-------

-------
                                   TECHNICAL REPORT DATA
                            (Pl\xtt tttd Inttnifixiit on tit rcttnr
 . REPORT NO.

   EPA/600/A-86/01S
                             a.
4. TITLE ANDSUBTITL6
      EPA Method Study 33: Ignitabllity Characteristics
         of Solids                             ,
                                                           ». REPORT OATS
                                                               February 1986
                                                           1. PlflPORMINO ORGANIZATION CODE
7. AUTHOR1SI                           .
     Handy, R.W.; Michael,  L.C., Mclaughlin, C.E.
     and Pellizzari.  E.D.
                                                            I. PIRFORMINO ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
     Research Triangle Institute
     P.O. Box 12194
     Research Triangle Park, NC  27709
                                                            10. PROGRAM ELEMENT NO.
                                                              BSD1A
                                                            It. CONTRACT/GRANT NO.
                                                              68-03-3099
 12. SPONSORING AGENCY NAME AND ADDRESS
     Environmental  Monitoring and Support Laboratory
     Office  of Research and Development
     U.S..Environmental Protection Agency
     Cincinnati,  OH  45268
                                                            13. TYPE OF REPORT AND PERIOD COVERED
                                                            14. SPONSORING AGENCY CODE
                                                              EPA600/6
 15. SUPPLEMENTARY NOTES
 16. ABSTRACT
         The objective of  this  research was to conduct an  interlaboratory study to
    estimate the precision of three 1gnitab1l1ty test methods  for solid wastes:
    radiant heat ignition, flame propagation and water extinguishabilfty.  The study
    consisted of two phases.  The first phase involved the replicate measurements of
    two well characterized test materials and a reference  material  by nine laboratories.
    The second phase included similar replicate measurements of seven varied test
    materials by up to five  laboratories.
         The study results indicated that not all sample types are amenable to
    measurement by each test.   Although all of the procedures  showed some deficiency,
    the test results indicated  that, with some changes 1n  test chamber design they are
    potentially useful as  routine methods.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.lDENTIFIERS/OPEN ENDED TERMS  C.  COSATI Field/Croup
18. DISTRIBUTION STATEMENT

     Release to public
EPA Farm 2220-1 (R.». 4-77)   *HC»IOU1 EDITION IS OBlOLtTC
                                               19. SECURITY CLASS (This Report!
21. NO. OF PAGES
                                                  Unclassified
                                               20. SECURITY
                                                                          22. PRICE

-------

-------
                                   •FOREWORD

    Environmental measurements are required to determine the quality of
ambient waters and the character of wastes and waste effluents.  The
Environmental Monitoring and Support Laboratory - Cincinnati (EMSL-C1nc1nnat1)
conducts research to:

         Develop and evaluate techniques to measure the presence and
         concentration of physical, chemical, and radiological pollutants 1n
         water, wastewater, bottom sediments, and solid wastes.

         Investigate methods for the concentration, recovery, and
         Identification of viruses, bacteria, and other microorganisms 1n
         water.

         Provide for an Agency-wide quality assurance program to assure
         standardization and quality control of systems for monitoring water
         and wastewater.

    This publication of the EMSL-Clnclnnatl, entitled:  "USEPA Method Study
33. Ign1tab1l1ty Characteristics of Solids" reports the results of a study of
three  separate  1gn1tab111ty parameters; radiant heat Ignition, flame
propagation, and water ext1ngu1shab1!1ty.  Federal agencies, states,
municipal Hies, universities, private laboratories, and Industry should find
this evaluative study helpful 1n their efforts to establish 1gn1tab111ty
criteria for solid and hazardous wastes.
                                      Robert L. Booth
                                      Director, EMSL-C1nc1nnat1
                                      ill

-------
                                    ABSTRACT    ;;;  ,,-v,.


    The objective of this research was to'conduct  an  Inter!aboratory study to
estimate the precision of three 1gn1tab111ty  test  methods for solid wastes:
radiant heat Ignition, flame propagation and  water ext1ngu1shab1Hty.  The
study consisted of two phases.  The first phase Involved the replicate
measurements of two well characterized test materials and a reference material
by nine laboratories.  The second phase Included similar replicate
measurements of seven varied test materials by up  to  five laboratories.

    The study results Indicated that not all  sample,types are amenable to
measurement by each test.  Although all of the procedures showed some
deficiency, the  test results Indicated that, with some changes 1n test
chamber design they are potentially useful as routine methods.
                                       iv

-------
                                    Contents

Section                                             ,ti;                Page

u av. i aimer. •••••••••.«•«,,««..,,«,..,,»»,,».,««»«»,,««,»««...,•••.,•    11
Foreword [[[   ill
Hus tract • ««••»«««•»•»,«,«,,,,.,,,,.,,,.,,,,,,,,«,,,,««••••.«««,«....    lv
I aU I cS •«•••«««••••••.•«»«,,
1.  INTRODUCTION .............................. '. .....................     1

2.  SUMMARY [[[     2

3.  RECOMMENDATIONS .................................................     4

4.  DESCRIPTION OF STUDY ....................................... .....     S

          Test  Design ...................... ....... ...................     5
          Methodology  ........... ................. ,... ...... . ........     S
          Participating Laboratories .................................     S
          Phase 1 [[[     6
          Selection and Preparation of Wastes ........................     6

S.  TREATMENT  OF  DATA ...............................................     8

6.  DISCUSSION AND CONCLUSIONS ............................ . .........    10

          Phase I  Results ........................ . ...................    10
          Outliers [[[    17
          Phase II Results ...........................................    17
          Statistical  Summary and Precision Statement ................    17
          Evaluation of Methods ............... . ......................    17


-------
                                     TABLES
Number
 1
 2
 3
 4
 5
 6
 7
 8
 9
 10
 11
 12
 13
 14
 !5
 16
 17
 18
 19
 20
 21
 22
                             r. »«•«•*«<
Phase
Phase
Phase
Phase
Phase
Phase
Phase
Phase
Phase
Phase
Phase
Phase
Phase
Phase
Phase
Phase
Phase
Phase
Phase
                                            i •••••••i
Phase II - Statistical Summary
Physical Description of Test.Materials	
Phase I Results - Radiant Heat Ignition Test.
      I Results - Flame Propagation Test	
      I Results - Water Ext1ngu1shab1l1ty Test
      II Results - Radiant Heat Ignition Test
                           Heat
                           Heat
      »••*•••««•>•«•««>«•••
      Test Material
      Test Material
      Test Material
      Test Material
      Test Material
                No.
II Results - Radiant Heat Ignition Test
II Results - Radiant Heat Ignition Test
II Results - Radiant Heat Ignition Test
II Results - Radiant Heat Ignition Test
II Results • Flame Propagation Test - Test Material
II Results - Flame Propagation Test - Test Material
II Results - Flame Propagation Test - Test
II Results - Flame Propagation Test
II Results - Flame Propagation Test
II Results - Water Ext1ngu1shab1l1ty Test
II Results - Water Ext1ngu1shab1Hty Test
II Results - Water Ext1ngu1shab1l1ty Test
II Results - Water Ext1ngu1shab1l1ty Test
II Results - Water Ext1ngu1shab1l1ty Test
II Results - Water Ext1ngu1shab1l1ty Test
II Results - Water Ext1ngu1shab1Hty Test
       Material
- Test Material
- Test Material
        Test
        Test
        Test
        Test
        Test
        Test
        Test
   No.
   No.
   No.
   No.
Material
Material
Material
Material
Material
Material
Material
No.
No.
No.
No.
No.
3..
4..
5.,
8..
9..
           5,
           6,
           7,
           8,
           9,
                                                                No.
                                                                No.
                                                                No.
                                                                No.
                                                                No.
                                                                No.
                                                                No.
Page

  3
  7
 11
 13
 15
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
                                        vi

-------
                                   SECTION 1

                                  INTRODUCTION

    The United States Environmental  Protection Agency (USEPA) has specified 1n
the Federal Register (1) that a solid waste .exhibits  the characteristic of
1gn1tablllty if "it 1s capable of causing'fire through friction, absorption of
moisture or spontaneous chemical changes and, when Ignited, burns so
vigorously and persistently that 1t creates a hafard." At the  present time,
there are no suitable, validated procedures for determining the 1gn1tab1l1ty
characteristics of solid (non-11qu1d) wastes.

    The objective of this Interlaboratory study was to assess precision and
potential of application of three 1gn1tab1l1ty tests; radiant heat  Ignition,
flame propagation, and water extlngulshabHlty.  This report contains a
description of this study, an assessment of the three tests of  1gn1tab1l1ty
characteristics and recommendations for further development of  the  methods.

-------
                                   SECTION 2

                                    SUMMARY

    This evaluation, the first for evaluation of potential  testing methods  for
sol Ids, was performed at the Research Triangle Institute (RTI)  (2).
Techniques for 1gn1tab1Hty measurements were developed and tested on  a
variety of waste materials.  Three 1gn1tab1l1ty characteristics,  radiant  heat
ignition time, flame propagation rate and water ext1ngu1shab1!1ty were tested
on seven waste materials by three to five laboratories.  Measured values  for
each test were statistically analyzed.  The overall means,  precisions  and
percent relative standard deviations for all  waste types are shown 1n  Table 1.

