A Laboratory Column Study on Particles Release in Remediation of Seawater Intrusion Region

ZHOU Jun, LIN Guoqing, LIU Jianbo, ZHANG Peidong, and GONG Lei

?

A Laboratory Column Study on Particles Release in Remediation of Seawater Intrusion Region

ZHOU Jun1), LIN Guoqing2), *, LIU Jianbo1), ZHANG Peidong1), and GONG Lei1)

1)College of Environment and Safety Engineering, Qingdao University of Science and Technology, Qingdao 266042, P. R. China?2) Key Laboratory of Marine Environment and Ecology, Ministry of Education, Ocean University of China,Qingdao 266100, P. R. China

In coastal areas, excessive exploitation of groundwater causes seawater intrusion. In artificial recharge of aquifer remediation process, the replacement of saltwater and freshwater with each other causes colloid release, and permeability also decreases. In this paper, the aquifer samples containing minimal clay mineral (mainly illite) in Dagu River aquifer were used. Adopting horizontal column experiments, we studied the influences of seepage velocity and ionic strength onparticle release, as well as the relationship between them. In the column experiments, the Critical Salt Concentration (CSC) of the Dagu River aquifer was determined as 0.05molL?1approximately. This result was basically consistent with the DLVO theoretical calculation. For the constant seepage velocity, the salinity descending rate and partical release were slower, and the peak of particle concentration was lower. However, the total amount of released particles was almost constant at different salinity descending rate. For constant salinity descending rates, the peak of particle concentration decreased as seepage velocity increased, but the total amount of released particles rose up. The experiments also indicated the existence of a critical seepage velocity, which dropped with the decrease of salt concentration. When the concentration of NaCl solution decreased from 0.17molL?1to 0.06molL?1, the critical seepage velocity decreased from 3cmmin?1to 2.5cmmin?1, which is consistent to the results predicted by DLVO theory.

seawater intrusion; particles release; critical salt concentration; salinity descending rate; critical seepage velocity

1 Introduction

In the coastal region, the excessive exploitation of groundwater leads to seawater intrusion (Khublaryan 2008). In China, seawater intrusion phenomenon is very serious. In China, seawater intrusion phenomenon is very serious (Zhang and Li, 2012 Wang., 2012 Huang., 2013 Yang., 2012). Taking Qingdao City as an example, the seawater intrusion area reached 300km2(Sun., 2012).The preventionand recovery ofseawater intrusion havebeen a research hotspot in the environmental field. At present, the main recovery method of seawater intrusion are underground anti-seepage walls and artificial recharge (Shammas, 2008). However, due to the water sensitivity of porous medium, when high salinity seawater is replaced byfreshwater, fine particles will be released, which are then transported, re-captured and re-deposited. These processes can lead to the reduction of aquifer permeability, which in turn can render restoration efforts ineffective. Fine particles can be released as a result of change in the hydrochemistry condition and the hydrodynamic condition.

In the aspects of hydrochemistry, water sensitivity can be found in salt-fresh water transition zone (Khilar., 1983). Wang foundCritical Salt Concentration (CSC) according to the gradual changes of solution salinity. When the salinity is lower than CSC, the electrostatic repulsion between the surface of particles and that of matrix was stronger than van der waals force, leading to the release of particles from the pore surfaces (Wang., 2009). Not only the CSC, but also salinity descending rate affects the reduction of aquifer permeability. The salinity descending rate has an influence on both the amount and rate of colloids release. Abrupt changeof salt concentration causes heavy particles release, while gradual changeof salt concentration causes less release (Lin., 2012). Zhou. (2009) found that the released particles produced a series of complex particle processes, such as flocculation, deposition, and transport, thereby blocking pores of porous media and decreasing the permeability.

Hydrodynamic condition is another important factor leading to clay release. The particles adhering to the pore surfaces of the loose porous medium could be released by hydrodynamic forces. Gruebeck and Collins (1982) found the ‘critical velocity’ that leads to particle release. When it is higher than ‘critical velocity’, particles are released from the pore surfaces under the shear stress, and then transported as liquid. Arulanandan. (1980) showed that the ‘critical velocity’ was influenced by the type of clay, the ionic strength,temperature and pH. Shang. (2008 found that particles adsorbed on the infiltration wet front were released, and migrated with the infiltration wet front. In addition, the released amount of colloidal particles increased with soil water content increase. In fact, these parameters are the same set of parameters that influences the ‘critical salt concentration’, indicating that both hydro-chemically induced release and hydrodynamically induced release can occur synchronously and have interacting influences.

