 

LIBRARY
Michigan State
University

 

 

 

PLACE It RETURN BOXtoromavothb Mount“ yuxncord.
TO AVOID FINES Mum on or baton duo duo.

DATE DUE DATE DUE DATE DUE

      

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

usu IoAn Affirmative man/Equal Opponunlty Int-tum
W

”3-9.1

 

EVALUATION OF CONTROLLED FREEZING TO REMOVE TRAPPED

RESIDUAL NAPL

by

Craig A Lehner

A THESIS

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of
MASTER OF SCIENCE

Department of Civil and Environmental Engineering

1995

ABSTRACT
EVALUATION OF CONTROLLED FREEZING TO REMOVE TRAPPED
RESIDUAL N APL
by
Craig A. Lehner

Groundwater contamination continues to pose a threat to the drinking water supply.
Contaminants, namely non-aqueous phase liquids (NAPLs), seep into the soil from
hazardous waste sites and underground storage tanks. Once in the soil, 10 to 50 percent
of the contaminant remains trapped in both the saturated and unsaturated soil. This
trapped residual contaminant then slowly and continually dissolves, contaminating the
groundwater. No current methods for removing this trapped residual contaminant are
100 percent eﬁ‘ective and cost eﬂ'ective. Thus, new innovative remedial techniques are
needed to eﬂ‘ectively address this problem.

The present study focuses on removing trapped residual through controlled
freezing. It is also focused on gathering frozen and unfrozen gradient data with and
without dodecane, the representative NAPL used in the bench scale study. For the
experiments, a cell was developed using glass beads to represent the porous media,
thermistors within the cell for temperature measurement, and ﬁ'eeze pipes placed

vertically on the exterior of the cell to create the freeze front. The experiments consisted

of setting up a water saturated only cell and also a water-residual NAPL saturated cell.
The freeze pipes were turned on, while simultaneously starting a ﬂow of water into the
cell to ﬂush any ﬁ'eed residual away from the freeze front. Periodic temperature
measurements were taken throughout the cell to obtain gradient values andthe amount of
trapped residual removed was measured to determine the effectiveness of this procedure.
The data from the gradients revealed two main trends. The two main trends were:
1) the frozen gradient was greater than the unfrozen gradient and 2) the unfrozen to
frozen gradient ratio remained fairly constant throughout all of the experimental runs.
The results of the removal eﬂ‘ectiveness varied greatly, ranging ﬁ'om 76 percent to 26
percent removal effectiveness. The results clearly demonstrate that this method of
trapped residual contaminant removal is eﬂ‘ective, although not 100 percent emcient.
The results ﬁ’om this experimentation indicate that further research using variations of

some or all of the parameters used in this study is needed.

ACKNOWLEDGMENTS

I would like to thank Dr. Orlando B. Andersland and Dr. David C. Wiggert for
giving me the opportunity to research this topic and for their continued assistance and
guidance. I would also like to thank Lizette Chevalier for her help and encouragement
and 1C. Brenton for his aid in designing and constmcting some of the needed devices for
this research as well as taking care of many of the technical problems that occurred

during this experimentation.

iv

TABLE OF CONTENTS

Page
List of Tables .................................................. vi
List of Figures .................................................. vii
Nomenclature .................................................. ix
CHAPTER
1. Introduction ............................................. 1
1.1 Description of the Problem ............................... 1
1.2 Current Methods of Treatment ............................ 3
1.3 Effect of Soil Freezing on Contaminants ..................... 4
1.4 Scope of Investigation .................................. 6
2. System Characterization .................................... 8
2.1 Soil Characterization ................................... 8
2.2 Contaminant Characterization ............................ 9
3. Experimental Setup ........................................ 13
3.1 Description of Apparatus and Instrumentation ................ 13
3 .2 Experimental Procedure ................................. 19
4. Discussion of Results ....................................... 23
4.1 The Frozen and Unfrozen Gradient ........................ 23
4.2 Rate of Freeze Front Advancement ........................ 30
4.3 Contaminant Removal .................................. 36
4.4 Uncertainty .......................................... 41
5. Field Applications ......................................... 44
5.1 Field Setup and Arrangement ............................. 44
6. Further Research. ......................................... 47
6.1 Modiﬁcations to this Experiment .......................... 47
6.2 Additional Variables for Future Research .................... 48
References. ................................................ 49
Appendix A. ............................................... 50
Appendix B ................................................ 100

LIST OF TABLES

Table Page
1 Capillary Numbers and Corresponding Flushing Rates .......... l9
2 Frozen and Unfrozen Gradients, “nth and Mthout Dodecane . . . . 25

3 Capillary Number, Avg. Gradient, and Ratio Between Gradients . . 27

4 Freeze Front Advancement Rate for Capillary Numbers of 0
and 0.00028 ........................................ 32

5 Freeze Front Advancement Rate For Capillary Number of 0.00030 . 33

6 Freeze Front Advancement Rate For Capillary Numbers of 0.00037

and 0.00045 .......................................... 34
7 Removal Eﬂ‘ectiveness .................................. 3 7
Al - A16 Data From Experimentation ........................... 50
B1 - B3 Data From Experimentation ........................... 100

Figure

7a,b

8a,b,c

10

11
12
13
14
15

16

LIST OF FIGURES

Page
Sources and Processes By Which The Soil
Becomes Contaminated ........................... 2
Water Wetted and Air Saturated Entrapment
\Vrthin a Soil Pore ............................... 2
Residual Oil Saturation Versus Capillary Number ........ 10
Singlet Blob Trapping, Constant Size Pore Space, Bypass
Trapping .................................. 12
Schematic of the Experimental System ................ 14
Proﬁle of Copper Tubing Placement .................. 15
Schematic of Cell with Thermistors and W1re Screen and
Photo of Cell with Thermistors and “ﬁre Screen ......... 17
Setup of Residual ................................ 22
Temperature versus Location for Flow equal to 0 ml/min
with Dodecane "Entire". ........................... 25
Temperature versus Location for Flow equal to 0 ml/min
with Dodecane "Close-Up". ........................ 26
Capillary Number versus Frozen Temperature Gradient . . . 28

Capillary Number versus Unfrozen Temperature Gradient . 28

Advancement Rate vs Time for Flow = 53 ml/min ........ 35
Picture of Residual Before Freezing .................. 39
Picture of Residual Aﬁer Freezing ................... 40
Percent Removed vs Capillary Number ................ 41

vii

LIST OF FIGURES

Figure Page
17 Possible Setup for Field Decontamination After

After Andersland, 1991 .......................... 45
18 Alternate Setup for Field Decontamination ............. 45
19a,b Plan Views of Freeze Pipe Spacing ................... 46
Al-A16.l Figures From Gradient Data ........................ 50
B1-B10 Figures From Gradient Data ........................ 100

viii

NOMENCLATURE

Total Cross Sectional Area
Acceleration Due to Gravity
Length of the Cell

Bond Number

Capillary Number

Porosity

Flow Rate

Particle Radius

Width of the Cell
Displacing Fluid V1scosity
Displacing Fluid Velocity
Seepage Velocity

Fluid Density Diﬁ‘erence

Interfacial Tension

ix

[L2]
[LT'zl
[L]

[1]

[1]

[1]
[L’T‘]
[L]
[L]
[MT‘U]
[LT‘]
[11“]
[MIR]
[MTZ]

CHAPTER 1
INTRODUCTION
1.1 Description of the Problem

Leakage that occurs from hazardous waste sites and underground storage tanks
continues to pose a threat to the drinking water supply. When these hazardous materials,
namely non-aqueous phase liquids (NAPLs), seep into the soil, they migrate down
through the porous media, leaving a residual as they progress. When they reach the water
table, the NAPLs continue to sink if they are more dense than water (DNAPL), or ﬂoat
on the water table if they are less dense than water (LNAPL) as shown in Figure 1
(\Vrlson, et al, 1990). In both cases, between 10 and 50 percent of the contaminant is left
behind in the form of a residual which continually contaminates the groundwater by
dissolving slowly over a long period of time. In the case of LNAPL, a large percentage
of the contaminant can easily be removed by saturating the soil and allowing the LNAPL
to ﬂoat to the surface of the water table where it can be disposed. When this saturation
occurs, generally by a rise in the water table, 50 to 90 percent of the contaminant is
removed. The remaining 10 to 50 percent becomes trapped in the soil by capillary forces
present in the porous media. The entrapment within a water wetted saturated soil pore
and an air saturated soil pore is illustrated in Figure 2 (Hunt, et al, 1988).

The capillary force, or pressure, is the result of interfacial forces acting between the
water, contaminant, and soil pores. Compared to buoyant, viscous, and gravity forces
present in the porous media, capillary forces are the strongest and thus prevent the
residual contaminant from being released from the soil pore (Corey, 1986). The concept
behind residual trapping is discussed in greater detail in Chapter 2.

1

 
 
 
     
   
 

hazardous waste site
2 ground surface

leaking tank
vapor ‘ f
“ ‘ _ VADOSE

ZONE

SATURATED
ZONE

  

Figure 1: Sources and Processes by Which the Soil Becomes Contaminated

WATE R

   

(a) (b)

Figure 2: (a) Water wetted saturated and (b) air saturated entrapment within a
soil pore

3

1.2 Current Methods of Treatment

There are several methods currently used to cleanse contaminated soil in both
saturated and unsaturated media. However, no method currently used is economical and
100 percent eﬁ'ective in removing residual contamination in either type of media.

Soil washing is the most eﬁ‘ective method of treating contaminated soil and can be
used in either a saturated or unsaturated instance. This method involves excavating the
soil and then cleaning it. Two commonly used cleaning methods are to heat the soil to
very high temperatures to burn away the contaminant or to wash the soil. After it has
been cleaned, it is then replaced in its previous location, or dumped somewhere else.
This method is 100 percent emcient, however, it is very expensive and is therefore
usually only used on smaller areas of contamination (VVrlson, et al, 1990). In unsaturated
soil, vapor extraction is one of the more eﬁ‘ective methods used to remove residual
contaminants. It involves forcing hot gases down through the soil, which volatize the
residual in the soil. The resultant gases are then vacuumed out. This method can reduce
the residual contamination by up to 95 percent, however, is again only eﬁ‘ective in
unsaturated conditions. In saturated soil, surfactant ﬂushing is a method that is currently
being examined. This technique attempts to remove the residual contaminant by altering
the interfacial tension between the residual contaminant and the porous media, thus
allowing residual mobilization to occur easier (Wilson, et al, 1990). Surfactant ﬂushing
has proven to be successful in laboratory experiments, however, it has been less
successﬁil in the ﬁeld. This occurs for several reasons. First, in experimental
applications it is easier to ensure that the surfactant will come into contact with all the

residual contaminant in the soil, thus, optimizing the use of the surfactant. However, in

4

ﬁeld applications, this becomes much more diﬁ‘ucult to do, mainly because it is not
possible to see down into the soil and determine which portions have been in contact with
the surfactant and which areas need to be treated. Another factor to consider when
using surfactants is that in experiments the surfactant can be obtained from the soil

easily, while In the ﬁeld, it is diﬁcult to retrieve the surfactant once it has entered the

soil (Andersland and Wiggert, 1994; erson, et al, 1990). Another method currently
used mainly in saturated media is biodegradation, which is the injection of bacteria into
the contaminated soil. These bacteria then feed upon the contaminant, while not
contaminating the soil or groundwater. However, this method is generally not used alone
since the bacteria most often feed only upon the contaminant that is dissolved in the

water leaving the undissolved residual (Wilson, et al, 1990).

1.3 Effect of Soil Freezing on Contaminants

Previous experimentation has been conducted on the eﬁ‘ect of ﬁ'eezing on
contaminant immobilization. Iskander (1986) showed experimentally that artiﬁcial
freezing can be used to successfully immobilize contaminants in the soil without any ill
eﬁ‘ects to the environment (Iskander, Jenkins, 1985). Iskander also showed that creating
an artiﬁcial freezing ﬁ'ont in the soil could separate contaminants and possibly act as an
alternative to slurry walls or other types of contaminant barriers. Further research on the
eﬁ‘ectiveness and integrity of using frozen soil as a contaminant barrier is currently being
conducted. It was also noted in Iskander's research that repeated ﬁ'eezing and thawing of
a contaminated area of soil tended to consolidate the hazardous waste material (Iskander,

1986). This leads to the concept that controlling the rate at which the soil is frozen could

5
release the contaminant which is trapped and allow an eﬁ‘ective cleanup of the
contaminated area.

Based on Iskander's research and other similar experiments, researchers at Michigan
State University (MSU') began to examine how and if controlled ﬁ'eezing could remove
trapped residual contamination through a series of experiments conducted by Soehnlen
(1991). The experiments were run using 1 mm diameter glass beads to simulate the
porous media These beads were placed in a sealed tank 15 cm wide by 25 cm high with
a thickness of 1.5 mm. The cell was saturated with water, followed by a NAPL
(dodecane), and then resaturated with water, leaving a residual in the porous media. The
cell was then lowered into a coolant tank to create a freeze front. Many diEerent
lowering rates were attempted until an optimum rate, a rate at which the residual
contaminant is most eﬁ‘ectively removed from the media, was found. At this optimum
rate of 0.8 cm/day, it was noted that as the freeze ﬁ'ont "pushed" the residual out in front
causing it to build up, the amount of water present at the ﬁ'eeze front would decrease. As
a result of the limited water at the ﬁost line, the ﬁ'eeze front could not continue and
would then ”jump" over the built up residual and continue ﬁeezing on the other side
(Soehnlen, 1991). The experiment showed that the formation of ice will force residual
that is trapped in the porous media out of the pore. This phenomena of exclusion from
the ice formation can be explained by the way ice crystals form. Ice crystals form very
pure and do not accept any substitute ions into the structure. When ice forms, its
formation possesses much stronger forces than the capillary forces which trap the
residual and can thus overcome them. Also, when ice forms it increases in volume by

approximately 9%. This increase also drives the residual from the soil pore (Pounder,

5
release the contaminant which is trapped and allow an eﬁ‘ective cleanup of the
contaminated area.

Based on Iskander's research and other similar experiments, researchers at Michigan
State University (MSU) began to examine how and if controlled ﬁeezing could remove
trapped residual contamination through a series of experiments conducted by Soehnlen
(1991). The experiments were run using 1 mm diameter glass beads to simulate the
porous media These beads were placed in a sealed tank 15 cm wide by 25 cm high with
a thickness of 1.5 mm. The cell was saturated with water, followed by a NAPL
(dodecane), and then resaturated with water, leaving a residual in the porous media. The
cell was then lowered into a coolant tank to create a ﬁeeze ﬁ'ont. Many diﬁ‘erent
lowering rates were attempted until an optimum rate, a rate at which the residual
contaminant is most eﬁ‘ectively removed from the media, was found. At this optimum
rate of 0.8 cm/day, it was noted that as the ﬁeeze ﬁ'ont "pushed” the residual out in ﬁont
causing it to build up, the amount of water present at the freeze front would decrease. As
a result of the limited water at the frost line, the ﬁ'eeze front could not continue and
would then "jump" over the built up residual and continue freezing on the other side
(Soehnlen, 1991). The experiment showed that the formation of ice will force residual
that is trapped in the porous media out of the pore. This phenomena of exclusion from
the ice formation can be explained by the way ice crystals form. Ice crystals form very
pure and do not accept any substitute ions into the structure. When ice forms, its
formation possesses much stronger forces than the capillary forces which trap the
residual and can thus overcome them. Also, when ice forms it increases in volume by

approximately 9%. This increase also drives the residual from the soil pore (Pounder,

6

1965). The experimentation by Soehnlen (1991) left many unanswered questions, some

of which are examined in this research.