    Based on the data acquired, the following conclusions were  made.

         Major problems with equipment and procedural details were discovered
         in this collaborative effort.  These deficiencies  must be corrected
         before the methods can be collaboratlvely tested or recommended  for
         use as standard procedures.

         Two of the largest sources of error arise from the recording  methods
         used and lack of precise, beginning and end points  of the tests.

         The three methods tested are potentially useful for evaluating the
         ignitability hazards posed by solid wastes, however, considerable
         standardization and automation of the test procedures  will be
         required.

         Exposure of some waste types (cotton fiber, polyurethane foam) to  a
         radiant heat source Is not sufficient to Ignite the samples.

         The flame propagation and water extlngulshablHty  tests  could not  be
         performed on petroleum-based wastes, as the wastes were  not readily
         ignitable by a flame as specified 1n the method.

-------
                                             TABLE  1.   Phase  II Statistical Summary
CJ
Radiant Heat
Test Material
No.
No.
No.

No.


No.

No*

No.

3 Cotton Fiber
Day 1
Day 2
4 Polyurethane Foam
Day 1
Day 2
5 Paint Waste
Day 1
Day 2
6 Waste 011
Day 1
Day 2
7 Waste 011 /Sand
Day 1
Day 2
8 Waste Solvent
Day 1
Day 2
9 Sawdust/Kerosene
Day 1
Day 2
X (Sec)
ND
ND
ND
ND

98(n=5)
104(n=-4)

119(n=4)
151(n»4)

151(n»4)
127(n=4)

49(n-4)

16(n-5)
14(n=4)
Sr


52
25

17
23

39
6
,.>•-;•;
'*T6

6
3
Ignition
S


56
63

38
152

68
63

23
9

6
3
%RSD


58
60

32-
101

45
50
.
38
19

35
24
Flame Propagation
S (cm/sec) Sr S %RSD
4.3 (n=3) 1.4 2.3 53.8
2.6 (n»l) * * *
0.3 (n-3) 0.05 0.07 21.4
0.3 (n-1) * * *

0.04(n-4) 0.1 0.01 28.8
0.03(n»2) 0.005 0.004 13.6

ND
ND

ND
ND

0.1 (n-3) 0.1 0.1 100.0
0.1 (n-1) * * : *

OJl(n»4) 0.04 0.11 104.3
0.20(n»2) 0.03 0.02 12.3
Water Ext1ngu1shab1lity
2 (mL)
4.3(n=4)
2.4(n»3)
33.3(n-3)
4.0(n»2)

2.3(n»4)
2.8(n»3)

4.9(n-1)
2.8(n-l)

2.5(n-1)
2.9(n»l)
*
4.9(n-3)
5.2(n=2)

2.7(n-4)
2.9(n-3)
Sr
0.5
1.5
2.3
0.5

0.7
0.8

*
*

" *
*

0.8
0.6

1.1
0.8
S
3.0
1.3
2.6
2.4

1.1
1.3

*
*

*
*

2.5
2.2

1.4
1.3
%RSD
69.4
55.3
77.6
61.0

49.2
44.8

*
*

*
*

50.8
41.9

51.8
43.2
     *   a value cannot be calculated from existing data
     n   a number of laboratories reporting data
     ND  » No data reported

-------
                                   SECTION 3        :

                                RECOMMENDATIONS!  §v^  '

    Based on the study results, observations regarding the methodology,  and
comments from study participants, the following recommendations  are  presented
for future work:

         For the radiant heat Ignition test, develop an  automated system which
         activates the radiant heat source when the chamber Interior reaches
         the desired temperature.

         Introduce an elapsed time device to accurately measure beginning and
         end points of the radiant heat and flame propagation tests.

         To avoid false starts, lengthen the sample trough used 1n the flame
         propagation test so that sample may be Ignited without activating the
         first  thermocouple.  A narrower trough would also minimize
         misdirected propagation, giving more consistent results.

         Develop a mechanism for generating a uniform spray pattern for the
         water  ext1ngu1shab111ty test.  Automate the spray pulsing system to
         maximum extent possible.  Incorporate a more accurate means of
         measuring water delivery volume.

         Define the exact endpolnt of the ext1ngu1shab1!1ty test.

-------
                                    SECTION 4

                              DESCRIPTION OF STUDY

Test Design

    The Interlaboratory study was conducted to measure the precision of three
measurements of 1gn1tab1l1ty characteristics of solids.  The study was
composed of two phases: an Initial familiarization phase and a second phase 1n
which an Increased number of wastes was evaluated.  For each waste type and
1gn1tab1l1ty characteristic, participants were required to perform six
replicate measurements—three on each of two days.  Analysis of the reference
material was required  before and after each triplicate run.

Methodology

    The test methods evaluated  1n this study, 1gn1tab1Hty, flame propagation
and water ext1ngu1shabH1ty, were developed as a means of judging the
combustion characteristics of waste materials.  The details of these methods
are presented  1n  Appendix A.  The test chamber described 1n Appendix B was
used  for'all evaluations.   In the radiant heat Ignition test a radiant heat
source  is placed  6 cm  from the  surface of a waste material.  Ignition is
detected by a  thermocouple sensor placed above the sample surface and the time
required for ignition  is recorded on a strip chart recorder.  The rate of
flame propagation 1s measured by recording the time required for a flame to
burn  a  premeasured distance between two thermocouple sensors placed a fixed
distance above the sample surface.  Results are then expressed as cm/sec.  The
axtinguishability characteristic is measured by Igniting the material and,
when  the surface  of the sample  is completely aflame., extinguishing with a
calibrated spray  of water, the  volume of which is used as the measure of
extinguishability.

Participating  Laboratories

    Approximately 20 laboratories were contacted by Research Triangle
Institute (RTI) for an indication of interest.  The initial list of contacts
was provided jointly by the Office of Solid Waste (OSW), the Environmental
Monitoring and Support Laboratory, Cincinnati (EMSL-C1ndnnati), and the
American Society  for Testing and Materials (ASTM).  Representatives of each
laboratory were informed of the nature and objective of the study and notified
that  participation would be on  an unpaid basis.  Eight laboratories agreed to
participate, however,  three laboratories of the eight that participated in
Phase I withdrew  from  Phase II  of the study.  The five remaining laboratories
submitted data from the radiant heat ignition test.  Four of the five also
returned results  from  the flame propagation and water extinguishability
tests.  The final study, therefore, Included the five laboratories which are
1isted below:

-------
    Dow Chemical  Company
    Texas Division
    Freeport,  TX   77541

    Gull ford Laboratories
    827 Huffman Street
    Greensboro, NC  27405

    Research Tr1angle Institute
    P.O. Box 12194
    Research Triangle Park, NC  27709

    Safety Consultants Engineering, Inc.
    5240 Pearl Street               .              .
    Rosemont, IL  60018

    Twin City Testing and Engineering Laboratory, Inc.
    662 Cromwell Avenue
    St. Paul, MN  55114

Phase  I

    In  this study, an Initial Phase I was Included to Increase the reliability
of Phase II measurements by 1) familiarizing the analysts with the ignita-
bility methodologies and 2) insuring accurate calibration and consistent
operation of the ignitability apparatus prior to Initiation of Phase  II.  Each
of nine laboratories war. provided an RTI-built.testing apparatus  (Appendix  B)
a reference material, and aliquots of two wastes, chosen because  both had been
characterized previously in a single-laboratory evaluation (2).  Wood wool
excelsior was chosen as the reference material because of its use by
Underwriters' Laboratories as a standard for class A fire extinguishers and
its consistent behavior in the single-laboratory evaluation.  A summary of  the
Phase  I data 1s presented in Section 5, Discussion and Conclusions.

Selection and Preparation of Hastes        .

    The two materials used in Phase I were lignite-coal/xylene and red oak
sawdust/kerosene mixtures.  The lignite coal was dried in a fume  hood, ground,
and sieved to 40-60 mesh and placed 1n containers.  The material  was  then
spiked with reagent grade xylene and mixed on a rotary blender for 20 minutes
to give a 25% (w/w) xylene concentration In the mixture.  Red oak sawdust from
a sawmill was collected, air dried, placed In containers, spiked  with reagent
grade kerosene, and mixed on a rotary blender for 20 minutes, yielding a 20%
kerosene concentration 1n the mixture.

    Test materials 3 through 9 were used 1n Phase II of the study. All wastes
are described 1n  Table 2.

-------
              Table 2.  Physical Description of Jest Materials
                              Phase  I Materials

1.  Coal/Xylene.  Lignite coal was ground to approximately 10 mesh and
    combined with reagent grade xylenes to yield a 25% (w/w) mixture of
    xylenes and coal.

2.  Sawdust/Kerosene.  Red oak sawdust was dried for 24 hours at ambient
    temperature.  A mixture of 20X  (w/w) reagent grade kerosene and dried
    red oak sawdust was prepared.

                              Phase  II Materials

3.  Cotton Fiber.  Commercially available.

4.  Polyurethane Foam.  A one-Inch  thick foam of the type used for padding
    on furniture or in pillows.  Submitted as a single piece.  Commercially
    available.

