SEPA
United States
Environmental Protection
Agency
FINAL
Life Cycle Assessment of Upgrade
Options to Improve Nutrient Removal
for the City of Santa Fe, NM,
Paseo Real Wastewater Treatment Plant
Prepared for:
U.S. Environmental Protection Agency
Standards and Health Protection Division
Office of Water, Office of Science and Technology
1200 Pennsylvania Avenue NW (4305T)
Washington, DC 20460
Prepared by:
Eastern Research Group, Inc.
110 Hartwell Ave
Lexington, MA 02421
January 5, 2023
EPA Document Number 820-R-23-001
-------
Executive Summary
Executive Summary
Nutrient pollution of waterbodies across the United States is one of the most pervasive
environmental issues facing the country today.1 In partnership with states, tribes, and other
federal agencies, the U.S. Environmental Protection Agency (EPA) has led efforts to address
nutrient pollution by providing scientific and technical assistance for implementing nutrient-
based policies and regulations, including numeric nutrient water quality criteria, total maximum
daily loads, and effluent limits for point source dischargers.
Recently, wastewater treatment plant (WWTP) operators and stakeholders have
expressed concern over the potential for significant environmental and health implications
associated with treatment technologies required to achieve more stringent effluent concentrations
for nutrients (i.e., nitrogen and phosphorus) (Falk et al., 2013; U.S. EPA, 2022a). For example,
greater use of materials and energy results in potentially greater emissions of toxic chemicals and
greenhouse gases. Studies are beginning to suggest there could be a point of diminishing returns
where the economic and environmental consequences of advanced treatment begin to outweigh
the benefits of greater nutrient removal (Falk et al., 2013; Foley et al., 2010).
The Paseo Real WWTP (PR WWTP), which serves the City of Santa Fe, New Mexico, is
faced with the challenge of balancing the need for improved nutrient removal while limiting
additional environmental impacts. The city recently commissioned a Nutrient Loading and
Removal Optimization Study, which developed and evaluated several options for process
optimization and upgrading to meet more stringent effluent nutrient limitations. That study
identified reverse osmosis (RO) as the technology that would result in the lowest effluent
nutrient concentrations. However, the city has expressed the same concerns as others (Falk et al.,
2013; Foley et al., 2010) related to the cost, practicality, and coincident environmental impacts
associated with an RO system.
The objective of this study is to conduct a life cycle assessment (LCA) of the upgrade
options available to the PR WWTP. LCA is a widely accepted, systematic technique to assess
and quantify the holistic environmental aspects and potential impacts associated with individual
products, processes, or services. In 2021, EPA completed an LCA of generalized WWTP
configurations (U.S. EPA, 2021a) that demonstrated the potential for considerable increase in
environmental impacts associated with technologies designed to achieve the highest level of
nutrient removal. This study uses a similar methodology applied to an actual case study system.
The treatment configurations evaluated by this study were designed specifically for the PR
WWTP (Carollo Engineers, 2018) and are described in Table ES-1.
1 United States Environmental Protection Agency. 2022 EPA Nutrient Reduction Memorandum. Accelerating
Nutrient Pollution Reductions in the Nation's Water (April 2022).
-------
Executive Summary
Table ES-1. Summary of Treatment Scenarios Evaluated for this Study.
Proposed Scenarios
Target Eff
(m
luent Cone.
2/L)
Description
Total
Nitrogen
Total
Phosphorus
Baseline
5
1
The Baseline Scenario is the anticipated state of the
facility following implementation of all currently planned
facility upgrades, including upgraded aeration system, a
combined heat and power system, and partial effluent
diversion to the Rio Grande.
Scenario 1 -
Sidestream Filtration
4.5
0.7
Scenario 1 is the Baseline configuration with the addition
of sidestream filtration, which includes treatment of the
high nutrient concentration filtrate that is generated from
sludge dewatering processes.
Scenario 2 - Tertiary
Filtration
3
0.05
Scenario 2 is the Baseline configuration with the addition
of tertiary deep bed media filters and new chemical feed
facilities for enhanced nutrient removal.
Scenario 3 - Reverse
Osmosis
2
0.05
Scenario 3 is the Baseline configuration with the addition
of a microfiltration/reverse osmosis system downstream
of the secondary clarifiers.
Scenario 4 - Zero
Discharge
5
1
Scenario 4 assumes the same facility configurations as
the Baseline Scenario, with no discharge to the Santa Fe
River. All current effluent discharges to the Santa Fe
River would instead be diverted to the Rio Grande using
a larger pipeline than currently planned under the
Baseline Scenario. The city would continue serving its
non-potable reuse customers' needs.
This study uses 12 standard LCA metrics that describe potential environmental, energy
and climate, water, and toxicity impacts, as well as cost estimates for each configuration. Life
cycle inventories (LCIs2) of each configuration were developed in collaboration with the study
workgroup, which includes staff from EPA, the New Mexico Environment Department, the City
of Santa Fe (including staff from the PR WWTP), Carollo Engineers, and Eastern Research
Group, Inc. (ERG). Where possible, uncertainty ranges in LCI inputs were defined and used in
subsequent Monte Carlo simulations to describe ranges of uncertainty in study results. The
study's system boundary includes all relevant details of the wastewater treatment processes,
environmental releases from each process, and the supply chains associated with inputs to each
process. Study results are provided on the basis of a standard volume of water treated by each
configuration to different effluent nutrient concentration targets.
LCA results across all scenarios and metrics are provided in Figure ES-1. Results for
each metric have been standardized to a common scale of-1 to 1 by dividing results by the
maximum and minimum values across all Scenarios. Results show that Scenario 3 (Reverse
Osmosis) has the lowest eutrophication potential impacts but the highest impacts across all other
metrics. Additionally, water scarcity impacts, which consider life cycle water use as well as local
water scarcity, suggest that Reverse Osmosis would result in much greater impacts than all other
scenarios due to the brine disposal process, which renders water associated with the injected
2LCIs provide a list of all input and output flows to the system under investigation. Inputs may include raw
materials, energy or water, and outputs may include emissions to water, land or air.
-------
Executive Summary
brine unavailable for other purposes. Monte Carlo uncertainty results indicate that, within the
range of uncertainty of the treatment performance assumptions, Scenario 2 (Tertiary Filters)
eutrophication potential impacts are comparable to those of Reverse Osmosis. Figure ES-1
shows that Tertiary Filters would result in lower potential impacts across all other metrics.
A*
Energy and
Climate
4
w\/
Water
Quantity
Environment
HHPM
CED
HHNC
HHC
^ ^ Toxicity
GWP
— Baseline
SI
S3
Sidestream Filtration
Reverse Osmosis
S2
S4
Tertiary Filters
Zero Discharge
EP-Eutrophicatiori
HHC-Cancer Tox.
AP-Acidifi cation
HHNC-Non-cancer Tox.
SFP-Smog
HHPM-Particulates
CED-Energy Demand
WD-Water Depletion
FFD-Fossil Fuel
WS-Water Scarcity
GWP-Global Warming
ET-Ecotoxicity
Figure ES-1. Standardized Results from Each Study Treatment. A value of 1 (i.e., toward the
outer edge of the plot_ reflect the greatest environmental harm, while a value of -1 (i.e., toward
the center of the plot) reflects the least environmental harm.
in
-------
Executive Summary
LCA results also show that Scenario 1 (Sidestream Filtration) can achieve about a 17%
improvement in eutrophication potential relative to the Baseline Scenario, while potential
impacts across all other metrics result in increases ranging from 1% to 6% relative to the
Baseline Scenario. This suggests that, in terms of impact per unit of nutrient removed,
Sidestream Filtration may be more efficient than the other evaluated technology options, which
is also supported by a nutrient removal standardization analysis performed in the study (Section
3.5.3). Despite resulting in greater impacts across some of the other metrics, Tertiary Filters and
Reverse Osmosis result in eutrophi cation potential reductions of approximately 57% and 63%,
respectively, relative to the Baseline Scenario.
Scenario 4 (Zero Discharge), which accounts for the additional energy required to divert
the majority of PR WWTP effluent to the Rio Grande, results in similar impacts to the Baseline
Scenario for eutrophication potential (this study assumes eutrophi cation impacts of effluent
discharge do not depend on discharge location), water depletion, and water scarcity. The Zero
Discharge scenario results in slightly higher impacts than the Baseline Scenario for all other
metrics, owing to the minor increases in material and energy requirements of full effluent
diversion compared to partial effluent diversion.
Results normalization, standardization, and sensitivity analyses were performed to
contextualize the study results. Normalization is an optional step in life cycle impact assessment
that indicates the significance of impact category results by calculating their contribution to total
category impact on a regional or per capita basis. Normalized results (Section 3.5.1) show that as
a share of average U.S. per capita impacts, eutrophication potential impacts are larger than
contributions from all other impact categories, ranging from 2% to 5% for each scenario. The
water depletion category also has relatively high normalized impacts, ranging from -2% for the
Baseline Scenario and Scenario 4 (due to water reuse) to 1.2% for Scenario 3. Contributions
from other impact categories range from 0.01% to 0.42% of average per capita burdens across all
scenarios.
Standardizing impacts to units of nutrients removed (a proxy for nutrient removal
efficiency—see Section 3.5.3) showed no changes to the relative rankings of alternatives under
baseline study results but showed progressively decreasing efficiency with increasing levels of
treatment, with the largest decreases mostly occurring between Tertiary Filters and Reverse
Osmosis.
Sensitivity analyses (Section 4) examine the influence of key parameters, eutrophication
potential characterization factors, global warming potential characterization factors, electricity
grid mix, and sludge management on the environmental performance of treatment scenarios.
Compared to baseline results, sensitivity results show that relative rankings between scenarios
generally remain unchanged across the range of sensitivity assumptions; however, the magnitude
of difference in impacts between scenarios is affected. For example, the eutrophication potential
sensitivity analysis accounts for bioavailability of organic nitrogen, which is the dominant
effluent nutrient contributor to eutrophication potential impacts for the more advanced nutrient
removal scenarios (Tertiary Filtration and Reverse Osmosis). Under conditions where organic
nitrogen may be less bioavailable, eutrophication potential impacts of all scenarios are reduced,
and the relative difference between Tertiary Filters and Reverse Osmosis is lessened. Impacts of
Tertiary Filters were found to be sensitive to alum dosing. Using more alum than anticipated
iv
-------
Executive Summary
could result in water depletion impacts for Tertiary Filters comparable to Reverse Osmosis. The
electricity grid sensitivity analysis shows that if a greater fraction of solar energy were used,
impacts across all scenarios would be reduced, though reductions for eutrophication potential
and water depletion would be minor. Impacts for the particularly energy-intensive Reverse
Osmosis scenario would, for some metrics (e.g., cancer and noncancer toxicity, smog formation,
fossil fuel depletion), be more comparable to, and sometimes less than, other treatment
configurations.
Results of this study, summarized in Table ES-3, reinforce the findings of previous
research (Falk et al., 2013; U.S. EPA, 2022a), showing that increasingly advanced levels of
nutrient removal lead to improved water quality while producing greater environmental impacts
in other categories and at higher costs. Sidestream Filtration (Scenario 1) would result in small
improvements to nutrient removal with correspondingly small increases in potential
environmental impacts. Reverse Osmosis (Scenario 3) offers the greatest potential for improved
nutrient removal but does so at the expense of potentially greater environmental impacts
compared to all other scenarios being considered in this analysis. Zero Discharge (Scenario 4)
would result in comparable nutrient emissions to the Baseline Scenario and only small increases
in environmental impacts associated with diverting effluent to the Rio Grande.
Table ES-2. Summary of Study Results.
LCA
Results
SI - Sidestream
Filtration
S2 - Tertiary Filters
S3 - Reverse Osmosis
S4 - Zero Discharge
Impact
Small increases in
potential
environmental impacts
Small to moderate
increases in potential
environmental impacts
Except for
eutrophication
potential, potential
environmental impacts
generally much
greater than other
scenarios considered
Small increases in
impacts associated
with full effluent
diversion
Benefit
Small improvement to
nutrient removal
Large improvement to
nutrient removal
Largest improvement
to nutrient removal
Eutrophication
potential impacts
diverted from Santa Fe
River to Rio Grande
-------
notice
NOTICE
This document was produced by the U.S. Environmental Protection Agency (EPA). It has been
subjected to EPA's administrative review process and has been approved for publication.
Mention of trade names, technologies and processes, or commercial products does not constitute
endorsement or recommendation for use.
The facility operating information and related analyses in this document are based on data
received from the facility featured in this document. While EPA has reviewed and evaluated
these data, EPA does not assume responsibility for the accuracy of the data used in the analyses.
Neither the data used in this report nor the technology evaluations provided here nor the
conclusions or results reported in this document substitute for site-specific analysis needed when
considering the use of these technologies at other facilities.
Technology performance and variability in effluent concentrations, particularly for nutrient
removal, is affected by site-specific factors such as process design, wet weather flow, variability
in influent flow and concentrations, process control capabilities, presence of biological inhibitors
or toxics, presence of equalization tanks, sidestreams, and many other factors. In addition, a
plant's actual flow and nutrient loading relative to the design capacity could be a significant
factor that impacts performance. As such, the information in this report can be viewed as a guide
based on the investigated plant's actual operation over 36 months but should not be used to
translate performance or variability to other plants without careful consideration of the plant's
site-specific conditions.
This document is intended to be solely informational and does not impose legally binding
requirements on EPA or other U.S. federal agencies, states, local, or tribal governments, or
members of the public.
vi
-------
Acronyms and Abbreviations
Acronyms and Abbreviations
AP
Acidification potential
AS
Activated sludge
AWARE
Available WAter REmaining method
AZNM
Arizona/New Mexico eGRID subregion
BFP
Belt filter press
BOD
Biochemical oxygen demand
CBOD
Carbonaceous biochemical oxygen demand
CED
Cumulative energy demand
CHP
Combined heat and power
CO
Carbon monoxide
COD
Chemical oxygen demand
DAF
Dissolved air flotation
DAP
Diammonium phosphate
DBP
Disinfection byproduct
DQI
Data quality indicator
EF
Emission factor
eGRID
Emissions & Generation Resource Integrated Database
eLCI
Electricity LCI
EON
Effluent organic nitrogen
EP
Eutrophication potential
EPA
U.S. Environmental Protection Agency
ERG
Eastern Research Group, Inc.
ET
Ecotoxicity
FP
Formation potential
GHG
Greenhouse gas
GT
Gravity thickener
GWP
Global warming potential
H2S
Hydrogen sulfide
HAB
Harmful algal blooms
HH
Human health
HHC
Human health cancer potential
HHNC
Human health noncancer potential
HHPM
Human health particulate matter formation potential
HHV
High heating value
ICE
Internal combustion engine
ISO
International Organization for Standardization
LCA
Life cycle assessment
LCCA
Life cycle cost analysis
LCI
Life cycle inventory
LCIA
Life cycle impact assessment
3
m
cubic meter
MBR
Membrane bioreactor
MCF
Methane conversion factor
MF
Microfilter
MGD
Million gallons per day
N
Nitrogen
NM
New Mexico
NMED
New Mexico Environment Department
-------
Acronyms and Abbreviations
NNC
Numeric nutrient criteria
NOM
Natural organic matter
NOx
Nitrogen oxides
ORD
U.S. EPA Office of Research and Development
P
Phosphorus
PM
Particulate matter
PPCP
Pharmaceuticals and personal care products
PR
Paseo Real
QAPP
Quality Assurance Project Plan
RO
Reverse osmosis
SFP
Smog formation potential
SFR
Santa Fe River
TKN
Total Kjeldahl nitrogen
TN
Total nitrogen (Total Kjeldahl Nitrogen + Nitrate/Nitrite)
TP
Total phosphorus
TRACI
Tool for the Reduction and Assessment of Chemical and Environmental
Impacts
UF
Ultrafiltration
UIC
Underground injection control
UNFCCC
United Nations Framework Convention on Climate Change
US LCI
United States Life Cycle Inventory Database
UV
Ultraviolet
VOCs
Volatile organic compounds
WD
Water depletion
WECC
Western Electricity Coordinating Council
WQS
Water quality standard
WS
Water scarcity
WWT
Wastewater treatment
WWTP
Wastewater treatment plant
viii
-------
Acknowledgements
Acknowledgements
This work was overseen by the following workgroup members:
• EPA Office of Water, Office of Science and Technology: Mario Sengco, Jennifer Brundage,
Danielle Anderson, Jim Keating, Melissa Dreyfus, Kelly Gravuer, Menchu Martinez,
Anthony Tripp
• EPA Region 6: Forrest John, Russell Nelson, Jasmin Diaz-Lopez
• EPA Office of Research and Development: Cissy Ma
• New Mexico Environment Department: Shelly Lemon, Heidi Henderson, Jennifer Fullam,
Kristopher Barrios, Sarah Holcomb, Susan Lucas-Kamat
• City of Santa Fe and Paseo Real Wastewater Treatment Plant: Michael Dozier, Efren
Morales, Shannon Jones, John Del Mar, Brian Snyder, Angela Bolton, Sandra Gabaldon
• Carollo Engineers (contractors for the City of Santa Fe): Becky Luna, Tanja Rauch-
Williams, Andrew Henderson
Contractor support to EPA was provided by Eastern Research Group (ERG): Sam Arden, Ben
Morelli, and Sara Cashman. This work was conducted under EPA Contract No. EP-C-17-041,
Work Assignment 4-77.
IX
-------
Table of Contents
TABLE OF CONTENTS
Page
1. Introduction and Objective 1-1
1.1 Case Study System 1-2
1.2 Paseo Real WWTP Background 1-4
1.3 Wastewater Treatment Scenarios 1-4
1.4 Metrics and Life Cycle Impact Assessment 1-5
2. LCA Methodology 2-1
2.1 Goal and Scope Definition 2-1
2.1.1 Functional Unit 2-1
2.1.2 System Definition and Boundaries 2-1
2.2 Life Cycle Inventory 2-2
2.2.1 Introduction 2-2
2.2.2 Foreground LCI Data 2-5
2.2.3 LCI Background Data Sources 2-14
2.2.4 Process GHG Emission Estimation Methodologies 2-19
2.2.5 LCI Limitations 2-19
2.3 Life Cycle Impact Assessment Model 2-20
2.3.1 LCIA Limitations 2-21
3. Life Cycle Impact Baseline Results 3-1
3.1 Environment 3-3
3.1.1 Eutrophication Potential 3-3
3.1.2 Acidification Potential 3-5
3.1.3 Smog Formation Potential 3-7
3.2 Energy and Climate 3-8
3.2.1 Cumulative Energy Demand 3-8
3.2.2 Fossil Fuel Depletion 3-9
3.2.3 Global Warming Potential 3-10
3.3 Water 3-12
3.3.1 Water Depletion 3-12
3.3.2 Water Scarcity 3-13
3.4 Toxicity 3-15
3.4.1 Ecotoxicity 3-15
3.4.2 Human Health—Particulate Matter Formation 3-17
3.4.3 Human Health Toxicity—Cancer Potential 3-18
3.4.4 Human Health Toxicity—Noncancer Potential 3-20
3.5 Normalization and Standardization 3-21
3.5.1 Standard Normalization 3-21
3.5.2 S anta F e GHG Inventory 3-22
x
-------
Table of Contents
TABLE OF CONTENTS (Continued)
Page
3.5.3 Results Standardized to Nutrient Removal 3-23
4. Sensitivity Analysis Results and Discussion 4-1
4.1 Important Parameters 4-1
4.1.1 Main Electricity 4-2
4.1.2 Compost Emissions 4-2
4.1.3 Biogas Production 4-3
4.1.4 Nutrient Emissions 4-4
4.1.5 Land Application Emissions 4-5
4.1.6 Biological GHG Emissions 4-5
4.1.7 Scenario 2 (Tertiary Filters) Alum 4-6
4.1.8 Scenario 3 (RO) Electricity 4-6
4.2 Eutrophication Potential 4-6
4.3 Global Warming Potential Characterization Factors 4-11
4.4 Electricity Grid Mix 4-13
4.5 Sludge Management 4-16
5. Conclusions 5-1
6. References 6-1
XI
-------
Appendix A
Appendix B
Appendix C
Appendix D
Appendix E
Appendix F
APPENDICES
Life Cycle Impact Assessment
Life Cycle Inventory Data
LCIA Results
Data Quality Assessment
Determination of Metals Removal Performance
Parameter Sensitivity Results
-------
List of Tables
LIST OF TABLES
Page
Table ES-1. Summary of Treatment Scenarios Evaluated for this Study ii
Table ES-2. Summary of Study Results v
Table 1-1. Total Nitrogen (TN) and Total Phosphorus (TP) Causal Thresholds by Site
Class 1-2
Table 1-2. Proposed Study Scenarios 1-5
Table 1-3. Metrics Included in the LCA 1-7
Table 2-1. Foreground Unit Processes Included in Each Wastewater Treatment
Configuration 2-5
Table 2-2. Influent and Estimated Effluent Water Quality for each Wastewater Treatment
Configuration 2-7
Table 2-3. Allocation of Biogas to Onsite Combustion Processes 2-10
Table 2-4. Background Unit Process Data Sources.a 2-14
Table 2-5. Arizona/New Mexico Average Electrical Grid Mix 2-16
Table 2-6. Life Cycle Inventory Data for Biogas Cleaning 2-17
Table 2-7. Characteristics of Digestate, Yard Waste, and Finished Compost 2-17
Table 2-8. Key Landfill Modeling Parameters 2-18
Table 3-1. Description of Impact Contribution Categories 3-1
Table 3-2. Summary of average annual COD and nutrient discharges across treatment
scenarios 3-5
Table 3-3. U.S. Per Capita Normalization Factors (Lippiatt et al., 2013; Ryberg et al.,
2014) 3-21
Table 3-4. Normalized impact results, expressed as the percent of per capita impacts
allocated to wastewater treatment 3-22
Table 3-5. Summary of treatment scenario GHG emissions, compared to Santa Fe per
capita emissions 3-23
Table 3-6. Nutrient removal performance of treatment scenarios expressed as total
nitrogen (TN) removal, total phosphorus (TP) removal, and total nitrogen
equivalents (N eq.) removal 3-23
Table 4-1. Summary of Measured Effluent Organic Nitrogen Bioavailability 4-7
Table 4-2. Comparison of IPCC Assessment Report 4 and Assessment Report 5 20- and
100-year characterization factors 4-12
Table 4-3. Global Warming Potential (GWP) Sensitivity Analysis Results 4-12
Table 4-4. Change in impacts as a function of electricity grid 4-15
Xlll
-------
List of Tables
LIST OF TABLES (Continued)
Page
Table 4-5. Change in impacts as a function of solids handling assumptions 4-17
Table 5-1. Standardized Baseline Impacts for Each Study Treatment Scenario 5-1
xiv
-------
List of Figures
LIST OF FIGURES
Page
Figure ES-1. Standardized Results from Each Study Treatment iii
Figure 2-1. Subset of LCA model structure with example unit process inputs and outputs 2-4
Figure 2-2. System diagram of the existing PR WWTP following upgrades to the
biological treatment system 2-10
Figure 2-3. System diagram of Scenario 1 - Sidestream Filtration 2-11
Figure 2-4. System diagram of Scenario 2 - Tertiary Filters 2-12
Figure 2-5. System diagram of Scenario 3 - Reverse Osmosis 2-13
Figure 2-6. System diagram of Scenario 4 - Zero Discharge to Santa Fe River 2-14
Figure 3-1. Summary of baseline LCIA results for the Baseline Scenario and Scenarios
1-4 (S1-S4) 3-1
Figure 3-2. Eutrophication potential results for each treatment scenario, including
uncertainty ranges as the 5th and 95th percentile results from Monte Carlo
simulations 3-4
Figure 3-3. Summary of annual nutrient mass discharges across treatment scenarios.
COD not shown to not overwhelm nutrient visibility 3-5
Figure 3-4. Acidification potential results for each treatment scenario, including
uncertainty ranges as the 5th and 95th percentile results from Monte Carlo
simulations 3-6
Figure 3-5. Smog formation potential results for each treatment scenario, including
uncertainty ranges as the 5th and 95th percentile results from Monte Carlo
simulations 3-8
Figure 3-6. Cumulative energy demand results for each treatment scenario, including
uncertainty ranges as the 5th and 95th percentile results from Monte Carlo
simulations 3-9
Figure 3-7. Fossil fuel depletion results for each treatment scenario, including uncertainty
ranges as the 5th and 95th percentile results from Monte Carlo simulations 3-10
Figure 3-8. Global warming potential results for each treatment scenario, including
uncertainty ranges as the 5th and 95th percentile results from Monte Carlo
simulations 3-11
Figure 3-9. Water depletion results for each treatment scenario, including uncertainty
ranges as the 5th and 95th percentile results from Monte Carlo simulations 3-13
Figure 3-10. Water scarcity results for each treatment scenario 3-15
Figure 3-11. Ecotoxicity results for each treatment scenario, including uncertainty ranges
as the 5th and 95th percentile results from Monte Carlo simulations 3-16
xv
-------
Introduction and Objective
LIST OF FIGURES (Continued)
Page
Figure 3-12. Human health—particulate matter formation results for each treatment
scenario, including uncertainty ranges as the 5th and 95th percentile results
from Monte Carlo simulations 3-18
Figure 3-13. Human health toxicity—cancer potential results for each treatment scenario,
including uncertainty ranges as the 5th and 95th percentile results from Monte
Carlo simulations 3-19
Figure 3-14. Human health toxicity—noncancer potential results for each treatment
scenario, including uncertainty ranges as the 5th and 95th percentile results
from Monte Carlo simulations 3-21
Figure 3-15. Impact results standardized to 1 cubic meter of wastewater treated 3-24
Figure 4-1. Sensitivity of top two important parameters for each impact category 4-1
Figure 4-2. Eutrophication potential sensitivity analysis results including uncertainty
ranges as the 5th and 95th percentile results from Monte Carlo simulations 4-9
Figure 4-3. Sensitivity of global warming potential results to selection of characterization
factors 4-13
Figure 4-4. Illustration of electricity grid sensitivity analyses 4-15
Figure 4-5. Illustration of solids handling sensitivity analyses 4-17
xvi
-------
Section 1: Introduction and Objective
1. Introduction and Objective
Nutrient pollution of waterbodies across the United States is one of the most pervasive
environmental issues facing the country today. Whether in lakes or reservoirs, rivers or streams,
estuaries or marine coastal waters, the human health, environmental, and economic impacts from
excessive amounts of nitrogen (N) and phosphorus (P) continue to rise every year. Communities
struggle with nutrient-fueled harmful algal blooms (HABs), which produce toxins that can sicken
people and pets, contaminate food and drinking water sources, destroy aquatic life, and disrupt
the balance of natural ecosystems. HABs can raise the cost of drinking water treatment, depress
property values, close beaches and fishing areas, and negatively affect the health and livelihood
of many Americans (U.S. EPA, 2015). Global climate change is only expected to exacerbate
eutrophication even as federal, state, and local governments struggle to address the sources of
nutrient pollution (USGCRP, 2016).
In partnership with states, tribes, and other federal agencies, the U.S. Environmental
Protection Agency (EPA) has led the effort to address nutrient pollution by helping states
prioritize waters; providing scientific and technical assistance with developing water quality
standards for total nitrogen (TN) and total phosphorus (TP); and helping to guide implementation
of nutrient criteria including total maximum daily loads for impaired waters and water quality-
based effluent limits for point source dischargers. Given the urgency of the problem, EPA's
Office of Water plans to accelerate progress in controlling nutrient pollution in the nation's
waters by scaling up existing, foundational approaches and more broadly deploying new data
assessments, tools, financing approaches, and implementation strategies (U.S. EPA, 2022b).
Additionally, EPA plans to integrate the objectives of both the Safe Drinking Water Act and
Clean Water Act in a One Water approach to find durable solutions to the challenges and costs
associated with reducing nutrient pollution. At the same time, EPA foresees incorporating
promising innovations, creative partnerships, and unprecedented opportunities to invest in clean
and safe water in the Bipartisan Infrastructure Law to accelerate progress in reducing nutrient
pollution/
EPA has assisted states in translating their narrative criteria to protect waters from
eutrophi cation.4 In New Mexico, for example, the state's water quality standards (WQS)
regulations5 include a narrative criterion to protect aquatic life from nutrient conditions that
contribute to production of undesirable or nuisance aquatic life. The criterion states, "Plant
nutrients from other than natural causes shall not be present in concentrations that will produce
undesirable aquatic life or result in a dominance of nuisance species in surface waters of the
state" (20.6.4.13.E New Mexico Administrative Code). In other words, non-zero nutrient
concentrations that will not produce undesirable effects are acceptable. The state translates this
narrative criterion using numeric threshold values in its Comprehensive Assessment and Listing
3 For more information, see the 2022 EPA Nutrient Reduction Memorandum website.
4 "Eutrophication is defined as an increase in nutrient input to surface waters to the extent of over enrichment, with a
corresponding increase in primary productivity and related negative effects" (Serediak et al., 2014).
5 Codified at 20.6.4 NMAC.
1-1
-------
Section 1: Introduction and Objective
Methodology (CALM),6 which are based on reference conditions and applied to specific site
classes in perennial, wadable streams. These numeric thresholds then become the basis for
reasonable potential analyses and the development of water quality-based effluent limits in
permits for point source dischargers. In most cases, this means potentially more stringent effluent
limits for NM dischargers with the implementation of numeric thresholds requiring additional
treatment to meet new limits.
Recently, operators and other stakeholders have expressed concern that there may be
significant environmental and health implications when facilities move towards treatment
technologies that remove more TN and TP to attain very low nutrient targets (e.g., Falk et al.,
2013; U.S. EPA, 2022a). For example, potential impacts other than eutrophication are associated
with greater use of chemicals, disposal of biosolids and brine (e.g., from reverse osmosis [RO]),
increased energy demands, and greater release of greenhouse gases (GHGs). Studies in other
countries also suggest a point of diminishing returns where the economic and environmental
consequences begin to outweigh the benefits (e.g., Foley et al. 2010; Falk et al. 2013).
1.1 Case Study System
The Paseo Real (PR) wastewater treatment plant (WWTP), which is owned and operated
by the City of Santa Fe, New Mexico, is one such facility challenged with balancing the need for
improved nutrient removal while limiting additional environmental impacts. The PR WWTP
discharges its effluent into the Santa Fe River, which is listed as "impaired" for nutrients and
Escherichia coli (E. coli) bacteria. During certain parts of the year, flow in the Santa Fe River is
almost entirely composed of discharge from the PR WWTP, which means the river's nutrient
dynamics are highly sensitive to effluent concentrations at the PR WWTP. The New Mexico
Environment Department (NMED) has developed numeric TN and TP thresholds to translate its
narrative nutrient criteria, as shown in Table 1-1.7"8 The PR WWTP discharges to the Cienega
Creek to Santa Fe WWTP portion of the Santa Fe River, which is characterized as site class TN
Moderate and TP Flat-Moderate.
Table 1-1. Total Nitrogen (TN) and Total Phosphorus (TP) Causal Thresholds by Site Class.
Parameter and Site
Class
Site Median Threshold
(90th quantile) (mg/L)
TN Flat
0.69
TN Moderate
0.42
TN Steep
0.30
TP High-Volcanic
0.105
TP Flat-Moderate
0.061
TP Steep
0.030
Note: Thresholds that apply to Paseo Real WWTP
are italicized.
6 2021 CALM, https://www.env.nm.gov/surface-water-qualitv/calm/
7 Table 3, P.8 of Appendix C of NMED's 2021 CALM https://www.env.mn.gov/surface-water-qualitv/calm/.
8 New Mexico and EPA apply the thresholds for permitting purposes as 30-day average values.
1-2
-------
Section 1: Introduction and Objective
Although these thresholds are not WWTP effluent criteria, surface waters, and
particularly those that are effluent-dominated, will generally not meet these thresholds if effluent
nutrient concentrations are much higher. Moreover, these thresholds are lower than most
facilities in New Mexico can currently achieve end-of-pipe, including the PR WWTP. This has
led the City of Santa Fe and the State of New Mexico to evaluate operational and technological
options for improving removal of nitrogen and phosphorus from the PR WWTP's effluent.
In 2018, the city completed a Nutrient Loading and Removal Optimization Study to
examine the facility's options for process optimization and upgrading to meet one of several
effluent "tiers"9 for N and P removal (Carollo Engineers, 2018). The study identified several
options to reduce effluent nutrient discharges. The identified options include optimization of the
existing biological process and treatment of the filtrate return flow (Tier 1); installation of a
membrane bioreactor (MBR) with chemical addition (Tier 2); installation of tertiary treatment
with chemical addition (Tier 3); and installation of RO (Tier 4), which the study estimated could
achieve the lowest effluent concentrations of all the options investigated. All proposed options
would reduce nutrient releases relative to the 2018 status quo but would vary in their cost and
ability to achieve the numeric nutrient thresholds for the Santa Fe River. Capital cost estimates
provided in the Nutrient Loading and Removal Optimization Study ranged from $8.6 million for
Tier 1 to $87 million for Tier 4.
While RO comes closest to achieving New Mexico's numeric nutrient thresholds, the city
has expressed the same concerns as others (Falk et al., 2013; Foley et al., 2010) related to the
cost, practicality, and cross-media environmental impacts10 of an RO system.
The objective of this study is to conduct a life cycle assessment (LCA) on the PR WWTP
in order to ascertain and quantify the potential environmental harms and benefits of various
options for improving the removal of nutrients. LCA is a widely accepted, standardized,
systematic technique to assess the holistic environmental aspects and potential impacts
associated with individual products, processes, or services that can be applied to these kinds of
issues. Often referred to as a "cradle-to-grave" analysis, LCAs reveal the presence of
environmental trade-offs between "comparable" options, which indicates that no single option is
typically capable of providing the best potential environmental performance across diverse
impact categories. In 2021, EPA completed an LCA using generalized WWTP configurations
titled, Life Cycle and Cost Assessments of Nutrient Removal Technologies in Wastewater
Treatment Plants (U.S. EPA, 2021a). That study demonstrated the potential for a considerable
increase in cross-media environmental impacts (e.g., energy demand, climate change potential)
for technologies and treatment configurations designed to achieve the highest levels of nutrient
removal. Building upon this earlier work, this current LCA study will provide data that can be
useful to local, state, and federal decision-makers and other stakeholders make informed choices
based on environmental considerations. These choices could potentially include informing
treatment technology selection, balancing nutrient-water-energy nexus, future development of
9 Note: "tiers," as used throughout this document, refers specifically to the treatment levels developed in Carollo
Engineers 2018. This is not related to the term "tiers" as used in the context of antidegradation in water quality
standards.
10 "Cross-media" refers to the broad scope of LCA studies, considering the whole environment and not a single
media (e.g., air, water, soil) or impact category.
1-3
-------
Section 1: Introduction and Objective
revised water quality standards such as discharger-specific nutrient "temporary standards,"11
revisions to the designated use, or revisions to site-specific criteria12 for discharge into the Santa
Fe River by the PR WWTP. This report only focuses on the technical analysis (i.e., the life cycle
assessment itself) and does not address the policy implications of the results or future regulatory
processes.
1.2 Paseo Real WWTP Background
The PR WWTP has been in operation since 1963, discharging treated effluent to the
Santa Fe River. Its current design capacity is 13 million gallons per day (MGD) average
maximum month flow or 12 MGD average day annual flow, with an average annual flow of 4.85
MGD. It serves approximately 85,000 residential customers, in addition to an unknown quantity
of tourists and visitors (Carollo Engineers, 2018). In 2020, in an effort to update outdated
equipment and improve their level of treatment, the facility began to implement a series of
relatively low-cost upgrades including the installation of an upgraded aeration system with more
energy-efficient blowers to allow for better control of dissolved oxygen levels. At the same time,
the City of Santa Fe was in the process of installing a combined heat and power system to
expand energy recovery from the biogas produced in the anaerobic digesters. The facility is also
reviewing other options that provide a trade-off between nutrient removal and factors such as
cost, operational complexity, and infrastructure requirements (see Section 1.3).
In addition to operational upgrades, the facility is planning to implement partial diversion
of plant effluent from the current outfall on the Santa Fe River to a new outfall on the Rio
Grande. This would allow the city to exchange PR WWTP effluent for additional water
withdrawals from the Rio Grande without reducing flow in the Rio Grande. The additional
diversions for potable water supply would help accommodate anticipated population growth and
reduce water supply shortages under projected climate change conditions. The discharge of PR
WWTP effluent to the Rio Grande is expected to begin operation in about five years.
The facility also sends a portion of its treated effluent to customers in the city to be used
as non-potable water.
1.3 Wastewater Treatment Scenarios
The wastewater treatment scenarios proposed in this study, while mostly derived from the
Nutrient Loading and Removal Optimization Study (Carollo Engineers, 2018), were refined in
consultation with the project workgroup.13 The workgroup proposed scenarios based on their
relevance to the PR WWTP and their ability to produce differentiated effluent quality and
potential environmental impacts. The membrane bioreactor (MBR) with chemical addition (Tier
2) from the Nutrient Loading and Removal Optimization Study was excluded from the list due to
11 As provided in by 20.6.4.10. NMAC. which is equivalent to a "water quality standard variance" under federal
regulations at 40 CFR § 131.14.
