310 likes | 458 Views
RECENT ADVANCES IN OUR UNDERSTANDING OF SEDIMENT-TO-WATER CONTAMINANT FLUXES: THE SOLUBLE RELEASE FRACTION. Louis J. Thibodeaux Jesse Coates Professor Gordon A. & Mary Cain Department of Chemical Engineering, Louisiana State University Baton Rouge LA. 70803 5 Feb. 2003 SEMINAR PRESENTATION
E N D
RECENT ADVANCES IN OUR UNDERSTANDING OF SEDIMENT-TO-WATER CONTAMINANT FLUXES:THE SOLUBLE RELEASE FRACTION Louis J. Thibodeaux Jesse Coates Professor Gordon A. & Mary Cain Department of Chemical Engineering, Louisiana State University Baton Rouge LA. 70803 5 Feb. 2003 SEMINAR PRESENTATION C S M E University of Delaware, Civil and Environmental Engineering, Newark, DE.
INTRODUCTION and OUTLINE • The sediment bed chemical release process is a key factor effecting water quality. • Getting the process correct is needed for confident forecasting. • Water quality models are undergoing reformulation and restructuring of both particle and soluble processes plus procedural changes in the calibration hierarchy. • Focus of this presentation is on the soluble release fraction.
INTRODUCTION and OUTLINE continued: • After covering the definitions and origins of the soluble release concepts some laboratory and field data will be presented followed by a ranking of the “likely theoretical suspects” to explain the mechanism. • A coupled bioturbation driven, water-side boundary regulated process is offered as the likely mechanism. • Field data from the Hudson River Thompson Island Pool(TIP) will be presented in some detail. • Closure will cover model successes, outstanding uncertainties and needs for further investigations.
REVIEW OF THE PARTICLE RESUSPENSION PROCESSIt occurs during storm events, primarily. The easily erodable material is quickly suspended and the bed surface becomes armored. Little more if any further erosion of the surface occurs even if the storm persist for a long time-period.
FAILURE OF THE PARTICLE RESUSPENSION MODEL AT LOW FLOWS • An early model hypothesis: since the hydrophobic organic chemicals(HOCs) are strongly bound to solids only the particles need to be tracked in the system. • Soluble release was included as a bed-side molecular diffusion process. • No muddy water present and the total suspended solids (TSS) concentration were low. The chemical release remained significant. • High flow events are few in number and endure for brief time-periods whereas the low flows endure for very long time-periods.
COMING TO GRIPS WITH THE NEW PROCESS • During low-flow time-periods adjust the “particle re-suspension” model parameters using the total chemical concentration in the water column. (USGS,WDNR). • Adopt a strict calibration hierarchy that decouples the particles and the chemical processes (Connolly. et al.). • Change from one particle size to multiple(>=3)size classes including very fine ones(Ziegler, et al., USACE). • Introduce the chemical dissolution theory rate equation to quantify the soluble fraction and obtain on-site data to quantify the mass-transfer coefficients(MTCs).
FIELD MEASURED Kf VALUES(CM/DAY) • Graphical data of measured Kf values vs. Julian day follow for the Grasse River, the Hudson River and the Kalamazoo. • Notice that the Kf vs. J-day function for each river is unique as are the ranges of the numerical values. • Water quality modelers input the Kf vs. J-day function to drive the MTC variability in the soluble release rate equation.
KALAMAZOO RIVER Kf RIVERINE RIVERINE IMPOUNDMENT LAKES
LABORATORY EXPERIMENTS WITH OLIGOCHAETES Chemical Kd(L/Kg) Kf(cm.day) Dibenzofuran 105 1.4 - 2.2 Phenanthrene 330 1.6 - 2.9 Trifluarlin 840 0.34 - 5.9 Pentachlorobenzene 1120 4.3 - 7.0 Pyrene 1230 3.3 - 6.2 Hexachlorobenxene 2240 6.8 - 8.9
DETAILED PCB FIELD STUDY • The Thompson Island pool, a six mile section on the Upper Hudson River. • Chemistry on 12 congeners over a Koc range of log(4.40) to log (6.18). • 512 observations on Kf. • Data collected over a four year time period(1996-1999). • Observations on Kf for “clear water flows”. This means those flow rates <= 10,000cfs and TSS in range of 1 to 10 mg/L.
SUMMARY OF LAB. AND FIELD EVIDENCE ON Kf • Field values range from 1 to 100 cm/day. • They increase in magnitude with increasing chemical hydrophobicity. • Generally the numerical values higher for rivers than lakes(?). • Each aquatic system has a a yet-to-be-fully explained unique annual cycle behavior pattern.
THE BUTCHER/GARVEY PROCESS MODEL(20th SETAC Conf.,1999) • They observed the field measured Kf increased with increasing Koc of the congener. • Proposed a simultaneous release model with a pore-water term and a particle release term. • The contributions(Kpw) and(Kp) in the rate equation were linear and additive. • It provided a reasonable correlation of the data but the algorithm curvature was problematic and the rate equation was without theoretical support.
TRANSPORT PARAMETERS-TIP DATA REGRESSION DERIVED________________ Season B Db RR Linear RR (cm/day) (cm2/day) ____________________________________________________ Early Spring 18.7 .0302 0.77 0.54 Spring 32.4 .0191 0.96 0.82 Summer 51.8 .00956 0.99 0.96 Fall 10.5 .00336 0.74 0.27 Winter 35.5 .00898 0.78 0.80
SUMMARY OF THEORY AND PROPOSED MODEL • Rate equation is one of chemical solubilization based of Ficks first law of diffusion. • Model fabricated from existing, well known individual processes: particle bioturbation, LEA at S/W interface and transport through waterside boundary layer. • Transparent algorithm with algebraic coupling of transport coefficients(Db &B) and thermodynamic parameter (Kd=Koc*foc).
SUMMARY OF CONSISTANCY BETWEEN MODEL AND DATA • The thermodynamic functions are consistent: Kf increases with increasing Koc and then flatten out. • The extracted transport parameters, Db for particle biodiffusion and B for the water-side boundary layer, are in good agreement with literature reported values.
UNCERTAINTIES • Other benthic processes may explain the same data. For example gas generation and some macro fauna inject fine particles directly into the boundary layer. • The cause of the annual cycle Kf behavior is unknown. Enhancing and attenuating factors include: SAV emergence, bloom and die-off, bottom feeding on algae, sunlight and temperature on formation of algal mats, seasonal flow variations, ice cover, etc.
CONCLUDING REMARKS • Modelers must use the correct process mechanism and algorithm in order to make creditable long-term concentration predictions. • The bioturbation driven process model explains many key observations. • We are generally ignorant about many aspects of the chemical release processes in aquatic ecosystems. • More lab. and field data is need; alternative models are needed as well. • We need to de-mystify the Kf annual cycle behavior patterns.