Abstract

Soil vapor extraction (SVE) is a common and typically effective physical treatment process for remediation of volatile contaminants in unsaturated soils. SVE process was evaluated in this study to determine its effectiveness for gasoline removal using column tests with real soils. This paper serves five main purposes: it evaluates the influence of air injection and air extraction, continuous and intermittent air extraction, initial concentration of gasoline, soil water content and air flow rate on SVE for gasoline removal from sandy soils. Comparison of injection methods indicated that extraction was more efficient when air introduced from the top of the column rather than the bottom. The initial concentration of gasoline in the soil had a significant effect on the rate of extraction and the overall removal of TPH, with reduced removal efficiency observed when the initial gasoline concentration increased. It was found that continuous air extraction has the highest efficiency for gasoline removal from sandy soils. Higher venting velocities led to more rapid removal of gasoline from sandy soil columns. In addition, increased soil moisture content led to faster gasoline extraction rates. It was found that SVP has the highest efficiency for gasoline removal from sandy soils and can remediate the vadose zone effectively to meet the soil guideline values for protection of groundwater.

1. Introduction

In situ remediation of petroleum hydrocarbon contamination in soil and groundwater is an important approach for restoration of environmental quality in affected soils. Organic compounds have many adverse effects on the environment and human health due to their neurotoxic, carcinogenic, and teratogenic properties ¹. Toluene is a well-known constituent of BTEX, which is rapidly absorbed through respiratory and gastrointestinal tracts; the exposure to which results in various symptoms including weakness, headache, vertigo, and ataxia ². Remediation of contaminated soil can be achieved using in situ and ex situ techniques. Ex situ methods are prohibitively expensive in many cases due to the high cost of excavation and the difficulty in reaching contaminants found in deep soils and under buildings. The existing and potential in situ technologies for remediation of soil contaminated by petroleum include soil washing, soil vapor extraction, landfarming, soil flushing, solidification/stabilization, asphalt batching, thermal desorption, biopiles, phytoremediation, bioslurry systems, bioventing, encapsulation, and aeration ³. Soil venting, also known as soil vapor extraction, is recognized as a cost-effective technology for remediation of unsaturated soils contaminated with volatile organic compounds (VOCs) and semi-volatile organic compounds (SVOCs) ³. In addition to SVE, bioventing is also one of the standardand cost-competitive techniques ⁴ for cleaning petroleum-contaminated vadose zone soils ^{5, 6}.

SVE is one of the most common treatment procedures to remove the gasoline compounds from soil with medium to high porous media ^{7, 8}. The advantages that SVE system holds over the other used technologies are its high decontamination in short term ^{6, 9}, cost-effectiveness ^{3, 9}, simplicity of equipment, system operation, and maintenance ⁷.

Fischer et al., ¹⁰ found out that the SVE can remove toluene with an efficiency of 97% after 25 h extraction in dry soil. Qin et al., ⁹, reported that the maximum removal efficiency of chlorobenzene via SVE with initial concentration of 1.1 mg/g soil at a gas flow rate of 0.3 m3/h was 95% during an 80-hour operation. The integration of bioventing (BV) and soil vapor extraction (SVE) appears to be an effective combination method for soil decontamination ¹¹.

In this study, column tests using a natural soil were conducted to represent the current SVE technology applied to the remediation of contaminated soil. Parameters relevant to SVE operating conditions varied during these tests to elucidate the influence of air injection and air extraction, continuous and intermittent air extraction, extraction airflow rate and the initial gasoline concentration on the efficiency of soil venting.

2. Materials and Methods

The sandy soil used in this study was taken from a depth of 2.5 m the Water Science Institute Experimental Base in Yongle, Beijing. It was repeatedly washed with water to get clear water. Then it was air-dried, first at room temperature and after, it was autoclaved at 121°C, 15 psi, for 45 min. The size distribution of soil is shown in Table 1. As the 0.25-0.075 mm particles represented more than 50% of the total mass, it was classified as silt sandy soil according to soil standard 102-85 ^{12, 13}. The organic carbon content was 0.3421‰.

Figure Figure 1 shows the main component of the experimental setup. The apparatus was a Perspex column with an inner diameter of 8cm and height of 50 cm. Six 10-mm holes were drilled into the column to provide monitoring points with a distance of between holes. At the 5 cm height of the column, a perforated stainless steel plate was inserted to maintain the soil and distribution of the inlet gas uniformly.

