EPA/600/R-12/730
                         September 2012
    Environmental Technology
    Verification Report
    UNDERGROUND STORAGE TANK
    AUTOMATIC TANK GAUGING
    LEAK DETECTION SYSTEMS

    FRANKLIN FUELING SYSTEMS
    TSP-IGF4 WATER FLOAT
    AND
    TSP-IGF4P FLOAT

              Prepared by
              Baneiie
             7/70 Business oj Innovation
          Under a cooperative agreement with


            U.S. Environmental Protection Agency
ET1/ET1/ET1/

-------
                                EPA/600/R-12/730
                                September 2012
Environmental Technology Verification
                Report

   ETV Advanced Monitoring Systems Center

      UNDERGROUND STORAGE TANK
       AUTOMATIC TANK GAUGING
       LEAK DETECTION SYSTEMS

       FRANKLIN FUELING SYSTEMS
           TSP-IGF4 FLOAT
                  AND
           TSP-IGF4P FLOAT
                  by
          Anne Marie Gregg, and Amy Dindal, Battelle
             John McKernan, U.S. EPA

-------
                                        Notice

The U.S. Environmental Protection Agency, through its Office of Research and Development,
partially funded and collaborated in the research described herein.  This report has been
subjected to the Agency 'speer and administrative review. Any opinions expressed in this report
are those of the author (s) and do not necessarily reflect the views of the Agency,  therefore, no
official endorsement should be inferred. Any mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
                                           11

-------
                                      Foreword

The U.S. Environmental Protection Agency (EPA) is charged by Congress with protecting the
nation's air, water, and land resources. Under a mandate of national environmental laws, the
Agency strives to formulate and implement actions leading to a compatible balance between
human activities and the ability of natural systems to support and nurture life. To meet this
mandate, the EPA's Office of Research and Development provides data and science support that
can be used to solve environmental problems and build the scientific knowledge base needed to
manage our ecological resources wisely, understand  how pollutants affect our health, and prevent
or reduce environmental risks.

The Environmental Technology Verification (ETV) Program has been established by the EPA to
verify the performance characteristics of innovative environmental technology across all media
and report this objective information to permitters, buyers, and users of the technology, thus
substantially accelerating the entrance of new environmental technologies into the marketplace.
Verification organizations oversee and report verification activities based on testing and quality
assurance protocols developed with input from major stakeholders and customer groups
associated with the technology area.  ETV consists of six environmental technology centers.
Information about each of these centers can be found on the Internet at http://www.epa.gov/etv/.

Effective verifications of monitoring technologies are needed to assess environmental quality
and to supply cost and performance data to select the most appropriate technology for that
assessment. Under a cooperative agreement, Battelle has received EPA funding to plan,
coordinate, and conduct such verification tests for "Advanced Monitoring Systems for Air,
Water, and Soil" and report the results to the community at large. Information concerning this
specific environmental technology area can be found on the Internet at
http ://www. epa.gov/etv/centers/centerl .html.
                                           in

-------
                                 Acknowledgments

The authors wish to acknowledge the contribution of the many individuals without whom this
verification testing would not have been possible. Quality assurance (QA) oversight was
provided by Michelle Henderson, Laurel Staley, Teri Richardson, and Lauren Drees, EPA, and
Rosanna Buhl, Zack Willenberg, and Kristin Nichols, Battelle. We gratefully acknowledge the
Xerxes Corporation for providing a 6-foot diameter fiberglass underground storage tank shell,
BP for donating 3,000 gallons of unleaded gasoline plus transportation and Tanknology for
donating tank fittings. Additionally, we truly appreciate the in-kind analytical support from
Marathon Corporation. Finally, we want to thank Dr. Samuel Gordji of the University of
Mississippi and SSG Associates and Mr. Randy Jennings of the Tennessee Department of
Agriculture for their review of the Quality Assurance Project Plan (QAPP) and this verification
report.
                                          IV

-------
                                      Contents
                                                                                  Page
Foreword	iii
Acknowledgments	iv
List of Abbreviations	ix
Chapter 1 Background	1
Chapter 2 Technology Description	2
Chapters Test Design and Procedures	4
3.1 Test Overview	4
3.2 Test Site Description	6
   3.2.1  JS-20 Building	6
   3.2.2  Test Vessel	6
   3.2.3  Fuel Storage Tanker	10
   3.2.4  Waste Fuel Storage	10
3.3 Experimental Design	10
   3.3.1  Test la Continuous Water Ingress Test-Minimum Detection Height	12
   3.3.2  Test Ib  Continuous Water Ingress Test-Smallest Detection Increment	13
   3.3.3  Test 2 Water Ingress Detection of a Quick Water Dump, Then a Fuel Dump (Quick
   Dump)	13
3.4 Experimental Procedures	15
   3.4.1  Pre-Run Preparations	15
   3.4.2 Water Preparation andRotameter Checks	17
   3.4.3  Pre-Run Readings and Samples	18
   3.4.4 Water Ingress	18
   3.4.5 Run Observations	19
   3.4.6 Data Logging	19
   3.4.7 Run Termination	19
   3.4.8 Post-Run Sampling Analysis	19
   3.4.9 Post-Run Activities	19
3.5 Monitoring	20
3.6 Operational Factors	20
Chapter 4 Quality Assurance/Quality Control	21
4.1 Data Collection Quality Control	21
4.2 Audits	23
   4.2.1  Performance Evaluation Audit	24
   4.2.2  Technical  Systems Audit	25
   4.2.3  Data Quality Audit	26
4.3 Quality Assurance/Quality Control  Deviations	26
Chapters Statistical Methods	28
5.1 Accuracy	28
5.2 Sensitivity	28
   5.2.1  Tolerance Limit	29
   5.2.2  Minimum Detectable Level Change	30
5.3 Precision	31
5.4 Phase Separation	32

-------
5.5 Operational Factors	32
Chapter 6 Test Results	33
6.1 Accuracy	33
6.2 Sensitivity	36
   6.2.1 Tolerance Limit	36
   6.2.2 Minimum Detectable Level Change	37
6.3 Precision	38
6.4 Phase Separation, Mixing, and Float Response	39
   6.4.1 Mixing and Float Response with EO Fuel	39
   6.4.2 Mixing and Float Response with El 5 Fuel	42
   6.4.3 Mixing and Float Response with Flex Fuel	45
6.5 Operational Factors	49
Chapter 7 Performance Summary for the Franklin Fueling Systems TSP-IGF4 (First
Generation) Water Float	50
7.1 Performance Summary for the Franklin Fueling Systems First Generation Water Float	50
Chapter 8 Performance Summary for Franklin Fueling Systems TSP-IGF4P (Second
Generation) Float	53
8.1 Performance Summary for Franklin Fueling Systems Second Generation Float	53
Chapter 9 References	56
                                    Appendices

Appendix A Summary of Deviations from the QAPP
Appendix B Tank Volume Chart
Appendix C Barometric Pressure and Temperature Data
Appendix D Franklin Fueling Systems Test Data
                                         VI

-------
                                       Tables

Table 1. Summary of Verification Tests and Performance Parameters	5
Table 2. Tests 1 and 2 Run Matrix	11
Table 3. Run Summary for the Continuing Water Ingress and Water Dump Tests	14
Table 4. Analytically Determined Ethanol Content of Fuels	16
Table 5. Continuous Water Ingress Test Flow Rates	17
Table 6. Other Independent Variables Monitored During Testing	20
Table 7. Data Collection Quality Control Assessments for the ATG Verification Tests	21
Table 8. Differences from Target Fuel Heights for Continuous and Dump Test Runs	23
Table 9. PEA Results for ASTM Methods D4815 and D5501 for Ethanol Content
Determination	24
Table 10. PE Audit Results for Karl-Fischer Titration Method for Water Content
Determination	25
Table 11. Accuracy Results for the Franklin Fueling Systems First Generation Water Float	34
Table 12. Accuracy Results for the Franklin Fueling Systems Second Generation Float	34
Table 13. Accuracy Results for the Franklin Fueling Systems Technologies at the Start of the
Incremental Run (Time 0)	35
Table 14. Accuracy Results for the Franklin Fueling Systems Technologies at Run End
(Time 100)	36
Table 15. Tolerance Limit for All Test 1 Runs	37
Table 16. Tolerance Limit for Only the EO Runs	37
Table 17. Tolerance Limit for Only the El5 Runs	37
Table 18. Minimum Detectable Level Change for All Test 1  Runs	38
Table 19. Minimum Detectable Level Change for Only the EO Runs	38
Table 20. Minimum Detectable Level Change for Only the E15 Runs	38
Table 21. Precision Results for the First Generation Water Float	39
Table 22. Precision Results for the Second Generation Float	39
Table 23. Water Content and Density of Dense Phase at Completion of EO and El 5 Test 1
Runs	40
Table 24. Water Content and Density of Fuel at Completion of EO and El 5 Test 1 Runs	40
Table 25. Summary of Franklin Fueling Systems First  Generation Water Float Dump Test
Observations	48
Table 26. Summary of Franklin Fueling Systems Second Generation Float Dump Test
Observations	48
Table 27. Summary of Franklin Fueling Systems First  Generation Water Float Accuracy	50
Table 28. Summary of Franklin Fueling Systems First  Generation Water Float Precision and
Sensitivity	52
Table 29. Summary of Franklin Fueling Systems First  Generation Water Float Dump Test
Observations	52
Table 30. Summary of Franklin Fueling Systems Second Generation Float Accuracy	53
Table 31. Summary of Franklin Fueling Systems Second Generation Float Precision and
Sensitivity	55
Table 32. Summary of Franklin Fueling Systems Second Generation Float Dump Test
Observations	55
                                         vn

-------
                                       Figures

Figure 1. The Franklin Fueling Systems First Generation Water Float (TSP-IGF4) with fuel
float	2
Figure 2. The Franklin Fueling	3
Figure 3. Photographs of the test vessel at the Battelle West Jefferson facility. Top photo is
an exterior view test vessel with scaffolding platform.  The vessel is holding EO at 65% full.
Bottom photo shows the technologies during an El5 continuous water ingress run	8
Figure 4. Test vessel  schematic	9
Figure 5. EO-25% Full With Splash Duplicate - Graphical display of water detection
technology response	41
Figure 6. EO Dump Test - Graphical display of water detection technology response	41
Figure 7. E15-65% Full-With Splash - Graphical display of water detection technology
response	43
Figure 8. E15 Dump  Test - Graphical display of water detection technology response	43
Figure 9. El 5 Dump  Test -  Before water dump (initial condition)	44
Figure 10. E15 Dump Test - After fuel dump (final condition)	44
Figure 11. E85-25% Full-With Splash- Graphical display of water detection technology
response	46
Figure 12. E85 Dump Test - Graphical display of water detection technology response	46
Figure 13. E85 Dump Test - After the water dump	47
Figure 14. E85 Dump Test - After fuel dump (final condition)	47
                                          Vlll

-------
                              List of Abbreviations

AMS      Advanced Monitoring Systems
ASTM     American Society for Testing and Materials
ATG      automatic tank gauging
D         difference between measured and technology increments
dm         measured incremental change in water level
dt         technology-reported incremental  change in water level
DQA      data quality audit
DVR      digital video recorder
EO         100% gasoline
El 5        fuel that is 15% ethanol and 85% gasoline, by volume
E85        fuel that is 85% ethanol and 15% gasoline, by volume; also referred
           to as flex fuel
EPA       Environmental Protection Agency
ETV       Environmental Technology Verification
flex        flex fuel, or E85
FRP       fiberglass-reinforced plastic
ft         foot or feet
gal/hr      gallon/hour
k          tolerance coefficient
LD        leak detection
LRB       laboratory record book
mL/min    milliliter/minute
ml         milliliter
MLC      minimum water level change/minimum detectable water level change
NIST      National Institute of Standards and Technology
NWGLDE  National Work Group  on Leak Detection Evaluations
ORD      Office of Research and Development
OUST     Office of Underground Storage Tanks
PEA       performance evaluation audit
QA        quality assurance
QAPP     Quality Assurance Project Plan
QC        quality control
QMP      Quality Management Plan
SD        standard deviation
SOP       standard operating procedure
SRM      standard reference material
TL         tolerance limit
TSA       technical systems audit
x         mean
UST       underground storage tank
Var        variance
VTC       Verification Test Coordinator
                                         IX

-------
                                      Chapter 1
                                     Background


The EPA supports the ETV Program to facilitate the deployment of innovative environmental
technologies through performance verification and dissemination of information.  The goal of the
ETV Program is to further environmental protection by accelerating the acceptance and use of
improved and cost-effective technologies.  ETV seeks to achieve this goal by providing high-
quality, peer-reviewed data on technology performance to those involved in the design,
distribution, financing, permitting, purchase, and use of environmental technologies.

ETV works in partnership with recognized testing organizations; with stakeholder groups
consisting of buyers, vendor organizations, and permitters; and with the full participation of
individual technology developers. The program evaluates the performance of innovative
technologies by developing test plans that are responsive to the needs of stakeholders,
conducting field or laboratory bench tests (as appropriate), collecting and analyzing data, and
preparing peer-reviewed reports.  All evaluations are conducted in accordance with rigorous QA
protocols to ensure that data of known and adequate quality are generated and that the results are
defensible. The definition of ETV verification is to establish or prove the truth of the
performance of a technology under specific, pre-determined criteria or protocols and a strong
quality management system.  High-quality data are assured through implementation of the ETV
Quality Management Plan (QMP). ETV does not endorse, certify, or approve technologies.

The EPA's National Risk Management Research Laboratory and its verification organization
partner, Battelle, operate the Advanced Monitoring  Systems (AMS) Center under ETV.  The
AMS Center recently evaluated the performances of two Franklin Fueling Systems technologies:
TSP-IGF4 Water Float and TSP-IGF4P Float.

-------
                                      Chapter 2
                               Technology Description
This report provides results for the verification testing of the Franklin Fueling Systems
TSP-IGF4 Water Float and the Franklin Fueling Systems TSP-IGF4P Float. The following is a
description of the technologies based on information provided by the vendor. The information
provided below was not verified in this test.

The Franklin Fueling Systems TSP-IGF4 Water Float (this standard float is referred to as "First
Generation"  throughout this report for clarity) was designed to detect and measure the level of
water present at the bottom of a fuel storage tank in conjunction with a magnetostrictive level
probe and automatic tank gauge (ATG) system.  Figure 1 presents a picture of the Franklin
                                              Fueling Systems First Generation Water Float
                                              installed on a magnetostrictive probe that also
                                              contains an upper fuel inventory float. The
                                              probe is installed in the storage tank by
                                              suspending it from a chain such that the
                                              bottom of the probe is near the bottom of the
                                              tank. Specific versions of the water float are
                                              available for use in diesel fuel and  (non-
                                              ethanol blended) gasoline. This float is
                                               ballasted to have a net density intermediate
                                               to that of water and their respective fuels
                                               such that it is intended to float at the water-
                                               fuel interface.
 Figure 1.  The Franklin Fueling Systems
First Generation Water Float (TSP-IGF4)
             with fuel float.
The Franklin Fueling Systems TSP-IGF4P Float (referred to as "Second Generation" throughout
this report) is a float system designed specifically for low-ethanol blend gasoline containing up
to 15% ethanol (El 5). Figure 2 presents pictures of this float. The float buoyancy is such that it
is intended to respond to water and water-rich compositions of phase separation.

Information acquired during operation of these water detection technologies is transmitted from
the floats via a two-conductor signal cable to a data recording and display console.  A single
console can compile data for several individual floats, and the Franklin Fueling Systems TS-550
console was used for this purpose during the verification test.  The TS-550 has a touch screen
interface that continuously displays fuel levels and water  levels graphically in the display.  An
optional printer is also available and was used during the  test. The console also generates an
electronic data file and can be connected to a computer using a lObaseT ethernet connection,
which enabled data downloads and use through an internet browser.

-------
From the National Work Group on Leak Detection Evaluations' (NWGLDE) August 2011
revised certification, previous Franklin Fueling Systems water detection technology testing
reported (http://www.nwglde.org/evals/franklin_fueling_e.html):
   •  Minimum detectable water level in the tank is 0.208 inch using the First Generation float
      (TSP-IGF4) and 0.44 inch using the Second Generation model (TSP-IGF4P).
   •  Minimum detectable change in water level is 0.011 inch using the First Generation float
      (TSP-IGF4) and 0.013 inch using the Second Generation model (TSP-IGF4P).

                                  The total cost of the Franklin Fueling Systems technology
                                  that was used for testing was $5,450.70. This setup
                                  included the TS-550 console with printer (p/n T550DP,
                                  $2,149.20), two magnetostrictive inventory probes (p/n
                                  TSP-LL2, $1,424.70 each), the First Generation TSP-
                                  IGF4 Water Float for water detection in gasoline (p/n
                                  TSP-IGF4, $142.00), and the Second Generation TSP-
                                  IGF4P Float for separated phase detection in alcohol-
                                  blended fuel (p/n TSP-IGF4P, $314.10).

                                  The total cost for the First Generation Water Float
                                  assembly alone would be  $3,715.90, which includes the
                                  console, probe and float ($1,566.70 not including the
                                  console), while the cost for the Second Generation Float
                                  assembly would be $3,888, which includes the console,
                                  probe and float ($1,738.80 not including the console).
 Figure 2. The Franklin Fueling
   Systems Second Generation
      Float (TSP-IGF4P).

-------
                                      Chapter 3
                            Test Design and Procedures
3.1 Test Overview
This verification test was conducted according to procedures specified in the QAPP,1 including
deviations as described in Appendix A, and adhered to the quality system defined in the ETV
AMS Center Quality Management Plan (QMP)2. A technical panel of stakeholders was
specifically assembled for the preparation of the QAPP.  A list of participants in the technical
panel is presented in the QAPP.  The panel included representatives from industry associations,
state and federal governments, including representatives of the NWGLDE, and users. The
responsibilities of verification test stakeholders and/or peer reviewers included:

   •   Participate in technical panel discussions (when available) to provide input to the test
       design;
   •   Review and provide input to the QAPP; and
   •   Review and provide input to the verification report(s)/verification statement(s).

The QAPP and this verification report were reviewed by experts in the fields related to
underground storage tank (UST) leak detection (LD) and statistics.  The following experts
provided peer review:

   •   Randy Jennings, Tennessee Department of Agriculture  and
   •   Samuel Gordji, University of Mississippi and SSG Associates.

Battelle conducted this verification test with funding support from the EPA's Offices of
Research and Development (ORD) and Underground Storage Tanks (OUST) and the technology
vendors, with in-kind support from the Xerxes Corporation (a 6-foot [ft]  diameter fiberglass UST
shell), Tanknology (tank fittings), BP (provided 3,000 gallons of unleaded gasoline plus
transportation) and analytical support from Marathon Corporation.

-------
This verification test evaluated the performances of the Franklin Fueling Systems First
Generation Water Float and Second Generation Water Float. One goal of this verification test
was to provide information on the operability of ATG systems when used with ethanol-blended
fuel. To accomplish this goal, the experimental design included the following four options for
testing:

       1.  Water ingress detection  of continuous water ingress with a splash or without a splash
          (Continuous);
       2.  Water ingress detection  of a quick water dump, then a fuel dump (Quick Dump);
       3.  Water ingress and fuel leak detection during water ingress and fuel egress (Water
          Ingress + LD); and
       4.  Fuel leak detection (LD).

Franklin Fueling Systems chose to have its technologies tested using options 1 and 2 as
described in the QAPP for water ingress detection. These tests were performed in a controlled
test vessel that simulated the storage tank environment.  The verification testing was conducted
in a research building at Battelle's West Jefferson, OH facility between September 13  and
September 30, 2011. The technologies were challenged with fuel of differing ethanol
compositions, fuel heights within the test vessel, and water ingress methods/rates. The resulting
water detection data were used to calculate the accuracy, sensitivity, and precision, where
appropriate. Operational factors such as maintenance needs, data output, ease of use, and repair
requirements were also assessed based on technical staff observations. These performance
parameters were evaluated quantitatively using the statistical methods in Chapter  5 and
qualitatively through recorded observations. Temperature and density within the test vessel were
monitored throughout testing, and the water content of the fuels and dense phases were
analytically determined after testing using Karl Fisher titration. All testing was captured using
one or more digital video recorders (DVRs). Table 1 presents a summary of the tests performed,
and Section 3.3 presents the experimental design.
Table 1. Summary of Verification Tests and Performance Parameters
Test
1a: Continuous
Water Ingress
Test-Minimum
detection
height
1b: Continuous
Water Ingress
Test-Smallest
detection
increment
2: Quick Dump
Test Description
Continuous water ingress
detection with or without a
splash to determine the
minimum water level that the
ATG can detect
Continuous water ingress
detection with or without a
splash to determine the
smallest change in water level
that the ATG can detect
Quick water ingress detection,
then a fuel dump to induce and
observe phase separation
Performance
Parameter
• Accuracy
• Sensitivity
• Precision
• Operational
factors
• Sensitivity
• Phase
separation
• Operational
factors
Independent
Variables
• Water ingress
method/rate
• Fuel height in
tank
• Fuel type
• Water ingress
method/rate
• Fuel height in
tank
• Fuel type
• Water dump
• Fuel dump
• Fuel type
Number of
Runs
12 Runs + 4
Duplicates
Continuation of
runs in Test 1a
while observing
10 incremented
measurements
3 Runs

-------
A representative of Franklin Fueling Systems installed the two technologies in the test vessel and
trained Battelle technicians at the West Jefferson test facility on the afternoon of August 30 and
the following day on August 31, 2011.

At the end of August/beginning of September, 3,000 gallons of EO and 1,500 gallons of E85
were delivered to the test site. These fuels were stored in separate compartments of a three-
compartment fuel tanker that had been leased during testing. Portions of the EO and flex fuel
received at the test facility were blended in the third tanker compartment on September 5, 2011
and again on September 8, 2011 to produce E15 for the initial runs, and another batch was
blended on September 22, 2011.  A sample was taken from each compartment after the initial
blend was made and from only the El5 compartment after the second blend was made. The
samples were analyzed to verify that they contained ethanol within 10% of the target ethanol
contents of 0%  (EO), 15% (El 5),  and 85% (flex fuel).
3.2 Test Site Description

The interior of an existing research building, JS-20, at Battelle's West Jefferson, OH south
campus and the exterior area surrounding the building were modified to accommodate a
specially fabricated test vessel and support items. The test vessel was fabricated from a 6-ft
diameter piece of a fiberglass storage tank shell which was fitted with glass ends to allow visual
observation of the conditions within the vessel during testing. Exterior storage facilities were
made available for fuel and waste storage. Detailed descriptions of the research test site and
equipment items are provided below.

3.2.1  JS-20 Building

JS-20 is a large, high-bay building on the south property of Battelle's West Jefferson, OH
campus.  When last used, the building was operated as an intrinsically safe structure for gas
pipeline research.  The building has four large bay doors along the south side and a walk-through
entry door at the east and west ends. Two large louvered vents are located on the wall opposite
the bay doors in the northwest corner to allow air infiltration. The building is equipped with a 5-
ton (although only certified to 1 % tons) manually-operated overhead crane that was used to
assist in placing the test vessel in its desired location.  Equipment located inside JS-20 during the
verification tests included the test vessel  with scaffolding, vendor-supplied LD equipment and
consoles, fuel transfer hoses, two large ventilation fans, computers, assorted wet sampling
devices and monitors, and DVRs. The building and the exterior areas surrounding the building
are connected to a common grounding grid, and all metal equipment items used during testing
were connected to this grid.  Fuel and waste storage areas were located outside of JS-20 (see
Sections 3.2.3 and 3.2.4).

3.2.2  Test Vessel

Battelle staff designed and oversaw fabrication of the test vessel used for verification testing.
This vessel provided visualization of the behavior of the technologies, as well as the behavior of
ethanol-blended fuels  when water was introduced. Figure 3 presents photographs and Figure 4

-------
depicts the schematics of the test vessel showing the features described below and the installed
technologies.

The test vessel was constructed from a 6-ft diameter shell section of a fiberglass-reinforced
plastic (FRP) storage tank.  The section was cut to 4 ft 3 inches in length yielding a maximum
volume of 880 gallons. Appendix B presents the tank chart for the fabricated test vessel and lists
the volume at various fill heights. Glass bulkheads  were installed at each end of the test vessel to
allow observation of the interior during the runs.  The vessel was checked for leaks at the
fabrication shop and again after being placed in JS-20 by filling it with water to 94% of its
capacity (approximately 830 gallons).  At the fabrication shop the leak test lasted for 2 hours
(hr), while at the field test site the leak test lasted  overnight.  The test vessel was equipped with
four fiberglass ports to allow placement of the LD equipment to be tested.  These ports were
constructed of 4-inch FRP couplers with 12-inch risers installed along the top surface centerline
of the test vessel.  The top of the test vessel was also fitted with a 2-inch fuel filler cap and port.
A fuel filler riser pipe extended down from this fuel filler port to a point approximately 14 inches
from the bottom of the test vessel. A vent line was  installed near the top of the test vessel to
transport vapors displaced during filling and operation to the outside of the JS-20 structure. A 2-
inch drain, two 4-inch sampling ports, and a 4-inch  water ingress port were also fitted to the test
vessel. Approximately 5 quarts of resin were added to the bottom of the test vessel to level the
base and raise the interior shell to the height of the drain line, thus allowing complete draining of
the test vessel between runs. Finally, four 2-inch  thermometer wells were installed at
approximately the 25% height and 50% height levels for holding thermometers. A  containment
system was constructed around the test vessel which was capable of retaining the complete
volume of the test vessel should it leak. The containment was constructed of 2-inch by 4-inch
lumber covered with several layers of polyethylene  sheets.