1.4 Scope of Investigation

In light of previous research, the present study focused on creating a controlled ﬁ'eeze
front to remove trapped residual NAPL from a porous media. In conducting this
research, two main objectives were established. The ﬁrst was to obtain information on
temperature gradients at the ﬁ'ost line, and the second was to determine NAPL removal
eﬁ'ectiveness by this controlled freezing technique. These goals, equipment design and
experimental work are related to the aforementioned research by Soehnlen (1991)
conducted at MSU. This experiment was directed toward gaining a more precise and
accurate understanding of parameters reported in previous work, while further
developing the theories behind this new method of treatment.

To gain a greater understanding of the gradient data and removal eﬂ‘ectiveness, extra
devices were included in the design of the system. These devices include therrnistors, a
ﬂushing water system, entrance and exit valves, and a horizontal freezing system.

Thermistors were added uniformly in the cell to obtain temperatures throughout the
cell and to ensure that temperature measurements could be taken at and around the ﬁost
line regardless of where it was in the cell. There was limited interference in the
continuity of the porous media due to the small size of the thermistors. The
speciﬁcations and accuracy of the therrnistors and the data logger, used to read the

temperatures, are included in Chapter 3.

7

A ﬂushing water system was also added to the cell design. This was added to prevent
the contaminant from building up at the ﬁ'eeze front as it was released. If the freed
contaminant is not ﬂushed away from the ﬁost line the ﬁeeze front will "jump" over the
buildup of contaminant because of the high concentration of residual and low
concentration of water. This was one of the main observations from Soehnlen's (1991)
experiments. The ﬂushing water is also provided to ﬂush the ﬁeed residual to the top of
the cell where it is removed through exit valves.

Finally, the system was designed to represent one that might be used in the ﬁeld.
This included having a horizontal ﬁ'eezing system with the water ﬂushing system
previously described. The horizontal ﬁ'eeze front was created by placing two copper
tubes, on each side of the cell. These tubes were placed up against the glass of the cell,
spaced 10 inches (25.4 cm) apart. More precise details and a schematic showing the
ﬁeezing system can be found in Section 3.1. This is fairly representative of vertical
freeze pipes placed in the ground at a ﬁeld site. The ﬂushing water is also somewhat
representative of a water injection system that would be used in the ﬁeld. The
experimental ﬂushing water enters freely along the bottom of the cell under a screen,
migrates horizontally and vertically up through the soil. In the ﬁeld, the ﬂow would be
directed into the media at an upward angle and would progress in the same manner as the

current experimentation.

CHAPTER 2
SYSTEM CHARACTERIZATION
2.1 Soil Characterization

In this experiment, glass beads were used to represent soil. They were selected for
these experiments primarily due to their homogeneity, known properties, such as
porosity, and visualization advantages (Morrow, et al, 1988). These properties aid in
setting up a homogeneous residual and allowed the freezing to be performed in a
relatively predictable environment. The glass beads used measured between 0.92 and
1.24 mm in diameter. This small size was chosen because of the narrow cell that was
selected for the current experimentation (4 mm cell opening). The porosity of the glass
beads in the cell was approximately 0.39. This was determined by ﬁlling a graduated
cylinder to a speciﬁed level with glass beads and then compacting them with a small
narrow metal rod as they would be compacted in the cell. A measured amount of water
was then added until the water was level with the glass beads. The cylinder was
examined to ensure that the beads were saturated and there was no air present in the
media. The volume of voids was taken as the volume of water poured into the graduated
cylinder. The porosity could then be determined by dividing the volume of voids by the
total volume. This process was repeated three times with very consistent results. The

seepage velocity is calculated once the porosity is known and is given by v,:

" =q/4" (Eq. 1)

where q is the ﬂow rate, A is the total cross sectional area, and n is the porosity (Potter,

Wiggert, 1991). The seepage velocity was used extensively in calculating the changing

8

9

ﬂow rate as the ﬁeeze ﬁ'ont advanced and was also used in relating these ﬂow rates to the

capillary number.

2.2 Contaminant Characterization
To understand more about the contaminant that was selected for this experiment, it is
necessary to examine how NAPLs are trapped in a porous media. After the contaminant
has seeped into the soil, it has inﬁltrated the pore spaces of the soil. The capillary forces,
which are much stronger than the buoyant, viscous or gravity forces present, then cause
entrapment of the contaminant to occur. The capillary number is given by N,:

1‘.

0' (Eq. 2)

NC:

where v is the displacing ﬂuid velocity, u is the displacing ﬂuid viscosity, and o is the
interfacial tension (in this case between the dodecane and water) (Hunt, et al, 1988). The
capillary number is the ratio of viscous to capillary forces. It is when the capillary force
becomes larger than the viscous, gravity, or buoyant forces present in the porous media
that the residual is trapped.

The initial capillary number was selected based on graphs of residual oil saturation
versus capillary number (Figure 3). The capillary number selected was the one which
just begins mobilization of the residual. In Figure 3, this is the point on the curve where
the slope begins to increase rapidly. Using these graphs in Figure 3, a ﬂow rate can be
calculated which will change as the length of the cell changes due to the advancement of

the freeze ﬁ'ont. It is given as:

 

” (Eq- 3)

10

where l is the length of the cell (which will vary depending upon the advancement of the
ﬁ'eeze ﬁ'ont), w is the width of the cell which is constant, and n is the porosity of the

media which also remains constant.

 

 

 

 

 

 

 

 

 

 

    

 

 

 

 

 

 

’00 v va'v?!’ v v vvv-Vj Vvvvvv‘t r rv‘rvv—‘ﬁ vvvvv—l Y 77"" 20° v Vva W vvvvvvv‘r Y rvvvva v vvvvvw 1 vvvrvw' v vvv
'00 l .0 b
so » 9° norm.
0...”
..° . 'O'"m 1 ..o b
00"...”
12 O O' I: o I!!! I m
l I"! 8 aaa
800 ¢ $.00 TESTS". 4
3'
a 00 .0
u
3 r
‘ so ' so
an L '°
t
to > ‘ 2° ' 't
0
0° no A LAAL‘J - AAA-Au! A A‘IA.‘ A L‘AI’AJ A AAA-lad a ALIA]
wand—“MW.
0000‘! cocoo- 0000 occ- 0.. 0' IO 00m cocoa 0000' 0” 00 0! '0
(INLLOIV NWOEI
(mum M.
‘0) lb)
300 v 1 800 —1 1
I00 } .0 'l
ISO '- K0
.......... In!!!“
"° c-rum ‘ “.0 “Na-revolve "
L L 0.1..”
" lac int 1 can ’ 13.0
J "ST ' .. . J i ‘7. "ET I .0.
300% 113')... _, i- 11811000
1
g no r .1 3 ac
I’
N ‘0 00 p
QQ :- 0,0 '-
3° !- R. P
0° _1 u_a *1 I .0 r ._.a “—1 ‘ a
00000! 000000 ram I” M 0.! L0 0000” cocoon 0.0000 0.00 0" 0| [.0
mm W W“! "I

(c) (d)

Figure 3: Residual Oil Saturation Versus Capillary Number
One way the contaminant can become trapped is by a large variation in the size of the
pore space through which the contaminant and water travel, due to the ﬂow ﬁ'om an
aquifer. When this occurs, the smaller pores drastically reduce the capillary pressure
between the NAPL and water causing the NAPL to snap oﬁ‘ (Wilson et al, 1990). This is
referred to as singlet blob trapping and is illustrated in Figure 4a. Ifthe pore space that
the NAPL and water travel through is relatively constant in size, then the capillary force

does not change signiﬁcantly and the oil and water each remain continuous. This is

Sill

T6511

its 11
Vola
Were
that .
dOde

{amp

Wise

11

shown in Figure 4b. The NAPL can also be trapped when a pore is bypassed. This
simply means there is more than one path through which the contaminant and water can
travel. The contaminant and water enter the smaller pore path ﬁrst due to capillary forces
(Corey, 1986)(I-Iunt, et al, 1988). The ﬂuids may then follow the same path, or enter
through another pore path. After the NAPL and water have entered both paths, the larger
path becomes less resistant to the ﬂuids and thus, the smaller pore path is bypassed
leaving the residual trapped in the pore. This is shown in Figure 4c. The two diﬁ‘erent
types of trapping can also occur together. The bypassing could take place in the larger
path with the pore size varying, snapping oﬁ‘ the NAPL (\thson, et al, 1990)(Hunt, et al,
1988). It is important to understand that the percentage of residual that will be trapped is
determined by these two trapping mechanisms and the variability of soil from place to
place. Thus, by using uniform, spherical glass beads, the variability in soil ﬁ'om place to
place is practically eliminated and a more accurate estimate of the percentage of trapped
residual contaminant can be made.

The contaminant used in the experiment was normal dodecane. This was used due to
its low solubility, its relative safety compared to other types of LNAPL, and its low
volatility. The dodecane was dyed red with Oil Red 0, biological stain. Glass beads
were soaked in this oil-dye mixture for three days and then visually examined to ensure
that the dye would not color the beads red; it did not. The interfacial tension between the
dodecane and water is 52.8 mN/m at 25 °C. This number will vary some at a
temperature of 0 ° C (next to the freeze ﬁ'ont), however, will not change signiﬁcantly.
This value is used in the equation for capillary number (Eq. 2) to gain an estimate of the

velocity required for the residual to begin mobilization.

12

 

pore

Figure 4a: Singlet Blob Trapping

 

 

Figure 4c: Bypass Trapping

of
1911;
tub
COp
ﬂan

in d11

CHAPTER 3
EXPERIMENTAL SETUP

To determine how the experimental system would be set up and what would be
included in it's design, the objectives stated in Section 1.4 were examined closely. In
addition to the objectives, there were physical features that were also desirable. These
objectives include designing the cell so that pictures could be taken to show the exclusion
process of the ice on the residual dodecane, and creating a horizontal ﬁ'eeze ﬁ'ont similar
to that applied in the ﬁeld. Ideal components for the system were then investigated to
determine their feasibility of using them. Limitations on these components resulted
mainly ﬁ'om excessive cost. In these cases alternatives had to be created and/or found

and used. Speciﬁcs are explained in the following sections.

3.1 Description of Apparatus and Instrumentation

The entire system consisted of a cooling process, a temperature measurement system,
the cell itself, and a water ﬂushing system. A schematic of the experimental system is
shown in Figure 5. The cooling system consisted of a coolant tank and four 2 inch
diameter copper tubes with valves. The coolant tank held 48 gallons of a 50/50 mixture
of antifreeze and water and was maintained at a temperature of -24.5 ° C (+/- 2 °C). A
pump was included on the coolant tank to pump the coolant mixture through the copper
tubing. The copper tubing was ﬂattened, providing a width of 3 1/8 inches (7.9 cm) of
copper to be in direct contact with the glass of the cell. This tubing couuld not be
ﬂattened perfectly, thus, allowing only 50 percent of the copper (1.56 in or 3.95 cm) to be
in direct contact with the glass. This is the reason that such a large diameter tube

13

lb:

TF1
the:
Obie

the:

14

 

 

 

 

 

 

 

Coolant Tank

 

 

 

 

 

 

 

Valva- O

Figure 5: Schematic of Experimental System
was chosen. The ends of the tubing were kept round to allow a ﬁtting to be placed on
them easier. The tubing was placed on both sides of the cell with two tubes per side and
each tube directly opposite another as shown in Figure 6. The copper tubes were bent
out at the ends because the cell was so narrow that the tubes could not be placed up
against the glass without hitting the one directly opposite. Valves were included in the
copper tubing system to allow coolant to be ﬂushed through either the ﬁrst set of tubes,
the second set, or both. The entire cooling system was insulated with therma-cel
insulation, 0.5 inches (1.27 cm) thick which ﬁt snugly around all the copper piping and
Tygonm tubing. The temperature measurement system consisted of 20 YSI precision
thermistors (44033 with resistance of 2252 OHMS at 25 °C). The therrnistors were
obtained from Newark Electronics in Grand Rapids, Michigan. Three conductor

thermistor wire was used with the thermistors. It was obtained from S.W. Controls in

1 5
Plymouth, Michigan. The wire was cut into 20-65 foot (2.2 m) pieces, the ends stripped,
the wires ohmed to determine which two of the three wires would be used, and then the

ends soldered to the thermistors. A heat sink was used when soldering the wire to the

    

 

 

C, l

111%? at... m”;

Figure 6: Proﬁle of the Copper Tubing Placement
thermistors so that the thermistors would not be damaged. A water proof shrink tube
was then placed over the wires to aid in protecting them from the water and ice. An oil
resistant silicon sealant was then used to ensure the thermistor and wire would remain
waterproof. The thermistors were calibrated using a 0 °C ice bath. Because the
resistance measured by the thermistors was linear, only this calibration was required. To
ensure this linearity, a thermistor reading, with the calibration from the 0°C ice bath
already applied, was taken at room temperature and was then compared with two
diﬁ‘erent thermometers. The three temperatures read to within 0.2 °C of each other,

supporting that this linearity was indeed correct. During the experimentation, a logging

ho

the

seal
Sinc
and

resis
third
auror.

lempe

mind.

16

multimeter was used to read the thermistors. This machine read to the nearest 0.1 ° C and
was accurate to plus or minus 0.3 °C for temperature ranges of -80 °C to + 80 °C.

The cell consisted of two glass sides 1/4 inch thick, two plexiglass ends variable in
thickness, and a plexiglass bottom also variable in thickness. Two plexiglass sides that
ﬁt around the copper tubing were then placed against the glass on each side to help
support it. An inﬂow valve was provided on each side of the cell, while two outﬂow
valves were included on one side of the cell. Along the bottom of the cell, a wire screen
was placed (Figure 7a and b) to allow the water entering the cell to pass ﬁ'eely into the
cell. This wire mesh was ﬁne enough to keep the glass beads ﬁ'om penetrating through it,
but still allowed water to pass ﬁ'eely through. This ﬁ'ee ﬂow of water into the cell
allowed for an even distribution across the cell. This, in turn, allowed a relatively
homogeneous residual to be set up as well as consistent ﬂushing from one experiment to
the next.

Three diﬁ‘erent sealants were used to try and seal the cell. The ﬁrst was a silicon
sealant (aquarium sealant) which appeared to seal the cell until the NAPL was added.
Since the NAPL was oil based, it deteriorated the adhesiveness of the sealant to the glass
and plexiglass. The second sealant used was Leak Lock”. This was chosen due to its
resistance to oil, however, it did not adhere well to the glass or plexiglass. Finally, a
third sealant, Ultra Blue Silicon Sealant, was used. This sealant is used primarily in the
automotive industry, is readily available, adheres to most anything, and can withstand
temperatures of -80 to 500 °F. This last sealant eﬂ‘ectively sealed the cell.

The water ﬂushing system used for this cell was designed with three primary needs in

mind. The ﬁrst was to provide an eﬁ‘ective way to saturate the cell with water, dodecane,

Co per Tubln\g

 

 

 

 

 

\‘L
A-——3sa «21—» \
Elli
Valves
f O O O O O
.77": k ,1 I A' K /4
" ‘ \/ \.
' ,: Thermistors ‘ Thermistors
., ' E trance
‘_ Wire Cloth . Valve
3
( ~A—W . 80 cm ~ ,A ~——_>

Entrance Valve

Figure 7a: Schematic of the Cell with Thermistors and Wire Screen

 

Figure 7b: Photo of Cell with Thermistors and Wire Screen

ex.

USE

the

em;
betv
the 1

becal

Wale:

ﬂ’JShir.