5.  Paint Waste.  A yellow-colored  composite paint waste obtained from a
    hazardous waste disposal  facility.  This material had the consistency of
    putty and a strong acetate odor.  There was.no apparent separated
    solvent on surface.  Characteristic of paint manufacturing residual.

6.  Waste Oil.  A composite of used oil collections from local service
    stations.  This material  had the consistency of molasses.

7.  Waste Oil/Sand.  A mixture of a light waste oil, the same as used in
    Waste No. 6, (60%) and builder's sand (40%).  Some stirring needed to
    maintain sample homogeneity.

8.  Waste Solvent.  A viscous orange-colored composite of waste materials
    obtained at Triangle Resources  Industries, Reidsville, NC  27320.

9.  Same as No. 2.

-------
                                   Section 5

                               TREATMENT OF DATA

    The purpose of the data generated 1n this  collaborative study was to
estimate the single and overall precision of measurements of the 1gn1tab111ty
characteristics of solid wastes.  Because there are no true values-associated
with the measurements for each material tested, statistical analyses are
necessarily limited to estimates of precision.  Prior to statistical analyses,
all data were -tested for outliers using an Iterative Cochran's test for
homogeneity of variances at the 95* confidence leve.1. 'According to Cochran,
If a standard deviation of one set of replicates Is significantly different
from the other standard deviations 1n the group, then that set belongs to a
separate population and can be rejected subject to the significance level
criteria below.  The criteria for the rejection was 0.05 significance level
for Cochran's "C" given by the formula shown below:
                largest
                 "1

               Si = standard deviation of the 1th pair of duplicates

 Data rejected  using Cochran's test are denoted by "**" following  the results
 in  Phase  II data tables  (Appendix C).  No further attempts  were made to
 identify  outlying data.  With the unusually small amount of data  returned by
 study participants 1t was feared that elimination of erratic but  valid data
 points would occur, biasing the final results.

    For each laboratory  and waste material, ,the traditional mean  and standard
 deviation of daily triplicate measurements were calculated.  Single-laboratory
 precision estimates (Sr) were calculated from these,data as follows:


 Single-Laboratory/Analyst Standard Deviation:
                    „   c2   . n   c2  .        4. „   c2
                    "LI \1   nL2 \2   * * *   nL1  b1
                        nLl + nL2 * • •  • * njj



    where: nL-f = number of replicate determinations  performed by  laboratory 1

            2
               = variance of replicate measurements  by laboratory 1
                                     8

-------
To characterize the multllaboratory performance of the methods, the overall
meanX, the overall precision (S) and the percent relative standard deviation
XRSDV were computed as follows:

     Overall Mean:        X •    1   "    v
                                 ST   *    ««          11
                                 "   1-1   i

    where:   n • number  of laboratories
             T » mean of measurements reported by laboratory 1


 Overall  Precision:
                            (V
      where S- » standard deviation of means  of n laboratories

                  p
 For this study, So has been estimated from means of 3 replicate observations in

 each laboratory, thus 1t Includes all the variability common between
                p
 laboratories (Sg)» but only one third of the variability common within each

 laboratory (S /3).  Since we want S to Include all variability for a single

 analysis in any laboratory, therefore for this study;


              99???       ?       ??
             r£ _  r*-  t  <•£•  _  et  .  ^C   .  jfC.   _  rC.   . y^C.
             S  -  SB  +  Sr  -  SB  +  Sr   + 25r   -  5X  + 25r
                                               ~r

 Percent Relative Overall Standard Deviation:
     %RSD =  4  x  100

-------
                                    Section 6

                           DISCUSSION AND CONCLUSIONS

Phase I Results                           •

    The Importance of requiring participants to analyze preliminary samples
prior to the formal study has been shown 1n previous method validation
studies.  Such analyses help provide more realistic data by: 1) familiarizing
analysts with analytical methods; 2) allowing resolution of equipment or
procedural problems; and 3) screening laboratories that are either unwilling
or unable to solve analytical difficulties.  For this collaborative effort, no
attempt was made to disqualify laboratories, but only to familiarize them with
the test chambers and samples.  Each laboratory received sufficient amounts of
each of two test materials for measurement of the three 1gn1tab1l1ty
characteristics.  The results of  Phase  I measurements are shown 1n Tables 3-5.

    The mean  Ignition times reported by six laboratories for test material 1
 (coal/xylene) was 24.0  seconds on day 1 and 24.5 on day 2.  Overall precision
S of those measurements was 11.1  and 7.8 seconds, respectively, representing
relative standard deviations  (XRSD) of  46.3 and 32.0 percent.  Ignition time
measurements  on test material 2  (sawdustAerosene) were times of 16.0 and 23.5
seconds for day 1 and day  2 measurements, respectively, with relative standard
deviations of 27.2 and  47.4 percent.  The reference material, wood wool
excelsior, measured before and after each set of three samples, performed
consistently  on both days, giving means of 49.5 and 51.4 seconds and relative
standard deviations of  14.1 and 5.7 percent.  The results obtained from test
materials 1 and 2 do not appear to have been influenced significantly,
therefore, by fluctuations 1n the operating characteristics of the test
chambers.                        '   •              "

    Results of flame propagation  tests  reported by seven laboratories for test
material 1 yielded a mean  propagation rate of 0.49 and 0.57 cm/sec for day 1
and day 2 measurements, respectively, with relative standard deviations of
43.3 and 25.45L  Precision for test material 2 was slightly poorer, as
indicated by  day 1 and  day 2 relative standard deviations of 40.9 and 50.0*,
respectively.  Only four laboratories reported flame propagation data for test
material 2.   All of the remaining laboratories reported that a flame would not
propagate over the entire  surface of the samples.  Since this problem did not
occur  in the  single laboratory evaluation of these methods (2) or in the
analysis of the reference material (Table 4), a probable cause of non-
propagation was non-homogeneity of the test material.  At least one laboratory
reported that the contents of separate containers of the same test material
differed in appearance.

    In the water extingu1shabH1ty test, the volume of water needed to
extinguish a flame on test material 1 was 17.8 mL for both days 1 and 2, with
overall precision (S) of 14.5 ml  (XRSD 81.3X) and 17.6 ml (98.8*),
respectively.  Test material 2 fires required an average of 5.9 ml of water in
day 1 measurements and  3.0 ml 1n  day 2 measurements.  Percent relative
standard deviations were 79.1 and 55.IX on day 1 and day 2, respectively,
while those for the reference material were 102.9 and 53.3X.
                                       10

-------
                          Table  3.   Phase  I  Results  -  Radiant  Heat  Ignition  Test

                                      Test Material  1:  Coal/Xylene

                                              Laboratory No.
Summary Statistics

Day 1
Day 2


Day 1

Sample Mean (Sec) 7
Std. Deviation S
Variance S?
Sample Mean (Sec) 7
Std. Deviation S
Variance S2


Sample Mean (Sec) 7
Std. Deviation S
Variance S2
Day 2 Sample Mean (Sec) 7
Std. Deviation S
Variance $2
101
19.0
1.6
2.3
22.2
1.2
1.4
. Test
.
101
13.2
2.9
8.4
'16.7
3.8
14.8
102 105 106
22.7 23.0 16.5
1.3 1.0 1.7
1.6 1.0 3.0
28.0 25.3 18.8
0.6 5.7 1.2
0.4 32.3 1.6
107 111
45.8 17.3
3.8 1.7
14.4 2.9
36.7 16.2
5.8 1.2
33.9 1.4
S Sr
24.0
2.0
24.5
3.4
S XRSO
11.1 46.3
7.8 32.0
Material 2: Sawdust/Kerosene
Laboratory No.
102 105 106
21.5 87.2* 13.3
2.2 68.3* 2.9
4.8 4664* 8.3
24.2 40.0 20.3
3.8 11.0 4.8
14.6 121 22.6

107 111
37.5* 16.2
22.7* 1.2
515* 1.4
ND 16.2
6.1
37.9
Summary
X S,.
16.0-
2.4
23.5
6.5
Statistics
S XRSO
4.3 27.2
11.1 47.4
*  *  Data rejected as outliers by Cochran's test
ND ' No Data reported or sample did not Ignite

-------
                       Table 3:   Phase I Results - Radiant Heat Ignition Test

                                         Reference Material

                                           Laboratory No.
Summary Statistics

Day 1


Sample Mean (Sec)
Std. Deviation
Variance

I
S
S2
101
(n-2)
54.2
2.3
5.4
102
(n-4)
49.3
0.9
0.8
105
(n-3)
53.7
11.0
121.0
106
(n-4)
43.0
7.4
55.1
107 111
(n-4)
ND 47
4
20
.1
.5
.2
^
Laboratory No.