The QAPP originally specified that a grid pattern  would be placed on the bottom of the tank to
enhance visualization of the dense phase.  However, the entire  interior of the test vessel was
coated with a white resin to provide  a contrast with  the liquid in the vessel such that the grid was
not necessary. Rulers were also placed vertically  into the resin at each end of the test vessel to
measure the observed dense phase height to the nearest millimeter (mm).  For further
information see the documentation on Deviation Number 8 in Appendix A.

As part of the verification, water was allowed to enter the test vessel in one of three ways: as an
ingress that produced a splash, as an ingress that did not produce a splash, and as a large volume
water dump. A system for water delivery into the test vessel was fabricated to accommodate
controlled ingress of water to satisfy each of these ingress methods. The water delivery system
consisted  of a 5-gallon bucket that delivered water to either a rotameter with a range of 0 to 300
milliliters  per min (mL/min) or a 2-inch valve.  The rotameter led to a three-way valve that could
be toggled between a splash-ingress tube and a no-splash ingress tube. The splash-ingress tube
discharged straight into the test vessel, while the no-splash-ingress tube delivered water that
trickled down the fuel filler pipe  and into the test vessel without causing a splash. The 2-inch
valve, when opened, allowed rapid introduction of water into the research vessel. A constant
pressure head was maintained in the supply bucket to ensure that the rotameter flow rate did not
fluctuate during the verification run. The constant head was established by filling the bucket

-------
through a water float valve from a separate 18-gallon reservoir.  Scaffolding was erected around
the test vessel to provide access to the sampling ports and the water delivery system.
Figure 3.  Photographs of the test vessel at the Battelle West Jefferson facility. Top photo is
an exterior view test vessel with scaffolding platform. The vessel is holding EO at 65% full.
Bottom photo shows the technologies during an E15 continuous water ingress run.

-------
                        6 feet
Sampling
Ports
         N
            Fuel
            Filling ->O
            Port
                               Additional
                               ATG Ports
                                    Vent
^
Water
Ingress
Port
                  Franklin Fueling
                                                  4.25 feet
           Technology Ports
                  Top View
                    Additional ATG Ports
                                                                                Franklin Fueling
                                                                               Technology Ports
       N^ Thermometer
          wells (4)
                          Front View
                                                                         Side View
Figure 4.  Test vessel schematic.

A representative of Franklin Fueling Systems installed the two technologies to be verified using
two of the 4-inch ports provided on top of the test vessel. The Second Generation Water Float,
designed for observing a separated phase, was installed in the port nearest the glass, and the First

-------
Generation Water Float was installed in the port nearest the middle of the test vessel. A standard
installation was performed, and the signal conductors were run down the scaffolding and across
to building columns where they were strung through wire clips to the technology console in the
intrinsically safe (i.e., protection technique for safe operation of electronic equipment in
explosive environments) portion of the building.

3.2.3  Fuel Storage Tanker

Fuel was temporarily stored in a leased three-compartment fuel tanker certified for fuel service.
The storage compartments consisted of one 2,000-gallon and two 2,500-gallon compartments.
Testing required delivery of EO and flex fuel and blending to produce El5. EO and flex fuel
were delivered and placed in the two 2,500-gallon compartments. Fuels from these two
compartments were then blended to produce El 5 which was stored in the third (2,000-gallon)
tanker compartment. After the proper quantities of EO and flex fuel were added to the
compartment, the compartment contents were circulated to blend the  mixture by withdrawing
fuel from the bottom valve and pumping it back into the top  hatch of the compartment.
Recirculation continued until the entire volume of the compartment was turned over at least
twice.  The verification test schedule required that El 5 be blended on two separate occasions.

The fuel  storage tanker was placed in a large, impervious containment system constructed of a
rubber-coated tarp capable of retaining the complete volume of the largest compartment should it
leak.  Fuel was transferred during blending and between fuel and waste storage areas and the test
vessel using an air-driven pump. A gasoline-powered air compressor was located in  front of JS-
20 in a safe area for supplying the air.

3.2.4  Waste Fuel Storage

The fuel  and water mixture in the test vessel at the completion of each run was drained from the
vessel into one or more of the 275-gallon polyethylene totes  that were located on a concrete pad
outside JS-20.  A total of 10 totes were available and placed  within a  containment system that
was capable of retaining the entire volume of the largest tote (10 percent of the total potential
volume stored in all totes). The containment area was constructed of 2-inch by 4-inch lumber
covered with polyethylene sheets.  Fuel was transferred from the test  vessel to the totes using an
air-driven pump.  Waste fuel accumulated in the totes was periodically pumped from the totes
into a vacuum truck for disposal by a commercial hazardous waste treatment firm.
3.3 Experimental Design

This verification test was designed to evaluate the functionality of the ATG systems when in
ethanol-blended fuel service. Both technologies were tested simultaneously to ensure the testing
conditions were the same and to minimize waste fuel. The technologies were installed at the
testing facility by the vendor, and Battelle staff was trained on the proper use of the technologies
as it pertained to the QAPP. Battelle staff checked the technology console for status messages
continuously until an initial float response was indicated, recorded several instrument parameter
values at the time of initial float response and every 10 minutes thereafter during the increment
                                           10

-------
runs, and backed up the collected data each day. No on-site calibrations were necessary for the
technologies.

The characteristics of independent variables were selected and established during the runs to
determine the response of the dependent variables. Performance parameters were evaluated
based on the responses of the dependent variables and used to characterize the functionality of
the ATG systems.  Table  2 is the matrix of the Test 1 and 2 runs.
Table 2. Tests 1 and 2 Run Matrix
Fuel Type
EO
E15
ESS
Test 1 Runs
Fill Height
25%
Without Splash
X
XX
X
With Splash
XX
X
X
65%
Without Splash
XX
X
Not Conducted
With Splash
X
XX
Not Conducted
Test 2 Runs
Dump
X
X
X
X indicates runs performed during verification testing. XX indicates where duplicate runs were conducted.

Dependent Variable Responses—The ATGs were evaluated with respect to their ability to
properly respond to the presence of water. Detection of water ingress represents the dependent
variable for these tests.

Independent Variable Levels—The levels of the independent variables were established to
simulate conditions expected to be found in operating USTs. The water ingress detection tests
consider different independent variables.

The independent variables included in the runs and the levels for each variable depended on the
environment the run was simulating.  The variables were altered to achieve different conditions
for the ATGs to operate within.  All water ingress tests were performed at the test facility in the
test vessel described in Section 3.2, thus preserving important physical tank features that impact
ATG technology response.  The independent variables that were varied for the test runs are:

   •   Fuel ethanol content;
   •   Fuel height; and
   •   Water ingress method/rate.

The first independent variable comprised fuels of three different ethanol concentrations (0%,
15%, and 85%). The EO fuel served as an operational baseline for the ATGs. The low end
represented EPA El5 Waiver fuel (http://www.epa.gov/otaq/regs/fuels/additive/el5/index.htm).
The flex fuel represented an existing high-end blend in use. Prior to beginning  the verification
test, the ethanol content was confirmed analytically using American Society for Testing  and
Materials (ASTM) D48153 for EO and E15, and ASTM D55014 or an equivalent method for flex
fuel. As stated in the QAPP, ethanol results were required to be within 10% of the nominal
concentration before each test run (e.g., an acceptable ethanol content of E15 would be between
                                           11

-------
13.5% and 16.5%).  For the initial runs (using EO), however, testing was started before receiving
the analytical laboratory results on ethanol content (Deviation Numbers 4 and 5).  Waiting for
the analytical results to come back would have delayed testing several days. The ethanol content
results instead returned the day after testing began and had no impact on testing results.  The
water content of the fuel and the ethanol-water mixture (dense phase) were determined using a
Karl-Fischer titration method.

The second independent variable was defined as fuel height and consisted of two heights (a 25%
full tank and a 65% full tank) during verification testing.  These two heights were chosen to
represent reasonable fuel levels that could be expected in operating tanks.  The lower fuel height
yielded the greater splash mixing potential but shorter diffusion columns through which the
water could flow. Conversely, the higher fuel height yielded the lower splash mixing potential,
but the higher diffusion column.  The fill heights were established to ± 10% of the target height
of either 25% or 65%. At 25% and 65% of the height of the test vessel, 170 and 610 gallons,
respectively, of fuel were in the test vessel.

The QAPP called for testing the technologies at 25% full  and 90% full.  Instead of a 90% full
height, testing was performed at a 65% full height. The change of testing at 65% instead of 90%
was made as a result of laboratory bench tests showing that flex fuel has the potential to hold a
large amount of water.  Testing at a 90% height would have potentially resulted in insufficient
space in the test vessel to complete water ingress testing for flex fuel. The change in fuel fill
height is the subject of Deviation Number 1. Additionally, this change resulted in less fuel waste
and safer conduct of testing due to the smaller amounts of fuel needed for the respective  tests
while maintaining a substantially higher diffusion column than possible with the 25% height.

The third independent variable was water ingress method/rate. Water ingress was either
continuous or rapid. Continuous ingress was performed with or without a splash on the surface
of the fuel. Water was fed into the test vessel at a constant rate which was controlled using a
constant pressure-head reservoir metered through a rotameter. The location of the continuous
water ingress was either straight onto the fuel surface or down the surface  of the fuel filler riser.
These two methods were selected to simulate the types of continuous water ingress that might
occur in an operating UST.  A rapid water ingress method was also devised, wherein a 2-gallon
dump of water was rapidly dropped into the test vessel as might occur when water was present in
the fuel delivery tanker or present in the spill bucket prior to opening.

3.3.1  Test la Continuous Water Ingress Test-Minimum Detection Height

The water ingress tests were focused on the mixing method of water addition into the test vessel.
In the first test,  a continuous stream of water was introduced into the field test vessel to produce
a splash on the surface of the fuel or to not produce a splash by trickling the water along  the
surface of the fuel filler riser pipe to slowly meet the surface of the fuel. These runs were
performed using the three different ethanol blends at two  different fill heights described above.
The independent variables and levels  for the continuous water ingress test were:

   •   Fuel ethanol content (three levels): EO, El5, and flex fuel;
   •   Fuel height (two levels): 25% and 65% full; and
   •   Water ingress method/rate (two levels): with splash and without splash.


                                           12

-------
The water ingress method/rate was selected to establish conditions that impact the degree of
mixing that occurs in a tank using the three ethanol blends.  In these runs, the true ingress rate
ranged from a minimum of 152 mL/min to 188 mL/min for both the with- and without-splash
ingress rates. The rate was established to accumulate enough water to generate a technology
response within 1 hour.  In some cases the water addition continued beyond 1 hour to ensure
observing a response in the water detection technology. If a response was not observed in 3
hours, the run was terminated. Run termination times were established to be the same for the
two ingress methods because it was expected that this time interval encompasses the potential for
the technologies to detect the water with both ingress methods. With these methods of water
ingress, some mixing occurred due to splash mixing (depending on the height of fuel in the
vessel) and some mixing occurred by diffusion (no splash).  Introducing water with a splash was
accomplished by positioning  a water tube such that water droplets would free-fall to the fuel
surface below. Introducing water without a splash was accomplished by positioning the water
tube such that surface tension allowed the water to flow along the outside of the fuel filler riser
pipe with minimal agitation to the surface of the fuel.

3.3.2 Test Ib Continuous Water Ingress Test-Smallest Detection Increment

To address the second requirement of water detection, once the water detection technologies
reacted to the minimum water height, the  smallest increment in water height that can be
measured was determined.  The ingress rate of 200 mL/min was calculated to produce a height
increase at the bottom of the tank of approximately 1/16* of an inch in 10 minutes. Readings
were taken from the technology, as well as visually, 10 minutes after the increment portion of the
run started. Both the technology readings and the manually-measured water levels were
recorded.  Readings/measurements were taken after ten, 10-minute increments for each  replicate
of Test 1 (to produce a minimum of 100 measurements).

3.3.3 Test 2 Water Ingress Detection of a Quick Water Dump, Then a Fuel Dump (Quick
Dump)

The second test focused on the potential to detect phase separation in an UST. The test was
designed to simulate a quick water ingress rate followed by a  high degree of mixing such as
might occur if the spill bucket was dumped into the tank at a 25% fill height and then fuel was
dumped to fill the tank to a 65% fill height. This test was mainly observational in that the test
vessel was disturbed quickly  with water and fuel, and the response of the technology was
recorded throughout the test.  Three runs of this type were conducted for Test 2, one for each of
the fuel types being evaluated in this verification test. The EO run was conducted first and used
as the baseline for the technology responses to establish the minimum wait time of 30 minutes
with El5 and flex fuel.  The independent variables and levels  for Quick Dump water ingress test
were:

   •  Fuel ethanol content (three levels): EO, El5, and flex fuel;
   •  Fuel height started at 25% and was filled after water detection to 65% full; and
   •  Water ingress method/rate:  2  gallon water dumps until the technologies detected water
       ingress
                                           13

-------
Table 3 presents the run summary and sequence for Tests 1 and 2. The QAPP called for data to
be gathered under every combination of levels between all variables.  However, when
performing the flex fuel runs at 25% of the tank height, the run termination time was reached
before the technology responded to the water ingress or any clear presence of water or a
separated phase was produced.  (After 3 hours at 200 mL/min, approximately 7.5 gallons of
water would have been added, which should have produced a water level of approximately 2
inches.) The time to produce a response in a 65% full vessel would need to be even greater than
this time.  As a result, and per the QAPP, the two 65% full runs were removed from the test
design since it was believed that no usable quantifiable data would be generated, and a large
amount of contaminated fuel would have been produced.
Table 3.  Run Summary for the Continuing Water Ingress and Water Dump Tests
Test Day
1
2
3
4
5
5
6
7
7
8
9
10
11
12
12
13
14
Date (2011)
9/13
9/14
9/15
9/16
9/19
9/19
9/20
9/21
9/21
9/22
9/23
9/26
9/27
9/28
9/28
9/29
9/30
Not conducted*
Not conducted*
Fuel
Type
EO
EO
EO
EO
EO
EO
E15
E15
E15
Flex
Flex
Flex
E15
E15
EO
E15
E15
Flex
Flex
Fill Height,
percent
25
25
25 then 65
65
65
65
25
25
25
25
25 then 65
25
65
25 then 65
25
65
65
65
65
Ingress Method
Without Splash
With Splash
Dump
Without Splash
With Splash
Without Splash
Without Splash
With Splash
Without Splash
Without Splash
Dump
With Splash
With Splash
Dump
Without Splash
Without Splash
With Splash
Without Splash
With Splash
Run ID
EO-25-wo
EO-25-w
EO-dump
EO-65-wo
EO-65-w
EO-65-wo-DUP
E15-25-WO
E15-25-W
E15-25-WO-DUP
Flex-25-wo
Flex-dump
Flex-25-w
E15-65-W
E15-dump
EO-25-wo-DUP
E15-65-wo
E15-65-W-DUP
Flex-65-wo
Flex-65-w
w - with
wo - without
DUP - duplicate/replicate run
*Runs not conducted because the results from the flex fuel runs at 25% full were terminated after 3 hours without
responses from the technologies.
                                           14

-------
3.4 Experimental Procedures

3.4.1 Pre-Run Preparations

A number of pre-run preparations were performed to ensure data quality and consistency. Pre-
run preparations included fuel blending and transfer, preparation of the water distribution system,
and introduction of water to the drain in the test vessel.

Table 3 presents the fuels and ingress methods that were used for the various runs during testing.
Some run conditions listed in Table 3 differ from the conditions discussed in the QAPP. For
instance, the flex fuel test with a fill height of 65% never occurred because testing at the 25%
level generated inconclusive data. If the flex fuel  tests had also been run, the results would have
been just as inconclusive and would have wasted several hundred gallons of fuel. Also EO-25-
wo-DUP was run as a duplicate of EO-25-wo instead of Flex-25-w as was stated in the QAPP,
because the Flex-25-wo and Flex-25-w runs were terminated after 3 hours of inconclusive
results.  For further information refer to Deviation Number 12.

Fuel deliveries included EO and flex fuel.  These fuels were used in the runs and they were also
used to prepare two volumes of El 5 (VEIS).  The amounts of EO (VEO) and flex fuel (Fpp), which
was presumed to contain 85% ethanol, needed for the blend were calculated using the equations
shown below:
                                   VEIS * 0. 15 = l^Etoh                       Equation  1
                                                                            Equation 2
                                     0.85
                                   VEIS — VFF = I^EO                          Equation 3

Two batches of El 5 blend were produced during the tests.  For the initial batch, the calculated
volumes of EO and flex fuel were measured by the pump gauge on the delivery tanker and
pumped into the fuel storage tanker or one of the 275-gallon totes. For the second batch, the
calculated volumes of EO and flex fuel were placed into one or more 275-gallon totes and
measured using the graduation marks on the totes. After the corrected volumes were measured,
both the EO and flex fuel were added to one compartment in the fuel storage tanker. The pump
was then set up to circulate the contents of the bottom of the compartment to the top of the
compartment to mix the solution.  The contents of the compartment were mixed for roughly an
hour, or long enough for the pump to circulate the volume two times. After mixing, a 50 mL
sample was collected to determine the actual ethanol content using the quick test described in the
next paragraph. If the quick test results came back low, more  flex fuel would have been mixed
in, while if the quick test results came back high, more EO would have been added (although
neither of these ever occurred during testing).  The quick test was then repeated and the process
continued until the desired ethanol content was established.

Prior to collecting a sample for laboratory analysis, the ethanol content of each bulk fuel was
tested using a method published in Appendix E of the "Guidebook for Handling, Storing, &
Dispensing Fuel Ethanol."5 This quick test was performed by adding 50 mL of water and 50 mL
                                          15

-------
of ethanol fuel to a 100 mL graduated cylinder, capping the cylinder, and shaking it until the
contents were fully mixed. The volume of the dense phase and the volume of the light phase
(Fip) were both recorded after mixing.  The ethanol content (Etoh) was calculated using the
following equation:

                              Etoh = 98.69 - [1.97 * (Tip)]                   Equation 4


After the ethanol content was determined using the quick test and found to be in the requisite
concentration range, roughly 100 mL of each fuel blend was collected and sent to an analytical
laboratory for analysis. The QAPP indicated that these samples would be stored and shipped at
0° to 5°C (32° to 40°F), but after discussion with the analytical laboratory it was determined that
shipping and handling at ambient temperature would be adequate (Deviation Number 6). Table
4 presents the amount of ethanol in the fuels used for this verification test.
Table 4. Analytically Determined Ethanol Content of Fuels
Ethanol
Blend
EO
E15
E15 Duplicate
E85
E85
E15
E85
E85 Duplicate
Sampled
Date
8-Sep-1 1
8-Sep-1 1
8-Sep-1 1
8-Sep-1 1
8-Sep-1 1
26-Sep-11
29-Sep-11
29-Sep-11
Analysis
Date
9-Sep-1 1
9-Sep-1 1
9-Sep-1 1
9-Sep-1 1
9-Sep-1 1
28-Sep-11
4-Oct-1 1
4-Oct-1 1
Batch
1
1
1
1
1
2
1 -rerun
1 -rerun
Analytical
Method
D4815
D4815
D4815
D5501
D5501
D4815
Modified D5501
Modified D5501
% Volume
Ethanol
0.11
13.76
13.79
74.54*
74.65*
14.46
79.66
79.44
*Results not within acceptance criteria, rerun using a modified D5501 method by another laboratory.

Deviation 3 stated that the original analytical laboratory determination of the flex fuel ethanol
content was not within ±10% of 85 percent ethanol as was specified in the QAPP. This
laboratory also used a calibrated range outside the acceptable target range of the sample but
within the stated ASTM method. However, both the fuel terminal mix ticket from the fuel
supplier and the quick test described in Section 3.4.1 determined the fuel ethanol content to be
acceptable. Because of this information and due to time constraints, the verification testing
continued as scheduled, and another flex fuel sample was sent to a different laboratory.  This
second laboratory performed a modified D5501 method that expanded the calibration range to
encompass the targeted range of this technology evaluation. This laboratory determined the
sample to contain 79.55% ethanol, which was within ±10% of the expected value.

After the ethanol content of the fuel was determined, fuel was transferred from the storage area
to the test vessel.  An air driven pump and several sections of transfer hose were used to transfer
fuel into the test vessel. The suction hose was first used to connect the correct tanker
compartment to the pump inlet, and the discharge hose was  used to connect the  pump outlet into
the test vessel. After the proper vent lines and valves were open, the air line that supplied
compressed air to the pump was opened to allow fuel to flow from the tanker into the test vessel.
The 25% and 65% fill levels were marked on the outside of the test vessel with  a measuring tape,
                                           16

-------
and when the fuel was at the desired level, the tanker discharge valve was closed.  The discharge
hoses were completely cleared of fuel between runs. The hoses were then disconnected, coupled
to themselves (end-to-end) and stored until the next transfer.

3.4.2  Water Preparation and Rotameter Checks

Water used for the ingress tests was colored with food dye, placed in a two-reservoir distribution
system with a constant head, and fed to the test vessel through a rotameter or dump valve. Tap
water from the site was placed in an 18-gallon reserve bucket, and several drops of food dye
were added to the reserve bucket until the water was a vibrant color.  Blue food dye was used to
produce the best contrast between the fuel and the water.  The reserve bucket fed the constant-
head reservoir that discharged directly into the test vessel (for quick dump runs) or through a
rotameter into the test vessel (for continuous ingress runs). The rotameter flow rate was checked
several times each day. For this check, the rotameter was set to the desired flow rate, and a
sample was collected in a graduated cylinder as the elapsed time was measured. Typically, a
sample was collected for 20 seconds in a graduated cylinder so that the volume of the sample
collected could be easily measured. Three such checks were performed each day,  and the results
were recorded in the Laboratory Record Book (LRB).  Table 5 presents the measured flow rate
data from the continuous water ingress test.
Table 5. Continuous Water  Ingress Test Flow Rates
Run ID
EO-25-w
EO-25-wo
EO-65-w
EO-65-wo
EO-25-wo-DUP
EO-65-wo-DUP
E15-25-W
E15-25-WO
E15-65-W
E15-65-WO
E15-25-WO-DUP
E15-65-W-DUP
E85-25-W
E85-25-WO
Rotameter setting
(ml/min)
200
200
200
200
220
200
200
200
200
220
200
220
200
200
Determined Ingress
rate (ml/min)
177
182
183
179
181
183
176
183
152
188
176
156
160
153
% Difference
-11%
-9%
-8%
-10%
-18%
-8%
-12%
-9%
-24%
-14%
-12%
-29%
-20%
-24%
The QAPP specified that water would be added to the test vessel until the water depth reached
75% of the vendor-stated detection level prior to initiating each run.  This preparation step was
specified so as to shorten the time needed for the technology to initially detect water.  However,
technologies for two different vendors were installed in the same test vessel, and the differences
in the detection thresholds of each vendor's technology were such that this criterion could not be
achieved.  In addition, the vender-stated detection levels were low enough that it was not
necessary to establish a water layer before starting ingress testing (Deviation Number 7). Due to
the fact that water added to the test vessel for most of the runs would sink to the bottom of the
vessel, however, it was still necessary to add water to fill the drain pipe prior to beginning each
                                           17

-------
run.  Otherwise, the water added during the first 10 to 15 minutes of an ingress run would
displace the fuel already in the drain and confound the observations.  After the initial fuel level in
the test vessel was established, water was added to the drain by lowering a clear pipe into the
drain from the top of the test vessel and pouring water down the pipe until it appeared that the
drain was full of water. This was not done for runs with flex fuel because the water would mix
directly into the fuel.

3.4.3 Pre-Run Readings and Samples

Samples were collected throughout testing to determine the water content and the density of the
material in the test vessel. A 50 mL or smaller sample was withdrawn from specific spots in the
test vessel through the sampling ports provided on top of the vessel.  Sample information and
density results were recorded on the Sample Conditions and Chain-of-Custody log.  Roughly 2 to
4 mL of each sample were separated into a vial and delivered to Battelle's laboratory where the
water content was determined using a Karl-Fischer titration method.  The remainder of the
sample was passed through a flow-through density meter that displayed the density and
temperature of the sample.

An 8-ft long "thief sampler was used to collect samples from the test vessel.  Between each
sample, the sampler and the containers  were decontaminated using methanol as a rinse agent.
The sampler was allowed to air dry before collecting the next sample.

Various readings were taken and data were recorded before and during  every run. These
readings included start times, end times, temperatures, heights, etc. These readings were
recorded  on Water Ingress and the Fuel Dump data sheets.  In  addition, at certain intervals, data
sheets were printed from the technology console.

Two QAPP deviations occurred related to these readings.  The QAPP stated that the water height
would be measured to the nearest l/32nd of an inch using a standard ruler. However, the  scale
installed in the bottom of the test vessel was graduated in millimeters; thus the water height was
measured to the nearest millimeter (1/25.4 inch) or 0.5 millimeter (1/50.8 inch) instead
(Deviation Number 9).  Another deviation (Deviation Number 10) from the QAPP was that
instead of continuously monitoring the  density of fuel, grab samples were obtained from  the tank
and tested at certain intervals.  The original plan to continuously withdraw a sample using a
peristaltic pump would have generated  static electricity, thus, producing an explosion hazard.
This deviation, therefore, was  needed to eliminate safety concerns of having pumping and
electrical equipment near the test vessel.

3.4.4 Water Ingress

Three types of water ingress methods were tested: continuous water ingress with splash,
continuous water ingress without splash, and a quick dump.  The two different continuous
methods were introduced into  the test vessel using a rotameter. One outlet led to a fill tube that
allowed the water to run down the fuel  filler riser pipe without creating a splash, while the other
outlet led into the test vessel and allowed the water to fall several feet to create a splash.  A
three-way valve was used to connect the rotameter discharge to the proper outlet. The valve on
the rotameter was adjusted until the desired flow rate was achieved.
                                           18

-------
The rotameter was not used for the quick dump runs.  The water reservoir bucket was marked at
the 2-gallon level. After the water reservoir bucket was filled to this mark with water, a ball
valve was quickly turned to allow the full contents of the reservoir to enter the test vessel at
once.

