VOCsLouisville 5 18a AHewitt

Multi-increment TCE Vadose Zone Investigation : 

Multi-increment TCE Vadose Zone Investigation Alan D. Hewitt, Charles A. Ramsey, and Susan R. Bigl Support: Army Environmental Center USACE-EQT Environmental Quality and Technology

UST site: two tanks positioned on concrete pad  6 m (20 ft) bgs : 

UST site: two tanks positioned on concrete pad  6 m (20 ft) bgs

UST Site History: 

UST Site History 1960 - TCE and #2 fuel oil tanks installed 1972 - TCE tank removed, 2nd #2 fuel tank installed 1989 - Both fuel tanks removed 1992 - TCE detected in vadose zone 2003 - Specified subsurface area treated with 1 to 3% KMnO4 solution 2004 - Concrete pad removed and area backfilled with clean material

Site Investigation Objectives: 

Site Investigation Objectives Average TCE concentration (mg/kg) in area treated by in-situ oxidation with 1 to 3% solution of KMnO4 Distribution of TCE in treated subsurface zone Two decision units - undisturbed zone and backfilled zone

Surface view of the two decision units and borehole locations: 

Surface view of the two decision units and borehole locations SB 1 SB 10 SB 9 SB 8 SB 6 SB 5 SB 2 SB 4 SB 7 SB 3 SB 11 5 m Building Pavement and Sidewalks Undisturbed Area Backfilled Area N Soil Borings

Hollow stem coring: 1.2 m (4 ft) lifts: 

Hollow stem coring: 1.2 m (4 ft) lifts

Subsampling core with small coring tool: 

Subsampling core with small coring tool

Transferring 5-g sample into sample bottle containing 5-mL of methanol : 

Transferring 5-g sample into sample bottle containing 5-mL of methanol Discrete Sample

Transferring 5-g sample into sample bottle containing 200-mL of methanol: 

Transferring 5-g sample into sample bottle containing 200-mL of methanol Multi-increment Sample

Discrete and Multi-increment samples sent for analysis: 

Discrete and Multi-increment samples sent for analysis 1, 5 g plug of soil / 5 mL MeOH ≈ 40, 5 g plug of soil / 200 mL MeOH

1.2-m (4-ft) core after sampling: 

1.2-m (4-ft) core after sampling

3-D view of discrete sample distribution : 

3-D view of discrete sample distribution Building 0 m 12 m Undisturbed Zone Backfilled Zone Discrete Samples SB-1 SB-4 SB-8 SB-9 SB-11 SB-7 SB-3 SB-6 SB-5 SB-10 SB-2 Outlined volumes delineate subsurface zones of concern

Multi-increment sampling strategy and design in the two decision units: 

Multi-increment sampling strategy and design in the two decision units

Sampling and Analysis: 

Sampling and Analysis Eleven soil borings obtained by hollow stem coring Sampling and analysis Contractor collected 213 discrete 5-g samples for analysis by Method 5035A ERDC/CRREL collected 57 Multi-increment samples using both a vertical and horizontal sampling design. Analysis was by HPLC-UV (1828, 5-g increments)

Slide15: 

Results – Undisturbed Zone NS – not sampled, ND –not detected, NR – no recovery of soil core, * average of field duplicates, † average of field triplicates

Average TCE Concentrations: 

Average TCE Concentrations Backfilled Zone Discrete 14 mg/kg 4.4 mg/kg* Multi-increment - vertical 4.2 mg/kg - horizontal 5.6 mg/kg Undisturbed Zone Discrete 170 mg/kg 20 mg/kg* Multi-increment - vertical 180 mg/kg - horizontal 270 mg/kg * with Q test removal of highest discrete value

Slide17: 

Step 1 – Fill cells with the smallest horizontal or vertical MI average concentration where the column or row has a MI sample concentration less than 100 mg TCE/kg Delineate high concentration zones from MI samples

Slide18: 

Step 2 - Enter order of magnitude concentrations for TCE in empty cells to cause the calculated averages to approach, but not exceed, the MI sample concentrations.

Surface view of the two decision units and borehole locations: 

Surface view of the two decision units and borehole locations SB 1 SB 10 SB 9 SB 8 SB 6 SB 5 SB 2 SB 4 SB 7 SB 3 SB 11 5 m Building Pavement and Sidewalks Undisturbed Area Backfilled Area N Soil Borings

3-D view of discrete sample distribution : 

3-D view of discrete sample distribution Building 0 m 12 m Undisturbed Zone Backfilled Zone Discrete Samples SB-1 SB-4 SB-8 SB-9 SB-11 SB-7 SB-3 SB-6 SB-5 SB-10 SB-2 Outlined volumes delineate subsurface zones of concern

Slide21: 

KMnO4 distribution within the vadose zone

Summary: 

Summary Discrete and MI sampling strategies estimated similar average TCE concentrations in both decision units (provided all the data are retained) MI sampling strategy and two sampling designs (vertical and horizontal) estimated similar average TCE concentrations in both decision units

Summary: 

Summary MI sampling strategies allow for many more increments (sample mass) to be included in the analysis. Far fewer MI samples obtained the same average concentration and source zone distribution of TCE as compared to the discrete samples.

Summary: 

Summary Discrete sampling strategy detected one location where TCE is > 1% w/w MI sampling strategy and two sampling designs forecast three locations where TCE is > 1% w/w

Slide25: 

Questions ?

QA Samples: 

QA Samples Contractor Co-located duplicate samples in 10 of 11 boreholes RPD averaged 40% ERDC-CRREL Co-located EnCore duplicates adjacent to contractor co-located samples. RPDs averaged 53% among EnCore samplers (n=9) RPD of means of duplicates averaged 55% between contractor and ERDC values (n=9)

Sampling QA:  Multi-increment field Triplicates: 

Sampling QA: Multi-increment field Triplicates Vertical Backfilled - 5.2, 3.5, 5.2 Undisturbed - 13, 11, 11 / 0.92, 0.79, 0.84 / 910, 710, 530 Horizontal Backfilled - 21*, 3.2, 3.2 (* piece of Styrofoam found in sample) Undisturbed - 4.2, 3.9, 4.0 / 2400, 1200, 1700 RSDs averaged 17% (minus set with Styrofoam)

Additional Multi-increment QA: 

Additional Multi-increment QA Each multi-increment collection vessel (4, 8 and 16 oz) was spiked with TCE in the field. Recoveries of 98.6, 104, and 102% Each multi-increment collection bottle was repeatedly opened in field as if used to collect a sample, on three separate days. No loss of MeOH from narrow mouth 8- and 16-oz bottles 7% loss in MeOH / day from wide mouth 4-oz jar

Slide29: 

Step 1 – Fill cells with maximum concentration values from the highest MI sample Determine High Concentrations – Most Conservative

Slide30: 

Step 2 – Enter order of magnitude concentrations for TCE in empty cells to cause the calculated averages to approach, but not exceed, the MI sample concentrations. Delineate high concentration zones from MI samples