Subscriber access provided by ORTA DOGU TEKNIK UNIVERSITESI KUTUPHANESI
Article
Investigating the Energy-Water Usage Efficiency of the Reuse of Treated Municipal Wastewater for Artificial Groundwater Recharge Eric Daniel Fournier, Arturo A. Keller, Roland Geyer, and James Frew Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.5b04465 • Publication Date (Web): 13 Jan 2016 Downloaded from http://pubs.acs.org on January 13, 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 34
Environmental Science & Technology
Investigating the Energy-Water Usage Efficiency of the Reuse of Treated Municipal Wastewater for Artificial Groundwater Recharge Eric D. Fournier, Arturo A. Keller,∗ Roland Geyer, and James Frew
1
Bren School of Environmental Science & Management, University of California at Santa Barbara, Santa Barbara, CA, USA, 93106-5131 E-mail:
[email protected] Phone: +001 805 8937548. Fax: +001 805 8937612
Abstract
2
3
This project investigates the energy-water usage efficiency of large scale civil infras-
4
tructure projects involving the artificial recharge of subsurface groundwater aquifers via
5
the reuse of treated municipal wastewater. A modeling framework is introduced which
6
explores the various ways in which spatially heterogeneous variables such as topogra-
7
phy, landuse, and subsurface infiltration capacity combine to determine the physical
8
layout of proposed reuse system components and their associated process energy-water
9
demands. This framework is applied to the planning and evaluation of the energy-
10
water usage efficiency of hypothetical reuse systems in five case study regions within
11
the State of California. Findings from these case study analyses suggest that, in certain
12
geographic contexts, the water requirements attributable to the process energy con-
13
sumption of a reuse system can exceed the volume of water that it is able to recover
14
by as much as an order of magnitude. ∗
To whom correspondence should be addressed
1 ACS Paragon Plus Environment
Environmental Science & Technology
15
Introduction
16
In recent years, the experience of persistent severe drought conditions in the Western United
17
States has motivated regional water resource management authorities to seriously pursue a
18
number of alternative sources of freshwater supply. 1–5 Prominent among these alternative
19
sources is reclaimed or recycled water; synonyms, which refer to the water that can be ob-
20
tained from a municipal sewage waste stream through the successive application of advanced
21
tertiary treatment processes. 6,7
22
Historically, the majority of projects involving the reuse of treated municipal wastewater
23
have been limited to relatively low quality end use applications such as the irrigation of
24
urban landscape features (e.g. golf courses, cemeteries, etc.) or agricultural crops that are
25
not destined for human consumption (e.g. turf, animal feedstocks, cover crops, etc.). 8–10 This
26
limitation has not been so much the consequence of any technical inability of wastewater
27
treatment processes to supply highly purified, even potable water, but rather due to the high
28
economic cost or strong social objections to the implementation of projects oriented towards
29
higher quality end uses. 11–18 Indeed, it is only in the past two decades that the reuse of
30
treated wastewater for relatively higher quality end use applications has become both more
31
socially acceptable and economically viable. 19–22
32
One of the fastest growing of these high quality end use applications for the reuse of
33
treated municipal wastewater is the artificial recharge of subsurface groundwater aquifers. 23–25
34
This process involves taking the reclaimed water from its point of production – the wastew-
35
ater treatment plant [WWTP] – and physically transporting it to another location where
36
it can be returned to the subsurface either through a pump driven direct injection well or,
37
more commonly, through surface infiltration basins. 26 The surface infiltration basins are es-
38
sentially just small to medium sized (∼1-10 hectares) engineered ponds whose beds have been
39
lined with highly permeable materials that are meant to facilitate the vertical movement of
40
water downward, under the force of gravity, into the underlying aquifer. 27
41
The use of reclaimed water for the artificial recharge of subsurface aquifers is meant to 2 ACS Paragon Plus Environment
Page 2 of 34
Page 3 of 34
Environmental Science & Technology
42
provide at least two categories of benefits. The first category relates to the various ways in
43
which the activity might improve the health of the aquifer being recharged. These benefits
44
range from the mitigation of overdraft conditions to the prevention of salt-water intrusion in
45
coastal areas. 28–30 The second category of benefits relates to the choice of reclaimed water
46
as the feedstock for the recharge activity. The main perceived benefit of this choice stems
47
from the commonly held belief that this type of reuse activity provides a net savings of the
48
water that is being constantly produced as treated effluent by wastewater treatment plants;
49
water that would otherwise be ”lost” to the environment. 28–30 It should be noted that this
50
belief implicitly discounts the value of any potential ecosystem services that such effluent
51
discharges might provide. 31
52
However, the assertion that the use of treated municipal wastewater for aquifer recharge
53
actually amounts to a net water savings is one which has not yet been rigorously proven by
54
an analysis of the net consumptive water use for an existing reuse system. 32–35 Furthermore,
55
we hypothesize that the validity of this assertion is likely to depend heavily upon the local
56
geographic context of the hydrologic basin in which a reuse project is to be implemented.
57
This hypothesis is rooted in an awareness of the dynamic interrelationship of our energy and
58
water systems, sometimes referred to as the Energy-Water Nexus. 36–39
59
Each different geographic region is characterized by a different portfolio of energy gen-
60
eration technologies that is responsible for the production of the electricity being locally
61
consumed. 40,41 This portfolio is typically referred to as the local grid mix. It has been
62
observed by others that, depending upon the composition of generation technologies com-
63
prising this grid mix, the consumption of electricity within a region can be associated with
64
non trivial withdrawals and or consumption of freshwater due to such processes as the cool-
65
ing of thermoelectric coal and natural gas fired power plants or the evaporative losses from
66
reservoirs associated with the operation of hydroelectric dams. 42–44
67
We observe that the quantity of process energy consumed by the reuse of treated wastew-
68
ater for artificial recharge applications is likely to strongly depend upon local geographic con-
3 ACS Paragon Plus Environment
Environmental Science & Technology
69
text, with characteristics such as the length and the elevation profile of the pipeline corridor
70
connecting the WWTP to the recharge site being intrinsically linked to regional topography,
71
existing landuse, etc. 45,46 In light of this observation, we further hypothesize the possibility
72
that, in certain settings, the reuse of treated municipal wastewater as a source of new alter-
73
native water supply might be associated with process water consumption demands that are
74
in excess of the quantity of water being locally recovered within a given span of time.
75
If such a situation were found to exist it would mean that these types of reuse activities
76
do not actually save water, in a net sense, but rather amount to little more than a mechanism
77
for the regional importation of virtual water: one wherein water is transported, virtually, in
78
the form of electricity, from the hydrologic basin housing the electricity production facilities
79
to the hydrologic basin housing the water reuse facility. 47 It is important to recognize that
80
such a situation, in and of itself, would not necessarily nullify the perceived environmental
81
benefits of these types of water reuse projects. Rather, it would only suggest that these
82
assumed benefits be more critically evaluated and in a manner which takes into consideration
83
the spatial relationship between the water scarcity of the watershed in which the reuse
84
project is being implemented relative to the water scarcity within the powershed in which
85
the corresponding energy is being produced.
