Emerging Contaminants
Program
2008-2011 Summary Report
Overview of Presentation

Monitoring Program
 Why,

Who, Where
Analytical Methods
 QAQC

Study Results
 2008-2011
Monitoring Program Overview
Why Monitor?
 Shift
from agriculture to municipal water use
 Increased awareness of water quality
 Increased awareness of emerging contaminants in
water supplies
 Antidepressants,
Endocrine disruptors
 Impacts on aquatic life
 Strengthening
of the regulatory framework
 Baseline of data for drinking water supplies
 Assess changes in the future
Monitoring Program Overview
Cooperative Monitoring Program
 Began
in 2008
 Co-funded
 Boulder,
Broomfield, Fort Collins, Greeley, Longmont,
Loveland, Estes Park, Northern Water and U.S. Bureau of
Reclamation
 Pharmaceuticals,
Personal Care Products, Endocrine
Disruptors, Pesticides and Herbicides
 Analysis done at CU Boulder, Center for Environmental
Mass Spectrometry
Sampling Sites
 19
Sites
 Reservoirs,
canals,
streams, raw and
finished water
 C-BT and non-C-BT
water supplies
 Evolving
 Four
events yearly
 Seasonal
influences
 Runoff
 Recreation
 Reservoir
processes
Monitoring Program Overview
Sample Collection
Shared by co-funders
 Protocol

Use of gloves
 No DEET
 No caffeine
 No nicotine


Collection Method
Grab Sample
 Kemmerer

QAQC

Trip and field blanks
Analytical Methods

Sample Preparation by Solid Phase Extraction

Low Level Method by LC/MS-MS(EPA 1694)

Hormone Method by LC/MS-MS (EPA 1698)

Differences of Hormone Method with UL labs

Screening Method by LC/TOF-MS

QA/QC Program
CU (CEMS)
Low Level Analysis (LC-MS-MS)
Liquid Chromatography (HPLC)
Mass Spectrometer (MS)
Selective analysis of individual
compounds, based on:
Some new compounds were added to the low level method in 2011: lamotrigine,
sucralose, triclosan, venlafaxine… In 2012: atrazine…
Hormone Method (LC-MS-MS)
E th in yl estrad io l-1 7 alph a
T estosteron e
OH
OH
H
H
H
H
HO
H
H
O
C 20 H 23 O 2 E xact M ass: 29 5 .1 70 4
C 19 H 2 9 O 2 +
E x act M ass: 28 9.2162



