©EPA
United States
Environmental Protection
Agency
Office of Water
(WH-552)
EPA-821-R-93-011
Revision 1
September 1993
Preliminary Report of EPA Efforts to
Replace Freon For The Determination
of Oil and Grease
-------
-------
DISCLAIMER
This report has been reviewed by the Analytical Methods Staff of the EPA Office of Water. It is a
preliminary report intended to provide information on the current status of this study. Evaluation of
the data collected under this study is ongoing. Mention of company names, trade names or
commercial products does not constitute endorsement or recommendation for use.
Questions or comments regarding this report should be addressed to:
William A. Telliard, Chief
Analytical Methods Staff
Engineering and Analysis Division (WH-552)
Office of Science and Technology
U.S. EPA Office of Water
401 M Street, S.W.
Washington, D.C. 20460
ACKNOWLEDGEMENTS
This study was the result of a cooperative effort between the EPA Office of Water, Office of Solid
Waste and Emergency Response, Office of Air and Radiation, and Office of Research and
Development. This report was prepared under the direction of William A. Telliard with assistance
from the Environmental Services Division of DynCorp Viar, Inc. (under EPA contract no. 68-C9-
0019), and its subcontractors: Interface Inc., Environmental Survey Associates, SRI International,
Skinner and Sherman Laboratories, ETS Analytical Services, and the Geochemical and Environmental
Research Group of Texas A&M University. Other contributors to this study included the EPA
Central Regional Laboratory in Annapolis, MD, the 3M Corporation, Horiba Instruments, and Varian
Sample Preparation Products.
-------
-------
TABLE OF CONTENTS
Executive Summary 1
Section 1 Background 3
Section 2 Phase I Study Design 5
2.1 Study Objectives 5
2.2 Sample Source Selection 5
2.3 Analytical Study Design 6
Section 3 Field Sampling 11
Section 4 Data Validation and Statistical Analysis 13
4.1 Data Validation 13
4.2 Statistical Analysis 13
Section 5 Discussion 19
5.1 Separatory Funnel Extraction of Aqueous Samples 19
5.2 Soxhlet and Sonication Extraction of Solid Samples 20
5.3 Solid Phase Extraction (SPE) of Aqueous Samples 20
5.4 Non-Dispersive Infrared Analysis of Aqueous Samples 22
5.5 Graphical Presentation of RMSD Versus Acceptance Limit Results 22
5.6 Graphical Presentation of the Solvent-Freon Ratios 22
Section 6 Phase I Study Conclusions 39
6.1 Alternative Solvents for Existing Methods 413.1, 9070 and 9071A 39
6.2 Sonication as an Alternative Technique to Soxhlet Extraction 40
6.3 Alternative Techniques for Aqueous Samples 40
6.4 Retention and Elimination of Solvents for Further Study 40
6.5 Retention of Alternative Techniques for Further Study 41
Section 7 Follow-Up and Possible Phase II Activities 43
Appendix A Site Summary
Appendix B Data Summary
Appendix C Discussion of Statistical Techniques Used in the Preliminary Data
Evaluation
-------
LIST OF EXHIBITS
Exhibit 1: Solvent RMSDs for Conventional Techniques
Exhibit 2: Relative Percent Deviation vs. Concentration
Exhibit 3: Summary Statistics for Alternative Solvents in the Determinatinon of Oil and
Grease - Aqueous Waste Stream, Separatory Funnel Extraction
Exhibit 4: Summary Statistics for Alternative Solvents in the Determination of Oil and
Grease - Solid Waste Stream, Soxhlet Extraction
Exhibit 5: Summary Statistics for Alternative Techniques in the Determination of Oil
and Grease - Solid Waste Stream, Sonication Extraction
Exhibit 6: Summary Statistics for Alternative Techniques in the Determination of Oil
and Grease - Aqueous Waste Stream, 90 mm Solid Phase Extraction Disk
Exhibit 7: Summary Statistics for Alternative Techniques in the Determination of Oil
and Grease - Aqueous Waste Stream, 45 mm Solid Phase Extraction Disk
Exhibit 8: Summary Statistics for Alternative Techniques in the Determination of Oil
and Grease - Aqueous Waste Stream, SPE Column (old version)
Exhibit 9: Summary Statistics for Alternative Techniques in the Determination of Oil
and Grease - Aqueous Waste Stream, SPE Column (new version)
Exhibit 10: Summary Statistics for Alternative Techniques in the Determination of Oil
and Grease - Aqueous Waste Stream, Infrared Analysis
Exhibit 11: Summary of Solvents and Conventional Techniques with RMSDs Within the
Acceptance Limit
Exhibit 12: Root Mean Square Deviations - Aqueous Waste Stream, Separatory Funnel
Extraction
Exhibit 13: Root Mean Square Deviations - Solid Waste Stream, Soxhlet Extraction
Exhibit 14: Root Mean Square Deviations - Solid Waste Stream, Sonication Extraction
Exhibit 15: Root Mean Square Deviations - Aqueous Waste Stream, Alternative
Techniques, Non-Petroleum Samples
Exhibit 16: Root Mean Square Deviations - Aqueous Waste Stream, Alternative
Techniques, Petroleum Samples
Exhibit 17: Mean Solvent-Freon Ratios for All Techniques Tested
Exhibit A-l: Site Summary Freon Replacement Study
Exhibit A-2: Wastewater Sources Freon Replacement Study
11
-------
EXECUTIVE SUMMARY
The Clean Air Act Amendments of 1990 (CAAA) require that use of Class I
chlorofluorocarbons (CFCs) which deplete the ozone layer be phased out by 1996. Freon 113
(trichlorotrifluoroethane) is a Class I CFC that is required by present U.S. Environmental Protection
Agency (EPA) wastewater and solid waste methods for measurement of the conventional pollutant "oil
and grease". This report provides the results of Phase I of EPA's study to replace Freon 113 in the
determination of oil and grease.
The objective of Phase I was to find a solvent or solvents that would produce results nearly
identical to the results produced by Freon 113 in the oil and grease measurement. Solvents evaluated
were n-hexane, «-hexane plus methyl tertiarybutyl ether (MTBE) in an 80/20 mixture, DuPont 123
(2,2-dichloro-l,l,l-trifluoroethane), methylene chloride, and perchloroethylene (tetrachloroethene).
Solvents were evaluated by comparing results from triplicate extraction of 40 wastewaters and 28
solid/sludge wastes, from 39 industrial facilities in 24 industrial categories, to the results produced by
extraction of these same samples with Freon 113.
In' addition to testing the solvents listed above, sonication extraction, solid phase extraction
(SPE) using cartridges and disks, and a proprietary solvent/non-dispersive infrared technique were
tested on a subset of samples.
Alternative Solvents
Results were stratified according to extraction solvent and technique, and within this
stratification, into three categories: all samples, petroleum-based samples, and non-petroleum based
samples, depending on whether the oil and grease was of animal, vegetable, or mineral origin.
Data from solvent comparisons in the Phase I study are summarized in Exhibit 1 and show
that the results produced by n-hexane, perchloroethylene, and the 80/20 mixture of «-hexane and
MTBE are not statistically different from the results produced by Freon 113 (i.e., within the
Acceptance Limit) for some sample strata, whereas the results produced by methylene chloride and
DuPont 123 are not within this limit.
Based on these results and on the results of a preliminary study by EPA's Environmental
Monitoring Systems Laboratory in Cincinnati, Ohio (EMSL-Ci), EPA will retain or eliminate certain
solvents and techniques from further consideration as candidates for replacement of Freon 113 in the
oil and grease measurement, as follows:
1. n-hexane will be retained because the results for petroleum-based solid samples and non-
petroleum aqueous samples are within the Freon 113 Acceptance Limit and because n-
hexane was used in the oil and grease measurement prior to the advent of Freon 113.
2. Perchloroethylene will be retained because the results for non-petroleum aqueous samples
are within the Freon 113 Acceptance Limit and because perchloroethylene can be used hi
the measurement of oil and grease by infra-red techniques.
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
3. Although the results for petroleum-based solid samples extracted with n-hexane/MTBE
(80/20) are within the Freon 113 Acceptance Limit, 80/20 will be eliminated from further
study due to concerns about laboratory safety and solvent composition change during
storage. Such concerns were raised during Oil and Grease Workshops held by EPA in
Norfolk, Virginia and Boston, Massachusetts to allow regulated industries, laboratories
and other interested parties the opportunity to discuss the status of the Agency's Freon
replacement efforts.
4. Methylene chloride will be eliminated because the results produced are far from results
produced by Freon 113.
5. DuPont 123 will be eliminated because results produced using this solvent are not within
the Freon 113 Acceptance Limit for any category of solid or aqueous samples and because
DuPont 123 is a Class II CFC that will need to be phased out eventually.
6. Cyclohexane, which was not formally evaluated in the first phase of the study, will be
considered in future evaluations. The decision to evaluate cyclohexane is based on
concerns that have been raised at the EPA workshops and elsewhere concerning the
neurotoxicity of /z-hexane.
Alternative Techniques
Of the alternative techniques evaluated in this study, only sonication extraction of non-
petroleum solid samples produced results equivalent to existing techniques with Freon 113. The use of
smaller solvent volumes required by solid phase extraction (SPE) and the increased sensitivity of the
non-dispersive infrared technique might warrant further study.
Exhibit 1.
Solvent Root Mean Square Deviation (RMSDs) for Conventional Techniques
SMO
DuPoM
Pcittitor
MtOl
Noraillud Root M«a Sqiura DnlalJoa
NOTE: Acceptance Llmiu for Freon 113 in each stratum (petroleum or non-petroleum) are shown by dotted vertical lines. RMSDs to the left of the respective Acceptance Limit indicate
that the alternative solvent produces results equivalent to Freon 113 for that statutn, whereas RMSDs to the right of the respective Acceptance Limit indicate degrees of lesser agreement,
with the least agreement being farthest to the right (highest RMSD).
September 1993
-------
SECTION 1
BACKGROUND
Chlorofluorocarbons (CFCs) have been shown to be primary contributors to the depletion of
the earth's stratospheric ozone layer. The United States, as a party to the Montreal Protocol on
Substances that Deplete the Ozone Layer and as required by law under the Clean Air Act
Amendments of 1990 (CAAA), is committed to controlling and eventually phasing out CFCs. Under
both the Montreal Protocol and the CAAA, Class I CFCs will be phased out by January 1, 1996. To
be consistent with its commitment to control and phase out CFC use, the U.S. Environmental
Protection Agency (EPA) is investigating the use of alternative solvents in lieu of Freon 113, a CFC
that is mandated in EPA methods for the determination of "oil and grease".
Currently, three Agency methods are used for regulatory compliance monitoring
determinations of oil and grease content in environmental samples. They are Method 413.1,
promulgated at 40 CFR Part 136, and Methods 9070 and 9071A, promulgated at 40 CFR Parts 260-
270 (EPA Publication SW-846 by reference). Method 413.1 is used in Clean Water Act (CWA)
programs to determine total oil and grease content in surface and saline waters and in industrial and
domestic wastes. This gravimetric method involves acidification of the sample, serial extraction of the
oil and grease with Freon 113 in a separatory funnel, evaporation of the solvent from the extract, and
weighing of the residue. Method 9070 is used in programs administered under the Resource
Conservation and Recovery Act (RCRA) and is essentially the same as Method 413.1. RCRA Method
9071A is used to recover low levels of oil and grease from sludges, soils, other solid matrices, and
some industrial wastewaters. The method involves acidification or drying, Soxhlet extraction of oil
and grease with Freon 113, and weighing of the residue after evaporation of the solvent.
In all three methods described above, the result, termed "total recoverable oil and grease", is
a method-defined parameter. This means that the result depends totally on how the measurement is
made. Therefore, changes to the specific analytical protocols have the potential of changing the
numerical value of the results for a given sample.
Clean Water Act effluent guidelines for 25 major industries include limitations on the
discharge of oil and grease (40 CFR 4Q7-471). Further, oil and grease is a regulated pollutant in over
10,000 National Pollutant Discharge Elimination System (NPDES) permits and in many RCRA
operating permits. Civil penalties for violation of NPDES and RCRA permits can be severe. The
measurements on which these guidelines and permits are based were made with Freon 113 or, before
1978, with n-hexane as the solvent. The regulated community is concerned that a change in the
solvent used for oil and grease determinations could cause a change in the results and put their
facilities in violation of NPDES or RCRA permits. In response to this concern, EPA initiated efforts
to identify a replacement solvent or an alternative measurement technique that gives oil and grease
results as close as possible to those obtained with Freon 113.
The Agency's initial efforts to find a suitable replacement solvent for Freon 113 were
conducted by the Office of Research and Development's Environmental Monitoring Systems
Laboratory in Cincinnati, Ohio (EMSL-Ci). EMSL-Ci first used laboratory-prepared, synthetic
samples containing materials that represent "oil and grease" compounds covering extremely wide
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
boiling ranges, such as No. 2 fuel oil, No. 6 fuel oil, Prudhoe Bay crude oil, animal lard, and wheel-
bearing grease. Reagent water was spiked with these materials dissolved in an organic solvent to
simulate real-world samples. These samples were then extracted using several different solvents in
place of Freon 113, and the residues were determined gravimetrically.
Subsequent evaluations by EMSL-Ci used a limited number of actual industrial waste samples.
The results of EMSL-Ci's work were presented in A Study to Select a Suitable Replacement Solvent
for Freon 113 in the Gravimetric Determination of Oil and Grease, by F. K. Kawahara, October 2,
1991. This study resulted hi the preparation of draft Method 413.3, a modification of 413.1 which
utilizes an 80/20 mixture of «-hexane and methyl tertiary butyl ether (MTBE) instead of Freon 113.
On July 3, 1991, the EPA Office of Air and Radiation (OAR) proposed (56 FR 30519) to
amend CWA and RCRA analytical methods for oil and grease to require the use of the 80/20 mixture
in lieu of Freon 113. This proposal was based on the findings of the EMSL-Ci studies. Based on
comments received on this proposal, OAR has delayed any solvent replacement to allow additional
time to study alternatives. In late 1991, the Office of Water (OW) and the Office of Solid Waste
(OSW) began planning a comparative study to collect data in support of OAR's efforts to replace
Freon 113 in Agency oil and grease methods. The remainder of this document provides a report on
the first phase of that study.
4
September 1993
-------
SECTION 2
PHASE I STUDY DESIGN
The first phase of the cooperative Office of Water (OW) and Office of Solid Waste (OSW)
effort involved evaluation of alternative solvents and measurement techniques. The original study
design is described fully in a study plan dated December 19911. The final Phase I study design is
summarized below.
2.1 Study Objectives
The purpose of the first phase of the study was to continue EMSL-Ci's investigation of
replacement solvents to identify solvents or solvent/extraction systems that provide equivalent
performance to Freon 113. Specific objectives were to:
• Evaluate alternative solvent and solvent/extraction system equivalency across a range of
real world effluent and solid waste samples from a variety of facility types
« Evaluate a series of solvents posing a lower potential risk to stratospheric ozone
• Provide clear direction for further study of one or two solvents or solvent/extraction
systems across an even broader range of effluent and solid waste samples that are
regulated under the Clean Water Act (CWA) and the Resource Conservation and
Recovery Act (RCRA).
This effort was intended to be used as a screening study that would be followed by a Federal
Register notice of the availability of resulting data, development of revised analytical procedures, and
a confirmation study to support the recommended method revisions.
2.2 Sample Source Selection
The kinds of sample matrices included in this first phase of the study represented wastewaters
and solid wastes from a variety of industrial categories and facilities. A total of 18 facilities were
originally selected for sampling based on the following considerations:
• industrial category
• sample matrix (e.g., wastewater, sludge)
lDraft Study Plan For Sampling and Analysis Activities to Support Freon Replacement Method Study, December
1991. Copies of this study plan are available from the Sample Control Center (operated by DynCorp Viar), 300
North Lee Street, Alexandria, Virginia 22314, (703) 557-5040.
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
• expected oil and grease concentration levels
• geographic location
• accessibility of waste streams
• limited co-mingling of wastes
• willingness of facilities to participate
• cooperation and assistance of facility personnel.
As the study progressed, 21 additional sites were included to expand the number of industrial
categories and waste types, as budget constraints would allow. Considerations for selecting additional
sites involved extending the range of oil and grease concentrations being studied, focusing on
categories with existing oil and grease effluent limitations and on categories with petroleum-based
waste types, and including waste streams with known analytical interference problems.
The Office of Water coordinated all facility contacts and planning for the sample collection
efforts. Facilities volunteering to participate were selected primarily from EPA Regions I, II, and HI
to minimize the costs of travel and equipment transportation.
Aqueous and solid/sludge waste samples were collected from a total of 39 industrial and
commercial facilities. These 39 facilities represent 24 industrial categories, 40 aqueous waste streams
and 28 solid/sludge waste sources. A site summary is presented in Appendix A, Exhibit A-l, and
shows the facility types, industrial categories, and waste streams included in the study. Appendix A,
Exhibit A-2, presents additional information about the wastewater sources; it identifies petroleum-
based streams and groups the samples into one of four treatment categories. Samples identified as
having no treatment include raw process wastewaters. Primary treatment includes simple detention for
oil/water separation or solids removal, or physical/chemical treatment designed for something other
than oil removal. Secondary oil/water separation includes systems such as dissolved air flotation,
dispersed gas flotation, and filtration intended for oil removal beyond simple detention. Biological
treatment includes activated sludge and aerated lagoon systems.
2.3 Analytical Study Design
Initially, the study focused on evaluation of six solvents, including Freon 113, using
conventional separatory funnel extraction for aqueous samples and Soxhlet and sonication extraction
for solid samples, all followed by gravimetric determination. Interest in EPA's study was widespread,
however, and shortly after initiating the first phase of the study EPA was approached by several
manufacturers of alternative extraction devices and measurement techniques. These manufacturers
agreed to analyze splits of the EPA samples by their techniques and to provide the results of their
analyses to EPA at no cost to the Agency. The approaches to evaluation of each of the measurement
techniques are presented below.
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Separatory Funnel Extraction of Aqueous Samples
The performance of five alternative extraction solvents was compared with Freon 113 hi
actual sample matrices. The five solvents were:
• n-hexane and methyl tertiary butyl ether (MTBE) in an 80/20 mixture
• «-hexane alone
• DuPont 123 (2,2-dichloro-l,l,l-trifluoroethane)
• methylene chloride (dichloromethane)
• perchloroethylene (tetrachloroethene).
Wastewater samples were extracted with each of these solvents and Freon 113 using the
separatory funnel techniques described hi Method 413.1.2 Since the densities of the solvents range
from lighter than water to heavier than water, the specifics of the extraction procedures in Method
413.1 were adjusted to explicitly deal with removal of each solvent from the separatory funnel and
appropriate treatment for emulsions that might form.3
Three aliquots of each sample weref extracted with each solvent and were analyzed hi a single
laboratory only. Although the initial plan was to have all samples analyzed hi one laboratory to
eliminate a potential source of differences, three laboratories were ultimately required to meet
schedule, cost, and quality considerations. Having multiple laboratories was thought to be a negligible
source of variability because all comparisons between solvents were performed on a single sample
basis.
A total of 40 triplicate aqueous sample sets were collected and sent to contract laboratories for
analysis. Twenty-one triplicate sample sets were sent to Skinner and Sherman, Inc. in Waltham, MA,
13 triplicate sample sets were sent to ,ETS Analytical Services in Roanoke, VA and eight triplicate
sample sets were sent to the Geochemical and Environmental Research Group of Texas A&M
University.
Soxhlet and Son/cation Extraction of Solid Samples
Sonication was tested as an alternative to the Soxhlet extraction required by Method 9071A.
In addition, the performance of five alternative extraction solvents was compared with Freon 113 hi
actual sample matrices. The five alternative solvents tested were the same as those tested hi aqueous
samples using conventional separatory funnel techniques.
2Method 9070 is essentially the same as Method 413.1 and therefore was not formally evaluated.
3A total of five alternate versions of Method 413.1 were drafted for use by laboratories participating hi this
study. Copies of these methods are available from the Sample Control Center, 300 North Lee Street, Alexandria,
Virginia 22314, (703) 557-5040.
September 1993 7
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Solid samples were extracted with each of the five alternative solvents and Freon 113 using
the Soxhlet extraction techniques described in Method 9071A. Splits of the solid samples were also
analyzed at the same laboratory using all six solvents and sonication procedures. The specifics of the
extraction procedures hi Method 9071A were adjusted to explicitly deal with the use of alternative
solvents and the use of sonication extraction techniques.4
A total of 28 triplicate solid sample sets were collected and sent to contract laboratories for
analysis. Twelve triplicate sample sets were sent to Skinner and Sherman, Inc. in Waltharn, MA, rune
triplicate sample sets were sent to ETS Analytical Services in Roanoke, VA and nine triplicate sample
sets were sent to the Geochemical and Environmental Research Group of Texas A&M University.
Solid Phase Extraction (SPE) of Aqueous Samples
The performance of solid phase extraction (SPE) techniques was evaluated by the EPA's
Central Regional Laboratory (CRL) hi Annapolis, MD and by two manufacturers of SPE devices. The
SPE techniques tested on aqueous samples have the advantage of using significantly less solvent than
conventional separatory runnel extraction techniques.