12 As provided in 20.6.4.10 NMAC and 40 CFR 131.11(b¥iD.
13 The project workgroup consists of members from EPA, the State of New Mexico, the City of Santa Fe, Eastern
Research Group (ERG) (contractor to EPA), and Carollo Engineers (Carollo) (contractor to the City of Santa Fe).
See Acknowledgements for details.
1-4
-------
Section 1: Introduction and Objective
it producing similar effluent quality to the installation of tertiary treatment with chemical
addition (Tier 3) (now Scenario 2 in Table 1-2) but with a higher cost. The final list of proposed
scenarios is provided in Table 1-2, with individual scenarios discussed further in Section 2.2. It is
important to note that these scenarios are not sequential but standalone alternatives. For example,
Scenario 2 is not the result of Baseline plus sidestream filtration (Scenario 1) plus tertiary
filtration. Instead, each scenario is a unique process or combination of processes that are added
separately to the Baseline configuration.
Table 1-2. Proposed Study Scenarios.
Scenario
Effluent Cone. (mg/L)a
Description
Total
Nitrogen
Total
Phosphorus
Existing Site Thresholds
Thresholds
0.42
0.061
See Table 1-1.
Existing Conditions
Status Quob
5-7
1-5
Based on the analysis of effluent concentrations discussed in
Carollo Engineers (2018).
Proposed Scenarios
Baseline0
5
1
The Baseline Scenario refers to the anticipated state of the
facility following implementation of all currently planned
facility upgrades and partial effluent diversion to the Rio
Grande.
Scenario 1 -
Sidestream Filtration
4.5
0.7
Scenario 1 refers to the Baseline configuration with the
addition of filtrate return flow treatment.
Scenario 2 - Tertiary
Filtration
3
0.05
Scenario 2 includes the Baseline configuration with the
addition of tertiary deep bed media filters and new chemical
feed facilities. Note that Scenario 2 is equivalent to Tier 3 of
Carollo Engineers (2018).
Scenario 3 - Reverse
Osmosis
2
0.05
Scenario 3 includes the Baseline configuration with the
addition of a microfiltration/reverse osmosis system
downstream of the secondary clarifiers. Note that Scenario 3
is equivalent to Tier 4 of Carollo Engineers (2018).
Scenario 4 - Zero
Discharge (to Santa
Fe River)
5
1
Scenario 4 assumes the same facility configurations as the
Baseline Scenario, with no discharge to the Santa Fe River.
All current effluent discharges to the Santa Fe River would
instead be diverted to the Rio Grande, and the city would
continue serving its non-potable reuse customers' needs.
a Concentrations are estimates of average conditions as provided by Carollo on April 9, 2021. See Table 2-2 for
more detailed information.
bEffluent concentration ranges from Table 2-3 of Santa Fe, 2018.
0 When capitalized, "Baseline" refers to the Baseline Scenario. When not capitalized, "baseline" refers to baseline
LCA results.
1.4 Metrics and Life Cycle Impact Assessment
Table 1-3 summarizes the metrics assessed for each system configuration, together with
the method and units used to characterize each. Abbreviations are included for each metric,
which are used throughout this report.
1-5
-------
Section 1: Introduction and Objective
Most of the life cycle impact assessment (LCIA) metrics are estimated using EPA's Tool
for Reduction and Assessment of Chemicals and Other Environmental Impacts (TRACI) version
2.1 (Bare, 2012; Bare, 2011). TRACI includes a compilation of methods representing current
best practices for estimating ecosystem impacts based on U.S. conditions in conjunction with
information from life cycle inventory (LCI) models. Global warming potential (GWP) is
estimated in the baseline results using the 100-year characterization factors provided by the
Intergovernmental Panel on Climate Change (IPCC) Fifth Assessment Report (IPCC, 2013). A
sensitivity analysis is presented in Section 4.2 using 20-year GWPs. In addition to TRACI, the
ReCiPe LCIA method is used to characterize water depletion and fossil resource use (Huijbregts
et al., 2017), impacts which are not included in the current version of TRACI. Water scarcity is
evaluated in terms of relative water stress related to water withdrawal using the Available WAter
REmaining (AWARE) Method.14 Cumulative energy demand, including the energy content of
all non-renewable and renewable energy resources extracted throughout the supply chains
associated with each configuration, is estimated using a method adapted from one provided by
the Ecoinvent Centre (Ecoinvent Centre, 2010). Cumulative energy demand is an aggregated
reporting of LCI flows associated with energy inputs and, unlike LCIA categories, does not
attempt to characterize potential environmental impact.
Toxicity impacts, including human health cancer and noncancer potential and ecotoxicity
of waste streams generated under each scenario, are calculated using the USEtox™ model,
which is incorporated in TRACI 2.1. EPA's report Life Cycle and Cost Assessments of Nutrient
Removal Technologies in Wastewater Treatment Plants (U.S. EPA, 2021a) included the
evaluation of WWTP-based toxicity impacts derived from metals, disinfection byproducts
(DBPs), and trace organics. Like the current study, the primary goal of the prior study was to
perform an LCA of WWTPs that provided different levels of nutrient removal; the inclusion of
metals, DBPs, and trace organics was done to quantify those systems' ancillary impacts on non-
nutrient water quality parameters. Across a range of different treatment systems, metals were
shown to have the largest effect on toxicity impact results. Impacts from DBPs were found to be
the next most influential (albeit to a far lesser extent than metals), however, the PR WWTP does
not chlorinate its effluent, which minimizes the potential for DBP formation. Impacts from toxic
organics were found to be small in comparison to metals. Moreover, no site-specific data on
toxic organic concentrations are available at the PR WWTP. Therefore, this study only evaluates
toxicity impacts derived from metals in order to quantify ancillary impacts or benefits of the
study scenarios, following the methodology developed in EPA's nutrient removal LCA report
(U.S. EPA, 2021a).
In an LCA, environmental impacts are a function of various air and water emissions (e.g.,
nitrous oxide) and characterization of those emissions into a common unit that is representative
of a potential to cause impact (e.g., nitrous oxide expressed in carbon dioxide equivalents with
the potential to cause global warming). Emissions can occur in different environments around the
world—they can come from anywhere in upstream supply chain and production processes or
they can come from the study system itself. Characterization factors must be able to account for
this generality and be geographically or environmentally specific enough to reasonably capture
14 AWARE scores can be found at the Water Use in Life Cycle Assessment (WULCA) website: httos://wulca-
waterlca. org/aware/.
1-6
-------
Section 1: Introduction and Objective
the potential for impact. For example, the metrics included in this study quantify potential
impacts that can range in geographic scale from global (e.g., GWP and fossil fuel depletion
potential) to regional (e.g., smog formation potential, eutrophication potential).
The geographic scale of impacts therefore varies and is not always clearly defined in the
LCA. For example, impact categories modeled using EPA's TRACI method rely on U.S. average
characterization factors. This means that even though emissions can occur at the local (e.g.,
burning natural gas at a facility), regional (e.g., burning natural gas at a power plant that feeds
into the regional electricity grid), or national (e.g., burning natural gas at multiple factories that
manufacture components of a WWTP) scale, the characterization factors used to translate those
emissions to potential impacts assume national average conditions. Cumulative energy demand
and fossil fuel depletion are inventory metrics that are largely domestic, as the majority of the
U.S.'s energy supplies are sourced internally. Water scarcity characterization factors are
determined at the watershed15 level, which is the smallest scale of all Table 1-3 metrics. The
three toxicity categories utilize global characterization factors with detailed context information
such as "urban air," "rural air," or "indoor air." These contexts communicate information related
to human exposure potential of emissions, providing an indirect means of modeling regional
impact potential. Additional discussion regarding development methods of each metric is
included in Appendix A, while additional discussion of results and their geographic context is
provided in Sections 3 and 4.
Table 1-3. Metrics Included in the LCA.
Metric
Abb-
reviation
Method
Unit3
Description
Eutrophication
Potential
EP
TRACI 2.1
kg N eq.
Assesses impacts from excessive load of
macro-nutrients to the environment. Important
emissions include NH;,. COD and BOD, and
N and P compounds. The influence of each
compound is translated to an equivalent
quantity of nitrogen.
Acidification
Potential
AP
TRACI 2.1
kg S02 eq.
Quantifies the acidifying effect of substances
on their enviromnent. Important emissions:
S02, NOx, NH3, HC1, HF, H2S.
Cumulative
Energy Demand
CED
Ecoinvent
MJ-eq.
Measures the total energy from point of
extraction in nature; results include both
renewable and non-renewable energy sources.
Global
Warming
Potential
GWP
IPCC
kg CO2 eq.
Represents the heat-trapping capacity of
greenhouse gases over a 100-year time
horizon. Important emissions: CO2, CH4, N20.
Fossil Fuel
Depletion
FFD
ReCiPe
kg oil eq.
Captures the consumption of fossil fuels,
primarily coal, natural gas, and crude oil. All
fuels are standardized to kg oil eq based on
the heating value of the fossil fuel.
15 Watershed boundaries are based on a global dataset and are unique to the method. They do not necessarily
correspond with a specific Hydrologic Unit Code level. For additional information on method development, see
Boulay et al. (2018) and Miiller Schmied et al. (2014).
1-7
-------
Section 1: Introduction and Objective
Metric
Abb-
reviation
Method
Unit8
Description
Smog
Formation
Potential
SFP
TRACI 2.1
kg O3 eq.
Determines the formation of reactive
substances (e.g., tropospheric ozone) that
cause harm to human health and vegetation.
Important emissions: NOx, BTX, NMVOC,
CH4, C2H6, C4H10, C3H8, CY,H 14. acetylene,
EtOH, formaldehyde.
Human
Health—
Particulate
Matter
Formation
HHPM
TRACI 2.1
kg PM2.5 eq.
Results in health impacts such as effects on
breathing and respiratory systems, damage to
lung tissue, and other human health concerns.
Primary pollutants (including PM2.5) and
secondary pollutants (e.g., SOx and NOx) lead
to particulate matter formation.
Human Health
Toxicity—
Cancer Potential
HHC
USEtox™
2.02
CTUh
The comparative toxic unit (CTU)
characterizes the probable increase in cancer
related morbidity (from inhalation or
ingestion) for the total human population per
unit mass of chemical emitted.
Human Health
Toxicity—
Noncancer
Potential
HHNC
USEtox™
2.02
CTUh
A CTU for noncancer characterizes the
probable increase in noncancer related
morbidity (from inhalation or ingestion) for
the total human population per unit mass of
chemical emitted.
Ecotoxicity
ET
USEtox™
2.02
CTUe
Assesses potential fate, exposure, and effect of
chemicals on the enviromnent. Like the
human toxicity category, the CTUe unit
assesses the potential fraction of species
affected (i.e., disappearing) per unit mass of
chemical emitted.
Water Scarcity
WS
AWARE
m3 world
equivalents
Scales water depletion results by a range of
0.1 (no water stress at location of withdrawal)
to 100 (very high water stress). The water
stress factors are based on the available water
remaining in a watershed after the demands of
humans and the aquatic ecosystem have been
met.
Water Depletion
WD
ReCiPe
m3
Freshwater withdrawals which are evaporated,
incorporated into products and waste,
transferred to different watersheds, or
disposed into the sea after usage.
Table abbreviations: BOD = biochemical oxygen demand; BTX = aromatic hydrocarbons including benzene,
toluene and xylene isomers; CH4 = methane; C2H6 = ethane; C4H10 = butane; C3H8 = propane; CY,H 14 = hexane;
CO2 = carbon dioxide; COD = chemical oxygen demand; CTUe = comparative toxicity units for enviromnent;
CTUh = comparative toxicity units for humans; eq. = equivalents; EtOH = ethanol; HC1 = hydrochloric acid; HF =
hydrofluoric acid; H2S = hydrogen sulfide; m3 = cubic meter; MJ = megajoules; N = nitrogen; NH3 = ammonia;
NMVOC = non-methane volatile organic compounds; NOx = Nitrogen oxides; N20 = nitrous oxide; O3 = ozone; P
= phosphorus; PM2.5 = particulate matter 2.5; SO2 = sulfur dioxide; SOx = sulfur oxides.
Table Data Source: LCIA characterization factors were drawn from an EPA effort to harmonize flows for the
Federal LCA Commons: https://www.lcacommons.gov/lcia-methods-without-flows.
a Equivalents refers to characterized impact results, where all pollutants have been transformed to a single unit, or
reference substance (e.g., nitrogen equivalents for eutrophication potential), to be on a consistent basis in terms of
their contribution to category impacts.
1-8
-------
Section 1: Introduction and Objective
1-9
-------
Section 2: LCA Methodology
2. LCA Methodology
This study design follows the guidelines for an LCA provided by the International
Organization for Standardization (ISO) standards titled Environmental management—Life cycle
assessment—Principles and framework (14040) (ISO, 2006a) and Environmental management—
Life cycle assessment—Requirements and guidelines (14044) (ISO, 2006b). The following
subsections describe the scope of the PR WWTP study and the functional unit (defined below)
used for comparison, as well as the system boundaries, inventory data, and modeling procedure.
2.1 Goal and Scope Definition
2.1.1 Functional Unit
A functional unit is a "quantity of interest" that provides the basis for comparing results
in an LCA (e.g., a gallon of treated wastewater). The key consideration in selecting a functional
unit is to ensure the wastewater treatment configurations are compared based on equivalent
performance. In other words, an appropriate functional unit allows for an "apples-to-apples"
comparison. The primary functional unit for this study is the treatment of a cubic meter (m3) of
municipal wastewater such that it meets one of several effluent quality targets. Differentiated
effluent qualities are a critical component of the analysis and will be captured in the reported
environmental impact results, leading to differentiated environmental performance. Other
functional units are used for comparison purposes and are discussed further in Section 3.5.
2.1.2 System Definition and Boundaries
The system boundary includes all relevant details of the wastewater treatment processes,
environmental releases from each process, and the supply chains associated with the inputs to
each process. The analysis will estimate the impacts of electricity consumption using the
electricity mix of Arizona and New Mexico's Emissions & Generation Resource Integrated
Database (eGRID) subregion (US EPA, 2020). Chemical use associated with system operation
and periodic cleaning of equipment (e.g., membranes) are within the system boundary. The
analysis also includes impacts associated with consumable materials used in the filter systems.
Environmental impacts associated with release of effluent to the receiving water and brine
disposal are also considered.
Production of the influent and the wastewater collection system are excluded from the
system boundaries. It is assumed that these elements would be equivalent for all examined
treatment configurations and, therefore, can be excluded from the scope of the analysis.
Mechanical systems and electronics are excluded from the LCA study boundary due to lack of
detailed information. Past analyses have shown the contribution of infrastructure to the overall
results of an LCA for a WWTP to be relatively insignificant (Emmerson et al., 1995; Xue et al.,
2019). In general, these types of capital equipment are used to treat large volumes of wastewater
over a useful life of many years. Thus, energy and emissions associated with producing these
facilities and equipment generally become negligible.
Downstream impacts associated with effluent release (i.e., eutrophication and toxicity
impacts) are accounted for using the methods and metrics discussed in Appendix A, which
EP-C-17-041; WA 3^77
2-1
-------
Section 2: LCA Methodology
quantify the potential for impact. These methods are based on the mass of nutrients or toxic
substances released into the environment and their potential to lead to eutrophication or toxicity
impacts—they do not account for interactions with the receiving environment, which determine
if and how that potential is realized. For the current study, this means that eutrophi cation and
toxicity impacts are based on the mass of pollutants discharged at points of direct effluent release
(which include the Santa Fe River and proposed Rio Grande discharges) as well as any emissions
associated with processes at the WWTP or the supply chains of inputs to the processes.
Flow diagrams of system boundaries for each scenario are provided in Section 2.2.2. As
previously noted, the proposed scenarios in this study were selected based on their relevance to
the PR WWTP and their ability to produce differentiated effluent quality and potential
environmental impacts. Although only 4 scenarios (plus baseline) were evaluated, they represent
a range of environmental outcomes that will provide insights to potential intermediary options
(i.e., between the baseline and reverse osmosis).
2.2 Life Cycle Inventory
2.2.1 Introduction
An LCI is a comprehensive list of inputs and outputs to and from the system across the
entire life cycle of the product or process. It accounts for the flows to and from nature (e.g.,
emissions to air, water discharges) and between related processes in the technosphere (e.g.,
material and energy requirements) for each process in the assessed life cycle (ISO, 2006b). The
LCI for the Baseline Scenario is based on historical, average conditions of the PR WWTP, as
well as estimates of changes that may occur due to upgrades that are currently being installed.
Operational calculations are based on average annual data (where available) and standardized to
a cubic meter basis using the total volume of water treated in the years for which data are
available. Environmental impacts of infrastructure are allocated to wastewater treated over the
lifetime of individual components.
LCI data are the foundation of any LCA study. Every element included in the analysis is
modeled as its own LCI unit process entry. The connection of LCI unit process data constitutes
the LCA model. A simplified depiction of a subset of this structure for this study is shown in
Figure 2-1. The overall system boundaries include all unit processes associated with plant
operations and disposition of sludge. Each box in the figure represents an LCI unit process. The
full system is a set of nested LCIs where the primary outputs (in red) of one process serve as
inputs (in blue) to another process. Within each nested level, there can be flows both to and from
the environment. Flows from the environment are written in orange in Figure 2-1 and are
represented by the thin black arrows crossing the system boundary from nature. Emissions to the
environment are listed in green, and these flows are tabulated in the calculation of environmental
impacts. Intermediate inputs (in blue) are those that originate from an extraction or
manufacturing process within the supply chain.
The distinction between the foreground and background systems is not a critical one. The
foreground system tends to be defined as those LCIs that are the focus of the study. In this case,
that is the WWTP itself. Background LCI information is comprised of extractive and
2-2
-------
Section 2: LCA Methodology
manufacturing processes that create material and energy inputs required by the wastewater
treatment systems.
2-3
-------
Section 2: LCA Methodology
Raw wastewater and
intermediate inputs
Background System
Coal power
Inputs
Processed coal (kg)
Transport (tkm)
Grid electricity (kWh)
Outputs:
Coal electricity (kWh)
Hg to water (kg)
CO 2 to air (kg)
~
SOk to air (kg)
Coal eictraction
inputs
Raw coal (kg)
Grid electricity (kWh)
Diesel (L)
Water (L)
Outputs:
Processed coal (kg)
PM2.5 to air (kg)
As to water (kg)
Electricity Mix
Inputs
Coal electricity (kWh)
Gas electricity (kWh)
Nuclear electricity (kWh)
Hydro electricity (kWh)
Line losses (kWh)
Potable water [mJ]
Outputs.'
Grid electricity (kWh)
C02 to air (kg)
PM2.5 to air (kg)
Treated wastewater
Primary Treatment
| inpu ts
Influent (m3)
Grid electricity (kWh)
| Ou tpu ts:
Primary effluent (m3)
Foreground System
Biological Treatment
Inputs
Primary effluent (m3)
Grid electricity (kWh)
Outputs:
Secondary effluent (m5)
CH4 to air (kg)
H?Q to air (kg)
Post-Biological Treatment
1 J
Inputs
\J
Secondary effluent (m3)
\W
Grid electricity (kWh)
Receiving Stream
Outputs:
Inputs
Effluent (m3)
~
Treated HzO (m3)
Outputs:
Nature
NzO to air (kg)
HH3 to water (kg)
KEY
Atetes:
Blue teHt
Intermediate inputs
Background system
Each individual box represents an example unit process.
Green text
Emissions to environment
Foreground system
Inputs and outputs as well us unit processes listed are provided
Fled text
Primary process output
Flow between unit processes
as an example, and are not exhaustive.
Raw inputs from nature
—*¦
Flow to or from nature
Figure 2-1. Subset of LCA model structure with example unit process inputs and outputs.
EP-C-17-041; WA 3-77
2-4
-------
Section 2: LCA Methodology
2.2.2 Foreground LCI Data
As discussed earlier, the foreground system for this study is defined as the PR WWTP
itself. For each of the five wastewater treatment configurations evaluated, PR WWTP staff or
their consulting engineers (Carollo Engineers) provided foreground information. The foreground
LCI unit process data developed for this study for all levels are summarized in Appendix B in
Table B-l through Table B-5. All data collected for this study were subject to quality assurance
project plan (QAPP; available upon request from EPA) requirements for completeness,
representativeness, accuracy, and reliability. A description of overall data quality results for the
LCI is provided in Appendix D.
Table 2-1 provides an overview of the foreground unit processes that make up each of the
wastewater treatment configurations evaluated in this study. Many unit processes are common to
all configurations with inputs and outputs remaining consistent across scenarios. Scenarios are
primarily differentiated based on additional sidestream or tertiary unit processes. Energy demand
and process GHG emissions (introduced in Section 2.2.3.1) of the biological treatment process
are adjusted to consider lower nutrient and biochemical oxygen demand (BOD) loading resulting
from installation of sidestream treatment in Scenario 1. Operation of secondary treatment
processes for Tertiary Filtration, Reverse Osmosis and Zero Discharge (Scenarios 2, 3, and 4) are
consistent with Baseline performance. Sludge and biogas production and treatment are expected
to remain consistent across scenarios but are assessed in sensitivity and uncertainty assessments.
Energy, chemical, and material inputs (e.g., background unit processes) to each of the
unit processes are tracked in terms of energy, mass, or volume units. Releases to air and water
for each unit process are tracked together with information about the environmental compartment
to which they are released to allow for appropriate impact characterization. Waste streams are
connected to supply chains associated with providing waste management services, such as
landfilling.
Table 2-1. Foreground Unit Processes Included in Each Wastewater Treatment Configuration.
Unit Process
Wastewater Treatment Configuration
Baseline
(B)
Scenario 1
B +
Sidestream
Filtration
Scenario 2
B +
Tertiary
Filters
Scenario 3
B +
Reverse
Osmosis
Scenario 4
B +
Zero
Discharge
(Full
Diversion3)
Core facility
V
V
V
V
y
Preliminary treatment: screening
and grit removal
V
V
V
V
V
Secondary treatment: biological
V
V
V
V
V
Tertiary treatment: disk filtration
y
y
y
Sidestream treatment: filtrate
y
Tertiary treatment: deep bed
media filters
V
Tertiary treatment: microfiltration
V
Tertiary treatment: reverse
osmosis
V
EP-C-17-041; WA 3-77
2-5
-------
Section 2: LCA Methodology
Unit Process
Wastewater Treatment Configuration
Baseline
(B)
Scenario 1
B +
Sidestream
Filtration
Scenario 2
B +
Tertiary
Filters
Scenario 3
B +
Reverse
Osmosis
Scenario 4
B +
Zero
Discharge
(Full
Diversion3)
Chemical post-treatment
V
Disinfection: ultraviolet
V
V
V
y
V
Wastewater treatment plant
(WWTP) effluent: discharge
V
V
V
V
WWTP effluent: reuse
V
V
V
V
V
WWTP effluent: partial
diversion3
V
V
V
V
WWTP effluent: full diversion3
V
Sludge: dissolved air flotation
V
V
V
V
y
Sludge: anaerobic digestion
y
y
y
y
y
Sludge: belt filter press
y
y
y
y
y
Sludge: landfilling
y
y
y
y
y
Sludge: composting
y
y
y
y
y
Sludge: land application
y
y
y
y
y
Biogas: cleaning
V
V
V
V
V
Biogas: flaring
y
y
y
y
y
Biogas: boiler
y
y
y
y
y
Biogas: combined heat and power
y
y
y
y
y
Brine: underground inject
y
•J Indicates the unit process is relevant for select wastewater treatment configuration.
3 Refers to full diversion of PR WWTP effluent from the primary discharge point in the Santa Fe River to the
secondary outfall in the Rio Grande.
Detailed water quality data were compiled from a range of sources, including historical
monitoring data from the PR WWTP, estimates of anticipated treatment performance made by
Carollo Engineers, and treatment performance from similar systems. Table 2-2 summarizes
influent and effluent water qualities used for this study. Influent data are based on historic
(2015-2020) data and are mainly illustrated for comparison purposes; they are only used for
calculations discussed in Section 3.5.3. Effluent data for organics and nutrients are based on
performance estimates provided by Carollo for each of the treatment configurations. For metals,
this study assumes past performance, as measured by 2015-2020 observed effluent quality, to be
representative of the Baseline configuration. This study assumes Scenario 1 would have a
negligible effect on metals removal and uses performance data from similar systems to estimate
metals removal by Scenarios 2 and 3, as discussed further in Appendix E.
2-6
-------
Section 2: LCA Methodology
Table 2-2. Influent and Estimated Effluent Water Quality for each Wastewater Treatment Configuration.
Parameter
Influent"
Baseline/Scenario 4 -
Zero Discharge
Scenario 1 -
Sidestream
Filtration"1'e
Scenario 2 - Tertiary
Filters6'h
Scenario 3 - Reverse
Osmosis0''
Value
Value
Range
Valueb
Range0
Valueb
Range0
Valueb
Range0
Organics and Nutrients, mg/L
Total suspended solids (TSS)
353
5.0
<10
5
<10
5
<10
3
<5
Volatile suspended solids
(VSS)
NA
3.5
<1
3.5
<1
3.5
<1
2.1
<4
carbonaceous biochemical
oxygen demand (5-day)
(cBOD5)
340
5.0
<10
5
<10
3
<5
3
<5
Chemical oxygen demand
(COD)
1015
30
<50
30
<50
20
<30
20
<30
Total Kjeldahl nitrogen (TKN)
78.6
3.5
<5
3
<5
2.5
<3
1.5g
<3
Nitrate/nitriteJ
NA
1.5
NA
1.5
NA
0.5
NA
0.43
NA
Total nitrogen (TN)
78.6
5.0
<10
4.5
<7
3
<5
2
<5
Organic nitrogen (Org-N)k
28.1
2.5
<3
2.5
<3
2.5
<3
1.5
<2
Ammonium-nitrogen (NH4-N)
50.5
0.1
<1
0.1
<1
0.1
<1
0.1
<1
Total phosphorus (TP)
13.8
1.0
<2.5
0.7
<1
0.05f
<0.1f
0.05
<0.2
Orthophosphate (OP)
NA
0.1
<0.2
0.05
<0.2
0.02g
<0.05g
0.02
<0.2
Metals, jig/Lk
Arsenic
3.0
1.0
0-200
1.0
0-200
0.90
0-180
0.10
0-20
Cadmium
0.21
0.021
.021-0.20
0.021
.021-0.20
0.019
0.019-0.18
0.00021
Chromium
1.6
0.045
0.045-100
0.045
0.045-100
0.045
0.045-100
0.045
0.045-100
Copper
103
4.0
2.1-39
4.0
2.1-39
3.4
1.8-33
0.24
0.13-2.3
Lead
2.6
0.34
0.20-192
0.34
0.20-192
0.27
0.16-150
0.0034
0.0020-1.9
EP-C-17-041; WA 3-77
2-7
-------
Section 2: LCA Methodology
Parameter
Influent"
Baseline/Scenario 4 -
Zero Discharge
Scenario 1 -
Sidestream
Filtration"1'e
Scenario 2 - Tertiary
Filters6'h
Scenario 3 - Reverse
Osmosis0''
Value
Value
Range
Valueb
Range0
Valueb
Range0
Valueb
Range0
Mercury
0.082
0.0019
8E-4-2.7
0.0019
8E-4-2.7
1.73E-03
7E-4-2.4
0.00010
41 >4 0.14
Nickel
4.6
2.5
1.3-6.2
2.5
1.3-6.2
2.4
1.3-6.0
0.23
0.12-0.56
Silver
1.7
0.015
0.015-0.34
0.015
0.015-0.34
0.015
0.015-.34
0.015
0.015-0.34
Zinc
165
64.9
0-118
64.9
0-118
46.1
0-84
1.9
0-3.5
Table abbreviations: NA = not available.
a Average of 2020 PR WWTP data.
b Expected average annual effluent quality.
0 Can be +/- % or concentration range.
d Sidestream filtration assumes aeration system improvements are in place and includes filtrate return flow treatment for nitrogen removal (DEMON®
Annamox) and phosphorus removal (AirPrex®/MagPrex™ from digestate).
e These are estimated values based on experience from other similar installations. These values are not to be understood as technology performance
guarantees.
f This indicates limit of technology. Concentrations depend on chemical dose addition.
g Estimated. Depends largely on size distribution of soluble organic nitrogen (unknown).
h Assumes tertiary nitrogen and phosphorus filters in series. Assumes that sidestream treatment for nitrogen and phosphorus removal (Scenario 1) would not
be installed in this scenario.
1 Assumes microfilter and reverse osmosis treatment downstream of secondary treatment. Assumes that sidestream treatment for nitrogen and phosphorus
removal (Scenario 1) and tertiary filtration for nitrogen and phosphorus removal (Scenario 2) would not be installed.
J Calculated as the difference between TKN and NH4-N.
k Influent and Baseline effluent concentrations determined from historic (2015-2020) metals data. See Appendix E for methods to determine removal rates
from Scenarios 1-4.
2-8
-------
Section 2: LCA Methodology
2.2.2.1 Baseline Scenario - Planned Upgrades
The PR WWTP is in the process of upgrading and expanding several system components,
including the existing biological treatment process for enhanced nutrient removal and process
control, and the capacity of their anaerobic digesters. The PR WWTP is also adding a combined
heat and power (CHP) system to convert methane produced by the digesters into usable heat and
electricity. Figure 2-2 shows the layout of the wastewater treatment facility, which reflects the
anticipated state following all currently planned upgrades.
Preliminary treatment at the PR WWTP includes bar screens and aerated grit traps.
Collected grit and screenings are trucked offsite to the local landfill. Primary clarifiers precede
the plant's biological process, which includes an optional anoxic selector preceding a pair of
aeration basins that are configured as "four-pass carrousel oxidation ditches" (Carollo Engineers,
2018). Primary effluent is blended with return activated sludge, mixed liquor, filtrate, and
dissolved air flotation (DAF) underflow before secondary treatment. Secondary effluent is
treated with disc filters prior to ultraviolet (UV) disinfection, post-aeration, and release.
Currently, most wastewater effluent is discharged to the Santa Fe River. However, the
municipality is proceeding with a plan to build a diversion pipeline to route most of the PR
WWTP effluent to the Rio Grande (see Section 1.2 for additional discussion). The city does not
know exactly how much effluent will eventually be diverted. The Baseline Scenario (as well as
Scenarios 1-3) assumes that an annual average flow of 1 MGD (range of 0.5-2 MGD) will be
diverted to the Rio Grande, with the remainder continuing to satisfy existing non-potable reuse
demand or to be discharged to the Santa Fe River. Actual daily diversion flows will vary
seasonally, with more diversion occurring in the winter months. The diversion flow rate (1
MGD) is an assumed value for the purposes of conducting this LCA, as the city is still
conducting planning and permitting efforts and has not committed to any particular flow.
The PR WWTP pumps a portion of its treated effluent offsite for non-potable reuse as
irrigation water, which defers pumping and consumption of 0.127 m3 of groundwater per m3 of
treated effluent after accounting for volume losses during treatment. Although this water reuse is
seasonal, as it is used mainly for irrigation purposes, this study accounts for the reuse flow on an
average annual basis. Offsite pumping of non-potable water for reuse requires an estimated 0.075
kilowatt hours per cubic meter (kWh/m3) of treated wastewater and avoids 0.028 kWh of
electricity that would be used to pump groundwater, leading to a net increase in electricity
demand. Emission of nutrients, chemical oxygen demand (COD), and metals that would
otherwise have been emitted to surface water with the rest of treatment plant effluent are instead
applied to land in the irrigation water. No avoided fertilizer benefit is assessed for the reused
water.
EP-C-I7-04I; WA 3^77
2-9
-------
Section 2: LCA Methodology
\
Figure 2-2. System diagram of the existing PR WWTP following upgrades to the biological
treatment system.
Waste activated sludge from the secondary clarifiers is thickened in a DAF process and
mixed with primary solids before anaerobic digestion. A portion of sludge is stabilized using
lime addition. Digestate is thickened in a belt filter press (BFP) and stabilized in an onsite
windrow composting facility (50% of digestate) with regional green waste or is sent to the local
landfill (50% of digestate). Section 2.2.3 briefly describes the scope of air emissions modeling
performed for each process. The quantity of digestate produced and its fate is consistent across
scenarios. LCI data for these processes are available in Table B-l.
The PR WWTP installed a CHP system as part of their plant upgrades. Biogas produced
in the anaerobic digesters is combusted onsite in the CHP engines, boilers, or flares, as illustrated
in Table 2-3. The facility provided emissions data for the combustion processes for nitrogen
oxides (NOx), carbon monoxide (CO), and volatile organic compounds (VOCs), as reported in
Table B-l. Another LCA considering beneficial use of anaerobic digestion biogas was used to
provide supplementary estimates of other air emissions for the CHP engines, boiler, and flare
(Morelli et al., 2019).
Table 2-3. Allocation of Biogas to Onsite Combustion Processes.
Combustion Process
Baseline
Value
Minimum
Value
Maximum
Value
Combined heat and power engine
83%
74%
95%
Boiler
15%
2.1%
24%
Flare
1.9%
1.9%
2.9%
2-10
-------
Section 2: LCA Methodology
2.2.2.2 Scenario 1 - Sidestream Filtration
Figure 2-3 shows the WWTP layout for Sidestream Filtration (Scenario 1). Operation of
preliminary, primary, secondary, and sludge treatment processes remain the same in Sidestream
Filtration (Scenario 1) as in the Baseline Scenario (Section 2.2.2.1). Additional nitrogen and
phosphorus removal is achieved by installing equalization and sidestream treatment processes on
the flow of BFP filtrate before it returns to secondary treatment. DEMON® Anammox and the
MagPrex™ struvite processes are used to remove ammonia and phosphorus from the filtrate
return flow, respectively. A magnesium chloride salt is added to the filtrate to precipitate
phosphorus as struvite. Struvite is assumed to be used as an agricultural fertilizer displacing the
production of diammonium phosphate (DAP). Both processes require additional electricity
consumption. Filtrate treatment reduces the load of nitrogen to secondary treatment, however,
the emission of nitrous oxide from the DEMON® process is similar to those of conventional
nitrification/denitrification biological processes (Weissenbacher et al., 2010). LCI data for
Sidestream Filtration (Scenario 1) are available in Table B-2.
Figure 2-3. System diagram of Scenario 1 - Sidestream Filtration.
2.2.2.3 Scenario 2 - Tertiary Filtration
Figure 2-4 shows the WWTP layout for (Scenario 2). Operation of preliminary, primary,
secondary, and sludge treatment processes remain the same in Scenario 2 as in the Baseline
Scenario (Section 2.2.2.1) with the exception of disc filtration, which would be eliminated. In
this scenario, additional nitrogen and phosphorus removal are achieved through the installation
of sequential deep-bed media filters. Methanol is added to secondary effluent before the first
deep-bed media filter to assist in denitrification. Alum is added for the removal of phosphorus in
the second filter. Additional pumping energy is required to move wastewater through the filters.
The effect of filter operation on secondary treatment processes is assumed to be negligible. LCI
data for these processes are available in Table B-3.
2-11
-------
Section 2: LCA Methodology
Mix Tank
E
ae
¦
ge
Key:
Liquid Stream
Solid Stream
Chemical Input
Figure 2-4. System diagram of Scenario 2 - Tertiary Filters.
2.2.2.4 Scenario 3 - Reverse Osmosis
Figure 2-5 shows the WWTP layout for Reverse Osmosis (Scenario 3). Operation of
preliminary, primary, secondary, and sludge treatment processes remain the same in Reverse
Osmosis (Scenario 3) as in the Baseline Scenario (Section 2.2.2.1) with the exception of disc
filtration, which would be eliminated. Additional nitrogen and phosphorus removal are achieved
through installation of an RO filter. A microfilter (MF) is also installed before RO to reduce
fouling and prevent membrane damage. The total quantity of MF and RO membrane units are
installed based on a design flow of 9 MGD and the production of 2 MGD of brine relative to a
total facility design flow of 12 MGD. The remaining 25% of design flow bypasses the MF/RO
process.
Carbon dioxide and sodium bisulfite are added to the wastewater in a chemical post-
treatment process. Of the 9 MGD of wastewater treated by the MF/RO process under design
conditions, 2 MGD becomes brine and is disposed of via onsite deep well injection. Based on
discussions with PR WWTP staff and their consulting engineers, other options for disposal of
RO brine, such as mechanical or pond evaporation, are less feasible for this project and are not
modeled in this analysis. LCI data for these processes are available in Table B-4.
2-12
-------
Section 2: LCA Methodology
Figure 2-5. System diagram of Scenario 3 - Reverse Osmosis.
2.2.2.5 Scenario 4 - Zero Discharge to Santa Fe River
Figure 2-6 shows the WWTP layout for Zero Discharge (Scenario 4). Operation of all
treatment processes remain the same in Scenario 4 as in the Baseline Scenario (Section 2.2.2.1).
