4.6.0 Micro-emulsions (Cell 8)

The University of Florida operated this cell in cooperation with the US EPA.  The cell was approximately 3m by 5m in surface area and was instrumented as shown in  Figure 4.6.0.1 .  The injection wells are along the south side of the cell and labeled 284x.  The extraction wells are along the north side of the cell and labeled 285x.  The numbers in red in Figure 4.6.0.1 are the elevation of the top of the casing.  The locations labeled 289x are the locations where post-remediation cores were collected.  Multilevel samplers are designated 281x, 282x, and 283x.  Objects with the designation 286x are small diameter screened wells installed to monitor the head during the experiments.

Seven tasks were performed as a part of this technology demonstration. These tasks are shown in Table 4.6.0.3. Initially, the formation was sampled by coring the formation during the installation of wells and multilevel samplers.  Descriptive logs of the wells cored are available in the database in the table labeled BoringLogs.  The core samples were analyzed for a group of chemicals believed to be representative of the NAPL and of regulatory concern.  The selected contaminants represented several classes of chemicals. Following installation of the wells and monitoring equipment, the formation was allowed to re-equilibrate before any additional measurements were made.  Ground-water samples were collected under both static and dynamic conditions.  Injecting uncontaminated ground water into the formation, simulating a forced gradient flush or water flood, created the dynamic flow conditions.  Static samples were collected to assess the maximum concentration that would be anticipated in water samples from this location.  Ground-water sampling was followed by a pre-remediation tracer study.  The pre-remediation tracer study was designed to estimate the mass and distribution of NAPL in the formation and to characterize hydrodynamic performance.  Once the pre-remediation tracer study was completed, the remedial technology was implemented and its performance monitored.  After termination of the remedial demonstration, another tracer test was conducted, followed by post-demonstration ground water sampling and core sampling.  A time line for these activities is shown in Table 4.6.0.3.  Analysis of these data permits an evaluation of the technology.

The objective of this remedial approach is to extract the NAPL by bulk partitioning into stable micro-emulsions and avoid creating macro-emulsions that could dramatically reduce the permeability of the formation.  From a theoretical standpoint due to the high solubilization power of the micro-emulsion and the mass of NAPL present in the test cell, only one pore volume of remedial fluid should be needed for a homogeneous isotropic media that does not have significant dispersion. The maximum efficiency that can be achieved by piston displacement is approximately 20 l of NAPL/1000 l of this precursor fluid. In practice, the extraction process will be less efficient because of heterogeneity of the test cell and nonuniform NAPL distribution. Therefore, nine pore volumes of remedial fluid were used in the test.  This mass of remedial fluid was selected based on the heterogeneity observed from tracer breakthrough curves.

Table 4.6.0.1 Properties of the Remedial Fluid
Property  Polyoxyethylene (10) oleyl ether n-pentanol Mixture
Weight % in mixture
3.5
2.5
  
Density (g/cm3)  
0.811
  
Viscosity (cP @ 25oC)  
3.5128
  
Boiling Point (oC)  
137.3
N.D.
Flash point (oC)  
38
N.D.
Molecular weight (g/mol)  
85.15
...
Aqueous solubility (wt % @ 25oC)
Completely miscible
2.19
completely miscible

 

 Table 4.6.0.2 Operating Conditions for the Experiments
Parameter
Activity
Pre-Remediation Tracer Demonstration Post-Remediation Tracer
Average Saturated Thickness (m)  3.0 3.0 3.0
Average Head Across Cell (m) 0.2 0.3 0.2
Average Influent Flow rate (lpm) 3.6 3.6 or 1.8 3.6
Average Effluent Flow rate (lpm) 3.6 3.6 or 1.8 3.6

 

