Subscriber access provided by Fudan University
Article
Spatial models of sewer pipe leakage predict the occurrence of wastewater indicators in shallow urban groundwater Patrick R. Roehrdanz, Marina Feraud, Do Gyun Lee, Jay C Means, Shane A. Snyder, and Patricia A. Holden Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.6b05015 • Publication Date (Web): 20 Dec 2016 Downloaded from http://pubs.acs.org on December 29, 2016
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 35
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 27 28 29 30 31 32
Environmental Science & Technology
Spatial models of sewer pipe leakage predict the occurrence of wastewater indicators in shallow urban groundwater Patrick R. Roehrdanz1,2, Marina Feraud1,2, Do Gyun Lee1,2,3, Jay C. Means1,2, Shane A. Snyder4, Patricia A. Holden1,2* 1
Bren School of Environmental Science & Management, University of California, Santa Barbara, CA 93106, USA
2
Earth Research Institute, University of California, Santa Barbara, CA 93106, USA
3
Incheon National University, Incheon 22012, Korea
4
Department of Chemical and Environmental Engineering, University of Arizona, Tucson, AZ, USA
*Corresponding author: Patricia A. Holden, 3508 Bren Hall, Bren School of Environmental Science & Management, University of California, Santa Barbara, CA 93106-5131 E-mail:
[email protected]; phone: +1 805 893 3195; fax: +1 805 893 7612.
33
ACS Paragon Plus Environment
1
Environmental Science & Technology
34 35 36
Page 2 of 35
TOC Art
37
ACS Paragon Plus Environment
2
Page 3 of 35
38 39
Environmental Science & Technology
ABSTRACT Twentieth century municipal wastewater infrastructure greatly improved U.S. urban
40
public health and water quality. However, sewer pipes deteriorate and their accumulated
41
structural defects may release untreated wastewater to the environment via acute breaks or
42
insidious exfiltration. Exfiltrated wastewater constitutes a loss of potentially reusable water and
43
delivers a complex and variable mix of contaminants to urban shallow groundwater. Yet,
44
predicting where deteriorated sewers impinge on shallow groundwater has been challenging.
45
Here we develop and test a spatially explicit model of exfiltration probability based on pipe
46
attributes and groundwater elevation without prior knowledge of exfiltrating defect locations. We
47
find that models of exfiltration probability can predict the probable occurrence in underlying
48
shallow groundwater of established wastewater indicators including the artificial sweetener
49
acesulfame, tryptophan-like fluorescent dissolved organic matter, nitrate, and a stable isotope of
50
water (δ18O). The strength of the association between exfiltration probability and indicators of
51
wastewater increased when multiple pipe attributes, distance weighting, and groundwater flow
52
direction were considered in the model. The results prove that available sanitary sewer databases
53
and groundwater digital elevation data can be analyzed to predict where pipes are likely leaking
54
and contaminating groundwater. Such understanding could direct sewer infrastructure
55
reinvestment towards water resource protection.
56 57
Keywords: exfiltration, groundwater, sewer, wastewater, geographic information system
58 59 60
ACS Paragon Plus Environment
3
Environmental Science & Technology
61
Page 4 of 35
INTRODUCTION
62
Well-functioning sanitary sewer systems collect and convey urban wastewater to
63
centralized treatment plants where nutrients are sequestered into biosolid residues, and pathogen
64
concentrations are decreased significantly by disinfection before treated effluents discharge to
65
rivers and oceans. Collecting and treating sewage has profoundly improved water quality and
66
diminished the prevalence of infectious diseases in the 20th century urban U.S.1 The associated
67
infrastructure is by now immense: approximately 800,000 miles of publicly owned sewer mains2
68
convey sewage to wastewater treatment plants for 75% of the nation’s population.3 However,
69
many sewers worldwide are aged and deteriorated, with non-catastrophic accumulated defects
70
allowing sewage to leak, or exfiltrate, into surrounding soils.4-6 As a consequence of exfiltration,
71
nutrients, pathogens,4, 7-9 and legacy and emerging contaminants including pharmaceuticals10 can
72
enter shallow groundwater.11-14 Exfiltration rates vary8 and may total 10% or more of dry
73
weather collection system flow;4, 12, 15-18 thus, where wastewater is intended for recharge or reuse
74
post-treatment,19 exfiltrated sewage is a lost water resource. Exfiltration can contaminate
75
drinking water sources, recreational waters, and aquatic habitats, since shallow groundwater
76
migrates into deep aquifers via preferential pathways4, 20 and to surface waters via base flow
77
recharge21-24 or compromised storm drains.25, 26 Disinfectants and antimicrobial agents in
78
exfiltrated sewage contribute to the evolution and dissemination of antibiotic resistance genes in
79
the environment.27 As groundwater supplies shift with climate change,28 wastewater
80
infrastructure should be managed to reduce subsurface contamination.