Effluent discharge (the portion that is not pumped to non-potable reuse customers) is diverted
completely to the Rio Grande via pipeline, with no discharge into the Santa Fe River. Partial
diversion of effluent to the Rio Grande is included in all scenarios, and the necessary
infrastructure to achieve full diversion will be installed regardless of the quantity of effluent
ultimately diverted (PVC piping for the pipeline, which will be 30-inch pressure pipe, is included
in all scenarios). Increased electricity demand for additional effluent pumping is the only
additional requirement for the full diversion scenario. LCI data for this process are available in
Table B-5.
2-13
-------
Section 2: LCA Methodology
Liquid Stream
Solid Stream
Chemical Input
Figure 2-6. System diagram of Scenario 4 - Zero Discharge to Santa Fe River.
2.2.3 LCI Background Data Sources
The supply chains of inputs to the wastewater treatment processes are represented where
possible using publicly available data from the Federal LCA Commons (Federal LCA Commons,
2021). Within the Federal LCA Commons, background material, fuel, and transport datasets are
sourced from the National Renewable Energy Laboratory's U.S. LCI database (NREL, 2019).
Where required background datasets were not available from the Federal LCA Commons, the
Ecoinvent version 3.7 database is used (Frischknecht et al., 2005). Ecoinvent is a widely used
global LCI database available by paid subscription. Table 2-4 lists background unit processes
used in the LCA model and their source databases. The environmental flow inputs and outputs
for the selected background databases were harmonized using EPA's Federal LCA Commons
Elementary Flow List (Edelen et al., 2019). Using this standardized list ensures that all the
environmental flows in the LCA are properly captured in the impact assessment results.
Table 2-4. Background Unit Process Data Sources."
Background Input
Original Unit Process Name
LCI Database
Alum
Aluminum sulfate production, powder aluminum sulfate,
powder Cut-off, S
Ecoinvent 3.7
Carbon dioxide
Carbon dioxide production, liquid carbon dioxide, liquid
Cut-off, S
Ecoinvent 3.7
Citric acid
Citric acid production citric acid Cut-off, S
Ecoinvent 3.7
Fertilizer, nitrogen
Urea production urea Cut-off, S
Ecoinvent 3.7
Fertilizer, nitrogen and
phosphorus
Diammonium phosphate production diammonium
phosphate Cut-off, S
Ecoinvent 3.7
Fertilizer, phosphorus
Single superphosphate production single superphosphate
Cut-off, S
Ecoinvent 3.7
Fertilizer, potassium
Potassium sulfate production potassium sulfate Cut-off, S
Ecoinvent 3.7
2-14
-------
Section 2: LCA Methodology
Table 2-4. Background Unit Process Data Sources."
Background Input
Original Unit Process Name
LCI Database
Filter nozzles, steel
Casting, steel, lost-wax casting, steel, lost-wax Cut-off, S
Steel production, chromium steel 18/8, hot rolled steel,
chromium steel 18/8, hot rolled Cut-off, S
Ecoinvent 3.7
Membrane,
microfilter/reverse
Polyvinylfluoride production polyvinylfluoride Cut-off, S
Ecoinvent 3.7
osmosis
Phosphoric acid
Phosphoric acid production, dihydrate process phosphoric
acid, fertilizer grade, without water, in 70% solution state
Cut-off, S
Ecoinvent 3.7
Polymer
Polyacrylamide production polyacrylamide Cut-off, S
Ecoinvent 3.7
Proprietary cleaning
solution
Citric acid production citric acid Cut-off, S
Ecoinvent 3.7
Residuals to landfill
Treatment of inert waste, inert material landfill inert
Ecoinvent 3.7
waste, for final disposal Cut-off, S
Sodium hypochlorite
Sodium hypochlorite production, product in 15% solution
state sodium hypochlorite, without water, in 15% solution
state Cut-off, S
Ecoinvent 3.7
Electricity
Electricity, AC, 120 V (from 2019 AZNM grid)
eLCI
Anthracite
Anthracite coal, at mine
USLCI
Caustic soda
Sodium hydroxide, production mix, at plant
USLCI
Diesel, combusted
Transport, passenger truck, diesel powered
Diesel, combusted in industrial equipment
USLCI
Filter pads, polyester
Unsaturated polyester, UPR, resin, at plant
USLCI
Gravel
Gravel, at mine
USLCIb
Lime
Quicklime, at plant
USLCI
Methanol
Methanol, at plant, kg
USLCI
Natural gas, anaerobic
digestion
Natural gas, combusted in industrial boiler
USLCI
Natural gas, compost
Natural gas, combusted in industrial equipment, 1.357
m3/kg, 52.13 MJ/kg
USLCI
Sand
Sand, at mine
USLCIb
Sulfuric acid
Sulfuric acid, at plant
USLCI
a The label "Cut-off, S" refers to system processes from Ecoinvent's cut-off system model. A system process
aggregates all allocated upstream and process elementary flows within an single inventory, providing
confidentiality for upstream data providers and data portability for LCA practitioners. System models refer to
treatment of recycled content across life cycles. In the cut-off system model, the enviromnental impacts of material
extraction and processing are allocated to the material's first user, allowing recycled material to enter subsequent
life cycles without enviromnental burden.
b Adapted from USLCI's limestone mining unit process.
Electricity is a key background unit process for all the wastewater treatment
configurations investigated. Table 2-5 displays the Arizona/New Mexico (AZNM) subregion
generation resource mix applied in the foreground LCA model and the U.S. average generation
resource mix used in the electricity sensitivity analysis (U.S. EPA, 2021b). The AZNM resource
mix provides 83% of the electricity for the AZNM consumption mix used in the analysis. The
remaining 17% of consumed electricity is provided by neighboring exporting regions.
2-15
-------
Section 2: LCA Methodology
Consumption mixes consider trading that occurs due to grid complexity, and is expected to
provide a more accurate estimate of environmental impact associated with grid-based electricity
consumption in eGRID subregions (Hottle and Ghosh, 2021). These data are based on eGRID
resource mix information from 2019 and were generated using EPA's Electricity LCI (eLCI) tool
within the Federal LCA Commons (U.S. EPA, 2020a). A loss factor of 5.3% is applied to
account for electricity losses during distribution to the final consumer (i.e., PR WWTP). Section
4.4 presents a sensitivity analysis for the electricity used by the facility, modeling a scenario that
uses U.S. average electricity, as well as a scenario that uses 100% solar electricity, to meet the
WWTP electricity requirements.
Table 2-5. Arizona/New Mexico Average Electrical Grid Mix.
Fuel
Regional Grid
(%)
U.S. Average
Grid (%)
Natural gas
44.9%
38.4%
Coal
22.3%
23.3%
Nuclear
18.8%
19.6%
Solar
4.50%
1.74%
Geothermal
3.60%
0.37%
Hydro
3.10%
6.83%
Wind
2.20%
7.15%
Biomass
0.40%
1.56%
Oil
0.10%
0.61%
Other
0%
0.44%
2.2.3.1 Biogas Cleaning
Basic biogas cleaning processes, including iron sponge scrubbing, moisture removal,
compression, and siloxane removal, are modeled for the portion of biogas combusted in the
onsite boiler and CHP engines. Iron sponge scrubbing uses iron oxide impregnated-wood chips
to remove hydrogen sulfide (H2S) from produced biogas. Iron oxide media can be regenerated
several times by air purging, releasing adsorbed H2S as elemental sulfur. Modeling assumes that
the concentration of H2S is reduced from 500 (Wiser et al., 2010) to 1 part per million volume
(Ong et al., 2017). The process requires electrical energy to circulate air for the media
regeneration step. Spent media is disposed of in an inert material landfill. Moisture is removed
from produced biogas by chilling and condensation, assuming an electrical energy requirement
equivalent to 2% of produced biogas energy content (Ong et al., 2017). Biogas is compressed to
4 pounds per square inch gauge prior to combustion. Siloxane removal using activated carbon is
the final biogas cleaning step. The quantity of activated carbon required for siloxane adsorption
is estimated assuming a biogas siloxane content of 100 milligrams per cubic meter and a mass
loading rate of 10% (siloxane mass/activated carbon mass). Table 2-6 presents LCI data for the
biogas cleaning processes. Biogas production and cleaning is consistent across the considered
scenarios.
2-16
-------
Section 2: LCA Methodology
Table 2-6. Life Cycle Inventory Data for Biogas Cleaning.
Process Name
Input Name
Mean Value
Units8
Biogas cleaning—iron sponge
Electricity
3.5E-5
kWh/m3
Iron sponge
5.4E-4
kg/m3
Biogas cleaning—moisture removal
Electricity
0.04
kWh/m3
Biogas cleaning—compression
Electricity
3.4E-3
kWh/m3
Biogas cleaning—siloxane removal
Activated carbon
3.5E-4
kg/m3
a Biogas cleaning inventory data is normalized to the average annual flow of wastewater
treated at the PR WWTP.
2.2.3.2 Digestate Composting
Half of produced digestate is composted with yard waste at an onsite windrow
composting facility. The facility reports that approximately 0.39 kilogram (kg) of yard waste is
composted per kg of digestate. The LCI for the composting process includes electricity and
natural gas consumption and process emissions of ammonia, carbon monoxide, methane, nitrous
oxide, and non-methane volatile organic compounds. Only process emissions attributable to the
digestate are included in the LCA model, as yard waste is a separate material and its emissions
are not attributable to the wastewater system. Characteristics of the digestate, yard waste, and
finished compost (presented in Table 2-7) are used to estimate process emissions. Compost is
produced in an indoor facility and leachate production is assumed to be negligible. LCI data for
the compost process is available in Table B-l. Digestate production and composting is consistent
across the considered scenarios.
Table 2-7. Characteristics of Digestate, Yard Waste, and Finished Compost.
Characteristic
Digestate
Yard Wastec
Compost
Units
Moisture content
87%a
48%
32%d
% of wet mass
Nitrogen content
5.8%a
1.5%
2.4%d
% of dry mass
Phosphorus content
1.9%b
0.20%
0.90%e
% of dry mass
Potassium content
3.1%b
1.3%
0.48%f
% of dry mass
Carbon content
41%b
43%
36%d
% of dry mass
a PR WWTP average values for 2020.
b (Nkoa, 2014).
0 (Yoshida et al., 2012).
dPR WWTP compost chemical analysis for 2019.
e (Morelli et al., 2019)
f(Keng et al., 2020).
2.2.3.3 Compost Land Application
Finished compost is a Class A material that is assumed to be sold locally and used as a
soil amendment on home gardens or agricultural crops (U.S. EPA, 2002). The LCA model
assumes that using compost in these applications displaces the use of chemical fertilizers such as
2-17
-------
Section 2: LCA Methodology
urea, single superphosphate, and potassium sulfate, based on the nitrogen, phosphorus, and
potassium content of the compost (Table 2-7). Nutrients in organic amendments, such as
compost, are typically less plant-available than similar quantities of nutrients in chemical
fertilizers (Rigby et al., 2016). Fertilizer substitution rates are applied to estimate the quantity of
plant-available nutrients in land-applied compost that can reasonably be assumed to displace the
production and use of chemical fertilizers. Average fertilizer substitution rates of 30%, 73%, and
80%) were used for nitrogen, phosphorus, and potassium respectively (see Table B-l for more
information).
The analysis assumes that 12% of land-applied carbon is sequestered beyond the 100-
year time horizon considered in the baseline LCA model based on literature values indicating a
range of between 9% (Boldrin et al., 2009) and 15% (Yoshida et al., 2012). The model also
includes estimates of emissions to air and water that would accompany land application of
composted digestate. These emissions are assumed to be similar in magnitude to emissions that
would result from use of chemical fertilizers, leading to no net change in agricultural emissions.
These emissions are included in the analysis to demonstrate their scale relative to other aspects
of the system. LCI data for the land application process is available in Table B-l. Compost land
application is consistent across the considered scenarios.
2.2.3.4 Digestate Landfilling
Half of produced digestate is trucked offsite (64 km) and disposed of in the local landfill.
The landfill's gas capture system is assumed to have the national average gas capture rate of
68.2% over the facility's lifespan. The landfill does not have an energy recovery system and
flares the captured landfill gas. Emissions data from the biogas flare are used to estimate flare
emissions (Table B-l). A first order decay equation is used to estimate the quantity of degradable
carbon that is converted to methane over the 100-year analysis period as a function of the values
reported in Table 2-8. Produced methane that is not captured, and either flared or oxidized in the
landfill cover, is released to the atmosphere. Non-degradable carbon and the fraction of
degradable carbon that is not decomposed within the 100-year analysis period are sequestered,
providing a climate benefit. Leachate treatment and emissions are included in the LCA model
based on LCI data from (Righi et al., 2013). Leachate is assumed to be produced at a rate of 145
liters per metric ton of organic waste landfilled. LCI data for the landfill and leachate treatment
process is available in Table B-l. Digestate landfilling is consistent across the considered
scenarios.
Table 2-8. Key Landfill Modeling Parameters.
Piiriimolor
Value
I nils
Source
Landfill gas capture rate
68%
% of produced gas
"Typical collection" for decay factor of
0.02 (U.S. EPA, 2020b).
Degradable organic carbon (DOC)
5%
% of wet mass
(RTI International, 2010).
Non-degradable organic carbon
0.4%
% of wet mass
Calculated based on digestate DOC and
carbon content in Table 2-7.
Fraction of degradable carbon
decomposed (DOCf)
65%
% of DOC
£SYLYIS_._20JI)
Decay factor (k)
0.02
unitless
Factor for arid area (LandGEM).
2-18
-------
Section 2: LCA Methodology
Parameter
Value
Units
Source
Fraction of degraded carbon
converted to methane
50%
% of decomposed
carbon
(RTI International, 2010).
Oxidation factor
10%
% of produced methane
(IPCC, 2006).
2.2.4 Process GHG Emission Estimation Methodologies
Estimates of onsite, process-based GHG emissions are made for methane (CH4)
production from biological treatment, anaerobic digestion, and landfill disposal of biosolids.
Estimates of nitrous oxide (N2O) emissions from biological treatment and receiving waters are
also included in the analysis (IPCC, 2006). Carbon dioxide (CO2) emissions from wastewater
treatment processes are not included in the inventory of GHG emissions, in alignment with IPCC
Guidelines for national inventories (IPCC, 2006) as they are biogenic in origin and do not
contribute to GWP. The methodology for calculating GHG emissions associated with wastewater
treatment is generally based on guidance provided in the IPCC Guidelines for national
inventories; however, more specific emission factors for CH4 and N2O are used based on site-
specific emissions for representative biological treatment processes. A detailed presentation of
the calculations used to estimate process GHG emissions is provided in Appendix Section B.2.
2.2.5 L CI Limitations
Some of the main limitations that readers should understand when interpreting the LCI
data and findings are as follows:
• Support personnel requirements. Support personnel requirements are excluded
from the LCA model. The energy and wastes associated with research and
development, sales, administrative personnel, or related activities are not included, as
energy requirements and related emissions are assumed to be quite small for support
personnel activities.
• Representativeness of background data. Background processes are representative
of either U.S. average data (in the case of data from the Federal LCA Commons) or
European or global average (in the case of Ecoinvent) data. In some cases, European
Ecoinvent processes were used to represent U.S. inputs to the model (e.g., for
chemical inputs) due to lack of available representative U.S. processes for these
inputs. The background data, however, met the criteria listed in the project QAPP for
completeness, representativeness, accuracy, and reliability. The overall data quality
results for the LCI are provided in Appendix D.
• Full LCI model data accuracy and uncertainty. In a complex study with thousands
of numeric entries, the accuracy of the data and how it affects conclusions is a
difficult subject. The reader should keep in mind the uncertainty associated with LCI
data when interpreting the results. Comparative conclusions should not be drawn
based on small differences in impact results.
• Transferability of results. The LCI data presented here are specific to the PR
WWTP. LCI results may vary substantially for other case-specific operating
conditions and facilities.
2-19
-------
Section 2: LCA Methodology
2.3 Life Cycle Impact Assessment Model
The model used to conduct the Life Cycle Impact Assessment was constructed in
openLCA version 1.10.3, an open-source LCA software package developed by GreenDelta
(GreenDelta, 2020). This open-source format allowed project team members to seamlessly share
the LCA model.
Appendix B presents LCI data originally developed in Excel and transferred into the
OpenLCA model. Tables in Appendix B present LCI data according to the treatment processes
included in the LCA model, noting which processes are relevant for each treatment
configuration. LCI flow labels correspond to the "background input" names in Table 2-4.
Interpretation of LCIA results requires understanding the uncertainty associated with
inventory data. A Monte Carlo approach was used to estimate uncertainty ranges for the baseline
results presented in Section 3. The model was parameterized in OpenLCA to allow uncertainty
data attached to each parameter to propagate through the model. Results uncertainty associated
with impact assessment was not included in the analysis and is expected to affect the treatment
configurations similarly, as the drivers of impact are common across scenarios (See Section 3 for
more detail).
A Monte Carlo analysis randomly samples the constructed LCA model based on
uncertainty data attached to global parameters, process parameters, and inventory flows. By
carrying out this sampling procedure over a large number of model runs (1,000 in this analysis),
OpenLCA constructs a histogram of model results (Clavreul et al., 2012). The 5th and 95th
percentile values from these model runs were used to establish uncertainty bounds around the
Monte Carlo mean. Lognormal distributions were typically used to represent emissions to nature.
The lognormal distribution is the default distribution used to model environmental flows in the
Ecoinvent 2 database (Ciroth et al., 2012). Triangular distributions are used to define uncertainty
for material, energy, and chemical inputs and outputs using minimum, mean, and maximum
identified values to define the distribution vertices. Appendix B.l documents inventory values,
associated uncertainty data, supporting assumptions, and sources.
At the analysis level, it is important to consider that uncertainty in inventory or
characterization is not purely multiplicative when considering differences between systems
(Hong et al., 2010). For many LCA analyses, many background and some foreground processes
will be shared between systems. For example, background electricity generation is often shared,
and chemical additives or concrete could be shared foreground processes for wastewater
treatment. Such shared processes allow for fewer confounding factors when comparing results.
Once all necessary data were input into the openLCA software and reviewed, a system
model was created for the parameterized treatment configuration. The models were reviewed to
ensure that each elementary flow (e.g., environmental emissions, consumption of natural
resources, energy demand) was characterized under each impact category for which a
characterization factor was available. LCIA results were then calculated by generating a
contribution analysis for the product system based on the defined functional unit of treatment of
one-cubic meter of wastewater. Appendix A discusses the detailed LCIA methods used to
translate the LCI model in openLCA into the impact results assessed in this study.
2-20
-------
Section 2: LCA Methodology
ERG compiled LCI data in a central Excel spreadsheet and included a data quality index
(DQI) matrix to evaluate the quality of the LCI data. A DQI matrix evaluates data based on five
criteria: source reliability, completeness, temporal correlation, geographical correlation, and
technological correlation. ERG adhered to EPA guidance for assessing LCI data quality when
scoring the DQI (Edelen and Ingwersen, 2016). The results of this evaluation indicate LCI data
quality is sufficient for use. A DQI matrix for LCI data can be found in Appendix D.
2.3.1 L CIA Limitations
While limitations of the LCI model are specifically discussed in Section 2.2.5, some of
the main limitations that readers should understand when interpreting the LCIA findings are as
follows:
• Transferability of results. While this study is intended to inform decision-making
for a wide range of stakeholders, the impacts presented here relate to a specific
WWTP in Santa Fe, New Mexico.
• Site specificity. Although the study refers to a specific WWTP, some metrics are not
able to provide site-specific results. For example, eutrophication potential,
particularly with respect to direct effluent emissions, only captures a direct
relationship between potentially eutrophying pollutants that is based on the Redfield
ratio (see Appendix A. 1 for method description) and does not describe local water
quality dynamics.
• LCIA method uncertainty. In addition to the uncertainty of the LCI data, there is
uncertainty associated with applying LCIA methodologies and normalization factors
to aggregated LCI data. For example, two systems may release the same total amount
of the same substance, but one quantity may represent a single high-concentration
release to a stressed environment while the other quantity may represent the aggregate
of many small dilute releases to environments that are well below threshold limits for
the released substance. The actual impacts would likely be very different for these
two scenarios, but the LCI does not track the temporal and spatial resolution or
concentrations of releases in sufficient detail for the LCIA methodology to model the
aggregated emission quantities differently. Therefore, it is not possible to state with
complete certainty that differences in potential impacts for two systems are
significant differences. Although there is uncertainty associated with LCIA
methodologies, all LCIA methodologies are applied to different wastewater treatment
configurations uniformly. Therefore, comparative results can be determined with a
greater confidence than absolute results for one system.
2-21
-------
Section 3: Life Cycle Impact Baseline Results
3. Life Cycle Impact Baseline Results
An overview of LCIA results are provided in Figure 3-1. For each metric, results have
been standardized by dividing each result by the maximum absolute value across all scenarios so
that each can be expressed on a scale of -1 to 1. A value of 1 represents the scenario with the
largest impact within a category, and -1 represents the smallest impact. No weighting factors are
applied, which implicitly gives equal weight to each of the 13 metrics. Figure 3-1 shows that
Scenario 3 (RO) results in the largest impacts across all metrics except eutrophication potential.
The remainder of this section illustrates and discusses these results in greater detail.
Figure 3-1. Summary of baseline LCIA results for the Baseline Scenario and Scenarios 1-4
(S1-S4). For each metric, results were standardized by dividing each result by the
maximum absolute value across all scenarios so that each metric can be expressed on a
scale of -1 to 1, where 1 indicates the greatest impact among all scenarios. Metric
abbreviations are provided in Table 1-3.
In the following sections, baseline LCIA results are presented in greater detail in four
groups based on whether the results pertain broadly to 1) environmental quality, 2) energy and
climate, 3) water, or 4) toxicity. For all metrics, impact contributions are presented according to
treatment processes or major drivers. Refer to Appendix A for more information on individual
impact categories, their underlying environmental issue, and the pollutants that contribute to each
impact. A description of treatment processes and major drivers is provided in Table 3-1 and
applies to Figure 3-2 through Figure 3-14.
Table 3-1. Description of Impact Contribution Categories.
Category
Description
Processes
Main Plant—Energy Use
Includes electricity and diesel fuel consumption that cannot be
allocated to individual unit processes (due to a lack of meter data).
Primary and Secondary
Treatment
Includes landfill disposal of screenings/grit and process air emissions
from the biological treatment system.3
EP-C-17-041; WA 4-77
3-1
-------
Section 3: Life Cycle Impact Baseline Results
Post-Secondary Treatment
Process grouping includes disk filtration,3 ultraviolet disinfection,3 and
tertiary treatment processes for scenarios 1-3.
Biogas Cleaning and
Combustion
Process grouping includes biogas cleaning and combustion in the
combined heat and power (CHP) system, boiler, or flare.
Sludge Processing and Disposal
Process grouping includes dissolved air flotation,3 belt filter press,3
anaerobic digestion,3 digestate composting, digestate landfilling, and
compost land application. Includes environmental credits associated
with avoided energy and fertilizer.
Brine Injection
Includes energy and water consumption associated with reverse
osmosis brine deep well injection.
Effluent Reuse
Includes energy and avoided energy consumption and emissions to
water associated with wastewater effluent reuse.
Effluent Diversion
Includes energy and infrastructure inputs required for effluent
diversion to the Rio Grande. All effluent emissions are reflected in the
"Effluent Release" process group.
Effluent Release
Includes emissions to surface water from treated wastewater effluent.
Drivers
WWTP Process Emissions
Direct greenhouse gas emissions from the secondary biological
treatment process and anaerobic digesters.
Energy
Net consumption of electricity, diesel, and natural gas at the PR
WWTP. Avoided heat and electricity are included in this category and
reduce net impact attributed to energy consumption.
Transport
Includes the share of diesel combustion impacts allocated to vehicle
use.
Chemicals
Includes all chemicals used at the PR WWTP.
Landfill
Includes all impacts associated with landfilling of digestate and
subsequent leachate treatment.
Biogas Combustion
Driver grouping includes biogas cleaning and combustion in the CHP
system, boiler, or flare.
Composting
Includes energy and emissions to air associated with digestate
composting.
Water Reuse
Includes energy and avoided energy consumption and emissions to
water associated with wastewater effluent reuse.
Effluent Diversion
Includes energy and infrastructure inputs required for effluent
diversion to the Rio Grande. All effluent emissions are reflected in the
"Effluent Release" process group.
Effluent Release
Includes emissions to surface water from treated wastewater effluent.
Land Application
Includes emissions to air and water associated with compost land
application.
Avoided Product
Includes avoided fertilizer production. Avoided energy products are
included in the "Energy" driver grouping.
Materials
Includes all consumable infrastructure materials modeled for the
tertiary treatment processes and diversion pipeline.
a Energy consumption is reflected in "Main Plant—Energy Use."
In Sections 3.1 through 3.4, panel "a" in each figure presents net environmental impact
results according to treatment process, as well as two sets of uncertainty ranges developed using
Monte Carlo analysis. Each set of uncertainty bars was developed based on the 5th and 95th
percentile values that result from 1,000 iterations of the LCA model. In cases where negative
impacts occur, the total net impact, as well as the uncertainty ranges about the total net impact,
may be wholly within the columns in the figures below. The black set of uncertainty bars
includes uncertainty estimates associated with all processes in the treatment system (see Table
EP-C-I7-04I; WA 4^77
3-2
-------
Section 3: Life Cycle Impact Baseline Results
2-1 for a list of processes in each scenario). These uncertainty ranges include uncertainty that is
unique to each scenario, as well as uncertainty estimates for processes that are common to all
assessed treatment scenarios (termed "shared uncertainty"). When uncertainty is associated with
processes that are common to all treatment systems, it has bearing on the absolute magnitude of
impact realized by each treatment system but cannot be used to differentiate between treatment
scenario environmental performance.
The blue set of uncertainty bars includes uncertainty estimates associated only with
treatment processes that are unique to individual treatment scenarios (e.g., sidestream filtration,
MF/RO) or processes in which treatment performance varies according to scenario (e.g., effluent
release, effluent reuse). The portion of analysis uncertainty unique to each treatment scenario
affects both the absolute magnitude of impact that is potentially realized by each scenario and
provides an opportunity to differentiate between scenarios based on independent sources of
uncertainty. In situations where blue uncertainty bars do not overlap (even if black bars do), we
can be more confident that the mean impacts of each alternative are different.
3.1 Environment
3.1.1 Eutrophication Potential
Figure 3-2 presents eutrophication potential results by treatment process (a) and major
drivers (b). Figures a and b both show that effluent release is the predominant contributor to
eutrophication potential. Land application of compost is the second largest contributor to
eutrophication for all scenarios but contributes a larger relative share of impact for Scenarios 2
and 3, as the contribution from direct effluent release is smaller for those scenarios.
The uncertainty bounds in Figure 3-2 indicate that Tertiary Filtration and Reverse
Osmosis (Scenarios 2 and 3) have similar eutrophication impacts and are likely to yield reduced
impacts relative to the Baseline Scenario, Sidestream Filtration (Scenario 1), and Zero Discharge
(Scenario 4) across the range of conditions described in Appendix B. The largest reduction in
eutrophication potential, relative to the Baseline Scenario impact, is achieved in the RO scenario
(Scenario 3). Impact uncertainty ranges, particularly the upper bounds, are mostly due to the
range of effluent pollutant concentrations illustrated in Table 2-2, and are driven by sources of
uncertainty that are unique to each treatment scenario.
As noted in Section 2.2.3.3, nutrient emissions that contribute to eutrophication potential
during land application are expected to be similar in magnitude to emissions that would occur in
an alternate scenario where chemical fertilizer is used instead of compost. Given this, the
eutrophication potential associated with compost land application could reasonably be allocated
to the agricultural production system, reducing the net eutrophication potential of all scenarios,
thereby eliminating the contribution from land application in panel "b".
EP-C-I7-04I; WA 4^77
3-3
-------
Section 3: Life Cycle Impact Baseline Results
2.5E-02
H 2.0E-02
1.0E-02
5.0E-03
1.34E-02
£
* O.OE+OO
-5.0E-03
12E-02
r 1.34E-02
£
93E-03
Baseline
SI - Sidestream S2 - Tertiaiy Filters
Filtration
S3 - Reverse
Osmosis
S4 - Zero Discharge
~ Main Plant - Energy Use 0 Effluent Reuse
~ Effluent Release ~ Post-Secondary Treatment
~ Primary and Secondary Treatment ~ Brine Injection
O T otal O Monte C arlo Mean
~ Biogas Cleaning and Combustion
~ Sludge Processing and Disposal
0 Effluent Diversion
1.5E-02
1.0E-02
% 5.0E-03
5
6
fc 0.0E+00
-5.0E-03
Baseline
SI - Sidestream S2 - Tertiaiy Filters S3 - Reverse S4 - Zero Discharge
Filtration Osmosis
~ WWTP Process Emissions
~ Chemicals
E3 Composting
~ Effluent Release
~ Materials
E3 Energy
~ Landfill
0 Water Reuse
E3 Land Application
• Total
El Transport
~ Biogas Combustion
0 Effluent Diversion
~ Avoided Product
Figure 3-2. Eutrophication potential results for each treatment scenario, including uncertainty
ranges as the 5th and 95th percentile results from Monte Carlo simulations. Black uncertainty
bars represent combined uncertainty estimates for both shared and unique LCI data for
each scenario, and blue bars include uncertainty estimates only for the LCI inputs that are
unique to individual scenarios. Non-overlapping areas of the two uncertainty bars indicate
uncertainty that is shared across scenarios. Panels a and b show results aggregated
according to the different categorizations described in Table 3-1.
LCIA methods are capable of capturing regional differences in potential impact. But they
do not model watersheds in enough detail to distinguish potential or actual impact associated
with emission of the same quantity of nutrients in either the Rio Grande or Santa Fe Rivers,
which leads to the identical potential impacts shown in Figure 3-2 for the Baseline Scenario and
Scenario 4. As detailed in Appendix A.l, TRACI eutrophication potential characterization
factors are intended to capture the relative influence that each pollutant (nutrients and COD)
could have on algae growth in the photic zone of an aquatic ecosystem when released to an
environment where it is the limiting nutrient (Norris, 2002). Pollutants in effluent release that are
captured in Figure 3-2 include COD, nitrate, ammonia, organic nitrogen, and phosphorus. The
influence of additional factors on eutrophication potential, including bioavailability of organic
EP-C-17-041; WA 4-77
3-4
-------
Section 3: Life Cycle Impact Baseline Results
nitrogen and receiving environments, is discussed further in a eutrophication potential sensitivity
analysis in Section 4.2.
To provide additional context for the trends illustrated in Figure 3-2 from direct effluent
release, Table 3-2 summarizes characterization factors and average annual mass discharges of
the major nutrient forms that contribute to the eutrophi cation potential of direct effluent
discharges. Average annual mass discharges are also illustrated in Figure 3-3.
Table 3-2. Summary of average annual COD and nutrient discharges across treatment
scenarios.
Pollutant
Charact.
Factor (kg
N eq./kg
pollutant)
Baseline
SI -
Sidestream
Filtration
S2 -
Tertiary
Filtration
S3 -
Reverse
Osmosis
S4 - Zero
Discharge
Average kg/yr Discharged
COD
0.05
201,171
201,171
134,114
134,114
201,171
Nitrate (created)
0.24
10,059
10,059
3,353
3,353
10,059
Ammonia
0.78
671
671
671
671
671
Nitrogen,
organic
0.99
16,764
16,764
16,764
10,059
16,764
Phosphorus
7.30
6,706
4,694
335
335
6,706
40,000
35,000
30,000
25,000
20,000
15,000
10,000
5,000
0
Baseline
~ Nitrate (created)
SI - Sidestream
Filtration
M Ammonia
S2 - T ertiary
Filters
S3 - Reverse
Osmosis
S4 - Zero
Discharge
~ Nitrogen, organic
~ Phosphorus
Figure 3-3. Summary of annual nutrient mass discharges across treatment scenarios. COD not
shown to not overwhelm nutrient visibility.
3.1.2 Acidification Potential
Figure 3-4 presents acidification potential results by treatment process (a) and major
driver (b). The figures reveal that biogas combustion, energy consumption (primarily electricity),
and process emissions from digestate composting contribute considerable shares of acidification
impact. Biogas combustion releases NOx and SO2 emissions that contribute to acidification
potential. As illustrated in Figure 3-4, panel b shows a reduced contribution from energy,
compared to main plant energy use in panel a, because that category shows the net effect when
considering both plant energy consumption and avoided energy produced by the CHP system.
EP-C-17-041; WA 4-77
3-5
-------
Section 3: Life Cycle Impact Baseline Results
Chemical use contributes moderately to net impact in Scenario 2, increasing the mean impact
relative to the Baseline, Scenario 1, and Scenario 4. RO has the highest mean acidification
potential due primarily to increased electricity demand from RO operation and deep well
injection.
The uncertainty bars in panel a show considerable overlap across the five scenarios,
which is almost completely dominated by sources of uncertainty that are common to all
treatment scenarios (i.e., blue bars are barely visible). For example, the compost process is the
same across all scenarios. While this uncertainty does affect the magnitude of acidification
potential within the demonstrated range, it is not independent across scenarios and therefore does
not minimize the difference in mean impact. This finding gives greater confidence that
differences in mean impact consequentially differentiate treatment scenarios. Parameter
uncertainty results, presented in Appendix F, indicate that results are most sensitive to facility
energy consumption and compost emissions. The fact that the mean impact for Scenario 3 is at or
near the upper bound of the uncertainty range for Baseline, Scenario 1, and Scenario 4 gives
reasonable confidence that the RO treatment scenario would lead to increased acidification
potential.
4.0E-3
3.5E-3
3.0E-3
2.5E-3
2.0E-3
1.5E-3
1.0E-3
5.0E-4
0.0E+0
-5.0E-4
Baseline
SI - Sidestream
Filtration
S2 - Teitiaiy Filters
S3 - Reverse
Osmosis
S4 - Zero Discharge
H Main Plant - Energy Use E3 Effluent Reuse
~ Effluent Release ~ Post-Secondary Treatment
~ Primary and Secondary Treatment ~ Brine Injection
OTotal OMonte Carlo Mean
0 Biogas Cleaning and Combustion
~ Sludge Processing and Disposal
0 Effluent Diversion
3.5E-03
3.0E-03
2.5E-03
2.0E-03
1.5E-03
1.0E-03
' 5.0E-04
1.0E-18
-5.0E-04
aa
WiwiifiwS
SSiSSSiSSa
III
V/#///
¦
m
wmm.
SI - Sidestream
Filtration
S2 - Tertiary Filters
S3 - Reverse
Osmosis
S4 - Zero Discharge
~ WWTP Process Emissions
~ Chemicals
~ Composting
~ Effluent Release
~ Materials
0 Energy
H Landfill
B9 Water Reuse
E3 Land Application
O T otal
E3 Transport
~ Biogas Combustion
~ Effluent Diversion
~ Avoided Product
Figure 3-4. Acidification potential results for each treatment scenario, including uncertainty
ranges as the 5th and 95th percentile results from Monte Carlo simulations. Black uncertainty
bars represent combined uncertainty estimates for both shared and unique LCI data for
EP-C-17-041; WA 4-77
3-6
-------
Section 3: Life Cycle Impact Baseline Results
each scenario, and blue bars include uncertainty estimates only for the LCI inputs that are
unique to individual scenarios. Non-overlapping areas of the two uncertainty bars indicate
uncertainty that is shared across scenarios. Panels a and b show results aggregated
according to the different categorizations described in Table 3-1.
3.1.3 Smog Formation Potential
Figure 3-5 presents smog formation potential results by treatment process (a) and major
driver (b). Biogas combustion and grid electricity consumption are the primary drivers of smog
formation impact. As illustrated in Figure 3-5, avoided energy products, which are included in
the "sludge processing and disposal" treatment group in panel a, generate a considerable avoided
burden credit that reduces the net smog formation potential of all treatment configurations.
Chemical production and transportation contribute minorly to impact in this category.
Mean estimates of smog formation are similar for the Baseline, Scenario 1, Scenario 2,
and Scenario 4, with nearly complete overlap of both uncertainty ranges. The lower bound of the
combined uncertainty range for Scenario 3 is at or near the upper bound of uncertainty estimates
for the other treatment options, giving high confidence that the RO treatment option will lead to
a significant increase in smog formation impact.
a) 1.0E-01
T3 8.0E-02
b 6.0E-02
-------
Section 3: Life Cycle Impact Baseline Results
Figure 3-5. Smog formation potential results for each treatment scenario, including uncertainty
ranges as the 5th and 95th percentile results from Monte Carlo simulations. Black uncertainty
bars represent combined uncertainty estimates for both shared and unique LCI data for
each scenario, and blue bars include uncertainty estimates only for the LCI inputs that are
unique to individual scenarios. Non-overlapping areas of the two uncertainty bars indicate
uncertainty that is shared across scenarios. Panels a and b show results aggregated
according to the different categorizations described in Table 3-1.
