Relative roles of resuspended particles and pore water in release of contaminants from sediment

SEWater Science and Engineering, 2014, 7(3): 344-350do:03882/.issn.1674-2370.2014.03.009http://www.waterjournal.cne-mail: wse2008@vip. 163.comRelative roles of resuspended particles and pore water inrelease of contaminants from sedimentHong-wei ZHUl,2, Peng-da CHENG, Dao-zeng WANG*I1. Shanghai Key Laboratory of Mechanics in Energy Engineering, Shanghai Institute of Applied Mathematicsand Mechanics, Shanghai University, Shanghai 200072, P. R. China2. Key Laboratory of Waterway Dredging Technology, Ministry of Transport, National Engineering ResearchCenter of Dredging Technology and Equipment, Shanghai 201208, P R. China3. School of Mechatronical Engineering and Automation, Shanghai University, Shanghai 200072, P. R. ChinaAbstract: Sediment layers containing contaminants play a significant role in environmentalhydrodynamics. Experiments were conducted in order to characterize the relative roles ofresuspended particles and pore water under different flow and sediment conditions. A conservativetracer (NaCI) and a reactive tracer (phosphate) were used as contaminants in the bottom sediment ina laboratory flume. The mixing between the overlying water and pore water occurred over a shorttime while the desorption of contaminants from fine-grained resuspended particles lasted arelatively long time. The effects of resuspended particles and pore water on the variations of releaseflux and concentration of contaminants in water with time under different hydrodynamic conditionswere quantified. The results show that pore water dominated the initial release flux, which couldbe several orders of magnitude greater than the flux due to molecular diffusion. Flux contribution ofdesorption from sediment particles in the latter release could be equal to what was seen from porewater in the initial stage.Key words: sediment resuspension; resuspended particle; pore water; release of contaminants;release fux1 IntroductionSediment-water interactions in lakes and rivers have become very important since bottomsediments are large repositories of contaminants (Corbett 2010). When sediment resuspensionoccurs, the contaminants may not be permanently stored in the bottom sediments and mayrepeatedly be recycled (Chung et al. 2009). Resuspended particles may be an additional sourceof solutes if they react in the water column (Fig. 1) (Tengberg et al. 2003). If these particlesstay in flowing water, they can be oxidized, releasing dissolved contaminants into the watercolumn (Kalnejais et al. 2007, 2010). The impact of resuspension on water quality depends onboth hydrodynamic conditions and sediment features (Li et al. 2002). The sediment's cohesiveThis work was supported by the National Natural Science Found中国煤化工)972134 and11032007).YHCNMHG*Corresponding author (e-mail: dzwang@staff.shu.edu.cn)Received Feb. 27, 2013; accepted Jun. 6, 2013strength depends on multiple factors, including grain size, mineralogical properties, and theactivity of benthic organisms, which determine the ability of sediment to resist the shear stressimposed by the overlying water (Dey and Papanicolaou 2008). Contaminants exist in bothdissolved and solid phases. The conservative solutes have nothing to do with sediment particles.For reactive solutes, the degree of particle association has to be taken into consideration aschemical reactions can contribute to medium-term or long-term immobilization (Li et al.2004).Aquatic environmentWater flowAdvection 一+Resuspended●.