Accepted Manuscript Cost, energy, global warming, eutrophication and local human health impacts of community water and sanitation service options Mary E. Schoen, Xiaobo Xue, Alison Wood, Troy R. Hawkins, Jay Garland, Nicholas J. Ashbolt PII:
S0043-1354(16)30895-8
DOI:
10.1016/j.watres.2016.11.044
Reference:
WR 12525
To appear in:
Water Research
Received Date: 15 June 2016 Revised Date:
31 October 2016
Accepted Date: 14 November 2016
Please cite this article as: Schoen, M.E., Xue, X., Wood, A., Hawkins, T.R., Garland, J., Ashbolt, N.J., Cost, energy, global warming, eutrophication and local human health impacts of community water and sanitation service options, Water Research (2016), doi: 10.1016/j.watres.2016.11.044. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Cost, Energy, Global Warming, Eutrophication and
2
Local Human Health Impacts of Community Water
3
and Sanitation Service Options
SC
RI PT
1
Mary E. Schoen*a, Xiaobo Xueb, Alison Woodc, Troy R. Hawkinsd, Jay Garlande, Nicholas J.
5
Ashboltf
6
a.
7
[email protected]
8
b.
9
Albany, State University of New York, 1 University Place, Rensselaer, NY 12144;
M AN U
4
Soller Environmental, Inc., 3022 King St., Berkeley, CA 94703;
TE D
Department of Environmental Health Sciences, School of Public Health, University at
[email protected]
11
c.
12
Engineering, 301 E. Dean Keeton St. C8600, Austin, TX 78712-8600,
[email protected]
13
d.
14
Lexington, MA 02421;
[email protected]
15
e.
16
45268;
[email protected]
EP
10
AC C
The University of Texas at Austin, Dept. of Civil, Architectural and Environmental
Franklin Associates, a Division of Eastern Research Group, 110 Hartwell Avenue,
U.S. Environmental Protection Agency, 26 W. Martin Luther King Drive, Cincinnati OH
1
ACCEPTED MANUSCRIPT
17
f.
Rm. 3-57D South Academic Building, School of Public Health, University of Alberta,
18
Edmonton AB T6G 2G7;
[email protected]
RI PT
19
ABSTRACT
21
We compared water and sanitation system options for a coastal community across selected
22
sustainability metrics, including environmental impact (i.e., life cycle eutrophication potential,
23
energy consumption, and global warming potential), equivalent annual cost, and local human
24
health impact. We computed normalized metric scores, which we used to discuss the options’
25
strengths and weaknesses, and conducted sensitivity analysis of the scores to changes in variable
26
and uncertain input parameters. The alternative systems, which combined centralized drinking
27
water with sanitation services based on the concepts of energy and nutrient recovery as well as
28
on-site water reuse, had reduced environmental and local human health impacts and costs than
29
the conventional, centralized option.
30
advantages of the alternative community water systems (compared to the conventional system)
31
were in terms of local human health impact and eutrophication potential, despite large,
32
outstanding uncertainties.
33
energy recovery technologies had the least local human health impact; however, the cost of these
34
options was highly variable and the energy consumption was comparable to on-site alternatives
35
without water reuse or energy recovery, due to on-site water treatment. Future work should aim
36
to reduce the uncertainty in the energy recovery process and explore the health risks associated
37
with less costly, on-site water treatment options.
TE D
M AN U
SC
20
EP
Of the selected sustainability metrics, the greatest
AC C
Of the alternative options, the systems with on-site water reuse and
2
ACCEPTED MANUSCRIPT
38
KEYWORDS
39
Sustainability; water; wastewater; LCA; QMRA
RI PT
40 1.0 Introduction
42
Planning for a sustainable community water system requires a comprehensive understanding
43
and assessment of the integrated source water, drinking water, and sanitation services over their
44
life cycles. In previous work, we described the need for and use of integrated sustainability
45
assessment to evaluate community water systems within a stakeholder-driven framework (e.g.,
46
integrated municipal water management (Thomas and Durham 2003)). In addition, we selected a
47
set of technical metrics and tools which we consider critical to evaluate built water services, but
48
also of reasonable effort to calculate (Xue et al. 2015). Then, we evaluated a selection of water
49
service options for the coastal community of Falmouth, MA, using the proposed technical
50
metrics, including environmental impacts (Xue et al. 2016), local human health impacts (Schoen
51
et al. 2014), cost (Wood et al. 2015), and technical resilience (Schoen et al. 2015). In this
52
companion paper, we summarize the strengths and weakness of the selected community water
53
systems across the previously calculated, technical sustainability metrics using newly calculated
54
normalized scores and discuss insights that can only come from looking at these metrics
55
together.
AC C
EP
TE D
M AN U
SC
41
56
Throughout, we refer to metrics, defined as a measurable value of an attribute (e.g., equivalent
57
annual cost), as well as the various input parameters (e.g., discount rate), which were used to
58
calculate the metrics. An input parameter, metric, or score is referred to as variable if the
59
variation in value cannot be reduced with collection of additional information; whereas
60
uncertainty can be better estimated with collection of more or better data (Vose and Vose 2000). 3
ACCEPTED MANUSCRIPT
The metrics previously described include: local human health impact from pathogen and
62
chemical exposures resulting from community-wide water system use; equivalent annual cost
63
(EAC), which quantifies the monetary costs and benefits of each system; life cycle energy
64
consumption; life cycle global warming potential (GWP) from on-site and supply chain
65
greenhouse gas emissions including CO2, CH4, and N2O; life cycle eutrophication potential,
66
which is based on on-site and supply chain releases of aqueous and atmospheric nitrogen and
67
phosphorus; and technical resilience, which qualitatively evaluates the water system’s capacity to
68
deal with potential future event and climatic challenges. Based on stakeholder input, only a
69
selection of the available life cycle analysis impact categories was included in the evaluation of
70
environmental impacts. Resilience was not included in the following comparative analysis
71
because we were unable to differentiate the selected water system options (Schoen et al. 2015).
M AN U
SC
RI PT
61
This assessment is the first we are aware of to evaluate both water (i.e., potable and non-
73
potable) and sanitation services (i.e., septic/sewage and greywater) across cost, environmental,
74
and local human health impacts. Portions of community water systems (i.e., either water or
75
sanitation) have been assessed by others using integrated or sustainability assessments for water
76
supply (Lai et al. 2007, Rygaard et al. 2014), energy and water recovery options (Lee et al.
77
2013), and firefighting flows (Aydin et al. 2014). These studies rarely include metrics that span
78
health, environment, economic, and technological aspects (Malmquist 2006), especially the local
79
human health impacts (Lai et al. 2007, Rygaard et al. 2014) and resilience metrics (Rygaard et al.
80
2014). A further common deficiency is the lack of systematic consideration of variability and
81
uncertainty across metrics when comparing system options (although, the variability in a subset
82
of quantitative metrics was discussed by Fagan et al. (2010) and Rygaard et al. (2014)).