3.4.5 Run Observations

Observations were taken throughout each run to record the characteristics and reactions in the
test vessel. A notation was made on the run data sheets any time that an interaction, reaction, or
mixing characteristic was witnessed in the test vessel.

3.4.6 Data Logging

A lObaseT ethernet connection was used to connect the Franklin Fueling Systems console with a
laptop computer for the purpose of logging the data from the probes. The console connected to a
computer through an internet browser.  Some adjustment to computer/tank gauge IP addresses,
subnet masks, and other network parameters  was required to achieve proper communication with
the computer. The browser interface was set to display in the desired inches and refresh every 30
seconds.  Once computer connection had been established, parameters/status was able to be
viewed by anyone or altered by an administrator using a password.

3.4.7Run Termination

The continuous water ingress runs were terminated after the 100-minute incremental ingress
portion was completed, or when there were no changes indicated by the probes, or after 3 hours
if no reaction. Three hours was chosen because at the flow rates used in the testing close to 6
gallons would have been introduced to the tank in that time period, assuming complete
separation this would have created a dense phase of more than 2 inches. The quick dump runs
were terminated no sooner than 30 minutes after fuel addition to the tank had stopped.  Fuel
addition occurred at between 50 and 70 gallons per minute. When no changes in reading from
the probes were observed, the run was terminated.

3.4.8 Post-Run Sampling Analysis

The same types of readings were taken and samples were collected after the runs as for the pre-
run readings and samples discussed in Section 3.4.3. A sample  of the dense phase in the test
vessel was also collected through the drain.  Similarly, it was analyzed for density and water
content.

3.4.9 Post-Run Activities

At completion of each run, fuel was transferred into the waste totes. The process for transferring
fuel from the test vessel to the totes was similar to that used to transfer fuel,  except for the
suction and discharge locations.
                                           19

-------
3.5 Monitoring
Other variables may influence the operability of ATGs during the evaluation; therefore,
information on these other variables was collected during the testing but not controlled. Table 6
presents a list of these other variables, their measurement methods, and monitoring frequencies.
Appendix C presents the barometric pressure and ambient temperature conditions during the test
period.
Table 6. Other Independent Variables Monitored During Testing
Variable
Barometric pressure
Ambient temperature
Fuel temperature
Fuel density
Tank size, geometry, and
material of construction
Measurement Method
Barometer
Thermometer
Thermometer
Density meter
Construction specifications
Monitoring Frequency
Semi-continuous from Battelle
Weather Station
Semi-continuous from Battelle
Weather Station
Periodically during testing when
samples were taken
Periodically during testing when
samples were taken
Once prior to tank use
3.6 Operational Factors

Operational factors such as maintenance needs, data output, and sustainability factors such as
ease of use, and repair requirements were noted when observed. Battelle testing staff
documented observations in the LRB and data sheets. Examples of recorded information include
the daily status of diagnostic indicators for the technology, the effort associated with any repair,
vendor effort (e.g., time on site) for setup, the duration and causes of any technology downtime
or data acquisition failure and operator observations on many other related items (i.e., technology
startup, ease of use, and user-friendliness of the software).
                                           20

-------
                                      Chapter 4
                        Quality Assurance/Quality Control
QA/quality control (QC) procedures were performed in accordance with the QMP for the AMS
Center and the QAPP for this verification test.  QA/QC procedures and results are described in
the following subchapters.

4.1 Data Collection Quality Control

Table 7 presents a list of parameters that were proposed to be measured during the ETV tests and
the QA criteria established for them in the QAPP. Some deviations to these specified procedures
were observed during testing and noted during audits of the test.  Further discussion of this
aspect of the ETV test is provided below.

Table 7. Data Collection Quality Control Assessments for the ATG Verification Tests
Measured
Parameters
Induced water
ingress rate
Ethanol content
of fuel
Water content
of fuel and
dense phase
Fuel height
Dense phase
height
Method of
Assessment
Verify metered
rates in
triplicate using
stopwatch and
graduated
cylinder
ASTMD4815or
D5501 or
equivalent
method
ASTM E203 or
E1064: Karl-
Fischer Titration
or equivalent
method
1/8-inch
graduated scale
on the exterior
of the vessel
Standard ruler
with 1-mm and
0.1-inch
graduations
Frequency
Performed at least
once each day,
prior to testing
Once for each
batch delivered or
prepared
Once before and
after each water
ingress run
Once prior to and
during each run,
as required
At the intervals
specific to the run
being performed
Acceptance
Criteria
As determined by
assessment
method
± 1 0% of target
ethanol content
As determined by
assessment
method
± 1 0% of either
25% or 65%
height, run
dependent
As determined by
assessment
method
Corrective Action
Verified flow rate used to
calculate an average
error, which was applied
to the rotameter setting
used during a run
Review data to
troubleshoot results and
adjust as necessary
Review data to
troubleshoot results and
adjust as necessary
Adjust fuel level in vessel
as necessary
Review data and adjust
as necessary
                                          21

-------
The initial approach for the water ingress method was to use a peristaltic pump as the means of
controlling the water ingress. Due to safety concerns, a different system for water delivery into
the test vessel was fabricated to accommodate controlled ingress of water (Deviation Number
13).  Continuous water ingress was supplied using a gravity feed apparatus that maintained a
constant pressure by providing a secondary reservoir  and a float valve that controlled the water
level in a primary reservoir.  The primary (constant pressure) reservoir fed into a rotameter
which was verified prior to testing at least once each day of testing. This water feed  system was
used to supply a constant-rate water supply in lieu of a peristaltic pump due to safety concerns
associated with the static electricity build up and having an electricity source near the fuel-
containing test vessel.  To evaluate the flow rate prior to each day's testing, water was collected
in triplicate for a 20-second duration, measured with a stop watch, into a graduated cylinder at a
given flow rate as read from the rotameter.  The resulting flow rate (Fiow) for each replicate was
compared to the rotameter reading (.Rota) to calculate a percent error (Err):

                                      r,    Flaw—Rote                             -r-t    ..    —
                                      Err = —	                             Equation 5
                                             «ota

These individual replicate errors were averaged and the average applied to the rotameter setting
recorded for testing performed on a particular date. The resulting flow rate was used to calculate
volumes of water added to the vessel for a given experiment.

Experimental starting fuel heights were established at 25% or 65% of the test vessel height.
These corresponded to 17 13/ie inches and 46 5/ie inches, respectively, as read from rulers (with
1/8-inch graduations) applied to the glass sides of the test vessel. The rulers were attached to the
vessel prior to the beginning of testing and remained until all runs  were completed. Readings
between the 1/8-inch graduations were estimated to the nearest 1/16 inch.  The total interior
height of the vessel was 71 Vi inches due to the %-inch of resin added to the bottom of the  vessel
to allow the probes to sit  on a flat surface.  As presented in Table 8, starting fuel heights were
within 10% of the 25% or 65% height for all runs except the E15-25-wo-DUP run, which
happened to be 10.5% below the 25% fill height.
                                            22

-------
Table 8. Differences from Target Fuel Heights for Continuous and Dump Test Runs
Run ID
EO-25-w
EO-25-wo
EO-65-w
EO-65-wo
EO-25-wo-DUP
EO-65-wo-DUP
E15-25-W
E15-25-WO
E15-65-W
E15-65-WO
E15-25-WO-DUP
E15-65-W-DUP
E85-25-WO
E85-25-W
EO Dump
E15 Dump
E85 Dump
% Difference from Target
Fuel Height
2.1%
2.8%
-3.2%
-3.9%
-6.3%
-3.4%
-5.6%
0%
-2.7%
-1 .8%
-11%
-2.6%
-3.2%
-6.0%
-0.7%
-3.2%
-9.8%
For each test, once fuel was added to the test vessel, and prior to beginning water ingress, one
sample from the center of the test vessel next to the fill riser pipe (for EO tests) or one sample
from between the ATG probes (for El5 and flex fuel tests) was taken to determine initial density
and water content.  An aliquot (approximately 2 to 4 mL) of each sample was placed in a 4-mL
dram vial for water content analysis and an aliquot was analyzed on site for density as soon as
practical after sampling.

Dense phase height was measured using standard stainless steel rulers incorporated into the resin
placed inside the test vessel to level the bottom of the vessel. The rulers, with 1 mm and 0.1 inch
graduations, were placed inside the test vessel during construction (one at each end of the test
vessel) with the zero graduation of the ruler flush with the resin bottom of the test vessel.  During
test vessel placement at the test site, the vessel containing several inches of water was leveled as
closely as possible to within 2 mm as noted on the rulers. The north end of the test vessel was
approximately 2 mm higher than the south end, which resulted in dense phase readings on the
north  end approximately 2 mm less than those from  the south end of the vessel.
4.2 Audits

Three types of audits were performed during the verification test: a performance evaluation audit
(PEA) of the analytical methods, a technical systems audit (TSA) of the verification test
procedures, and a data quality audit (DQA). Audit procedures are described further below.
                                           23

-------
4.2.1  Performance Evaluation Audit

A PEA was conducted to assess the quality of the analytical measurements made in this
verification test.  National Institute of Standards and Technology (NIST) traceable standards
were used to evaluate all of the analytical methods.  These Standard Reference Materials (SRMs)
were analyzed directly (i.e., without preparation); the SRMs fell in the middle of the calibration
ranges of the analytical methods.

The two methods identified in the QAPP for ethanol analysis were D48153 for the lower
percentages and D55014 for the high percentages. The acceptable criterion for the audit was for
the result to be within 10% of the certified value.  Table 9 presents the results of the PEA and
concluded that these methods produced acceptable results.
Table 9.  PEA Results for ASTM Methods D4815 and D5501 for Ethanol Content
Determination
Method
D4815
D5501
Analysis
Date
8/12/11
8/12/11
Sample
Description
SRM2287E10
SRM 2900
Ethanol-Water
Certified
Ethanol
Concentration,
percent
10.1
95.6
Analytical
Ethanol
Concentration,
percent
9.58
96.6
Recovery
95%
101%
The method used for the determination of water content is a Battelle Standard Operating
Procedure (SOP)6 that follows method EPA E203 and El064. The same Ethanol-Water
SRM 2900 used to perform the ethanol PEA was used to evaluate the water method. In addition
to the SRM, two certified calibration check standards were analyzed. The criterion for this
method was within 5% of the certified concentration.  As shown in Table 10, the SRM result was
not within these bounds; however, the other two check standards were within the criteria.  The
Battelle Verification Test Coordinator (VTC) and the laboratory representative discussed the
results and determined the method acceptable (Deviation Number 11).  The certified level for
SRM 2900 is ± 1.9%. The PEA results of the SRM are acceptable if this variation is taken into
consideration.
                                          24

-------
Table 10. PE Audit Results for Karl-Fischer Titration Method for Water Content
Determination
Analysis
Date
8/23/1 1
8/23/1 1
8/23/1 1
Sample ID
53358-25-6
Water
Standard 1
Water
Standard 10
Sample
Description
SRM 2900 Ethanol-
Water-Replicate 1
Water Standard 1
Water Standard 10
Certified
Water
Content, %
4.4
4.4
4.4
Average
0.10
0.10
Average
1.0
1.0
Average
Analytical
Water Content,
%
5.26
5.31
5.25
5.27
0.093
0.115
0.104
0.988
0.997
0.993
% Error
20%
3.1%
1.1%
4.2.2  Technical Systems Audit

The Battelle AMS Center QA Officer for this verification test performed a TSA during the
laboratory bench-test portion of this verification test to ensure that the verification test was
performed in accordance with the QMP for the AMS Center and the QAPP.  On September 14
and 15, 2011 this same person conducted a TSA to verify that field testing was being conducted
according to the QAPP requirements. The September 14th TSA was conducted at the field test
site to observe the run with EO fuel at 25% full, and with a splash. Ms. Jennifer Redmon (RTI
International) conducted a simultaneous TSA on behalf of EPA under contract to Neptune and
Co. during the Battelle audit. The TSA included a review of documents available at the test site
for reference and records being maintained by the testing staff; observations of the test vessel
water delivery and measurement  system; the initiation of several splash runs; and the real time
data recording practices during each run. A debriefing was conducted with the Battelle VTC,
Battelle Verification Testing Leader, Battelle AMS Center Manager, EPA AMS Center Project
Officer and QA Manager, and Ms. Redmon.

On September 15, 2011, a TSA was conducted to  review the water content analytical procedures
at one of Battelle's analytical laboratories.  The results of the TSAs indicated that testing was
conducted according to the QAPP with minor exceptions.

Three observations were noted during the audit: 1) the laboratory analysis of the flex fuel was
greater than 10% different than the nominal concentration because the laboratory calibration
range did not include a standard at or below 85%;  2) the test vessel was not pre-filled with water
up to 75% of the vendor-stated threshold level prior to test initiation; and 3)  a peristaltic pump
was not used to deliver water to the test vessel due to concerns about the use of electrical
equipment around the test vessel. Battelle's assessment was that the noted deviations did not
negatively impact the quality of data being generated for the test, but it was agreed during the
                                          25

-------
debriefing that the VTC would attempt to identify another analytical laboratory to perform
confirmation analysis.  One finding for both the field test site and laboratory bench testing was
related to data or sample identification and the need for clear and direct links to the technology
data on site and the titrameter data in the laboratory.  TSA observations included an observation
on the need for more coordinated recording practices and an observation that supporting
measurements were not being collected at the frequency specified defined in QAPP Table 9.

4.2.3  Data Quality Audit

Records generated in the verification test received a one-to-one review before these records were
used to calculate,  evaluate, or report verification results. Data were reviewed by a Battelle
technical staff member involved in the verification test.  The person performing the review added
his/her initials and the date to a hard copy of the record being reviewed.

One hundred percent of the verification test data were reviewed for quality by the VTC, and at
least 25% of the data acquired during the verification test and 100% of the calibration and QC
data were audited by Battelle's QA Reviewer.  The data were traced from the initial acquisition,
through reduction and statistical analysis, to final reporting to ensure the integrity of the reported
results. All calculations performed on the data undergoing the audit were checked.

The DQA included a review of the raw data in comparison to the data calculation spreadsheets,
through final reporting in the report tables.  All imbedded calculations in the spreadsheets were
verified for accuracy to the QAPP, and all QC  results were reviewed. The report was reviewed
against the QAPP, and all deviations to testing were reported.  The text was reviewed against the
data tables to ensure the discussion was consistent with the data.  Minor transcription and
calculation errors  were noted and brought to the attention of the VTC for correction.  A data
audit report was prepared, and a copy was distributed to the EPA.
4.3 Quality Assurance/Quality Control Deviations

Appendix A presents a list of all deviations found during the QA/QC checks performed. Specific
deviations are discussed throughout this verification report where appropriate. The remaining
deviations are discussed below.

Deviation Number 2, the first QA deviation, stated that the calibration procedures for ethanol
blends and analysis of the PEA samples did not follow the methods defined in the QAPP. This
deviation occurred because certified standards needed for the calibration were not available.
However, the analytical laboratory routinely analyzes fuels according to the ASTM standard that
was used.

Deviation Number 12 discusses the changes made to the test run matrix to maximize data
collection and minimize fuel waste. The two E85 runs at  65% full were not conducted because
the parallel runs at 25% full were inconclusive.  In addition, since the technologies did not
respond to the E85 test runs, the incremental sensitivity tests were not conducted. Finally, the
duplicate run of the E85 fuel was changed to a duplicate of EO fuel.
                                           26

-------
Deviation Number 14 stated that many of the QC requirements listed in the QAPP were different
than those actually implemented by the analysis laboratories.  This deviation occurred due to
using several different laboratories and defining several different test methods during the
verification test design phase.  These variations in implemented QA procedures are expected to
have little or no impact on the verification test results, as the labs followed the ASTM
requirements that are widely accepted.

Deviation Number 15 was a consequence of the corrective action for Deviation Number 3. No
PEA occurred when the second flex fuel sample was sent for ethanol determination at the second
laboratory.  The second analysis of the E85 fuel was performed in-kind from the only laboratory
identified to use a modified D5501 method that fit the technology evaluation parameters. The
PEA sample was actually sent to the selected laboratory, but the laboratory never analyzed the
sample.
                                          27

-------
                                       Chapter 5
                                 Statistical Methods
The statistical methods used to evaluate the quantitative performance factors listed in Section 3.3
are presented in this chapter. Qualitative observations were also used to evaluate verification test
data.  The following subchapters describe each performance parameter evaluated.
5.1 Accuracy

Accuracy is the measure of the degree of agreement between the technology reading and the
independently-measured reading. Accuracy, as evaluated in this verification test, is the degree to
which the initial technology dense phase (i.e., water or phase separation) height measurement in
the test vessel agrees with the height measurement taken using the ruler installed in the vessel.
Bias was calculated to derive an estimate of accuracy by comparing the technology
measurements with the observed ruler measurements at the time of the initial response for each
run as shown in Equation 6.

                           „-     V"  DPHT-initi-DPHo-initj                     _,    ..   ,
                           Bias = >    - ! - l-                     Equation 6
           where:   n             = the number of runs,
                            t      = the technology-measured water or separated dense phase
                                    height at the time of initial technology response, and
                            t      = the independently-observed water or separated dense
                                    phase height at the time of initial technology response.
5.2 Sensitivity

Sensitivity is a measure of the extent to which the methods and instrumentation associated with a
given technology are able to detect the event of interest when in fact the event has occurred. A
technology is determined to have higher sensitivity as the event becomes more difficult to detect
with a certain degree of confidence.  Sensitivity differs according to the nature of the test and
type of event. Two measures of sensitivity were evaluated in the continuous water ingress
verification tests: 1) the minimum detectable height of water or separated dense phase in the test
vessel, and 2) the smallest detectable change in the height of water or separated dense phase.
                                           28

-------
5.2.1  Tolerance Limit

For this verification test, the first part of sensitivity was quantified by the minimum value for
water or separated dense phase height at which the probability is at least 0.95 (95%) that the
technology detected the presence of either water or a separated dense phase in the bottom of the
vessel. This estimate of sensitivity was based on the average of the technology measurements
acquired at the time of initial response from each run, to which a one-sided tolerance interval
was applied to derive the 95% probability. Tolerance Limit (TL) was calculated to derive an
estimate of the sensitivity  of each technology to detect water or separated dense phase using
Equations 7, 8, and 9 in the following steps:

       1.    The mean  W  of the measured water or separated dense phase heights when the
            technology first responded to continuous water ingress was calculated using
            Equation 7.
                                       =r
                                         ^->i=i
Equation 7
           where:   n             = the number of runs (12), and
                    DPHj-init     = the technology-measured water or separated dense phase
                                    height at the time of initial technology response.

       2.    The standard deviation (SD) of the measured heights was calculated using
            Equation 8.
                                                     1/2
                                   2f=1(DPHT_initj-  "'
                            SD =
                                          n-l
Equation 8
           where:    n           = the number of runs (12),
                    DPHj-init     = the technology-measured water or separated dense phase
                                    height at the time of initial technology response, and
                     x           = the mean of the initial technology responses.

       3.    The tolerance coefficient (K) for a one-sided normal tolerance interval with a 95%
            probability level and a 95% coverage for the number of runs (n=12) was obtained
            from a tolerance factors table.7

       4.    Finally, the TL was calculated using Equation 9.

                                    TL =  x + k SD                         Equation 9


            where the terms are defined as above.
                                          29

-------
5.2.2  Minimum Detectable Level Change

Sensitivity also is quantified by the smallest detectable change in the water or dense separated
phase level height once water or a dense phase is detected with at least a 95% probability of
detecting the change. This aspect of sensitivity was based on comparing the paired measurement
values (technology readings versus manual observations) for each of the 10 incremental
differences established after the challenged technologies responded to the presence of the dense
phase in the continuous water ingress runs. The minimum detectable level change (MLC) in
water height was calculated to estimate the sensitivity for each technology to detect a change in
water height using Equations 10 through 15 in the following steps:

       1.  For each technology, the incremental differences were calculated between the
          technology-measured water or separated dense phase heights (difrrinc) for each of the
          10 consecutive time increments (inci through incio) for all EO and El 5 continuous
          water ingress runs (r).

       2.  The incremental differences were calculated between the independently-observed
          water or separated dense phase heights (diforinc) for each of the 10 consecutive time
          increments (inci through incio) for all EO and El5 continuous water ingress runs (r).

       3.  For each technology and each run, the paired differences (Deltarmc) were calculated
          for each pair of technology-measured and independently-observed incremental
          changes (hi r inc- horinc) as in Equation 10:

                             Deltarinc =  difTrinc —  dif0rinc               Equation 10


           where:   r      = a specific EO or El 5 run,
                    inc    = a specific 10-minute time increment within run r,
                    difp   = the incremental difference in the technology-measured dense
                             phase height, and
                    difo   = the incremental difference in the independently-measured dense
                             phase height.

       4.  For each technology and each run, the average of all paired differences over the run
          (Deltar) was calculated as in Equation 11.
                             Deltar= Zrnc=1     rmc                        Equation 11
           where:   inc    = a specific 10-minute time increment within run r, and
                    Deltar = the paired incremental difference.

       5.  The run variance (Varrn) of the 10 paired differences was calculated separately for
          each run as in Equation 12.
                                           30

-------
                                                f         	\2
                                           ,.    (Deltar. -Deltar)
                                 Varr  = Z/L-i-	—	—             Equation 12
                                    in   ^-Ufit — ±       ti—1
           where:   n      = the number of runs (6),
                    ti      = the number of time increments (10),
                    inc    = a specific time increment, and
                    Deltar.ncand Deltar are defined in Equations 10 and 11, respectively.

       6.  The pooled variance (Varp) between all runs was calculated as in Equation 13.

                                        (tir1-l)Varr1+ ...+(tir -llVaiy             ^
                                Varp =  —^	yn  (ti -iT——             Equation 13


           where:   n      = the number of runs (12),
                    ti      = the number of time increments (10),
                    r      = the run designation, and
                    Var   = the run variance.

       7.  The pooled standard deviation (SDP) was calculated as in Equation 14.

                                                                            Equation 14


       8.  The tolerance coefficient (K) for two-sided tolerance intervals with a 95% probability
          and a 95% coverage for the number of pairs (5 * 10 and 7 *  10) was obtained from a
          tolerance factor table.8

       9.  Finally, the MLC that the technology can detect was calculated using Equation 15.

                                 MLC = k SDp                              Equation 15


           where terms are as defined above.
5.3 Precision

Precision is a measure of the extent to which the methods and instrumentation associated with a
given technology yield results that are reproducible.  For a given set of test conditions, precision
is characterized by the ratio of the mean W of a technology-measured value to its SD.  For the
continuous water ingress runs, precision corresponds to the ratio of the mean associated with the
technology-measured water or separated dense phase height at the time of initial technology
response (from Equation 7) to the SD of the technology-measured water or separated dense
phase height at that same point in the time (from Equation 8).  Precision could only be based on
the initial response heights because these are the only technology readings that can be considered

-------
reproducible. Heights measured during the increment phase of the runs varied because the
height for the first time increment was interdependent and recorded only after all technologies in
the test vessel had responded to the water ingress.
5.4 Phase Separation

Phase separation during water ingress tests was defined as formation of a separate dense phase,
other than water, that appeared in the lower portion of the liquid in the test vessel and was
recognizable when a change in appearance of the vessel contents occurred. This change resulted
in differentiation of a separate liquid layer that formed below the fuel during water ingress. This
occurrence was observed visually and recorded using a DVR during testing. Test conditions
leading to phase separation were documented to define the testing environment in which phase
separation occurred (i.e., the phase separation layer height, fuel temperature and density, etc.).
The water introduced to the test vessel during the ingress periods was dyed blue with food dye to
aid in the visualization of the phase separation.
5.5 Operational Factors

Operational factors such as maintenance needs, calibration frequency, data output, ease of use,
and repair requirements were evaluated and summarized based on technical staff observations for
all runs.
                                           32

-------
                                       Chapter 6
                                     Test Results

This chapter provides results of the quantitative and qualitative evaluations of this verification
test for the Franklin Fueling Systems First Generation Water Float and the Franklin Fueling
Systems Second Generation Water Float.  Appendix D presents the run data that were collected
and used to provide these results.
6.1 Accuracy

The accuracies of the Franklin Fueling Systems floats are shown by the differences that occurred
between the observed dense phase height and the dense phase height reported by the technology.
Bias represents the average accuracy over all of the runs.  A difference of 0.0 inches indicates
that the heights were the same for the two methods (most accurate). A bias of 0.0 inches
indicates that the technology measurement is either very accurate or produces the same number
of overestimates as underestimates.

Tables 11 and 12 present both the differences and technology bias that were calculated based on
the initial detections of water for the Franklin Fueling Systems First and Second Generation
water floats, respectively.  Table 13 presents the bias results for the beginning of incremental test
runs (Time 0). Time 100 measurements of the observed and technology measured dense phase
heights and bias results are presented in Table 14. Results for the two E85-25 runs were not
included in the bias estimate because no separated dense phase was produced when testing with
flex fuel. Consequently, the E85-65 runs were not performed, and therefore no results from
these runs could be included in the bias estimate.