▼particlesReleasingAdsorptionDesorptionDorewaterpore water↑↓SttlingSedimentMolecularmovementResuspension'」diffusionRiver bedintertace g 先Dissolved -一Adsorbed(Sediments, pore water,contaminants)contaminantFig.1 Mechanisms of interaction between particles and contaminants in conceptual modelThe purpose of this study was to investigate the characteristics of conservative andreactive contaminants released from cohesive and non-cohesive sediments over a range 0bottom shear stress values. Laboratory experiments were conducted to predict the release flux.This combination of contaminants, sediment properties, and hydrodynamics will enable a morecomplete understanding of roles of resuspended particles and pore water in the release ofpollutants from sediment.2 Materials and methods2.1 Experimental apparatusA circulating flume (Li et al. 2008) consisting of a rectangular test section with 6 m inlength, 0.2 m in width, and 0.45 m in depth was used to investigate the response of bottomsediments over a range of flow velocities (Fig. 2). Surface water flow was controlled by avariable frequency pump and a tail valve. The flow velocity varied from 0.1 m/s to 0.3 m/s.The maximum design water depth was 0.4 m.Energy6m__dissipationGlass channelSampling portsTail valveequipments ;i1mi.Im。OutleInleReservoirVariablet←一Bedform 4m_VentfrequencyValveReturm pipepumpFig. 2 Schematic diagram of experime中国煤化工YHCNMH GHong-wei ZHU et al. Water Science and Engineering, Jul. 2014, Vol. 7, No. 3, 344-35045A micro-propeller velocimeter was used to measure the flow velocity. A layer of sedimentcontaining tracers of about 5 cm in thickness was gently and uniformly laid on the bottom ofthe flume. Three horizontal arrays of sampling ports with 1 m intervals were installed throughthe sides of the flume for parallel sampling.2.2 Field sites and samplingThe sediment used in the experiment was collected from Dianshan Lake, China (Lin et al.2010). Natural sediments with dso < 0.05 mm and dso = 0.35 mm (dso is the sediment mediansize) were respectively used as fine-grained and coarse-grained particles, and also as cohesivesilt and non-cohesive sand. Prior to the experiment, the sediment was prepared following athree-step procedure: (1) the sediment was washed three times with hydrogen peroxide anddistilled water to remove impurities, (2) the sediment was washed in a concentratedhydrochloric acid for approximately 12 hours to remove adsorbed contaminants on particles,and (3) the sediment was washed again with distilled water. After being filtered and dried, thesediment was then stored for the laboratory experiment. A DDS-11A conductivity meter wasused to measure the concentration of NaCl. A 722N visible spectrophotometer was used tomeasure phosphorus concentrations following the method proposed by Zhu et al. (2011).2.3 Experimental methodsIn this study, phosphate was used as a reactive tracer (Su et al. 2011). The fate ofphosphate in aquatic environments is highly dependent on the sorptive behavior of sediment.The equilibrium adsorption capacity of sediment should be measured experimentally. Theadsorption capacity C is defined as the mass of phosphorus adsorbed per unit mass ofsediment, with a unit of mg/g. The adsorption capacity depends on the surface area and qualityof sediment. The adsorption and desorption experiments were first conducted in glass vessels.A monopotassium phosphate solution was first added to the sediment. Samples were taken atdifferent time intervals. An agitator was used during the experiments to make sure that thesediment stayed in suspension. The dimensionless concentration Cw/Co was used in this study,where Co is the initial concentration of the contaminant in pore water and Cw is theconcentration of the contaminant in the overlying water.3 Results and discussion3.1 Adsorption and desorption experimentsMigration of phosphorus in the sediment depends on the interactions between sorptioncharacteristics of sediment particles and pore water. The results of the sorption experimentsare shown in Fig. 3 and Fig. 4.中国煤化工MHCNM HG346Hong-wei ZHU et al. Water Science and Engineering, Jul. 2014, Vol. 7, No. 3, 344-3500.350.70-。0.2管08.....00:.8.------------ .己0.20- t Co= 2.5 mg/L, silt+C;=0.5g/L,sit0.60 i-0- Co= 1.5 mg/L, sit0.15-0- G=10gL,sit▲Co= 2.5 mg/L, sand▲Gs-0.5g/L, sand .