The soil venting experiments were conducted in three stages: (1) introduction of soil and gasoline to the column; (2) an equilibration period to allow for distribution of the gasoline throughout the column; and (3) the soil venting tests. Test soil with an initial gasoline concentration of 6% (Mgasoline:Msoil) and adjusted moisture content with sterilized water was cautiously and rapidly placed into the column. Once preparation of the soil column was completed, the column was allowed to equilibrate for 24 to 48 hours. Following the equilibration step, the soil venting process was started. Finally, in order to avoid atmospheric contamination an activated carbon filter was used to absorb the the off-gas exiting from columns before reaching the atmosphere. Five sets of soil venting experiments were conducted to investigate the effects the of the following variables on the removal efficiency: (1) air injection vs. air extraction, (2) continuous vs. pulsed extraction, (3) rate of airflow, (4) initial soil gasoline concentration, and (5) soil moisture content. Detail information for each venting condition is shown in Table 2. The venting efficiency (η) of each experiment was calculated from the initial gasoline concentration (C0) and the gasoline concentration (Ct) of soil after venting using the equation η=1-Ct/C0 are also shown in Table 2.

The concentration of gasoline in the soil was quantified by an Agilent gas chromatography equipped with flame ionization detector (Agilent GC, 6820, United States). The GC-FID procedure was optimized as follows. The amount of 1 μL of extracted liquid sample was injected into the instrument. Helium with flow rate of 1.11 mL/min was used as the carrier gas and N2 with flow rate of 30 mL/min as the makeup gas. Air at 300 mL/min and H2 at 30 mL/min were used as flame gases. The characteristic of GC column was Agilent a HP-5 capillary column (30 m, 0.32 i.d). The column temperature was programmed to start at 30°C (3.5 min) then increase to 100°C at a rate of 6°·min-1, and then from 100°C to 200°C at a rate of 25°·min-1. The temperatures of the injector and detector were held fixed at 210, and 280°C, respectively. Toluene quantification was determined by a previously prepared calibration curve with correlation coefficient of 0.999. Toluene showed a retention time of 9.29 min under the experimental condition.

3. Results and Discussion

The efficiency of gasoline removal by air injection and vacuum extraction venting for a soil column with desiccated soil at an airflow of 60 mL•min-1 are shown in Table 2 (Group I). Only 49.54% of the gasoline in the soil column was removed by air-injection from the column bottom after 95 hours of venting when the initial soil gasoline concentration was 1.128 mg•g^-1. In contrast, 73.52% of the gasoline was removed by vacuum extraction from the top of the column after only 90 hours of venting with a high initial soil gasoline concentration of 2.937 mg•g-1. The efficiency of gasoline removal during injection of air into the bottom of the column was affected by airflow distribution, which is dependent on the operating conditions and capillary pressure-related characteristics of the medium and are sensitive to relatively subtle changes in soil structure ¹⁴. Air injection can also lead to inefficient removal due to the formation of preferential flow paths and in practice can cause contaminants to be trapped under impenetrable surfaces ¹⁵. While for air extraction, the volatility organic pollution and air ware extracted firstly, which caused negative pressure in soil column, then the air flew into the column due to the pressure grade between inside and outside of column. Extraction-injection and injection-extraction maybe more efficient for the smooth air flow, thus the homogeneous air distribution would be form.

The results of tests using a flow rate of 60 mL•min-1 and different initial concentrations are shown in Group II in Table 2. With an initial soil gasoline concentration of 2.934 mg g-1, 80.49% was removed after 104.5 hours. This decreased to 65.38% after 163 hours when the initial gasoline concentration increased to 9.177 mg•g-1. This decreasing trend was further proved when the initial gasoline concentration increased to 76.40 mg•g-1, for the gasoline removal efficiency was only 47.43% after 163 hours at such higher initial concentration. From these results, it was obvious that vacuum extraction efficiency decreased significantly as initial soil gasoline concentrations increasing. To reach a particular remediation goal, longer remediation times will be required for sites with higher initial contaminant concentrations than comparable sites with lower pollutant concentrations treated using the same conditions. Gasoline will be present as light non-aqueous phase liquid (NAPL) in soil when the concentration is high enough to reach residual saturation ¹⁶. This will lead to the accumulation of gasoline at the bottom of the column, reduce the amount of gasoline extracted effectively from the column, and therefore reduce the overall efficiency of soil venting. Kaslusky and Udell ¹⁷ have determined that with sufficient NAPL accumulation, gravitational forces could overcome trapping forces leading to downward flow of the liquid. When the initial gasoline concentration in column was increased to 76.40 mg•g-1, we did find NAPL concentrated at the bottom of column in this study.

