Volatilization Rates From Dredged Material and Soils Literature Review. Indiana Harbor Canal

1
Volatilization Rates From Dredged Material and
Soils - Literature Review.Indiana Harbor Canal

• Authors Louis J. Thibodeaux (thibod _at_
  che.lsu.edu)
• R.Ravikrishna, Kalliat T. Valsaraj
• Gordon A. Mary Cain Department of Chemical
  Engineering
• and (www.che.lsu.edu)
• Hazardous Substance Research Center
  (South/Southwest)
• (www.hsrc.org/hsrc/html/ssw)
• Louisiana State University, Baton Rouge, LA 70803
• Presentation at Pan American Advanced Studies
  Institute
• in Rio de Janeiro, Brazil
• Sponsors US Environmental Protection Agency,
  Hazardous Substance Research Centers and Nation
  Science Foundation, USA

2
Outline

• Chemical Volatilization from dredged material
  beds and soil surfaces - definition
• The Volatilization process from sediment in
  piles/ layers
• Results of the Literature Review
• Modeling the process mathematically
• Implications of model/data results for Exposure
  Assessment
• Real CDF operations - Attenuating and Enhancement
  factors
• Questions and Answers

3
What is Volatilization ?

• Volatilization - Transfer of a material from a
  non-vapor phase(such as soils/ sediment/ liquids)
  to the bulk vapor phase.
• With respect to dredged material in a CDF,
  volatilization refers to the release of
  chemicals contained in the sediment matrix
  (solids/water/air) to the atmosphere.
• Volatilized chemicals can be dispersed away from
  the CDF by surface winds

4
Contaminant Loss Pathways from a Confined
Disposal Facility
5
Schematic of Contaminant Transport in Dredged
Sediment
6
Literature Findings

• Theory and models
• drying of solids 1920s, pesticide evaporation
  1970, hydrocarbon evaporation 1970
• Reports, manuscripts, other reviews, etc.
• Lab, pilot scale , field
• Tabulation of data
• PAHs, pesticides, PCBs, BTX, metals,
  dioxins/furans
• Critical evaluation

7
Benzo(a)pyrene- Comparison of Model Predicted
Fluxes and Reported Fluxes
8
Dieldrin- Comparison of Model Predicted Fluxes
and Reported Fluxes
9
Aroclor 1248- Comparison of Model Predicted
Fluxes and Reported Fluxes
10
Benzene- Comparison of Model Predicted Fluxes and
Reported Fluxes
11
Literature FluxesIHC-CDF Chemicals.
Instantaneous Measured Fluxes vs. Model Predicted
Fluxes

• 61 of time, the model predicts or over predicts
  the measured flux
• 39 of time, the model underpredicts the measured
  flux
• Measured fluxes Normalized to IHC-CDF
  concentrations only. Others parameters not
  normalized such as organic carbon content,
  effective diffusivity and wind speed.

12
Theoretical Model

• Processes accounted for
• Chemical concentration in soil
• Equilibrium desorption from solid to air in pore
  spaces
• Chemical vapor-phase molecular diffusion within
  soil air-filled pore spaces
• Chemical transport through air-side surface layer
• Chemical delivery as a flux to the atmospheric
  boundary layer

13
Transport model and Parameters

• Semi-infinite layer model derived from past
  analysis by Thibodeaux and Jury
• Surface flux depends on sediment side and
  air-side mass transfer resistance
• Nomenclature

14
Transport model and Parameters

• Semi-infinite layer model derived from past
  analysis by Thibodeaux and Jury
• Surface flux depends on sediment side and
  air-side mass transfer resistance

15
Transport model and Parameters
Mean Surface Flux

• Semi-infinite layer model derived from past
  analysis by Thibodeaux and Jury
• Surface flux depends on sediment side and
  air-side mass transfer resistance

16
Schematic of Laboratory Flux Chamber

• Surface Area exposed 375 cm²
• Air Flow Rate 1700 mL/min
• PAHs on XAD-2 resin desorbed with acetonitrile
  and analyzed using HPLC
• Flux Mass trapped / (Area Time)

17
Laboratory Simulation

• Apparatus Description
• A device 10 cm in depth with area 375 cm²
• Containing Dredged material
• Air stream 1.7 L/min flowing over the bed
• High flow rate eliminates most of air-side
  resistance to mass transfer
• Operating Conditions
• Change in humidity of incoming air from 100 to
  dry (0)
• Rewet with water to field capacity
• Rework or windrow the surface

18
Laboratory Simulation continued

• Vapor Capture and Analysis
• Solid Adsorbents used
• Organic specific tubes attached to exit ports
• XAD-2 resin filled sampling tubes
• EPA method of analysis
• Ammonia H2SO4 coated silica gel
• Hydrogen Sulfide Activated coconut charcoal
• Sediment/Dredged material source
• Indiana Harbor south of Chicago, IL.

