Citation
Sustainability assessment of wastewater treatment plants with water reuse for urban agriculture

Material Information

Title:
Sustainability assessment of wastewater treatment plants with water reuse for urban agriculture a case study in Hyderabad, India
Creator:
Miller, Leslie
Publication Date:
Language:
English
Physical Description:
xv, 130 leaves : illustrations ; 28 cm

Subjects

Subjects / Keywords:
Sewage disposal plants -- Case studies -- India -- Hyderabad ( lcsh )
Sewage irrigation -- Case studies -- India -- Hyderabad ( lcsh )
Urban agriculture -- Case studies -- India -- Hyderabad ( lcsh )
Sustainable agriculture -- Case studies -- India -- Hyderabad ( lcsh )
Sewage disposal plants ( fast )
Sewage irrigation ( fast )
Sustainable agriculture ( fast )
Urban agriculture ( fast )
India -- Hyderabad ( fast )
Genre:
Case studies. ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )
Case studies ( fast )

Notes

Bibliography:
Includes bibliographical references (leaves 120-130).
General Note:
College of Engineering and Applied Science
Statement of Responsibility:
by Leslie Miller.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
747428759 ( OCLC )
ocn747428759
Classification:
LD1193.E52 2011d M54 ( lcc )

Full Text
SUSTAINABILITY ASSESSMENT OF WASTEWATER TREATMENT PLANTS
WITH WATER REUSE FOR URBAN AGRICULTURE: A CASE STUDY IN
HYDERABAD, INDIA
By
Leslie Miller
B.A., Biology, Boston University 2004
M.Eng., Civil Engineering, University of Colorado Denver 2008
A thesis submitted to the
University of Colorado Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy (Ph.D.)
College of Engineering and Applied Sciences
May 2011


This thesis for the Doctor of Philosophy
degree by
Leslie Ann Miller
has been approved by
Anu Ramaswami
Priyanie Amerasinghe
Ram Ramaswami
Joann Silverstein


Miller, Leslie A. (Ph.D., College of Engineering and Applied Sciences)
Sustainability Assessment of Wastewater Treatment Plants with Water Reuse for
Urban Agriculture in India
Thesis directed by Professor Anu Ramaswami.
ABSTRACT
Wastewater reuse in urban agriculture is a widespread practice in many developing
world cites that has many advantages (water savings, nutrient cycling, and
livelihoods) and disadvantages (pathogen health risk). Energy use and associated
greenhouse gas (GHG) emissions in centralized wastewater treatment plants
(WWTPs) can mitigate some of the health risks, however these tradeoffs have not
been quantified. The objective of this thesis is to conduct a sustainability assessment
of WWTPs with water reuse for urban agriculture in India. Three stages of work were
included.
1) The role of water and wastewater (W/WW) infrastructures in urban energy
metabolism was explored first. W/WW infrastructures were found to contribute 3 to
16% of community-wide electricity use and greenhouse gas (GHG) emissions for 16
Indian cities; for another 23 the proportion was less than 3%. Energy intensity for
drinking water supply and wastewater treatment averaged 1.30.7 Wh/gal (n=7) and
0.40.2 Wh/gal (n=5), respectively. Energy intensity for water pumping/treatment
was more than double that for wastewater, the reverse of cities in Colorado, USA,
likely due to poorer source water quality in India. 2
2) A life cycle assessment (LCA) was conducted of Nallacheruvu (8MGD) WWTP in
Hyderabad, India that used upflow anaerobic sludge blanket reactor followed by
oxidation ponds, yielding 99% and 81% removal of fecal coliforms and BOD5,
respectively. The LCA showed energy use and GHG emissions of 0.7Wh/gal and
lgC02e/gal, 48% of the later from on-site electricity use, 41% from methane process
emissions, 10% from embodied energy in infrastructure, and 1% from nitrous oxide
process emissions. A consequential LCA, conducted using the DAYCENT model for
wastewater reuse in urban agriculture, showed only 1% of the nitrogen in treated
effluent was reused in urban agriculture, due to land constraints along the flow path
of the wastewater. As a result, annual system-wide GHG emissions for untreated and
treated wastewaters releases to the riverine system were similar at 2,463mtC02e and
2,819mtC02e, respectively. Avoided impacts due to reuse of biogas for electricity
and avoided fertilizer each accounts for 5% reductions from the total for treated
water.


3) An urban agriculture site study was conducted to assess the impact of pathogen
reduction in WWTP on spinach. This was explored in a site study using three
different waters: groundwater, treated effluent from WWTP, and untreated water.
While E.coli in the waters consistently differed by 2-3 orders of magnitude between
the three plots, the E.coli in the crop measured at the endpoint of the study (harvest
conditions) was not significantly different between the groundwater and WWTP plots
(t-test P>0.1), while the untreated water was slightly higher (P<0.025). For Ascaris,
qualitative results showed little difference between Ascaris on crop grown with
treated and untreated waters (26-36 eggs/lOOg spinach), while groundwater-irrigated
spinach had lower Ascaris levels (9 eggs/lOOg spinach). Unexpectedly, when the
researcher carefully took crop samples, the E.coli results compared to farmer-
harvested crop were one order of magnitude lower, suggesting recontamination of the
crop from farmer handling and contact with soil and contaminated water. The
recontamination hypothesis was confirmed (P<0.1) by sampling each of the three
plots, comparing sanitized handling (n=3) versus farmer handling (n=3) in each plot.
Using data from WWTP LCA and urban agriculture site study, a sustainability
assessment showed that treated effluent and untreated surface water were similar in
the case of GHG emissions, pathogen risk (15% and 18% probability of disease over
one year based on E.coli results), yield (20kg/m and 23kg/m for one year) and water
saved (0 gallons groundwater used), but varied in terms of economic investment
($97,093 vs $0 per year). The groundwater site had lower GHG emissions, energy
use, pathogen risk, but consumed a lot of water (1,125 gallons/m2/year), yielded only
10% crop at harvest (2kg/m2) compared to the other plots, and cost approximately
$600 per year. If groundwater tables are at risk, then wastewater reuse offers an
alternative. While the WWTP technology did not provide as many benefits to urban
agriculture as expected, there may be benefits towards cleaning up the Musi River
and avoiding groundwater contamination that are beyond the scope of this work.
More studies with social actors and institutions are needed to identify sustainability
priorities in each community.
This work contributes to:
The urban metabolism literature by examining energy use and intensity in
water/wastewater infrastructures in developing country cities,
WWTP LCA literature by conducting a first LCA using operating data from an
Indian WWTP with water reuse in urban agriculture,
The urban agriculture literature by completing a first field study of pathogen
impacts from treated and untreated wastewater use, and


The sustainability assessment literature as it links water/wastewater, energy,
infrastructure capital investments, urban agriculture, and health.
This abstract accurately represents the content of the candidates thesis. 1 recommend
its publication.
Signed
Anu Ramaswami


DEDICATION
To the strong women who have inspired me.
Also, to my father and mother, who never put limits on what I could do.


ACKNOWLEDGEMENT
First, Id like to recognize the National Science Foundations Interdisciplinary
Graduate Education and Research Traineeship Grant (IGERT; Grant # DGE-
0654378), which funded this project. There are many individuals who made this
dissertation possible. I sincerely thank my advisor, Dr. Anu Ramaswami, who spent
countless hours with me to conceptualize, articulate, organize, and carry out this
project. She has been a valued mentor in my graduate education and her support has
been invaluable. Dr. Tanya Heikkila was extremely helpful and patient concerning
the interdisciplinary pieces of this project. I am appreciative of Dr. Priyanie
Amerasinghe, of IWMI, who offered indispensable advice and facilitated the project
in Hyderabad. Every one of my committee members at the University of Colorado,
Dr. John Brett, Dr. Ram Ramaswami, and Dr. JoAnn Silverstein, have been a
pleasure to work with and have offered many constructive insights into this project.
Many thanks to Luann Rudolph and Meghan Bernard for their help with numerous
logistical issues.
Site Study, Hyderabad: Mr. Prasanna Kumar was a crucial resource before, during,
and after the site study and has offered cherished advice, thank you. Aruna, the
translator who I worked with, went above and beyond her duties and was a friend. Id
also like to thank Urmilla, who I never met in person, but whose work was very
skilled. Thank you to Chandriah and his family for all of the hard work and I hope to
meet you again someday. Mr. Jagadeeshwar, Lakshmi, Mr. Santosh Kumar, Mr. PN
Raju, and Srinivas of the HMWSSB and Nallacheruvu STP were very gracious in
their cooperation with this project. All of the workers at the Nallacheruvu STP were
willing to help daily with anything that I needed for this site study.
ICRISAT, Hyderabad: Id like to acknowledge the partnership with
IWMI/ICRISAT that enabled lab measurements in India, and that was facilitated by
Mrs. Rosana Mula. I am appreciative of Dr. Gopalakrishnan and everyone in his lab
who made room for me and were a great help. There were many others at ICRISAT
who reliably carried out some of the lab testing for this project: Dr. Sahrawat, Dr.
Pathak, Mr. Murthy, Mr. Sudi, Mr. Andre and everyone in the physical and chemical
soil laboratories. Also, Id like to acknowledge the team at EPTRI and Dr. Srinivas
for their flexibility and carrying out the required tests. My sincere gratitude to
researchers in Hyderabad who offered knowledge and friendship: Belko and
Mainassara for use of the temperature gun, Emilie for joining me on a trip to the site,
Paul for helpful discussions on hydrology, Kiara for helpful discussions on site
selection, and Rafael for help concerning volume measurements over email. Thank


you to Mr. Prabakar for your help with the Indian customs and to Mr. Kumar at
ICRISAT for your help with a lengthy registration process.
Delhi, India: Additionally, the extent of the urban metabolism data would have been
greatly limited without the willingness of Mr. Emani Kumar, Ravi Ranjan, Ashish
Rao, Ashish Verma, Laasya Bhagavatula, Snigda Garg and the entire ICLEI South
Asia team. Also, Mr. Ramesh Negi and Ajay Kumar of the Delhi Pollution Control
Board provided detailed data and Mr. Dwarakanath of the Delhi Government was
very helpful in coordinating meetings. Mr. Reddy and Mr. Rajiv of the Delhi
Pollution Control Committee took time to have many good discussions with me.
Thank you to everyone who helped to make my stay in India an exciting and pleasant
one.
Id like to thank my fellow classmates in the Sustainable Urban Infrastructure
Program at UCD, in particular Abel Chavez, Dr. Tim Hillman, and Dr. Mark Pitterle,
whose work has greatly enhanced my own. I am forever grateful to my parents, my
family, and my friends for all of their continued support throughout this entire
process. Last, but not least, thank you to my fiance, Christian, who understood the
daily happenings of this project more than anyone else and whose optimism
supported me greatly.


TABLE OF CONTENTS
LIST OF FIGURES................................................................x
LIST OF TABLES..............................................................xiii
Chapter
1. Introduction...............................................................1
1.1 Advantages of Wastewater Reuse for Urban Agriculture..................2
1.2 Disadvantages of Wastewater Reuse for Urban Agriculture...............4
1.3 Wastewater Treatment Plants (WWTPs) for Sustainable Water Reuse for
Urban Agriculture...........................................................5
1.3.1 Sustainability assessment of wastewater treatment plants with water
reuse in urban agriculture................................................6
1.3.2 Case study selection...............................................7
1.4 Research Goals and Objectives..........................................8
1.4.1 Phases of Work.....................................................8
2. Role of Water and Wastewater Infrastructures in Urban Energy Metabolism...9
2.1 Introduction...........................................................9
2.1.1 Urban Metabolism: Water, Nutrients, and Energy.....................9
2.2 Data sets used for this study.........................................11
2.2.1 Basic Infrastructure Data.........................................11
2.2.2 Energy use in water systems and in cities overall.................12
2.2.3 Intersection of two data sets.....................................13
2.3 Metrics............................................................. 14
2.4 Results for Indian cities by three metrics............................16
2.4.1 Carbon emissions for water infrastructure as a proportion of overall in-
boundary community emissions........................................16
2.4.2 Electricity use in water and wastewater infrastructure............18
2.4.3 Energy use for municipal water supply treatment and pumping per gallon
and per capita...........................................................20
2.4.4 Energy use for wastewater treatment and pumping per gallon, per capita,
and per BOD load removed.................................................22
2.5 Comparison between India and US.......................................24
2.6 Insights and Recommendations for Future Work..........................27
3. Understanding Case Study Area and Infrastructure Components...............28
3.1 City Description of Hyderabad, India..................................28
3.1.1 Population and Demographics.......................................28
3.1.2 Climate...........................................................29
3.2 Water Supply in Hyderabad.............................................29
3.2.1 Musi River........................................................29
3.2.2 Cross Watershed Transfers.........................................30
3.2.3 Groundwater Withdrawals...........................................31
vii


3.3 Wastewater Infrastructure............................................31
3.3.1 Wastewater Treatment Plant Technology............................31
3.4 Urban Agriculture....................................................34
3.4.1 Soil Characteristics.............................................34
3.5 Discussion...........................................................35
4. Wastewater Treatment Plant Life Cycle Assessment: Nallacheruvu WWTP in
Hyderabad, India............................................................36
4.1 Introduction.........................................................36
4.2 Wastewater Treatment Plant Description...............................40
4.2.1 Physical Description.............................................41
4.2.2 Energy Recovery..................................................42
4.2.3 Process Insight..................................................42
4.2.4 NWWTP Performance Data from HMWSSB...............................44
4.2.5 UASB WWTP Performance............................................45
4.2.6 Consequential Life Cycle Assessment..............................46
4.3 Life Cycle Assessment Methodology....................................47
4.3.1 Goal and Scope of WWTP LCA.......................................49
4.3.2 WWTP LCA Data....................................................49
4.4 LCA Results..........................................................62
4.4.1 On-site Energy Use and GHG Emissions.............................62
4.4.2 Consequential LCA Results........................................65
4.5 Insights and Recommendations for Future Work.........................67
5. Measuring Water and Food Relationships: A Site Study....................69
5.1 Introduction.........................................................69
5.2 Site Selection and Study Design......................................69
5.2.1 Site Selection...................................................69
5.3 Study Design.........................................................73
5.3.1 Composites and Replicates........................................74
5.4 Site Study Methodology...............................................76
5.4.1 Site Preparation and Pre-Analyses................................76
5.4.2 Dynamic Measurements.............................................78
5.4.3 Farming, Irrigation, and Harvesting Practices During Study.......84
5.4.4 Sustainability Assessment Methodology............................86
5.5 Results..............................................................91
5.5.1 Pre-Analysis Results.............................................91
5.5.2 Dynamic Measurements.............................................91
5.6 Insights and Recommendations for Future Work........................101
5.7 Sustainability Tradeoffs and Discussion.............................102
6. Conclusion..............................................................109
6.1 Major Contributions and Conclusions.................................109
6.2 Recommendations for Future Work.....................................Ill
viii


APPENDIX A: DAYCENT METHODOLOGY......................112
APPENDIX B: PATHOGEN TEST METHODOLOGY................113
APPENDIX C: ABBREVIATIONS............................119
References...........................................120
IX


LIST OF FIGURES
Figure 1-1: Urban agriculture in Hyderabad, India................................2
Figure 1-2: A local farmer heading to market with newly harvested spinach grown
with untreated surface water.....................................................3
Figure 1-3: Flood irrigation with untreated surface water in Hyderabad, India....4
Figure 1-4: Sustainability pentagon for evaluating the tradeoffs of the coupled water-
wastewater-urban agriculture system..............................................7
Figure 2-1: Percentage of community-wide GHG emissions associated with in-
boundary energy use that are attibuted to water and wastewater infrastuctures...16
Figure 2-2: Percentage of emissions from water and wastewater infrastructure versus
A) population of Indian cities; B) number of households in Indian cities; C) average
household income in Indian cities ($USD); D) percentage of households with
electricity (Indian state average); or E) human development index (Indian state
average)........................................................................17
Figure 2-3: Water and wastewater flows through Hyderabad........................18
Figure 2-4: Percentage of electricity used for water supply (WS) infrastructure in
comparison to wastewater (WW) infrastructure....................................20
Figure 2-5: Energy use per gallon of municipal water supply (treatment and pumping
separated where available)......................................................21
Figure 2-6: Energy use per gallon of wastewater (treatment and pumping separated
where available)................................................................22
Figure 2-7: Community-wide emissions for Broomfield, Colorado, USA and Bhopal,
India...........................................................................26
Figure 3-1: Map of India........................................................28
Figure 3-2: Urban agriculture adjacent to Nallacheruvu WWTP.....................33
Figure 3-3: Musi River in Hyderabad, India near the outfall of Amberpet WWTP... 35
Figure 4-1: Entrance to NWWTP...................................................40
Figure 4-2: Process flow diagram of Nallacheruvu WWTP. The missing data has been
requested from NWWTP............................................................41
Figure 4-3: UASB with biogas capture (yellow pipes) at NWWTP....................45
Figure 4-4: NWWTP effluent discharge to the stream, with urban agriculture plots
nearby..........................................................................53
Figure 4-5: Results from Beaulieu et al 2011. Percent conversion of N inputs to N2O
via denitrification versus the nitrogen yield from catchment to stream (indicator of
river network length) of 866 rivers worldwide...................................57
Figure 4-6: Buffaloes grazing at NWWTP..........................................59
Figure 4-7: Output of total inorganic nitrogen added in treated effluent irrigation
water as compared to nitrous oxide flux from 10 growing cycles of spinach-cultivated
land............................................................................60
x