86
Methods
87
In order to develop a better quantitative understanding of the expected net energy-water
88
usage efficiency of proposed new water reuse projects involving significant artificial ground-
89
water recharge components we developed a three part numerical modeling framework that
90
consists of both planning and assessment components. The overarching goal of this modeling
91
effort was to provide a generic platform from which the following three challenges could be
92
addressed for any proposed geographic context.
93
1. Given a discrete bounded geographic domain, rank the aggregate suitability of each 4 ACS Paragon Plus Environment
Page 4 of 34
Page 5 of 34
94
Environmental Science & Technology
site within that domain for the purpose of constructing a surface spreading basin.
95
2. Given a discrete bounded geographic domain, propose a connected corridor of sites
96
within that domain that minimally traverses the space between the site of an existing
97
WWTP and a proposed surface spreading basin relative to one or more continuously
98
defined measures of cost/distance.
99
3. Given a valid corridor specification, evaluate the water withdrawals/consumption re-
100
quired to generate the quantity of electricity that is consumed for the purpose of moving
101
a fixed volume of water between the two endpoints of the corridor at a rate specified
102
by the permitted maximum flow capacity of the existing WWTP.
103
Locating Suitable Sites for Artificial Groundwater Recharge Basins
104
Our approach to Challenge-1 is based upon a well established cartographic modeling tech-
105
nique known as Weighted Overlay Analysis [WOA]. 48–51 In a WOA, multiple separate raster
106
layers, each corresponding to multiple independent components of aggregate site suitability
107
for a given landuse application, are separately derived from raw spatial data inputs. 52 In
108
each of these layers, every raster grid cell is encoded with an ordinal ranking corresponding
109
to its component suitability for the landuse activity in question with no weighting applied. 52
110
Once these various layers have been appropriately defined, they are then be combined, on a
111
cell for cell basis, to generate a uniform representation of aggregate site suitability. 52
112
The implementation of any WOA begins by first defining a discrete area to function as
113
the spatial unit for the analysis. For the purposes of this investigation, this spatial unit was
114
defined as the United States Geologic Survey [USGS] hydrologic unit code system’s [HUC]
115
level-4 sub-basin area containing the source WWTP of interest. 53 Within California, on
116
average, each of these level-4 sub-basins comprises a closed contiguous area of roughly 1,800
117
square kilometers. 53 For our investigation, these spatial units were discretized into raster
118
grids with a cell resolution of 100 by 100 meters. This resolution was carried out throughout
5 ACS Paragon Plus Environment
Environmental Science & Technology
119
all of the other spatial modeling activities that will be subsequently discussed.
120
As mentioned previously, the WOA approach involves parsing the complex concept of
121
aggregate site suitability into its individual component parts such that is possible to define
122
relative suitability rankings from available spatial data layers. In our WOA implementation,
123
we decomposed the composite suitability for the construction of a surface spreading basin
124
into the following three suitability components: a slope based ranking, an expected hydraulic
125
conductivity based ranking, and an existing landuse based ranking. 46,54–57
126
The raw input geospatial source data that was used to develop the slope based site
127
suitability ranking was the 30 meter National Elevation Dataset [NED] hosted by the USGS
128
through the National Map Server. 58 From this raw digital elevation model an output slope
129
layer was computed using a standard eight-neighbor kernel technique. This intermediate
130
slope layer, consisting of percentage slopes ranging from 0-100%, was then processed into
131
a layer of ordinal numerical rankings, ranging from 0-10, using a histogram equalization
132
procedure. Through this procedure the counts of actual slope percentages in the input
133
layer were transformed such that the frequency of 1’s, 2’s, 3’s, etc. in the output layer
134
conformed to a normal distribution. This procedure has the net effect of producing a non-
135
linear transformation of slope percentages to slope rankings; penalizing the ranking of steeper
136
slopes more greatly than shallow ones.
137
The raw input geospatial source data used to develop the hydraulic conductivity based
138
ranking was the STATSGO2/SURGO surface soil characteristics dataset hosted by the USDA
139
through the National Geospatial Data Gateway. The raw STATSGO2/SURGO dataset exists
140
as a set of vector geometries linked to an underlying database of attribute fields documenting
141
key physical characteristics of various different soil types. The specific attribute used to
142
derived the hydraulic conductivity ranking used in our WOA approach was the hydraulic
143
group categorization. This categorization scheme divides soils into seven groups on the basis
144
of their infiltration rate and overall drainage characteristics. First this input vector layer is
145
converted to a raster output of desired resolution. Then the rasterized version is transformed
6 ACS Paragon Plus Environment
Page 6 of 34
Page 7 of 34
Environmental Science & Technology
146
into the same ordinal ranking scheme, ranging from 0-10 as that used for the other suitability
147
criteria using a transformation scheme which preferentially ranks soils with higher degrees
148
of hydraulic conductivity as being more suitable. For more details on both the hydraulic
149
group categorization scheme as well as the exact transformations used please consult the SI.
150
The raw input geospatial source data used to develop the landuse base ranking was
151
the National Landuse Landcover Dataset hosted by the USGS through the National Map
152
Server. The NLCD encodes each pixels using one of twenty predefined landuse-landcover
153
classifications. First the NLCD raster was clipped and resampled to the desired output
154
spatial resolution, only this time, using a nearest neighbor sampling technique to preserve
155
the integrity of the classification scheme. Next the raster classification scheme used to encode
156
the data was converted into the same ordinal, 1-10 ranking scheme used elsewhere in the
157
analysis using a transformation scheme which designated more highly developed landscapes
158
and those which would present an intrinsic engineering challenge to the implementation of
159
an infiltration basin as being fundamentally less suitable than undeveloped open land. Here
160
again, for more details on the exact transformations used please consult the SI.
161
It is conceivable that there might be other additional dimensions of suitability that might
162
be relevant to the evaluation of aggregate site suitability for this particular landuse. And
163
indeed, various other data sources, in various different combinations, have been proposed in
164
previous studies with similar goals. 49–51,59,60 However, it was only by limiting the scope of
165
the required spatial data inputs to these three components that we were able to collect and
166
process raw data inputs covering the breadth desired geographic domain (the entire state of
167
California) at the desired spatial resolution (100 x 100 meters).