EPA's CRL extracted aqueous samples with EnvirElute SPE columns (also known as SPE
cartridges) supplied by Varian Sample Preparation Products, using Freon 113 and all five solvents
listed above and gravimetric determination. Each of the samples sent to the EPA laboratory was
analyzed in triplicate. Due to the high solids content in most of the aqueous samples, the EPA CRL
was able to analyze only four of the eight sample sets they received.
Varian Sample Preparation Products extracted 20 aqueous samples with a newer version of the
SPE columns and a refined technique that allowed testing of samples containing higher concentrations
of suspended solids than the previous version tested by the EPA CRL. Varian tested Freon 113, n-
hexane, and the 80/20 hexane/MTBE mixture as solvents and determined the results gravimetrically.
Varian did not test methylene chloride or perchloroethylene as extraction solvents. Each of the
samples sent to Varian was analyzed in triplicate.
The 3M Corporation extracted 28 aqueous samples, using Empore SPE disks, with the
following five solvents: Freon 113, methylene chloride, «-hexane, 100% MTBE, and
perchloroethylene. The results were determined gravhnetrically. The 3M Corporation did not test the
80/20 mixture of /z-hexane and MTBE. Splits of each sample were analyzed using 47 mm SPE disks
and 90 mm SPE disks. Each of these sample splits was analyzed in triplicate. The 90 mm SPE disks
required the use of a full 1-liter sample volume, which is consistent with traditional oil and grease
measurement techniques and with all other techniques evaluated in this study except the 47 mm SPE
disks and the infrared technique described below. Both of these techniques utilized 250 mL sample
volumes.
All samples extracted with SPE devices were splits of samples analyzed by separatory funnel
extraction.
4A total of eleven alternate versions of Method 9071A were drafted for use by laboratories participating in this
study. A separate version was prepared for Soxhlet extraction of each of the five alternative solvents. Separate
versions were also prepared for sonication extraction of each alternate solvent and Freon 113. Copies of these
methods are available from the Sample Control Center, 300 North Lee Street, Alexandria, Virginia 22314, (703)
557-5040.
8
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Non-Dispersive Infrared Analysis of Aqueous Samples
Horiba Instruments Inc. analyzed 36 aqueous samples by a non-dispersive infrared (NDIR)
analyzer technique using S-316 (a chlorofluorocarbon also known as Flon and specifically developed
for the Horiba analyzer) as a solvent. An advantage of this infrared technique is that it offers lower
levels of detection than those afforded by conventional separatory funnel extraction followed by
gravimetric determination. As with the SPE analyses, the Horiba results were compared statistically to
the results produced on splits of the same samples with separatory funnel and Freon 113 extraction.
September 1993
9
-------
-------
SECTION 3
FIELD SAMPLING
Sample collection and handling efforts were carried out by Office of Water (OW) staff,
contractors, and facility staff. Activities included:
• communications and coordination with representatives of selected sampling sites and
assigned laboratories
• logistics planning
• preparation of field equipment, sample containers, shipping containers, and documentation
• collection, handling, and transport of samples
• preparation of sample collection documentation
• follow-up communications with receiving laboratories for sample tracking.
For safety purposes, arrangements were made in advance to have plant personnel accompany
and assist U.S. Environmental Protection Agency (EPA) and contractor representatives while on site.
To minimize sampling costs where possible, EPA also asked some of the selected facilities to have
plant personnel collect samples. In those cases, EPA sent prepackaged sampling kits directly to the
sites with instructions for use and shipment to assigned laboratories.
Since the primary purpose of this data gathering effort was to investigate the use of alternative
oil and grease extraction solvents with real world samples, and not to characterize selected
wastewaters and solid wastes for regulation development, short-term grab sampling of the selected
sources was deemed sufficient to meet the requirements of the study.
Sampled material was mixed in a collection vessel, and transferred by siphoning or scooping
into 1-liter, wide-mouth, clear glass bottles with PTFE-lined caps. (Samples sent to 3M for extraction
with the 47 mm disks and to Horiba for infrared analysis were collected in 250 mL bottles.) Aqueous
samples were preserved on-site with HC1 (1:1) to pH less than two, and all aqueous and solid/sludge
samples were cooled to between 0 and 4°C during storage and shipment. Each sample bottle was
labeled with a unique EPA sample number and identifying information, including bottle number,
source location, collection date, and preservatives used. EPA sample numbers were the primary
method of identifying and tracking samples. These numbers were pre-assigned and recorded to ensure
proper control over samples. EPA traffic reports were completed for each site or sampling episode,
and accompanied each shipment to the appropriate laboratories. Copies of these reports were used by
EPA contractor and field personnel for tracking purposes.
Information regarding site-specific sampling activities was recorded in an on-site log for each
sample location, and included EPA sample numbers, collection date and time, description of sample
location, collection procedure, sample pH and temperature, and preservatives used. Field personnel
double-checked all labels, traffic reports, and log entries to ensure accuracy and consistency.
September 1993
11
September 1993
13
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Workshop held by the Office of Water (OW) in Norfolk, VA on May 4, 1993. This report includes a
revised statistical analysis of those results as well as a statistical analysis of the results generated using
solid phase extraction (SPE) and infrared measurement techniques. A table containing all results for
the study in unreduced form is provided in Appendix B.
Data Stratification
Initial statistical analyses of all samples within each extraction technique yielded no solvents
that were within the respective RMSD Acceptance Limits. Further stratification of the study database
was undertaken to determine whether results equivalent to Freon 113 could be achieved by testing
subclasses of the data. Stratification by industry type was attempted, but resulted in too many small
data sets. Because it was expected that oil and grease of biological (animal or vegetable) origin might
behave differently than oil and grease of mineral origin (petroleum), the data were stratified into
"petroleum" and "non-petroleum" samples. (If necessary, personnel from the sampled facilities were
questioned as to the origin of oil and grease in their effluents or wastes). For example, the effluent
from a meat packing plant was categorized as non-petroleum, whereas the effluent from a refinery
was categorized as petroleum.
Solvent-to-Freon 113 Ratio
A solvent-to-Freon 113 ratio (solvent-Freon ratio) was computed to allow comparison of the
average amount of oil and grease extracted by various solvents and measurement techniques with the
amount extracted by the approved techniques using Freon 113.
The solvent-Freon ratio for each sample was formed by dividing the mean of the triplicate
results produced with the alternative solvent or technique by the mean of the triplicate result produced
with the approved technique using Freon 113. The mean, standard deviation (SD), relative standard
deviation (RSD), and the median of the solvent-Freon ratios were calculated across all samples in
each of the eight data sets representing an alternative solvent or technique (Exhibits 3 through 10).
These results are discussed hi Section 5.
Because averaging gives no indication of the variability of the data, the solvent-Freon ratio is
a less powerful measure of the agreement between an alternative solvent (or technique) and Freon 113
than the normalized RMSD described below. This normalized RMSD was used as the main criterion
of similarity. Median, mean, and standard deviation values of the solvent-Freon ratio were used as an
aid hi describing the distribution of the data.
Logarithmic Data Transformation
The triplicate analysis of each sample allows the error associated with each measurement to be
modeled as a function of concentration. Analytical data normally have a proportional error structure
over the calibration range of the measuring technique (i.e., as the parameter being measured
increases, the standard deviation will increase in proportion to the measurement. This proportional
error structure normally results hi a constant relative standard deviation [RSD; coefficient of
variation].)
14
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the De-termination of Oil and Grease
To determine the nature of the error in this study, the RSD was plotted as a function of
concentration. Exhibit 2 shows this plot for aqueous samples. Similar results occurred with
solid/sludge samples. As can be seen from Exhibit 2, the plot has a proportional error structure above
a concentration of approximately 35 mg/L. Below 35 mg/L, however, the RSD rises rapidly and
asymptotically approaches a very large value at the detection limit of the method (5 mg/L).
The data were examined by ANOVA statistical techniques. The heteroscedasticity (non-
uniform standard deviations) of these data violates the statistical requirements for use of ANOVA
techniques. The heteroscedasticity was overcome by modeling standard deviation as a function of
concentration, as discussed in Appendix C, and then subjecting the data to an appropriate variance-
stabilizing transformation according to the equation:
z = log (x + c) - log c
where
x is the concentration (in mg/L or mg/kg), and
c is a constant.
The constant (c) was used to prevent samples at low concentrations from having exaggerated
influence and to allow negative values to be transformed to a logarithmic scale. The constant (c) was
set to 100 mg/L for aqueous samples and 10,000 mg/kg for solid samples. The transformed data (z)
were then subjected to ANOVA.
To assure that the results below the detection limit were not unduly influencing the
comparison between solvents, the data were treated in three ways before transformation: (1) the
results were tested "as is" (i.e., without alteration), (2) all results below the nominal detection limit
were set to the detection limit, and (3) results were eliminated if they were below the nominal
detection limit using the currently approved extraction technique (e.g., separatory funnel extraction
with Freon 113 or Soxhlet extraction with Freon 113). After transforming the data using the equation
cited above, all three treatments yielded nearly identical results: the solvents that produced results
closest to Freon 113 were not changed by the three data treatments. This conclusion substantiates the
belief that the transformation diminishes the effect of the results below and near the nominal detection
limit. After reaching this conclusion, all data were treated with the second option for the remainder of
the study. This is consistent with the EPA Engineering and Analysis Division's standard reporting of
results that are below method detection limits.
Root Mean Square Deviation (RMSD)
As was noted above, the primary measure of similarity used to compare Freon 113 with each
of the other solvents and techniques was the RMSD of the alternative results for each sample around
the Freon 113 results. For each technique, the RMSD represented the standard deviation of the
differences between the alternative solvent-determined concentrations for each sample and the Freon
113-determined concentrations using the appropriate approved technique. A smaller RMSD indicated
better agreement with Freon 113.
September 1993
15
-------
-------
SECTION 5
DISCUSSION
The statistical results presented in Exhibits 3 through 10 are discussed in Sections 5.1 through
5.4 below. As explained in Section 4, the root mean square deviation (RMSD) was used as the main
criterion of equivalence between the results obtained with established techniques (Freon 113 with
separatory funnel or Soxhlet extraction) and all other solvents and techniques. Other descriptive
statistics (median, mean, standard deviation [SD], and relative standard deviation [RSD] of solvent-
Freon ratios) are provided in this report only to demonstrate the distribution of the data.
In these tests, some solvents and techniques extracted more oil and grease than Freon 113.
Although achieving higher results than the Freon 113 methods might be seen as a method
"improvement", such enhanced recovery of an operationally-defined parameter presents significant
problems for the implementation of the "improved" method under the National Pollutant Discharge
Elimination System (NPDES), as thousands of NPDES permits contain oil and grease limits based on
the present methods. Therefore, the objective of this study was to find a solvent or technique that
yields oil and grease results equivalent to rather than better than those obtained with Freon 113.
Exhibits 3 through 10 are arrayed to provide a comparison of each alternative solvent or
technique tested, for the entire group of samples analyzed, for petroleum-based samples, and for non-
petroleum samples. Solvents or techniques that have a normalized RMSD within the Acceptance Limit
shown for the established method (separatory funnel or Soxhlet extraction with Freon) yield results
that are not statistically different from Freon 113.
5.1 Separatory Funnel Extraction of Aqueous Samples
Results of the statistical analysis of data for separatory funnel extraction and gravimetric
determination of oil and grease in aqueous samples are presented in Exhibit 3. The results are based
on analysis of 25 petroleum-based samples and 13 non-petroleum samples for a total of 38 aqueous
samples. (Not all of the 40 aqueous samples taken were successfully analyzed.)
Exhibit 3 shows that when all samples were examined as a group, none of the solvents tested
yielded results within the Acceptance Limit. When non-petroleum and petroleum samples were
examined separately, however, n-hexane and perchloroethylene were both within the Acceptance
Limit in the non-petroleum group.
Mean solvent-Freon ratios for the separatory funnel extractions ranged from 1.02 (for
petroleum-based samples extracted with w-hexane) to 2.61 (for non-petroleum extracted with DuPont
123). Relative standard deviations of the solvent-Freon ratios ranged from 51% (for petroleum-based
samples extracted with methylene chloride) to 230% (for non-petroleum samples extracted with
DuPont 123). Median solvent-Freon ratios were in the range of 0.81 to 1.38.
September 1993
19
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
The data in Exhibit 3 suggest that although some alternative solvents were capable of
producing average and/or median results similar to Freon 113 hi each category of samples analyzed,
the variability of these data was extremely high. As a result, and based on the RMSD, the only
alternative solvents that yielded results equivalent to Freon 113 were n-hexane and perchloroethylene,
and the equivalency of these solvents was demonstrated for non-petroleum samples only.
5.2 Soxhlet and Sonication Extraction of Solid Samples
Results of the statistical analysis of data for 28 solid samples are presented in Exhibits 4 and
5. Exhibit 4 presents the results of data generated using the established Soxhlet extraction and
gravimetric determination with six solvents; Exhibit 5 presents the results generated using the
alternative sonication and gravimetric determination with the same six solvents.
Exhibit 4 indicates that relatively good agreement between median solvent-Freon ratios was
obtained using alternative solvents and Soxhlet extraction of solid samples. Means of solvent-Freon
ratios ranged from 0.99 to 2.14, and RSDs ranged from 22% to 160%. The results in Exhibit 4
further indicated that none of the alternative solvents was within the Acceptance Limit when all
samples were evaluated together. Similarly, none of the solvents was within the Acceptance Limit
when only non-petroleum samples were examined. Two of the solvents, however, were within the
Acceptance Limit when only petroleum-based samples were evaluated (n-hexane and 80/20.)
In Exhibit 5, all six solvents (including Freon 113) used with sonication extraction of solid
samples were compared to Freon 113 using Soxhlet extraction. This comparison was made because
the purpose of the study was to compare the performance of alternative solvents and techniques with
currently approved solvents and techniques. As noted earlier, the Agency method for determination of
oil and grease in solids (Method 9071A) specifies the Soxhlet extraction using Freon 113 as a solvent.
Exhibit 5 shows poor agreement between the mean and median results obtained with Freon
113 using the Soxhlet extraction and with all solvents using sonication. Further, variability (standard
deviation and relative standard deviation) of these data was high. (Exhibit 5 shows mean solvent-
Freon ratios that range from 0.49 to 3.69, with RSDs of 37% to 370% and median solvent-Freon
ratios from .43 to .81.) Aside from methylene chloride, all solvents tested with sonication yielded
mean and median solvent-Freon ratios below those obtained using Soxhlet extraction with Freon 113.
As a result of the high variability between the sonication data and the Soxhlet data with Freon 113,
only one RMSD value was observed within the Acceptance Limit hi Exhibit 5. That value was for
perchloroethylene when only non-petroleum samples were examined.
5.3 Solid Phase Extraction (SPE) of Aqueous Samples
Exhibits 6 through 9 present the results of the statistical analysis of data generated from
aqueous samples using solid phase extraction (SPE) and gravimetric determination. The format of
these tables is the same as that of Exhibits 3 through 5. As with the results of sonication extraction of
solid samples, all results generated using SPE techniques were compared to the performance of
approved test methods (in this case, Freon 113 using separatory funnel techniques). Results of
alternative techniques were not compared with each other, nor were they compared to the
performance of Freon 113 with the alternative technique.
20
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Exhibit 6 summarizes the statistical calculations of data generated by 3M using 90 mm SPE
disks. Exhibit 7 presents the results generated by 3M using 47 mm SPE disks. Nineteen of the 25
aqueous/petroleum and nine of the 13 aqueous non-petroleum samples in the study were tested using
each type of disk.
Mean solvent-Freon ratios in Exhibits 6 and 7 for the SPE data using alternative solvents were
all above 1.0 (they ranged from 1.09 to 2.78 for 90 mm disks and from 1.19 to 3.5 for 47 mm
disks). The ratios between Freon 113 with SPE and Freon 113 with separatory funnel extraction were
.94 and 1.11 for the 90 mm and the 47 mm disks, respectively. Variability with the SPE disks was
generally high, with RSDs ranging from 57 to 110% for the 90 mm disks and from 60 to 130% for
the 47 mm disks. Median solvent-Freon ratios using the 90 mm disks ranged from 0.94 to 1.83. The
highest mean results were obtained with methylene chloride and 100% MTBE. Median solvent-Freon
ratios for alternative solvents using the 47 mm disks were higher than for the 90 mm SPE disks,
ranging from 1.26 to 3.65.
The RMSD values in Exhibits 6 and 7 further show that the use of separatory funnel
extraction with Freon 113 and the use of either 47 mm or 90 mm disks with alternative solvents do
not produce equivalent results. The only RMSD value in these tables within the Acceptance Limit is
for Freon 113 using the 90 mm disks when non-petroleum samples were examined alone.
Exhibit 8 presents the results obtained by the EPA Central Regional Laboratory (CRL) using
an early version of the Varian SPE column. Due to technical difficulties (primarily clogging of the
columns due to high percent solids in the samples) results for only two petroleum and two non-
petroleum samples were obtained. Mean and median solvent-Freon ratios were less than 1.0, and
RSDs were lower than for other techniques tested. Normalized RMSD results indicate that all solvents
tested were equivalent to separatory funnel extraction with Freon 113. A review of the individual
results for the samples tested, however, shows that oil and grease concentrations in three of four
samples is near or below the detection limit.
The setting of results below the detection limit to 5 mg/L was considered as a possible reason
why the CRL SPE data showed all solvents to be equivalent to Freon 113. Tests with the three data
treatments described in the section on "Logarithmic Data Transformation" in section 4.2, however,
yielded equivalent results, indicating that this was not the sole reason. Because these concentrations
are low, the addition of the constant in the log-transformation reduces the effect of the differences
between solvents. In addition, the results are fairly precise, with RSDs ranging from 19 to 55 percent.
This combination of low concentrations and precise data makes the RMSDs small and consistent, so
that differences between solvents are not discernable. Although it would be possible to reduce or
eliminate the constant from the equation for transforming the CRL SPE data, the net result would be
that the conclusions would be based on a single sample.
Given all of the above, the final conclusion concerning the CRL SPE data is that there are
insufficient data to provide a rigorous comparison between solvents.
Exhibit 9 summarizes Varian's data for fifteen of the 25 petroleum samples and five of the 13
non-petroleum samples tested using a later version of the their SPE column. As noted earlier, only
Freon 113 and two alternative solvents (n-hexane and the 80/20 mixture) were evaluated by Varian.
Mean solvent-Freon ratios ranged from 0.93 to 1.13, and the median ratios ranged from 0.93 to 1.19.
Variability, as represented by the RSD of the solvent-Freon ratio, was relatively low (21% to 49%.)
Neither «-hexane nor the 80/20 mixture yielded RMSDs within the Acceptance Limit for all samples
September 1993
21
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
taken together, or for the petroleum subset. Both «-hexane and the 80/20 mixture had RMSDs within
the Acceptance Limit in the non-petroleum category. This result must be viewed with caution due to
the small number of samples involved (n = 5).
5.4 Non-Dispersive Infrared Analysis of Aqueous Samples
The statistical results obtained with data generated by Horiba Instruments using an infrared
analyzer are presented in Exhibit 10. These results represent 23 of the 25 petroleum-based aqueous
samples and all 13 of the non-petroleum aqueous samples in Phase I of the study. The mean solvent-
Freon ratio found using the Horiba method was 1.2 to 2.3 times the result for separatory funnel
extraction using Freon 113. Variability was high (58 to 98% RSD.) The normalized RMSDs for the
Horiba results were much higher than the Acceptance Limits in all three categories analyzed,
indicating that results wqre not equivalent to Freon 113.
5.5 Graphical Presentation of RMSD Versus Acceptance Limit Results
To give the reader a better understanding of the relative performance of all of the alternative
solvents tod techniques evaluated, the RMSD and Acceptance Limit data are presented graphically in
Exhibit 1, which was presented in the Executive Summary of this report, and in greater detail in
Exhibits 11 through 15.
In Exhibits 11 through 15, the RMSD for a particular solvent or technique is represented by a
solid or hollow circle, and Acceptance Limits are indicated by horizontal lines. Where these RMSD
circles fall within the relevant Acceptance Limit line, that solvent or technique is equivalent to
separatory funnel or Soxhlet extraction with Freon 113, as appropriate. Where no solvents or
techniques yield results equivalent to Freon 113, no circles are shown within the Acceptance Limit
line.
5.6 Graphical Presentation of the Solvent-Freon Ratios
Exhibit 16 summarizes all of the solvent-Freon ratios on one graph. In this graph, the mean
solvent-Freon ratio (computed prior to log-transformation of the data) is plotted on a logarithmic scale
so that amounts of oil and grease greater or less than the amount extracted by Freon 113 are
equidistant from 1.00.
