Investigation and Remediation of Groundwater Contamination at a Pesticide Facility

1
Investigation and Remediation of Groundwater
Contamination at a Pesticide Facility

• Carter et al., 1995. Water Quality Research
  Journal of Canada, pp. 469-491.

2
Site History

• Since 1972 Ciba-Geigy Canada has operated an
  agrichemical formulating and warehousing facility
  in Cambridge, Ont.
• In 1991 soil impacted with Dinoseb was discovered.

3
Preliminary Investigation

• Wells installed just below the water table
  confirmed that dinoseb was present in groundwater
  at about 400 mg/L (40 times the ODWO).
• Other organic compounds were also found.

4
Environmentally Sensitive Site

• About 30 of Cambridge drinking water is pumped
  from 8 wells located between 600 and 2200 m from
  the site in the underlying bedrock.

5
Location of Site and Municipal Supply Wells
Hiway 401
Sheldon Dr.
Hiway 24
Franklin Boulevard
6
Further Investigation

• Two pesticides (metolachlor and dinoseb) have
  been measured above ODWO in deep groundwater in
  the bedrock about 30 m below the site.

7
Pesticide Characteristics

• Significant impacts to groundwater are not
  expected under typical agricultural use.
• Recent well surveys indicate that deep migration
  of pesticides is uncommon.

8
Ciba initiated a four-phase soil and groundwater
remediation program.

• Phase 1 Background information review.
• Phase 2 Remedial investigation.
• Phase 3 Remedial alternative analysis.
• Phase 4 Remedial action plan implementation.

9
All 4 Phases have been initiated.

• Localized excavation of impacted soils and
  implementation of interim measures to control and
  treat contaminated groundwater.
• Site remediation for 10-20 years.

10
Investigation Methodology

• Process and Waste Management Audit
• Private Well Survey
• Soil Vapour Survey
• Drilling Program

11
Methodology...

• Geophysics
• Groundwater Flow Modeling
• Regional Well Testing
• Monitoring Program

12
Methodology...

• Treatability Studies/Interim Groundwater Control
• 0 Regulatory/Public Liaison

13
Process and Waste Management Audit

• Current and historical site operational
  practices.
• Interviews with current and former employees.
• Records of all chemical compounds handled at the
  site.

14
Private Well Survey

• Review water well records on file with MOE.
• Door-to-door survey within a 1 km. radius.
• Only two private wells discovered, neither were
  registered with MOE.

15
Soil Vapour Survey

• Volatile organic solvents were used at the plant.
• Shallow soil vapour survey conducted and
  information used to identify areas for drilling
  and soil sampling.

16
Drilling Program

• 87 monitoring wells installed as of Aug. 1994.
• Using mainly 10.8 cm hollow stem augers with
  continuous coring.
• Wells constructed of 5 cm PVC pipe and slotted
  well screens set in silica sand filter pack.

17
Drilling Program...

• Shallow wells near water table are constructed
  with 3m screens.
• Deeper wells have 1.5 m screens sealed with
  bentonite grout.
• Measure K in situ.

18
Groundwater Flow Modeling

• Used a 2-D model (FLOWPATH) to estimate
  groundwater flow velocities, regional groundwater
  flow conditions in the bedrock, and the potential
  effectiveness of a groundwater pump and treat
  system.

19
Regional Well Testing Program

• How to use nearby municipal wells?
• Shut-down municipal wells for 48 hours and
  measure water level response in on-site wells.
• Measurable drawdown of 0.1 to 1 m at a distance
  of 1 km.

20
Monitoring Program

• Wells sampled up to 4 times per year until 1994.
• Use Waterra foot valve samplers.
• Sample protocol according to MOE standards.

21
Treatability Studies

• Interim groundwater pump and treat facility with
  two granular activated carbon units in series.
• Alternatives are ultraviolet-peroxide destruction
  and in situ /ex situ bioremediation .

22
Regulatory/Public Liaison

• Regular meetings with municipal, regional, and
  provincial regulators and the general public.
• Major technical reports in the local library.

23
Results Hydrogeologic Setting
Unit Description Thickness (m) K (cm/s)
1 Silty sand/sandy silt 10 10-2 - 10-5 2
Silty clay/clayey silt 3 10-5 - 10-7
3 Sand 12 10-3
- 10-4 4 Silt till 4
10-4 5 Sand, gravel 1
not tested 6 Dolomite 100
10-2 - 10-5
24
Summary
Surficial Sand Aquifer
Clay and Silt Aquitard
Lower Sand/Till Aquifer
Bedrock Aquifer
Unit 3 is confined, semi-confined, and
unconfined on site.
25
Pesticide Distribution on Site

• Metolachlor has the greatest exceedance of ODWO
  (50 mg/L).
• Up to 10,000 mg/L in surficial sand aquifer.

26
Pesticide Distribution...

• Maximum concentration in bedrock is 14 mg/L.
• Mixing/dispersion is suspected as the principal
  process of reduction.
• Some metabolites are present.
• Ciba-derived compounds have not been detected in
  the municipal wells.

27
On-Site Groundwater Flow Patterns

• Shallow groundwater flow from east to west.
• Velocities from FLOWPATH 0.4 to 75 m/yr.
• Maximum migration distance of 250m.

28
Water Table Configuration
306
305
304
303
302
29
Groundwater Flow Directions...

• Lateral flow direction in lower aquifer and
  bedrock is south and southwesterly.
• Dilute concentrations of pesticides in bedrock
  south and southwest of the site.

30
Equipotential Lines Bedrock
299
298
297
296
31
Results from Shut-Down Tests

• Water levels in bedrock and lower aquifer are
  influenced by municipal pumping.
• Vertical gradients in lower aquifer increased up
  to a factor of about 2.

32
Municipal Wells that Influence Water Levels in
Bedrock Wells on Site
33
Interim Groundwater Control/Treatment

• Install pumping wells in upper and lower aquifers
  to control lateral migration off site.
• Capture zone analysis with FLOWPATH indicates
  that 2 wells at 49 and 65 m3/day are enough.

34
On-Site Water Treatment

• Two 454 Kg. granular activated carbon units in
  series.
• Reduce pesticide concentrations to below
  detection limits of 1 mg/L.

35
Conclusions

• Unplanned release of pesticides and other organic
  compounds.
• Pesticide levels in shallow aquifers exceed ODWO.

36
Conclusions...

• Lateral and vertical transport of pesticides a
  function of source area, physical/chemical
  properties, and groundwater flow conditions.

37
Conclusions

• Six pesticides have migrated 30 m downward
  through five layers and laterally approximately
  200 m.
• Concentrations of five pesticides in bedrock
  aquifer are below ODWO.

38
Conclusions...

• Downward movement of pesticides largely
  influenced by nearby municipal wells.
• Variations in vertical gradients caused by
  irregularities in hydraulic connection between
  overburden and bedrock fracture system.

39
Conclusions...

• Groundwater pump and treat system to control
  further migration of pesticides.
• Activated carbon units efficiently remove
  contamination.
• System operates at 114 m3/day.