Figure 4-8: On-site energy-related and process GHG emissions at NWWTP. Error
bars show the range in emissions..............................................64
Figure 4-9: Wastewater nitrogen flows in the two core scenarios...............65
Figure 4-10: Land easily accessible for NWWTP treated effluent reuse in urban
agriculture....................................................................66
Figure 4-11: Consequential LCA results comparing the GHG emission impact from
releasing untreated wastewater to various interventions........................66
Figure 5-1: Aerial view of NWWTP showing co-location of urban agriculture plots.
1: groundwater; 2: NWWTP effluent; 3: untreated surface water..................70
Figure 5-2: Farmer, Chandriah, and translator, Aruna, at the untreated surface water
plot...........................................................................72
Figure 5-3: Crop pathogen measurement overview: challenge of dilution versus
concentration..................................................................75
Figure 5-4: Soil texture pyramid. Groundwater plot: blue; treated effluent plot:
orange; untreated surface water plot: green....................................77
Figure 5-5: Preparation of the channel, with help of the farmer, Chandriah, leading to
the untreated surface water plot...............................................79
Figure 5-6: Lysimeters installed in groundwater plot...........................83
Figure 5-7: Harvesting spinach, which is bundled and placed under a cloth wetted
with nearby water, usually untreated surface wastewater, to prevent wilting...85
Figure 5-8: Comparison of beta-Poisson and exponential dose-response functions.
Source: Adapted by Asano et al. 2007 from Haas and Eisenberg 2001..............89
Figure 5-9: Initial measurements of E. coli (A) and NPK nutrients (B) in irrigation
water..........................................................................91
Figure 5-10: E. coli (A) and Ascaris (B) in irrigation water over the one month
growing cycle..................................................................92
Figure 5-11: Total nitrogen (A), soluble phosphorus (B), and soluble potassium (C) in
irrigation water for the three plots over one growing cycle....................93
Figure 5-12: E. coli (A) and Ascaris (B) content of soil throughout one growing cycle
(this study)...................................................................94
Figure 5-13: Total nitrogen in soil over one growing cycle. Soil sampling depths are
noted..........................................................................95
Figure 5-14: Total phosphorus in soil over one growing cycle. Soil sampling depths
are noted......................................................................95
Figure 5-15: Total potassium in soil over one growing cycle. Soil sampling depths are
noted..........................................................................96
Figure 5-16: The plots at time of harvest......................................97
Figure 5-17: Harvested bundles of spinach and their roots. The plots from left to
right: groundwater, treated effluent, and untreated surface water.............98
Figure 5-18: E. coli (A) and Ascaris ova (B) on spinach in this study.........99
xi


Figure 5-19: After harvest, bundles stacked together and wetted with nearby water
(often wastewater) to prevent wilting........................................99
Figure 5-20: Sampling at the groundwater plot................................100
Figure 5-21: E. coli found on crop at mid- (farmer harvest) and end-point (both
sanitized and farmer) to test for recontamination during farmer harvest......100
Figure 5-22: E. coli averaged over one growing cycle for all media: irrigation water,
soil, and crop harvested in two different ways...............................101
Figure 5-23: Probability of contracting a disease from ingesting lOOg of raw spinach
per week for one year........................................................104
Figure 5-24: Tradeoffs between groundwater use, energy and GHG, 1/food produced,
and health risk in this case study. Units were normalized as noted above.....106
Figure 5-25: Sustainability pentagon showing tradeoffs between energy use/GHG
emissions, health risk, infrastructure capital investment, groundwater use, and the
inverse of food produced. All sites are compared to one another as a percentage. ..106
Figure 6-1: Farmer, Chandriah, and family helping with harvest at the groundwater
plot.........................................................................110
Xll


LIST OF TABLES
Table 1-1: Examples of cities, crops, and pathogens associated in the practice of
wastewater irrigation worldwide.............................................1
Table 2-1: Information on metrics and data used for this urban energy metabolism
study......................................................................14
Table 2-2: Municipal water supply profile for 6 Indian cities (Source: Ministry of
Urban Development 2010)....................................................19
Table 2-3: Wastewater profile for 6 Indian cities (Source: Ministry of Urban
Development 2010)..........................................................19
Table 2-4: Energy use for municipal water supply for 13 Indian cities......22
Table 2-5: Energy use for wastewater treatment and pumping for 13 Indian cities... 24
Table 2-6: Energy use for total water infrastructure (water supply+wastewater) for 13
Indian cities..............................................................24
Table 2-7: Total energy use in water and wastewater utilities in the US and India.... 25
Table 3-1: Average Musi River quality (2009-10) compared with India Central
Pollution Control Board water quality criteria for drinking water sources before
treatment..................................................................30
Table 3-2: Capacity and actual treatment volumes of WWTPs in Hyderabad, India. 33
Table 4-1: Denitrification and nitrification pathway description and formation of N2O
................................................................................38
Table 4-2: Typical composition of high strength untreated domestic wastewater
(Metcalf and Eddy 2003) compared to influent wastewater to NWWTP reported from
March 2009- March 2010 (Kumar 2010).............................................42
Table 4-3: NWWTP measured parameters and treatment efficiencies for March 2009-
March 2010 (Kumar 2010)..........................................................45
Table 4-4: N2O emissions from WWTP processes: a comparison of results using
IPCC methodology to findings by Ahn et al 2010..................................51
Table 4-5: Example of DAYCENT outputs using agricultural plots for one growing
cycle only.......................................................................61
Table 4-6: On-site energy use and GHG emissions for NWWTP.......................63
Table 5-1: Tests done over one crop growing cycle. Gray: tests outsourced to a lab;
white: tests done by this researcher. All sampling, transport, and sample preparation
was also done by this researcher.................................................74
Table 5-2: Summary of dose-response slope parameter for various enteric pathogen
ingestion studies. Source: Adapted by Asano et al. 2007 from Regli et al. 1991, Ward
et al. 1986 and Black et al. 1988................................................89
Table 5-3: Volumes of irrigation water resulting from flood irrigation of the different
plots............................................................................92
Table 5-4: Total loading of nutrients and pathogens over one growing cycle (this
study)
Xlll
94


Table 5-
Table 5-
Table 5-
5: Soil water nutrient content from each plot.............
6: Crop nutrient content of leaf and root.................
7: Statistical tests of E.coli on crop at time of harvest
.97
.98
105
xiv


1.
Introduction
As city populations grow, their urban metabolism (resource consumption,
energy use, and waste generation) also increases (Wolman 1965; Kennedy,
Cuddihy et al. 2007). Often in developing nations, cities displace surrounding
agricultural land and fresh irrigation water, forcing agriculture downstream of
urban riverine/wastewater discharges. This nutrient-rich resource is valuable
to farmers who are seeking a widely-available and consistent source of
irrigation water for their crops. This practice is not new or rare; in fact, it
stems from ancient Greece and today, an estimated 200 million farmers
irrigate at least 20 million hectares with raw or partially treated wastewater
(Raschid-Sally and Jayakody 2008). This accounts for approximately 8% of
total worldwide irrigated land (263 million hectares in 1996), of which two-
thirds lies in Asia (Howell 2001). This amount of farmers represents
approximately 15% of the total amount of people economically active in
agriculture worldwide (FAOSTAT 2009). Some cities where the practice
actively takes place today, along with some crops grown and some pathogens
associated with wastewater irrigation are listed in Table 1-1 (van der Hoek
2004).
Table 1-1: Examples of cities, crops, and pathogens associated in the practice
of wastewater irrigation worldwide
Cities Crops Grown Pathogens
Mexico City, Mexico Spinach Roundworm (Ascaris)
Lima, Peru Lettuce Hookworm
Hyderabad, India Parsley E. coli
Ho Chi Minh, Vietnam Cilantro Giardia (Giardia lamblia)
Teheran, Iran Tomatoes Hepatitis A Virus
Nairobi, Kenya Potatoes Typhoid (Salmonella typhi)
Kumasi, Ghana Berries Cholera (Vibrio cholerae)
1


Because this practice is widespread and legislation is difficult, the question is
no longer if wastewater should be used for irrigation, but how it can be made
more sustainable and safe (Scott, Faruqui et al. 2004; Van Rooijen, Turral et
al. 2005). The next two sections briefly describe the advantages and
disadvantages of wastewater reuse for urban agriculture.
1.1 Advantages of Wastewater Reuse for Urban Agriculture
Wastewater reuse for urban agriculture is often considered to provide many
benefits. Qualitatively, these advantages are reported to be:
Conservation of water: Water reused for urban agriculture means that less
freshwater is needed, which is important as water scarcity is increasing (van
der Hoek, Hassan et al. 2002).
Nutrient recycling: Wastewater contains nutrients, leading many farmers
to prefer wastewater for irrigation because it is thought to increase
productivity (Qadir, Wichelns et al. 2007).
Figure 1-1: Urban agriculture in Hyderabad, India
2


Avoided fertilizer (Asano 1998): Nutrients in wastewater could save the
farmers money and could have the indirect impact of saving energy and
greenhouse gases (GHG) (Pitterle and Ramaswami 2009).
Land treatment of wastewater: Without other treatment options, land
application may provide some decrease in surface freshwater contamination
(Raschid-Sally and Jayakody 2008).
Spatial and temporal accessibility of irrigation water: Oftentimes, farmers
have better access to wastewater as a source of irrigation water because it is in
constant supply in urban and
peri-urban areas, even in the
dry season. This is because
cities are drawing municipal
drinking water from outside
their boundaries and it is being
discharged as wastewater after
use (Qadir, Wichelns et al.
2007).
Decreased need for
expensive refrigerated transport
or storage facilities: This is
most valued in developing
countries with hot climates
(Qadir, Wichelns et al. 2008).
Nutrition: Urban agriculture
(which is facilitated by
wastewater reuse in many
developing world cities) provides both farmers and consumers with a local,
fresh supply of vegetables (Qadir, Wichelns et al. 2008).
Figure 1-2: A local farmer heading to market
with newly harvested spinach grown with
untreated surface water
3


Better livelihoods: Wastewater is an inexpensive source of water and
nutrients allows farming families to grow high-value and high-demand crops
like vegetables (Kilelu 2004), which generates more income and raises living
standards, therefore allowing for indirect benefits like education (Raschid-
Sally and Jayakody 2008).
For these reasons, wastewater is considered a valuable resource for many.
The articles/reports above are largely qualitative studies. Many of these
benefits, along with savings in energy, greenhouse gas emissions, and water,
have not been measured.
1.2 Disadvantages of Wastewater Reuse for Urban Agriculture
On the other hand, while there
are many advantages, the
practice of wastewater reuse in
urban agriculture poses public
health and environmental
problems as water, soil, and
crops become increasingly
contaminated. Wastewater
contains a variety of pollutants
such as: salts, metals,
metalloids, pathogens, residual
drugs, organic compounds,
endocrine disruptor
compounds, and active
residues of personal care
products (Qadir, Wichelns et
al. 2007). Farmers in developing countries often use water from a polluted
stream, diluted wastewater, or untreated sewage directly on crops.
Wastewater from any source is seldom treated before being applied to crops
Figure 1-3: Flood irrigation with untreated
surface water in Hyderabad, India
4


(Qadir, Wichelns et al. 2007). Pollution in wastewater frequently affects soil
quality and/or causes acute or chronic diseases. Pathogens, specifically
intestinal nematode infections, have been identified as the main threat to
human health in the short term (Ensink, Blumenthal et al. 2008).
1.3 Wastewater Treatment Plants (WWTPs) for Sustainable Water
Reuse for Urban Agriculture
WWTPs are effective in removing pathogens and other harmful substances
from water and have been shown to decrease health risk (Asano 1998). Cities
in the developing world are implementing WWTP infrastructure to address
this need for treatment of sewage-polluted water. But WWTPs are themselves
energy intensive. A joint study by the American Water Works Association
Research Foundation (AWWARF) and the California Energy Commission,
compiled data on wastewater utilities in the US (Association of Metropolitan
Sewerage Agencies, now the National Association of Clean Water Agencies
or NACWA and a study in Iowa) and these ranged from 0.8-3.5 Wh/gal in
energy use (Carlson and Walburger 2007). Conventional wastewater treatment
alone is estimated to consume 3-5% of U.S. electricity (Shizas and Bagley
2004; US EPA 2006). For a typical U.S. municipal energy budget, wastewater
treatment is one of the largest at 23% (Means 2003). These energy
investments are expected to offer various benefits in terms of pathogen
reduction and they may help in more sustainable wastewater reuse for
agriculture. Also, overall greenhouse gas (GHG) emissions reductions may be
achieved due to WWTP processes and subsequent application of effluent to
farmlands.
No literature has been published that explores the fate of GHG emissions
when wastewater is reused in agriculture. Some studies have computed GHG
benefits of avoided fertilizer when nutrients are reused in agriculture (Pitterle
2008). However, there is high uncertainty associated with direct nitrous oxide
(N2O) emissions from wastewater (IPCC 2006; Del Grosso, Ojima et al.
5


2009). The DAYCENT model, which has been verified to yield reliable
results in N2O emissions from cropped fields (Del Grosso, Mosier et al. 2005;
Jarecki, Parkin et al. 2007), has not been used to estimate emissions from
wastewater application to land. Thus, the energy investments in WWTP-
related GHG emissions and associated pathogen risk reduction for urban
agriculture are unknown.
1.3.1 Sustainability assessment of wastewater treatment plants with
water reuse in urban agriculture
Based on the above review, wastewater infrastructure can have multiple and
conflicting sustainability impacts: economic benefits to farmer (food
production), health benefits to society (pathogen risk reduction in food), GHG
emissions (not known whether it will increase or decrease), water reuse (water
savings) and monetary cost (increases with more infrastructure). This
sustainability quadrant (Figure 1-4) is one way of weighing tradeoffs and has
been used by other authors (Pearce and Vanegas 2002; Aubin, Papatryphon et
al. 2009). This paper will evaluate these tradeoffs for the three sites used in
this study, differing sources of irrigation water: groundwater, treated effluent
from a WWTP, and untreated surface water representive of the sewage-
contaminated riverine system.
The five comers represent quantifiable environmental and social parameters
pertaining to sustainability:
Irrigation water saved will be measured by the amount of water reused for
irrigation and therefore, the equivalent ffeshwater/groundwater saved.
Food produced is a parameter that is closely linked with nutrient delivery
as well as with farmer livelihoods. It is shown as the inverse as higher
impacts are shown as worse.
Pathogen risk reduction was determined from lab tests on pathogen
content of vegetables.
6


Greenhouse gas reduction is based on the full life cycle GHG emissions
for delivering the irrigation water to each site.
Monetary investment is based on infrastructure investments, amortized
over their lifetimes.
All plots will be
measured against one
another for each
parameter as a
percentage.
Although WWTP
infrastructure may
offer many benefits,
its overall role in
urban metabolism has
not been measured.
Therefore,
Energy Use/GHG
Emissions
100% V
80%

Figure 1-4: Sustainability pentagon for evaluating
the tradeoffs of the coupled water-wastewater-urban
agriculture system
quantification of sustainability tradeoffs to the coupled water-wastewater-
urban agriculture system is important at this time when WWTP infrastructure
is being built.
1.3.2 Case study selection
Because many location-specific factors affect these tradeoffs, a case study
approach was necessary. Hyderabad, India as chosen for the following
reasons: centralized WWTP infrastructure is newly implemented (secondary
treatment within the last 5 years), wastewater contamination of surface water
is ubiquitous, and wastewater-polluted water is reused for urban agriculture.
From the case study, data was obtainable on livelihoods (food production),
health risks (pathogen content), wastewater reuse (water use/savings),
7


wastewater treatment (energy use and GHG emissions), and wastewater
infrastructure (cost).
1.4 Research Goals and Objectives
This study aims to quantify multiple and competing sustainability impacts of
implementing WWTP infrastructure in developing world cities with
subsequent reuse of wastewater in urban agriculture. The five sustainability
metrics are: water use/savings, nutrient supply/food production, pathogen
content, energy use/greenhouse gas emissions, and cost of infrastructure.
1.4.1 Phases of Work
This work is implemented in four steps:
1) Role of Water and Wastewater in Urban Energy Metabolism (Ch. 2): a field
analysis for India on water supply and wastewater treatment infrastructure-
related urban metabolism (energy use and energy-related GHG emissions);
2) Understanding Case Study Area and Infrastructure Components (Ch. 3):
description of Hyderabad, India, and its water supply, wastewater treatment,
and urban agriculture infrastructures;
3) Wastewater Treatment Plant Life Cycle Assessment (Ch. 4): a WWTP life
cycle assessment (LCA) measuring energy and GHG investments in WWTP
with and without subsequent water reuse for urban agriculture;
4) Measuring Water Quality and Food Quality Relationships: A Site Study
(Ch. 5): a farming urban agriculture study near the Musi River to determine
water use, nutrient delivery, and pathogen risk at three sites: one site irrigated
with freshwater/ groundwater, one irrigated with effluent from the WWTP,
and one irrigated with untreated surface water;
8


2. The Role of Water and Wastewater Infrastructures in Urban Energy
Metabolism
2.1 Introduction
Throughout history and in all parts of the world, it has been well-documented
that sewage-contaminated water poses problems to human and environmental
health (World Health Organization 2006; Laine, Huovinen et al. 2010).
Today, as access to basic sanitation is lacking in many developing countries,
practices, such as open defecation and uncontrolled release of sewage
(blackwater) to rivers, are polluting riverine systems in major cities of the
world (Van Rooijen, Turral et al. 2005; Qadir, Wichelns et al. 2008; Raschid-
Sally and Jayakody 2008).
Wastewater treatment plants (WWTP) or sewage treatment plants (STP) are
useful in removing pathogens and other pollutants from water. Many rapidly
growing cities are therefore installing WWTPs, providing primary, secondary,
and sometimes even tertiary treatment, for their communities. WWTPs have
an important role in a citys metabolism of water, nutrients, and energy.
2.1.1 Urban Metabolism: Water, Nutrients, and Energy
Urban metabolism studies often focus on a flow across city boundaries of:
water, substances (e.g. nutrients, chemicals, food, construction materials, etc.),
or energy. Some studies, like Kennedy et al., considers on all of these flows
and observes that city metabolism is generally increasing worldwide
(Kennedy, Cuddihy et al. 2007).
Some authors have focused on water foot-printing in order to track the inflow
and outflow of water in urban regions, incorporating water use for
consumption and production of goods and services (Allan 1998; Luck,
9


Jenerette et al. 2001; Jenerette, Wu et al. 2006; Yu, Hubacek et al. 2010;
Water Footprint Network 2011). Water footprint methods are being refined to
incorporate embodied water in energy flows into cities.
Substance flow analysis can be done for macronutrients like nitrogen (N) and
phosphorus (P). A nitrogen balance was done by Lawrence Baker et al. for the
central Arizona-Phoenix (USA) ecosystem. They identified natural inputs of
fixed N, deliberate human-mediated N inputs (mainly agricultural-related,
including fertilizers), and inadvertent human-mediated N inputs (combustion-
derived NOx) (Baker, Hope et al. 2001). Nitrogen was followed through
transfers within and among subsystems, and accumulation within the
ecosystem. Then, the outputs of nitrogen as deliberate exports (in food and
wastewater) and inadvertent exports (NOx, N2O, NH3, N2, and surface water)
were tracked. The overwhelming majority, 88%, of total nitrogen inputs to the
ecosystem were found to be human-mediated: deliberate inputs accounted for
52%, while inadvertent inputs made up the additional 36% (Baker, Hope et al.
2001). The largest deliberate effort in removing nitrogen was done by
WWTPs, which removed 10% of the input nitrogen.
A few groups are studying urban energy metabolism (Wolman 1965; Huang
1998; Hillman and Ramaswami 2009; Zhang, Zhang et al. 2010). Many
sectors consume energy and release GHG emissions, and energy flows are
multiplied by emission factor to determine the GHG impacts. Some sectors
generate a portion of energy themselves, for example, in the production of
biogas from WWTPs. Methods for studying energy flows are also being
refined and are including embodied energy and transboundary energy use.
Accounting for transboundary items is important to the entire life cycle and
scopes are used to categorize in-boundary versus out-of-boundary items.
Many groups use scopes to inventory transboundary items (Hillman and
Ramaswami 2009; ICLEI 2009; Kennedy, Steinberger et al. 2009).
10


