Human viruses in deep groundwater in Wisconsin:

Implications for public health, water treatment, and groundwater protection

Kenneth R. BradburyWisconsin Geological and Natural History Survey

University of Wisconsin-Extensiony

Wisconsin Soil and Water Conservation SocietyAnnual Conference

2/9/12

Waterborne Disease Outbreaks in the USA

• From 1971 to 2002 there were 716 outbreaks associated with an infectious agent in drinking water

• 60% of the outbreaks were attributable to groundwater

• Pathogen in more than 50% of outbreaks is unknown and assumed viral

Summarized from CDC reports, e.g., MMWR, 2004, 53(SS08);23-45

Examples of Groundwater-borne Disease Outbreaks

• South Bass Island, OH 2004• Campylobacter jejuni,

Salmonella typhimurium, norovirus, and Giardia in private and municipal wells

• 1,450 persons ill

• Missouri 1999• Salmonella typhimurium in

municipal well• 124 persons ill

• Walkerton, Ont. 2000• E. coli O157:H7 in municipal

well• Of 4,800 residents, 2,300

became ill• 7 died

• Wyoming 2001• Norovirus in non-

community, transient well• At least 84 patrons ill

• Door County WI 2007• Norovirus in restaurant well• At least 229

patrons/employees ill• 6 hospitalized

Groundwater supplied from deep, confined bedrock aquifers is usually assumed to be protected from contamination and free of pathogenic organisms, but……the presence of viruses in deep wells in Madison, Wisconsin shows that pathogenic surface contaminants can reach deep aquifers over relatively short time spans.

Take-home messages…

furthermore…

• Viruses were found in every well tested• Leaky sewers are a likely virus source• Virus transport to wells appears to be much

more rapid than standard groundwater models suggest

• Rapid flow through discrete geologic features (fractures) might explain virus occurrence

Pathogens in Human WastewaterBacteria (e.g., Campylobacter,Salmonella, E. coli O157:H7, Aeromonas)

Viruses (e.g., adenovirus, enterovirus, rotavirus)

Protozoa, (e.g., Cryptosporidium,Giardia, Toxoplasma)

1,438 pathogens infect humans of which 32 are waterborne in the U.S.

Particulate transport

• Particulates and common diameters• Colloids (<0.2 um)• Bacteria (~.2 – 20 um)• Viruses (~50 nm) (1000 nm= 1um; 1nm is one billionth of

a meter)

• For comparison• Human hair (~50-100 um)• Rock fracture (1 um – 1 mm - ?)

Viruses are often much smaller than fracture apertures or rock pores.

Public supply wells in

Wisconsin

Currently, 66 municipal water systems in Wisconsin do not

disinfect water

Background• Initial virus sampling conducted as part of

AWWARF (now Water Research Foundation)- funded study of aquitards

• We wanted to find out whether viruses were present in a deep, confined bedrock aquifer

• We did not expect to find any because the travel times were supposedly much longer than the virus lifetimes

Acknowledgements: co-authors, field workers, and cooperators

• Madeline Gotkowitz, Dave Hart, Pete Chase, Andrew Aslesen, Jake Krause, Wisconsin Geological and Natural History Survey

• Mark A. Borchardt, Susan K. Spencer, Marshfield Clinic Research Foundation

• Chris Gellasch, Jean Bahr, University of Wisconsin-Madison

• Randy Hunt, U.S. Geological Survey

• John Cherry, Beth Parker, University of Waterloo, Ontario

• Madison Water Utility

• Madison Municipal Sewerage District

• Wisconsin DNR – Groundwater Joint Solicitation Program

Study Objective Evaluate the occurrence of human viruses in the Mount Simon aquifer, a confined, deep sandstone aquifer in south-central Wisconsin.

Wisconsin Geology

bedrock glacial

Wisconsin’s geologic history created important bedrock and glacial aquifers. Aquitards control groundwater flow and quality in both

settings.

Dane County geology...

The next slide shows a cross

section along this line...

glacial geology

Granite

Mt Simon &Eau Claire sandstones

Eau Claireshale

Sandstone& dolomite

BlueMounds

Mt Horeb Verona MadisonGlacial deposits

sandstone aquifer

water table

citywell

Hydrogeologic Cross Section

Groundwater moves downward and laterally through Dane County’s aquifers...

