AAnacostia Active Cappingia A tive C inacost c app ng DDemonstrationemonstration StStatusatus

Danny Reible, PhD, PE, DEE, AN ESmith Chair of Environmental eH alth

EngineeringUniversity of Texas at Austin

NATO CCMS- Ljubljana, SloveniaJune 19, 2007

2

Remedy Performance and EffectivenessRemedy Performance and Effectiveness

Short-term remedy performance. Have sediment cleanup levels been achieved afterimplementation?

Long-term remedy performance. Have sediment cleanup levels been maintained for at least 5 years, and thereafter as appropriate?

Short-term risk reduction. Have remedial-action objectives been achieved? Do data demonstrate or at least suggest a reduction in fishtissue concentrations, a decrease in benthic toxicity, or anincrease in species diversity or other community indexes after5 years?

Long-term risk reduction. Have remedial-action objectives been maintained for at least 5 years, and thereafter as appropriate? Has the predicted magnitude and timing of risk reductionbeen achieved or are they likely to be achieved?

Ideal Recovery ScenarioIdeal Recovery ScenarioActive Remediation

100

80

60

40

20

0

-10

Qua

lity

Waterway modifications Regional contamination Baseline

Remedy

-5 0 5 10 15 20Years after start of active

remediation

3

4

BackgroundBackgroundDredging effectiveness uncertain

Sediment Dredging at Superfund Megasites:Assessing the Effectiveness, National Research Council Report 2007 Site factors reduce effectiveness (e.g. debris) Monitoring has been inadequate to demonstrate effectiveness

Capping with sand easy to implement but may not be sufficiently protective (e.g. groundwater upwelling) Alternative – “active” capping

Capping with sequestering or reactive components to aid cap effectiveness Demonstration of placement and containment effectiveness in Anacostia River, Washington DC Demonstration of organoclay for NAPL containment

Creosote- Portland, Oregon Manufactured gas plant wastes – New York

EPA Hazardous Substance Research CenterEPA Hazardous Substance Research Center

South and SouthwestLouisiana State University Georgia Institute of Technology Rice University Texas A&M University University of Texas www.hsrc-ssw.org www.sediments.org

• Research and Technology Transfer • Contaminated sediments and dredged material • Historically focused on in-situ processes and risk management • Unique regional (4&6) hazardous substance problems

• Outreach • Primarily regional in scope • Driven by community interests and problems

• Courses - Application of US Sediment Remediation Guidance • Next Course, September 5-7,2007 Portland, Oregon, USA

Study Area Location

Anacostia River

Washington DC6

7

Project TasksProject TasksLab Testing/Selection of materials (2002-2003) Site Characterization (2003) Cap Placement (March-April 2004)

Apatite, Aquablok, Sand placed via clamshell Goal – place thin (6”) active layer overlain by 6” sand with conventional equipment using gravity settling to controldisturbance of soft sediment

Coke Breeze placed in laminated mat Goal- test placement of neutrally buoyant or expensivematerials in controlled manner Coke originally chosen as one of few bulk carbon sourceseconomically feasible

Also employed activated carbon and organoclays both in bulk and in mat

Monitoring performance (Ongoing through 2008)

8

0

500

1000

1500

2000

2500

3000

3500

4000

4500

0 500 3500

C /foc (ppb)

Ct/f

(

Little or no correlation with bulk solid phase concentration

R2

Bioaccumulation

Significant correlation with

concentration

1000 1500 2000 2500 3000

sed

lippp

b)

y = 1.8x = 0.78

0.00

1.00

2.00

3.00

4.00

5.00

6.00

7.00

0.00 0.50 1.00 1.50 2.00 2.50

Predicted BSAF

Mea

sure

d B

SAF

dissolved phase

9

Sorbents for Sequestration andSorbents for Sequestration and Bioavailability ReductionBioavailability Reduction

1.4 4.7 4.9 6.2Log Kf

Coke

Sand

AC/O

CSo

il

PAHs/PCBs sorbed to sorbing organic phase is less bioavailable Reduces porewater concentrations Reduces potential for accumulation in organisms

R.G. Luthy – developing method of direct addition of activated carbon (AC) to sediments Our work -use of organoclays (OC) & other amendments in caps

Measured PCB sorption coefficient (Kf)

Considerable flux reduction with AC

10

2,4,52,4,5--PCB Isolation Provided by SorbentPCB Isolation Provided by Sorbent--amended Thin Layeramended Thin Layer (1.25(1.25--cm) Capscm) Caps

0.E+00

2.E-04

4.E-04

6.E-04

8.E-04

1 100 10000 1000000 100000000

Time (years)

Flux (mg/cm2-yr)

Coke

Sand

Soil

AC

Long isolation times

0

0.1

0.2

0.3

0.4

1 10 100 1000 10000

Time (yrs)

Flux (mg/cn2-yr)

Sand Coke

Soil

AC

Diffusion Only

Advection and Diffusion

(Murphy et al., 2005)