3.2 Energy and Climate
3.2.1 Cumulative Energy Demand
Figure 3-6 presents cumulative energy demand (CED) results by treatment process (a)
and major driver (b). The avoided energy credits associated with anaerobic digestion, which are
included in the "sludge processing and disposal" treatment group in panel a, considerably reduce
the net CED of all treatment systems, leading to a small net positive energy demand for the
Baseline and Scenario 1. Combined uncertainty ranges for these scenarios show the potential to
achieve a net zero energy demand. The CEDs of Scenario 1 and Scenario 2 are similar to the
Baseline scenario, with the minor CED increases in Scenario 2 being attributable to alum
production. The unique uncertainty range for Scenario 4 has no overlap with the Baseline,
Scenario 1, or Scenario 2, indicating a consequential increase in CED attributable to the energy
demand of effluent diversion. The CED of Scenario 3 is significantly greater than that of the
other scenarios due to the energy intensity of RO and brine deep well injection. As biogas enters
the treatment plant as a waste product, the energy content of the wastewater and the resulting
biogas is excluded from CED estimates.
EP-C-I7-04I; WA 4^77
3-8
-------
Section 3: Life Cycle Impact Baseline Results
35
30
25
20
15
10
5
0
-5
-10
-15
2.43
2.98
3.88
18.56
Baseline
SI - Sidestream
Filtration
S2 - Tertiary Filters
H Main Plant - Energy Use E3 Effluent Reuse
~ Effluent Release ~ Post-Secondary Treatment
~ Primary and Secondary Treatment ~ Brine Injection
O T otal O Monte C arlo Mean
S3 - Reverse S4 - Zero Discharge
Osmosis
SBiogas Cleaning and Combustion
~ Sludge Processing and Disposal
~ Effluent Diversion
35
i 30
I 25
; 20
! 15
j 10
! 5
i 0
r-5
j-10
-15
¦rrmrrr
till
V"*"* 1"*"!
Baseline
SI - Sidestream
Filtration
S2 - Tertiary Filters
WWTP Process Emissions
Chemicals
Composting
Effluent Release
Materials
S Energy
0 Landfill
S Water Reuse
0 Land Application
• Total
S3 - Reverse
Osmosis
0 Transport
B Biogas Combustion
E2 Effluent Diversion
~ Avoided Product
S4 - Zero Discharge
Figure 3-6. Cumulative energy demand results for each treatment scenario, including
uncertainty ranges as the 5th and 95th percentile results from Monte Carlo simulations. Black
uncertainty bars represent combined uncertainty estimates for both shared and unique
LCI data for each scenario, and blue bars include uncertainty estimates only for the LCI
inputs that are unique to individual scenarios. Non-overlapping areas of the two
uncertainty bars indicate uncertainty that is shared across scenarios. Panels a and b show
results aggregated according to the different categorizations described in Table 3-1.
3.2.2 Fossil Fuel Depletion
Figure 3-7 presents fossil fuel depletion results by treatment process (a) and major driver
(b). Trends are similar to those discussed for CED with slightly more pronounced benefits for the
avoided products that result from sludge processing and disposal, which include energy and
fertilizer. The reduced impact associated with avoided products leads to net benefits for the
Baseline, Scenario 1, Scenario 2, and Scenario 4. Chemical use contributes moderately to impact
in Scenario 2 and Scenario 3. Energy demand associated with effluent diversion also contributes
moderately to fossil fuel depletion in Scenario 4. Avoided fertilizer production provides a minor,
but non-negligible, reduction in net fossil fuel depletion. As with CED, there is minimal
EP-C-17-041; WA 4-77
3-9
-------
Section 3: Life Cycle Impact Baseline Results
uncertainty associated with the fossil fuel depletion results, and Scenario 3 demonstrates
significantly greater impact than the other four scenarios.
2.0E-01
1.5E-01
1.0E-01
5.0E-02
0.0E+00
-5.0E-02
-1.0E-01
-1.5E-01
er
-2.39E-02
a-
-2.18E-02
"5=
-8.49E-03
6.64E-02
7.54E-03
Baseline
~ Main Plant - Energy Use
~ Effluent Release
~ Primary and Secondary Treatment ~ Brine Injection
O T otal O Monte C arlo Mean
SI - Sidestream S2 - Tertiaiy Filters
Filtration
S Effluent Reuse
E Post-Secondaiy Treatment
S3 - Reverse
Osmosis
S4 - Zero Discharge
~ Biogas Cleaning and Combustion
~ Sludge Processing and Disposal
0 Effluent Diversion
b) 1.0E-01
t 5.0E-02
<2 0.0E+00
1
-5.0E-02
Baseline SI - Sidestream S2 - Tertiary Filters
Filtration
M WWTP Process Emissions £3 Energy
~ Chemicals E3 Landfill
B Water Reuse
E3 Composting
~ Effluent Release
~ Materials
E3 Land Application
• Total
S3 - Reverse
Osmosis
0 Transport
HBiogas Combustion
~ Effluent Diversion
¦ Avoided Product
S4 - Zero Discharge
Figure 3-7. Fossil fuel depletion results for each treatment scenario, including uncertainty
ranges as the 5th and 95th percentile results from Monte Carlo simulations. Black uncertainty
bars represent combined uncertainty estimates for both shared and unique LCI data for
each scenario, and blue bars include uncertainty estimates only for the LCI inputs that are
unique to individual scenarios. Non-overlapping areas of the two uncertainty bars indicate
uncertainty that is shared across scenarios. Panels a and b show results aggregated
according to the different categorizations described in Table 3-1.
3.2.3 Global Warming Potential
Figure 3-8 presents GWP results by treatment process (a) and major driver (b). Process
GHG emissions from secondary treatment and main plant electricity demand are the largest
contributors to GWP impact (panel a). Sludge processing and disposal registers a net reduction in
GWP, but the effect is muted compared to other impact categories such as CED and fossil fuel
depletion, as process GHG emissions from anaerobic digestion, composting, and land application
EP-C-17-041; WA 4-77
3-10
-------
Section 3: Life Cycle Impact Baseline Results
counteract the benefit of avoided fertilizer and energy products. Digestate landfilling contributes
moderately to gross impact (panel b) for all treatment scenarios.
o
U
1.5
1.0
0.5
-0.5
t
0.79
0.80
0.83
i
1.29
0.88
Baseline
SI - Sidestream S2 - Tertiary Filters
Filtration
S3 - Reverse
Osmosis
S4 - Zero Discharge
~ Main Plant - Energy Use E3 Effluent Reuse
~ Effluent Release ~ Post-Secondary Treatment
~ Primary and Secondary Treatment ~ Brine Injection
O Total O Monte Carlo Mean
~ Biogas Cleaning and Combustion
~ Sludge Processing and Disposal
~ Effluent Diversion
2.0
1.5
1.0
0.5
0.0
-0.5
//<&//<
=
O
u
ct
Baseline
S1 - Sidestream
Filtration
S2 - Tertiary Filters
S3 - Reverse
Osmosis
S4 - Zero Discharge
~ WWTP Process Emissions
~ Chemicals
~ Composting
~ Effluent Release
~ Materials
03 Energy
0 Landfill
~ Water Reuse
E2 Land Application
• Total
~ Transport
~ Biogas Combustion
0 Effluent Diversion
~ Avoided Product
Figure 3-8. Global warming potential results for each treatment scenario, including uncertainty
ranges as the 5th and 95th percentile results from Monte Carlo simulations. Black uncertainty
bars represent combined uncertainty estimates for both shared and unique LCI data for
each scenario, and blue bars include uncertainty estimates only for the LCI inputs that are
unique to individual scenarios. Non-overlapping areas of the two uncertainty bars indicate
uncertainty that is shared across scenarios. Panels a and b show results aggregated
according to the different categorizations described in Table 3-1.
Most of the uncertainty in GWP impact is associated with treatment processes that are
shared across scenarios, giving greater confidence that true differences in mean impact exist
across scenarios. However, the unique uncertainty ranges overlap for the Baseline and Scenario
1, indicating that there is no meaningful difference in GWP between these scenarios. The Monte
Carlo mean for Scenario 3 is 37% greater than the Monte Carlo mean of the next most impactful
scenario (Scenario 4), giving high confidence that the RO treatment scenario would lead to a
considerable increase in GWP relative to the other treatment options. The uncertainty ranges are
skewed towards greater impact, pulling the Monte Carlo mean higher than the analysis (best
EP-C-17-041; WA 4-77
3-11
-------
Section 3: Life Cycle Impact Baseline Results
estimate) mean. This skew in the uncertainty results demonstrates the potential for considerably
greater impact if management practice encourages process emissions and system performance in
the upper end of the uncertainty ranges described in Appendix B.
3.3 Water
3.3.1 Water Depletion
Figure 3-9 presents water depletion results by treatment process (a) and major driver (b).
Water depletion refers to consumptive uses of water as described in Appendix A.8. Its unit of
cubic meter per cubic meter of wastewater treated (m3/m3 wastewater treated) can be interpreted
as the cubic meters of water depleted, or consumed, for every cubic meter of water treated.
Effluent reuse is one of the main offsets of water depletion and provides a persistent benefit to
the wastewater treatment facility regardless of the treatment option pursued. Chemical
consumption and upstream production are a moderate or major sources of water depletion for
Scenario 1, Scenario 2, and Scenario 3. Production of alum in Scenario 2 uses a considerable
quantity of water, rivaling the benefit of effluent reuse and introducing uncertainty into the
Scenario 2 water depletion results.
The uncertainty depicted in this figure is predominantly due to the contribution of alum
and uncertainty in how much of it is needed to reduce effluent phosphorus concentrations. The
amount of phosphorus that will need to be removed in the tertiary filters depends on the
performance of the secondary biological process. The Scenario 2 unique uncertainty range
overlaps with the mean water depletion estimate for the RO treatment scenario. While the
realization of this situation is possible, it is not expected under average operating conditions if
the biological treatment system is performing according to design standards.
Deep well injection of RO brine removes nearly 30% of wastewater treated by RO from
active circulation in the watershed, resulting in depletion of approximately 0.17 m3/m3
wastewater treated. Tertiary treatment filters do not process the full quantity of secondary
effluent, as described in Section 2.2.2.4. This process has the benefit of sequestering pollutants
away from humans and sensitive ecosystems but depletes available water resources. Brine
disposal is labeled as WWTP process emissions in panel b.
EP-C-I7-04I; WA 4^77
3-12
-------
Section 3: Life Cycle Impact Baseline Results
a) 0.30
0.20
0.10
0.00
-0.10
-0.20
-0.13
-0.04
Baseline
SI - Sidestream S2 - Tertiary Filters
Filtration
& Main Plant - Energy Use 0 Effluent Reuse
~ Effluent Release m Post-Secondary Treatment
~ Primary and Secondary Treatment ~Brine Injection
O Total OMonte Carlo Mean
0.08
-0.14
S3 - Reverse S4 - Zero Discharge
Osmosis
~ Biogas Cleaning and Combustion
~ Sludge Processing and Disposal
12 Effluent Diversion
b) 0.30
0.20
p 0.10
0.00
-0.10
I ¦
I
Baseline SI - Sidestream S2 - Tertiary Filters
Filtration
O WWTP Process Emissions 0 Energy
~ Chemicals El L andfill
El C omposting BB W ater Reuse
El Effluent Releas e E2 L and Application
~ Materials • T otal
S3 - Reverse
Osmosis
E3 Transport
~ Biogas Combustion
~ Effluent Diversion
n Avoided Product
S4 - Zero Discharge
Figure 3-9. Water depletion results for each treatment scenario, including uncertainty ranges
as the 5th and 95th percentile results from Monte Carlo simulations. Black uncertainty bars
represent combined uncertainty estimates for both shared and unique LCI data for each
scenario, and blue bars include uncertainty estimates only for the LCI inputs that are
unique to individual scenarios. Non-overlapping areas of the two uncertainty bars indicate
uncertainty that is shared across scenarios. Panels a and b show results aggregated
according to the different categorizations described in Table 3-1.
3.3.2 Water Scarcity
Figure 3-10 presents water scarcity results by treatment process (a) and major driver (b).
Water scarcity builds on water depletion results, where contributions to water depletion are
further characterized depending on where that depletion occurs and how scarce water is in that
location relative to the rest of the world. Water scarcity characterization factors are on a scale of
0.1 to 100 with units of cubic meter world equivalents per cubic meter (m3 world
equivalents/m3), representing areas with no water stress (0.1) to areas with very high water stress
(100) (Boulay et al., 2018). For example, if 1 cubic meter of water were depleted in Orlando,
Florida, where water is less scarce and water scarcity factors generally range from 1-5, its water
scarcity impact would be 1-5 m3 world equivalents/m3. Conversely, water scarcity factors in
EP-C-17-041; WA 4-77
3-13
-------
Section 3: Life Cycle Impact Baseline Results
Santa Fe are 100, as water is most scarce there according to the AWARE method (Boulay et al.,
2018). Therefore, if 1 cubic meter of water were depleted in Santa Fe, its water scarcity impact
would be 100 m3 world equivalents/m3. For more information on method development and
interpretation, see Appendix Section A.9.
The general trends and drivers of water scarcity are the same as those of water depletion;
however, water scarcity results are weighted to highlight the burdens and benefits of water use
and reuse in water-scarce regions such as north-central New Mexico. The water scarcity metric
highlights the benefits of reusing wastewater effluent in Santa Fe (results in a large offset, or
negative value in Figure 3-10, due to the high-water scarcity factor of 100 in Santa Fe), while
drawing attention to the issues surrounding brine water disposal (results in a large impact, or
positive value in Figure 3-10, again due to the high water scarcity factor of 100 in Santa Fe).
Based on the current model, the primary impact of brine water disposal is captured by water
scarcity as we assume the brine that is injected (design flow of 2 MGD) sequesters the associated
water indefinitely, removing it (and any co-occurring contaminants) from the hydrologic cycle.
Water is less scarce nationally than it is in the Santa Fe region, and therefore this water scarcity
analysis minimizes the scarcity concerns associated with water use in, for example, chemical
production supply chains and electricity production. The net effect, due to brine injection, in
Scenario 3 is that loss of brine water leads to a world equivalent loss of 5 m3 of water per m3 of
treated wastewater, highlighting the importance of this loss in the Santa Fe region.
EP-C-I7-04I; WA 4^77
3-14
-------
Section 3: Life Cycle Impact Baseline Results
20
15
10
5
0
-5
-10
-15
-12.97
¦12.72
1-11.12
5.03
Baseline
SI - Sidestream S2 - Teitiary Filters
Filtration
S3 - Reverse
Osmosis
-12.96
S4 - Zero Discharge
~ Main Plant - Energy Use E3 Effluent Reuse
~ Effluent Release ¦ Post-Secondary Treatment
~ Primary and Secondary Treatment ~ Brine Injection
O T otal
~ Biogas Cleaning and Combustion
~ Sludge Processing and Disposal
~ Effluent Diversion
b> 20
Baseline SI - Sidestream
Filtration
~ WWTP Process Emissions 0 Energy
~ Chemicals
E3 Composting
~ Effluent Release
~ Materials
S2 - Tertiary Filters
Landfill
Water Reuse
Land Application
Total
S3 - Reverse
Osmosis
0 Transport
~ Biogas Combustion
0 Effluent Diversion
~ Avoided Product
S4 - Zero Discharge
Figure 3-10. Water scarcity results for each treatment scenario. Panels a and b show results
aggregated according to the different categorizations described in Table 3-1.
3.4 Toxicity
3.4.1 Ecotoxicity
Figure 3-11 presents ecotoxicity results by treatment process (a) and major driver (b).
Electricity consumption and production are the primary drivers of ecotoxicity across all
treatment scenarios. Detailed review of model results shows that ecotoxicity of electricity
production is dominated by the presence of nuclear energy in the Arizona/New Mexico regional
grid mix and release of vanadium during the fuel extraction process. Nuclear energy contributes
nearly 19% of the region's fuel resources (Table 2-5).
Effluent release is a minor, but non-negligible, contributor to ecotoxicity impact for the
Baseline Scenario, Scenario 1, Scenario 2, and Scenario 4. Zinc is the primary pollutant
contributing to ecotoxicity of effluent release. Scenario 3 (RO) nearly eliminates ecotoxicity
impacts related to effluent discharge. However, the increased energy demand of the RO
treatment scenario significantly increases net ecotoxicity impact relative to the other treatment
EP-C-17-041; WA 4-77
3-15
-------
Section 3: Life Cycle Impact Baseline Results
configurations. Toxicity impacts associated with brine disposal are not included here, as it is
assumed deep well injection sequesters brine away from any receiving environment. However,
this is a limitation of the current model and should be evaluated in future work.
The increased electricity demand of the full diversion scenario (Scenario 4) contributes to
a moderate increase in ecotoxicity impact. This increase is significant enough that it minimizes
overlap of uncertainty bounds for Scenario 4 with the Baseline, Scenario 1, and Scenario 2. The
latter three scenarios have similar ecotoxicity impacts. Scenario 3 (RO) has the highest
ecotoxicity impact across all scenarios.
40
30
20
10
0
-10
-20
-24.3
11.6
Baseline
S1 - Sidestream
Filtration
S2 - T ertiary Filters
S3 - Reverse
Osmosis
S4 - Zero Discharge
& Main Plant - Energy Use 0 Effluent Reuse
~ Effluent Release H Post-Secondary Treatment
~ Primary and Secondary Treatment ~ Brine Injection
O Total O Monte Carlo Mean
SBiogas Cleaning and Combustion
~ Sludge Processing and Disposal
~ Effluent Diversion
b) 30
25
I 15
Baseline
SI - Sidestream S2 - Tertiary Filters
Filtration
S3 - Reverse
Osmosis
S4 - Zero Discharge
U WWTP Process Emissions
~ Chemicals
E3 Composting
~ Effluent Release
~ Materials
E3 Energy
E£ Landfill
H Water Reuse
E Land Application
• Total
E3 Transport
~ Biogas Combustion
H Effluent Diversion
~ Avoided Product
Figure 3-11. Ecotoxicity results for each treatment scenario, including uncertainty ranges as
the 5th and 95th percentile results from Monte Carlo simulations. Black uncertainty bars
represent combined uncertainty estimates for both shared and unique LCI data for each
scenario, and blue bars include uncertainty estimates only for the LCI inputs that are
unique to individual scenarios. Non-overlapping areas of the two uncertainty bars indicate
uncertainty that is shared across scenarios. Panels a and b show results aggregated
according to the different categorizations described in Table 3-1.
EP-C-17-041; WA 4-77
3-16
-------
Section 3: Life Cycle Impact Baseline Results
3.4.2 Human Health—Particulate Matter Formation
Figure 3-12 presents particulate matter formation potential results by treatment process
(a) and major driver (b). Panel a indicates that main plant electricity consumption and post-
secondary treatment processes are the main contributors to particulate matter formation. Panel b
illustrates that the majority of post-secondary treatment impact comes from chemical
consumption. Biogas combustion contributes moderately to particulate matter impact for all
treatment scenarios.
Sludge processing and disposal shows a minor net reduction in impact in panel a. The
results in panel b demonstrate that composting process emissions are one of the main drivers of
particulate matter impact and offset the benefit of avoided energy and fertilizer production.
Avoided fertilizer production (labeled avoided product in panel b) yields a larger benefit here
than has been demonstrated for other impact categories.
The uncertainty assessment results indicate that most uncertainty is attributable to
treatment processes that are common to all scenarios. The upper end of the uncertainty range for
Scenario 2 is an exception to that and is attributable to chemical consumption. Given this, the
analysis is unable to distinguish the particulate matter formation impact of Scenarios 2 and 3.
However, the lower bound of Scenario 3 is higher, giving us greater confidence that the Baseline,
Scenario 1, and Scenario 4 will yield reduced impact in this category relative to the RO treatment
option.
EP-C-I7-04I; WA 4^77
3-17
-------
Section 3: Life Cycle Impact Baseline Results
1.4E-04
1.2E-04
1.0E-04
8.0E-05
6.0E-05
4.0E-05
2.0E-05
O.OE+OO
-2.0E-05
^8.93E-5
1 ™
&
4.52E-5
4.67E-5
i :
mk
5.09E-5
Baseline
S1 - Sidestream
Filtration
S2 - Tertiary Filters
~ Main Plant - Energy Use 0 Effluent Reuse
~ Effluent Release ~ Post-Secondary Treatment
~ Primary and Secondary Treatment ~Brine Injection
O T otal O Monte C arlo Mean
S3 - Reverse S4 - Zero Discharge
Osmosis
~ Biogas Cleaning and Combustion
~ Sludge Processing and Disposal
~ Effluent Diversion
Baseline SI - Sidestream S2 - Tertiary Filters
Filtration
~ WWTP Process Emissions ~ Energy
~ Chemicals B Landfill
E3 C om posting 9 W ater Reuse
~ Effluent Release ~ Land Application
~ Materials • T otal
S4 - Zero Discharge
El Transport
~ Biogas Combustion
~ Effluent Diversion
~ Avoided Product
Figure 3-12. Human health—particulate matter formation results for each treatment scenario,
including uncertainty ranges as the 5th and 95th percentile results from Monte Carlo
simulations. Black uncertainty bars represent combined uncertainty estimates for both
shared and unique LCI data for each scenario, and blue bars include uncertainty estimates
only for the LCI inputs that are unique to individual scenarios. Non-overlapping areas of
the two uncertainty bars indicate uncertainty that is shared across scenarios. Panels a and
b show results aggregated according to the different categorizations described in Table 3-1.
3.4.3 Human Health Toxicity—Cancer Potential
Figure 3-13 presents human health toxicity—cancer potential results by treatment process
(a) and major driver (b). Taken together, panels a and b demonstrate that electricity consumption
is the predominant driver of toxicity cancer impact. With the exception of chemical use in
Scenario 2, all of the processes and driver categories depicted in Figure 3-13 are linked to
electricity consumption or production. As with ecotoxicity results, the nuclear fuel extraction
process contributes most of the impact associated with electricity production. Emissions of
arsenic (V), lead, and mercury are responsible for this impact.
EP-C-17-041; WA 4-77
3-18
-------
Section 3: Life Cycle Impact Baseline Results
While effluent release does not strongly contribute to baseline results, detailed review of
the openLCA model reveals that the positively skewed uncertainty range for the Baseline
Scenario and Scenarios 1, 2, and 4 is strongly influenced by the outlier metal concentrations in
historical water quality data. Triangular distributions are used as a conservative estimate of
uncertainty for metal effluent concentrations and lead to uncertainty ranges that skew towards
higher values for the non-RO scenarios. Arsenic released to water is the primary pollutant
contributing to toxicity cancer potential in the higher end of the uncertainty range. It should be
noted that the influence of these outliers would likely be minimal under average operating
conditions and is likely enhanced by the Monte Carlo simulation approach. The minimal
uncertainty associated with the RO treatment process indicates the effectiveness of this
membrane technology in removing pollutants that contribute to human toxicity cancer potential.
f—
V
1.0E-07
8.0E-08
6.0E-08
4.0E-08
2.0E-08
0.0E+00
-2.0E-08
-4.0E-08
1.20E-8
1.36E-8
1.71E-8
Baseline
H Main Plant - Energy Use El Effluent Reuse
~ Effluent Release ~ Post-Secondary Treatment
~ Primary and Secondary Treatment ~ Brine Injection
O Total O Monte C arlo Mean
5.11E-8
1.98E-8
SI - Sidestream S2 - Tertiary Filters
Filtration
S3 - Reverse S4 - Zero Discharge
Osmosis
13 Biogas Cleaning and Combustion
~ Sludge Processing and Disposal
~ Effluent Diversion
6.0E-08
5.0E-08
4.0E-08
3.0E-08
2.0E-08
1.0E-08
0.0E+00
-1.0E-08
Baseline
SI - Sidestream S2 - Tertiary Filters S3 - Reverse S4 - Zero Discharge
—Filtration Osmosis—
iWWTP Process Emissions
~ Chemicals
~ Composting
~ Effluent Release
~ Materials
ED Energy
E3 Landfill
~ Water Reuse
E3 Land Application
• Total
~ Transport
~ Biogas Combustion
~ Effluent Diversion
~ Avoided Product
Figure 3-13. Human health toxicity—cancer potential results for each treatment scenario,
including uncertainty ranges as the 5th and 95th percentile results from Monte Carlo
simulations. Black uncertainty bars represent combined uncertainty estimates for both
shared and unique LCI data for each scenario, and blue bars include uncertainty estimates
only for the LCI inputs that are unique to individual scenarios. Non-overlapping areas of
the two uncertainty bars indicate uncertainty that is shared across scenarios. Panels a and
b show results aggregated according to the different categorizations described in Table 3-1.
EP-C-17-041; WA 4-77
3-19
-------
Section 3: Life Cycle Impact Baseline Results
3.4.4 Human Health Toxicity—Noncancer Potential
Figure 3-14 presents human toxicity—noncancer results by treatment process (a) and
major driver (b). Taken together, panels a and b demonstrate that electricity consumption is the
predominant driver of toxicity noncancer impact. All the processes and driver categories
depicted in Figure 3-14 are linked to electricity consumption or production. Emissions of lead,
mercury, and arsenic (V) are responsible for this impact.
As described for human toxicity—cancer, while effluent release does not strongly
contribute to baseline results, detailed review of the open LCA model reveals that the positively
skewed uncertainty range is strongly influenced by the outlier metal concentrations in historical
water quality data. Triangular distributions are used as a conservative estimate of uncertainty for
metal effluent concentrations and lead to uncertainty ranges that skew towards higher values for
the non-RO scenarios. Arsenic released to water is the primary pollutant contributing to toxicity
noncancer potential in the higher end of the uncertainty range It should be noted that the
influence of these outliers would likely be minimal under average operating conditions and is
likely enhanced by the Monte Carlo simulation approach. The minimal uncertainty associated
with the RO treatment process indicates the effectiveness of this membrane technology in
removing pollutants that contribute to human toxicity noncancer potential.
2.0E-05
1.5E-05
-5.0E-06
Baseline
SI - Sidestream S2 - Tertiary Filters S3-Reverse S4 - Zero Discharge
Filtration Osmosis
~ Main Plant - Energy Use El Effluent Reuse
El Effluent Release ~ Post-Secondary Treatment
~ Primary and Secondary Treatment ~ Brine Injection
O T otal O Monte C arlo Mean
~ Biogas Cleaning and Combustion
~ Sludge Processing and Disposal
0 Effluent Diversion
1.5E-05
1.0E-05
5.0E-06
-5.0E-06
0.0E+00 &
Baseline
L_
m
SI - Sidestream S2 - Tertiary Filters S3 - Reverse S4 - Zero Discharge
Filtration Osmosis
~ WWTP Process Emissions
~ Chemicals
~ Composting
~ Effluent Release
~ Materials
E3 Energy
0 Landfill
H Water Reuse
Ei Land Application
• Total
E3 Transport
~ Biogas Combustion
13 Effluent Diversion
~ Avoided Product
EP-C-l 7-041; WA 4-77
3-20
-------
Section 3: Life Cycle Impact Baseline Results
Figure 3-14. Human health toxicity—noncancer potential results for each treatment scenario,
including uncertainty ranges as the 5th and 95th percentile results from Monte Carlo
simulations. Black uncertainty bars represent combined uncertainty estimates for both
shared and unique LCI data for each scenario, and blue bars include uncertainty estimates
only for the LCI inputs that are unique to individual scenarios. Non-overlapping areas of
the two uncertainty bars indicate uncertainty that is shared across scenarios. Panels a and
b show results aggregated according to the different categorizations described in Table 3-1.
3.5 Normalization and Standardization
3.5.1 Standard Normalization
Normalization is an optional step in LCIA that aids in understanding the significance of
impact assessment results. Normalization is conducted by dividing the impact category results by
a normalization factor. The normalization factor is typically the environmental burdens of the
region of interest either on an absolute or per capita basis. The results presented in this study are
normalized to reflect impacts on the basis of per person equivalents in the U.S. using TRACI2.1
normalization factors (Ryberg et al., 2014) and data from the Building Industry Reporting and
Design for Sustainability Database (Lippiatt et al., 2013) (Table 3-3). Some impact categories are
not included due to lack of available normalization factors.
Table 3-3. U.S. Per Capita Normalization Factors (Lippiatt et al., 2013; Ryberg et al., 2014).
lmp;icl ( .lienor\
I nil
Norniiili/iiliun
hiclor (I .S.. 200X)
lmp;icl per
Person-1
Source
1 !iiii\>pliic;ilimi
kg \ eq \ i-
(. (.1: l)
IMvigcl al, 2<) 14
Global warming
kg CO2 eq/yr
7.4E+12
24,334
Ryberg et al., 2014
Acidification
kg SO2 eq/yr
2.8E+10
92
Ryberg et al., 2014
Smog
kg O3 eq/yr
4.2E+11
1,381
Ryberg et al., 2014
Particulate Matter Formation
kg PM2 5 eq/yr
7.4E+9
24
Ryberg et al., 2014
Water Depletion
L H20 eq/yr
1.7E+14
559,027
Lippiatt et al., 2013
a Impact per person calculated using 2008 U.S. population of 304,100,000 (World Bank, 2017).
By multiplying impact results calculated in this study (impact per m3) by the annual
volume of domestic wastewater treated each year at the PR WWTP (4.85 MGD or 6,705,006
m3/yr [Section 1.2]), dividing by the service population (85,000 residential customers [Section
1.2]), and dividing by per capita normalization factors, it is possible to calculate the approximate
annual contribution of wastewater treatment to the total per capita impact of a Santa Fe resident
in each impact category. This calculation excludes impacts from commercial, public, and
industrial sources, and therefore overestimates the impact from individuals. The results of this
calculation for the five treatment scenarios and environmental impact in six categories are
presented in Table 3-4.
EP-C-I7-04I; WA 4^77
3-21
-------
Section 3: Life Cycle Impact Baseline Results
Table 3-4. Normalized impact results, expressed as the percent of per capita impacts allocated
to wastewater treatment.
Impsicl Ciiloyorv
liiisolino
SI -
Sidcslrcsim
lilliitlion
S2 -
Tertisirv
Killers'
S3 -
Uc\ crsc
Osmosis
S4 - Zero
Discharge
Lulioplncalion IVik'iHiul
4^",.
4 1".j
\ r\>
IS",,
4.9%
Global Warming Potential
0.25%
0.26%
0.27%
0.42%
0.29%
Acidification Potential
0.18%
0.18%
0.20%
0.25%
0.19%
Smog Formation Potential
0.28%
0.29%
0.30%
0.42%
0.30%
Particulate Matter Formation
0.01%
0.02%
0.03%
0.03%
0.02%
Water Depletion
-2.0%
-1.8%
-0.52%
1.2%
-2.0%
Normalized results show that, of the impacts for which normalization factors are
available, eutrophication impacts make up the largest contribution to typical per capita impacts,
ranging from 2% to 5%. Impacts of GWP, acidification, and smog formation make up less than
1%, ranging from 0.2% to 0.4%. Normalized water depletion results demonstrate the widest
variability across treatment scenarios with a minimum normalized impact of -2% for Scenarios 1
and 4, and a maximum normalized impact of 1.2% for Scenario 3. Normalized results show that
impacts associated with water depletion are comparable to those of eutrophi cation for Scenario 3.
Water depletion results do not account for local water scarcity, placing further emphasis on the
importance of this inventory metric in the Santa Fe region.
The greater proportion of impacts made up by eutrophi cation is reasonable, as the direct
discharge of nutrients and other eutrophying constituents is one of the main components of a
WWTP. Similarly, because most of the wastewater is returned to the environment and is not
depleted (except for certain unit processes in Scenarios 2 and 3), the relatively small fractions of
per capita water depletion identified in these results are reasonable in the context of typical water
consumption. For reference, 559,027 liters of water per person per year (Table 3-3) equates to
405 gallons per person per day, which represents both direct use of water (e.g., drinking, bathing)
and indirect use associated with production of products and services used by the typical person
each day (e.g., agriculture).
3.5.2 Santa Fe GHG Inventory
The City of Santa Fe's Environmental Services Division analyzed GHG emissions from
all sources within city limits following a protocol developed by the Compact of Mayors (City of
Santa Fe, 2017). They found the average per capita GHG emissions for a Santa Fe resident to be
10 metric tons per year, compared to a New Mexico state average of 32 metric tons per year and
a U.S. average of 17 metric tons per year.
Table 3-5 shows the portion of Santa Fe resident per capita GHG emissions that would be
attributed to study treatment scenarios. Contributions range from 0.62% for Baseline and
Scenario 1 to 1.0% for Scenario 3. These contributions are higher than normalized results based
on TRACI 2.1 normalization factors (Table 3-3), as the City of Santa Fe (2017) estimates per
capita GHG emissions of 10 metric tons per year, compared to the 24 metric tons per year
national average estimated by Ryberg et al. (2014).
EP-C-I7-04I; WA 4^77
3-22
-------
Section 3: Life Cycle Impact Baseline Results
Table 3-5. Summary of treatment scenario GHG emissions, compared to Santa Fe per capita
emissions.
Parameter
Baseline
SI - Side-
stream
Filtration
S2 -
Tertiary
Filters
S3 -
Reverse
Osmosis
S4 - Zero
Discharge
Source
Wastewater-Based GHG Emissions, This Study
kg CO2 eq./m3 treated
0.79
0.80
0.83
1.29
0.84
This study
m3 treated per year
6.71E+06
6.71E+06
6.71E+06
6.71E+06
6.71E+06
This study
(4.85 MGD)
kg CO2 eq./year
5.27E+06
5.37E+06
5.58E+06
8.63E+06
5.65E+06
Calculated
Population served
85,000
85,000
85,000
85,000
85,000
Carollo, 2018
kg CO2 eq ./person/year
62
63
66
102
66
Calculated
Santa Fe GHG Emissions, City of Santa Fe
kg CO2 eq ./person/year
10,000
10,000
10,000
10,000
10,000
Citv of Santa Fe.
2017
WWTP fraction
0.62%
0.63%
0.66%
1.02%
0.66%
Calculated
3.5.3 Results Standardized to Nutrient Removal
Generally, model results throughout this study are standardized to the study's functional
unit, which is a cubic meter of treated wastewater. In studies such as this, however, standardizing
to a different unit of measure can provide a different perspective and help results interpretation.
Given the importance of nutrient removal for the PR WWTP, this study compared impact results
when standardized to the removal rates for TN, TP, and total nitrogen equivalents (N eq.) (i.e.,
using eutrophication potential characterization factors). Table 3-6 shows the removal rates
achieved for each of these three quantities by the treatment scenarios, both in terms of annual
mass removal and percent removal. The difference in removal rates between each scenario is
generally only 3-4% (the highest difference is for TP removal, where Scenario 2 and Scenario 3
achieve 7% better removal than Baseline), despite Scenarios 2 and 3 achieving effluent nutrient
concentrations that are generally less than half of the Baseline Scenario effluent nutrient
concentrations (Table 2-2). As such, when impacts are standardized to 1 kilogram of nutrient or
nutrient equivalent removed (instead of 1 cubic meter of water treated) the resulting trends are
largely unaffected, as illustrated in Figure 3-15.
Table 3-6. Nutrient removal performance of treatment scenarios expressed as total nitrogen
(TN) removal, total phosphorus (TP) removal, and total nitrogen equivalents (N eq.) removal.
Treatment Performance
Metric
Baseline
SI -
Sidestream
Filtration
S2 -
Tertiary
Filters
S3 -
Reverse
Osmosis
S4 - Zero
Discharge
TN (kg/yr removed)
499,576
499,576
506,281
512,987
499,576
TP (kg/yr removed)
85,833
87,845
92,204
92,204
85,833
N eq. (kg N eq./yr removed)3
1,387,989
1,402,675
1,439,455
1,446,094
1,387,989
TN (% removal)
95%
95%
96%
97%
95%
TP (% removal)
93%
95%
99.6%
99.6%
93%
N eq. (% removal)
95%
96%
98%
99%
95%
EP-C-17-041; WA 4-77
3-23
-------
Section 3: Life Cycle Impact Baseline Results
SI -
S2 -
S3 -
Treatment Performance
Sidestream
Tertiary
Reverse
S4 - Zero
Metric
Baseline
Filtration
Filters
Osmosis
Discharge
a Refer to Appendix A. 1 for method description and characterization factors.
1
0.8
0.6
0.4
0.2
0
AP
B SI S2 S3 S4
EP
i
0.8
0.6
0.4
0.2
0
B SI S2 S3 S4
SFP
i
0.8
0.6
0.4
0.2
0
B SI S2 S3 S4
CED
i
0.8
0.6
0.4
°o m nn m
B SI S2 S3 S4
FFD
i
0.8
0.6
0.4
0.2
0
-0.2
-0.4
-0.6
S2 S3 S4
GWP
ET
HHPM
i
0.8
0.6
0.4
0.2
0
1
0.8
0.6
0.4
0.2
0
nil
H
0.8
0.6
0.4
0.2
B SI S2 S3 S4
B SI S2 S3 S4
II |
B SI S2 S3 S4
1
0.8
0.6
0.4
0.2
0
HHC
HHNC
WS
WD
i
0.8
0.6
0.4
0.2
0
B SI S2 S3 S4
B SI S2 S3 S4
-1
rn
IHi
||^ S3 |
¦ Base Model ~ Per kg N removed ~ Per kg P removed ~ Per kg N eq. removed
Figure 3-15. Impact results standardized to 1 cubic meter of wastewater treated (black), 1
kilogram of nitrogen removed (yellow), 1 kilogram of phosphorus removed (green), or 1
kilogram of N eq. removed (blue). All results have been normalized to the absolute value of
the maximum impact/benefit for each metric/standardization approach combination, so
that the largest value is 1 and the smallest value is -1.