As discussed previously, wastewater treatment alone is estimated to consume
3-5% of U.S. electricity (Shizas and Bagley 2004; US EPA 2006), but this
proportion may be even higher in developing countries where electricity is
less common. Due to a lack of energy data, the role of water and WWTP
infrastructure in the energy metabolism of developing world cities is not well-
known. To achieve a holistic view, urban metabolism studies are needed to
quantify the proportion of energy use in urban systems that goes towards
water and wastewater infrastructure.
2.2 Data sets used for this study
In order to understand the role of water and wastewater infrastructure on
energy use in Indian cities, primary data was obtained from two different
sources.
2.2.1 Basic Infrastructure Data
The Indian Ministry of Urban Development (MoUD), Government of India
(Gol) completed a study for 2008-9 on service level benchmarking for water
supply, wastewater treatment, storm water drainage and waste management.
The purpose of the study was to understand what improvements have been
made due to Gols financial assistance towards infrastructure improvements
for delivery of municipal services to city residents (Ministry of Urban
Development 2010). 28 cities from 14 states participated in this study.
The MoUD study includes the following indicators for water supply: number
of connections (both residential and non-residential), volume water produced,
source of water supply (% from groundwater and surface water), volume
water consumed (both residential and non-residential), water treatment
capacity of drinking water treatment plants, volume of treated water storage,
distribution pipe length, average pressure, and number of water samples
passing the standards tests (Ministry of Urban Development 2010). The water
supply data does not include: the type of treatment, the energy used for
11


pumping and treating the water, or the specific sources of water (or distances
it travels/is pumped).
For wastewater treatment, the following indicators are included in the MoUD
study: the number of properties with access to toilets and the number that are
connected to sewers (as only a portion of toilets are connected to the sewer
system), the area covered by the sewerage network, the number of WWTPs,
the volume of sewage treated, the volume of treated water reused, and the
number of effluent samples passing the disposal standards tests (Ministry of
Urban Development 2010). The wastewater data in this study lacks: the type
of treatment, and the energy used for pumping and treating the wastewater.
The report does not specify the proportion of the citys energy or the
proportion of the citys GHG emissions that are caused by the water supply
and wastewater treatment sectors.
2.2.2 Energy use in water systems and in cities overall
For energy data, a report from ICLEI (Local Governments for Sustainability)
was used. ICLEI gathered data on city-wide energy use and GHG emissions
within the geographic boundary for 54 South Asian cities, 41 of which are in
India (ICLEI- South Asia 2009). This data set used the World Resources
Institutes (WRI) scopes 1, 2, and 3 to inventory greenhouse gas emissions.
Briefly, scopes 1 and 2 refer to direct GHG emissions and indirect GHG
emissions primarily from electricity, respectively, and encompass the
traditional accounting method. Scope 3 focuses on indirect GHG emissions
and includes critical urban materials that are needed by the city but are
produced outside of the boundary.
This data set included scope 1 community-wide energy consumption and
related emissions in the residential, commercial, industrial, transportation,
waste, and other sectors including the following fuels: liquefied petroleum
gas, fuel wood, kerosene, diesel, petrol, light diesel oil, compressed natural
12


gas, auto gas, municipal solid waste, and coal. Also, scope 2 community-wide
electricity consumption and related emissions were quantified for residential,
commercial, and industrial sectors, as well as for the municipal services of
building and facilities, street lighting, water supply and wastewater treatment
plant, and others. Data on all of these were not available for every city. More
specific data was also provided by ICLE1 South Asia and was used to separate
energy use for water supply from wastewater treatment.
2.2.3 Intersection of two data sets
To link the water infrastructure with energy infrastructure, data from both
reports was needed. Of the 41 cities from the ICLEI report, and the 26 cities
from the MoUD report, 11 cities were common. Then, ICLEI had separate
water supply for only 6 of these cities and separate wastewater data for only 4
cities. Data from the ICLEI report was obtained for 2007-08 while data from
the MUD report was for 2008-09, making them one year different.
Additionally, data from Hyderabad and Delhi was obtained during visits to
those two cities. For Hyderabad, data from the Hyderabad Municipal Water
Supply and Sewerage Board (HMWSSB) website was used to estimate the
energy consumption for pumping of municipal water supply in Hyderabad
(HMWSSB 2008). Wastewater data for March 2009- March 2010 was
provided by Mr. M. L. Prasanna Kumar at the HMWSSB on one of its
WWTPs (Nallacheruvu). This data was scaled up to estimate the total energy
consumption of Hyderabads WWTPs (Kumar 2010). For Delhi, Mr. Ramesh
Negi and Mr. Ajay Gupta of the Delhi Jal Board provided data for 2008-09
and 2009-2010 on energy consumption for treatment and distribution
(pumping) of the municipal water supply as well as for treatment and pumping
of wastewater (Negi and Gupta 2011). In total, un-separated energy data for
water infrastructure (water supply and wastewater) was obtained for a total of
13 Indian cities. Separated data will be discussed later.
13


2.3 Metrics
The three overall metrics used to describe the role of water supply and
wastewater infrastructure in urban energy metabolism are found in table 2-1.
Data needed to calculate/inform these metrics and the data set compiled are
also included.
Table 2-1: Information on metrics and data used for this urban energy
metabolism study_________________________________________________
Metric Data Needed to Estimate Sub-data Needed Data set
1 Carbon emissions for water infrastructure as a proportion of overall community emissions Energy-related GHG emissions for water supply and wastewater infrastructure 39 Indian Cities
Overall energy-related GHG emissions for community/city
2 Energy use for municipal water supply pumping and treating per gallon Energy data on pumping and treating reported separately Percentage of municipal supply that is groundwater 8 Indian Cities
per capita Volume of municipal water provided per capita How many people are without connections
3 Energy use for wastewater pumping and treating per gallon Energy data on pumping and treating reported separately 6 Indian Cities
per capita Volume of wastewater collected per capita How many people with access to toilets and connected to sewerage system
per BOD load removed Biochemical oxygen demand removal efficiencies t *
The methods for these metrics were:
Metric #1. Carbon emissions for energy use in water infrastructure as a
proportion of overall community emissions: ICLEI-South Asia (ICLEI-SA)
collected data (2007-08) from the engineering and administrative departments
of the participating Urban Local Bodies (ULBs) on community-wide scope 1
and 2 energy use (ICLEI- South Asia 2009). Then, ICLEI-SA used this energy
data in the Harmonized Emissions Analysis Tool (HEAT) to calculate the
equivalent carbon emissions for each sector. ICLEI calculated the percentage
of community-wide emissions that were from corporation or municipal
14


services provided to the city, of which water infrastructures were a subset
(ICLEI- South Asia 2009). Electricity, the primary form of energy used for
treatment and pumping of drinking water and wastewater, was the only form
used to estimate emissions from water infrastructures.
Metric #2. Energy use for municipal water supply treatment and pumping per
gallon and per capita: To determine the energy use per gallon of water,
volumes of municipal water provided to city residents was needed and was
found in the MoUD report for 2008-09 (Ministry of Urban Development
2010). Electricity data for municipal water supply combing treatment and
pumping was provided by ICLEI-SA and 11 of their cities overlapped with
data needed in the MoUD report. Electricity data for Delhi was provided by
the Delhi Jal Board and was separated for treatment and pumping. For
Hyderabad, only water pumping data was available to this researcher (figure
2-5 and table 2-4). Of these 13 Indian cities that data could be gathered for, 8
cities provided separate electricity data for municipal water supply combining
treatment and pumping (ICLEI- South Asia 2009). The other relevant sub-data
was the proportion of total number of city residences to those that have water
tap connections, also given in the MoUD report. The total populations (2008-
09) for the ULBs were used to calculate the per capita energy use for
municipal water supply (Ministry of Urban Development 2010).
Metric #3. Energy use for wastewater treatment and pumping per gallon, per
capita, and per biochemical oxygen demand (BOD) load removed: To
determine the energy use per gallon of wastewater, the volume of wastewater
collected was obtained from the MoUD report for 2008-09 (Ministry of Urban
Development 2010). Electricity data for wastewater combing treatment and
pumping was provided by ICLEI-SA and 11 of their cities overlapped with
data needed in the MoUD report. Electricity data for Delhi was provided by
the Delhi Jal Board and was separated for treatment and pumping. For
15


Hyderabad, only wastewater treatment data was available to this researcher
(figure 2-6 and table 2-5). Of these 13 Indian cities that data could be gathered
for, 6 cities provided separate electricity data for municipal water supply
combining treatment and pumping (ICLEI- South Asia 2009). To inform
energy use per capita for wastewater treatment and pumping, the proportion of
properties with connections to the sewer was important (not all toilets are
connected to sewers). This data was also found in the MoUD report. The
amount of BOD removed could be calculated for both Delhi and Hyderabad as
BOD measurements were provided by those two cities.
2.4 Results for Indian cities by three metrics
2.4.1 Carbon emissions for water infrastructure as a proportion of
overall in-boundary community emissions
The electricity-related emissions from municipal water supply and wastewater
infrastructure as a
£ u
n
s
18%
16%
14%
12%
10%
8%
6%
4%
2%
0%
1.1
tl
,1.
I I
111 I
III
percentage of the
total carbon
emissions from 39
communities (or
ULBs) are shown
in figure 2-1. For
the majority of
^ ^ ^ these communities,
4^// *
s-
V Figure 2-1: Percentage of community-wide GHG
emissions associated with in-boundary energy use that
are attributed to water and wastewater infrastructure
the proportion of
the total electricity-
related emissions
arising from water infrastructure was less than 6%. However, a few cities had
much higher proportional emissions: Lucknow (16%), Shimla (12%), and
16


Trivandrum (13%) (1CLEI- South Asia 2009). Trends that may explain these
results were explored in figure 2-2 A-E.
The percentage of emissions from water and wastewater infrastructure were
plotted against: city population in 2001 (figure 2-2A) (ICLEI- South Asia
2009), number of households in the city (figure 2-2B) (Ministry of Home
Affairs 2001), average household income (figure 2-2C) (TRENDSnIFF 2008),
percentage of households with electricity (state percentages) (figure 2-2D)
(Ministry of Health and Family Welfare 2007), human development index
18% 18%
f
3 16% - | 16% *
E 1
£ 14% £ 14% -
i
6 12% * | 12%
1 10% 1 10% - -
5 5
6 8% - - - fc 8%
5 1
5 6% * # 5 6% -
3 * - *
? 4% g 4% :
** ; *
I 2* ******* * S 2% jj - *-; - * i *** * *


0 1 2 3 4 S 6 0 250,000 500,000 750,000 1,000,000 1,250,000 1,500,000 1,750,000 2,000,000
Population (Millions in 2001) Number of Households in the City (2001)
18%
1
a 16%
E
£ 14% :
i 12%
3
Average Annual Household Income USD (2004'5)
- !8%
I
S 16%
E
£ 14% -
i
112% '
I 10%
*
a 8%
20 40 60 80 100
% Households with Electricity (State)
120
Figure 2-2: Percentage of emissions
from water and wastewater
infrastructure versus A) population of
Indian cities; B) number of households
in Indian cities; C) average household
income in Indian cities ($USD); D)
percentage of households with
electricity (Indian state average); or E)
human development index (Indian
state average).
17


(state data from 1981) (figure 2-2E) (Government of Meghalaya Shillong
2008).
No significant trends emerged. More reliable household-level data may offer
more answers, but currently, the proportion of energy-related GHG emissions
for water supply and wastewater infrastructure in cities can not be well-
explained by the proxies used here to represent urbanization.
Although these carbon emissions are large, they did not include WWTP
process emissions. As described later, process emissions can also be large.
Therefore, emissions from WWTP could be doubled.
2.4.2 Electricity use in water and wastewater infrastructure
Electricity use and energy-related GHG emissions for the combined water and
wastewater infrastructures can be affected by: the amount of water pumped
from various sources, coverage of water supply and wastewater infrastructure,
percent of homes served by taps and sewers, and the amount of wastewater
collected for treatment. Figure 2-3 describes the water and wastewater
infrastructure for Hyderabad, India, while tables 2-2 and 2-3 describe the 13
cities for which the same infrastructure data was available (population and
city area for these cities are shown in table 2-4). The Ministry of Urban
Development report provides more information on data quality ratings and
URBAN AREA
Treated and
pumped to
city
properties
397 MGD
Properties
with access
to toilets:
58%
Properties
connected
to sewer:
27%
Collected and treated by
WWTPs, then released
115 MGD

Released directly to
the environment
175 MGD
Figure 2-3: Water and wastewater flows through Hyderabad
18


more information on water and wastewater infrastructure.
Figure 2-3 shows that far more water is supplied to a city than is treated in
WWTPs. This explains why, on an aggregate basis, the majority of energy is
used for water supply treatment and pumping.
Table 2-2: Municipal water supply profile for 6 Indian cities (Source: Ministry
of Urban Development 2010),_____________________________________
Community 2008-09 Surface Water Ground- water Residential Water Connections Non-residential Water Connections Volume Water Produced Water Supply Consumed
% of Municipal Water Supply %of Residences % of Non- residences Million US Gallons/ day
Ahmedabad 88% 12% 45% 9% 244 171
Delhi 88% 12% 47% 16% 971 809
Guntur 100% 0% 48% 15% 20 9
Hyderabad 100% 0% 42% 12% 397 289
Shimla 97% 3% 32% 117% 9 7
Tiruchirapalli 100% 0% 45% 11% 24 16
Table 2-3: Wastewater profile for 6 Indian cities (Source: Ministry of Urban
Development 2010). ____________________________________________
Community 2008-09 Properties with Access to Toilets Properties Connected to Sewer Sewerage Network Coverage Volume Wastewater Produced Volume Wastewater Collected/ Treated Wastewater Collected/ Treated
% of Total Properties %of Community Area Million US Gallons/ day %of Wastewater Produced
Ahmedabad 78% 60% 74% 137 89 65%
Delhi 59% 41% 48% 743 467 63%
Guntur 72% 12% 25% 7 nil nil
Hyderabad 58% 27% 48% 290 115 40%
Shimla 97% 74% 79% 5 1 16%
Tiruchirapalli 91% 23% 25% 22 15 67%
The water flows and infrastructure profiles (figure 2-3 and tables 2-2 and 2-3)
can help to explain the differences in energy consumption for water supply
and wastewater in cities. Figure 2-4 displays cities with separated data and
compares the total electricity used for water supply to that used for wastewater
(Note: Data for Hyderabad includes energy use for water supply pumping and
19


wastewater treatment only, while data for Guntur and Tiruchirapalli includes
energy use for wastewater pumping only).
In the case of
the two cities
using the
lowest
relative
amount of
energy for
wastewater
treatment,
Guntur and
Shimla, this
result is
expected per
the infrastructure profiles. For Guntur, their WWTP is commissioned but not
yet running (as of 2008-9), and therefore, no wastewater was being collected
or treated. For Shimla, only 16% of the total produced wastewater was
collected and treated, which is surprising due to their higher amount of
properties connected to sewers and larger percentage of community area
covered by the sewerage network coverage.
2.4.3 Energy use for municipal water supply treatment and pumping
per gallon and per capita
This section focuses on the energy used for both treatment and pumping of
municipal water supply that is distributed to the city residents. Often, data was
not available for both treatment and pumping, or it was aggregated together as
a total. Figure 2-5 represents the electricity consumption per gallon of
municipal water supply. When the outlier (Shimla) was removed from Figure
Figure 2-4: Percentage of electricity used for water supply
(WS) infrastructure in comparison to wastewater (WW)
infrastructure.
20


2-5, average total energy use for water supply treatment and pumping was
1.260.68 Wh/gal (n=7).
The cities for
which there
was not
separate data
available for
water supply
treatment and
pumping are
shown in
blue. The two
cities for
which there
was separate data were Delhi and Shimla. Because Shimla is a hill station in
the Himalayan Mountains, there is a high amount of energy consumed to
pump water uphill to homes. Also, water is sourced from long distances
(Ranjan Kumar Guru 2011). In the case of Delhi, treatment energy dominates
the total electricity use per gallon for municipal water supply. This could be
explained by the large population living upstream from Delhi that is suspected
to be the contributing to the pollution in the surface streams. Unfortunately,
the types of treatment at the municipal water plants are not included in the
ICLEI or MoUD reports.
Again, Shimla stands out with the highest energy user per capita for water
supply in table 2-4. Because Shimla is in a mountainous area and water must
be pumped up steep gradients, the energy use for supply per person is
expected to be high.
Treatment
> Pumping
Total:
Treatment
and
Pumping
Energy Not
Separated
Figure 2-5: Energy use per gallon of municipal water supply
(treatment and pumping separated where available)
21


