, the scoping tests  (i.e., Phase I
tests at NREL) will provide preliminary data that can be used to determine if there is
no need to conduct subsequent EPA-method testing at the low-flow condition.
                                       34

-------
Replication of Results

      Duplicate tests were carried out in the single-chamber mode at the low-flow
condition (Runs 31 and 34) and high-flow condition (Runs 32 and 33).  Results for
these two pair of replicate runs and blanks are shown in Table 6 in concentration units
(as calculated from data in Table 4).

      Results in Table 6 show that the duplicates generally agree within a factor of
about 2 or less for the duplicate runs at low-flow and high-flow conditions.  Agreement
within a factor of 2 or less was considered to be reasonably good agreement
considering the variability of the test conditions and the sampling/analysis methods.
The only exception is toluene, which was different by a factor of 10 for the low-flow
condition (0.31 ng/L versus 3.26 ng/L).  However, the low value for toluene of
0.31 ng/L in Run 31 is very close to the detection limit of 0.2 ng/L

          TABLE 6. COMPARISON OF REPLICATE RUNS (SINGLE-CHAMBER MODE)

CC!4 (ng/L)
Toluene (ng/L)
V-PICs (ng/L)
Dichlorobenzene (u.g/dscm)
Naphthalene (ng/dscm)
SV-PICs (ng/dscm)
a Concentrations shown for
and 0.699 dscm for MM5.
Blank3
< 0.2
< 0.2
4.5
0.086
6.18
320
blanks are

Run
1
0
258
0
Low flow
31 Run 34
21 1.98
31 3.26
365
36 0.59
6.00 12.6
2,530
based
5,310
High
Run 32
5.98
1.13
231
1.15
10.7
5,350
on average sample volumes; 19.1 L
flow
Run 33
2.50
2.22
311
0.83
9.47
2,470
for VOST
                                        35

-------
Comparison of Emissions for Different Operating Modes and Flow Conditions

      Results for the low-flow and high-flow conditions, in both single- and
dual-chamber modes, are shown in Table 7, excluding tests with UV light.
Comparison of these results is discussed below, bearing in mind that duplicates
generally agreed only within a factor of 2 (per the previous paragraphs).

      In the non-UV tests in the dual-chamber mode, the POHCs and PICs are
always higher for the high-flow condition than for the low-flow conditions, but are within
a factor of 2,  except for CCI4.  In the single-chamber mode, the only emissions that
were consistently higher for the high-flow condition than for the low-flow condition were
CCI4 and DCB.

      Dioxins and furans are also higher in the high-flow condition than in the low-flow
condition (single chamber), but not by more than a factor of 2. Interestingly, the
highest concentration of dioxins  and furans was found in the dual-chamber, high-flow
condition (Run 37).

      The parameter most affected by different chamber modes and flow rates is
residence time.  High-flow, single-chamber provides the shortest residence time, and
low-flow, dual-chamber provides the longest  residence time.  The two intermediate
conditions of  low-flow,  single-chamber and high-flow, dual-chamber provide approxi-
mately the same residence time. It would be expected that the emissions would
decrease with increasing residence, time (since the operating temperatures were
comparable).  Results shown in  Table 7 indicate that CC14 and DCB follow this
residence time pattern,  but that toluene, naphthalene, and PICs do not follow the
pattern.
                                      36

-------
CO
LU
Q
O
CC
LU
d.
O
CO

o
LU
CC
LU
LL
LL
Q
CC
O
LL



i
CO
LU
CC
o
2:
LL
O


O
CO
CC
O
O
LU
_l
m


CD
"E
o
~co
Q















Q)
^)
CO

o
0
D5
C
CO









5
_o
c»
±

_o
1








£
.0

D)
3-








o

i
o





CO
c
CC

00
CO
c
CC





fV^
CO
c
CC



CM
CO
c
CC




CO
c
•3
CC




CO
c
CC




CO
C
co
CO













O CD
CD T—
CM •* CM O CD
T-; T- CD in O O CO
Cvi T-" CD CD CM O O CD r^
•* T- CO
°l

i- co in co
r-^ r-~ co T- co
CD O T- CD 03 O CO"
•* CD T-
m
co"



co co
CO CD
O CM CO t"- O CM
in CM co *3- o o CM
c\i CM T- CD o> o do rr
CO *3-
CM"


oo co m
en i- T- r-. o
iri T- t- T-" o" o -^r
co T— in
CM CO
m"


CO CM
co CD en o T-
cn CM m CD o o o
i- co" in ci CM o o CD r^
CO T~ "T-
CO CO
in



i- T- co o
CM CO CO O i-
^ CD 00 CD CO O 1 K
in co 1
CM LO
CM"


00 T-
00 T-
co o o
00 CD O O
cvi CM in p i- o o
CD o •* d co o CD d <
V V • $ 2
0

E ~*-
o 0
eo . — . o
~° E — . _ • c
0 ^ o E E~ -S
c n to o o • ~

~o>S c 0"S) -5^5 ®
^^, d CO ^ O t~ C_5 co en CO
^"^ 0 C5 T- ~~ ^" d ^5
— ^ ^ S! "^ rti CL ' x co o "o1
Ooi •— S5 co > .2 ^ 030
O 1 — > Q ^ 2! CO O LL O <2-



iri
i
E
o
to
-a
CO
CO
co
d
TJ
co
1 —
co

o
**"
— 1
CO
1—
co"
0
E
3
O
0
Q.
E
CO
0
O5
£
03
CO
c
o
T3
0
to
5
fli
CO
Si
J
L
C
0
C"
CO
CO
.1

2
0
8
0
o

tB
                                         37

-------
      It is also interesting to note that the dual-chamber mode tests have significantly
lower concentrations of V-PICs at both flow conditions than did any of the single-
chamber mode tests.  But, as noted previously, the dual-chamber mode test had the
highest concentration of dioxins and furans.

      These comparisons of results for the two operating modes and two flow
conditions tend to support the previous suggestion that future testing be directed
primarily to the high-flow (lower residence time) condition.  They also suggest that
future solar testing with EPA methods might possibly be directed primarily to the
single-chamber mode, but this must be contingent on first determining if dual-chamber
operation is necessary for solar effectiveness (i.e., flame in the single chamber may
interfere with solar photolytic affects).
                                       38

-------
Effect of Artificial UV Light

      The series of 10 runs done using the EPA sampling and analysis methods
included four runs with the quartz window in the reactor and an artificial UV light.  This
ultraviolet lamp with a parabolic reflector was  positioned just outside the quartz
window to direct the UV into the chamber (or  into the second chamber for the dual-
chamber mode).  The purpose of these tests was to determine if the artificial  UV had
any significant effect on  the POHC and/or PIC levels, compared to the levels  in the
thermal destruction tests done without UV light.

      When the quartz  window was in place for the UV light tests, the heat losses
from the chamber were  considerably higher. To maintain the 1400°F temperature, fuel
oil was fired as an auxiliary fuel in one of the  burner guns when operating in the
single-chamber mode with quartz window and UV light (i.e., Runs 36 and 35). When
the two UV light tests were conducted in the dual-chamber mode (Runs 39 and 40),
propane was fired in the second chamber during Run 39 to maintain the 1400°F
temperature.  No propane was fired in the second chamber in Run 40, which  allowed
the temperature to decrease in the second chamber to 1064°F (even though it had
been preheated to 1400°F).  Since the two UV light tests in the single-chamber mode
involved firing fuel oil and the one dual-chamber test with UV light involved firing extra
propane in the second chamber, more oxygen was consumed in these three tests, so
the O2 levels were lower than  in the corresponding tests done without UV light.

      Results for all the tests, including UV tests, are summarized in Table 4. As
shown in that table, two UV tests were done in the single-chamber mode, with one at
low flow (Run 36) and one at high flow (Run 35).  Two UV tests were also done in
dual-chamber mode  (Runs 39 and 40), but both were done at high flow, and Run 40
was  done without firing  of propane in the second chamber.  The results for these tests
are discussed below.
                                      39

-------
      The V-PICs decrease in the single-chamber UV tests by a factor of at least 3.4
in the low-flow test and a factor of at least 2.6 in the high-flow test.  No decrease in
V-PICs is evident for the dual-chamber UV tests, but V-PICs are significantly lower in
all the dual-chamber tests than in the single-chamber tests. The data do not show
any significant decrease in emissions of SV-PICs or POHCs for UV light tests.

      As noted previously, the UV light tests involved higher heat losses (through the
quartz window) requiring use of additional fuel oil or propane to maintain the desired
temperatures, which reduced the O2 levels.  It has been suggested that the lower O2
levels would increase flame temperatures, even though the final gas temperature
remained the same, which may account for the decrease  in V-PIC during the single-
chamber UV tests.  However, the fact that the SV-PICs did not decrease makes this
suggestion seem an unlikely explanation of the decrease  in V-PIC emissions.