19
Experimental Run Protocol

• RUN I Wet sediment with dry air flow
• Provide maximum initial fluxes from wet sediment
  under dry condition
• RUN II Air flow switched to 99 relative
  humidity
• Provide maximum initial fluxes under humid
  conditions
• RUN III Sediment rewetted to initial water
  content
• Provide a measure of flux expected after a rain
  event
• RUN IV Sediment remixed with dry air flow
• Provide a measure of flux from reworked sediment
  under dry conditions
• RUN V Sediment rewetted to field capacity
• Provide a measure of flux from rewetted sediment
  under dry conditions

20
Experimental Run Protocol
21
Model behavior of Aroclor 1248 flux for
conditions in experimental Runs
22
PCB Fluxes
23
PAH Fluxes
24
PAH Fluxes
25
Other Fluxes
26
CONCLUSIONS

• Small quantities of toxic substances lost to air.
• Flux starts high but falls rapidly. Within days
  to one week, flux very low and near zero
• The weathered-out surface layer creates a
  barrier that retards the escape of deeper
  originating volatiles
• Theoretical model mimics data to a high degree.
• It is better in a qualitative sense than in a
  quantitative one

27
CONCLUSIONS Contd.

• Bed Reworking returns the flux to hih value
  approaching that of fresh material
• Re-wetting or increasing humidity in the air
  produces some minor flux increases

28
Supplemental Slides
29
Indiana Harbor Canal - Laboratory Studies at
Waterways Experiment Station

• Laboratory flux chambers used to measure flux
  from Indiana Harbor Canal dredged materials.
• Effects of various event such as drying,
  re-wetting and reworking.
• Fluxes of 91 compounds (PAHs and PCBs) analyzed
  at 24 different time intervals.
• Modeled fluxes exceeded measured fluxes by upto 1
  order of magnitude (for 95 of the cases).
• Measured fluxes exceeded modeled fluxes for 5 of
  the cases.
• Model served as a good screening level estimate.

30
Surface Emission Attenuation Factors
Instantaneous Surface Flux

• Surface Drying
• Increase in sediment sorption capacity for
  chemical leads to higher sediment-air partition
  constant (Kd )leads to lower pore air
  concentrations ? LOWER EMISSIONS (self capping?)
• Rainfall
• When surface is not dry rainfall fills pore
  spaces with water. Diffusion in pore water ltlt
  Diffusion in pore air
• Effective Diffusivity (DA3 ) ? LOWER
  EMISSIONS
• Snow
• Serves as a clean surface cap. Increases surface
  mass transfer resistance (KG ) ? LOWER
  EMISSIONS

31
Surface Emission Attenuation Factors
Instantaneous Surface Flux

• Consolidation
• Decrease in dredged material porosity can occur
  due to dewatering or due to consolidation from
  sediment weight ? leads to decrease in Effective
  Diffusivity (DA3 ) ? LOWER EMISSIONS
• Cracking
• Dredged material surface must be dry to cause
  significant cracking. Dry surfaces mean high
  partition constant and therefore low fluxes.

32
Surface Emission Enhancement Factors
Instantaneous Surface Flux

• Surface Re-wetting (humidity effect)
• Rewetting dry soil surface leads to decrease in
  sediment sorption capacity for chemical ? lower
  sediment-air partition constant (Kd )leads to
  higher pore air concentrations ? EMISSIONS
  (spike)
• Surface Reworking or Winrowing
• Any mechanism causing reworking of dredged
  material results in exposure of unexposed
  sediments to air. Results in EMISSIONS .
  Resets dynamics to initial conditions (i.e. t0)
  

33
Late Breaking calculations or findings

• Dont double count chemical exposure with vapors
  and dust
• Snow cover of 1 to 3 cm reduces flux by 90 and
  96. A 11 cm layer results in 99 reduction. This
  suggests that on snow days, the volatile
  emissions is reduced dramatically

34
CLOSURE - Literature Review

• The accumulated knowledge paints a consistent
  picture and reveals the key chemodynamic
  processes that explain chemical vaporization from
  soils and sediments
• The theory and models (mathematical) are at an
  advanced state of development and verification
• There is much good and excellent data on chemical
  fluxes for many chemicals based on laboratory
  experiments primarily
• Many of the studies simulate exactly or
  approximate the volatilization process on CDFs

35
CLOSURE - Literature Review (continued)

• Mathematical models
• capture the qualitative behavior patterns of
  vaporization very nicely
• are capable of best quantitative predictions in
  the laboratory where all parameters are precisely
  known
• The quantitative features of the mathematical
  models provide the best algorithms making
  prediction
• for extending existing data
• in the absence of data
• Large-scale field studies are needed as the
  capstone quantitative test of the chemical
  emission flux algorithms

36
Effect of soil and sand caps on phenanthrene
emissions
Sand Organic Carbon 0 Sediment organic
Carbon 3
37
Comparison of Volatile losses from dredged
sediment
38
Effect of relative humidity of phenanthrene
emissions
Humid
Humid
Dry