3.2.3. Steam Remediation of Nitrobenzene-Contaminated Heterogeneous Aquifer

This part of the experiment was conducted in a 2-D simulation tank with the dimensions of 60 cm (length) × 50 cm (height) × 4 cm (width) (Figure 2). The steam injection well installed at the bottom of the tank for stem injection comprised a cylinder with a diameter of 4 cm and height of 7.5 cm, having many tiny pores at the top. The tiny pores were covered with gauze to prevent sand from entering the steam injection well. There were 24 sampling ports in front of the tank to monitor the temperature using a multi-channel temperature detector (Shenzhen Hua Xin Measuring Instrument Company, Shenzhen, China). Figure 2a,b illustrates the experimental schematic design and the layered heterogeneous aquifer device, and Figure 2c shows the lens heterogeneous aquifer. The physical properties of the quartz sand samples used in the tests are presented in Table 4. The

experimental filling medium of the layered heterogeneous aquifer is presented in Table 5. The purpose of the experimental design was to simulate the influence of the interface on steam migration when the upper layer was a low-permeability medium and the lower layer was a high-permeability medium in the actual site. The R values of the two groups differed by an order of magnitude, which depicts the scenario of stratified heterogeneous aquifers.

**Figure 1.** Schematic of the azeotropic test device.

**Table 3.** Specific scheme of the experiment.


**Figure 2.** Schematic (**a**) and photo (**b**) of the experimental setup for the layered heterogeneous aquifer, and schematic of the experimental device for the lens heterogeneous aquifer (**c**).

**Table 4.** Properties of quartz sand samples in the tests.



**Table 5.** Specific experimental plan for the layered heterogeneous aquifer experiment.

Note: R = hydraulic conductivity of lower layer medium Kl/hydraulic conductivity of upper layer medium Ku.

The background medium of the lens heterogeneous aquifer was filled with coarse sand and the lens was filled with fine sand. Before the experiment, tap water was continuously injected at a rate of 0.5 m d−<sup>1</sup> by a peristaltic pump to ensure uniformity of the porous media and to remove entrapped air. Subsequently, 300 mg/L NB was injected into the lower left water inlet until the concentration detected at the upper right water outlet was the same as the water inlet, which simulated the formation and stabilization of the NB contaminant plume. Steam was generated at a constant rate by a 3 kW steam generator (Norbest Machinery Manufacturing Co. Ltd., Wuhan, China), which was supplied by a constant flux of water producing saturated steam at a temperature of 120 ◦C and pressure of 200 kPa. The steam flow was set to 1 kg h−1. The temperatures were continuously measured in the sandbox. Water samples were collected at selected intervals (0–270 min) using a disposable syringe. The concentration of NB in water was analyzed using liquid chromatography (HPLC, Agilent 1260, Santa Clara, CA, USA). The hydraulic conductivity ratio R is defined in the layered heterogeneous aquifer as follows:

$$\mathbf{R} = \mathbf{K}\_{\mathrm{l}} / \mathbf{K}\_{\mathrm{u}} \tag{2}$$

where Kl is the hydraulic conductivity of the lower layer medium and Ku is the hydraulic conductivity of the upper layer medium.

3.2.4. Ethanol-Enhanced Steam Remediation of Nitrobenzene-Contaminated Layered Heterogeneous Aquifers

The experimental device shown in Figure 2 was the same as the previous experimental device. An experimental setting with a hydraulic conductivity ratio R of 381 in the layered heterogeneous aquifer remediation experiment was selected to accelerate the remediation rate experiment. On the completion of the experiment, samples were analyzed for the distribution of pollutants in the aquifer. Furthermore, 500 mL of ethanol at a flow rate of 5 mL/min with a peristaltic pump was injected above the interface to enhance the repair efficiency of the interface in the upper low-permeability zone. The injection positions are shown in Figure 2. In the lens heterogeneous aquifer, 300 mL ethanol was injected into the lens. The ethanol injection volume was calculated using Equation (3). Ethanol distribution stabilization for 2 h was conducted after injecting the ethanol. Later, the concentration distribution of the pollutants before remediation was analyzed. The concentration of NB was determined using liquid chromatography. After injecting hot steam, the sandbox temperature was continuously measured. Water samples were collected at 0–270 minintervals using a disposable syringe. Ethanol injection volume was calculated using the equation:

$$\mathbf{V} = \mathbf{A} \times \mathbf{D} \times \boldsymbol{\sigma} \times \mathbf{0}.3 \tag{3}$$

where A is the affected area of the low-permeability (cm2); D is the thickness of the simulated tank (cm); σ is the porosity of the medium; and 0.3 is the volume fraction of ethanol.