The technology results were compared to a human visual measurement and cannot be considered
more accurate than the mm marks on the ruler (1/25.4 inch).  In addition, the separation of the
dense phase is more distinctly visible in the EO runs than the El5 runs as observed by the
verification staff.  This inherently added more variability among the El5  observed results. Given
this, the accuracy results were not compared by variable.  The First and Second Generation
Floats returned negative bias results for all test runs.  The First Generation Float became slightly
less accurate as the test runs progressed (starting at -1.09, to -1.10, and ending at -1.21);
however, the  Second Generation Float was relatively steady with -0.70 bias.
                                           33

-------
Table 11.  Accuracy Results for the Franklin Fueling Systems First Generation Water Float
Run ID
EO-25-w
EO-25-wo
EO-65-w
EO-65-wo
EO-25-wo-DUP
EO-65-wo-DUP
E15-25-W
E15-25-WO
E15-65-W
E15-65-WO
E15-65-W-DUP
E15-25-WO-DUP
E85-25-W
E85-25-WO
E85-65-W
E85-65-WO
Independently-
Observed Dense
Phase Height
(inches)
0.91
**
0.83
0.91
0.91
0.83
1.5
0.87
1.6
1.3
1.5
1.1
0
0.00
Technology-
Measured Dense
Phase Height
(inches)
0.01
**
0.01
0.01
0.01
0.01
0.01
0.01
0.03
0.01
0.02
0.01
o.ooa
o.ooa
Difference
(inches)
-0.90
**
-0.82
-0.90
-0.90
-0.82
-1.5
-0.86
-1.6
-1.3
-1.5
-1.1
0.00a
0.00a
Not Conducted0
Not Conducted0
              Bias (inches)
I
-1.09
a. Data points were not included in the bias calculation because a separated phase did not form.
b. Flex-65 runs were not performed because a separated phase did not form in the Flex-25 runs.
** Floats were not enabled.

Table 12. Accuracy Results for the Franklin Fueling Systems  Second Generation Float
Run ID
EO-25-w
EO-25-wo
EO-65-w
EO-65-wo
EO-25-wo-DUP
EO-65-wo-DUP
E15-25-W
E15-25-WO
E15-65-W
E15-65-WO
E15-65-W-DUP
E15-25-WO-DUP
E85-25-W
E85-25-WO
E85-65-W
E85-65-WO
Independently-
Observed Dense
Phase Height
(inches)
0.63
**
0.63
0.59
0.63
0.63
0.87
0.67
0.87
0.83
0.87
0.71
0
0
Technology-
Measured Dense
Phase Height
(inches)
0.01
**
0.01
0.02
0.01
0.02
0.02
0.01
0.01
0.04
0.02
0.01
0.11a
13a
Difference
(inches)
-0.62
**
-0.62
-0.57
-0.62
-0.61
-0.85
-0.66
-0.86
-0.79
-0.85
-0.70
0.11a
13a
Not Conducted0
Not Conducted0
             Bias (inches)
                    -0.70
a. Data points were not included in the bias calculation because a separated phase did not form.
b. Flex-65 runs were not performed because a separated phase did not form in the Flex-25 runs.
** Floats were not enabled.
                                                 34

-------
Table 13. Accuracy Results for the Franklin Fueling Systems Technologies at the Start of
the Incremental Run (Time 0)
First Generation Water Float
Run ID
EO-25-w
EO-25-wo
EO-65-w
EO-65-wo
EO-25-wo-DUP
EO-65-wo-DUP
E15-25-W
E15-25-WO
E15-65-W
E15-65-WO
E15-65-W-DUP
E15-25-WO-DUP
E85-25-W
E85-25-WO
E85-65-W
E85-65-WO
Observed Dense
Phase Height (in)
0.94
1.5
0.83
0.91
0.91
0.83
1.5
0.94
1.6
1.3
1.5
1.1
0
0.00
Dense Phase
Height (in)
0.02
**
0
0.03
0.02
0.01
0.03
0.02
0.03
0.04
0.03
0.03
0
0.00
Difference (in)
-0.92
**
-0.83
-0.88
-0.89
-0.82
-1.4
-0.92
-1.6
-1.3
-1.5
-1.1
0.00a
0.00a
Second Generation Water
Float
Dense Phase
Height (in)
0.35
**
0.26
0.35
0.36
0.28
0.57
0.29
0.64
0.51
0.61
0.44
0.00
0.00
Difference
(in)
-0.59
**
-0.57
-0.56
-0.55
-0.55
-0.89
-0.65
-0.97
-0.79
-0.89
-0.66
0.00a
0.00a
Not Conducted15
Not Conducted15
               Bias
-1.1
-0.70
a. Data points were not included in the bias calculation because a separated phase did not form.
b. Flex-65 runs were not performed because a separated phase did not form in the Flex-25 runs.
** Floats were not enabled.
                                                35

-------
Table 14. Accuracy Results for the Franklin Fueling Systems Technologies at Run End
(Time 100)
First Generation Water Float
Run ID
EO-25-w
EO-25-wo
EO-65-w
EO-65-wo
EO-25-wo-DUP
EO-65-wo-DUP
E15-25-W
E15-25-WO
E15-65-W
E15-65-WO
E15-65-W-DUP
E15-25-WO-DUP
E85-25-W
E85-25-WO
E85-65-W
E85-65-WO
Observed Dense
Phase Height (in)
1.89
2.32
1.81
1.81
1.81
1.81
3.23
2.20
3.74
3.15
3.66
2.28
0.00
0.00
Dense Phase
Height (in)
0.92
2.2
0.83
0.87
0.90
0.86
1.6
1.1
1.6
1.6
1.6
1.2
0.00
0.00
Difference (in)
-0.97
-0.10
-0.98
-0.94
-0.91
-0.95
-1.6
-1.1
-2.1
-1.5
-2.1
-1.1
0.00a
0.00a
Second Generation Water
Float
Dense Phase
Height (in)
1.29
2.45
1.18
1.22
1.24
1.21
2.30
1.50
2.64
2.22
2.60
1.62
0.00
0.00
Difference
(in)
-0.60
0.13
-0.63
-0.59
-0.57
-0.60
-0.93
-0.70
-1.1
-0.93
-1.1
-0.66
0.00a
0.00a
Not Conducted15
Not Conducted15
               Bias
-0.69
a. Data points were not included in the bias calculation because a separated phase did not form.
b. Flex-65 runs were not performed because a separated phase did not form in the Flex-25 runs.
** Floats were not enabled.
6.2 Sensitivity

6.2.1 Tolerance Limit

The tolerance limit predicts the minimum detection height (in inches for these test runs) that the
technologies can detect with a 95% confidence. Table 15 presents the TLs for the technologies
over all of the EO and El 5 runs. Tables 16 and 17 show the data for the TL calculations by
ethanol blend, as EO and El 5, respectively. For this test, the TL was a function of the separation
distance between the bottom of the test vessel and the technology probe.  These results  show that
the two technologies were installed about 0.5 inch from the bottom of the test vessel. For the
same reason as identified previously for accuracy, TLs could not be defined for the flex fuel runs
because the runs were terminated after 3 hours when no dense phase had  formed.

As installed into the test vessel, the TL for the First Generation Float was 0.03 inches and the TL
for the Second Generation Float is 0.04 inches.  The TLs for the EO runs were lower than the El 5
runs for both technologies.  This is due to the lower mean values for the EO runs over the El 5
runs.
                                           36

-------
Table 15. Tolerance Limit for All Test 1 Runs
Statistic
Mean (x) (inches)
Standard deviation (SD) (inches)
Number of runs (n)
Tolerance coefficient (k)
Tolerance Limit (TL) = x+ k SD (inches)
First Generation
Water Float
0.01
0.01
11
2.8
0.03
Second Generation
Float
0.02
0.01
11
2.8
0.04
Table 16. Tolerance Limit for Only the EO Runs
Statistic
Mean Ov (inches)
Standard deviation (SD) (inches)
Number of runs (n)
Tolerance coefficient (k)
Tolerance Limit (TL) = x+ k SD
First Generation
Water Float
0.01
0
5
4.2
0.01
Second Generation
Float
0.01
0.01
5
4.2
0.04
Table 17. Tolerance Limit for Only the E15 Runs
Statistic
Mean (x) (inches)
Standard deviation (SD) (inches)
Number of runs (n)
Tolerance coefficient (k)
Tolerance Limit (TL) = x+ k SD
First Generation
Water Float
0.02
0.01
6
3.7
0.05
Second Generation
Float
0.02
0.01
6
3.7
0.06
6.2.2 Minimum Detectable Level Change

The minimum detectable level change in water height is used to estimate the smallest change (in
inches for these runs) that the evaluated technology can read. Like above, Table 18 presents the
combined EO and El 5 results for the technologies while Tables 19 and 20 present the separate EO
and El 5 data, respectively.  An MLC value near 0.0 indicates that the technology is able to
detect very small changes in the level of the dense phase. Once again, this parameter could not
be defined for the flex fuel runs.

The overall MLC for the First Generation Float was 0.129 inches (approximately 1/8 inch).
They both also had lower MLC values with the EO test runs at approximately 1/15* of an inch
(0.067 inch).
                                          37

-------
Table 18. Minimum Detectable Level Change for All Test 1 Runs
Statistic
Varp (inches)
SDP (inches)
Number of pairs (n)
Tolerance coefficient (k)
Minimum Level Change (MLC) = kSDp
First Generation
Water Float
0.003
0.059
119
2.2
0.13
Second Generation
Float
0.001
0.032
119
2.2
0.07
Table 19. Minimum Detectable Level Change for Only the EO Runs
Statistic
Varp (inches)
SDP (inches)
Number of pairs (n)
Tolerance coefficient (k)
Minimum Level Change (MLC) = kSDp (inches)
First Generation
Water Float
0.00089
0.030
59
2.3
0.07
Second Generation
Float
0.00047
0.022
59
2.3
0.05
Table 20. Minimum Detectable Level Change for Only the E15 Runs
Statistic
Varp (inches)
SDP (inches)
Number of pairs (n)
Tolerance coefficient (k)
Minimum Level Change (MLC) = kSDp (inches)
First Generation
Water Float
0.00564
0.075
60
2.3
0.18
Second Generation
Float
0.00151
0.039
60
2.3
0.09
6.3 Precision

Tables 21 and 22 present a ratio of the mean to the SD that is used to help determine the
precision of the collected data.  These tables show the overall precision for each Franklin Fueling
Systems technology and precision results for the individual variables.  A high-precision value
signifies a high degree of reproducibility, whereas a low precision values signifies the opposite.
Overall these results indicate that the variables did not affect the precision of the technologies;
however, there were slight differences ranging from 1.6 to 2.7 for the First Generation Water
Float and 1.6 to 2.6 for the Second Generation Float.
                                           38

-------
Table 21. Precision Results for the First Generation Water Float
Test 1 Runs
Mean W
(inches)
Standard
Deviation (SD)
(inches)
Precision (X
/SD)
Overall
0.013
0.006
2.0
EO
0.01
0

E15
0.015
0.008
1.8
25% Full
0.012
0.004
2.7
65% Full
0.020
0.011
1.8
With
Splash
0.014
0.005
2.6
Without
Splash
0.018
0.012
1.6
— Not devisable by zero

Table 22. Precision Results for the Second Generation Float
Test 1 Runs
Mean Ov
(inches)
Standard
Deviation (SD)
(inches)
Precision (x
/SD)
Overall
0.016
0.009
1.8
EO
0.014
0.005
2.6
E15
0.018
0.012
1.6
25% Full
0.01
0

65% Full
0.015
0.008
1.8
With
Splash
0.016
0.009
1.8
Without
Splash
0.01
0

— Not devisable by zero
6.4 Phase Separation, Mixing, and Float Response

In the process of conducting Test 1 and Test 2 as described herein, the technologies were
challenged to detect water that had been added to fuels with differing alcohol contents. The
ability of the technology to detect water added to the test vessel was in part affected by the
interaction of water and the fuel in the test vessel and was markedly different for each type of
fuel tested. This interaction was influenced by the amount of alcohol in the fuel as well as the
mixing taking place between the water, alcohol, and gasoline.  Test 1 introduced two types of
mixing, and Test 2 introduced a third type. In general for all fuels, water splashing on the
surface of the fuel resulted in tiny water droplets with increased surface area compared to ingress
without a splash for a respective fuel height.  The ingress without splash resulted in larger water
droplets in the fuel, with less surface area, which produced less mixing within the fuel layer.
Observations on the degree of phase separation, mixing, and float responses were documented
during each run using one or more DVRs. The phase separation and mixing effects are discussed
below for each fuel along with general observations on technology response. In addition,
Appendix D presents graphical displays of the data generated over the entire run time for each
dump test run using the three fuels along with the other data from the runs.

6.4.1  Mixing and Float Response with EO Fuel

When water was mixed with the EO fuel, it immediately settled to the bottom of the test vessel.
The tests with EO showed no qualitative difference in mixing based on the water ingress method
                                           39

-------
as all of the water added appeared to collect on the bottom of the test vessel in the dense layer.
Table 23  shows the average water content of the dense phase was 98.4% and Table 24 shows that
there was no water detected in the fuel above the detection limit of the method. These two tables
also include a summary of the density measurements taken during verification testing. Water
ingress with a splash produced a wide and turbulent mixing area just below the surface of the
fuel that resulted in a large area (an approximately 12-inch circle) of splash-down near the
bottom of the test vessel, whereas water ingress without a splash resulted in very little surface
turbulence. No entrainment of water in the fuel was visible at any time, regardless of the type of
water ingress employed. Both of the Franklin Fueling System technologies responded to the
presence  of water at the bottom of the test vessel during all EO runs.

Figure 5 is a graph of the EO-25% full-with splash duplicate run. Both of the technologies
detected the water in the initial ingress detection section of the graph. The First Generation Float
was the last of all of the technologies evaluated to detect the water, marking the end of Test la
for all of the runs. Then the graph levels off horizontally in between the initial detection test and
the contentious ingress test, because no water was being added to the test vessel during that time.
Finally, the upward similarly sloped lines in the incremental ingress test section of the graph
show how the technologies tracked the ingress of water. The observed measurements and the
technology recorded results have similar slopes during the incremental test.

During the EO  dump test, mixing did not impact the ability of either technology to detect water
that had entered the test vessel during the EO dump run, and the floats returned to essentially the
same height after the fuel dump as had been recorded prior to the fuel dump. These observations
are depicted in Figure  6 and presented in more detail in Appendix D, Test Day 3,  and Run
Number 2.
Table 23. Water Content and Density of Dense Phase at Completion of EO and E15 Test 1
Runs
Test 1 Runs
EO
E15
n
6
7*
Average % Water
in Dense Phase
98.4
68.4
Standard Deviation
of % Water
in Dense Phase
1.46
13.2
Density,
g/ml
0.994
0.946
Standard
Deviation of
Density, g/mL
0.002
0.020
*Includes a duplicate sample
Table 24. Water Content and Density of Fuel at Completion of EO and E15 Test 1 Runs
Test 1 Runs
EO
E15
n
7*
6
Average % Water
in Fuel
<0.101
0.514
Standard Deviation
of % Water
in Fuel
0
0.084
Density,
g/ml
0.745
0.754
Standard
Deviation of
Density, g/mL
0.003
0.005
*Includes a duplicate sample
                                           40

-------
               Intial Detection
                   Ingress
                                                                    Second
                                                                    Generation
                                                                    Observed Water
                                                                    Heieht
Figure 5.  EO-25% Full With Splash Duplicate - Graphical display of water detection
technology response.
                                                                 First Generation

                                                                 Second Generation
                                         Time
Figure 6.  EO Dump Test - Graphical display of water detection technology response.
                                           41

-------
6.4.2  Mixing and Float Response with El5 Fuel

Similar fuel-water droplet formation interaction was noticed for the El 5 tests as for the EO runs,
with the exception that diffusion was visible throughout the fuel layer for all El 5 runs. Diffusion
currents were observed in the fuel where the water entered as tiny water droplets appeared to
drift away laterally from the water droplet column while others drifted upward.  Many of these
tiny water droplets were observed to dissolve into the fuel while the larger droplets continued to
the bottom of the test vessel to collect with the dense layer. Laterally-drifting small water
droplets were more apparent in the 65% fuel height runs than in the 25% runs.

Between the two tests, mixing was greatest for the Test 1 runs (continuous ingress), and mixing
was higher in the with-splash runs than in the without-splash runs. The fine bubbles that were
produced in the with-splash runs increased the surface area available for mixing, and as the water
fell through the fuel column, this high amount of surface area allowed the water to "pull" ethanol
from the El5 fuel. For both the 25% and 65% full runs,  enough alcohol had diffused into the
water that the water/alcohol mixture, being denser than the gasoline,  readily settled to the  bottom
of the test vessel.  Tables 23 and 24 support these observations showing that the dense phase for
the El 5 runs contains less water than the measurements for the EO runs plus the fuel had
measureable amounts of water in it for the El 5 runs where none was detected in the EO.

Like the with-splash runs, the without-splash runs produced a large surface area for alcohol
diffusion into water, but not as large as the with-splash runs. Also like the with-splash runs, the
water/alcohol mixture produced in the without-splash runs at the 25% and 65% full levels readily
settled to the bottom of the test vessel.  At completion of the runs, a greater amount of separated
phase was detected in the with-splash runs at 25% than the without-splash runs at 25% full level,
thus indicating a greater amount of alcohol being removed from the El 5. Observed dense phases
were on average more than twice as deep for El 5 runs than for EO runs when normalized for
water volume added, indicating that a substantial volume of ethanol was absorbed into the dense
phase. When the fuel height variable was also taken into account, the El5 tests showed a  greater
observed dense phase height at 65% than 25% height.  This was observed in the technology
responses where the volume of water added to attain initial water detection was on average less
in the 65% height tests than the 25% tests. Figure 7 shows the graphical representation of the
El5-65% Full-With Splash with sections separating the initial detection and the incremental
ingress portions of the verification test run.

Both technologies responded to water that had entered the test vessel during the continuous water
ingress runs with El5 and the El5 Dump run.  However, fuel added during the El5 dump test
caused both technologies to drop to 0 inches from the bottom as the separated phase dissolved
into the fuel, thus  masking the fact that water had leaked into the test vessel. The initial test
condition was yellow fuel in the test vessel at 25% full. The blue water dump settled to the
bottom of the test vessel, removing ethanol along the way, resulting in a green separated phase.
These observations are depicted in Figures 8, 9, and 10 and presented in more detail in Appendix
D, Test Day 12, Run Number 16.
                                           42

-------
        V
V
\
Time
                                                                 First Generation




                                                                 Second Generation
Figure 7.  El5-65% Full-With Splash - Graphical display of water detection technology


response.
     01
    cc
     ra

1 A



Ofi -



n



Water
Dump
(2 gal)
8:22


;
Fuel Dump
start \
8:56
\
^^— First Generation
^^— Second
1 Generation



I Fuel
Dump
9:12

                                       Time
Figure 8.  E15 Dump Test - Graphical display of water detection technology response.
                                          43

-------
Figure 9. E15 Dump Test - Before water dump (initial condition).
Figure 10. E15 Dump Test - After fuel dump (final condition).
                                         44

-------
6.4.3  Mixing and Float Response with Flex Fuel

Mixing was nearly instantaneous during all types of runs using flex fuel.  The same fine and
large bubble patterns as had been observed during the runs using El5 were also visible
immediately below the surface at the water ingress location during the continuous ingress runs,
but the water soon dissolved into the fuel (or the fuel dissolved into the water). With the splash
ingress runs, the water dispersed in a cloudy fashion into many fine droplets and was visible until
approximately 6 inches below the fuel surface before dissipating into the fuel.  The without
splash ingress runs caused the water to enter the fuel in a plume, then continued to approximately
10 inches below the fuel surface before dissipating. In both cases, the added water did not
appear to reach the bottom of the test vessel with flex fuel. After the diffusion took place, subtle
changes in fuel appearance were notable until the entire contents of the test vessel were changed
to the same green color.  Figure 11 presents the technology output for this test run. No visible
separated phase was observed, and the First Generation Water Float showed no response. At
times, the Second Generation Water Float responded to water ingress as if it was neutrally
buoyant with the high alcohol content fuel. This would occur if the fuel density and the float
density were very close to one another.  During the continuous ingress runs, the Second
Generation Water Float rose and sank in the fuel mixture following eddies that were induced by
the water mixing with the fuel. For several of the runs, the float was manually pushed down to
the bottom of the test vessel using the PVC sampling tube. Eventually, the Second Generation
Water Float rose to such a height within the test vessel that the top of the float was lodged
against the bottom of the fuel float above the liquid (i.e., it was pegged against the fuel float).  At
this point, the technology was no longer capable of providing accurate height information.
Because the technologies did not respond properly at the 25% full level, the 65% level runs were
omitted from the test.

The Second Generation Water Float responded similarly during the quick dump run  as it did in
the continuous ingress runs. After the quick dump of the first 4 gallons of water into 170 gallons
of flex fuel, a separated phase was clearly visible. At that time, the vessel contents were multi-
colored:  the separated phase was deep  blue (indigo blue), and the remaining contents were a
graduation of green. The green was due to the dyed water mixing with the alcohol in the fuel.
Mixing was best at the bottom, and gradually decreased as the height of the fuel column
increased. Within 60 seconds of starting the fuel dump, however, the separated phase became
completed engulfed  in the flex fuel and disappeared, and the entire contents of the vessel became
uniformly green in color.  Both floats rose in response to mixing in the tank. The dense phase
never returned after the fuel dump, and the level previously reported by the First Generation
Water Float eventually decreased to zero, and the Second Generation Water Float became
pegged against the fuel float. These observations are depicted in Figures  12 13, and 14 and
presented in more detail in Appendix D, Test Day 9, and Run Number 19. Tables 25 and 26
summarize the observations of the Test 2 Dump runs for the First and Second Generation Floats,
respectively.
                                           45

-------
     14
T7T
5l2
.§
01
Q.
vt
01
QC
i; R
u
01
n c
Water
i C
> ^
CuO
o
o
= 2
o5
—•—First
Generation
• Second
Generation





^^^H






"£> 'Vb '^o "^o ^ '^
Figure 11.  E85-25% Full-With Splash- Graphical display of water detection technology

response.
  01
  V)


  I
  V)
  01
  ce.

  o


  I

  tt
  Q
  ra


  §3
  _o
  o
First

Generation

Second

Generation
                                      Time
Figure 12.  E85 Dump Test - Graphical display of water detection technology response.
                                          46

-------
 Figure 13. E85 Dump Test - After the water dump.
Figure 14. E85 Dump Test - After fuel dump (final condition).
                                        47

-------
Table 25. Summary of Franklin Fueling Systems First Generation Water Float Dump Test
Observations



Test 2 Dump
Runs
EO



E15

ESS

After Water Dump

Was Phase
Separation
Observed?
Yes, it was blue


Yes, it was dark
blue-green
No clear
separation, but
stratification of
green in middle
and dark blue on
bottom (Figure 13)
Was Phase
Separation
Detected by the
first generation
float?
Yes



Yes

Yes

After Fuel Dump

Was Phase
Separation
Observed?
Yes, it was blue
No clear
separation, but
stratification of
yellow on top
and green below
(Figure 10)
No, fuel became
uniform green
color
(Figure 14)
Was Phase
Separation
Detected by the
first generation
float??
Yes



No

No
Note: Initial color of the fuel blends were yellow, initial color of water was dyed blue.
Table 26. Summary of Franklin Fueling Systems Second Generation Float Dump Test
Observations
Test 2 Dump
Runs
EO
E15
ESS

Was Phase
Separation
Observed?
Yes, it was blue
Yes, it was dark
blue-green
No clear
separation, but
stratification of
green in middle and
blue on bottom
(Figure 13)
Was Phase
Separation
Detected by the
second
generation
float?
Yes
Yes
Yes

Was Phase
Separation
Observed?
Yes, it was blue
No clear
separation, but
stratification of
yellow on top
and green below
(Figure 10)
No, fuel became
uniform green
color
(Figure 14)

Was Phase
Separation
Detected by the
second
generation
float?
Yes
No
No, float moved
up the ATG
probe to the fuel
float
Note: Initial color of the fuel blends were yellow, initial color of water was dyed blue.
                                            48

-------
6.5 Operational Factors

The Franklin Fueling Systems technologies were installed at the testing facility by the vendor,
and Battelle staff was trained on the proper use of the technologies as it pertained to this testing
design.  Battelle staff checked the technology console for status messages continuously until an
initial float response was indicated, recorded several instrument parameter values at the time of
initial float response and every 10 minutes thereafter during the increment runs, and backed up
the collected data each day. No on-site calibrations were performed.

Signals from the two Franklin Fueling Systems probes equipped with water floats were
transmitted to a console  that was used to display and record the water heights at various times
during each run. Several adjustments to the console programming supplied with the technology
were required after the vendor representative left the test site. No printer was initially installed
in the console; however, one was shipped later for installation by Battelle staff. Also, at the
beginning of testing, the console tank level display screen was frozen with the message "loading
user interface," and the console would not initiate communication with the computer.  After
several attempts to cold reboot the system (by  unplugging it for several minutes) and after
several technical assistance calls, the system interface loaded and the console communicated
properly with the computer.  At this point, the  system had lost the setup program with the test
vessel parameters (dimensions and volume) and  showed only one probe. A corrected file was e-
mailed and installed on the gauge by Battelle technicians.  Later,  during the first initiated run, the
water float portion of the probes appeared to be floating (after over an inch of water had been
added to the test vessel), but the gauge reading for both probes remained at 0.00 inches.  During
a technical assistance call, the problem was diagnosed as a failure to enable the water floats.  A
new setup file was written and saved, and subsequent testing showed the water probes to be
operating correctly. Following swap-out of the controller board and installation of updated
system parameters, communication between the  tank gauge and computer still often required  a
computer reboot whenever the ethernet cable (Cat 5e) was unplugged.