Table 4.6.0.3 Study Sequence
Test Activity Fluid injected * Total flux rate (lpm)- volume (l)(pv) Duration
Core and install instrumentation Collect soil samples None  NA  12/15/95 - 12/18/95
Ground-water sampling Collect water samples None  NA  no data
Pre-remediation tracer Establish flow field Water  3.6 lpm 5/16/96
  Inject tracer suite 1  Meth, dmp, hex, methep  3.6 lpm
 934 l		 5/17/96
  Maintain flow field Water  3.6 lpm 7/8/96
Remedial technology Establish flow field  Water   7/16/96
  Inject remedial fluid  3% polyoxyethylene(10) oleyl ether,  2.5% pentanol  54.3 kl (9 pv)  7/17/96 - 8/3/96
  Remove pentanol  3% polyoxyethylene(10) oleyl ether  6 kl (1 pv)  8/3/96 - 8/4/96
  Remove surfactant Water  39 kl (6.5 pv)  8/4/96 - 8/20/96
Post-remediation tracer Establish flow field  Water  3.6 lpm  9/13/96
  Inject tracer suite 2  Meth, dmp, hex, methep 3.6 lpm 
900 l		  9/14/96
  Maintain flow field  Water  3.6 lpm  9/21/96
Ground-water sampling Collect water samples None  NA  no data 
Core  Collect soil samples None  NA  9/28/96
* meth = methanol, dmp = 2,2-dimethyl-3-pentanol, hex = hexanol, methep = 6-methyl-2-heptanol.
 

4.6.1    Core Analysis

Core data are plotted by constituent as a function of depth without regard for spatial location within the test cell (Figures 4.6.1.1 through 4.6.1.10).  Black data points represent samples collected prior to remediation and blue data points represent samples collected after remediation.  The vertical bars show mean concentrations for each depth zone.  Confidence limits at the 95% level are denoted by the small + at the center of the bar.  As described in Section 4.1, several methods were used to interpret the core data. Total mass of contaminant in the treatment zone, estimated by each of these methods, is displayed at the top of each figure.  The two magenta lines depict the vertical extent of the target zone and the surface area within the cell is printed at the top of the figure. Core data are displayed in the following figures: Figure 4.6.1.1 dichlorobenzene, Figure 4.6.1.2  decane, Figure 4.6.1.3  ethylbenzene, Figure 4.6.1.4  meta-xylene, Figure 4.6.1.5  naphthalene, Figure 4.6.1.6  ortho-xylene, Figure 4.6.1.7  1,1,1-trichloroethane, Figure 4.6.1.8  trimethylbenzene, Figure 4.6.1.9  toluene, Figure 4.6.1.10  undecane. If one ignores compounds with an initial concentration less than 0.1 mg/kg, it appears that NAPL removal was spatially uniform in the treatment zone with reductions of an order of magnitude or more. Since the error associated with low concentrations becomes large when one attempts to take differences, it seems logical to exclude these data.  Table 4.6.1.1 summarizes the core data results.  The performance is extremely consistent for all of the  compounds  suggesting we were successful in creating a micro-emulsion with the contaminants and move all of the compounds at one time.  If the micro-emulsion had not solubilized all of the NAPL constituents, one would anticipate seeing chromatographic separation of the compounds.  No evidence of chromatographic separation was observed. There is no indication of downward or upward migration of contaminants suggesting NAPL mobilization was not a problem. The removal fraction seemed to be approximately the same throughout the profile.  Uniform removal is important to a robust treatment system.  Successful removal in one portion of the treatment zone while concentration of the contamination in another portion of the treatment zone could lead to a more difficult problem for the treatment train in the long run.

Table 4.6.1.1  Summary of Core Data, Based on Boxcar Averaged Results
Chemical Pre-remediation concentration (mg/kg) Post-remediation concentration (mg/kg) Fraction removed
dichlorobenzene 0.93 0.027 0.97
1,1,1-trichloroethane 0.0083 0.01 -0.2
toluene 0.02 0.012 0.4
o-xylene 0.2 0.07 0.92
m-xylene 0.14 0.015 0.89
naphthalene 0.85 0.1 0.89
trimethylbenzene 2.7 0.2 0.93
decane 22 1.1 0.95
undecane 56 2.6 0.98
ethylbenzene 0.057 0.0081 0.86

 

4.6.2 Tracer Analysis


Table 4.6.2.1  A summary of the tracer activities is shown in the following table:
Tracer Pre-Remediation Post-Remediation
Concentration (mg/l) Total Mass Injected (g) Mass recovered 

(g)*

 Concentration (mg/l) Total Mass Injected (g) Mass recovered 

(g)*
Methanol  2700  2546  2436  2700  2430  2148
2,2-dimethyl-3-pentanol  930  877  822  930  837  827
Hexanol  862  813    911  859  
6-methyl-2-heptanol  526  496    627  591  
Tracer Volume (l)  943  900
Injection Time (min)  262  250

* Based on zero moment of extrapolated data.