EP-C-17-041; WA 4-77
3-24
-------
Section 4: Sensitivity Analysis Results and Discussion
4. Sensitivity Analysis Results and Discussion
4.1 Important Parameters
ERG performed a general parameter sensitivity analysis to evaluate the model parameters
that contribute most to impacts, characterize their relative importance, and provide further
context to their baseline values. This section further explains differences in impact across
scenarios and, where applicable, discusses how comparable this study's results are to other,
similar systems. Sensitivity results are placed in the context of parameter uncertainty ranges used
in the Monte Carlo analysis (which is introduced in Section 2.3). Details on specific parameter
uncertainty and uncertainty distributions can be found in Appendix B.l.
To identify important parameters, ERG reviewed the detailed contribution results
illustrated in Section 3. Large impact contributions (generally >10% of total impact across
multiple metrics) were traced back to individual model parameters or groups of parameters. ERG
then varied each parameter individually by +/- 10% (this range is an arbitrary threshold used to
test sensitivity) and recalculated impact results, isolating the effect of each parameter and
providing an indication of its relative importance. Abbreviated sensitivity results are summarized
in Figure 4-1 (only the top two parameters for each impact category are displayed) while full
results are provided in Appendix F. Results represent the absolute value of the change in baseline
environmental impact associated with each +/- 10% change in parameter value. Where possible,
text in the following sections provides context on the realistic range of parameter variability for
study systems.
Acidification Potential
Main Electricity
Compost Emissions
Smog Formation Potential
0% 2%
Biogas Production
Main Electricity
Fossil Fuel Depletion
0% 20% 40% 60%
80% 100%
Main Electricity
Biogas Production
Water Depletion
0% 10% 20%
S2 Alum
Main Electricity
Human Health Particulate Matter Formation
0% 2% 4% 6'
Main Electricity
Compost Emissions
Eutrophication Potential
0% 2% 4%
Nutrient Emissions ¦
Land Application... J
Cumulative Energy Demand
0% 20% 40%
Main Electricity
Biogas Production
Global Warming Potential
0% 2% 4%
Biological GHG... J
Main Electricity
Ecotoxicity
0% 10%
Main Electricity
Biogas Production
Human Health Toxicity Cancer Potential
0% 10% 20% 30%
Main Electricity
Biogas Production
Human Health Toxicity Noncancer Potential
0% 10% 20%
Main Electricity
Biogas Production
¦ Baseline
¦ S2 - Tertiary Filters
S4 - Zero Discharge
¦ SI - Sidestream Filtration
S3 - Reverse Osmosis
Figure 4-1. Sensitivity of top two important parameters for each impact category. For full list
of parameters by impact category, see Appendix F.
EP-C-17-041; WA 4-77
4-1
-------
Section 4: Sensitivity Analysis Results and Discussion
4.1.1 Main Electricity
Electricity (energy use) from the main plant, consisting of electricity consumption for the
core treatment facility is an important driver for nine of the 12 environmental metrics. The core
treatment facility includes all primary, secondary, and sludge treatment processes. Electricity
demand of tertiary treatment processes, such as deep bed media filters and MF/RO, are not
included in the designated core treatment facility and are evaluated separately.
The importance of facility-wide electricity demand is not surprising and suggests that
increased process control and aerator efficiency resulting from recent upgrades to the biological
process support controls could have a meaningful effect on reducing impacts. The effect of core
facility energy use is common to all treatment options, and therefore does not provide an
opportunity to distinguish between treatment scenarios. Instead, this sensitivity result highlights
the importance of maintaining or improving core facility electrical efficiency and the potential
benefits available from reduced electricity consumption or reduced environmental impact of the
electrical grid.
The contractor compared the energy consumption estimate for the main plant (not
including additional unit processes included in Scenarios 1-4) presented in this case study
against previous studies of comparable systems. Falk et al. (2013) indicates energy demand
ranging from 0.5 kWh/m3 wastewater treated for a conventional activated sludge design to 1.4
kWh/m3 wastewater treated for an activated sludge with enhanced settling and RO. This study's
LCI baseline design is similar to the conventional activated sludge design outlined in that study.
PR WWTP records indicate an electricity use of 0.73 kWh/m3 wastewater treated, falling in a
similar range to Falk et al. (2013). Similarly, EPA's Life Cycle Cost Assessments of Nutrient
Removal Technologies in Wastewater Treatment Plants study (U.S. EPA, 2021a) estimated a
range of electricity use from 0.20-0.57 kWh/m3 wastewater treated for systems ranging from a
conventional activated sludge (Level 1) to systems that incorporate different types of biological
nutrient removal (Levels 2 and 3). These estimates are lower than the 0.73 kWh/m3 value used
for this study, which indicates the potential to optimize electricity use at the PR WWTP.
These comparisons confirm the magnitude of electricity consumption used in this case
study and suggest the results from this study may be comparable to similar facilities around the
country.
4.1.2 Compost Emissions
Process emissions released during the composting process were identified as important
parameters for the acidification and particulate matter formation potential impact categories. Half
of produced digestate is composted with yard waste at an onsite windrow composting facility.
However, only process emissions attributable to the digestate are included in the LCA model, as
yard waste is a separate material, and its emissions are not attributable to the wastewater system.
Ammonia is the main pollutant contributing to impact in these categories and contributes to
result uncertainty.
Emission factors in the literature for ammonia range from 1 .OE-4 to 0.12 kg NH3-N per
kg of feedstock N (Fukumoto et al., 2003; Hellebrand, 1998; Maulini-Duran et al., 2013). The
EP-C-I7-04I; WA 4^77
4-2
-------
Section 4: Sensitivity Analysis Results and Discussion
baseline ammonia emission factor, which is the middle of the three identified values (Hellebrand,
1998), indicates that 0.044 kg of NH3-N will be released per kg feedstock N. Another study,
which presents emission factors specifically for mixtures of dewatered biosolids and green or
woody waste, reports values in a comparable range but was not directly used here, as it pertains
to forced-air systems (Roe et al., 2004). Variability in compost emission factors is attributable to
variations in feed materials, management practices, and environmental conditions, not to system
type or technology.
Uncertainty in this parameter was assessed using a lognormal distribution with a
geometric standard deviation of 1.69, which results in a 95th percentile emission factor of
approximately 0.1 kg NH3-N per kg feedstock N. This emission factor is 130% greater than the
baseline emission factor. Results of the sensitivity analysis indicate that if emission levels
consistently fall in the high end of the range, acidification and particulate matter formation
potential impacts would increase by approximately 40% and 60%, respectively, for the Baseline
Scenario. Compost emissions are not affected by the selection of treatment scenario.
Compounding uncertainty in the emission factor is uncertainty in the quantity of nitrogen
present in digestate entering the compost process. Plant records indicate that on average, 5.8% of
digestate dry mass is nitrogen. The Monte Carlo analysis (results presented in Section 3) includes
the effect of varying nitrogen content between 1.7% and 8.2% of digestate dry mass.
Acidification and particulate matter formation results are less sensitive to digestate nitrogen
content than they are to emission factors, which have a wider range of potential values. However,
if high nitrogen contents (8.2%) and emission factors (0.1 kgNHa-N per kg feedstock N)
coincide, this can lead to an increase in acidification potential of nearly 70%. This value can be
compared to the 40% increase discussed in the previous paragraph that is due only to a high
emission factor. Higher digestate nitrogen content also increases the potential for land
application emissions and avoided fertilizer benefits.
The included estimates of ammonia emissions are expected to be representative of typical
windrow composting systems. However, given the contribution of ammonia emissions to
environmental impact and the fact that they are the primary pathway for nitrogen loss during
composting (Wong and Selvam, 2017), scientists are looking for ways to minimize compost
ammonia emissions. A review of gaseous composting emissions indicates that use of biofilters,
certain bulking agents (e.g., straw, sawdust, biochar), and lowering pH by adding
phosphogypsum all hold potential to reduce ammonia emissions (Sayara and Sanchez, 2021).
4.1.3 Biogas Production
This parameter sensitivity analysis indicates that the energy and climate and toxicity
metrics are sensitive to changes in biogas production. The main implication of biogas production
on the LCA model relates to the potential to increase or decrease energy recovery from produced
biogas. The model is informed by 47 days' worth of daily biogas production records from the PR
WWTP (2019). The 25th and 75th percentile values over that period are 223,835 and 238,875
standard cubic feet per day, respectively. Both values are within 4% of mean daily biogas
production, indicating relatively stable production and low potential for considerable changes in
impact resulting from this source. Moreover, biogas production is expected to be consistent
EP-C-I7-04I; WA 4^77
4-3
-------
Section 4: Sensitivity Analysis Results and Discussion
across treatment scenarios and therefore has more bearing on the absolute magnitude of impact
results than on the relative environmental performance between treatment scenarios.
Even if biogas production remains stable across scenarios and over time, the balance of
associated benefits and impacts will change according to changes in the displaced energy mix.
Improvements in grid electricity environmental performance will reduce the benefits of
anaerobic digestion, while displacement of dirtier electricity sources would enhance system
environmental benefits.
Sensitivity results show that a 10% increase in biogas production leads to decreases of 6-
44% and 8—74% in total cumulative energy demand and fossil fuel depletion impact,
respectively, depending on treatment scenario. More advanced treatment scenarios, with higher
energy demand, are less sensitive to changes in biogas production, as biogas energy production
represents a smaller portion of total energy demand. Large changes in impact, such as the 74%
decrease in the Scenario 4 FFDP, resulting from modest (10%) changes in biogas production are
a result of small values of net impact. This results from a balance between impacts associated
with process operation and benefits resulting from avoided product credits (see Figure 3-7 for
illustration). When net impacts are small, even small changes in impact can have a large effect
on impact potential. Given this and the small reported variability in daily biogas production, this
parameter is not expected to strongly influence environmental impacts at the PR WWTP.
4.1.4 Nutrient Emissions
Nutrient emissions are the most sensitive parameter for the eutrophication potential
impact category. The sensitivity of other impact category results to nutrient emissions is
negligible. Eutrophi cation sensitivity results in Figure 4-1 (full results in Figure F-l) show that in
Scenario 1 and the Baseline, a 10% change in TN and TP emissions leads to 7% and 7.5%
changes in eutrophi cation potential, respectively. Scenarios with more advanced nutrient removal
processes are less sensitive to a 10% change in nutrient emissions, as the absolute change in
emitted nutrients is lower and other sources share more of the eutrophi cation burden.
For the Baseline scenario, 5 mg/L and 1 mg/L are the expected effluent concentrations
for TN and TP. Expected maximum effluent concentrations are <10 mg/L and <2.5 mg/L (or
<100% and <150% greater than expected average concentrations). Given the results of the
sensitivity analysis reported above, if effluent concentrations of nearer to 10 mg/L and 2 mg/L
are sustained over time, this would lead to an approximate 75% increase in estimated baseline
eutrophi cation potential. However, sustained effluent concentrations in this range are not
expected. In comparison, nutrient emission uncertainty in the LCI model is estimated using a
lognormal distribution with a geometric standard deviation of 1.54 for all scenarios. In the
Baseline Scenario, this distribution produces an 95th percentile emission value of approximately
10 mg/L for TN. Uncertainty in the technological performance of treatment processes is
quantified in the blue error bars in Figure 3-2, which at their maximum represent a 46% increase
relative to median eutrophi cation potential for the Baseline Scenario.
The LCA results presented in Section 3.1.1 use the TRACILCIA method, which assesses
generalized eutrophication applicable to both freshwater and marine environments. More detail
on TRACI eutrophication modeling can be found in Section A. 1. The model uses TN and TP
EP-C-I7-04I; WA 4^77
4-4
-------
Section 4: Sensitivity Analysis Results and Discussion
when characterizing impact, aggregating the more specific chemical forms of both nutrients.
While the current version of TRACI does not consider an availability factor (effect factor), it has
been recognized that such a factor can influence specific estimates of eutrophication potential
(Norris, 2002). The availability factor is a measure of bioavailability and represents the fraction
of a specific chemical compound that is plant-available and therefore capable of contributing to
eutrophi cation in a defined time period. The results of a eutrophi cation potential sensitivity
analysis that consider nutrient bioavailability are presented in Section 4.2.
4.1.5 Land Application Emissions
Land application emissions are the second most important parameter group, next to
nutrient emissions in wastewater effluent, for the eutrophi cation potential impact category. A
10% change in land application emissions leads to a 1.4-3.9% change in eutrophi cation
potential, with Scenarios 2 and 3 being more sensitive. Aqueous emissions of phosphorus and
nitrate contribute equally to land application eutrophi cation potential impact.
Generalized emission factors were used to estimate field emissions resulting from
compost land application. Uncertainty exists regarding actual field emissions that would occur as
a function of application rate, method, timing, and subsequent weather conditions. It is expected
that this uncertainty would affect all treatment scenarios equally. The Monte Carlo analysis
assesses uncertainty in land application field emissions by using identified emission factors as
the mean and applying a geometric standard deviation of 1.69. Using this distribution, the 95th
percentile emission factor estimates are approximately 2.5 times higher than the mean.
4.1.6 Biological GHG Emissions
GHG emissions from the biological treatment processes were identified as one of two
parameter groups with the greatest influence on global warming potential. Sensitivity results
show that a 10% change in process biological GHG emissions, which come from nitrous oxide
and methane, lead to a 3—5% change in net global warming potential. Emissions from the
biological process are constant across scenarios but have wide ranges of uncertainty. The nitrous
oxide emission factor in the Baseline Scenario assumes that 0.36% of total Kjeldahl nitrogen
(TKN) influent to the biological process is released as nitrous oxide. The limited sample size in
Chandran (2012) found this value to vary between 0.09% and 0.62%. Baseline methane
emissions are estimated using a methane correction factor (MCF) of 0.05, which represents the
degree to which a system is anaerobic and capable of producing methane, with a potential range
of 0 to 0.1 (IPCC, 2006).
The Monte Carlo analysis applies a lognormal distribution with a geometric standard
deviation of 1.69 to the baseline estimate of nitrous oxide and methane emissions. This
distribution produces a 95th percentile estimate of nitrous oxide emissions that is approximately
25% greater than the upper bound nitrous oxide emission factor reported by Chandran (2012),
and a 95th percentile estimate of methane emissions that approximates the value associated with a
an MCF of 0.1. As these values are 130% and 100% greater than baseline inputs, corresponding
increases to GWP would be on the order of 40-70%, depending on scenario, if these emission
levels were sustained over time. The described variability in emission factors indicates that
EP-C-I7-04I; WA 4^77
4-5
-------
Section 4: Sensitivity Analysis Results and Discussion
uncertainty in GHG emissions has the capacity to considerably influence net GWP impact
results.
4.1.7 Scenario 2 (Tertiary Filters) Alum
Alum used in the tertiary deep bed filters presented in Scenario 2 was identified as an
important contributor to the water depletion and particulate matter formation impact categories.
Sensitivity results show that a 10% change in alum dose leads to a 4% and 30% change in
particulate matter formation and water depletion, respectively.
Considerable variation in the necessary alum dose is possible depending on the quantity
of phosphorus that needs to be removed. The Baseline Scenario requires 0.95 mg/L of
phosphorus removal with an uncertainty range of 0.85-2.45 mg/L using a triangular distribution,
which represents a 150% increase in alum consumption at the high end of the range. In situations
where prolonged use of elevated alum doses is required, impacts for the mentioned impact
categories would be considerably increased.
4.1.8 Scenario 3 (RO) Electricity
Electricity use from MF and RO presented in Scenario 3 was identified as an important
parameter, driving metrics such as acidification potential, particulate matter formation potential,
global warming potential, ecotoxicity, human health toxicity (cancer and noncancer potential)
and smog formation potential. Baseline electricity estimates for MF, RO, and brine injection are
assumed to be within +/- 20% of the actual value based on the estimates provided by Carollo
Engineers, which is double the +/-10% range used in the sensitivity analysis. While
environmental impact results are sensitive to RO electricity demand compared to other
parameters, the expected variability in electricity consumption is low compared to other
important parameters (e.g., alum use, GHG emissions, nutrient emissions).
The electricity demand of RO and ancillary processes is comparable to similar systems
from the literature. The estimated electricity input to the MF and RO processes in this study is
0.33 kWh/m3 treated, while electricity input to deep well injection is 0.61 kWh/m3 treated. Falk
et al. (2013) reports electricity demand estimates for RO systems, including deep well injection,
that are approximately 31% lower than estimates used in this study. Energy demand of the RO
unit with deep well injection in the Falk et al. study can be roughly estimated by subtracting the
energy demand of Level 4 (0.72 kWh/m3 treated) from the energy demand of Level 5 (1.4
kWh/m3 treated), resulting in an estimate of 0.65 kWh/m3 treated. The Falk et al. study only
assumes that 50% of effluent is treated in the MF/RO unit processes, which contributes to the
observed difference in energy demand. In this study, two thirds of plant influent is sent to the
MF/RO unit processes.
4.2 Eutrophication Potential
The baseline eutrophication potential analysis discussed in Section 3 uses average U.S.
characterization factors from the TRACI 2.1 eutrophication method. These characterization
factors are based on the amount of algal growth that could be caused by each nutrient if it were
to reach a water body where it was limiting, assuming full bioavailability (Norris, 2002).
However, research performed in recent decades suggests that a fraction of nutrient compounds
EP-C-I7-04I; WA 4^77
4-6
-------
Section 4: Sensitivity Analysis Results and Discussion
found in WWTP effluent—organic nitrogen compounds in particular—may not be fully
bioavailable and would thus not lead to eutrophication of receiving waters (Bronk et al., 2010;
Filippino et al., 2011; Liu et al., 2012; Sattayatewa et al., 2009; Simsek et al., 2013; Urgun-
Demirtas et al., 2008).
Organic nitrogen present in WWTP effluent (hereafter referred to as effluent organic
nitrogen, or EON) can exist in a range of forms depending on the composition of WWTP
influent, the microbial community within the WWTP, and the specific biological processes used
by the WWTP. Although there is still uncertainty related to the exact source and composition of
EON (Mesfioui et al., 2012; Pehlivanoglu-Mantas and Sedlak, 2008), it is likely that a
considerable fraction is composed of the metabolic products of biological activity within the
WWTP itself (Parkin and McCarty, 1981a, 1981b). Of those metabolic products, a portion
consists of labile nitrogen-containing compounds including urea, dissolved free amino acids, and
nucleic acids, which can turn over on the order of seconds to days (Bronk, 2002; Bronk et al.,
2007). The more recalcitrant compounds are not as well-characterized, but research into similar
marine organic nitrogen pools suggests they may persist on the order of months to years (Benner,
2002; Bronk, 2002).
EON bioavailability also depends on the range of complex interactions that can occur
between it and a receiving environment. Nitrogen in otherwise recalcitrant forms can be
biotically mobilized when exposed to different microbial communities, or abiotically mobilized
when exposed to light or high salinities (Bronk et al., 2010; Filippino et al., 2011; Mesfioui et al.,
2012). For the PR WWTP, this suggests that EON compounds that may be recalcitrant in the
Santa Fe River may, over time, be biotically or abiotically acted upon in the Rio Grande,
becoming bioavailable.
To determine how bioavailability of organic nitrogen may affect eutrophication potential
results, a sensitivity analysis was performed. First, estimates of EON bioavailability were
compiled from the literature, as shown in Table 4-1. For this analysis, we used data only from
experiments that lasted 14 days or more (italicized values in Table 4-1) to be more representative
of the travel time from the PR WWTP to the Gulf of Mexico, which is likely on the order of
months to years. As most studies only report a range of results, we approximated the central
tendency as the average of each study's minimum and maximum value. This results in an overall
EON bioavailability average of 47% and range of 18-71%, compared to an assumption of 100%
for baseline model results.
Table 4-1. Summary of Measured Effluent Organic Nitrogen Bioavailability.
Study
WWTP Type
Test Length
(days)
Bioavailability
Ave.
Range
Filippino et al., 2011
BNR and five-stage Bardenpho
2
64%
31-96%
Bronk et al., 2010
Two different advanced BNR
plants
2
16%
9-23%
Liu et al., 2011
Eight different BNR plants
14
50%
32-68%
Urgun-Demirtas et al.,
2008
Pilot scale nitrification and TN
plant
14
40%
18-61%
EP-C-17-041; WA 4-77
4-7
-------
Section 4: Sensitivity Analysis Results and Discussion
Study
WWTP Type
Test Length
(days)
Bioavailability
Ave.
Range
Simsek et al., 2013
AS (min) and trickling filter
(max)
28
59%
47-71%
Sattayatewa et al., 2009
Four-stage Bardenpho
14
38%
28-48%
Sensitivity Analysis3
47%
18-71%
a Minimum and maximum values from italicized values, or studies with a test length of 14 days or greater.
Table abbreviations: AS = activated sludge, BNR = biological nutrient removal, TN = total nitrogen
Following the framework introduced in Norris (2002) and Seppala et al. (2004), the
bioavailability factors in Table 4-1 were applied to the original TRACI 2.1 eutrophication
potential characterization factors in the LCA model. Modified characterization factors were only
applied to effluent emissions from the PR WWTP and not to other nutrient emissions in the LCI.
Figure 4-2 shows these results in two formats. Panel a shows the contribution of major treatment
processes, along with the mean (white dot), 5th percentile (bottom error bar), and 95th percentile
(top error bar) results from the Monte Carlo simulation. For this sensitivity analysis, the Monte
Carlo simulation accounted for all previously discussed base model uncertainty data, in addition
to the range of bioavailability factors from Table 4-1 (a triangular distribution was assumed for
EON bioavailability, as no study identified a specific distribution type). Panel b shows the
contribution of individual chemical species to eutrophi cation potential impact. Baseline model
results from Section 3 are shown as black dashes for comparison purposes.
EP-C-17-041; WA 4-77
4-8
-------
Section 4: Sensitivity Analysis Results and Discussion
OX)
u
2.5E-02
2.0E-02
1.5E-02
1.0E-02
5.0E-03
0.0E+00
-5.0E-03
- 1.2E-02
- 1.2E-02
1.0E-02
T
Jr^ 4.61,-03 T^4.3E-03
1
—v—
Baseline
SI - Sidestream S2 - Tertiaiy Filters
Filtration
S3 - Reverse
Osmosis
S4 - Zero Discharge
B Main Plant - Energy Use
~ Effluent Release
~ Primary and Secondaiy Treatment
O Total
~ Effluent Reuse
~ Post-Sec ondaiy Treatment
~ Brine Injection
OMonte Carlo Mean
~ Biogas Cleaning and Combustion
~ Sludge Processing and Disposal
~ Effluent Diversion
—Baseline Model Results
1.6E-02
1.4E-02
1.2E-02
1.0E-02
8.0E-03
6.0E-03
4.0E-03
2.0E-03
0.0E+00
-J¦ 1.0E-02
>E-03
T4-
3E-03
Baseline
~ Phosphorus
~ Organic Nitrogen
• Total
S1 - Sidestream
Filtration
S2 - Tertiaiy Filters
S3 - Reverse
Osmosis
S4 - Zero Discharge
~ Nitrogen, other
~ Land Application
— Baseline Model Results
~ Chemical Oxygen Demand
¦ Other
Figure 4-2. Eutrophication potential sensitivity analysis results including uncertainty ranges as
the 5th and 95th percentile results from Monte Carlo simulations. Panel a shows results
aggregated according to major plant process. Panel b shows contributions to
eutrophication impacts from individual nutrient species.
A comparison of baseline model results (black dashes in Figure 4-2; 100% of nitrogen
bioavailable) to sensitivity analysis results (47% of nitrogen bioavailable) shows that reducing
the bioavailable fraction of nitrogen from 100% to 47% in the TRACI model decreased the
eutrophication potential impacts for all scenarios. Decreases are largest for Scenarios 2 and 3 at
19% and 13%, respectively, while decreases for the Baseline/Scenario 4 and Scenario 1 are only
8% and 10%, respectively, as EON makes up a smaller proportion of total impacts for those
scenarios. A decrease in overall impacts is important if these results are to be incorporated into a
wider normalization analysis; normalized eutrophication potential impacts presented in Section
3.5 (Table 3-4, first row) would be reduced by the same amounts listed above.
Baseline model results show Scenario 3 to result in the lowest eutrophication potential
impacts (4.9E-3 kg N eq./m3), with impacts from Scenario 2 being only 16% greater (5.7E-3 kg
N eq./m3) and considerable overlap between each scenario's uncertainty range. Because organic
nitrogen is the biggest difference between the effluent of Scenarios 2 and 3, applying a
EP-C-17-041; WA 4-77
4-9
-------
Section 4: Sensitivity Analysis Results and Discussion
bioavailability factor to EON reduces the difference in impacts between these two scenarios to
just 9%. These results suggest that as the bioavailability of EON decreases, eutrophication
potential impacts between Scenarios 2 and 3 become more similar.
Some research on the bioavailability of phosphorus in treated wastewater has shown
reduced bioavailability of certain phosphorus compounds (Li and Brett, 2015; Qin et al., 2015)
and even lower bioavailability for treatment systems that use phosphorus precipitation or
chemical removal processes, such as alum used in Scenario 216 (Ekholm and Krogerus, 1998; Li
and Brett, 2012). However, bioavailability depends on the type of phosphorus compounds
present in effluent (e.g., dissolved/particulate, reactive/non-reactive, hydrophobic/hydrophilic [Li
and Brett, 2015; Qin et al., 2015]), the determination of which is beyond the scope of this study.
Still, it is possible that the small amount of phosphorus in Scenario 2's effluent (phosphorus
represents 6.5% of the total impact for Scenario 2 in Figure 4-2b) would be in a stable metal
complex and relatively less bioavailable than phosphorus in other scenario effluents. This would
further reduce the eutrophi cation impacts of Scenario 2 relative to other scenarios.
In recent decades, considerable research into the bioavailability of EON has been
performed (studies reviewed here are not exhaustive). This research has been motivated, in part,
by the continual advancement of wastewater nutrient removal technologies and their
encroachment on technological limits of organic nitrogen removal in particular (Lewis et al.,
2011). Still, methods for determining EON bioavailability are not perfect, as noted by most
authors cited in Table 4-1. First, bioavailability is generally determined by a measurement of the
net change in organic nitrogen concentration. Over test durations of days or weeks, researchers
acknowledge there is likely continual turnover of some portion of the organic nitrogen pool
(representing a contribution to biological activity) that cannot be quantified without more
advanced and rarely used measurement techniques such as molecular tracking (Bronk et al.,
2010; Mesfioui et al., 2012). For example, Fourier transform ion cyclotron mass spectrometry
was used to show that under one 14-day bioassay experiment, 79-100% of the compounds
present at the start of the experiment were replaced with new compounds produced during the
experiment (Bronk et al., 2010; Mesfioui et al., 2012). Organic nitrogen can also be highly
abiotically reactive (Bronk et al., 2010; Filippino et al., 2011; Mesfioui et al., 2012), resulting in
partial mobilization of nitrogen under conditions that may not be simulated using standard
bioassay methods. These factors suggest that the ranges of EON bioavailability reported in the
literature (e.g., Table 4-1) may be underestimating its full bioavailability.
At its point of discharge, PR WWTP effluent often makes up the majority of flow in the
Santa Fe River, as upstream water allocations often exceed natural flow and have resulted in the
Santa Fe River being characterized an unclassified intermittent stream (NMED, 2012). As such,
the ecology of the Santa Fe River downstream from the PR WWTP is highly dependent on the
quality and quantity of PR WWTP effluent. An assimilative capacity study was performed from
2017 to 2018 to develop a water quality model to understand how changes in PR WWTP effluent
quality would affect Santa Fe River water quality below the point of discharge (Leonard Rice
Engineers, 2018). As part of the study, water quality was sampled over several seasons along a
transect in the Santa Fe River. However, sampling was conducted at a time when PR WWTP
16 Scenario 1 also uses a chemical precipitation process for phosphorus removal, but uses magnesium to precipitate
phosphate ions to create struvite, which can be used directly as a bioavailable fertilizer.
EP-C-17-041; WA 4-77
4-10
-------
Section 4: Sensitivity Analysis Results and Discussion
effluent TN concentrations ranged from 4 to 7 mg/L and were at times dominated by ammonia,
representing conditions unlike any that would be encountered in current study scenarios (where
TN concentrations range from 2 to 5 mg/L with negligible ammonia concentrations). Sampling
and modeling results of the assimilative capacity study suggested that the Santa Fe River
downstream of the discharge point was highly dynamic because of a reach of restored wetlands,
lending considerable uncertainty to how the Santa Fe River would respond under the much lower
nutrient loading regimes considered in that study and here. In terms of organic nitrogen, one
sample transect showed a small spike in concentration through the wetland with subsequent
declines, consistent with heightened biological activity and generation of organic material in the
wetland. No discussion as to the reactivity, persistence, or bioavailability of this organic nitrogen
pool was provided.
Ultimately, the results of this analysis and evidence from the literature suggests that the
eutrophication potential impacts from organic nitrogen could be variable in time and space. In
the Santa Fe River just downstream of the PR WWTP discharge, it is possible that EON may be
less bioavailable as it travels through a limited range of biotic and abiotic conditions in that river
reach. Eutrophi cation potential impacts could therefore be towards the lower end of the range
displayed in Figure 4-2, and impacts between Scenarios 2 and 3 would be more similar.
However, as those compounds travel through different environments (including wetlands, the
range of conditions along the length of the Rio Grande, and ultimately the Gulf of Mexico), those
EON compounds will have been exposed to countless microbial consortiums, light conditions,
and salinity regimes, all of which have been shown to make EON more bioavailable. In the
context of a global LCA, EON bioavailability may therefore be towards the upper end of the
range in Table 4-1 (71%), or greater still given the limitations of standard bioassay methods.
From this wider perspective, eutrophi cation potential impacts may be more closely approximated
by the baseline model results that assume all nutrients will, at some time, become bioavailable.
4.3 Global Warming Potential Characterization Factors
In this sensitivity analysis, the effect of using GWP factors from the two most recent
IPCC Assessment Reports—the Fifth Assessment Report (IPCC, 2013) and Fourth Assessment
Report (Pachauri and Reisinger, 2007)—was evaluated. GWP factors are the values used to
transform the emission of all molecules that have heat trapping potential into a standardized unit.
The standardization process takes CO2 as its reference value setting its value to 1, with all other
factors being set relative to that standard (i.e., kg CO2 eq.). There are many parameters that
determine CO2 eq. values, and the scientific basis for this determination process continues to
evolve, with the IPCC reviewing and updating factors as the evidence improves. Table 4-2 shows
both the 2007 (AR4) and 2013 (AR5) factors for the primary GHGs resulting from the life cycle
of wastewater treatment.
EP-C-17-041; WA 4-77
4-11
-------
Section 4: Sensitivity Analysis Results and Discussion
Table 4-2. Comparison of IPCC Assessment Report 4 and Assessment Report 5 20- and 100-
year characterization factors.
Compound
Units
AR4
A]
R5
20-yr
100-yr
20-yr
100-yr
O
O
kg CO2 eq./kg
1
1
1
1
ch4
kg CO2 eq./kg
72
25
84
28
N20
kg CO2 eq./kg
289
298
264
265
The effect of different GWP factors on net GWP impacts depends on the relative
contribution of each GHG to the total GWP impacts of each treatment scenario. For example,
GWP impacts for the Baseline Scenario are fairly evenly mixed between CO2 emissions from
electricity use and methane (CH4) and nitrous oxide (N2O) emissions from the biological
treatment process (Section 3.2.3, Figure 3-8). Conversely, a greater proportion of impacts for
Scenario 3 come from CO2 emissions from electricity use. Therefore, total impacts from the
Baseline Scenario are likely to be more sensitive to the selection of GWP factors, given the
higher factor values for CH4 and N2O (Table 4-2).
GWP impacts for the different GWP factor scenarios are provided in Table 4-3 and
illustrated in Figure 4-3. Compared to base model results, 20-year factors produce the largest
increases in GWP impacts given the difference in 20-year vs. 100-year factors for CH4. Recently,
municipalities and states that track their GHG emissions have begun using 20-year factors (e.g.,
Howarth 2020) given the importance of methane emissions on GWP. Twenty-year factors also
tend to reduce the relative difference between treatment scenarios (e.g., the relative difference in
impacts between the Baseline Scenario and Scenario 3), mainly owing to the relative
contributions of CH4 and CO2 to net impacts. Still, the relative ranking of treatment scenarios
remains unchanged regardless of GWP factor selection, with the Baseline Scenario resulting in
the lowest GWP impacts and Scenario 3 resulting in the highest GWP impacts.
Table 4-3. Global Warming Potential (GWP) Sensitivity Analysis Results.
GWP Model
Net Impact (kg CO2 eq./m3 wastewater treated)
Baseline
SI
S2
S3
S4
Baseline model results
0.79
0.80
0.83
0.83
0.88
AR4: 100-year
0.60
0.62
0.65
0.65
0.69
AR5: 100-year
0.62
0.64
0.67
0.67
0.72
AR4: 20-year
1.38
1.40
1.44
1.44
1.49
AR5: 20-year
1.57
1.58
1.62
1.62
1.67
EP-C-17-041; WA 4-77
4-12
-------
Section 4: Sensitivity Analysis Results and Discussion
Global Warming Potential
"O
a;
+->
ro
oj
- 1.5
ro
5
OJ
to
ro
ro
cr
QJ
(N
o
u
ClD
0.5
h N m ^
oo
Baseline AR4:100-year AR5:100-year AR4: 20-year AR5: 20-year
Figure 4-3. Sensitivity of global warming potential results to selection of characterization
factors.
4.4 Electricity Grid Mix
In this sensitivity analysis, scenario models were run using different electrical grids to
determine the sensitivity of impacts to this model parameter selection. Models were run using
either a U.S. average grid mix or a 100% solar grid mix and compared to base model results,
which were calculated using a regional grid reflective of local electricity production. The specific
composition of the Arizona/New Mexico grid is provided in Table 2-5.
When conducting the sensitivity analysis, the electrical grid mix that serves the WWTP is
varied for each treatment scenario, while the electrical grid mix associated with background
processes remain constant. This is reasonable, since it is likely that background chemicals and
fuels are not produced in the same region of the U.S. that they are utilized. Results for all impact
categories were reproduced and compared to base model values (regional grid mix). Table 4-4
provides the results of the analysis, where the value is the percent change from the base model
results. Figure 4-4 illustrates the comparisons, but instead shows results on a scale that is
normalized to the absolute value of the maximum value for each metric across all scenarios, so
that results can be presented on a scale of -1 to 1.
In some cases, such as the cumulative energy demand change for Scenario 4 on solar, the
percent change is unusually high, which reflects a relatively large change standardized to an
original net impact that was close to zero (i.e., small). Results should therefore be interpreted in a
relative sense.
Changing from the regional grid to the U.S. average increases impacts for all metrics
across all scenarios, except for a minor decrease (<1%) of global warming potential for Scenario
3. The largest increases result for toxicity metrics and cumulative energy demand. Conversely,
changing from the regional grid to one entirely driven by solar uniformly decreases impacts.
These decreases are mostly much larger in magnitude than the changes that result from switching
EP-C-17-041; WA 4-77
4-13
-------
Section 4: Sensitivity Analysis Results and Discussion
to a U.S. average grid, illustrating the magnitude of improvements that could result from using
solar electricity.
Improvements from switching to 100% solar are likely underestimated here as well, as
the current modeling approach assumes that any electricity produced by the CHP system offsets
solar. If, for example, only the plant was run on solar and any electricity produced from the CHP
system fed into the existing regional grid, electricity offsets would be greater and net impacts
would be reduced.
EP-C-I7-04I; WA 4^77
4-14
-------
Section 4: Sensitivity Analysis Results and Discussion
Table 4-4. Change in impacts as a function of electricity grid.
Metric
Baseline
SI
S2
S3
S4
U.S.
Ave.
Solar
U.S.
Ave.
Solar
U.S.
Ave.
Solar
U.S.
Ave.
Solar
U.S.
Ave.