The tank gauge system was designed to acquire information using the thermal paper printer
(small 2 inch wide  strips), and it was also connected to the internet using an ethernet cable.
During the current tests,  the printer was used to capture snapshots of information by periodically
using the "print" button to obtain a hardcopy of measurements, while the ethernet connection
was used to record real-time data directly to a laptop computer. The console transferred
information to the computer at 30-second intervals.  In general, the system was easy to use as
intended and, after the initial setup issues were rectified, the system required no maintenance  or
repair once testing began.  The continuous data capturing process was relatively easy to use.
                                           49

-------
                                     Chapter 7
      Performance Summary for the Franklin Fueling Systems TSP-IGF4
                          (First Generation) Water Float


7.1 Performance Summary for the Franklin Fueling Systems First Generation Water Float

The performance of the Franklin Fueling Systems First Generation Water Float was evaluated
for its accuracy, sensitivity, precision, phase separation detection, and operational factors. The
ethanol content, fill height of fuel, and water ingress method/rate were varied to challenge the
water detection technology under a variety of simulated UST conditions. The First Generation
Water Float responded to the continuous water ingress when the test fuel was EO and El 5. No
response was observed using flex fuel. As a result, the performance parameters defined in the
QAPP, and summarized below, could not be determined for this technology when flex fuel was
employed. Tables 27 and 28 present the performance parameters determined using the data from
the EO and E15 Test la and Ib water ingress runs.

The accuracy of the Franklin Fueling Systems First Generation Water Float is shown by the
differences that occurred between the observed dense phase height and the dense phase height
reported by the technology.  Bias represents the average accuracy over all of the runs. A
difference of 0.0 inches indicates that the heights were the same for the two methods (most
accurate).  A bias of 0.0 inches indicates that the technology measurement is either very accurate
or produces the same number of overestimates as underestimates. The First Generation Franklin
Fueling Systems water float had a negative bias that moved farther from 0.0 as the tests
progressed through the incremental ingress portion of the test runs.

Table 27.  Summary of Franklin Fueling Systems First Generation Water Float Accuracy
Parameter
Accuracy (Bias) (inches)
Initial Response
(inches)
-1.09
Initial Increment
Reading
(Time 0)
(inches)
-1.10
Final Increment
Reading
(Time 100)
(inches)
-1.21
This verification test evaluated sensitivity by calculating the TL and the MLC. The TL predicts
the minimum detection height (in inches for these test runs) that the technologies can detect with
a 95% confidence. Table 28 presents the sensitivity as expressed in the TL and MLC and the
precision of the technology by each variable.  The TL for the First Generation Water Float was
0.03 inches. There was no difference in the TL of the technology when separated by variable.
                                          50

-------
The MLC in water height was used to estimate the smallest change (in inches for those runs) that
the evaluated technology can read. An MLC value near 0.0 indicates that the technology is able
to detect very small changes in the level of the dense phase. The First Generation Water Float
had very similar MLC results, with an overall MLC of 0.13 inches (approximately 1/8 inch).
Therefore, this technology detected the minimum 1/8*  inch change in dense phase height in EO
and in El 5 blends.

Tables 28 also presents a ratio of the mean to the SD that was used to help determine the
precision of the collected data.  Again, this parameter was summarized by the overall precision
for the First Generation Water Float and by the individual variables. A high-precision value
signifies a high degree of reproducibility, whereas a low precision value signifies the opposite.
Overall these results indicate that the variables did not affect the precision of the technologies;
however there were slight differences. The First Generation Water Float was more precise with
25% full over 65% full, and with a splash over without a splash.

Finally, Table 29 summarizes the observations during the Test 2 Dump runs. The water dump
was detected by the First Generation Water Float in all three fuel blends; however, once the fuel
was dumped in, the water was only detected in EO.

In general, the system was easy to use as intended.  Once an ATG water detection technology is
installed, operation of the console involves following the prompts on the console screen.

The Franklin Fueling First Generation Water Float responded to the water ingress when the test
fuel was EO and El5, but showed no response when flex fuel was used as the test fuel. The
reason for the lack of response was that no clear separated dense phase was formed in the flex
fuel when water was added to the test vessel.  As a result, the performance parameters defined in
the QAPP could not be determined for this technology when flex fuel was employed. The
calculated performance parameters were determined using the pooled data from the EO and El 5
water ingress runs.

Currently 40 CFR, Section 280.43(a) states water detection technologies should detect "water at
the bottom of the tank," which does not address water entrained in the fuel due to increased
miscibility with the presence of ethanol. The water sensor, tested according to "EPA's Standard
Test Procedures for Evaluating Leak Detection Methods: Automatic Tank Gauging Systems,"
did not detect water in the test vessel containing either intermediate  (El 5) or high (E85) ethanol
blends if the water was suspended in the product or the water did not reach the  bottom of the
tank. Because of this, there is not sufficient data to evaluate whether this technology, when used
with UST systems containing intermediate or high ethanol blends, would indicate a potential
release under every circumstance.
                                           51

-------
Table 28. Summary of Franklin Fueling Systems First Generation Water Float Precision
and Sensitivity
Test 1 Runs
EO Runs (n=5)
E15Runs(n=6)
25% Full Runs (n=5)
65% Full Runs (n=6)
With Splash Runs (n=5)
Without Splash Runs (n=6)
All Runs (n=11)
Statistics
Mean (x)
(inches)
0.01
0.02
0.01
0.02
0.01
0.02
0.01
SD
(inches)
0.000
0.008
0.004
0.011
0.005
0.012
0.006
Precision
(X/SD)
—
1.8
2.7
1.8
2.6
1.6
2.0
Sensitivity
TL (inches)
0.01
0.05
0.03
0.06
0.04
0.06
0.03
MLC
(inches)
0.07
0.18
0.10
0.16
0.19
0.08
0.13
— Not devisable by zero
Table 29. Summary of Franklin Fueling Systems First Generation Water Float Dump Test
Observations




Test 2 Dump
Runs
EO





E15



ESS




Was Phase
Separation
Observed?
Yes, it was blue




Yes, it was dark
blue-green
No clear
separation, but
stratification of
green in middle
and dark blue on
bottom (Figure 13)

Was Phase
Separation
Detected by the
first generation
float?
Yes





Yes



Yes




Was Phase
Separation
Observed?
Yes, it was blue
No clear
separation, but
stratification of
yellow on top and
green below
(Figure 10)

No, fuel became
uniform green color

(Figure 14)
Was Phase
Separation
Detected by
the first
generation
float?
Yes





No



No
Note: Initial color of the fuel blends were yellow, initial color of water was dyed blue.
                                           52

-------
                                     Chapter 8
       Performance Summary for Franklin Fueling Systems TSP-IGF4P
                            (Second Generation) Float


8.1 Performance Summary for Franklin Fueling Systems Second Generation Float

The performance of the Franklin Fueling Systems Second Generation Float was evaluated for its
accuracy, sensitivity, precision, phase separation detection, and operational factors.  The ethanol
content, fill height of fuel, and water ingress method/rate were varied to challenge the water
detection technology under a variety of simulated UST conditions. The Second Generation Float
responded to the continuous water ingress when the test fuel was EO and El5, but became
pegged under the upper fuel float when tested in flex fuel. No clear separated dense phase was
formed in the flex fuel when water was added to the test vessel. As a result, the performance
parameters defined in the QAPP,  and summarized below, could not be determined for this
technology when flex fuel was employed. Tables 30 and 31 present the performance parameters
determined using the data from the EO and El5 Test la and Ib water ingress runs.

The accuracy of the Franklin Fueling Systems Second Generation Water Float is shown by the
differences that occurred between the observed dense phase height and the dense phase height
reported by the technology.  Bias represents the average  accuracy over all of the runs. A
difference of 0.0 inches indicates that the heights were the same for the two methods (most
accurate). A bias of 0.0 inches indicates that the technology measurement is either very accurate
or produces the same number of overestimates as underestimates. The Second Generation
Franklin Fueling Systems water float had a negative bias of-0.7 throughout Tests la and Ib.

Table 30. Summary of Franklin Fueling Systems Second Generation Float Accuracy
Parameter
Accuracy (Bias) (inches)
Initial Response
(inches)
-0.70
Initial Increment
Reading
(inches)
-0.70
Final Increment
Reading
(inches)
-0.69
This verification test evaluated sensitivity by calculating the TL and the MLC. The TL predicts
the minimum detection height (in inches for these test runs) that the technologies can detect with
a 95% confidence. Table 31 presents the sensitivity as expressed in the TL and MLC and the
precision of the technology by each variable. The TL for the Second Generation Water Float
was 0.04 inches.  There was little difference in the TL of the technology when separated by
variable.
                                          53

-------
The MLC in water height was used to estimate the smallest change (in inches for those runs) that
the evaluated technology can read. An MLC value near 0.0 indicates that the technology is able
to detect very small changes in the level of the dense phase.  The Second Generation Water Float
had very similar MLC results, with an overall MLC of 0.07 inches (approximately 1/15 inch).
Therefore, this technology detected the minimum 1/8* inch change in dense phase height in EO
and in El 5 blends.

Table 31 also presents a ratio of the mean to the SD that was used to help determine the precision
of the collected data.  Again, this parameter was summarized by the overall precision for the
Second Generation Water Float and by the individual variables. A high-precision value signifies
a high degree of reproducibility, whereas a low precision value signifies the opposite.  Overall
these results indicate that the variables did not affect the precision of the technologies.

Finally,  Table 32 summarizes the observations during the Test 2 Dump runs. The water dump
was detected by the Second Generation Water Float in all three fuel blends; however, once the
fuel was dumped in, the water was only detected in EO.

In general, the system was easy to use as intended. Once an ATG water detection technology is
installed, operation of the console involves following the prompts on the console screen.

The Franklin Fueling Second Generation Water Float responded to the water ingress when the
test fuel was EO and El5,  but showed no response when flex fuel was used as the test fuel.  The
reason for the lack of response was that no clear separated dense phase was formed in the flex
fuel when water was added to the test vessel.  As a result, the performance parameters defined in
the QAPP could not be determined for this technology when flex fuel was employed. The
calculated performance parameters were determined using the pooled data from the EO and El5
water ingress runs.

Currently 40 CFR, Section 280.43(a) states water detection technologies should detect "water at
the bottom of the tank," which does not address water entrained in the fuel due to increased
miscibility with the presence of ethanol. The water sensor, tested according to "EPA's Standard
Test Procedures for Evaluating Leak  Detection Methods: Automatic Tank Gauging Systems,"
did not detect water in the test vessel containing either intermediate (El 5) or high (E85) ethanol
blends if the water was suspended in  the product or the water did not reach the bottom of the
tank. Because of this, there is not sufficient data to evaluate whether this technology, when used
with UST systems containing intermediate or high ethanol blends, would indicate a potential
release under every circumstance.
                                           54

-------
Table 31. Summary of Franklin Fueling Systems Second Generation Float Precision and
Sensitivity
Test 1 Runs
EO Runs (n=5)
E15 Runs (n=6)
25% Full Runs (n=5)
65% Full Runs (n=6)
With Splash Runs (n=5)
Without Splash Runs (n=6)
All Runs (n=11)
Statistics
Mean (x)
(inches)
0.01
0.02
0.01
0.02
0.02
0.01
0.02
SD
(inches)
0.005
0.012
0.000
0.008
0.009
0.000
0.009
Precision
(X/SD)
2.6
1.6
—
1.8
1.8
—
1.8
Sensitivity
TL (inches)
0.04
0.06
0.01
0.05
0.05
0.01
0.04
MLC
(inches)
0.05
0.09
0.05
0.09
0.09
0.06
0.07
— Not devisable by zero
Table 32. Summary of Franklin Fueling Systems Second Generation Float Dump Test
Observations
Test 2 Dump
Runs
EO
E15
ESS

Was Phase
Separation
Observed?
Yes, it was blue
Yes, it was dark
blue-green
No clear
separation, but
stratification of
green in middle and
blue on bottom
(Figure 13)

Was Phase
Separation
Detected by the
second
generation
float?
Yes
Yes
Yes

Was Phase
Separation
Observed?
Yes, it was blue
No clear
separation, but
stratification of
yellow on top and
green below
(Figure 10)
No, fuel became
uniform green color
(Figure 14)
Was Phase
Separation
Detected by
the second
generation
float?
Yes
No
No, float
moved up the
ATG probe to
the fuel float
Note: Initial color of the fuel blends were yellow, initial color of water was dyed blue.
                                           55

-------
                                     Chapter 9
                                    References
1.     Quality Assurance Project Plan for Verification of Underground Storage Tank Automatic
      Tank Gauging Leak Detection Systems. U.S. Environmental Technology Verification
      Program, Battelle, 2011.
2.     Quality Management Plan for the ETV Advanced Monitoring Systems Center, Version 8.
      U.S. EPA Environmental Technology Verification Program, Battelle, April 2011.
3.     ASTM, D4815-09: Standard Test Method for Determination ofMTBE, ETBE, TAME,
      DIPE,  tertiary-Amyl Alcohol and Ci to €4 Alcohols in Gasoline by Gas Chromatography.
      November 2009.
4.     ASTM, D5501-09: Standard Test Method for Determination ofEthanol Content of
      Denatured Fuel Ethanol by Gas Chromatography. May 2009.
5.     Guidebook for Handling, Storing, and Dispensing Fuel Ethanol. Center for
      Transportation Research Energy Systems Division Argonne National Laboratory, U.S.
      Department of Energy.
6.     Battelle, Standard Operating Procedure (SOP) Operation and Maintenance of Karl-
      Fischer Water Analysis Instrumentation in SOP No. III-009-OO. 1992, Organic and
      Polymer Chemistry Department.
7.     Lieberman, G., Tables for One-Sided Statistical Tolerance Limits, ed. I.Q. Control.  Vol.
      Vol. XIV, No 10. 1958.
8.     CRC Handbook of Tables and Probability and Statistics, ed. W.H.B. (ed.). 1966: The
      Chemical Rubber Company.
                                         56

-------
           Appendix A
Summary of Deviations from the QAPP
               A-l

-------
Deviation
  (Date)
              Description
             Cause
ETV Report
 Location
 QAPP
Location
  No. 1

(8/31/11)
Page 28 and following pages of the QAPP
indicated that two fuel fill heights (25% and
90%) would be established during the
water ingress and fuel leak tests.  This
was revised to indicate that the fuel fill
height for water ingress tests would be
25% and 65%.  The deviation happened
because if the test vessel had been filled
to 90%, a test run would require a
condition in the test vessel that may not  be
achievable, because the vessel capacity
would possibly been exceeded.
Bench test results showed that E85
had the potential of holding a large
amount of water. The QAPP
indicated that the 95% fill height for
E85 would be reduced to a lesser
level or the two 90% full runs would
be removed from the tests.

The 90% level is a remnant of the
original EPA ATG protocol for
testing the leak rate of fuel out of a
storage tank.  It was not intended
to apply to water leaking into a
tank.  When the  draft ETV QAPP
was separated into four separate
test types, the 90% level was not
revised for the water ingress tests.
                                                                                      Section 3.3
                                                                                      Page 12

                                                                                      Section 3.4.1
                                                                                      Page 17
              Section B1
              Page 28
           Table 12 and Section C1.1: calibration
           procedures for ethanol blends and
           analysis of performance evaluation audit
           (PEA) samples do not follow the QAPP.

           The PEA samples for the two methods
           used to determine the fuel ethanol content,
           D4815 (for E15 and EO) and D5501 (for
           E85), were analyzed according to the
           calibration procedures specified in the
           ASTM  methods instead of the procedures
           defined in the QAPP.  This deviation also
           applies to the samples that were collected
           during testing, which were also analyzed
           following the same ASTM methods.

           D4815- Standard Test Method for
           Determination of MTBE, ETBE,  TAME,
           DIPE, tertiary-Amyl Alcohol and Ci to C4
           Alcohols in Gasoline by Gas
  No. 2    Chromatography uses a calibration curve
           as specified  in the QAPP Table 12;
 (9/2/11)   however, the curve was not analyzed once
           every 30 samples as specified.  Once the
           calibration curve was  established, it was to
           be verified every run or every 10 samples
           whichever was more frequent, with a
           continuing calibration standard.

           D5501- Standard Test Method for
           Determination of Ethanol Content of
           Denatured Fuel Ethanol by Gas
           Chromatography uses a one-point
           standard response factor for calibration
           instead of a calibration curve  as stated in
           the QAPP Table  12. D5798-10 Standard
           Specification for Fuel Ethanol (Ed70-Ed85)
           for Automotive Spark-ignition Engines
           specifies that Method D5501 should be
           used for determination of ethanol content
           in E85. The analytical laboratory,  Intertek,
           used a 96%  ethanol certified standard for
           the one-point instrument calibration.	
                                        The QAPP contained errors related
                                        to calibration procedures. Certified
                                        standards for fuel ethanol contents
                                        between 70% and 85% that would
                                        expand the calibration range are
                                        not available.
                                  Section 4.3
                                  Page 26
              Section
              C1.1
              Page 55-56

              Section B5
              Table 12
              Page 51
                                                      A-2

-------
Deviation
  (Date)
              Description
             Cause
ETV Report
 Location
 QAPP
Location
  No. 3

 (9/21/11)
QAPP Table 11  in Section B5: The true
value of the high ethanol blend may not
meet the acceptance criteria of +/-10% of
the target ethanol content of 85%. The
analytical method results were lower than
expected when compared to the metered
mix ticket received from the blender and
the Method to Determine the Total
Hydrocarbon Content of Alcohol Fuel from
the U.S. Department of Energy's (DOE)
Guidebook for Handling, Storing,  and
Dispensing Fuel Ethanol. The accuracy of
the analytical method is questioned for the
following reasons:

.   ASTM Method D5501-Standard Test
    Method for Determination of Ethanol
    Content of Denatured Fuel Ethanol by
    Gas Chromatography uses a one-
    point calibration standard response
    factor based on a 96% ethanol
    standard. The defined range for the
    method is 93% to 97% ethanol
    content, and it is not proven to have a
    linear response lower than 93%.

.   Although ASTM Method D5798-
    Standard Specification for Fuel
    Ethanol (Ed70-Ed85) for Automotive
    Spark-ignition Engines specifies that
    Method D5501 should be used for
    determining ethanol content in E85,
    this method may not have the
    necessary accuracy.	
Based on the recommendation of
ASTM Method D5798-10, Battelle
included use of ASTM Method
D5501 in the QAPP as the
verification method for E85 ethanol
content. However, upon receipt of
the analytical results, this method
no longer appears reliable for test
purposes. The analytical results
returned a value of 75% ethanol for
the E85 blend.  This is outside of
the acceptable criteria stated in
Table 11. The mix ticket supplied
with the fuel defined the mixture as
85%.  In addition, the method from
the DOE Guidebook that is readily
used by the industry resulted in an
ethanol concentration of 86.87%,
corroborating the mix ticket value.
                                                                                      Section 3.4.1
                                                                                      Page 16
              Section B5
              Table 11
              Page 50
            QAPP Section B and B2.1 stated that the
            fuel ethanol content determination would
            be performed before testing to verify that
  No. 4     the ethanol concentration is within +/-10%
            of the target level.  A sample from the
 (9/26/11)    second blended batch of E15 was
            analyzed as soon as possible by Method
            D4815; however, the testing was not
            delayed to await the results.	
                                         Due to a change in the anticipated
                                         run order for various reasons
                                         (waste considerations, technology
                                         communications issues, etc.), the
                                         lag time allowed for the return of
                                         analytical results was removed
                                         from the schedule.
                                  Section 3.3
                                  Page 12
              Section B
              Page 28

              Section B2.1
              Page 47
            The QAPP stated that the percent ethanol
            should be tested and confirmed to be
  No. 5     ±10% of nominal concentration prior to
            each run. The EO fuel was not analyzed
(11/15/11)   prior to testing.
                                         Due to a change in the anticipated
                                         run order for various reasons
                                         (waste considerations, technology
                                         communications issues, etc.), the
                                         lag time allowed for the return of
                                         analytical results was removed
                                         from the schedule.
                                  Section 3.3
                                  Page 12
              Section B
              Page 28

              Section B2.1
              Page 47
            The QAPP stated that a 10 to 250 ml_
            sample of fuel would be collected into a
            glass sampling jar with a Teflon-lined cap
  No. 6     and sent to an analytical laboratory for
            analysis of ethanol content at 0° to 5°C
(11/15/11)   (32°to40°F). The samples for ethanol
            content analysis were  stored and shipped
            at room temperature.
                                         Battelle determined that cooling
                                         during storage and shipping was
                                         not necessary after discussing the
                                         issue with the analytical laboratory.
                                         This requirement was included in a
                                         previous version of the QAPP
                                         intending to use a different ASTM
                                         method for the fuel ethanol content
                                         determination.
                                  Section 3.4.1
                                  Page 16
              Section B2.1
              Page 47
                                                      A-3

-------
Deviation
  (Date)
              Description
             Cause
ETV Report
 Location
 QAPP
Location
  No. 7

(11/15/11)
The QAPP stated that the test vessel was
to be pre-filled with water to approximately
75% of the vendor-stated amount needed
to trigger a response prior to initiating the
water ingress runs.  The test vessel was
not pre-filled with water because four
different floats were installed in the same
test vessel, and the float nearest the
bottom of the vessel was less than this
stated amount.  Pre-filling the test vessel
relative to one of the other floats would
have caused the lowest float to respond to
water before the run had begun.
This pre-filling was thought to be
needed to shorten test times, but
the requirement was specified
before actual information was
received from the participating
vendors.  After the technologies
were installed by the vendors, the
estimated time to detect water for
the most sensitive technology was
calculated using the tank volume
chart, and the time was determined
to be a manageable duration
without pre-filling the tank.  In
addition, the conditions better
mimicked the actual ingress
scenario of a LIST.
                                                                                        Section 3.4.2
                                                                                        Page 17
              Section B1
              Page 28
  No. 8
(11/15/11)
            The QAPP stated that a grid with
            incremental pattern spacing would be
            placed within the view area of the test
            vessel to clearly display the height(s) and
            width(s) of various liquid phases in the
            tank.  A tape measure was mounted
            vertically on the end of the vessel, but no
            horizontal tape was installed
                                         The horizontal lines were initially
                                         included in the test vessel design to
                                         enhance observation of the test run
                                         and not to collect measured data.
                                         Because the shell used to construct
                                         the test vessel was deep red in
                                         color, visualization of a water or
                                         dense phase separation was
                                         expected to be difficult.  The
                                         proposal to add a white grid to the
                                         bottom of the test vessel to provide
                                         a strong contrast for viewing test
                                         conditions and dense phases was
                                         modified by fully coating the vessel
                                         interior with a white resin, thus
                                         providing a better contrast than
                                         would have been provided  had just
                                         a portion of the vessel bottom been
                                         coated with a white resin and grid.
                                   Section 3.2.2
                                   Page 7
              Section B2.2
              Page 48
  No. 9

(11/15/11)
The QAPP stated that water height would
be measured by standard ruler to at least
the nearest 1/32nd of an inch.  Neither the
external tape measure or internal ruler was
incremented to 1/32nd; however the
internal increment was to the nearest
millimeter (1/25.4 inch).	
When fabricating the test vessel,
the internal metal rulers were
identified for fuel compatibility and
ease of readability.
                                                                                        Section 3.4.3
                                                                                        Page 18
              Section B5
              Table 11
              Page 50
  No. 10
(11/15/11)
The QAPP stated that fuel density/specific
gravity would be monitored semi-
continuously. However, density was
monitored prior to each run, at the
midpoint of each run, and at the end of the
each run.
When the QAPP was written,
Battelle anticipated using a
continuous density monitor
installed in the test vessel.  During
the job hazard analysis for the ETV
test,  however, the method
proposed for continuously
extracting a sample for density
measurement was found to
represent a safety/explosion
hazard. Battelle also determined
that continuous density monitoring
was  unnecessary.	
                                                                                        Section 3.4.3
                                                                                        Page 18
              Section B1
              Table 9
              Page 38
                                                       A-4

-------
Deviation
  (Date)
              Description
             Cause
ETV Report
 Location
 QAPP
Location
  No. 11

(11/15/11)
A Battelle laboratory ran the Karl-Fisher
Titration Method for water content. The
PEA results for NIST SRM 2900 (4.4%
water by mass) did not meet the
acceptance criteria of ±5% of the certified
value. The actual recovery for the SRM
was 120%. The PEA does  not confirm
that Battelle's laboratory is able to
accurately measure low concentrations of
water in fuel.  Water standards analyzed
with the PEA indicated  that the analytical
method was not accurate at 0.1% water
although it was accurate at  1% and 10%.

The QAPP specified that if the PEA results
are not acceptable that the  PEA would be
repeated.  However, the water PEA was
not repeated as specified in the QAPP.
                                                    The PEA SRM sample was over
                                                    recovered; however, the analyst
                                                    included three other independent
                                                    NIST-traceable standards to verify
                                                    the method. Two of these
                                                    standards were within the
                                                    acceptable criteria.
                                  Section 4.2.1
                                  Page 24
              Section
              C1.1
              Page 56
            Table 4 of the QAPP listed the run to be
            performed during the field portion of the
            ETV test.  Changes were made to the
            Runs conducted, including:
  No. 12     •    Runs 11 and 15 were not conducted.
                Run 10 was conducted as a duplicate
(11/15/11)       of Run 3 rather than a duplicate of
                Run 6.
                The detailed 10-minute incremental
                sensitivity tests were not conducted
                for Runs 5 and 6.
                                        Changes to the runs were made
                                        after data were collected from the
                                        initial test runs. These data
                                        indicated that following the design
                                        in some cases would result in
                                        inconclusive observational data
                                        and an accumulation of
                                        unnecessary fuel waste.
                                  Section 4.3
                                  Page 26
              Section B1
              Table 4
              Page 30
  No. 13

(11/15/11)
Water ingress was not controlled by
peristaltic pump as described in the QAPP
but rather by gravity feed with an in-line
flow meter.  Three flow measurements
were taken with a graduated cylinder and
stop watch prior to each run. The average
flow rate was used as a correction factor
for the nominal flow rate setting.
This change was due to the same
issue as defined in Deviation
Number 10. During the job hazard
analysis for the ETV test, the
peristaltic pump proposed for
adding water to the test vessel was
found to represent a
safety/explosion hazard because
the plastic tubing used  in the pump
could build a static charge when
the steel rollers traversed the tubes
during use.  The gravity feed option
was devised just before testing
began.	
                                                                                      Section 4.1
                                                                                      Page 22
              B1.1.3
              Page 40
                                                      A-5

-------
Deviation
  (Date)
              Description
Cause
ETV Report
 Location
 QAPP
Location
  No. 14

(11/15/11)
Quality control requirements for the
analytical data (Table 12) are different
than the QC data collected by the
analytical laboratories.