      UV light did not generally increase POHC emissions in the single-chamber
tests. CCI4 emissions do, however, increase by a factor of 3 in the high-flow, dual-
chamber tests (Run 39 versus Run 37), and there also is  some increase  in  D/F
emissions.

      The UV light test done in the dual-chamber mode,  without firing of propane in
the second chamber (Run 40), indicates an increase in emissions for SV-PICs and for
two of the four POHCs—CCI4 and DCB.  These increases possibly reflect the effect of
a lower temperature in the second chamber (1064°F) rather than an effect of UV light.
However, it is difficult to explain why the emissions in this dual-chamber test would be
higher than those in single-chamber tests.  That is, in Run 40, the first chamber is
operating at conditions similar to those in two of the single-chamber tests (Runs 32
and 33). Reformation of the two POHCs and SV-PICs in  the second chamber seems
unlikely, but the effect of the residence time in the second chamber (4.9 s) at the
intermediate temperature  (1400° to 1064°F),  along with UV light, is unknown.
                                     40

-------
Characteristics of V-PIC and SV-PIC Emissions

      The previous sections discuss the total amount of V-PICs and SV-PICs that
represent the sum of several different compounds found in the analysis (e.g., the
10 largest components). These compounds could only be semiquantitated, assuming
a relative response factor (RRF) of 1.0. In several cases the compound could not be
specifically identified.  The following two subsections provide some additional
information about the PIC analysis results. More details are contained in Appendix C.

      V-PICs—Data presented in Table 4 show that the concentration of V-PICs was
lower than the concentration of SV-PICs  by at least a factor of 10, in all runs. A
tabulation of the V-PIC data, given in Table 8, shows considerable diversity in the
V-PICs from run to run. That is, except for benzene, the V-PICs found in samples
from one run are generally not the same as those found in other runs, even for
duplicate runs (e.g., Runs 31 and 34).  The reason for this lack of consistency in the
V-PICs is unknown, but it might reflect a variety of components present at low
concentrations.

      The volatile PIC data do not include very many chlorinated compounds, which
is somewhat surprising considering the high Cl content of the liquid feed.  Likewise,
the SV-PICs do not include any chlorinated compounds (see Table 9).

      It  should be noted that the V-PIC data presented in Table 8 do not include
several silane-containing constituents that were found in all runs in relatively large
amounts (see Appendix C).  These silane-containing PICs were not included  in the
summation of V-PICs, because they are  believed to have been produced from the
reactor's internal insulation.  Thus it did not seem appropriate to include them as
"products of incomplete corfibustion."
                                       41

-------
TABLE 8. AMOUNT OF VOLATILE PICs FOUND IN A VOST TRAP PAIR SAMPLE (ng)
Compound
Unknown
Sulfur dioxide
Unknown
Carbon disulfide
Unknown
Dichlonomethane
C6H12 Hydrocarbon
C6H12 Hydrocarbon
Unknown
C6H10-C7H12 Hydrocarbon
Butanal
Jnknown
2-Butanone
Trichloromethane + unknown
Tetrahydrofuran and chloroform
C6H100
Unknown
C6H12 Hydrocarbon
Benzene
C8H18 Hydrocarbon
Cy-Cg Hydrocarbon
2-Propylfuran
C6H12O (Unknown)
Octene (C8H16)
Octane
Unttnown
C6H120 Aldehyde
Unlcnown
Unknown
1,1,1 -Trichloro-2-propanone
C8H140 (Furan)
Chlorohydrocarbon (chloroalkane)
CyH^O Aldehyde
Hydrocarbon (Cg-C13)
Hydrocarbon (Cg-C13)
Benzaldehyde
Unknown
C12H26 Hydrocarbon
CyH^O-CjjH^O Ketone
C8H12O-C8H16O Aldehyde
C10-C13 Hydrocarbon
Octanal
CgH2O Hydrocarbon

Scan
No.
29
29
31
35
36
38
38
45
48
51
55
55
59
63
63
63
64
71
76
81
82
110
134
166
171
172
199
200
218
224
224
225
266
279
286
307
313
314
318
318
321
322
331
Run No.
Blk


















35
50























31



32

3





1,670
19

138






327











1,020
155
42

320


454

463
32



62

21




208






423
726
423










178

201



281



1,550
13

33
359

368
444

1,081

1,010





453



788
93

437

11









409







374


34



74

135
570
1,660


16




563

1,170
20

869


375
346



417











647


35







364
136
165
13


203



315
42

188











376








11

36

35

22



274
50



22



213
176


157
















113


166

148
37

13










25

17


91
25













570





141




38
44
141


68







53

41

63
201
78








24




41







25


39



19








14
42
13


213
202






34



15


339


i



109



40

19










19
36




149







19




32
404


18




191


                               42

-------
Table 8 (continued)
Compound
Dichlorobenzene
Unknown
^10^22"<-'12^24 Hydrocarbon
(alkene)
Unknown
C10H22-C13H28 Hydrocarbon
Unknown
Phenyl alkyl ketone
C9-C15 Aldehyde
Unknown
ClO^cP'Cu^sO
C12-C1S Hydrocarbon

Total Amount
Scan
No.
335
336
341
342
351
351
364
372
389
414
426


Run No.
Blk












85
31






42





4,685
32

268



216

13
223

18

4,824
33












5,827
34












6,862
35









10


1,823
36












1,376
37



13








895
38







"




779
39


48

31







1,079
40
99


32



51




1,069
Note: See Appendix D for data on chlorinated dibenzo dioxins and furans.
       43

-------
                TABLE 9. AMOUNT OF SEMIVOLATILE PICs FOUND IN THE MM5 SAMPLES (ng)

Compound
Alcohol
C6H100
Alcohol
C6H140
c6Hio°
C6H100
C6H12°2
C8H10 (subst. benzene)
Unknown
C6H100
Unknown ketone
p^jQ
[Unknown
C7H1002
C9H10°
[Benzole acid
| Unknown acid ester
J1 ,3-lsobenzylfurandione
kHioPa
C10H1002
(Unknown
Ci5H3o02
Phthalate, subst.
Unknown acid ester
Unknown acid ester
Unknown
C19H2602
C19H3802
Bis(2-ethylhexyl)
jphthalate

IJTotal Amount
Scan
16
112
122
142
181
202
250
290
301
321
342
396
492
567
640
655
704
791
808
948
955
1099
1209
1223
1235
1268
1347
1359
1610


Run No.
Blk






150















33





42

22
31


308

20

190






18

796

94
16
68





20


30

1,56
32
72

360

960

600


140
128



240
1,120
190
108






100





4,01
33


340

190

440


44





800

78






70



52

2,01
34


380

580

1,200


58





800

86






88



56

3,248
35


360

900

1,380


30





780

72





30
69

33

60

3,714
36


300

189

480








630

75



72

99
246

105
54
84

2,334
37


390

930

660








1,170

84


99



132


45
69

3,579

38


285

249

720




27



540

60




36
27
90
24
36

54

2,148

39


330

330

630





21


780

66

30
36



60



48

2,33

40

1,710
111
165

9,300
660

600



168


1,410

96






51





14,271
Deleted C7H8 (toluene) since it was present in XAD/filter blanks and may be a contaminant in XAD or extractant solvent
(also is a volatile POHC).

Deleted C13H28 (tridecane) and C9H20 (nonane) because these were due to the D/F spiking and were present only in the
(samples analyzed for D/F.

(Note:  See Appendix D for data on chlorinated dibenzo dioxins and furans.
                                                     44

-------
      SV-PICs—Results show that the concentrations of SV-PICs were much higher
than those of the V-PICs.  Unlike the V-PICs, the SV-PICs were fewer in number, and
those representing the larger amounts were usually present in all runs.  As can be
seen in Table 9, the predominant SV-PICs (in terms of frequency of occurrence and
concentrations) consisted of:

       •     Alcohol (Scan  122)
            C6H10O (Scan 181)
            C6H12O2 (Scan 250)
       •     Benzoic acid (Scan 655)

       As noted previously, the total SV-PICs were much  higher in Run 40 than in any
 other runs.  This run contained amounts of additional PICs that were not present in
 other runs, including large amounts of C6H10O at Scans 112 and 202.
                                       45

-------
                                SECTION 6

                  CONCLUSIONS AND RECOMMENDATIONS
      Evaluation of the data from the non-solar tests at MRI leads to the following
conclusions and recommendations.
                                                                      s
CONCLUSIONS

1.     DREs for the four POHCs were all above 99.999%.  Naphthalene exhibited the
      lowest DRE,  ranging from 99.99926% to 99.99966%. The highest DREs were
      exhibited by DCB, with all being above 99.99999%.  No practical method of
      operating the system at reduced DRE, near 99.99%, was possible, at least
      within the system's present capabilities and design.  Changes could be made in
      the system that might provide lower DREs, such as  reducing reactor internal
      volume or increasing gas flow rates to shorten residence time, but were beyond
      the scope of the present project.

