Spatial Models of Sewer Pipe Leakage Predict the ... - ACS Publications

Dec 20, 2016 - Shane A. Snyder,. ∥ and Patricia A. Holden*,†,‡. †. Bren School of Environmental Science & Management, University of California...
0 downloads 0 Views 1MB Size
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