Sediment

Sand

Sorbent x

11

12

13

Reactive Core Mat (RCM)Reactive Core Mat (RCM) ProductionProduction

14

RCM PlacementRCM Placement

15

OrganoclayOrganoclay filled mats for oil seepfilled mats for oil seep controlcontrol-- M&B SiteM&B Site

16

AquaBlokAquaBlok --Clay Polymer MaterialClay Polymer Material for Permeability Controlfor Permeability Control

139.8 140 140.2 140.4 140.6 140.8 141 141.29

10

11

Sur

face

W See e Rates – Post Placementpag

20

/d)

sch

arg

e (

cm 15

Uncapped10

i 5

fic D Sand

i 0

Sp

ec

Aquablok

-5 139.8 140 140.2

J140.4

li140.6

2004140.8 141 141.2

u an Day

S 1mith,7 2004

18

Uplift in direct response to tidal forcing with little or no lag time

19

Sediment Camera ImageSediment Camera Image –– Anacostia RiverAnacostia River

Bubble

Gas Voids

20

FIGURE 2

.

ject

0

10

20

30

0 5 10 15 20 25

04-16-04

04-23-04

04-30-04

05-07-04

05-14-04

05-20-04

hrs

Cap Deformation During the Period 4/16/4004 through 5/25/04 1500 hrs

HydroQual, Inc

Anacostia River Sediment Capping Research ProWashington, D.C.

-30

-20

-10

Meters

Def

orm

atio

n in

mm

05-25-04 at 1500

21

Gas related uplift of impermeable capGas related uplift of impermeable cap ((AquaBlokAquaBlok))

22

23

AquaBlokAquaBlok CapCapSuccessful at diverting groundwater upwelling Some heaving due to tidal pressure variations

Uplift ~ 1 mm No observable impacts

Gas accumulation led to cap uplift Uplift approximately 20 mm before rapid release Accumulation and release on 14-60 day cycle Gas release decreased significantly by second season (labile organic carbon reduction) Suggests 2 stage capping could be effective

Sand capping to exhaust labile organic carbon Clay placement in 2nd season to divert upwelling from contaminated sediment

Effectiveness of other caps?Effectiveness of other caps?

Basic Question

Are organic and metal sequestration layers more effective than sand?

24

25

Caps Effective but ContinuingCaps Effective but Continuing SourcesSources have led tohave led to SurfaceSurface RecontaminationRecontamination

CB-C1-S-06

CB-C1-S-04

CB-C1-S-02

CB-C1-SD-04

CB-C1-SD-04-D

0

1

2

3

4

5

6

7

8

9

10

11

12

Dep

th B

elow

Cap

-Wat

er In

terf

ace

(inch

es)

S ­ Sand Layer SD - Sediment Layer Month 18 Monitoring Event

Coke Breeze Mat

Native Sediment

Sand

Coke Breeze Cap Profile

Apatite-Sediment Interface

But all caps effectiveBut all caps effective -- hard tohard to differentiatedifferentiateCap PAHs Percent Sediment and C/C0 versus Depth

15

17

19

21

23

25

27

0% 50% 100% 150% C/C0 and Percent Sediment

Dep

th (c

m)

Phen Pyr B[a]A Chrys B[b]F B[k]F B[a]P % Sediment

27

Solid PhaseSolid Phase MicroExtractionMicroExtractionSorbentSorbent PolymerPolymer

PDMS (poly-dimethylsiloxane) Thickness of glass core: 114-108 µm Thickness of PDMS coating: 30-31 µm Volume of coating: 13.55 (±0.02) µL PDMS per meter of fibre

x

Field Deployment SystemField Deployment System

28

29

0

SPME Measured Porewater Profile Surface mean

5

10

15

20

25

30

Dep

th c

m

Phenanthrene Surficial Average

100 200 300 400 500 600

Concentration ng/L

30

Lessons being LearnedLessons being LearnedInnovative cap materials possible to place in thin layers (15cm) using conventional equipment with experienced contractor A laminated mat provides opportunities for controlled placement of light and/or high value materials Low permeability AquaBlok cap

Evidence of “heaving” with tidal fluctuations- no apparent impact Effectively diverted seepage further into river Trapped gas leading to irregular release

Conventional sand caps very effective Difficult to differentiate effectiveness of active caps Current pore water sampling initiative expected to better demonstrate effectiveness of active caps

Surficial sediments can be recontaminated w/o source control

31

Active Capping StatusActive Capping Status SummarySummaryActive capping can provide greater effectiveness forspecific problems

Mobile dissolved contaminants Activated carbon/coke Organoclay

Mobile NAPL Organoclay

Control of upwelling Clay polymer (AquaBlock or benthonite in mat)

Effectiveness likely better measured by dissolved concentrations, not bulk sediment concentrations Conventional sand capping effective for typicalsediment contaminants

Hydrophobic, strongly sorbed contaminants Summary in Journal Remediation Dec 06