2.     Preliminary testing did not indicate any detrimental effects of water injection on
      DRE. If water injection had decreased the DRE from 99.9995% to 99.9985%
      (i.e., increased the CCI4 emissions by a factor of 3), this would have been
      clearly discernible.

3.     The  lowest level of POHCs, PICs, and D/F in the emissions always occurred in
      the low-flow  rate runs.  However, the lowest levels were  not always associated
                                     46

-------
      with one test mode (single- or dual-chamber).  Conversely, the highest level did
      not always occur in high-flow rate tests.

4.    When on-sun or solar tests are subsequently done using the Minipilot System,
      a significant decrease in emissions should be quantifiable for V- and SV-PICs,
      ranging from a factor of 6.7 to 6.9 for low-flow rates to 8.9 to 10.5 for high-flow
      rates.  Also,  a significant decrease in emissions should be quantifiable for two
      of the POHCs (CCI4 and DCB) at the high-flow rate condition (by a factor of 6.3
      and 7.5, respectively).  A lesser decrease could be determined  for these two
      POHCs  at the low-flow rate condition (by a factor of 2.1 and 3.5, respectively).
      It would not be possible to determine any decrease for the other two POHCs
      (toluene and naphthalene) at the low-flow condition, and only a limited decrease
      at the high-flow condition (i.e., by a factor of 2.8 and 1.8, respectively).  Lastly,
      a significant decrease  in emissions can be determined for dioxin at the high-
      flow condition (by a factor of 5.1).  Only a limited decrease (a factor of 2.6) can
      be determined at the low-flow condition. A  large decrease in emissions can be
      determined for furans, at either  low- or high-flow conditions.

 5.     Results from two pairs of duplicate tests showed that they generally agreed
      within a factor of 2. The only exception was for toluene in one pair of tests, but
       it was near the detection limit in one run.

 6.     Evaluation of data relative to  possible effects on emissions of flow  rate and
       operating modes did not, in general, indicate any dramatic differences,
       considering that replicate runs generally agreed only within a factor of 2.  The
       only exception to this  was V-PICs, which were significantly higher in the single-
       chamber tests than in the dual-chamber tests.

 7.     Tests with artificial UV light indicated a significant decrease for  V-PIC emissions
       in the single-chamber mode (by a factor of at least 3.4 in the low-.flow test and
                                         47

-------
     at least 2.6 in the high-flow test) but no significant decrease for SV-PICs or for
     the four POHCs.  Excluding Run 40, the UV light tests did not indicate any
     significant increase in emissions, except that CCI4 appeared to increase in the
     dual-chamber test along with some increase in D/F.

     One UV light test (Run  40) was done in the dual-chamber mode without firing
     any propane in the second chamber, allowing  the temperature in the second
     chamber to decrease (to 1064°F).  Results from this test showed a significant
     increase in emissions of two of the four POHCs and SV-PICs.  Even though
     this test involved a lower temperature in the second chamber, it is  difficult to
     understand why these emissions would be higher than those in the correspond-
     ing high-flow, single-chamber mode tests without UV light (Runs 32 and 33).
     That is, in Run 40 the effluent from the first chamber would be expected to be
     essentially the same as in a single chamber test done at the same high flow
     rate (Runs 32  and 33).  That effluent from the first chamber in Run 40 would
     then pass through the second chamber. Thus the emissions from  the second
     chamber in Run  40  should not be significantly higher than those from the first
     chamber (Runs 32 and 33), regardless of the temperature or UV light in the
     second chamber.

RECOMMENDATIONS

1.     Considering the previous discussion of the data and the conclusions, it is
      recommended that work proceed on testing at NREL with solar input. The
     testing at NREL  should be done at the same temperature and flow rates of
      secondary air  and liquid organic as the non-solar tests at MRI. Use of these
      same  conditions and the same EPA sampling and analysis methods is neces-
      sary for quantitatively determining the decrease in emissions that should be
      provided by solar input (as EPA's and NREL's earlier bench-scale  research
      would indicate).
                                     48

-------
2.    The results of the tests at MR! showed that a greater decrease in emissions
      can be determined for the high-flow rate condition than for the low-flow rate
      condition. It is, therefore, recommended that EPA-method tests at NREL be
      directed primarily to tests at the high-flow condition.  However, this recommen-
      dation should be implemented only if preliminary solar testing (e.g., Phase I
      .scoping tests) indicates that the longer residence time provided by the low-flow
      condition is not necessary to produce a significant effect on emissions.
      Decreases in emissions will likely depend on the length of time the gases are
      exposed to solar input, but  may not show any effect on emissions within the
      range of residence times associated  with the two flow conditions.  In fact, those
      times may be much longer  than what is  needed. Such  a result would indicate
      the need for further testing  at considerably higher flow rates to determine the
      optimum residence time, so that the  size of the solar reactor for  larger systems
      could be minimized.

 3.    The evaluation of the data, as previously discussed, suggests that future testing
       could be limited to single-chamber mode (as well as limiting them to the high-
       flow  condition).  It is, therefore, recommended that single-chamber mode be
       implemented, if preliminary solar testing (e.g.,  Phase I scoping tests) indicates
       that the presence of flame  (in the single chamber) does not adversely impact
       the effect of solar input on  the emissions.

 4.    Further testing with the artificial  UV light at different conditions (i.e.,
       temperature, exposure time,  lamp configuration, etc.) may be warranted, to
       determine if reduction in SV-PICs (or POHCs) can be achieved  along with the
       indicated reduction in  V-PICs that occurred  in the single chamber tests.
                                        49

-------
                                 SECTION 7

                                REFERENCES
1.     Conceptual Designs of Solar Systems for Solid Waste Destruction.  Draft Final
      Report prepared for U.S. Environmental Protection Agency by Midwest
      Research Institute. September 28, 1990.

2.     Feasibility Study for Bench-scale Testing of Desorption of Organics from Soil
      with Destruction in a Solar Furnace.  Draft Final Report prepared for U.S.
      Environmental Protection Agency by Midwest Research Institute.
      September 24,  1991.

3.     Graham, John L,  et al. "The High Temperature Thermal/Photolytic Oxidation of
      Monochlorobenzene."  University of Dayton Research Institute.  Paper
      submitted for publication in the Journal of Photochemistry and Photobiology,
      1992.

4.     Quality Assurance Project Plan for Testing of the Minipilot-Scale Solar Reactor.
      Prepared for U.S. Environmental Protection Agency by Midwest Research
      Institute. March 20, 1992.
                                      50

-------
             APPENDIX A
   POHC INPUTS AND VOST RESULTS WITH
CALCULATION OF DREs FOR CCI4 AND TOLUENE
                 51

-------









CO
LLJ
cc
i 	
1 —
Q_

«••
O
I
o
Q_
Q
Z
^
CO
LLJ

^^
DC
^
Q
_J
LL
Q
LLJ
UJ
i *
LL.
III
UJ
CO
<
£
•7
^
UJ

CD
H















1
OT
CD
"5
s
"3
Q.
.=
O
O
Q.







•£
CO
Z

CD
CD
"Q
L_



^
O
0



CD
0
Q







CD
CO
5
_o
-a
CD
CD
%*~
o
CO
s>
0
cr
_j




^
c
E
•o)




as
"E
E




CD CD O)
{0°^5> .£=
in E 12
> -*S CD
—
NM*


c
13
CC

CO

CM
CD


CO
CM
CD




CO
•sf
0



f2







CM
CO
^~






O
Tj-







p
T—
CO




CO

•~>
"o
CD
z
CO

0^
°
Q
8
CO
"Trt
.Q
CD
C
"CO
£
JZ
CO
c
_
0?
in
•a

CO
CD"
CD
~o
v§
o
_
Q
S"
o

CD"
. 0

0)0
^- "*
S "
• c

H^IZS
~
1 1
CD O
Q O
.a o
52

-------
TABLE A-2. ADJUSTED VOST VOLUMES
Run
31
31
31
32
32
32
33
33
33
34
34
34
35
35
35
36
36
36
37
37
37
38
38
38
39
39
39
40
40
40
Trap
pair
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
1
2
3
Volume
sampled
(L)
18.93
20.84
19.74
21 .72
20.19
22.31
19.82
20.95
20.08
20.04
20.00
20.09
20.14
19.96
20.05
19.86
20.06
19.89
19.99
20.02
19.99
19.94
19.94
19.88
19.85
20.03
20.00
19.93
20.09
19.98
Meter
temp.
(F)
74
73
74
72
71
71
72
73
73
74
74
74
72
72
71
73
73
74
61
59
57
63
61
63
59
59
60
52
51
51
BP
("Hg)
29.60
29.60
29.60
29.60
29.60
29.60
29.16
29.16
29.16
29.00
29.00
29.00
28.96
28.96
28.96
28.96
28.96
28.96
29.10
29.10
29.10
29.16
29.16
29.16
29.00
29.00
29.00
28.93
28.93
28.93
Meter
coef.
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
, 0.9788
0.9788
0.9788
\
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
0.9788
Adjusted
volume
(L)
18.12
19.99
18.90
20.87
19.44
21.48
18.76
19.80
18.98
18.80
18.76
18.85
18.94
18.77
18.89
18.64
18.83
18.63
19.29
19.39
19.44
19.20
19.28
19.15
19.16
19.33
19.27
19.45
19.65
19.54
               53