Day 2


Sample Mean (Sec)
Std. Deviation
Variance

co coxl
ro
101
(n-3)
51.0
1.7
2.8
102
(n-4)
50.6
4.3
18.6
105
(n-3)
50.7
3.8
14.3
106 '
(n-4)
94.4*
8.4*
70.6*
107 111
(n-4)
ND 53
1
36
.4
.9
X Sr S XRSD
49.5
6.3 7.0 14.
Summary Statistics
1

X ST S *RSD
51.4
3.2 2.9 5.7

 *   Data rejected by Gochran's test
ND - No data; sample ,d1d not Ignite

-------
Table 4.  Phase I Results - Flame Propagation Test
          Test  Material  1:   Coal/Xylene
                Laboratory No.
Summary Statistics

Day 1
Day 2


Sample Mean (cm/Sec)
Std. Deviation
Variance
Sample Mean (cm/Sec)
Std. Deviation
Variance


I
tz
S2

101
0.33
0.02
0.0005
0.39
0.05
0.002

102
0
0
0
0
0
0

.54
.18
.03
.68
.04
.002
Test
105
0.57
0.13
0.02
0.63
0.15
0.02
Material
106
0.31
0.11
0.01
0.65
0.16
0.02
107
0.80
0.26
0.07
0.62
0.22
0.05
111
0.49
0.09
0.01
0.52
0.03
0.001
112
0.36
0.09
0.01
0.48
0.10
0.01
% Sf S XRSD
0.49
0.15 0.21 43.3
0.57
0.12 0.14 25.4
2: Sawdust Kerosene
. Laboratory No.

Day 1
Day 2
* a
ND *

Sample Mean (cm/Sec)
Std. Deviation
Variance
Sample Mean (cm/Sec)
Std. Deviation
Variance

S2
y
§2
Data rejected by Cochran's
No cjata reported. No flame
101
ND
ND
test
propagj
102
0
0
0
0
0
0
it1<
.10
.01
.0001
.08
.01
.0001
on
105
ND-
'ND

106
ND
ND

107
0.06
0.04
0.001
0.01
0.003

111
0.03
0.01
0.002
0.14*
0.20*
0.04*

112
0.07
0.03
0.001
0.02
0.003

. Summary Statistics
X Sp S XRSD
0.06 -
0.03 0.02 40.9
0.04
. 0.02 0.02 50.0


-------
Table 4.  Phase I Results - Flame Propagation Test
                Reference Material
               Laboratory No.
Summary Statistics

Day 1

Sample Mean (cm/Sec) I
Std. Deviation S
Variance S2
101
(n-2)
0.48
0.03
0.001
102
(n-4)
0.46
0.03
0.001
105
(n=4)
0.51
0.05
0.003
106
(n-2)
0.32
0.014
0.002
107
(n-4)
0.40
0.01
0.0001
111
(n-4)
0.33
0.04
0.001
112
(n-4)
0.38
0.04
0.001
X
0.41
Laboratory No.

Day 2

Sample Mean (cm/Sec) J
Std. Deviation S
Variance S2
101
(n-2)
.0.43
0.11
-0.01
102
(n-4)
0.50
0.07
0.004
105
(n-4)
0.52
0.03
0.001
106
(n-2)
0.41
0.11
0.01
107
(n-4)
0.40
0.04
0.002
111
(n-4)
0.37.
0.05
0.003
112
(n-4)
0.35
0.025
0.001
X
0.42
Sr
0.03
Summary
Sr
0.06
S XRSD
0.08 19.0
Statistics
S XRSD
0-.08 19.1

-------
F*
01
                                     Table 5.   Phase I  Results  -  Water  Ext1ngu1shabH1ty  Test



                                                  Test Material 1:  Coal/Xylene
                                                                                                     Summary  Statistics

Day 1
Day 2

Day 1
Day 2
* a
NO »

Sample Mean (ml)
Std. Deviation
Variance
Sample Mean (ml)
Std. Deviation
Variance
-
Sample Mean (ml)
Std. Deviation
Variance
Sample Mdan (ml)
Std. Deviation
Variance
Data rejected as
No data reported

I
S2
I
la

7
S
S2
I
S2
outliers
. Sample
101
11.3
0.7
0.5
9.6
7.1
50.4

101
ND
ND
102
39.2
7.7
58.8
41.9
2.1
4.5
Test
102
3.8
0.6
0.4
3.9
0.6
0.3
105
35.4
8.7
74.9
38.3
5.4
29.6
106
6.3
1.7
2.8
8.7
3.0
9.1
107
17.3
5.9
35.6
78.0*
32.3*
1045*
111
7.5
2.6
7.0
7.1
2.7
7.5
11
7
1
3
1
0
0
2
.6
.9
.5
.4
.3
.1
Material 2: Sawdust/Kerosene
Laboratory- No.
105
5.3
2.7
7.1
5.1
0.9
0.9
106
5.3
2.1
4.3
3. '6
1.7
2.8
107
- 8.6*
8.1*
66.8*
1.4
0.5
0.3
111
1.8
1.0
1.0
2.5
1.3
1.6
112
13
2
7
1
0
0
.2
.71
.3
.4
.2
.04
% Sr S XRSD
17.8
5.1 14.5 81.3
17.8
1.8 17.6 98.8
Summary Statistics
X Sr S XRSD
5.9
2.1 4.7 79.1
3.0
« 0.9 1.7 55.1
by Cochran's test
did not Ignite

-------
Table 5.  Phase I Results - Water Extingu1shab111ty Test
                   Reference Material
                   Laboratory No.
Summary Statistics
101
(n-2)
Day 1


Day 2


Sample Mean (ml)
Std. Deviation
Variance
Sample Mean (mL)
Std. Deviation
Variance
I
S
S2
7
so
S2
11.
7.
52.
5.
0.
0.
4
2
3
1
3
1
102
(n-4)
4.1
0.5
0.3
3.5
0.5
0.3
105
(n-4)
3.5
0.7
0.6
3.7
1.1
1.3
106
(n-4)
3.6
0.3
0.1
3.0
1.3
1.7
107
(n=4)
8.7
8.1
65.3
3.8*
3.7*
13.4*
111
(n-4)
1.2
0.3
0.1
1.2
0.6
0.3
112
(n-4)
1.2
0.2
0.03
1.5
0.4
0.2
X Sr S XRSD
4.8
3.8 4.9 102.9

3.0
0.8 1.6 53.3


-------
Outliers

    Because of the small amount of data received for Phases I and II, minimal
outlier testing was performed.  During Phase I, the number of data sets ranged
from 4 to 7.  For these data, only Cochran's test was applied to eliminate
laboratories with the worst precision.  Those laboratories had 1n most cases,
also returned study data that were abnormally high as compared to other
laboratories.   In Phase I, 5 of 71 triplicate measurements were rejected while
in Phase  II, 1 of 98 triplicates were rejected.

Phase  II  Results

    A  compilation of the raw data obtained during Phase II of the study can be
found  1n  Tables  6-22  (Appendix C).   Only  five Phase II laboratories reported
radiant heat  Ignition  data.  Four of those five also reported flame
propagation  and  water  ext1ngu1shabH1ty results.  Ideally, each of the
responding  laboratories should have  returned complete data sets for seven
waste  types  for each of the three Ignltablllty tests.  Test materials 3 and 4,
cotton fiber  and polyurethane foam,  respectively, failed to Ignite within five
minutes under  the radiant  heat source.  The radiant heat Ignition procedure
was  performed,  therefore,  on only five sample types by five laboratories.

     The flame  propagation  test yielded even fewer useable data than the
radiant heat  test.   Test materials 6 and  7 could not be analyzed by the
procedure.   For this  test, five  test materials were examined by an average of
only three  laboratories, making  statistical evaluation of the methods
difficult.   Likewise,  1n the water extlngulshablllty test, three to four
laboratories  reported  data from  five test materials.

Statistical  Summary and Precision Statement

     For ease  of discussion of the radiant heat Ignition, flame propagation,
and  water extinguishabilHy tests of Phase II, the statistical portions of
Tables 6-22  (Appendix  C) have been combined In Table 1.  Five test materials
were  analyzed  by each  of the three tests. Since only one laboratory reported
extinguishability data for test materials 6 and 7, those data are not being
considered  in  this  discussion.  Radiant heat ignition measurements yielded
overall percent  relative standard deviations 1n the range of 19-101X.  Data
obtained  from  flame propagation and  extinguishability tests, had relative
standard  deviations of 12.3-100.OX and 41.g-77.6X, respectively.

Evaluation of Methods

    The quantity of data from Phase  II laboratories was considerably smaller
than required for a meaningful statistical evaluation of the 1gn1tab1lity
methods.  Thorough correlation of results of the three methods for each test
material  is not  possible because 1n  most  cases only two tests were performed
on each waste.   Examination of the data 1n Table 1, however, reveals
considerable variation 1n both single-laboratory and overall precisions for
all three tests.  It 1s not possible to state which test exhibits the greatest
                                        17

-------
variability either day-to-day or across waste types.   The most reliable
evaluation of the methods 1s likely therefore, to be  qualitative.   As  part  of
the study, participants were encouraged to submit narratives  of their
experiences with the test apparatus, and to provide comments  as to  the
applicability of the methods to waste characterization.   One  laboratory
observed while conducting the radiant heat Ignition test that the relationship
between the rheostat setting and the source temperature  was not constant.   As
a result, replicate radiant heat measurements were highly variable  even at  the
same rheostat setting.  Another laboratory noted that the Interior  temperature
of the chamber changed as a function of the length of time the Instrument was
In use and soot deposited on the chamber walls and heat  source.  It was
further noted that fluctuations In Interior chamber temperature resulted when
the chamber door was opened to Introduce samples.  Since Ignition times are
dependent on chamber temperature at the beginning of  the test, a second heat
source should be Introduced to the chamber accompanied by automatic activation
of 1) the radiant heat source at a predetermined chamber temperature and 2) a
timing device that would measure and record time elapsed between activation of
the heat source and a signal from the flame-detecting thermocouple. This
would eliminate the need for a strip chart recorder,  which in this  study was
itself the source of considerable error.  The test procedure  requires  a chart
speed of 0.5 inches/minute (0.02 cm/sec).  Since the  chart distances recorded
in the study were commonly 1n the 0.2-0.4 cm range, a possible 10-20%  error
was introduced by the recording step alone.