-7--Co= 1.5 mg/L, sand-v-Cs= 1.0g/L, sand0.100.5520 305010203040506070 80Time (h)Time (h) .Fig. 3 Dynamic adsorption experiments with respect to Fig. 4 Dynamic desorption experiments with respectdifferent initial concentrations of phosphate Coto different sediment concentrations C;From the results shown in Fig. 3, it can be seen that the initial concentration of phosphatehad a large effect on the adsorption process. The adsorption capacity of sediment increasedwith the initial concentration of phosphate. Generally, both sand and silt would take two tofour hours to reach the equilibrium condition, and the adsorption capacity still slightlyincreased after a period of time. Interestingly, because silt was made of smaller particles,which had larger specific surface areas, it reached equilibrium more slowly than sand, and hada relatively high phosphate content. This may be explained by the fact that adsorption can bedivided into two steps: the first step is rapid surface adsorption (over minutes to hours); andthe second step is slow reaction (over days to months), which is a process of adsorbedphosphate in the solute phase diffusing to the inner grains (Wang et al. 2007). In the dynamicdesorption experiments, described in Fig. 4, the time to reach the equilibrium state wasapproximately one more hour at the sediment concentration of 0.5 g/L than the time at 1 g/L,which indicated that, with the same grain size, a higher sediment concentration correspondedto a faster arrival at the equilibrium state. Meanwhile, with the same sediment concentration, asmaller particle size corresponded to more phosphate being desorbed from the sediment, as itdid with adsorption.3.2 Molecular diffusive flux and resuspension fluxThe flux across the sediment-water interface has traditionally been assumed to bedominated by molecular diffusion when the sediment is at a steady state or in steady flow(Steinberger and Hondzo 1999). Diffusion experiments were conducted using sand and siltwith different NaCl concentrations. The variation of solute density due to the increase of theNaCl concentration was not considered in this study.Fig. 5 shows that the concentrations of NaCl in the overlying water increased with time.As the molecular diffusion was very slow, only 30% of the initial quantity was released intothe water in 24 hours. A comparison of the solute proportions in sand and silt showed thatalthough the general trend was almost the same, solutes wre r1ogned fnntar fonm silt thanfrom sand. Silt may have low porosity, resulting in a higl中国煤化工across theTYHCNMHGHong-wei ZHU et al. Water Science and Engineering, Jul. 2014, Vol. 7, No. 3, 344-35034sediment-water interface. This indicates that the physicochemical characteristics of sedimentmay also play an important role in the contaminant release process.Based on the results shown in Fig. 6, the fluxes due to bulk resuspension could be severalorders of magnitude greater than those due to molecular diffusion. Almost all the conservationsolute (NaCI) in pore water entered the overlying water when sediment resuspension occurred.As phosphorus can be stored in the aqueous phase (dissolved total phosphorus, DTP) and solidphase, there was an equilibrium sorption process, and only 12% of the phosphorus, with theresuspension of pore water, contributed to the release flux to the overlying water. Theresuspension was a significant term in release of solutes from sediment. However, some strongreactive solutes might not have been affected because most of them were stored in solid phasein sediments.035 r7 0.030 .1.2p 口 Molecular dfision (NaCI)0.30--2---40.02510 BBulk resuspension (DTP)- 0.020 .0.8-; 0.20-I 0.015宁0.15+ t0.0100.4-0.10_△. Sand0.0050.05-+- DifferenceJ051152025Time (h)Fig. 5 Non-dimensional concentration of NaCl at steadyFig. 