AC C
EP
TE D
72
4
ACCEPTED MANUSCRIPT
The options considered here, described in the following section, include novel treatment and
84
energy recovery elements not yet widely implemented or evaluated across the cost, local human
85
health, and environmental metrics. As such, there remains considerable uncertainty associated
86
with the input parameters used to calculate the metrics. The objectives of this work are to
87
identify system options with clear advantages across the sustainability metrics while accounting
88
for natural variability and/or uncertainty; and identify results that may change with collection of
89
additional data to guide future information collection efforts for these novel technologies. While
90
this paper focuses on the technical sustainability assessment results, and not the entire decision-
91
making process, our discussion emphasizes how the results could be used in a stakeholder-
92
preferred decision approach (e.g., Multi-Criterion Decision Analysis [MCDA](Belton and
93
Stewart 2002)).
94
2.0 Approach
95
2.1 Case Study
96
The case study town of Falmouth, MA, faces expanding urbanization (with a population of
97
31,500 in 2011) and seasonal tourism, yet the predominating septic systems have resulted in
98
excessive nutrient exports and coastal eutrophication (Cape Cod Commission 2015).
99
evaluated five community water and wastewater service options to replace current traditional
SC
M AN U
TE D
EP
septic systems.
We
AC C
100
RI PT
83
101
The business-as-usual (BAU) system consisted of a conventional, centralized drinking water
102
system and a centralized wastewater treatment system, referred to here as the conventional
103
system (see Supporting Information Figure S1 for diagrams of the BAU treatment technology).
104
The Falmouth community consumes about 4.6 million gallons per day (MGD) of water,
105
approximately 60% of which is extracted from surface sources (Falmouth Department of Water 5
ACCEPTED MANUSCRIPT
2013). Considering the byproducts from wastewater treatment, the effluent and sludge, the
107
former entered the groundwater through filtration basins and the latter was transported (after
108
dewatering) out of the watershed to a management facility. There was no additional treatment of
109
the byproducts or reuse. The following “alternative” options maintained the centralized drinking
110
water system, but replaced the centralized wastewater treatment system. Two alternatives using
111
septic systems were proposed by the stakeholders and two additional options were selected based
112
on the concepts of energy recovery and water reuse.
SC
RI PT
106
The first alternative included dry composting toilets and on-site greywater treatment by the
114
existing septic system (CT-SS) that utilize absorption trenchs, where “greywater” refers to non-
115
toilet wastewater from sinks, showers, washing machines, etc., within households (refer to
116
Supporting Information Figure S2 for CT-SS technology diagrams). In the second alternative,
117
the centralized wastewater treatment was replaced with urine-diverting toilets and on-site fecal
118
solids (and greywater) treatment by the septic system (UD-SS) absorption trench system (refer to
119
Supporting Information Figure S3 for UD-SS technology diagrams). For these septic-based
120
options, the generated compost or urine was collected and transported out of the watershed to a
121
less nutrient-sensitive area for use as soil amendments. No additional treatment of the generated
122
byproducts was considered. All potable and non-potable household water uses were assumed to
123
be supplied by the existing centralized drinking water system.
AC C
EP
TE D
M AN U
113
124
In the third alternative, a low-volume flush toilet and blackwater pressure sewer were modeled
125
to provide the community with energy recovery via a combined heat and power (CHP) anaerobic
126
digester system in combination with community food residuals co-digestion. Blackwater was
127
assumed to be supplied only via the toilets and kitchen food-waste grinders, hence containing
128
more concentrated organic material and nutrients than traditional sewage, which is roughly 70% 6
ACCEPTED MANUSCRIPT
greywater. The dewatered digestate was assumed to be applied to local agricultural fields in the
130
environmental assessments, but shipped out of the watershed in the EAC assessment (discussed
131
in Sections 2.2.2 and 2.2.3). The on-site greywater was assumed to be collected; treated using
132
biological sand filtration followed by ultraviolet disinfection; and reused for toilet flushing,
133
outside irrigation, and watering homegrown salad crops, hence providing blackwater energy with
134
greywater reuse (BE-GR). The final alternative was identical to BE-GR with the addition of on-
135
site rainwater collection, treatment by in-line filtration and ultraviolet disinfection, and use as a
136
hot-water supply for showering (BE-GRR) (refer to Supporting Information Figure S4 for BE-
137
GR/R technology diagrams). The Cape Cod region has an annual precipitation average of 123
138
inches (based on the last 50 years) (NOAA 2013). Falmouth has an existing separate stormwater
139
system; therefore, stormwater was not addressed in this comparative analysis.
M AN U
SC
RI PT
129
2.2 Metrics
141
2.2.1 Local Human Health Impact
142
The local human health impact from the operation and community–wide use of each option
143
was estimated using quantitative risk assessment including both operating and possible failing
144
conditions (Schoen et al. 2014). The resulting key pathogen and chemical risks were translated
145
into DALYs (Disability-Adjusted Life Years), i.e., the sum of years of life lost due to premature
146
mortality and years lived with disability (Murray and Acharya 1997). The cumulative DALYs
147
for each option were expressed as the 100-year DALY per 10,000 people. The human health
148
impact was not assessed for global health burdens associated with the life cycle of the systems
149
(Heimersson et al. 2014, Kobayashi et al. 2015a, Kobayashi et al. 2015b). While we recognize
150
the relevance of the global human health life cycle impacts, particularly impacts from fine
151
particulate matter such as released during power production, these tools are currently under
AC C
EP
TE D
140
7
ACCEPTED MANUSCRIPT
152
development (Fantke et al. 2014) and were not included due to uncertainty associated with their
153
calculation. The exposure pathways for the conventional option included the dominant exposure of
155
ingestion of contaminated drinking water from a cross-connection of a drinking water main with
156
sewage and the accidental ingestion of contaminated recreational water. All but the rainwater
157
harvesting alternative (BE-GRR) included shower exposures to the surrogate disinfection by-
158
product, chloroform. For the energy recovery digester options (BE-GR and BE-GRR), we
159
modeled exposures to a suite of pathogens from ingestion of treated greywater from non-potable
160
home and garden use and from ingestion of treated rainwater while showering for BE-GRR. The
161
septic-based systems (CT-SS and UD-SS) included the exposure to contaminated recreational
162
water from tank effluent and leakage. The application of byproducts (e.g., compost or sludge) as
163
soil amendments was not assessed. The health impact assessment system boundary and
164
assumptions are provided in the Supporting Information, Figure S5 and Section S1, as well as
165
Schoen et al. (2014).
TE D
M AN U
SC
RI PT
154
2.2.2 Economic: Equivalent Annual Cost
167
The equivalent annual cost (EAC) of each system quantified the monetary costs and benefits of
168
each system with the reference year of 2014 (Wood et al. 2015). EAC provides a means of
169
combining one-time and ongoing costs and benefits into an annualized cost stream to allow for
170
comparison across alternatives with disparate temporal distributions of costs and benefits. All
171
EACs were calculated on a per-household basis.