Table 2-4: £nergy use: or municipal water supply for 13 Indian cities.
Community Popu-lation Million 2008-09 City Area Municipal Water Supply
Ground -water in Muni- cipal Water Supply Electricity Use Volume Water Prod- uced Electricity per Gallon
For Treat- ment For Pump- ing Total Per Capita Treat- ed Pump- ed Total
km2 % Million kWh kWh/ capita Million US Gallons / day Wh/gallon
Ahmedabad 5.6 466 12% 102 102 63 244 1.14
Bengaluru 7.8 793 0% 246
Bhopal 1.8 284 5% 78 78 42 79 2.68
Bhubaneswar 1.1 135 16% 71
Delhi 17.8 1,397 12% 274 23 297 17 971 0.77 0.06 0.84
Guntur 0.6 63 0% 6 6 11 20 0.90
Hyderabad 7.6 617 0% 1 91 91 12 397 0.63 0.63
Indore 2.0 214 12% 21 21 11 49 1.18
Nashik 1.6 259 0% 91 v-'. \
Raipur 1.0 154 15% 39 ' -
Shimla 0.2 20 3% 0.8 74 75 393 9 0.23 21.9 22.1
Tiruchirapalli 0.8 147 0% 13 13 16 24 1.48
Trivandrum 1.0 142 0% 59
2.4.4 Energy use for wastewater treatment and pumping per gallon, per
capita, and per BOD load removed
This section focuses on the energy use for both treatment and pumping of
wastewater. For some communities, data was not available for both treatment
and pumping, or it was aggregated together as a total. Figure 2-6 and Table 2-
5 represent the electricity data collected. When the outlier (Shimla) was
removed from Figure 2-6, average total energy use for wastewater treatment
and pumping was 0.410.18 Wh/gal (n=4).
The cities for which there was not separate data available are shown in blue.
The two cities from which there was separate data are Ahmedabad and Delhi.
Guntur and Tiruchirapalli only had wastewater pumping data, while
Hyderabad only had wastewater treatment data. Unfortunately, the types of
treatment at the wastewater treatment plants are not included in the ICLEI or
MoUD reports.
For Delhi, the energy use for wastewater is highest overall, but not per capita
or per gallon (Table 2-5). Per capita, Ahmedabad has the highest energy use
22


for
wastewater,
largely due
to a high
amount
used in
treatment.
This is to
be expected
because
Ahmedabad
collects
65% of its
produced wastewater (one of the highest in table 2-3). Per gallon wastewater
treated, Shimla has the highest overall, and may be explained by a high energy
requirement for pumping up hills, as with drinking water. Because the data is
not separated for Shimla, this cannot be further described at this time.
Last, energy use per milligram of BOD (biochemical oxygen demand) was
calculated for Delhi and Hyderabad. These energies were 0.7 and 0.4 Watt-
hours/gram (Wh/g) BOD removed, respectively (table 2-5). To benchmark
these studies, WWTPs in the US used between 1.5 to 9.8 Wh/g BOD removed
using various technologies such as trickling filters, attached growth, activated
sludge, and advanced wastewater treatment processes with and without
nitrification (Pitterle 2008). Therefore, energy use for removal of BOD was
much lower in Indian WWTPs when compared to US WWTPs.
= i
O '
G T3
Q. 0>
£
Treatment
Pumping
u 1 -
> .. Total:
u 0 mm I i Treatment
01 1 and
UJ ^ / / / / Pumping Energy Not
Separated
Figure 2-6: Energy use per gallon of wastewater (treatment and
pumping separated where available)
23


Table 2-5: Energy use for wastewater treatment and pumping for 13 Indian
cities.
Community Wastewater
Electricity Use Volume Treated Electricity Use per Gallon BOD load removal effi- ciency Energy use per BOD load removal
For Treat- ment For Pump- ing Total Per Capita Treated Pumped Total
Million kWh kWh/ capita Million US Gallons / day Wh/gallon mg/ gallon kWh/ mg BOD removed
Ahmedabad 16 i 17 10 89 0.48 0.03 0.51
Bengaluru ' V'.-.r 97
Bhopal / [r 7
Bhubaneswar 1 * -rK ...i-V.;
Delhi 32 23 55 3 467 0.19 0.13 0.32 438 0.73
Guntur ; 0.07 0.07 0.11 0 '
Hyderabad 9 9 1 115 0.21 0.21 522 0.40
Indore : 22 jk.
Nashik V T V . . :. ; 37 he T-X*. x-i j
Raipur A 0
Shimla 2 2 9 1 \ 5.31 '
Tiruchirapalli 3 3 4 15 0.60 0.60
Trivandrum 0
Total electricity use and per capita electricity use for both water and
wastewater infrastructures is shown in table 2-6.
Table 2-6: Energy use for total water infrastructure (water
supply+wastewater) for 13 Indian cities._____________
Total Water Infrastructure
Community Total Electricity Use Million kWh Per Capita Electricity Use kWh/ capita
Ahmedabad 123 22
Bengaluru 320 41
Bhopal 78 42
Bhubaneswar 10 10
Delhi 351 20
Guntur 7 11
Hyderabad 100 13
Indore 21 11
Nashik 42 26
Raipur 22 22
Shimla 76 402
Tiruchirapalli 16 20
Trivandrum 46 48
24


2.5 Comparison between India and US
A wide variation in energy use for drinking water supply and wastewater
utilities is seen in both India and the US. An AWWARF study compiled data
from other studies (AWWA Water:/Stats database and studies done in Iowa
and Wisconsin) and found that total energy use for drinking water utilities
ranges from 0.3-3.8Wh/gallon (Carlson and Walburger 2007). As mentioned
in Chapter 1, this group also compiled data on wastewater utilities in the US
and found that energy use for wastewater utilities ranged from 0.8 to 3.5 Wh
per US gallon wastewater treated (Carlson and Walburger 2007). However,
when energy use in water supply and wastewater treatment is compared city to
city, energy use for drinking water utilities is usually about half of that for
wastewater utilities. A research team at UCD has compiled a fairly good data
set on drinking water and wastewater treatment for cities in Colorado.
Durango and Westminister are shown in Table 2-7.
Table 2-7: Total energy use in water and wastewater utilities in the US and
India
Utility Water Supply Wastewater
City Wh/ gallon
OO Durango, CO 0.9 3.4
O Westminister, CO 0.6 1.0
India Ahmedabad 1.1 0.5
Delhi 0.8 0.3
A lower proportional energy use per gallon for drinking water treatment in the
Colorado cities is reflective of the source water quality. In India, it is thought
that more pumping is needed for chlorination at the drinking water plants to
disinfect a higher concentration of pollution in the water. Also, in India, there
has been a preference for wastewater treatment technologies that require low
energy, like the upflow anaerobic sludge blanket (UASB) (Tare and Nema
n.d.). In the US, a higher energy use in wastewater treatment could be
25


attributed to more energy consuming processes and higher levels of treatment,
when compared to India.
When total community emissions from a city in the US (Broomfield, CO) is
compared to a city in India (Bhopal), the larger proportion of energy use for
water and wastewater is evident for India.
Figure 2-7: Community-wide emissions for Broomfield, Colorado, USA and
Bhopal, India.
26


While there are many differences between these two cities, one notable
difference is that water and wastewater contributes a larger proportion of
community-wide emissions for Bhopal, India when compared to Broomfield,
Colorado, USA.
2.6 Insights and Recommendations for Future Work
From this chapter, we can conclude that energy use and related GHG
emissions are a large proportion of the community-wide total. For most
communities, more than 1% of the total energy-related emissions were from
water and wastewater infrastructures, and this proportion was as high as 16%.
With process emissions, this sector could double these emission to contribute
2-32% to community-wide GHGs. On a per capita basis, more energy is
invested in drinking water treatment than wastewater treatment overall in
India cities. In the next chapters, a case study approach will be used in which
a life cycle assessment of a WWTP with linkages to urban agriculture will be
discussed.
Future recommendations that would enhance this study are to gather more
information on: distances over which water is supplied (both horizontal and
vertical), types of treatment used, source water quality, and treated effluent
wastewater quality.
27


3. Understanding Case Study Area and Infrastructure Components
This chapter describes the setting of the study in Hyderabad, India, its riverine
system, its WWTPs, and its areas of urban agriculture.
3.1 City Description of Hyderabad, India
3.1.1 Population and Demographics
Hyderabad, located in southern India, is the capital city of the state of Andhra
Pradesh. The twin cities of Hyderabad-Secunderabad (herein referred to as
Hyderabad) sit roughly in the middle of the country at 526 meters above sea
level. Hyderabad is ranked as the 6th largest city in India (World Gazetteer
2010) and the 36th largest city in the world (City Mayors Statistics 2009). The
population in 2010 (calculated based on the 2001 census) was 4.1 million
(World Gazetteer 2010), while greater Hyderabad was estimated at 7.6 million
people (Jacobi 2009; Ministry
of Urban Development 2010).
The greater Hyderabad area is
expected to house 10.5
million residents by 2015
(Sustainable Hyderabad
2006). The urban population
of the state is 27.08%, similar
to that of India as a whole at
27.78% (Centre for Good
Governance 2008).
Consequently, 70% of the
population is scattered
throughout rural areas of the
state and make their living as farmers (Mercy Corps 2004). The per capita
'1ST'-.
Udapur
Kandla
*
Ahmad abad
Srinagar
Amritsar
NEW DELHI - * i
,* S-EPAli

Jaipur Agra Kanpur
,,f INDIA Varanasi
Bhopal
G*ogts
Calcutta*
ARABIAN .Mumba)
*ahaiush7RA Nagpur
Hyderabad'* Vlshakhapamam
BfriTAN
x.___.Imphal -
O'
-. A I
k
Panatt
Marmagao#
ft* 10t
Bangalore
Calicut*

Chennai
Pondicherry
Cochin (Kochi) Madurai
KfRAiA
Trivandrum.^
Kovalam ,8*) LANKA
Figure 3-1: Map of India
Andaman
Islands
Port Blair
Nicobar /stands
28


income is approximately $1.75/day and rising (+13.04% in 2006-2007) while
farmer income has seen negative growth (-6.56% in 2006-2007) (The Hindu
2007).
3.1.2 Climate
Hyderabad sits on the Deccan Plateau in the center of India. From May 1,
2009-April 30, 2010, Hyderabad received an average yearly rainfall of 1.6
mm/day (range 0-101.1 mm/day) (Rao 2010). Over the same time period,
Hyderabads average minimum daily temperature was 20.7C (range 10.0-
28.7C) and average maximum daily temperature was 34.1C (range 26.0-
43.4C) (Rao 2010). The monsoon season is approximately June-August.
3.2 Water Supply in Hyderabad
3.2.1 Musi River
Hyderabad sits close to the border of the Krishna and Godavari River basins.
The Musi River, which flows through the urban center of Hyderabad, is a
minor tributary of the Krishna River. West of the city are two dams, Himayat
Sagar and Osman Sagar, which were constructed in the early 1950s to regulate
the upper catchment of the Musi and provides the 1.1 million residents with
3.5 million cubic meters (MCM) per month of water (Van Rooijen, Turral et
al. 2005). Virtually all of this supplied water is consumed within Hyderabad,
so the river has little natural flow downstream of the city, but contains a large
amount of wastewater. About 150 km downstream of Hyderabad, the Musi
River sub-basin drains into the lower potion of the Krishna River (Van
Rooijen, Turral et al. 2005). Within the urban area of Hyderabad, the Musi has
poor water quality: it contains raw sewage, chemicals, oils, and other trash,
it has no plant or animal life, it has a bad odor, and contact with it is
harmful to human health (Devi, Samad et al. 2009). Hyderabad Municipal
Water Supply and Sewerage Board (HMWSSB) authorities measure Musi
water quality parameters, and data from 33 tests was provided for the time
29


period between July 2009- March 2010. In table 3-1, this data is compared
with the water quality criteria from the Central Pollution Control Board
(Ministry of Environment and Forests, Government of India) for the river to
be a drinking water source after treatment and disinfection (Central Pollution
Control Board 2008). The Musi is not meeting these standards and is not even
suitable for boating (Devi, Samad et al. 2009).
Table 3-1: Average Musi River quality (2009-10) compared with India
Central Pollution Control Board water quality criteria for drinking water
sources before treatment
Parameter/ Water Source pH DO lmg/L| BODj COD TSS VSS lmg/Ll [mg/L| (mg/LJ lmg/L] standard deviation (n=# tests) Coliforms* [MPN/ lOOmLj
Musi River 7.3 0.5 (n=33) 2.1 2.2 (n=33) 155 71 (n=33) 402 176 (n=33) 163 40 (n=16) 5322 (n=16) FC: 418,176 139,915 (n= 16)
Drinking water source used in conventional treatment and disinfection 6-9 >4 <3 N/A N/A N/A TC: < 5,000
*FC= Fecal Coliforms; TC= Total Coliforms; MPN= most probable number
Downstream of Hyderabad, the water quality of the Musi River improves
significantly due to natural processes that occur during settling, aeration,
microbial activity, etc., over long distances (Van Rooijen, Turral et al. 2005;
Ensink, Blumenthal et al. 2008).
3.2.2 Cross Watershed Transfers
In the 1960s, water from outside of the local catchment area was diverted
from the Godavari River Basin, specifically from its tributary, the Manjira
River (Celio, Scott et al. 2010). The Singur Dam was built in 1991 to regulate
Manjira River water for Hyderabad and it doubled the volume of available
water to 18 MCM per month (Van Rooijen, Turral et al. 2005). In 2003,
Hyderabad began receiving Krishna River water pumped over 135 km and
30


400 m in elevation from the Nagarjuna Sagar reservoir, which provided 10
MCM per month (Van Rooijen, Turral et al. 2005; Celio, Scott et al. 2010).
Future plans by the HMWSSB, who govern water supply and sewerage for
Hyderabad, include further extracts from the Krishna and Godavari Rivers.
3.2.3 Groundwater Withdrawals
Currently, private groundwater withdrawals is estimated to be about 10% of
the urban water supply (about 3.3 MCM per month) (Van Rooijen, Turral et
al. 2005). However, the municipal water supplied by the HMWSSB does not
include groundwater (Ministry of Urban Development 2010). Rapidly
growing competition from agriculture and the urban-industrial sector will
continue to put stress on the already scarce surface and groundwater resources
for Hyderabad and surrounding regions (Van Rooijen, Turral et al. 2009;
Celio, Scott et al. 2010; Venot, Bharati et al. 2010; Venot, Reddy et al. 2010).
3.3 Wastewater Infrastructure
For Hyderabad, 80% of the water supply used by people is released as sewage
(Ramachandraiah and Vedakumar 2007). According to a Ministry of Urban
Development report, 40% of the produced wastewater in Hyderabad is being
collected and treated before discharged into the Musi (Ministry of Urban
Development 2010). This leaves an average of 175 million gallons of
untreated wastewater entering the riverine system every day. For most of the
year (in the dry season), the Musi River would not flow without the input of
sewage water (Van Rooijen, Turral et al. 2005; Ramachandraiah and
Vedakumar 2007).
3.3.1 Wastewater Treatment Plant Technology
Within the last 5 years, Hyderabad has been implementing sewage treatment
plants in efforts to treat all of the water entering the Musi to secondary level
(Devi, Samad et al. 2009). India is unique in that it has favored UASB
technology more than any other country in the world (Khalil, Sinha et al.
31


2008). UASB reactors were selected in India for the following unique
characteristics when compared to other WWTP technology: low capital costs,
low energy requirements, low O&M costs, lower skills required for O&M,
lower sludge production, and potential for energy recovery and biosolids
generation (Khalil, Sinha et al. 2008; Tare and Nema n.d.). However, these
characteristics were originally determined for treatment of high strength
industrial effluents and may not be as attractive for domestic sewage
treatment, as UASB reactors only partially treat the wastewater and may make
more problems for the next steps in the WWTP (Tare and Nema n.d.). As
described by Heffeman et al, shortcomings in design, construction, and
operator maintenance greatly contribute to inferior UASB performance in
treating sewage (Heffeman, Lier et al. 2011). Additionally, discharges of
industrial effluents into sewage drains flowing into the WWTPs offer many
challenges due to the toxic materials and sulphate (contributing to immediate
oxygen demand) content of the wastewater (Tare and Nema n.d.).
Among other issues described by Tare and Nema, two notable characteristics
of UASB effluent are: BOD will not be lower than 70-100mg/L due to
limitations of the UASB reactor, and UASB effluent is highly anoxic and
exerts a high immediate oxygen demand on the receiving water body or land.
A second stage of aerobic treatment, like that found at Nallacheruvu (the case
study WWTP), can lower the BOD and COD by 50% and can increase the
dissolved oxygen by 50%, but costs in infrastructure and operations are
increased (Walia, Kumar et al. 2011; Tare and Nema n.d.).
Three completed WWTPs that use UASB reactors are located in the south-
east area of Hyderabad (Amberpet, Nagole, and Nallacheruvu, while Attapur
is not yet complete and is south-central) (see table 3-2). Nallacheruvu WWTP
is the case study site for this study.
32


Table 3-2: Capacity and actual treatment volumes of WWTPs in Hyderabad,
India
WWTP Annual Average Capacity MGD Annual Average Actual MGD Ratio
Amberpet 90 66* 0.57
Attapur 13 10* 0.09
Nagole 45 34* 0.29
Nallacheruvu 8 5 0.05
TOTAL 156 115
* Actual MGD for Ambertpet, Attapur, and Nagole was estimated based on
ratio of each capacity to the totals and data from Nallacheruvu WWTP
The ratios in table 3-2, determined from data provided by NWWTP
authorities, were used in Chapter 4 to scale up data to the city level when
totals for Hyderabad were not known.
The building of the Nallacheruvu WWTP in 2007 displaced urban farmers
that had been farming in the area for up to 40 years (McCartney, Scott et al.
Figure 3-2: Urban agriculture adjacent to Nallacheruvu WWTP
33


2008). Because these farmers used surface water to irrigate their crops, this
area has a long history of wastewater contamination in both soil and
groundwater. Today, adjacent to the Nallacheruvu WWTP, farmers grow
crops such as spinach, coriander, mint, chilies, papaya, amaranth, fenugreek,
fennel, and others.
3.4 Urban Agriculture
Downstream of Hyderabad, the Musi River is used extensively for irrigation,
with nearly 40,000 hectares of farmland being irrigated from the river
(Hamilton, Stagnitti et al. 2007). This has resulted in severe groundwater
pollution and an overall long-term decline in the productivity of wastewater-
irrigated lands by more than 50 percent (Devi, Samad et al. 2009). A few
researchers have studied wastewater reuse in Hyderabad and the effect on the
environment and the people (Gopal 2004; Sustainable Hyderabad 2006;
Srinivasan and Reddy 2009). The International Water Management Institute
has pioneered much of this work in Hyderabad and throughout the world
(Buechler and Devi 2002; Devi, Samad et al. 2009; Jacobi 2009). The
Resource Centres on Urban Agriculture and Food Security (RUAF) are also
active in Hyderabad and globally, with the primary aim to promote and
institutionalize urban agriculture processes in cities (RUAF 2010).
3.4.1 Soil Characteristics
Indian soils are generally grouped into two types, referred to as red and black.
Red soils, or alfisols (lixisols by FAO classification), are mineral soils with
low silt to clay ratio due to a history of strong weathering in wet tropical and
subtropical regions (Blokhuis, Bouma et al. 1991; Bhattacharyya, Chandran et
al. 2007). Today, they predominately occur in monsoonal and semi-arid
regions.
The soils present in this site study were black soils, or vertisols. These
mineral soils were conditioned by their parent material, expanding clay, and
34


most occur in semi-arid tropics. They are known to be finely textured and
have poor internal drainage, which makes them productive only when
managed (Blokhuis, Bouma et al. 1991).
3.5 Discussion
Hyderabad was an ideal location for this study because this researcher had
access to urban agriculture, researchers and laboratory facilities at the
International Water Management Institute (IWMI) and the International Crop
Research Institute for the Semi-Arid Tropics (ICRISAT), and newly
implemented WWTPs, making a study based on WWTPs and urban
agriculture possible. Next, the case study at a WWTP in Hyderabad will be
discussed.
Figure 3-3: Musi River in Hyderabad, India near the outfall of Amberpet
WWTP
35