This graph shows that the mean solvent-Freon ratios range from approximately 0.4 to 3.3,
depending on the solvent, technique, and type of sample being extracted. Of interest in this graph is
the influence of the solvent or technique on the average amount of material extracted. For example,
Soxhlet extraction with the alternative solvents seems to extract somewhat more oil and grease than
the reference Soxhlet extraction with Freon 113, whereas less oil and grease is extracted with
sonication rather than with Soxhlet using Freon 113 unless methylene chloride is used in the
sonication extraction. It must be remembered that these ratios represent the overage amount of
material extracted, whereas the RMSD represents the deviation from the amount extracted by Freon
113 on an individual sample basis. Therefore, although these averages are interesting, they cannot be
relied upon as the final criterion for equivalence.
22
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
The use of Freon 113 as the extracting solvent in the alternative techniques allowed the
variables of solvent and technique to be separated. Using cartridges (columns), the amount of oil and
grease extracted by Freon 113 and the alternative solvents is less than the amount extracted using the
reference separatory funnel method for all solvent-technique combinations except 80/20 with the new
cartridges. However, all solvents extract more material than Freon 113 when extracted by the same
cartridge technique.
Using 47 mm disks with Freon 113, nearly twice as much oil and grease is extracted than is
extracted by Freon 113 in the reference separatory funnel method. The amounts of oil and grease
extracted by the alternative solvents using the 47 mm disk are even greater than the amount extracted
by Freon 113. For 90 mm disks, the amount of oil and grease extracted by Freon 113 is slightly less
than the amount extracted using the reference separatory funnel method, but the alternative solvents
extract more than Freon 113 by either the reference separatory funnel extraction or 90 mm disk
extraction.
The effects of solvent versus technique cannot be separated for Flon-316 and NDIR because
Freon 113 was not tested with the NDIR technique. Therefore, it is not known whether the larger
amount of oil and grease indicated by the Flon/NDIR technique results from the use of Flon-316 or
results from the way in which the NDIR is calibrated.
It must be emphasized that the solvent to Freon ratios are averaged, and that the differences
may not be statistically significant due to the high variability of the data. Although significance tests
could be performed, the RMSD is a more reliable indicator of the difference between alternative
solvents/techniques and the reference Freon 113 separatory funnel method.
September 1993
23
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Exhibit 3.
Summary Statistics For Alternative Solvents in the Determination of Oil and Grease
Aqueous Waste Stream, Separatory Funnel Extraction
All Samples (N=38)
Solvent
Freon
Hexane
MeClj
Perchlor
DuPont
80/20
Mean
1.00
1.17
1.57
1.65
1.85
1.38
SD
-
1.26
1.84
0.99
3.66
2.17
RSD
-
110
120
60
200
160
Median
1.00
0.87
1.00
1.37
1.08
0.94
RMSD
1.4'
1.7
3.3
1.7
4.5
3.7
Non-Petroleum (N = 13)
Solvent
Freon
Hexane
MeClj
Perchlor
DuPont
80/20
Mean
1.00
1.45
2.33
1.65
2.61
2.00
SD
-
1.89
3.00
1.01
6.00
3.62
RSD
-
130
130
61
230
180
Median
1.00
0.94
1.32
1.34
1.03
0.98
RMSD
1.8'
1.3"
3.2
1.1"
4.7
4.1
Petroleum (N=25)
Solvent
Freon
Hexane
MeCIj
Perchlor
DuPont
80/20
Mean
1.00
1.02
1.20
1.65
1.47
1.07
SD
-
0.77
0.60
1.00
1.61
0.74
RSD
-
75
51
61
110
69
Median
1.00
0.81
1.00
1.38
1.15
0.92
RMSD
1.5'
2.6
3.6
2.9
3.5
2.0
* Acceptance Limit
** Value Within Acceptance Limit
Mean = Mean of Solvent to Freon Ratios
SD = Standard Deviations of Solvent to Freon Ratios
RSD = 100 x SD/Mean
Median = Median of Solvent to Freon Ratios
RMSD = Normalized Root Mean Square Deviation of Sample x Solvent Means
24
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Exhibit 4.
Summary Statistics For Alternative Solvents in the Determination of Oil and Grease
Solid Waste Stream, Soxhlet Extraction
All Samples (N=28)
Solvent
Freon
Hexane
MeCl2
Perchlor
DuPont
80/20
Mean
1.00
1.18
1.63
1.75
1.37
1.29
SD
-
1.07
1.56
2.40
1.03
1.31
RSD
-
91
95
140
75
100
Median
1.00
0.95
1.06
1.20
1.04
1.01
RMSD
1.5'
2.2
2.9
3.5
2.5
2.4
Non-Petroleum (N = 1 1)
Solvent
Freon
Hexane
MeCl2
Perchlor
DuPont
80/20
Mean
1.00
1.48
2.05
2.14
1.52
1.66
SD
-
1.69
2.27
3.44
1.52
2.07
RSD
-
110
110
160
100
120
Median
1.00
0.95
1.11
1.22
0.94
1.03
RMSD
1.9*
2.6
3.1
3.5
2.8
2.8
Petroleum (N=17)
Solvent
Freon
Hexane
MeCl2
Perchlor
DuPont
80/20
Mean
1.00
0.99
1.36
1.50
1.28
1.06
SD
-
0.23
0.83
1.46
0.56
0.23
RSD
-
23
61
98
44
22
Median
1.00
0.96
1.04
1.18
1.08
1.00
RMSD
1.7*
1.1"
2.5
3.5
2.0
1.1"
* Acceptance Limit
** Value Within Acceptance Limit
Mean = Mean of Solvent to Freon Ratios
SD = Standard Deviations of Solvent to Freon Ratios
RSD = 100 x SD/Mean
Median = Median of Solvent to Freon Ratios
RMSD = Normalized Root Mean Square Deviation of Sample x Solvent Means
September 1993
25
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Exhibit 5.
Summary Statistics For Alternative Techniques in the Determination of Oil and Grease
Solid Waste Stream, Sonication Extraction
All Samples (N=27)
Solvent
Soxhlct Frcon
Hexane
McCl,
Perchlor
DuPont
80/20
Freon
Mean
1.00
0.50
2.51
0.82
0.72
0.61
0.62
SD
-
0.32
9.25
0.57
0.55
0.61
0.42
RSD
-
65
370
69
77
99
68
Median
1.00
0.43
0.57
0.73
0.52
0.50
0.50
RMSD
1.5*
3.3
3.8
2.0
2.6
3.4
2.4
Non-Petroleum (N= 10)
Solvent
Soxhlct Ficon
Hexane
MeClj
Pcrchlor
DuPont
80/20
Frcon
Mean
1.00
0.49
0.63
0.89
0.64
0.53
0.58
SD
-
0.34
0.38
0.33
0.34
0.34
0.31
RSD
-
70
60
37
53
64
53
Median
1.00
0.48
0.62
0.81
0.70
0.56
0.54
RMSD
1.9*
3.1
2.6
1.3"
2.5
3.1
2.5
Petroleum (N-17)
Solvent
Soxhlct Ficon
Hexane
McClj
Peichloi
DuPont
80/20
Freon
Mean
1.00
0.50
3.69
0.79
0.76
0.66
0.64
SD
-
0.33
11.77
0.68
0.66
0.73
0.49
RSD
-
65
320
86
86
110
76
Median
1.00
0.43
0.57
0.70
0.49
0.47
0.48
RMSD
1.7'
3.5
4.6
2.5
2.7
3.7
2.4
• Acceptance Limit
" Value Within Acceptance Limit
Mean - Mean of Solvent to Freon Ratios
SD - Sundud Deviations of Solvent to Freon Ratios
RSD - 100 x SD/Mcan
Median - Median of Solvent to Freon Ratios
RMSD ** Normalized Root Mean Square Deviation of Sample x Solvent Means
26
September 1993
_
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Exhibit 6.
Summary Statistics For Alternative Techniques in the Determination of Oil and Grease
Aqueous Waste Stream, 90 mm Solid Phase Extraction Disk
All Samples (N=28)
Solvent
Sep. Funnel Freon
Hexane
MeCl2
Perchlor
MTBE
Freon
Mean
1.00
1.15
1.95
1.26
2.71
0.93
SD
-
1.03
1.75
1.01
2.81
0.62
RSD
-
90
90
80
100
66
Median
1.00
0.99
1.26
1.11
1.62
0.94
RMSD
1.5'
6.0
7.1
5.8
7.9
5.0
Non-Petroleum (N=9)
Solvent
Sep. Funnel Freon
Hexane
MeCl2
Perchlor
MTBE
Freon
Mean
1.00
1.09
2.27
1.44
2.57
0.94
SD
-
0.68
1.71
0.81
2.09
0.35
RSD
-
63
76
57
81
38
Median
1.00
1.00
1.83
1.36
1.67
1.00
RMSD
2.0'
3.2
6.3
3.7
5.4
1.3"
Petroleum (N=19)
Solvent
Sep. Funnel Freon
Hexane
MeCl2
Perchlor
MTBE
Freon
Mean
1.00
1.18
1.80
1.19
2.78
0.93
SD
-
1.18
1.80
1.08
3.15
0.70
RSD
-
100
100
91
110
76
Median
1.00
0.94
1.16
0.99
1.58
0.87
RMSD
1.7*
7.3
7.6
6.8
9.2
6.3
* Acceptance Limit
** Value Within Acceptance Limit
Mean = Mean of Solvent to Freon Ratios
SD = Standard Deviations of Solvent to Freon Ratios
RSD = 100 x SD/Mean
Median = Median of Solvent to Freon Ratios
RMSD = Normalized Root Mean Square Deviation of Sample x Solvent Means
September 1333
27
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Exhibit 9.
Summary Statistics For Alternative Techniques in the Determination of Oil and Grease
Aqueous Waste Streams, SPE Column (new version)
All Samples (N=20)
Solvent
Sep. Funnel Freon
Hexane
80/20
Freon
Mean
1.00
0.96
1.13
0.76
SD
-
0.38
0.49
0.33
RSD
-
39
43
44
Median
1.00
0.95
1.16
0.78
RMSD
1.7'
3.5
4.6
4.5
' Non-Petroleum (N=5)
Solvent
Sep. Funnel Freon
Hexane
80/20
Freon
Mean
1.00
1.07
\1-13
0.72
SD
-
0.38
0.24
0.22
RSD
-
36
21
31
Median
1.00
1.08
1.19
0.77
RMSD
2.6'
1.6"
1.8"
6.2
Petroleum (N= 15)
Solvent
Sep. Funnel Freon
Hexane
80/20
Freon
Mean
1.00
0.93
1.13
0.78
SD
-
0.39
0.55
0.37
RSD
-
42
49
47
Median
1.00
0.93
1.13
0.80
RMSD
1.8"
3.7
4.8
4.3
* Acceptance Limit
** Value Within Acceptance Limit
Mean = Mean of Solvent to Freon Ratios
SD = Standard Deviations of Solvent to Freon Ratios
RSD = 100 x SD/Mean
Median = Median of Solvent to Freon Ratios
RMSD = Normalized Root Mean Square Deviation of Sample x Solvent Means
30
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of OH and Grease
Exhibit 10.
Summary Statistics For Alternative Techniques in the Determination of Oil and Grease
Aqueous Waste Stream, Infrared Analysis
All Samples (N=36)
Solvent
Sep. Funnel Freon
Flon
Mean
1.00
1.62
SD
-
1.55
RSD
-
96
Non-Petroleum (N-13)
Solvent
Sep. Funnel Freon
Flon
Mean
1.00
2.33
SD
-
2.29
RSD
-
98
Median
1.00
1.01
Median
1.00
2.05
RMSD
1.5"
8.3
RMSD
2.0*
9.4
Petroleum (N =23) "
Solvent
Sep. Funnel Freon
Flon
Mean
1.00
1.21
SD
-
0.71
RSD
, ..
58
Median
"f.i.oo
1.00
RMSD
1.7*
6.4
* Acceptance Limit
** Value Within Acceptance Limit
Mean = Mean of Solvent to Freon Ratios
SD = Standard Deviations of Solvent to Freon Ratios
RSD = 100 x SD/Mean
Median = Median of Solvent to Freon Ratios
RMSD = Normalized Root Mean Square Deviation of Sample x Solvent Means
September 1993
31
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Exhibit 11.
Normalized Root Mean Square Deviations
Aqueous Waste Stream, Separatory Funnel Extraction
o
CO
«
5.0 -
4.5 -
4.0 -
3.5 -
3.0 -
2.5 -
2.0-
1.5 -
1.0-
0.5 -
r> n -
O
O
o
•
o
O Non-Petroleum
• Petroleum
Non-Petroleum
Petroleum
Acceptance Limit
Hexane MeCl2 Perchlor DuPont
Solvent
80/20
NOTE: Points below the respective Acceptance Limit are not significantly different from separatory funnel extraction with Freon
32
September 1993
_
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Exhibit 12.
Normalized Root Mean Square Deviations
Solid Waste Stream, Soxhlet Extraction
5.0 -|
4.5 -
4.0 -
3.5 -
3.0 -
2.5 -
2.0
1.5
1.0
0.5
0.0
O
o
O
DuPont
80/20
O Non-Petroleum
• Petroleum
Non-Petroleum
Acceptance Limit
Petroleum
Acceptance Limit
Hexane MeCl2 Perchlor
Solvent
Note: Points below the respective Acceptance Limit are not significantly different from separatory funnel extraction with Freon
September 1993
33
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Exhibit 13.
Normalized Root Mean Square Deviations
Solid Waste Stream, Sonication Extraction
Q
|
5.0 -
4.5 -
4.0 -
3.5 -
3.0 -
2.5-
2.0 -
1.5 -
1.0 -
0.5 -
0.0 -
*
•
0
• 0
0
O Non-Petroleum
• Petroleum
O
•
Non-Petroleum
Acceptance Limit
Petroleum
Acceptance Limit
Hexane MeCl2 Perchlor
Solvent
DuPont
80/20
Freon
NOTE: Points below the respective Acceptance Limit are not significantly different from separatory funnel extraction with Freon
34
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Exhibit 14.
Normalized Root Mean Square Deviations
Aqueous Waste Stream, Alternative Techniques
Non-Petroleum Samples
15 -i
12 -
9-
6 -
v SPE 47 mm Disk
T SPE 90 mm Disk
D SPE Column (new)
• Infrared
D
V
Acceptance Limit
SPE Column
...... SPE Disk
Infrared
1 1 : 1 1
Hexane MeCl2 Perchlor 80/20
Solvent
Freon
MTBE
Flon
NOTE: Points below the respective Acceptance Limit are not significantly different from separatory funnel extraction with Freon
September 1993
35
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Exhibit 15.
Normalized Root Mean Square Deviations
Aqueous Waste Stream, Alternative Techniques
Petroleum Samples
8-
6-
4-
2-
v
D
v SPE 47 mm Disk
T SPE 90 mm Disk
n SPE Column (new)
• Infrared
Acceptance Limit
SPE Column
~T... SPE Disk
Infrared
Hexane MeCl2 Perchlor 80/20
Solvent
Freon
MTBE
Flon
36
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Exhibit 16.
Mean Solvent-Freon Ratios for All Techniques Tested
MTBE
Flon
Freon
^ DuPont
I
80/20
Perchlor
MeCl,
Hexane o
0.4
Ideal
Ratio
O D A
CD
A Aqueous
• Soxhlet
o Sonication
a SPE Column (old)
A SPE Column (new)
o SPE Disk 47 mm
* SPE Disk 90 mm
• Infrared
3.3
Solvent to Freon Ratio
September 1993
37
September 1993
39
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
6.2 Sonication as an Alternative Technique to Soxhlet Extraction
For sonication as an alternative extraction technique, perchloroethylene was equivalent to
Soxhlet extraction with Freon 113 for petroleum samples. Therefore, perchloroethylene is a leading
candidate replacement solvent if sonication is considered on an equal footing with the currently-used
Soxhlet extraction technique. Perchloroethylene and n-hexane would both have two acceptable
RMSDs out of a possible six.
6.3 Alternative Techniques for Aqueous Samples
None of the alternative techniques for aqueous samples yielded results with RMSDs within the
Acceptance Limits when an adequate number of samples was tested. Some of the results using the
Varian SPE columns, particularly the newer version, are promising. It remains to be seen, however,
whether solid phase extraction (SPE) is applicable to samples with high dissolved solids. Without
further work, none of the alternative techniques for aqueous samples can be recommended as a
replacement for methods 413.1 and 9070 at this time.
6.4 Retention and Elimination of Solvents for Further Study
Based on these results, the results of the ealier study by EMSL-Ci, and comments received by
EPA, the following solvents will be retained or eliminated from further consideration as candidates
for replacement of Freon 113 in oil and grease measurement as follows:
1. /z-hexane will be retained because the results for petroleum-based solid samples and non-
petroleum aqueous samples are within the Freon 113 Acceptance Limit and because n-
hexane was used in the oil and grease measurement prior to the advent of Freon 113.
2. Perchloroethylene will be retained because the results for non-petroleum aqueous samples
are within the Freon 113 Acceptance Limit and because perchloroethylene can be used in
the measurement of oil and grease by infra-red techniques.
3. Although the results for petroleum-based solid samples extracted with w-hexane/MTBE
(80/20) are within the Freon 113 Acceptance Limit, 80/20 will be eliminated from further
study due to concerns about laboratory safety and solvent composition change during
storage. Such concerns were raised during Oil and Grease Workshops held by EPA in
Norfolk, Virginia and Boston, Massachusetts to allow regulated industries, laboratories,
and other interested parties the opportunity to discuss the status of the Agency's Freon
113 replacement efforts.
4. Methylene chloride will be eliminated because the results produced are far from results
produced by Freon 113.
40
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
5. DuPont 123 will be eliminated because results produced using this solvent are not within
the Freon 113 Acceptance Limit for any category of solid or aqueous samples and because
DuPont 123 is a Class II CFC that will need to be phased out eventually.
6. Cyclohexane, which was not formally evaluated during the first phase of the study, will
be considered in future evaluations. The decision to evaluate cyclohexane is based on
concerns that have been raised at the EPA workshops and elsewhere concerning the
neurotoxicity of n-hexane.
6.5 Retention of Alternative Techniques for Further Study
Of the alternative techniques evaluated hi this study, only sonication extraction of non-
petroleum solid samples produced results equivalent to existing techniques with Freon 113. The use of
smaller solvent volumes required by solid phase extraction (SPE) and the increased sensitivity of the
non-dispersive infrared technique might warrant further study.
September 1993
41
-------
-------
SECTION 7
FOLLOW-UP AND POSSIBLE PHASE II ACTIVITIES
A presentation of preliminary results was made at the Pittsburgh Conference on Analytical
Chemistry and Applied Spectroscopy in Atlanta, Georgia on March 4, 1993. Additional presentations
were made at U.S. Environmental Protection Agency (EPA) workshops held on May 4, 1993 in
Norfolk, Virginia and on June 30, 1993 in Boston, Massachusetts. The purpose of these workshops
was to provide a forum in which all interested parties could discuss the preliminary results of Phase I
and possible options for Phase II.
A notice of the availability of the results of the first phase of the study will be published in
the Federal Register, with a request for public comment. This report, with subsequent revisions, if
any, will be mailed to those responding to the notice. The notice may also include a study plan for the
second phase of the study and a request for regulated industries to produce data using the one or two
most promising alternative solvents.
If EPA proceeds with a second study phase, it will likely be designed to assess the precision,
accuracy, and comparability of the one or two most promising alternative solvent/extraction systems
and measurement techniques from the Phase I study results. The range of industrial effluents might
also be expanded in the second phase, in part through cooperative efforts sponsored by various
regulated industries. Alternatively, the second phase of the study may involve side-by-side testing of
an alternative solvent versus Freon 113, so that each permittee can develop its own correction factor
(see below) pending renewal of their NPDES or RCRA permit.
Options for Replacement of Freon 113 in the Oil & Grease Method
The Phase I results have shown that no solvent produces results identical to the results
produced by Freon 113, but that «-hexane, perchloroethylene, and the 80/20 mixture of n-hexane and
MTBE produce results equivalent to Freon 113 for some samples. These results suggest that if an
immediate decision to replace Freon 113 needs to be made, one or more of these solvents should be
selected, and that any Phase H effort should concentrate on these solvents. Concerns have been
raised, however, about laboratory safety and solvent storage problems associated with handling the
80/20 mixture. In addition, the reduction in solvent use afforded by solid phase extraction and the
lowered detection limit attained with non-dispersive infrared determination, provide compelling
reasons for further study of these techniques. EPA therefore desires to be as comprehensive as
possible in exploring options for any possible Phase n study, but needs to narrow the focus of these
options once they have been explored.
The range of options under consideration at this time is given below.
• File for an exemption under the Clean Air Act Amendments for the use of Freon 113 in
the oil and grease method.
September 1993
43
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
• Use recycled Freon. Laboratories would reuse Freon 113 recovered from earlier testing.
• Choose one new solvent.
• Choose more than one new solvent, if different solvents are found to work best on
different kinds of samples.
• Use the Total Petroleum Hydrocarbon (TPH) method for petroleum-contaminated
samples.
• Switch from gravimetric to infrared spectroscopy-based methods.
• Use solid phase extraction (SPE) rather than solvent extraction.
• Develop a correction factor or factors to be applied to data generated by solvents other
than Freon 113.
«.