AC C
EP
166
172
The scope of the EAC assessment included the material/energy inputs during the installation
173
(i.e., not manufacturing), operation and maintenance of water service starting from water
174
extraction and ending with the end-of-life discharge/reuse of wastewater byproducts. The local 8
ACCEPTED MANUSCRIPT
costs of transportation of byproducts out of the watershed were included but not the costs and
176
benefits of using the byproducts from novel treatment systems as fertilizers or soil conditioners
177
because these products are generally not legal in the U.S., and thus there exists no market. We
178
implicitly assumed replacement of systems/components in perpetuity, though costs become
179
irrelevant as they are further in the future because of discounting at a positive rate. The range of
180
discount rates explored comes from recommendations made by the National Center for
181
Environmental Economics' to use 3% and 7% for the public and private rates of return on
182
investments (we selected 5% as a midpoint) (National Center for Environmental Economics
183
2010).
M AN U
SC
RI PT
175
All data and assumptions were based on the current scenario in Falmouth, MA, whenever
185
possible. For example, we assumed that the water treatment and distribution systems currently in
186
place will remain in operation, so fixed costs associated with operating and maintaining
187
associated infrastructure were the same across all scenarios. Also, the sludge generated by BAU
188
was transported out of the watershed to a management facility (no additional treatment or soil
189
application was considered). The benefits captured include the cost savings due to water supply
190
reduction, as well as the benefit from sale of energy generated by the anaerobic digester. Cost
191
savings due to reduced water usage by certain technologies were accounted for by adjusting
192
varying supply costs according to the water demand for each scenario, based on Falmouth’s
193
existing block pricing structure for water supply. Since approximately 95% of homes in
194
Falmouth use septic systems, and approximately 20% of these systems are not functioning
195
properly, results reported here incorporate both functional and failing cases for the entire service
196
area, assuming the systems were retrofitted in the calculation of cost.
AC C
EP
TE D
184
9
The EAC system
ACCEPTED MANUSCRIPT
197
boundary and assumptions are in the Supporting Information, Figure S7 and Section S2, as well
198
as Wood et al. (2015). 2.2.3 Environmental Impact
200
Due to the local stakeholders’ interests, the life cycle assessment (LCA) was focused on the
201
energy consumption, global warming and eutrophication potentials of the water and wastewater
202
technology options (Xue et al. 2016). The scope of the LCA metrics included energy and
203
material inputs and associated emissions during the construction and operation/maintenance
204
stages of water services, assuming nominal operating conditions (not including impacts from
205
treatment failures), starting from water extraction and ending with wastewater discharge/reuse
206
(the system boundaries are depicted in Supporting Information Figure S8). The specific metrics
207
included: life cycle energy consumption (MJ per household.day); life cycle global warming
208
potential (kg CO2-equivalent per household.day); and life cycle eutrophication potential (g N-
209
equivalent per household.day). The metrics were computed based on U.S.-specific inventory and
210
using the Tool for the Reduction and Assessment of Chemical and Other Environmental Impacts
211
(TRACI) impact assessment (EPA 2014).
TE D
M AN U
SC
RI PT
199
The functional unit (i.e., a measure of the equivalent function of the studied systems) was the
213
water and ‘waste’ service requirement to meet a household’s water and sanitation needs. In order
214
to ensure a fair comparison, system expansion was conducted in order to account for the nitrogen
215
and phosphorus fertilizers and electricity provided by utilizing the household outflow of
216
compost, urine, and blackwater (refer to Supporting Information Table S3). Options that did not
217
supply energy or fertilizer were augmented with additional grid electricity (i.e., equivalent
218
electricity production) and synthetic fertilizer, so that each scenario had an equivalent amount of
219
electricity and fertilizer.
AC C
EP
212
10
ACCEPTED MANUSCRIPT
The life cycle eutrophication potential of fertilizers (or soil amendments) was estimated based
221
on the following assumptions. Similar to EAC, the dewatered sludge generated by BAU was
222
transported out of the watershed to a management facility (no additional treatment or soil
223
application was considered) and the urine and compost from CT-SS and UD-SS were transported
224
to a less nutrient-sensitive watershed where they were used as soil amendments; this was
225
possible because the energy savings of displacing synthetic fertilizer with urine was large enough
226
to transport urine via a diesel truck up to 70–120 km, depending on the truck characteristics.
227
Whereas, unlike the EAC assumptions, the dewatered digestate generated by BE-GR/R was used
228
locally for soil amendment due to the high energy used to transport it.
229
about the environmental metrics, please refer to Section S3 in the Supporting Information as well
230
as Xue et al. (2016).
M AN U
SC
RI PT
220
For more information
2.3 Uncertainty and Variability
232
In previous publications, we used three approaches to characterize the uncertainty and
233
variability in each metric: 1) the LCA-based metrics combined the variability and uncertainty of
234
input parameters into a Monte Carlo approach (not the life cycle impact characterization factors)
235
and conducted a separate parametric uncertainty analysis based on the 5th, 50th, and 95th
236
percentiles of the input parameters’ distributions (Xue et al. 2016); 2) the human health metric
237
captured variability of exposure input parameters using a Monte Carlo approach and generated
238
separate health impact distributions given different assumptions for a key uncertain input
239
parameter, which was identified as part of a separate parametric uncertainty analysis (Schoen et
240
al. 2014); and 3) the EAC metric used a parametric sensitivity analysis based on high- and low-
241
input parameter estimates to characterize variability and uncertainty (Wood et al. 2015). The
AC C
EP
TE D
231
11
ACCEPTED MANUSCRIPT
242
parameter/s with the most influence on each option’s metric variation using the previously
243
performed parametric sensitivity and uncertainty analyses are listed in Table S6. We selected a subset of the variable or uncertain parameters: parameters that when varied,
245
resulted in overlapping 5th and 95th percentiles of metric distributions across options (or the low
246
to high predicted range for cost and the human health impact of BAU) in the above mentioned
247
sensitivity analyses. These “key” parameters are discussed in the results section and were used
248
in a sensitivity analysis of the normalized scores, described in the following section.
SC
RI PT
244
249 2.4 Normalized Scores
251
In order to identify the best (and worst) performing system for each metric and compare the
252
relative difference between the best and worst performing options across metrics, we calculated
253
normalized scores of each metric (Linkov and Moberg 2011). We selected a zero-max scoring
254
method, where the options were scored relative to the best performing option, which was given a
255
score of one, based on the guidance provided by Rowley et al. (2012) (please see Supporting
256
Information Section 4 for the score calculation).
TE D
M AN U
250
To highlight the options with clear advantages versus options with a wide range of
258
performance due to natural variability and/or uncertainty, a sensitivity analysis of score was
259
performed to show how the relative performance of the systems change due to variations in the
260
selected, key input parameters.
AC C
EP
257
261
If the input parameter/s were uncorrelated among options, we recalculated all the options’
262
scores using the high and low metric values identified in the sensitivity analyses for the option
263
with uncertainty or variability and the best-estimate metric results for all other options (Tables
264
S7-11, Supporting Information). If the variability or uncertainty was correlated, then the set of 12
ACCEPTED MANUSCRIPT
affected options was varied while the other options were set at their best estimates. This resulted
266
in several scenarios for each metric, based on variations in key uncertain or variable input
267
parameters, each with a set of associated scores. The 5th and 95th values of the predicted metric
268
distribution were used instead of the predicted range from the parametric sensitivity analyses for
269
the LCA-based metrics to simplify the presentation after a comparison of scores using both
270
approaches showed little difference (results not shown).
