Assessing waterborne health risks through Quantitative Risk Assessment Models

1
Assessing waterborne health risks through
Quantitative Risk Assessment Models

• Benoit Barbeau
• École Polytechnique de Montréal
• Victoria, BC - June 25, 2008

2
Presentation Outline

• The place of QMRA in 2008
• Limitations of QMRA
• Source water
• Treatment modules
• Infectivity
• Relationship with disease in the population
• Project objectives
• Research team and collaborators
• Outcomes/Conclusions

3
Background

• Quantitative Microbial Risk Assessment increased
  worldwide recognition over the last decade
• USEPA have initiated its use in 1989 (SWTR) and
  has refined its application for the LT2SWTR
• Several provincial DW regulations have been
  derived from the USEPAs work
• Recently, Health Canada has been promoting a QMRA
  approach and is even developing a web-based tool
  for small systems
• In Europe, the Netherlands have incorporated the
  need to perform a QMRA as part of Safety Water
  Plans and have adopted the 10E-4 risk level
  originally proposed by the USEPA

4
The basic equations behind QMRA
5
Limitations of QMRAMicrobial Occurrence Data
(CRW)
Montreal source water
1 000

• Analytical issues
• Low occurrence leads
  to numerous
  non detects
• Total vs infectious counts
• Genetic profiling of Cryptosporidium
• C. parvum and C. hominis main risks
• Some watersheds are dominated by C. andersoni
  (cattle) and C. baileyi (birds)
• 10-14 of fresh oocysts are infectious in mice

31 below detection limit
100
10
Giardia/100L
1
0.1
0.01
0,0
0,2
0,4
0,6
0,0
1,0
Centiles
6
Limitations of QMRATreatment modules

• Physical removal
• Large variability in the expected
  removal through
  conventional
  treatment
• Removal might be a function
  
  of the pathogen load
• How do we model membrane
  treatment ?
• Disinfection
• The use of the USEPA CT10 approach is misleading
• A new player UV disinfection

7
Limitations of QMRATreatment modules (cntd)

• Accounting for treatment variability
• Process variability due to OM

Ozone
Cryptosporidium inactivation
Water temperature (oC)
8
Limitations of QMRATreatment modules (cntd)

• Accounting for treatment variability
• Impact of downtime

1E00
Cryptosporidium
1E-01
Giardia
Mean annual risk of infection
Risque annuel moyen (inf./ an)
Annual risk of infection
1E-02
1E-03
1E-04
Ozone
UV




9
Limitations of QMRARelationship with disease in
the population ?
Cryptosporidium_WTP B
1.E00
3/10 000
1.E-01
1.E-02
1.E-03
1.E-04
Mean annual risk of infection
1.E-05
1.E-06
1.E-07
1.E-08
Eau brute
FDSCO3Cl2
FDCOO3Cl2
TCO3UVCl2
FDCSOO3Cl2
FDSCO3UVCl2
FDCOO3UVCl2
FDCSOO3UVCl2
10
Project activities

• Select a QMRA model
• Comparing the different approaches currently
  developed in the US, Canada and Europe
• Validate key elements of the QMRA model
• Monitoring of several microbially- challenged
  full-scale drinking water treatment plants
• Focus on Giardia /Crypto using large volume
  sampling
• Measure indicators of treatment performance
• Linkage with GI data
• Where possible, health data from sentinel sites
  (C-Enternet) could be used

11
Research team
12
Research team (cntd)
13
Collaborators
14
Outcomes

• Improve of our understanding of some key
  variables used in QMRA.
• Establish more appropriate goals to drive
  management technology/decisions ()
• Link Canadian researchers and utilities with
  international experts in the domain of QMRA
• Provide stakeholders with a better understanding
  of QMRA