1 8
and then again with water. The second was to ﬂush the dodecane away ﬁ'om the freeze
ﬁ'ont as it advanced, and to provide enough water for the freeze ﬁont to remain
continuous. The third need was to allow drainage of the cell if required. The ﬁnal
system, as used in the experimentation, can be seen in Figure 5. This system which was
created pumped deionized water (using a Cole-Partner variable speed drive with pressure
loaded pump head, model number G-07144-05) into a direct reading ﬂowmeter (Cole
Parmer, model number G-32012-33), through a three-way valve, and ﬁnally through a T-
valve which led to both cell entrances. The water was pumped through the three-way
valve and allowed to ﬂow either into the cell or into a collection reservoir by simply
turning the valve. This three-way valve also allowed for drainage of the cell when setting
up the residual and when completing the experiment and emptying the cell.

Deionized water was used in the experiments because of results ﬁom previous
experiments. In the experiment conducted by Soehnlen (1991), regular tap water was
used and the minerals in the water were expelled along with the residual contaminant as
the freeze front advanced. In eﬁ‘ect, this created two contaminants in the experiment. To
simplify the problem and concentrate on removal of the NAPL, deionized water was
employed. The water was pre-cooled in a cold room and then moved back and forth
between a cooler with ice and a ﬁ'eezer to maintain the temperature. The temperature of
the water was maintained at 10.5 °C (+/- 1.5 °C). This temperature range was selected
because it is close to the temperature of ground water which would be the most preferable
water source in a ﬁeld situation because it is the most inexpensive. The last experimental
run was performed with a water temperature of 0 ° C to examine the effect of a colder

ﬂushing water temperature on the effectiveness of removing the trapped residual NAPL.

the

19

3.2 Experimental Procedure

To accomplish the stated objectives, contaminant and non-contaminant experiments
were mn using seven diﬁ‘erent capillary numbers and corresponding ﬂushing rates as
shown in Table 1. Contaminant and non-contaminant runs were used for several
diﬁ'erent reasons. The main reason was to compare the ﬂow and unﬁ'ozen gradients in
both cases and determine if they were the same or if the presence of a NAPL aﬁ‘ected
them. Another reason for using both cases was to gain gradient data more quickly. In the
case without NAPLs present, it was easy to run the experiments quickly because no clean
up was necessary after each experimental run. The only time necessary between
experiments was the time it took the porous media to thaw. However, in the case of a
contaminant run, the cell and the glass beads had to be emptied and cleaned in between
each experimental run. It was also beneﬁcial to run the non-contaminant experiments
ﬁrst to get the ”bugs'" out of the system before introducing the NAPL, which would
cause a greater amount of time to be spent not only ﬁxing the problem, but, also cleaning

it up.

 

[Capillary Number I 0.00000 0.00011 0.00020 0.00023 0.00030 0.00037 0.00045
Inow Rate (ml/min) ] 0 19 35 49 53 65 79

 

 

 

 

 

 

 

 

 

Table l: Capillary Numbers and Corresponding Flushing Rates
The ﬁrst step in running a non-contaminant experiment was to place 18 thermistors in
the cell, 1 in the antifreeze, and 1 in the ﬂushing water. These thermistors were

uniformly spaced 1 to 1 1/2 inches (2.5 to 3.75 cm) apart in the cell. The glass beads

We
the
eqL
equ
and
rtad

met.

Cell“

20

were then poured into the cell around the thermistors. Once the cell was ﬁll] of dry glass
beads, the cell was saturated using the pump and ﬂushing water system and prodded with
a narrow metal rod to eliminate any air pockets. Glass beads were then added as needed
to ﬁll the cell to the ﬁrst exit valve. Once the media was set up in the cell, initial
thermistor readings were taken, including the temperature of the antifreeze and the
ﬂushing water. Next, Styrofoam insulation (1 1/2 inches, 3.81 cm, thick) was placed
completely around the cell in two layers, including over the top and bottom of the cell.
F‘mally, the pummng of the coolant and the ﬂushing water were started simultaneously.
Thermistor and ﬂowmeter readings were taken more frequently at the beginning of the
experiment and less ﬁequently as the ﬁeeze front advanced. This was done because the
freeze ﬁ'ont advanced very quickly at ﬁrst and then continually slowed down. To
maintain a constant ﬂow rate, the ﬂow was reduced proportionally as the ﬁ'eeze ﬁont
advanced and reduced the area over which the ﬂow could enter the cell. This reduction
was based on determining the seepage velocity from a capillary number. The readings
were stopped when either the freeze ﬁ'ont advanced the entire length of the cell, or until
the ﬁeeze front reached a point where it could not advance any further, i.e. the point of
equilibrium. In the initial experiments, the ﬂow rate was decreased once this point of
equilibrium was reached in order to gain an understanding of the speed of the ﬁ'eeze ﬂoat
and how much variations in the ﬂow rate would affect it. At each different ﬂow,
readings were taken for a minimum of two days to ensure equilibrium had indeed been
met.

For a contaminant run, the thermistors and glass beads were placed in the cell as the

cell was saturated as in the non-contaminant run, and the locations of the thermistors

res
wa
Fig
{811

ﬂus

21

were again measured. The dodecane was then evenly applied along the top of the cell as
the water table was slowly lowered. Once the cell was saturated with dodecane, the
water table was slowly raised. The rate at which the water table was decreased and
raised were monitored. The tree oil was removed by raising the water table allowing the
oil to ﬂoat to the surface of the water and then collecting it so that a percent of residual
could be found. This value was then compared with tabulated values of percent residual
remaining depending on the speed of the water acting as the water table. Again
thermistor readings were taken periodically depending on the rate of ﬁeeze front
advancement. Figure 8a shows the initial saturation of the cell with dodecane. The water
table is being lowered and the residual is being added ﬁ'om the top of the cell. The
residual is less dense than water, and thus, will only travel as far down in the cell as the
water has gone. Figure 8b shows the cell completely saturated with dodecane while
Figure 8c shows the residual being leﬁ behind as water is pumped back into the cell
removing the ﬁee dodecane. Next, in the contaminant experiment, the freeze pipes and
ﬂushing water were then started simultaneously. Again, the rate of ﬂushing water was
determined based on the capillary number and the initial capillary number value was
based on the graphs in Figure 3. The capillary number values were then varied to allow
different ﬂow rates to be attempted. These values were the same ones used for the non-

contaminant runs (Table l).

macaw “adage—EU Ransom ”om Semi

 

{1

th

th1

3a

obt

Gra
repu
be 01
PIOp.
Heces

Coma

CHAPTER 4
DISCUSSION OF RESULTS

The two main objectives of this project, as stated in the scope of the investigation,
were to gain an understanding of the frozen and unfrozen gradients in the media with and
without residual, and to determine how eﬂ‘ective the controlled freezing process is at
removing residual contaminant.

4.1 The Frozen and Unfrozen Gradient

The gradient data gathered for all experimental runs is presented in Table 2. This
table lists the capillary number (used to determine the ﬂow rate and seepage velocity of
the ﬂushing water), whether dodecane residual was set up for the experimental run, and
the measured frozen and unﬁozen gradients for the "Close-up" and "Entire" graphs in
° Cl'm and ° C/cm. The range of capillary numbers used was based upon data ﬁ'om Figure
3 as explained in Chapter 3. The capillary number of 0.00030 was used primarily
because this is the average value calculated ﬁ'om the graphs in Figure 3. The speciﬁcs of
obtaining this number are discussed in greater detail in Chapter 3.

Obtaining frozen and unﬁ'ozen gradient'data is an important aspect of this research.
Gradient data can be used in a controlled freezing remediation ﬁeld application to aid in
replicating a successﬁil experimental cleansing process. If accurate gradient values can
be obtained through experimentation, the speed at which the freeze front should
propagate to successfully remove the contaminant can be reproduced. Also, it is
necessary to understand the seepage velocity of water required to ﬂush the freed residual
contaminant to the ground surface. This seepage velocity also aﬁ‘ects the ﬁ'eezing rate.
Therefore, many different factors are occurring which all interact and ultimately deﬁne

23

24

the gradients. Thus, to eﬁ‘ectively use the gradient in a ﬁeld application, a common
correlation is needed that relates the parameters to each other and to the gradient. These
successful gradient parameters become essential in a ﬁeld application because the entire
process will be done completely under the ground and out of view. For example, it is
desirable to not only determine at what freezing rate residual contaminant is successfully
removed, but, also to understand whether this rate should be maintained by a very cold
ﬁeeze pipe with a very high ﬂushing water seepage velocity or with a freeze pipe that is
not as cold with ﬂushing water that has a lower seepage velocity, or some other optimum
combination.

The gradient values were determined by graphing the temperature measurements of
each thermistor versus the thermistor locations in the cell. The readings taken were read
from a data logger which read to within +/- 03°C. The thermistors themselves had a
negligible error, while the thermistor wire was not long enough to aﬁ‘ect the data.
Therefore, the uncertainty for the gradient values was based on only one factor. Thus,
the uncertainty is small for the gradient data portion of the experiment. In Figures 9 and
10, the thermistor locations are represented by the individual marks (i.e., circles, squares,
x's, etc). These values were then plotted over time as shown by each line in Figures 9
and 10. The headings ”Close-up" and "Entire" listed in Table 2 refer to two separate
graphs that were plotted for each experimental run. Figure 9 shows the experimental data
plotted over the entire cell length for a particular run, while Figure 10 shows only data
close to the 0° C isotherm (i.e. data over a smaller range of cell length) plotted for a
particular run. Examples of this are shown in Figure 9 which shows the experimental

data over the entire cell length (30 in., 76 cm) plotted for a run with no dodecane and no

25

 

 

Capillary Dodecane Frozen Frozen Unfrozen Unfrozen
Number (dT/dx) in (deg Cftn) (dT/dx) in (deg Clem) (dT/dx) in (deg Cfrn) (dT/dx) in (deg Clem)
(Y or N) Close-up Entire Close-up Entire Close-up Entire Close-up Enu're

0.00045 Y 8.16 8.40 3.21 3.31 5.90 5.53 2.32 2.18
0.00037 N 7.85 8.00 3.09 3.15 5.85 5.82 2.30 2.29
0.111037 Y 6. 16 6.41 2.43 2.52 4.67 4.47 1.84 1.76
0.10137 Y 7.74 7.83 3.05 3.08 6.25 4.27 2.46 1.68
0.11130 N 3.43 4.33 1.35 1.70 2.62 3.16 1.03 1.24
0.1!)030 N 5.71 5.44 2.25 2.14 4.13 4.08 1.63 1.61
0.00030 Y 7.14 7.14 2.81 2.81 4.79 4.33 1.89 1.70
0.111030 Y 4.44 4.40 1.75 1.73 3.64 3.30 1.43 1.30
0.111030 Y 4.64 5.00 1.83 1.97 2.32 3.08 0.91 1.21
0W Y 3.09 2.71 1.22 1.07 2.19 2.1 1 0.86 0.83
0.111028 N 3.76 4.44 1.48 1.75 3.26 3.12 1.28 1.23
0111028 Y 7.22 7.29 2.84 2.87 6.29 5.93 2.48 2.33
0.00020 N 2.91 2.91 1.15 1.15 2.17 2.17 0.85 0.85
0.01111 1 N 2.32 2.32 0.91 0.91 1.21 1.21 0.48 0.48

 

 

 

 

 

 

 

 

 

 

 

 

* Indicates ﬂushing water temperature used in experiment was 0° C
Table 2: Frozen and Unﬁ'ozen Gradients, With and Mthout Dodecane

seepage velocity (Nc = 0), and Figure 10 which uses the same data from the same
experimental run, however, only experimental data around the 0°C isotherm (12 in, 30.5
cm) is plotted. These two graphs were plotted to gain a more accurate estimate of the
gradient. As the freeze front reached equilibrium, the coolant temperature ﬂuctuated as

discussed in Chapter 3, thus, the last reading was not necessarily the ﬁnthest distance the

 

 

 

 

 

 

 

—’— t=15.5hr

20
‘5 ’23)”. +t=0hr
2G

 

 

—‘°'— t=22.5hr

 

 

 

Temperature (C)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

-5 f/ ‘\ I -__‘__ t: 27.5 hr
-10 ”5’ —o—— t = 38.25 hr
“5 a —0— 1: 46.75 hr
‘20 .1 —0— t: 625 hr
’25 ' —><— I: 67.75 hr

0 5 10 15 20 25 30
Location (in)

Figure 9: Temperature vs Location for Flow = 0 ml/min with Dodecane

 

 

 

._————I +r=0hr

 

15
—‘C*_ t=12hr

 

10

—°— t=15.5hr

 

—°— t=22.5hr
—*"— t=27.5hr

0
/,o: ——o—t=3s.25hr

——‘— t=46.75hr

 

1h. 1 .
Nth
M1

 

 

 

 

—O— t=62.5hr

 

 

 

 

 

 

 

 

"5 —><-—- t= 67.75 hr
15 17 19 21 23 25 27

Location (in)

 

 

 

Figure 10: Temperature vs Location for Flow = 0 mllmin with Dodecane
ﬁeeze front had reached. For example in Figure 10, the line for t=62.5 hours crossed the
0°C mark at the furthest distance ﬁom the ﬁ'eeze pipe (22.7 in, 57.7 cm), however, the
last reading taken was at t = 67.75 hours and crosses 0°C closer to the ﬁ'eeze pipes (22.2
in., 56.4 cm). Therefore, the two gradient values for a ”Close-up" and "Entire" graph
were averaged for each experimental run and both sets of graphs and the tabulated data
are included in Appendix A.

The ”Entire" and "Close-up" results ﬁom Table 2 were averaged and are shown in
Table 3 along with a frozen to unfrozen gradient ratio and the capillary numbers. Many
interesting trends can be seen in this table. The most obvious and consistent trend is that
the frozen gradient is larger than the unﬁozen gradient. This trend supports the fact that
ice has a greater thermal conductivity than water, and therefore, the temperature change
over a distance, x, should be greater in ice than the temperature change through water
over the same distance, x. The ratio of unfrozen gradient to ﬂow gradient on average is

0.72, as shown in Table 3. This ratio remains fairly consistent throughout the data

al

‘13

int.

dis

H0
intr
leng

Simi

27

indicating the change in capillary number does not eﬁ‘ect the ratio between the two
gradients. This relatively constant ratio can also be seen in Figures 11 and 12. The data
for the ﬁozen and unﬁozen gradients is scattered over approximately the same total
length only shifted over about 1 inch. Also, the distance between each individual point
is similar for the frozen and unfrozen data.

The next consistent trend that can be seen ﬁom Table 2 and Figures 11 and 12, is that
a higher capillary number gives a higher gradient value for both the ﬁozen and unfrozen
gradient. This can be explained because as the capillary number increases, the
seepage velocity of the water entering the cell to ﬂush the ﬁeed residual increases, thus,
introducing more heat into the system. Therefore, the temperature will vary more over a
distance, x. It would be logical to assume that the Mom gradient would vary more
than the ﬂow gradient because of the ﬂushing water entering only on the unfrozen side.
However, this is not the case, possibly due to the fact that although the ﬂushing water
introduces more heat into the system, it was introduced fairly uniformly over entire the
length of the unfrozen side. Therefore, each location on this side was aﬁ‘ected in a

similar manner. The ﬂushing water process and design is explained in detail in Chapter 3.

