Case study 3: Irrigation of food crops with treated wastewater
Problem Formulation (Background, goal, and breadth)
The wastewater treatment plant Harnaschpolder near Delft is one of the largest in Europe. It
collects and treats the wastewater of more than 1 million people, treats with conventional
biological treatment and discharges 10.000 m³ secondary effluent per hour (dry weather flow).
The wastewater effluent could be an alternative source of fresh water for the greenhouse sector
– after further treatment. The greenhouse sector does not want to compromise the safety of
greenhouse produce for consumers and also for the greenhouse workers. The owners of the
wastewater treatment are conducting a research project to look at treatment options.
Currently they suggest 2 different treatment lines, 1: use sand filtration and multimedia
filtration (for N/P removal) followed by ultrafiltration and reverse osmosis, but this is expensive
(different lines shown in Figure 1).
Objective:
 Assess the risk/probability that lettuce consumers and workers in the greenhouses will get
an infection after consuming the lettuce and after exposure to wastewater during irrigation.
 Propose the control measures to reduce the risk to consumers and workers and evaluate
the effectiveness of these measures.
Stakeholders:
1. Wastewater treatment plant operators and administrators
2. Farmers/ workers of green houses
3. Consumers
4. Technology vendors
5. Health authorities
6. National/ international regulatory authorities
7. Residents of the region
Figure 1: Treatment lines.
LINE 1: Reference treatment (conventional + membranes): no chemical disinfection
WWTP
effluent
VIRUS
2-4 Log
VIRUS
1-2 Log
Microbiological barriers
SAND
FILTRATION
MULTIMEDIA
FILTER
100um-10um
100um-1um
NO3-
PO4- & Susp. solids
UF
RO
100-10nm
<1nm
Susp. Solids
Salts & Dissolved
Compounds
LINE 2: Innovative line (membranes): no chemical disinfection
VIRUS
0.2-1 Log
Microb. barriers
BiopROtector
NO3- & Nutrients
3FM /
DiscFilter
Suspended
solids
VIRUS
2-4 Log
AiRO
<1nm
Salts & Dissolved
Compounds
Only 1
effective
barrier
against
virus at
the end
TREATMENT
Sand Filtration
Microfiltration (MF)
Ultrafiltration (UF)
Nanofiltration (NF)
Reversal Osmosis (RO)
Reversal Osmosis (RO)
UV irradation
LOG REMOVAL
VIRUSES
0.1-2
0.2-2
1-4
3-7
1.4-3.6
3-7
2-3
REFERENCE
Shirasaki et al. Water Research (2010)
B. Zhu et al. Water Research (2005)
Madaeini S.S. Water Research (1999)
O'Grady. Dev. Biol. Stand. (1996)
EPA/NSF (2006)
Madaeini S.S. Water Research (1999)
Hijnen et al. Water Research (2006)
Available data:
No data is available concerning the treatment efficiencies and the microbial concentrations of
the secondary sewage, and therefore should be obtained from literature or by estimation.
Chemical parameters in the secondary treated water are available but this risk assessment
focuses on the microbial contamination of the water to assess the public health risk.
Necessary data:
Risk assessment paradigm:
1. Hazard Identification
To perform a microbial risk assessment several pathogens could be used. Because
enteric viruses are often found in sewage, even in high concentrations, and are difficult