168
The additive combination of these three component measures of overall site suitability
169
results in the output of a single aggregate suitability layer. In order to select a single site
170
as the choice of the surface spreading basin from within this layer, the following additional
171
processing steps are taken. First the aggregate suitability ranking layer is converted to a
172
binary format by applying a minimum aggregate suitability ranking threshold. Once this
7 ACS Paragon Plus Environment
Environmental Science & Technology
173
thresholded layer is produced, the closed connected regions of contiguously high aggregate
174
site suitability are ranked relative to one another on the basis of their size, with larger
175
areas being considered most favorable since they can store more water. Finally, the single
176
largest area within this set is chosen and its corresponding centroid cell computed as the final
177
destination site for the location of the surface spreading basin. The size of the final selected
178
suitable regions used in the five case studies were: Santa Barbara: 0.75 km2 , Oxnard: 3.63
179
km2 , Fresno & Tulare: 19.07 km2 , Santa Ana & San Bernadino: 15.90 km2 , San Diego 3.95
180
km2 .
181
Figure 1 provides a graphical illustration of our WOA analysis based site suitability mod-
182
eling framework for a sample case study region encompassing the Southern Californian cities
183
of San Bernadino and Santa Ana. Figure 1 elements 1 - 2 - 3 depict the ordinal rank-
184
ings for the slope, permeability, and existing landuse suitability components, respectively.
185
Element 4 depicts the intermediate output aggregate suitability layer. Element 5 depicts
186
the binary encoded layer mask depicting connected areas of high aggregate suitability. And
187
finally, element 6 shows the final selected destination site for the surface spreading basin,
188
relative to that of the source WWTP (magnified for visual clarity). The sources of the
189
raw data inputs used for the WOA site suitability model are provided in the supplementary
190
information [SI].
191
Locating Optimal Corridors for Water Distribution Infrastructure
192
Once a single site has been proposed as the location for the destination surface spreading
193
basin, the process of addressing challenge Challenge-2, involving the search for an optimal
194
corridor for the water distribution pipeline that will connect it to the site of the source
195
WWTP, can begin. The naive solution to this problem would be to simply select the raster
196
grid cells which fall along the Euclidean shortest path linking the source and the destina-
197
tion. Such an approach however, would ignore the possibility there might be many different
198
types of cost/distance relationships associated with the location of a pipeline corridor within 8 ACS Paragon Plus Environment
Page 8 of 34
Page 9 of 34
Environmental Science & Technology
Figure 1: Graphical illustration of the weighted overlay analysis based site suitability modeling framework implemented for a case study region encompassing the Southern California cities of San Bernadino and Santa Ana. Figure elements include: slope score (1), hydraulic conductivity score (2), existing landuse score (3), composite suitability score (4), binary mask (5), reference source with selected destination (6)
9 ACS Paragon Plus Environment
Environmental Science & Technology
199
the spatial domain and the likelihood that these costs each be continuously non-uniformly
200
distributed within the domain. 56,61,62
201
The canonical name for this problem is the multi-objective corridor location [MOCL]
202
problem. 63 It exists as a sub-set of the generic category of graph theoretical shortest path
203
[SP] problems and has been shown to be NP-hard in its construction; meaning that there
204
are no known deterministic algorithms capable of generating optimal solutions in a runtime
205
that is some polynomial function of problem size. 64 As such, the development of effective
206
heuristic approaches to its solution is an important and practically useful topic of research
207
within the areas of combinatoric optimization and operations research.
208
The combination of the broad extent and the high resolution of the raster grids being
209
used for this analysis translated into the proposition of MOCL problem statements whose
210
search domains comprised on the order of n = 106 to 107 nodes. As it turns out, the
211
number of unique self-avoiding corridors which must be evaluated to deterministically solve
212
the MOCL is thought to scale as 2.638n . 65 Clearly then, in the context of such vast problem
213
specifications, the use of conventional deterministic approaches would be computationally
214
infeasible.
215
In order to broach this issue, a novel parallel implementation of an existing genetic
216
algorithm for the MOCL problem was developed which enabled us to leverage modern multi-
217
core processing hardware to generate near optimal solutions with a reasonable amount of
218
time/effort. 63 This algorithm was deployed in the cloud using the Google Compute Engine
219
platform to allow for simultaneous and continuous solution runs. An in depth discussion of
220
the mechanics of this algorithm as well as benchmark analysis of its runtime performance and
221
output solution quality can be found in the SI. The Go programming language source code
222
for this new implementation of Zhang et al.’s algorithm has been made freely available by the
223
authors at the following host repository: https://github.com/ericdfournier/corridor.
224
Each MOCL problem specification requires three different cost surfaces: a slope based
225
cost, an existing landuse disturbance based cost, and an accessibility based cost. Each
10 ACS Paragon Plus Environment
Page 10 of 34
Page 11 of 34
Environmental Science & Technology
226
surface was generated using an ordinal scale with values ranging from 1 - 10. The slope
227
cost surface imposes higher cost values on areas with steeper slopes. This incentivizes the
228
minimal accumulation of slope along the corridor’s length, to minimize the energy required
229
to deliver reclaimed water up gradient. The existing landuse disturbance cost layer imposes
230
higher cost values on areas with sensitive or intensive existing land uses. This incentivizes
231
the avoidance of areas that might be significantly negatively impacted by the introduction
232
of a major new water distribution pipeline. Finally, the accessibility cost imposes higher
233
costs on areas that are located away from access roads and throughways. This incentivizes
234
the routing of corridors along roads and highways that would make the construction and
235
maintenance of a pipeline easier and more cost effective for its managing agency.
236
The slope based cost surface was derived from the same NED source dataset used to gen-
237
erate the slope based site suitability score used in the WOA model. Similarly, the landuse
238
disturbance based cost surface was derived from the same NLCD dataset used to produce
239
the existing landuse disturbance score used for the WOA model; albeit using a modified
240
transformation scheme. The accessibility based cost surface was derived from the U.S. Cen-
241
sus Bureau’s TIGERTM road network dataset using the follow cost transformation scheme:
242
Major Roads = 10, Streets = 5, and Off-Road = 1.
243
Interestingly, the accessibility and disturbance objectives were observed to be fairly
244
strongly inversely spatially correlated with one another within the case study regions in-
245
vestigated. The computed correlation coefficients (ρ) between the two layers for each case
246
study site were: Fresno & Tulare: -0.39, Oxnard: -0.35, Santa Barbara: -0.41, Santa Ana
247
& San Bernadino: -0.47, San Diego: -0.52 [p-values = 0.0000]. This feature challenges the
248
corridor selection algorithm to make explicit tradeoffs between the two during its search for
249
the best possible corridor.
250
Figure 2 provides a graphical illustration of the data inputs and a set of output corridor
251
solutions generated by the genetic algorithm used in the MOCL problem specification for
252
the same sample case study region comprising the cities of Santa Ana and San Bernadino.
11 ACS Paragon Plus Environment
Environmental Science & Technology
253
Figure 2 elements 1 - 2 - 3 depict the three input cost surfaces for slope, existing
254
landuse disturbance, and accessibility objectives, respectively. Element 4 depicts a set
255
of corridor solutions generated as output for a single run of the genetic algorithm. These
256
corridor solutions link the endpoint WWTP source to the surface spreading basin destination
257
referenced previously in Figure 1.