-------
TABLE A-3.  VOST ANALYSIS RESULTS
Sample
volume
(L)
Run 31
Field blank pair
Pair 1 18.12
2 19.99
3 18.90
Total 57.01
Run 32
Field blank pair
Pair 1 20.87
2 19.44
3 21 .48
Total 61 .79
Run 33
Field blank pair
Pair 1 18.76
2 19.80
3 18.98
Total 57.54
Run 34
Field blank pair
Pair 1 18.80
2 18.76
3 18.85
Total 56.41
Run 35
Field blank pair
Pair 1 18.94
2 18.77
3 18.89
Total 56.60
Carbon Tetrachloride
Amount found (ng) Avg.
cone
front back (ng/Lj

<2.0
15.6
15.5
16.3
66.2
<2.0
77.8
124.0
124.8
437.2

<2.0
35.5
42.3
73.7
193.5
<2.0
28.0
20.9
17.0
85.7
<2.0
118.8
118.1
67.0
389.8

<2.0
6.3
7.3
5.2
1.16
< 2.0
47,1
28.6
34.9
7.08

<2.0
11.4
10.9
19.7
3.36
<2.0
9.2
6.0
4.6
1.52
<2.0
39.9
28.4
17.6
6.89
Toluene
Amount found (ng)
front

<2.0
2.4
9.0
27.8
49.8
<2.0
14.4
31.6
51.6
110.7

<2.0
37.7
8.1
12.6
67.2
<2.0
58.8
9.9
11.5
87.0
<2.0
15.1
5.0
10.5
38.8
back

<2.0
3.2
<2.0
5.4

<2.0
9.1
<2.0
<2.0


<2.0
3.9
2,3
2.6

3.2
2.4
<2.0
2.4

2.4
4.1
1.9
2.2


Avg. Surrogate
cone. recovery
(ng/L) (%)

91/74
71/126
120/122
141/92
0.874
58/44
122/102
146/95
162/57
1.79

49/68
85/79
74/78
105/72
1.17
66/94
116/92
86/89
129/106
1.54
74/101
101/111
74/101
108/108
0.686
               54

-------
TABLE A-3  (continued)
Carbon Tetrachloride
Sample Amount found (ng) Avg.
volume ~™v~
(L)
Run 36
Field blank pair
Pair 1 18.64
2 18.83
3 18.63
Total 56.10
Run 37
Field blank pair
Pair 1 19.29
2 19.39
3 19.44
Total 58.12
Run 38
Field blank pair
Pair 1 , 19.20
2 19.28
3 19.15
Total 57.63
Run 39
Field blank pair
Pair 1 19.16
2 19.33
3 19.27
Total 57.76
Run 40
Field blank pair
Pair 1 19.45
2 19.65
3 19.54
Total 58.64
front back {ng/L)

3.2
17.8
17.1
10.6
62.9

<2.0
27.9
34.1
24.9
114.2

<2.0
10.0
7.0
11.5
39.8

<2.0
73.5
25.4
19.4
191.8

<2.0
172.2
283.0
248.4
970.3

3.3
7.4
6.2
3.8
1.12

< 2.0
13.0
7.3
7.0
1.96

< 2.0
3.6
2.8
4.9
0.691

<2.0
52.3
12.7
8.5
3.32

<2.0
57.8
91.9
117.0
16.5
Toluene
Amount found (ng) Avg.
front back (ng/L)

2.6
22.7
6.3
15.4
51.1

<2.0
10.6
13.2
32.2
72.7

4.2
11.4
4.7
10.2
36.5

4.6
20.8
9.9
16.0
65.7

3.0
10.9
12.1
18.2
51.7
•
3.2
2.7
<2.0
<2.0
0.911

<2.0
11.4
3.3
<2.0
1.25

3.4
3.7
< 2.0
4.5
0.633

3.1
6.8
6.8
5.4
1.14

<2.0
3.8
<2.0
4.7
0.882
Surrogate
recovery
(%)

79/84
80/111
89/112
81/106


38/96
125/113
136/111
140/110


81/81
93/111
96/95
99/109


87/76
90/86
93/106
102/108


94/91
115/112
115/99
107/107

          55

-------
TABLE A-4.  VOLATILE POHC EMISSIONS

Run 31

Run 32

Run 33

Run 34

Run 35

Run 36

Run 37

Run 38

Run 39

Run 40

Effluent
gas flow
(dscm/min)
0.120

0.221

0.208

0.117

0.211

0.121

0.213

0.119

0.209

0.209

POHC concentration •
(ng/L or ng/dscm)
CCI4
Toluene
CCI4
Toluene
CCI4
Toluene
CCI4
Toluene
CCI4
Toluene
CCI4
Toluene
CCI4
Toluene
CCI4
Toluene
CCI4
Toluene
CCI4
Toluene
1.16
0.874
7.08
1.79
3.36
1.17
1.52
1.54
6.89
0.686
1.12
0.911
1.96
1.25
0.691
0.633
3.32
1.14
16.5
0.882
POHC
Carbon tetrachloride
(g/min)
0.000000139

0.000001562

0.000000699

0.000000,178

0.000001454

0.000000136

0.000000418

0.000000083

0.000000694

0.000003448

emissions
Toluene
(g/min)

0.000000105

0.000000395

0.000000243

0.000000181

0.000000145

0.000000110

0.000000266

0.000000076

0.000000238

0.000000184
               56

-------
TABLE A-5. DREs FOR VOLATILE POHCs
Carbon tetrachloride

Run 31
Run 32
Run 33
Run 34
Run 35
Run 36
Run 37
Run 38
Run 39
Run 40

Run 31
Run 32
Run 33
Run 34
Run 35
Run 36
Run 37
Run 38
Run 39
Run 40
Input
(g/min)
0.432
0.827
0.785
0.487
0.861
0.445
0.819
0.432
0.810
0.760

Input
(g/min)
0.216
0.414
0.393
0.244
0.430
0.223
0.409
0.216
0.405
0.380
Emissions
(g/min)
0.000000139
0.000001562
0.000000699
0.000000178
0.000001454
0.000000136
0.000000418
0.000000083
0.000000694
0.000003448
Toluene
Emissions
(g/min)
0.000000105
0.000000395
0.000000243
0.000000181
0.000000145
0.000000110
0.000000266
0.000000076
0.000000238
0.000000184
ORE
(%)
99.999968
99.99981 1
99.999911
99.999963
99.999831
99.999970
99.999949
99.999981
99.999914
99.999546

ORE
(%)
99.999951
99.999905
99.999938
99.999926
99.999966
99.999950
99.999935
99.999965
99.999941
99.999952
               57

-------
                      APPENDIX B
                   MM5 RESULTS AND
CALCULATION OF DREs FOR DICHLOROBENZENE AND NAPHTHALENE
                          58

-------
TABLE B-1.  MM5 SUMMARY DATA

Run 31
Run 32
Run 33
Run 34
Run 35
Run 36
Run 37
Run 38
Run 39
Run 40
Sampling
time
(min)
120
120
120
120
120
120
120
120
120
. 120
Gas
volume
(dscm)
0.617
0.751
0.815
0.612
0.824
0.494
0.726
0.604
0.801
0.802
Orsat analysis
%O2
9.0
12.0
12.0
8.8
8.0
3.4
8.4
5.0
7.0
11.3
% CO2
11.0
6.6
6.4
8.4
9.6
12.4
8.8
10.8
9.6
6.7
Percent
water
7.6
8.7
10.0
11.5
10.6
17.5
10.9
12.2
12.3
8.8
Average
stack
temp
232
304
299
198
302
210
324
221
300
266
% Iso-
kinetic
100.9
99.8
100.5
101.4
100.7
108.2
101.0
102.5
102.5
96.2
             59