to remove with conventional treatment processes viruses could be considered as a
conservative choice. Rotaviruses are causing gastrointestinal disease and adenoviruses
are causing both gastroenteritis and acute respiratory illnesses.
2. Hazard characterization
a. Dose response analysis
For both rotavirus and adenovirus a dose response is available
b. Health effects/ health outcomes/ type of disease
3. Exposure Assessment
a. Exposure routes
i. Direct consumption of lettuce by consumers
ii. Inhalation of water by workers
iii. Accidental ingestion of water by worker
iv. Direct contact between worker and water: ingestion of water by uptake
the aerosols and surface to hand to mouth contact
b. Exposure frequency and duration
c. Exposure volume (ml)
d. Breath volume and frequency
e. Number of people exposed / annual infection risk
Exposure pathways:
Table 1: Consumer Exposure Pathway 1
Parameter name
[rotavirus] in sewage
effluent
Value
0.873
Unit
pfu/ml
Assumption
 The data was derived from different
treatment processes, and a average was
calculated
Reference
 Oragui et al. (1989)
 Smith and Gerba (1982)
 Bates et al. (1984)
Recovery of the method
0.5
%
Removal efficiency of
the treatment
Total concentration of
viruses in 7 days
Volume of water in
leafs after irrigation
Number of irrigations
per day
4 or 6
ulog1
0.0004738
85
10.8
pfu/ml
2
n
Reduction/die-off
0.69
ulog/day
Time between harvest
and consumption
2
days
Washing procedure at
home
Diary lettuce
consumption
2
ulog2
22.5
g/person/
day
ml
 Log normal distribution
Assumption = 50 % recovery with the applied
detection method
4 or 6 log reduction obtained with the
applied treatment
Accumulation of the viruses on the lettuce
after 7 days
Described in literature
Normal distribution
Assumption = normally there will be 2
irrigations per day, under dry conditions it
will be 3 times and under more humid
conditions it will be 1 time
Custom distribution
Described in literature
Normal distribution
Assumption = 2 days after harvesting the
lettuce is comsumed
Triangular distribution
Described in literature
Normal distribution
Described in literature
Normal distribution
 Rao et al. (1988)
??
Hamilton et al. 2006
Hamilton et al. 2006
P.Gale et al 2005
GEMS/Food Regional
Diets, 1998
Table 2: Rotavirus single and annual risks for consumers through ingesting lettuce
Risk
Annual Risk
mean (SE) median (SE) 90% (SE) 99.99% (SE) mean (AR)
median (AR) 90% (AR) 99.99% (AR)
4.70E-05
3.96E-09 2.62E-06
4.05E-02
2.23E-03
2.77E-07 1.84E-04
9.45E-01
6.40E-04
3.53E-07 1.90E-04
2.40E-01
1.90E-02
2.50E-05 1.40E-02
1
Treatment line 1
Treatment line 2
SE: Single event
AR: Annual Risk
Median and 90 Percentile Rotavirus Risk and Annual risk for
Consumers- Lettuce Consumption
Treatment line 1
1.00E+00
1.00E-01
1.00E-02
1.00E-03
1.00E-04
1.00E-05
1.00E-06
1.00E-07
1.00E-08
1.00E-09
Treatment line 2
median (SE)
90% (SE)
median (AR)
90% (AR)
SE: Single event
risk
AR: Annual
Risk
Table 3: Worker Inhalation Exposure Pathway 2
Parameter Name
Concentration of
Adenovirus in effluent
Recovery efficiency
Value
20
Unit
Pfu/mL
Assumptions