In the pre-remediation tracer test, 96% of the non-retarded tracer (methanol) was recovered while in the post-remediation study only 88% was recovered. Errors associated with a 4% mass loss are not likely to generate major errors in the analysis of the pre-remediation conservative tracer.  Corrections to the 12% mass loss observed in the post-remediation conservative tracer, however, may need correction to provide optimal evaluation.  Corrections for mass loss have not been made in this interpretation. Analyzing the post-remediation tracer as being conservative may lead to an overestimation of the NAPL remaining since the retarded tracer was more conservative (98% recovered in the post-remediation study).  Even with this potential uncertainty, the data appear to be of a quality that quantitative statements can be made based on the tracer results. Four different approaches are presented to interpret the data. Two of the approaches evaluate the data using moment analysis calculating the first and second moments.  The first approach used all of the data, as available in the database.  The second approach extrapolated the last 25% of the data greater than the quantification limit. The breakthrough data was assumed to follow a log-linear relationship. The data was extrapolated until the concentration was 0.001 mg/l.  The remaining two methods fit the data to simple one-dimensional models.  The first model is a simple convective dispersive model, as described in Section 4.1.2.  The second model, also described in 4.1.2, is a stochastic one- dimensional model, which assumes dispersion is a result of the heterogeneity in the flow field. This approach assumes the flow field is log-normally distributed and that the contaminant is uniformly distributed within the flow field. Based on a visual examination of the plotted data, the stochastic model does not appear to be appropriate under these field conditions.

The actual NAPL mass estimate from the data set is inversely proportional to the NAPL:water partition coefficient.  To maintain internal consistency, the partition coefficients listed in Table 4.1.1 were used  for all the studies.  This was done even though the NAPL composition was quite variable across the site.  In this test cell, the NAPL was reasonably uniformly distributed through the cell with a Sn = 0.092 ± 0.011 (ignoring the estimate made by assuming a log normal distribution of the hydraulic conductivity with the stochastic model).  The post-remediation NAPL saturation was 0.022 ± 0.016.  There was more variability in the post-remediation NAPL saturation than the pre-remediation NAPL saturation estimate.  Using these estimates of NAPL saturation one would estimate 76% of the NAPL was removed.  The removal fraction based on tracer analysis is less than the removal fraction observed for individual components in the core data.  It is possible that the micro-emulsion left residual coatings on the soil particles and did not solubilize some of the very heavy tar and "pitch" compounds in the NAPL.  The compounds left behind may not be of environmental significance and behave much as naturally occurring organic carbon.  Only data for the retarded tracer 2,2-dimethyl-3-pentanol are summarized in the tables below.  Results for the other tracers are included in the hyperlinked images.

Table 4.6.2.2. Summary of Extraction Well 1 Tracer Analysis
Extraction Well 1
Pre-Remediation
Post-Remediation
  Conservative (methanol) 2,2-dimethyl 
-3-pentanol		 NAPL Saturation Conservative (methanol) 2,2-dimethyl 
-3-pentanol		 NAPL Saturation
Pulse duration (days)  0.15  0.15    0.15 0.15  
zero moment data  463.908  156.608    400.412  155.906  
zero moment extrapolated  466.144  159.195    401.529  162.626  
first moment / zero moment data (days)  1.311  2.171  0.094  1.251  1.5528  0.032
first moment extrapolated / zero moment (days)  1.356  2.342  0.103  1.267  1.899  0.071
Convective Dispersive Model      0.074     0.015 
dispersivity (m)  0.904  1.075    0.270  0.287  
est. initial conc. (mg/l)  2668  887    2423  900  
mean time of travel (days)  0.726  1.124    0.918  1.015  
Stochastic model      0.073      0.015
variance in travel time  0.592  0.635    0.342  0.352  
travel time (days)  0.734  1.132    0.920  1.021  
est. initial conc. (mg/l)  2628  868    2426  902  

Figure 4.6.2.1  Extraction Well 1 Pre-remediation tracer analysis 2,2-dimethyl-3-pentanol
Figure 4.6.2.1.b  Extraction Well 1 Pre-remediation tracer analysis 2,2-dimethyl-3-pentanol log
Figure 4.6.2.2  Extraction Well 1 Post-remediation tracer analysis 2,2-dimethyl-3-pentanol
Figure 4.6.2.2.b  Extraction Well 1 Post-remediation tracer analysis 2,2-dimethyl-3-pentanol log
Figure 4.6.2.3  Extraction Well 1 Pre-remediation tracer analysis hexanol
Figure 4.6.2.3.b  Extraction Well 1 Pre-remediation tracer analysis hexanol log
Figure 4.6.2.4  Extraction Well 1 Post-remediation tracer analysis hexanol
Figure 4.6.2.4.b  Extraction Well 1 Post-remediation tracer analysis hexanol log
Figure 4.6.2.5  Extraction Well 1 Pre-remediation tracer analysis methylheptanol
Figure 4.6.2.5.b  Extraction Well 1 Pre-remediation tracer analysis methylheptanol log
Figure 4.6.2.6  Extraction Well 1 Post-remediation tracer analysis methylheptanol
Figure 4.6.2.6.b  Extraction Well 1 Post-remediation tracer analysis methylheptanol log
 