 

 

 

Frozen Frozen Unfrozen Unfrozen Capillary Ratio of
Gradient Gradient Gradient Gradient Number Unfrozen
(dT/dx) (dT/dx) (dTIdx) (dT/dx) to Frozen
(deg Cftn) (deg Clem) (deg Cftn) (deg Clcm) Gradient
8.28 3.26 5.72 2.25 0.00045 0.69
7.93 3.12 5.84 2.30 0.00037 0.74
6.29 2.47 4.57 1.80 0.00037 0.73
7.79 3.06 5.26 2.07 0.00037 0.68
3.88 1.53 2.89 1.14 0.00030 0.74
5 .58 2.19 4.1 1 1.62 0.00030 0.74
7.14 2.81 4.56 1.80 0.00030 0.64
4.42 1.74 3.47 1.37 0.00030 0.79
4.82 1.90 2.70 1.06 0.00030 0.56
2.90 1.14 2.15 0.85 0.00030 0.74
4.10 1.61 3.19 1.26 0.00028 0.78
7.26 2.86 6.1 1 2.41 0.00028 0.84
2.91 1.15 2.17 0.85 0.00020 0.75
2.32 0.91 1.21 0.48 0.00011 0.52

 

 

 

 

 

 

Table 3: Capillary Number, Average Gradients, and Ratio between Gradients

 

(‘npillnry Nurnhcr

( ‘npillnr'y Nun-her

O
A

Q

C
A

Q
(‘1

Q
C)

lo
a?

28

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00
Frozen Temperature Gradient (dT/dx in deg Cfrn)

Figure 11: Capillary Number vs Frozen Temperature Gradient

9.00

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0.00 1.00 2.00 3.00 4.00 5.00 6.00
Unfozen Temperature Gradient (dT/dx in deg Clin)

Figure 12: Capillary Number vs Unﬁozen Temperature Gradient

7.00

era
chh

mes

dod
capi
1101
vﬂu

bytl

4001
beca
Used

ﬂush

vahle
1u030

5UbSt.‘

29

Now that the basic trends have been examined and established, it is desirable to
examine the different variations in the experimental procedure that may have affected
either the ﬂow or unﬁ'ozen gradient. The ﬁrst of these variations is whether the
presence of dodecane residual had any eﬂ‘ect on the frozen or unﬁ'ozen gradients. This
can be examined by looking at Table 2, which indicates which experiments used
dodecane. Comparing the experimental runs with and without dodecane for the same
capillary number, no distinct diﬁ'erence can be seen. There is some diﬁ‘erence, however
not substantial or consistent enough to conclude that its presence aﬁ‘ects the gradient
values. For the most part, the diﬁ‘erences appear to be minor and could easily be caused
by the minor variances which make exact reproduction of the experimental runs
extremely diﬁcult to achieve. Also, the presence of dodecane residual in the system is
small when compared to the volume of water and glass beads used, and would not be
expected to noticeably aﬁ‘ect the gradient. The results seem to support this.

Another variation in the experimental procedure was to use ﬂushing water at 0° C.
Although a temperature of 9 to 12 ° C is much more reasonable to use in this experiment,
because this is closer to the temperature of ground water which would most likely be
used in a ﬁeld situation involving this process, it is still desirable to examine how other
ﬂushing water temperatures effect the results of this experiment. Only one experimental
run, shown in Table 2, was performed using a ﬂushing water of 0° C. The frozen gradient
values (in °C/cm) are 1.22 for the "Close up" and 1.07 for the ”Entire", while the
unfrozen values are 0.86 and 0.83 respectively for this particular run. The values are

substantially below the other frozen and unﬁozen gradient values obtained for this

capi

into

1111's
the l
obse

W35

deter
5,1518
the p.
ﬂushi
freezi
freeze
deer-e;
from 1
any ﬁt;
Media
the £10.
can be

Pipes. '

30

capillary number. This occurs because a colder ﬂushing water temperature is introduced

into the system and is closer to the ﬁ'eezing temperatures that are ultimately achieved.

4.2 Rate of Freeze Front Advancement

From previous research by Soehnlen (1991), an optimum rate of ﬁ'eeze ﬁ'ont
advancement was obtained. The cell in Soehnlen's experiments was slowly lowered into
the ﬁeeze tank at diﬁ‘erent rates. The optimum rate was determined by visually
observing how much residual was pushed out in front of the freeze ﬁ'ont and how much
was left behind in the ﬁozen media. The setup and process of Soehnlen's experiment is
explained in greater detail in Section 1.2. The optimum rate ﬁ'om his research was
determined to be 0.8 cm/day (0.033 mm/hr). In the current experimentation, the ﬁeezing
system was set up to freeze horizontally across the cell, instead of vertically up it as in
the past research. This diﬁ‘erence in ﬁ'eezing technique, and the introduction of a
ﬂushing water system, did not allow for precise control of the ﬁ'eeze ﬁ'ont rate. When the
freezing system was ﬁrst turned on, up to 5 cm of the media on each side of the copper
freeze pipe would freeze very quickly. However, over time, the ﬁ'eeze rate would
decrease until equilibrium was met, meaning all of the heat possible was being extracted
from the media with the ﬁ'eezing system and therefore, the freeze front could not progress
any further. It would have been theoretically possible to control the rate at which the
media froze simply by controlling the ﬂushing water entering the system. By increasing
the ﬂow, which means an increase in seepage velocity and capillary number, more heat
can be introduced into the system and thus balance the heat being extracted by the ﬁ'eeze

pipes. This seepage velocity could then slowly, at any desirable rate, be reduced, thus

31
allowing the freeze ﬁ'ont to advance. However, to equalize the heat extraction energy of
the ﬁ'eeze pipes at the beginning of the experiment, a large ﬂow would be required. The
pump chosen for the experiment could not supply the ﬂow required. The pump variable
speed is explained in the experimental setup in Chapter 3. Its range was chosen based on
what capillary numbers would most eﬁ‘ectively aid in removing the residual contaminant,
not for controlling the rate of freezing. There were no pumps available that covered the
broad range that would be required and that were also within the budget of this research.
Rates of ﬁeeze ﬁ'ont advancement were determined for this experiment anyway so that a
comparison with the results of Soehnlen could be made. These rates are shown in Tables
4, 5, and 6 both in cm/hr and in/hr for all of the experimental runs and one nm is shown
graphically in Figure 13. The data in Tables 4, 5, and 6 is all graphed and included in
Appendix B. These freeze rates were obtained by looking at the net movement of the
frost line for each timed reading. The exact location of the frost line at the time each
temperature reading was taken was determined by interpolation ﬁ'om the thermistor
readings. The rate of freeze ﬁont advancement varied over each experimental run. The
advancement rates were fast at ﬁrst and then slowed down. It is not until the rates slow
down substantially, i.e. towards the end of the experiments, that they approach the 0.8

cm/day rate that Soehnlen, determined to be the most effective. Figures 9, 10, all of the
ﬁgures in Appendix A, and all of the ﬁgures in Appendix B also show this in different
ways. In ﬁgures from Appendix A, the lines that indicate each time temperature readings
were taken start ﬁrrther apart and slowly become grouped together. Therefore, it is still
desirable to create a freeze ﬁ'ont traveling horizontally at around 0.8 cm/day with a

system that will ﬂush the freed residual. This should be examined to determine whether

32

0500qu 505:5 E... 5;)
380.0 E... o be 9.3:...2 b.3230 8.. .coEoo....>u< EB"— ennui

e 2.5

.8 BE... .88... 23 2.... 88: E... .3... 3538 59m 0 3
on... 88c 9.03. .a .388. canon n .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

88... as... 8.2 8.8 8....
_ 88... 88... 8.» 82. SN...
88... 85... n... «a... 8.... :88. 88.... 8... and. 8.8.
«:8 8.8... n...“ .8... 83 2.3 88... an... 9... :3
8.... 8.... can a.“ 8.... «2 .5... an .2. a:
a a o o o 8
e653 8...... En. E19 84.. 9.2.6. 9.5.. E... 95. 3.
2...... .5695.)— ...oEo>c_z 25,—. 2.2—.262 .5532
0...: 0.3. .02 .02 .oz 33. 83. .02 .02 .92
38300 565.3 3 2.30.60 .53 .
«386 a 20 .8 32:85:; 2.2.... 38.".
3.8... 88... n2. 8.. 8n...
8.... 88... 8... 5... 8N...
$8... 28... 8... an... 8.... 33 3.8... 2.... S.“ 03...
o. .8 5o... 2... 5.... 8.... 2.... 88... 8.. 8:. 82
a: .c 88... on... 53 23 a... :8... 8.. R. 82
8.... 88... 8... 2.. 83 as... 8.... 8.2 .3 8." 8...... 8n... 8.. a... 8..
:2. 82 S... 8... N...” :8... as... can 88 88... .8... n2... 8.: 2k 8..
o o c c o o c o 8
32.3 8...... E... a a... :58 881.. a% Ea a... see... 3.5. E... 2.3 a:
GEE. EoEo>Oz EOE—>2 08:. 305252 3059;02 05:. “co—=25: 2.05252
23. 23. .oz .02 .oz 03. 23— .oz .92 .oz 23. Sam .92 .oz .92
9. :3. 0588a .85.; . 2 8.. o588o .85.? : ... a... 8.880 .23 .

 

 

 

 

e u .6 a... .8535»? .8... as...

 

1—1

 

1 av 1'1

 

I zjj

 

33

Freeze Front Advancement for CN 2 0.00030 with Dodecane

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Run #1 Run #2
Net Net Net Rate Rate Net Net Net Rate Rate
Movement Movement Time Movement Movement Time
(in) (cm) (hrS) (in/hr) (cm/hr) (in) (cm) (hrs) (in/hr) (cm/hr)
0 0 0 0 0 0
1.84 4.67 1.42 1.30 3.29 2.31 5.87 1.00 2.31 5.87
0.510 1.30 1.58 0.323 0.820 0.440 1.12 1.00 0.440 1 . 12
0.510 1.30 3.67 0.139 0.353 1.89 4.80 7.25 0.261 0.662
0.234 0.594 5.50 0.0425 0.108 0.400 1.02 1.75 0.229 0.581
0.240 0.610 5.33 0.0450 0.1 14 0.200 0.508 5.50 0.0364 0.0924
-0.0400 -0.102 3.33 00120 00305
Run #3 Run #4
Net Net Net Rate Rate Net Net Net Rate Rate
Movement Movement Time Movement Movement Time
(in) (an) 016) (in/hr) (cm/hr) (in) (an) 013) (In/hr) (emit)
0 0 0 0 0 0
1.95 4.95 2.00 0.975 2.48 3.59 9. 12 5.00 0.7 18 1.82
1.51 3.84 7.00 0.216 0.548 0.760 1.93 6.00 0.127 0.322
-0.120 -0.305 3.50 -0.0343 -0.0871 0.610 1.55 7.50 0.0813 0.207
0.640 1.63 4.75 0.135 0.342
0.0900 0.229 6.25 0.0144 0.0366
Freeze Front Advancement for CN 0130.00030 Without Dodecane
Run #1 Run #2
Net Net Net Rate Rate Net Net Net Rate Rate
Movement Movement Time Movement Movement Time
(in) (an) (Inn) 0an (cm/hr) (in) (an) (MS) (in/hr) (cm/hr)
0 0 0 0 0 0
2.15 5.46 2.83 0.760 1.93 3.77 9.58 6.50 0.580 1.47
0.580 1.47 2.92 0.199 0.505 0.360 0.914 5.17 0.0696 0.177
0.590 1.50 3.92 0. 15 1 0.382 -0.0300 -0.0762 4.50 -0.00667 -0.0169
0.930 2.36 5.33 0.174 0.443
0.880 7.24 5.58 0.158 0.401
0.390 0.991 5.67 0.0688 0.175
0.100 0.254 7.33 0.0136 0.035

 

‘ = Began readings at second freeze pipe
” = Began readings after ﬁrst freeze pipe was already turned on

Table 5:

Freeze Front Advancement for Capillary Number of 0.00030
With and Without Dodecane

 

 

 

34

23880 265;? c5 53»
9.806 93 >886 .«o EBE=Z .02:an .8 EoEooaa>c< 2on 082m 6 03am.

.8 BE... $8.? 83 on... 88: .cc coca 3538 595 u 2.
RE 302.. 9.8»... .a 3532 Suom a .

 

cNNd Sued cod no; curd

nvud 38.0 on.“ an; Omnd

rd.— auhd and «06 N».—
c o c

 

eéss e52: J35 EE «a:
2.5. .5526: .5532
Sam 83. .02 .02 .oz
2888 5.3 .
$266 u 20 .8 aol.E8§>v< 89."— 88.“—

 

 

 

 

 

Rbd. $8.0. omé 36. 8.6-

 

 

 

 

 

 

ES. 488.? 2:. 23- 88...- 2.5 33 and 8.2 02 23 83 n; on." 82.
£3 23 S.» a.“ a." 3.: $3 8... 2.: n2 :3 2.3 8.0 32 VS
a o o o o o o o o
3:5 ES: 35 Ea as See . Age Ed Elem En 35$ :55 ~25 Eﬂ ca
05:. 808262 288962 08F 82:26:— 808262 05;. E2582 E2582
2am 2.x :2 .02 .oz 83. 25. 32 .oz .oz 83. 33. .02 .oz .02
z. :3. 2888 .85.? . u. :3. 05889 5.3 . z. :3. 05880 5;» .

 

>886 a 20 he .eﬂeomﬂ>u< 38m 308m

 

35

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 900
0.800
0 700

8 8 8 8
\O in V M
:5 d :5 c5

(IQ/U!) mu mowoompv

0 200
0.100
0 00

Figure 13: Advancement vs Time for Flow = 53 ml/min

20.00 25.00 30.00 35.00 40.00 45.00
Time (hrs)

15.00

5.00 10.00

0.00

35

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8 8 8 8 8 8 8 8 8 8
0: °° ". ‘0. V”. V. '9. N "2 o
o o o o o o o o o

(ﬁt/U!) mu mmouupv

Figure 13: Advancement vs Time for Flow = 53 mllmin

5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00

0.00

Time (hrs)

‘8

eff

{”01
1101

WC!

36

or not this rate is the only one that will prove to be highly eﬁcient, or even if it will
improve the success of the controlled freezing technique over the results of this research.
The effectiveness of removing trapped residual contaminant was not compared to or
graphed against the ﬁ'eeze front advancement rate because it varied for each experiment.
Instead, the capillary number was compared with the removal effectiveness. Removal

eﬂ‘ectiveness is the topic of the next section and will be discussed in greater detail there.

4.3 Contaminant Removal

The second and ultimate goal of this research was to determine if this proposed
technique in contaminant removal was indeed eﬁ‘ective and if so, how eﬁ'ective and
under what conditions. The results of the research indicate that this method is eﬁ‘ective,
although not 100 percent eﬁicient. There are many factors which give the results of this
procedure a moderate range of uncertainty. While the percentage of removal results do
not have to be precise to give a general idea of the success of this procedure, the inherent
uncertainty is an important aspect in examining the true value of this removal technique.
The uncertainty and the parameters aﬁ‘ecting it will be examined later in this chapter.