4. Wastewater Treatment Plant Life Cycle Assessment: Nallacheruvu
WWTP in Hyderabad, India
4.1 Introduction
Many developing cities are currently installing centralized wastewater
treatment plant (WWTP) infrastructure. As discussed in chapter 2, WWTP
processes can be resource intensive in terms of energy use and energy-related
greenhouse gases (GHGs), as well as methane (CH4) and nitrous oxide (N2O).
Direct emissions of N2O and CH4, both potent greenhouse gases, can vary by
the processes used within the WWTP and subsequent emissions can vary by
whether the water is reused in agriculture or released to rivers. A life cycle
assessment (LCA) is needed to quantify the water quality improvements
achieved with various levels WWTP infrastructure investments that consume
energy and release GHGs.
Many life cycle assessments of WWTPs have been done in developed
countries, e.g. Australia (Foley, Haas et al. 2005), Canada (Sahely, MacLean
et al. 2006), France and Switzerland (Houillon and Jolliet 2005), Germany
(Remy and Ruhland 2006), Portugal (Machado, Urbano et al. 2006), Spain
(Vidal, Poch et al. 2002; Hospido, Moreira et al. 2004; Munoz, Peral et al.
2007), Sweden (Palme, Lundin et al. 2005), UK (Dixon, Simon et al. 2003),
and USA (Murray, Horvath et al. 2008; Pitterle 2008). The focus of these
studies range from process specific to whole plant energy and GHG LCA. In
general for these developed world LCAs, embodied energy and related GHG
emissions are low in comparison to the end-use energy and related GHG
emissions. The only LCA study from the above group to use actual WWTP
operating data to quantify end-use, embodied, and avoided energy impacts
was done by Pittlerle (2008). From Pitterles work, annual end-use energy for
36


a WWTP in Denver, CO, USA, was 2.3Wh/gal (1.3gC02e/gal), while
embodied energy was 1.7Wh/gal (0.4gCO2e/gal). A WWTP benchmarking
study in the US also reported end-use electricity at 1 to 5 Wh/gal (Pitterle
2008).
In contrast, for the developing world, there have been very few LCAs of
WWTPs. Murray et al. published a hybrid-LCA of sewage sludge treatment in
China, but did not include processes to treat wastewater before or after sludge
removal (Murray, Horvath et al. 2008). A study for a South-Asian WWTP
included energy use for construction activities, process operations, and
materials production, but did not quantify GHG emissions (Khan, Aijun et al.
2008) . For India, the only study on WWTP energy consumption used model
data to determine energy use for a WWTP serving 2,670 residents of a rural
Indian village, who generate 39,626 total US gallons wastewater per day.
They found that the WWTP would use 62.5kWh/day, or approximately 8.5
kWh/person/year or 0.6 Wh/gallon treated (Devi, Dahiya et al. 2007). When
compared to the wastewater infrastructure electricity use in chapter 2, which
ranges from 3 to 10 kWh/person/year, or 0.2 to 5 Wh/gal, the result from Devi
et al. is realistic. As WWTPs are being implemented in countries like India,
Brazil, and Colombia, UASB technology is being recommended because they
cost less and use less energy (Khalil, Sinha et al. 2008; Bdour, Hamdi et al.
2009) . Consequentially, other components, such as N2O and CH4 process
emissions, can become very important in life cycle impacts for developing
countries.
N2O emissions from wastewater and WWTP are of interest because of their
high global warming potential (298mtC02e/mtN20) and the high uncertainty
in their release. N2O is emitted naturally from wastewater due to microbial-
facilitated nitrification and denitrification. WWTPs commonly utilize
nitrification and denitrification processes to biologically remove nitrogen from
37


wastewater. Table 4-1 describes the pathways and conditions by which these
processes occur.
Table 4-1: Denitrification and nitrification pathway description and formation
of N2Q____________________
Process Pathway Optimal Conditions for Microbes n2o formation Conditions that Promote N20 Formation
Denitrification (reduction of N03~) Anoxic: DO ideally less than 0.2 mg/L Reduction of High DO (1)
no3 -*no2 -> pH: optimum is 7.0-8.5 NO to N2, Low pH (2)
no-^n2o-+n2 Temperature: 5-30C with N20 High Nitrite (1)
Organic carbon required intermediary Low COD/N ratio (1)
Aerobic: DO ideally From
Nitrification nh3^nh2oh greater than 1 mg/L oxidation of Low DO (1)
(oxidation of >NOH>NO pH: optimum is 7.5 to 8.5 NH2OH or
NH3) N02_Nor Temperature: 10-40C reduction of High Nitrite (1)
Low food to organism ratio no2_
(1) : Kampschreur, Temmink et al. 2009
(2) : Thom and Sorensson 1996
In the denitrification step in WWTPs, N20 formation increases with high
dissolved oxygen, high nitrite concentration, and low COD/N ratio
(Kampschreur, Temmink et al. 2009). All three of these factors are linked to
decreased denitrification rates and lead to reduced N20 emission. Wastewater
pH of less than 6.8 is also expected to increase N20 formation (Thom and
Sorensson 1996), increasing the yield of N20, i.e. the percentage of N2 that is
N20. This ratio of denitrified nitrate that becomes N20 as compared to that
which becomes N2 partially determines the amount of N20 emitted from the
denitrification process (Beaulieu, Tank et al. 2011).
In the nitrification step (aerobic oxidation), low dissolved oxygen causes local
oxygen limitation and an increase in a nitrifier denitrification pathway, in
which nitrifying bacteria oxidize ammonia (NH3) to nitrite (N02 ), and
subsequently reduce NO to N20 to N2 (Wrage 2003; Kampschreur, Temmink
et al. 2009). High nitrite accumulation leads to N20 emissions due to the same
nitrifier denitrification pathway.
38


After wastewater is treated in a WWTP, in the case that effluent is
subsequently reused in urban agriculture, there are few models describing the
fate of nitrogen. According to 1PCC guidelines, nitrogen released to surface
water is assumed to be converted to N2O at an emission rate of 0.005 (range
0.0005-0.25) kg N20-N/kg N. Limited field data and assumptions about
nitrification and denitrification in riverine systems were used to determine this
emission factor. These assumptions are that all of the nitrogen is discharged
with the wastewater and that N2O production in the riverine system is directly
related to nitrification and denitrification of the nitrogen in the wastewater
(IPCC 2006). More recently, Beaulieu et al. carried out extensive experiments
in US rivers to find that 0.75% of dissolved inorganic nitrogen inputs to
riverine systems was converted to N2O emissions by a combination of
denitrification and nitrification (Beaulieu, Tank et al. 2011).
For soil application, the IPCC gives default emission factors for N2O
emissions from managed soils. These range from 0.003 kg N20-N/kg N for
flooded rice fields to 16 kg NiO-N/kg N for tropical organic crop and
grassland soils (IPCC 2006). Wastewater reuse in cropped soils is not
included. Understanding the fate of nitrogen in wastewater reuse for urban
agriculture will also help in understanding the positive impacts of avoided
fertilizer.
Therefore, a WWTP LCA using reported WWTP operations data along with
modeling impacts external to WWTP is needed for a specific site in India. The
objective of this study is to carry out a full systems life cycle assessment of
Nallacheruvu WWTP in Hyderabad, India, including WWTP process
emissions, end-use energy, embodied energy, and consequential emissions
from off-site N2O and CH4 when WWTP effluent is reused in urban
agriculture.
39


4.2 Wastewater Treatment Plant Description
The Nallacheruvu Wastewater Treatment Plant (NWWTP) was commissioned
by the HMWSSB in 2007 after which it was inaugurated on February 25,
2009. The influent wastewater to this plant is thought to be domestic and it
originates from open drains and closed pipes that are connected to the sewer
network. It is difficult to estimate the service area and households served by
NWWTP. Because NWWTP is estimated to treat 5% of the total wastewater
treated in
Hyderabad (table
3-2), the total area
covered by the
sewer network
(294 km2) was
scaled down
proportionally
HYDERABAD METROPOLITAN
WATER SUPPLY*SEWERAGE
JO ML!) TM1TKFM! AU1
At KAUACtttiJVU VWU. ,
ceST9fTOJSCT< BttCnm
**-<** miWft tBOfC* WT its

"it11
*T **** * ?" ' A
(Ministry of Urban Figure 41]; Entrance to NWWTP
Development
2010). From this calculation, the NWWTP serves approximately 15 square
kilometers in and around the area of Uppal, Hyderabad. The total amount of
properties in Hyderabad that are connected to sewers (551,026) (Ministry of
Urban Development 2010) was also scaled down by the 5% proportion to
result in 22,796 properties that are served by NWWTP. To estimate the
amount of people served by NWWTP, the total population (7,597,058) was
divided by total properties (2,028,435) in Hyderabad (Ministry of Urban
Development 2010). Then, the amount of properties that are served by
NWWTP was multiplied by 3.75 people/property, resulting in 85,377 people.
The maximum capacity that can be handled by the NWWTP is 8 million
gallons (US) per day, but actual treatment was as low as 3 MGD in the dry
40


season (March-May 2010) and the average annual flow was 5 MGD for the
first year of operation.
4.2.1 Physical Description
The processes at the NWWTP are in the following order (from inflow to
outflow): coarse screen channel (20mm), pumped to inlet chamber, fine screen
channel (6mm), detritor tank (settling and grit removal), upflow anaerobic
sludge blanket reactors (UASB), facultative aerated lagoon, polishing pond,
chlorination (as needed), and sludge drying. Biogas is collected at the UASB,
scrubbed for H2S removal, and flared. All data shown in figure 4-2 were
provided by NWWTP authorities, except the change in chemical oxygen
demand (COD), biochemical oxygen demand (BOD), and total organic carbon
(TOC) across the UASB, which were independently measured by the
researcher.
Sewer
Open
and
Closed
Drains
'S
/
/
/
\
COD: 497 mg/L
BOD,: 170 mg/L
TOC: mg/L
Ammonia-N; 24 mg/L
Nitrate-N: 28 mg/L
Nitrite-Si: mg/L
Ofganic-N: mg/L
FD 573,277 MPN/lOOmL
/ V
Coarse
and Fine
Screens
and
Detritor
s Tank
" s pH: - 1
D0:mg/l
COD: 537 mg/L
BOD,. 178 mg/L
TO C: 80 mg/L
Ammonia-N: mg/L
Nitrate-N: mg/L
Niinte-N: mg/l.
Orgamr-N; mg/L
. FQ MPN/lOOmL
f
Facultative
; Aerated
Lagoon
DT: 1 day
\
BOD,; mg/L
TOC: mg/L
Atnmnnia-N: mg/L
Nltnitr-N; mg/L
Nitrtte-N: mg/L
Polishing
Pond
DT: 0.S day
V
s
X
pH: 7.8
DO; 5.3 mg/L
COD: 183 mg/L
BOD.: 32 mg/L
TOC: mg/L
Ammonla-N 7 mg/L
Nitrale-N: 15 mg/L
Nitnte-N: 0.1 m*/L
Oiganic-N: 3.5 mg/L
FC: 5,478 MPN/lOOmL
UASB Reactors
SRT: 33 days
HRT: 9 hours
DO: mg/L
COD: 231 mg/L
BOD-,: 82 mg/L
TOC: 56 mg/L
Ammenia-N mg/L
Nitratfr-N: mg/L
Nitiitc-N: mg/L
Organic*#: mg/L
FC: MPN/lOOmL
\
\
Urban Agriculture
and Evapotranspiration
Figure 4-2: Process flow diagram of Nallacheruvu WWTP. The missing data
has been requested from NWWTP. (SRT: solids retention time; HRT:
hydraulic retention time; DT: detention time)
41


4.2.2 Energy Recovery
Biogas is produced and captured at the UASB. At full capacity, this plant is
expected to generate 512 cubic meters biogas/day, which is expected to
contain 60-65% methane (310 cubic meters methane/day) (Kumar 2010). In
the future, power will be generated from a dual fuel genset. The expected
power generation at full capacity of the plant is 705.6 kWh/day (Kumar 2010).
Currently, power generation has not been started. Fugitive emissions of
biogas from the UASB have not been measured.
Sludge is currently being generated and dried at the plant. At full capacity, 4
cubic meters grit/day, 12.6 cubic meters wet sludge/day and 6.28 metric tons
dried sludge/day is expected (Kumar 2010). In the first year, only about 1 mt
dried sludge/day was being generated (Kumar 2010). The plant had not yet
decided the fate of the dried sludge, but it is dried and held on-site.
4.2.3 Process Insight
The influent wastewater to the NWWTP can be compared to the typical
composition of high strength (60 gal/capita/day) untreated domestic
wastewater, given by Metcalf and Eddy (Table 4-2).
Table 4-2: Typical composition of high strength untreated domestic
wastewater (Metcalf and Eddy 2003) compared to influent wastewater to
NWWTP reported from March 2009- March 2010 (Kumar 2010). TBD=to be
determined.
Parameter Untreated Domestic Wastewater Influent WW to NWWTP
COD 800 497
bod5 350 170
TOC 260 TBD
Total Nitrogen 70 TBD
Ammonia 45 24
Nitrate 0 28
Nitrite 0 TBD
Organic 25 TBD
Fecal Coliform 105-108 6*105
TSS 85 353
VSS 315 128
42


4.2.3.1 Sewerage System
The dissolved oxygen in the influent wastewater to NWWTP (1.9 mg/L)
suggests that the sewerage system is an aerobic environment. This is expected
as the network of open drains and closed pipes is cascading, turbulent flow,
which introduces oxygen into the wastewater. The high nitrate (NO3 ) in this
influent water may result from a combination of: nitrification occurring in the
aerobic environment within the sewerage system, runoff from agriculture, and
industrial effluents. Some nitrification is evident in the lower amount of
ammonium (NH/) and higher amount of NO3 (table 4-2), and is expected
due to the optimal pH, DO, and temperature (table 4-1). However, the short
retention time may hinder a large amount of nitrification.
Ammonification, or the conversion of organic nitrogen (NH2) to NH3 or NH4+,
is also occurring within sewers. Biodegradable soluble organic nitrogen, a
major component (90-95%) of total organic nitrogen, is converted to NH/
when the pH is lower than the pKa (9.25), as is seen in NWWTP influent. At
pH higher than pKa, biodegradable soluble organic nitrogen is expected to be
converted to NH3 and lost as a gas (Metcalf and Eddy 2003).
4.2.3.2 Screens and Detritor Tank
The screens remove some solids and provide more aeration. The settling in the
detritor tank removes more solids and insoluble particulate organic nitrogen.
The DO is expected to decrease here and the water is expected to start
approaching an anoxic environment. Also, settling is expected to remove
some portion of fecal coliforms and nematode eggs (George, Crop et al.
2002).
4.2.3.3 Upflow Anaerobic Sludge Blanket
Microbes in the wastewater use organic carbon as their food source and prefer
to use oxygen as an electron acceptor until the DO is below 0.2 mg/L. Then,
the microbes are forced to use NO3 ) as an electron acceptor. NO3 is expected
43


to be reduced as shown in table 4-1. Denitrification causes the pH to increase,
and may result in a deprotonation of NH/ to NH3 and resulting loss of NH3
gas. N2O yield (percentage of denitrification) is expected to decrease to
undetectable levels (~0) when pH is higher than 6.5 to 7 (Thom and
Sorensson 1996). Finally, organic nitrogen is not expected to change much
due to UASB processes (Arceivala and Asolekar 2007)
4.2.3.4 Facultative Aerated Lagoon
The aerobic environment allows for more extensive nitrification as conditions
are optimal for nitrifiers to oxidize NH4+ to NO3 In the earlier part of the
basin, heterotrophic organisms feed on organic matter and consume oxygen,
while nitrifiers grow in the later part where little organic matter is present and
there is less competition for oxygen. Nitrite is not expected to accumulate at
any point in the WWTP, and if it did, it would mean that toxic conditions have
killed the necessary microbes (Novotny 2006). BOD, COD, and TOC should
all decrease due to the cascading aeration. However, at NWWTP, the
detention time of 1 day may be too low for significant nitrification. We have
requested two sets of in-plant measurements to verify this.
4.2.3.5 Polishing Pond
The final step for NWWTP is often the polishing pond. Removal of additional
BOD, COD, TOC, nutrients as well as fecal coliforms and nematode eggs is
expected here through settling (Spellman 2009). A chlorination bed follows
but is not often used.
4.2.4 NWWTP Performance Data from HMWSSB
The performance data provided by the Hyderabad Municipal Water Supply
and Sewerage Board was averaged for one year (March 2009-March 2010)
and is shown in table 4-3.
44