EPA solicits comments on these options and seeks any other options for resolution of this
issue. As comments and other information become available, the Agency will continue to attempt to
keep all interested parties informed. Because Freon 113 will not be commercially available after
January ,1, 1996, and because it is desirable to phase out the use of all Class I CFCs as expeditiously
as possible, the first two options listed above may not be viable.
44 September 1993
-------
APPENDIX A
SITE SUMMARY
-------
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Exhibit A-1.
Site Summary
Freon Replacement Study
WASTE STREAMS
FACILITY TYPE
Paper «Ul - -- -;,
Oil Production Site
Leather Tannery
rito;-" ,~:7-7-,,-
Petroleum Refinery
rial, Laundry ;\;,
Textile Hill
Metal Finish ing Plant ••
Fish Oil Plant
Rendering; Plant ' ,.
Coke Plant
Siati^h'terhouse ,' *
Wood Preserving Plant
Br-ilUw Fjlttid, supplier '
Sofia
Poultry Plant
Rolling Hill
'
Mayonnaise Plant
jSwfoori pJwnV ,
Abrasives Plant
Seafood Plant
Poultry Plant:,
Heat Packing Plant
Can Manufacturing Plant
•
Oi ly Water Treatment Plant
Can/*(a«fcifa<:turfns Plant,,
Can Manufacturing Plant
Handling Facility
Polymer Plant
Restaurant ,' "'' ' ,;;
Industrial Laundry
formulating Plant '
Leather Tannery
Petroleum Refinery
INDUSTRIAL
CATEGORY/ACTIVITY
, Pulp fc Paper , - ,.
Oil & Gas Extraction
POTW ' •• •• * * ""*
Leather Tanning
' POTtf; *
Petroleum Refining
Industrial ,LauMries, ' '-.'
Textile Manufacturing
Metal f injshing
Soap & Detergent Manf .
Iron & Steel Manufacturing
Heat Products & Rendering -
Timber Products
Oil & Gas Extraction •"
Contaminated Soils
Poultry Processing- %
Iron & Steel Manufacturing
Petroleum, Stttrage' Facilities
Miscellaneous Foods
seafood Processing ' s ;
Abrasives Manufacturing
Petroleum Stora'geJFactlltlei
Seafood Processing
- Poultry Processing
Meat Products & Rendering
" transportation Facil.i t ies
Coil Coating
- Miscellaneous Foods ' "
Shore Reception Facilities
', Cpil Coating
Coil Coating
Drum Reconditioning
Organic Chemicals
AQUEOUS
Bleach Slant Efftyeot:
Dispersed Gas Flotation Eff.
" Mfcary Effluent
"•Jsfondary Effluent
Primary Effluent
API Separator Effluent
Secondary Effluent
, 'Dissolved A$* jrtota'tion Eff/
Lagoon Effluent
Separator Effluent
Primary Effluent
Secondary Effluent
Primary Efftuent
" SecorxJary Effluent
Secondary Effluent
' - ieeoncfepy'Ef
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Exhibit A-2.
Wastewater Sources
Freon Replacement Study
LEVEL OF TREATMENT
UPSTREAM OF SAMPLE POINT
FACILITY TYPE
AQUEOUS
WASTE STREAMS
Oil Production Site
Leather Tannery
Petroleum Refinery
Industrial Laundry
Textile Ml I v ;
Metal Finishing Plant
Fi$ OH Plant '
Rendering Plant
Cok» Plant" "
Slaughterhouse
Wood Preserving Plant
Poultry Plant
OH Terminal
HayonnaisVPUnt n ^
Seafood Plant
Abrasives Plant
Oil Terminal
Seafood Hant
Poultry Plant
Meat Packing Plant
Jtaflroad Yard
Jjanj^nufacturing Kant ,
Soup Plant
"pXly Water^TreatmenJ^plant
Can Manufacturing Plant
Can Manufacturing Pt ant
Drum Handling Facility
Industrial Laundry
Formulating Plant
Leather Tannery
petroteu* Refinery
Dispersed Gas Flotation Eff.
: ~-secondary Effluent <•
Primary Effluent
API Separator Effluent
Dissolved Air Flotation Eff.
LageojfEf fluent' „,
Separator Effluent
-
Primary Effluent
, ^ecojxfary Effluent
Primary Effluent
Secondary Effluent
Se£5j|idary fefflusnt
Secondary Effluent
Filter Effluent
Separator Effluent
>,tf$erjf fluent
Secondary Effluent
~: Process Uastewater
Separator Effluent
Prfihafy Affluent
Dissolved Air Flotation Inf.
LagOOft Effluent
Dissolved Air Flotation Eff.
^ t»riroafy Affluent ' _,
Dissolved Air Flotation Eff.
< fepaVaf6> Eff tuenie
Dissolved Air Flotation Inf.
l Air Ftotation Inf.
j *ir Ftotation Eff.
Filter Effluent*
Primary Effluent
Prttnary Effluent
Primary Effluent
API Separator
interceptor Effluent
BASIS
N
P
P
P
N
P
N
N
P
N
~ P
P
*N
P
N
P
k
N
NONE PRIM. SO/W BIOL.
P
K
N
P
f
X
X
LEGEND
BASIS:
P - Petroleum
N - Non-Petroleum
LEVEL OF TREATMENT:
NONE - No Treatment
PRIM - Primary Treatment
SO/U - Secondary Oil/Water Separation
BIOL - Biological Treatment
A-4
September 1993
-------
APPENDIX B
DATA SUMMARY
-------
-------
r*
cuco
LUUUI—I
oo s:
O.D
C to
CU to--
O 3T3
CZ O *r—
O CUr—
C_) 3 O
crto
CU fO
to E
03 CT-
C£3 i ^
O
£
&
V V
1
CO
O_
LU
o:
a
LLJ
CD CD
V V V
CD CD r-H CD
LOLOLOLOLOLO
V V
t-H
o_
LU
| —
LU
^»
8
LU
B
DC
CJ
1 1 1
1—
^
cl
00
LU
'
FACILITY
CD CD
V V V
CD CD t — ICD
LOLOLOLOLOLO
CO
Q_LU
cr
1 — 1
— 1
1 — i
I
t
LU
g
0_
_J
CO
t-H
CM
CM
CM
CM
_,
a-L
UJ
1
LU
C3
3
O
O
LU
x F5a:5!
LiJ tJ.J iii ^3
I-)
UO
CM
CM
LO
CM
CD
U_
3
S
i — i
u_
u_
UJ
!g
o_
co
t— 1
co ro ro ch CTI PH
cr> to T-H CM to co
r^* co ^D LO T-H r-—
to e~^% I-H to vo to
t-H -^J-i-H
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CM l~- CM f~~ i-H CO
CM CM T-H LO CO T-H
CD CD CD CD CD CD
CD CD CD CD CD CD
CO CD CO CO O^ CTl
CM CM T-H CO CM T-H
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CM CD CD
f"— ^* r-H CO LO ^^
CMCOi-H t— ICM
CD CDCDCDCD
CD CDCDCDCD
CD CDCDCDCD
r~» CMCOCOi— 1
LU
1=1
§
S
i-H LUO
— 1 rv LU UJLLJCD
C3 Ll_ DT 2: CL. CO
OO
•z.
o
I-H
t_J
1
LU
CO
0
LU
ce
LU
f—
1
CM
CM
CM
CM
CM
'LOCMrHOirvJiTl
CD to CD CM to r-H
CM«S- LOCM
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD tO CD IO CD
T-HCOVOtOCHCO
T-H CM
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
•— 1 i-HCM
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CO "S- CD CD LO
CMLrtCMVOLOtO
i-H CM
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD to CD CD CD
T-H T-H tO LO IO T-H
i-Hi-H CMi-Hi-H
LU
0
|
£
CO LU
CM ~Z.CZ.
T-H LUO
I— LU 1 1
C3CD rv 1 1 1 1 1 1 1 1 i f—^
OO
tj
0
OO
LU
CO
o
S
1
CM1
oo ^f vo^j*
CD CD CD CD
V V
i-H CD CD CO *3- CO
r^UDLOCOtOLO
V V
CO CD CD t-H CO CO
tOLOLOCOLOLO
V V V
r^ooenmo
V V V
"* CD CD i—l CD CD
UJ
O
|
O
T-H LUO
cr
i
I-H
£
^
o.
rn
1
en
co
CM
vo
CD
r^
co
T-H
LO
3
u.
i
t—4
fe
^
J
CO
R
OOCMCOOOC*}
CSOi— ICOO
CD CDCDCDCD
•SJ-LOCOCOi-H
i-H r-H T-H tO i-H
CD CD CD CD CD
i-H i-H i-H tO T-H
CD CD CD CD CD
^ LO CM T-H i-H
S
f*"1
LU
S
—1 LU
1
en
oo
»-H
O
LU
t
LU
^
5
_j
CO
1 —
i-HTtmrxico
T-T-<
CD CD CD CD CD
i-H CO CD CD P—
CM CM CO CO CM
CD CD CD CD CD
CM CM CO CO CM
CD CD CD CD CD
CM CM en to co
CM cocMr^* CM
LU
1=1
1— 1
§
5
5
— 1 LU
1
1—4
o
oo
Lu
to
IP
:c
1
1
ct!
-------
00
coV
*
Si
CO
g
..SI
r-.
o
o.cooo*3"co
CMt-!t-H»3-LOCM
CM
O
OOCMIOCMCM
3 O. CO CO .-^ t—1 LO LO CO CO ^ CD
•» t—I ^J- VO CO t^ t—IO t—I •* *d-O
3t-.CTlO.OCO -a* ^" CO
CM CM'a-COt-l
OOOOOO
CD
LO
OCOOOOO
t-ioicdcovdid
OC3CDCDCDCD _____
CDCDCDCDCDCD OOOOOO
^- f^* t-H ^D ^O ^O ^D t^ ^D ^D ^D CM
COCOCMf CMCM t-H OO OO ID t-H I--
OOOOOO
O CD CD <=> O d
OOOOOO
CD CD CD CDC
CDCDCDCDC
SC3^^C3O CDCDCDCDCDCD
*"^ h^> LO CO CT> ^~* t-H ^O ^D CD t—I
CO-=l-CM-*CMt-. t-HCOOO'S-VOOO
t-H t-Ht-l
CDCDOC3C3O OC3C3CDC5O
^ CO CO «4" i— 1
COt— *^J-CMCO
3CDCDC3
CMCT>t-ICM«a-0
t—I t—It—It—It—I
CO
03
3
T3
ct
t—4
£
5 z
o S
. -
u. u_
LLJ LU
U- Lu
tti
>-
Lt
LU
OO
^^
fc
I
co
LU
01
S
1
>
"o
.0
CO
a
CM
*t CM
CO CM
CM CM
CO CO
CM CM
CM CM
CM CM
CM CM
CM
CM
S
-------
U)
.
OJLO
J^
E O)
4-> ro
orv
(u
oo
•
§
0
h-
00
^
OO
1/5
..s§ i
t/) ro
c to to
0 CO
•r-i — Q)
•f-> CLCT
£- ro 13
+J tor—
§ to-^
o zs-o co
EC O 'i— Ct_
5QJi — LU
=S O Qi
crto
o> ro
to c
s- "~ a
CD— l±t
(= £ CSI
i — O_
•r- LU
o a:
t-H
c\-
LU
ce
5
~-*i
oo
LU
^
3:
LU
I—
oo to
CD t^1*
V V V V
CD CD t-H CD CD CD
LOLOLOLOLOLO
V V V V V
CD CD CD CD CD CD
LOLOLOLOt-HLO
t-H
V V V V V
CD CD CD CD OO CD
LOLOLOLOCnLO
V V V V
oo-3-ooo
LOLOLOLOOOLO
CSI
LU
O
5
CD
3:
CO LU
fez^irio
23x^5!
ID O£ UJ LU LU CD
ou_:i:s:a-oo
^_^
s^ ^
o*
i — i
cr
i — i
'-o^-r^t-H'ooo
i-H OO CO CO r-i LO
CO i— 1 <— « «— 1 CM VQ
co •<=}•
i— 1
CD CDCDCDCDCDCD
CD CD^-^-LO^"CD
r-. CD
t— i i— i
0 CD CD CD CD CD
CD CO CO O"l CD CD
CD LO C\J CO GO CD
CNJ LO
CO t-H
CD CDCDCDCDCDCD
CD • CD CD O"» t — ILOCD
CD OO •«=!• LO VO CO
CM LO
CD CDCDCDCDCDCD
•xl" OO CO I""*-*. »~H CXJ
LO ;t-n cn
LU
Q
5
3
o
CO LU
csj :zoi
t-H LUC)
1 — LU 1 1
=2 12: 2: >- 3: CD
2: o o - 3:0
oocsi
— ) rv* LU LULU CD
^-^
OO
h-
LU
3:
X
M-LncsiLocnco
t-H cn t-H -^t- r^. co
co crt ' ' to cn LO
CD to t-H CD CSJ r-~
CSICSJCSJ CSI
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
*5t"CO CO OO i-H to
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CSI CSI CD lO CD CSJ
CSIOOLOLOv-lt— 1
t-H
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CD CD CD CD CD cn
CD to CD to CO CO
LOCO CO to to
CDCDCDCDCDCD
CD CD CD CD CD CD
CDCDCDCDCDCD
CD CD CSJ CD CD t— 1
i-H CD CSJ CO CD «3"
VOI-— LOCD
r-H
LU
g"""1
_,
3:
CJ
CO LU
csj 2:0;
T-H LUC)
oo§3:cjcsi
Q-LUXI — Qi — .
ZDOiLULULUCD
t/1
*
2:
O
1—
S
t — 1
p
•3-«3CJ\CO!-^r-<
*=t- r^ to r^ to CD
co csj t— itocnco
t-H CD LO Cn CD LO
CD CO -3- "3- LOCD
COCOt— ICO LOt— 1
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CD CO cn i—t CD to
to cr> LO 01 ** r~-
r-H i— 1
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
csj vo *3- cn cn to
CSJ t— 1
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
000000
CDI — tO*3-CSJOO
cncnr^-Locoi —
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
tO CO LO CSJ VO OO
t-H i— It-H
LU
Q
££
o
'g
CO LU
zzrz: 5^5:0
Q-LUXI — Qi---,
— > CCCLU LU LU CD
^^
OO
1—
LU
3:
X
v
to
a>
-a
c
ro
ai
"o
•§
LU Lu
LL.
>- LU
LU
LU
a:
a.
LU
§
oo
a
oo
LU
CD
oo
CD
oo
LU
CD
to
s
cu
LO
CSJ
CSJ
CSI
CSJ
LO tO
CSI CSI
CSI CSI
CSI CSI
CSI CSI
to
CSJ
CSJ
CSJ
CSJ
CSJ
CSI
CSJ
Esj
CSJ
CSJ
CSI
t
o
u
I
la
CO
-------
e»j
s.
co«t
t^
_
1
s
4J
1
CJ
Ol
I-
Cfl
o
i
JE!
X"N
to
_ffl 5
col?
"if
S)l—
l/)-^
S3-0 CO
O"T"* *^
CUf— ' ' '
30 a:
O"CO
18 c
O
-^ii
gi s
s
!
O-
§
1
LLJ
Zp
s
tj
r^
S
£E
£
2
Hfc
C
Q
to
>2
H
^^
g
i— ICOCnVOLOCM
LOLOCMI — CTlh-
CD CD CD CD CD CD
CD co CM CM co cn
CO T-H CO t-H CM CM
V
CD CD CD CD CD CD
LOI — CT> LOCO CO
CMT— ICM CMCM
CD CD CD CD CD CD
CO «-H CO r-H CM CO
CD CD CD CD CD CD
LO^~ CM CD CM CO
CO CM OO CM T-H CM
S
I
O
CO LU
CM -ZLCX.
t— 1 LUO
|— LU_1_J
^^
<
or
_j
o
HH
_J
Lt
LU
gj
t
|
1— 1
<
CO
CM
CM
CM
§3
>— <
E
LU
d
6
vo
t-~ LO"*
.-i co^t-
V V V
0,LOLOLOI^CO
V V V V
LO CD CD CD CD CD
r-LOLOLOLOco
V V V V
CD CD CD CD CD CD
i-) i— 1
V V V V
LO CD CD CD CD CD
O*» LO LO LO i— 1 LO
i— 1
0
i
g
CO LU
CM zor
t— 1 LUO
^^
<
cy
_J
^f
t-H
_J
fe
•°^
LL_
LL.
LU
1
CP
a
00
en
CM
CM
CM
CM
^1-
VO
M-
VO
^t-
**D
u_
^
^-^
CD
UJ
^?
Lu
t— *
UJ
=>
t
UJ
1
LU
OO
O"»
CM
CM
r-HCO
CD CD
V V V
CDCDCDLOLO
LOLOLOLOLO
V V V
CD CD CD ^J- CO
LOLOLOLOLO
V V V
CD CD CD VO CM
LOLOLOLOLO
V V V
LOLOLO LO LO
S
|
5
CO LU
CM 3
.— 1 LU
1— LU_1
2Sx^5!
ISOiLUUJCD
^
^
J
o
u
LU
O.
OO
fe
"*^
— J
t
LU
|
LU
OO
cn
CM
CM
cr> CM CD o-> CM
co r^-. cr> co co ^
"^- r^« CD LO OO CO
i— 1 1— 1 VO t— 1 CO LO
CDVOCn^t-COLO
1— l1^- I— 1 1—1
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
t— ICTli— tLOl — .CD
CM t-H CM i—t t-H CM
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CM CO CD -set* CO CM
CM CM CM t-H t-H CM
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD OS CM VO VO CD
CM t-H CM t-H t-H CM
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
t-H •"* CD VO CO Ol
CM i—t CM t-H t-H t-H
LU
O
g
5
OO UJ
CO ZQ±
«-H UJO
|||||S
ZPO^LULULUCD
ca u. 3: E: o- oo
s
"^.
I-H
t_J
1 — 1
OO
LU
OO
gj
Ij:
O-
LU
OO
1 — 1
"-
CD
CO
CM
CM
CM
CDCMVOCMVOCO
r^. r^. I-H co vo LO
COLO CM CO COLO
t-H CM OO CD CD LO
CM CO CO LO t-H »*•
CO LO^" "3" LO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CO t— 1 r-. OO ^- CD
CM CO CM CM CM CO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CM CM CM CM CM CO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
T-H LO CM CO CO LO
CO CO CO CM CM CM
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CM CO CM CM CM CM
LU
O
§
CJ
CO LU
1— LLJ_J_J
^-^
t/)
tj
rn
o
OO
Lu
ta
3
OO
CD
1—
Q_
LU
OO
^1
*"•
f— t
CM
CM
CM
LO t-H CM «^f CO CM
CMt— ICMt-H t-H
CD CD CD CD CD CD
CD r^ LO cn cn LO
t-H
CD CD CD CD CD CD
CD -3- VO t-~ CM Ol
co os vo en cn co
CD CD CD CD CD CD
CD "3- CD CO t-H VO
CD r*> co co cn i^-
CD CD CD CD CD CD
CDCMVOCDCOI--
COVOI — CMOOVO
t-H t-H
S
01
0
J
3:
CJ
CO LU
1— LU_I_I
Q-LUXh-Oi--,
~~) r*s 1 1 1 1 1 fl 1 1 i r~~*
^-^
3
?
C
to
-a
-^
CO
a>
"o
CO
f
1
t-
5
T3
CO
"o
CO
[ *
t-
T3
CU
N
>,
C
to
1
cu
£_
^
CO
C
to
s
0)
H
CO
73
CU
+->
O
t
o
CJ
f
c:
to
^
o
c
cu
to
CO
CU
tc
o
LU
a.
OO
00
o>
0)
fe
-Q
Q.
of
CO
(0
OQ
-------
•
LU
a
i
00
oo
<1 E
CO fO
o a>
•r- 1 — d)
4-> Q.O
TO HT3
d *w
§3 w-<
0 gT3 CO
O CUr^ LU
c_> rs o a:
crco
cu to
CO
CD CD CD CD CD CD
fetittSiSS
CD CD CD CD CD CD
CD CD CD CD CD CD
C\J CO CD -^f LO t-H
••Q^-O ^~ *d~ ^"LO
LU
a
rv*
O
1
CO LU
(M -Z.C£
«— 1 LUO
|~ L^j — J — J
— i eg; [_U LU LUCD
a u_ in s: a_ co
^-^
^£
or
t— H
O1
1 — 1
^
LU
3
u_
u_
LU
§
oo
s§
CO
CM
^
CO
r-H
CD
CD
CO
LO
CD
CD
ts
CD
CD
CT>
LO
|
oo
1 — 1
1=1
LU
O-
00
1 —
3
Lu
Lu
LU
s
CO
ID
CO
CM
CO «3" CO CO CM CT»
LO CO *^t *^J- CO
OO
;^
i — i
h-
S
(_H
g
OO
LU
g
oo
§
10
CO
CM
r^cocooMOcn
'3-VOCOCMVOLO
CM i-H rH CO CO CO
Oir^COCMUDCD
CD i-H CD CD LO rH
r— 1
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
I^O1OO«3-LOCO
CM CM CM CM r-l CM
CD CD CD CD CD CD
CD C5 CD CD CD CD
CD CD CD CD CD CD
8 CD CD CD CD CD
CD CD CD CD CD
CO *3- CO CM CO CO
CM CM rH CM i-H CM
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CO rH U3 *st- LO CO
CM CO CO r-H rH CM
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
t£> rH CP» VO CO CO
CM CO CM CO rH CM
LU
Q
s
o
1
o
CO LU
t — 1 LUCD
1 — LU — 1 — I
go: LULU LUCD
^^
oo
1—
LU
1
oo
LU
_J
00
g
1
CO
CM
r^ co cr> crt co LO
CMCMCMlOt— 1
CD CD CD CD CD CD
CMrHCOCDrHI^
COLOCMCMLOrH
CD CD CD CD CD CD
CM r^ CM *3" CO rH
CD CD CD CD CD CD
COVOLOCDCO'3-
•— 1 "3" LO IO LO rH
rH
CD CD "=f CD CD CD
CO'^-VOCD'3-LO
LOCO ^flOCM
rH
LU
a
r^f
f^
5
CO LU
Cxi ZQ:
r-H LUO
1 — LU — 1 _I
CL.LUXI— G£*^.