The eﬂ‘ectiveness of removal was obtained for ﬁve runs, three with a capillary
number of 0.00030, one with a capillary number of 0.00037 and one with a capillary
number of 0.00045. The results are presented in Table 7. The results vary widely,
ranging ﬁ'om 76 percent down to 25 percent of the trapped residual contaminant that was
removed after this controlled freezing technique was applied. The ﬁrst of these ﬁve runs
was the second overall experimental run performed using dodecane. The initial
experimental run using dodecane provided no removal effectiveness results. This was
due to minor problems throughout the experiment which did not allow collection of the

residual contaminant that was removed by the ﬁeezing process to be done in a controlled

37

manner. While no data could be attained for this ﬁrst trial, it appeared to effectively
remove the residual contaminant by pushing it out in front of the ﬁeeze front. Also, very

little residual appeared to remain in the ﬁozen media. Thus, by visual observation,

 

 

 

Capillary Residual Total Width Residual Residual Percent
Number Initial Width Frozen Within Wid Removed Removed
Total Frozen From Width
(ml) (in) (in) (m1) Frozen (m1)
0.00045 120 30 2.01 8.1 2.0 25
0.00030 120 30 4.12 16.5 8.0 49
0.00030 95 30 2.50 7.9 6.0 76
0.00037 95 30 2.14 6.8 2.5 37
0.00030 125 30 4.56 19.0 5.0 26

 

 

 

 

 

 

 

 

Table 7: Removal Eﬁ‘ectiveness
the removal of the contaminant seemed to be very eﬁ‘ective. However, the second and
third trials, with capillary numbers of 0.00045 and 0.00030 respectively, as shown in
Table 7, did not perform in the same impressive manner that the ﬁrst experimental run
did. The removal effectiveness was only 25 and 49 percent respectively. After these two
experiments, which obviously lacked repetitiveness, the method of cleaning the glass
beads was examined and then modiﬁed. The cleaning method was altered to include a
thorough cleaning in acetone that ensured the entire surface area of each glass bead came
into contact with the acetone (for a more detailed description of the cleaning process, see
the experimental procedure in Chapter 3). The fourth run was then made using the
modiﬁed cleaning technique with the capillary number equal to 0.00030. Based on visual
appearance, this run performed closely to the initial experimental run, although it was

still not quite as visually impressive as the initial trial. This run had a removal

3 8
eﬁ‘ectiveness of 76 percent. Figures 14 and 15 show this trial before and after freezing.
Figure 14 shows the residual set up before the freezing process was started, while Figure
15 shows the ﬁeeze front after 2 hours. In Figure 15, the ﬁ'ost line is indicated by an
arrow pointing to the leading edge of it. The buildup of residual just past the ﬁost line
into the unﬁ'ozen media indicates that the freeze ﬂoat is indeed ﬂeeing the trapped
residual and pushing it out in ﬁ'ont. Some residual is still apparent in the ﬁ'ozen media,
however, it is considerably less than the initial amount on the unﬁ'ozen side.

The results ﬁ'om Table 7 appear to demonstrate only one somewhat predictable
occurrence. This is the improved removal effectiveness of the ﬁeezing process after the
glass beads were cleaned in acetone for the ﬁrst time. This improved cleaning technique
seemed to greatly improve the removal eﬁ‘ectiveness. However, the next two runs did
not show the same results. From this observation, the cleanliness of the glass beads
becomes an extremely important variable. It would be impossible to determine how
clean the beads actually were before each experiment without a great deal of time and
extra equipment. It is also diﬁicult to determine how much the experiments were
affected by the adsorption of dodecane by the glass beads. The presence of dodecane
coating the beads initially would alter the interfacial tensions that normally occur in the
porous media. This fact alone could have made this controlled freezing process much
less eﬁ‘ective. The total uncertainty of the experiment should be examined to determine
how much the removal effectiveness may have been affected.

The results ﬁ'om Table 7 are graphed in Figure 16. This ﬁgure illustrates that none of
the capillary numbers used had a conclusive effect on the removal effectiveness under

these experimental conditions and constraints. It is important to remember that these

39

 

Figure 14 Picture of Residual Setup Before Freezing

4O

Frost Line

 

Figure 15 Picture of Residual Setup Aﬁer Freezing

41

results include two experimental runs in which there is uncertainty in the cleanliness of
the beads. This ﬁgure does not show anything else unusual that could not be seen in
Table 7. It does reveal the fact that this method has some effectiveness on the removal of
residual contaminant from a porous media and presents the results in an illustrative

manner.

 

 

 

 

 

 

Percent Removed
8 8 8 as 8

 

10

 

 

 

 

 

 

 

0.00026 0.00031 0.00036 0.00041 0.00046
Capillary Number

Figure 16: Percent Removed versus Capillary Number

4.4 Uncertainty
As mentioned at the beginning of this chapter, there is a moderate uncertainty
attached to these results. Because uncertainty is based on the uncertainty of the
individual items that are used to determine the end value, the uncertainty of each item
used in the equations or process of obtaining the removal eﬁ‘ectiveness results needs to be

examined. The most basic uncertainty in the experimental process that eﬁ‘ects the

42

removal eﬂ‘ectiveness is transferring the dodecane ﬁom container to container and
accurately measuring its volume. The beakers and graduated cylinders used to measure
the dodecane were graduated and could be read to the nearest 0.5 ml. However, when
retrieving the dodecane that had been removed by the ﬁeezing process, it was collected
with large amounts of water. The water and dodecane were then separated and the
dodecane measured in a graduated cylinder. This process involved collecting the water
and dodecane in a 5 gallon plastic container, pouring this into a glass separator and then
pouring whatever dodecane was removed into a graduated cylinder. Warm water was
used to rinse each container to remove any remaining dodecane, however, because of the
small amount of dodecane recovered and the large container, any extra that remained on
the sides of the plastic or glass could greatly effect the ﬁnal measurement.

The next place there is uncertainty is in the ﬁnal calculation for percent removal.
This calculation is the amount of dodecane recovered divided by the amount of residual
contaminant within the area of media ﬁ'ozen during the experiment. In order to
determine the amount of residual within the frozen area, the assumption must be made
that the residual contaminant is homogeneously distributed throughout the cell. This
assumption must be made because a known amount of residual was set up in the cell over
the entire area. There is no way to determine exactly how much is in each area of the
cell. From visual observation, this assumption of homogeneity appears to be true,
however, it is not realistic to assume that the residual contaminant is perfectly
homogeneous. However, it is probably se and would thus, not affect the percent removal

results substantially.

43

The controlled freezing process does remove residual contaminant from a porous
media. However, the extent of its eﬁiciency is not totally understood. From the
uncertainty variables, it appears that the results are within a small uncertainty range.
Only a vague value can be placed on the uncertainty because the uncertainty of the
variables are diﬁcult to place values on. However, as an estimate, an uncertainty of 10
percent seems reasonable. This uncertainty does not aﬁ‘ect the results of this research
signiﬁcantly because the values obtained are only indications of the eﬂ‘ectiveness of this

ﬁ'eezing process, and they are by no means ﬁnal answers to a complex problem.

CHAPTER 5
FIELD APPLICATIONS
The goals of this research were to determine how eﬁ‘ective this controlled ﬁeezing
method is at removing trapped residual contaminant and to obtain gradient information.
The gradient information is what will allow the successﬁrl experimental procedure to
applied eﬁ'ectively in the ﬁeld. The utilization of vertical ﬁeeze pipes to create a
horizontal ﬁeeze wall and the use of water to ﬂush the freed contaminant makes the

gradient data even more similar to what would be used in the ﬁeld.

5.1 Field Setup and Arrangement

There are a couple possible setups for a ﬁeld site as shown in Figures 17 and 18. In
Figure 17, the horizontal ﬁ'eeze pipes are placed throughout the bottom of the
decontamination area and around the sides, forming a box. Freeze pipes are then placed
symmetrically throughout the box along with injection wells in a pattern as shown in
Figure 19a. The pipes around and under the box would then be ﬂow and have water
added into the enclosed box until it is saturated. At this point the interior freeze pipes
sould be turned on along with the injection wells. Figure 19b shows how the area would
be frozen over time. The contaminant would then be collected at the surface as it is
removed.

Figure 18 shows another possible setup which is similar to Figure 17, except that the
sides that enclose the contaminated area are angled toward each other forming a point at
the bottom. The ﬁ'eeze pipes and injector wells are placed throughout the contaminated

area again as in Figure 19b.

44

45

New Ground Water Elevation
v 4/

[—

 

Contaminant

   
 

Reﬁigeration
Pumps

+—— Vertical Freeze Pipes
@®@@9@®®@0
\ Horizontal Freeze Pipes
Figure 17: Possible Field Setup For Decontamination
After Andersland 1991

  
  

 

 

 
 

Freeze Pipes Spaced
Evenly Across Each Side

New Water Table
Contaminant

 

 
    
 
   
 

Reﬁigermon Pumps Initial Water Table

Freeze Pipes

Figure 18: Alternate Field Setup for Decontamination, Shape and Proﬁle Diagrams

46

 

   

   

     

 

 

. Freeze Pipes
- Injection Wells

Figure 19a: Plan View of Freeze Pipe Locations

is) Q) (a) Q) Q]
Q
()0

Freeze Pipes
- Injection Wells
Figure 19b: Plan View of Freeze Fronts After Some Time, t

 

   

 

 

   

 

 

CHAPTER 6
FURTHER RESEARCH
The experimentation performed here answered some questions, however, raised a
great deal more. There were many things that came up during this research which would
be very desirable to examine. However, they were not within the time ﬂame or budget of
the current research. Thus, there is a great deal more that can be examined to more ﬁrlly

understand what the true beneﬁts of this controlled freezing procedure are.

6.1 Modiﬁcations to The Current Experimentation

There are some factors within this experimentation that created uncertainties which
sould be corrected in further research. First, the freezing system should be designed so
that the cell will ﬁeeze completely. As mentioned earlier in Chapter 4, the cell in this
experimentation would not ﬁ'eeze completely, thereby, forcing assumptions to be made
about the homogeneity of the trapped residual. Thus, a cell designed to freeze
completely would take out this uncertainty. Secondly, the coolant used to create the
ﬁeeze ﬁ'ont varied in temperature by up to 4° C which aﬁ‘ected the amount of porous
media that would remain frozen. A temperature diﬁ‘erence of +/- 1° C should be
maintained to achieve more precise results. Finally, the ﬂushing water used to ﬂush the
ﬁ'eed residual to the surface of the porous media was not speciﬁcally directed at the area
of interest, i.e. directly ahead of the ﬁ'eeze ﬂoat on the unfrozen side. While this models
a ﬁeld application more closely, it is desirable in this initial experimentation to have

more control over where the ﬂushing water is applied in order to ensure that the residual

47

48

that has been freed by the ﬁeeze ﬂoat is ﬂushed away from the front to allow the freezing

to remain continuous.

6.2 Additional Variables for Future Research

While this experimentation covered a great deal, there is much more that can be done
to improve the method and the knowledge of the method. The ﬁrst thing that should be
done is to try higher capillary numbers than those used in this experimentation. This
should be done because of the uncertainty in the basis for selecting a capillary number,
in order to use a velocity in which mobilization of the residual is just beginning. The
values used in this research were only approximate and were used only as a starting
point.

The next variable that should be modiﬁed is using a uniform sand, such as an Ottawa
Sand, in the research instead of glass beads. This should be done because the sand could
be replaced for each experiment and thus take away the uncertainty in the cleanliness of
the porous media as was the case in this experimentation with the glass beads.

Finally, the possibility of using a surfactant in the ﬂushing water was brought up
during this research, however, was unable to be examined due to time constraints. This

idea could be very beneﬁcial and would be relatively easy to attempt.

10.

11.

LIST OF REFERENCES

Andersland, OB. and “riggert, D.C. (1993,1994), Personal Communication.

Corey. AT. (1986). W Water

Resources Publications, Littleton, Colorado.

Hunt, JR, Sitar, N., Udell, K.S., "Non Aqueous Phase Liquid Transport and
Cleanup 1. Analysis of Mechanisms", Water Resources Research, Vol. 24,
August 1988, pp. 1247-1258.

Iskandar, I.K. (1986), "Eﬁ‘ects of Freezing on the Level of Contaminants in
Uncontrolled Waste Sites", Cold Regions Research and Engineering
Laboratory, US. Army Corps of Engineers, Special Report 86-19.

Iskandar, I.K., Jenkins, T.F. (1985), ”Potential Use of Artiﬁcial Ground
Freezing for Contaminant Immobilization", Cold Regions Research and
Engineering Laboratory, US. Army Corps of Engineers,

Hanover, NH, Sept.

Morrow, NR, Chatzis, 1., Taber, JJ. 1988. Entrapment and Mobilization of
Residual Oil in Bead Packs, SPE Reservoir Engineering, Society of Petroleum
Engineering, V013122927-934.
pp 927-934.

Potter. MC and Wiggert. DC. (1991), Winds
Prentice Hall, Inc.

Pounder, ER (1965), W, Pergamon Press, London.

Soehnlen, Greg, Cleansing Contaminated, Granular Soils by Controlled
Freezing, Masters Report, Michigan State University, 1991

Tumeo, Mark A, and Davidson, Bret. 1993. Hydrocarbon exclusion from ground
water during freezing. J. of Environmental Engineering, American Society of Civil
Engineers, Vol 119:4:715-724.

Wilson, J.L., Conrad, S.H., Hanson, W.R., Peplinski, W. Hagen, E., 1990,

"Laboratory Investigation of Residual Liquid Organics From Spills, Leaks,
and the Disposal of Hazardous Wastes in Groundwater", EPA/600/6-90/004.

49

50

APPENDIX A

51

058.com .23 >6: Bow ”2 29¢

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

me he 06 «.c 9.x ad nd ed 0.2 m«.e«
we ﬁn he 3‘ to ed as 9w ﬁc— an.v«
ed «6. 9— a._ a.« ﬁn ed. ﬁn ma; em.««
n.«- :1 h.«- o.«- c. ~ - @6- c._ w; w.m_ «c.o«
2.- cd- 56- _.w- ve- m.n- ﬁv- «.v- «.«~ and.
odT a.«- c._ T «.«T 2:- _.c_- «.a- cd- 9.: 5.:
GE. Q:- 067 «.2- E:- w.n_- 02- 02- to «c.0—
n.««- v.m«- v.««- a.««- n.««- ~.««- ~.««- a.««- Q6 8.3
v.:- «.3- QB- 02- «.2- 93- _.e- bd- hd— cm.«_
v.«_- «.2- «.2. _.«_- n.«- ad- m._ he 2: «:—
_.w- w.m- e6- «.6- «.w- c4.- mﬁ we 52 and
«.0. od- a.v- ad- v6- o.«- ma ed «.2 omd
ﬁv- vd- «.m- me. Qw- 04- ed ed «.2 :3.
w.m- Ev.- e.«- n.m- Yv- v.v- m6 _.w ﬁn and
w.«- od- h._- o.«- v6- «6- On nd ed 36
QM- :t c.«- w.«- Wm- >.«_- we- c.« _.«- and
o.«- ed- hA- v.«- ﬁn- saw—- ﬁwT ﬁc- _.««- cc.«
.2 2.60". an: “.3". 2 2.6V". 2 n«.wmua 9:. n.5«u. mu: n.««u. a: 02". E: «W. E: on.
E:
0v BEEQQEoF eon—woo;

 

 

 

 

 

 

 

 

 

52

 

 

E 3.60 n. [X]
E m.«c u. lol
E when“ lol-
E n«.wm uz Iliall
E n.~.« u. '1'
E n.«« u. l-QII
E On— ". Ilol
E «_ u. .IIDl

 

 

cm

ocaoueoa 53 5.5:: o u 30E .8 55304 m> 2282‘th "_< Bewﬁ

m«

c«

E: 5:83
n _

c—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m

 

 

 

 

 

 

 

 

 

c—

n—

o«

53

 

 

E n55 u. lxil
E m.«c u. lQl.
E nude u. Ilol
E n«.wm u. '0'
E n.e« u. lsll
E n.«« u. '0'
E n.n_ u. '9'
E «_ u. [T

.Ecﬂarlll-l

 

 

0:80on .23 5:58 o n 30E 3.. =23er m> 32825.8. ”:4. unswE

 

 

 

 

 

 

 