Intertek (ASTM Method D4815) and the
Marathon (ASTM Method Modified D5501)
analytical laboratories followed the ASTM
method requirements. Intertek QC data
were received from the analytical lab
during the PEA but not for subsequent
samples.

The QC data from the Karl-Fisher Titration
method included three control standards
every batch instead of a control standard
every sample.
                                                    The analytical laboratories and the
                                                    analytical methods changed
                                                    multiple times during the design
                                                    phase.  ASTM QC requirements
                                                    were not incorporated  into the final
                                                    version of the QAPP.
                     Section 4.3
                     Page 26
              Section B5
              Table 12
              Page 51
  No. 15

(11/15/11)
           A PEA was not conducted for the E85
           ethanol analysis performed by the
           Marathon laboratory.
                                        Many laboratories and methods
                                        were investigated as an alternative
                                        to using ASTM D5501 for the
                                        ethanol determination at the high
                                        end. The second analysis of the
                                        E85 fuel was performed in kind
                                        from the only laboratory identified
                                        to use a modified D5501 method
                                        that fit our parameters (Marathon
                                        laboratory). The PEA sample was
                                        sent for analysis with the E85
                                        mixture; however, the laboratory
                                        did not analyze  it.	
                     Section 4.3
                     Page 27
              C1.1
              Pages 55-56
                                                      A-6

-------
   Appendix B
Tank Volume Chart
       B-l

-------
                 Area of Circle Segments for a 6-ft Diameter Tank
Area/D  values taken from:
      Concrete Pipe Design Manual, American Concrete Pipe Association,
      Arlington, VA, 1974, p 397

Diameter (D) = 71.25 inches
Gallons = gal/linear foot x 4.25 ft - 1.25 gallon + 0.25 gallon

Highlighted rows were the 25% and 65% levels at which testing was performed.
Tape
Height,
inches
-0.04
0.68
1.39
2.10
2.81
3.53
4.24
4.95
5.66
6.38
7.09
7.80
8.51
9.23
9.94
10.65
11.36
12.08
12.79
13.50
14.21
14.93
15.64
16.35
17.06
17.78
18.49
19.20
Depth (d),
inches
0.71
1.43
2.14
2.85
3.56
4.28
4.99
5.70
6.41
7.13
7.84
8.55
9.26
9.98
10.69
11.40
12.11
12.83
13.54
14.25
14.96
15.68
16.39
17.10
17.81
18.53
19.24
19.95
d/D
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.11
0.12
0.13
0.14
0.15
0.16
0.17
0.18
0.19
0.2
0.21
0.22
0.23
0.24
0.25
0.26
0.27
0.28
Area/D2
0.0013
0.0037
0.0069
0.0105
0.0147
0.0192
0.0242
0.0294
0.0350
0.0409
0.0470
0.0534
0.0600
0.0668
0.0739
0.0811
0.0885
0.0961
0.1039
0.1118
0.1199
0.1281
0.1365
0.1449
0.1535
0.1623
0.1711
0.1800
Area, ft2
0.046
0.130
0.243
0.370
0.518
0.677
0.853
1.036
1.234
1.442
1.657
1.883
2.115
2.355
2.605
2.859
3.120
3.388
3.663
3.941
4.227
4.516
4.812
5.108
5.411
5.722
6.032
6.346
Gallons
0.46
3.15
6.73
10.77
15.47
20.52
26.12
31.95
38.23
44.84
51.67
58.85
66.24
73.86
81.82
89.89
98.18
106.70
115.44
124.30
133.37
142.56
151.98
161.39
171.03
180.89
190.76
200.73
                                        B-2

-------
Tape
Height,
inches
19.91
20.63
21.34
22.05
22.76
23.48
24.19
24.90
25.61
26.33
27.04
27.75
28.46
29.18
29.89
30.60
31.31
32.03
32.74
33.45
34.16
34.88
35.59
36.30
37.01
37.73
38.44
39.15
39.86
40.58
41.29
42.00
42.71
43.43
44.14
44.85
45.56
Depth (d),
inches
20.66
21.38
22.09
22.80
23.51
24.23
24.94
25.65
26.36
27.08
27.79
28.50
29.21
29.93
30.64
31.35
32.06
32.78
33.49
34.20
34.91
35.63
36.34
37.05
37.76
38.48
39.19
39.90
40.61
41.33
42.04
42.75
43.46
44.18
44.89
45.60
46.31
d/D
0.29
0.3
0.31
0.32
0.33
0.34
0.35
0.36
0.37
0.38
0.39
0.4
0.41
0.42
0.43
0.44
0.45
0.46
0.47
0.48
0.49
0.5
0.51
0.52
0.53
0.54
0.55
0.56
0.57
0.58
0.59
0.6
0.61
0.62
0.63
0.64
0.65
Area/D2
0.1890
0.1982
0.2074
0.2167
0.2260
0.2355
0.2450
0.2546
0.2642
0.2739
0.2836
0.2934
0.3032
0.3130
0.3229
0.3328
0.3428
0.3527
0.3627
0.3727
0.3827
0.3927
0.4027
0.4127
0.4227
0.4327
0.4426
0.4526
0.4625
0.4723
0.4822
0.492
0.5018
0.5115
0.5212
0.5308
0.5404
Area, ft2
6.663
6.987
7.312
7.640
7.967
8.302
8.637
8.976
9.314
9.656
9.998
10.343
10.689
11.034
11.383
11.733
12.085
12.434
12.787
13.139
13.492
13.844
14.197
14.549
14.902
15.254
15.603
15.956
16.305
16.650
16.999
17.345
17.690
18.032
18.374
18.713
19.051
Gallons
210.82
221.13
231.44
241.86
252.28
262.93
273.58
284.34
295.09
305.97
316.84
327.82
338.80
349.79
360.88
371.98
383.18
394.28
405.49
416.69
427.90
439.11
450.31
461.52
472.73
483.94
495.03
506.24
517.33
528.32
539.41
550.40
561.38
572.25
583.12
593.88
604.64
B-3

-------
Tape
Height,
inches
46.28
46.99
47.70
48.41
49.13
49.84
50.55
51.26
51.98
52.69
53.40
54.11
54.83
55.54
56.25
56.96
57.68
58.39
59.10
59.81
60.53
61.24
61.95
62.66
63.38
64.09
64.80
65.51
66.23
66.94
67.65
68.36
69.08
69.79
70.50
Depth (d),
inches
47.03
47.74
48.45
49.16
49.88
50.59
51.30
52.01
52.73
53.44
54.15
54.86
55.58
56.29
57.00
57.71
58.43
59.14
59.85
60.56
61.28
61.99
62.70
63.41
64.13
64.84
65.55
66.26
66.98
67.69
68.40
69.11
69.83
70.54
71.25
d/D
0.66
0.67
0.68
0.69
0.7
0.71
0.72
0.73
0.74
0.75
0.76
0.77
0.78
0.79
0.8
0.81
0.82
0.83
0.84
0.85
0.86
0.87
0.88
0.89
0.9
0.91
0.92
0.93
0.94
0.95
0.96
0.97
0.98
0.99
1
Area/D2
0.5499
0.5594
0.5687
0.578
0.5872
0.5964
0.6054
0.6143
0.6231
0.6318
0.6404
0.6489
0.6573
0.6655
0.6726
0.6815
0.6893
0.6969
0.7043
0.7115
0.7186
0.7254
0.732
0.7384
0.7445
0.7504
0.756
0.7612
0.7662
0.7707
0.7749
0.7785
0.7816
0.7841
0.7854
Area, ft2
19.386
19.721
20.049
20.377
20.701
21.025
21.343
21.656
21.967
22.273
22.577
22.876
23.172
23.461
23.712
24.026
24.301
24.568
24.829
25.083
25.333
25.573
25.806
26.031
26.247
26.455
26.652
26.835
27.012
27.170
27.318
27.445
27.554
27.643
27.688
Gallons
615.28
625.93
636.35
646.78
657.09
667.40
677.48
687.46
697.32
707.07
716.71
726.24
735.65
744.84
752.80
762.77
771.51
780.03
788.32
796.39
804.35
811.97
819.37
826.54
833.38
839.99
846.27
852.09
857.70
862.74
867.45
871.48
874.96
877.76
879.21
B-4

-------
              Appendix C
Barometric Pressure and Temperature Data
                  C-l

-------
                Baroneter (in)
                      2011
          T^     \^    \^    \^     \^    \^    \
      Sat   Sun   Mon   Tue   Wed   Thu   Fri
      9/10  9/11  9/12  9/13  9/14  9/15  9/16
90.OT
50.0--
40.0-
           Outside Tenperature 
-------
          Baroneter  (in)
                2011
          T      \     I      I      T
Sat   Sun   Mon   Tue   Wed   Thu   Fri
9/17  9/18  9/19  9/20   9/21  9/22 9/23
     Outside Tenperature (F>
                2011
     \     \^    \^     \^    \^     \^    \
Sat   Sun   Non   Tue   Wed  Thu   Fri
9/17  9/18  9/19  9/20  9/21  9/22  9/23
                    C-3

-------
          Baroneter  (in)
                2011
     \     \^    \^    \^    \^    \^    \
Wed   Thu   Fri   Sat   Sun   Non   Tue
9/28  9/29  9/30  10/1  10/2  10/3  10/4
      Outside  Tenperature (F>
                 2011
          i     i      i     i      r
Wed   Thu   Fri   Sat   Sun   Mon   Tue
9/28  9/29   9/30  10/1  10/2  10/3  10/4
1
                    C-4

-------
       Appendix D
Franklin Fueling Test Data
           D-l

-------
TEST DAY 1
Run Number
1
Test Day
1
Date
9/13/11
Fuel
EO
Fuel Level,
percent
25
Ingress Method
Without splash
Run ID
EO-25-wo

Time
Fuel height FFS side (inches)
DP FFS side (mm)
Temp 1 (°C) (corrected)
Temp 2 (°C) (corrected)
Temp 3 (°C) (corrected)
Temp 4 (°C) (corrected)
Ingress rate (ml/min)
Ingress rate determined (ml/min)
Start of Incremental Test (T=0)
Initial
11:30 AM
18-7/16
trace
21.9
21.7
NA
NA
200
Mid-Point
1:45 PM
18-7/8
37.5
23.0
23.1
NA
NA
200
Final (T=1 00)
3:20 PM
19-1/4
59
23.8
24.0
NA
NA
200
182a
NAb
a.  Determined using the average determined flow rate and applying the average % error.
b.  Unclear when ingress was stopped between minimum detect & incremental tests.

Rotameter Calibration
Observed
Flow Rate,
cm3/min
190
200
200
Volume
Collected,
ml
59.0
61.0
59.0
Collection
Time,
sec
20.0
20.0
20.0
Determined
Flow Rate,
ml/min
177.0
183.0
177.0
Error
-7%
-9%
-12%


Average
Standard deviation
Observed
Flow Rate,
cm3/min
197
6
Volume
Collected,
ml
59.7
1.2
Collection
Time,
sec
20.0
0.0
Determined
Flow Rate,
ml/min
179.0
3.5

Error
-9%
2%
                                         D-2

-------
Minimum Detection Height Test
Technology
First Generation
Second
Generation
Time at
Alarm
Observed
Dense Phase
Height,
mm
Observed
Dense Phase
Height,
inches
Technology
Dense Phase
Height,
inches
Elapsed Time
to Alarm,
min
Water
Volume to
Alarm,
ml
Water Volume
to Alarm,
gal
Floats were not enabled
Smallest Detection Increment Test
Elapsed
Time,
min
0
10
20
30
40
50
60
70
80
90
100
Observed
Dense
Phase
Height,
mm
38
40
42.5
44.5
47
49
51
53
55
57
59
Observed
Dense
Phase
Height,
inches
1.50
1.57
1.67
1.75
1.85
1.93
2.01
2.09
2.17
2.24
2.32
First
Generation
Technology,
inches
Second
Generation
Technology,
inches
Water floats not enabled
1.45a
1.59
1.69
1.77
1.86
1.94
2.01
2.08
2.14
2.22
1.66a
1.80
1.89b
1.98
2.06
2.14
2.22
2.30
2.37
2.45
Cumulative
Water Added
Since T = 0,
ml
0
1,821
3,642
5,463
7,284
9,105
10,926
12,747
14,568
16,389
18,211
Cumulative
Water Added
Since T = 0,
gal
0.00
0.48
0.96
1.44
1.92
2.41
2.89
3.37
3.85
4.33
4.81
Cumulative
Total Water
Added to Test
Vessel,
gal
6.49
6.98
7.46
7.94
8.42
8.90
9.38
9.86
10.34
10.82
11.31
      a.  These reading are from when the floats were first turned on at 1:50,
      b.  Reading was changed from 1.90 to correspond with the raw data.
not at the +10 min increment of 1:55.
                                                        D-3

-------
Smallest Detection Increment Difference
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
1.50
1.57
1.67
1.75
1.85
1.93
2.01
2.09
2.17
2.24
2.32
Measured
Incremental
Change,
inches

0.079
0.098
0.079
0.098
0.079
0.079
0.079
0.079
0.079
0.079
Second
Generation
Technology
Depth,
inches
**
1.66
1.80
1.89
1.98
2.06
2.14
2.22
2.30
2.37
2.45
Second
Generation
Technology
Incremental
Change,
inches
**

0.140
0.090
0.090
0.080
0.080
0.080
0.080
0.070
0.080
Delta
Incremental
Change,
(Technology
- Measured)
inches
**

0.042
0.011
-0.008
0.001
0.001
0.001
0.001
-0.009
0.001
Shading indicates that no measurement was taken.
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
1.50
1.57
1.67
1.75
1.85
1.93
2.01
2.09
2.17
2.24
2.32
Measured
Incremental
Change,
inches

0.079
0.098
0.079
0.098
0.079
0.079
0.079
0.079
0.079
0.079
First
Generation
Technology
Depth,
inches
**
1.45
1.59
1.69
1.77
1.86
1.94
2.01
2.08
2.14
2.22
First
Generation
Technology
Incremental
Change,
inches
**

0.140
0.100
0.080
0.090
0.080
0.070
0.070
0.060
0.080
Delta
Incremental
Change,
(Technology
- Measured)
inches
**

0.042
0.021
-0.018
0.011
0.001
-0.009
-0.009
-0.019
0.001
**Probes were not enabled.
Shading indicates that no measurement was taken.
                                        D-4

-------
TEST DAY 2
Run Number
3
Test Day
2
Date
9/14/11
Fuel
EO
Fuel Level,
percent
25
Ingress
Method
With splash
Run ID
EO-25-w

Time
Fuel height FFS side (inches)
DP FFS side (mm)
Temp 1 (°C) (corrected)
Temp 2 (°C) (corrected)
Temp 3 (°C) (corrected)
Temp 4 (°C) (corrected)
Ingress rate (ml/min)
Ingress rate determined (ml/min)
Start of incremental test (T=0)
Initial
12:40 PMa
18-3/8
trace
20.7
20.3
NA
NA
200
Mid-Point
1:44 PMa
18-7/16
23
20.8
20.5
NA
NA
200
Final (T=1 00)
3:56 PMa
18-15/16
48
21.0
20.7
NA
NA
200
177b
2:16 PM
       a. The clock was set 37 min fast, however this data has been corrected.
       b. Determined using the average determined flow rate and applying the average % error.

Rotameter Calibration
Observed
Flow Rate,
cm3/min
195
190
180
Volume Collected,
ml
58.0
56.0
53.0
Collection
Time,
sec
20.0
20.0
20.0
Determined
Flow Rate,
ml/min
174.0
168.0
159.0
Error
-11%
-12%
-12%

Average
Standard deviation
Observed
Flow Rate,
cm3/min
188
8
Volume
Collected,
ml
55.7
2.5
Collection
Time,
sec
20.0
0.0
Determined
Flow Rate,
ml/min
167.0
7.5
Error
-11%
0%
                                          D-5

-------
Minimum Detection Height Test
Technology
First Generation
Second Generation
Time at
Alarm
1:44 PMa
1:18 PMa
Observed
Dense Phase
Height,
mm
23
16
Observed
Dense Phase
Height,
inches
0.91
0.63
Technology
Dense Phase
Height,
inches
0.01
0.01
Elapsed Time
to Alarm,
min
64
38
Water
Volume to
Alarm,
ml
1 1 ,349
6,738
Water Volume
to Alarm,
gal
3.00
1.78
      a. The clock was set 37 min fast, however these data have been corrected.



Smallest Detection Increment Test
Elapsed
Time,
min
0
10
20
30
40
50
60
70
80
90
100
Observed
Dense
Phase
Height,
mm
24
27
29.5
32
34.5
37
39.5
41.5
43.5
46
48
Observed
Dense
Phase
Height,
inches
0.94
1.06
1.16
1.26
1.36
1.46
1.56
1.63
1.71
1.81
1.89
First
Generation
Technology,
inches
0.02
0.09
0.20
0.30
0.39
0.49
0.58
0.67
0.76
0.85
0.92
Second
Generation
Technology,
inches
0.35
0.44
0.55
0.65
0.75
0.85
0.93
1.03
1.11
1.19
1.29
Cumulative
Water Added
Since T = 0,
ml
0
1,773
3,546
5,320
7,093
8,866
10,639
12,413
14,186
15,959
17,732
Cumulative
Water Added
Since T = 0,
gal
0.00
0.47
0.94
1.41
1.87
2.34
2.81
3.28
3.75
4.22
4.68
Cumulative
Total Water
Added to Test
Vessel,
gal
3.00
3.47
3.93
4.40
4.87
5.34
5.81
6.28
6.75
7.21
7.68
                                                     D-6

-------
Smallest Detection Increment Difference
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
0.94
1.06
1.16
1.26
1.36
1.46
1.56
1.63
1.71
1.81
1.89
Measured
Incremental
Change,
inches

0.118
0.098
0.098
0.098
0.098
0.098
0.079
0.079
0.098
0.079
Second
Generation
Technology
Depth,
inches
0.35
0.44
0.55
0.65
0.75
0.85
0.93
1.03
1.11
1.19
1.29
Second
Generation
Technology
Incremental
Change,
inches

0.090
0.110
0.100
0.100
0.100
0.080
0.100
0.080
0.080
0.100
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.029
0.012
0.002
0.002
0.002
-0.018
0.021
0.001
-0.018
0.021
Shading indicates that no measurement was taken.
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
0.94
1.06
1.16
1.26
1.36
1.46
1.56
1.63
1.71
1.81
1.89
Measured
Incremental
Change,
inches

0.118
0.098
0.098
0.098
0.098
0.098
0.079
0.079
0.098
0.079
First
Generation
Technology
Depth,
inches
0.02
0.09
0.20
0.30
0.39
0.49
0.58
0.67
0.76
0.85
0.92
First
Generation
Technology
Incremental
Change,
inches

0.066
0.110
0.100
0.090
0.100
0.090
0.090
0.090
0.090
0.070
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.052
0.012
0.002
-0.008
0.002
-0.008
0.011
0.011
-0.008
-0.009
Shading indicates that no measurement was taken.
                                        D-7

-------
TEST DAY 3
Run Number
2
Test Day
3
Date
9/15/11
Fuel
EO
Fuel Level,
percent
25 then 65
Ingress
Method
Dump
Run ID
EO-dump





Time
Fuel height FFS
side (inches)
DP FFS side
(mm/inches)
1st Generation
Reading (inches)
2nd Generation
Reading (inches)
Temp 1 (°C)
Temp 2 (°C)
Temp 3 (°C)
Temp 4 (°C)
Initial-
Water

Dump
(2 gal)
12:10 PM
17-3/4

Trace
0.00
0.10
19.4
19.6
NA
NA
Reading
SMin

After
Dump 1
12:15 PM


18/0.71
0.00
0.11


NA
NA
Reading
10Min

After
Dump 1
12:20 PM


18/0.71
0.00
0.11


NA
NA

Water

Dump 2

12:28 PM


18/0.71
0.18
0.49


NA
NA
Reading
SMin

After
Dump 2
12:33 PM


31 71.22
0.25
0.61


NA
NA
Reading
10Min

After
Dump 2
12:38 PM


31 71.22
0.25
0.61


NA
NA
Mid
Waliipc


Fuel
Dump
12:39 PM
18

31 / 1 .22
0.25
0.61
19.2
19.5
NA
NA
Shading indicates that no measurement was taken.

Time
Fuel height FFS
side (inches)
DP FFS side
(mm/inches)
1st Generation
Reading (inches)
2nd Generation
Reading (inches)
Temp 1 (°C)
Temp 2 (°C)
Temp 3 (°C)
Temp 4 (°C)
Fuel
Dump
1:21 PM

31 71.22
0.26
0.73


Not
recorded
Not
recorded
Reading
SMin
After
Fuel
Dump
1:26 PM

31 71.22
0.27
0.64




Reading
10Min
After
Fuel
Dump
1:31 PM

31 71.22
0.27
0.64




Reading
15Min
After
Fuel
Dump
1:36 PM

31 71.22
0.27
0.64




Reading
20Min
After
Fuel
Dump
1:41 PM

31 71.22
0.27
0.64




Reading
25Min
After
Fuel
Dump
1 :46 PM

31 71.22
0.27
0.64




Reading
SOMin
After
Fuel
Dump
1:51 PM
44-1/2
31 71.22
0.27
0.64
Not
recorded
Not
recorded
Not
recorded
Not
recorded
Shading indicates that no measurement was taken.
                                         D-8

-------
TEST DAY 4
Run Number
8
Test Day
4
Date
9/16/11
Fuel
EO
Fuel Level,
percent
65
Ingress Method
Without splash
Run ID
EO-65-wo

Time
Fuel height FFS side (inches)
DP FFS side (mm)
Temp 1 (°C) (corrected)
Temp 2 (°C) (corrected)
Temp 3 (°C) (corrected)
Temp 4 (°C) (corrected)
Ingress rate (ml/min)
Ingress rate determined (ml/min)
Start of incremental test (T=0)
Initial
12:13 PM
44-5/8
trace
18.5
17.9
19.2
18.8
200
Mid-Point
1:17 PM
44-13/16
23
18.3
17.9
18.8
18.3
200
Final (T=1 00)
3:23 PM
45-1/8
46
18.2
17.8
18.5
18.0
200
179a
1:43 PM
      a. Determined using the average determined flow rate and applying the average % error.



Rotameter Calibration
Observed
Flow Rate,
cm3/min
200
200
200
Volume
Collected,
ml
55.5
63.0
62.0
Collection
Time,
sec
20.3
20.3
19.9
Determined
Flow Rate,
ml/min
164.0
186.2
186.9
Error
-18%
-7%
-7%

Average
Standard deviation
Observed
Flow Rate,
cm3/min
200
0
Volume
Collected,
ml
60.2
4.1
Collection
Time,
sec
20.2
0.2
Determined
Flow Rate,
ml/min
179.1
13.0
Error
-10%
7%
                                        D-9

-------
Minimum Detection Height Test
Technology
First Generation
Second
Generation
Time at
Alarm
1:17 PM
12:50 PM
Observed
Dense Phase
Height,
mm
23
15
Observed
Dense Phase
Height,
inches
0.91
0.59
Technology
Dense Phase
Height,
inches
0.01
0.02a
Elapsed Time
to Alarm,
min
64
37
Water
Volume to
Alarm,
ml
1 1 ,460
6,625
Water Volume
to Alarm,
gal
3.03
1.75
      a.  Reading is not exactly at initial times, however within a couple hundredths of an inch.