Table 4.6.2.3. Summary of Extraction Well 2 Tracer Analysis
Extraction Well 2
Pre-Remediation
Post-Remediation
  Conservative (bromide) 2,2-dimethyl 
-3-pentanol		 NAPL Saturation Conservative (bromide) 2,2-dimethyl 
-3-pentanol		 NAPL Saturation
Pulse duration (days)  0.150  0.150    0.150 0.150   
zero moment data  523.389  171.952    430.483  158.347  
zero moment extrapolated  524.958  173.709    431.414 159.943   
first moment / zero moment data (days)  1.006  1.641  0.092  1.116  1.224  0.014
first moment extrapolated / zero moment (days)  1.025  1.704  0.096  1.126  1.266 0.018
Convective Dispersive Model      0.103      0.014
dispersivity (m)  0.300  0.378    0.213 0.221   
est. initial conc. (mg/l)  2997  1009    2509  930  
mean time of travel (days)  0.663  1.171    0.822  0.910  
Stochastic model      0.103     0.014 
variance in travel time  0.359  0.400    0.304  0.310  
travel time (days)  0.665  1.176    0.824  0.911  
est. initial conc. (mg/l)  2994  1009    2512  931  

Figure 4.6.2.7  Extraction Well 2 Pre-remediation tracer analysis 2,2-dimethyl-3-pentanol
Figure 4.6.2.7.b  Extraction Well 2 Pre-remediation tracer analysis 2,2-dimethyl-3-pentanol log
Figure 4.6.2.8  Extraction Well 2 Post-remediation tracer analysis 2,2-dimethyl-3-pentanol
Figure 4.6.2.8.b  Extraction Well 2 Post-remediation tracer analysis 2,2-dimethyl-3-pentanol log
Figure 4.6.2.9  Extraction well 2 Pre-remediation tracer analysis hexanol
Figure 4.6.2.9.b  Extraction well 2 Pre-remediation tracer analysis hexanol log
Figure 4.6.2.10  Extraction well 2 Post-remediation tracer analysis hexanol
Figure 4.6.2.10.b  Extraction well 2 Post-remediation tracer analysis hexanol log
Figure 4.6.2.11  Extraction Well 2 Pre-remediation tracer analysis methylheptanol
Figure 4.6.2.11.b  Extraction Well 2 Pre-remediation tracer analysis methylheptanol log
Figure 4.6.2.12  Extraction Well 2 Post-remediation tracer analysis methylheptanol
Figure 4.6.2.12.b  Extraction Well 2 Post-remediation tracer analysis methylheptanol log

Table 4.6.2.4. Summary of Extraction Well 3 Tracer Analysis
Extraction Well 3
Pre-Remediation
Post-Remediation
  Conservative (bromide) 2,2-dimethyl 
-3-pentanol		 NAPL Saturation Conservative (bromide) 2,2-dimethyl 
-3-pentanol		 NAPL Saturation
Pulse duration (days)  0.150 0.150     0.150  0.150   
zero moment data  416.990  138.972    408.086  151.002  
zero moment extrapolated  419.777 141.825     410.567  154.643  
first moment / zero moment data (days)  1.299  1.976 0.074   1.199  1.330  0.016
first moment extrapolated / zero moment (days)  1.366  2.141 0.081   1.223  1.428 0.024 
Convective Dispersive Model      0.098      0.012
dispersivity (m)  0.440  0.736    0.220  0.228  
est. initial conc. (mg/l)  2425  861    2356  863  
mean time of travel (days)  0.810  1.399    0.916  0.998  
Stochastic model     0.100      0.012
variance in travel time  0.430  0.544    0.311  0.315  
travel time (days)  0.814  1.420    0.919  1.000  
est. initial conc. (mg/l)  2423  861    2366  866  