Log normal distribution
0.5
%
Treatment removal
efficiency T1
Treatment removal
efficiency T2
6
Log10



50% recovery efficiency is assumed
Triangular distribution
Triangular distribution
4
Log 10

Triangular distribution
Reference
Fong et al. (2009)
Humidity
0.8
%



Breathing rate
23
L/min


Time
120
min

Volume of water inhaled
2.208
L



Percentage of particles
generated by the spray
device in respirable range
75
%




Percentage of particles
deposited in lungs
4.1
%


Normal distribution
Maximum 1.0 and Minimum is 0.8
High humidity reflective of hot house
conditions
Adult male light physical activity
Log normal distribution and breathing
rate for 20 year old during heavy
exercise specified as maximum
Assumed that viruses would remain in
the air for a period of 2 hours following
spray irrigation and/or worker in vicinity
during and post spray for a period of 2
hours
No die-off of viruses
Triangular distribution
Computation based on 120 min 80%
humidity and breathing rate
No distribution specified
High pressure hose device used for
irrigation
Percentage based on a range of
experiments showing 69-96% for trigger
nozzles and humidifiers
No distribution specified point estimate
used
Only a proportion of aerosols of size 5
microns are deposited in the alveoli
No distribution specified point
estimated used



Haas et al 1999
Finlay 1994
Anderson et al 2007


Haas et al 1999
Anderson et al 2007


Anderson et al 2007
O’Toole et al 2009

ICRP Dosimetric
model
Table 4: Single and annual risks for workers exposed to adenovirus through inhalation
Treatment line 1
Treatment line 2
Risk
Annual Risk
mean (SE) median (SE) 90% (SE) 99.99% (SE) mean (AR)
median (AR) 90% (AR) 99.99% (AR)
3.95E-03
1.81E-06 1.78E-03
7.85E-01
9.30E-02
3.97E-04 3.24E-01
7.85E-01
7.45E-03
4.91E-05 8.78E-03
8.62E-01
1.75E-01
1.08E-02 8.56E-01
1.00E+00
Median and 90 Percentile Adenovirus Risk and Annual risk for
Workers- Inhalation
Treatment line 1
Treatment line 2
1.00E+00
1.00E-01
1.00E-02
median (SE)
90% (SE)
1.00E-03
median (AR)
90% (AR)
1.00E-04
1.00E-05
1.00E-06
Table 5: Worker Ingestion Exposure Pathway 3
Parameter
Concentration of rotavirus
Value
0.873
Unit
ufp/ml
Assumption

Taking the average of the highest
Reference

Oragui et al. (1989)
in effluent sewage

Recovery method
Treatment removal
efficiency
0.5
%


6
ulog1




concentrations’ results in considering
the worst case (conservative action).
They derived from different treatment
methods.
Using log normal distribution.
Assuming 50 % recovery with the
applied detection method.
Using triangular distribution.
Using multiple barriers before using
that water in the irrigation of lettuce I n
greenhouses.
6 log reduction obtained with the
applied treatment
Using triangular distribution.
& Rao et al. (1988)





Volume aerosols ingestion
per event
1
ml


Numbers of events per day
2
No.


Volume on hands
0.01
ml


Volume transfer to mouth
0.33
%


Number of hand to mouth
120
No.



Based on estimates by NSW firefighters
about volume exposure per event. This
is mean value
Assumed high pressure irrigation
system
Using log normal distribution.
Assuming 2 irrigations per day as a
normal situation, 3 times under dry
conditions and 1 time under more
humid conditions.
Using custom distribution.
Based on volume transfer obtained for
surface to hand transfer
Using log normal distribution.
Based on transfer of 33.9% observed for
hand to mouth transfer of PRD-1
bacteriophage
Using normal distribution.
15 times of the workers hands’ touching
their mouths multiplied by 8 working
hours.
Using normal distribution.

Shirasaki et al. Water
Research (2010),
B.Zhu et al. W.
Research (2005),
Madaeini S.S. Water
Research (1999),
O'Grady. Dev. Biol.
Stand. (1996),
EPA/NSF (2006),
Madaeini S.S. Water
Research (1999)
Hijnen et al. W,
Research (2006)
Dan Deere 2004
-

O’Toole et al 2009

Rusin et al 2002

Nicas, M, & Jones,
RM, (2009)
Table 6: Single and annual risks for workers exposed to rotavirus through ingestion
and surface-mouth contact
Treatment line 1
Treatment line 2
Risk
Annual Risk
mean (SE) median (SE) 90% (SE) 99.99% (SE) mean (AR)
median (AR) 90% (AR) 99.99% (AR)
5.40E-04
1.86E-07 2.27E-04
1.90E-01
3.74E-02
4.09E-05 4.86E-02
1
1.08E-03
5.35E-06 9.54E-04
2.45E-01
7.34E-02
1.17E-03 2.13E-01
1.00E+00
Median and 90 Percentile Rotavirus Risk and Annual risk for
Workers- Ingestion & Hand- Mouth Contact
Treatment line 1
Treatment line 2
1.00E+00
1.00E-01
1.00E-02
median (SE)
1.00E-03
90% (SE)
1.00E-04
median (AR)
90% (AR)
1.00E-05
1.00E-06
1.00E-07
SE: Single event
risk
AR: Annual
Risk
Table 7: Consumer Exposure Pathway 4 – Accidental exposure
Parameter name
[rotavirus] in sewage
effluent
Value
0.873
Unit
pfu/ml
Recovery of the method
0.5
%
Removal efficiency of the
treatment
4 or 6
ulog1
Frequency of event
Volume of water
consumed
Pfu = plaque forming units
Ulog1 = log removal
N = number
1
100
days
mL
Assumption
 The data was derived from different
treatment processes, and a average was
calculated
 Log normal distribution