258
Estimating Reuse System Pump Energy Requirements
259
After a valid corridor has been generated linking the site of the existing WWTP to that of the
260
proposed site for the surface spreading basin, the process of addressing Challenge-3, involving
261
the evaluation of the proposed system’s net energy-water usage efficiency, can finally proceed.
262
One of the primary benefits associated with investing all of the effort required to generate
263
near optimal corridors is that the resulting detailed route information can then be used to
264
derive estimates for the system’s pump energy requirements using the well established civil
265
engineering computational techniques. 66
266
These techniques begin with assumptions about the basic characteristics regarding the
267
pipeline’s physical structure such as pipe section length, cross sectional area, materials, etc.
268
For the purposes of this analysis the pipelines were modeled as being constructed of standard
269
concrete pipe sections with a cross sectional area that was sufficiently large so as to ensure
270
that the maximum flow speed through the pipeline did not exceed a maximum velocity of 7
271
m/s. The number and location of fitting elements modeled as part of each pipeline system
272
were derived from each case study’s detailed pipeline corridor specification. All other relevant
273
assumptions as well as a detailed overview of the computation procedure itself are provided
274
in the SI.
275
Estimating Reuse System Process Energy Related Water Demands
276
The corresponding water required to generate the energy needed to pump the water (P )
277
can be estimated by referencing the technology composition of the local grid mix to liter12 ACS Paragon Plus Environment
Page 12 of 34
Page 13 of 34
Environmental Science & Technology
Figure 2: Graphical illustration of a multi-objective corridor location problem specification with input cost surfaces and output corridor solutions for the case study region comprising the cities of San Bernadino and Santa Ana. Figure elements include: Slope cost (1), disturbance cost (2), accessibility cost (3), solution corridors (4)
13 ACS Paragon Plus Environment
Environmental Science & Technology
278
ature values for the water intensity of various electricity generation technologies. For the
279
purposes of this analysis, the composition of the local average grid mix in each case study
280
region was taken from values published in the UC Berkeley Water for Energy Sustainability
281
Tool [WWEST] web based life cycle inventory modeling tool. 45,67–70 Based upon the relative
282
geographic location of the case study sites, three different grid mix values, corresponding to
283
three separate utility providers operating within the state of California, were ultimately used.
284
The water intensity of electricity generation for each of the different production technologies
285
were taken from a National Renewable Energy Laboratory [NREL] review of the operation
286
water consumption and withdrawal factors for electricity generating technologies. 44 The val-
287
ues contained in this study were selected because they reflect the most recent consensus
288
based assessment of a dynamic and complicated field, one which is the focus of continued
289
active research. 71–75
290
Results and Discussion
291
Case Study Corridor Solutions
292
We investigated the net energy-water usage characteristics of five hypothetical new reuse
293
systems, assuming 100% reuse of the existing WWTP’s National Pollutant Discharge Elim-
294
ination System [NPDES] permitted maximum daily flow capacity; an assumption which
295
corresponds to the worst case scenario in terms of system scale. Five hydrologically distinct
296
regions distributed throughout Central and Southern California were selected, each of which
297
is depicted graphically in the map set at the top portion of Figure 3. These five regions com-
298
prise the level 4 HUC sub-basin zones containing the following cities (each referenced to their
299
corresponding figure elements): Santa Barbara 1 , Ventura & Oxnard 2 , Fresno & Tulare
300
3 , Santa Ana & San Bernadino 4 , and San Diego 5 . These locations were selected
301
on the basis of the dissimilarity of their local geographic contexts (i.e. endpoint separation
302
distance, regional topography, region landuse, etc.) as well as the demonstrated interest of 14 ACS Paragon Plus Environment
Page 14 of 34
Page 15 of 34
Environmental Science & Technology
303
their corresponding water management authorities in the reuse of treated wastewater for the
304
artificial recharge applications.
305
At the lower portion of Figure 3 is a plot illustrating the elevation profiles associated
306
with the corridor solutions generated by the genetic algorithm for each of the five case study
307
regions. The color of each plot series corresponds to the color of each of the case study
308
regions depicted in the map panel. The vertical axis references the elevation relief which
309
occurs along the length of each proposed corridor in meters. The horizontal axis references
310
the along-path distance of each corridor in kilometers.
311
In four of the five case study regions, corresponding to Figure 3 elements 1 - 2 -
312
4 - 5 , the slope of the along-corridor elevation profile is net positive. This is because,
313
each of these HUC zones are coastal and their corresponding WWTP are situated at the
314
lowest elevation point in the basin to take advantage of gravity in the collection of waste
315
sewage. In the HUC sub basin containing Fresno & Tulare however, element 3 , the slope
316
of the along-corridor elevation profile is net negative. This is because this HUC sub basin is
317
landlocked (i.e. internally drained) and the WWTP is not positioned at the lowest elevation
318
point within the basin.
319
As the elevation profile plot illustrates the length of each corridor and thus the distance
320
that the water must be pumped depends in part upon the absolute area of the sub-basin
321
and in part upon the relative position of the source WWTP relative to the site selected for
322
the surface spreading basin. The Ventura & Oxnard case study region, plotted in orange,
323
comprises a relatively large land area. However, the site deemed to be most suitable for the
324
location of a surface spreading based was positioned fairly close to the site of the WWTP.
325
Therefore, the total along-corridor distance associated with the Ventura & Oxnard case study
326
region is shorter than that for the San Diego region which, in terms of area, is comparatively
327
larger in size.
328
Another interesting observation which can be made from this figure is the relative rough-
329
ness (i.e. the second derivative of slope) associated with each of the proposed solution corri-
15 ACS Paragon Plus Environment
Environmental Science & Technology
Figure 3: Overview map depiction of five case study regions plotted relative to the alongcorridor elevation profiles associated with each region’s corresponding output solution corridor. Santa Barbara (1), Oxnard & Ventura (2), Fresno & Tulare (3), Santa Ana & San Bernadino (4), San Diego (5)
16 ACS Paragon Plus Environment
Page 16 of 34
Page 17 of 34
Environmental Science & Technology
330
dors. For example, the elevation profile for the Santa Barbara case study region’s corridor is
331
characterized by a high degree of roughness. The explanation for this feature has to do with
332
the structure of the different objective variables within the search domain area. In Santa
333
Barbara for example, the WWTP is separated from the proposed surface spreading basin
334
site by a large amount of high value land-uses. These land-uses therefore correspond to high
335
disturbance costs which must be weighed relative to the slope based costs during the multi-
336
objective optimization process. Thus, the high degree of roughness in the Santa Barbara
337
watershed along the corridor elevation profile indicates the algorithm trades off a smoother
338
elevation profile for one that is characterized by a lower degree of landuse disturbance and/or
339
a higher degree of accessibility.
