Subscriber access provided by Brought to you by ST ANDREWS UNIVERSITY LIBRARY
Article
Reduction of Diffusive Contaminant Emissions from a Dissolved Source in a Lower Permeability Layer by Sodium Persulfate Treatment Bridget Ann Cavanagh, Paul Carr Johnson, and Eric Daniels Environ. Sci. Technol., Just Accepted Manuscript • Publication Date (Web): 11 Nov 2014 Downloaded from http://pubs.acs.org on November 11, 2014
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 28
Environmental Science & Technology
1
Reduction of Diffusive Contaminant Emissions from a Dissolved Source in a Lower
2
Permeability Layer by Sodium Persulfate Treatment
3
Bridget A. Cavanagh*, Paul C. Johnson*, and Eric J. Daniels+
4
*
School of Sustainable Engineering and the Built Environment, Ira A Fulton Schools of
5 6
Engineering, Arizona State University, Tempe, AZ 85287 +
Chevron Energy Technology Company, San Ramon, CA 94583
7 8 9 10 11 12 13 14 15 16 17 18 19 20 1
ACS Paragon Plus Environment
Environmental Science & Technology
21
ABSTRACT
22
Residual contamination contained in lower permeability zones is difficult to remediate and can,
23
through diffusive emissions to adjacent higher permeability zones, result in long-term impacts to
24
groundwater. This work investigated the effectiveness of oxidant delivery for reducing diffusive
25
emissions from lower permeability zones. The experiment was conducted in a 1.2 m tall x 1.2 m
26
wide x 6 cm thick tank containing two soil layers having three orders of magnitude contrast in
27
hydraulic conductivity. The lower permeability layer initially contained dissolved methyl tert-
28
butyl ether (MTBE), benzene, toluene, ethylbenzene, and p-xylenes (BTEX). The treatment
29
involved delivery of 10% w/w non-activated sodium persulfate (Na2S2O8) solution to the high
30
permeability layer for 14 days. The subsequent diffusion into the lower permeability layer and
31
contaminant emission response were monitored for about 240 days. The S2O82- diffused about 14
32
cm at 1% w/w into the lower permeability layer during the 14 day delivery, and continued
33
diffusing deeper into the layer as well as back toward the higher-lower permeability interface
34
after delivery ceased. Over 209 days the S2O82- diffused 60 cm into the lower permeability layer,
35
the BTEX mass and emission rate were reduced by 95-99%, and the MTBE emission rate was
36
reduced by 63%. The overall treatment efficiency was about 60 - 110 g- S2O82-delivered/g-
37
hydrocarbon oxidized, with a significant fraction of the oxidant delivered likely lost by back-
38
diffusion and not involved in hydrocarbon destruction.
2
ACS Paragon Plus Environment
Page 2 of 28
Page 3 of 28
39
Environmental Science & Technology
INTRODUCTION
40
According to USEPA (1), the most frequently used in situ soil and groundwater
41
remediation methods at Superfund sites between 2008 and 2011 include soil vapor extraction,
42
pump and treat, and chemical treatment. These emphasize the treatment of more permeable
43
subsurface zones at cleanup sites as they rely on fluid flow for removal of contaminants and/or
44
delivery of reactants. If the lower permeability zones are not also fully remediated, groundwater
45
may still be impacted for decades by diffusive emissions, (2 – 11).
46
Complete remediation of lower permeability layers is impracticable with current
47
technology at many sites. Alternative strategies involve: a) partial source treatment into the
48
lower permeability layer and b) active treatment at the higher-lower permeability interface, both
49
of which are intended to reduce emissions to the permeable zones to acceptable levels. For
50
example, Marble et al. (12) report on TCE flux reduction from lower permeability soils after
51
partial treatment by permanganate and Clifton et al. (11) report on experiments in which the
52
diffusive emission of benzene from a nonaqueous phase liquid-impacted lower permeability
53
layer was reduced by about 90% by aerobic biodegradation while dissolved oxygen was being
54
delivered by water flowing through the adjacent permeable layer.