Figure 4.6.2.13  Extraction Well 3 Pre-remediation tracer analysis 2,2-dimethyl-3-pentanol
Figure 4.6.2.13.b  Extraction Well 3 Pre-remediation tracer analysis 2,2-dimethyl-3-pentanol log
Figure 4.6.2.14  Extraction Well 3 Post-remediation tracer analysis 2,2-dimethyl-3-pentanol
Figure 4.6.2.14.b  Extraction Well 3 Post-remediation tracer analysis 2,2-dimethyl-3-pentanol log
Figure 4.6.2.15  Extraction Well 3 Pre-remediation tracer analysis hexanol
Figure 4.6.2.15.b  Extraction Well 3 Pre-remediation tracer analysis hexanol log
Figure 4.6.2.16  Extraction Well 3 Post-remediation tracer analysis hexanol
Figure 4.6.2.17.b  Extraction Well 3 Post-remediation tracer analysis hexanol log
Figure 4.6.2.17  Extraction Well 3 Pre-remediation tracer analysis methylheptanol
Figure 4.6.2.17.b  Extraction Well 3 Pre-remediation tracer analysis methylheptanol log
Figure 4.6.2.18  Extraction Well 3 Post-remediation tracer analysis methylheptanol
Figure 4.6.2.18.b  Extraction Well 3 Post-remediation tracer analysis methylheptanol log
 

4.6.3    Extraction Analysis

Mass extracted from the test cell was quantified for several chemicals as follows: dodecane , naphthalene , para-xylene , pentanol , total decane , trichlorobenzene , trimethylbenzene , undecane. Each of the figures shows the concentration of the individual contaminant in the produced water for each of the extraction wells and the measured flux from each of the wells.  The mass flux was evaluated by fitting a spline through the measured data and then interpreting the concentration and fluid flux at the same time.  The mass flux was summed over the total time and reported as total mass removed, as shown at the bottom of the figure.  Two alternatives could be utilized to estimate the initial mass of contaminant in the test.  First, if there were a complete analysis of the NAPL one could estimate the initial mass of a particular contaminant based on NAPL saturation from tracer data and the mole fraction of the individual contaminant in the NAPL.  This would likely give the best estimate if the flow field during the tracer experiments and the extraction test were the same since the formation volume "seen" by the tracer is likely very similar to the volume "seen" by the remedial fluid. Samples of NAPL were collected and analyzed from several locations around the site. Samples collected in this manner represent readily mobile NAPL.  It is necessary to assume that this NAPL is in equilibrium (i.e., having the same composition) with the rest of the NAPL in the immediate area.  If there is a "pitch" portion to the NAPL this fraction may not be in equilibrium with the energetically free NAPL creating a potential bias to the results.

Second, core data from the pre-remediation sampling can be used to estimate initial contaminant mass. Using core data creates a different set of  biases.  The core samples that were analyzed represent only a sub-sample of the total core removed from the formation.  Much of the formation is occupied by cobbles that were too large to be included in the subsample; therefore, the estimated mass is likely smaller than the actual mass.  This is not a significant problem when comparing core data to core data but may be a significant problem when using core data as an estimate for the starting mass in the individual test cell.  In addition to this problem, the geometry of the test cell creates dead zones or at least zones that are difficult to access.  Thus, the remedial fluid may not have swept all of the contamination.  In the analysis presented here, core data were used as a starting point. The removal fractions determined from elution analysis appear to be less than the removal fractions estimated from core data.  These discrepancies are likely due to the overestimation of the initial mass, as discussed above.

As can be seen in the table below, the injected pentanol was essentially removed from the formation. The good closure on this measurement supports good quality control for the flux measurements and suggests little loss of fluids from the cell during the remediation technology demonstration.

Table 4.6.3.1 Contaminant Extraction
Chemical
Initial mass *(g)
Mass extracted (g)
Fraction removed
Trimethylbenzene
160
1238**
 
decane
1900
695
 0.37
naphthalene
74
65
 0.88
para-xylene
 
274
 
undecane
4700
3117
 0.66
pentanol (injected)
1358000***
1259000
 0.93

* Initial mass estimate based on geostatistical analysis of core data.
** 1,3,5-trimethylbenzene was analyzed in the elution curve while 1,2,4-trimethylbenzene was analyzed in the core analysis.
*** Pentanol was not initially present as a contaminant; this is the mass injected in the cell as a component of the remedial fluid.