Hormones ionize poorly in LC-MS.
Need a more sensitive method (low ng/L).
Low level method was modified (adding NH3 to
mobile phase) to increase detection.
Hormone Analysis
Difference in reporting limits between CEMS and UL
Compound
17 alpha-Ethynyl
estradiol
17 beta-Estradiol
Reporting Limit
(ng/L)
CEMS
UL
Hormone Type
Birth Control Pill
10
0.5
5
0.5
20
100
Estrone
1
0.5
Phenolic Endocrine
Disruptor
Natural Metabolite
Progesterone
1
0.1
Natural Female
Testosterone
1
0.1
Natural Male
Bisphenol A
Natural Female
Comparison of findings between CEMS and UL
Value (ng/L)
Compound
BT-DLU
June 2011
BT-UTD
August 2011
17 alpha-Ethynyl estradiol
17 beta-Estradiol
Bisphenol A
Estrone
Progesterone
Testosterone
Testosterone (cis)
Testosterone (trans)
17 alpha-Ethynyl estradiol
17 beta-Estradiol
Bisphenol A
Estrone
Progesterone
Testosterone
Testosterone (cis)
Testosterone (trans)
CEMS
<40
<40
<20
<5
<1
<1
<10
<5
<20
<1
<1
<1
UL
<0.5
<0.5
<100
<0.5
0.2
<0.1
0.1
<0.5
<0.5
<0.1
0.6
0.1
<0.1
<0.1
QA-QC Data (LC-MS-MS)
Compound
Linear Regression Equation
2,4-D
y = 0.0489x
Acetaminophen
y = 0.1197x
Atenolol
y = 0.0728x
Bisphenol A
y = 0.0021x
Bupropion
y = 1.7004x
Caffeine
y = 0.1285x
Carbamazepine
y = 0.7066x
Clarithromycin
y = 0.9714x
Cotinine
y = 0.0522x
DEET
y = 0.9445x + 0.0609
Diazinon
y = 2.1603x
Diltiazem
y = 5.4689x
Diphenhydramine
y = 5.0469x
Diuron
y = 0.0695x
Erythromycin
y = 0.3208x
Fluridone
y = 1.3361x - 0.0299
Gemfibrozil
y = 0.0877x
Lamotrigine
y = 0.1375x
Metoprolol
y = 0.4529x
Propranolol
y = 0.5392x
Sucralose
y = 0.0095x
Sulfamethoxazole
y = 0.0675x
Triclosan
y = 0.0039x + 3.21x10-4
Trimethoprim
y = 1.1399x
Venlafaxine
y = 3.2319x
17-a-Ethinylestradiol y = 23.28x - 686.33
17-b-Estradiol
y = 35.10x - 1023.53
Estrone
y = 166.86x - 1089.71
Progesterone
y = 8203.15x
Testosterone
y = 6313.95x - 33533.14
R2
0.998
0.999
0.999
0.953
0.996
0.998
0.997
0.991
0.998
0.999
0.995
0.995
0.995
0.996
0.991
0.999
0.996
0.997
0.996
0.995
0.999
0.997
0.997
0.995
0.995
0.999
0.999
0.989
0.985
0.999
Spiked Sample 1 Spiked Sample2 Spiked Sample 3
Average
Feb 2011
June 2011
August 2011 Concentraton
43.0
41.4
42.2
42.2
47.1
47.0
51.3
48.5
42.4
41.4
46.3
43.4
45.2
40.9
50.2
45.4
38.8
35.0
41.2
38.3
40.4
39.6
44.2
41.4
43.3
40.9
36.2
40.1
41.6
37.8
34.6
38.0
42.4
39.6
43.1
41.7
46.6
44.1
47.7
46.1
40.8
40.2
47.1
42.7
42.9
39.7
42.3
41.6
41.8
38.9
42.2
41.0
40.0
37.5
43.3
40.3
42.7
39.2
40.4
40.8
38.9
37.2
42.2
39.4
40.3
38.2
45.3
41.3
34.4
33.5
37.5
35.1
55.9
52.0
60.3
56.1
50.0
45.7
50.0
48.5
n.a.
39.4
41.2
40.3
41.8
39.6
44.9
42.1
n.a.
34.9
37.5
36.2
47.8
43.5
50.1
47.2
37.8
34.5
38.2
36.9
n.a.
81.3
80.8
81.1
n.a.
81.3
79.7
80.5
n.a.
83.3
81.2
82.2
n.a.
84.0
79.8
81.9
n.a.
81.9
78.6
80.3
Standard
Deviation RSD (%)
0.80
1.9
2.43
5.0
2.56
5.9
4.67
10.3
3.12
8.1
2.45
5.9
3.61
9.0
3.50
9.2
1.82
4.4
1.83
4.0
3.84
9.0
1.67
4.0
1.81
4.4
2.89
7.2
1.78
4.4
2.54
6.4
3.62
8.8
2.06
5.9
4.17
7.4
2.51
5.2
1.30
3.2
2.67
6.4
1.80
5.0
3.36
7.1
2.04
5.5
0.35
0.4
1.15
1.4
1.52
1.8
2.92
3.6
2.32
2.9
Field blanks
also included
in each
sampling set
(< LOD)
Field samples were spiked at 40 ppt and extracted and analyzed with the LC-MS-MS Method to calculate
Standard Deviations of the methodology. EDCs were spiked at 80 ppt to be above LOD's.
n.a. = not analyzed for that sampling event
Screening Method by LC/TOF-MS
 Analyze
4,000,000,000 times/second
Vial in TOF Mode
 Database Search Pesticides
 Database Search Pharmaceuticals
 Verify Positives by Manual Inspection
 Search for Unknowns—Fluridone, Lamotrigine…
 Standard Analysis where Necessary
 The LC/TOF-MS instrument (Agilent 6220) is our tool for
identification of non-target compounds and unknowns.
 It provides accurate masses below 2 ppm error and it has
been succesfully used for many of our current projects.
 The high resolving power of 4 GHz allows the separation of
isobaric interferences from compounds of interest.
 Linearity of response is over 2 orders of magnitude.
Study Results Overview


Characterization of WWTP Influence
Pharmaceuticals and Personal Care Products
 WWTP

inputs
Herbicides and Pesticides
 Canal
maintenance
 Weed control adjacent to waterways

Recreational Influences
 Streams
and reservoirs
 Summer months
WWTP Inputs
 Fraser
River
 Four
WWTPs above
Windy Gap
 Willow
 Three
 Big
Creek
Lakes SD
Thompson River
 Estes
Park SD
 Upper Thompson SD
 Barker
Reservoir
 Nederland
WWTP
Characterization of WWTP Influence
Sampling of WWTP effluent
 Sample
representative of all WWTP
in study area
 Provide baseline of compounds
found in study area
 June 2009 and August 2010
 June
2009 – LC/TOF method for
analysis
 August 2010 – LC/MS low-level
method for analysis
Compound Type
Herbicides and
Pesticides
Household Products
Personal Care
Products
Endocrine Disruptor
Pharmaceutical and
Drugs
Compound
Detection Limit
(ng/L)
Classification
2,4-D
Diazinon
Diuron
Fluridone
5
1
5
5
Herbicide
Insecticide
Herbicide
Herbicide
Sucralose
DEET
15
20
Artificial Sweetener
Bug Repellant
Triclosan
20
Antibacterial
Bisphenol A
20
Plasticizer
Acetaminophen
Atenolol
Bupropion
Caffeine
5
5
1
10
Analgesic
Blood Pressure
Antidepressant
Stimulant
Carbamazepine
Clarithromycin
Cotinine
Diltiazem
2
2
1
5
Antidepressant
Antibiotic
Stimulant
Blood Pressure
Diphenhydramine
Erythromycin
Gemfibrozil
Lamotrigine
Metoprolol
Propranolol
5
10
5
5
1
1
Antihistamine
Antibiotic
Analgesic
Antidepressant
Blood Pressure
Blood Pressure
Sulfamethoxazole
Trimethoprim
Venlafaxine
5
5
1
Antibiotic
Antibiotic
Antidepressant
Characterization of WWTP Influence
Endocrine Disrupting Compounds
Analysis at Underwriters Laboratory