 

 

 

 

 

$w
\N
\t\v\\
s&\
at .
\\

 

 

 

 

 

 

 

 

2-

2-

o—

n—

c«

(3) armeradrnol

54

05889 .835 sea 03 ”2 use.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«.«- «.«- «.«- «.«- «.««- v.3- «.««- «.««- v.««- _«.«
«.«- «.«- «.«- «.«- «.2- 2.2- «.2- «.2- «.2. _«.v
v.«- «.«- «.«- «.«- «.«T «.2 T «.2- «.«- «.«- ««.«
«.«- «.«- «.«- «.«- «.«- «.«- «.«- «.2- «.2- ««.«
«.2- «.2- «.2- «.2- v.«- «.«- «.«- «.« «.c ««.«
«.2- «.2- «.2- v.2- «.«- «d- c._- «.« _.« ««.«
«.2- _.2- «.3- «.3- «.2- «.« «d «d «.v ««.2
«.«T «.«T «.2- _.2- «.2- «.c «.o «.« «.« _«.«_
«.««- «.««- «.««- «._«- «.2. v6- «6 «.« «.« «2.2
_.««- _.««- c.««- ««.«- «.««- «.2- «.2 «.« «.« ««.3
«.2- «.2- «.2- «.2- «.3- «.« v.« «.« «.« ««.2
«.«- «.v- «.v- «.«- «.c- «.« «.« «.2 «.2 ««.2
_.c «d «._ «.« «.« «.2 «.« «.3 2.3 _«.«_
«.« «.« v6 ~.« _.2 «.2 «.2 «.2 «.2 v«.c«
«.« «.« _.« «.« v.2 «.2 «.2 «.2 «.2 ««.««
«.2 v.2 _.: «.2 «.3 «.2 «.2 2.2 «.2 8.3
«.«. «.2 «.2 «.3 «.2 2.2 «.3 «.c« «.o« ««.««
«.3 «.3 _.2 v.2 «.2 «.2 «.2 «._« c._« ««.««
E «.«vn. E «.««uu E ««.««u2 E ««.««u. E «.««u. E ««.««u. E «.«—u. E «.«u. E an.
E:
CV 2382—th :eueooq

 

 

 

 

 

 

 

 

 

55

0:30on 8953 30E EoN ”«< 032.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«.«- «.«- «.«- «.v- «.«- «.v- 2.».- _.«- _«.«
«.v- «.«- «.«- _.«- «.«- _.«- «.v- «.«- _«.v
«.v- «.«- «.«- «.«- «.«- «.«- «.«- «.«- ««.«
«.«- «.v- «.«. _.«- «.«. «.«- «.«- «.«- ««.«
«.«- «.«- «.«- «.«- _.«- «.«- «.«- «.2- ««.«
«.2- «.«- «.«- 2 _- 2 T «.2- v..:- v.2- ««.«
«.2- «.2 T «.2 2- «.2- «.2- «.2- v2- «.3- ««.2
«2. «.2- «.2- e2- «2- «2- «2- «.2- .«2
o.o«- 92- v.«_- «.««- «._«- «._«- «.o«- «.2«- «3.2
v.««- «.««- _.««- «.««- «.v«- o.v«- _.««- v.««- ««.3
«.2- v2- «.«_- «.2- «.2- «.2. v2- «2- ««.2
«.«- «.«- «.«- «.«- v.«- «.«- «.«- «.«- ««.2
«.«- «.3- «.2- _.«- «.2- «.2- «.«- «.«- «2

«2 _.« «.« «.« «.« «.« «.« «.v v«.««

«4. v.0 26 «.v «.« .6 «.« «.« «c.««

«.« v.« «.« «.« «.« «.« v.« «.« ««.«
.2 «.2 «.2 «.2 «.« «.2 «.2 2.: ««.««
v2 v3 «.3 «.2 «.2 «.2 «.2 _.2 ««.««

E «.««u. E ««.««u. E ««.v«u. E ««.«eu. E 2.3". E «—.««u. E ««.««u. E ««.««u.
«5
6v 83823.2. cognac-2

 

 

 

 

 

 

 

 

 

S6

 

 

E c.6«u. lxl
E «23.1.... le
E «ﬂucua IIIOII
E «cén-I-a -||O||I

E «.«vu. IT:
E ««.««u. .Ililln

E «.VNE IOII
E ««.cnua loll

E «.«—u. IIDI

E an. IT

 

 

38300 25.23 5:58 o u 30E .8 52804 m> 25anth ”«< oEmE

o« ««

$5 .8330..—
o« 2

2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

W.

 

 

 

 

 

 

o«-

««-

ON-

2-

2-

2

2

o«

««

(3) 01111910de .I.

57

 

 

E c.c« u. IIOII
««.««ua ll-OII-
E ««.v« u. IIIT
E ««.«c u. ll.-
E «_.«c u“ I|.¢||I
E «_.«« "a Il-oI-l
E «c.¢« u. IIDII

a 2.8 u. lol-

 

 

0:80on 30:55 5:58 o u 32..» 3.3 .8283 m> 88822th :.«< 2=«E

:5 :33qu

8.«« ooéN 8.«« co.«« 8.

_« 8.

c« 8.2 8.2

 

 

o.«-

 

c.0-

 

od-

 

o.«-

 

\
\‘\\Q\N

ed

 

\

 

 

 

o.«

 

oé

 

 

9.0

 

 

 

 

 

 

 

 

 

c.«

 

 

9.2

«.2-

(3) mmmodmal

58

288qu 3923 5:582 n 30E «.1 032.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«.««- v.««- «.««- ««.«- c.««- v.««- ««.«- _«.«
«.2- 2.2- «.2- «.2- «.3- «.2- «.2- 2«.v
«.2- «.2- 2 T _.2- 2- «.2- Z- ««.«
«- «é- «.«- «.«- ﬁv- «.v- «.v- ««.«
«.«- «.«- «.«- «.«- o..- «.2- «.2. ««.«
to «.« «d- «d «.2 «.2 «.« ««.«
« «.« «.2 «._ «.« «.« «.« ««.2
«.« «.v «.« «.« _.« 3. «.« _«.2
«.v _.« «.v «.« «.« «.v «.v 2.2
«.« «.« «.« «.« «.« «.v v ««.3
v6 «.« «.« «.« « «.« «.v ««.2
«.2 «.2 «.2 «.« 2.: «.« «.« 8.2
«.2 «.2 «.2 2 «.2 «.2 2.: _«.2
«.3 «.2 «.2 «.2 «.2 «.2 v.2 v«.o«
v2 «2 2.2 «.2 «.2 v.2 «.3 ««.««
«2 «.2 «.2 v.3 «.2 «.2 «.2 v«
«.«. «.2 _.2 «.3 «.2 «.2 «.2 ««.««
«.2 «.2 3.2 «.3 «.2 «.2 «.2 ««.««
E «_.o«u. E ««.««u. E ««.««u. E «.«.«u. E ««.2 2". E ««.«-L. E «u.
GEE:
ADV 23a..o«Eu.—. 5:804

 

 

59

 

 

E 2 an". Ill:
2 Nina. IT
2 2.8". IITI
E : emu. IT
2 2... E Ill-

... «3". lol-

E cu. Ill

 

 

c«

380on 505.3 Ease 2 u 30E “8 .8280..— m> 23quth "«< 23E

««

c«

«5 8:83
2

2

 

 

 

 

 

 

 

 

 

 

 

 

'\\
12‘)

 

w J‘\

 

 

 

 

 

 

 

 

 

 

««-

o«-

2-

2-

2

2

ON

(3) amendmol

 

6O

canon—yon 255$» 5:58 2 n 305 .8 5:80... m> 23825.2. ”_.«< E=«E

«5 .8583

2 2 « «

 

 

 

 

 

2 2.2". ll
2 «Ian. ILTI
2 2.8". If
2 :3". IT
a m: .u. IT:

2 :2". Incl.

E cu. 'T

 

 

\\

\ ‘
\
‘\
.
‘-

 

 

 

\‘\

 

 

:\

 

 

 

 

 

 

 

 

 

 

 

 

 

2

(3) amwadmal

61

8888 .855 case 2 u 3E “3. 2.5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«.v«- «.««- «.««- «.««- v.v«- «.««- «.v«- ..««- «.««- «.««- ..««- .«.«
«.2- «3- «.2- «.2- «.2- «2- «.2- ..2- «.3- «.2- «.3- .«.v
Q. T «.2- Q. T .2- «.«T «.. T «.. T «.2- «.«- «.«- «.«- ««.«
«.v- .6. v.«- «.«- «.0- c.«- «.«- v.«- «.«- «.«- «.«- ««.«
v. T v. T «.«- «.«- «.«. «.«- «.«- «.« «.« «.« «.« ««.«
«.« «.« «.c- «.«- v.«- .d- «.« «.« «.« «.v «.« ««.«
«.« «.v «.. «.o t. «.. «.« «.« «.« «.« «.« ««.2
«.« «.0 o... «.« ..« «.« v.« .6 ..« «.« «.« .«2
«.v «.0 ..« «.« «.« «.« «.« ..« «.« «.« «.« 2.2
c... «.0 ..« «.« ..« «.« «.. «.« «.« «.v «.v ««.3
«.v «.0 «.0 c... «.« v.« «.« «.« .6 v.« «.« ««.2
«.« .... «.2 «.« «.« v.« «.« «.« .2 «.« «.« ««.«.
.... «.2 «.«. «.2 «.« «.« «.« .2 «.2 o... «.2 .«2
v2 «.« . «.« . «.2 «.2 .2 «.2 «.«. «.3 «.3 «.2 ««.««
«.3 «.«. «.2 «.3 «.2 <2 «... «.2 «.2 v2 «.3 ««.««
«.2 «.«. «.2 «.3 «.«. «.«. v.2 «.«. «.2 .2 «.3 ««.««
«.«. .2 .2 «.3 v2 .3 «.«. .3 «.2 «.2 «.3 ««.««
«.«. «.«. «.2 «.3 .2 «.2 «.«. «.«. «.2 «.2 «.3 ««.««
E .«.««n. E ««.««u. E .«.v«u. E ««u. E ««.««n. E «.««u. E «.««u. E «.2". E ««.. .u. E .«6u. E on.
.5.
6v 888995... .5289.

 

 

 

 

 

 

 

 

 

 

 

62

 

 

E .«.««u. |Tl
E ««.««u. l-xI
E .«.v«u. llxll
E ««u. '0'

E ««.cvu. '0':
E «.««u. Ilqu
E «..««u. llql
E 3.". '0'

E ««.. .u.

 

.l

E .«du. [IDI-

... O". llllll

 

 

2:30on 52...? 5:55 «« u 30.". 8.. 5.88.. 9» 8332.80... “3. 2%.".

25 .5539.
c« «« c« 2 2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

««-

««-

o«-

«.-

2-

2

2

o«

(3) amuodmal

63

0:80.009 505.3 5:58 «« u 30.... 5.. 5.800.. 2, 2882.80... u..v< 2%.".

.8 888..
8.. _ 8.2 8.8 8... 8.. 8... 8w

 

 

 

 

2 8.2.". lxl:

 

 

2 2.8". -|x||

 

E .«.v«u. IOI

 

E ¢«u. loll

 

2 8.3». IIT

 

 

E «.««u. IT

 

 

E 2.8". IT

 

2. 3:". II..|.

E ««...u. I'D-l

 

 

E .«du. III-l

 

 

 

 

 

 

 

 

 

 

 

r c.2-

«.«-

o6-

Qw-

c.«-

a...

c.«

o...

0.0

o.«

Q2

(3) ”mandala;

05880 .883 case a. u .35 ”2 2%..

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

..«N- «N. «.NN- w.NN- VN- N.VN- w.MN.. QVN- ..«N- .M.N
«6.- «.m.- w.N.- «.M.- 0.v.- 0.V.- v.- ...»... N.N.- .M.g
v.w- ...... h- «...- «.w- «- v.w- w- ..«- wM.«
md- ... N... ...- N- M.N- «. . n «.G- M «N...
«.M v.« «.v «.N .... ... 3. N.N «.« «Nd
«.... ..« N.« v.5 . .0 « «.v 0.? 0.. N0.«
«.w #6. «d. v.« w... 0.0 M.0 v.« «... «0.0.
N.« «d. «... #6. Na .- m.0 M.« M... .M.N.
0K. « c. N.« «.0 «.« 0 «.v ..0 «. M.
«6 #0 a... N... « «..v M.« «.M M.« «0....
N.« «.0 v.” w «.0 «..0 m N... 0.w «N0.
«.« M... «N. «.N. h... N.N. M.M. M.N. M.M. 00.x.
N. 0.M. ”.9. «av. «.«. 0.M. «.v. QM. ..v. .M.«.
...M. N.«. M0. ..0. 06. M6. «.9. 0.M. N6. v«.ON
0.... ..0. ..b. «.0. ..«. «..v. N.«. «M. We. «0.NN
«6. N0. N... b. ..«. 06. N.«. «.«. Md. «N
«.V. M.0. M... ..h. N.«. «6. M.«. «.M. 3?. ««..«N
06. ..0. M... «.0. «. 0.... ..«. 0.M. M6. «.....N
E «0.«vﬂ. E «Ndﬁﬂ. E «.0Mﬂ. E NM". E «c.0Nu. E «.«.H. E «N. . .H. .E «.«". E O".
a...
.0. 8.5.8.50... 5.2.00.

 

 

65

 

 

E «0.2:... [XI-
E ««.ovu. IIIOI
E «.««u. lol
E ««u. Ll

E ««.««u. ITI
E «.2". ll
E ««.. .n. llol
E «.«u. '0'

E «u. l-Iull

 

 

o«

0500qu 505.3 5:58 «v .0. 5.500. 9» 05.88.80... ”«< 053".

««

o«

3.. 5.500..
« .

2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2

2

«N

(3 30(1) amuadmal

 

......
mmmmmm

 

V3 l‘
a N o.

00

 

 

 

 

 

 

 

 

 

 

\\\\

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10

9.5

8.5

8

7.5
Location (in)

7

6.5

5.5

Location for 49 mllmin Without Dodecane

Figure A5.1: Temperature vs

67

05880 .23 case a. u 32... 6.4 2%...

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«.«- «.T «.T ....- «..- c..- «.«- .«.v
0...- ..T «..- «.«. ..«- «.0 v...- ««.«
«.T «.«- ..«- «..- ..«- «.« «.« ««.«
«.«- «.«- «.«- «.«- «. T «.v «... ««.«
«.«- «.0- «.0- «.0- «.«- «.« «.«. ««.«
«.«- «.«- «.«- «.«- «.«- «.. «.3 «0.2
«.2- e3- «.2- «.3- «3- «.0- «.3 .«.2
«.2- «.«T «.2. v2- «.2- «.3- «.«. «..«.
«.««- «.««- c.««- ..««- «.««- «..«- «.2 «0.3
0.2- «2- «.« T 0.2. «.« T «.2- 0.2 ««.2
«.T «.T «.0- ..0 ..«. ..« «.3 8.2
Q... «.0 «.« ..« «.« «.«. 0.2 .«2
«.2 «... «.«. «.«. v.2 v.2 «.«. ««.««
0.2 «.2 «.«. ..2 «.3 «.2 «.«. «0.««
«.2 «.2 «.2 «.2 «.2 «.2 «.«. ««.««
«.«. «.2 «.«. «.«. «.«. «.«. «.«. ««.««
v2 «.«. «.«. «.«. «.2 ..«. v.2 ««.««
E .vu. E «.««u. E «.««u. E «.«.u. E ««.. .u. E «.«u. E «u. a...
.0. 8:52.80... 5.500..

