thesis defense 4

Center for Environmental Security

1

Methods and devices for assessment of fiprole pesticides in engineered waterways

Sam Supowit, environmental engineeringPhD thesis defense9/25/2015

Committee: Rolf Halden, chair; Paul Westerhoff; Paul C. Johnson


2

Importance

Methods• Sampling• Analysis

Identify contributing contaminants

DATA


3

Major contributions• Method for quantifying fipronil and byproducts in

wastewater/sludge• Time-integrating sampler that samples trace level

contaminants (e.g., fipronil) across the sediment-water interface

• Data on occurrence of fipronil in the environment• Fate of fipronil in a wastewater treatment plant

and wetland via a mass balance approach


IntroductionFipronil• Highly toxic to invertebrates• Occurs at trace levels (ng/L)• Lipophilic• Emerging contaminant• Byproducts equally/more toxic

4

Fipronil Sulfide Sulfone Amide Desulfinyl

Fiproles


5

Introduction – Rationale

Waterway

WWTP

Fish

Aquatic insects

Angiosperms

Pollinators

Birds


6

Introduction – Rationale • Only two prior studies assessed fiproles in

wastewater matrices

• Fiproles may be resistant to degradation in WWTPs (Heidler & Halden, 2009)

• Improved precision/sensitivity necessary to assess fate of fiproles in WWTPs and downstream waters


7

Introduction – Hypotheses

1.

2.

3.

Total fiprole mass balance


8


1.

2.

3.

≤


9


1.

2.

3. WWTP mass balance


10

• Fiproles are largely resistant to degradation in treatment.Hypothesis


11

• Fiproles are largely resistant to degradation in treatment.Hypothesis


12

• Fipronil implicated in colony collapse disorder

• Highly toxic to bees LD50 = 1-6 ng/bee


Compound Procambarusa Hyalella aztecab Diphetor hagenib 33 OC urban

water conc. (µg/L)

Half-life

31 LC50 (µg/L) 30 LC50 (µg/L)

30 EC50 (µg/L)

30 LC50 (µg/L)

30 EC50 (µg/L)

34 Silt loam (d)

35 Facultative conditions (d)

Fipronil 14.3-19.5 1.3-2.0 0.65-0.83 0.20-0.57 0.11-0.21 0.05-0.39 21±0.15 -

-desulfinyl 68.6 - - - - 0.05-0.13 - 217-497

-sulfide 15.5 1.1-1.7 0.007-0.003 - - ND >200 195-352

-sulfone 11.2 0.35-0.92 0.12-0.31 0.19-0.54 0.055-0.13 0.05-0.19 >200 502-589 aProcambarus species were clarkii and zonangulus. bValues for H. azteca and D. hageni are the 95% confidence interval. OC – Orange County, California ND – non detect

1


13


http://www.actbeyondtrust.org/wp-content/uploads/2013/07/IUCN2013sympo03_sluijs.pdf


14

≠


=


15

Methods1. Review active sampling plans– Identify and critique various means of sampling for

ascertaining fate of trace level HOCs 2. Optimize sampling, extraction, and analysis for precise,

trace-level detection of fiproles3. Design, develop, and deploy a sampler to assess fiproles

in pore water and surface water4. Assess the mass of fiproles in wastewater process

streams and wetland– Apply appropriate sampling strategy, and analysis method


16

1. Review of “active” sampling strategies

Snapshot in time

Volume sampled proportional to flow

Constant sample rate Average concentrations


17

1. Review of “active” sampling strategies: pore water sampling

SPE

Automatic water collection

Discrete grab sampling

In situ extraction


18

Feature

Grab Water collection In situ extraction

Bottle/bailer

Kemmerer sampler

MINI-POINT Rhizon

Thin layer film

sampler

Syringe In situ SPE

Active GFF/PUF sampling

ISCO/Sigma

Gulper samplers

Osmo-Sampler CFIS CSS Automatic

SPE units* CLAM PROFEXS SAMOS

Sampling specification

Time integration None0

None0

Potentiala

0 Potentiala

0None

0 Potentiala

0 None

0Excellent

0 None

0Good294,95

Good159

Good159

Potential196

Good17

Potential0

Potential0

Flow-weighting None0

None0

None0

None0

None0

None0

None0

Excellent13097-99

None0

None0

Potential0

None0

Potential0

None0

Potential0

Potential0

Time discrete Yes12100,101

Yes70102

Yes3284

Yes70102

Yes3989,103

Yes262,104

Yes187

Yes1105

Yes1106

No0

No0

No0

Yes30107,108

No0

Yes4769

Yes3989,103

Sample handlingVolatile loss reduction

Poor0

Poor0

Potential0

Potential0

Potential0

Potential0

Potential0

Very poor0

Potential0

Potential0

Potential0

Potential0

Potential0

Fair17

Potential0

Potential0

Adsorption loss reduction (HOCs)