    Operational difficulties were also experienced In the flame propagation
test.  The leading cause for the small amount of data was non-propagation of
the flame or non-ignition.  One laboratory reported that the  propagation rate
for some samples depended on the depth of the sample  in  the trough. One
analyst reported that flames had a tendency to propagate erratically or in  the
opposite direction of the second thermocouple.  A narrower trough would
minimize the possibility of misdirected propagation.   Some false starts were
reported when ignition of the sample was attempted.  This problem could be
solved by lengthening the trough so that the sample can  be ignited  without
activating the first thermocouple.

    In the water extinguishability test, several factors undoubtedly
contributed to the overall error of the measurements.  Perhaps the  greatest
source of error was the confusion over whether extinguishment of only  visible
flames or also the embers constituted the endpoint of the analysis. If the
latter endpoint is assumed, larger amounts of water would be  used.
Modifications in the water delivery system would solve some problems which
were reported by analysts.  The system 1s designed to deliver a cone-shaped
spray, the diameter of which 1s the same as that of the  sample dish.  More
accurate results might be achieved by using a spray pattern that delivers
water to the entire surface of the sample. .Direct measurement by weight, of
the amount of water delivered to the flaming sample would be  preferable to
calculating the value from time and delivery rate, both  of which contribute
separately to the total error of the extinguishability measurement.
                                    18

-------
    The three 1gn1tab1l1ty tests studied 1n this collaborative effort have the
potential of being useful and precise techniques for determining the potential
Ignition hazard of solid wastes.  Before the methods, particularly the water
ext1ngu1shabH1ty test, can be considered fully reliable,  substantial
modifications and refinements of the test chamber Itself are mandatory.  Many
of those modifications have been discussed 1n this section,  but only further
testing on actual waste samples will determine the extent  of the changes
ultimately required.

     In the current study, although eight laboratories agreed to participate 1r
the  full study, data/test were reduced to as few as three  sets, resulting 1n
only limited  statistical treatment of the results and evaluation of the
methods.  Once the method 1s  refined, the 1gn1tab1l1ty tests should again be
collaboratlvely tested with sufficient laboratories that are paid,  1f
necessary,  to assure full participation.
                                      19

-------
                                   REFERENCES

1.  Federal Register,  Vol.  45,  No.  98, May 19, 1980, p. 33122

2.  Michael, L.C., PerHtt, R.L.,  and PelHizarl, E.D.  Single-Laboratory
    Evaluation of IgnltabllUy and Related Methods.  Environmental  Protection
    Agency, Las Vegas, NV,  1984, 49 pp.

3.  McFarren, E.F., Llshka, R.J.,  and Parker, J.H.  Criterion for Judging
    Acceptability of Analytical Methods.  Analytical Chemistry 42 (3),  358,
    1970.
                                      20

-------
                                   APPENDIX A
                              IGNITABILITY METHODS
                      RADIANT HEAT IGNITION TEST PROCEDURE

1.0 SUMMARY OF THE METHOD

    A sample 1s placed 6 cm beneath a preheated radiant heat source  and  the
time to Ignition detected by a thermocouple sensor and recorded  on a strip
chart recorder.

2.0 APPARATUS AND REAGENTS

    2.1  Test chamber - as shown 1n Figure 1..
    2.2  Sample container - aluminum weighing pan, 6 cm diameter X  1.7 cm deep.
    2.3  Controller/sensor - as shown 1n Figure 1.
    2.4  Recorder - strip chart, variable speed and Input Impedance.
    2.5  Safety Equipment
         2.5.1  Flameproof gloves  (Lab Safety Supply 1915M or equivalent)
         2.5.2  Tongs - 53 cm  (Fisher, 15-207 or equivalent)
         2.5.3  Respirator (Fisher, 13-995-11 or equivalent)
    2.6  Balance  (minimum accuracy to nearest 100 mg)

3.0 TEST PROCEDURE

    3.1  Locate the test chamber in a fireproof fume hood with the  exhaust fan
         turned on.
    3.2  If the hood 1s large  enough, locate the controller/sensor  and the
         recorder outside the  hood but 1n proximity to the test  chamber.
    3.3  With an empty sample  container on the sample platform,  (Figure  1;
         part no. 5) position  the  sample platform to provide a heating element
         (Figure 1; part no. 1) to sample container (top r1m) distance of 6  cm.
    3.4  Position the thermocouple (Figure 1; part no. 7) 1 cm above the top
         rim and centered over the sample container.
    3.5  Connect the thermocouple  to position 33 (Figure 1) on the  controller/
         sensor using the cable supplied.
    3.6  Connect the heater/solenoid cable (Figure 1; part no. 16,  17) to
         position 34 (Figure 1) on the controller/sensor.
    3.7  Connect the power cord on the controller/sensor to a 110 VAC power
         source.
    3.8  Set the variable transformer to 0.
    3.9  Position a dummy (blackened with carbon soot) sample container
         (centered) under the  heat source.  Close the chamber door.
    3.10 Turn on the controller/sensor.
    3.11 Turn on the heater.   Set  the variable transformer to the setting
         determined during the temperature adjustment procedure.
    3.12 Preheat the radiant heat  source for 30 minutes.
    3.13 Transfer sample to the sample container to a depth of approximately
         1 cm, making sure that the sample surface 1s level and  smoothed.
                                    21

-------
                                WARNING
          Extreme care must be exercised In testing materials known or
          suspected of being extremely flammable.   Preliminary  tests using
          greatly diminished sample sizes should be conducted prior to
          performing the actual test to Insure the safety of the analyst.
          A reduced test sample depth should be used  In cases where sample
          Ignition 1s extremely rapid and/or violent.
3.14 Place the filled sample container outside, but In proximity to, the
     test chamber.
3.15 Start the recorder at a chart speed of 0.5 1n/m1n and  a full-scale
     sensitivity of 1 volt/Inch.
3.16 Open the chamber door and remove the dummy sample container.
3.17 Using tongs and flameproof gloves, pick up the filled  sample
     container and place 1t, centered, under the radiant  heat source.
     Immediately mark the recorder chart by gently touching the tip of the
     thermocouple with the tongs.  Lower the door  on  the  test chamber.
                                 CAUTION
          The placement of the sample beneath the  radiant heat  source, the
          marking of the recorder chart and the closing of  the  chamber
          door must be executed as quickly as possible  to optimize the
          precision of the test results.
3.18 Raise the chamber door and remove the Ignited sample from  beneath the
     radiant heat source and carefully place 1t 1n the  bottom of  the
     chamber.
                               •  WARNING
          Raise the fume hood and test chamber doors  just to a  level that
          facilitates removal of the sample from beneath  the radiant heat
          source.  Flameproof gloves should be used 1n  combination with
          tongs to protect the analyst from the burning sample.
3.19 Extinguish the fire by smothering.
3.20 Prepare the chamber for the next sample by positioning the dummy
     sample container beneath the radiant heat source.
3.21 Lower the fume hood door to one-half the height of the chamber.
3.22 Measure the distance  (cm) from the Initial mark  to Ignition  on the
     recorder.  Calculate the time to Ignition from the calibrated
     recorder speed.
3.23 Proceed with the analysis of subsequent samples  by repeating steps
     3.13 through 3.22.
                                 22

-------
                        FLAME PROPAGATION TEST PROCEDURE

1.0 SUMMARY OF THE METHOD

    A sample 1s placed 1n an aluminum trough and Ignited at one end.   The  time
required to burn a premeasured distance 1s sensed by thermocouples  and
recorded on a two-pen strip chart recorder.