Table 4-3: NWWTP measured parameters and treatment efficiencies for
March 2009- March 2010 (Kumar 2010)__________________________________________
Parameter (Units) Average Influent to Average Effluent from NWWTP NWWTP Average Standard Deviation (n= # of data points) Removal % for Pollutants Disposal Standards in appropriate units
pH 7.2 0.7 (n=46) 7.8 0.9 (n=46) 5.5-9.0
DO (mg/L) 1.9 2.1 (n=44) 5.3 2.4 (n=44) ---
BOD5 (mg/L) 170 87 (n=46) 32 43 (n=46) 81% 30
COD (mg/L) 497 207 (n=45) 183 189 (n=45) 62% 250
TSS (mg/L) 353 251 (n=27) 23 43 (n=27) 93% 100
VSS (mg/L) 128 77 (n=24) 8.1 17 (n=24) 94%
Fecal Coliforms (MPN/lOOmL) 573,277 300,837 (n=26) 5,478 2,206 (n=26) 99% 10,000
The average effluent discharges at NWWTP are reported to meet disposal
standards set by the Indian Central Pollution Control Board. Chlorination is
available at NWWTP, but it is rarely used because the effluent fecal coliform
concentrations meet disposal standards. There is evidence that chlorination
produces N2O during decompostition of monochloramine (NH2CI) at neutral
pH (Hashimoto 1981). Because chlorination is not often utilized in this
WWTP, these emissions are thought to be negligible.
4.2.5 UASB WWTP
Performance
The NWWTP
performance is in line
with the literature.
Concerning pathogen
treatment efficiency,
UASB paired with a
polishing unit alone can
remove 99.8% nematode
eggs (Tyagi, Sahoo et al.
2010). UASB
technology is useful in meeting the maximum permissible limits of fecal
coliform for disposal (10,000 MPN/lOOmL), but the suggested desirable limit
45


of 1,000 MPN/lOOmL is not often met (Khalil, Sinha et al. 2008; Tare and
Nema n.d.). Therefore, tertiary treatment is needed and chlorination is often
planned, but it is only used to meet the maximum permissible limits.
In the case of resource recovery, which was a major attraction in choosing
UASB implementation, many plants that planned to utilize biogas for
electricity generation have not yet started and sludge is not yet being utilized
in a significant way (Kumar 2010; Tare and Nema n.d.). For power
generation, this delay could be for many reasons, including low biogas
generation in small and medium sized WWTPs and in cold months in
Northern India. Also, there is very little incentive for WWTP to start energy
recovery from biogas for the following reasons: the UASB WWTP has a low
energy requirement, power outages do not greatly affect the technology, the
energy bill is linked to the installed load of the WWTP, and there are upfront
costs and risks to starting and maintaining power generation from biogas
(Tare and Nema n.d.). Utilizing dried sludge has also been problematic as
there is not a reliable or lucrative market for the sale of sludge (Kumar 2010;
Tare and Nema n.d.).
4.2.6 Consequential Life Cycle Assessment
Consequential LCA is used to determine how flows and related impacts will
change as a result of decisions and actions taken outside the WWTP
boundary. Consequential LCA has been used to quantify GHG impacts due to
changes in flows in response to policy decisions, such as the ramping up of
com-based biofuels in the US (Zhang, Spatari et al. 2010). At a time when
some studies (Farrell, Plevin et al. 2006; Argonne National Laboratory 2008)
suggested that biofuels could reduce atmospheric GHGs as growing feedstock
sequesters carbon, Searchinger et al. used consequential LCA to determine
that GHGs may actually increase as a result of land use change. This group
modeled land use change due to a worldwide farmer response to growing
46


demand and prices for biofuels, and predicted that they would subsequently
convert grasslands, forests, etc, to cropland (Searchinger, Heimlich et al.
2008). Consequently, they modeled GHGs to double in the coming years.
In this WWTP LCA, decisions about both in-boundary and out-of-boundary
flows will be considered. Because the WWTP does not have direct control
over what happens to the released water after treatment, GHG impacts from
WWTP effluent when it is released to the environment or reused in urban
agriculture is considered out-of-boundary, and could have land use change
implications. Two other consequential impacts will be included: avoided
fertilizer due to the nutrient content of WWTP effluent reused in urban
agriculture, and avoided electricity from the grid when biogas is reused to
generate electricity.
4.3 Life Cycle Assessment Methodology
The following equation (4 components) is used to quantify total on-site life-
cycle energy use and GHG emissions. The four components include:
End-use energy in WWTP operations,
Process emissions of methane (CH4),
Process emissions of nitrous oxide (N2O), and
Embodied energy of infrastructure.
1CCH0W - * GWc, >+ )
n
+(WN20 GWPNfi) + ((Wc_, EFCeaJITUfetiJ\
(Equation 4-1)
where: Eon-site= on-site energy use at the WWTP (only electricity);
EFEicctricity= average CO2 emission factor from thirteen thermal power
plants (Coal, Oil, and Gas Fuels) in the State of Andhra
Pradesh (2008-2009);
47


WCH4-uASBFugitive= kg methane lost from UASB as fugitive emissions
per year;
GWPch4= global warming potential of methane: 24 kgCChe/kgCtB;
WcH4-CaPturcdLcak= kg methane leaked after being captured per year;
WCemcnt= total kg cement used in WWTP infrastructure;
EFccment= CO2 emission factor for cement manufacture;
TLifctimc= average lifetime of WWTP infrastructure: 30 years;
Wn20= kg nitrous oxide emitted from nitrification and denitrification
processes on-site at NWWTP per year;
GWPn20= global warming potential of nitrous oxide: 298
kgC02e/kgN20.
Equation 4-1 is applied to assess energy and GHG emissions for LCA of
WWTP in India with emphasis on UASB technology.
For the WWTP LCA, scope 1 emissions come directly from the wastewater in
the form of N2O and CH4, scope 2 emissions come from end-use electricity,
and scope 3 emissions are the embodied energy of infrastructure.
Consequential LCA addresses the full system inside and outside the WWTP
boundary comparing various scenarios with and without wastewater
agriculture, and will be discussed in a separate section. This consequential
LCA is important, as little is known about life cycle GHG impacts of WWTP
when the wastewater is subsequently used for agriculture.
Greenhouse gas impacts of WWTPs are not well characterized in developing
countries. The IPCC Guidelines for National Greenhouse Gas Inventories
provides a method to calculate GHG emissions from untreated wastewater for
India (Doom, Towprayoon et al. 2006), although is has high uncertainty.
When WWTPs are built in developing countries, there is an expected increase
in energy use and greenhouse gas emissions from WWTP electricity use.
However, if CH4 capture is used in the WWTP, significant GHG emissions
can be mitigated. These consequential LCA impacts will be quantified in
section 4.4.2, following WWTP LCA.
48


4.3.1 Goal and Scope of WWTP LCA
LCA goal and scope definition serve the purpose of setting the boundaries of
the project and help to determine the most appropriate functional unit. The
scope for this LCA includes WWTP process boundaries, with consequential
LCA representing various scenarios outside the boundary. The functional unit
used was million gallons wastewater treated per year (sometimes expressed as
per gallon since this LCA was done as NWWTP had been operating for one
year).
4.3.2 WWTP LCA Data
4.3.2.1 Scope 1 Direct Emissions from WWTP Processes
4.3.2.1.1 Fugitive emissions of methane from UASB
Even though methane capture is an integral part of the UASB, fugitive
emissions from the water surface are expected. Biogas yield can be estimated
from the amount of COD removed during the treatment process. This
theoretical biogas yield is between 0.35 and 0.5 cubic meters biogas/kg of
COD removed (IPCC 2006; Tare and Nema n.d.). However, for UASB
technology the actual yield is expected to be only 25-30% of this value or
0.08-0.1 cubic meters biogas/kg of COD removed (Tare and Nema n.d.). The
majority of the biogas remains dissolved in the effluent, and increases its
BOD and COD. It is assumed that this biogas is made up 65% methane and
32% carbon dioxide by volume (Monteith, Sahely et al. 2005; Kumar 2010);
0.651 kgCH4/cubic meter CH4 was used for the density of methane (at latm
and the average daily temperature of Hyderabad, 27.4C (Rao 2010)). The
difference between the amount produced and the amount biogas captured was
then calculated and assumed to be lost to the environment.
W
VyCH4-UASBFuf.ilh,e
(V -V
k Biogas-Yield Biogas-Captured
MG
\ p n
> 1 CII, ^CH,
(Equation 4-2)
49


where: VBi0gas-Yieid= cubic meters of theoretical biogas yield from the UASB
per year, based on COD reduction;
VBiogas-CaPturcd= cubic meters of biogas captured from UASB per year;
Pch4= proportion of biogas that is methane: 0.65 cubic meters
CH4/cubic meters biogas;
Dch4~ density of methane: 0.651 kg CH^cubic meter CH4 at average
daily temperature for Hyderabad (27.4C);
MG= million gallons: 1,736 million gallons wastewater treated per
year.
4.3.2.1.2 Methane leakage after capture
Because the volume of biogas currently captured is not measured, the direct
emissions from biogas were estimated as a proportion of that expected for full
capacity of the plant. Even though biogas is flared at this plant, incomplete
combustion and leaks are expected and a 5% undetected biogas leak rate can
be assumed (Sahely, MacLean et al. 2006).
W,
CII4 -Captured Ultli
y */? * P *D
r Biogas-Captured 'Leak 1 CIIA LJdl4
MG
(Equation 4-3)
where: R.Leak= undetected biogas leak rate due to incomplete combustion and
leakage: 5% according to Sahely et al. 2006.
4.3.2.1.3 Fugitive emissions of nitrous oxide from UASB and Oxidation
Pond
Few people have measured N2O and CH4 emissions from UASB and
oxidation ponds. Although it is thought that denitrification in WWTP anoxic
zones is the largest contributor to N2O emissions (US EPA 2009), nitrification
in WWTP aerobic zones has been found to be significant, especially in
WWTP where both anoxic and aerobic processes are used (Kampschreur, van
der Star et al. 2008; Ahn, Kim et al. 2010). Kampschreur et al 2009 completed
a comprehensive review of many different types of WWTP processes and
their related emissions (Kampschreur, Temmink et al. 2009) and Ahn et al
measured emissions from many different technologies.
50


For this WWTP LCA, fugitive emissions from the anaerobic UASB reactors
and the aerobic facultative aerated (oxidation) pond were estimated from
similar technologies described in Ahn et al: the anaerobic portion of a
separated biological nutrient removal (BNR) WWTP at approximately
0.001% as kgN20-N/kg total nitrogen removed (10% of 0.01%), and an
oxidation ditch at 0.03% kgN20-N/kg total nitrogen removed (Ahn, Kim et al.
2010). Therefore, nitrous oxide emissions were estimated at 0.031% (range
0.02%-0.04%) as kgN20-N/kg total nitrogen removed. Ahn et al also showed
the % of influent TKN that becomes N2O. In order to make comparisons with
the percent of influent nitrogen in IPCC methodology, this is the number
shown in table 4.4.
Table 4-4: N2O emissions from WWTP processes: a comparison of results
using IPCC methodology to findings by Ahn et al 2010._________
Source Method Equation mg N20/ person/ year (Range) EF: % as kg N20-N/kg influent N (range)
IPCC 2006 Calculated for Hyderabad, India N2Owwtp= Population (HYD) *degree of utilization of modem, centralized WWTP % (used % ww collected in HYD) fraction of industrial and commercial co- discharged protein (default 1.25 given by IPCC) EF kgN20/person/year (0.0036) 628 (201- 2,715) 0.02 (0.01- 0.04)
Whole Plant 280 (150- 410) 0.01
Ahn et al. 2010 Anaerobic portion of BNR Process Researchers in the US measured mass flux from each zone in the WWTP and normalized to the daily influent total Kjeldahl 28(15-41) 0.001
Aerobic Oxidation Ditch nitrogen (TKN) loading 1800 (1030- 2570) 0.03
4.3.2.1.4 Scope 1 Direct Emissions from On-site WWTP Operations
Emissions from vehicle transport operations are very low because there are no
municipally-owned or company-owned cars specifically for WWTP use.
51


Motorbikes are used by a few workers, but mainly for transport to and from
work, and not often for transport around the WWTP grounds. Human-
powered pushcarts are the most common form of transport around the WWTP.
Electricity was the major form of energy used at the Nallacheruvu WWTP;
diesel and natural gas use were reported to be very little to none (Kumar
2010). Therefore, emissions from gasoline, diesel, and natural gas are not
included.
4.3.2.2 Scope 2 Emissions from Electricity Use for WWTP Operations
The actual on-site electricity use for the plant was 3.3 MWh/day for the first
year of operation, equivalent to 0.7Wh/gal (6.2 MWh/day is expected at full
capacity). An average of emission factors for 13 Thermal Power Plants in the
state of Andhra Pradesh (Coal, Oil, and Gas Fuels) (2008-2009) is 0.7470.3
mtC02/MWh (only carbon dioxide emissions were reported here) (India
Central Electricity Authority 2009).
4.3.2.3 Scope 3 Emissions from Embodied Energy in WWTP Materials
On-site building materials are dominated by concrete use for construction of
the WWTP. Concrete is used for: coarse screen channel, main pumping
station, inlet chambers, fine screen channel, detritor tank, division box 1&2,
distribution box, UASB reactors, facultative aerated lagoon, polishing pond,
chlorine mixing tank, chlorine contact tank, sludge pump house, sludge drying
beds, gas holder, gas scrubber/blower room, biogas genset room, biogas flare
unit, chlorination room, mechanical, electrical and plumbing room, and
administration block house. On-site piping includes: raising main (one cast
iron), distribution lines (two cast iron and one high density polyethylene),
sludge lines (two cast iron), filtrate line (one stainless steel), and gas line (one
fiberglass reinforced plastic).
Because incomplete data was provided for piping infrastructure, the
infrastructure material emissions are calculated only for concrete use. An
52


estimated total 17,022 metric tones of concrete was calculated from widths
and lengths provided by Mr. Kumar of NWWTP, and some heights were
estimated by this researcher. An emission factor of 0.33 mt CC^e/mt concrete
(Pitterle 2008) was used. A typical lifetime of a WWTPs pumps, tanks, and
other technical parts is fifteen years while buildings, filter beds and pipes is 30
years (Lundin, Bengtsson et al. 2000; Foley, Haas et al. 2005). Therefore, the
emissions from cement were divided by 30 years.
4.3.2.4 Consequential Evaluation of GHG Emissions
N2O and CH4 emissions from WWTP effluent released to both urban
agriculture plots and the riverine system were quantified as out-of-boundary
emissions. For urban agriculture, these emissions can be quantified with
greater certainty because N2O emissions from managed soils is well-studied
(Del Grosso, Ojima et al. 2009) and CH4 is expected to be very low due to the
aerobic environment of land application. However, these emissions from
rivers and streams are more uncertain.
The impacts of two main scenarios were evaluated:
Uncontrolled release of wastewater without treatment and
53


Treatment at NWWTP with partial reuse in urban agriculture with the
remainder flowing into the riverine system downstream of NWWTP.
For NWWTP treated effluent release to urban agriculture, the amount of land
adjacent to the WWTP was estimated using Google Earths ruler tool. The
amount of water that would be used by the total area over 10 months
(agriculture was estimated to stop for 2 months per year due to monsoon rains
as per interviews with local farmers) was estimated from the volumes used in
this reports site study (see Chapter 5).
In addition to the two core scenarios, avoided GHGs were computed from:
Avoided fertilizer and
Avoided electricity due to the reuse of biogas.
Avoided GHG emissions from avoided fertilizer and electricity from biogas
reuse were calculated in the following ways. For urban agriculture, avoided
emissions from fertilizer use that wasnt needed due to the nutrient content of
treated wastewater were credited to the off-site N2O emissions.
Concentrations of total soluble nitrogen, phosphorus, and potassium (N, P, K),
the major components of synthetic fertilizers, in treated effluent water were
measured at the site (see chapter 5). These amounts were scaled up for use in
the adjacent area to NWWTP over 10 months per year. Then, to determine the
avoided GHG emissions from not using synthetic N, P, and K fertilizers, the
amounts were multiplied by emission factors compiled by Pitterle and totaled
(Pitterle 2008).
CHC _ (('
Avoided Pmiliztr \ lrrigaiionWalerNiogtn
EF )+(C
Fertilizerplwspholm > \ lmgationWaterPmssim
*FF 1+fC
Fertilizer^,rog{ > V lrngationWaterPhosphoms
* EF 1
FertilizerPplassium >
*
(Equation 4-4)
where: GHGAvoided Fertiizcr= avoided CO2 equivalent emissions from fertilizer;
Cin-igation Water (N, p, k.) = concentration of nutrients in irrigation water
54


(based on treated effluent water measurements from Chapter 5);
EFfertiiizer (n, p, K)= emission factor: 4.57 kg C02e/kg-N, 1.25 kgCC^e/
kg-P, or 1.29 kg CCFe/kg-K fertilizer (Pitterle 2008).
These avoided emissions were then credited to the off-site N2O emissions or
N2O from agriculture in the results.
As stated earlier, biogas that is currently flared on-site at NWWTP could be
used to generate electricity. This potential electricity credit was calculated by
using a net calorific value of 23.3 MJ/m3 for biogas with 65% methane from
anaerobic digestion in a WWTP (Bonnier 2008). A 25% power plant
efficiency for biogas to electricity from Pitterles work at a US WWTP is used
here (Pitterle 2008).
GHG
Avoided r
= V
Biogas- Captured
* NCV,
Biogas
* *7pp-Biogas XkWh-MJ
* EF
Electricity
(Equation 4-5)
where: GHGAvoided Electricity = avoided CO2 equivalent emissions from electricity
generated from biogas;
NCVBiogas= net calorific value of biogas: 23.3 MJ/m3 (Bonnier 2008);
Tlpp-Biogas= US average power plant efficiency: 25% (Pitterle 2008);
Xkwh-Mj= conversion: 0.2778 kWh/MJ;
EFEiectridty= average CO2 emission factor from thirteen thermal power
plants (Coal, Oil, and Gas Fuels) in the State of Andhra
Pradesh (2008-2009) (India Central Electricity Authority 2009).
This result was credited to on-site WWTP emissions.
Finally, to yield total emissions per million gallons, these avoided emissions
were divided by 1,736 million gallons of wastewater treated by NWWTP per
year, then multiplied by the amount of water appropriate for each scenario in
the results.
55


4.3.2.4.1 Methods for the Two Core Scenarios
4.3.2.4.1.1 Untreated Base Case
GHG emissions from direct release of untreated wastewater was estimated
using IPCC methodology for both CH4 and N2O releases to streams (IPCC
2006). For methane, the IPCC methodology and variation was used and is
described in Equation 4-6.
GHGRiverineCHi = C,nfluen,-COD (Bo MCF) GWPCHi (Equation 4-6)
where: GHGRiverinc ch4= CO2 equivalent emissions from methane released
from riverine systems per million gallons wastewater treated;
Cinfiuent-coD= concentration of COD in influent wastewater: 1,881kg
COD/million gallons (Kumar 2010);
B0= maximum CH4 producing capacity: 0.25 kgCFE/kgCOD (IPCC
2006);
MCF= methane correction factor for rivers and lakes, an indicator on
the degree of which the system in anaerobic: 0.1 (IPCC 2006);
GWPch4= global warming potential of nitrous oxide: 24
kgC02e/kgCH4.
The variation CH4 emissions were calculated from the variation in MCF,
which ranged from 0 to 0.2 for rivers and lakes (IPCC 2006). In IPCC
methodology, the emission factor for CH4 from riverine systems is given by
B0*MCF.
To estimate N2O emissions from wastewater in riverine systems, IPCC
methodology along with a PNAS study was used. IPCC assumptions result in
the estimation that N2O emissions from nitrification are double those from
denitrification in streams (Mosier, Kroeze et al. 1998; Beaulieu, Tank et al.
2011). In a PNAS study by Beaulieu et al. 2011, researchers carried out an
extensive study in order to improve the estimate of the total amount of
nitrogen that is converted to N2O in streams.
56