Domestic wells

the lakes

16

aquifer

aquifer

aquitard

Units in the “sandstone”

aquifer

Core of Cambrian sandstone from the Madison area. Sand

is clean, well-sorted, well rounded, predominately

quartz.

Microscope image of sandstone from a Madison supply well. Clean, well-rounded quartz grains are

generally non-reactive.

18

Core sample of the Eau Claire shale (aquitard)

1919

Eau Claire aquitard

Wells sampled

2007-2009We began with 10 wells in a variety of locations to be sure we had virus-positive wells to work with.

We later reduced the sampling to 6 wells due to budget and logistic considerations.

Virus Sampling

• Virus samples collected monthly. All samples taken prior to chlorination

• qRT-PCR/qPCR detection of enterovirus,rotavirus, hepatitis A virus, norovirus, and adenovirus

•Enterovirus and adenovirus serotype determined by nucleic acid sequencing; infectivity determined by cell culture

• Isotopes (3H, 2H, 18O) analyzed at the University of Waterloo Environmental Isotope Lab

Study Methods

Construction of Madison supply wells

deep aquifer

aquitard

shallow aquifer

glacial materials

casi

ngop

enho

lewell

crystalline rock

50

300

320

800

rela

tive

dept

hs,f

eetb

elow

surfa

ce

not to scale

water table

potentiometricsurface

Typical (newer wells)

deep aquifer

aquitard

shallow aquifer

glacial materials

casi

ngop

enho

le

well

crystalline rock

50

220230

730

rela

tive

dept

hs,f

eetb

elow

surfa

ce

not to scale

water table

potentiometricsurface

147

Some older wells (1928)Note shallow casing

rela

tive

dept

hs,f

eetb

elow

surfa

ce

casi

ngop

enho

le

well 7

30

215225

725

237.6

not to scale

casi

ngop

enho

le

well 30

23

266277

800

312

casi

ngop

enho

le

well 19

15

255

718

260

245

casi

ngop

enho

le

well 13

60

780

260

370

casi

ngop

enho

le

well 12

35

404418

986

260

rela

tive

dept

hs,f

eetb

elow

surfa

ce

well 11

18

752

casi

ngop

enho

le

111

deep aquifer

aquitard

shallow aquifer

glacial materials

casi

ngop

enho

le

well

crystalline rock

50

300

320

800

rela

tive

dept

hs,f

eetb

elow

surfa

ce

not to scale

water table

potentiometricsurface

Construction details of the municipal wells sampled throughout the project. Diagram at upper left shows typical hydrostratigraphy and well construction for the Madison area.

Wells sampled

2007-2009We began with 10 wells in a variety of locations to be sure we had virus-positive wells to work with.

We later reduced the sampling to 6 wells due to budget and logistic considerations.

Overall findingsWe found viruses in every well sampled in 2007-2009, though not every time sampled

Summary of virus detections by water source, wells 7, 11, 12, 13, 19, 30 only.

virus detection (gc/l)

water source number of samples percent positive min max mean

wells148 46.6 0.00 6.27 0.65

Lake Mendota17 82 0.00 532 44

sewage influent26 100.0 12,900 36,310,000 2,010,000

Most frequently detected viruses, by rank. Number of detections in parentheses. Multiple

viruses indicate a rank tie.

Rank wells sewage lakes

1 A41 (38) A41 (18) A41 (12)

2 A31 (12) A31 (15) A2, E3 (3)

3 A2 (11) A2, E3 (5) A31, E11, E30 (2)

4 E3 (5) E9 (4) --

5 E30, A7 (4) A6, E11 (3) --

6 Cox A16, Cox B3, Cox B4, E11 (3) Cox A16, Cox B4 (2) --

7 A5, E6, E9, E71 (1) A5, A7, E6, E30, E71, E? (1) --

2007-2009 was an unusually wet period; large storms in August 2007 and June 2008

Coincidence with time The coincidence of virus detections in wells, lakes, and

sewage is surprising.sewage is surprising.g g

Well Virus Concentrations - Coincidence with Time

percentages indicate positive detections

note scale changes

Coincidence of peaks in wells miles apart

suggests a regional problem

16 different species (serotypes) of viruses were identified in wells, sewage, and lake water during this study, and in many cases wells and sewage contained identical virus serotypes. Detected viruses include Enteroviruses echovirus 3, echovirus 11, Coxsackie A16, B3, and B4, Adenoviruses 2, 7, 31, and 41.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

