4.5.0 Surfactant Mobilization (Cell 5)

Essentially seven studies were performed. 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 by Michigan Technological University using GC/MS for a group of chemicals initially believed to be representative of the site and of regulatory concern.  The selected chemicals represented several classes of contaminants. Following the installation of the instrumentation, the formation was allowed to reequilibrate 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.  The static samples would be representative of 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 of NAPL in the formation.  Once the pre-remediation tracer study was completed, the remedial technology was implemented and its performance monitored.  At the termination of the remedial demonstration, the three characterization steps were repeated in reverse order. A time line for the activities is shown in table 4.5.0.3. An analysis of this data permits an evaluation of the technology.

The objective of the study was to mobilize the contaminants by reducing the interfacial tension and solubilize the contaminants by increasing the solubility in the remedial fluid. The remedial fluid was a solution of 2.1% Tween^TM 80 and 2.2% Aerosol^TM OT.  From a theoretical standpoint, only one pore volume of remedial fluid should be needed for a homogeneous isotropic media that does not have significant dispersion. Since the actual test cell is heterogeneous and the NAPL distribution is also hetrerogeneous within the test cell, more than one pore volume is actually required for a successful remediation.  In this study, approximately 2 pore volumes were used.  It was hoped that two pore volumes would be adequate to demonstrate the effectiveness of the surfactant mobilization system.

Table 4.5.0.1 Operating Conditions for the Experiments

Parameter	Activity
	Pre-Remediation Tracer	Demonstration	Post-Remediation Tracer
Average Saturated Thickness (m)	5.14	5.41	5.21
Average Head Across Cell (m)	0.37	2.32	0.47
Average Influent Flow rate (lpm)	6.70	6.06	6.25
Average Effluent Flow rate (lpm)	6.89	6.02	6.32

The high head gradient during the experiment showed a significant effect of increased viscosity of the fluids used during the demonstration phase of the work.  In addition to the remedial fluid viscosity, it is possible that liquid crystals or macromolecules were developed when the remedial fluid combined with the contaminants. This head loss through the cell could have serious implications to the application of the technology full scale. It would be necessary to use either very low injection rates or space wells close together to permit implementation.

Table 4.5.0.2 Properties of the Remedial Fluid

Property	TweenTM 80	AerosolTM OT	Mixture
Weight % in mixture	2.1	2.2
Density (g/cm3)	ND*	ND
Viscosity (cP @ 25oC)	ND	ND
Boiling Point (oC)	ND	ND
Flash point (oC)	ND	ND
Molecular weight (g/mol)	ND	ND	...
Aqueous solubility (wt % @ 25oC)	ND	ND

* Not determined

Table 4.5.0.3 Study Sequence

Test	Activity	Fluid injected *	Total flux rate (lpm)- volume (l)	Duration
Core and install instrumentation	Collect soil samples	None		11/28/95 - 12/1/95
Ground-water sampling	Collect water samples	None
Pre-remediation tracer	Establish flow field	Water
	Inject tracer suite 1	br, dmp, hex	5.9 (lpm)  1499 (l) Note difference in flow rate from table above	2/12/96
	Maintain flow field	Water	6.89 (lpm)?	3/12/96-3/18/96
Remedial technology	Establish flow field	Water
	Inject remedial fluid	aot100, tween80	6 (lpm)	8/13/96 -  8/20/96
	Remove surfactant	Water	6 (lpm)	8/20/96- 8/25/96
Post-remediation tracer	Establish flow field	Water
	Inject tracer suite 2	br, dmp, hex, methep	6.36 (lpm)  1419 (l)	9/10/96
	Maintain flow field	Water	6.32 (lpm)	9/10/96 - 9/18/96
Ground-water sampling	Collect water samples	None
Core	Collect soil samples	None		9/26/96 - 9/27/96

* Br = bromide, dmp = 2,2-dimethyl-3-pentanol, hex = hexanol, methep = 6-methyl-2-heptanol.

4.5.1 Core Analysis

The pre-remediation pattern of decane and undecane and trimethylbenzene is quite similar to the pattern of dichlorobenzene, as can be seen in Figure 4.5.1.2  decane, Figure 4.5.1.9  undecane, and Figure 4.5.1.7  trimethylbenzene. Observations of decane, undecane, and trimethylbenzene after remediation show similar removal as the dichlorobenzene.  Since the relatively soluble compound dichlorobenzene and relatively insoluble compounds have the similar extraction patterns, it is likely the remedial process was dominated by mobilization rather than solubilization.