2.0 APPARATUS AND REAGENTS
                                     '             *
    2.1  Test chamber - as shown 1n Figure 1.
    2.2  Sample container - prepared by molding heavy-duty aluminum foil  1n a
         20.5 X 6.6 X 3.5 cm (Inside dimensions) stainless steel trough.
    2.3  Controller/sensor - as shown 1n Figure 1.
    2.4  Recorder - two-pen, strip chart, variable speed and Input  Impedance.
    2.5  Safety Equipment                        i
         2.5.1  Flame proof gloves (Lab Safety Supply, 1915M or equivalent)
         2.5.2  Tongs - 53 cm  (Fisher, 15-207 or equivalent)
         2.5.3  Respirator (Fisher, 13-995-11 or equivalent)
    2.6  Wood splints

3.0 TEST PROCEDURE

    3.1  Locate the test chamber in a fireproof fume hood with the  exhaust fan
         turned on.
    3.2  If the hood is large  enough, locate the controller/sensor  and the
         recorder in the hood.  If sufficient space Is not available in the
         hood, locate the controller/sensor and/or the recorder outside the
         hood but in proximity to the test chamber.
    3.3  Connect the thermocouples, located in the bottom of the test chamber,
         to the. strip chart recorder.       •
    3.4  Prepare the sample container by molding heavy-duty aluminum foil  in
         the stainless steel trough.'
    3.5  Transfer sample to the sample container to a depth of approximately
         1 cm, making sure that the sample surface is even and smoothed.
                                     WARNING
              Extreme care must be exercised in testing materials known or
              suspected of being extremely flammable.  Preliminary  tests  using
              greatly diminished sample sizes should be conducted prior to
              performing the actual test to insure the safety of the analyst.
              A reduced sample depth should be used 1n cases where  sample
              ignition is extremely rapid and/or violent.
    3.6  Place the filled sample container centered between the two thermo-
         couples in the bottom of the test chamber.
    3.7  Start the recorder at a chart speed of 0.5 in/min and a full-scale
         sensitivity of 100 millivolts/Inch.   •
    3.8  With a lighted wood splint, Ignite the sample at one end of the
         trough.  Immediately  lower the test chamber door.
    3.9  Lower the fume hood door to one-half the height of the test chamber.
                                     23

-------
3.10 Record the distance (cm) between the two thermocouples and the linear
     distance traveled by the recorder (cm) peri during the burn period.
                                  HOTE       ••% .•.&•** < A
          If the flame does not propagate the full distance between the
          thermocouples, manually measure the distance propagated and the
          distance travelled by the recorder pen.
3.11 Raise the fume hood and test chamber doors slightly.  Cover the
     sample with the stainless steel trough I1d.
                                 WARNING
          The fume hood and test chamber doors should be raised only as
          high as required to cover the burning sample.  Flameproof gloves
          should be used In combination with tongs to protect the analyst
          from the burning sample.
 3.12 Prepare  the chamber for the next sample by removing the extinguished
     sample.
                                  24,,

-------
                     WATER EXTINGUISHABILITY TEST PROCEDURE

1.0 SUMMARY OF THE METHOD

    An Ignited sample 1s placed 6 cm beneath a sprayer nozzle and  extinguished
by delivering water at a calibrated rate to the sample.  The volume  of water
required to extinguish the sample 1s determined from the duration  of the  spray
and the flow rate of water.

2.0 APPARATUS AND REAGENTS

    2.1  Test chamber - as shown 1n Figure 1.
    2.2  Sample container - aluminum "tart" pan, 1.5 cm diameter (top,  Inside)
         X 3.2 cm deep.
    2.3  Controller/sensor - as shown 1n Figure 1.
    2.4  Recorder - strip chart, variable speed and Input Impedance.
    2.5  Safety Equipment
         2.5.1  Flameproof gloves (Lab Safety Supply, 1915M or equivalent)
         2.5.2  Tongs - 53 cm  (Fiaher, 13-995-11 or equivalent)
         2.5.3  Respirator (Fisher, 13-995-11 or equivalent)
    2.6  Balance (minimum accuracy to nearest 100 mg)
    2.7  Wood splints
    2.8  Water - municipal supply

3.0 TEST PROCEDURE

    3.1  Locate the test chamber 1n a fireproof fume hood with the exhaust  fan
         turned on.
    3.2  If the hood 1s large  enough, locate the controller/sensor and  the
         recorder in the hood.  If sufficient space 1s not available in the
         hood, locate the controller/sensor and/or the recorder outside the
         hood but in proximity to the test chamber.
    3.3  Connect the copper tubing containing the water pressure regulator  and
         gauge to a suitable supply source using the brass fitting (3/8"  male
         NPT - 1.4" tube) provided.
                                     CAUTION
              This tubing should be connected to a water supply source  with
              the regulator and gauge located close to the source.  If  the
              fitting supplied for this connection is incorrect, substitute
              the appropriate  alternative.
    3.4  Connect the long end  of the copper tubing to the solenoid valve
         (Figure 1, part no. 13).
    3.5  Install the sprayer nozzle (Figure 1,'part no. 10) in the bulkhead
         fitting on the inside right side of the test chamber.
    3.6  Turn on the water at  the source and test for leaks.  Tighten fittings
         as required.
    3.7  Set the water pressure to the solenoid valve to 10 psi.
    3.8  Position the sample platform centered and exactly 6 cm under the
         sprayer nozzle.
                                     25

-------
3.9  Place the aluminum divider between the support brackets  located  near
     the bottom of the test chamber.
3.10 Turn on the power to the controller/sensor.
3.11 With an empty sample container on the sample platform, activate  the
     sprayer continuously for 1 minute to eliminate air  from  the  copper
     tubing.
3.12 Read the water pressure (1n the absence of flow)  from the  pressure
     gauge.  Adjust to 10 ps1, 1f necessary.
3.13 Activate the sprayer 1n several short bursts and  again note  the  water
     pressure.  Adjust, 1f necessary.
3.14 Repeat step 3.12 until the pressure gauge reads 10  ps1 without flow/
3.15 Using an empty sample container, activate the sprayer 1n twenty  short
     bursts (approx. 100 millisecond), approximately 1 second apart.
     Record the cumulative time.
3.16 Carefully transfer the water 1n the sample container to  a  25 ml
     grduated cylinder.  Record the volume delivered.
3.17 Calculate the water flow rate 1n m1H1l1ters per  second.
3.18 Transfer sample to the sample container to a depth  of 1  cm,  making
     sure that the sample surface 1s level and even.
                                 WARNING
          Extreme care must be exercised 1n testing materials known or
          suspected of being extremely flammable.  Preliminary  tests  using
          greatly diminished sample sizes should be conducted prior to
          performing the actual test to Insure the safety of  the  analyst.
          A reduced test sample depth should be used 1n  cases where sample
          Ignition 1s extremely rapid and/or violent.
3.19 Place the sample on the aluminum divider near the bottom of  the  test
     chamber.
3.20 Ignite the sample, 1n the center, with a lighted  wood splint. Lower
     the test chamber door completely and the fume hood  door  to one-half
     the height of the chamber.
3.21 Allow the fire to burn until the entire surface of  the sample is
     burning.
3.22 Open the fume hood and chamber doors.  Using tongs, center the
     ignited sample on the sample platform directly beneath the sprayer
     nozzle.
3.23 Lower the test chamber door completely and the fume hood door to
     one-half the height of the test chamber.
3.24 Depress the switch which activates the sprayer and  timer,  in short
     bursts (approximately 100 milliseconds), approximately 1 second  apart
     and observe for continued burning between bursts.
                                 CAUTION
          This delivery process mayn require practice  before  1t can be
          executed successfully.
3.25 Record the time to extinguishment to the nearest  0.01 seconds.
3.26 If the water was not effective in extinguishing the fire,  extinguish
     the flame by smothering.
3.27 Prepare the test chamber for the next sample by resetting  the timer
     and removing the extinguished sample.
                                  26

-------
                       APPENDIX B
                IGHITABlllTY TEST CHAMBER


















u
0
0
a
D
20
°rs
o L5
30
Ds 9
0*
3 i i.


Figure 1. Ignitmbllity Test Chanber and Controller/Sensor.
                          27

-------
                                  PARTI LIST
Pm *°-                           Description
    1        Heater,  750 watt,  120 volt (American  Scientific Products,
             61856);  lever clay aaaeably forPrecision KJeldahl flask
             heater;  element cemented to ceramic base with Savereisen
             sealing  cement (No. 93 powder).
    2        Heater support, Uoistrut P-1000.
    3        Aluminum rod-, 1/2 in diameter (Fisher,  14-666).
    4        Right angle connector (Fisher, 14-666-20).
    S        Plate support (Fisher, 14-666-24).
    6        Thermocouple protection tube, Omegatite 450  , 3.16 x
              1/4 x 12 in (Omega Engineering, PTRA31614-12); with
              closed end saved off.
    7        Thermocouple, sub-miniature, Type'K,  Inconel  sheith,
              0.020 x 12 in (Omega Engineering, SCAIN-020U-12);
             Thermocouple extension (Omega Engineering, EXTT-K-24);
             Thermocouple connectors.
    8        Foot plate (Fisher, 14-666-25)(Omega  Engineering, SMP-K-MP).
    9        Transite* box, 45 x 45 x 60 cm, hinged  top.
    10        Hollov spray nozzle, stainless steel, 1/4 x  1-5/16 in
             male (McMaster-Carr Supply- 34505K76);  female connector,
              1/4 x 1/4 in NPT (Swagelok ,.SS-400-7-4).
    11        Bulkhead union, stainless steel, 1/4" (Swagelok9,
              SS-400-61).
    12        Tubing,  stainless steel, 1/4 in O.D.
    13        Solenoid valve, 1/4 in, normally closed (McHaster-Carr
             Supply,  4639R58).
             Not shown:  Copper tubing, 1/4 in O.D.; pressure gauge,
             0-15 psi (Matheson, 63-3115); vater pressure regulator
              (Fisher, 15-529); Female branch tee,  1/4 x 1/4 in NPT
              (Swagelok , SS-400-3-4TTF).
    14        Terminal board, double row (Kulka, 601-GP-2).
    15        High temperature cable, copper, 14AWG (W.K. Hile, 9G-PUT-1);
             solderless connector, spade terminals.
    16        Cable, 2 conductor, 18AVG copper (power supply to heater).
    17        Cable, single conductor, 18AVG.
    18        Chromatography trough (Fisher, 15-729,54).
    19        Door, Transite, sliding; thermal-resistant  glass vindow.
    20        Angle bracker, 4 hole.
                                  28