14-S

ep-0

7

24-O

ct-0

7

26-N

ov-0

7

19-D

ec-0

7

24-J

an-0

8

26-F

eb-0

8

24-M

ar-0

8

28-A

pr-0

8

27-M

ay-0

8

7-Ju

l-08

28-J

ul-0

8

25-A

ug-0

8

17-S

ep-0

8

29-S

ep-0

8

13-O

ct-0

8

27-O

ct-0

8

10-N

ov-0

8

2-D

ec-0

8

16-D

ec-0

8

7-Ja

n-09

23-J

an-0

9

3-Fe

b-09

16-F

eb-0

9

11-M

ar-0

9

31-M

ar-0

9

27-A

pr-0

9

A41 x x x x x x NS x x x x x x x x x x x x xA31 x x x x NS x x x x x x x x x x x xA2 x x x x x NS xE3 x x NS x x xE30 NS x xA7 x NSE11 x x x x NSCoxA16 x NS xCoxB3 NSCoxB4 x NS x

A41 x x x x x x x x x x x x x x x x x xA31 x x x x x x xA2 x x x x x x xE3 x x xE30 x xA7 x x x xE11 x x xCoxA16 x xCoxB3 x xCoxB4 x

A41 x NS NS NS NS NS NS NS x NS x x x x x x x xA31 NS NS NS NS NS NS NS NS x xA2 x NS NS NS NS NS NS x NS x NSE3 NS NS NS NS NS NS NS x NS x xE30 NS NS NS NS NS NS NS x NSA7 NS NS NS NS NS NS NS NSE11 NS NS NS NS NS NS NS NS x xCoxA16 NS NS NS NS NS NS NS NSCoxB3 NS NS NS NS NS NS NS NSCoxB4 NS NS NS NS NS NS NS NS

Sample round

all

wel

lsla

kes

sew

age

med

ian

sam

ple

date

There is coincidence between serotypes present in sewage and serotypes present in groundwater.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

14-S

ep-0

7

24-O

ct-0

7

26-N

ov-0

7

19-D

ec-0

7

24-J

an-0

8

26-F

eb-0

8

24-M

ar-0

8

28-A

pr-0

8

27-M

ay-0

8

7-Ju

l-08

28-J

ul-0

8

25-A

ug-0

8

17-S

ep-0

8

29-S

ep-0

8

13-O

ct-0

8

27-O

ct-0

8

10-N

ov-0

8

2-D

ec-0

8

16-D

ec-0

8

7-Ja

n-09

23-J

an-0

9

3-Fe

b-09

16-F

eb-0

9

11-M

ar-0

9

31-M

ar-0

9

27-A

pr-0

9

A41 x x x x x x NS x x x x x x x x x x x x xA31 x x x x NS x x x x x x x x x x x xA2 x x x x x NS xE3 x x NS x x xE30 NS x xA7 x NSE11 x x x x NSCoxA16 x NS xCoxB3 NSCoxB4 x NS x

A41 x x x x x x x x x x x x x x x x x xA31 x x x x x x xA2 x x x x x x xE3 x x xE30 x xA7 x x x xE11 x x xCoxA16 x xCoxB3 x xCoxB4 x

A41 x NS NS NS NS NS NS NS x NS x x x x x x x xA31 NS NS NS NS NS NS NS NS x xA2 x NS NS NS NS NS NS x NS x NSE3 NS NS NS NS NS NS NS x NS x xE30 NS NS NS NS NS NS NS x NSA7 NS NS NS NS NS NS NS NSE11 NS NS NS NS NS NS NS NS x xCoxA16 NS NS NS NS NS NS NS NSCoxB3 NS NS NS NS NS NS NS NSCoxB4 NS NS NS NS NS NS NS NS

Sample round

all

wel

lsla

kes

sew

age

med

ian

sam

ple

date

Sewers are ubiquitous in urban areas.