Their results
showed no
correlation to river
network length,
even though longer
water residence
times were
expected to
increase the
percentage of N
being denitrifed.
Regardless, the
global range was
0% to 0.9% conversion of N inputs to N20 via denitrification (figure 4-5).
Ultimately, Beaulieu et al. estimates that the percentage of dissolved inorganic
nitrogen was converted to N2O via denitrification and nitrification in rivers is
0.75% (Beaulieu, Tank et al. 2011). For this WWTP consequential LCA, the
average value reported was using the IPCC emission factor
(0.005kgN20/kgN), and shown in equation 4-7.
GHCjRiVerine G' InfluentNitrogen EFRiverine ^ GWPN2o (Equation 4-7)
where: GHGRiVerine N20= CO2 equivalent emissions from nitrous oxide released
from riverine systems per million gallons wastewater treated;
Cinfiucnt-Nitrogen= concentration of inorganic nitrogen in influent
wastewater to NWWTP (Kumar 2010);
EFRivcrincN20= the default IPCC emission factor for N2O emissions
from domestic wastewater nitrogen effluent from nitrification
and denitrification in rivers and estuaries: 0.005 kgN20/kgN
(IPCC 2006);
GWPN2o= global warming potential of nitrous oxide: 298
kgC02e/kgN20.
0.1 1 10 100 1000 10000
N yield from catchment to stream (kg km'2 y'1)
Figure 4-5: Results from Beaulieu et al 2011. Percent
conversion of N inputs to N2O via denitrification versus
the nitrogen yield from catchment to stream (indicator of
river network length) of 866 rivers worldwide
57


To calculate the variation in N2O emissions from riverine nitrogen, the
emission factor of 0.0075 kgN20/kg dissolved inorganic nitrogen from the
PNAS article was used and the range is shown in Figure 4-11.
These N2O and CH4 emissions per gallon of untreated wastewater are
multiplied by the amount of water appropriate for each scenario in the results.
4.3.2.4.1.2 DAYCENT for estimating of N2O emissions from agriculture
Because few people have measured N2O directly from urban agriculture with
wastewater irrigation, this researcher had planned to measure N2O emissions
in the site study (chapter 5). However, the equipment in India did not have the
necessary parts, and permission to use equipment in the US was not given due
to instrument contamination associated with these studies. Therefore, the
DAYCENT model was used.
DAYCENT, developed by a group at the Natural Resource Ecology
Laboratory at Colorado State University (CSU), is a well-documented and
widely used model for estimating GHG emissions from cropped fields (Del
Grosso, Mosier et al. 2005; Del Grosso, Ojima et al. 2009). It has been used
and validated by researchers as well as the US Environmental Protection
Agency (Jarecki, Parkin et al. 2007; US EPA 2011). It is most often used to
estimate N2O emissions for major crops (wheat, com, soybeans, etc) with
commercial fertilizer use. It has not been used for wastewater agriculture, for
vegetables, or for India.
Inputs to the model included (and detailed in Appendix A):
Weather specific to Hyderabad: obtained for 2000-2010 from Dr. Kesava
Rao, a scientist of Agroclimatology at ICRISAT;
Historical data: assumptions were made on agriculture frequency, type,
nutrient delivery, etc from year 1 until this study started. It is known that
58


agriculture began in the area around the late 1960s and continued until the
wastewater treatment plant was built. Grazing of buffaloes in the area has
been occurring and
fire is often used to
clear grasses to
start cultivation.
Soil
characteristics:
physical and
chemical
Figure 4-6: Buffaloes grazing at NWWTP.
parameters that
were determined from lab tests described in chapter 5;
Crop characteristics: growth type, fraction of carbon allocated to roots,
weather conditions appropriate for growth, etc. which were discussed with
Dr. Parton at CSU. Also, grams carbon per kg of spinach was needed to
determine net primary productivity (National Council on Radiation
Protection and Measurements 1983).
Nitrogen and organic matter delivered in the water were scaled from year
1 to 2011 from water nutrient and organic matter test results in the site
study (for further description of sites and nutrients, see chapter 5). The
seasonal distribution was based on measured amounts of riverine nitrogen
at the basin month of the Ganges (Green, Vorosmarty et al. 2004). Then,
based on the change in river nitrogen load from 1970-2010 (estimated at
50% for Brazil-Russia-India-China (BRIC)), nitrogen was scaled
accordingly for the last 40 years (Bakkes, Bakkes et al. 2008). Previous
nutrients were scaled linearly back to year 1.
59


Fertilizer and organic matter addition events were scheduled along with
irrigation events, as these components were delivered in the irrigation
water.
In the model, all study activity took place on the same days as they
actually occurred. After the site study in March-April 2010, a monsoon
season was simulated and cultivation began again in late June 2010 and
continued for 9 additional identical growing cycles through February
2011. For yearly data, the emissions from these 10 growing cycles were
summed. Input data are detailed in Appendix A.
Output N2O emissions from the model were summed through the end of 2011,
as fluxes were seen, and expected, for many months after the study ended. The
first growing cycle was based identically on the actual field study. For the 9
additional growing cycles, modeled irrigation, nitrogen fertilizer, and organic
Total Inorganic Nitrogen Added g/ha N20 flux g N/ha
1,000,000
Figure 4-7: Output of total inorganic nitrogen added in treated effluent
irrigation water as compared to nitrous oxide flux from 10 growing cycles of
spinach-cultivated land
60


matter addition events occurred once every three days Fluxes of N20, that are
higher than the baseline, can be seen for about 9 months after the last addition
of nitrogen (figure 4-7).
In figure 4-7 a steady baseline concentration of 1,9gN/hectare for N20 flux
can be seen. This background was subtracted from the total N20 flux in the
consequential LCA results.
In the DAYCENT model, net primary productivity depends on a temperature
curve representing the conditions for which the crop grows best and other
growth parameters that were refined for the site irrigated with untreated water
(chapter 5). The modeled output for productivity at the site irrigated with
treated water was also similar to the actual measured productivity. However,
the site irrigated with groundwater had higher modeled productivity than the
actual. This is thought to be due to nutrients in lower soil levels that the model
is including, but which the actual crop could not access (table 4-5) (Del
Grosso, Parton et al. 2011). Background concentrations are not subtracted
from values in table 4-5.
Table 4-5: Example of DAYCENT outputs using agricultural plots for one
growing cycle only (chapter 5)________________________________________________
Output
Units
Source
Irrigation Water at Site
_ . Untreated
. Treated e
Groundwater ...... t Surface
Effluent ... ^
Water
Net Primary Productivity gC/ha Actual 67 759 948
DAYCENT Model 485 671 950
N20 nitrification gN/ha 2,082 2,235 1,291
N20 denitrification gN/ha 25 0 0
N20 flux gN/ha 2,108 2,235 1,292
NO flux gN/ha 3,821 11,399 9,766
N2 denitrification gN/ha 16 0 0
For this WWTP consequential LCA, N20 emissions per square meter from
DAYCENT were divided by the gallons of water used per square meter per
year (determined from water quantity used for the treated effluent plot in
61


chapter 5), and multiplied by the amount of water that could be applied to
agriculture for the different scenarios. N2O is the only GHG considered here
because the aerobic environment of agriculture oxidizes more CH4 than it
releases. Therefore, CH4 emissions fall to zero when the treated wastewater is
reused in agriculture.
4.4 LCA Results
4.4.1 On-site Energy Use and GHG Emissions
The in-boundary results of this study can be described by the following
efficiency metrics for water use, energy use and related operating costs, and
GHG emissions for the NWWTP.
4.4.1.1 Wastewater Generated
Wastewater collected for NWWTP is 56 gallons (US) per person per day as a
yearly average. In 2006, Indias average per capita wastewater generation was
36 gallons/day (United Nations Development Programme 2006). Because the
sewer network is a combination of closed pipes and open drains, they are
expected to collect stormwater during the rainy season and other
water/effluents that are dumped by residents.
4.4.1.2 Energy Use
The actual electricity use at this WWTP was calculated for the average
5MGD (18MLD) over the first year of operations. The average electricity to
treat one gallon of wastewater at this plant was 0.7Wh/gal, much less than that
reported in the US due to UASB selection. Finally, the cost of electricity only
for treating wastewater for the Nallacheruvu WWTP comes to approximately
$0.0001 USD/gallon wastewater treated (or 0.004 INR/gallon wastewater
treated).
62


4.4.1.3 GHG Emissions
The greenhouse gas emissions from this WWTP come from electricity use,
infrastructure, CH4 leakage before and after capture from the UASB, and N2O
emissions from the denitrification and nitrification processes in NWWTP. The
total emissions from CH4 were 435mgC02e/gallon wastewater treated and the
total N2O emissions were 9mgCC>2e/gallon wastewater treated as seen in table
4-6.
Table 4-6: On-site energy use and GHG emissions for NWWTP
Item Nominal (Average) Flow Parameter (A) Comparative Benchmark (USA) Emission Factor or Global Warming Potential (B) GHG Contribution (=A*B)
Operations Electricity 0.7 Wh/gallon (1) 2.4 Wh/gallon (5) 747 mgC02e/Wh (6) 518 mgC02e/ gallon
COD Reduction in UASB for Methane Production 1,158 mg COD reduced/gallon (1) 17 mg CH4 fugitive emissions/gallon (2,7) 211 mg CH4 captured/gallon (5) 24 mgCOie/ mgCH4 (7) 435 mgC02e/ gallon
27 mg CH4 captured/gallon (1) 1.4 mg CH4 leaked from capture/gallon (3)
Nitrous Oxide Production as a Percentage of Total Nitrogen Removal 99 mgTN removed/ gallon (1) Anaerobi c Process 0.001% as mg N20-N/mg N (4) 111 mgTN/ gallon (5) 0.077% mgN20- N/mgN (5) 298 mgCOie/ mgNiO (7) 9 mgC02e/ gallon
Aerobic Process 0.03% (0.02- 0.04) as mg N20-N/mg N (4)
103 mgC02e/gallon for NWWTP (7)
Concrete ! 327 mg concrete/annual gallon (l) 163 mg concrete/ annual gallon (5) 0.33 mgC02e/ mg concrete (5) 108 mgC02e/ annual gallon
(1) : Kumar 2010
(2) : Tare and Nema n.d. See section 4.3.2.1.1
(3) : Sahely, MacLean et al. 2006
(4) : Ahn, Kim et al. 2010
(5) : Pitterle 2008
(6) : India Central Electricity Authority 2009
(7) IPCC 2006
The emissions per gallon of water treated for each item as well as the total on-
site emissions are shown in figure 4-8.
63


On-site On-site On-site Nitrous On-site Total On-site
Electricity Use Methane from Oxide from Embodied Emissions
Processes Processes Energy in
Infrastructure
Figure 4-8: On-site energy-related and process GHG emissions at NWWTP.
Error bars show the range in emissions.
The ranges shown for on-site electricity-related emissions are due to the range
in the emission factor (standard deviation of 300mgC02e/Wh) (India Central
Electricity Authority 2009), for on-site methane from processes are due to the
3
range in COD conversion to biogas (0.08-0.1 m /kg COD removed) (Tare and
Nema n.d.), for on-site nitrous oxide from processes are due to the range in
total nitrogen converted to N2O in aerobic processes (Ahn, Kim et al. 2010),
and for embodied energy in infrastructure are due to the range in the emission
factor (0.25-0.33 mgC02e/mg reinforced concrete) (Pitterle 2008).
4.4.1.4 Limitations of WWTP Data
While the on-site energy use from NWWTP (0.7Wh/gal) is within the range
for Indian cities shown in Chapter 2 (0.2-5Wh/gal), there was no way to verify
this data independently. Data reliability is an ongoing issue that is found in
many LCA studies and is acknowledged here. As mentioned previously,
further measurements of flows throughout NWWTP (figure 4-2 and table 4-2)
have been requested and are currently being measured. However, the results
have not been obtained at the time of this publication.
64


4.4.2 Consequential LCA Results
Figure 4-10 shows the results from various modeled efforts towards reducing
greenhouse gas impacts of wastewater. In this case study, when treated
wastewater is reused in urban agriculture on readily accessible land, the
impact in terms of GHGs is not significantly different when compared to
uncontrolled release of untreated wastewater. As seen in figure 4-9, only 1%
of the nitrogen is being captured in available agricultural land. This is due to
results being highly sensitivity to the amount of land available. Sensitivity to a
different variety of crop (paragrass, a tall grass used as animal feed) was also
tested. However, the resulting emissions were not much different and are not
shown here.
99kgTN/MG Removed in
WWTP
9 kgC02e/MG
Uncontrolled Release of
Untreated Wastewater to
Surface Water
1,719MGY
Release of Treated
Wastewater to
Surface Water
17MGY
Reuse of Treated
Wastewater for
Urban Agriculture
195kgTN/MG
290 kgC02e/MG
504 mtC02e/year
95 kgTN/MG
141 kgC02e/MG
242 mtC02e/year
Figure 4-9: Wastewater nitrogen flows in the two core scenarios
The amount of land easily accessible (with minimal infrastructure) for treated
effluent reuse in urban agriculture was estimated to be 5,525m and indicated
in figure 4-10. The total treated effluent released from this plant would need
approximately 561,838m land for total reuse in urban agriculture. There is
this much land available near to NWWTP (in figure 4-10), but mixing with
nearby streams would be difficult to avoid.
65


Avoided fertilizer
and avoided
electricity due to
biogas reuse
decreases the total
GHG impact by
5% each. In the
case that all
NWWTP effluent
could be directly
reused in urban
agriculture, the
GHG impact could
Figure 4-10: Land easily accessible for NWWTP treated be decreased by
effluent reuse in urban agriculture.
about 20% (figure 4-
11).
N20 from Agriculture Off-site CH4 Off-site N20 On-site WWTP
5,000
Wastewater
without NWWTP
Stream
Accessible
Agriculture and
Remaining
Accessible
Agriculture and
Remaining
Agriculture
Released to Stream Released to Stream
with credits for
Avoided Fertilizer
and Electricity
Figure 4-11: Consequential LCA results comparing the GHG emission impact
from releasing untreated wastewater to various interventions
66


In figure 4-11, in last two cases, avoided emissions from electricity as a result
of biogas reuse was subtracted from the on-site emissions. However, in the
second to last case, avoided emissions due to no fertilizer use was subtracted
from the off-site N2O emissions, while in the last case, the avoided emissions
were subtracted from N2O from agriculture.
4.5 Insights and Recommendations for Future Work
The following insights were seen from the results in this chapter:
Wastewater treatments plants are effective at removing pathogens
(99%), BOD (81%), and solids (93%) from influent wastewater while
retaining a high amount of nutrients and using a low amount of
electricity per gallon wastewater treated.
Energy recovery potential exists for WWTPs but is not used.
A system-wide analysis shows that nutrient recovery from wastewater
agriculture is highly dependent on the flow rate of wastewater and
associated land available. Nutrient recovery was relatively small in
terms of the percentage nutrients used versus the total nitrogen
discharged. Dried sludge, in contrast, could be a more successful
avenue for nutrient recovery as it can be distributed more safely and
easily. However, a market for the sludge must first be established.
System-wide greenhouse gas emissions with and without the use of
WWTP were not very different, given the range of model uncertainty.
However, energy investments did reduce BOD, COD, and pathogens.
Further exploration of many of the parameters would be useful for this study.
Notably, the reasons for the unusually high influent nitrate concentration are
not well understood. Measurements of influent nitrate at all four WWTPs
would help to understand the variation in the sewer system. The same
67


measurements taken before and after a storm could provide insight to the
sources of the high nitrate concentrations. Also, the flows throughout the
WWTP, that have been requested, would be useful for further quantifying
process emissions.
In the next chapter, the fate of pathogens from wastewater reuse for urban
agriculture will be discussed.
68


5. Measuring Water and Food Relationships: A Site Study
5.1 Introduction
As discussed in chapter 1, the life-cycle benefits and costs arising from net
energy investments and net greenhouse gas (GHG) emissions in newly
implemented WWTP infrastructure have not been quantified for WWTP
effluent reuse in urban agriculture.
This study takes an urban agriculture perspective to evaluate water savings,
nutrient delivery, and pathogen reduction achieved for irrigation of three
different urban agriculture plots. The three irrigation waters of differing
nutrient and pathogen qualities were sourced from:
1) Groundwater from a borewell (50 feet deep),
2) Treated effluent from the wastewater treatment plant, and
3) Untreated water in surface streams.
5.2 Site Selection and Study Design
Several sites where wastewater agriculture was used were visited within
Hyderabad by a team of researchers including Miller, Ramaswami, and
Amerasinghe, to assess crops being grown and to speak with farmers. Sites
visited included a borewell irrigated site in Kachivani and a wastewater site in
Peerzadiguda and finally, the Nallacheruvu site.
5.2.1 Site Selection
The farming site at Nallacheruvu was chosen for the following reasons: 1) its
co-location of WWTP and urban agriculture, 2) ready access to three different
qualities of water, 3) the availability of an experienced farmer, and 4) the
HMWSSB gave permission for use of the study site and were willing to share
data for NWWTP. The study took place during the dry-season in March- May
69


2010, when water levels were at their lowest, and wastewater was the least
diluted with storm water.
Figure 5-1: Aerial view of NWWTP showing co-location of urban agriculture
plots. 1: groundwater; 2: NWWTP effluent; 3: untreated surface water.
Source: Google Earth, Imagery date April 5, 2010.
The crop of interest was spinach because it is frequently grown in this area
and is a leafy green vegetable commonly eaten by people in India. Other crops
grown in the region include mainly paragrass, an animal feed. The WHO
irrigation guidelines are often most stringent for leaf crops because they are
eaten raw in many parts of the world (World Health Organization 2006).
However, it is recognized in this study that most vegetables in India are
cooked and not eaten raw (Khanum, Siddalinga Swamy et al. 2000). Palak
70