Fair0

Potential0

Potential0

Poor0

Potential0

Potential0

Potential0

Fair0

Potential0

Potential0

Potential0

Potential0

Potential0

Good17

Potential0

Potential0

Applications

Mass balances Fair3900109-111

Potential0

N/A0

N/A0

N/A0

Potential0

Potential0

Excellent13997,112

N/A0

Potential0

Potential0

Potential0

Potential0

Potential0

Potential0

Potential0

Spacial characterization

Fair50113,114

Good12115

Good0

Good12115

Potential0

Good0

Good0

Good10116

Good1117

Potential0

Potential0

Potential0

Good0

Excellent2118

Good0

Good0

MatricesSediment pore water

Difficult860119,120

No0

Yes3284,121

Yes130102,122

No0

Potential0

No0

No0

No0

No0

No0

No0

Potential0

No0

No0

No0

Wastewater Yes1990123,124

Yes50125,126

No0

No0

Yes0

Yes0

Yes0

Yes273127,128

No0

No0

Yes0

Yes0

Yes0

Yes0

Yes0

Yes0

Cost range $ $$ $$ $$ $$ $ $ $$$ $$$ $$$ $$ $$ $$$ $$$ $$$ $$$$

Automation N/A Potentiala Potentiala Potentiala None Potentiala None High High Moderate Moderate Moderate Moderate Moderate High High

Commercially available Y N N Y N Y Y Y N Y Y N N Y N N

a – If an automatic pump were to be integrated (not present in current design)

*Any device aside from the CLAM that automatically extracts water using a pump and SPE resin.

WWTPMass balance + =

New wetland sampler

1. Review of “active” sampling strategies: comparisons


19

2. Water analysis method1000 mL

RAS & PS

500 mg/3 mL Strata XL 4 mL eluate x 2

Concentrations calculated by both standard addition and isotope dilution


20

2. Sludge analysis method

Surrogate addition

Acetone extraction Shake Centrifuge

Solvent switch to hexane

Cleanup on Florisil

Analyze by LC-MS/MS


21

3. New sampling device testingFiproles in a wetland• Sample across the sediment-water interface


22

4. Fiproles in a WWTP

PP

Wetland

River

==

Primary sedimentatio

n basins

Secondary sedimentatio

n basinsHeadworks Aeration

basins

PS Thickening Centrifuge

WAS Thickening Centrifuge

Acid Phase

Methane Phase

DS Thickening Centrifuge

Centrate Treatment

Disinfection


23

SamplingWWTP assessment• Sample all process streams

using most appropriate sampling method


24

Results – Analytical method

MRM chromatograms for spiked and unspiked biosolids

• 20 ppb spike• High intensities (> 106 cps)• High S/N


25

Table 3. Method detection limits for fiproles from various studies in water and sludge matrices.

Water (ng/L) Sludge (pg/g)

Source Schlenck et al, 2001

Heidler & Halden, 2009 Hladik et al, 2008 Weston &

Lydy, 2014b This study Heidler & Halden, 2009 This study

Matrix Surface water Wastewater inf/eff River water Wastewater

influentSurrogate

wastewater Sludge Surrogate sludge

ExtractionLLE (pentane) +

normal phase SPE (Florisil)

Reversed phase SPE

(HLB)

Reversed phase SPE (HLB)

LLE (DCM) + filtration

+ GPC

Reversed phase SPE

(StrataTM-X)

SLE (MeOH/acetone)

SLE (acetone) + normal phase SPE

(Florisil)

Analysis GC-ECD LC-MS/MS GC-ion trap MS GC-MS LC-MS/MS & GC-MS/MS LC-MS/MS LC-MS/MS &

GC-MS/MS

Fipronil 500 10-20 2-2.9 0.88-1.49 0.045 400 20

-sulfide 1000 N/A 1.8-2.2 0.88-1.49 0.16 N/A 140

-sulfone 2000 N/A 3.5-7.0 0.88-1.49 0.07 N/A 100

-amide N/A N/A N/A 0.88-1.49 0.3 N/A 90

-desulfinyl 500 N/A 1.6-2.7 0.88-1.49 0.77 N/A 240

a The range shown (0.88-1.49) was given for all fiproles, and no individual MDLs are published

HLB, hydrophilic-lipophilic balance; DCM, dichloromethane; MeOH, methanol; GPC, gel permeation chromatography; SPE, solid phase extraction; LLE, liquid-liquid extraction; SLE, solid-liquid extraction

Results – Analytical method


26

Applying the method• Use modified method to validate new biphasic

sampling device

• Use method to perform WWTP mass balance


27

Sampler development–the in situ sampler for biphasic water sampling (IS2B)