Group III (Table 2) shows gasoline removal efficiencies for pulsed and continuous vacuum extraction under identical initial conditions. 89.29% of the gasoline in the soil column was removed after 264 hours under continuous vacuum extraction with an initial soil gasoline concentration of 9.177 mg•g-1, the ratio of Mass_gasoline to Mass_soil was 6%, and the airflow rate was 100 mL•min-1. Only 7.37% of the gasoline in the control column (no flow) was removed during the same period. The fraction of gasoline removed using a 12 hour•day^-1 pulsed vacuum extraction scheme was 41.82%. This is nearly half of the total removal during the eleven days continuous extraction, indicating that a significant fraction of the gasoline present in the soil was very difficult to remove using continuous extraction.

Thornton and Wootan ¹⁸ have determined that the rate of gasoline removal is dependent on the soil venting airflow rate. Higher venting velocities will cause the extraction front to reach the extraction point faster, thus lead to more rapid pollutant breakthrough. While Wilson et al. ¹⁹ have found that there is an optimal venting airflow rate and vapor flux. When the venting airflow rate increases beyond a point, the venting flux does not increase continuously. The influence of airflow rate on the efficiency of soil venting in our experiments showed in Group IV in Table 1. Only 31 hours were required to remove 96.40% of the gasoline from the soil column when the venting airflow rate was 800 mL•min-1, compared to 96 hours for 96.30% removal at 400 mL•min-1 and 142 hours for 92.00% removal at 200 mL•min-1. When the venting airflow rate was further reduced to 100 mL•min-1, it took 264 hours to remove only 89.29% of the gasoline present in the soil column. This confirms the relationship between soil venting airflow rate and the rate of gasoline removal that higher venting velocities led to more rapid removal of gasoline from sandy soil columns.

Harper et al. ²⁰ have considered that soil moisture content is one of the factors mostly affecting remediation time due to its influence on contaminant availability and soil permeability. The effect of moisture content on the efficiency of soil venting is shown in Group V of Table 2. It took 264 hours to remove 89.29% of the gasoline present in the column when soil moisture content was 0% at an initial gasoline concentration of 6% and venting airflow rate of 100 mL•min-1. While when the soil moisture content increased to 5%, 10%, and 15% and the other conditions were all like above, it only took 138 hours to remove >90% of the gasoline in columns. And the highest efficiency of 98.74% for 138 hours was achieved when the soil moisture content reached 15%. This indicates that increasing the soil moisture content in a certain range will lead to accelerated gasoline removal by soil venting. While the soil moisture content reduces the void space (porosity attributed to air) in soil, makes it more difficult for air to move into the soil and may potentially have a negative influence on the remediation process ²¹. This is because increased moisture content leads to a decrease in the mass transfer coefficient between the NAPL and gas phases by reducing the interfacial surface area between them ^{20, 22}. However, the presence of water enhances the volatilization of organic compounds by displacing them from particle surfaces, therefore increasing the rate of vapor transfer and the efficiency of remediation by reducing the vapor sorption capacity of the soil matrix. This is further enhanced by the competition between water and VOCs for mineral surfaces, which is illustrated by the ability of dry soils to retain large quantities of sVOC for long time, releasing them if there is an increase in the soil moisture content ^{23, 24}.

4. Conclusions

From column tests using a real soil, we have evaluated the efficiency of SVE under different conditions and drawn the following conclusions: i. Air extraction was more efficient than air injection. ii. The initial soil gasoline concentration affected the gasoline removal rate. When residual saturation occurred, the dominant phase of petroleum hydrocarbons was a NAPL phase and the efficiency of its removal reduced significantly. iii. There was a positive correlation between the soil venting airflow rate and rate of gasoline removal. Higher venting velocities led to more rapid gasoline removal. The presence of water enhanced the volatilization of organic compounds in gasoline, increasing the rate of vapor transfer and increasing remediation efficiency. As the moisture content of a sandy soil increased, the rate of gasoline removal and efficiency increased.

Acknowledgements

The studies described in this paper were funded jointly by the 863 Program (2007AA06A410) of the Chinese Ministry of Science and Technology and the studies of safety evaluation and pollution prevention technology and demonstration for groundwater resources in Beijing (D07050601510000).

References