(spinach in Hindi) seeds were bought from a local agricultural store in Uppal
(south-east neighborhood of Hyderabad). The farmer preferred an All
Green variety that is known to grow well in hot temperatures. The species is
actually Beta vulgaris and a heirloom variety of chard that is native to India
(BackyardGardener.com 2010; EvergreenSeeds.com 2010; Singh and
Agrawal 2010), but will herein be referred to as spinach.
The pathogens of interest in the crop are Escherichia coli (E. coli) and
nematode ova (eggs) (Roundworm Ascaris lumbricoides (Ascaris); and
Hookworm: no distinction was made between Old World, Ancylostoma
duodenale, and New World Necator americanus hookworm). E. coli and
nematode eggs are commonly used to indicate wastewater contamination and
associated health risks (Cifuentes 1998; An, Yoon et al. 2007; Mara, Sleigh et
al. 2007; Ensink, Blumenthal et al. 2008). Nematodes pose a high health risk
when compared to other pathogens due to their infective dose being small,
their ability to live longer in the environment, and the fact that humans
generally have low immunity to them (Gaspard, Ambolet et al. 1997; World
Health Organization 2006).
For nutrients, the focus was on the primary macronutrients (nitrogen,
phosphorus, potassium) that are essential for plant development. Nitrogen (N)
is utilized for: all proteins, enzymes, metabolic processes involved in the
synthesis and transfer of energy; chlorophyll; rapid growth, increasing seed
and fruit production; and improving the quality of leaf and forage crops.
Phosphorus (P) is utilized for: photosynthesis; formation of all oils, sugars,
starches; transformation of solar energy into chemical energy; proper plant
maturation; withstanding stress; and rapid growth. Finally, potassium (K) is
utilized for: building of protein, photosynthesis, fruit quality and reduction of
diseases; and K levels are usually higher than others to reflect parent material
71


(igneous rocks ~50,000ppm) (North Carolina Department of Agriculture and
Consumer Services).
Furthermore, organic carbon is an important part of soil for: crop yield, soil
fertility, soil moisture retention, aeration, nitrogen fixation, mineral
availability, disease suppression, soil composition, and general soil structure
(Leu 2007). Total organic carbon can also be measured in water and is
expected in water with sewage contamination. BOD and COD were also
measured in irrigation water.
pH, electrical conductivity (EC), and total suspended solids (TSS) were of
interest in this study because they can affect and inform on other parameters.
Levels of pH in water and
soil are known to affect
availability of nutrients: at
low pH, macronutrients
tend to be less available,
while at high pH,
micronutrients tend to be
less available. EC, or the
capacity of the media to
conduct electrical current,
is directly related to the
amount of solids dissolved
in that media (soil or
water). EC can affect soil
texture, cation exchange
capacity, drainage
conditions, organic matter
level, salinity, and subsoil
Figure 5-2: Farmer, Chandriah, and translator,
Aruna, at the untreated surface water plot
72


characteristics (Grisso, Alley et al. 2009). TSS is simply a measure of the
amount of solids that are not dissolved in the water, and can relate to oxygen
demand, turbidity, and organic matter in water.
5.3 Study Design
The same crop was grown in three different sites with three widely
varying source water qualities to compare the impacts of water quality on
crop quality (food pathogens) and productivity.
Plots were co-located so that the same farmer could cultivate them and the
same practices could be used.
The crops were grown over the same time period and irrigated at regular
intervals by the farmer.
The researcher observed irrigation events at least twice per week and
communicated with the farmer using a translator. The farmer was
instructed to treat all three plots in the same manner, and in particular, not
to fertilize any one plot if it was doing poorly.
The researcher measured flow rate at the start of each irrigation event, and
sampled the water throughout irrigation events. Water from the
groundwater plot was sampled near the end of the irrigation event to
ensure proper purging of stagnant well water.
Some water, soil, and crop samples were delivered to local laboratories for
analysis as shown in gray below. Other samples were analyzed by the
researcher at lab facilities provided by the International Crop Research
Institute for the Semi-Arid Tropics (ICRISAT) with a cooperative
agreement with UC Denver.
73


The researcher tested for pH, electrical conductivity, and total suspended
solids in water, and E. coli, total coliforms, and Ascaris and Hookworm
ova in water, soil, and crop.
In addition, the researcher had purchased Hach kits in the US to complete
nitrogen, phosphorus, and potassium testing. But because of customs
regulations banning transport of chemical reagents, these tests had to be
outsourced to a lab in India.
Table 5-1: Tests done over one crop growing cycle. Gray: tests outsourced to
a lab; white: tests done by this researcher. All sampling, transport, and sample
preparation was also done by this researcher._________________________________
Groundwater Plot Treated Effluent Untreated Surface
Pre-analysis Tests done
Soil Physical Characteristics X X X
Soil Nutrients X X X
Irrigation Water Pathogens X X X
Irrigation Water Nutrients X X X
Dynamic Monitoring Frequency of tests over study
Water Quantity (volume) 9* 9* 9*
Water Pathogens 9* 9* 9*
Water Nutrients 9* 9* 9*
Soil Pathogen (E. co/;7Nematode) mid, end mid, end mid, end
Soil Nutrients 3 3 3
Soil Water Nutrients mid mid mid
Crop Quantity (weight) End end end
Crop Pathogen (E. co/i/Nematode) mid, end mid, end mid, end
Crop Nutrients mid, end mid, end mid, end
* Farmer irrigated about every two days, depending on weather. Water quantity,
pathogens, and nutrients were recorded by the researcher every alternate irrigation
event
5.3.1 Composites and Replicates
The researcher gathered composite water samples during the observed
irrigation events (over 5-15 minutes depending on water flow rate) for
water quality analysis, and crop and soil samples before irrigation events.
74


Crops were sampled at midpoint and endpoint (n=3 from each plot). The
endpoint represents harvest conditions.
Duplicates were run by the researcher for E.coli at every tenth sample,
regardless of the media, throughout the study as a method of doing quality
assessment/quality control (QA/QC). Control plates (run with dilution
water) were prepared along with every set of samples in order to ensure
that sterile techniques were used.
Independent laboratories did their own QA/QC with blanks and standards.
When weighing crop bundles at harvest, a composite sample of 20 bundles
was taken to the lab and weighed. Then, an average weight per bundle
could be determined.
Three grab samples (lOOg spinach each) from the larger composite sample
from each plot were taken for E.coli testing, then composited again for
nematode testing.
Figure 5-3: Crop pathogen measurement overview: challenge of dilution
versus concentration
75


Next, the methodologies to carry out this study will be discussed.
5.4 Site Study Methodology
Pre-analyses of water and soil quality were done to determine the most-
suitable plots for this study. The purpose was to have certain factors for the
three plots to be as similar as possible, such as soil characteristics, location,
and farmer practices, thus isolating the impact of water quality on urban
agriculture. Plots were chosen according to pre-analysis results. After proper
preparation of the plots, including wetting, weeding, and plowing, the seeds
were planted and routine testing was done throughout one growing cycle. Pre-
analysis is described first, then dynamic measurements.
5.4.1 Site Preparation and Pre-Analyses
Before choosing the exact location of the plot, water and soil parameters were
tested. Water tests were taken from many locations to get initial measurements
of nutrients and pathogens in the area. Soil samples were taken from each
potential plot location and tested for physical characteristics and nutrient
content.
For analysis of soil physical characteristics, samples were taken from four
random spots within each potential plot with the appropriate soil core tools
throughout from 0-15cm and 15-30cm, then combined to form a composite
sample for each soil layer. Soil texture (distribution of particle sizes), bulk
density, wilting point and field capacity, porosity, and particle density were
measured at ICRISAT using sieve analysis, moisture retention at pressures of
0.33 and 15 bar, bulk weight per volume, and particle size tests and
calculations. Soil type for the three plots were similar as determined by the
soil texture pyramid (figure 5-4) (National Resources Conservation Service
2011).
76


The plot locations were chosen so that the soil textures were similar and the
three plots were co-located with their individual source of water. Their
orientation to stream bed, delivery of water, direction of water flow, size,
slope, and elevation were taken into consideration and were made as similar
as possible. All plots were oriented with their longest side parallel to the
closest stream bed (approximately 10m away for each) and every plot was
irrigated by opening a channel and allowing the plot to flood lengthwise. Each
plot was 12 m, but the untreated surface water plot was 6m by 2m while the
groundwater and treated plots were 4m by 3m, due to space constraints and an
effort to maintain as shallow a slope as possible. The elevation for each plot
was roughly the same at 470lm above sea level. Plots were leveled but had a
slight gradient to facilitate gravity-flow irrigation.
Figure 5-4: Soil texture pyramid. Groundwater plot:
x; treated effluent plot: o; untreated surface water
plot: 0.
Small channels were
dug to deliver water to
the groundwater and
treated effluent water
plots, while the
untreated surface
water plot had an
established network of
channels that diverted
water from the nearby
surface stream. The
groundwater and
treated effluent water
sites were wetted with
the appropriate water
and newly plowed 10
77


days before seeding. The untreated surface water plot had been cultivated with
spinach approximately 1 month earlier and laid without a crop for that time.
For this study, there was much discussion on whether to amend the soils with
lower nutrient content, to make them all similar. It was decided that the soils
would not be amended, as the farmer, Chandriah, was confident that the crops
would grow in all plots. The farmer prepared soils as per his normal practice
with the appropriate water for site and plowing. He did not use additional
fertilizers throughout the study.
5.4.2 Dynamic Measurements
The following parameters were measured periodically throughout one crop
growth cycle from March 26- April 28, 2010:
Irrigation water quantity
Irrigation water quality
o pH, EC, TSS
o BOD, COD, TOC
o Nutrients: Nitrogen, phosphorus, and potassium
o Pathogens
E. coli and Total coliform
Nematode ova (Ascaris and Hookworm)
Soil quality
o pH, EC, TOC
o Nutrients: Available and total nitrogen, phosphorus, and
potassium
o Soil water nutrients: Nitrogen, phosphorus, and potassium
o Pathogens (same as in water)
Crop quantity
Crop quality
o Nutrients: Nitrogen, phosphorus, and potassium
o Pathogens (same as in water)
The methods for each of these tests are described next.
78


5.4.2.1 Water Quantity
In order to standardize the volumes of water over all three plots during flood
irrigation (by both the researcher and the farmer), various methods were
attempted. Because the groundwater and treated effluent water plots had water
being delivered through a pipe, which flowed into a channel, a time/volume
method was used. In this method, the amount time that it took to fill a known
volume (bucket) was noted at least twice a week at time of irrigation. The
flows were not too fast and this method worked well.
For the untreated surface water plot, the water was delivered in a channel. To
measure volume here, a float
(a cap from a water bottle) was
used and the time that it needed
to travel a known distance was
measured. Then the flow was
calculated by multiplying
velocity by cross-sectional area
of the channel. Each time this
method was used, the banks and
the bottom of the channel were
cleaned from vegetation to
ensure that the initial cross-
sectional area measurements
were retained. This method was
always done three times and the
average was taken.
For the groundwater and treated effluent water plots, the bucket method was
compared to the float method within the channel leading the plot. With
these rough calculations, the float method was found to be, on average, within
Figure 5-5: Preparation of the channel,
with help of the farmer, Chandriah, leading
to the untreated surface water plot
79


9% (standard deviation of 8%) by volume of the bucket method for
measuring water volume.
SA.2.2 Water Quality
Sampling of water was done in the same way throughout the study. Any
materials for E. coli detection were sterilized prior to sampling or testing.
Sterile procedure and techniques were strictly followed for all E. coli tests. All
other sampling vessels and materials were always cleaned with soap and water
and well rinsed with distilled water prior to sampling or testing. Water
samples were always taken at points where the water was well mixed. This
was determined visually by fast moving, turbulent flow. Samples were taken
from below the surface of the water at 40-60% of the water depth to minimize
settling of solids. Care was also taken not to disturb sediment before or during
sampling. Sampling vessels were always filled to the top so as to minimize air
space. All samples for E. coli testing were kept on ice and the tests were done
the same day, as soon as possible. Samples for nutrient testing were kept on
ice, transferred to a refrigerator in the appropriate lab, and the test was done as
soon as possible.
Generally, irrigation water may not be considered as a significant source of
nutrients or pathogens. However, because the irrigation water in this study
contains wastewater, it is an important part of the agricultural system.
The following tests were done on water:
pH, EC, TSS: Measurements of pH and electrical conductivity were done by
this researcher using a Oakton Multi-Parameter Meter. The instrument was
periodically calibrated for temperature and pH standards. TSS was measured
by filtering water through a filter paper and then drying it at 3-4 hours at 65C
(the only oven available) (Standard Method 2540 D).
80


TOC, BOD, and COD: Total organic carbon was tested to determine the
amount of organic matter in the irrigation water. The test was carried out on a
Total Organic Carbon Analyser TOC-Vcpn (manufactured by Shimadzu
Corporation). BOD5 (Standard Method 5210 D) and COD (Standard Method
5220 B Open Reflux Method) were also done 2 and 3 times, respectively, for
irrigation water. All these tests were contracted to the Environmental
Protection Training and Research Institute (EPTRI) labs in Hyderabad.
Nutrients NPK: Water was tested for plant available dissolved inorganic
nutrients such as: inorganic soluble nitrogen (NH^-N Standard Method 4500-
NH3 C and NCE'-N Standard Method 4500-N03- E), soluble phosphorus
(Standard Method 4500-P D), and soluble potassium (Standard Method 3500-
K B). All tests were carried out using Standard Methods for Examination of
Water and Wastewater (American Public Health Association, American Water
Works Association et al. 2006) at ICRISAT. Nitrite (Standard Method 4500-
N02 B) in all three irrigation waters was tested once to confirm that nitrite
concentrations were low and could subsequently be ignored as insignificant.
Pathogens: Hachs membrane filtration method for E. coli and total coliform
detection was the most accessible and precise method for fecal indication.
This method involved diluting the water sample by factors of 10,10,10 ,
104, 105, 106, until each pathogen could be assumed to develop into separately
visible colonies over an incubation period of 20-24 hours. The filter paper on
which the pathogens are filtered is placed into a sterile petri dish with nutrient
both that stains E. coli red/purple and total coliforms blue. Then, the dilution
factors can be applied to scale back to the original concentration of these
bacteria in source water. See Appendix B for method details.
For detection of Ascaris and hookworm eggs, a method developed by IWMI
(which was adapted from Ayers and Myer 1996) was used. This involved
gathering large samples, at least 5 liters of each irrigation water and leaving
81


them to settle overnight in the lab. The next day, a multi-step centrifugation
protocol was done to separate sediment and eggs from the water. Then, the
sediment containing the eggs was suspended in zinc sulfate, and a McMaster
slide was used to identify and count the eggs under a microscope (Ayres and
Mara 1996). A more detailed description of these methods can be found in
Appendix B.
5.4.2.3 Soil Quality
Soil samples were taken from random spots within the plot (first carefully
removing the top 3cm and extracting the sample). For chemical and physical
testing, soil was extracted with a soil core tool. For E. coli testing, soil was
extracted with a small shovel. Soil samples for E. coli testing were kept on ice
and the tests were done the same day, as soon as possible.
pH, EC, TOC: Total organic carbon was also done to determine the amount
of organic matter in the soils. Soil pH and EC were also done by ICRISAT.
Methods for all of these tests can be found in Soil Science Society of America
and American Society of Agronomy 1996.
Nutrients: Available and Total NPK: The soil quality parameters tested
were for both plant available and total nutrients such as: mineral (NH/-N and
NO3 -N) and total nitrogen, Olsen and total phosphorus, and exchangeable and
total potassium (Soil Science Society of America and American Society of
Agronomy 1996). ICRISAT carried out these test and included QA/QC with a
known standard sample from the International soil analytical exchange. The
results from this standard sample were always found to be within the expected
range for nitrogen, phosphorus, and potassium. Also, duplicates run were
always found to be within 99% of each other.
Lysimeter Soil Water and NPK: Soil water was collected with two
lysimeters per plot, randomly placed near the inlet of water to the plot and
82


another randomly placed
farthest from the inlet to
plot (Soilmoisture
Equipment Corp.
1900L12-B02M2).
These soil water
samplers were installed
according to the method
provided by
Soilmoisture and placed
under a vacuum for over a period of 8 hours in each plot. The samples were
extracted and taken to the EPTRI lab for total nitrogen (KEL Plus Classic DX
TN Analyzer), total phosphorus (Standard Method 4500-P D), and total
potassium (Standard Method 3500-K B) tests.
Figure 5-6: Lysimeters installed in groundwater
plot.
Pathogens: E. coli testing was done by slightly modifying of Hachs
membrane filtration method. Using sterilized materials, 10 g of soil was added
to 95mL of 0.8% NaCl, covered with parafilm, and placed on a shaker for 60
minutes. Under a sterile vacuum hood, this liquid was then filtered through a
strainer to remove large particles that would clog the membrane. With this
solution, appropriate dilutions were done and the regular method was
followed as described in section 5.4.2.2.
The method for nematode ova identification in soil was adapted from Zenner
et al. for this study (Zenner, Gounel et al. 2002). 50g soil was added to 200mL
distilled H2O. This was mixed thoroughly with a metal rod for 20 seconds and
filtered through a coarse sieve. The solution was divided into tubes and
centrifuged. Then, sediment from one tube was resuspended in the floation
solution, magnesium sulfate, and the top of the tube was covered with a glass
cover slip. During a second centrifugation step, the eggs floated to the top of
83


the tube and stuck to the coverslip. Eggs on the coverslip could be counted
under a microscope. This second centrifugation step was repeated until eggs
were no longer found.
5.4.3 Farming, Irrigation, and Harvesting Practices During Study
After the plots were chosen, the source of irrigation water was the major
difference in practice between the plots. The farmer, Chandriah, cultivated all
three plots and his normal practices were followed for plowing, seeding,
irrigation, pest control, and harvesting. Plowing was done by hand with a hoe.
Seeds were broadcasted by sweeping motions over the plot, then raked into
the soil, then flooded with water. The plots were irrigated about every 2-4
days, as was deemed appropriate by the farmer, for weather and soil
conditions.
On the 10th day of the study (about 1/3 into the study) a worm was attacking
and causing a lot of damage to the treated effluent plot and the untreated
surface water plot. An insecticide (phorate:
http://www.hyderabadchemicals.com/hyfort.htmn (US EPA 2010) was used
on this day, once only, to kill this insect and the farmer was sure that the crop
would be lost if this action was not taken. This action may have affected the
nematode population in the soil, but it is unknown whether it affected
nematode ova. Because the farmer uses this chemical regularly when there are
insect problems, it is likely that it does not kill nematode ova, because there
are many in the soil. Soil samples were taken for E. coli one week later, for
nutrients one and a half weeks later, and for nematode ova two weeks later.
This treatment will not affect the water samples as the water is taken at the
entrance to the plot, before flowing across the plot surface.
At time of harvest, the leaves were cut at the base (before the roots) and
gathered into bundles. To hold them together, they were wrapped with a
84