This graphic shows only regional connector sewers. There are hundreds of miles of local sewers.

The older sewers are nearly 100 years old.

Source: http://www.surrey.ca

Cross sections showing the possible locations of a sewer relative to the water table. A: gravity-drain sewer below water table; B: gravity-drain sewer above water table; C: force main above water table; D: force main below water table. H1 and H2 represent hydraulic head inside and outside the sewer. Arrows show directions of potential sewer leakage.

Sewers are a potential virus source. Some Madison sewers are in hydraulic positions that allow them to leak to groundwater.

Sewer Locations Relative to the Water Table

Is t1 + t2 + t3 < Virus survival time (~ 2 years) ?

How Did the Viruses Reach the Deep Aquifer? (assume continuous aquitard)

Kbulk-vert ~ 2 x 10-4 cm/s

� �

Kbulk-vert ~ 2 x 10-4 cm/s

�

� �

Kbulk-horiz ~ 3.5 x 10-3 cm/s

�

� �

Is t1 + t2 + t3 < Virus survival time (~ 2 years) ?

If we pick the right parameters, transport to the aquifer could occur with the virus lifetime

But, other transport pathways are also possible

Possible Transport Pathways Through Aquitard

1) fractures

2) depositional or erosional stratigraphic windows

3) down cross-connecting open wells or boreholes

4) along damaged, deteriorated, poorly sealed well annulus or breaches in well casings

We still have not determined the pathway – research continues!

Not to scale

UnitWell 7

Mt. Simon aquifer (confined)

Eau Claire aquitard

sanitary sewer (nonpoint source)

monitoring wells

water table

open borehole

well casing

6m

Wonewoc Formation

Tunnel City Group

glacial sediments

CURRENT WORK

Geophysical logs of monitoring well adjacent to Madison well #7

A major fracture exists at 80 feet.

Large horizontal fracture at 80 ft

Dipping fracture at 165 ft

Installing a FLUTE liner at well 7Installing a FFLUTE liner at wwell 7

So what, do viruses matter?Consider-

In a recent (2006-2007) evaluation of 14 Wisconsin communities using untreated water, viruses were found in wells from every system.

And…

Using a crossover-control study design, Dr. Mark Borchardt and colleagues (Marshfield Clinic) estimate that about 15% of the annual gastro-intestinal illness in these communities can be attributed to consuming untreated water.

Should communities be required to disinfect public water??

• Following our studies in 2009, the DNR decided to require all communities to disinfect public water supplies.

• Last year (2011) the Legislature voted to make it illegal to require disinfection.• Reasons given included cost, taste, and a desire for local

decision-making• Cost of illness was not considered

Implications/Recommendations

Human viruses can be commonly present in groundwater in deep bedrock wells.

Communities that use groundwater for a drinking water source should ensure that chlorination or other disinfection levels are sufficient to deactivate viruses.

Sampling for viruses requires a time series approach because virus concentrations, and virus species, vary with time in individual wells.

Untreated sewage contains very high concentrations of viruses and should be considered a source of groundwater contamination.

Human enteric viruses potentially represent a powerful new tracing tool for hydrogeologic studies. Fundamental and applied research on the use and effectiveness of viruses as tracers should be undertaken.

Recent Publications

Borchardt, M. A.; Bradbury, K.R.; Gotkowitz, M. B.; Cherry, J. A.; Parker, B. L.. 2007. Human enteric viruses in groundwater from a confined bedrock aquifer. Environmental Science and Technology. 41(18); 6606-6612

Bradbury, K.R., M.A. Borchardt, M. Gotkowitz, and R.J. Hunt. 2008. Assessment of Virus Presence and Potential Virus Pathways In Deep Municipal Wells. WGNHS Open-File report 2008-08. 48 p.

Bradbury, K.R., M.A. Borchardt, M. Gotkowitz, and S. Spencer. 2010. Human viruses as tracers of wastewater pathways into deep municipal wells. Final report to the Wisconsin Department of Natural Resources. 47 p.

Thank You

Questions?