55
This work focuses on partial source treatment of lower permeability layers containing
56
dissolved and sorbed contaminants. Situations like this exist in dissolved plumes down-gradient
57
of NAPL source zones and in source zones where the NAPL has been largely removed by prior
58
treatment or natural attenuation. The concept studied involves dissolved oxidant delivery to
59
higher permeability zone(s), which leads to diffusive loading of oxidant in the lower
60
permeability zones, eventual destruction of the contaminants, and reduced emissions.
3
ACS Paragon Plus Environment
Environmental Science & Technology
61
In situ chemical oxidation (ISCO) use has increased since early applications in 1984 (13,
62
14). The oxidant is typically dissolved in liquid solution and injected under pressure into wells,
63
but may also be delivered in gas phase. Its distribution is controlled by soil structure (14) and
64
changes in soil properties resulting from chemical oxidation reactions, such as gas generation
65
and precipitate formation. With increasing subsurface heterogeneity, remediation becomes more
66
difficult (15). ISCO delivery by diffusion into lower permeability layers has been discussed, with
67
emphasis on permanganate to treat dense non-aqueous phase liquid sources (7, 12, 16-19).
68
For petroleum hydrocarbon-impacted sites that are the focus of this work, the most
69
relevant oxidants are hydrogen peroxide (H2O2), ozone (O3), oxygen (O2), and persulfate (S2O82)
70
(14). Permanganate is used for ISCO treatment, but was not considered because it does not treat
71
benzene (14). Dissolved contaminants driving current groundwater remediation decisions are
72
MTBE (methyl tert-butyl ether) and BTEX (benzene, toluene, ethyl benzene, xylenes). Of the
73
oxidants, H2O2 and O3 are unlikely to be of use for diffusive delivery given their fast and non-
74
specific reaction rates (20).
75
Sodium persulfate has a desirable combination of high solubility and moderate reaction
76
rate. Its aqueous solubility is about 11,000 times greater than O2 (550,000 mg/L as Na2S2O8 or
77
about 440,000 mg/L S2O82-) and it is relatively more persistent and selective for petroleum
78
hydrocarbon compounds than O3 and H2O2 (21). Persulfate decomposition also leads to sulfate
79
(SO42-) production (22), which can be an electron acceptor in petroleum hydrocarbon
80
biodegradation reactions.
81 82
Persulfate can oxidize organics through “unactivated” electron transfer, although at a relatively slow rate (14). The reaction rate of S2O82- with dissolved hydrocarbons can be
4
ACS Paragon Plus Environment
Page 4 of 28
Page 5 of 28
Environmental Science & Technology
83
increased by “activation” strategies, which include elevating the temperature, or adding
84
chemicals such as highly alkaline fluid, ferrous iron, and hydrogen peroxide (22).
85
Use of S2O82- for diffusive delivery has been of recent interest. For example, Johnson et
86
al. (22) modeled heat-activated S2O82- diffusion into lower permeability layers and concluded
87
that at temperatures >50 oC, relatively short lifetimes of sulfate radicals limit diffusive delivery
88
distances to a few cm. Tsitonaki et al. (18) suggested that non-activated applications of S2O82-
89
may be preferable for remediation of lower permeability zones because a slower reaction rate
90
favors increased oxidant penetration by diffusion. Merker (23) found that S2O82- was able to
91
diffuse 6 cm into Palouse loess in 85 d after interface treatment with 0.1 M S2O82- solution.
92
This work is complementary in that it focuses on treatment of lower permeability layers
93
containing dissolved and sorbed contaminants, and is unique in that effects of treatment on
94
diffusive contaminant emissions are quantified.
95 96 97
EXPERIMENTAL DESIGN Treatment and control experiments were conducted using the two-dimensional physical
98
model shown schematically in Figure 1. The tank was prepared with two horizontal water-
99
saturated layers: an upper higher-permeability layer and a lower lower-permeability layer source
100
zone. The aquifer materials were selected to achieve about a 1000X permeability contrast so that
101
lateral flow would occur through the higher permeability layer with insignificant flow through
102
the lower permeability layer. Influent, effluent, and in-tank dissolved concentrations of
103
contaminants and oxidant, as well as flow rate and hydraulic head difference are monitored with
104
time. The diffusive emission flux [mg/cm2-d] from the lower- to higher-permeability layer was
105
determined by mass balance as in Clifton et al. (11) knowing the effluent flow rate and dissolved 5
ACS Paragon Plus Environment
Environmental Science & Technology
106
hydrocarbon concentration, where the emission flux = effluent flow rate [L/d] × effluent
107
dissolved concentration [mg/L]/interface area [cm2].