There are rare studies on the simultaneous influence of the salinity descending rate and seepage velocity on particles release, and the relationship between them. Our re- search used samples with the sand on Dagu aquifer containing fine clay (mainly illite), taking NaCl solution as the permeating liquid. Through DLVO theory calculation and column experiments, we determined the CSC of the sediments.Based on this, we studied the influence of the seepage velocity and the salinity descending rate on particles release, as well as the relationship between them. We made the quantitative analysis of the influence of salt concentration on critical seepage velocity.The conclusion can provide new ideas for the seawater intrusion treatment.

2 Materials and Methods

2.1 Aquifer Materials

We collected aquifer materials from the lower reachof the Dagu River near the coast of Qingdao City (Shandong Province, China). The aquifer materials consisted of unconsolidated fluvial deposits. It was then air-dried, sieved through a 2-mm screen, and stored in sealed containers.

The basic aquifer materials characterization included particle size and mineralogical analysis by x-ray diffraction. The samples consisted mainly of sand, having a mean particle size of 0.38mm and porosity of 0.23.The clay consisted of illite,kaolinite, and montmorillonite, with illite as the main component (Table 1).

Table 1 Grain size distribution of the aquifer materials

2.2 Aqueous Solutions

We used solutions of controlled salt concentrations with the fresh-water and seawater collected from the field. Standard electrolyte solutions were also prepared using NaCl at concentrations of 0.4, 0.23, 0.17, 0.11, 0.06, 0.003molL?1. Freshwater was collected from the Dagu River where the sediment samples were taken, and seawater was collected from the Jiaozhou Bay near Qingdao City. Fresh-water and seawater were both filtered through a 0.45μm membrane to remove suspended materials. The chemical composition of the water samples is shown in Table 2.

Table 2 Chemical composition of freshwater and seawater

2.3 Experiment Set-Up

Dry sediments were packed in glass columns (3.0cm i.d., 25cm length). The bulk density of the packed columns was 1.6gcm?3. The columns were horizontally arranged in order to eliminate the effect of gravity on particles transport.Thenthe columns were saturated with 0.4 molL?1NaCl solution under a partial vacuum (Fig.1). Column effluent was collected with a fraction collector, and analyzed for electrical conductivity, particle concentrations (using a spectrophoto-meter at a wavelength of 600nm). A calibration curve was developed to convert absorption units to particle mass concentrations. All experiments were conducted at ambient laboratory temperature of 23℃ (Fig.1).

2.4 Critical Salt Concentration Experiments

The saturated column was first displaced with 0.4molL?1NaCl solution, followed by 0.23, 0.17, 0.11, 0.06, and 0.03molL?1NaCl solutions successively. Column effluents were collected at a 5min interval.

Fig.1 Schematic diagram of the experimental set-up.

2.5 Salinity Descending Rate Experiments Under Different Seepage Velocities

For the salinity descending rate experiments, different seepage velocities (8.8cmh?1and 18.1cmh?1) were established with a peristaltic pump. The salinity descending rate depended on the ratio of the volume of the mixer to the velocity of inflow.The space velocity is, whereis the flow rate,is the volume of the mixer. Therefore, the rate of salt decrease is defined as

The columns were filtered with different salinity descending rates (0.11, 0.36, 0.54, 1.08h?1).

2.6 Critical Seepage Velocities Experiments Under Different Salt Concentrations

The columns were saturated by different concentrations of NaCl solution (0.17molL?1and 0.06molL?1). Then the columns were filtered by the same solution at different velocities. The velocity of inflow was from low to high.

2.7 DLVO Theory

Goldenberg (1983) indicated that when ionic strength reached tothe CSC, both interaction energy and total force acting on the fine particles are zero,sothe particle attached on the pore wall would be released. Relative to the pore wall, the particles was very small. Therefore, the particles could approximately be considered as sphere.

The van der Waals interaction energy was calculated as (Gregory, 1981)

whereis the effective Hamaker constant of particle-matrix system; λ0is characteristic length of 100nm.

The effective Hamaker constant (123) was calculated using individual Hamaker constants of particle, water, and matrix:

where11,22, andAare the Hamaker constants of the particles, fluid and matrix respectively.

The DLVO profiles for particles and their electrostatic interaction energy with pore wall were calculated as (Gregory, 1975):

whereis the dielectric permittivity of media;is the radius of the particle,is theBoltzmann constant;is the absolute temperature;is the ion valence;is the electron charge;ψ,ψare the surface potentials of the particles and the pore wall respectively;is the inverse Debye length.

3 Results and Discussion

3.1 Critical Salt Concentration

In order to study the influence of the salinity descending rate and seepage velocity on particles release, first of all, we determined the CSC of the particles in the sediment.