 

 

 

 

 

 

 

0500.05. 5.3 ...—5.5 «v ....- 30.--. 5.. 5.500. m> 05.82.80... 6< 05$”.

8 .833
8...« 8.«« 8.o« 8.2 8.2 8.« 8d

68

 

 

 

 

 

... .VH. I‘ll.

\

 

 

 

E «.««u. Ll

E «.««u. 1'-

 

 

2. we". Iloll

 

E ««...u. II-ol

 

 

E on". lol

... an. IT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«.««-

«.««-

«.2-

«.2-

c.«-

a...

«.«

«.2

«.2

«.8

(3) 01mm: .1.

69

0500.009 5.? c828 «v u 30.... .0. 5.500.. m> 05.82.80... “..«... 05w...

.5. .5500.
8.«« 8. .« 8.o« 8.2 8.2 8.2

\\

8.2

 

\

«.3-

«.2-

 

 

 

 

c.2-

«.«-

 

 

 

\

 

... :ﬁ. l-TI \

 

E «.««u. Ilol

 

 

 

E «.0«u. loll

 

2 «an IT \

«.«

 

 

 

E ««...n. Ill-l \

a...

 

 

 

c0

«.«

 

 

 

 

 

 

 

 

 

«.2

(3) amnuodmal

70

2 =3. .05an 5;» case m u 30E “2 2.5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m..«- v..«- «..«- o.««- 5..«- 5.5«. «..«- v.«« .m.«
ﬁx. in. 5.»- 5.»- 5.0. 5.«- w.. 5.«« .m...
c..- c..- «..- ..c- 9. 56 .... «.«« and
«.5 5.5 v.5 «.5 5.5 5.«. 5.2 ««.«« n«.5
«... c... 5.5. to. c... 9... 0.5. v.«« n«.w
v... . ..2 5.«. w... 5.«. 9m . 5.2 v.«« «0.5
5.3 e2 «.2 ..«. ..2 5.2 5.2 v.«« 50.2
«.2 v.2 ..2 c... n.«. «.2 «.5. m.«« .m.«.
v.«. 5... o... ..o. 5.5. v.2 5.2 «.«« 5. .m.
0.5 «.5 5.5 «.5 5.x .... 5.3 n.«« 59v.
..5. 5.5 n.5 ..m 5.» ..«. 5.2 v.«« n«.2
n... ..v. 5.2 «..«. 5.2 «.2 c2 «.«« 8.2
..2 5.2 Q2 5.2 0.3 ..5. ..5. n.«« .m.5.
«.2 «.2 «.2 m6. «.2 v.5. «.2 m.«« 95.8
v.5. 0.2 ..2 5.... 5.2 5.5. m5. n.«« 5o.««
«.5. 0.2 «.2 5.2 5.2 5.5. «.5. v.«« 8.3
n5. 5.2 «.2 5.3 5.2 «.5. «.5. 5.«« n5.n«
«.5. v.5. n.2 5.2 5.2 5.5. 0.5. 5.«« n5.5«

E n.0«u. E 3.8". E 05.“. E 5..«.». E 590". E «H. E «VAN. E cu. .5.
6. 2882.50... .5289.

 

 

 

 

 

 

.5 :3. .238on 5.? 5:5:— mn u 3a.”. 8.. 5.38.. m> 8382.th u5< 05m."—

Ec 5:83
8.8 8.2. 8.8 8.2 8.... 8d 8d
c.m«-

 

 

 

71

 

 

% o.o«-
o6 . -

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2 SN". Iol. \\
E 3.8". lol 2:-
2 wt". IIT \\ 3.-
a 2.2.... ll \\
- od
2 Sen“ '0' \ Q.
E mu. '0' \\\ ed
2 3..-. Ammhy. \. \ 3:
.. .lol \\ /lu” m \u
,0 .x/ﬂ'f’tv.\u.
E C". III um [Arm -«WW _. a o . lu\0\n\ Qw—
vlanl-llD‘\.LUlln. I I I I .
..TIIIIDI 6 ON
I I it‘ll-Ill; IIIIIIJIIIIIIIIII

 

 

o.n«

(3) oxtmmdmol

72

 

 

a new". lol.
2 3.8.". Ill
2 0:". loll

a tin. Il-II

 

 

.5 :3. .288on 5.3 ...:SE mm u 32... 8.. 5.309. m> 95.82.th ”..5< 8:9".

:5 .5330...
8.» cm... 8.5 and 86 and

8.n and 86
8.3-

 

c.«.-

 

 

92-

 

 

 

\ qvu

 

c.«..

c...

 

 

 

 

c.«

 

 

OJ.

 

 

c6

.8

 

 

 

 

 

 

 

 

 

 

 

ed.

(3) ammadmal

73

Q :3. .2888 .23 5:58 m u 30E H? 22.5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«.3 5.5. n... 5.«. 5.5. . v.2 m«
5.3 .. c... 2 3 n2 wa..«
0.3 5.5. n... «.«. 5.2 m2 3.8
«.3 ad. 0... 5.«. n2 v.2 5.
«. v5 2.. m... ..m . 5.3 «0.5.
cd n6 5.5 o. ..«. 5.3 002
m « c.« 5 ed. v.3 2
v.5. 5.5 5 5... c2 v.3 «0.5.
w n6 w c... 0.2 5.3 05
n... ..v ..c .... m2 Q3 wmd
.- m...- m.. «.5 e.«. 5.3 5
5.n- n6- «.m- o «.5. «.3 86
e. .- 5.. .- «.2- 5.5- Om 5.3 «9v
2- 5.2- v.2- n.2- w.5- m3 mad
«.««- n.m«- o.««- 5..«- «..«- 0.3 «
m2: 2:". ...E ..u. mE n«.5u. mE «u. «E .u. mE an.
55
6. 2382.50... 5.38...

 

 

«a :3. 6:3...qu 5.3 5:55 mm u 30.... B. 5:89. m> 2550550... um< 28E

.5. 8:83
mm 8 2 2 n

74

 

 

 

 

 

8; 2:". Incl

 

8.. :8 'fl

 

2.. 2.0.... lol

...... mu. Il-ul

 

m:— —"~ IUII

\\

 

5..E On. IIT

 

 

 

\\/

 

 

 

 

 

 

 

 

 

 

 

n«-

o«-

2-

c.-

c.

2

c«

(3) amendmal

75

«.. :3. 658.59 5.3 5:58 mm H 30E 8.. 5.88.. m> 2282.th u..w< 2:»...

W5 5

n6

.5. 8:83

0

n6

m...

 

 

 

\\\

 

 

 

 

8.. 2:". .l
2.. :8 IT

...... 2.8". ITI

 

WY

 

\
\k

 

 

 

\

 

 

 

 

 

 

 

 

 

 

 

 

 

 

«.-

o.-

(3) mmmadmal

76

2 =3. 65880 5:5 5:52 an n 32". 5< 03¢

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

9m— aé ma. ad a9 8mm
a.n_ 3.— 02 2: ma: No.8
v.2 5.2 NS ad «.2 3.2
ad— 0: 2: ad 5.— 2.:
3: «6 as ta m.a_ 2.2
2 ad— a.a ed a_ 3.:
«.2 2: o ha va— 2.2
3: 9a as ad ma— aha
as ﬁn m4. n.a a.a_ 2.
a 3 ad 5 a. and
3- «.m- ed- v a.a_ 3mm
ad. 2- 2:- ..m- «.2 «3.
«.2- n5- ad? a.. mi Sum
3: nus—n. 2: 02". a: mu. a: an“ 5 an.
:5
av 8382—th 5:83

 

 

 

77

mt cam— .ugooﬁon 5—3 Fat—E mm H 30E haw GOENOOJ m> Quasar—huh. d< Gasman-

95 .8383
mm on n _ 3 n

 

 

 

 

 

E 3:". ll \

 

 

\
3.. SE '0' \ \
a; on. Ill

 

a: an. IT

 

2.1+ \
\

 

 

ITIHMMkaI/I\

 

 

 

 

 

 

 

 

 

ON-

2.

0..

c—

m—

cm

(3) ammdmal

78

an :3. .ocaoooon— 53> 5:58 an n Boﬁ .8 cos—83 m.» 23.38th ”~.o< «SaE

 

 

 

 

 

 

 

 

 

 

c5 5:33
a 3 a 2 u 3 c on n 3 e
2-
\m! 2-
\\ n
a; 2.2". IT! \ u
E 02". IT um\\\\ o W
a: on. II-l \ )
%m\\ n o
\
o“\\\
2
2

 

 

 

 

 

 

 

 

 

 

 

 

79

a :3. 65889 .23 case 2 u 3: 52 £3

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Wm _ ad 2: c.: _.N_ SN dodm
#3 ed ad d9 9: EN _ad_
d3 n.a a.a dd ad. ddm ama—
d a.a ..c I. Nd add aad.
ﬁn in v n6 N6 Now .26—
Né d.m On he a.a a.a nd
n._ o._ a.a dd as 08 nma
2- «.7 no tn m6 EN «2.
Ni. a.a- ON- n6- dd 3N dﬁo
vd- vd- a.a- d6- tn- ddu aaé
3:- mi- 3- 0.2- dd- 5.8 V
am. Wow. add- dd? dd? cu m
a: ndmu. 2: nudmn. 9:. n.a_u. 9:. In. E: mu. a: on.
2:
CV 83anth c2384

 

 

 

 

 

 

 

80

wt :3. 6:80on .23 5:58 an n 30E 5. .5283 9, 2282.th 52 Bani

:5 .5239—
mm on n—

o—

 

 

 

 

\m!

 

9:. ndnu. IT

\\

 

 

8: nNdNu. ll

 

2.. 2:". lol

\x

 

9:. :u. '0'.

 

 

\
\

 

 

mu: Wﬂu U

 

 

Ecua'lnl

 

 

 

 

 

 

 

 

 

 

 

 

3.

cm-

2-

2-

c—

m—

cm

mm

(3) ammodma 1,

81

c—

9: 3:83

h

E :3. 6:80qu 5:5 £E=E an n 30E 8.. 5:89— m> 2223509 :62 Paar;

 

 

 

 

 

\\

 

3.. nan“ |.|.

a; 3:". Incl

\w

 

9:. 03"“ Illl

 

 

 

 

 

 

W

 

 

 

 

 

 

 

(3) madam;

82

3 =3— .ocaoovoo 505:5 5:58 an n BOE a _< 035.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E 2.. do 2 G 3. 3. 3 3 0.: «.3 0.2 ZN So
3 2: 2: «a as 3 3 an 3. 5.: mil. ....2 :1.th -- Im»..€:--
S. 2: n: I: 3 as a.a 3 2 S: q: -IQEI OS; .2.
5 ed 2: do a.» an 3 3 3 2 2. a.a SN 2 .2
3 E as 2 ca 5. an a.a an a.a 2: S: a.a of;
an 3 3 a.a no he a.a 2. 3 n: ..2 ..2 EN 2.2
E 2. “.2 nd n: «.2 a: as n: 2: ”.2 2m a.a 2::
ed 0.: 3. S; n: a.a n: 3: 2; E. q: q: as 3.2
E. a: S: G. 3.. n: 3.. 0.2 a: 3: 2: can 2N 3.8
3; I: ..2 a2 ..2 2.. «.2 a: v.2 2: 3: «.2 EN 32
a: 3. at q: :2 3; Ni 5.2 n: 2: 2: 82 NS 8.3
.Jﬂl i3. 2. E «a a: n2 3: 3.. 2: 2: n2 «.2 3.2
i W! 79 «a: a: a.a 0.: ..2 0.: n: S: 2: 3m «.8 2.:
.3 33$ 2 new". 2 ans". 2 0.5". E 2.3.... 2 ans". 2 2.8". 238.... ... of". 2%.? :33". 2 23.... E on,
$6.5

 

 

 

 

 

 

 

 

 

CV 388%th

 

 

 

 

 

 

c2684

 

 

g and .958qu 3223 5:55 mm H Bo: 5.. .5233 m> oSEuQEu... z _< uSwE

a: .233
cm mm cm 2 S n

 

 

 

83

 

 

a $3". loll

 

E ”2?. loll . \\\\

 

E 5.3.". IT V\\
E maﬁa". lain

 

... 92". Ilol \K

 

... gnu. l2] \\

 

(3) mmmodmol

 

E mad". llUl

 

 

\

... on. If...

 

 

 

lb
\Eg
//
x.

 

0
0
0

D
1

 

 

 

 

 

 

 

 

 

 

 

...”an

gann—sseas
We ...
a—nc‘ggﬁh""

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(3) modulo;

lion (in)

= 53 mllmin Without dodecane, Run # l

 

 

d «a a H H H “ a a U

wwwwwwwww
hanqn—nnmq

gnawed—35.45

 

 

 

 

 

 

\\\\\

 

 

NM

 

\M

 

\ \\

 

 

\

\\

 

\

\ m

 

 

 

 

 

 

 

15

10

In

C

(3) man;

-10

10

6.5 7 5 8.5 9.5
Location (in)

5.5

Figure A1 1.1: Temperature vs Location for Flow = 53 mllmin Without dodecane. Run # l

85

2 ea .8888 3223 case mm u >55 “NE 33.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E.— n.: 2 2 _.v~ mm
as— n.2 3: N6— nén wmgm
2.— 2.— o.n_ e. ..v~ 3.8
S «.3 0.2 ww— ~.v~ 3
5.3 03 On— 3: SN 8.:
c. 2 v.2 0.2 m6— adu end—
md a.a 2. We 9mm 2
5.— ni v.2 E New No.3
02 ~.m_ 0N— ~.m_ min nd
..2 ad— 2: w.: New mm.»
m6 v...“ «6 es «in b
2- mg. n._. ha «ﬁn and
..m- a.a- vd- 2.- ..vm mod
2:- a.a? «.3- v.2- 9mm and
EN. haw- 8- an. Emu N
E 5.3". E 2.2"“ .2 be; .u. 2 no". E on.
2:
CV 2380mth guano;

 

 

2 SE 65895 305$» 5:58 mm u 32..”— .8 5:33 m> Baﬂuqsok ”a: Esmi

E: .5333
3 cm 2 c—

86

 

 

 

 

 

2 3.8". III
... 2.2.... IATI
... a: E IITI

2 3". IT

5 an. in!

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mm-

ON.

2.

on-

o—

m—

cm

mm

(3) axruuadmol

87

 

 

E SNNn. llolll
E 2.2"” I'nTII

E 50.2". Ill

 

 

at :3— .ocaoovoo 5223 582:. mm u 26E .8 8:83 m> 2282.th ”_.~_< 2sz

Wm w n.

N.

5

E: 8:83

n6

6

On

We

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A:

~—

(3) modulo;

88

3 =3. 658:5 .23 5:55 8 u sea a? 2.5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

:— «.2 3.— mm
2: Q: 3.— a.a—N
as 3: s; 3.8
3- a.a «.2 E
a.a- v.7 2 «0.2
NS. Q2- a.a end—
na- Qua- n n—
ag T m- v.2 no.2
a.a- he. a a.a
ms- ﬁe. né an.»
«.0. m6- T 5
n6- 2- En. wad
n6- ﬂu- o; 7 mod
ﬂ? ﬁn- 2- and
0. n6. Nam- n
12 3.3"“ a: on. a: an.
_ ea
_ ADJ 833K805. 8:83

 

89

as 8:83

ON 2 c—

3 =3— .oswoovoa 55» 5:58 3 u 30E .8”. 8:83 m> BEER—Eon. ”n; ﬂawE

 

 

 

 

 

3.. 3.3". llll
mar— 0"“ '0'

9:. an. IT

 

 

 

//.m
«(Ki

 

 

 

\.