-------
TABLE B-2. SV-POHC EMISSION RESULTS
Run
31

32

33

34

35

36

37

38

39

40

Blank train

Weight Volume POHC
found sampled concentration ,
(fig) (dscm) (ng/dscm)
1 ,2-Dichlorobenzene
Naphthalene
1 ,2-Dichlorobenzene
Naphthalene
1 ,2-Dichlorobenzene
Naphthalene
1 ,2-Dichlorobenzene
Naphthalene
1 ,2-Dichlorobenzene
Naphthalene
1 ,2-Dichlorobenzene
Naphthalene
1 ,2-Dichlorobenzene
Naphthalene
1 ,2-Dichlorobenzene
Naphthalene
1 ,2-Dichlorobenzene
Naphthalene
1 ,2-Dichlorobenzene
Naphthalene
1 ,2-Dichlorobenzene
Naphthalene
0.22 0.617
3.70
0.86 0.751
8.00
0.68 0.815
7.72
0.36 0.612
7.72
0.30 0.824
7.89
0.36 0.494
6.75
0.45 0.726
9.09
0.21 0.604
5.55
0.15 0.801
5.91
1 .04 0.802
5.75
0.06
4.32
0.357
5.997
1.145
10.652
0.834
9.472
0.588
12.614
0.364
9.575
0.729
13.664
0.620
12.521
0.348
9.189
0.187
7.378
1.297
7.170
-

Stack POHC
flow emission rate
(dscm/m) (g/min)
0.120 0.000000043
0.00000072
0.221 0.000000253
0.00000235
0.208 0.000000174
0.00000197
0.117 O.OOOO00069
0.00000148
0.211 0.000000077
0.00000202
0.121 0.000000088
0.00000165
0.213 0.000000132
0.00000267
0.119 0.000000041
0.00000109
0.209 0.000000039
0.00000154
0.209 0.00000027
0.00000150
_

                60

-------
  TABLE B-3. DREs FOR SEMIVOLATILE POHCs
1 ,2-Dichlorobenzene

Run 31
Run 32
Run 33
Run 34
Run 35
Run 36
Run 37
Run 38
Run 39
Run 40

Run 31
Run 32
Run 33
Run 34
Run 35
Run 36
Run 37
Run 38
Run 39
Run 40
input3
(g/min)
1.73
3.31
3.14
1.95
3.44
1.78
3.27
1.73
3.24
3.04

Input3
(g/min)
0.216
0.414
0.393
0.244
0.430
0.223
0.409
0.216
0.405
0.380
Emissions
(g/min)
0.000000043
0.000000253
0.000000174
0.000000069
0.000000077
0.000000088
0.000000132
0.000000041
0.000000039
0.00000027
Naphthalene
Emissions
(g/min)
0.00000072
0.00000235
0.00000199
0.00000148
0.00000205
0.00000165
0.00000267
0.00000109
0.00000154
0.00000150
ORE (%)
99.9999975
99.9999924
99.9999945
99.9999965
99.9999978
99.9999950
99.9999960
99.9999976
99.9999988
99.999991 1

DRE(%)
99.99966
99.99943
99.99949
99.99940
99.99952
99.99926
99.99935
99.99949
99.99962
99.99961
a Inputs from Appendix A, Table A-1.
                    61

-------
            APPENDIX C






PIC RESULTS FOR VOST AND MM5 SAMPLES
                62

-------
TABLE C-1. AMOUNT OF VOLATILE PICs FOUND IN A VOST TRAP PAIR SAMPLE (rig)
Compound
Jnknown
Sulfur dioxide
Unknown
Carbon disulfide
Unknown
Dichloromethane
C,H12 Hydrocarbon
C8H,2 Hydrocarbon
Unknown
CjHto-CjH,, Hydrocarbon
Butanal
Unknown
2-Butanone
Trichloromethane + unknown
Tetrahydrofuran and chloroform
CSH100
Unknown
C,H12 Hydrocarbon
Benzene
CeH1t Hydrocarbon
Cy-Cj Hydrocarbon
2-Propylfuran
CeH12O (Unknown)
Octene(C,Hl6)
Octane
Unknown
CeH,2O Aldehyde
Unknown
Unknown
1 ,1 ,1-Trichloro-2-propanone
C8H14O (Furan)
Chlorohydrocarbon (chloroalkane)
C7HUO Aldehyde
Hydrocarbon (C8-C)3)
Hydrocarbon (C9-C13)
Benzaldehyde
Unknown
C,2H2e Hydrocarbon
C^O-C^eO Ketone
CeH,2O-C8H,8O Aldehyde
C)0-C,3 Hydrocarbon
Octanal
CjjHjo Hydrocarbon
Scan
No.
29
29
31
35
36
38
38
45
48
51
55
55
59
63
63
63
64
71
76
81
82
110
134
166
171
172
199
200
218
224
224
225
266
279
286
307
313
314
318
318
321
322
331
Run No.
Blk


















35
50























31



32

3





1,670
19

138






327











1,020
155
42

320


454

463
32



62

21




208






423
726
423










178

201



281



1,550
13

33
359

368
444

1,081

1,010





453



788
93

437

11









409







374


34



74

135
570
1,660


16




563

1,170
20

869


375
346



417











647


35







364
136
165
13


203



315
42

188











376








11

36

35

22



274
50



22



213
176


157
















113


166

148
37

13










25

17


91
25













570





141




38
44
141


68







53

41

63
201
78








24




41







25


39



19








14
42
13


213
202






34



15


339






109



40

19










19
36




149







19




32
404


18




191


                               63

-------
TABLE C-1 (continued)
Compound
Dichlorobenzene
Unknown
C,0H22-C12H24 Hydrocarbon (alkene)
Unknown
CIOH22-C,3H28 Hydrocarbon
Unknown
Phenyl alkyl ketone
C,-C,S Aldehyde
Unknown
CJ^O-C^O
C12-C15 Hydrocarbon

Total Amount
Scan
No.
335
336
341
342
351
351
364
372
389
414
426


Run No.
Blk












85
31






42





4,685
32

268



216

13
223

18

4,824
33












5,827
34












6,862
35









10


1,823
36












1,376
37



13








895
38












779
39


48

31







1,079
40
99


32



51




1,069
Note: See Appendix D for data on chlorinated dibenzo dioxins and furans.
        64

-------
TABLE C-2.  AMOUNT OF SEMIVOLATILE PICs FOUND IN THE MM5 SAMPLES (jig)

Compound
Alcohol
C,H1C0
Alcohol
C.H,p
C.H100
C,H100
C«HI202
C,H10 (subst. benzene)
Unknown
CsHtoO
Unknown ketone
C7H120
Unknown
C7H,002
C,H100
Benzole acid
Unknown acid ester
1 ,3-lsobenzylfurandione
CSH,002
C10H,002
Unknown
C15H3002
Phthalate, subsL
Unknown acid ester
Unknown acid ester
Unknown
C18HM02
C,,HM02
Bis(2-ethylhexyl)
phthalate

Total Amount
Scan
16
112
122
142
181
202
250
290
301
321
342
396
492
567
640
655
704
791
808
948
955
1099
1209
1223
1235
1268
1347
1359
1610


Run No.
Blk






150















33





42

225
31


308

20

190






18

796

94
16
68





20


30

1,560
32
72

360

960

600


140
128



240
1,120
190
108






100





4,018
33


340

190

440


44





800

78






70



52

2,014
34


380

580

1,200


58





800

86






88



56

3,248
35


360

900

1,380


30





780

72





30
69

33

60

3,714
36


300

189

480








630

75



72

99
246

105
54
84

2,334
37


390

930

660








1,170

84


99



132


45
69

3,579
38


285

249

720




27



540

60




36
27
90
24
36

54

2,148
39


330

330

630





21


780

66

30
36



60



48

2,331
40

1,710
111
165

9,300
660

600



168


1,410

96






51





14,271
Deleted C7H, (toluene) since it was present in XAD/filter blanks and may be a contaminant in XAD or extractant solvent
(also is a volatile POHC).
Deleted C^Ha, (tridecane) and C,Hm (nonane) because these were due to the D/F spiking and were present only in the
samples analyzed for D/F.
Note: See Appendix D for data on chlorinated dibenzo dioxins and furans.
                             65

-------
	TABLE C-3.  POHCs AND PICs IN VOST AND MMS BUNKS

                                                       Amount (ng)
                                             Scan    (on field blank pair
	Compound	No.	from Run 31)