All of the remaining monitored compounds (Figure 4.5.1.3 ethylbenzene, Figure 4.5.1.4  naphthalene, Figure 4.5.1.5  ortho-xylene, Figure 4.5.1.6  1,1,1-trichloroethane, and Figure 4.5.1.8  toluene) showed good removals. Consistently, a large number of samples were below quantification limits. The remaining samples appeared to be reduced in concentration. This reduction may be due to the method used for sampling where uncontaminated material was mixed with contaminated material, or some of the material was not able to be mobilized because it was too tightly bound to the formation. The first option would suggest that the delivery efficiency was not optimal (not delivering the remedial fluid to all of the points of contamination). The second alternative would suggest that the composition of the NAPL was variable even over this relatively short distance.  Overall, the removal efficiency is excellent.  The only significant concern based on core data would be the viscosity issue mentioned above.

Table 4.5.1.1 Summary of Core Data Based on Boxcar Averaged Results

Chemical	Pre-remediation concentration (mg/kg)	Post-remediation concentration (mg/kg)	Fraction removed
dichlorobenzene	7.9	0.52	0.93
1,1,1-trichloroethane	0.88	0.27	0.69
toluene	3.3	0.07	0.98
o-xylene	3.9	0.099	0.97
m-xylene
naphthalene	2.6	0.14	0.95
trimethylbenzene	3.4	0.15	0.96
decane	43	3	0.93
undecane	58	7.4	0.87
ethylbenzene	0.72	0.022	0.97

4.5.2 Tracer Analysis

* Based on zero moment of extrapolated data.

In the pre-remediation tracer test, 99% of the non-retarded tracer bromide was recovered  while 94% of the non-retarded tracer was recovered in the post-remediation study.  This would be considered good closure for the study and should make the non-retarded tracer data valuable  in analyzing flow in the test cell.  Note that the saturated heads and flux and hydraulic gradient during both tests were essentially the same, supporting the usefulness of the data.  In the pre-remediation tracer test, only 45% of the retarded tracer was recovered.  Either there was a significant amount of degradation of the DMP or there was  an analytical problem.  Several of the test cells had a significant loss of DMP that makes analysis of the data problematic.  Not including some type of loss mechanism in the analysis will yield an underestimation of S_n.  All of the approaches presented assume no transformation.  There was a significant problem with well 2 data in that there were no samples collected on the rising side of the breakthrough curve. Interpretation of this data is questionable.  The models did not converge for the conservative tracer even though the data is presented in the figure.  Even with these problems, the zero moment  analysis suggests good closure and appears more forgiving than the inversion approach. It would be possible to "tweak" the model manually by constraining the inversion approach but this was not done on any of the other evaluations and it was felt that this would be inappropriate for this document.  No real conclusions can be made on the performance for  this study based on the tracer data because of the problems mentioned.  The data is presented only for completeness.

Table 4.5.2.2. Summary of Extraction Well 1 Tracer Analysis

Extraction Well 1	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)	0.178	0.178		0.250		0.250
zero moment data	128.917	73.172		131.898		81.083
zero moment extrapolated	140.388	77.576		135.729		82.698
first moment / zero moment data (days)	0.908	1.683	0.127	2.386		2.761	0.022
first moment extrapolated / zero moment (days)	1.079	1.920	0.114	2.448		2.866	0.024
Convective Dispersive Model			0.159
dispersivity (m)	1.518	4.936		0.200*		1.76*
est. initial conc. (mg/l)	754	512		933*		727*
mean time of travel (days)	0.553	1.209		1.995*		1.738*
Stochastic model			0.233
variance in travel time	0.754	1.198		1.0*		0.1
travel time (days)	0.584	1.598		1.995*		2.215
est. initial conc. (mg/l)	760	529		933*		217

*Large standard deviation not converged.