340
Case Study Energy-Water Usage Efficiency Estimates
341
The electric utility provider which services this region is Southern California Edison [SCE].
342
The portfolio of electric generation technologies in use by this utility provider in 2011 is
343
depicted graphically by the pie chart in the upper portion of Figure 4. As the plot shows,
344
the dominant production technologies are Natural Gas (48%), Coal (12%), Nuclear (19%),
345
Hydro (9%), and Geothermal (9%); with the remainder (∼4%) comprised of a mixture of
346
renewable generation technologies. 70
347
In terms of the water intensity of electricity production it has been observed that gener-
348
ation technologies which require process water for cooling tend to have higher water with-
349
drawal and consumption factors than for renewable processes. The water intensity, measured
350
in terms of water withdrawals and consumption, for SCE’s grid mix are described graphi-
351
cally by the combination of plots contained in Figure 4 using data from the NREL literature
352
review. The absence of a plot series for any column in either these figures indicates an es-
353
timated consumption/withdrawal factor of 0 m3 /M W h. For the purposes of this analysis,
354
mean factor values (horizontal red bars) were used to compute the net energy-water usage
355
efficiencies for proposed reuse systems in each of the five case study regions. 17 ACS Paragon Plus Environment
Environmental Science & Technology
Figure 4: Breakdown of the production technology mix in use by the local electric utility provider: Southern California Edison plotted relative to the ranges of water consumption and withdrawal factors associated with each production technology category
18 ACS Paragon Plus Environment
Page 18 of 34
Page 19 of 34
Environmental Science & Technology
356
Figure 5 plots the estimated ratios of energy derived water expenditures, measured in
357
terms of both withdrawals and consumption, to the volume of water recovered for proposed
358
reuse systems in each of the five case study regions. The horizontal red line depicts the
359
net water savings threshold for a proposed reuse system; defined as the point at which the
360
volume of water recovered is equal to the volume of water withdrawn/consumed to supply
361
the process energy for the reuse activities. The range of ratios is determined by propagating
362
the uncertainty regarding the power efficiency of the pump systems in use for each system
363
(50% - 90%).
364
Of the five case study regions investigated only one reuse system, located in the inland
365
sub-basin containing the cities of Fresno and Tulare, was found to provide an estimated net
366
water savings when the embedded water consumption associated with the systems’ process
367
energy were taken into account. This result can be explained, in large part, by the unique
368
spatial orientation of the pipeline corridor endpoints in this case study region. For, in this
369
region alone, the elevation of the proposed site for the surface spreading was lower than that
370
of the WWTP; a feature which allowed the water being sent along the pipeline corridor to
371
be modeled as flowing downhill under the force of gravity, thus requiring only minimal pump
372
energy to overcome frictional losses.
373
In two of the five case study regions, those containing the cities of San Bernadino & Santa
374
Ana and San Diego, the water consumption requirements associated with each systems es-
375
timated pump energy demands were sufficiently high to negate the water savings associated
376
with the reuse activities nearly ten times over. The range of water withdrawals is even worse,
377
exceeding the volume of water recovered by reuse in these locations by nearly two orders of
378
magnitude. The explanation for the large size of these values again stems from the charac-
379
teristics of the proposed pipeline corridor at the case study sites: with each exhibiting long
380
along-corridor distances and large net elevation gradients between the corridor endpoints.
19 ACS Paragon Plus Environment
Environmental Science & Technology
Figure 5: Ranges for the ratios of energy derived water expenditures, measured in terms of both withdrawals and consumption, to water recoveries for proposed reuse systems in each of the five case study regions
381
Model Component Validation
382
A model validation exercise was undertaken to evaluate the effectiveness of the first two
383
components of our modeling approach - the WOA site suitability model and the corridor
384
location model. The researchers were unable to validate the method for estimating the
385
net energy-water usage efficiencies of the proposed reuse systems due to the lack of public
386
availability of real world facility operational data for existing systems. For both components
387
of this validation exercise, publicly available information about the structural components
388
of an existing reuse system with the Santa Ana & San Bernadino case study basin were used
389
for reference. The first validation component involved an investigation of the quality of the
390
WOA model predicted destination site selections for the location of an artificial groundwater
391
recharge infiltration basin.
392
Figure 6 depicts the results of this first model component validation exercise. At the
393
top of the figure a map of the composite site suitability scores generated from the WOA
20 ACS Paragon Plus Environment
Page 20 of 34
Page 21 of 34
Environmental Science & Technology
Figure 6: Weighted Overlay Analysis model component validation results. The map plotted at top shows the relative location - with point to point euclidean distances - of the reference source, the predicted destination, and the reference destination for the siting of a groundwater infiltration basin within the Santa Ana & San Bernadino case study region. The grayscale shading of the mapped area depicts the composite site suitability scores (0-30) for the entire case study region. Likewise, the histogram plot at bottom depicts the distribution of these same composite site suitability scores graphically.
21 ACS Paragon Plus Environment
Environmental Science & Technology
394
model is shown overlaid with a set of three connected points plotting the reference source,
395
the reference destination, and the predicted destination. The euclidean distances separating
396
each of these three points of interest are given. Additionally, at the bottom of Figure 6 is a
397
histogram plot showing the distribution of the composite site suitability scores for the case
398
study basin. The red and blue vertical lines on this plot depict the composite site suitability
399
scores for the reference destination and the predicted destination, respectively. As this figure
400
shows, according to the suitability ranking criteria introduced in our WOA approach, the
401
predicted destination site score is significantly higher than of the reference destination.
402
The second validation component involved an investigation of the quality of the corridor
403
location model solutions relative the location of an existing pipeline corridor supplying an
404
artificial groundwater recharge operation within the Santa Ana & San Bernadino case study
405
region. The detailed specification of this corridor was obtained by the researchers through a
406
manual digitization process which involve the use of high resolution satellite imagery capable
407
of resolving individual pipeline components.
408
Figure 7 depicts the results of the second model component validation exercise. At the
409
top portion of Figure 7 are a set of map panels depict the location of the top 100 output
410
corridor solutions generated by the model (shaded in blue) relative to the reference corridor
411
(in red). The plot at center graphically depicts the various objective scores computed for each
412
of these top 100 solution corridors (solid lines) relative to the objective scores computed for
413
the reference corridor (broken lines). As this plot shows, while the slope scores are roughly
414
the same, the solution corridors tend to have lower accessibility scores but higher disturbance
415
scores than the reference corridor. The plot at bottom of Figure 7 shows a histogram of the
416
total aggregate objective scores for the top 100 solution corridors generated by the model.
417
The blue and red vertical lines illustrate the finding that the top solution corridor exhibits
418
a 3% improvement over the reference corridor, as measured in terms of total aggregate
419
objective score.