6 Non-dimensional concentrations of NaClstate in sand and silt and their differenceand DTP at steady and suspended states3.3 Contributions from suspended particles and pore waterA series of experiments were conducted under typical flow conditions using fine- grainedsediment. When the bottom shear stress increased, more particles and pore water would enterthe overlying water. As can be seen from Fig. 7, the concentration of contaminants dependedon the shear stress. The release of NaCl was faster than that of phosphate with the increase ofshear stress. The NaCl concentration in the water body increased immediately when shearstress exceeded the critical shear stress. The contaminant release flux might have been due tothe contaminants in the liquid phase stored in the surficial sediment. Pore water might havedominated the release at the initial stage of resuspension. The DTP concentration also increasedat the early stage, but its ratio to its initial value was less than that of the NaCl concentration.Phosphates existed both in solid and liquid phases in the bottom sediment. Only thosestored in pore water were released into the overlying water at first, while others were stilladsorbed in resuspended particles. Time is needed for the adsorption-desorption process toreach an equilibrium condition. As shown in Fig. 6, when bulk resuspension occurred, theNaCl concentration did not increase much after one hour iplr rTR annuttion was as中国煤化工1.5 times that of ten hours before.TYHCNMH G .348Hong-wei ZHU et al. Water Science and Engineering, Jul. 2014, Vol. 7, No. 3, 344-3503.4 Long-term release from suspended particlesThe concentration distribution of suspended sediment shows two different mechanisms.Coarse-grained sediment may resuspend many times, if the properties of the bottom sedimentdo not change with time. For cohesive, fine grained sediment, the distribution of sedimentparticle sizes with depth and the time-changed coherency must be considered (Rubin andAkinson 2001).A typical flow condition was applied to the fine-grained sediment. Fig. 8 shows thereleased amount of total phosphorus (TP) and sediment concentration under a flow rate of 10 cm/sat water depths of h = 0.1 m and h = 0.2 m. Sediment concentration and TP concentrationslowly approached a stable value within about 12 hours. The TP concentration rapidly declinedwhile sediment concentration declined in the overlying water. After about 6 to 8 hours, therelease amount gradually stabilized, which means that all release processes reached stablestates. It can also be seen that under a specific flow rate, the depth of the overlying water alsoinfluence the release of pollutants. The resuspension release flux of pollutants decreased withincreasing water depth. This might be due to the fact that deeper water resulted in a smallerbottom shear stress, which caused a decrement in the amount of suspended particles.- + NaCl三6000.30宣0.08- ---- DTP500-ξ 0.060.20]*300.t 0.15g200H--- mat h=0.2mt 0.10- math-0.1m0.05:-△-Csath=0.2m00.050.10 0.15246810Shear stress (Nm2 )Time (h)Fig.7 Variations of non-dimensional concentrations Fig. 8 Released amount of TP and sedimentof NaCl and DTP with shear stressconcentration at certain flow conditions4 ConclusionsThrough experimental studies and theoretic analyses, the mechanisms of release ofcontaminants from bottom sediment were examined deeply. The adsorption capacities of siltwere larger than those of sand because silt has a larger specific surface area. When thesediment was static, only molecular diffusion on the sediment-water interface was considered.When the sediment resuspension occurred, all the conservative contaminants were releasedfrom pore water resuspended with sediment. When reactive tracer was contained alone, porewater played a dominant role at the early stage of pollutant release, and then the adsorbedtracer was desorbed from resuspended particles, controlling successive pollutant release. Thmixing of overlying water and pore water occurred in a hant tim"hile tho adorrption and中国煤化工desorption of contaminants from resuspended particles ocng period.FHCNMH GHong-wei ZHU et al. Water Science and Engineering, Jul. 2014, Vol. 7, No. 3, 344-350349The release flux of contaminants depended on hydrodynamic conditions, including flowvelocity and water depth. Variations in velocity and depth, which lead to an increase in bottomshear stress, will result in an increase in a large quantity of contaminants.ReferencesChung, E. G，Bombardelli, F. A, and Schladow, S. G 2009. Modeling linkages between sedimentre-suspension and water quality in a shallow, eutrophic, wind-exposed lake. Ecological Modelling,220(9-10), 1251-1265. [doi:10. 