 Volatile POHCs
      Carbon tetrachloride                          -               < 4.0
      Toluene                                    -               < 4 0
 Volatile PICs
Benzene
CSH,8 Trimethyl pentane




Q!l«*«M .•.«.•.•..:.»:_._ — — * 	 * ^

Total Vol PICs (ng) =
76
81
464-
360-
377-
.
35
50
8-
5-
4-
85
                                                       Amount (ug)
                                             Scan     (in blank MMS
                Compound                   No.          train)

 Semivolatile POHCs
      Dichlorobenzene                            -               0.06
      Naphthalene                                -               4.32
 Semivolatile PICs
      C,H12O2                                250                150
      Phthalate, subst                         1209                 33
      Bis(2-ethylhexyl) phthalate                 1610                 42
	Total SV PICs (ug) =	225

 Note:    Silane compounds have not been included in totals, because they
         were likely produced from the reactor's internal insulation and were
         therefore not considered to be 'products of incomplete combustion."
                                66

-------
                        TABLE C-4. RUN 31 POHCs AND PICs

Compound
Volatile POHCs
Carbon tetrachloride
Toluene
Volatile PICs
CS2 (carbon disulfide)
Dichloromethane
Unknown
Silane/methy ethyl ketone
Tetrahydrofuran and chloroform
2-Propylfuran
Hydrocarbon (C9-C13)
Hydrocarbon (C,-C13)
Benzaldehyde
C,2HM Hydrocarbon
Hydrocarbon (C,,-CI3)
Cs,H20 hydrocarbon
Phenyl alkyi ketone
Siljnp
Silgno
Silano containing oontaminanto
Silano containing oontaminanto
Silano containing oontaminanto
Silano containing oontaminanto
Silano containing oontaminanto
Silano containing oontaminanto
Total Vol PICs (ng) =

Compound
Semivolatile POHCs
Dichlorobenzene
Naphthalene
Semivolatile PICs
Alcohol
C,H,00
CeH1202
C7H1002
Benzoic Acid
1 ,3-lsobenzylfurandione
C8H,002
C,0H,002
Unknown
Bis(2-ethylhexyl) phthalate
Total SV PICs (ng) =
Scan
No.

-
-

36
38
55
59
63
110
279
286
307
314
321
331
364
247-
442-
484-
364-
377-
484-
386-
378-

Scan
No.

-
-

122
181
250
567
655
791
808
948
1268
1610

Amount (ng)
(on one trap pair)

21.9
5.6

32
3
1,670
19
138
327
1,020
155
42
320
454
463
42
459-
443-
4,640
14,700
8r430-
223-
466-
48-
4,685
Amount (ug)
(in MM5 train)

0.22
3.70

308
20
190
18
796
94
16
68
20
30
1,560
Note:     Silane compounds have not been included in totals, because they were likely produced
         from the reactor's internal insulation and were therefore not considered to be "products
         of incomplete combustion.'
                                          67

-------
                    TABLE C-5.  RUN 32 POHCs AND PICs
                 Compound
Scan
 No.
  Amount (ng)
(on one trap pair)
Volatile POHCs
        Carbon tetrachloride
        Toluene
Volatile PICs
        Carbon cfisulfide
        Dichloromethane
        Butanal
        C,H,2 Hydrocarbon
        Benzene
        C,H1t Hydrocarbon
        C,HI40 (Furan)
        C7HM0 Aldehyde
        Unknown
        Unknown
        Octanal
        Unknown
        Unknown
        C9-C1S Aldehyde
        Unknown
        C12-CIS Hydrocarbon

        Silano containing
   35
   38
   55
   70
   76
   80
  224
  266
  313
  321
  322
  336
  351
  372
  389
  426
  442-
  485-
           124.9
            23.5

              62
              21
             208
             423
             726
             423
             178
             201
             281
           1,550
              13
             268
             216
              13
             223
              18
        Silano containing oontaminanto
           no containing i
        Silano containing oontaminanto
        Silano containing oontaminanto
  382-
  483-
  288-
                Total Vol PICs (ng)
           6,610
          15 TOO
          13 300
            330-
            286-
             35-
           4,824
                 Compound
Scan
 No.
  Amount (ng)
 (in MM5 train)
Semivolatile POHCs
        Dichlorobenzene
        Naphthalene
Semivolatile PICs
                       0.86
                       8.00
Alcohol
Alcohol
C,H100
C,H1202
C,H100
Unknown (ketone)
C,H100
Benzole Acid
Unknown (acid ester)
1 ,3-lsobenzylfurandione
Unknown (acid ester)
Total SV PICs (ng) =
16
122
181
250
321
342
640
655
704
791
1235

72
360
960
600
140
128
240
1,120
190
108
100
4,018
Note:     Silane compounds have not been included in totals, because they were likely
         produced from the reactor's internal insulation and were therefore not
         considered to be 'products of incomplete combustion.*
                                    68

-------
                   TABLE C-6.  RUN 33 POHCs AND PICs
Compound
Volatile POHCs
Carbon tetrachioride
Toluene
Volatile PICs
Unknown
Unknown
Carbon disulfide
Dichloromethane
C8H,Z Hydrocarbon
Trichloromethane + unknown
C7-Ct Hydrocarbon
Benzene
Cj-C, Hydrocarbon
C,H12O (Unknown)
C7HM (aldehyde)
C10-C13 Hydrocarbon





Total Vol PICs (ng) =
Compound
Semivolatile POHCs
Dichlorobenzene
Naphthalene
Semivolatile PICs
Alcohol
C,H100
C.H.A
C,H100
Benzole acid
1 ,3-lsobenzylfurandione
Unknown (acid ester)
Bis(2-ethylhexyl) phthalate
Total SV PICs (jig) =
Scan
No.
29
31
36
38
44
63
69
76
82
134
265
320
494-
302-
380-
484-
360-
OQ-l...
Scan
No.
122
181
250
321
655
791
1235
1610
Amount (ng)
(on one trap pair)
46.9
41.6
359
368
444
1,081
1,010
453
788
93
437
11
409
374
790-
11 300
o]l40
477-
447-
445-
5,827
Amount (ug)
(in MM5 train)
0.68
7.72
340
190
440
44
800
78
70
52
2,014
Note:     Silane compounds have not been included in totals, because they were
         likely produced from the reactor's internal insulation and were therefore not
         considered to be 'products of incomplete combustion.'
                                    69

-------
                  TABLE C-7. RUN 34 POHCs AND PICs
Compound
Volatile POHCs
Carbon tetrachioride
Toluene
Volatile PICs
Carbon disulfide
Dichloromethane
CSH,2 Hydrocarbon
C,H12 Hydrocarbon
Butanal
C8H100
CeH12 Hydrocarbon
Benzene
C7-C, Hydrocarbon
Octene (C,H,«)
Octane
Unknown
C10-C)3 Hydrocarbon
Silnno





Total Vol PICs (ng) =
Compound
Semivolatile POHCs
Dichlorobenzene
Naphthalene
Semivolatile PICs
Alcohol
C,H100
C.H.A
C,H100
Benzoic acid
1 ,3-lsobenzylfurandione
Unknown (acid ester)
Bis(2-ethylhexyl) phthalate
Total SV PICs (>ig) =
Scan
No.
36
38
38
45
54
63
70
76
84
166
171
218
320
442-
494-
384-
484-
360-
378-
Scan
No.
122
181
250
321
655
791
1235
1610
Amount (ng)
(on one trap pair)
37.2
61.2
74
135
570
1,660
16
563
1,170
20
869
375
346
417
647
1 450
1-4- 500
12 200
458-
407-
34-
6,862
Amount (jig)
(in MM5 train)
0.36
7.72
380
580
1,200
58
800
86
88
56
3,248
Note:     Silane compounds have not been included in totals, because they were
         likely produced from the reactor's internal insulation and were therefore
         not considered to be "products of incomplete combustion."
                                   70

-------
            TABLE C-8. RUN 35 POHCs AND PICs
           Compound
Scan
 No.
                                               Amount (ng)
                                             (on one trap pair)
Volatile POHGs
     Carbon tetrachloride
     Toluene
Volatile PICs
     C,H12 Hydrocarbon
     Unknown
     C,H10 - C7H12 Hydrocarbon
     Butanal
     Trichloromethane
     CeH12 Hydrocarbon
     Benzene
     Cy-C, Hydrocarbon
     C,H,4O aldehyde
     Octanal
     Silano unknown
     Silano containing oontaminanto
     Silono containing oontaminanto
     Silano containing oontaminanto
     Silano containing oontaminanto
     Silano containing oontaminanto
     Silano containing oontaminanto
              Total Vol PICs (ng)
   45
   48
   51
   54
   64
   71
   76
   84
  265
  322
  414
  442-
  484-
  886-
  378-
  484-
  299-
  377-
                                                       158.7
                                                        19.2