^~~i fy^ 1 1 1 1 1 | j.<_f f~^
^.^
^^
O*
i — i
t— H
_J
1
Lu
Lu
LU
CD
1
CM
CM
CM
1
_J
(-H
s:
LU
X
LU
H-
OO
1^
CD
CD
CD
t— 1
C3
CD
i — |
r— |
O
CD
rH
r— 1
_J
^^
^^
f~^
LU
U_
LU
LL.
LU
LU
0
1
CM
CO
CM
CM
CM
_j
i
*-H
y
CO
O
~
*5
D»
^^
CZ
(n
jQ
•o
ta
JC=
u>
•)->
I
o
CO
, g
'E
•f~*
cu
5
"O
c
ru
CO
1
'o
to
1
+J
t_
o
it-
•§
B
*ro
cu
£_
i
^
C
IO
s
CO
1
£_
O
O
1
-^
c:
s
i
Jai
CO
S
LU
55
rs
CQ
PJ
0)
OJ
«^
1
«Q
t
•K*
Q.
1
-------
JOO
CD
si
§^
^g-'co
ro
rei c*r-
ta—iS
.a'gVg
o
8!
O Ot— ICOLO
CM CO CO CM|-^r~-t-HCOCO .—I
COCM.— IIOCOIO
t— t co CM in
h^ r^»
uncvj
O^ L
IO CO
HCOVJ-t-ICMCM
^O CMi—li—I
r-t CTlCOrH
CM
"CMr-ir-ICM *"" CM^CM COCoSoCMCO §
CDC3CSOOO
<=><=> OOC3C
OC9
O C=>
(=> CZ9 O O
C71 C7) CO COC
LOCOLOC3LOCM
CO LOCMi—I
O
CTi
I~»L
CM
OO
-COCTl
CMCM^COOOCO
C3 O O !=><=> C=> OC30C3OO
C3 C3C3^ C3C3^COC3C3 C3CDOOOCD LO^OLO^LOLO
_. __ CT> r-ivo-* oocacaeaea oczjooocs LO CM
IOLOLOI--.VDLO CM LO-ct-LO C3C3C3C3C3C5
OC3C3OOO
CO CO C7) CO CO CO _
CMCMCOCOCOCM CO C
CM
IO
I
**.
I
CO
v
CO
o>
1C
s
T3
OS
LLJ
=:
LLJ
_1LU
o-
hH
_J
CJ
co
»—i
Q
LU
O.
CO
cr
H-4
C^
I
u.
<
s
s
O
LU
g
>it
r55
o
LU
s
>—I
gg
CO
Jn
£=%£^
OC)«ZCJCM
.mg
Q u_ 3r 5: o_ co
CJ
§
t—I
I
a
I—I
g
i gg
:=^:f!=>
^SS^^^!
00
v-*
fc
1
CO
CM
i±.
a
i—*
|
CJ
LU
gg
:£^
i£3
T3
o
10
a
c
re
s
g
3 3
t t
S
*-*
O
CO
3
t
LU
CO
O)
CM
CM
CM
CM CM
1
gf
1
s
-------
CO
CM
CM
,-ILO
CM
C01-
CUCM
el) cn
cu
00
1
00
^^
00
LO
cu z
^" o 1 1 1
..O (= 5:
C 00 to
O rjj
•r-r— CU
+J O.CT
ra Et>
S- ro 3
4-> OOr—
C CO
CU UT^
O 3T3 CO
f~~ o •*-" *"*
Srjji — LU
3 O Ct:
crto
CO ro
ro CT-
£?'"" O
CD— l-ii
E E CM
O rv
i— 1
O.
| —
LU
>
I
0
OO
LU
O
i — 1
1
] —
^
a
OO
LU
<
£
>-
__
3
^^
ov
^
CD
CO
CD
CD
CO
CD
1 —
CO
Lu
5
S
• j|
1 i ,
i — t
| —
ki |
— I
u_
t '
UJ
c2
f-H
£
«
~
1 CM
COOOLO
i-H
or-.o
LO LO
CD COO
^J- LO
Or— 1CD
88-s
\ ij
Q
ctL
S
O
LU
LU
_ILU
1— XOQ
^
<
E
CD
CM
OO
LU
Cu
OO
£
1 , 1
:D
_i
Lu
1 '
LU
>;
i — i'
c£
Cu
co
CM
CM
CM
LTJ
r-t
V V V V V
CD CD CD CD CD CD
LOLOLOLOCMLO
t— 1
V V V V V V
CD CD CD CD CD CD
LOLOLOLOLOLO
V V V V V V
CD CD CD CD CD CD
LOLOLOLOLOLO
V V V V V
CD CD CD CD CD CD
LOLOLOLOLOLO
CM
LU
D;
3
o
CO LU
CM ^" i-V
,-1 LUO
O CD rv* LULU LU CD
CDLuDZSIQ-OO
^
^
—J
O"
1 — 1
— 1
t—
LU
^-^
3
LU
Lu
LU
>;
g
8
LU
OO
CO
CM
CM
CM
1—
^i
Cu
LU
O
CM
g—j
r-\
°
to
"*
LO
-*
to
^
o
r
Lu
s
LU
1
1 1
|
UJ
1
Lu.
LU-
LU
§
a
o
CJ
LU
OO
a->
CO
CM
CM
CXI
CF% CD CM CO *=3- CO
^O OO CD CTl ^1" (TV
CD CT> CO CO CM CO
l^Q LO^O ^3 I**1" CD
LOCO CO
O CD O OCD CD
CDCDCDOCDCD
CD CD CD O OCD
CM^f*!*^- CD **D
CO UJ
CM ^Oi
t— I LUO
h- LU_J_JO
ZD Qi LU LU LU CD
LO
^
I — i
! —
S
i — i
^
o
00
LU
§
I
OO
>
p
"*
LU
1—
OO
CD
CM
CM
CM
CM1-^- CDCDi — 1 CM
cn LO CM •— i r~- oo
CM CD CO "3" CM CO
OOLOI — <-HCDf^
COLO CM CM CM CM
CO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
r-. t-H to r-- CO CM
CM t-H «— 1 CM f^. i— t
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CO OO LO r-. LO t— 1
CM t—t i— 1 CM ^J- i-H
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
.— 1 CO CM CM CD LO
CO . i-HCMCMi— I
,— 1
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CO CD CD CD OO
r^- co ^- LO CM CM
CM i— tCM tp
LU
a
a;
3
nr
CO LU
CM 'Z.CZ.
^H LUO
1— LU_1— 1
CULUXI— CC"- .
^CCILULULUCD
OO
(—
LU
1
X
o
oo
LU
CD
§
I
OO
>
b
"^
LU
00
S
CD
,-ltOOO «5l-|^-
»— ^CMLO LOCM
V
tOLOCDOCM^H
LO user. LOCM r--
V V V
CD LO CD CD CD CD
LOCMLOLOLOCD
r-1 i— 1
V V V V
CD CD LO CD CD CD
LOLOLOLOLOLO
t— 1 ,
V V
CO CD ^f CD CO CM
to LO tO LO P-. tO
LU
(=1
|
O
co LU
Iz^fecD
CL.LUXI — Qi-->
— ^ rv' i < 1 1 i 1 1 1 i CD
C3Lu:3=^cuao
<
cr
_i
i — i
— i
h-
ra
Lu
1 1
LU
g;
§
Q_
t — I
CM
CM
LU
OO
— >
CD
£
LU
1—
Fn
=
OO
CO
i—j
^
CD
CD
to
•5J-
CD
^
CD
to
|
^^
s
s
g
LU
i — i
fe
LU
=>
^_J
Lu
1 >
LU
§;
s
Cu
i — i
Si
CM
CM
r-l CO
CD CD
V V V V
CD .— 1 CD CD CD CD
LOLOLOLOCDLO
V V V V
CD CM CD CD CD CD
LOLOLOLOCDLO
t— 1
V V V V V
CD CD CD CD CD CD
LOLOLOLOCDLO
V V V V V
CD CD CD CD CO CD
LO LO LO LO CO LO
LU
O
0
CJ
CO LU
CM ^SLCtL
^H LUO
CD O CM
^
i— t
3
8
LU
Q_
^"
UJ
U_
U_
LU
^
§
O
CJ)
UJ
OO
CM
CM
CM
i
i— i
V
cu
3
r~~
ra
>
rp
JO
-a
rc>
JIT
l/>
•g
I
"o
00
O)
c:
E
§ir~
s_
a>
•fj
•a
ra
oo
%
>
^
oo
4.)
t-
£
"S
N
ro
c
TC
|
a>
L.
3.
t/)
_^
c
ro
_Q
S-
ra
t/
a
—
K
j«;
I/
C3
UJ
CV_
CO
-------
CO O
\
^
52
i/)
cu z;
Is?
•r-r— Q)
4-> CLD
2 il
111 1
a> ro
|c =
2*^ o
•a"o>c5
m£ CM
c— a.
•r— 1 1 i
0 a:
i-H
O_
g
1
LU
S
IS
1
i
3
1
CO CMCDCO
CO
V V
OLO CD CO CD t^
LOVOLOVOCMCO
CO
V V V V V
CD CD CD CD CD CD
LOLOLOLOJ5;LO
v v v v
CD ^f CD CD CD CD
i-H
V V
CD CD CD CO CD CTl
*""*
S
|
S ^Oi
r* LUO
S
1
f-H
LU
— 1
U_
LU
CM
CM
CV4
a.
HH
1
1
g
r^
10
vo
i-H
LO
r-H
1
S
t— t
t EFFLUENT
1
CO
CM
CM
CM
CO CO VO CT) ^f CO
r~- LO CTl i-H CM «3
LO CT> CD i-H CM r^
vo co cn LO en i— i
LOi-HVOCOCMLO
CD CD CD CD CD CD
CD CD CD CD CD CD
CO CO LO t-H CO LO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
^3- CO CT1 CD CD i-H
CO CO LO i-H CO VO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
i— 1
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
1 — i-H LO CD CD CD
CO GO'S- CM CM LO
a
s
o
CO LU
CM -Z.CZ
pH LUO
0_ LU X F= Oi --x
LO
3Z
O
1— 1
1
LU
O
*— 1
8
SI
CM
CO CTl •<=*• r-H CD CM
CO t-H f-» CO O^ CO
CD LO LOCO t-H CO
CTt CO WO LOCO CO
t-H t-HWO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD Cn CD
LO *=f CO LO CO CO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
*5f «sj- CO WQ r-. CO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD t-H CD
r— Jt-HrHt-H r-H
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
O CD CD CD O CD
t-Ht-Ht-Hr-H r-H
LU
1
g
i— 1 LUO
Z3 a: LU LU LU CD
a u_ n: 2: a_ co
S
b
1
LU
1—
a
i — i
SI
CM
LOCO^OCMLOCM
CO LO CM t-H CO *3-
t-H r-H i-H CT> ^" r^*.
CO rH f-H f— 1 CO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
Cn CO WO 'd- i-H t-H
CM r- 1 r-H "g- t-H CO
CD CD CD CD CD CD
CD CD CD CD CD CD
Cn CO LO C\J i— 1 LO
CM i — It — 1 ^ t — 1 CO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD t-H CD
C\1CM«-H^- CM
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
Cn^OWOWOCMCO
CM t-H t-H «5t* t-H CM
S
3:
CO LU
CM 3O£
r-l LUO
O O^ "T~ LJ CM
D_ LU >< ^ Oi ^^
a LI_ in s: a- co
LO
I—H
LO
1 — i
1 — 1
g
LU
CM
SI
s
1 — 1
1
LO
t-H
LJ_
_J
1 — 1
g
CO CD ID O1 IO CO
^D r*^ ^D^t" ^O CM
CO LO CD CM i-H t--
i— ICMi— ICM
CD CD CD CD CD CD
CD CD CD CD CD CD
LO CD CD CO CD CD
^J" ^•^'Cn CO CD
i-Hi-H t-HCM
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
-^- CO IO r-H LO CD
i-H i-H r-H i-H CM
CD CD CD CD CD CD
CD CD CD CD CD CD
•"d-CDCDLOCDCD
LOVOCMO1CMO1
i-Hi— 1 i-Hi-H
CD CD CD CD CD CD
CD CD CD CD CD CD
i-H CD CD CD CD CD
^f CO CO CO i-H CD
i-Hi-H i— ICM
LU
O
a;
CO LU
CM Z.OZ.
rH LUCD
h- LU_I_J
2
fc
_i_
x
o
CO
_J
1 — 1
g
s
LO
CM
CM
CO CO CO *£> I^. t-H
CM CD CM CO CD CO
CDCDOCDCDCD
CD CD CD CD CD CD
t-H co cn t— ivo r-^.
CM t— 1 1— 1 CO r-H t-H
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
t— I r— 1 rH CO t-H i— 1
0 CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
t— 1 ^Q CM LO CM •«}•
CM t— t CM CM t-H i-H
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
co CD cn CTI CD ^r
CM t-H t-H CO CM CM
LU
CD
t— H
g
CO LU
CM -Z.OZ.
CULUXh— QC' —
rDC£LULULUCD
Q LU zc s: a. co
LO
t— 4
I
?
8
LU
1
t— 1
1
t-H
1
3
I
»
V
0)
re
JQ
•3
8
2
ra
i
Q_
LO
-------
•
LU
CO
CO
t/7
0) 2:
lOi —
CD tO' — .
o zs-a co
O COr^ LU
c_> =s o a:
a> fa "^
ro c. 'T—
CD i w
O rv
,-H
Q_
LU
t
LU
O
CO
OO COCO CD I OO
CO CM CM i— ICMO1
*3- LO CD CM CO CO
t— 1 CO -=t CM CM CO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
t— tr-cococnoo
CM i— 1 CM i— 1 •— 1 i-H
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CM CM
CO VO
V V V V
CD CD CD CM CD CD
LO LO LO CO LO Cf>
V V V V V
CDCDCDCDCDCD
LOLOLOI^LOLO
V V V V V
CDCDCDCDCDCD
LOLOLOCMLOLO
T— 1
V V V V V
CDCDCDCDCDCD
LO LO LO LO LO r--.
t— <
LU
3
rc
c_>
CO LU
CM za:
i-H LUO
h- LU_J_J
^^ />^ 1 1.1 1 1 1 1 1 I C"}
tO« — li — 1 CD 1 tO
i— 1 CO CVJ tD *£> LO
^)* COr— 1 LO LO CO
t— l^-f-HCOO^
IO
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CMCD^COCDf— 4
VOLOi— ILOrHCvl
•-I LO
CD CD CD CD CD
CD CD CD CD CD
CD CD CO CD CD
LOCDf— 1CDCD
t— 1 I—I
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
i— 1 CD OO CM CD i— 1
r-~a->.— ci— i CM cvj
t-H r-H
CDCDCDCDCDCD
CDCDCDCDCDCD
CD CD CD CD CD CD
rH CD CO CO CD O
«* CD CVJ ID CD CM
i— 1 CO
i— 1
LU
1
6
CO LU
CVJ ^ Oi
oo5n:ocM
Q_ LUX I — rv*^^
CD Lt_ rc SICL.OO
C? OO CO tO C7 CVJ
CO CM LO CO O1 LO
Ot VO CTi CD *JD LO
i— ICVJCMi— 1
CDCDCDCDCDCD
8 CD CD CD CD CD
CD CD CD CD CD
CDCDCDCDCDCD
.— ICOCMCO'3't— 1
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CD CD CD CO CD CD
i-H^l-^- i— ICM
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CD CD CD CD CD CO
CM i— 1 1— 1 CM CO t^
CM i-H i— 1 LO LO
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CD CD CT% CD CD CD
t-H CO LO CM C^ CD
CMCO CO«*i-H
LU
O
5
CO LU
CM ZC£
c— 1 LUO
^^ rv r
^"^ Ll_ ~*~ ^r" n CO
cv/ cvr
CM U3
V V V V
CDCDCDCDCDCD
LOCMLOLOCMLO
V V V V V
LOLOLOLOCOLO
i— 1
V V V V
CDCDCDCDCDCD
LOlOLOLOr^LO
CM ,— 1
V V V V V V
CDCDCDCDCDCD
LOLOLOLOLOLO
»
g
g
CO UJ
S LU^S
t~> c^ •<£_ *T* c^) c\]
Cl U_ ^ ^Q_ OO
o
F
LU
X
o-
Of
co
h-
LU
•g
LU
CL.
CO
uu
LU
CD
CM
CD
CO
CM
3
Q_
§
*«.
I
-------
00
•
I
co
..sf I
lii
+3 S?r2
C (/)
C
O^j5
r— O.
5t 1 1
ce
t-H
CL
LU
CC.
IS
1
ll.l
Qj
1
I
U1
J_
K-
£
V V V V V
CD CD CD CD O
LOLOLOLOLO
V V V V V
CD CD CD CD CD
LOLOLOLOLO
V V V V V
CD CD CD CD CD
LOLOLOLOLO
a
1
pro ^ LLJC_>
^-%
r-H r-H r-H r-H CM t-H
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CO CO I---* "3- LO
r-H t-H r-H r-H CM r-H
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CMIOI — .LOCO»i>
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CMr-H«Dl — COLO
t — tOO t — I t — 1 r-Hr-H
LU
C3
C£.
31
O
CO LU
CM Z0±
r-H LUO
=DC£LULUUJCD
CO
^
P?
i — i
CO
1—
LU
CO
in
1
§
r-H
CO
(NJ
CD co ^=f oo-^r^c
CO -
r^. oo oo oo <*o CD
rH OO O-» LO OO r*-
COCO t-HC\Jr- 1
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
t£> -sf LO r^ o^ oo
CD CD CD CD CD CD
OO CD CD CD CD
CDOOOCDCD
CDCDCDOOO
co r^. to oo CM to
t— 1 1— 1 1— i t— 1 C\J r-1
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
co -cfcj- LO r-. 0-v
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
^•t — l^T^-CO CO
LU
1
g
CO LU
CM ^ fV
r-H LUO
\— LU_I_J
^:s:^>-3icD
oo SLU LU LU CD
i/>
fc
X
CD
CO
j—
LU
CO
S
LU
1
CD
t-H
CO
CM
CM •* r-~ cn r~~
CO CD CO CO P"-
V
CD Cn CO CD CD ^~
LO^OLOCMLOCn
V V
CD CD CO CD CD CD
LO r-H LO "=1- LO CO
t-H r-H r-H
V V V
CD CD r-H r-H CD CD
LO LO LO CO r-H LO
CM
V V V
CD CD T-H CD CD CD
LOLOLOLOCDLO
r-H CM
LU
O
CD
CO LU
CM -ZiOi.
r-H UJO
(— LU 1 1
ooSaEScM
Q_LUXt-— O£ — -
— i rt* LULULU CD
2
a
i — i
(—
13
U_
U_
LU
LU
CO
CD
T-H
T-H
CO
CM
i — i
r-H
O
CD
t-H
CD
T-H
CD
T-H
—1
1 1
2
Q
S
:s
u_
Lti
SEPARATOR
CD
T-H
t-H
CO
CM
CD UDCD
r-H CDCM
V V
CO CD CD CD «^-
r--LOLOLOUD
T-H
V V V
IO CD CD CD CD
T-H
V V
CD CD CD CD CO
COLOLOLOr*.
T-H
LU
in
a;
3
C_3
LU
LU
_1 LU
HI DiUJ E LU
J^ 1 i — T" g^ Q
i
en
CO
i — i
m
LU
to-
co
5
^3
Lu
Lu
LU
1—
Q-
LU
CO
CD
T-H
T-H
CO
CM
CO ^f *^-
CM CDLO
V V
CO CD CD CD Cn
CnLOLOCMCO
t-H
V V
CD CD CD CD CD
T-HLOLOCMCO
T-H r-Hi-H
V V V
CO CD CD CD CD
r-H
LU
m
1
o
LU
_l LU
LU fV LU |— LU
3
E
S5=
CO
r— 1
LU
a.
CO
1—
^>
Lu
LU
LU
SEPARATOR
CD
r— 1
T-H
CO
CM
I
GO
J3
•a
|
O)
£_
ra
CM
-------
CO
cn
CftLO
rHOJ
i!'
oo
LU
a:
^_^
00
Ol
s @
in m
c: to co
O CLO
n3 ET3
t- (O 3
-t-J tOi —
O) 10"^!