Solar
Eutrophication Potential
-0.1%
-1%
-0.1%
-1%
0.0%
-1.1%
0%
-2.7%
-0.1%
-0.6%
Acidification Potential
3.5%
-36%
4.0%
-37%
3.5%
-33%
10%
-49%
5.0%
-40%
Cumulative Energy Demand
3%
-269%
3%
-227%
2%
-176%
1.5%
-69%
2%
-141%
Global Wanning Potential
-0.2%
-59%
-0.2%
-60%
-0.2%
-59%
-0.5%
-71%
-0.3%
-63%
Fossil Fuel Depletion
6.0%
-348%
7.4%
-396%
19%
-1023%
8.7%
-245%
31%
-1320%
Smog Formation Potential
0.8%
-42%
0.9%
-43%
0.9%
-41%
2.3%
-55%
1.3%
-46%
Ecotoxicity
17%
-242%
18%
-230%
18%
-236%
22%
-153%
19%
-196%
Human Health—Cancer Potential
23%
-332%
23%
-302%
19%
-243%
22%
-153%
23%
-240%
Human Health—Noncanccr Potential
22%
-307%
22%
-285%
22%
-284%
22%
-153%
22%
-229%
Human Health—Particulate Matter Formation
16%
-43%
17%
-43%
10%
-26%
27%
-43%
23%
-46.0%
Water Depletion
0.2%
-0.8%
0.3%
-1%
1.0%
-3.1%
1.7%
-2.6%
0.4%
-0.9%
Figure 4-4. Illustration of electricity grid sensitivity analyses. Scale is normalized to the maximum value across all scenarios for
each metric.
EP-C-17-041; WA 3-77
4-15
-------
Section 4: Sensitivity Analysis Results and Discussion
4.5 Sludge Management
The final sensitivity analysis evaluates the importance of solids handling approaches.
Under base model conditions, the PR WWTP is assumed to send 50% of digestate to its
composting facility and the other 50% to the local landfill. Each disposal route entails a mix of
impacts and benefits. For example, composting digestate produces a usable product (compost)
that can offset fertilizer production and result in lower net impacts for some metrics, including
water depletion and particulate matter formation. However, the composting process produces
GHGs, including N2O and CH4. Similarly, when digestate is landfilled, anaerobic decomposition
produces CH4 emissions contributing to global warming potential, but prevents the release of
emissions contributing to eutrophication, acidification, and particulate matter formation
potential.
To evaluate impact tradeoffs that may occur from the PR WWTP either composting or
landfilling 100% of their digestate, ERG ran all model scenarios under both conditions. Table
4-5 summarizes the change in impacts relative to the base assumption of 50% composting and
50%) landfilling. Red shading indicates an increase in impact potential relative to the baseline
results, while green shading indicates improved environmental performance. Figure 4-5
illustrates the results of the sensitivity analysis, where results for each metric have been
standardized to the absolute value of the largest result across all scenarios (i.e., largest positive or
negative result) so that all values for that metric can be translated to a scale of -1 to 1.
Results show that impact sensitivities to the digestate processing approach are variable,
as 100%) composting improves global warming potential, fossil fuel depletion, smog formation
potential, ecotoxicity, and water depletion; while 100%> landfilling improves eutrophi cation
potential, acidification potential, cumulative energy demand, and human health toxicities.
Compared to composting, landfilling requires less energy and results in fewer land and
water impacts from nutrients or toxic pollutants if one assumes that the contents of the landfill
stay sequestered indefinitely. The LCA model assumes collection and offsite treatment of landfill
leachate. However, landfill liners and leachate collection systems tend to degrade over time,
resulting in slow leaks and potential impacts to groundwater resources.
Although composting does require additional energy, the resulting compost is assumed to
offset the need to produce traditional fertilizer, which is an energy- and water-intensive process.
Accordingly, composting 100% of the PR WWTP digestate is enough to reduce net impacts for
global warming potential by 4—7%, fossil fuel depletion by 2—7%, smog formation potential and
ecotoxicity by 7—13%, and water depletion by 21-88%).
EP-C-I7-04I; WA 3^77
4-16
-------
Section 4: Sensitivity Analysis Results and Discussion
Table 4-5. Change in impacts as a function of solids handling assumptions.
Metric
SI
S2
S3
S4
100%
Compost
100%
Landfill
100%
Compost
100%
Landfill
100%
Compost
100%
Landfill
100%
Compost
100%
Landfill
100%
Compost
100%
Landfill
Eutrophication Potential
14%
-14%
17%
-17%
33%
-33%
39%
-39%
14%
-14%
Acidification Potential
22%
-22%
22%
-22%
19%
-19%
15%
-15%
20%
-20%
Cumulative Energy Demand
20%
-20%
16%
-16%
12%
-12%
3%
-3%
9%
-9%
Global Wanning Potential
-7%
7%
-7%
7%
-7%
7%
-4%
4%
-6%
6%
Fossil Fuel Depletion
-4%
4%
-4%
4%
-11%
11%
-1%
1%
-13%
13%
Smog Formation Potential
-8%
8%
-8%
8%
-8%
8%
-5%
5%
-7%
7%
Ecotoxicity
-8%
8%
-8%
8%
-8%
8%
-5%
5%
-7%
7%
Human Health—Cancer Potential
13%
-13%
11%
-11%
9%
-9%
3%
-3%
8%
-8%
Human Health—Noncancer Potential
14%
-14%
13%
-13%
12%
-12%
4%
-4%
9%
-9%
Human Health—Particulate Matter
Formation
22%
-22%
21%
-21%
12%
-12%
11%
-11%
19%
-19%
Water Depletion
-21%
21%
-24%
24%
-83%
83%
-37%
37%
-21%
21%
Figure 4-5. Illustration of solids handling sensitivity analyses. Scale is normalized to the maximum value across all scenarios for
each metric. See Table 1-3 for abbreviation descriptions.
EP-C-17-041; WA 3-77
4-17
-------
Section 5: Conclusions
5. Conclusions
This study compares the environmental impact of the optimized PR WWTP (Baseline
Scenario) against four potential scenarios intended to reduce nutrient pollution in the Santa Fe
River. Baseline results present a best estimate of environmental performance for each treatment
scenario across 12 environmental impact categories. A Monte Carlo uncertainty assessment was
performed to quantify uncertainty in baseline LC A results. A parameter sensitivity analysis was
carried out to identify key parameters influencing impact results in each category (Section 4.1).
Additional sensitivity analyses were conducted to examine how impact results are affected by
selection of eutrophication and global warming potential characterization factors, electricity grid
mix, and sludge management practices.
Table 5-1 presents a summary of baseline LCA results from Section 3. For each metric,
results have been standardized by dividing each result by the maximum absolute value across all
scenarios so that each can be expressed on a scale of -1 to 1, where a value closest to -1 (1)
represents the scenario with the best (worst) performance in a particular impact category. No
weighting factors are applied in Table 5-1 or throughout this study, which implicitly gives equal
weight to each of the 12 metrics.
Table 5-1. Standardized Baseline Impacts for Each Study Treatment Scenario".
Metric
Standardized Impact Results
Baseline
Scenario
Sl-
Sidestream
Filtration
S2 -
Tertiary
Filters
S3 - Reverse
Osmosis
S4 - Zero
Discharge
Eutrophication Potential
1
0.83
0.43
0.37
1
Acidification Potential
0.7
0.7
0.79
1
0.76
Cumulative Energy Demand
0.13
0.16
0.21
1
0.3
Global Wanning Potential
0.61
0.62
0.65
1
0.68
Fossil Fuel Depletion
-0.36
-0.33
-0.13
1
-0.11
Smog Formation Potential
0.68
0.69
0.72
1
0.73
Ecotoxicity
0.32
0.35
0.35
1
0.48
Human Health Toxicity—
Cancer Potential
0.24
0.27
0.34
1
0.39
Human Health Toxicity—
Noncancer Potential
0.25
0.28
0.29
1
0.41
Human Health—Particulate
Matter Formation
0.51
0.52
0.9
1
0.57
Water Depletion
-1
-0.9
-0.26
0.57
-1
Water Scarcity
-1
-0.98
-0.86
0.39
-1
a - See Section 2 for a definition of metrics. Standardized baseline impacts obtained for each metric by dividing
each result by the maximum absolute value of that metric across all scenarios. For each metric, the value closest
to -1 represents the scenario with the best performance in a particular category, while the value closest to 1
represents the scenario with the worst performance. Standardized scales are meant to convey a measure of the
relative performance of scenarios across individual metrics. For full, unstandardized results, see Section 3.
EP-C-17-041; WA 4-77
5-1
-------
Section 5: Conclusions
Project goals emphasize the importance of eutrophication potential impacts among the 12
considered environmental metrics. LCA results show that Scenario 3 (RO) results in the lowest
eutrophi cation potential impacts. Baseline eutrophi cation potential results for Tertiary Filtration
indicate similar performance to RO, especially when considering results of the Monte Carlo
uncertainty assessment. As expected, eutrophi cation potential impacts are greatest for the
Baseline and Zero Discharge Scenarios, which represent current, optimized operation of the PR
WWTP and are only differentiated in terms of their discharge locations, not in the amount of
nutrient removal they provide. The Sidestream Filtration scenario realizes a 17% improvement in
eutrophication potential impact relative to the Baseline. The eutrophi cation potential sensitivity
analysis, which examines the influence of assumptions related to organic nitrogen
bioavailability, shows that the ranked performance of treatment scenarios remains unchanged.
However, the difference in impacts between Tertiary Filtration and RO is reduced, indicating
that when lower bioavailability of EON is assumed, the relative performance of Tertiary
Filtration improves.
Table 5-1 shows that reductions in nutrient pollution and eutrophication potential
associated with RO come at the expense of higher environmental impacts in all other
environmental categories (relative to the Baseline Scenario). While the same is true for Tertiary
Filtration, the magnitude of increases in environmental impact (relative to the Baseline Scenario)
are considerably reduced. Zero Discharge only results in a small increase in impacts relative to
the Baseline Scenario, owing to the additional energy that would be required to pump all effluent
that is not reused to the Rio Grande, but results in no reduction in eutrophication potential. The
moderate reduction in eutrophication potential associated with Scenario 1 (filtrate treatment)
comes at the expense of only minor increases in other environmental impacts. In terms of impact
per unit of nutrient removed, Scenario 1 is most similar to the baseline and is more efficient
across most metrics than Scenarios 2-4 (this can also be seen in Section 3.5.3, Figure 3-15).
The uncertainty assessment identifies several items not captured in the baseline results
presented in Table 5-1. It was stated above that Scenario 3 has higher impacts than all other
Scenarios for all impact categories except for eutrophication potential. However, Scenario 2
could result in comparable impacts to Scenario 3 for water depletion and particulate matter
formation due to uncertainty in the amount of chemicals needed to sufficiently reduce
phosphorus effluent concentrations. The relative similarity of water depletion impacts between
Scenarios 2 and 3 is, however, eliminated, when local water scarcity is considered. Water
scarcity impact results show much greater potential impacts for Scenario 3 than all other
scenarios due to brine disposal in water-scarce New Mexico, which renders water associated
with the injected brine unavailable for other purposes.
For human health toxicity cancer and noncancer potentials, which are driven by metal
discharges, there is uncertainty regarding the expected metals removal performance of the
Baseline Scenario, Scenario 1, Scenario 2, and Scenario 4 that suggest impacts could be higher in
all those scenarios compared to Scenario 3. Data on metal effluent concentrations (Table 2-2)
show maximum concentrations that are typically 1 to 3 orders of magnitude greater than average
concentrations. The influence of these outliers would likely be minimal under average operating
EP-C-I7-04I; WA 4^77
5-2
-------
Section 5: Conclusions
conditions, and toxicity cancer and noncancer potential impacts would likely be closer to the
expected baseline value for each scenario.
In Section 3.5.1, this study normalized LCA results based on U.S. per capita impacts
(Lippiatt et al., 2013; Ryberg et al., 2014) for the subset of metrics for which normalization
factors are available. Normalization is one way to identify the impact categories that are most
strongly influenced by the study system relative to typical emission rates for the wider region or,
in this case, country. Eutrophication potential impacts make up the largest contribution compared
to typical U.S. per capita impacts, ranging from 2 to 5%. This indicates that 2-5% of per capita
eutrophying emissions are attributable to the wastewater treatment services as provided by the
scenarios considered in this study. Normalized water depletion results demonstrate the widest
variability across treatment scenarios, with a minimum normalized impact of -2% for Scenarios 1
and 4, and a maximum normalized impact of 1.2% for Scenario 3. Water depletion results do not
account for local water scarcity, which would place further emphasis on the importance of this
inventory metric in the Santa Fe region. Normalized global warming potential, acidification
potential, and smog formation potential contribute less than 1% to total per capita impact in each
category. Normalization results therefore suggest that the choice of treatment system will be
most consequential for eutrophi cation and water depletion impacts.
Additional sensitivity analyses performed confirm the relative performance of treatment
scenarios that are reflected in the baseline results. However, the magnitude of differences
between treatment scenario impacts are influenced by sensitivity assumptions. The electricity
grid mix sensitivity analysis shows that if each scenario used electricity generated from 100%
solar power, potential impacts of Scenario 3 are much more comparable to other scenarios across
most metrics (potential impacts remain highest for water depletion and water scarcity).
Standardizing impact results to different measures of nutrients removed (Section 3.5.3) also
results in negligible change to the relative ranking of impacts across scenarios.
Certain potential environmental impacts associated with the RO treatment process
(Scenario 3) were not captured in this study but are worth noting. First, the RO treatment process
produces treated wastewater (RO permeate) with low levels of total dissolved solids that, without
modification, can be corrosive to equipment, leach metals from geological substrates, and be
toxic to aquatic organisms. This is of particular concern when RO permeate constitutes a
considerable share of total flow as would be the case seasonally in the Santa Fe River. Treated
RO effluent should be blended with natural waters (e.g., Rio Grande water) to increase total
dissolved solids and base ion concentrations such that effluent does not negatively impact
receiving environments. While beyond the scope of this study, other options exist for the reuse
and management of treated effluent, such as direct potable reuse or aquifer recharge. These
effluent reuse scenarios would likely reduce flow to the Santa Fe River and entail additional
impacts and benefits not considered here.
The second potential environmental impact of RO that was not captured in this study is
that the LCA model for Scenario 3 only considers impacts of brine disposal associated with the
energy required for deep well injection and water depletion. Water depletion impacts assume that
disposed brine is taken out of the local water cycle and does not mix with groundwater resources.
RO brine can be highly corrosive and is, by definition, a concentrated form of the constituents
that have been separated from the permeate that could lead to additional impacts to groundwater
EP-C-I7-04I; WA 4^77
5-3
-------
Section 5: Conclusions
resources (Ahmed et al., 2001; Chelme-Ayala et al., 2009). If RO brine intrudes into
groundwater aquifers used for drinking water or is otherwise mobilized in the environment,
current impact results would likely underestimate environmental impacts associated with brine
disposal.
One of the key challenges of the interpretation phase of an LCA study is consideration of
environmental/economic tradeoffs and how an individual or institutional decision-maker can or
should weigh impacts across multiple metrics. Weighting of environmental impacts can be used
to synthesize LCA results and determine the best option among alternatives. This study does not
apply weighting factors, nor should the implicit equal weighting in Table 5-1 or elsewhere in this
report be taken as an endorsement of equal importance across economic and environmental
metrics. As a next step, stakeholders should explore the incorporation of weighting factors so
that the results presented in this study can be used more directly within a decision-making
framework.
EP-C-I7-04I; WA 4^77
5-4
-------
Section 6: References
6. References
Ahmed, M., Shayya, W.H., Hoey, D., Al-Handaly, J., 2001. Brine disposal from reverse osmosis
desalination plants in Oman and the United Arab Emirates. Desalination 133, 135-147.
https ://doi .org/10.1016/S0011-9164(01 )80004-7
Bare, J., 2012. Tool for the Reduction and Assessment of Chemical and Other Environmental
Impacts (TRACI): User's Manual Version.
Bare, J., 2011. TRACI 2.0: The Tool for the Reduction and Assessment of Chemical and other
Environmental Impacts. Clean Technologies and Environmental Policy 13, 687-696.
https://doi.org/10.1007/sl0098-010-0338-9
Benner, R., 2002. Chapter 3 - Chemical Composition and Reactivity, in: Hansell, D.A. &
Carlson, C.A. (Eds.), Biogeochemistry of Marine Dissolved Organic Matter. Academic
Press, San Diego, pp. 59-90. https://doi.org/10.1016/B978-012323841-2/50005-l
Boldrin, A., Andersen, J.K., Miller, J., Christensen, T.H., Favoino, E., 2009. Composting and
compost utilization: accounting of greenhouse gases and global warming contributions.
Waste Management & Research 27, 800-812.
https://doi.org/10.1177/0734242X09345275
Boldrin, A., Neidel, T.L., Damgaard, A., Bhander, G.S., Miller, J., Christensen, T.H., 2011.
Modelling of environmental impacts from biological treatment of organic municipal
waste in EASEWASTE. Waste Management 31, 619-630.
https://doi.Org/10.1016/j.wasman.2010.10.025
Boijesson, G., Svensson, B.H., 1997. Nitrous oxide emissions from landfill cover soils in
Sweden. TellusB 49, 357-363. https://doi.org/10.3402/tellusb.v49i4.15974
Boulay, A., Bare, J., Benini, L., Berger, M., Lathuilliere, M.J., Manzardo, A., Margni, M.,
Motoshita, M., Nunez, M., Pastor, A.V., Ridoutt, B., Oki, T., Worbe, S., Pfister, S., 2018.
The WULCA consensus characterization model for water scarcity footprints: assessing
impacts of water consumption based on available water remaining (AWARE). The
International Journal of Life Cycle Assessment 23, 368-378.
https://doi.org/10.1007/sll367-017-1333-8
Bronk, D.A., 2002. Chapter 5 - Dynamics of DON, in: Hansell, D.A. & Carlson, C.A. (Eds.),
Biogeochemistry of Marine Dissolved Organic Matter. Academic Press, San Diego, pp.
153-247. https://doi.org/10.1016/B978-012323841-2/50007-5
Bronk, D.A., Roberts, Q.N., Sanderson, M.P., Canuel, E.A., Hatcher, P.G., Mesfioui, R.,
Filippino, K.C., Mulholland, M.R., Love, N.G., 2010. Effluent organic nitrogen (EON):
bioavailability and photochemical and salinity-mediated release. Environmental Science
& Technology 44, 5830-5835. https://doi.org/10.1021/esl01115g
Bronk, D.A., See, J.H., Bradley, P., Killberg, L., 2007. DON as a source of bioavailable nitrogen
for phytoplankton. Biogeosciences 4, 283-296. https://doi.org/10.5194/bg-4-283-2007
Carollo Engineers, 2018. Nutrient loading and removal optimization study. City of Santa Fe.
Carollo Engineers, 2017. Santa Fe water reuse feasibility study. City of Santa Fe and Santa Fe
County.
Chandran, K., 2012. Greenhouse nitrogen emissions from wastewater treatment operation: Phase
I. Water Environment Research Foundation.
Chelme-Ayala, P., Smith, D.W., El-Din, M.G., 2009. Membrane concentrate management
options: a comprehensive critical review. Canadian Journal of Civil Engineering 36,
1107-1119. https://doi.org/10.1139/L09-042
Ciroth, A., Muller, S., Weidema, B., 2012. Refining the pedigree matrix approach in ecoinvent.
EP-C-I7-04I; WA 4^77
6-1
-------
Section 6: References
City of Santa Fe, 2017. Greenhouse gas emissions in Santa Fe [WWW Document], URL
https://www.santafenm.gov/santafe_emissions
Clavreul, J., Guyonnet, D., Christensen, T.H., 2012. Quantifying uncertainty in LCA-modelling
of waste management systems. Waste Management 32, 2482-2495.
https://doi.Org/10.1016/j.wasman.2012.07.008
Daelman, M.R., van Voorthuizen, E.M., van Dongen, L.G., Volcke, El., van Loosdrecht, M.C.,
2013. Methane and nitrous oxide emissions from municipal wastewater treatment-results
from a long-term study. Water Science Technology 67, 2350-2355.
https://doi.org/10.2166/wst.2013.109
Ecoinvent Centre, 2010. Cumulative Energy Demand (CED) Method implemented in ecoinvent
data v2.2.
Edelen, A., Ingwersen, W., 2016. Guidance on data quality assessment for life cycle inventory
data (No. EPA/600/R-16/096). U.S. Environmental Protection Agency, National Risk
Management Research Laboratory.
Edelen, A., T. Hottle, S. Cashman, W. Ingwersen., 2019. The federal LCA commons elementary
flow list: background, approach, description and recommendations for Use [WWW
Document], URL
https://cfpub.epa.gov/si/si_public_record_Report.cfm7dirEntryIrN34725 l&Lab=NRMR
L (accessed 8.18.21).
Ekholm, P., Krogerus, K., 1998. Bioavailability of phosphorus in purified municipal
wastewaters. Water Research 32, 343-351.
Emmerson, R.H.C., Morse, G.K., Lester, J.N., Ph.D., Edge, D.R., 1995. The life-cycle analysis
of small-scale sewage-treatment processes. Water and Environment Journal 9, 317-325.
https://doi.Org/10.llll/j.1747-6593.1995.tb00945.x
Falk, M.W., Reardon, D.J., Neethling, J.B., Clark, D.L., Pramanik, A., 2013. Striking the balance
between nutrient removal, greenhouse gas emissions, receiving water quality, and costs.
Water Environment Research 85, 2307-2316.
https://doi.org/10.2175/106143013X13807328848379
Federal LCA Commons, 2021. Federal LCA Commons | Life Cycle Assessment Commons
[WWW Document], URL https://www.lcacommons.gov/ (accessed 7.20.21).
Filippino, K.C., Mulholland, M.R., Bernhardt, P.W., Boneillo, G.E., Morse, R.E., Semcheski,
M., Marshall, H., Love, N.G., Roberts, Q., Bronk, D.A., 2011. The bioavailability of
effluent-derived organic nitrogen along an estuarine salinity gradient. Estuaries and
Coasts 34, 269-280. https://doi.org/10.1007/sl2237-010-9314-l
Foley, J., De Haas, D., Hartley, K., Lant, P., 2010. Comprehensive life cycle inventories of
alternative wastewater treatment systems. Water Research 44, 1654-1666.
Frischknecht, R., Jungbluth, N., Althaus, H.-J., Doka, G., Dones, R., Heck, T., Hellweg, S.,
Hischier, R., Nemecek, T., Rebitzer, G., Spielmann, M., 2005. The ecoinvent database:
overview and methodological framework. The International Journal of Life Cycle
Assessment 10, 3-9. https://doi.org/10.1065/lca2004.10.18Ll
Fukumoto, Y., Osada, T., Hanajima, D., Haga, K., 2003. Patterns and quantities of NH3, N2O
and CH4 emissions during swine manure composting without forced aeration—effect of
compost pile scale. Bioresource Technology 89, 109-114. https://doi.org/10.1016/S0960-
8524(03)00060-9
GreenDelta, 2020. OpenLCA. Berlin, Germany.
EP-C-I7-04I; WA 4^77
6-2
-------
Section 6: References
Hellebrand, H.J., 1998. Emission of nitrous oxide and other trace gases during composting of
grass and green waste. Journal of Agricultural Engineering Research 69, 365-375.
https://doi.org/10.1006/jaer.1997.0257
Hong, J., Shaked, S., Rosenbaum, R.K., Jolliet, O., 2010. Analytical uncertainty propagation in
life cycle inventory and impact assessment: application to an automobile front panel. The
International Journal of Life Cycle Assessment 15, 499-510.
https://doi.org/10.1007/sll367-010-0175-4
Hottle, T., Ghosh, T., 2021. Regional electricity consumption mixes using trade data for
representative inventories. The International Journal of Life Cycle Assessment 26, 1211—
1222. https://doi.org/10.1007/sll367-021-01876-3
Howarth, R.W., 2020. Methane emissions from fossil fuels: Exploring recent changes in
greenhouse-gas reporting requirements for the State of New York. Journal of Integrative
Environmental Sciences 17, 69-81. https://doi.org/10.1080/1943815X.2020.1789666
Huijbregts, M.A.J., Steinmann, Z.J.N., Elshout, P.M.F., Stam, G., Verones, F., Vieira, M., Zijp,
M., Hollander, A., van Zelm, R., 2017. ReCiPe2016: a harmonised life cycle impact
assessment method at midpoint and endpoint level. International Journal of Life Cycle
Assessment 22, 138-147. https://doi.org/10.1007/sl 1367-016-1246-y
IPCC, 2013. Climate Change 2013: The physical science basis. Contribution of working group I
to the fifth assessment report of the Intergovernmental Panel on Climate Change. Stocker,
T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V.
Bex and P.M. Midgley (Eds). Cambridge University Press, Cambridge, United Kingdom
and New York, NY, USA.
IPCC, 2006. IPCC guidelines for national greenhouse gas inventories. Intergovernmental Panel
on Climate Change, National Greenhouse Gas Inventories Programme, IGES, Japan.
ISO, 2006a. ISO 14040: Environmental management — Life cycle assessment — Principles and
framework (No. ISO 14040:2006(E)). The International Organization for
Standardization.
ISO, 2006b. ISO 14044: 2006 Environmental management — Life cycle assessment —
Requirements and guidelines (No. ISO 14044:2006(E)), ISO 14044. The International
Organization for Standardization.
Keng, Z.X., Chong, S., Ng, C.G., Ridzuan, N.I., Hanson, S., Pan, G.-T., Lau, P.L.,
Supramaniam, C.V., Singh, A., Chin, C.F., Lam, H.L., 2020. Community-scale
composting for food waste: A life-cycle assessment-supported case study. Journal of
Cleaner Production 261, 121220. https://doi.org/10.1016/jjclepro.2020.121220
Leonard Rice Engineers, Inc., 2018. City of Santa Fe nutrient loading and optimization study.
Phase 3: assimilative capacity study. City of Santa Fe, New Mexico.
Levis, J.W., Barlaz, M.A., 2011. What is the most environmentally beneficial way to treat
commercial food waste? Environmental Science & Technology 45, 7438-7444.
https://doi.org/10.1021/esl03556m
Lewis, W.M., Wurtsbaugh, W.A., Paerl, H.W., 2011. Rationale for control of anthropogenic
nitrogen and phosphorus to reduce eutrophication of inland waters. Environmental
Science & Technology 45, 10300-10305. https://doi.org/10.1021/es202401p
Li, B., Brett, M.T., 2015. The relationship between operational and bioavailable phosphorus
fractions in effluents from advanced nutrient removal systems. International Journal of
Environmental Science and Technology 12, 3317-3328. https://doi.org/10.1007/sl3762-
015-0760-y
EP-C-I7-04I; WA 4^77
6-3
-------
Section 6: References
Li, B., Brett, M.T., 2012. The impact of alum based advanced nutrient removal processes on
phosphorus bioavailability. Water Research 46, 837-844.
https://doi.Org/10.1016/j.watres.2011.ll.055
Lippiatt, B.C., Kneifel, J., Lavappa, P., Suh, S., Greig, A.L., 2013. Building industry reporting
and design for sustainability (BIRDS): Technical manual and user guide (No. NIST
Technical Note 1814). National Institute of Standards and Technology.
Liu, H., Jeong, J., Gray, H., Smith, S., Sedlak, D.L., 2012. Algal uptake of hydrophobic and
hydrophilic dissolved organic nitrogen in effluent from biological nutrient removal
municipal wastewater treatment systems. Environmental Science & Technology 46, 713—
721. https://doi.org/10.1021/es203085y
Maulini-Duran, C., Artola, A., Font, X., Sanchez, A., 2013. A systematic study of the gaseous
emissions from biosolids composting: raw sludge versus anaerobically digested sludge.
Bioresource Technology 147, 43-51. https://doi.Org/10.1016/j.biortech.2013.07.118
Mesfioui, R., Love, N.G., Bronk, D.A., Mulholland, M.R., Hatcher, P.G., 2012. Reactivity and
chemical characterization of effluent organic nitrogen from wastewater treatment plants
determined by Fourier transform ion cyclotron resonance mass spectrometry. Water
Research 46, 622-634. https://doi.Org/10.1016/j.watres.2011.ll.022
Morelli, B., Cashman, S., Arden, S., Ma, M., Xin (Cissy), Turgeon, J., Garland, J., Bless, D.,
2019. Life cycle assessment and cost analysis of municipal wastewater treatment
expansion options for food waste anaerobic co-digestion (No. EPA/600/R-19/094). U.S.
Environmental Protection Agency, Washington, D.C.
Miiller Schmied, H., Eisner, S., Franz, D., Wattenbach, M., Portmann, F.T., Florke, M., Doll, P.,
2014. Sensitivity of simulated global-scale freshwater fluxes and storages to input data,
hydrological model structure, human water use and calibration. Hydrology and Earth
System Sciences 18, 3511-3538. https://doi.org/10.5194/hess-18-3511-2014
Nemecek, T., Kagi, T., 2007. Life cycle inventories of agricultural production systems (No. 15).
Ecoinvent Centre, Diibendorf and Zurich, Switzerland.
Nkoa, R., 2014. Agricultural benefits and environmental risks of soil fertilization with anaerobic
digestates: A review. Agronomy for Sustainable Development 34, 473-492.
https://doi.org/10.1007/sl3593-013-0196-z
NMED, 2012. Santa Fe river from Nichols Reservoir to the outfall of the Santa Fe wastewater
treatment facility use attainability analysis. New Mexico Environment Department.
Norris, G.A., 2002. Impact characterization in the Tool for the Reduction and Assessment of
Chemical and other environmental Impacts: methods for acidification, eutrophication,
and ozone formation. Journal of Industrial Ecology 6, 79-101.
https://doi.org/10.1162/108819802766269548
NREL, 2019. U.S. life cycle inventory database.
Ong, M.D., Williams, R.B., Kaffka, S.R., 2017. Comparative assessment of technology options
forbiogas clean-up (No. 500-11-020). California Energy Commission, Davis, California.
Pachauri, R.K., Reisinger, A., 2007. IPCC fourth assessment report. Intergovernmental Panel on
Climate Change.
Parkin, G.F., McCarty, P.L., 1981a. Sources of soluble organic nitrogen in activated sludge
effluents. Journal (Water Pollution Control Federation) 53, 89-98.
Parkin, G.F., McCarty, P.L., 1981b. Production of soluble organic nitrogen during activated
sludge treatment. Journal (Water Pollution Control Federation) 53, 99-112.
EP-C-I7-04I; WA 4^77
6-4
-------
Section 6: References
Pehlivanoglu-Mantas, E., Sedlak, D.L., 2008. Measurement of dissolved organic nitrogen forms
in wastewater effluents: Concentrations, size distribution and NDMA formation potential.
Water Research 42, 3890-3898. https://doi.Org/10.1016/j.watres.2008.05.017
Qin, C., Liu, H., Liu, L., Smith, S., Sedlak, D.L., Gu, A.Z., 2015. Bioavailability and
characterization of dissolved organic nitrogen and dissolved organic phosphorus in
wastewater effluents. Science of the Total Environment 511, 47-53.
http s: //doi. org/10.1016/j. scitotenv .2014.11.005
Rigby, H., Clarke, B., Pritchard, D.L., Meehan, B., Beshah, F., Smith, S.R., Porter, N.A., 2016.
A critical review of nitrogen mineralization in biosolids-amended soil, the associated
fertilizer value of crop production and potential for emissions to the environment. Science
of The Total Environment 541, 1310-1338.
https://doi.Org/10.1016/j.scitotenv.2015.08.089
Righi, S., Oliviero, L., Pedrini, M., Buscaroli, A., Delia Casa, C., 2013. Life Cycle Assessment
of management systems for sewage sludge and food waste: Centralized and decentralized
approaches. Journal of Cleaner Production 44, 8-17.
https://doi.Org/10.1016/j.jclepro.2012.12.004
Roe, S.M., Spivey, M.D., Lindquist, H.C., Thesing, K.B., Strait, R.P., 2004. Estimating ammonia
emissions from anthropogenic nonagricultural sources - draft final report. Emissions
Inventory Improvement Program, U.S. Environmental Protection Agency.
RTI International, 2010. DRAFT - Greenhouse gas emissions estimation methodologies for
biogenic emissions from selected source categories: solid waste disposal wastewater
treatment ethanol fermentation. U.S. Environmental Protection Agency.
Ryberg, M., Vieira, M.D., Zgola, M., Bare, J., Rosenbaum, R.K., 2014. Updated US and
Canadian normalization factors for TRACI 2.1. Clean Technologies and Environmental
Policy 16, 329-339.
Sattayatewa, C., Pagilla, K., Pitt, P., Selock, K., Bruton, T., 2009. Organic nitrogen
transformations in a 4-stage Bardenpho nitrogen removal plant and
bioavailability/biodegradability of effluent DON. Water Research 43, 4507-4516.
https://doi.Org/10.1016/j.watres.2009.07.030
Sayara, T., Sanchez, A., 2021. Gaseous emissions from the composting process: controlling
parameters and strategies of mitigation. Processes 9. https://doi.org/10.3390/pr9101844
Seppala, J., Knuuttila, S., Silvo, K., 2004. Eutrophication of aquatic ecosystems a new method
for calculating the potential contributions of nitrogen and phosphorus. The International
Journal of Life Cycle Assessment 9, 90-100. https://doi.org/10.1007/BF02978568
Serediak, N.A., Prepas, E.E., Putz, G.J., 2014. Eutrophi cation of freshwater systems, in: Treatise
on Geochemistry. Elsevier Science, pp. 305-323.
Simsek, H., Kasi, M., Ohm, J.-B., Blonigen, M., Khan, E., 2013. Bioavailable and biodegradable
dissolved organic nitrogen in activated sludge and trickling filter wastewater treatment
plants. Water Research 47, 3201-3210. https://doi.Org/10.1016/j.watres.2013.03.036
Slorach, P.C., Jeswani, H.K., Cuellar-Franca, R., Azapagic, A., 2019. Environmental
sustainability of anaerobic digestion of household food waste. Journal of Environmental
Management 236, 798-814. https://doi.org/10.1016/jjenvman.2019.02.001
SYLVIS, 2011. Biosolids Emissions Assessment Model (BEAM): Version 1.1.
Urgun-Demirtas, M., Sattayatewa, C., Pagilla, K.R., 2008. Bioavailability of dissolved organic
nitrogen in treated effluents. Water Environment Research 80, 397-406.
https://doi.org/10.2175/106143007X221454
EP-C-I7-04I; WA 4^77
6-5
-------
Section 6: References
U. S. EPA, 2022a. Reports on nutrient pollution [WWW Document], URL
https://www.epa.gov/nutrient-policy-data/reports-nutrient-pollution (accessed 2.7.22).
U.S. EPA, 2022b. Accelerating Nutrient Pollution Reductions in the Nation's Waters. Memo by
RadhikaFox, Assistant Administrator, Office of Water. April 5, 2022.
https://www.epa.gOv/system/files/documents/2022-04/accelerating-nutrient-reductions-4-
2022.pdf
U.S. EPA, 2021a. Life cycle and cost assessments of nutrient removal technologies in
wastewater treatment plants. (No. EPA 832-R-21-006). U.S. Environmental Protection
Agency.
U.S. EPA, 2021b. eGRID Summary Tables 2019.
U.S. EPA, 2020a. Electricity Life Cycle Inventory.
U.S. EPA, 2020b. Documentation for greenhouse gas emission and energy factors used in the
waste reduction model (WARM): management practice chapters. U.S. Environmental
Protection Agency.
U.S. EPA, 2015. A compilation of cost data associated with the impacts and control of nutrient
pollution (EPA 820-F-l5-096). Office of Science and Technology, Washington, DC.
U.S. EPA, 2003. Wastewater technology fact sheet, screening and grit removal.
U.S. EPA, 2002. Biosolids technology fact sheet use of composting for biosolids management.
EPA 832-F-02-024.
US EPA, O., 2020. Emissions & Generation Resource Integrated Database (eGRID) [WWW
Document], URL https://www.epa.gov/egrid (accessed 8.18.21).
USGCRP, 2016. The impacts of climate change on human health in the United States: a
scientific assessment. United States Global Change Research Program.
Weissenbacher, N., Takacs, I., Murthy, S., Fuerhacker, M., Wett, B., 2010. Gaseous nitrogen and
carbon emissions from a full-scale deammonification plant. Water Environment Research
82, 169-175. https://doi.org/10.2175/106143009X447867
Wiser, J.R., P.E., Schettler, J.W., P.E., Willis, J.L., P.E., 2010. Evaluation of combined heat and
power technologies for wastewater treatment facilities. U.S. Environmental Protection
Agency.
Wong, J.W.C., Selvam, X.W.A., 2017. Improving compost quality by controlling nitrogen loss
during composting. In: Current Developments in Biotechnology andBioengineering:
Solid Waste Management. Elsevier Science, pp. 59-82. https://doi.org/10.1016/B978-0-
444-63664-5.00004-6
Woon, K.S., Lo, I.M.C., Chiu, S.L.H., Yan, D.Y.S., 2016. Environmental assessment of food
waste valorization in producing biogas for various types of energy use based on LCA
approach. Waste Management 50, 290-299.
https ://doi. org/10.1016/j. wasman.2016.02.022
World Bank, 2017. Population, Total [WWW Document], URL
https://data.worldbank.org/indicator/SP.POP.TOTL (accessed 3.16.18).