Figure 4.5.2.1  Extraction Well 1 Pre-remediation 2,2-dimethyl-3-pentanol
Figure 4.5.2.1.b  Extraction Well 1 Pre-remediation 2,2-dimethyl-3-pentanol log
Figure 4.5.2.2  Extraction Well 1 Post-remediation 2,2-dimethyl-3-pentanol
Figure 4.5.2.2.b  Extraction Well 1 Post-remediation 2,2-dimethyl-3-pentanol log
Figure 4.5.2.3  Extraction Well 1 Post-remediation methylheptanol
Figure 4.5.2.3.b  Extraction Well 1 Post-remediation methylheptanol log

Table 4.5.2.3. 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)	0.178	0.178		0.250		0.250
zero moment data	73.744	52.808		113.769		63.582
zero moment extrapolated	105.382	54.976		113.869		66.721
first moment / zero moment data (days)	1.098	1.187	0.012	2.360		2.886	0.032
first moment extrapolated / zero moment (days)	1.951	1.324	-0.045	2.361		3.155	0.048
Convective Dispersive Model	N.C.
dispersivity (m)		1.574		0.200*		1.481*
est. initial conc. (mg/l)		310		789*		668*
mean time of travel (days)		0.858		2.0*		1.462*
Stochastic model	N.C.
variance in travel time		0.760		1*		0.1
travel time (days)		0.915		2.0*		2.188
est. initial conc. (mg/l)		314		789*		193

*Large standard deviation
N.C. Model did not converge

Figure 4.5.2.4  Extraction Well 2 Pre-remediation 2,2-dimethyl-3-pentanol
Figure 4.5.2.4.b  Extraction Well 2 Pre-remediation 2,2-dimethyl-3-pentanol log
Figure 4.5.2.5  Extraction Well 2 Post-remediation 2,2-dimethyl-3-pentanol
Figure 4.5.2.5.b  Extraction Well 2 Post-remediation 2,2-dimethyl-3-pentanol log
Figure 4.5.2.6  Extraction Well 2 Post-remediation methylheptanol
Figure 4.5.2.6.b  Extraction Well 2 Post-remediation methylheptanol log

Table 4.5.2.4. 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)	0.178	0.178		0.250		0.250
zero moment data	154.327	82.876		115.424		59.234
zero moment extrapolated	213.138	85.523		116.486		60.763
first moment / zero moment data (days)	1.128	2.173	0.135	2.389		2.653	0.016
first moment extrapolated / zero moment (days)	2.018	2.268	0.017	2.398		2.735	0.020
Convective Dispersive Model			0.199	N.C.		N.C.
dispersivity (m)	1.701	3.315
est. initial conc. (mg/l)	989	722
mean time of travel (days)	0.958	1.621
Stochastic model			0.220	N.C.
variance in travel time	0.783	0.980				0.1
travel time (days)	1.014	2.672				2.309
est. initial conc. (mg/l)	986	640				212

Figure 4.5.2.7  Extraction Well 3 Pre-remediation 2,2-dimethyl-3-pentanol
Figure 4.5.2.7.b  Extraction Well 3 Pre-remediation 2,2-dimethyl-3-pentanol log
Figure 4.5.2.8  Extraction Well 3 Post-remediation 2,2-dimethyl-3-pentanol
Figure 4.5.2.8.b  Extraction Well 3 Post-remediation 2,2-dimethyl-3-pentanol log
Figure 4.5.2.9  Extraction Well 3 Post-remediation methylheptanol
Figure 4.5.2.9.b  Extraction Well 3 Post-remediation methylheptanol log
 

4.5.3    Extraction Analysis

 There are significant problems in interpreting the extraction data if one believes the core data. The core data are probably the most reliable since it was performed in an internally consistent manner and did not have problems with interference created by the remedial fluids or the potential loss of mass during analysis by the separation of the liquid phases after or during sampling.  The  dichlorobenzene data suggests that 51% of the initial mass was removed; decane and undecane show a 23% and 12% removal, respectively.  The dichlorobenzene should be more easily solubilized than either the decane or undecane, but the data is totally inconsistent with the removal fractions observed in the core data.  Most likely, this is due to sampling and analysis problems. Naphthalene  and  ortho-xylene  showed production of the compounds.  Most likely, the problem was a co-eluted peak and insufficient methods development prior to the field activities.
 

Table 4.5.3.1 Summary of the Mass Extracted.

Chemical	Initial mass *(g)	Mass extracted (g)	Fraction removed
dichlorobenzene	540	277	0.51
decane	3900	897	0.23
naphthalene	300	1079	3.60
o-xylene	380	707	1.86
undecane	5000	579	0.12

*Initial mass estimate based on geostatistical analysis.