17 compounds analyzed, 5 compounds detected
Low number of detections at very low concentrations at all
sites during study period
Compound
Classification
17 alpha-Estradiol
Estrone
Nonylphenol isomer mix
Phenylphenol
Progesterone
Reproductive Hormone
Reproductive Hormone
Phenolic Endocrine Disruptor
Phenolic Endocrine Disruptor
Reproductive Hormone
June
2009
(ng/L)
Aug
2010
(ng/L)
0.8
12
900
100
<0.1
<0.5
6.5
1100
100
0.2
Characterization of WWTP Influence
 Samples
of WWTP effluent no longer collected
 New sampling site in 2011
1
mile downstream of Lake Estes
 Influence from Estes Park and Upper Thompson SDs
 LC/MS
low-level method can capture compounds of concern
Upstream to Downstream
 August 2010 WWTP effluent
 August 2011 site in Big
Thompson River below WWTP
 In general, same compounds at
a lower concentration
 August 2011 added
lamotragine, sucralose, triclosan
and venlafaxine
 Compounds that persist
downstream: Carbamazepine,
gemfibrozil, lamotrigine,
metoprolol, sucralose,
sulfamethoxazole and
venlafaxine
Pharmaceuticals and Personal Care Products
Horsetooth Reservoir
 Samples collected at
three depths: Top,
Metalimnion, and
Bottom
 Compounds consistent
with those found in
the Hansen Feeder
Canal
• Metoprolol detected
at every sampling
event
• Elevated sucralose
concentrations
 No apparent spatial variation, compounds stable in reservoir due to
constant influx
 Some variability with depth but no apparent pattern
Pharmaceuticals and Personal Care Products
Boulder Reservoir
 Less
detections and
lower concentrations
 More detections of
cotinine compared to
other sites
 Cotinine and
sucralose persist in
the finished water
 Very
low
concentrations
Herbicides and Pesticides

Widely used throughout study area
Agriculture, private land owners, parks and golf
courses
 2,4-D most common ingredient
 Often applied adjacent to waterbodies


Northern Water Operations
Control of terrestrial weed, algae and aquatic
plants in and around canals
 Hansen Feeder Canal, Saint Vrain Supply Canal and
Boulder Feeder Canal
 Weed control around Carter Lake and
Horsetooth Reservoir

Herbicides and Pesticides
Big Thompson River
 Some influence from
WWTP for 2,4-D diazinon
and diuron
 2,4-D most frequently
detected
 August 2011 at BT-DLU,
highest 2,4-D
concentration for all the
sampling sites for the
study period
Low concentration upstream
 BT-DLU next to park

Herbicides and Pesticides
Boulder Reservoir
 Different
weed control
efforts in Boulder
Feeder Canal
 Earth-lined
 Vascular
aquatic plants
 Sonar©
Applied in fall when
canal is dewatered
 Active ingredient
fluridone

Herbicides and Pesticides
Boulder Reservoir
 2,4-D,
diuron and fluridone all detected
 Most detections with highest concentrations at sites closest to
Boulder Reservoir
 Detections in finished water
Recreational Activities



Detections of caffeine, DEET and
triclosan can be attributed to
WWTP but can be increased by
recreational activities
Boulder Reservoir, Carter Lake
and Horsetooth Reservoir
Cache la Poudre and Big
Thompson Rivers
Recreational Activities
Big Thompson River
 Caffeine
and DEET in
WWTP effluent
 Caffeine
highest
concentrations

Generally higher than
PPCPs and herbicides
 Downstream
concentrations that are
higher than or equal to
the WWTP sites
suggest additional
source
Recreational Activities
Horsetooth Reservoir
 Caffeine detected at all
sampling events
 No apparent spatial
pattern
 No pattern at different
depths
Boulder Reservoir
 Spike in caffeine in BFC in August
2011 may suggest additional
source
 Caffeine and DEET detected in the
finished water at very low
concentrations
Conclusions





Drinking water sources in study area have very clean
water
Some compounds were found to persist throughout the
study area
Influence of WWTPs is apparent but dilution is
significant
Boulder Feeder Canal is impacted by herbicide
applications
The program should continue to evolve to include more
compounds currently being considered for regulation by
the EPA
Thank You

Co-Funders
City of Boulder
 City of Broomfield
 City of Fort Collins
 City of Greeley
 City of Longmont
 City of Loveland
 Town of Estes Park
 US Bureau of Reclamation


CU Boulder Center of Environmental Mass Spectrometry
Imma Ferrer
 Michael Thurman

Questions?
Jen Stephenson
Northern Water
jstephenson@northenwater.org
970 (622-2334)
Report available online at:
http://www.northernwater.org/WaterQuality/WaterQualityReports