RI PT
265
3.0 Results
272
The results for each metric are summarized below (Figure 1).
The dominating process
contributing to each metric is listed in Table S12, Supporting Information.
M AN U
273
SC
271
3.1 Local Human Health Impact
275
The energy recovery digester and greywater reuse option (BE-GR) had the overall best local
276
human health performance (due to the elimination of the exposure pathways of wastewater
277
contaminated recreational water and a cross-connection between a water main and sewage). The
278
alternative option that added rainwater use (BE-GRR) followed, and then the septic-based,
279
composting toilet option (CT-SS) (Figure 2a). The conventional option (BAU) risk was
280
attributed almost entirely to the cross-connection event of a water main with sewage. The BAU
281
had two key, uncertain variables that resulted in a range of predicted performance, the dilution of
282
the cross-contamination, and the frequency of sewer-to-drinking water cross-connection events.
AC C
EP
TE D
274
283
For the BAU median bar in in Figure 2a, a high dilution of sewage was assumed (i.e., 0.001),
284
and a frequency of cross-connection events of every other year. The error bars addressed event
285
frequencies of every year and 10 events in 100 years. We speculate that cross-connection events
286
happen quite frequently (EPA’s Office of Ground Water and Drinking Water 2001); therefore,
13
ACCEPTED MANUSCRIPT
287
we calculated scores for the scenarios of 10, 50, and 100 events in 100 years for BAU (Table
288
S13, Supporting Information). 3.2 Economic: Equivalent Annual Cost
290
The alternative options had lower equivalent annual cost (EAC) than the conventional option
291
(BAU) and these results were robust against the range of variability and uncertainty examined,
292
except for the energy recovery digester and rainwater use option (BE-GRR) whose range of
293
expected cost overlapped with the cost range of BAU (Figure 2b). The BE-GRR option may or
294
may not be lower in cost than the conventional option, depending on variable and uncertain
295
parameters. The parameters with the largest impact on BE-GRR’s costs were the 1) operation
296
and maintenance (O&M) costs for the greywater recycling system, which vary dramatically
297
depending on the specific system installed, and 2) the benefit from sales of electricity/heat
298
generated in the digestion process. We calculated scores using the high and low metric results
299
for the BE-GR and BE-GRR options, holding all other options at their best estimates (Table S14,
300
Supporting Information).
TE D
M AN U
SC
RI PT
289
Costs for both the BE-GR and BE-GRR systems being (generally) lower than BAU might
302
surprise professionals who know that pressure sewer systems have historically not been cost-
303
competitive with gravity sewer systems. The combined system components resulted in the lower
304
costs for the BE- systems. Key among these components are that extremely low-flush toilets
305
were used, greywater was treated onsite, and systems were at the neighborhood scale rather than
306
the city scale: these greatly reduced the volume of water being pumped and the distance of
307
pumping, which greatly reduced the energy costs associated with pressure sewers.
AC C
EP
301
308
The septic system capital cost was a variable parameter shared by both septic-based options
309
(UD-SS and CT-SS). This cost was extremely variable, depending on the specific installation
14
ACCEPTED MANUSCRIPT
and whether the existing septic system was usable or needed replacement. Still, if the septic costs
311
were high (low) for a composting system in a given home, then the costs were similarly high
312
(low) for a urine-diversion system in the same home. We calculated scores for the high and low
313
septic capital cost scenarios for these options, holding all other options at their best estimates
314
(Table S14, Supporting Information).
RI PT
310
3.3 Environmental Impact
316
3.3.1 Life cycle eutrophication potential
317
The conventional option (BAU) had the highest median eutrophication potential (Figure 2c).
318
The septic-based options (CT-SS and UD-SS) had the least eutrophication potential over the
319
range of variability and uncertainty explored. This was due to the assumption that urine and
320
compost were collected from households and transported to a less nutrient-sensitive watershed
321
where they could be used as soil amendments; whereas, we assumed that the digestate was used
322
locally for soil amendments due to the high energy use to transport it.
TE D
M AN U
SC
315
The eutrophication potential from the energy recovery digester options (BE-GR and BE-GRR)
324
was highly variable and uncertain, depending on the composition of feed to the digester, and on
325
various operation conditions such as temperature and retention time. We calculated scores using
326
the 5th and 95th percentile eutrophication potential for the BE-GR and BE-GRR options, holding
327
all other options at their best estimates (Table S15, Supporting Information).
AC C
EP
323
328
3.3.2 Life cycle energy consumption
329
The conventional option (BAU) was the most energy intensive over the range of variability
330
and uncertainty considered (Figure 2d), driven by the equivalent electricity production, sewage
331
treatment, and drinking water treatment and distribution. Conversely, the blackwater energy
15
ACCEPTED MANUSCRIPT
332
recovery sewer options (BE-GR and BE-GRR), because of the associated electricity output, had
333
the lowest median net energy requirement. Looking only at the septic-based options, UD-SS was always more energy intensive in the
335
Monte Carlo iterations than CT-SS since the former demanded more treated drinking water than
336
the latter. For the energy recovery digester options, BE-GRR was more energy intensive than
337
BE-GR, since the rainwater treatment was more energy intensive than treating and supplying
338
shower water with a centralized drinking water system.
SC
RI PT
334
There was a great deal of overlap in the predictions of the septic-based options and the energy
340
recovery digester options, and the dominating sources of variability and uncertainty were
341
uncorrelated. Variation in the direct electricity consumption of greywater treatment and reuse
342
had a great impact on metric variation for the BE-GR and BE-GRR alternatives. To estimate the
343
impact of this variation on the score, we calculated scores using the 5th and 95th percentiles of the
344
energy consumption distributions for the BE-GR and BE-GRR options, holding all other options
345
at their best estimates (Table S16, Supporting Information).
TE D
M AN U
339
Uncertainty in the septic-based options’ energy consumption was introduced from the
347
equivalent electricity production estimate. The energy produced from the bioreactors, which
348
sourced from household food waste and restaurant grease traps, was uncertain because of
349
uncertain compositions of digestion feed, which influenced the modeled biogas generation rate
350
and the methane content. We calculated scores using the 5th and 95th percentiles of the energy
351
consumption distributions for the CT-SS, UD-SS, and BAU options, holding the energy recovery
352
digester options at their best estimate metric values (Table S16, Supporting Information).
353
AC C
EP
346
3.3.3 Life cycle global warming potential (GWP)
16
ACCEPTED MANUSCRIPT
The options with the energy recovery digester (BE-GR and BE-GRR) had much lower life
355
cycle greenhouse gas emissions over the range of variability and uncertainty explored (Figure
356
2e), mainly because the conventional (BAU) and septic-based options (CT-SS and UD-SS)
357
shared an additional equivalent electricity production. Although the BAU, CT-SS, and UD-SS
358
options had large, overlapping ranges of variability and uncertainty, the CT-SS and UD-SS
359
options had less GWP potential than BAU over the Monte Carlo iterations, given that all share
360
the equivalent electricity production; however, BAU also included the energy intensive
361
centralized wastewater treatment and sewerage collection system.