                                                        364
                                                        136
                                                        165
                                                         13
                                                        203
                                                        315
                                                         42
                                                        188
                                                        376
                                                         11
                                                         10
                                                       1,470
                                                        356-
                                                      13,600
                                                      10,000
                                                         47-
                                                        427-
                                                        435-
                                                       1,823
           Compound
Scan
 No.
                                               Amount (fig)
                                               (in MM5 train)
Senwolatfle POHCs
     Dichlorobenzene
     Naphthalene
Semivolatile PICs
                   0.30
                   7.89
Alcohol
C,HI00
C8H,Z02
C,HI00
Benzoic acid
1 ,3-lsobenzylfurandione
Unknown (acid ester)
Unknown (acid ester)
CtAA
Bis(2-ethylhexyl) phthalate
Total SV PICs (ug) =
122
181
250
321
655
791
1223
1235
1347
1610

360
900
1,380
30
780
72
30
69
33
60
3,714
Note:     Silane compounds have not been included in totals,
         because they were likely produced from the reactor's
         internal insulation and were therefore not considered to be
         'products of incomplete combustion.'
                             71

-------
                    TABLE C-9.  RUN 36 POHCs AND PICs
                                                Scan        Amount (ng)
                 Compound                      No.       (on one trap pair)

Volatile POHCs
       Carbon tetrachloride                           -                 25.2
       Toluene                                      -                 25.4
Volatile PICs
       Sulfur dioxide                                29                   35
       Carbon cfisulfide                              35                   22
       C,H,j Hydrocarbon                           45                  274
       Unknown                                    47                   50
       2-Butanone                                  59                   22
       Unknown                                    64                  213
       C,H,2 Hydrocarbon                           71                  176
       C7-C, Hydrocarbon                           84                  157
       C12HM Hydrocarbon                          315                  113
       C,0-C13 Hydrocarbon                         321                  166
       CjHjo-C^Ha Hydrocarbon                     331                  148
       Silano unknown                             442-                  924-
       Silano containing contaminants                494-
             i containing oontaminanto                363-
                                                  378               -|Q 6QQ
       Silano unknown                             443-                  74-
       Silano containing oontaminanto                494-                  48-
             i containing oontaminanto                300                  413
                      I oontaminanto                378-                 344-
	Total Vol PICs (ng) =	1,376

                                                Scan        Amount (ng)
                 Compound                     No.         (in MM5 train)

 Semivolatile POHCs
        Dichlorobenzene                              -                 0.36
        Naphthalene                                  -                 6.75
 Semivolatile PICs
Alcohol
C6H100
C9H1202
Benzoic acid
1 ,3-lsobenzylfurandione
C,*H*,02
Unknown (acid ester)
Unknown (acid ester)
C18HM02
C.oH^O,
Bis(2-ethylhexyl) phthalate
Total SV PICs (ug) =
122
181
250
655
791
1099
1223
1235
1347
1359
1610

300
189
480
630
75
72
99
246
105
54
84
2,334
 Note:     Silane compounds have not been included in totals, because they were
          likely produced from the reactor's internal insulation and were therefore not
          considered to be 'products of incomplete combustion.'
                                     72

-------
                   TABLE C-10. RUN 37 POHCs AND PICs
                 Compound
Scan
 No.
  Amount (ng)
(on one trap pair)
Volatile POHCs
        Carbon tetrachloride
        Toluene
Volatile PICs
        Sulfur dioxide
        2-Butanone
        Tetrahydrofuran
        C6H,2 Hydrocarbon
        Benzene
        C7H14O aldehyde
        C7H14O-C,H,,O ketone
        Unknown
        Silano unknown
        Silano unknown
        Silano unknown
        Silano unknown
        Silano unknown
        Silano unknown
        Silano unknown
        Silano containing contaminanto
        Silano containing contominanto
        Silano containing contaminants)
        Sib no
   44
   60
   65
   71
   78
  265
  318
  342
  322-
  336-
  462-
  443-
  442-
  462-
  426-
  494-
        Silnno
        Silano containing oontaminanto
        Silano containing contaminants
        Silano containing oontaminanto
                Total Vol PICs (ng) =
  364-
  442-
  482-
  482-
  485-
  366-
  378-
             40.9
             22.0

              13
              25
              17
              91
              25
             570
             141
              13
             c(9O"
             467-
             453-
              82-
           7 Q10
              67-
              46-
             386-
          •|Q "7QQ
          ^•« 3QQ
             '448-
              66-
              44-
              27-
             326-
             664-
             895
                 Compound
Scan
 No.
  Amount (ng)
 (in MM5 train)
Semivolatile POHCs
        Dichlorobenzene
        Naphthalene
Semivolatile PICs
                       0.45
                       9.09
Alcohol
C.H100
C,H1202
Benzoic acid
1 ,3-lsobenzylfurandione
Unknown
Unknown (acid ester)
ClsHgoOj
Bis(2-ethylhexyl) phthalate
Total SV PICs (ng) =
122
181
250
655
791
955
1235
1359
1610

390
930
660
1,170
84
99
132
45
69
3,579
Note:    Silane compounds have not been included in totals, because they were likely
         produced from the reactor's internal insulation and were therefore not
         considered to be "products of incomplete combustion.'
                                    73

-------
                   TABLE C-11. RUN 38 POHCs AND PICs
                 Compound
                                               Scan
                                               No.
  Amount (ng)
(on one trap pair)
Volatile POHCs
        Carbon tetrachloride
        Toluene
Volatile PICs
            13.6
            15.1
Unknown
Sulfure dioxide
Unknown
2-Butanone
THF (tetrahydrofuran)
Unknown
C,H12 Hydrocarbon
Benzene
Unknown
C7HMO Aldehyde
C10-C1S Hydrocarbon
Silano unknown
Silano unknown
Silano unknown
Silano unknown






Silano unknown
Silano unknown
Silano unknown





Total Vol PICs (ng) =

Compound
29
35
36
57
64
64
71
76
200
266
320
318
462-
443-
462-
4S4-
362-
377^-
442-
462-
494-
464-
OQQ
37?-

Scan
No.
44
141
68
53
41
63
201
78
24
41
25
74-
50-
3 50Q
65-
479-
3 2 "900"'
•"13-QQQ-
'364-
472-
75-
64-
283-

779
Amount (ug)
(in MM5 train)
 Semivolatile POHCs
        Dichiorobenzene
        Naphthalene
 Semivolatile PICs
             0.21
             5.55
Alcohol
C,H100
C,H,A
C7H,20
Benzoic acid
1 ,3-lsobenzylf urandione
Unknown (acid ester)
Unknown (acid ester)
Unknown
C, H O2
Phthalate, subst.
Bis(2-ethylhexyl) phthalate
Total SV PICs (ug) =
122
181
250
396
655 .
791
1223
1235
1268
1347 '
1209
1610

285
249
720
27
540
60
27
90
24
36
36
54
2,148
 Note:    Silane compounds have not been included in totals, because they were likely
          produced from the reactor's internal insulation and were therefore not
          considered to be 'products of incomplete combustion.'
                                   74

-------
                   TABLE C-12. RUN 39 POHCs AND PICa
                 Compound
                                        Scan
                                         No.
           Amount (ng)
         (on one trap pair)
Volatile POHCs
        Carbon tetrachioride
        Toluene
Volatile PICs
        Carbon disulfide
        2-Butanone
        Trichloromethane
        C4H8O Tetrahydrofuran + chloroform
        C,H,2 Hydrocarbon
        Benzene
        Unknown
        1,1,1-trichloro-2-propanone
        C7H120 Aldehyde
        C,H120-C,H,,0 Aldehyde
        CH>Hz2-C12H24 Hydrocarbon (alkene)
        C^Hjj-C^Hj, Hydrocarbon
        Silano unknown
        Silano unknown
        Silano unknown
        Silano containing oontaminanto
              containing oontaminanto
              containing
Silano unknown
Silano unknown
Silano unknown
Silano unknown
Silano containing oontaminanto
Silano containing oontaminanto
Silano containing oontaminanto
         Total Vol PICs (ng) =
  36
  59
  64
  64
  71
  77
 172
 224
 265
 318
 341
 351
 336-
,462-
 444-
 494-
 364-
 377-
 364-
                                                               125.8
                                                               27.6