420
As mentioned previously, the authors were unable to perform a similarly detailed valida-
22 ACS Paragon Plus Environment
Page 22 of 34
Page 23 of 34
Environmental Science & Technology
Figure 7: Corridor location model component validation results. The map panels plotted at top show the location of the top 100 output solutions generated by the corridor location algorithm relative to that of the reference corridor. The plot at center depicts the objective scores for each corridor across each of the three separate objectives. These can be compared to the scores for the reference corridor plotted as broken horizontal lines. The plot at bottom depicts the distribution of total aggregate objective scores for the top 100 solution corridor generated.
23 ACS Paragon Plus Environment
Environmental Science & Technology
421
tion of their predictions for the net energy-water usage efficiency of the existing reuse facility
422
in the Santa Ana & San Bernadino case study basin. This inability was due to the lack of
423
publicly available data for that facility’s operations. Were this facility were to be structured
424
in accordance with the assumptions used throughout this analysis the authors would antic-
425
ipate, generally, for it to be operating as a new water consumer due to the high predicted
426
pump energy demands associated with the physical layout of its structural components.
427
Caveats and Considerations
428
We believe that both the modeling framework and the case studies’ results presented in con-
429
junction with this research significantly advance our understanding of the net environmental
430
performance of water reuse systems involving artificial groundwater recharge components.
431
Specifically, they underscore the importance of the linkages between local geographic con-
432
text, pump energy consumption, and the net water usage efficiencies of water reuse systems.
433
While there are likely important benefits to the watershed in which water is reused, the
434
following issues must be better understood.
435
Firstly, within the state of California a non-trivial number of the power plants are strate-
436
gically positioned along the coast to take advantage of the use of available seawater to supply
437
the water intensive process of cooling. In the case of these facilities, the net energy-water
438
usage efficiency of reuse activities can and likely should be interpreted differently from sit-
439
uations in which the water being used for cooling is being sourced from more conventional
440
freshwater surface flows or groundwater reservoirs. Indeed, in these unique circumstances,
441
it may be more appropriate to consider the net effects of large scale reuse activities as being
442
akin to a form of virtual ocean desalination.
443
Another potential concern is that there are a number of modeling assumptions inherent to
444
this analysis which stem from the core assumption that both the surface spreading basin and
445
the underlying aquifer which it feeds would be capable of supporting both a rate and a total
446
volume of recharge commensurate with 100% reuse of the WWTP’s maximum permitted 24 ACS Paragon Plus Environment
Page 24 of 34
Page 25 of 34
Environmental Science & Technology
447
throughput capacity. On the one hand, it is entirely possible that the existing WWTP
448
may regularly operate at a level below this maximum permitted capacity, thus reducing the
449
total volume destined to be recharged. On the other hand, it is also possible the either the
450
proposed site of the surface spreading basin and/or the underlying aquifer may not be able to
451
accommodate either the designated rate of recharge or the total recharge volume. If this were
452
to be the case, the system must be either scaled back or decomposed into multiple separate
453
recharge sites. In any case, addressing these issues would require highly context specific
454
information regarding the daily treatment flow of the existing WWTP, the precise scale and
455
infiltration rate of the proposed new surface spreading basin, the hydraulic storage capacity
456
of the adjacent subsurface aquifer, and the nature of the vertical connectivity between the
457
infiltration basin and the subsurface aquifer.
458
Finally, it is clear from this analysis that the large scale reuse of treated wastewater for
459
artificial groundwater recharge would impose significant new electricity demands on the re-
460
gional power grid. For the sake of simplicity, our framework assessed the water requirements
461
associated with this marginal electricity demand using average grid mix data. Given recent
462
trends towards the decarbonization of electricity production technologies and the ongoing
463
push for the adoption of more renewable energy technologies, it is also possible that the wa-
464
ter intensity of the marginal grid mix may be significantly different/lower than the average.
465
Such a situation, while difficult to accurately forecast, would lead to reduced evaluations
466
for the water intensity of these types reuse systems, thus making them appear to provide
467
greater net water savings.
468
Acknowledgement
469
The authors thank the Walton Family Foundation Sustainable Water Markets (SWM) Fel-
470
lowship program for its generous financial support of this project.
25 ACS Paragon Plus Environment
Environmental Science & Technology
471
References
472
(1) Aghakouchak, A.; Feldman, D.; Stewardson, M. J.; Saphores, J.-D.; Grant, S.;
473
Sanders, B. Australia’s drought: lessons for California. Science 2014, 343, 1430–1431.
474
(2) Diffenbaugh, N. S.; Swain, D. L.; Touma, D. Anthropogenic warming has increased
475
drought risk in California. Proceedings of the National Academy of Sciences 2015, 112,
476
3931–3936.
477
(3) Lo´aiciga, H. a. Managing Municipal Water Supply and Use in Water-Starved Regions:
478
Looking Ahead. Journal of Water Resources Planning and Management 2015, 141,
479
01814003.
480
481
(4) Stoutenborough, J. W.; Vedlitz, A. Public Attitudes Toward Water Management and Drought in the United States. Water Resources Management 2014, 28, 697–714.
482
(5) Swain, D. L.; Tsiang, M.; Haugen, M.; Singh, D.; Charland, A.; Rajaratnam, B.; Diff-
483
enbaugh, N. S. The extraordinary California drought of 2013-2014: Character, context,
484
and the role of climate change. Bull Am Meteorol Soc 2014, 95, S3–S7.
485
(6) Asano, T.; Burton, F.; Leverenz, H.; Tsuchihashi, R.; Tchobanoglous, G. Water Reuse:
486
Issues, Technologies, and Applications, 1st ed.; McGraw-Hill Professional: Boston, MA,
487
2007; p 1570.
488
(7) USEPA, 2012 Guidelines for Water Reuse; 2012; pp 1–643.
489
(8) California Department of Water Resources Recycled Water Task Force, Water Recycling
490
2030 ; 2003; p 300.
491
(9) Bixio, D.; Thoeye, C.; Wintgens, T.; Ravazzini, a.; Miska, V.; Muston, M.; Chikurel, H.;
492
Aharoni, a.; Joksimovic, D.; Melin, T. Water reclamation and reuse: implementation
493
and management issues. Desalination 2008, 218, 13–23.
26 ACS Paragon Plus Environment
Page 26 of 34
Page 27 of 34
Environmental Science & Technology
494
(10) Lautze, J.; Stander, E.; Drechsel, P.; da Silva, A. K.; Keraita, B. Global experiences
495
in water reuse; International Water Management Institute (IWMI); CGIAR Research
496
Program on Water, Land and Ecosystems (WLE), 2014.