Xue, X., Cashman, S., Gaglione, A., Mosley, J., Weiss, L., Ma, X.C., Cashdollar, J., Garland, J.,
2019. Holistic analysis of urban water systems in the Greater Cincinnati region: (1) life
cycle assessment and cost implications. Water Research X2, 100015.
https://doi.Org/10.1016/j.wroa.2018.100015
Yoshida, H., Gable, J. J., Park, J.K., 2012. Evaluation of organic waste diversion alternatives for
greenhouse gas reduction. Resources, Conservation and Recycling 60, 1-9.
https://doi.Org/10.1016/j.resconrec.2011.ll.011
EP-C-I7-04I; WA 4^77
6-6
-------
Appendix A - Life Cycle Impact Assessment
APPENDIX A
LIFE CYCLE IMPACT ASSESSMENT
EP-C-I7-04I; WA 4^77
-------
Appendix A - Life Cycle Impact Assessment
Appendix A - Life Cycle Impact Assessment
LCIA is defined in ISO 14044 section 3.4 as the "phase of life cycle assessment aimed at
understanding and evaluating the magnitude and significance of the potential environmental
impacts for a product system throughout the life cycle of the product (ISO, 2006b)." Within
LCIA, the multitude of environmental LCI flows throughout the entire study boundaries (e.g.,
raw material extraction through chemical and energy production and through wastewater
treatment and effluent release) are classified according to whether they contribute to each of the
selected impact categories. Following classification, all the relevant pollutants are normalized to
a common reporting basis, using characterization factors that express the impact of each
substance relative to a reference substance. One well known example is the reporting of all GHG
emissions in C02-eq. The LCI and LCIA steps together compromise the main components of a
full LCA.
ISO 14040 recommends that an LCA be as comprehensive as possible so that "potential
trade-offs can be identified and assessed (ISO, 2006a)." Given this recommendation, this study
applies a wide selection of impact categories that encompass both environmental and human
health indicators. The selected LCIA categories address impacts at global, regional, and local
scales.
This study considers 12 impact categories in assessing the environmental burdens of the
nine wastewater treatment configurations. The majority of impact categories address air and
water environmental impacts, while three of the selected impact categories are human health
impact indicators. There are two main methods used to develop LCIA characterization factors:
midpoint and endpoint. The impact categories selected for this study are all midpoint indicators.
Midpoint indicators are directly associated with a specific environmental or human health
pathway. Specifically, midpoint indicators lie at the point along the impact pathway where the
various environmental flows that contribute to these issues can be expressed in a common unit
(e.g., C02-eq). Units such as CO2 equivalents express a relevant environmental unit, in this case
radiative forcing (W-yr/m2/kg), in the context of a reference substance. This is mentioned to
reinforce the fact that there are physical mechanisms underlying all of the impact assessment
methods put forward. Endpoint indicators build off of these midpoint units and translate them
into impacts more closely related to the final damage caused by the substance, which include: (1)
human health, (2) man-made environment, (3) natural environment, and (4) natural resources
(Udo de Haes et al., 1999). It is commonly believed that endpoint indicators are easier for many
audiences to understand, but suffer due to the fact that they significantly increase the level of
uncertainty associated with the results because the translation to final damage are typically less
understood and lack data. To reduce uncertainty of the results, this work generally focuses on
indicators at the midpoint level.
The LCIA method provided by the Tool for the Reduction and Assessment of Chemical
and Environmental Impacts (TRACI), version 2.1, developed by the U.S. EPA specifically to
model environmental and human health impacts in the U.S., is the primary LCIA method applied
in this study (Bare, 2012). Additionally, the ReCiPe LCIA method is recommended to
characterize fossil fuel depletion and water depletion (Goedkoop et al., 2009). Energy is tracked
based on point of extraction using the cumulative energy demand method developed by
ecoinvent (Ecoinvent Centre, 2010a).
EP-C-17-04I: WA 4-
A-8
-------
Appendix A - Life Cycle Impact Assessment
Summaries of each of the 12 impact categories evaluated as part of this study are
provided in the subsequent sections. Each summary includes a table of the main substances
considered in the impact category, associated substance characterization factor, and the
compartment (e.g., air, water, soil) the substance is released to or extracted from (in the case of
raw materials). These tables highlight key substances but should not be considered
comprehensive.
A.l Eutrophication Potential
Eutrophication occurs when excess nutrients (e.g., nitrogen or phosphorus) are introduced
to surface and coastal water causing the rapid growth of aquatic organisms. This growth
(generally referred to as an "algal bloom") reduces the amount of dissolved oxygen in the water,
thus decreasing oxygen available for other aquatic species. Eutrophication can lead to several
negative endpoint effects on human and ecosystem health. Oxygen depletion or changing
nutrient availability can affect species composition and ecosystem function. Additionally, the
proliferation of certain algal species can result in toxic releases that directly impact human health
(Henderson, 2015).
Table A-l provides a list of common substances that contribute to eutrophication, along
with their associated characterization factors. As indicated in the table, air emissions can also
contribute to eutrophication, through the atmospheric deposition of nitrogen compounds. The
TRACI 2.1 eutrophication method considers emissions to both fresh and coastal waters. TRACI
2.1 characterization factors for eutrophication are the product of a nutrient factor and a transport
factor (Bare et al., 2003). The nutrient factor is based on the amount of potential algae growth
caused by each pollutant. The relative eutrophying effect of a nitrogen or phosphorus species is
determined by its stoichiometric relationship to the Redfield ratio (Norris, 2002). The Redfield
ratio is the average C:N:P ratio of phytoplankton, and describes the necessary building blocks to
facilitate algal growth and reproduction (Redfield, 1934). The transport factor accounts for the
likelihood that the pollutant will reach a body of water based on the average hydrology
considerations for the U.S. The transport factor is used to account for the fact that not all the
nutrient released will reach aquatic systems and supply limiting nutrients. Both air and water
emissions have the potential to contribute to eutrophication; however, the fraction of air
emissions which make their way into bodies of water is often lower, which is reflected in a
smaller transport factor, and the correspondingly lower characterization factors of nitrogen oxide
air emissions in Table 4-1.
Both BOD and COD are also shown in Table A-l as contributing to eutrophication
impacts. Although the mechanism of oxygen consumption differs from that associated with
nutrient emissions of nitrogen and phosphorus, the result remains the same. Only COD (and not
BOD) values are characterized in this study to avoid double-counting (Norris, 2002).
In this study, U.S. average characterization factors are used, which are created as a
composite of all water basins in the U.S. For a discussion of the procedure used to produce
composite U.S. characterization factors, see Norris (2002). It must be recognized that context
specific features of an individual WWTP and the hydrology and ecology of the watershed in
which it is located could serve to ameliorate or increase site-specific impacts. In addition, water
EP-C-I7-04I; WA 4^77
A-9
-------
Appendix A - Life Cycle Impact Assessment
body-specific nutrient limitations and local transport characteristics tend to be the most decisive
factors in determining regional differences in eutrophication impacts (Henderson, 2015).
Table A-l. Main Pollutants Contributing to Eutrophication Potential Impacts
(kg N eq/ kg Pollutant).
Pollutant
Chemical Formula
Compartment
Characterization Factor
Biological oxygen demand (BOD5)
N/A
Water
0.05
Chemical oxygen demand (COD)
N/A
Water
0.05
Ammonia
nh3
Water
0.78
Nitrate
no3-
Water
0.24
Nitrogen dioxide
no2
Air
0.04
Nitrogen monoxide
NO
Air
0.04
Nitrogen oxides
NOx
Air
0.04
Nitrogen organic bound
N/A
Water
0.99
Phosphate
PO,3
Water
2.4
Phosphorus a
P
Water
7.3
Selected Method-
TRACI 2.1
a Represents phosphorus content of unspecified phosphorus pollutants (e.g., "total phosphorus" in effluent
composition).
A.2 Cumulative Energy Demand
The cumulative energy requirements for a system can be categorized by the fuels from
which energy is derived. This method is not an impact assessment, but rather is a cumulative
inventory of all energy extracted and utilized. Energy sources consist of non-renewable fuels
(natural gas, petroleum, nuclear and coal) and renewable fuels. Renewable fuels include
hydroelectric energy, wind energy, energy from biomass, and other non-fossil sources.
Cumulative energy demand (CED) includes both renewable and non-renewable sources as well
as the embodied energy in biomass and petroleum feedstocks. CED is measured in MJ/kg.
Energy is tracked based on the higher heating value (HHV) of the fuel at the point of extraction.
Table A-2 includes a few examples of fuels that contribute to CED in this project and their
associated characterization factors.
Table A-2. Main Energy Resources Contributing to Cumulative Energy Demand.
Energy Resource
Compartment
Units
Characterization
Factor
Energy, gross calorific value, in biomass
Resource (biotic)
MJ/MJ
1.0
Coal, hard, unspecified, in ground
Resource (in ground)
MJ/kg
19
Gas, natural, in ground
Resource (in ground)
MJ/kg
47
Oil, crude, in ground
Resource (in ground)
MJ/kg
46
Selected Method-
Ecoinvent
EP-C-17-041; WA 4-77
A-10
-------
Appendix A - Life Cycle Impact Assessment
A.3 Global Warming Potential
Global warming refers to an increase in the earth's temperature in relation to long-
running averages. In accordance with IPCC recommendations, TRACTs GWP calculations are
based on a 100-year time frame and represent the heat-trapping capacity of the gases relative to
an equal weight of carbon dioxide. Relative heat-trapping capacity is a function of a molecule's
radiative forcing value as well as its atmospheric lifetime. Table A-3 provides a list of the most
common GHGs along with their corresponding GWPs, or CO2 equivalency factors, used in
TRACI 2.1. Contributing elementary flows can be characterized using GWPs reported by the
IPCC in either 2007 (Fourth Assessment Report) or in 2013 (Fifth Assessment Report) (IPCC,
2007; IPCC, 2013). While the 2013 GWPs are the most up-to-date, the 2007 GWPs have been
officially adopted by the United Nations Framework Convention on Climate Change (UNFCCC)
for international greenhouse gas reporting standards and are used by EPA in their annual
greenhouse gas emissions report. The baseline results in this study apply the 2007 GWPs, but
results with the 2013 GWPs are provided in a sensitivity analysis in Section 4.2.
Table A-3. Main Greenhouse Gas (GHG) Emissions Contributing
to Global Warming Potential Impacts
(kg CO2 eq/kg GHG).
C.IIC.
( Ik-iii ic;il
l-'oi'llllllil
( (impiirlmoiil
c;\\i> (IPC ( 2oo7)
C;\\ P (IPCC 2013)
Carbon dioxide
CO2
Air
1.0
l.U
Nitrous oxide
N20
Air
3.OE+2
2.7E+2
Methane
ch4
Air
25
28
Sulfur
hexafluoride
sf6
Air
2.3E+4
2.4E+4
Selected Melliod—
IPCC 2UU~ or Jul3 lUUa
A.4 Acidification Potential
The deposition of acidifying substances such as those listed in Table A-4 have an effect
on the pH of the terrestrial ecosystem. Each species within these ecosystems has a range of pH
tolerance, and the acidification of the environment can lead to shifting species composition over
time. Acidification can also cause damage to buildings and other human infrastructure (Bare,
2012). The variable buffering capacity of terrestrial environments yields a correspondingly
varied response per equivalent unit of acidification. Due to a lack of data, the variable sensitivity
of receiving regions is not captured in TRACI characterization factors (Norris, 2002). The
acidification method in TRACI utilizes the results of an atmospheric chemistry and transport
model, developed by the US National Acid Precipitation Assessment Program (NAPAP), to
estimate total North American terrestrial deposition of expected SO2 equivalents due to
atmospheric emissions of NOx and SO2 and other acidic substances such as HC1 and HF, as a
function of the emissions location (Bare et al., 2003). Emissions location is modeled in this study
as average U.S. using TRACI's composite annual North American emissions average of U.S.
states.
EP-C-I7-04I; WA 4^77
A-ll
-------
Appendix A - Life Cycle Impact Assessment
Table A-4. Main Pollutants Contributing to Acidification Potential Impacts
(kg SO2 eq/kg Pollutant).
Pollutant
Chemical Formula
Compartment
Characterization
Factor
Sulfur dioxide
S02
Air
1.0
Ammonia
nh3
Air
1.9
Nitrogen dioxide
NO2
Air
0.70
Nitrogen oxides
NOx
Air
0.70
Hydrogen chloride
HC1
Air
0.88
Hydrogen fluoride
HF
Air
1.6
Hydrogen sulfide
H2S
Air
1.9
Selected Method-
TRACI 2.1
A.5 Fossil Depletion
Fossil depletion is a measure of the study systems demand for non-renewable energy
resources. As non-renewable resources, the availability of fossil energy will not change (i.e., new
fossil energy will not be produced) on relevant human timescales. When these resources are
depleted and resource quality declines, the cost and environmental impact of accessing a given
quantity of energy increases. Fossil depletion is measured in kg oil equivalent based on each
fuel's heating value. Renewable energy systems and uranium are not included in the fossil
depletion metric, but are assessed within the CED methodology previously discussed. Table A-5
presents common fossil fuel flows and their associated characterization factors for this impact
category.
Table A-5. Main Fossil Fuel Resource Contributing to Fossil Depletion (kg oil eq/kg Fossil
Fuel Resource).
Fossil Fuel Resource
Compartment
Characterization Factor
Oil, crude, 42 MJ per kg
Resource (in ground)
1.0
Coal, 18 MJ per kg
Resource (in ground)
0.43
Coal, 29.3 MJ per kg
Resource (in ground)
0.70
Gas, natural, 30.3 MJ per kg
Resource (in ground)
0.72
Gas, natural, 35 MJ per m3
Resource (in ground)
0.83
Methane
Resource (in ground)
0.86
Selected Method—
ReCiPe
A.6 Smog Formation Potential
The smog formation impact category characterizes the potential of airborne emissions to
cause photochemical smog. The creation of photochemical smog occurs when sunlight reacts
with NOx and volatile organic compounds (VOCs), resulting in tropospheric (ground-level)
ozone (O3) and particulate matter. Potential endpoints of such smog creation include increased
EP-C-17-041; WA 4-77
A-12
-------
Appendix A - Life Cycle Impact Assessment
human mortality, asthma, and deleterious effects on plant growth. Smog formation potential
impacts are measured in kg of O3 equivalents. Table A-6 includes a list of smog forming
chemicals expected to be associated with this project along with their characterization factors.
Table A-6. Main Pollutants Contributing to Smog Formation Impacts (kg O3 eq/kg
Pollutant).
Polliiliinl
( homiciil
I'ormuhi
( ompiirimonl
( hiii'iKicri/iilidii l-'iicloi'
Sulfur monoxide
SO
Air
1.0
Carbon monoxide
CO
Air
0.06
Methane
ch4
Air
0.01
Nitrogen dioxide
no2
Air
17
Nitrogen oxides
NOx
Air
25
Volatile organic compounds
(VOCs)
N/A
Air
3.6
Selected Method—
TRACI 2.1
A.7 Human Health—Particulate Matter Formation Potential
Particulate matter (PM) emissions have the potential to negatively impact human health.
Respiratory complications are particularly common among children, the elderly, and individuals
with asthma (U.S. EPA, 2008a). Respiratory impacts can result from a number of types of
emissions including PM10, PM2.5, and precursors to secondary particulates such as sulfur
dioxide and nitrogen oxides. Respiratory impacts are a function of the fate of responsible
pollutants as well as the exposure of human populations. Table A-7 provides a list of common
pollutants contributing to impacts in this category along with their associated characterization
factors. Impacts are measured in relation to PM2.5 emissions.
Table A-7. Main Pollutants Contributing to Human Health—Particulate Matter Formation
Potential
(kg PM2.5 eq/kg Pollutant)
Polliiliinl
( homiciil
I'ormuhi
( onipiirlim-nl
('hiii'iiclcri/iilion l-'iicloi-
Particulates, <2.5 um
N/A
Air
1.0
Particulates, >2.5 um, and <
lOum
N/A
Air
0.23
Ammonia
nh3
Air
0.07
Nitrogen oxides
NOx
Air
7.2E-3
Sulfur oxides
sox
Air
0.06
Selected Method—
TRACI 2.1
EP-C-I7-04I; WA 4^77
A-13
-------
Appendix A - Life Cycle Impact Assessment
A.8 Water Depletion
Water depletion results are displayed on a consumptive basis (i.e., depletion). When
water is withdrawn from one water source and returned to another watershed this is considered
consumption, as there is a net removal of water from the original water source. For instance, it is
assumed that deepwell injection of the brine fluid from RO is consumptive water depletion, since
water is being diverted from a watershed making it unavailable for subsequent environmental or
human uses. Consumption also includes water that is withdrawn and evaporated or incorporated
into the product. Cooling water that is closed-loop circulated, and does not evaporate, is not
considered consumptive use. Water depletion is only included as an inventory category in this
study, which is a simple summation of water inputs. The analysis does not attempt to assess
water-related damage factors. For instance, there is no differentiation between water depletion
that occurs in water-scarce or water-abundant regions of the world. Water depletion in this study
includes values for upstream fuel and electricity processes. In addition to water depletion
associated with thermal generation of electricity from fossil and nuclear fuels, the water
depletion for power generation includes evaporative losses due to establishment of dams for
hydropower. Table A-8 shows some of the common flows associated with water depletion along
with their characterization factors. Section 3.3.1 also discusses some of the uncertainty
associated with calculating water depletion in LCA.
Table A-8. Main Water Flows Contributing to Water Depletion.
Water Flow
Compartment
Units
Characterization Factor
Water, lake
Resource (in water)
m3 H20/m3
1.0
Water, river
Resource (in water)
m3 H20/m3
1.0
Water, unspecified natural origin
Resource (in water)
m3 H20/m3
1.0
Water, well, in ground
Resource (in water)
m3 H20/m3
1.0
Water, unspecified natural origin/kg
Resource (in water)
m3 H20/kg
1.0E-3
Selected Method-
ReCiPe
A.9 Water Scarcity
The AWARE method is used to assess water scarcity impact. The water scarcity indicator
seeks to answer the question, "what is the potential to deprive another freshwater user (human or
ecosystem) by consuming freshwater in this region?" (Boulay et al., 2018). AWARE water
scarcity factors, depicted in Table A-9, are applied on top of the water depletion inventory values
that result from application of the method described in Section A.8. Water scarcity factors are
1
developed based on the inverse of Availability Minus Demand AMD subtracts human and
ecosystem water requirements from the total availability of water in a region and divides the
resulting quantity by the area of that region. Characterization factors are developed by dividing
the regional AMD inverse by the corresponding world average value, resulting in a
dimensionless value termed m3 world equivalents. When the demand for water exceeds
availability in a given region the value of the characterization factor is set at a maximum value of
100, as is the case for Santa Fe. Physical interpretation of this maximal value for the Santa Fe
EP-C-17-041; WA 4-77
A-14
-------
Appendix A - Life Cycle Impact Assessment
region means that the Santa Fe region has 100 "times less water remaining per area within a
certain period of time as the world average" (Boulay et al., 2018).
Table A-9. Main Water Flows Contributing to Water Depletion.
Region of Withdrawal
Compartment
Units
Characterization
Factor
Santa Fe, New Mexico
Resource (in water)
m3 world equivalents/in3
100
AZNM Electrical Grid, Generation
Resource (in water)
m3 world equivalents/in3
80.3
WECC Electrical Grid, Generation
Resource (in water)
m3 world equivalents/in3
42.2
AZNM Electrical Grid, Consumption
Resource (in water)
m3 world equivalents/in3
52.0
U.S., National Average
Resource (in water)
m3 world equivalents/in3
17.3
Selected Method-
AWARE
A. 10 Human Health—Cancer Potential
Carcinogenic human health results in this study are expressed on the basis of
Comparative Toxic Units (CTUh) based on the USEtox™ method (Huijbregts et al. 2010).
Characterization factors within the USEtox™ model are based on fate, exposure, and effect
factors. Each chemical included in the method travels multiple pathways through the
environment based on its physical and chemical characteristics. The potential for human
exposure (e.g., ingestion or inhalation) varies according to these pathways. The effect factor
characterizes the probable increase in cancer-related morbidity for the total human population
per unit mass of a chemical emitted (i.e., cases per kg) (Rosenbaum et al., 2008). The full
USEtox™ model contains over 3,000 chemicals of global relevance, and is the product of an
international project to harmonize the approach to evaluation of toxicity effects. The USEtox™
model develops characterization factors at the continental and global scale. The exclusion of
more localized parameters is justified in that it was found during the harmonization process that
site-specific parameters have a far lower impact on results than do the substances themselves.
Global midpoint characterization factors are employed from the most recent version of
USEtox™ available in OpenLCA, version 2.02. An updated version of USEtox™, version 2.11,
was released in April 2019. Characterization factors for the heavy metals, toxic organics and
DBPs were updated in the OpenLCA USEtox™ LCIA method to match version 2.11. All other
characterization factors remain at the default value for OpenLCA's USEtox version 2
(recommended+interim) database. Not all heavy metals, toxic organics and DBPs have
established characterization factors in the USEtox™ method. Several additional sources were
used to identify appropriate characterization factors. When no appropriate characterization factor
was identified, the pollutant was assigned a characterization factor equal to the median
characterization factor for its trace pollutant group. For illustration purposes, Table A-10 lists
five of the primary chemicals that contribute to cancer, human health impacts in the US and
Canada (Ryberg, 2013) along with their associated characterization factors.
EP-C-17-041; WA 4-77
A-15
-------
Appendix A - Life Cycle Impact Assessment
The developers of the USEtox™ method are clear to point out that some of the
characterization factors associated with human health effects should be considered interim,
owing to uncertainty in their precise values ranging across one to three orders of magnitude.
Sources of uncertainty are often attributable to the use of one exposure route as a proxy for
another (route-to-route extrapolation). For a more detailed discussion of uncertainty present in
these models, see the USEtox™ User's Manual (Huijbregts et al., 2010). Appropriate
interpretation of results must consider the uncertainty associated with the use of interim
characterization factors.
Table A-10. Main Pollutants Contributing to Human Health—Cancer Potential Impacts
(CTUh/kg Pollutant).
Pollutant
Chemical Formula
Compartment
Characterization Factor
Arsenic
As
Soil
1.8E-43
Formaldehyde
ch2o
Air
2.5E-5
Chromium VI
Cr
Soil
5.0E-33
Chromium VI
Cr
Air, urban
3.8E-33
Chromium VI
Cr
Water
0.01a
Selected Method-
USEtox™ 2.11
a - Designates an interim characterization factor.
A. 11 Human Health—Noncancer Potential
Non-carcinogenic human health results in this study are expressed on the basis of
Comparative Toxic Units (CTUh) based on the USEtox™ method, which is incorporated in
TRACI 2.1. The impact method characterizes the probable increase in noncancer related
morbidity for the total human population per unit mass of a chemical emitted (i.e., cases per kg)
(Rosenbaum et al., 2008). These impacts are calculated using the same approach as that taken for
human health - cancer (Section A. 10).
Global midpoint characterization factors are employed from the most recent version of
USEtox™ available in OpenLCA, version 2.02. An updated version of USEtox™, version 2.11,
was released in April 2019. Characterization factors for the heavy metals, toxic organics and
DBPs were updated in the OpenLCA USEtox™ LCIA method to match version 2.11. All other
characterization factors remain at the default value for OpenLCA's USEtox version 2
(recommended+interim) database. Not all heavy metals, toxic organics and DBPs have
established characterization factors in the USEtox™ method. Several additional sources were
used to identify appropriate characterization factors. When no appropriate characterization factor
was identified, the pollutant was assigned a characterization factor equal to the median
characterization factor for its trace pollutant group. For illustration purposes, Table A-l 1 lists the
main chemicals contributing to noncancer, human health impacts (Ryberg, 2013) along with their
associated characterization factors.
As is discussed in Section A. 10, uncertainty in USEtox factors can range across one to
three orders of magnitude for interim characterization factors, which are identified in Table
EP-C-17-041; WA 4-77
A-16
-------
Appendix A - Life Cycle Impact Assessment
A-l 1. At the current time, all characterization factors for metal compounds are considered
interim. Appropriate interpretation of results must consider the uncertainty associated with the
use of interim characterization factors.
Table A-ll. Main Pollutants Contributing to Human Health—Noncancer Potential
Impacts (CTUh/kg Pollutant).
Pollutant
Chemical Formula
Compartment
Characterization Factor
Acrolein
C3H4O
Soil
3.4E-5
Zinc, ion
IN
Soil
1.4E-43
Arsenic, ion
As3+
Soil
0.01a
Zinc, ion
IN
Air, urban
5.7E-33
Mercuiy (+11)
Hg(H)
Air, urban
1.24a
Selected Method-
USEtox™ 2.11
a - Designates an interim characterization factor.
A. 12 Ecotoxicity Potential
Ecotoxicity is a measure of the effect of toxic substances on ecosystems. The effects on
freshwater ecosystems are used as a proxy for general ecological impact. Characterization factors
within the ecotoxicity model are based on fate, exposure, and effect factors. Each chemical
included in the method travels multiple pathways through the environment. As a result of these
pathways, various compartments (e.g., freshwater, terrestrial) and the species they contain will
have differing opportunities to interact with the chemical in question (exposure). The effect
factor refers to the potential negative consequences on ecosystem health when exposure does
occur (Huijbregts, 2010). The exclusion of more localized parameters is justified in that it was
found during the harmonization process that these parameters have a far lower impact on results
than do the substances themselves. Ecotoxicity impacts are measured in terms of the Potentially
Affected Fraction of species due to a change in concentration of toxic chemicals (PAF
m3.day/kg). These units are also known as comparative toxicity units (CTUe).
Global midpoint characterization factors are employed from the most recent version of
USEtox™ available in OpenLCA, version 2.02. An updated version of USEtox™, version 2.11,
was released in April 2019. Characterization factors for the heavy metals, toxic organics and
DBPs were updated in the OpenLCA USEtox™ LCIA method to match version 2.11. All other
characterization factors remain at the default value for OpenLCA's USEtox version 2
(recommended+interim) database. Not all heavy metals, toxic organics and DBPs have
established characterization factors in the USEtox™ method. Several additional sources were
used to identify appropriate characterization factors. When no appropriate characterization factor
was identified, the pollutant was assigned a characterization factor equal to the median
characterization factor for its trace pollutant group. For illustration purposes, Table A-12 lists
some of the main chemicals found to contribute to ecotoxicity impacts (Ryberg, 2013) and their
USEtox™ global characterization factors.
As is discussed in Section A. 10, uncertainty in USEtox factors can range across one to
three orders of magnitude for interim characterization factors, which are identified in Table
EP-C-17-041; WA 4-77
A-17
-------
Appendix A - Life Cycle Impact Assessment
A-12. At the current time, all characterization factors for metal compounds are considered
interim. Appropriate interpretation of results must consider the uncertainty associated with the
use of interim characterization factors.
Table A-12. Main Pollutants Contributing to Ecotoxicity Potential Impacts
(CTUe [PAF m3.day/kg Pollutant]).
Pollutant
Chemical
Formula
Compartment
Characterization Factor
Zinc, ion
Zn2+
Ground water
1.3E+53
Chromium VI
Cr(VI)
Ground water
1.0E+53
Nickel, ion
Ni2+
Ground water
3.0E+53
Chromium VI
Cr(VI)
River
1.0E+53
Arsenic, ion
As3+
Ground water
1.5E+43
Selected Method-
USEtox™ witliin TRACI 2.11
a - Designates an interim characterization factor.
EP-C-17-041; WA 4-77
A-18
-------
Appendix B - Life Cycle Inventory Data
APPENDIX B
LIFE CYCLE INVENTORY DATA
EP-C-I7-04I; WA 4^77
-------
Appendix B - Life Cycle Inventory Data
Appendix B - Life Cycle Inventory Data
B.l Life Cycle Inventory Data Tables
Table B-l. Life cycle inventory data for unit processes that are consistent across scenarios
Process
Name
Input Name
Mean
Value
Min
Max
Units2
Distribution
Type
Geometric
Standard
Deviation
Uncertainty Range Note
Core
Facility
Electricity,
grid
0.73
NA
NA
kWh/m3
None
NA
Mean value: facility data
Diesel,
combusted
6.4E-3
NA
NA
liters/m3
None
NA
Mean value: facility data
Lime
3.5E-3
1.8E-3
5.3E-3
kg/m3
Triangular
NA
Min, max and mean are based on 25%, 50%
and 75% of 2020 lime consumption for
which demand is expected to reduce
following plant upgrades.
Preliminary
Treatment -
Screening
and Grit
Removal
Residuals to
landfill
0.04
9.8E-3
0.11
kg/m3
Triangular
NA
Mean value is based on facility data.
Min and max values are based on survey
data from eight U.S. WWTPs (U.S. EPA,
2003)
Secondary
Treatment -
Biological
Methane, to
air
8.0E-3
NA
NA
kg CHi/m3
Lognonnal1
1.69
See Appendix Section B.2 for details on
process GHG emission estimation.
Nitrous oxide,
to air
4.4E-4
NA
NA
kg N2O/1113
Lognonnal1
1.69
Tertiary
Treatment -
Disk
Filtration
Filter pads,
polyester
5.6E-4
3.1E-4
8.2E-4
kg/m3
Triangular
NA
Facility replaced 560 filter pads in 2020.
Mean, min and max values based on total
estimated mass of 3788, 2104 and 5471 kg.
Filter nozzles,
steel
1.5E-6
1.1E-6
1.9E-6
kg/m3
Triangular
NA
Facility replaced 200 filter nozzles in 2020.
Mean, min and max values based on total
estimated mass of 10, 7.5 and 12.5 kg.
Citric Acid
7.4E-5
6.6E-5
8.1E-5
kg/m3
Triangular
NA
No data provided. Use microfilter quantities
as proxy.
Sodium
hypochlorite
5.1E-5
4.6E-5
5.6E-5
kg/m3
Triangular
NA
Mean value is based on facility data
assuming a sodium hypochlorite density of
1209 kg/m3. Min and max values estimated
assuming +/-10% of reported value.
Disinfection
- Ultraviolet
Phosphoric
Acid
3.4E-4
3.0E-4
3.7E-4
kg/m3
Triangular
NA
Facility used 385 gallons in 2020. Assumed
density of 1834 kg/m3. Min and max values
EP-C-17-041; WA 4-77
B-l
-------
Appendix B - Life Cycle Inventory Data
Process
Name
Input Name
Mean
Value
Min
Max
Units2
Distribution
Type
Geometric
Standard
Deviation
Uncertainty Range Note
estimated assuming +/-10% of reported
value.
Sludge -
Dissolved
Air
Flotation
Polymer
1.2E-3
9.5E-4
1.4E-3
kg/m3
Triangular
NA
Facility used 10,230 gallons of polymer in
2020. Assumed 50% used in DAF and 50%
used in belt filter press. Mass of
polyacrylamide estimated assuming a
polymer density of 8.6 pounds per gallon
with an active polymer concentration of
40% w/w. Min and max values estimated as
a function of +/-10% baseline consumption,
density range of 8.5-8.7 lbs/gallon and
active polvmer concentration range of 36-
43%.
Sludge -
Belt Filter
Press
Polymer
1.2E-3
9.5E-4
1.4E-3
kg/m3
Triangular
NA
Sludge -
Anaerobic
Digestion
Methane, to
air
2.8E-3
1.5E-3
7.9E-3
kg CHi/m3
Triangular
NA
Mean, min and max annual biogas
production estimated based on mean, 25th
and 75th percentile values of daily
production for 2020. Mean, min and max
estimates of fugitive methane leakage based
on biogas methane content, leakage rate and
biogas production.
Daily production: 230,114 (223,835-
238,875) standard cubic feet
Methane content: 59%v/v (55%-64%)
Leakage rate: 2% (1.2%-5%)
Biogas
(output)
0.35
NA
NA
m3/m3
None
NA
Allocation of biogas to combustion
processes is described in Table 2-3. The
effect of changes in biogas output on model
results is assessed in the sensitivity
assessment.
Electricity
(output)
0.63
0.61
0.65
kWli/m3
Triangular
NA
Allocation of biogas to combustion
processes is described in Table 2-3. Heat
content of produced biogas is estimated
assuming an LHV of 597 BTU/scf. Mean
electrical and thermal efficiencies of the
CHP system are 35% and 45%,
respectively. In the best guess model run
Heat (output)
0.07
0.06
0.07
m3/m3
Triangular
NA
EP-C-17-041; WA 4-77
B-2
-------
Appendix B - Life Cycle Inventory Data
Process
Name
Input Name
Mean
Value
Min
Max
Units2
Distribution
Type
Geometric
Standard
Deviation
Uncertainty Range Note
(all scenarios), 54% of CHP heat production
is utilized onsite, avoiding natural gas
consumption. Min/max values for avoided
energy consumption are estimated as a
function of the share of biogas combusted in
the CHP and boiler, CHP electrical
efficiency (30-40%), CHP total efficiency
(75-85%), boiler thermal efficiency (80-
97%) and biogas LHV (556-649 BTU/scf).
Biogas -
Flaring
Nitrogen
oxides, to air
2.1E-4
1.9E-4
2.3E-4
kg/m3 biogas
combusted
Triangular
NA
Emission factors for these air pollutants was
drawn from Paseo Real's air permit
application. Min and max values estimated
assuming +/-10% of reported value.
Carbon
monoxide, to
air
7.9E-4
7.1E-4
8.7E-4
kg/m3 biogas
combusted
Triangular
NA
VOCs, to air
5.2E-5
4.6E-5
5.7E-5
kg/m3 biogas
combusted
Triangular
NA
Sulfur
dioxide, to air
1.3E-3
1.1E-3
1.4E-3
kg/m3 biogas
combusted
Triangular
NA
Mean values were pulled from Morelli et al.
£2019). Min and max values estimated
assuming +/-10% of reported value.
Particulate
matter, to air
5.8E-4
5.2E-4
6.4E-4
kg/m3 biogas
combusted
Triangular
NA
Methane, to
air
3.9E-3
3.5E-3
4.3E-3
kg/m3 biogas
combusted
Triangular
NA
Biogas -
CHP
Nitrogen
oxides, to air
4.6E-3
3.0E-3
6.1E-3
kg/m3 biogas
combusted
Triangular
NA
Emission factors for these air pollutants was
drawn from Paseo Real's air permit
application. Min and max values estimated
assuming +/-10% of reported value.
Carbon
monoxide, to
air
1.1E-2
7.6E-3
1.5E-2
kg/m3 biogas
combusted
Triangular
NA
VOCs, to air
2.6E-3
2.1E-3
3.0E-3
kg/m3 biogas
combusted
Triangular
NA
Sulfur
dioxide, to air
1.4E-5
1.3E-5
1.5E-5
kg/m3 biogas
combusted
Triangular
NA
Mean values were pulled from Morelli et al.
£2019). Min and max values estimated
assuming +/-10% of reported value.
Particulate
matter, to air
3.4E-5
3.1E-5
3.8E-5
kg/m3 biogas
combusted
Triangular
NA
Ammonia, to
air
6.4E-5
5.8E-5
7.0E-5
kg/m3 biogas
combusted
Triangular
NA
EP-C-17-041: WA 4-77
B-3
-------
Appendix B - Life Cycle Inventory Data
Process
Name
Input Name
Mean
Value
Min
Max
Units2
Distribution
Type
Geometric
Standard
Deviation
Uncertainty Range Note
Methane, to
air
4.3E-3
3.9E-3
4.7E-3
kg/m3 biogas
combusted
Triangular
NA
Nitrous oxide,
to air
1.0E-4
9.2E-5
1.1E-4
kg/m3 biogas
combusted
Triangular
NA
Nitrogen
oxides, to air
2.1E-4
1.9E-4
2.3E-4
kg/m3 biogas
combusted
Triangular
NA
Emission factors for these air pollutants was
drawn from Paseo Real's air permit
application. Min and max values estimated
assuming +/-10% of reported value.
Carbon
monoxide, to
air
7.9E-4
7.1E-4
8.7E-4
kg/m3 biogas
combusted
Triangular
NA
Biogas -
Boilers
VOCs, to air
5.2E-5
4.6E-5
5.7E-5
kg/m3 biogas
combusted
Triangular
NA
Sulfur
dioxide, to air
5.2E-4
4.6E-4
5.7E-4
kg/m3 biogas
combusted
Triangular
NA
Mean values were pulled from Morelli et al.
£2019). Min and max values estimated
Particulate
matter, to air
1.2E-4
1.1E-4
1.3E-4
kg/m3 biogas
combusted
Triangular
NA
assuming +/-10% of reported value.
Methane, to
air
4.1E-5
3.7E-5
4.5E-5
kg/m3 biogas
combusted
Triangular
NA
Nitrous oxide,
to air
1.0E-5
9.2E-6
1.1E-5
kg/m3 biogas
combusted
Triangular
NA
Electricity
0.05
NA
NA
kWh/digestate
None
NA
The facility reports use of 239,485 kWh in
2020 at their compost facility. Value was
scaled up to 399,142 kWh to reflect future
increase in volume of digestate composted.
Sludge -
Composting
Natural Gas
9.8E-4
NA
NA
m3/digestate
None
NA
The facility reports using 277 Dekathenns
of natural gas in 2020. Assume this value
reflects building heat and will remain
constant when increasing quantity of
digestate composted.
Methane, to
air
6.0E-4
NA
NA
kg CHi/digestate
Lognonnal1
1.69
Mean estimate is based on a methane
emission factor of 8.2E-3 kg CH4-C/kg C in
compost feedstock. See supporting Excel
workbook for sources.