 

 

 

 

 

 

 

 

mm-

om-

o.

n—

(3) 01mm: .1.

9O

 

 

mE mud—n. '0'

ma..— Oﬂu I

 

 

«N

g and 6.8830 53» 5:55 no u 30E 8.. =23qu m> 82anth ;.m_< oEwE

a

on

a—

5 8:83
E S

c—

n—

 

 

\\\

 

\\\

 

‘

\.

\

 

 

 

 

 

 

 

 

 

 

 

 

8-
2-
S-
u
n- m
I
m
a )
O
n
E

n—

91

a :3. .0588: .23 5:55 8 u 3oz «2 22:.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

92 m.: 5.3 92 8mm
...: «.2 2 @6— No.3
a.a a.a 3: v.2 3.2
as- a.a- m6- «(:2 3.:
n9. ﬁt. :2- 0.2 2.2
vNT 0.2- n- 2 a9:
5. n6. h; «.2 NZ:
tn- ed- Wm 3: mp.»
a.a. a 1n ms 2.
c «6 «4. m and
:o. o :6 QT 3....
:6. n... 56. «d. «3.
«A. On. ”.7 «.2- 28
E Cu. 9:. 02". a: v". a: an.
E:
g BEER—EB. 8:804

 

 

92

 

 

2.. :u. loll
3.. wen. l..|.
mun— Vﬂa '0'

E... an. IIITI

 

 

nm

N: :3— .ocuouvoo 5:5 ESE: no u 30E :8 8:83 m> 858383. HE< 95mm”.

2: 8:83

2 @—

 

 

7
\7

//

 

\\

//

 

E

/

 

M

 

 

\\
\\

 

K

r
\

\\/

/¢I4\K

 

 

 

 

 

 

 

 

on-

2.

E.

c—

m—

cm

(3) mmmadmal

93

 

 

wE :u. '0'

a; 2.", IT!

 

 

cm

02 3

Wm—

w—

e: 8:83

Q: 2

Na :3— .oc80:oo 5:5 5.5:: no u 30E .8 8:803 m> 23quth u_.v_< 8:9".

QB

@—

On—

2

 

 

\\.u
\\

 

\\

 

 

 

 

\\

 

 

 

 

 

 

 

 

 

 

 

cm.

2.

2-

o—

(3) ammadmal

94

z :3. 6.3880 85;» 5.5.5 no u 8: «2 22¢

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

”.2 90— 5.: 5: mm
2 Wm— —.V— Q: «MAN
Wm— cd— cN— W: 3.6m
M6 06 —.n 2 a—
nd? 0.. 9n- 56— N9:
2. Wm? ~67 0.: 0nd—
a.—N.. OEN- h.NN- m.» n—
wd- Wm- n6. 56— N06—
bﬁ- Nd- Wm- n.m_ no
Wm- QN- _.¢- —.: and
97 ad- —.c ad 5
ad- Nd- _.O- —.—u wwﬁ
a.a- ed- _. — u —.w.. Noé
WT v.7 Wm- ~67 wad
v.7 MA- wN- _.—N- N
mE WNJNH. 8: We". 9:. Nod": P:— OH:
:5
CV 25800th 8:803

 

 

0:80:00 50:55 5:58 we u 30E 8.: 8:803 m> 08.8098“. ”n_< Sawi

:5 8:83
R on 2 2

95

 

 

 

 

 

2.. 2.8". loll
2: we". I]?!
2.. 8.x". '0'

a: 0". ll

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ON.

2.

2-

3

m—

cm

(3) ammadmol

96

_N

ON

:5 8:803
@—

0=80:0Q 5055» 5:55 no u 30E 8.: 8:83 9» 0.220080% 3 .n_< 083k

2

m:

n—

 

\

 

 

 

 

mE mm. _ Nu: loll

\\

 

2.. we". Ilol

E San. IT

 

 

 

 

 

 

\

 

 

 

 

 

 

 

m7

2.

c:
(3) mandala .1.

o—

n—

97

28080 5;» 5:55 2 u 30E 62 038

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

v.3 v.2 «:2 hi v.3 mm
3; v.2 n6— mi «.3 :05
w.~_ N.N_ 0: mi Wm— 3.8
wd md v.0 0.: v:— 2
..n- On. 03. ﬁn v.2 No.2
5.2. ad... 53 T vd- NE on:—
_.mm. 0.3- v3N- wdm- 0.2 n—
o.0- «.0. ad- n6 WE No.2
v.0- ﬁn- :3 a.a «.2 nd
0.M- a.a- a.a m3— ud _ and
N- v.0. _.v v.2 Em: N.
wd- a.a- v 0: ..2 mad
ET YT :.N 0.0 0.: ~06
m- Wm- vd 0.: 5.3 wad
«.m- m6. 2.- m6 9n _ a
8; can. 8: Sn. 8: an. 8: ON": 8: an.
8
CV 0.880985. 8:803

 

 

 

 

 

 

98

mm

om

 

CV 8880080...
2

008080 :23 58:8 2. u 30E 8.: 8:83 m> 8380080.: 63‘ 058.:

V3

 

 

 

 

 

 

mE emu. llfll

”Eh—"a IIIOII

 

w:— wﬂa lorl

 

 

3.. 3n. '0'

 

//..m

 

50“.!!!

 

 

 

 

 

 

 

 

 

 

 

 

mm-

cm.

3.

(‘1!) “093901

c—

n—

on

99

E

0:80qu 5:5 5:55 2‘ u 305 Eu :28qu w> 00=§0QE0H ”—62 05mm

cm

3: .8583
w—

2

 

\

 

m
\

\

 

 

 

E can.

 

.l

\

 

a; :3 [IT

m:— wﬂu I'll

 

 

\

 

 

 

 

 

 

 

 

 

 

2.

S.

o—

m—

(3) ammadmal

100

Appendix B

101

w~§d ﬁg O MO «03552 had—1960 00.“ ~=0800§>ﬁ< ace"— ORDOHn— .—m O—Dah.

0583a 505;.» 23 5;»

no .0083 0008.0 33 0&0 000000 E... 00.00 «9:30.. canon n 3

0000 0000: 28000 a $52.0. 503 u .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

888 28.0 8.: 81° 88
38.8 88.0 8.5 83 83
888 :58 n2 :3 88 38.0. 88.0. as $8. 88.
«:8 88.0 2.8 :38 83 2.3 38.0 and n3 :3
83 88 can 03 88 ««.« 2.3 3..“ .2. z.”
o o a o o o
9.25 ES: 35 Ea a: 3.50 SE: 8...: 83 05
0.50. 506032 508262 0E? 2.08032 EoE>2
0.3. 050 .02 .02 .02 0.3. 0.00 .02 .02 .02
050030 5923 3 0500.60 5;) .
”wooed a 20 .8 30635.6( 2.20 0000i
38° 88° 3.: 2.. 83
8. .o 23.0 $0 :8 88
$88 288 88 :3 8:. :3 £88 2.2 :8 82
o. 8 258 :e 58 88 «2.0 888 8.“ SS 88
2. .o 88.0 o? 33 23 :3 :88 8.” R; 83
S. .o 588 as N2 83 2.8 83 8.2 :3 a: 83 88 own “an 8..
:2 88 S.» as :a $8.0 28.0 02.. 83 88.0 is $3 8.2 o: 8..“
o o o c o o o o o
eées €85 35 E3 05 €6.50 €85 GE 85 05 @5500 95.5 35 Ass 2:
05:. 2.05952 EoE>2 06:. ...—052,02 “c0596: 05:. .coEo>oz EoEu>oz
0.3. 0.0.0 .02 .02 .02 0.0.0 080 .02 .02 .02 0.0x 03¢ .02 .02 .02
2 =3. 05808 505.3 . 2 :3. 8.808 .85.? : z :3. 2.888 5.3 .

 

 

 

o I 6'53 EoE.0|0wIa>v< 820 00080

 

 

102

Freeze ﬂoat Advancement for CN = 0.00030 with Dodecane

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Run #1 Run #2
Net Net Net Rate Rate Net Net Net Rate Rate
Movement Movement Time Movement Movement Time
(in) (cm) (hrs) (in/hr) (cm/hr) (in) (cm) (hrs) (in/hr) (cm/hr)
0 0 0 0 0 0
1.84 4.67 1.42 1.30 3.29 231 5.87 1.00 2.31 5.87
0.510 1.30 1.58 0.323 0.820 0.440 1.12 1.00 0.440 1.12
0.510 1.30 3.67 0.139 0.353 1.89 4.80 7.25 0.261 0.662
0.234 0.594 5.50 0.0425 0.108 0.400 1.02 1.75 0.229 0.581
0.240 0.610 5.33 0.0450 0.1 14 0.200 0.508 5.50 0.0364 0.0924
—0.0400 .0102 3.33 -0.0120 -0.0305
Run #3 Run #4
Net Net Net Rate Rate Net Net Net Rate Rate
Movement Movement Time Movement Movement Time
(in) (cm) (hrS) (in/hr) (cm/hr) (in) (cm) (115) (In/hr) (cm/hr)
0 0 0 0 0 0
1.95 4.95 2.00 0.975 2.48 3.59 9.12 5.00 0.718 1.82
1.51 3.84 7.00 0.216 0.548 0.760 1.93 6.00 0.127 0.322
-0.120 -0.305 3.50 -0.0343 -0.0871 0.610 1.55 7.50 0.0813 0.207
0.640 1.63 4.75 0.135 0.342
0.0900 0.229 6.25 0.0144 0.0366
Freeze Front Advancement for CN of 0.00030 Without Dodecane
Run #1 Run #2
Net Net Net Rate Rate Net Net Net Rate Rate
Movement Movement Time Movement Movement Time
(in) (cm) (115) (in/hr) (cm/hr) (in) (an) 016) (in/hr) (cm/hr)
0 0 0 0 0 0
215 5.46 2.83 0.760 1.93 3.77 9.58 6.50 0.580 1.47
0.580 1.47 2.92 0.199 0.505 0.360 0.914 5.17 0.0696 0.177
0.590 1.50 3.92 0.151 0.382 -0.0300 -0.0762 4.50 000667 -0.0169
0.930 2.36 5.33 0.174 0.443
0.880 2.24 5.58 0.158 0.401
0.390 0.991 5.67 0.0688 0.175
0.100 0.254 7.33 0.0136 0.035

 

 

‘ = Began readings at second freeze pipe
“ = Began readings after ﬁrst freeze pipe was already turned on

Table B2: Freeze Front Advancement for Capillary Number of 0.00030

With and Without Dodecane

 

 

 

103

05808 5223 05 as»
38.0 use 288.0 8 e352 5:96 5.. 28.8580. 32... 38.0 ”8 03.;

:0 02:2 300:0 23 090 0000.... E... .200 00503. 505 u 3
030 0000.: 0:000... .0 nus—008 500m u .

 

 

 

 

o-d $86 8.0 ”0; card
nvad 38.0 and a.a.— annd
nu; «and and no... Na.—

0 o c
GEES 935 4:5 €84 8:
05:. 206032 80.5.2
030 0.5. .02 .02 .02
2.8080 5.3 .

 

 

9.086 u 20 8.. 80800532 220 0000.0

 

 

 

 

 

 

2.06. 586. and 36. ON _ .o-
206. 0386. 3.0. E. _ .o. 88.0. 2.6 a: d on.» 8.2 on; mad 5— .o 2..» EH 000.0
036 chad no.» 006. 0nd 3.: Sad 2:. 2.0.0. «a.a mvod mnnd cod ooh vmd
o o o o o o o o o
525 985 :5 Es «.5 9:88 95: 35 89 En 333 985 A25 45% 95
0:5. 208262 E08262 250. £0E0>oz 8080.62 05:. E08262 5050.62
0.00 330 .02 .02 .02 0.00 0.00 .02 .02 .02 0.30 0.00 .02 .02 .02
:. 5.x 2.3080 52:3 . 0.. =3. 88008 5;» . z 5m 2.838 55 .

 

Raced .- 20 3.. 808005.42 68.... 0000.0

 

104

0:80.020 55 5:55 3. 30 0:5. m> 0.00 20800532 “mm 0530

85 05F
8.2 8.8 8.2 8.8 8.3 8.8 8.2 8.2 80 8.0

 

 

 

 

 

 

//
eee

 

 

o’
(III/U!) mu ammmpv

 

/
§

 

 

 

 

 

 

 

 

 

 

 

 

 

105

an

nu

0.800—Von 55 550:. mm 80 0:5. m> 200 E0500§>e< :Nm 0530

an

9.5 05:.
n _

o.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

86

cud

ovd

cod

and

ON.—

(mm) mm anamompv

106

8.3

050039 53 5:55 mm 08 08C. m> 0.0m 0:0500:e>e< ”mam 0.530

 

 

 

 

 

 

 

 

 

 

 

 

 

 

3.5 05E.
8.2 8.: 8d— 86— 8.x 8.0 84. 8d 86
f/ 86
/
/l|ll|l././
/

6nd
8.—
on.—
8N

 

 

OWN

(ﬁt/a!) awn Inomoompv

107

ood—

cod—

8.:

9:80qu 53, 5:55 mm 8.. 25... m> 2am EuEuocu>v< :ém BawE

8d—

8.2

2.5 as:
8.”

86

oo.~

cod

 

/

 

/

/

 

/

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0°. h. we 'n v, M. N. -.
o c o o o c o o
(nun!) can}; JWOOUBAPV

O!
o

108

8.8

383°C .23 5:55 mm 08:. m> 2am EoEooS>u< :.nm Bani

 

 

 

 

 

 

 

 

 

is us:
8.2 8.8 8.2 8.2 8.“ 8.8
c
3
z/ N e
V
D.
3 A
m
z. W.
m
3 W
// 8
r 5 c

 

 

 

 

 

 

 

a.a

109

 

058600 5223 5:55 mm 8.. 2:5. m> EoEuocgc< ”—3— 2:3".

3:: 25,—.
8. n». 8.3. 8.9“ 8.8 8.3 8.3 8.2 8.3 86 86

/

 

./

/ 8.8

 

 

 

I
8
M
o'

/
. 3
(Jill!!!) 018)] Juanaommpv

 

 

%/
§ v83

 

 

K 856

Swd

 

 

 

 

 

 

 

 

 

 

 

110

coda

ocdm

0:80on 505$» Eat—E mm .8 052. m> 0.0m E0E00=a>u< 3.5m 053m

8.2

ES 0:5

8.2

 

 

2:6

 

§
6

 

.3.
c5

 

8

V.

c
(nun!) mu mamasumpv

 

 

 

 

 

 

8nd

 

 

coed

111

8.0.

8.3

0:80on :55 5:58 no 8.. 08.... m> 0.3. 80800535.. ”..mm 08mE

8.2

8.3

is 05:.
8.0

86

86 .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

..o

“1
c

M
d
(III/ﬂ!) aura mamasumpv

V.
o

md

06

112

8....

8.0.

8.3

0:80.89 285.3 5:58 no 8.. 08.... m> 0.3. 808005>u< ".dm 08$...

8&—

A25 0.5...
8.2

8.»

86

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

..c

m 0!
c o
(III/U!) mu manooumpv

V.
a

nd

ed

113

0:80qu 5.? 5:58 an 8.. 08.... m> 0.3. 80800526... ”.6. m 08wE

E... 0&8
8.8 8.8 8.2 8.2 88 8.8

. /

/ chad

/ 8.6

 

 

 

 

8 8 8

N N c—I

c o c
(In/u!) meg Juanaommpv

 

 

 

j/ 82
/ .

8.2.