                                                                 19
                                                                 14
                                                                 42
                                                                 13
                                                                213
                                                                202
                                                                 34
                                                                 15
                                                                339
                                                                109
                                                                 48
                                                                 31
                                                                 43-
                                                                 66-
                                                                4€5-
                                                             11 400
                                                   462-
                                                   494-
                                                   494-
                                                   377-
                      469-
                      468-
                      463-
                       66-
                       39-
                      362-
                      767-
                     1,079
                  Compound
                                         Scan
                                         No.
            Amount (ug)
           (in MM5 train)
 Semivolatile POHCs
        Dichlorobenzene
        Naphthalene
 Semivolatile PICs
                                                                0.15
                                                                5.91
Alcohol
C,H100
C,H1202
Unknown
Benzoic acid
1 ,3-lsobenzylf urandione
C10H10O2
Unknown
Unknown (acid ester)
Bis(2-ethylhexyl) phthalate
Total SV PICs (fig) =
122
181
250
492
655
791
948
955
1235
1610

330
330
630
21
780
66
30
36
60
48
2,331
 Note:    Silane compounds have not been included in totals, because they were likely
         produced from the reactor's internal insulation and were therefore not
         considered to be "products of incomplete combustion."
                                    75

-------
                   TABLE C-13. RUN 40 POHCs AND PICs
                 Compound
Scan
 No.
  Amount (ng)
(on one trap pair)
Volatile POHCs
        Carbon tetrachloride
        Toluene
Volatile PICs
        Sulfur dioxide
        2-Butanone
        Trichloromethane + unknown
        Benzene
        CeH12O Aldehyde
        Chlorohydrocarbon (chloroalkane)
        C^p Aldehyde
        Benzaldehyde
        C,0-C,3 Hydrocarbon
        Dichlorobenzene
        Unknown
        C.H1.0-C10HM0 Aldehyde
        Silano unknown
        Silano unknown
        Silano unknown
        Silano unknown
        Silano containing oontaminanto
        Silano containing oontaminonto
        Silano containing oontaminanto
        Silano unknown
        Silano unknown
        Silano unknown
        Silano containing oontaminanto
        Silano containing oontaminanto
        Silano containing oontaminanto
                 Total Vol PICs (ng) =
   41
   59
   65
   77
  199
  225
  256
  306
  321
  335
  341
  342
  402-
  442-
  466-
  462-
  496-
  303-
  379-
  442-
  462-
  494-
  494-
  279-
  377-
           230.0
             14.7

              19
              19
              36
             149
              19
              32
             404
              18
             191
              99
              32
              51
              74-
           2,310
              37-
              46-
             234-
           -«j4 QQQ
           -)3 400
             357-
             252-
              23-
              24-
             382-
                      1,069
                 Compound
Scan
 No.
  Amount (pg)
  (in MM5 train)
Semivolatile POHCs
        Dichlorobenzene
        Naphthalene
Semivolatile PICs
                       1.04
                       5.75
C,H100
Alcohol
CSH140
C,H100
CSH12O2
Unknown
Unknown '
Benzoic acid
1 ,3-lsobenzylfurandione
Unknown (acid ester)
Total SV PICs (pg) =
112
122
142
202
250
301
492
655
791
1235

1,710
111
165
9,300
660
600
168
1,410
96
51
14,271
 Note:    Silane compounds have not been included in totals, because they were likely
         produced from the reactor's internal insulation and were therefore not
         considered to be "products of incomplete combustion."
                                    76

-------
                  APPENDIX D






TABULATION OF ANALYSIS RESULTS FOR DIOXINS/FURANS
                      77

-------
  (T
  .
B W
5 -
!i

       .             .
       « 8 S fe  SJS 5
       "    --  --  "
         r-- CM' tvi cvi° 5
       _ O O O O|O CM

         § S Si
         g? § 5 5 £ ®.
                   CO U>
         »- 
 co
            w -r- cnleo .
            °° si mlro
             TT
       CO" CM U) O> CO I CO

       Ji.X^.eo ~l"
       o Q a
I§1§
                             § § 8 § 81
                             g co (^ co_ -H

                             «? fc 8 8 S|
                      1181 s

                      of £ £ co
                                              10
                                             * i-^
                                              10
               o g

               o °
               co fcT
               Is- o>
                                            0 O
                                            or--
                                            1— U>
                             eo co o> os rtl   r~ 53
                                           •co Jt
o o o o ot   ^ ^


™"u,-rf^(   Set
                             v ^ v in -^:|  cq
                             a o z
                             z z
                                    COQ
                             Q O
                 O
                                                     N.
                                                     CO
                                                     to
                                              3 «

                                              co S.
                                              co" °
                                              co ^
                        O CM
                        in ro
                                                     CO


                                                      V
                         I
                         v
                               U. U. U- ,,
                             U- Q Q Q £•

                               ^ £8
                               Q. I I 0
                                            •
                                                   •
                                              s
                                                        I
                                                     -=-==
                                                        I
                                                               -r-CT>itCO»-
                                                                 corfcnincncocn
                                                               Of-COCOcOCMOJCOO

                                                               °°ro§^'-^™'-
                                                                    yjt^


                                                               COO)   r^*"O5O>CDCO
                                                               8
                                                     8
                                                                 T-cocMcnco^iocM
                                                             S?
                                                               U. 0
                                                               Q Q
                                                               0 0
                                                                              u. O
                                                                         u. Q Q Q

                                                                    |gSS||
                                  COCOCMOJCMC\fcMCMO
                                    ""''''"

                                               78

-------
            APPENDIX E
LIQUID ORGANIC FEED ANALYSIS RESULTS
        FOR Cl, ASH, AND HHV
                79

-------
   TABLE E-1.  LIQUID ORGANIC FEED ANALYSIS RESULTS

Run 31
Run 33
Run 35
Run 35 duplicate
Total organic
chlorine (%)
28.27
28.43
28.33
28.20
% Ash
< 0.006
< 0.006
< 0.006

Higher heating
value (Btu/lb)
13,064
12,830
12,871
12,954
TABLE E-2,.  SUMMARY OF GALBRAITH AUDIT SAMPLE RESULTS
Higher heating value (Btu/lb)
Sample 30001
Sample 30002
Sample 30003
Prepared value
11,906

Prepared value
1.03

Prepared value
24.29
Reported value
11,772
% Ash
Reported value
0.93
Organic chlorine (%)
Reported value
26.20
% Accuracy
99%

% Accuracy
90%

% Accuracy
108
                         80

-------
          APPENDIX F
O2, CO2, AND H2O ANALYSIS RESULTS
 FOR EFFLUENT GAS IN EACH RUN
              81

-------
    TABLE F-1.  O2, CO2, AND WATER ANALYSIS RESULTS

Run 31
Run 32
Run 33
Run 34
Run 35
Run 36
Run 37
Run 38
Run 39
Run 40
GEM
average
%O2
(dry basis)
9.3
12.3
11.9
9.0
7.9
3.2
8.7
4.2
6.7
11.8
Orsat analysis
%O2
9.0
12.0
12.0
8.8
8.0
3.4
8.4
5.0
7.0
11.3
% CO2
11.0
6.6
6.4
8.4
9.6
12.4
8.8
10.8
9.6
6.7
Percent3
moisture
7.6
8.7
10.0
11.5
10.6
17.5
10.9
12.2
12.3
8.8
a Percent moisture (by volume) determined from MM5
  sampling train.
                           82

-------
             APPENDIX G






SUMMARY OF QUALITY ASSURANCE RESULTS
                 83

-------
               SUMMARY OF QUALITY ASSURANCE RESULTS
      Several quality assurance procedures were used as part of the 10 tests done
with the EPA sampling and analysis methods. These included:

•     QA audit samples

•     Surrogates spiked into all VOST and MM5 samples

•     Analysis of a blank MM5 train and blank VOST traps

•     Analysis of POHCs in liquid organic feed

•     Calibration of CEMs

•     Calibration of sampling equipment


QA AUDIT SAMPLES

      The following audit results were all within QA objectives for accuracy, except for
one determination (naphthalene check standard).4
Description
Simulated liquid
organic feed (spiked
POHCs in fuel oil)
Semivolatile check
standard
VOST/VOA water
check standard
Identification
04797
04801
04796
Analyte
o-Dichlorobenzene
Carbon tetrachloride
Naphthalene
Toluene
o-Dichlorobenzene
Naphthalene
Carbon tetrachloride
Toluene
Spike
level
43.0% w/w
12.24%
2.23%
5.47%
29.5 ng/mL
67.6 ng/mL
87 u.g/mL
113 ng/mL
Found
level
48.1%
12.8%
2.3%
5.2%
30.3 ng/mL
94.1 pg/mL
66.8 )ig/mL
86.1 ng/mL
Accuracy
112%
105%
103%
95%
103%
139%'
77%
76%
a Value is outside QAP accuracy objective of 75% to 125%, However, standard calibrating data were validated
using another independent standard solution (agreement of 109%).
                                    84

-------