497
498
499
500
501
502
(11) Dolniˇcar, S.; Saunders, C. Recycled water for consumer markets a marketing research review and agenda. Desalination 2006, 187, 203–214. (12) Dolniar, S.; Sch¨afer, A. I. Desalinated versus recycled water: public perceptions and profiles of the accepters. Journal of environmental management 2009, 90, 888–900. (13) Dolnicar, S.; Hurlimann, A.; Gr¨ un, B. What affects public acceptance of recycled and desalinated water? Water research 2011, 45, 933–43.
503
(14) Harris-Lovett, S. R.; Binz, C.; Sedlak, D. L.; Kiparsky, M.; Truffer, B. Beyond User
504
Acceptance: A Legitimacy Framework for Potable Water Reuse in California. Environ-
505
mental Science & Technology 2015, 49, 7552–7561.
506
(15) Hurlimann, A.; Dolnicar, S.; Meyer, P. Understanding behaviour to inform water supply
507
management in developed nations–a review of literature, conceptual model and research
508
agenda. Journal of environmental management 2009, 91, 47–56.
509
510
511
512
513
514
515
516
(16) Marks, J.; Martin, B.; Zadoroznyj, M. How Australians order acceptance of recycled water: National baseline data. Journal of Sociology 2008, 44, 83–99. (17) Po, M.; Kaercher, J. D.; Nancarrow, B. E. Literature Review of Factors Influencing Public Perceptions of Water Reuse; 2003; pp 1–44. (18) Bridgeman, J. Public Perception Towards Water Recycling in California. Water and Environment Journal 2004, 18, 150–154. (19) Rodrigo, D.; Calva, E. J. L.; Cannan, A.; Roesner, L.; O’Conner, T. P. Total Water Management; 2012; pp 1–131.
27 ACS Paragon Plus Environment
Environmental Science & Technology
517
518
519
520
521
522
(20) NRC, Water Reuse: Potential for Expanding the Nation’s Water Supply Through Reuse of Municipal Wastewater ; National Academies Press, 2012. (21) Freeman, G.; Poghosyan, M.; Lee, M. Where Will We Get the Water? Assessing Southern Californias Future Water Strategies; 2008; pp 1–37. (22) WateReuse, National Database of Water Reuse Facilities Summary Report; 2008; p 121.
523
(23) Langridge, R.; Fisher, A.; Racz, A.; Daniels, B.; Rudestam, K.; Hihara, B. Climate
524
Change and Water Supply Security: Reconfiguring Groundwater Management to Reduce
525
Drought Vulnerability; 2012; pp 1–82.
526
527
528
529
(24) Levine, A. D.; Asano, T. Recovering Sustainable Water from Wastewater. Environmental science & technology 2004, 38, 201A–208A. (25) NRC, Prospects for Managed Underground Storage of Recoverable Water ; National Academies Press, 2008; pp 1–350.
530
(26) Asano, T. Artificial recharge of groundwater ; Elsevier, 2013; pp 1–748.
531
(27) Bouwer, H. In Hydraulic Design Handbook, 1st ed.; Mays, L. W., Ed.; McGraw-Hill
532
533
534
535
536
Professional: New York, USA, 1999; Chapter 24, pp 1–44. (28) Anderson, J. The environmental benefits of water recycling and reuse. Water Science and Technology 2000, 3, 1–10. (29) Toze, S. Reuse of effluent water - Benefits and risks. Agricultural Water Management 2006, 80, 147–159.
537
(30) Hamilton, A. J.; Versace, V. L.; Stagnitti, F.; Li, P.; Yin, W. E. I.; Hermon, K.;
538
Premier, R. R.; Ierodiaconou, D. Balancing Environmental Impacts and Benefits of
539
Wastewater Reuse. WSEAS Transactions on Environment and Development 2006, 2,
540
117–129. 28 ACS Paragon Plus Environment
Page 28 of 34
Page 29 of 34
541
542
Environmental Science & Technology
(31) Wilson, M. A.; Carpenter, S. R. Economic valuation of freshwater ecosystem services in the United States: 1971-1997. Ecological applications 1999, 9, 772–783.
543
(32) Kennedy, L. A.; Tsuchihashi, R. Is Water Reuse Sustainable? Factors Affecting its
544
Sustainability. The Arabian Journal for Science and Engineering 2005, 30, 3–15.
545
(33) Sala, L.; Serra, M. Towards sustainability in water recycling. Water Science and Tech-
546
547
548
549
550
551
552
553
554
555
556
nology 2004, 50, 1–8. (34) Dillon, P. Future management of aquifer recharge. Hydrogeology journal 2005, 13, 313– 316. (35) Chen, Z.; Ngo, H. H.; Guo, W. A critical review on sustainability assessment of recycled water schemes. Science of the total environment 2012, 426, 13–31. (36) McMahon, J. E.; Price, S. K. Water and Energy Interactions. Annual Review of Environment and Resources 2011, 36, 163–191. (37) Pate, R.; Hightower, M.; Cameron, C.; Einfeld, W. Overview of Energy-Water Interdependencies and the Emerging Energy Demands on Water Resources; 2007; pp 1–20. (38) Klein, G.; Krebs, M.; Hall, V.; O’Brien, T.; Blevins, B. B. California’s Water Energy Relationship. Energy 2005, 1–180.
557
(39) Lofman, D.; Petersen, M.; Bower, A. Water, Energy and Environment Nexus: The
558
California Experience. International Journal of Water Resources Development 2010,
559
18, 73–85.
560
(40) Weber, C. L.; Jaramillo, P.; Marriott, J.; Samaras, C. Life cycle assessment and grid
561
electricity: What do we know and what can we know? Environmental Science and
562
Technology 2010, 44, 1895–1901.
563
564
(41) G´omez-Exp´osito, A.; Conejo, A. J.; Ca˜ nizares, C. Electric energy systems: analysis and operation; CRC Press, 2008; pp 1–664. 29 ACS Paragon Plus Environment
Environmental Science & Technology
565
(42) Feeley, T. J.; Skone, T. J.; Stiegel, G. J.; McNemar, A.; Nemeth, M.; Schimmoller, B.;
566
Murphy, J. T.; Manfredo, L. Water: A critical resource in the thermoelectric power
567
industry. Energy 2008, 33, 1–11.
568
(43) Macknick, J.; Newmark, R.; Heath, G.; Hallett, K. C. A Review of Operational Water
569
Consumption and Withdrawal Factors for Electricity Generating Technologies; 2011;
570
pp 1–29.
571
(44) Averyt, K.; Macknick, J.; Rogers, J.; Madden, N.; Fisher, J.; Meldrum, J.; Newmark, R.
572
Water use for electricity in the United States: an analysis of reported and calculated
573
water use information for 2008. Environmental Research Letters 2013, 8, 015001.
574
(45) Horvath, A. Life Cycle Energy Assessment of Alternative Water Supply Systems in
575
California; 2005; pp 1–174.
576
(46) Aissi, H.; Chakhar, S.; Mousseau, V. GIS-based multicriteria evaluation approach for
577
corridor siting. Environment and Planning B: Planning and Design 2012, 39, 287–307.