Carbon
monoxide, to
air
5.1E-5
NA
NA
kg CO/digestate
Lognonnal1
1.69
Mean estimate is based on a CO emission
factor of 4E-4 kg CO-C/kg C in compost
feedstock (Hellebrand, 1998)
EP-C-17-041; WA 4-77
B-4
-------
Appendix B - Life Cycle Inventory Data
Process
N;inu'
Inpul Niimo
Mcsin
Value
Min
M;i\
I iiils-
Dislrihulion
1 J |)C
Ccomclric
Siiindiird
l)c\ iiilion
I ncer(;iin(\ Riiniic Nolo
Nitrous oxide,
to air
1.6E-4
NA
NA
kg N20/digestate
Lognormal1
1.69
Mean cslinialc is based on a i\2(J emission
factor of 0.0129 kg N20-N/kg N in compost
feedstock. See supporting Excel workbook
for sources.
Ammonia, to
air
4.1E-4
NA
NA
kg NH3/digestate
Lognormal1
1.69
Mean estimate is based on a NH3 emission
factor of 0.044 kg NH3-N/kg N in compost
feedstock (Hellebrand 1998)
NMVOC, to
air
1.0E-4
NA
NA
kg
NMVOC/digestate
Lognormal1
1.69
Mean estimate is based on a NMVOC
emission factor of 1.04E-4 kg NMVOC/kg
compost feedstock. (Maulini-Duran et al.,
2013)
Sludge -
Land
Application
Ammonia, to
air
7.5E-5
NA
NA
kg NH3/compost
Lognormal1
1.69
Mean estimate is based on NH3 emission
factor of 0.016 kg NH3-N/kg NH3-N in
compost (Boldrin et al., 2011).
Carbon
sequestration
0.10
0.08
0.13
kg CCVcompost
Triangular
NA
Mean estimate is based on the assumption
that 12% of land-applied carbon is
sequestered beyond 100 years (Boldrin et
al., 2009; Yoshida et al., 2012)
Nitrate,
groundwater
0.03
NA
NA
kg NCb/compost
Lognormal1
1.69
Mean estimate is based on a NO3 emission
factor (groundwater) of 0.2 kg NC>3-N/kg N
in compost (Boldrin et al., 2011).
Nitrate,
surface water
0.03
NA
NA
kg NCb/compost
Lognormal1
1.69
Mean estimate is based on a NO3 emission
factor (surface water) of 0.2 kg N03-N/kg N
in compost (Boldrin et al., 2011).
Nitrous oxide
7.0E-4
NA
NA
kg N20/compost
Lognormal1
1.69
Mean estimate is based on a N20 emission
factor of 0.015 kg N20-N/kg N in compost
(Boldrin etal., 2011).
Phosphorus,
groundwater
6.6E-5
NA
NA
kg P/compost
Lognormal1
1.69
Mean estimate is based on a P
(groundwater) emission factor of 0.005 kg
P/kg P in compost, which was calculated
based on method in (Nemecek and Kagi,
2007) using standard application rates.
Phosphorus,
surface water
1.8E-3
NA
NA
kg P/compost
Lognormal1
1.69
Mean estimate is based on a P (surface
water) emission factor of 0.133 kg P/kg P in
compost, which was calculated based on
HP
-041;
B-5
-------
Appendix B - Life Cycle Inventory Data
Process
Name
Mean
Value
Distribution
Type
Geometric
Input Name
Min
Max
Units2
Standard
Deviation
Uncertainty Range Note
method in (Nemecek and Kagi, 2007) using
standard application rates.
Fertilizer,
Nitrogen
0.02
0.01
0.03
kg Urea/compost
Triangular
NA
The mean quantity of avoided urea was
estimated based on the nitrogen content of
compost assuming that 30% of nitrogen is
displaces production of chemical fertilizer.
Min and max values are estimated using
20% and 40% substitution rates,
respectively.
Fertilizer,
Phosphorus
0.11
0.07
0.14
kg SSP/compost
Triangular
NA
The mean quantity of avoided single super
phosphate was estimated based on the
phosphorus content of compost assuming
that 73% of phosphorus displaces
production of chemical fertilizer. Min and
max values are estimated using 46% and
100% substitution rates, respectively.
Fertilizer,
Potassium
0.01
7.6E-3
0.01
kg KoSO i/compost
Triangular
NA
The mean quantity of avoided K2SO4 was
estimated based on the potassium content of
compost assuming that 60% of potassium
displaces production of chemical fertilizer.
Min and max values are estimated using
60% and 100% substitution rates,
respectively.
Diesel,
combusted
3.4E-3
NA
NA
Liters/m3
None
NA
NA
Sludge -
Landfilling
LFG Flaring
0.02
0.01
0.03
m3 biogas/m3
Triangular
NA
Base value is based on first order decay of
digestate in the landfill according to
parameters in Table 2-8. Min value is based
on DOCf of 0.5. Max value is based on
DOCf of 0.8 and a decay factor of 0.2.
Carbon
sequestered
0.08
0.04
0.09
kg CO2/1113
Triangular
NA
Base value is based on first order decay of
digestate in the landfill according to
parameters in Table 2-8. Min value is based
on DOCf of 0.8 and a decay factor of 0.2.
Max value is based on DOCf of 0.5.
EP-C-17-041; WA 4-77
B-6
-------
Appendix B - Life Cycle Inventory Data
Process
N;inu'
Inpul Niimo
Mcsin
Value
Min
M;i\
I iiils-
Disl rihu 1 icin
Tj |>c
Ccomclric
Siiindiird
l)c\ iiilion
I ncer(;iin(\ Riiniic Nolo
Methane
4.2E-3
3.2E-3
6.0E-3
kg CHi/m3
Triangular
NA
Base value is based on first order decay of
digestate in the landfill according to
parameters in Table 2-8. Min value is based
on DOCf of 0.5. Max value is based on
DOCf of 0.8 and a decay factor of 0.2.
Nitrous oxide
1.5E-4
NA
NA
kg N20/m3
Lognormal1
1.69
Mean estimate is based on a N20 emission
factor of 0.016 kg N20-N/kg N in landfilled
digestate (Borjesson and Svensson, 1997).
Electricity
0.56
NA
NA
kWh/m3 leachate
None
NA
LCI data extracted from (Righi et al., 2013).
Oxygen
0.03
NA
NA
kg/m3 leachate
None
NA
Uncertainty not assessed.
Alum
0.02
NA
NA
kg/m3 leachate
None
NA
Landfill
Leachate
Treatment
Sodium
hydroxide
2.0E-3
NA
NA
kg/m3 leachate
None
NA
Chloride, to
water
0.10
NA
NA
kg/m3 leachate
None
NA
COD, to water
0.04
NA
NA
kg/m3 leachate
None
NA
Nitrogen, to
water
0.01
NA
NA
kg/m3 leachate
None
NA
Ammonium,
to water
3.0E-3
NA
NA
kg/m3 leachate
None
NA
WWTP
Effluent -
Partial
Diversion
Electricity
0.05
0.02
0.10
kWh/m3
Triangular
NA
Electricity data provided by Carollo
Engineering. Base, min and max values
estimated assuming 1, 0.5 and 2 MGD of
wastewater diverted to Rio Grande,
respectively.
1 Geometric standard deviation assigned based on recommended value from the Ecoinvent data quality pedigree matrix for a 'qualified estimate.' (Ciroth et al.,
2012)
2 '/m3' notation in the unit denominator indicates that inventory data is normalized on the basis of 1 m3 of treated wastewater.
3 Disk filtration is only included for Baseline and Scenario 1.
EP-C-I7-04I; WA 4^77
B-7
-------
Appendix B - Life Cycle Inventory Data
Table B-2. Life cycle inventory data for unit process data specific to Scenario 1 - Sidestream Treatment
Input
Name
Original
Unit
Process
Name
Mean
Value
Min
Max
Units2
Distributio
n Type
Geometric
Standard
Deviation
Uncertainty Range Note
Secondary
Treatment -
Biological
Methane,
to air
8.0E-3
NA
NA
kg CHVm3
Lognonnal1
1.69
No change from baseline. See Appendix Section
B.2 for details on process GHG emission
estimation.
Nitrous
oxide, to
air
4.0E-4
9.5E-5
7.4E-4
kg N2O/1113
Lognonnal1
1.69
See Appendix Section B.2 for details on process
GHG emission estimation.
Sidestream
Treatment -
Filtrate
Electricity
0.03
0.03
0.04
kWh/m3
Triangular
NA
Electricity demand of the phosphorus and nitrogen
sidestream processes is estimated to be 1,000 and
500 kWh/day, respectively. Min and max values
estimated assuming +/-10% of reported value.
Magnesiu
m Chloride
3.7E-3
3.3E-3
4.1E-3
kg active
ingredient/
m3
Triangular
NA
Estimated use is 100 gallons per day of 33% MgCb.
Mass of active ingredient is estimated assuming a
density of 1.32 g/cm3. Min and max values
estimated assuming +/-10% of reported value.
1 Geometric standard deviation assigned based on recommended value from the Ecoinvent data quality pedigree matrix for a 'qualified estimate.' (Ciroth et al.,
2012)
2 7m3' notation in the unit denominator indicates that inventory data is normalized on the basis of 1 m3 of treated wastewater.
EP-C-17-041; WA 4-77
B-8
-------
Appendix B - Life Cycle Inventory Data
Table B-3. Life cycle inventory data for unit process data specific to Scenario 2 - Tertiary Filters
Input
Name
Original
Unit Process
Name
Mean
Value
Min
Max
Units1
Distribution
Type
Geometric
Standard
Deviation
Uncertainty Range Note
Tertiary
Treatment -
Deep Bed
Media
Filters
Electricity
0.04
0.04
0.05
kWh/m3
Triangular
NA
Estimated use to run both filters at design
capacity is 700,000 kWh/year. Scaled to average
annual flow. Min and max values estimated
assuming +/- 10% of reported value.
Sand
4.0E-3
3.4E-3
4.2E-3
kg/m3
Triangular
NA
The mass of sand required was estimated
assuming a sand volume of 6,250 ft3 and a
density of 1,522 kg/m3. A density range of 1,281
- 1,602 kg/m3 was used to estimate min/max
values. Media lifespan = 20 years.
Anthracite
2.1E-3
1.9E-3
2.3E-3
kg/m3
Triangular
NA
The mass of anthracite required was estimated
assuming an anthracite volume of 6,250 ft3 and a
density of 50 lb/ft3. Min and max values
estimated assuming +/-10% of base value.
Media lifespan = 20 years.
Gravel
1.7E-3
1.5E-3
1.9E-3
kg/m3
Triangular
NA
The mass of gravel required was estimated
assuming a gravel volume of 2,500 ft3 and a
density of 1,602 kg/m3. Min and max values
estimated assuming +/-10% of base value.
Media lifespan = 20 years.
Methanol
7.0E-3
6.0E-3
8.0E-3
kg/m3
Triangular
NA
Methanol demand is 3.5 lbs methanol/lb
Nitrogen removed. The quantity of nitrogen
removed was calculated based on difference in
reported effluent quality between baseline and
Scenario 2. Min/max values were calculated
using a range of methanol demands of 3-4 lbs/lb
Nitrogen.
Alum
0.03
0.02
0.07
kg/m3
Triangular
NA
The alum requirement was estimated based on
removal of 0.95 mg P/L (base value) with a
target molar ratio of 5 g Aluminum/g
Phosphorus. The aluminum content of alum
(Al2(S04)3) is 0.16 g/g. Min/max values were
calculated based on the potential range of
required phosphorus removal, 0.85-2.45 mg P/L.
1 7m3' notation in the unit denominator indicates that inventory data is normalized on the basis of 1 m3 of treated wastewater.
EP-C-17-041; WA 4-77
B-9
-------
Appendix B - Life Cycle Inventory Data
Table B-4. Life cycle inventory data for unit process data specific to Scenario 3 - Reverse Osmosis
Input Viiik'
()ri!jn;il
I nil
Process
N;inu'
Mo;iii
Value
Min
M;i\
I nils'
Dislrihulinn
1 J |H>
(iooniolric
Siiindiinl
Do iiilion
I iKvr(;iin(> Rimiic Nolo
Tertiary
Treatment -
Microfiltratio
n
Electricity
0.06
0.05
0.07
kWh/m3
Triangular
NA
Electricity data provided by Carollo Engineering. Min
and max values estimated assuming +/-10% of base
value.
Sodium
hypochlorite
1.0E-4
9.0E-5
1.1E-4
kg/m3
Triangular
NA
Chemical use data provided by Carollo Engineering.
Min and max values estimated assuming +/-10% of
base value.
Caustic soda
6.6E-5
5.9E-5
7.2E-5
kg/m3
Triangular
NA
Sulfuric
Acid
1.1E-5
3.0E-6
1.2E-5
kg/m3
Triangular
NA
Chemical use data provided by Carollo Engineering.
Min and max values estimated using range of sulfuric
acid concentrations 25%-98%.
Citric acid
7.4E-5
6.6E-5
8.1E-5
kg/m3
Triangular
NA
Chemical use data provided by Carollo Engineering.
Min and max values estimated assuming +/-10% of
base value.
Membrane,
MF/RO
6.5E-5
4.1E-5
1.4E-4
kg/m3
Triangular
NA
MF membrane material is modeled as
polyvinylfluoride. The quantity of membrane material
was estimated assuming installation of 123 (range: 82-
247) membrane units with an average lifespan of 9
years. Each unit has a membrane area of 77 m2.
Membrane fiber specifications used in base, min and
max calculations are as follows: pore diameter - 0.03
|im. outer diameter - 1.3E-3 m, inner diameter - 7.0E-4
m, circumference - 4.1E-3 m and a PVDF density of
1.8 g/cm3 (range: 1.68-1.97).
Tertiary
Treatment -
Reverse
Osmosis
Electricity
0.27
0.24
0.29
kWh/m3
Triangular
NA
Electricity data provided by Carollo Engineering. Min
and max values estimated assuming +/-10% of base
value.
Membrane,
MF/RO
2.6E-4
2.3E-4
2.9E-4
kg/m3
Triangular
NA
RO membrane material is modeled as
polyvinylfluoride. The quantity of membrane material
was estimated assuming a 7 MGD flowrate to the RO
unit at the facilities design flow of 12 MGD. Each RO
membrane unit has a flowrate of 1.6 m3/hr requiring
1,049 membrane units with a safety factor of 1.5.
Membrane units have an average expected lifespan of
9 years.
EP-C-I7-04I; WA 4^77
B-10
-------
Appendix B - Life Cycle Inventory Data
Input Name
Original
Unit
Process
Name
Mean
Value
Min
Max
Units1
Distribution
Type
Geometric
Standard
Deviation
Uncertainty Range Note
Proprietary
solution 1,
citric acid
2.5E-4
1.8E-4
3.1E-4
kg/m3
Triangular
NA
Chemical use data provided by Carollo Engineering.
No information is available on the composition of
proprietary cleaning chemicals. Citric acid is used as a
proxy data source in the LCA model. Min and max
values estimated assuming +/- 25% of base value, due
to uncertainty about chemical composition.
Proprietary
solution 2,
citric acid
4.6E-5
3.5E-5
5.8E-5
kg/m3
Triangular
NA
Chemical
Post-
Treatment
Carbon
dioxide
0.01
0.01
0.02
kg/m3
Triangular
NA
Chemical use data provided by Carollo Engineering.
Min and max values estimated assuming +/-10% of
base value.
Caustic soda
0.01
0.01
0.02
kg/m3
Triangular
NA
Brine -
Underground
Inject
Electricity
0.55
0.45
0.69
kWh/m3
Triangular
NA
Electricity consumption was estimated assuming a
brine flowrate of 0.088 m3/sec with a required pump
pressure and efficiency of 1300 psi and 75%,
respectively. Min and max values estimated assuming
a range in pump pressures of 1200-1400 psi and a
range of pump efficiencies between 65 and 85%.
Water, to
ground
0.17
NA
NA
m3/m3
NA
NA
The volume of produced brine is estimated assuming a
reject rate of 29%.
1 7m3' notation in the unit denominator indicates that inventory data is normalized on the basis of 1 m3 of treated wastewater.
EP-C-17-041; WA 4-77
B-ll
-------
Appendix B - Life Cycle Inventory Data
Table B-5. Life cycle inventory data for unit process data specific to Scenario 4 - Full Effluent Diversion
Input Name
Original Unit
Process
Name
Mean Value
Min
Max
Units
Distribution
Type
Geometric
Standard
Deviation
Uncertainty Range Note
WWTP
Effluent - Full
Diversion
Electricity
0.18
0.13
0.20
kWMn3
Triangular
NA
Electricity data provided by Carollo Engineering.
Min and max values estimated based on the range
in estimated baseline pumping.
EP-C-17-041; WA 4-77
B-12
-------
Appendix B - Life Cycle Inventory Data
B.2 Greenhouse Gas Analysis
This section details the calculations used to determine the process-level GHG emissions
from the wastewater treatment and sludge handling stages, from the effluent, and from landfilled
sludge.
B.2.1 Methane Emissions from Biological Treatment
The methodology for calculating CH4 emissions associated with the wastewater treatment
configurations evaluated as part of this study is generally based on the guidance provided in the
IPCC Guidelines for national inventories. CH4 emissions are estimated based on the amount of
organic material (i.e., BOD) entering the unit operations that may exhibit anaerobic activity, an
estimate of the theoretical maximum amount of methane that can be generated from the organic
material (B0), and a methane correction factor that reflects the ability of the treatment system to
achieve that theoretical maximum. In general, the IPCC does not estimate CH4 emissions from
well managed centralized aerobic treatment systems. However, there is acknowledgement that
some CH4 can be emitted from pockets of anaerobic activity, and more recent research suggests
that dissolved CH4 in the influent wastewater to the treatment system is emitted when the
wastewater is aerated. The PR WWTP includes an optional anoxic zone preceding the aerated
oxidation ditches.
The methodological equation, adapted from (IPCC, 2006; RTI International, 2010), is:
CH4 PROCESS = BOD (mg/L) x Flow (m3/yr) x lxlO3 L/m3 x lxlO"6 kg/mg x B0 x MCF
Equation B-l
where:
CH4 process = CH4 emissions from wastewater treatment process (kg CH4 /yr)
BOD = Concentration of BOD entering biological treatment process (mg/L)
Flow = Wastewater treatment flow entering biological treatment process (m3/yr)
Bo = maximum CH4 producing capacity, 0.6 kg CHVkg BOD (IPCC, 2006)
MCF = methane correction factor (fraction)
IPCC guidelines recommend an MCF of 0 for a well-managed aerobic treatment process
with an uncertainty range of 0-0.1. Daelman et al. (2013)evaluated emissions associated with a
municipal treatment plant with biological nutrient removal (specifically nitrification and
denitrification), resulting in an MCF of 0.05, which was used for this study due the presence of
an anoxic zone for nitrification and denitrification and the fact that this value falls in the middle
of the recommended IPCC range. This calculation estimates that approximately 53 metric tons of
methane may be released annually from the biological treatment process.
EP-C-I7-04I; WA 4^77
B-13
-------
Appendix B - Life Cycle Inventory Data
B.2.2 Nitrous Oxide Emissions from Biological Treatment
The methodology for calculating N2O emissions associated with wastewater treatment is
based on estimates of emissions reported in the literature. The guidance provided in the IPCC
Guidelines for national inventories does not provide a sufficient basis to distinguish N2O
emissions from varying types of wastewater treatment configurations, particularly related to
biological nutrient reduction. More recent research has highlighted the fact that emissions from
these systems can be highly variable based on operational conditions, specific treatment
configurations, and other factors (Chandran, 2012). For this analysis, the best available N2O
emission factor for the biological treatment plant at the PR WWTP is for a plug-flow activated
sludge treatment process. The reported emission factors indicate that between 0.09% and 0.62%
of TKN influent to the biological process will be released as N2O. The average of minimum and
maximum emission factors, 0.36% was used as the baseline value in this study with the full
range of emission factors informing the uncertainty assessment.
The methodological equation is:
N20 PROCESS — TKN (mg/L) x Flow (m3/yr) x lxlO3 L/m3 x lxlO6 kg/mg x EF% x 44/28
B. 2.3 Methane Emissions due to Anaerobic Digestion
Fugitive methane emissions from the anaerobic digesters are estimated as a function of
biogas production, biogas methane content, methane density and assumed biogas leakage. The
facility reports biogas production of 230,114 standard cubic feet/day (2,380,000 m3/yr). Methane
content of the produced biogas averaged 59.1% for the period from March 2020 to July 2021.
The calculation assumes a methane density of 0.0417 lb/ft3 (0.668 kg/m3) at normal temperature
and pressure. Fugitive methane leakage rates assessed in the LCA literature range from 0% to
5% of produced biogas CH4 (Levis and Barlaz, 2011; Woon et al., 2016) with most values falling
in the range of 1-3% (Slorach et al., 2019; Yoshida et al., 2012). A value of 2% was selected as
the base value of this study, while the uncertainty assessment reflects a range of fugitive methane
leakage between 1.2% and 5%. The base calculation estimates that approximately 19 metric tons
of methane will be released annually from the digesters (range of 10-53 metric tons).
B.2.4 Nitrous Oxide Emissions from Effluent Discharged to Receiving Waters
The methodology for calculating nitrous oxide emissions associated with effluent
discharge is based on the guidance provided in the IPCC Guidelines for national inventories.
EP-C-! 7-041; WA 4^77 B-14
Equation B-2
where:
N2O PROCESS
TKN =
Flow =
EF% =
44/28 =
= N2O emissions from wastewater treatment process (kg N2O /yr)
Concentration of TKN entering biological treatment process (mg/L)
Wastewater treatment flow entering biological treatment process (m3/yr)
average measured % of TKN emitted as N2O, %
molecular weight conversion of N to N2O
-------
Appendix B - Life Cycle Inventory Data
N2O emissions from domestic wastewater (wastewater treatment) were estimated based on the
amount of nitrogen discharged to aquatic environments from each system configuration, which
accounts for nitrogen removed with sewage sludge.
N2OEFFLUENT = Neffluent (mg/L) x Flow (m3/yr) x lxl03 L/m3 x lxlO"6 kg/mg x EF3 x 44/28
Equation B-3
where:
N20effluent
Neffluent
Flow
EF3
44/28
N2O emissions from wastewater effluent discharged to aquatic
environments (kg N20/yr)
N in wastewater discharged to receiving stream, mg/L
Effluent flow, m3/yr
Emission factor (0.005 kg N2O -N/kg sewage-N produced)
Molecular weight ratio of N2O to N2
EP-C-I7-04I; WA 4^77
B-15
-------
Appendix C - LCIA Results
APPENDIX C
LCIA RESULTS
EP-C-I7-04I; WA 4^77
-------
Appendix C - LCIA Results
Appendix C - LCIA Results
See the accompanying Excel file titled Appendix C - LCIA Results.
EP-C-17-04I: WA 4-
C-l
-------
Appendix D - Data Quality Matrix
APPENDIX D
DATA QUALITY ASSESSMENT
EP-C-I7-04I; WA 4^77
-------
Appendix D - Data Quality Assessment
Appendix D - Data Quality Assessment
D.l Data Quality Indicators Matrix
Table D-l. Data Quality Indicators Matrix
I nil Pmi'i-ss l);il;i Qu;ilil\ Indiiiiliir (1-5)
Si ill I'll'
Si hi I'l l-
Kili;il)ilil\
(iimpliii'lli'ss
I'hinl l);il;i
Ti-ni|>iir;il
(uiTi'hiliim
(;iii^l';i|)llii;il
(iiiTi'hiliun
li'iliniiln^ii;il
(iMTi'hiliim
Carollo 2021
Sidestream
Phosphorous
Removal, Deep
Bed Media Filters,
Micro filter,
Reverse Osmosis,
Chemical Post
Treatment,
Diversion Energy
Difference
2
1
1
1
1
Lemon 2021a
Diversion
1
1
1
1
1
Lemon 2021b
Diversion
1
1
1
1
1
Luna 2021a
Sidestream
Nitrogen Removal,
Deep Bed Media
Filters
1
1
1
1
1
Polymer
Belt Filter Press
1
1
1
1
PR Air Permit
Flare, CLIP, Dual-
fuel Boilers
1
1
1
1
PR Compost
Electricity
Composting
1
1
1
1
1
PR Flow Data
Influent
1
1
1
1
1
PR Headworks
Electricity
Facility Total
(Electricity)
1
1
1
1
PR Input Data
Facility Total
(Lime, Diesel)
1
1
1
1
1
PR Metals Data
Effluent
1
1
1
1
1
PR Natural Gas
Composting
1
1
1
1
1
PR Nonpotable
Electricity
Facility Total
(Electricity)
1
1
1
1
1
PR Renewable
Energy
Flare, CLIP, Dual-
fuel Boilers
1
1
1
1
1
PR Reuse Data
Water Reuse
1
1
1
1
1
PR SCADA Data
Biological
Treatment
1
1
1
1
1
PR Solid Waste
Composting
1
1
1
1
PR Staff 2021
Facility Total
(Lime)
1
1
1
1
1
PR Staff Email
2021a
Screenings and Grit
1
1
1
1
1
PR Staff Email
2021b
Composting
1
1
1
1
1
EP-C- / 7-04 /; WA 4-
D-l
-------
Appendix D - Data Quality Assessment
I nil Pniii'ss Diiiii ()u;ilil\ lii(lii;il(ir (1-5)
S< hi riv
Si ill I'l l'
ki'li;il>ilil>
(iimpli'li'iKss
1. ili-nil u iv
li'inpniiil
(iiiTi'kilinn
Ci'ii^nipliiml
(iii'ii'liiliiin
liiliniilii^ii;il
(iiiTi'hiliiui
Amlinger el al.
2008
Composting
1
1
4
3
1
Andreoli et al.
2007
Anaerobic
Digestion
1
1
4
1
1
Bastian et al.
2011
CHP
1
1
3
2
1
Boldrin et al.
2009
Composting
1
1
4
3
2
Boldrin et al.
2011
Land Application
1
1
4
3
1
Bonton et al.
2012
Reverse Osmosis
1
1
3
3
1
Chandran2012
Biological
Treatment
1
1
3
1
1
DEUSA 2021
Sidestream
Phosphorous
Removal
1
1
1
2
1
Disk Filter
Disk Filter
1
1
1
2
2
Favoino and
Hogg 2008
Composting
1
1
4
3
2
Filter Media
(Anthracite)
Deep Bed Media
Filters
1
1
3
2
2
Filter Media
(Gravel)
Deep Bed Media
Filters
1
1
3
2
2
Fukomoto et al.
2003
Composting
1
1
5
3
2
Gas Density
Anaerobic
Digestion
1
1
5
1
1
Gonzalez et al.
2020
Land Application
1
1
1
3
1
HDR2016
Biological
Treatment
1
1
2
1
1
Hellebrand 1998
Composting
1
1
5
1
Hellmann 1997
Composting
1
1
5
3
1
IPCC 2006
Biological
Treatment
1
1
5
3
2
K2S04
Properties
Land Application
1
1
1
Kaberline et al.
2017
Sidestream
Nitrogen Removal
1
1
2
2
1
Keng et al. 2020
Composting
1
1
1
3
1
Khoshnevisan et
al. 2018
Land Application
1
1
2
3
1
Maulini-Duran et
al. 2013
Composting
1
1
3
3
1
Membrane Filters
Microfilter
1
1
4
1
1
Morelli et al.
2019
Composting, Land
Application
1
1
1
1
1
KP
-041;
D-2
-------
Appendix D - Data Quality Assessment
Source
Unit Process(es)
Unit Process Data Quality Indicator (1-5)
Source
Reliability
Completeness
Temporal
Correlation
Geographical
Correlation
Technological
Correlation
Nemecek and
Kagi 2007
Laud Application
1
1
4
3
1
Nitto 2019
Reverse Osmosis
1
1
1
2
1
Nkoa et al. 2014
Anaerobic
Digestion
1
1
3
3
1
O'Kelly 2005
Anaerobic
Digestion
1
1
5
1
1
Phosphoric Acid
Disinfection (UV)
1
1
1
1
1
Razza et al. 2009
Composting
1
1
4
3
Richard 2014
Composting
1
1
3
1
1
ROU 2007
Composting
1
1
4
3
Saer et al. 2013
Composting
1
1
3
3
1
Salemdeeb et al.
2017
Land Application
1
1
2
3
1
Sulfuric Acid
Microfilter
1
1
3
1
1
SYLVIS 2011
Composting
1
1
4
3
1
Tiquia et al. 2002
Composting
1
1
5
2
1
U.S. EPA 2003
Screenings and Grit
1
1
5
2
1
Yoshida et al.
2012
Composting
1
1
3
2
1
EP-C-17-041; WA 4-77
D-3
-------
Appendix E - Determination of Metals Removal Performance
APPENDIX E
DETERMINATION OF METALS REMOVAL PERFORMANCE
EP-C-I7-04I; WA 4^77
-------
Appendix E - Determination of Metals Removal Performance
Appendix E - Determination of Metals Removal Performance
The metals removal performance of each treatment scenario was determined using a
combination of historical water quality data and expected performance of upgraded treatment
scenarios based on the performance of similar systems. Given the nature of the data and how
they were used to compute removal rates, tables are not shown in text form here. Instead, the
reader is referred to the Metals tab of the project LCI workbook.
Influent and effluent data were provided by the PR WWTP in a range of forms, including
long term effluent concentrations as well as a series of 4 quarterly samples, where pairwise
influent and effluent samples were taken. ERG use these monthly samples to determine historic
metals removal performance, which was assumed (conservatively) to be representative of the
Baseline Scenario. To calculate removal rates for the Baseline Scenario as well as Scenarios 1-3,
the following rules/methods were used:
• Where an influent concentration was measured, but the effluent sample returned a
non-detect, the effluent concentration was assumed to be one half of the minimum
detection limit (MDL), where the MDL was based on EPA Method 200.8 for metals
and 245.1 for mercury.
• Scenario 1 (Filtrate Treatment) uses an annamox-based process and struvite
production to remove nutrients, neither of which target metals. ERG therefore
assumed that the metals removal performance of SI was the same as the Baseline
Scenario,
• For Scenarios 2 (Tertiary Filters) and 3 (RO), performance of similar systems from
U,S_ _EP A_(202_1J were used as surrogates for determination of the percent removal
performance of individual metal species.
• Scenario 4 was assumed to perform the same as the Baseline Scenario.
EP-C-17-04I: WA 4-
E-l
-------
Appendix F - Parameter Sensitivity Results
APPENDIX F
PARAMETER SENSITIVITY RESULTS
EP-C-I7-04I; WA 4^77
-------
Appendix F - Parameter Sensitivity Results
Appendix F - Parameter Sensitivity Results
Acidification Potential
Main Electricity ¦
Compost Emissions ™
Biogas Production ¦
S3 Electricity
S2 Alum
51 Electricity
Reuse Electricity
Biogas to Flare
52 Electricity
Land Application... -
Diversion Electricity
Biological GHG...
Nutrient Emissions
Metal Emissions
i Baseline
l SI - Sidestream Filtration
i S2-Tertiary Filters
S3 - Reverse Osmosis
I S4 - Zero Discharge
Eutrophication Potential
0.0% 2.0% 4.0% 6.0
Nutrient Emissions
Land Application...
Main Electricity -
S3 Electricity
Biogas Production
S2Alum -
Compost Emissions -
Diversion Electricity
Biological GHG...
Biogas to Flare
51 Electricity
52 Electricity
Reuse Electricity
Metal Emissions
I Baseline
¦ SI - Sidestream Filtration
¦ S2-Tertiary Filters
S3 - Reverse Osmosis
IS4 - Zero Discharge
Figure F-l. Environmental metric (Acidification Potential and Eutrophication Potential)
sensitivity to important model parameters. Axis values represent the percent change in
baseline impact that results from a +/-10% change in parameter values.
Cumulative Energy Demand
0% 20% 40%
Main Electricity ¦
Biogas Production ¦
S3 Electricity
Reuse Electricity ¦
Diversion Electricity 1
51 Electricity "
52 Electricity -
S2 Alum ¦
Biogas to Flare
Biological GHG...
Compost Emissions
Land Application...
Nutrient Emissions
Metal Emissions
¦ Baseline
¦ SI - Sidestream Filtration
¦ S2-Tertiary Filters
S3 - Reverse Osmosis
¦ S4-Zero Discharge
Global Warming Potential
Biological GHG...
Main Electricity
S3 Electricity
Biogas Production
Compost Emissions
Reuse Electricity
Diversion Electricity
S2 Electricity
S2 Alum
SI Electricity
Land Application...
Biogas to Flare
Nutrient Emissions
Metal Emissions
i Baseline
l SI - Sidestream Filtration
l S2 - Tertiary Filters
S3 - Reverse Osmosis
i S4- Zero Discharge
Fossil Fuel Depletion
Smog Formation Potential
20% 40% 60% 80% 100%
Main Electricity
Biogas Production
S3 Electricity
Reuse Electricity t
S2Alum —
Diversion Electricity L
S2 Electricity —
SI Electricity '
Biological GHG Emissions
Biogas to Flare
Compost Emissions
Land Application Emissions
Nutrient Emissions
Metal Emissions
i Baseline
i SI - Sidestream Filtration
i S2- Tertiary Filters
S3 - Reverse Osmosis
i S4- Zero Discharge
Biogas Production
Main Electricity
S3 Electricity
Reuse Electricity
S2 Alum
Diversion Electricity
S2 Electricity
SI Electricity
Compost Emissions
Biological GHG...
Biogas to Flare
Land Application...
Nutrient Emissions
Metal Emissions
l Baseline
l SI - Sidestream Filtration
i S2 - Tertiary Filters
S3 - Reverse Osmosis
i S4 - Zero Discharge
Figure F-2. Energy and Climate metric (Cumulative Energy Demand, Global Warming
Potential, Fossil Fuel Depletion and Smog Formation Potential) sensitivity to important
model parameters. Axis values represent the percent change in baseline impact that results
from a +/-10% change in parameter values.
EP-C-17-041; WA 4-77
F-l
-------
Appendix F - Parameter Sensitivity Results
Water Depletion
0% 10% 20% 30% 40%
S2 Alum
Main Electricity
Biogas Production
S3 Electricity
Reuse Electricity
Biological GHG...
Biogas to Flare
51 Electricity
52 Electricity
Compost Emissions
Land Application...
Nutrient Emissions
Metal Emissions
Diversion Electricity
¦ Baseline
¦ SI - Sidestream Filtration
¦ S2-Tertiary Filters
¦ S3 - Reverse Osmosis
¦ S4 - Zero Discharge
Figure F-3. Water metric (Water Depletion) sensitivity to important model parameter. Axis
values represent the percent change in baseline impact that results from a +/-10% change
in parameter values.
Ecotoxicity
Main Electricity
Biogas Production
S3 Electricity
Metal Emissions ¦
Reuse Electricity ¦
Diversion Electricity 1
S2 Electricity ¦
SI Electricity "
S2 Alum
Biogas to Flare
Biological GHG...
Compost Emissions
Land Application...
Nutrient Emissions
¦ Baseline
¦ SI - Sidestream Filtration
¦ S2-Tertiary Filters
S3 - Reverse Osmosis
IS4 - Zero Discharge
Human Health Particulate Matter Formation
Main Electricity
Compost Emissions
S2 Alum
S3 Electricity
Biogas Production
Reuse Electricity
Diversion Electricity
51 Electricity
52 Electricity
Land Application...
Biogas to Flare
Biological GHG...
Nutrient Emissions
Metal Emissions
I Baseline
i SI - Sidestream Filtration
i S2-Tertiary Filters
S3 - Reverse Osmosis
I S4 - Zero Discharge
Human Health Toxicity Cancer Potential
0% 10% 20% 30%
Human Health Toxicity Noncancer Potential
0% 10% 20% 30%
Main Electricity ¦
Biogas Production [
S3 Electricity
Reuse Electricity ¦
S2 Alum ¦
Diversion Electricity 1
S2 Electricity -
SI Electricity "
Metal Emissions J
Biogas to Flare
Biological GHG...
Compost Emissions
Land Application...
Nutrient Emissions
i Baseline
i SI - Sidestream Filtration
i S2-Tertiary Filters
S3 - Reverse Osmosis
iS4-Zero Discharge
Main Electricity
Biogas Production
S3 Electricity
Reuse Electricity
Diversion Electricity
S2 Electricity
Metal Emissions
SI Electricity
S2 Alum
Biogas to Flare
Biological GHG...
Compost Emissions
Land Application-
Nutrient Emissions
i Baseline
¦ SI - Sidestream Filtration
¦ S2-Tertiary Filters
S3 - Reverse Osmosis
lS4-Zero Discharge
Figure F-4. Toxicity metric (Ecotoxicity, Human Health Particulate Matter Formation,
Human Health Toxicity Cancer Potential, Human Health Toxicity Noncancer Potential)
sensitivity to important model parameter. Axis values represent the percent change in
baseline impact that results from a +/-10% change in parameter values
EP-C-17-041; WA 4-77
F-2
------- |