1. Designed and built an “active” biphasic sampler (IS2B)

2. Applied analytical method involving in situ SPE using IS2B

3. Validated method using fiproles as targets– Deployed device and demonstrated utility

End cap Shell Peristaltic pump

Porewater inlet manifold

SPE cartridges

Outlet manifold

InletsBulk water inlet manifold End cap

Inlets/outlet

B - Offline

A - Online

Organic Eluent Non-volatile &semi-volatile organics0.2 µm Filter

LC-MS/MS

Bulk water

Pore- Water

Discharge into bulk water

× 3 SPE cartridges

Manifolds

× 3

Results – sampler


28

Features

• Time-integrated pore water sampling

• Biphasic sampling

• In situ SPE

• Enables low detection limits

Results – sampler


29

• Results similar to conventional method

• Successful deployments of 48 hrs

• Yielded time-integrated values

• Precision lower than grab sample field replicates

ResultsIS2B sampler – wetland test


30

ResultsIS2B sampler – wetland test

Chemical Fipronil -Sulfide -Sulfone -Amide -Desulfinyl Total fiproles

I

BWIS2B 14.1 ± 3.3 ND (<0.7) 4.0 ± 1.3 ND (<0.8) 0.04 ± 0.14a 18.1 ± 4.6

LEA 10.0 ± 0.8 ND (<0.7) 3.4 ± 0.5 ND (<0.8) ND (<0.05) 13.4 ± 1.3

PWIS2B** 7.5 ± 1.0 1.4 ± 0.4a 3.7 ± 0.7 ND (<0.8) ND (<0.04) 12.6 ± 2.1

LEA 5.3 ± 0.2 1.4 ± 0.5a 1.9 ± 0.7 ND (<0.8) ND (<0.05) 8.6 ± 1.4

IIBW

IS2B 5.0 ± 2.5 0.8 ± 0.5a 2.3 ± 0.9 1.4 ± 0.7a 0.35 ± 0.16a 9.9 ± 4.6

LEA 3.0 ± 0.1 2.8 ± 1.1 2.2 ± 0.1 2.3 ± 0.2a ND (<0.05) 10.3 ± 1.5

PW IS2B* 5.6 0.94a 2.9 2.0a 0.3 11.6

IIIBW

IS2B 5.4 ± 0.8 0.8 ± 0.1a 3.7 ± 0.9 2.4 ± 0.4a 0.06 ± 0.11a 12.4 ± 2.3

LEA 4.6 ± 0.2 0.8 ± 0.1a 3.3 ± 0.1 2.0 ± 0.1a ND (<0.05) 10.7 ± 0.5

PW IS2B 4.2 ± 1.4 ND (<0.7) 2.9 ± 1.0 1.9 ± 0.5a 0.09 ± 0.08a 9.1 ± 3.0

• Pore water and bulk water concentrations similar in 2/3 of locations• Low sediment OC (~1%) low adsorption of fiproles• Pore water grab sample results similar to IS2B pore water results

a - values are below the limit of quantitation, and are therefore estimatedSampling locations I, II, and III are those referenced in Figure 2BW, bulk water; PW, pore water; LEA, laboratory extraction apparatus (large volume)Standard deviations shown are calculated from n=3, except where indicated*n=1 field replicate (2-day, time-averaged composite)**n=2 field replicates (2-day, time-averaged composite; ± values provided represent maximum/minimum)

Anchor

Porewater inlet

Bulk water inlet

Water outlet

Anchor

Porewater inlet

Bulk water inlet

Water outlet


31

IS2BRedesign

Figure 5-2. Pictures showing the internal components (panel a), including the syringes, pump motor, battery, and pore water extraction cartridges. Panel b shows the completed construction with a clear PVC shell. Panel c shows several individual components, including (from bottom to top) the stainless steel bottom cap, acrylic top cap, step motor, interior chassis, O-rings, battery, and shell. Panel d shows the constructed top tap with Swagelok fittings (for bulk water intake).

Figure 5-1. Diagram of the mIS2B.