O
LU
cn
Q
H-
00
^
S
CO
UJ
oi
CM
Q_
UJ
Q_
LU
o;
1 —
^
LU
••>
3
00
LU
S
^
C-J
LU
1—
^
S
OO
H^
§
aflt
O
O
00
fn
1 — 1
CJ>
"-
Cft rH tOrHrH
r*-*. to to t — ICM
LO
V
CD CD CD CD CD CM
"=t- LO CO LO tO to
rH COrHrH
V V
CD CD CD CO CD CD
CM CD rH
rH
V V V
CD CD CD CD CD t^
LO LO LO •* CO CO
V V
CD CD LO CD CD CD
t~^ LO LO i-H to LO
rH CMrH
LU
§
nr
CO LU
CM za;
rH LUO
Q-LuSl— Oi^.
ID Q± UJ LU LU CD
g
o
i — i
i
i —
LU
^
LI-
LI-
LU
o;
LU
b
u_
rH
rH
rH
CO
CM
_
^
s^
UJ
oo
p
CD
CM
rH
=>
CD
CD
rH
CD
S
CD
CD
O
^
s
s
§=
u_
1 — 1
1 —
LU
I
LI-
LI-
LU
C£.
LU
1—
I
U_
, — I
rH
CO
CM
CD rHrH
^ 00
V V
CD CD CD CD CD
rHLOLOLOCM
CO COrH
V V
CD CD CD CD CD
COLOLOLOCM
CO CO^-I
V V
00000
COLOLOLOCM
CM COi— 1
LU
g
— [
nr
LU
LU
1 1 1 1
1— Luxcac£
LU CXL LU trLU
f~~*
S
I
CD
cn
OO
3
LU
Q-
OO
1—
LU
1
U_
u_
LU
Q:
LU
i — i
u-
rH
rH
CO
CM
OO^f -1I-.
00 CMO
V
OCOCDCDLO
VOLOLOCOLO
CM CM
V
OVOOOO
CM CO
V V V
ooooo
CM CM
LU
S
CD
— 1
o
UJ
LU
| | , [
rcogujo
^^
s
1
^
oo
cn
LU
0-
oo
1—
LU
3
u_
u_
LU
C£.
LU
LL.
i — 1
rH
rH
CO
CM
LO CM CO r-~ CTl CO
CO to CO *!f !"•*• LO
CO i-H to Cft CVJ Cft
t — 1 CM LO « — ir^ ^"
LO i — I to CMOOCft
CO
CD OO CD CD CD
CD CD CD CD CD CD
OOCDCDCDCD
OOOOOO
CD CD CD CD CD O
CO CO -3- CM LO 1^-.
tO LO •"=)• LO LO M-
O CD CD CD CD CD
OOOOOO
CD CD CD CD CD CD
CD CD CD CD CD CD
OOOOOO
t^^ rH CO LO ^ LO
000000
CD CD CD CD CD CD
OOOOOO
CD CD CD CD CD CD
OOOOOO
UD t<- ^- LO LO •^~
O CD CD CD CD CD
OOOOOO
OOOOOO
OOOOOO
CD CD CD CD CD CD
tO tO -^f LO tO CO
S
o:
0
— i
5
CO LU
CM •Z.CXL
rH LUO
Q-UJXh- C£--
rDQiLULULUCS
OO
g
E
s
1 — 1
CO
UJ
s
=3
_l
OO
— 1
CD
CM
rH
rH
CO
CM
LOtOCM'S-tOCM
r^^c\j c\j ^r cvi o^
OOOCMLDCTii— 1
•-HLncOt— IO1LO
t-H
OCJC3CDOC3
C3 O CD CD CD C3
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CM ^d- CO CT* CM CD
CT) CO <£> CO h*. CO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
^t- co co r^ r^. CD
cnr^-uo co^oco
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
<-HLOCOCD^OVQ
cncor^cn^or^
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
O"»OO LO CO CO CO
LU
S
— i.
oz,_-
CO LU
CM 'Z.BZ.
rH LlJCD
Q-LUXI— Qi — .
ZDQ1LLJUJLUO
Q LL. 3Z SI Q- OO
OO
|_
-J
X
o
oo
LU
§
OO
_J
o
CM
rH
rH
CO
CM
rH^rH^t-CO
rH -«^- rH Cft LO
CMrH
v
toooooo
rHrH rHrH
V V
O CD CD CD O O
LOC-.I — CO COLO
•^- CM CMrH
V V V V
CD 0 CD 0 0 CD
LOLOLOrHCOLO
rHCM
V V
cnooooo
tQ LOCO CO "3" LO
rHrHrH
S
oi
CD
5
CO LLJ
CM z:c^
—~i QiUJ UJ LlJ CD
<
Q.
1 1
1
1 1
t—
LU
^3
U_
U_
LU
S
g
UJ
oo
CO
rH
rH
CO
CM
1 —
Q-
CD
CD
UJ
00
rH
CM
CD
CM
CD
CM'
CD
CM
u_
s
cn
LU
1 — 1
1 —
LU
— 1
Li_
u_
LU
1
8
UJ
CO
CO
t— r
T-H
CO
CM
1
i
1 — )
v
3
"ro
>
ro
S
-a
co
_d
in
-4->
>
00
Ol
c:
c
s_
OJ
+J
__
i
to"
+->
s
>
"o
to
_j_j
£_
**-
XJ
QJ
CO
C
ro
+->
o
£Z
tf
i
i/j
^*;
£Z
n3
S
s_
!_
O
O
c
£
nj
ffi
3
Ji^
in
cn
LU
CL,
OO
00
S
<*5
2?
Oi
£
•Q
Cl
•K*
&
00
-------
CO
• **—
CO O
_§*"*
8 01
r
s
_S3
..sg
CM
lg-»
*J e/Jr—
CTl/)
LOLOCTtLO f~^ >"^ ^~^ f~* *~^ <~^
CD CD CD CD CD CD
000000
CM CTl CTl CO f~- CM
LO CO •— 1 VO UD ^f
CDOOOOO
CD CD CD CD CD CD
CD CD CD O O <±>
CD CD CD CO CO CD
V V V V V V
ipp e^t f^t QQ g^> VO^D^D^D^D g~^ t^) g^ e^ <^> «~^>
M3 CD CD CD CD CD
LO *3" «— 1 CO CO CM
rHCMlOCOCO
V V V V V
r-^CDCDCOCD* COCDCDCO^ CDCDCDCDCDCD
lOLOLOCTkLO VDLOLO(^»'^ CDCDCDCDCDCD
CD CD CD CD CD
CDCD CDCDCD
1^1^ LOLO^J-
COCnCDCMOMO >X>CD -«=J-
CO LO t^. LO CM CO r-HCO CD
COCMVDCDOOCO rH
LO LO OO CD CO I"*-
CD r^. co co co CM
t-H
V V
CDCDCDCDOO COCDCDCZ3CM
CDCDCDCDCDCD LO^fCOLOLO
CDCDCDCDCDCD i — It — 1
CDCDCDCDCDCD
CD CO i— 1 1--. CO LO
COLOCOCOCVJM-
C\J
V V V V
OOOOCDO 00001-^
CDCDCDCDCOCD LOLOLOLOLO
CDCDCDCDCDCD
OOOOCDCD
o co r~- cri CM -^f
CO LO CM CO t— 1 CP»
r—t
V V V
0 CDC5OC3 C3OC3C3C3
O OOOO LOr~-t— ILOLO
CD CDCDCDCD CMCM
CD CDOOO
r-~ COIOCOCM
t-H
V V V
CDCDCDCDCDCD CDCMCDCDCD
CDC±30CDCDCD .—1
CO
IO
CD
VVVVV VVV V
OOOOO OOOLOO
LOLOLOLOLO LOLOLOLOLO
VVVVV VVV V
OOOOO OOOCTlCD
LOLOLOLOLO LOLOLOLOLO
VVVVV VVVVV
OOOOO OOCDCDCD
LOLOLOLOLO LOLOLOLOLO
i
I
&
CO
v
UJ
>
n.
CO
I
I
to
HH
O
LU
d_
to
1
1
a>
1
to
to
LU
8
CO
CXJ
"S
_
o
O
CO
s
Q-
CO
s
£
03
f/J
a
"to
t
**»
ci
-------
CO
OILO
i-H CM
• ro
Tin
&
rS
2^
^
CHI —
LULUi— i
OO3C
«a*i i ii 1 1
o_a_
1
I-H:^
°s
<
a
Q
00
^
OO
CO
5*0 LU
co ra
•r-i — CU
+-> O-O
ra E "O
t- ra ZS
o 3-a co
O CUr^ LU
<_5 ZS O CXL
CU ra
CO C
EJ"1" 13
-^^
MM CM
i — Q.
•r- LU
O rv^
i-H
Q.
1 1 i
a;
fe
LU
t
°
LU
1 — 1
LU
1—
5
L§
00
LU
i
=tt=
O
00
_J
1 — 1
U_
LOlOlOrHlQ
.-.OOCM^I-
V
COCOCOOOO
1 — LOLOi— ICTILO
i-H
V V V
OOOOOO
OICOLOCTlLOLO
V V V
OOOOOO
i-H i-H
V V
OOOOOO
r^ LO co co co co
i-H
LU
O.
1
^_l
-I-
CO LU
pH nj^j0,
•z.'3i'Z'>-'3^c3
Z3 O£ LU LU LU 0
a LL. zc s: Q- co
5
Of
1 — 1
Of
— '
1—
U_
U.
LU
g
o_
UJ
OO
CD
jrtj
CO
g
^?
! —
1
O
CO
CM
(H
0
CO
CO
CO
CO
CM
CO
•z.
3
u_
<
s
^
u.
1 — 1
1—
1
LL_
LU
oi
g
Q_
LU
OO
O
CSJ
CO
CM
CD CMi— *
t-H C30
V V
ooooco
C3LDLOI — LO
t-H t-H
V V
CMOOOLO
i-H
V V
oooor*-.
i-H i-H
LU
o
§
_J
tJ
LU
•21
LU
— 1 LU
1 1 1 rv LU i—LU
s
1
00
1 — 1
CD
LU
Q_
OO
1—
1
_J
u_
Ll_
UJ
CtL
g
Q-
LU
OO
o
CM
CO
CM
•5J-0
t-Hi-
00
COO
i-H
OC1
r--c
i-H
CDC
OM-
i-H
LU
1
__!
-T-
LU
g
>-;
II
s
I
oo
i — i
LU
a.
oo
I—
LU
U_
Lc.
LU
P
Q.
LU
OO
O
CM
CO
CM
0<*C\l«!t-
HOOO
OCOC3CZ3
O LOCO CO
CM •— 1
SJLOOO
n LOCO CM
CSlt-H
V
OC3C3CZ3
~ LOCO CO
CM t-H
LU
<* "*
i-H i-H
OOO
CO CO CO
V
oo o
[--COLO
V
000
cocor-.
LU
CZXCCM
LL.ZCCO
s
1 — 1
3
8
UJ
Q_
oo
1—
1
u_
u_
LU
g
Q-
LU
OO
O
CM
CO
CM
MVDCJMTil^cM
CM C3 i-H *d- 'd- C3
C3C3C3CZ>CZ)C3
r-H t-H i-H CO CM i-H
OOOOOO
co co r^. oo o** I-H
i-H i-H t-H CO t-H r-H
0 O O O O O
i-H i-H CM CM CM t— 1
OOOOOO
CO ^ t— 1 1-- OO t-H
•-H i-H CM CO CM i— 1
LU
i
=^j
«->
CO LU
CM *ZC£.
r-H LUO
h- LU_J.IJ
Q-LUXI — Qi-^.
Z3 OZ. LU LU LU O
cziLL. zn s: a. oo
5
cr
5
Of
i — i
— i
t_
LU
ZD
U.
u_
UJ
^
1 — t
o;
Q_
i-H
CM
CO
CM
a.
o
e
s
oo
„_,.
CM
1^
•=)•
C=>
CO
CO
0
S3
CD
C3
;"'
O
c!
s
s
«|
Lc.
•z.
\ — 1
fe
LU
_J
U_
Lc.
LU
=2
s
oi
a.
i-H
CM
CO
CM
t^Lf>Or-H«*
00^-^. <=>
C3CZ>CZ)CZ)C3
•* r-H CM LO CM
OOC3C3C3
COCSCOCOi-H
TI— ICMLOCM
C30OC3C3
^-i-Hr^(-.C=>
•^- i-H T-H CO CM
a
o:
g
-T-
o
UJ
LU
_l LU
lisiS,
s
en J
CTl
(/)
1 — 1
Q
LU
O_
CO
fe '•
LU
U_
u_
LU
^
'1
cc:
i-H
CM
CO
CM
COCOCTlCO'3-
->-"=> -"=>
OC3C3C3C3
LOr-Ht— 1COCM
OC3OC3CZ5
COCOCTVLOLO
LOt— It— ICO CM
OOOC3C3
§ r-H COCO'*
i— 1 i-H CO CM
LU
O
1— H
g
5:
C_3
I , |
1 UJ
IfeiS
s
{
00
1 — 1
0
UJ
a.
00
s
u_
LL.
LU
^
1 1
0_
T-H
CSl
OO
CM
to
*•»
*
I
"o
CO
!_
cu
o
V)
•a
CL)
N
o
cu
8
§
LU
O.
OO
i
•S
&
<0
-------
CM CD co o*» I^N. r-*.
«3- LO i— 1 CO UD IO
t— i r-~ co r~- CD
CO
. C3
cMr--cric- .
LOCOCDLOCOr
r— vor-*c ~
CO t-H
CM LO LO •* r-. CM
CM GO en co <£> cr>
CD CM CM LO LO r--
CM CM UD LO Cn CD
CO CM I"-. CM
CM
CM
t-HCO CDLOCnCMCO
LO CM CMCMI--
5
CDLOCD
CMt-~CD
CDCDCDCDCDCD CDCDCDCDCDCD CDCDCDCDCDCD
*d- Cvl t-H T— 1 1^. CO
•^- co co -^-«^- co
C3C3C3
CIS OC±3
voi— icar^i — ^co
CO ^f •^ LO CO -"y-
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
LO CTi CD CD CD CD
CM CM cn co r-. co
i-HCMCOCM
CD CD CD CD CD CD
CD' CD CD CD CD CD
CD CD CD CD CD CD
co r^ t— i CM cr» r^
co to ir> -«^- co co
CD CD CD CD CD CD
OOCDOO
CD r-- ^ -<=t- «— i
(*^- r*-^. r^-. co CTI
CD CD CD CD CD
CD CD CD CD CD
•=ccoco'<=i-<3-co co ^r co 10 co •<*
CDCDCDCDCDCD
CO t-H CD CD CD CD
CM CO CD t-H "3- t-H
CM CM ^-CM
CD CD CD CD CD CD
coLOcocor-.cn
COVOUD^-COCM
V
CO
CU
CD CD CD CDCDCDCDCDCD CDCDCDCDCDCD CDCDCDCDCDCD CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCD CDCDCDCDCDCD
CD t-H CT! I-. CTl CM
"5t- •«*• CO LO CO ^3-
CDCDCDCDCDCD
CD CD CD CD CD C~
CDCDCDCDCDt
CD CD CD CD CDC
^J" CO CD CD CPi ^
co 10 to ^- cn -
3CD _ _ ._
3LO I—t-HLOCDt-H
Jen o*>CT>CO CD cn
CDCDCD
cor-^co
CDCDCDCDCDCD CDCDCDCDCDCD CDCDCDCDCDCD CDCDCDCDCDCD
CDCDCDCDCDCD CDCDCDCDCDCD
CDCDCDCDCDCD CDCDCDCDCDCD
CDCDCDCDCDCD CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
•«3- t-H CD CD CD CD
CM CO CM CM COLO
CM CM CO CM
CD CD CD CD CD O
"=f CO CO CM CM CD
co r-. LO ^r oo »^-
r^r^ rococo
%
LU
X
X
o
to
o
o
cn
to
LJ_ LU
Z. O.
t-H tO
to
I—t
Q
1
.0
s
to
s
LU
I
to
LU
CM
CM
t-H
CO
CM
CO
CM
t-H
CO
CM
"8
I
£
fO
c3
K
<0
s
-------
CO
01LO
t-HCM
P CD
CD cn
+j ro
<~*^
cu
oo
f
on
CO
01
• -S §
to ra
e: to co
0 CO
•i— . — CO
+-> O.CT
(O E "O
t- ra 3
-L-> COi —
< — CO
co to-—
CJ 3-O
C Ov-
O COi —
CJ 3 O
crco
co n3
co c:
to Cv-
CU •!—
S- O
C£J 1 w
<*g,g
c5
.__;
LL.
CD
Q
OO
Z
5:
CO
CU
LU
ct:
LU
o;
Q_
LU
Di
I —
LU
_J
O
LU
O
-5*
HI
CJ
| —
CTv t— 1 CD CO CD r-~
r- 1 CO CM CM CM CO
1 — CM "3" CD COLO
CO •— 1 CM CM CO M-
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CO CD 1— , LO CM
CM t— 1 «— 1 t— 1 CO t— 1
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CM CD CD CD
^" CT) cn f--. co r—
CMrH t-HCO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD LO CD CD CD
i— 1 CO CO LO Cn ^~
CMt— 1 »— iCOt-H
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
i~- cn co cn CM M-
t — It — 1 1 — 1 1 — I CO* — 1
LU
OL
0
— 1
5
CO LU
CM 'Z.CZ
t— 1 LUO
1— LU_J_J
oo^niocM
CL.LUXI — C£^,
ZDOiLULULUCD
Q U_ 31 S: Q- CO
s
0
| —
s
pg»
00
^.O CQd? t— H 1^^. O"t
U1 UT (Jl C*"J r— i r—t
t— 1 *3- ^3" LO O^ OJ
OLO^OOCNJLO
1-*- C3% CM CvJ C3 OO
^=r CMCMOO
OO>C3C3OO
CD O CD CD O CD
CD CD CD CD CD C3
CD CD CD CD CD CD
CDOCDCDCDCD
OO f— 1 CO LO ^D CO
LO ^O LO LO LO LO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
OO h-- CO CO CO LO
LO LOLOLOLOLO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
r--- CD co *=a- LO co
LO <*O LO LO "3- *3~
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD Q CD CD CD CD
CD CD CD CD CD CD
O"i r--. co co f— I CD
r— 1 r— ) T— 1 CO LO t— *
i— 1
LU
CD
CJ
CO UJ
CM -Z.CH
T— 1 ' UJO
oo
i — i
CD
LU
Q-
v°
CM LOr-LO
O t— ICOCD
v
CD CD CD CD CD
COLO rococo
CO CMCOt-<
' V
CD CD CD CD CD
co in co co co
CO CMCOi— 1
V
CD CD CD CD CD
cnincor-ico
CO CMCOt— 1
s
_J
UJ
z.
LU
—1 UJ
ino^Lucj
§ i
c/>
o
UJ
CO
•4-cn^i
rH^-CM
CD CD CD
COt^CO
CDCDCP
t-Ht— 1
CD CD CD
CM i— 1
LU
§
»_
*~~*
Tj
8
LU
Q_
OO
CM •— * CO r-l CO 1— .
t— ICO CO CD CO CO
CMCOt— It— ITJ-
CD CD CD CD CD CD
•a- CD co co CD CD
'S- co cn LO csi n-
t— 1 CM t*H
CD CD CD CD CD CD
Cn CD t— 1 CM CD CD
co co co T*- LO in
i — 1 CM t — 1
CD CD CD CD CD CD
CD CD CD CM CD CD
CO CM i— ICO«3-LO
t~-t t— 1 CN] t-H
CD CD CD CD CD CD
t— H r~l t^
C£J
Q
G£
CD
o
CO I it ]
CM Z. C£
*— 1 '.UJO
13 Q; LU LU LU CD
CD u. :n a: a. co
^^
Of
O1
I-H
_l
i
CD
—I
OO
(O
IJEi
cn
CM
•g
di
L.