-------
                       PACTS LIST (cont'd.)
21        Auto transformer.
22        Fust holder, povtr.
23        Fuse holder, meters.
24        Fuse holder, beater controller.
23        Fuse holder, heater.
26        Fuse holder, spray.
27        Neon lamp, power.
28        SPST toggle switch, power.
29        Neon lamp, heater  control.
30        SPST toggle switch, heater control.
31        Omega  Model 650, digital thermocouple thermometer.
32        IMC Model 402,  digital timer.
33        Miniature thermocouple jack.
34        Nine-pin Molax  connector/panel mount.
35        Neon lamp, sprayer.
36        Double pole, double throw, center off, momentary/on-on
          miniature toggle switch (for sprayer).
37        Minature push button switch, momentary switch (timer reset)
                          29

-------
                                        APPENDIX C
                                      PHASE II DATA
                                                I
                 Table 6.  Phase  II Results - Radiant Heat Ignition Test
                                         «             *

                            Test  Material No. 5 - Paint Waste

                               Laboratory No.                   Summary Statistics

Day 1
Mean (Sec)
Std. Deviation
Variance
Day 2
Mean (Sec)
Std. Deviation
Variance




7
s
S2

"X
S
S2


101

69
16
262

49
13
176


102

65
4
16

69
3
10


106

113
15
231

183
45
2033
Reference

111

155
114
12990

116
18
317
112

88
16
242

ND


X Sr S *RSD

98
52 56 58


104
25 63 60

Material
1

Laboratory No.

Day 1
Mean (Sec)
Std. Deviation
Variance
Day 2
Mean (Sec)
Std. Deviation
Variance


X
s
S2

7
S
S2
101

50
4
17

41
0
0
102

52
1
2

52
3
8
106

40
0
0

42
3
12
111

51
5
28

53
6
36
112"

44
55
3014

ND



Summary Statistics
X Sr S %RSD

47
25 21 45


47
3 7 15

ND = No data reported
                                           30

-------
Table 7.  Phase II Results - Radiant Heat Ignition Test
            Test Material No. 6 - Waste Oil
             Laboratory No.                Summary Statistics

Day 1
Mean (Sec) I
Std. Deviation S
Variance S2
Day 2
Mean (Sec) X"
Std. Deviation S
Variance S2

101

91
9
60

80
9
81

102

124
7
56

191
27
745

106

85
18
,. 325

85
23
525
Reference
111 X

'175 119
27
720 '

249 151
27
752
Material
Laboratory No.

Day 1
Mean (Sec) Y
Std. Deviation S
Variance S2

Day 2
Mean (Sec) "X
Std. Deviation S
Variance S2
101

47
0
0


47
8
69
102

43
4
18


55
1
2
106

27
3
12


27
3
12
111 K

52 42
15
231
I
•
47 44
0
0
Sr S XRSD


17 38 32



23 152 101


Summary Statistics
Sr S *RSD


8 13 30




5 13 29

                         31

-------
Table 8.  Phase II Results - Radiant Heat Ignition Test
        Test Material  7 - Waste Oil/Sand Mixture
                 Laboratory No.             Summary Statistics

Day 1
Mean (Sec) I
Std. Deviation S
Variance S2
Day 2
Mean (Sec) Y
Std. Deviation S
Variance S2
101

211
68
4622

210
18
323
102

122
35
1236

57
19
345
106

82
8
58

130
13
175
Reference
Laboratory

Day ]_
Mean (Sec) I
Std. Deviation S
Variance S2
Day 2
Mean (Sec) I
Std. Deviation S
Variance S2
101

43
6
39

46
10
108
102

38
8
72

41
10
98
106

25
0
0

27
3
13
111

191
12
152

113
100
9852
t
Material
No.
in

45
0
0

54 -
10
91
^ ^ Sr S XRSD

151
39 68 45

127
6 63 50

Summary Statistics
X Sr S XRSD

38
5 10 26

42
9 13 32
                            32

-------
               Table 9.  Phase II Results - Radiant Heat Ignition Test
                         Test Material No. 8 - Waste Solvent
                                Laboratory  No.
Summary Statistics

Day 1
Mean (Sec) 7
Std. Deviation S
Variance $2
Da_y 2
Mean (Sec) 7
Std. Deviation S
Variance $2

101

31
4
20

38
3
9

102

54
4
16

49
7
50

106

65
18
325

52
10
108
Reference
Laboratory

Day 1
Mean (Sec) 7
Std. Deviation S
Variance S^
Mean (Sec) 7
Std. Deviation S
Variance S^
101

40
2
4
37
2
4
102

37
10
98
38
6
32
106

27
' 3
13
27
3
13
111

72
22
477

56
2
3
Material
No.
Ill
i
49
5
28
56
12
153
11.2 B Sr S *RSD

84 61
12 14 23 38
150

ND 49
6 9 19

Summary Statistics
112 X Sr S %R5D

59 42
'8 6 13 31
69
40
7 13 33
ND = No data reported
                                        33

-------
              Table 10.  Phase II Results - Radiant Heat Ignition Test
                       Test Material No. 9 - Sawdust/Kerosene
                               Laboratory No.                 Summary Statistics

Day 1
Mean (Sec) I
Std. Deviation S
Variance S2
Day 2
Mean (Sec) I
Std. Deviation S
Variance S2
101

14
2
3

13
2
3
102

17
11
124

15
6
33
106

13
6
33

12
3
8
Reference
Laboratory

Day J
Mean (Sec) "X
Std. Deviation S
Variance S2
Day 2
Mean (Sec) J
Std. Deviation S
Variance S2
101

43
2
4

46
15
212
102

33
4
18

44
3
8
106

23
3
13

30
0
0
111

16
2
5
,
17
3
6
Material
No.
Ill

49 .
5
28

41
2
3
112

20
7
46

ND




112

71
25
623

ND


X Sr S *RSD

16
6 6 35


14
3 3 24


Summary Statistics
? c c tfDcn
A Ov« o *>i\oU

44
12 18 41


40
7 9 23

ND = No  data  reported
                                           34

-------
                Table 11.  Phase II Results - Flame Propagation Test
                                      i             *


                         Test Material No. 3 - Cotton Fiber



                                Laboratory No.               Summary Statistics

DayJ[
Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance


Y
S
S2

Y
S
$2
102

3.7
0.5
0.3

ND


106

2.6
0.9
0.8

2.65
0.92
0.84
112

6.5
2.2
5.0

ND


8 Sr S XRSD

4.3
1.4 2.3 53.8


2.6
* * *

  Reference Material
             I


Laboratory No.
                                                             Summary Statistics

Day 1
Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance


X
S
S2

Y
S
S2
102

0.2
0.01
-
-
0.2
0.1
0.01
106

0.3
0.02
-

0.3
0.08
-
112

0.5
0.08
0.006

ND


X Sr S XRSD

0.3
0.04 0.16 52.1


0.3
0.1 0.1 36.1

ND = No data reported

*  = Value cannot be calculated from existing data
                                          35

-------
                Table 12.   Phase II Results  - Flame  Propagation  Test
                      Test Material No. 4 - Polyurethane Foam
                              Laboratory No.            	Summary Statistics
106 111 112 1 Sr S *RSD
Day 1
Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance

Day 1
Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance
* 0.3 0.3 0.4 0.33
S 0.1 . - , - 0.05 0.07 21.4
5 • _ -
X" 0.28 ND ND 0.28
S 0.01 - * * *
O ™
Reference Material
Laboratory No. Summary Statistics
106 111 112 • X Sr S %RSD
I 0.4 0.2 0.5 0.4
S 0.10 0.1 0.1 38.2
S2 - - 0.01
I
, *
"X 0.4 ND ND 0.4
S - * * *
S2
ND = No data reported
*  = Value cannot be calculated from existing data
                                          36

-------
                Table 13.  Phase II Results - Flame Propagation Test
                          Test Material No.5 - Paint Waste
                              Laboratory No.            	Summary Statistics

Day 1
Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance


y
S

y
s
s2
102

0.02
0.01

0.03
0.007
-
106 111

0.04 0.04
0.004 0.008

0.03 ND
0.003
• I
112

0.04
0.005

ND

.
1 Sr S XRSD

0.04
0.01 0.01 28.8

0.03
0.005 0.004 13.6

Reference Material


Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance


X
s
S2

y
s
S2

102
0.2
0.02
-

0.2
0.04

Laboratory No.
106 111
0.3 0.4
0
0

0.3 ND
0
0

112
0.5
0.1
0.02

ND


Summary Statistics
X Sr S %RSD
0.4
0.1 0.1 38.2


0.3
0.02 0.07 24.2

ND = No data reported
                                         37

-------
                Table 14.  Phase II Results - Flame Propagation Test

                         Test  Material  No.  8  - Waste Solvent

                               Laboratory No.   	          Summary Statistics
102 106 111 112 X Sr S %RSD
Day 1
Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance

Day 1
Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance
I
S2
y
S2

Y
S2
I
s
S2
, •
0.1 NO 0.2 0.1 0.1
0.1 0.2 Q.01 0.1 0.1 100.0
0.1 ND ND NO 0.1
0.03 * * *
Reference Material
Laboratory No. Summary Statistics
102 106 111 112 I Sr S *RSD
0.4 ND 0.5 . 0.66 0.5
0.01 - - 0.01 0.13 '26.3
0.4 ND ND ND 0.4
0.1 * * *
ND = No data reported
*  = Value cannot be calculated from existing data
                                            38

-------
                 Table 15.  Phase II Results - Flame Progatlon Test

                       Test Material No. 9 - Sawdust/Kerosene

                                Laboratory No.                Summary Statistics

DayJL
Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance




I
S2

I
S
S2


102

0.1
0.1

0.2
0.04


102
106 111

0.2 0.1
0.1

0.2 ND
0.02
Reference
Laboratory
106 111
112
.
0.04
0.01

' ND
Material
No.
112
X Sr S XRSD

0.11
0.04 0.11 104.3

0.20
0.03 0.02 12.3
i

Summary Statistics
I Sr S *RSD
     Day 1
Mean (cm/sec)    X
Std. Deviation  S
Variance        S

     Day 2
Mean (cm/sec)   X
Std. Deviation  S
Variance        S
0.4    0.4      0.4      0.4
0.02   0.07
0.4    0.5      ND
0.03   0.05
ND
         0.4
0.4
                 0.04   0.03   8.2
                 0.04   0.08  19.5
ND = No data reported
                                           39

-------
             Table 16.   Phase II  Results  - Water Ext1ngu1shab1Hty Test

                         Test Material  No. 3'-  Cotton Fiber

                               Laboratory No.      *          Summary Statistics

DayJ.
Mean (cm/Sec)
Std. Deviation
Varianr.o
Day 2
Mean (cm/Sec)
Std. Deviation
Variance


7
S2

Y
S
S2
102

3.4
0.4
0.1

2.9
0.4
0.2
106

7.6
0.6
0.4

2.5
2.6
6.8
111

1.9
0.5
0.3

1.9
0.4
0.2
112 X

4.8** 4.3
4.0**
16.4**

ND 2.4

Sr S *RSD

0.5 3.0 69.4


1.5 1.3 55.3
                                  Reference  Material

                                Laboratory No.
Summary Statistics

Day 1
Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance


Y
S
S^

I
S
S2
102

3.7
0.4
0.2

2.9
0.2
0.05
106

4.2
-
-

2.8
0.3
0.07
111 , 112 2 Sr S %RSD
.
0.02 1.9 2.4
0.8 0.5 1.9 81.0
0.7

0.02 ND 1.9
0.2 1.9 100.0
—
** = Data  rejected  by Cochran's  test
ND = No data  reported
                                          40

-------
            Table 17.   Phase  II Results - Water ExUngu1shab1l1ty Test
                      Test Material  No.  4  - Polyurethane  Foam
                                .Laboratory No.              Summary Statistics

Day 1
Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance


Day 1
Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance


Y
S2

Y
S2



Y
S
S2

Y
S
S2
106

2.2
0.6
0.3

111 112 X Sr S XRSD

2.4 5.3 3.3
1.1 '3.8 2.3 2.6 77.6
1.2 14.1

5.7 2.3 NO 4.0
0.2 0.7 0.5 2.4 61.0
0.04 0.5


106

4.1
0.8
0.6

5.0
2.9
8.6
Reference Material
Laboratory No. Summary Statistics
111 112 I Sr S XRSD

1.2 1.7 2.3
0.2 0.7 0.6 1.6 70.7
0.05 0.5
)
0.9 ND 2.9
0.05 ' 2.1 3.4 116.2
NO =  No data reported
                                          41

-------
             Table 18.  Phase II Results  - Water Ext1ngu1shab1l1ty Test
                          Test Material No. 5 -  Paint Waste
                               Laboratory  No.                 Summary Statistics

Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance

7
S

7
s
S2
102
3.5
0.6
0.4

3.4
0.5
0.3
106
2.8
1.2
1.5

3.5
1.2
1.4
in
1.7
0.1
0.01

1.6
0.2
0.02
112
1.4
0.1

ND
X Sr S *RSD
2.3
0.7 1.1 49.2

2 8
0.8 1.3 44.8
Reference Material
i •
Laboratory No.

Day 1
Mean (on/Sec)
Std. Deviation
Variance
Day!
Mean (on/Sec)
Std. Deviation
Variance


X
s
S2

I
S
S2
102

2.8
0.2
0.05

2.7
0.2
0.05
106

3.7
0.7
0.4

3.9
1.1
1.3
111

0.9
0.1
0.01

0.9
112

1.7
0.7
0.4

ND
Summary Statistics
X Sr S %RSD

2.3
0.5 1.3 56.3

2.5
0.5 1.6 62.6
ND = No data reported
                                         tffia'iiii MJhiiit

-------
             Table 19.  Phase  II Results - Water Ext1ngu1shab11ity Test

                           Test Material No.  6 - Waste 011

                     _^	Laboratory No.              Summary Statistics
                     102
106
111
112
X
XRSD
     Day 1

Mean (cm/Sec)   I    4.9       ND       ND       ND      4.9
Std. Deviation  S    0.8
Variance        $2   0.6

     Day 2
Mean (on/Sec)   Y    2.8       ND       ND       ND      2.8
Std. Deviation  S    0.7
Variance        S2   0.4



Day^L
Mean (cm/Sec)
Std. Deviation
Variance
Day_2
Mean (cm/Sec)
Std. Deviation
Variance




X
S
S2

I
S
S2


102

4.2
1.1
1.1

2.4
0.2
0.03
Reference Material
Laboratory No. Summary Statistics
106 111 112 X Sr S XRSD

ND ND ND 4.2
* * *


ND ND ND 2.4
* * *

 ND = No data reported
 *  = Value cannot be calculated from existing data
                                           43

-------
             Table 20.   Phase II  Results  - Water Ext1ngu1shab1l1ty Test
                       Test Material  No.  7 - Sand/011  Mixture
                               Laboratory  No.          	Summary Statistics

Day 1
Mean (cm/ Sec)
Std. Deviation
Variance
Day 2

Mean (cm/Sec)
Std. Deviation
Variance



Day 1
Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance


y
S
S2


Y
S
S2




y
S
S2

y
S
S2
102

2.5
0.3
0.1


2.9
0.7
0.5


102

2.7
0.9
0.8

3.3
0.3
0.1
106 111 112 X Sr S *RSD

ND ND ND 2.5
* * *


i
ND ND ND 2.9
* * *

Reference Material
Laboratory No. Summary Statistics
106 111 112 X Sr S %RSD

ND ND ND 2.7
* * *


ND ND ND 3.3
* * *
,
ND = No data reported
*  = Value cannot be calculated from existing data
                                           44

-------
             Table 21.  Phase II Results - Water Ext1ngu1shab1l1ty Test


                         Test  Material  No.  8 -  Waste Solvent
                                           t

                                Laboratory  No.                 Summary Statistics

Day j^
Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance




Day 1
Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (on/Sec)
Std. Deviation
Variance


y
S
S2

y
s
S2





Y
S
S2

"X
S
S2
102

7.6
1.1
1.3

6.7
0.9
0.8


102


3.3
0.6
0.3

2.6
1.1
1.1
1C6 111

NO 3.9
0.2
0.04

NO 3.7
0.2
0.03
Reference
Laboratory
106 111


ND 1.0
0.04
-

ND 0.9
0.01
—
112

3.1
0.7
0.5

ND


Material
No.
112
t
»
2.1
0.3
0.1

ND


Jt Sr S XRSD

4.9
0.8 2.5 50.8


5.2
0.6 2.2 41.9


Summary Statistics
X ST. S %RSD


2.1
0.4 1.2 56.9


1.7
0.7 1.3 78.3

ND = No data reported
                                            45

-------
             Table 22.  Phase II Results - Water Ext1ngu1shab1l1ty Test
                       Test Material No. 9 - Sawdust/Kerosene
                                Laboratory No.                Summary  Statistics

Day 1
Mean (cm/Sec)
Std. Deviation
Variance
•Day 2
Mean (cm/Sec)
Std. Deviation
Variance


7
S
S2

7
S
S2
102

3.6
1.0
1.1

3.5
0.7
0.5
106

2.6
0.5
0.2

2.2
0.6
0.3
111

3.3
1.8
3.2

3.0
1.0
1.1
Reference
Laboratory

Day 1
Mean (cm/Sec)
Std. Deviation
Variance
Day 2
Mean (cm/Sec)
Std. Deviation
Variance


X
S2

7
S
S2
102

2.5
0.2
0.03

3.1
0.3
0.1
106

3.2
0.6
0.3

3.2
1.0
1.0
111

1.0
0.05

0.9
0.05
112

1 1.2
0.13

NO
Material
No.
112

2.5
0.2
0.05

NO
1 Sr S XRSD

2.7
1.1 1.4 51.8

2.9
0.8 1.3 43.2

Summary Statistics
I Sr S %RSD

2.3
0.3 1.0 41.7

2.4
0.6 1.4 57.9
NO =  No  data  reported
                                          46

-------