32

PP

Wetland

River

==

Primary sedimentatio

n basins

Secondary sedimentatio

n basinsHeadworks Aeration

basins

PS Thickening Centrifuge

WAS Thickening Centrifuge

Acid Phase

Methane Phase

DS Thickening Centrifuge

Centrate Treatment

Disinfection

Results – WWTP mass balance


33

Results – WWTP mass balanceConcentrations of fiproles in wastewater streams• Fipronil (parent) is dominant congener in most streams• Sulfone is higher in RAS, indicating probable aerobic degradation of parent• Sulfide is highest in DWS, indicating probably anaerobic degradation of parent

Stream Fipronil -sulfide -sulfone -amide -desulfinyl 1 Total fiproles (as fipronil)

Influent 22.5 ± 4.5 NP 6.7 ± 1.8 NP 0.5 ± 0.8 29.5 ± 4.8

Primary effluent 21.4 ± 3.4 0.9 ± 2.9 5.2 ± 2.6 NP 0.2 ± 0.2 27.6 ± 5.8

Primary sludge 99.7 ± 53.0 3.0 ± 6.7 13.8 ± 9.3 NP NP 107.0 ± 54.5

Return activated sludge 33.7 ± 8.7 7.8 ± 0.8 25.0 ± 3.8 3.0 ± 0.3 0.01 ± 0.02 76.9 ± 25.6

Secondary effluent 16.4 ± 2.6 2.0 ± 1.4 12.4 ± 11.9 NP 0.1 ± 0.1 30.5 ± 12.9

Chlorination basin effluent 16.2 ± 2.3 0.8 ± 0.6 7.3 ± 5.1 0.7 ± 0.9 0.1 ± 0.1 24.9 ± 5.6

Plant effluent 20.1 ± 3.7 0.6 ± 0.3 5.9 ± 3.9 1.1 ± 1.0 0.1 ± 0.2 27.6 ± 5.6

Wetland effluent 14.7 ± 2.7 0.8 ± 0.4 4.4 ± 2.9 1.1 ± 0.6 0.1 ± 0.2 21.0 ± 4.2

Dewatered sludge* 2.0 ± 0.6 8.4 ± 4.2 1.3 ± 0.9 0.1 ± 0.0 1.2 ± 1.8 3.2 ± 1.5

NP, no peaks detected

* concentrations expressed as ng/g dry weight sludge1 detected concentrations near the MDL, estimated


34



35

Individual fiprole mass loads

• Reduction in fipronil coincides with increase in byproducts.

• Total fiprole mass from primary influent through disinfection effluent is not discernably different.



36

• No attenuation of total fiproles in WWTP

• About half the fiprole mass is attenuated in wetland

47 ± 13% total fiprole reduction

No discernable changeResults – WWTP mass balance


37

WWTP fate study – take home points•Conventional wastewater treatment is not efficient at removing fiproles.•Reduction in parent compound mass may coincide with degradate formation (sulfone, in particular).•Total fiprole levels re-entering the environment from wastewater treatment are toxicologically relevant and may impact biota (bees?). •Fiproles are “lost” in the wetland at a rate of about 50%


38

Hypotheses revisited1. Assessing fipronil and its byproducts in wastewater and surface water

will generate data precise enough to perform total fiprole mass balances in engineered waterways. – Most streams %RPD (concentration) < 20 %

2. An automatic biphasic sampling tool capable of time-integrated sampling and in situ solid phase extraction (SPE) can produce quantitative data comparable to conventional methods. – Grab sample and IS2B data was similar– %RSD for IS2B was greater

3. Since fipronil is resistant to degradation, a mass balance conducted over a WWTP and wetland will show fiproles to be highly conserved.– True in WWTP– Not true in wetland


39

Research implications and recommendations

Sam Supowit, environmental engineeringPhD thesis defense9/25/2015

Committee: Rolf Halden, chair; Paul Westerhoff; Paul C. Johnson


Implications• Total fiprole mass discharge = 7.9 Σf g/day (0.017 lb/day)

into wetland = 6.3 lb/yr


41

ImplicationsThe entire volume of AG fipronil in the U.K. during peak use was about 124 kg/yr (273 lb/yr)

In Japan, it’s about 50,000 kg/yr

In California, it’s about20,000 kg/yr

The estimated, extrapolated discharge by US WWTPs is ~ 500 – 700 kg/yr (1100 lb/yr)

ExtrapolationEstimated ~0.4% of market volumedischarged by WWTPs


42

Implications & Recommendations

IS2B data to model contaminant availability

Determine where fiproles in wetland are going

Environmental impacts

pore water

bulk waterwater

column biota

benthic biota

Passive exchange

Passive exchange + particulate ingestion

Passive exchange + food ingestion

Waterway

WWTP

Fish

Aquatic insects

Angiosperms

Pollinators

Birds


43

Recommendation: expansion of the search

• Apply sampling/analysis methods to other emerging and legacy contaminants fill data gaps.– Triclosan– Triclocarban– Neonicotinoids– Metolachlor– Trifluralin– Pyrethroids


Acknowledgments

Committee:Rolf HaldenPaul JohnsonPaul Westerhoff

Team members:Akash SadariaIsaac RollArjun VenkatesanEdward Reyes

Collaborators:

Nancy DenslowViet DangKevin KrollTop secret logistics aids


45

QUESTIONS

Documents

thesis defense 4