578
(47) Ma, J.; Hoekstra, A. Y.; Wang, H.; Chapagain, A. K.; Wang, D. Virtual versus real wa-
579
ter transfers within China. Philosophical Transactions of the Royal Society B: Biological
580
Sciences 2006, 361, 835–842.
581
582
(48) Malczewski, J. GIS-based land-use suitability analysis: a critical overview. Progress in Planning 2004, 62, 3–65.
583
(49) Ghayoumian, J.; Mohseni Saravi, M.; Feiznia, S.; Nouri, B.; Malekian, A. Application
584
of GIS techniques to determine areas most suitable for artificial groundwater recharge in
585
a coastal aquifer in southern Iran. Journal of Asian Earth Sciences 2007, 30, 364–374.
586
(50) Rahman, M. A.; Rusteberg, B.; Gogu, R. C.; Lobo Ferreira, J. P.; Sauter, M. A new
587
spatial multi-criteria decision support tool for site selection for implementation of man-
588
aged aquifer recharge. Journal of environmental management 2012, 99, 61–75. 30 ACS Paragon Plus Environment
Page 30 of 34
Page 31 of 34
Environmental Science & Technology
589
(51) Pedrero, F.; Albuquerque, A.; Marecos do Monte, H.; Cavaleiro, V.; Alarc´on, J. J.
590
Application of GIS-based multi-criteria analysis for site selection of aquifer recharge
591
with reclaimed water. Resources, Conservation and Recycling 2011, 56, 105–116.
592
(52) Bolstad, P. GIS Fundamentals: A First Text on Geographic Information Systems.; Eider
593
594
595
596
597
598
599
600
601
Press, 2005; pp 1–620. (53) Seaber, P. R.; Kapinos, F. P.; Knapp, G. L. Hydrologic Unit Maps: US Geological Survey Water Supply Paper 2294 ; USGS, 1987. (54) ASME, Municipal Wastewater Reuse by Electric Utilities: Best Practices and Future Directions; 2013; Vol. 5; pp 1–41. (55) Sadek, B. S.; Bedran, M.; Kaysi, I. GIS Platform for Multicriteria Evaluation of Route Alignments. Journal of Transportation Engineering 1999, 144–151. (56) Huber, D. L.; Church, R. L. Transmission corridor location modeling. Journal of Transportation Engineering 1985, 111, 114–130.
602
(57) Church, R. L.; Loban, S. R.; Lombard, K. An interface for exploring spatial alternatives
603
for a corridor location problem. Computers & Geosciences 1992, 18, 1095–1105.
604
(58) USGS, 3DEP Data Product. 2015; http://nationalmap.gov/3dep{\_}prodserv.
605
html.
606
(59) Mahdavi, A.; Tabatabaei, S.; Mahdavi, R.; Emamzadei, M. N. Application of digital
607
techniques to identify aquifer artificial recharge sites in GIS environment. International
608
Journal of Digital Earth 2011, 1–21.
609
(60) Chowdhury, A.; Jha, M. K.; Chowdary, V. M. Delineation of groundwater recharge
610
zones and identification of artificial recharge sites in West Medinipur district, West
611
Bengal, using RS, GIS and MCDM techniques. Environmental Earth Sciences 2010,
612
59, 1209–1222. 31 ACS Paragon Plus Environment
Environmental Science & Technology
613
614
615
616
(61) Goodchild, M. F. An evaluation of lattice solutions to the problem of corridor location. Environment and Planning A 1977, 9, 727–738. (62) Lombard, K.; Church, R. L. The gateway shortest path problem: generating alternative routes for a corridor location problem. Geographical systems 1993, 1, 25–45.
617
(63) Zhang, X.; Armstrong, M. P. Genetic algorithms and the corridor location problem:
618
multiple objectives and alternative solutions. Environment and Planning B: Planning
619
and Design 2008, 35, 148–168.
620
(64) Skriver, a. J. V.; Andersen, K. a. A label correcting approach for solving bicriterion
621
shortest-path problems. Computers and Operations Research 2000, 27, 507–524.
622
(65) Domb, C. Self avoiding walks on lattices. Stochastic Processes in Chemical Physics
623
2009, 15, 229.
624
(66) Liu, H. Pipeline engineering; CRC Press, 2003.
625
(67) Stokes, J.; Horvath, A. LCA Methodology and Case Study Life Cycle Energy As-
626
sessment of Alternative Water Supply Systems. International Journal of Life Cycle
627
Assessment 2006, 11, 335–343.
628
629
630
631
632
633
(68) Stokes, J. R.; Horvath, A. Energy and Air Emission Effects of Water Supply. Environmental Science & Technology 2009, 43, 2680–2687. (69) Stokes, J. R.; Horvath, A. Supply-chain environmental effects of wastewater utilities. Environmental Research Letters 2010, 5, 014015. (70) Stokes, J.; Horvath, A. Life-Cycle Assessment of Urban Water Provision : Tool and Case Study in California. Journal of Infrastructure Systems 2011, 15–24.
634
(71) Grubert, E. A.; Beach, F. C.; Webber, M. E. Can switching fuels save water? A
635
life cycle quantification of freshwater consumption for Texas coal-and natural gas-fired
636
electricity. Environmental Research Letters 2012, 7, 45801. 32 ACS Paragon Plus Environment
Page 32 of 34
Page 33 of 34
Environmental Science & Technology
637
(72) Mutel, C. L.; Pfister, S.; Hellweg, S. GIS-Based Regionalized Life Cycle Assessment:
638
How Big is Small Enough? Methodology and Case Study of Electricity Generation -
639
Supporting Information. Environmental Protection 2012, 46 .
640
(73) Meldrum, J.; Nettles-Anderson, S.; Heath, G.; Macknick, J. Life cycle water use for elec-
641
tricity generation: a review and harmonization of literature estimates. Environmental
642
Research Letters 2013, 8, 15031.
643
(74) Scanlon, B. R.; Reedy, R. C.; Duncan, I.; Mullican, W. F.; Young, M. Controls on
644
water use for thermoelectric generation: case study Texas, US. Environmental science
645
& technology 2013, 47, 11326–11334.
646
(75) Baker, J.; Strzepek, K.; Farmer, W.; Schlosser, C. A. Quantifying the impact of re-
647
newable energy futures on cooling water use. JAWRA Journal of the American Water
648
Resources Association 2014, 50, 1289–1303.
649
Supporting Information Available
650
Multi-objective corridor location algorithm runtime performance metrics, GIS data process-
651
ing workflows, and detailed case study analyses result data are all available in the SI:
652
This material is available free of charge via the Internet at http://pubs.acs.org/.
33 ACS Paragon Plus Environment
Environmental Science & Technology
653
Graphical TOC Entry
654
34 ACS Paragon Plus Environment
Page 34 of 34