SC
RI PT
354
Variability for all options was due, in part, to natural variation in the life cycle GWP of the
363
electricity mix (i.e., the percentages for various energy sources such as coal, natural gas,
364
biomass, etc.).
365
introduced by the amount of electricity produced from co-digestion and CHP processes. We
366
calculated scores using the 5th percentile value of the GWP distribution for the BAU, CT-SS, and
367
UD-SS options due to low electricity produced from co-digestion and CHP processes, holding
368
the other options at their best estimate (Table S17, Supporting Information). We did not include
369
a scenario with high GWP since the higher values of the predicted GWP distributions likely
370
resulted from high GWP of the electricity mix, which is a factor shared by all options.
M AN U
362
EP
TE D
For the conventional and septic-based options, additional uncertainty was
3.4 Normalized Scores
372
The comparison of technical sustainability metric scores of the water and wastewater system
373
options is summarized as a performance matrix (Table 1). The conventional option (BAU) had
374
the poorest performance among options for all scenarios examined across the local human health
375
impact, economic, and environmental impact metrics and for the range of variability and
376
uncertainty explored (Table 1), with one exception: the energy recovery digester option with
AC C
371
17
ACCEPTED MANUSCRIPT
377
greywater reuse and rainwater use (BE-GRR) had variable relative performance compared to that
378
of BAU in EAC, depending on the assumptions. Hence, no alternative option dominated in performance across all metrics. Given the input
380
assumptions, the septic-based options (CT-SS and UD-SS) scored relatively better in life cycle
381
eutrophication potential; the energy recovery digester options (BE-GR and BE-GRR) scored
382
relatively better in human health impact and GWP.
RI PT
379
It was difficult to differentiate the alternatives in terms of EAC due to variability. The relative
384
performance of the alternatives in terms of EAC changed depending on assumptions about the
385
cost of the greywater treatment and septic components (Figure 2a). For example, the BE-GR
386
score ranged from 0.57 to 1.00, depending on the on-site greywater treatment costs. The relative
387
performance of the alternatives in terms of life cycle energy consumption were similar and had
388
less variability and/or uncertainty than the EAC scores; however, the alternative with the highest
389
score changed given assumptions about the energy requirements for greywater treatment and the
390
equivalent electricity production (Figure 2b).
TE D
M AN U
SC
383
If decision makers were selecting between alternative options, BE-GR always out-performed
392
BE-GRR, and CT-SS performed better than UD-SS on all metrics, except EAC. Looking closer
393
at the BE-GR and CT-SS scores, there were large differences in the local human health and
394
eutrophication potential scores between these two options, with CT-SS performing better for
395
eutrophication and BE-GR performing better for human health. Of course, these scores do not
396
account for the relative importance of the metrics for stakeholders.
397
potential was the most important metric in the decision process, as for the Falmouth
398
stakeholders, and the other metrics were roughly half as important or less, then CT-SS would
399
have a combined best-estimate score (i.e., the sum of the product of the best estimate normalized
AC C
EP
391
If the eutrophication
18
ACCEPTED MANUSCRIPT
score and a metric-specific importance weight for each metric) greater than BE-GR, even though
401
BE-GR scored better in local human health, EAC and GWP. However, the relative comparison
402
would likely change if our assumptions about the use of byproducts as soil amendments was
403
altered (see Section 4.2).
RI PT
400
4.0 Discussion
405
4.1 How could the results be used?
406
The normalized scores were computed to illustrate the strengths and weaknesses of the system
407
options across metrics. However, the scores can also be used in a decision process to further
408
explore tradeoffs. Previous work on integrated sustainability assessment of urban water systems
409
have identified the following general steps in a multiple criteria decision process: 1) structuring
410
the decision problem; 2) articulating and modeling the preferences; 3) aggregating the alternative
411
evaluations; and 4) developing recommendations (Guitouni and Martel 1998, Lai et al. 2008).
412
Decision makers interested in the options reported herein may choose to consider the
413
sustainability metric findings and normalized scores, along with other considerations (e.g., global
414
health impact (Fantke et al. 2014), water withdrawal, social, cultural, political and governance
415
aspects (Bertera 2013)), in steps 1, 3, and 4 using their preferred decision approach to select an
416
option.
EP
TE D
M AN U
SC
404
When aggregating the alternative evaluations and developing recommendations, we
418
recommend an approach that captures the variability and uncertainty in technical metric
419
performance. Many of the options had a wide range of metric performance due to outstanding
420
uncertainty and/or natural variability, particularly the local human health impact, EAC, and life
421
cycle energy use. This uncertainty (or variation) as well as the possible correlation among
422
system options can be included in the aggregating step using, at a minimum, the multiple zero-
AC C
417
19
ACCEPTED MANUSCRIPT
423
max score sets for EAC and life cycle energy use, based on variations in key uncertain or
424
variable parameters (see Supporting Information Tables S18-S23 for linear scores).
425
advanced stochastic methods for aggregation require Monte Carlo samples from metric
426
distributions or score distributions, which would could be generated from the high, low scenarios
427
for EAC (Tervonen and Figueira 2008).
RI PT
More
Given the selected community water system options and assumptions, the outstanding
429
variability and uncertainty associated with EAC and energy consumption would be important to
430
resolve if these metrics were important to stakeholders in a decision process. Additional data on
431
the electricity production from digestion/co-digestion may affect the rankings of the alternative
432
options in life cycle energy consumption. This uncertainty also affected the GWP predictions,
433
but did not greatly affect the normalized scores of the options. Additional data on GWP and
434
EAC would not influence the decision process if, for example, eutrophication potential was
435
highly important compared to the other metrics. In addition, improvements to the metrics (i.e.,
436
the models) could better characterize the life-cycle eutrophication potential, EAC, and technical
437
resilience of system options (discussed in the Supporting Information, Section S5).
TE D
M AN U
SC
428
4.2 Assumptions not explored through sensitivity analysis
439
For the septic-based options (CT-SS and UD-SS), we assumed that collected urine or compost
EP
438
was shipped out of the watershed and applied to a less nutrient-sensitive location.
441
assumption resulted in a relatively low eutrophication potential for the septic-based options. If
442
the byproducts were applied to soils within the watershed, perhaps due to geography or limited
443
access to transportation, the eutrophication potential of the septic-based options would increase
444
by an unknown magnitude, with likely small decreases in GWP and energy consumption.
AC C
440
This
20
ACCEPTED MANUSCRIPT
We assumed that a distribution-wide cross-connection between potable and non-potable water
446
was unlikely for CT-SS, UD-SS, and BE-GR/R because these systems eliminated the traditional,
447
centralized sewerage distribution system. For the local human health impact of BE-GR/R, we
448
accounted for non-potable exposures to poorly treated or untreated greywater and rainwater;
449
however, we did not include isolated potable exposures to the untreated, on-site collected waters.