Assumption = 50 % recovery with the
applied detection method

Triangular distribution

4 or 6 log reduction obtained with the
applied treatment

Triangular distribution

Uncommon event

Normal distribution

This is equivalent to a cup of water
Reference
 Oragui et al. (1989)
 Smith and Gerba (1982)
 Bates et al. (1984)
 Rao et al. (1988)
AGWR, 2006
Table 8: Annual risk for workers exposed to rotavirus through accidental ingestion
Treatment line 1
Treatment line 2
Annual Risk
mean
median
90% 99.99%
9.40E-03
1.09E-05 1.20E-02
4.86E-01
1.65E-02
2.64E-02 4.77E-02
4.90E-01
Median and 90 Percentile Rotavirus of Annual risk for WorkersAccidental Ingestion
Treatment line 1
Treatment line 2
1.00E+00
1.00E-01
1.00E-02
median
90%
1.00E-03
1.00E-04
1.00E-05
SE: Single event risk
AR: Annual Risk
Conclusions
- Lower viral health risk for Treatment 1 for both workers and consumers
- Treatment 1 leads to production of lettuce which meets drinking water standards (1/10000
infection per annum)
- Treatment 2 is a new technology and few data are available about its efficacy
Recommendations
– To assess the risk of using treated effluent from this location pathogens should be measured
in the secondary effluent over time (for instance during one year to cover also possible
seasonal distribution). Because it is probably difficult to detect these pathogens in small
volumes of water the concentrations in this secondary effluent should subsequently be used
in a QMRA (using the different log removals from the different treatment processes).
To estimate the viability of pathogens culture methods should be used.
– Indicator organisms, like bacteriophages, could be used to determine the removal in the
different treatments.
– To reduce the risk of formation of aerosols by drip irrigation.
– To reduce the risk of surface contamination of crops by using drip irrigation (drip irrigation
gives an additional 2log removal).
– Crop selection (e.g. non edible crops).
– To investigate a third treatment train which will be more cost effective (lower operation cost)
and will introduce a specific disinfection process to produce water suitable for edible crop
irrigation.
Datagaps
- Methods for isolation and detection of enteric viruses and other pathogens in water and
environmental surfaces, including hands
- Methods to determine viability of pathogens present in water and environmental surfaces,
including hands
- Assay methods to determine the percentage of aerosol mass transmitted via inhalation versus
ingestion during water using activities
- Data quantifying the transmission, persistence and inactivation of pathogens in a variety of
exposure scenarios
- Data quantifying the relationship between the transmission, persistence and inactivation
characteristics of pathogens and their surrogates used in experimental and field studies (i.e.
where no assay methods are available for pathogens or there are safety and/or logistical
reasons for not using pathogens).
- Interaction of foodborne pathogens with plant tissue
Risk communication:
- It is important that the stakeholders are consulted from the outset of the project, and that a
risk communication strategy is devised in advance of the implementation of any new scheme.
For example provision of information about water reuse and detailing benefits of the scheme
should be provided to green house farmers.
Study group members:
Fernando Valero Cervera: ATLL, Spain
Joanne E. O’Toole: Monash University, Australia
Mahjid M. Alkhalaf: Saudi Food and Drug Authority, Saudi Arabia
Michel Montemayor van Rooy: ATLL, Spain
Sahar S. Dalahmeh: SLU, Sweden
Willemijn J. Lodder: RIVM, the Netherlands
Supervisor: Dr. Joan Rose