Arsenic in Minnesota Groundwater

1
Arsenic in Minnesota Groundwater

• Mindy Erickson, P.E.
• Water Resources Science
• University of Minnesota
• Funding provided by CURA, MDH, WRC and USGS

2
Why Arsenic? Why Now?

• In 2001, MCL changed from 50 ug/L to 10 ug/L
• Public Water Suppliers (PWS) must comply by
  January 1, 2006
• About 40 MN PWS affected
• Treatment plants for arsenic are expensive

3
Why the Change?Arsenic and Health

• Exposure
• Living with a smoker
• Arsenic at 50 ug/L
• Contaminant (EPAs standard threshold)

• Deaths / 1,000
• 10
• 20
• 0.1

Arsenic is classified by EPA as a known human
carcinogen. It can cause skin, lung, prostate,
and bladder cancers skin lesions nerve damage
and other health effects.
4
Arsenic Occurrence

• Arsenic naturally present in rock and sediment at
  1 to 100s mg/Kg
• Crustal average is 1.8 mg/Kg
• At 1.8 mg/Kg, solubilization of lt0.1 yields 10
  ug/L arsenic in water
• Certain geochemical conditions leach arsenic into
  groundwater
• Arsenic in mid-western groundwater is widespread
  and naturally occurring

5
Arsenic Chemistry

• Arsenate (As5 ? H2AsO4-, HAsO4-2)
• Oxidized form
• Adsorbs to metal oxides
• Arsenite (As3 ? H3AsO3)
• Reduced form more toxic inorganic form
• Adsorbs to iron oxides
• Organic Arsenic (many forms)
• In foods highest in seafood
• Uncommon in groundwater

6
Arsenic Release Mechanisms

• Reductive Desorption
• Reductive Dissolution
• Anion Competition
• Mineral Oxidation (often pyrite)

7
Project Goals

• Characterize arsenic variability
• Spatial
• Temporal
• Investigate geochemical connections and
  mechanisms
• Provide information for developing regulations
  and guidelines

8
Research Tasks

• Database building
• PWS Sampling
• Sediment Sampling
• Private Well Sampling
• Statistical Analysis
• Geochemical/hydrological modeling

9
Database Building

• Quaternary geology information
• Water quality data
• Sediment geochemical data
• Well construction information

10
Minnesota Arsenic Data

• National Uranium Resource Evaluation
• Public Water Supplies
• Centers for Disease Control
• MN Pollution Control Agency GWMAPS
• Minnesota Arsenic Study (MARS)
• Minnesota Geological Survey
• Other

11
Minnesota Arsenic and Geology
12
Recent Glacial History
13
Regional Arsenic and Geology
14
Arsenic and the Des Moines Lobe Till

• 10 of PWS within the Des Moines lobe footprint
  exceed 10 ug/L
• 2 of PWS outside of Des Moines lobe footprint
  exceed 10 ug/L
• 3,200 measurements

15
Arsenic in Sediment

• Sediment arsenic concentrations vary from lt1 to
  50 mg/Kg in the region
• Average sediment arsenic concentrations are
  similar across the region
• No direct link between sediment arsenic
  concentration and water arsenic concentration
• Des Moines lobe till is not the source of arsenic
  in groundwater

16
As and Well Characteristics
Till
Sand
17
As and Well Characteristics
Description Avg As As gt 10 Count
Screen 8 feet Clay 4 feet 20 60 224
All others 17 49 606
Screen gt 8 feet Clay gt 4 feet 12 40 71
18
Research Tasks

• Database building
• PWS Sampling
• Sediment Sampling
• Private Well Sampling
• Statistical Analysis
• Geochemical/hydrological modeling

19
PWS Sampling

• Arsenic Variability
• Local Private well sampling
• Temporal sampling of city well(s)
• Filtered/Unfiltered pairs (selected sites)
• Water analyzed for metals, arsenic species, major
  anions, TOC, ammonium

20
Participating Communities
21
Nielsville SI Wells
22
Nielsville SI Wells
23
Long-Term Temporal Results
24
Long-Term Temporal Results
25
Short-Term Temporal Results
26
PWS Results

• No difference between raw water filtered and
  unfiltered arsenic concentrations
• Significant difference between chlorinated
  filtered/unfiltered arsenic concentrations
• No long-term trends
• Some significant short-term variability
• Low arsenic private wells identified near some
  high arsenic PWS wells

27
PWS Results Implications

• Most water arsenic is dissolved
• Four quarterly samples from an old well is likely
  representative
• Short-term temporal sampling is worthwhile for
  wells with As close to the MCL
• Site investigation is worthwhile
• Low cost
• Quick implementation
• Significant potential cost savings

28
Research Tasks

• Database building
• PWS Sampling
• Sediment Sampling
• Private Well Sampling
• Statistical Analysis
• Geochemical/hydrological modeling

29
Sediment Sampling

• Sediment and water samples from 8 new private
  wells, mud-rotary drilling
• Sediment and water samples from 2 new monitoring
  wells, rotasonic drilling
• Sediment analyzed for metals, organic carbon,
  carbonate
• Selected sediment samples sequentially extracted
• Water analyzed for metals, major anions, arsenic
  species, ammonium, and TOC

30
Mud-Rotary Water Results
31
Mud-Rotary Results

• No direct correlation between sediment arsenic
  concentration and water arsenic concentration
• Arsenic concentration in till averaged 6 mg/Kg,
  Fe 2
• Arsenic in aquifer sediment averaged 1.5 mg/Kg,
  Fe 0.8
• High arsenic concentration increases over time in
  new wells

32
Monitoring Wells
33
Monitoring Wells
34
Monitoring Wells
35
Preliminary Rotasonic Results

• Organic layer observed between the till and sand
• 0.5 0.7 mg/Kg arsenic is adsorbed to the
  sediment grains
• Broad sediment and water geochemistry yet to come

36
Research Tasks

• Database building
• PWS Sampling
• Sediment Sampling
• Private Well Sampling
• Statistical Analysis
• Geochemical/hydrological modeling

37
Private Well Sampling

• 60 private wells selected based on well
  characteristics and arsenic concentration
• Water analyzed for metals, major anions, arsenic
  species, ammonium, and TOC
• Filtered/unfiltered pairs at 20 wells

38
Private Well Sampling Results
39
Private Well As Speciation Results
40
Preliminary Private Well Sampling Results

• Arsenic concentration is not directly correlated
  to TOC or ammonium concentration
• Most arsenic is As3
• Most arsenic is dissolved, not particulate
• Well water with a rotten egg smell has
  relatively low arsenic about 10 ug/L

41
Research Tasks

• Database building
• PWS Sampling
• Sediment Sampling
• Private Well Sampling
• Statistical Analysis
• Geochemical/hydrological modeling

42
Outcomes

• Quantify arsenic phases
• Support or refute preliminary observation of
  increasing arsenic concentration in new wells
• Narrow potential geochemical mechanisms and
  geological/hydrogeological controls
• Characterize arsenic variability
• Provide MDH with info for developing guidelines
  and regulations