450
Using the results from the Supporting Information, Section S1, the 100-year human health
451
burden would increase by approximately 2 DALY per 10,000 people or less if one percent of the
452
population ingested of a mouthful (i.e., approximately 100 ml day-1) of contaminated water for a
453
duration of approximately a month each year. An increase in total BE-GR/R DALY of this
454
magnitude would not change the rankings of the options in terms of local human health impact.
M AN U
SC
RI PT
445
455
4.3 Scope Across Metrics
456
A key challenge was selecting the scope for each analysis.
For example, the scale of impacts
was different across metrics (e.g., local, non-local, and global), especially considering the
458
impacts from the application of byproducts., We accounted for the eutrophication potential from
459
both the local application of digestate generated from the co-digestion process and non-local
460
application of compost generated from the dry composting toilets.
461
assessment, we assumed that all of the options shared the same local health impacts from
462
agricultural applications given the fixed area of agricultural land, regardless if traditional
463
fertilizers or digestate was applied. We did not estimate non-local or global health impacts.
464
Whereas, for EAC, we did not quantify the costs and benefits of using fertilizer or soil
465
conditioner products generated from human waste because these products are generally not legal
466
in the U.S., and thus there exists no market and limited data to estimate the costs or benefits. We
467
recommend a comprehensive assessment to identify the tradeoffs between environmental,
For the human health
AC C
EP
TE D
457
21
ACCEPTED MANUSCRIPT
468
economic, and human health aspects of utilizing urine/digestate/compost in agriculture for future
469
work. For the LCA-based metrics, environmental impacts including life cycle energy consumption,
471
global warming and eutrophication potentials, were calculated assuming equal provision of water
472
and wastewater service in Falmouth across options. For the local human health impact, we did
473
not assume equal provision of water and wastewater service across options; rather, we captured
474
expected local health impacts to the exposed users. Therefore, our analysis did not include
475
health impacts resulting from the utilization of LCA-derived “equivalent” products like
476
fertilizers.
477
differences between LCA impact assessment and microbial risk assessment quantifying local
478
human health impact (Harder et al. 2015, Harder et al. 2014).
SC
RI PT
470
M AN U
Recent publications include a comprehensive discussion on system boundary
4.4 Use of results for other locations
480
The human health impact findings can be transferred to other coastal localities for the selected
481
options. Only the recreational water risks were specific to Falmouth. Although the recreational
482
risks may change for other locations based on treatment plant outfall locations, septic density,
483
and other assumptions, the overall ranking of the options would likely remain unchanged, given
484
the extremely high health burden associated with the BAU cross-connection event and the low
485
health burden associated with the BE-GR/R options.
AC C
EP
TE D
479
486
In terms of cost, the centralized systems will vary the most from location to location,
487
depending on factors such as housing density, soil types, and plant design. For decentralized
488
systems, regional effects are likely to be smaller than the variation and uncertainty already
489
incorporated in the cost estimates, especially for the most sensitive parameters. It is likely that in
22
ACCEPTED MANUSCRIPT
490
other coastal locations where septic systems are the current norm, the septic options would also
491
have lower costs than the centralized option and the digester would remain highly uncertain. Considering the life cycle energy consumption and GWPs, the ability to apply the results
493
specific to Falmouth to other coastal localities is limited by the use of specific values
494
representing local topography, water resources and quality, plant design, and climate. For
495
example, the energy consumption of wastewater and blackwater collection is highly dependent
496
on local factors such as topography, population density, and transport distance. However, the
497
energy consumption and GWP of the on-site greywater reuse system with a designated treatment
498
technology should vary less across different locations than the BAU would. The absolute values
499
of eutrophication potentials will vary based the digester characteristics, when applicable; the
500
nutrient transport and fate in the local environment; and the characteristics of the receiving water
501
body.
M AN U
SC
RI PT
492
4.5 Insights: Alternative community water systems for coastal communities
503
The community water systems that incorporated elements of water reuse and energy recovery
504
did not perform best across the selected sustainability metrics compared to options without these
505
elements. The local human health impact of options with community energy recovery and on-site
506
water reuse was lower than options without these elements (i.e., options incorporating septic-
507
based sanitation services and the conventional system); however, the cost of these energy
508
recovery and water reuse options was highly uncertain and variable, due in part to the variability
509
in cost of the on-site water treatment systems. In addition, the life-cycle energy consumption was
510
comparable for alternatives options with and without energy recovery, for the majority of the
511
sensitivity analysis runs. The sensitivity analysis of the cost and energy consumption scores
512
indicated that decreasing the on-site greywater treatment cost and energy use increases the
AC C
EP
TE D
502
23
ACCEPTED MANUSCRIPT
relative performance of systems that incorporated elements of water reuse and energy recovery.
514
Given this knowledge and the predicted low health risks associated with the reuse of treated
515
greywater and harvested rainwater for non-potable purposes (please refer to the review of health
516
risk from non-potable reuse (Schoen and Garland 2015)), future work should explore the health
517
risks of on-site treatment options that are less costly and less energy intensive than biological
518
sand filtration followed by conventional ultraviolet disinfection.
520
5.0 Conclusions •
SC
519
RI PT
513
Alternative community water systems, based on the concepts of energy and nutrient recovery as well as on-site water reuse, had reduced environmental and local human
522
health impacts and costs than a conventional, centralized system for the selected coastal
523
community.
524
•
M AN U
521
The sensitivity analysis of the normalize sustainability metric scores to changes in key, uncertain and variable parameters identified options with clear advantages for some
526
metrics and options with a wide range of performance due to natural variability and/or
527
uncertainty. •
Of the selected sustainability metrics, the greatest advantages of the alternative
EP
528
TE D
525
community water systems (compared to the conventional system) were in terms of
530
local human health impact and eutrophication potential, despite large, outstanding
531 532 533
AC C
529
uncertainties. Whereas, the outstanding variability and uncertainty made it difficult to differentiate the relative performance of the options in terms of cost and energy consumption.
24
ACCEPTED MANUSCRIPT
534
•
Additional research on the expected energy generation from community digesters could
535
be important to collect to differentiate the overall sustainably of the options, but only if
536
energy consumption is relatively important to stakeholders. •
The LCA practice of “system expansion” remains problematic when estimating the
RI PT
537 538
local human health impact given that the local risk assessment focused on expected
539
exposures to users. •
Ultimately, modifications to the technology and additional iterations of integrated
SC
540
sustainability assessment are required to find a water and sanitation service solution
542
that performs best across all technical sustainability metrics.