81
Remediation of exfiltration requires sewer rehabilitation or replacement which, at a full
82
cost in the U.S. alone of more than $50 billion,3 is overall not economically feasible. Eliminating
83
alum in potable water treatment may prevent concrete sewer “crown corrosion”, whereby pipes
ACS Paragon Plus Environment
4
Page 5 of 35
Environmental Science & Technology
84
deteriorate from the strong acid formed following organic matter-fueled bacterial sulfate
85
reduction.29, 30 However, alum is not universally used in water treatment; further, sewer materials
86
other than concrete are susceptible to failure and leakage.4, 6, 11 Sewer rehabilitation decisions
87
involve socio-economic considerations of system failure diagnoses and risk estimates.31 Interior
88
pipe defects are typically discovered visually using closed circuit televising (CCTV)32 for select
89
regions of the collection system, and CCTV data have been shown to correspond to
90
micropollutant indicators of wastewater exfiltrated into groundwater.13, 33 Pipe defects have been
91
explained by various deterioration modeling approaches,34 for example using decision trees,35
92
regression statistics,6, 36-39 multivariate probability analysis,40 random forests,41 neural network
93
analyses, or linear programming42 that relate characteristics such as pipe age, pipe material of
94
construction, diameter, length, and slope to observed CCTV data.35, 38 Such modeling results can
95
guide rehabilitation locations or additional CCTV efforts.34 However, to prioritize within
96
collection system areas where groundwater is most vulnerable to contamination, a modeling
97
approach that accounts for system attributes affecting exfiltration into shallow groundwater is
98
needed. Previously, Lee et al.43 showed that evidence of wastewater influence in shallow
99
groundwater wells corresponded to modeled exfiltration probability scores of nearby sewers. The
100
modeled exfiltration probability scores resulted from newly employing an existing probabilistic
101
model of sewer pipe deterioration37 in the context of spatial information of municipal sewers and
102
groundwater elevations, then examining the relationship between the calculated exfiltration
103
probability and measured frequencies of wastewater indicators in nearby shallow groundwater.
104
Herein, we demonstrate that the modeling approach of Lee et al.43 is predictive: the
105
approach can be used to discover where, within a city sewer network, exfiltration is likely
106
impinging on shallow groundwater quality. The predictive power of the method was tested by
ACS Paragon Plus Environment
5
Environmental Science & Technology
Page 6 of 35
107
selecting high probability regions, then sampling a new set of groundwater wells with subsequent
108
analysis for indicator compounds. Because the Lee et al.43 approach involves directly marrying a
109
published pipe leakage algorithm to the municipal geographic information system (GIS) sanitary
110
sewer database and to publically available groundwater digital elevation data, any city can
111
conceivably apply the approach. The results support the use of a spatial model to target where
112
shallow groundwater is susceptible to exfiltration, thus allowing sewer segment rehabilitation to
113
be prioritized towards protecting water resources and/or public health as part of an overall risk
114
assessment/reduction scheme.