As shown in Fig.2, when the NaCl concentration of inflow was higher than 0.06molL?1, no released clay appeared in the effluents. However, when the concentration of the inflow was switched from 0.06molL?1to 0.003 molL?1, the particles were flushed out, and the NaCl concentration of the outflow was 0.054molL?1. It suggested that 0.054molL?1corresponds to the CSC in our sediments. This conclusion was almost equal to the calculation results of DLVO theory (Fig.3). Because of its heterogeneity in nature, the CSC is normally a value within a certain range (Blume., 2005).

Fig.2 NaCl concentration of inflow (a) and outflow (b), and particle mass concentration (c) in column outflow.

Fig.3 Interaction energy between particles and solid under different NaCl concentrations.

3.2 Effects of the Interaction of the Salinity Descen- ding Rate and Seepage Velocity on Particles Release

For the same seepage velocity, the fasterthe salinity descending rate, the higher the particles concentration in the effluent was, and the total release timewas alsoshorter (Fig.4). Under a velocity of 8.8cmh?1, when the rate of salinity decrease droppedfrom 1.08h?1to 0.36 h?1, the peak of particle concentration in the effluent decreased from 12.1gL?1to 6.3gL?1, declining by almost two times. And the total release time increased from 70 to 230min.

According to the DLVO theory, under the initial state with higher salinity, particles and pore surfaces are at attraction state. During salinity decrease, diffusion layer is thicker, and van der Waals force decreases quickly, when the interaction energy drops to zero, the particles reach static balance, and will be released afterwards. But the required time for salinity concentration matched to the required salinity by particles to reach static balance has much to do with the distance between particles and pore surfaces (Fig.3). Thereby, when the salinity descending rate is higher, more particles reach static balance in unit time. That is to say, the salinity descending rate determines the amount of particles that can be released inunit time,and as a result, the peak of particles concentration is higher.

For the constant salinity descending rate, when the velocity rose from 8.8cmh?1to 18.1cmhr1, the peak of particles concentration fell from 12.1gL?1to 9.8gL?1. Meanwhile, accumulative release time increased from 70 min to 110min (Fig.4a and Fig.5). It can be explained that, under a certain particle release speed, if the flow rate is higher, the solution volume through the column in unit time is larger, thus the particle concentration is more lower.

As shown in Table 3, for the constant seepage velocity, though the total amount of particles released was generally equal, but it decreasedslightly as the salinity descending raterose. For example, when the salinity descending rate was 0.36h?1, the total amount of particles releasedis 455mg, which is a little more than 421mg at the salinity descending rate of 1.08h?1. It can be explained that when salinity descending rate is higher, more particles will be recaptured by pore throat,and the particles in the outflow are less. This phenomenon can be pro- venby the reduction of permeability. The reduction of permeability is more obvious in the case of abrupt change of salinity than that with gradual change (Lin., 2012).

For the constant salinity descending rate, when the seepage velocity was higher, the total amount of particles flushed was larger (Table 3). This is because when the seepage velocity is higher, the released particles are subjected to more hydraulic disturbance, which is unfavorable for the particles to be re-deposited, thus more particles flow out with solution.

Table 3 Mass of clay particles and peak of particle concentration at different velocities and salinity descending rate

3.3 Critical Seepage Velocity for Particles Release Under Different NaCl Concentrations

Fig.6 shows the influence of velocity on particle release at different salt concentrations. There was a critical release velocity, exceeding which, particles began to release. When injecting the NaCl solutions of 0.17molL?1and 0.06molL?1 in the column, the critical velocities of particle release were 3cmmin?1and 2.5cmmin?1respectively.

Moreover, the critical velocity roseas ionic strength increased. This reveals the relationship between critical velocity and salt concentration. We interpret this phenomenon as that when the ionic concentration in solution is low, the repulsive force of the electric double layer becomes stronger due to thicker diffusion double layer, double layer overlap, and the rise of ζ-potential. Meanwhile, asthe electric double layer thickens, the influences of the stress force on particles become stronger, and thus it is easier for particles to be detached.

We can also find that the total amount of particles re- leased was one order of magnitude higher in 0.06molL?1than in 0.17molL?1 in Fig.6.It was likely because the interaction energy reduced asionic strength decreased, so the particle release was more sensitive to hydraulic force.

Fig.6 Relationship between velocity and concentration of particles flowing through 0.17molL?1 and 0.06molL?1 NaCl solutions.