Smallest Detection Increment Test
Elapsed
Time,
min
0
10
20
30
40
50
60
70
80
90
100
Observed
Dense
Phase
Height,
mm
23
26
29
31
34
36
38
40
42
44
46
Observed
Dense
Phase
Height,
inches
0.91
1.02
1.14
1.22
1.34
1.42
1.50
1.57
1.65
1.73
1.81
First
Generation
Technology,
inches
0.03
0.08
0.18
0.27
0.37
0.44
0.35
0.63
0.70
0.79
0.87
Second
Generation
Technology,
inches
0.35
0.43
0.53
0.62
0.73
0.82
0.90
0.99
1.06
1.14
1.22
Cumulative
Water Added
Since T = 0,
ml
0
1,791
3,581
5,372
7,162
8,953
10,744
12,534
14,325
16,115
17,906
Cumulative
Water Added
Since T = 0,
gal
0.00
0.47
0.95
1.42
1.89
2.37
2.84
3.31
3.78
4.26
4.73
Cumulative
Total Water
Added to Test
Vessel,
gal
3.03
3.50
3.97
4.45
4.92
5.39
5.87
6.34
6.81
7.28
7.76
                                                      D-10

-------
Smallest Detection Increment Difference
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
0.91
1.02
1.14
1.22
1.34
1.42
1.50
1.57
1.65
1.73
1.81
Measured
Incremental
Change,
inches

0.118
0.118
0.079
0.118
0.079
0.079
0.079
0.079
0.079
0.079
Second
Generation
Technology
Depth,
inches
0.35
0.43
0.53
0.62
0.73
0.82
0.90
0.99
1.06
1.14
1.22
Second
Generation
Technology
Incremental
Change,
inches

0.080
0.100
0.090
0.110
0.090
0.080
0.090
0.070
0.080
0.080
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.038
-0.018
0.011
-0.008
0.011
0.001
0.011
-0.009
0.001
0.001
Shading indicates that no measurement was taken.
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
0.91
1.02
1.14
1.22
1.34
1.42
1.50
1.57
1.65
1.73
1.81
Measured
Incremental
Change,
inches

0.118
0.118
0.079
0.118
0.079
0.079
0.079
0.079
0.079
0.079
First
Generation
Technology
Depth,
inches
0.03
0.08
0.18
0.27
0.37
0.44
0.51
0.63
0.70
0.79
0.87
First
Generation
Technology
Incremental
Change,
inches

0.050
0.100
0.090
0.100
0.070
0.070
0.279
0.070
0.090
0.080
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.068
-0.018
0.011
-0.018
-0.009
-0.009
0.200
-0.009
0.011
0.001
Shading indicates that no measurement was taken.
                                       D-ll

-------
TEST DAY 5 (AM)
Run Number
9
Test Day
5
Date
9/19/11
Fuel
EO
Fuel Level,
percent
65
Ingress
Method
With splash
Run ID
EO-65-w

Time
Fuel Height FFS side (in)
DP FFS side (mm)
Temp 1 (°C) (corrected)
Temp 2 (°C) (corrected)
Temp 3 (°C) (corrected)
Temp 4 (°C) (corrected)
Ingress rate (ml/min)
Ingress rate determined (ml/min)
Start of incremental test (T=0)
Initial
10:14AM
45-1/16
trace
17.9
17.6
18.0
17.6
200
Mid-Point
11:11 AM
45-3/16
21
18.0
17.6
18.0
17.5
200
Final (T=1 00)
1:04 PM
45-7/16
46
18.1
17.8
18.1
17.7
200
183a
11:24 AM
      a. Determined using the average determined flow rate and applying the average % error.




Rotameter Calibration
Observed
Flow Rate,
cm3/min
210
210
200
Volume
Collected,
ml
64.0
61.5
63.5
Collection
Time,
sec
20.0
20.0
20.0
Determined
Flow Rate,
ml/min
192.0
184.5
190.5
Error
-9%
-12%
-5%

Average
Standard deviation
Observed
Flow Rate,
cm3/min
207
6
Volume
Collected,
ml
63.0
1.3
Collection
Time,
sec
20.0
0.0
Determined
Flow Rate,
ml/min
189.0
4.0
Error
-8%
4%
                                       D-12

-------
Minimum Detection Height Test
Technology
First Generation
Second
Generation
Time at
Alarm
11:11 AMa
10:53 AMa
Observed
Dense Phase
Height,
mm
21
16
Observed
Dense Phase
Height,
inches
0.83
0.63
Technology
Dense Phase
Height,
inches
0.01
0.01
Elapsed Time
to Alarm,
min
57
39
Water
Volume to
Alarm,
ml
10,432
7,138
Water Volume
to Alarm,
gal
2.76
1.89
      a. Readings are not exactly at initial times, however within several seconds and a couple hundredths of an inch.




Smallest Detection Increment Test
Elapsed
Time,
min
0
10
20
30
40
50
60
70
80
90
100
Observed
Dense
Phase
Height,
mm
21
24
27
29
31
34
36
39
41
43
46
Observed
Dense
Phase
Height,
inches
0.83
0.94
1.06
1.14
1.22
1.34
1.42
1.54
1.61
1.69
1.81
First
Generation
Technology,
inches
0.00
0.03
0.09
0.20
0.29
0.38
0.45
0.56
0.65
0.74
0.83
Second
Generation
Technology,
inches
0.26
0.33
0.45
0.55
0.65
0.74
0.84
0.93
1.02
1.10
1.18
Cumulative
Water Added
Since T = 0,
ml
0
1,830
3,660
5,491
7,321
9,151
10,981
12,812
14,642
16,472
18,302
Cumulative
Water Added
Since T = 0,
gal
0.00
0.48
0.97
1.45
1.93
2.42
2.90
3.38
3.87
4.35
4.83
Cumulative
Total Water
Added to Test
Vessel,
gal
2.76
3.24
3.72
4.21
4.69
5.17
5.66
6.14
6.62
7.11
7.59
                                                       D-13

-------
Smallest Detection Increment Difference
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
0.83
0.94
1.06
1.14
1.22
1.34
1.42
1.54
1.61
1.69
1.81
Measured
Incremental
Change,
inches

0.118
0.118
0.079
0.079
0.118
0.079
0.118
0.079
0.079
0.118
Second
Generation
Technology
Depth,
inches
0.26
0.33
0.45
0.55
0.65
0.74
0.84
0.93
1.02
1.10
1.18
Second
Generation
Technology
Incremental
Change,
inches

0.070
0.120
0.100
0.100
0.090
0.100
0.090
0.090
0.080
0.080
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.048
0.002
0.021
0.021
-0.028
0.021
-0.028
0.011
0.001
-0.038
Shading indicates that no measurement was taken.
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
0.83
0.94
1.06
1.14
1.22
1.34
1.42
1.54
1.61
1.69
1.81
Measured
Incremental
Change,
inches

0.118
0.118
0.079
0.079
0.118
0.079
0.118
0.079
0.079
0.118
First
Generation
Technology
Depth,
inches
0.00
0.03
0.09
0.20
0.29
0.38
0.45
0.56
0.65
0.74
0.83
First
Generation
Technology
Incremental
Change,
inches

0.030
0.060
0.110
0.090
0.090
0.070
0.110
0.090
0.090
0.090
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.088
-0.058
0.031
0.011
-0.028
-0.009
-0.008
0.011
0.011
-0.028
Shading indicates that no measurement was taken.
                                       D-14

-------
TEST DAY 5 (PM)
Run Number
12
Test Day
5
Date
9/19/2011
Fuel
EO
Fuel Level,
percent
65
Ingress Method
Without splash
Run ID
EO-65-wo-DUP

Time
Fuel height FFS side (inches)
DP FFS side (mm)
Temp 1 (°C) (corrected)
Temp 2 (°C) (corrected)
Temp 3 (°C) (corrected)
Temp 4 (°C) (corrected)
Ingress rate (ml/min)
Ingress rate determined (ml/min)
Start of incremental test (T=0)
Initial
1:55 PM
44-15/16
trace
18.2
17.9
18.2
17.7
200
Mid-Point
2:49 PM
45-1/8
21
18.2
17.9
18.2
17.8
200
Final (T=1 00)
4:35 PM
45-3/8
46
18.4
18.1
18.3
17.8
200
183a
2:55 PM
a. Determined using the average determined flow rate and applying the average % error.



Rotameter Calibration
Observed
Flow Rate,
cm3/min
210
210
200
Volume Collected,
ml
64.0
61.5
63.5
Collection
Time,
sec
20.0
20.0
20.0
Determined
Flow Rate,
ml/min
192.0
184.5
190.5
Error
-9%
-12%
-5%

Average
Standard deviation
Observed
Flow Rate,
cm3/min
207
6
Volume
Collected,
ml
63.0
1.3
Collection
Time,
sec
20.0
0.0
Determined
Flow Rate,
ml/min
189.0
4.0
Error
-8%
4%
                                       D-15

-------
Minimum Detection Height Test
Technology
First Generation
Second
Generation
Time at
Alarm
2:49 PMa
2:29 PMa
Observed
Dense Phase
Height,
mm
21
16
Observed
Dense Phase
Height,
inches
0.83
0.63
Technology
Dense Phase
Height,
inches
0.01
0.02a
Elapsed Time
to Alarm,
min
54
34
Water
Volume to
Alarm,
ml
9,883
6,223
Water Volume
to Alarm,
gal
2.61
1.64
      a. Readings are not exactly at initial times, however within several seconds and a couple hundredths of an inch.




Smallest Detection Increment Test
Elapsed
Time,
min
0
10
20
30
40
50
60
70
80
90
100
Observed
Dense
Phase
Height,
mm
21
25
28
30
32
35
37
39
41
43
46
Observed
Dense
Phase
Height,
inches
0.83
0.98
1.10
1.18
1.26
1.38
1.46
1.54
1.61
1.69
1.81
First
Generation
Technology,
inches
0.01
0.03
0.12
0.22
0.31
0.42
0.50
0.59
0.68
0.77
0.86
Second
Generation
Technology,
inches
0.28
0.35
0.47
0.57
0.67
0.77
0.87
0.95
1.05
1.13
1.21
Cumulative
Water Added
Since T = 0,
ml
0
1,830
3,660
5,491
7,321
9,151
10,981
12,812
14,642
16,472
18,302
Cumulative
Water Added
Since T = 0,
gal
0.00
0.48
0.97
1.45
1.93
2.42
2.90
3.38
3.87
4.35
4.83
Cumulative
Total Water
Added to Test
Vessel,
gal
2.61
3.09
3.58
4.06
4.54
5.03
5.51
6.00
6.48
6.96
7.45
                                                       D-16

-------
Smallest Detection Increment Difference
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
0.83
0.98
1.10
1.18
1.26
1.38
1.46
1.54
1.61
1.69
1.81
Measured
Incremental
Change,
inches

0.157
0.118
0.079
0.079
0.118
0.079
0.079
0.079
0.079
0.118
Second
Generation
Technology
Depth,
inches
0.28
0.35
0.47
0.57
0.67
0.77
0.87
0.95
1.05
1.13
1.21
Second
Generation
Technology
Incremental
Change,
inches

0.070
0.120
0.100
0.100
0.100
0.100
0.080
0.100
0.080
0.080
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.087
0.002
0.021
0.021
-0.018
0.021
0.001
0.021
0.001
-0.038
Shading indicates that no measurement was taken.
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
0.83
0.98
1.10
1.18
1.26
1.38
1.46
1.54
1.61
1.69
1.81
Measured
Incremental
Change,
inches

0.157
0.118
0.079
0.079
0.118
0.079
0.079
0.079
0.079
0.118
First
Generation
Technology
Depth,
inches
0.01
0.03
0.12
0.22
0.31
0.42
0.50
0.59
0.68
0.77
0.86
First
Generation
Technology
Incremental
Change,
inches

0.020
0.090
0.100
0.090
0.110
0.080
0.090
0.090
0.090
0.090
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.137
-0.028
0.021
0.011
-0.008
0.001
0.011
0.011
0.011
-0.028
Shading indicates that no measurement was taken.
                                       D-17

-------
TEST DAY 6
Run Number
4
Test Day
6
Date
9/20/1 1
Fuel
E15
Fuel Level,
percent
25
Ingress Method
Without splash
Run ID
E15-25-WO

Time
Fuel height FFS side (inches)
DP FFS side (mm)
Temp 1 (°C) (corrected)
Temp 2 (°C) (corrected)
Temp 3 (°C) (corrected)
Temp 4 (°C) (corrected)
Ingress rate (ml/min)
Ingress rate determined (ml/min)
Start of incremental test (T=0)
Initial
10:17AM
17-13/16
trace
18.2
17.9
NA
NA
200
Mid-Point
10:58 AM
17-7/8
20
18.4
18.1
NA
NA
200
Final (T=1 00)
12:57 PM
18-1/4
54
19.1
18.8
NA
NA
200
183a
11:17AM
      a. Determined using the average determined flow rate and applying the average % error.



Rotameter Calibration
Observed
Flow Rate,
cm3/min
200
200
200
Volume
Collected,
ml
62
61
60
Collection
Times,
sec
20.0
20.0
20.0
Determined
Flow Rate,
ml/min
186.0
183.0
180.0

Error
-7%
-9%
-10%

Average
Standard deviation
Observed
Flow Rate,
cm3/min
200
0
Volume
Collected,
ml
61.0
1.0
Collection
Time,
sec
20.0
0.0
Determined
Flow Rate,
ml/min
183.0
3.0
Error
-9%
2%
                                        D-18

-------
Minimum Detection Height Test
Technology
First Generation
Second
Generation
Time at
Alarm
10:58 AM
10:44AMa
Observed
Dense Phase
Height,
mm
22
17
Observed
Dense Phase
Height,
inches
0.87
0.67
Technology
Dense Phase
Height,
inches
0.01
0.01
Elapsed Time
to Alarm,
min
41
28
Water
Volume to
Alarm,
ml
7,503
5,051
Water Volume
to Alarm,
gal
1.98
1.33
      a. Readings are not exactly at initial times, however within several seconds and a couple hundredths of an inch.




Smallest Detection Increment Test
Elapsed
Time,
min
0
10
20
30
40
50
60
70
80
90
100
Observed
Dense
Phase
Height,
mm
24
28
31
35
39
42
45
48
51
53
56
Observed
Dense
Phase
Height,
inches
0.94
1.10
1.22
1.38
1.54
1.65
1.77
1.89
2.01
2.09
2.20
First
Generation
Technology,
inches
0.02
0.03
0.13
0.27
0.38
0.50
0.62
0.75
0.86
0.96
1.06a
Second
Generation
Technology,
inches
0.29
0.40
0.55
0.69
0.81
0.94
1.06
1.18
1.29
1.40
1.50a
Cumulative
Water Added
Since T = 0,
ml
0
1,830
3,660
5,490
7,320
9,150
10,980
12,810
14,640
16,470
18,300
Cumulative
Water Added
Since T = 0,
gal
0.00
0.48
0.97
1.45
1.93
2.42
2.90
3.38
3.87
4.35
4.83
Cumulative
Total Water
Added to Test
Vessel,
gal
1.98
2.47
2.95
3.43
3.92
4.40
4.88
5.37
5.85
6.33
6.82
      a. Reading was changed to correspond with the raw data.
                                                        D-19

-------
Smallest Detection Increment Difference
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
0.94
1.10
1.22
1.38
1.54
1.65
1.77
1.89
2.01
2.09
2.20
Measured
Incremental
Change,
inches

0.157
0.118
0.157
0.157
0.118
0.118
0.118
0.118
0.079
0.118
Second
Generation
Technology
Depth,
inches
0.29
0.40
0.55
0.69
0.81
0.94
1.06
1.18
1.29
1.40
1.50
Second
Generation
Technology
Incremental
Change,
inches

0.110
0.150
0.140
0.120
0.130
0.120
0.120
0.110
0.110
0.100
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.047
0.032
-0.017
-0.037
0.012
0.002
0.002
-0.008
0.031
-0.018
Shading indicates that no measurement was taken.
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
0.94
1.10
1.22
1.38
1.54
1.65
1.77
1.89
2.01
2.09
2.20
Measured
Incremental
Change,
inches

0.157
0.118
0.157
0.157
0.118
0.118
0.118
0.118
0.079
0.118
First
Generation
Technology
Depth,
inches
0.02
0.03
0.13
0.27
0.38
0.50
0.62
0.75
0.86
0.96
1.06
First
Generation
Technology
Incremental
Change,
inches

0.010
0.100
0.140
0.110
0.120
0.120
0.130
0.110
0.100
0.100
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.147
-0.018
-0.017
-0.047
0.002
0.002
0.012
-0.008
0.021
-0.018
Shading indicates that no measurement was taken.
                                       D-20

-------
TEST DAY 7 (AM)
Run Number
7
Test Day
7
Date
9/21/11
Fuel
E15
Fuel Level,
percent
25
Ingress
Method
With splash
Run ID
E15-25-W

Time
Fuel height FFS side (inches)
DP FFS side (mm)
Temp 1 (°C) (corrected)
Temp 2 (°C) (corrected)
Temp 3 (°C) (corrected)
Temp 4 (°C) (corrected)
Ingress rate (ml/min)
Ingress rate determined (ml/min)
Start of incremental test (T=0)
Initial
9:36 AM
17
trace
18.9
18.6
NA
NA
200
Mid-Point
10:11 AM
17-1/8
37
19.2
18.9
NA
NA
200
Final (T=1 00)
12:15 PM
17-3/8
82
20.7
20.1
NA
NA
200
176a
10:35 AM
      a. Determined using the average determined flow rate and applying the average % error.



Rotameter Calibration
Observed
Flow Rate,
cm3/min
200
200
200
Volume
Collected,
ml
59.0
58.0
59.0
Collection
Time,
sec
20.0
20.0
20.0
Determined
Flow Rate,
ml/min
177.0
174.0
177.1
Error
-12%
-13%
-11%

Average
Standard deviation
Observed
Flow Rate,
cm3/min
200
0
Volume
Collected,
ml
58.7
0.6
Collection
Time,
sec
20.0
0.0
Determined
Flow Rate,
ml/min
176.0
1.8
Error
-12%
1%
                                       D-21

-------
Minimum Detection Height Test
Technology
First Generation
Second
Generation
Time at
Alarm
10:11 AM
9:56 AMa
Observed
Dense Phase
Height,
mm
37
22
Observed
Dense Phase
Height,
inches
1.46
0.87
Technology
Dense Phase
Height,
inches
0.01
0.02
Elapsed Time
to Alarm,
min
35
20
Water
Volume to
Alarm,
ml
6,161
3,521
Water Volume
to Alarm,
gal
1.63
0.93
      a. Readings are not exactly at initial times, however within several seconds and a couple hundredths of an inch.



Smallest Detection Increment Test
Elapsed
Time,
min
0
10
20
30
40
50
60
70
80
90
100
Observed
Dense
Phase
Height,
mm
37
43
50
56
61
65
69
72
76
79
82
Observed
Dense
Phase
Height,
inches
1.46
1.69
1.97
2.20
2.40
2.56
2.72
2.83
2.99
3.11
3.23
First
Generation
Technology,
inches
0.03
0.07
0.27
0.39
0.57
0.82
0.99
1.12
1.26
1.43
1.59
Second
Generation
Technology,
inches
0.57
0.77
1.00
1.22
1.40
1.59
1.74
1.89
2.04
2.17
2.30
Cumulative
Water Added
Since T = 0,
ml
0
1,760
3,521
5,281
7,041
8,802
10,562
12,322
14,082
15,843
17,603
Cumulative
Water Added
Since T = 0,
gal
0.00
0.47
0.93
1.40
1.86
2.33
2.79
3.26
3.72
4.19
4.65
Cumulative
Total Water
Added to Test
Vessel,
gal
1.63
2.09
2.56
3.02
3.49
3.95
4.42
4.88
5.35
5.81
6.28
                                                       D-22

-------
Smallest Detection Increment Difference
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
1.46
1.69
1.97
2.20
2.40
2.56
2.72
2.83
2.99
3.11
3.23
Measured
Incremental
Change,
inches

0.236
0.276
0.236
0.197
0.157
0.157
0.118
0.157
0.118
0.118
Second
Generation
Technology
Depth,
inches
0.57
0.77
1.00
1.22
1.40
1.59
1.74
1.89
2.04
2.17
2.30
Second
Generation
Technology
Incremental
Change,
inches

0.200
0.230
0.220
0.180
0.190
0.150
0.150
0.150
0.130
0.130
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.036
-0.046
-0.016
-0.017
0.033
-0.007
0.032
-0.007
0.012
0.012
Shading indicates that no measurement was taken.
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
1.46
1.69
1.97
2.20
2.40
2.56
2.72
2.83
2.99
3.11
3.23
Measured
Incremental
Change,
inches

0.236
0.276
0.236
0.197
0.157
0.157
0.118
0.157
0.118
0.118
First
Generation
Technology
Depth,
inches
0.03
0.07
0.27
0.39
0.57
0.82
0.99
1.12
1.26
1.43
1.59
First
Generation
Technology
Incremental
Change,
inches

0.040
0.200
0.120
0.180
0.250
0.170
0.130
0.140
0.170
0.160
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.196
-0.076
-0.116
-0.017
0.093
0.013
0.012
-0.017
0.052
0.042
Shading indicates that no measurement was taken.
                                       D-23

-------
TEST DAY 7 (PM)
Run Number
13
Test Day
7
Date
9/21/11
Fuel
E15
Fuel Level,
percent
25
Ingress Method
Without splash
Run ID
E15-25-WO-DUP

Time
Fuel height FFS side (inches)
DP FFS side (mm)
Temp 1 (°C) (corrected)
Temp 2 (°C) (corrected)
Temp 3 (°C) (corrected)
Temp 4 (°C) (corrected)
Ingress rate (ml/min)
Ingress rate determined (ml/min)
Start of incremental test (T=0)
Initial
3:08 PM
16-1/8
trace
20.7
19.4
NA
NA
200
Mid-Point
4:02 PM
16-5/16
28
20.7
20.4
NA
NA
200
Final (T=1 00)
6:09 PM
16-11/16
58
21.5
21.1
NA
NA
200
176a
4:29 PM
a. Determined using the average determined flow rate and applying the average % error.




Rotameter Calibration
Observed
Flow Rate,
cm3/min
200
200
200
Volume Collected,
ml
59.0
58.0
59.0
Collection
Time,
sec
20.0
20.0
20.0
Determined
Flow Rate,
ml/min
177.0
174.0
177.1
Error
-12%
-13%
-11%

Average
Standard deviation
Observed
Flow Rate,
cm3/min
200
0
Volume
Collected,
ml
58.7
0.6
Collection
Time,
sec
20.0
0.0
Determined
Flow Rate,
ml/min
176.0
1.8
Error
-12%
1%
                                       D-24

-------
Minimum Detection Height Test
Technology
First Generation
Second
Generation
Time at
Alarm
4:02 PM
3:39 PMa
Observed
Dense Phase
Height,
mm
28
18
Observed
Dense Phase
Height,
inches
1.10
0.71
Technology
Dense Phase
Height,
inches
0.01
0.01
Elapsed Time
to Alarm,
min
54
31
Water
Volume to
Alarm,
ml
9,506
5,457
Water Volume
to Alarm,
gal
2.51
1.44
      a. Readings are not exactly at initial times, however within several seconds and a couple hundredths of an inch.




Smallest Detection Increment Test
Elapsed
Time,
min
0
10
20
30
40
50
60
70
80
90
100
Observed
Dense
Phase
Height,
mm
28
32
35
39
42
45
48
51
54
57
58
Observed
Dense
Phase
Height,
inches
1.10
1.26
1.38
1.54
1.65
1.77
1.89
2.01
2.13
2.24
2.28
First
Generation
Technology,
inches
0.03
0.11
0.25
0.37
0.50
0.62
0.74
0.86
0.97
1.08
1.16
Second
Generation
Technology,
inches
0.44
0.53
0.68
0.82
0.95
1.06
1.18
1.29
1.40
1.52
1.62
Cumulative
Water Added
Since T = 0,
ml
0
1,760
3,521
5,281
7,041
8,802
10,562
12,322
14,082
15,843
17,603
Cumulative
Water Added
Since T = 0,
gal
0.00
0.47
0.93
1.40
1.86
2.33
2.79
3.26
3.72
4.19
4.65
Cumulative
Total Water
Added to Test
Vessel,
gal
2.51
2.98
3.44
3.91
4.37
4.84
5.30
5.77
6.23
6.70
7.16
                                                       D-25

-------
Smallest Detection Increment Difference
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
1.10
1.26
1.38
1.54
1.65
1.77
1.89
2.01
2.13
2.24
2.28
Measured
Incremental
Change,
inches

0.157
0.118
0.157
0.118
0.118
0.118
0.118
0.118
0.118
0.039
Second
Generation
Technology
Depth,
inches
0.44
0.53
0.68
0.82
0.95
1.06
1.18
1.29
1.40
1.52
1.62
Second
Generation
Technology
Incremental
Change,
inches

0.090
0.150
0.140
0.130
0.110
0.120
0.110
0.110
0.120
0.100
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.067
0.032
-0.017
0.012
-0.008
0.002
-0.008
-0.008
0.002
0.061
Shading indicates that no measurement was taken.
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
1.10
1.26
1.38
1.54
1.65
1.77
1.89
2.01
2.13
2.24
2.28
Measured
Incremental
Change,
inches

0.157
0.118
0.157
0.118
0.118
0.118
0.118
0.118
0.118
0.039
First
Generation
Technology
Depth,
inches
0.03
0.11
0.25
0.37
0.50
0.62
0.74
0.86
0.97
1.08
1.16
First
Generation
Technology
Incremental
Change,
inches

0.080
0.140
0.120
0.130
0.120
0.120
0.120
0.110
0.110
0.080
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.077
0.022
-0.037
0.012
0.002
0.002
0.002
-0.008
-0.008
0.041
Shading indicates that no measurement was taken.
                                       D-26

-------
TEST DAY 8
Run
Number
5
Test
Day
8
Date
9/22/201 1
Fuel
Flex Fuel
Fuel Level,
percent
25
Ingress
Method
Without
splash
Run ID
Flex-25-wo

Time
Fuel height FFS side (inches)
DP FFS side (inches)
Temp 1 (°C) (corrected)
Temp 2 (°C) (corrected)
Temp 3 (°C) (corrected)
Temp 4 (°C) (corrected)
Ingress rate (ml/min)
Ingress rate determined (ml/min)
Start of incremental test (T=0)
Initial
9:35 AM
17-3/8
Trace
18.2
18.0
NA
NA
200
Mid-Point
NA
NA
NA
NA
NA
NA
NA
NA
Final (T=1 00)
12:35 PM
17-15/16
17-3/43
19.5
19.1
NA
NA
200
153b
Not conducted
a.  Phase separation is not clearly defined and could just be mixing.
b.  Determined using the average determined flow rate and applying the average % error.