-------
i-H COLO CM CO CO LOOCOOCOCM «3" Cn LO ^)-CM t-H
t-H l-~ LO GO t—I
CM r-l-^COCOCO
»— i r— ICMi— 1
CTMJO CTl CO t— *
CO CO LO 1^^ CO
O^ ^" COO^ CXJ C
CO f^ CO CO CO L
C\J CO *«OCO C
CDOOCDCD
COIDCOOIP
CDCDCDCDCD
CDr-Io'oco
CO CM^T
CO CM LOCO CO"* t-H
CM t-H
OOOOOO OOOOOO
CDCDCDCDCD
CD CD CD CD CD CD
8 CD CD CD CD CD
CD CD CD CD CD
CD CD CD CD CD CD
COCDCTiCTtt— 1VO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
LO cn -^- CD a* CM
r^. co co co cr» co
CD CD CD CD CD CD
OOOOOO CD CDCDCDCDCD
co csi oo H3 cn cn co cor^i^cM*£>
CO CM t-H CO CO t-H CM COCOCM^CO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CO l_n CD Crt t-H t— i
LD -CO'=J-CT»
co
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
COCOt— ICMt-Ht-l
«=r co »— ico ^ co
CD CD
CD CD CD CD CD
(Jl CVJ «=J* CD CD VO
[^ CT) CO GO CD GO
OOOOCDO CD CDCDCDCDCD
cncMcouDcncn co voi^ocovo
COCMt—ICOCOt—I CM COCOCO"=i-CO
1
V-
£
•Q
&
CO
OOOOOO OOCDOCDO (OOCDCDCDCD
CDCDCDCDCD
OO
LO
CD CD CD CD CD
CD CD CD CD CD
CD CD ^DCD CD CD
cn co LO LO CD <*o
CM CO CM CO ^* t— 1
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
cnr — LOCDCD^O
r^ co GO co CD GO
CD CD CD CD
CD CM CO ••If
«S" CM t— 1 C
CD CD CDCDCDCDCD
cn co cr>t-~LOCMVO
CM COCOCM'^fCO
1
in
a>
"o
o
I-H
§
- s-
••3-
S3
"8
tl
00
»-,
CQ
-------
O)
CO
.—ICM
-1-
CO O
E CU
CD Ol
4-> ro
01
OO
£
&
^
LULUi— c
co si
§11
<=3LU
^
1— IZ
°s
g
>
oo
t/
to cc
c: to to
o cu
•r-i — CD
-t-) 0.0
ro E-g
£- CO 13
4-> OOi —
f— (/;
^-l
ooooo
OOOOO
LO VO f*-. O 1--.
*=i- co co LO ^J-
CDOOOO
OOOOO
Or-ir--.!— 1 LO
UJ
0
3
5
UJ
UJ
1 LU
:co
LUC£LUI — LU
s:u_:n£a-
<
g
l-~-
vf
c^n
3
LU
Cu
oo :
LU
:D
1
u_
U-
LU
>~
§
ce:
Cu
r-1
vo
-=t
co
CM
t— ICOVO
sss
CD CD CD
CO CD CO
CM CM CO
CD CD CD
CD CD CD
CM^HCO
CD CD CD
CD CD CD
OlCMCvJ
CM CM CO
LU
•Z.-Z.O
<
1-1
1Z.
•si
f>
8
LU
a.
00
t—
^L
LU
ZD
_l
U-
U_
LU
>-
o;
^
i — i
£
t— l
VQ
CvJ
OOOMp-^-^LO
f^. LO CT* *£> O ^J-
*.or--^rco to* — i
t — i i — i
000000
oooo o o
^O^CDOCS
I--.OOLOCOCJYC\I
COCOCO-^-LO.CO
CDCDOOC3CD
OOC3OCDO
OOOC3CDO
OOOOOO
CM -^3- OO VO -^ O^
^" OO CO *?}" ^" CO
OOOCZ3OCD
CDOCDCDC3G3
LOCO<£>aOCMO
CM co co ^ r*-.jco
C3C3CDOCDCD
C3CDOC3 OO
OOOOOO
CD O O O CD CD
CO CM rH i— 1 »— \ r^.
"=t- -^ CO LO ^O CM
i i i
Q
1
t_>
CO LU
CM 'Z.CC
El! LU^0
s::z:z>-3:cD
oo^^rocM
— > Q; UJLULU CD
OO
5
1 — 1
1
z:
to
oo
LU
^
CJ
Qi
LU
[_
_J
i — i
Ll_
CM
<^>
CO
CM
OOCSJrHr^CMOO
o ^o LO r*-. cru— i
co r-. r-i CM co cj*
<^> O1^ «^3 O^ CM r-
t — 1 CM) CTl » — 1
I— 1
OOOOOO
OOOOOO
OOOOOO
oo oooo
CO r-H ^O OO LO r^
LO vo vo r^. i^. vo
OOOOOO
OOOOOO
oooooo
OOOOOO
CD CD VO ^J- OO CTl
vo vo vo r^. vo vo
CDOOC3CJO
CD CD CD CD CD CD
CD CD CD CD CD CD
oooooo
i— 1 CM ^J- CM LO LO
CD CD O CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
OCDOOCDO
•^vovor^r^vo
I.I.I
Q
t— 1
3
5
CO LU
CM za£
i=! Lu!i!3
^2r^>-aio
^Q^LULULU CD
c^ u. m s: o_ oo
CO
1 —
LU
1
~T~
g
oo
LU
^
s
rt^
1—
_1
1 — 1
U-
s
CO
CM
OCM t-(
r-HCD O>
V V V V
VOr-HCDCDCDCD
LOLOLOLOLOLO
V V V V V V
CD CD CD CD CD CD
LOLOLOLOLOLO
V V V V V
t~- CD CD CD CD CD
V V V V
CD ^- CD CD i— 1 CD
LOLOLOLOLOLO
O
t— i
|
O
CO LU
CM ZQi
i— 1 LUO
ZZ;Z>-:I:CD
OO<:HOCM
Q-ujxPct:^-.
ZDQ^ LU LU 1 1 lt~>
§
1 1
1
o-
1 —
^
LU
=3
I
1 1
Ll_
LU
*
vo
CM
§5
S
C\!
c=>
t— 1
CO
cr>
CM
C\J
CO
_J
U-
5
a
LU
u_
•z.
1— 1
LU
ra
j
rr
1 1
UJ
u_
g
VO
CM
vOOMOVQ^
C3 CT>C3C3iC3
•^r^Loor^
t— H
OCOt-HOO"*
o^ r^* ^« f*> vo
t-H
OOOOOCD^
CT^ VO 1^ t-H VO
\ — 1
o
1 — 1
3
5
LU
LU
—J I ' 1
>-^^
n:o
i^Luxpaci
1 1 1 Q; 1 1 1 pz 1 1 1
s:u-3rso-
<
=
o%
^
oo
1=1
LU
Cu
oo
s
t
LU
*
VO
CM
vrwoocoa1!
C\/t — it — fU7CV
CDO^i— ICDCD
t— 1 I— It— 1
CDCr>^t-CDCD
CM r- It— 1
CDCttOOCDCD
(--ivo'vdcoco
t-H CM-— 1
' ' '
CU
1 — 1
a;
5
LU
LU
— 1 LU
>-;z:z
niO
-------
CO
23 c8
,
coo
I8
S en
rra
s
2-fe
%
^j
S
8
t/>
^"Sf—
-S|
C: tO tO
O Oi
ill
lii
ai-8
01 ro
in c
£ . >l"~
£7
CO— J-*l
.agg
o
1
o
i/5
^?
£
CO
*-»
LU
os
g
s
I— 1
LU
o;
g
1
LiJ
g
|
LtJ
g
S
fe
Gj
®
o
g
i
cMf-.
•3-0
V
OOLO
LOCO IO
V V
CD CD CD
LOLor-
V
CD CD CD
LOt— 11O
t— 1
ZS3ZC3
O f"> «^> ^3
CD CD CD CD CD CD
t-H CM LO CO CO CT*
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CM LO *=r CM cn cn
UD UD LD UD LO Ln
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
LOCO LOCO CM CO
CD CD CD CD CD CD
CD CD CD CD CIS CD
CD CD CD CD CD CD
COCO§LOLOCO
LOLOLOCO LOLO
LU
S
3
re
CO
CO LU
CM ^OZ
I^LU^jg^
l/J
*^"'
B
§
^
oo
UJ
g
3
oo
S
5S
CO
CM
CTii ^" CTt < — 1 VO I*1^
^•VOCDCOCOOt
CD VO tO <— 1 CO CT^
LOOCOCO^CD
« — 1^" ^" CvJ t — 1
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
OOOOOO
r-i CD f-* tO CO LO
tO VO LO tO [-- *&
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD i-H O CNJ r^- LO
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
i— icriCMco-^-r-.
to LO LO to r-- to
CD CD CD CD CD CD
§ CD CD CD CD CD
CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CM CD CO LO CD -^
tO tOLO tO I*-- to
UJ
o
1
g
CO LU
CM •Z.O^
t— \ LUO
fezllH5S
0- LU X, \— Cd ~~~
ID CC LU LU LU CD
cn LJ_ rc s: o_ co
^^
oo
tj
^£
o
oo
LU
1
00
1 1
g
ES
s
O-ICDCMLOCDCO
!— 1 !— < LO !— 1 CD CM
!— I
CD CD CD CD CD CD
•<=f CO CM CO CD 1^-
CM CM CM CM CM i— 1
OOCDOCDCD
LO CO CT» i-H CO »X)
CM CM CM CM ^H -—I
CDCDCDCDOO
LO r— i— 1 ^" CM LO
CM CM CM CM CO t-H
CD CD CD CD CD CD
COCTlLOCMVOi— 1
CM CM t— < CM i-< i— 1
LU
O1
|
CJ
CO LU
CM -Z.CZ
•-1 LUO
I— LU_I— 1
S=:z>-rco
ooj
1 — 1
O
CD
CO
CM
CM
CD
LO
CM
CD
CM
CD
Si
1
^
^-'
1
Ll_
1— 1
1—
LU
_J
1 1
Lu
LU
OS
P
§
LU
OO
CO
LO
CM
COt-HLOl — LO
r-HCDrH LOCO
CD r-H «3- CD CD
LOO-lCTlLO^H
rH CM^H
CD i— ICO CD CO
•=COMX>CMOO
«— t CM
CD CD CD CD CD
r*^. LTlt — 1 CD CO
LU
cn
S
g
LU
g
_J LU
1— LUXQ3QS
LU OSLU h — 1 1 1
^_^
^
2
OO
1 — 1
cn
LU
oo
jr;
LU
1
1 i
U_
LU
g
§
o_
LU
oo
CO
LO
CM
cnr--oocr!CM
CO^HCM^l-*
CD CD CD CD CD
CM LO CD LOCO
^J- CO "^" t — 1^"
CD CD CD CD CD
•^•LOCMLOCD
'q-CO-*^l'^-
CD CD CD CD CD
CTl-3- CO COLO
CO CO CO « — 1 "^~
LU
Q
OS
3
re
o
LU
_J LU
rco
•»£
O)
>
O
10
01
c
E
f'1"
t-
o>
^
"i
to
1
"o
tn
J
t—
•a
OJ
>fcj
s
ro
•g
c:
S
S
10
JX!
C
(t
5
aj
u
c
CO
"S
4->
<;
S
c
1
c:
n
S
|
a
CO
a
(T
>
b
•r—
CS
£
-------
CO
en
cntn
T-HCM
-
u_
a
i_
17
^
CO
^ cu ;z
* * *— ' E £
co ra
c: co co
0 CU
•1-1 — CU
4-J CXC
t- TO =
•*-> COi —
£Z V
CO co-<
O Z3"O CO
crco
cu ro
CO C
CU-r- "~
t- o
Cfl_l-^
e3 Cn O
,-EE g
O CC
1 — 1
1—
SL
j
O
OO
LU
g
ZIZ
o
LU
LU
1 —
OO
fe
<-5l
0
1—
—I
g
UD CD CO CO t-H LO
CD i— t CO CM t— t CTl
LO CO CT^ r-*-. T-H LO
^^CDf-CMLO
r--. co r— r CD CD en
i— Ir-ICM
CDCDCDCDCDCD
CD CD CD CD CD CD
CDCDCDCDCDCD
CDCDCDCDCDCD
i— 1 i-H CO LO tQ to
LO LO CM LO *^- LO
CDCDCDCDCDCD
CD CD CD CD CD CD
CD CD CD CD CD CD
CDCDCDCDCDCD
i— t OO •>- CD CO CO
LO^-CO^OCMLO
CDCDCDCDCDCD
CD CD CD CD CD CD
CDCDCDCDCDCD
CDCDCDCDCDCD
CO CM CM CO CM *O
LO LO CO <^> <^> "3-
CD CD CD CD CD CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CDCDCDCDCDCD
CO CD ^O CO CM LO
UJ
Q '
§
o
CO LU
CM :sa;
r-H LU CD
1— LU l_l
Q-LUXI— Q;^
ZDOi I i 1 LU LUCD
OO
^
1 — 1
LU
CJ3
— >
OO
_J
0
o-i
CO
CM
COlQCTiCOLO-*
cn ^t- CM ^- U3 en
LO r-H T-H T-H OO CTJ
T-H CM r~- -^- OO CD
CD COO T-H r-. CM
T-H T-H
CDCDCDCDCDCD
CDCDCDCDCDCD
CD CD CD CD CD CD
CD O CD CD CD O
i — 1
CD CD CD CD O CD
CDCDCDCDCDCD
CD CD CD CD CD CD
CD CD CD CD CD CD
r-H r-H
CD CD CD CD O CD
CD CD CD CD CD CD
CD CD CD CD CD CD
CDCDCDCDCDCD
CnLO CD CTJ i — 1 1£>
en t^ vo r^ en co
CD CD CD CD CD CD
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
T-H kO t-^ 1^ LO CM
LU
J
CO LU
T-H LUO
Q-LUXI— C£ —
ZZ) Qi LU LU LU CD
czi u_ zc :s o- co
^^
^
5
^
X
CO
LU
i
CO
1
1 — 1
o
cr>
CO
CM
r^-Crt LOO^ CM CO
to cOLOCJ^a^o^
CM r^-t— IrHCM
CDCDCDCDCDCD
CDCDCDCDCDCD
r-- LO CD i^. OMO
CO CO -^T CO CO CO
CDCDCDCDCDCD
CDCDCDCDCDCD
CD LO r^. LO CM CM
*^r co ^" co *^t" co
CDCDCDCDCDCD
CDCDCDCDCDCD
CO CO -3- CO CO CO
CD CD CD CD CD CD
CD CD CD CD CD CD
LO to CM i--*. co r--.
CO CO CO CO CO CO
UJ
Q
CD
I_LJ
5
CO UJ
s:s:s>-=n:cD
Q_LU>LOI — ^cno
T-H CO CM ^t- CM
CM T-H T-H
00000
COT— ICO -^- CM
CD CD CD CD CD
CD CD CD CD CD
i-HCDLO«*T— 1
CD CD CD CD CD
CD CD CD CD CD
LO T— 1 1--, M- CO
LU
3
zn
o
LU
LU
_J LU
~r~ O*3I! LU C_>
1 — Luxoaci;
g
^
^
en
oo
Q
LU
O-
oo
1—
LU
Lu.
i — i
s
CD
CO
CM
^t- co »>-. i— i ^t-
r-HCO CO
OCDCDCDCD
CD CD CD CD CD
|-v. *;o i — 1 CO t,O
CD CD CD CD CD
CD CD CD CD CD
CDCDCDC.DCD
CD CD CD CD CD
COCDt— It-HtO
^" CO ^f LO_^"
LU
i — )
§
zc
1 1 1
_J LU
1 — LUXCQQ;
LUC£LuFr.LU
S:LC.ZCS;Q.
3
^
^
•*
oo
LU
Q_
OO
,_
LU
U_
i — i
§
CO
CM
CMr-~C3^)-^Hri-
CO CD C^ CO t^1^ OO
i-Hi-H r-H
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
»^3 VO LO CO tO tO
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
to r*^^j* co F*-. to
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
r— i O1!) LO CO **d- 'i— 1
tO tO <«=J- CM t£) LO
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
CDCDCDCDCDCD
SCTi r--» co CM CD
Loto-^t- tor--
C3
i — i
§
*
zc
CJ
CO LI I
CM SE o;
OO
'o
15
•4-J
Ol
"S
.5
o
ro
S
I
CO
fe
&
CO
-------
CO
§3
• <*-
CO O
fc8
"§ Q> CO
33 en JCx
4-* If]
f2 8}
LO ^ril
V^-'B
O QJ
t—i —
O^OC3C3C3
CT* "^ CO CXJ ^ ^
COOt CO O"l ^3 C3
t— ti-i
CDCDOCDC3CD
**^> <^*> «^3 <^D t^P^P
O>COO^rHC-4t-H
t-Hi— If— 1
«
OCDOOOCP
§(±5<3C3C3CD
^3 ^3 C3 «— »»— J
CVJ *-5r «3 VO ^ ^
o%o>cocnc\ji-H
LU
Q
g
DC
t_3
CO LU
CM ZCe
i-H LUO
H- LU_1 1
SIZgS-DCCD
feccSCLUCD
au-zcsia-co
^
C/J
i
X
t/>
UJ
oo
>-
t— t
o
T-l
s
en encore CM r~-
CD CO CD CD'S- CD
CD CD CM CD CD 03
COt— 1 COCM
CDCDLOCDCDLO
LO *JD LO 'S* ^* f"^
COt— 1 COCM
V
CD CD CD CD CD t~^
COr-H COCM
V
CD CD CD CD CD CM
"d-OILOLOCniQ
CO COr— 1
LU
0
S
g
s
CO UJ
•^ LU_J^_J
SSujELUO
du_rc^Q-co
^
^
^-1*
o*
K-4
^^r
i — i
J£
LU
UJ
_J _I
LL- LL.
Ill |_i_|
Irf rf
0 0
CM CM
r^ i^.
CO CO
CM CM
CMCDLOOMO
'S" t — ICOCO^-
CD CD CD CD CD
CDLOrHOOLO
T— I
CDCDCDCDCD
CD'S- CO •— ICO
t-l
CDCDCDCDCD
CD LO COLO T-l
i-HCOCOlOLO
LU
Q
§
5
LU
•z.
LU
1— ujxcaa?
1 1 i rv* 1 1 i [— _ 1 1 i
s:u_:c CL.
-
,1
^j-
co
s
LU
CO
t;
_i
U-
U-
LU
^
C3-
CM
CM
•"S-'S'T— 1 CO OO LO VO CD T— 1
T-HT— ICM en co T— i CD en co
T-HCOCMCMLO
CDCDCD CDCDCDCDCDCD
CM U3CDCOCMCOLO
CDCDCDCDCDCD
CDCDCDCDCDCD
. LO VO LO CD VO LO
"S1 CO CO *«3" 'S' CO
CDCDCD CDCDCDCDCDCD
cnr-.r-. CDCDCDCDCDCD
V
CDCDCD CDCDCDCDCDCD
r*-» LO r~^ **~^ *""** f~^ **~^ *-~* r~^
LU
S
1
CO LU
CM 'Z.Ct.
T-H LUO
LU I — LU 1 1
CiLUCD =D;LuELUCD
u-rcoo ou-x: o-oo
-
1— 1
g <
3 2
1 1 1 QT
p\ 1 — 1
OO _l
fe fe
LU LU
=3 C3
_J _1
LU 1— 1
<£ "g
Q Q
CM CO
CM CM
^
LO
CD
CD
LO
CD
CD
"S-
T-H
CD
g
s
LL.
5
\^s
i
u_
1 — 1
fe
LU
=>
2:
i— 1
^
Q
CO
CM
•
T— 1
V
CO
a>
=3
"ro
^
c:
ro
S
•a
ro
j^
to
>
"o
to
I
E
*co
E
S!
T3
C
'o
to
4->
£_
O
,
ro
ro
•|
Q)
t-
%
in
±6
c:
ro
S
01
|
!/)
•a
a
5^.
o
u
±£
re
^
C
£
ra
g
i
^^
-------
APPENDIX C
DISCUSSION OF STATISTICAL TECHNIQUES
USED IN THE PRELIMINARY DATA EVALUATION
-------
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Data Transformations
The standard deviations of the measured sample concentrations varied greatly and were
roughly proportional to the mean sample concentrations, as is common with analytical data. This
heteroscedasticity violates the basic ANOVA assumptions, and was taken into account in the statistical
analysis.
The association between the standard deviations of sample concentrations and the mean
sample concentrations was easily demonstrated in this data set by plotting the replicate standard
deviations versus the mean of the replicate analyses for each sample/solvent combination, as is shown
in Attachment A. Linear regressions of the standard deviation of concentration versus the mean
concentration showed a very significant trend. Secondary predictors of the standard error including
solvent, industrial category, sampling point, sample kingdom, and laboratory, were checked by
including them as predictors along with the mean concentration. The only other factor that showed a
relationship with standard error is laboratory (which is described hi greater detail below.)
In order to allow the use of standard statistical tools, the concentration data were transformed
so that the data would have constant standard deviations. If the standard deviations were exactly
proportional to concentration, i.e., if there were no intercept in the regressions, then a logarithmic
transformation would have produced constant standard deviations [1]. Since, there was a positive
intercept in each of the standard deviation regressions, the data were transformed using the equation:
z=log(x+c)-Iog(c)
where c serves the dual purpose of adjusting for the non-zero intercept of the standard deviation curve
and allowing for a well-defined transformation that is robust to the negative concentration estimates
found hi this data. Examination of the data suggested a value of c=100 for the aqueous analyses and
c= 10,000 for both types of solid sample extractions. These values are on the same order of
magnitude as the negative of the x-intercept(= -intercept/slope) of each regression, and they are of
sufficient magnitude to keep even the most negative reported concentrations (-56 for aqueous, -1931
for sonication, -7556 for Soxhlet) a reasonable distance away from the pole at 0 hi the log
transformation. This distance prevents these points from gaining disproportionate leverage due to
being transformed far out to the left of the data. Following this transformation, the data no longer
showed a significant association between standard deviation and mean concentration. Therefore, the
transformation was suitable to the ANOVA analysis described below.