108
The use of non-activated S2O82- was evaluated based on the work discussed above and
109
wanting to maximize S2O82- delivery throughout the lower permeability layer. Although no
110
activator was added, it is possible that activation by ferrous iron occurred via release of ferrous
111
iron from minerals in the soil (14).
112
Separate treated and untreated (control) experiments were conducted using the same tank
113
and lower permeability layer to measure baseline emissions with no oxidant treatment, and
114
emissions after oxidant treatment.
115
Aquifer physical model design. The stainless steel tank is 122-cm tall, 122-cm wide, 5-
116
cm thick for the first 15-cm in from the edges, and 6.35-cm thick across the 92-cm wide window
117
installed to view packing conditions and flow. Sampling ports through the window allow water
118
collection for determining oxidant and chemical distribution. There is also a sampling port on the
119
effluent line. Geoprobe® stainless steel screen vapor/groundwater sampling implants (53-cm
120
long x 1.3-cm diameter) were used as flow manifolds at the influent and effluent sides of the
121
higher permeability layer. A constant 2.8 mL/min flow was controlled by a peristaltic pump at
122
the inlet and the saturated thickness was controlled by a constant head device at the outlet.
123
The higher-permeability layer was Quikrete® Play Sand® (99-99.9% by weight
124
crystalline silica retained in 50 mesh sieve) with measured hydraulic conductivity K = 6 x 10-2
125
cm/sec, porosity φT = 0.37 cm3-pores/cm3-soil, and fraction of organic carbon foc = 0.005 g-
126
carbon/g-soil. The lower-permeability bottom layer was 20% w/w 200 mesh silica sand mixed
127
with 80% w/w 120 mesh feldspar sand with measured K = 8 x 10-5 cm/sec, φT = 0.34 cm3-
128
pores/cm3-soil, and foc = 0.004 g-carbon/g-soil. The permeability contrast was about three orders 6
ACS Paragon Plus Environment
Page 6 of 28
Page 7 of 28
Environmental Science & Technology
129
of magnitude, and with the 0.16 cm head difference across the tank, the estimated average
130
horizontal velocity through the lower permeability layer was 15 d based on >10 m
141
treatment zone length and 90% of the initial dissolved BTEX mass
305
transformed by oxidation.
306
Intermediate and final breakdown products were not investigated in this work. Huang et
307
al. report intermediates that include tert-butyl formate, tert-butyl alcohol, methyl acetate, and
308
acetone from S2O82- degradation of MTBE (29) and acetone from other hydrocarbons (21). 17
ACS Paragon Plus Environment
Environmental Science & Technology
Page 18 of 28
309 1.5 1.25
1.5
Mass Remaining in Low K Layer Cumula ve Mass Removed via Effluent MTBE Control M(t= - 7d) = 0.74 g
1.25
1
1
0.75
0.75
0.5
0.5
0.25
0.25
0
1.25
Mass Remaining in Low K Layer Cumula ve Mass Removed via Effluent BTEX Control M(t= - 7d) = 0.51 g
-7d 29d 58d 89d 111d 141d 171d 209d 1.5
Mass Remaining in Low K Layer Cumula ve Mass Removed via Effluent
1.25
1
1
0.75
0.75
0.5
0.5
0.25
0.25
0
310
MTBE Treated M(t= - 7d) ) = 0.51 g
0
-7d 29d 59d 89d 118d 148d 178d 208d 1.5
Mass Remaining in Low K Layer Cumula ve Mass Removed via Effluent
BTEX Treated M(t= - 7d) = 0.53 g
0 -7d 29d 59d 89d 118d 148d 178d 208d
-7d 29d 58d 89d 111d 141d 171d 209d
311 312 313
Figure 4. Mass balance results for control (left plots) and treatment (right plots) experiments, with results normalized to the dissolved source mass at t = -7 d.
314
Partitioning experiments were conducted to determine sorption coefficients for MTBE
315
and the BTEX components in the lower permeability material. Those led to the conclusion that
316
the Ks values were at least an order of magnitude lower than expected from typical correlations,
317
and likely