115 116
MATERIALS AND METHODS
117
Study area
118
The study area is within a small city (population 89,000; area 108 km2; total sewer line
119
length approx. 470 km) on California’s central coast. Municipal wastewater is conveyed to a
120
wastewater treatment plant (WWTP) via a collection system including sanitary sewers that are
121
separate from storm sewers, and that range from 5 to 126 years old. Most sewers in the city are
122
constructed from vitrified clay pipe (VCP; 76.6% of system length) or polyvinyl chloride (PVC;
123
21.9%), with minor sections of reinforced concrete, cast iron and high-density polyethylene
124
(HDPE).
125
The specific study area within the city was chosen because its sewers have varying
126
characteristics (Fig. S1) and proximities to groundwater that could produce a range of predicted
127
probabilities of shallow groundwater contamination from sewer exfiltration. Previously, the
128
relationship between sewer-derived groundwater contamination probabilities versus measured
129
wastewater indicators in shallow groundwater was established opportunistically, i.e. by modeling
ACS Paragon Plus Environment
6
Page 7 of 35
Environmental Science & Technology
130
leakage of sewers in the vicinity of existing wells that had been sampled and analyzed for a
131
spectrum of contaminants.43, 44 Here, the objective was to determine the predictive capability of
132
the modeling approach. This required mapping contamination probabilities predicted from
133
wastewater exfiltration models, then siting wells—either existing or new—from which to sample
134
groundwater, for testing model efficacy.
135 136 137
Hydrogeology The study basin is a saddle-shaped depression between a prominent mountain range to
138
the north, a coastal ridge to the southwest, and the Pacific Ocean to the southeast. Consolidated,
139
mostly impermeable sedimentary rocks of marine origin (Rincon Shale, Sespe Formation,
140
Monterey Formation, Vaqueros Sandstone) form the boundaries of the basin and the surrounding
141
topography.45 Unconsolidated alluvium of marine and continental origin fills the basin and
142
overlies the consolidated rocks to a depth of 500 feet.46 The alluvium is the principal water
143
bearing formation in the basin. Groundwater sampled in this study is exclusively from the
144
youngest alluvium primarily of continental origin and is comprised primarily of sand, gravel, silt
145
and clay to a maximum depth of 100 feet.47 Rainfall is the only major source of groundwater
146
basin recharge. Average annual rainfall in the area ranges from 15 to 19 inches, although rainfall
147
is highly variable with some years receiving over 40 inches and some less than 5
148
inches. Thorough descriptions of the hydrogeologic setting are available elsewhere.47-49
149 150 151 152
ACS Paragon Plus Environment
7
Environmental Science & Technology
153 154
Page 8 of 35
Groundwater elevations Co-kriging of water table measurements over the period 2000-2013 was used to
155
interpolate groundwater elevation for the entire city, using publically-available
156
(http://nationalmap.gov/elevation.html) digital elevation data,50, 51 and following the method in
157
Chung and Rogers.52 The digital elevation dataset is based on groundwater surface elevation
158
measurements for 1,017 individual wells in the city region, comprising 15,671 unique
159
measurements.53 The median standard deviation of water level measurements in individual study
160
wells was 0.26 m, with 97% of wells displaying a standard deviation less than 0.9 m for the
161
analyzed period. Spatial interpolation was performed with ArcGIS version 10.1 Geostatistical
162
Analyst extension (ESRI, Redlands CA).
163 164 165
Spatially explicit model of sewer exfiltration To assess the likelihood of exfiltration in different regions comprising the city study area,
166
an exfiltration probability surface was generated that combines spatially-explicit information of
167
the sewer infrastructure, groundwater elevations, and previously documented sources of
168
groundwater contamination. Briefly, all public pipes in the sewer network were designated as
169
either potentially exfiltrating (above water table) or infiltrating (below water table) based on
170
sewer structure elevation data and interpolated mean water table elevation. Assuming that
171
contaminant attenuation in soils would increase with vertical migration distance, only pipes that
172
were determined to be