The results on the transformed variable were translated back to statements about concentration
by inverting the transformation, such that
x=exp(z+log(c))-c
September 1993
C-3
-------
Preffminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Results could then be expressed as approximate proportionate results, in concentration terms, if the
inverse calculations were applied at a representative level of concentration, such as the mean of the
transformed variable. For example, a limit value Zj and a mean z0 could be presented as a
proportionate limit in concentration terms by calculating
+log(c)) -exp^+logfr))
The analyses described in the remainder of this discussion should not be highly sensitive to the
particular values of c used, within the constraints described above. A sensitivity check could be done
to demonstrate this by testing alternative values of c.
Negative Concentrations
While it is counterintuitive to find negative measured concentrations, it can be expected to
occur, at least occasionally, due to analytical variability whenever very small concentrations are
measured via methods involving blank-subtraction. It is possible that some sample/solvent
combinations could be removed from this study by setting up statistical limits on whether the sample
can statistically be shown to contain a non-zero amount of oil and grease, for instance simply
computing a one-sided hypothesis (at say 5%) that the mean concentration for a sample/solvent
combination is greater than zero using the three replicate measurements (either with or without the
data transformation). However, it is desirable that the study remain balanced across solvents, and
since some solvents show concentrations clearly greater than zero while other solvents do not, such a
rule may not be useful. Since the data transformation method described above can cope with negative
measured concentrations, such samples were left in the study for the moment.
ANOVAs
The ANOVA model deemed best for this analysis was
y=a+bi+gj+du+euk
where i =!...! solvents (l=Freon), j = 1...J samples, and k=l...K (3) replicate analyses. Here a,
bj, gj, and dy are all fixed effects that were estimated5, and eijk is the random measurement error, with
mean 0 and variance a2. This model is a standard ANOVA model, and was fitted to obtain estimates
of each of these parameters and standard errors associated with each estimate.
Upon testing this model (see Attachment B), it was clear that the interaction term is
significant (i.e., the effect of solvents depends on the sample matrix).
5With £ bi= 0, £ gj=0, and £ ^=0 over i for fixed j and over j for fixed i
C-4
September 1993
-------
_ _ Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Evaluation of Solvents
The evaluation of individual sample-by-sample confidence limits on the difference between
each alternative solvent and Freon would have resulted in too many outcomes to consider and would
not have produced an overall picture of the performance of each alternative solvent. Instead, an
overall index of performance that summarizes the similarity of each alternative solvent to Freon,
while taking into account the differences in the outcomes for each solvent on different samples was
considered desirable. A natural measure for this is the root mean square deviation across samples
between the alternative solvent and the Freon results. In terms of the model above, the mean result
for each solvent for a sample is,6
«
so the root mean square deviation for solvent i is
This can also be computed as the root mean square deviation between the sample*solvent cell
means for the alternative solvent and Freon. The smaller this measure, the more closely the results
using the alternative solvent approximate the results using Freon. RMSD computes the squared
deviation of the average analytical results using alternative solvents on a sample from that of Freon on
the same sample, and accumulates these over all samples to provide an overall measure of agreement.
The data show significant interaction between solvent and sample in the statistical model, that is,
whether alternative solvents extract more or less oil and grease than Freon varies according to the
sample matrix. RMSD measures variations both above and below the Freon results, because we must
capture in our statistic the possibility that an alternative solvent extracts significantly less oil and
grease than Freon on some samples, and more on other samples. Having the same average results as
Freon across multiple samples is a desirable, but by no means sufficient, test of equivalence of the
solvents. As a rule of thumb, the RMSD that would be expected by chance alone, for instance if
Freon were tested in this protocol and compared with itself using separate analyses, can be computed.
Under the null hypothesis that there is no actual difference in the procedures, the square of RMSD,
appropriately normalized by the residual error estimate, will have an F distribution. Therefore,
K(RMSD)2
~
where s is the root mean square error (RMSE) of the model, and I, J, and K are as above.
6 For typographical simplicity, the carats over each parameter are omitted from this point on even though
the formulae refer to the sample estimates rather than the theoretical model values.
September 1993
C-5
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
A 95% acceptance region for the equality of the test is
IZMSIXj^Fr1
K.
or, hi terms of the normalized RMSD,
RMSD
Examples of these analyses are shown hi Attachment C.
Laboratory Comparisons
To test the sensitivity of the rankings to the effect of each laboratory, the above rankings were
rerun, dropping out each laboratory hi turn. These tests yielded no significant changes in the
comparability of the solvents to Freon.
Laboratory differences hi the analytical standard deviation were also explored. This was done
by an ANOVA of the within-sample*solvent standard deviation of the non-transformed concentration
versus laboratory. One laboratory showed significantly greater RSDs than the other two on soxhlet
and sonication extractions of solid samples. All labs had statistically similar RSDs on separatory
funnel extraction of aqueous samples.
Analysis by Type of Sample
Rankings were also run over subsets of the samples, including divisions by industrial source,
sampling point, kingdom, etc. to look for situations where the ranking changes or where all solvents
are markedly different from Freon. The results of these rankings as divided by industrial source
(petroleum vs. non-petroleum) are discussed hi the report of preliminary findings.
Alternative solvents which rank well overall could be examined for their deviation on
individual samples hi order to spot sample types showing larger deviations, which may indicate areas
for method improvement. Since none of the alternative solvents ranked well overall in the
preliminary analyses, no such examination of individual sample deviation was performed.
Review Reference
1. N.R. Draper and H. Smith, Jr.. Applied Regression Analysis, Second Edition. John Wiley &
Sons, New York, 1981, pp 238.
C-6
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Attachment A
September 1993
C-7
-------
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
AQUEOUS
Std(NET_AMOU) By Mean(NET_AMOU)
1400-
1300-
1200-
1100-
1000-
~ 900-
0 800-
< 700-
tn 600~
S, 500-
OT 400-
300-
200-
100-
o-
-100-
e
.
^ -
^ —
^ '
j&*f*T**"-''"
0 1000
Mean(NET_AMOU)
Fitting
Linear Fit
Linear Fit
Summary of Fit
Rsquare 0.30436
Root Mean Square Error 24.46967
Mean of Response 10.53657
Observations (or Sum Wgts) 139.2707
DF
Source
Model 1
Error 190
C Total 191
Analysis of Variance
Sum of Squares Mean Square
49775.18 49775.2
113765.33 598.8
163540.51
F Ratio
83.1297
Prot»F
0.0000
Parameter Estimates
Term Estimate Std Error t Ratio Prob>|t|
Intercept 3.9038199 2.19739 1.78 0.0772
Mean(NET_AMOU) 0.1751866 0.01921 9.12 0.0000
September 1993
C-9
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
SONICAT
Std(NET_AMOU) By Mean(NET_AMOU)
i
g"
CO
700000-
600000^-
500000-
400000-
300000-
200000-
100000-
o-
\ \ I T I I I i I
-100000 200000 400000600000 800000
Mean(NET_AMOU)
Fitting
Linear Fit
Rsquare
Linear Fit
Summary of Fit
0.353493
Root Mean Square Error 7629.233
Mean of Response 5671.682
Observations (or Sum Wgts) 40.93224
Analysis of Variance
Source DF Sum of Squares Mean Square
Model 1 4519150093 5e+9
Error 142 8265136937 58205190
C Total 143 1.28e+10
F Ratio
77.6417
Prob>F
0.0000
Parameter Estimates
Term Estimate Std Error t Ratio Prob>|t|
Intercept 1578.4888 1279.76 1.23 0.2195
Mean(NET_AMOU) 0.1625567 0.01845 8.81 0.0000
C-10
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
SOXHLET
Std(NET_AMOU) By Mean(NET_AMOU)
140000-
130000-
120000-
1 10000-
100000-
90000-
80000-
70000-
60000-
50000-
40000-
30000-
20000-
10000-
o-
-10000-
\ 1 I 1 I I I 1 I T
-100000 200000 500000 800000
Mean(NET_AMOU)
Fitting
Linear Fit
Linear Fit
Summary of Fit
Rsquare 0.341522
Root Mean Square Error 5148.764
Mean of Response 5813.688
Observations (or Sum Wgts) 35.17674
Analysis of Variance
Source DF Sum of Squares Mean Square F Ratio
Model 1 2034907888 2e + 9 76.7607
Error 148 3923446381 26509773 Prot»F
C Total 149 5958354269 0.0000
Parameter Estimates
Term Estimate Std Error t Ratio Prot»|t|
Intercept 3095.6208 921.88 3.36 0.0010
Mean(NET_AMOU) 0.0832695 0.0095 8.76 0.0000
September 1993
C-11
-------
_
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
Attachment B
-------
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
AQUEOUS
Response: LNP_AMT
Summary of Fit
Rsquare 0.97779
Root Mean Square Error 0.126681
Mean of Response 0.4824
Observations (or Sum Wgts) S87
Source Nparm
SAMPLE 33
SOLVENT 5
SAMPLE-SOLVENT 165
DF
33
5
165
Effect Test
Sum of Squares
235.42051
1.16299
33.66981
F Ratio
444.5370
14.4939
12.7155
Prot»F
0.0000
0.0000
0.0000
Whole-Model Test
3.0-
2.0-
1.0-
o.o-
-1.0-
-1.0 0.0 1.0 2.0
LNP AMT Predicted
3.0
Analysis of Variance
Source DF Sum of Squares Mean Square F Ratio
Model 203 270.59227 1.33297 83.0609
Error 383 6.14641 0.01605 Prob>F
C Total 586 276.73868 0.0000
September 1993
C-15
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
SAMPLE
3.0-
2.0-
1.0-
o.o-1
-1.0
SAMPLE Leverage
Effect Test
Sum of Squares F Ratio DF Prob>F
235.42051 444.5370 33 0.0000
Least Squares Means
Level Least Sq Moan Std Error Mean
22221 -0.081927652 0.0517172715 -0.08193
22223 0.166153922 0.0298589806 0.16615
22224 0.213805049 0.0517172715 0.21381
22225 1.143809535 0.0298589806 1.14381
22228 0.211932028 0.0298589806 0.21193
22229 0.045608445 0.0298589806 0.04561
22231 0.576576027 ' 0.0298589806 0.57658
22232 0.370878501 0.0298589806 0.37088
22233 0.129629655 0.0298589806 0.12963
22234 1.939240086 0.0298589806 1.93924
22236 0.358980370 0.0298589806 0.35898
22239 0.025785114 0.0298589806 0.02579
22241 0.060454805 0.0298589806 0.06045
22243 -0.008480414 0.0298589806 -0.00848
23106 0.040835588 0.0298589806 0.04084
23108 -0.052768754 0.0298589806 -0.05277
23110 0.026058421 0.0298589806 0.02606
23111 0.114974854 0.0298589806 0.11497
23113 0.082593153 0.0298589806 0.08259
23115 -0.091542461 0.0298589806 -0.09154
23116 -0.101100310 0.0298589806 -0.10110
23120 0.063497202 0.0298589806 0.06350
23121 0.177725090 0.0298589806 0.17773
23124 0.262299843 0.0298589806 0.26230
23457 1.823537819 0.0298589806 1.82354
23459 0.739219598 0.0298589806 0.73922
23461 1.321908021 0.0298589806 1.32191
23463 1.774738573 0.0298589806 1.77474
23466 0.021799847 0.0298589806 0.02180
23468 0.196272754 0.0298589806 0.19627
23470 1.549298928 0.0298589806 1.54930
23472 0.171667433 0.0298589806 0.17167
23473 1.630797456 0.0310782124 1.62519
23475 1.011166706 0.0298589806 1.01117
C-16
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
SOLVENT
3.0-
2.0-
9:' 1.0-
o.o-
-1.0-
0.30
0.40
SOLVENT
0.50
Leverage
0.60
Sum of Squares
1.1629909
Effect Test
F Ratio OF
14.4939 5
Prob>F
0.0000
Level
80/20
DUPONT123
FREON
HEXANE
METHYLENE CHLORI
PERCHLOR
Least Squares Means
Least Sq Mean Std Error Mean
0.4429666373 0.0132606094 0.460088
0.4689320069 0.0133475652 0.471056
0.4441710175 0.0132606094 0.460378
0.3996973354 0.0132606094 0.417870
0.5210694150 0.0132606094 0.537898
0.5317680406 • 0.0132606094 0.546996
September 1993
C-17
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
SAMPLE'SOLVENT
3.0-
2.0-
1.0-
o.o-
-1.0
^~\—'—I—'—I—'—I"
-1.0 -0.5 0.0 0.5 1.0
SAMPLE'SOLVENT
i—[—i—[—
1.5 2.0
Leverage
2.5
Sum ot Squares
33.669811
Least
Level
22221.80/20
22221 .DUPONT123
22221 .FREON
22221 .HEXANE
22221 .METHYLENE CHLORI
22221 .PERCHLOR
•22223.80/20
22223.DUPONT123
22223.FREON
22223.HEXANE
22223.METHYLENE CHLORI
22223.PERCHLOH
22224.80/20
22224.DUPONT123
22224.FREON
22224.HEXANE
22224.METHYLENE CHLORI
22224.PERCHLOR
22225.80/20
22225.DUPONT123
22225.FREON
22225.HEXANE
2222S.METHYLENE CHLORI
22225.PERCHLOR
22228.80/20
22228.DUPONT123
22228.FREON
22228.HEXANE
22228.METHYLENE CHLORI
22228.PERCHLOR
22229.80/20
22229.DUPONT123
22229.FREON
22229.HEXANE
22229.METHYLENE CHLORI
22229.PERCHLOR
22231.80/20
22231 .DUPONT123
22231,FREON
22231 .HEXANE
22231.METHYLENE CHLORI
22231 .PERCHLOR
22232,80/20
22232.0UPONT123
22232.FREON
22232.HEXANE
22232.METHYLENE CHLORI
Effect Test
F Ratio DF Prob>F
12.7155 165 0.0000
Squares Means
Least Sq Mean
0.030529205
-0.052346480
-0.073108472
-0.446287103
-0.000000000
0.049646940
0270663947
0.183009609
0.008045125
0242453319
0232777770
0.059973765
0.016463726
0258587912
0.167292518
0.355223892
0217527813
0267734435
2.339072219
2.647357710
0.549248083
0288558265
0.432379193
0.606241738
0255844448
0262488136
0.168376570
0276765722
0.100717481
0207399809
0.080562062
0.088903341
0.007156074
0.028209478
0.039780510
0.029039203
0.407664714
0.710641029
0.568079927
0.442544918
0.686321980
0.644203593
0.153214764
0270130900
0.408605785
0233756184
0.754308091
Std Error
0.1266809262
0.1266809262
0.1266809262
0.1266809262
0.1266809262
0.1266809262
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.1266809262
0.1266809262
0.1266809262
0.1266809262
0.1266809262
0.1266809262
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
c-w
September 1993
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of OH and Grease
22232.PERCW.OR
22233.80/20
22233.OUPONT123
22233.FREON
22233.HEXANE
22233.METHYLENE CHLORI
22233.PERCHLOR
22234,80/20
22234.DUPONT123
22234.FREON
22234.HEXANE
22234.METHYLENE CHLORI
22234.PERCHLOR
22236,80/20
22236.DUPONT123
22236.FREON
22236.HEXANE
22236.METHYLENE CHLORI
22236.PERCHLOR
22239,80/20
22239.DUPONT123
22239.FREON
22239.HEXANE
22239.METHYLENE CHLORI
22239.PERCHLOR
22241,80/20
22241 ,DUPONT123
22241,FREON
22241,HEXANE
22241 .METHYLENE CHLORI
22241 .PERCHLOR
22243,80/20
22243.DUPONT123
22243.FREON
22243.HEXANE
22243.METHYLENE CHLORI
22243.PERCHLOR
23106,80/20
23106 ,DUPONT1 23
23106.FREON
23106.HEXANE
23106.METHYLENE CHLORI
23106.PERCHLOR
23108,80/20
23108,DUPONT123
23108.FREON
23108.HEXANE
23108.METHYLENE CHLORI
23108.PERCHLOR
23110,80/20
23110.DUPONT123
23110.FREON
23110.HEXANE
23110.METHYLENE CHLORI
23110.PERCHLOR
23111,80/20
23111.DUPONT123
23111,FREON
23111 .HEXANE
23111,METHYLENE CHLORI
23111,PERCHLOR
23113,80/20
23113,DUPONT123
23113.FREON
23113.HEXANE
23113.METHYLENE CHLORI
23113.PERCHLOR
23115,80/20
2311S.DUPONT123
23115.FREON
2311S.HEXANE
23115.METHYLENE CHLORI
23115.PERCHLOR
23116,80/20
23116,DUPONT123
23116.FREON
23116.HEXANE
23116.METHYLENE CHLORI
23116.PERCHLOR
23120,80/20
23120.OUPONT123
23120.FREON
23120.HEXANE
23120.METHYLENE CHLORI
23120.PERCHLOR
23121.80/20
23121.DUPONT123
23121,FREON
23121 .HEXANE
23121 .METHYLENE CHLORI
23121 .PERCHLOR
23124.80/20
0.405255284
0.188988959
0565724967
-0.201028849
0.196219393
0.076333861
0.151539602
1.915012680
2.050117391
1.842390883
1.787820143
2.0834SS076
1.956644340
0.091455233
-0.079757881
0.315852264
0.008113392
1.610903084
0.207316127
0.021363971
0.006132858
0.035865335
-0.027923760
0.023174515
0.096097765
0.063867979
0.043815594
0.051997954
0.085105647
0.034613868
0.083327785
0.074858149
-0.397055696
0.041624164
-0.065607314
0.056849687
0.238448526
-0.037921896
0.009991076
0.090729899
0.013890435
-0.001218016
0.169542026
-0.304263947
0.012451833
0.107839250
0.010516094
-0.248037118
0.104881365
0.054514033
-0.112464383
-0.057731793
0.051891377
0.115524202
0.104617092
0.032430537
0.113213271
0.010865913
0.246631887
0.134357581
0.152349936
-0.012864876
-0.112727899
0.156503533
0.152200758
0.158711876
0.153735529
-0.353191414
-0.215308359
0.105828916
-0.000000000
0.107651494
-0.194235402
-0.031491895
-0.755272277
0.014470841
0.033139273
-0.002536661
0.135088860
0.035937835
0.070701751
0.039128838
0.045570111
0.107241338
0.082403339
0.106309145
0.142302937
0.128118963
0.177416900
0298465221
0.213737374
0.406740348
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
O.O731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
O.O731392668
0.0731392668
September 1993
C-19
-------
Preliminary Report of EPA Efforts to Replace Freon for the Determination of Oil and Grease
23124.DUPONT123
23124.FREON
23124.HEXANE
23124.METHYLENE CHLORI
23124.PERCHLOR
23457.80/20
23457.0UPONT123
23457.FREON
23457.HEXANE
23457.METHYLENE CHLORI
234S7.PERCHLOR
23459,80/20
23459.0UPONT123
23459.FHEON
23459.HEXANE
23459.METHYLENE CHLORI
23459.PERCHLOR
23461.80/20
23461 .DUPONT123
23461 .FREON
23461. HEXANE
23461 .METHYLENE CHLORI
23461 .PERCHLOR
23463.80/20
23463.DUPONT123
23463.FREON
23463.HEXANE
23463.METHYLENE CHLORI
23463.PERCHLOR
23466,80/20
23466.DUPONT123
23466.FREON
23466.HEXANE
23466.METHYLENE CHLORI
23466.PERCHLOR
23468.80/20
23468.DUPONT123
23468.FREON
23468.HEXANE
23468.METHYLENE CHLORI
23468.PERCHLOR
23470.80/20
23470.DUPONT123
23470.FREON
23470.HEXANE
23470.METHYLENE CHLORI
23470.PERCHLOR
23472,80/20
2347243UPONT123
23472.FREON
23472.HEXANE
23472.METHYLENE CHLORI
23472.PERCHLOR
23473,80/20
23473.DUPONT123
23473.FREON
23473.HEXANE
23473,METHYLENE CHLORI
23473.PERCHLOR
23475.80/20
23475.DUPONT123
23475.FREON
23475.HEXANE
23475.METHYLENE CHLORI
23475.PERCHLOR
0.142777666
0.122163816
0.133983982
0.337924452
0.430208791
1.548051893
1.467895079
2.037291907
1.950458735
1.643079228
2294450072
0.893192151
0.360133237
0.948661900
0.657482979
0.420806466
1.155040855
1.065287664
1.569397541
1.158550214
1.019989934
1.522559692
1.595663081
1.679632893
1.961525653
1.978669660
1.610839908
1.651522296
1.766241026
-0.020279108
0.032455987
0.035542717
0.014449181
0.029515790
0.039114518
0.128916135
0218894130
0.245849721
0.198197495
0202826321
0.182952721
1.499087286
1.546585294
1.507133089
1.603586213
1.541638107
1.597763579
0.065431592
0292055667
0.126754795
0.042734632
0293780572
0209247342
1.497531799
1.726140760
1.609112920
1.464292065
1.650423816
1.837283378
0.898253438
1.115189881
0.845932136
0.757485270
1.402980525
1.047158987
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
O.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0895769419
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
0.0731392668
------- |