1016/j.ecolmodel.2009.01.038]Corbett, D. R. 2010. Re- suspension and estuarine nutrient cycling: Insights from the Neuse River Estuary.Biogeosciences, (7), 3289-3300. [doi:10.5194/bg-7-3289 -2010]Dey, S., and Papanicolaou, A.2008. Sediment threshold under stream flow: A state-of-the-art review. KSCEJournal of Civil Engineering, 12(1), 45-60. [doi: 10.1007/s12205-008 -8045-3]Fang, H. W., Chen, M. H, and Chen, Z. H. 2009. Surface Characteristics and Model of the EnvironmentSediment. Beijing: Science Press. (in Chinese)Kalnejais, L. H, Martin, W. R., Signall, R. P., and Bothner, M. H. 2007. Role of sediment resuspension in theremobilization of particulate-phase metals from coastal sediments. Environmental Science and Technology,41(7), 2282- 2288. [doi: 10.102 1/es061770z]Kalnejais, L. H, Martin, W. R., and Bothner, M. H. 2010. The release of dissolved nutrients and metals fromcoastal sediments due to resuspension. Marine Chemistry, 121(1-4), 224-235. [doi:10.1016/j.marchem.2010.05.002]Li, B., Zhang, K., Zhong, B. C, and Wang, D. Z.2008. An experimental study on release of pollutants fromsediment under hydrodynamic conditions. Chinese Journal of Hydrodynamics, Ser. A, 23(2), 126-133.(in Chinese)Li, J. C., Chu, J. D., and Feng, H. L. 2002. Experimental investigation of impact way of scouring andsuspending of river bottom sediment on water quality. Resources and Environment in the Yangtze Basin,11(2), 137-140. (in Chinese) [doi: 1039/jissn.1048227.2002.02.08]Li, Y. P., Jiang, Y, Li, J, Zhang, G, Ding, L, Peng, J. P., Wang, C, and Fan, L. L.2004. On the relationbetween the release rate of TN, TP from sediment and water velocity. Journal of Lake Sciences, 16(4),318-324. (in Chinese) [doi:10.3321j.issn: 1003-5427.2004.04.005]Lin, W. Q. Lu, S. Q.. and Chen, Y. Z. 2010. An application of eco-dynamic model in evaluating eutrophicationcontrol measures for Dianshan Lake in Shanghai. Shanghai Environmental Science, 29(1), 1-10. (in chinese)Rubin, H, and Atkinson, J. 2001. Environmental Fluid Mechanics, 652-695. Abingdon: Marcel Dekker.Steinberger, N, and Hondzo, M. 1999. Difusional mass transfer at sediment-water interface. Journal ofEnvironmental Engineering, 125(2), 192-200. [doi: 10.1061/(ASCE)0733-93721999)125:2(192)]Su, N., Du,J. Z., Ji, T, and Zhang, J. 2011. 2Ra and 2Ra tracer study on nutrient transport in east coastal watersof Hainan Island, China: A case of nutrients. Water Science and Engineering, 4(2), 157-169. [doi:10.388/jissn.1674-2370.2011.02.004]Tengberg, A., Almroth, E, and Hall, P. 2003. Resuspension and its effects on organic carbon recycling andnutrient exchange in coastal sediments: In situ measurements using new experimental technology.Journal of Experimental Marine Biology and Ecology, 285-286, 119-142. [doi:10.1016/S0022-0981(02)00523-3]Wang, X. Q, Li, Z, Li, P. Y, and Guo, J. S. 2007. Adsorption and desorption of phosphorus on suspendedparticles in Three Gorges area. Resources and Environment in the Yangtze Basin, 16(1), 31-36. (in Chinese)[doi: 10.3969/jissn.1004-8227.2007.01.007]Zhu, H. W, Zhang, K, Zhong, B. C, and Wang, D. Z. 2011. Effects of particles and pore water in release ofpollutants due to sediment resuspension. Chinese Journal of Hydrodynamics, Ser. A, 26(5), 631-641. (inChinese) [doi: 10.3969/j.issn1000 4874.2011.05.015]中国煤化工Yun-li YU)MYHCNM HG350Hong-wei ZHU et al. Water Science and Engineering, Jul. 2014, Vol. 7, No. 3, 344-350