4 Conclusion

In this paper, both the effect of seepage velocity and salinity descending rate on particles release were invest- tigated. Moreover, the relationship between seepage velocity and rate of salt decreasewas researched. The critical salt concentration (CSC) was determined approximately as 0.054molL?1by both experiment and DLVO calculation. For the constant seepage velocity, the faster the salinity descending rate, the higher peak of particles concentration is. However, the total amount of particles released is a little decrease due to the recapture of pore throat. For the constant salinity descending rate, the peak of particle concentration decreases with seepage velocity increase, but the total amount of particles released rises. The critical velocity drops with the decrease of salt concentration. Moreover, when the salt concentration is equal to or lower than the critical salt concentration, the critical velocity is also relatively low; under this condition, the influence of velocity on particle releasecan not be ignored, especially, around recharge wells.

Acknowledgements

Funding for this research was provided by the Natural Science Foundation of Shandong, China, under Grant No. ZR2014DL005. Zhou Jun was supported by the China Scholarship Council. Funding for this research was also provided by the National Natural Science Foundation of China (No.40902066), and Key Project of Science and Technology of China (No. 2013ZX07202-007).

Arulanadan, K., Gillogley E., and Tulley, R.,1980.Development ofquantitativemethods to predict criticalshear and rate oferosion of natural undisturbedcohesive soils. TechnicalReport GL-80-5, USCOE Waterways, Vickshurg, 2pp.

Blume, T., Weisbrod, N., and Selker, J. S., 2005. On the critical salt concentration for particle detachment in homogeneous sand and heterogeneous Hanford sediments., 214: 121-132.

Goldenberg,L. C., and Magaritz, M., 1983. Experimental investigation on irreversible changes of hydraulic conductivity on the seawater-freshwater interface in coastal aquifer,19(1): 77-85.

Gregory, J., 1975. Interaction of unequal double layers at constant charge., 51: 44-51.

Gregory, J., 1981. Approximate expressions for retarded van der Walls interaction.,83: 138-145.

Gruesbeck, C., and Collins, R. E., 1982. Entrainment and deposition of fine particles in porous media., 22: 847-856.

Huang, X. Q., Lin, J. Q., Gan, H. Y., Xia,Z., Zheng, Z. C., and Pan, Y., 2013.Characteristics of groundwater chemical elements variation and seawater intrusion in east coast of Leizhou peninsula., 35(3):38-48.

Khilar, K.C., Fogler, H.S., and Ahluwalia, J.S., 1983. Sandstone water sensitivity: Existence of acritical rate of salinity decrease for particle capture.,5:789-800.

Khublaryan, M.G., Frolov, A.P., and Yushmanov, I.O., 2008.Seawaterintrusion into coastal aquifers., 35: 274-286.

Lin, G. G., Wang, F., Ding, J. D., and Zheng, X. L.,2012. Experimental study of permeability mutation on salt-fresh water transition zone mutations with permeability., 3:1944-1947.

Shammas, M.I., 2008. The effectiveness of artificial recharge incombating seawater intrusion in Salalah coastal aquifer, Oman., 55: 191-204.

Shang, J., Flury, M., Chen, G., and Zhuang, J., 2008.Impact of flow rate, water content, and capillary forces on in situ colloid mobilization during infiltration in unsaturated sediments.,44: W06411, DOI:10.1029/2007WR006516.

Sun, J. M., Li, J. J., and Gao, Z. J., 2012.Study on the status quo and prevention-control measures of seawater intrusion in Qingdao city., 40(33): 16330-16332.

Wang, Y., Han, Z. Y., Chen, J. H., and Li, M., 2009. Influence of hydro-chemical action on particle release., 35 (4): 57-60.

Wang, X., Han, G., and Qiao, Y. Q., 2012.Discussion on seawater intrusion and ecological restoration of southernLiao-dong Peninsula., 38(9):60-61.

Yang, J. L., Han, D. M., Su, X. S., Xiao G. Q., Zhao, C. R., Song, Q. C., and Wang, N., 2012.Environmental tracers ( δ2H-δ18O, δ34S, δ13C) as indicators of seawater intrusion processes in the coastal karst area.,27(12):1344-1352.

Zhang, L., and Li, W. M.,2012. Monitoring the seawater intrusion at the Lingjiang coast in eastern Zhejiang.,23 (3): 43-46.

Zhou, J., Zheng, X.L., Flury, M.,and Lin,G.Q., 2009.Per- meability changes during remediation of an aquifer affected by sea-water intrusion: A laboratory column study.,376:557-566.

(Edited by Ji Dechun)

DOI 10.1007/s11802-015-2850-3

ISSN 1672-5182, 2015 14 (6): 1013-1018

? Ocean University of China, Science Press and Springer-Verlag Berlin Heidelberg 2015

(January 28, 2015; revised June 16, 2015; accepted June 27, 2015)

* Corresponding author. E-mail:lingq@ouc.edu.cn