Rotameter Calibration
Observed
Flow Rate,
cm3/min
200
200
200
Volume
Collected,
ml
50.0
51.0
52.0
Collection
Times,
sec
20.0
20.0
20.0
Determined
Flow Rate,
ml/min
150.0
153.0
156.0
Error
-25%
-24%
-22%

Average
Standard deviation
Observed
Flow Rate,
cm3/min
200
0
Volume
Collected,
ml
51.0
1.0
Collection
Time,
sec
20.0
0.0
Determined
Flow Rate,
ml/min
153.0
3.0
Error
-24%
2%
                                          D-27

-------
Minimum Detection Height Test

Technology

First Generation
Generation

Time at
Alarm


Observed
Dense Phase
Height,
mm

Observed
Dense
Phase
Height,
inches

Technology
Dense Phase
Height,
inches


Elapsed
Time to
Alarm,
min


Water
Volume to
Alarm,
ml


Water Volume
to Alarm,
gal


Water Added at
Test
Termination,
gal

Did not respond
9:50 AMa
0
0
0.11b
15
2,295
0.61
7.28
a. This float began floating when the E85 was placed in the test vessel. Even after the float was manually pushed to the bottom of the tank the
float was still reading 10 inches. This reading was actually taken when the float first moved from the .10 in up to .11  in within a min.
b. Phase separation is not clearly defined and could just be mixing.
Smallest Detection Increment Test
                                               TEST NOT CONDUCTED
                                                         D-28

-------
TEST DAY 9
Run Number
19
Test Day
9
Date
9/23/1 1
Fuel
flex
fuel
Fuel Level,
percent
25 then 65
Ingress
Method
Dump
Run ID
Flex-dump

Time
Fuel height FFS
side (inches)
DP FFS side
(inches)
1st Generation
Reading (inches)
2nd Generation
Reading (inches)
Temp 1 (°C)
Temp 2 (°C)
Temp 3 (°C)
Temp 4 (°C)
Initial-
Water
Dump
(4 gal)
10:01 AM
16-1/4
0
0.03

18.8
18.2
NA
NA
Reading
16Min
After
Dump 1
10:17AM

2 to 3 in
0.03



NA
NA
Water Dump 2
(2 gal)
10:22 AM


0.01



NA
NA
Mid Values
Before Fuel
Dump
10:26 AM
16-5/8
3 to 4 in
0.01

17.7
18.0
NA
NA
Shading indicates that no measurement was taken.

Time
Fuel height FFS
side (inches)
DP FFS side
(inches)
1st Generation
Reading (inches)
2nd Generation
Reading (inches)
Temp 1 (°C)
Temp 2 (°C)
Temp 3 (°C)
Temp 4 (°C)
Fuel Dump
11:20 AM
16-5/8

0.30
12.38




Reading 30
Min After
Fuel Dump

44-7/16

0.00
40.22
18.3
18.7
18.8
19.4
Shading indicates that no measurement was taken.
                                        D-29

-------
TEST DAY 10
Run Number
6
Test Day
10
Date
9/26/201 1
Fuel
Flex Fuel
Fuel Level,
percent
25
Ingress
Method
With splash
Run ID
Flex-25-w

Time
Fuel height FFS side (inches)
DP FFS side (mm)
Temp 1 (°C) (corrected)
Temp 2 (°C) (corrected)
Temp 3 (°C) (corrected)
Temp 4 (°C) (corrected)
Ingress rate (ml/min)
Ingress rate determined (ml/min)
Start of incremental test (T=0)
Initial
9:51 AM
16-15/16
0
17.9
17.7
NA
NA
200
Mid-Point
NA
NA
NA
NA
NA
NA
NA
NA
Final (T=1 00)
12:52 PM
17-3/8
See Note 1
18.1
17.9
NA
NA
200
160a
Not conducted
1.  Phase separation is not clearly defined and could just be mixing.
a.  Determined using the average determined flow rate and applying the average % error.

Rotameter Calibration
Observed
Flow Rate,
cm3/min
200
200
200
Volume
Collected,
ml
53.0
54.0
53.0
Collection
Time,
sec
20.0
20.0
20.0
Determined
Flow Rate,
ml/min
159.0
162.0
159.0
Error
-21%
-19%
-21%

Average
Standard deviation
Observed
Flow Rate,
cm3/min
200
0
Volume
Collected,
ml
53.3
0.6
Collection
Time,
sec
20.0
0.0
Determined
Flow Rate,
ml/min
160.0
1.7
Error
-20%
1%
                                         D-30

-------
Minimum Detection Height Test
Technology
First Generation
Second
Generation
Time at
Alarm
Observed
Dense
Phase
Height,
mm
Observed
Dense
Phase
Height,
inches
Technology
Dense Phase
Height,
inches
Elapsed
Time to
Alarm,
min
Water
Volume to
Alarm,
ml
Water Volume
to Alarm,
gal
Did not respond
12:25 PMa
b
b
13.18
154
24,640
6.51
Water Added at
Test
Termination,
gal
7.65
a. This float began floating when the E85 was placed in the test vessel. Even after the float was manually pushed to the bottom of the tank the
float was still reading 10 inches. This reading was actually taken when the float first moved from the .10 in up to .11  in within a min.
b Phase separation is not clearly defined and could just be mixing
Smallest Detection Increment Test
                                               TEST NOT CONDUCTED
                                                         D-31

-------
TEST DAY 11
Run Number
14
Test Day
11
Date
9/27/201 1
Fuel
E15
Fuel Level,
percent
65
Ingress
Method
With splash
Run ID
E15-65-W

Time
Fuel height FFS side (inches)
DP FFS side (mm)
Temp 1 (°C) (corrected)
Temp 2 (°C) (corrected)
Temp 3 (°C) (corrected)
Temp 4 (°C) (corrected)
Ingress rate (ml/min)
Ingress rate determined (ml/min)
Start of incremental test (T=0)
Initial
9:54 AM
45-3/16
trace
16.2
15.6
16.6
16.0
200
Mid-Point
10:43 AM
45-5/16
41
16.3
15.7
16.5
15.9
200
Final (T=1 00)
1:04 PM
459/16
95
16.9
16.5
16.9
16.3
200
152a
1 1 :24 AM
 Determined using the average determined flow rate and applying the average % error




Rotameter Calibration
Observed
Flow Rate,
cm3/min
200
200
200
Volume Collected,
ml
50.5
50.5
50.5
Collection
Time,
sec
20.0
20.0
20.0
Determined
Flow Rate,
ml/min
151.5
151.5
151.5
Error
-24%
-24%
-24%

Average
Standard deviation
Observed
Flow Rate,
cm3/min
200
0
Volume
Collected,
ml
50.5
0.0
Collection
Time,
sec
20.0
0.0
Determined
Flow Rate,
ml/min
151.5
0.0
Error
-24%
0%
                                       D-32

-------
Minimum Detection Height Test
Technology
First Generation
Second
Generation
Time at
Alarm
10:43 AMa
10:18 AMa
Observed
Dense Phase
Height,
mm
41
22
Observed
Dense Phase
Height,
inches
1.61
0.87
Technology
Dense Phase
Height,
inches
0.03
0.01
Elapsed Time
to Alarm,
min
49
24
Water
Volume to
Alarm,
ml
7,424
3,636
Water Volume
to Alarm,
gal
1.96
0.96
      a. Readings are not exactly at initial times, however within several seconds and a couple hundredths of an inch.




Smallest Detection Increment Test
Elapsed
Time,
min
0
10
20
30
40
50
60
70
80
90
100
Observed
Dense
Phase
Height,
mm
41
48
54
60
67
70
76
81
87
90
95
Observed
Dense
Phase
Height,
inches
1.61
1.89
2.13
2.36
2.64
2.76
2.99
3.19
3.43
3.54
3.74
First
Generation
Technology,
inches
0.03
0.04
0.21
0.37
0.53
0.78
0.93
1.11
1.26
1.45
1.62
Second
Generation
Technology,
inches
0.64
0.81
1.02
1.24
1.46
1.66
1.89
2.09
2.26
2.47
2.64
Cumulative
Water Added
Since T = 0,
ml
0
1,515
3,030
4,545
6,060
7,575
9,090
10,605
12,120
13,635
15,150
Cumulative
Water Added
Since T = 0,
gal
0.00
0.40
0.80
1.20
1.60
2.00
2.40
2.80
3.20
3.60
4.00
Cumulative
Total Water
Added To Test
Vessel,
gal
1.96
2.36
2.76
3.16
3.56
3.96
4.36
4.76
5.16
5.56
5.96
                                                       D-33

-------
Smallest Detection Increment Difference
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
1.61
1.89
2.13
2.36
2.64
2.76
2.99
3.19
3.43
3.54
3.74
Measured
Incremental
Change,
inches

0.276
0.236
0.236
0.276
0.118
0.236
0.197
0.236
0.118
0.197
Second
Generation
Technology
Depth,
inches
0.64
0.81
1.02
1.24
1.46
1.66
1.89
2.09
2.26
2.47
2.64
Second
Generation
Technology
Incremental
Change,
inches

0.170
0.210
0.220
0.220
0.200
0.230
0.200
0.170
0.210
0.170
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.106
-0.026
-0.016
-0.056
0.082
-0.006
0.003
-0.066
0.092
-0.027
Shading indicates that no measurement was taken.
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
1.61
1.89
2.13
2.36
2.64
2.76
2.99
3.19
3.43
3.54
3.74
Measured
Incremental
Change,
inches

0.276
0.236
0.236
0.276
0.118
0.236
0.197
0.236
0.118
0.197
First
Generation
Technology
Depth,
inches
0.03
0.04
0.21
0.37
0.53
0.78
0.93
1.11
1.26
1.45
1.62
First
Generation
Technology
Incremental
Change,
inches

0.010
0.170
0.160
0.160
0.250
0.150
0.180
0.150
0.190
0.170
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.266
-0.066
-0.076
-0.116
0.132
-0.086
-0.017
-0.086
0.072
-0.027
Shading indicates that no measurement was taken.
                                       D-34

-------
TEST DAY 12 (AM)
Run Number
16
Test Day
12
Date
9/28/1 1
Fuel
E15
Fuel Level,
percent
25 then 65
Ingress
Method
Dump
Run ID
E15-dump

Time
Fuel height FFS
side (inches)
DP FFS side
(mm/inches)
1st Generation
Reading (inches)
2nd Generation
Reading (inches)
Temp 1 (°C)
Temp 2 (°C)
Temp 3 (°C)
Temp 4 (°C)
Initial-Water
Dump
(2 gal)
8:22 AM
17-7/16
31 71.22
0.00
0.00
15.6
15.7
NA
NA
Reading 5
Min After
Dump 1
8:27 AM

32 / 1 .26
0.03
0.42


NA
NA
Reading
10 Min
After
Dump 1
8:32 AM

32 / 1 .26
0.03
0.42


NA
NA
Mid Values
Before Fuel
Dump
not recorded

32 / 1 .26
0.03
0.42
15.8
15.8
NA
NA
Shading indicates that no measurement was taken.

Time
Fuel height FFS
side (inches)
DP FFS side
(mm/inches)
1st Generation
Reading (inches)
2nd Generation
Reading (inches)
Temp 1 (°C)
Temp 2 (°C)
Temp 3 (°C)
Temp 4 (°C)
Fuel
Dump
9:12 AM
44-8/16
32 / 1 .26
0.00
0.00


NA
NA
Reading
5 Min
After
Fuel
Dump
9:17 AM

not
recorded
0.00
0.00




Reading
10 Min
After
Fuel
Dump
9:22 AM

965 / 38a
0.00
0.00




Reading
15 Min
After
Fuel
Dump
9:27 AM

965 / 38a
0.00
0.00




Reading
20 Min
After
Fuel
Dump
9:32 AM

990 / 39a
0.00
0.00




Reading
27 Min
After
Fuel
Dump
9:39 AM

990 / 39a
0.00
0.00




Reading
30 Min
After
Fuel
Dump
9:42 AM
44-7/16
990 / 39a
0.00
0.00
16.5
16.6
16.5
16.6
a. Approximate values due to the dense phase being poorly defined.
Shading indicates that no measurement was taken.
                                         D-35

-------
TESTDAY12(PM)
Run
Number
10
Test Day
12
Date
9/28/201 1
Fuel
EO
Fuel Level,
percent
25
Ingress Method
Without splash
Run ID
EO-25-wo-DUP

Time
Fuel height FFS side (inches)
DP FFS side (mm)
Temp 1 (°C) (corrected)
Temp 2 (°C) (corrected)
Temp 3 (°C) (corrected)
Temp 4 (°C) (corrected)
Ingress rate (ml/min)
Ingress rate determined (ml/min)
Start of incremental test (T=0)
Initial
12:02 PM
16-7/8
0
15.2
14.5
NA
NA
220
Mid-Point
1:09 PM
17-1/16
23
15.4
14.8
NA
NA
220
Final (T=1 00)
2:58 PM
not recorded
46
15.6
14.9
NA
NA
220
181a
1:18 PM
a. Determined using the average determined flow rate and applying the average % error.



Rotameter Calibration
Observed
Flow Rate,
cm3/min
230
230
230
Volume Collected,
ml
63.0
63.0
63.0
Collection
Time,
sec
20.0
20.0
20.0
Determined
Flow Rate,
ml/min
189.0
189.0
189.0
Error
-18%
-18%
-18%

Average
Standard deviation
Observed
Flow Rate,
cm3/min
230
0
Volume
Collected,
ml
63.0
0.0
Collection
Time,
sec
20.0
0.0
Determined
Flow Rate,
ml/min
189.0
0.0
Error
-18%
0%
                                       D-36

-------
Minimum Detection Height Test
Technology
First Generation
Second
Generation
Time at
Alarm
1:08 PMa
12:41 PMa
Observed
Dense Phase
Height,
mm
23
16
Observed
Dense Phase
Height,
inches
0.91
0.63
Technology
Dense Phase
Height,
inches
0.01a
0.01a
Elapsed Time
to Alarm,
min
66
39
Water
Volume to
Alarm,
ml
11,932
7,051
Water Volume
to Alarm,
gal
3.15
1.86
a. Readings are not exactly at initial times, however within several seconds and a couple hundredths of an inch.




Smallest Detection Increment Test
Elapsed
Time,
min
0
10
20
30
40
50
60
70
80
90
100
Observed
Dense
Phase
Height,
mm
23
26
29
31
33
36
38
40
42
44
46
Observed
Dense
Phase
Height,
inches
0.91
1.02
1.14
1.22
1.30
1.42
1.50
1.57
1.65
1.73
1.81
First
Generation
Technology,
inches
0.02
0.09
0.19
0.29
0.39
0.47
0.56
0.64
0.73
0.82
0.90
Second
Generation
Technology,
inches
0.36
0.44
0.54
0.64
0.73
0.82
0.91
1.00
1.08
1.16
1.24
Cumulative
Water Added
Since T = 0,
ml
0
1,808
3,616
5,423
7,231
9,039
10,847
12,655
14,463
16,270
18,078
Cumulative
Water Added
Since T = 0,
gal
0.00
0.48
0.96
1.43
1.91
2.39
2.87
3.34
3.82
4.30
4.78
Cumulative
Total Water
Added to Test
Vessel,
gal
3.20
3.68
4.15
4.63
5.11
5.59
6.07
6.54
7.02
7.50
7.98
                                                       D-37

-------
Smallest Detection Increment Difference
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
0.91
1.02
1.14
1.22
1.30
1.42
1.50
1.57
1.65
1.73
1.81
Measured
Incremental
Change,
inches

0.118
0.118
0.079
0.079
0.118
0.079
0.079
0.079
0.079
0.079
Second
Generation
Technology
Depth,
inches
0.36
0.44
0.54
0.64
0.73
0.82
0.91
1.00
1.08
1.16
1.24
Second
Generation
Technology
Incremental
Change,
inches

0.080
0.100
0.100
0.090
0.090
0.090
0.090
0.080
0.080
0.080
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.038
-0.018
0.021
0.011
-0.028
0.011
0.011
0.001
0.001
0.001
Shading indicates that no measurement was taken.
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
0.91
1.02
1.14
1.22
1.30
1.42
1.50
1.57
1.65
1.73
1.81
Measured
Incremental
Change,
inches

0.118
0.118
0.079
0.079
0.118
0.079
0.079
0.079
0.079
0.079
First
Generation
Technology
Depth,
inches
0.02
0.09
0.19
0.29
0.39
0.47
0.56
0.64
0.73
0.82
0.90
First
Generation
Technology
Incremental
Change,
inches

0.070
0.100
0.100
0.100
0.080
0.090
0.080
0.090
0.090
0.080
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.048
-0.018
0.021
0.021
-0.038
0.011
0.001
0.011
0.011
0.001
Shading indicates that no measurement was taken.
                                       D-38

-------
TEST DAY 13
Run Number
17
Test Day
13
Date
9/29/201 1
Fuel
E15
Fuel Level,
percent
65
Ingress Method
Without splash
Run ID
E15-65-WO

Time
Fuel height FFS side (inches)
DP FFS side (mm)
Temp 1 (°C) (corrected)
Temp 2 (°C) (corrected)
Temp 3 (°C) (corrected)
Temp 4 (°C) (corrected)
Ingress rate (ml/min)
Ingress rate determined (ml/min)
Start of incremental test (T=0)
Initial
9:48 AM
45-3/8
0
15.6
14.9
15.9
15.4
220
Mid-Point
10:25 AM
45-7/16
32
15.7
15.1
15.9
15.2
220
Final (T=1 00)
12:29 PM
45-3/4
80
16.4
15.8
16.4
15.7
220
188a
10:49 AM
a. Determined using the average determined flow rate and applying the average % error.




Rotameter Calibration
Observed
Flow Rate,
cm3/min
250
250
250
Volume Collected,
ml
71.0
71.0
72.0
Collection
Time,
sec
20.0
20.0
20.0
Determined
Flow Rate,
ml/min
213.0
213.0
216.0
Error
-15%
-15%
-14%

Average
Standard deviation
Observed
Flow Rate,
cm3/min
250
0
Volume
Collected,
ml
71.3
0.6
Collection
Time,
sec
20.0
0.0
Determined
Flow Rate,
ml/min
214.0
1.7
Error
-14%
1%
                                        D-39

-------
Minimum Detection Height Test
Technology
First Generation
Second
Generation
Time At
Alarm
10:25 AMa
10:09 AMa
Observed
Dense Phase
Height,
mm
32
21
Observed
Dense Phase
Height,
inches
1.26
0.83
Technology
Dense Phase
Height,
inches
0.01
0.04a
Elapsed Time
to Alarm,
min
37
21
Water
Volume to
Alarm,
ml
6,968
3,955
Water Volume
to Alarm,
gal
1.84
1.04
a. Readings are not exactly at initial times, however within several seconds and a couple hundredths of an inch.




Smallest Detection Increment Test
Elapsed
Time,
min
0
10
20
30
40
50
60
70
80
90
100
Observed
Dense
Phase
Height,
mm
33
39
44.5
50
55
60
64
68
72
75
80
Observed
Dense
Phase
Height,
inches
1.30
1.54
1.75
1.97
2.17
2.36
2.52
2.68
2.83
2.95
3.15
First
Generation
Technology,
inches
0.04
0.13
0.32
0.51
0.70
0.88
1.03
1.19
1.33
1.47
1.61
Second
Generation
Technology,
inches
0.51
0.69
0.91
1.09
1.28
1.42
1.61
1.78
1.93
2.08
2.22
Cumulative
Water Added
Since T = 0,
ml
0
1,883
3,766
5,650
7,533
9,416
11,299
13,182
15,066
16,949
18,832
Cumulative
Water Added
Since T = 0,
gal
0.00
0.50
0.99
1.49
1.99
2.49
2.98
3.48
3.98
4.48
4.97
Cumulative
Total Water
Added To Test
Vessel,
gal
1.84
2.34
2.84
3.33
3.83
4.33
4.83
5.32
5.82
6.32
6.82
                                                       D-40

-------
Smallest Detection Increment Difference
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
1.30
1.54
1.75
1.97
2.17
2.36
2.52
2.68
2.83
2.95
3.15
Measured
Incremental
Change,
inches

0.236
0.215
0.219
0.197
0.197
0.157
0.157
0.157
0.118
0.197
Second
Generation
Technology
Depth,
inches
0.51
0.69
0.91
1.09
1.28
1.42
1.61
1.78
1.93
2.08
2.22
Second
Generation
Technology
Incremental
Change,
inches

0.180
0.220
0.180
0.190
0.140
0.190
0.170
0.150
0.150
0.140
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.056
0.005
-0.039
-0.007
-0.057
0.033
0.013
-0.007
0.032
-0.057
Shading indicates that no measurement was taken.
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
1.30
1.54
1.75
1.97
2.17
2.36
2.52
2.68
2.83
2.95
3.15
Measured
Incremental
Change,
inches

0.236
0.215
0.219
0.197
0.197
0.157
0.157
0.157
0.118
0.197
First
Generation
Technology
Depth,
inches
0.04
0.13
0.32
0.51
0.70
0.88
1.03
1.19
1.33
1.47
1.61
First
Generation
Technology
Incremental
Change,
inches

0.090
0.190
0.190
0.190
0.180
0.150
0.160
0.140
0.140
0.140
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.146
-0.025
-0.029
-0.007
-0.017
-0.007
0.003
-0.017
0.022
-0.057
Shading indicates that no measurement was taken.
                                       D-41

-------
TEST DAY 14
Run
Number
18
Test Day
14
Date
9/30/201 1
Fuel
E15
Fuel Level,
percent
65
Ingress
Method
With splash
Run ID
E15-65-W-DUP

Time
Fuel height FFS side (inches)
DP FFS side (mm)
Temp 1 (°C) (corrected)
Temp 2 (°C) (corrected)
Temp 3 (°C) (corrected)
Temp 4 (°C) (corrected)
Ingress rate (ml/min)
Ingress rate determined (ml/min)
Start of incremental test (T=0)
Initial
9:50 AM
45-3/8
0
15.2
14.5
15.6
14.8
220
Mid-Point
10:31 AM
45-7/16
38
15.2
14.4
15.5
14.7
220
Final (T=1 00)
12:47 PM
45-11/16
93
15.3
14.6
15.5
14.6
220
156a
11:07 AM
a. Determined using the average determined flow rate and applying the average % error.




Rotameter Calibration
Observed
Flow Rate,
cm3/min
220
220
220
Volume Collected,
ml
52.0
52.0
52.0
Collection
Time,
sec
20.0
20.0
20.0
Determined
Flow Rate,
ml/min
156.0
156.0
156.0
Error
-29%
-29%
-29%

Average
Standard deviation
Observed
Flow Rate,
cm3/min
220
0
Volume
Collected,
ml
52.0
0.0
Collection
Times,
sec
20.0
0.0
Determined
Flow Rate,
ml/min
156.0
0.0
Error
-29%
0%
                                        D-42

-------
Minimum Detection Height Test
Technology
First Generation
Second
Generation
Time at
Alarm
10:31 AM
10:11 AMa
Observed
Dense Phase
Height,
mm
38
22
Observed
Dense Phase
Height,
inches
1.50
0.87
Technology
Dense Phase
Height,
inches
0.02
0.02
Elapsed Time
to Alarm,
min
41
21
Water
Volume to
Alarm,
ml
6,396
3,276
Water Volume
to Alarm,
gal
1.69
0.87
a. Readings are not exactly at initial times, however within several seconds and a couple hundredths of an inch.




Smallest Detection Increment Test
Elapsed
Time,
min
0
10
20
30
40
50
60
70
80
90
100
Observed
Dense
Phase
Height,
mm
38
45
50
56
62
68
73
78
83
88
93
Observed
Dense
Phase
Height,
inches
1.50
1.77
1.97
2.20
2.44
2.68
2.87
3.07
3.27
3.46
3.66
First
Generation
Technology,
inches
0.03
0.05
0.22
0.24
0.53
0.67
0.92
1.09
1.24
1.43
1.60
Second
Generation
Technology,
inches
0.61
0.76
1.01
1.23
1.43
1.62
1.84
2.04
2.23
2.43
2.60
Cumulative
Water Added
Since T = 0,
ml
0
1,560
3,120
4,680
6,240
7,800
9,360
10,920
12,480
14,040
15,600
Cumulative
Water Added
Since T = 0,
gal
0.00
0.41
0.82
1.24
1.65
2.06
2.47
2.88
3.30
3.71
4.12
Cumulative
Total Water
Added to Test
Vessel,
gal
1.69
2.10
2.51
2.93
3.34
3.75
4.16
4.57
4.99
5.40
5.81
                                                       D-43

-------
Smallest Detection Increment Difference
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
1.50
1.77
1.97
2.20
2.44
2.68
2.87
3.07
3.27
3.46
3.66
Measured
Incremental
Change,
inches

0.276
0.197
0.236
0.236
0.236
0.197
0.197
0.197
0.197
0.197
Second
Generation
Technology
Depth,
inches
0.61
0.76
1.01
1.23
1.43
1.62
1.84
2.04
2.23
2.43
2.60
Second
Generation
Technology
Incremental
Change,
inches

0.150
0.250
0.220
0.200
0.190
0.220
0.200
0.190
0.200
0.170
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.126
0.053
-0.016
-0.036
-0.046
0.023
0.003
-0.007
0.003
-0.027
Shading indicates that no measurement was taken.
Increment
Number
Time 0
1
2
3
4
5
6
7
8
9
10
Measured
Depth,
inches
1.50
1.77
1.97
2.20
2.44
2.68
2.87
3.07
3.27
3.46
3.66
Measured
Incremental
Change,
inches

0.276
0.197
0.236
0.236
0.236
0.197
0.197
0.197
0.197
0.197
First
Generation
Technology
Depth,
inches
0.03
0.05
0.22
0.24
0.53
0.67
0.92
1.09
1.24
1.43
1.60
First
Generation
Technology
Incremental
Change,
inches

0.020
0.170
0.020
0.290
0.140
0.250
0.170
0.150
0.190
0.170
Delta
Incremental
Change,
(Technology
- Measured)
inches

-0.256
-0.027
-0.216
0.054
-0.096
0.053
-0.027
-0.047
-0.007
-0.027
Shading indicates that no measurement was taken.
                                       D-44

-------