M AN U
541
Abbreviations
544
EAC, equivalent annual cost which quantifies the monetary costs and benefits of each system;
545
GWP, life cycle global warming potential which is resulted from on-site and supply chain
546
greenhouse gas emissions including CO2, CH4, and N2O; BAU, the business-as-usual system,
547
consisted of a conventional, centralized drinking water system and a centralized wastewater
548
treatment system; CT-SS, composting toilets and on-site greywater treatment by the existing
549
septic tank absorption trench system; UD-SS, urine-diverting toilets and on-site fecal solids
550
treatment by the septic tank absorption trench system; BE-GR, non-potable greywater reuse
551
paired with a low-volume flush toilet and blackwater pressure sewer with energy recovery via a
552
combined heat and power anaerobic digester system in combination with community food
553
residuals co-digestion; BE-GRR, identical to BE-GR, with the addition of on-site rainwater use
554
as a hot-water supply for showering; MCDA, Multi-Criterion Decision Analysis; MGD, million
555
gallons per day; CHP, combined heat and power; DALYs, Disability-Adjusted Life Years;
AC C
EP
TE D
543
25
ACCEPTED MANUSCRIPT
TRACI, Tool for the Reduction and Assessment of Chemical and Other Environmental Impacts;
557
LCA, Life Cycle Assessment; and O&M, operation and maintenance.
558
Funding Sources
559
This project was supported by the U.S. Environmental Protection Agency Office of Research and
560
Development.
561
Acknowledgment
562
This project was supported by the U.S. Environmental Protection Agency Office of Research and
563
Development. The views expressed in this article are those of the authors and do not necessarily
564
reflect the views or policies of the U.S. Environmental Protection Agency. Any mention of
565
specific products or processes does not represent endorsement by the U.S. EPA. We thank our
566
peer reviewers, Dr. Jennifer Cashdollar, Dr. Cissy Ma, Dr. Michael Blackhurst, and Dr.
567
Desmond Lawler for adding insight into this assessment. We thank the Cape Cod community,
568
US EPA Region 1, and the Cape Cod Commission, as well as specific contributors (Marilyn ten
569
Brink, Hilde Maingay, Earle Barnhart, Gerald Potamis, Ken Moraff, Valerie Nelson and
570
Abraham Noe-Hays).
EP
TE D
M AN U
SC
RI PT
556
AC C
571 572
References
573 574 575 576 577 578 579 580
Aydin, N., Mays, L. and Schmitt, T. (2014) Technical and environmental sustainability assessment of water distribution systems. Water Resources Management 28(13), 4699-4713. Belton, V. and Stewart, T. (2002) Multiple criteria decision analysis: an integrated approach, Springer Science & Business Media. Bertera, W.J. (2013) Envision: A sustainability guide for water professionals (PDF). JournalAmerican Water Works Association 105(7), 50-53. Cape Cod Commission (2015) Cape Cod Area Wide Water Quality Management Plan Update, March 2015, Cape Cod Commission, Barnstable, MA.
26
ACCEPTED MANUSCRIPT
EP
TE D
M AN U
SC
RI PT
EPA, U.S. (2014) Tool for the Reduction and Assessment of Chemical and Other Environmental Impacts (TRACI), US EPA, http://www.epa.gov/ord/NRMRL/std/traci/traci.html. EPA’s Office of Ground Water and Drinking Water (2001) Potential contamination due to crossconnections and backflow and the associated health risks, EPA OGW DW, Washington, DC. Fagan, J., Reuter, M. and Langford, K. (2010) Dynamic performance metrics to assess sustainability and cost effectiveness of integrated urban water systems. Resources, Conservation and Recycling 54(10), 719-736. Falmouth Department of Water (2013) Water Treatmen and Distribution, http://www.falmouthmass.us/depart.php?depkey=water. Fantke, P., Jolliet, O., Evans, J.S., Apte, J.S., Cohen, A.J., Hänninen, O.O., Hurley, F., Jantunen, M.J., Jerrett, M. and Levy, J.I. (2014) Health effects of fine particulate matter in life cycle impact assessment: findings from the Basel Guidance Workshop. The International Journal of Life Cycle Assessment 20(2), 276-288. Guitouni, A. and Martel, J.-M. (1998) Tentative guidelines to help choosing an appropriate MCDA method. European Journal of Operational Research 109(2), 501-521. Harder, R., Holmquist, H., Molander, S., Svanström, M. and Peters, G.M. (2015) Review of environmental assessment case studies blending elements of risk assessment and life cycle assessment. Environmental Science & Technology 49(22), 13083-13093. Harder, R., Schoen, M.E. and Peters, G.M. (2014) Including pathogen risk in life cycle assessment of wastewater management. Implications for Selecting the Functional Unit. Environmental Science & Technology 49(1), 14-15. Heimersson, S., Harder, R., Peters, G.M. and Svanstrom, M. (2014) Including pathogen risk in life cycle assessment of wastewater management. 2. Quantitative comparison of pathogen risk to other impacts on human health. Environmental Science & Technology 48(16), 9446-9453. Kobayashi, Y., Peters, G.M., Ashbolt, N.J., Heimersson, S., Svanström, M. and Khan, S.J. (2015a) Global and local health burden trade-off through the hybridisation of quantitative microbial risk assessment and life cycle assessment to aid water management. Water Research 79, 26-38. Kobayashi, Y., Peters, G.M., Ashbolt, N.J., Shiels, S. and Khan, S.J. (2015b) Assessing burden of disease as disability adjusted life years in life cycle assessment. Science of The Total Environment 530-531(1), 120-128. Lai, E., Lundie, S. and Ashbolt, N.J. (2007) A new approach to aid urban water management decision making using trade-off sacrifice modelled by fuzzy logic. Water Science & Technology 56(8), 11-20. Lai, E., Lundie, S. and Ashbolt, N.J. (2008) Review of multi-criteria decision aid for integrated sustainability assessment of urban water systems. Urban Water Journal 5(4), 315-327. Lee, E.J., Criddle, C.S., Bobel, P. and Freyberg, D.L. (2013) Assessing the scale of resource recovery for centralized and satellite wastewater treatment. Environmental Science & Technology 47(19), 10762-10770. Linkov, I. and Moberg, E. (2011) Multi-criteria decision analysis: environmental applications and case studies, CRC Press. Malmquist, P.-A. (2006) Strategic planning of sustainable urban water management, IWA, London. Murray, C.J. and Acharya, A.K. (1997) Understanding DALYs (disability-adjusted life years). Journal of Health Economics 16(6), 703-730.
AC C
581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625
27
ACCEPTED MANUSCRIPT
RI PT
SC
M AN U
TE D
EP
657 658
National Center for Environmental Economics (2010) Guidelines for Preparing Economic Analyses, U.S. Environmental Protection Agency Office of Policy, http://yosemite.epa.gov/ee/epa/eerm.nsf/vwAN/EE-0568-50.pdf/$file/EE-0568-50.pdf. NOAA (2013) Cape Cod precipitation datasets. Rowley, H.V., Peters, G.M., Lundie, S. and Moore, S.J. (2012) Aggregating sustainability indicators: Beyond the weighted sum. Journal of environmental management 111, 24-33. Rygaard, M., Godskesen, B., Jørgensen, C. and Hoffmann, B. (2014) Holistic assessment of a secondary water supply for a new development in Copenhagen, Denmark. Science of The Total Environment 497–498(0), 430-439. Schoen, M., Hawkins, T., Xue, X., Ma, C., Garland, J. and Ashbolt, N.J. (2015) Technologic resilience assessment of coastal community water and wastewater service options. Sustainability of Water Quality and Ecology 6, 75-87. Schoen, M.E. and Garland, J. (2015) Review of pathogen treatment reductions for onsite nonpotable reuse of alternative source waters. Microbial Risk Analysis In Press. Schoen, M.E., Xue, X., Hawkins, T.R. and Ashbolt, N.J. (2014) Comparative human health risk analysis of coastal community water and waste service options. Environmental Science & Technology 48(16), 9728-9736. Tervonen, T. and Figueira, J.R. (2008) A survey on stochastic multicriteria acceptability analysis methods. Journal of Multi-Criteria Decision Analysis 15(1-2), 1-14. Thomas, J.-S. and Durham, B. (2003) Integrated water resource management: looking at the whole picture. Desalination 156(1), 21-28. Vose, D. and Vose, D. (2000) Risk analysis : a quantitative guide, Wiley, Chichester; New York. Wood, A., Blackhurst, M., Hawkins, T., Xue, X., Ashbolt, N. and Garland, J. (2015) Costeffectiveness of nitrogen mitigation by alternative household wastewater management technologies. Journal of Environmental Management 150, 344-354. Xue, X., Hawkins, T.R., Schoen, M.E., Garland, J. and Ashbolt, N.J. (2016) Comparing the life cycle energy consumption, global warming and eutrophication potentials of several water and waste service options. Water 8(4), 154. Xue, X., Schoen, M.E., Ma, X.C., Hawkins, T.R., Ashbolt, N.J., Cashdollar, J. and Garland, J. (2015) Critical insights for a sustainability framework to address integrated community water services: Technical metrics and approaches. Water Research 77, 155-169.
AC C
626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656
28
ACCEPTED MANUSCRIPT
Table 1. Best estimate zero-max scores and ranges (in parentheses)1 computed from variations in selected uncertain parameters (superscript) Centralized water and digester with greywater nonpotable reuse (BEGR)
Centralized water and digester with greywater reuse and rainwater for non-potable use (BE-GRR)
1.5×10-2
Centralized water and urinediversion toilet with septic (UDSS) 7.2×10-3
Human Health Impact
1.7×10-3 (8.9×10-4-7.1×10-3)2
1.00
5.6×10-2
Life Cycle Eutrophication Potential
4.5×10-2
1.00
0.34
6.1×10-2 (4.6×10-2-9.6×10-2)3
6.1×10-2 (4.6×10-2-9.6×10-2)3
Equivalent Annual Cost
0.38 (0.26-0.38)4-5
0.81 (0.55-0.85) 4-5
1.00 (0.68-1.00) 4-5
1.00 (0.57-1.00) 4-5
0.52 (0.38-0.52) 4-5
Life Cycle Global Warming Potential
0.20 (0.31)6
0.22 (0.37) 6
0.22 (0.35) 6
1.00
0.92
Life Cycle Energy Use
0.55 (0.41-0.61)7-8
0.83 (0.63-0.93) 7-8
1.00 (0.88-1.00) 7-8
0.93 (0.84-0.93) 7-8
RI PT
Centralized water and composing toilet with septic (CT-SS)
SC
Centralized water and sewage (BAU)
TE D
M AN U
Metric
0.90 (0.67-1.00) 7-8
EP
1. The range (in parentheses) was omitted if the variation in input parameter(s) resulted in no change in score
AC C
2. Frequency of cross-connection between a water main and sewage for BAU 3. Discharge from the digester digestate for BE-GR/R 4. Greywater treatment system operation and maintenance costs for BE-GR/R
5. Septic system costs of UD/CT-SS 6. Only one variation considered, low GWP for BAU, CT/UD-SS, due to low electricity
produced from the co-digestion and CHP process for BE-GR/R 7. Equivalent electricity required for BAU and CT/UD-SS 8. Energy consumption of greywater treatment and reuse for BE-GR/R
ACCEPTED MANUSCRIPT
b. Equivalent annual cost
1.E+03
5,000
1.E+02
4,000
1.E+01 1.E+00 1.E-01 1.E-02 UD-SS
BE-GR
BE-GRR
BAU
MJ per household.day
10 5 0
UD-SS
BE-GR
BE-GRR
1,400 1,200
M AN U
g N-eq per household.day
15
CT-SS
d. Life cycle energy consumption
20
1,000 800 600 400 200 0
BAU
CT-SS
UD-SS
BE-GR
BE-GRR
TE D
e. Life cycle global warming potential 200 150
EP
100 50 0
AC C
kg Co2-eq per household.day
3
1,000
SC
CT-SS
c. Life cycle eutrophication potential
2
2,000
0 BAU
1
3,000
RI PT
$ per household
DALYs per 10,000 people over 100 years
a. Local human health impact
BAU
CT-SS
UD-SS
BE-GR
BE-GRR
BAU
CT-SS
UD-SS
BE-GR
BE-GRR
ACCEPTED MANUSCRIPT
Figure 1. Sustainability metric results for (a) local human health impact (DALYs per 10,000 people over 100 years), (b) equivalent annual cost (2014 $ per household), (c) eutrophication potential (g N-equivalent per household.day), (d) energy consumption (MJ per household.day),
RI PT
and (e) global warming potential (kg CO2–equivalent per household.day). Bars represent median results and error bars present 5th and 95th percentiles (except for cost and human health impact for BAU, where the bar represents the best estimate and error bars present low and high
SC
scenarios). BAU: business-as-usual conventional, centralized sewer; CT-SS: composting toilet with greywater septic system; UD-SS: urine-diversion toilet with household sewer to septic
M AN U
system; BE-GR: blackwater sewer for energy recovery and household greywater treatment and reuse; BE-GRR: the same as BE-GR but also including rainwater harvesting, treatment, and use
AC C
EP
TE D
for hot-water.
ACCEPTED MANUSCRIPT
High septic system costs for CT/UD-SS
RI PT
Low septic system costs for CT/UD-SS BE-GRR
High greywater treatment costs for BE-GR/R
BE-GR UD-SS
Low greywater treatment costs for BEGR/R
CT-SS
Base Case Cost 0
0.2
0.4
0.6
SC
BAU
0.8
1
M AN U
a. Equivalent Annual Cost Score
Low energy requirements for greywater treatment for BE-GR/R High energy requirements for greywater treatment for BE-GR/R
TE D
Low equivalent electricity for BAU, CT/UD-SS
BE-GRR BE-GR UD-SS
High equivalent electricity for BAU, CT/UD-SS
CT-SS BAU
Base Case Energy Use
EP
0
0.2
0.4
0.6
0.8
1
b. Life Cycle Energy Consumption Score
AC C
Figure 2. Sensitivity Analysis of Zero-max scores for (a) equivalent annual cost and (b) life cycle energy consumption. The scores (x-axis) of the community water system options are presented for different input parameter assumptions (y-axis).
ACCEPTED MANUSCRIPT
•
Novel community water systems had reduced environmental and health impacts and costs than a conventional system. The greatest advantages of novel water systems were in local human health impact and
RI PT
•
eutrophication potential. •
There was large uncertainty associated with on-site water treatment and community
The novel water system options had different strengths and weaknesses across metrics.
AC C
EP
TE D
M AN U
•
SC
digesters.